{"id":"9f6f5631-396a-4cad-926e-d052c0a00366","arxiv_id":"2608.11988","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"In trilayer 3R-MoS2, the two zero-polarization stackings ABA and BAB show distinct low-frequency shear-mode Raman responses, contrary to the standard bond-polarizability model.","lead":"This paper shows that two electrically neutral stacking arrangements of a three-layer molybdenum disulfide crystal, which look identical in an electrostatic map, produce clearly different Raman light-scattering fingerprints. The result means no single measurement technique can reliably identify the stacking order that controls sliding ferroelectricity in these materials.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The central claim of a bond-polarizability model failure is asserted without showing the model's prediction; a symmetry analysis suggests the standard model predicts identical ABA/BAB shear activity, which would make the observed S1-active neutral domain a selection-rule inconsistency.","rationale":"The reader's formal weakest assumption was the tentative ABA/BAB labeling via the same-group PL hybridization model. My stress-test identifies a related but more load-bearing gap: the paper's central novelty, the failure of the bond-polarizability model, is never substantiated by an actual model prediction. The text asserts the failure twice (Results and Conclusion) and notes it is 'at variance with previous studies' (Refs. 27-29), but no calculation is provided. A symmetry-based reading of the standard model suggests it would predict identical shear-mode activity for ABA and BAB (only S2 active), which would make the observed S1-only spectrum in one neutral domain a selection-rule violation. That would not just weaken the 'mismatch' claim; it would suggest the S1-active domain is not ABA or BAB, destabilizing the entire stacking assignment. Alternatively, if a more complete model does predict different activities, the paper's claim is wrong. Either way, a concrete model calculation is required. I do not dispute the experimental observation itself: it is reproduced across flakes, survives annealing, and persists after encapsulation, so the underlying KPFM-Raman correspondence is likely real. The weakness is in the interpretation and labeling. Because the reader already assigned CONDITIONAL and flagged the missing model prediction in the rationale, my concern reinforces that verdict rather than changing it. Self-identified limitations in the manuscript—the 'tentatively assign' of ABA vs BAB and the absence of error bars on KPFM potentials—further support keeping the verdict at CONDITIONAL pending the model comparison.","tokens_in":8644,"tokens_out":35393,"duration_ms":357631,"concrete_test":"Implement the interlayer bond-polarizability model of Liang et al. (Nanoscale 2017) for a trilayer MoS2 using the four stackings ABC, CBA, ABA, and BAB, with the same interlayer force-constant and polarizability-derivative parameters used in Refs. 27-29. Compute the Raman intensities of the two shear modes S1 and S2 in the backscattering, cross-polarized geometry of Fig. 2b. Compare I(S1) and I(S2) for ABA versus BAB. If the model gives I(S1)=0 and I(S2)>0 for both stackings, then the experimentally observed S1-only neutral domain cannot be ABA or BAB, and the paper must re-examine the stacking assignment. If the model gives I(S1)>0 in one stacking and I(S2)>0 in the other, then the central 'mismatch' claim is refuted. This single calculation settles whether the claimed model failure is real or an artifact of the assignment.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's headline conclusion—that the standard bond-polarizability model cannot account for the differing shear-mode activity of ABA and BAB—rests entirely on an assertion; no model calculation or predicted Raman spectrum is shown anywhere in the main text or supplement. The relevant comparison is the predicted intensity ratio I(S1)/I(S2) for ABA and BAB in the experimental backscattering, cross-polarized geometry, and it is absent. This matters because the standard interlayer bond-polarizability model (Luo et al. 2015, Liang et al. 2017, Refs. 25,26) is a nearest-neighbor sum over interfaces. In ABA the two interfaces have stacking vectors (+v, -v), in BAB (-v, +v). Under the usual symmetry that the polarizability derivative changes sign when the stacking vector reverses, the antisymmetric shear mode S1 (outer layers moving in opposite directions) cancels to zero intensity in both stackings, while the symmetric mode S2 is active in both. If that is the model's prediction, then the S1-only spectrum of Region I in Fig. 2b is not merely 'beyond the model'—it violates the horizontal-mirror (σ_h) symmetry that both ABA and BAB possess, since S1 transforms as E'' and is forbidden in backscattering. That would call into question whether the S1-active neutral domain is ABA or BAB at all, undermining the central claim. Conversely, if a more complete version of the model does predict S1 in one stacking and S2 in the other, the paper's mismatch claim is refuted. Either way, the central claim is unanchored without the explicit model prediction, and the paper's own 'tentatively assign' language for the ABA/BAB PL labeling (Results, PL paragraph) does not resolve this.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript combines KPFM, low-frequency Raman spectroscopy, and low-temperature photoluminescence to characterize stacking domains in trilayer 3R-MoS2 flakes on hBN. The authors find that two KPFM-degenerate low-potential domains, which they assign to the zero-net-polarization ABA and BAB stackings, exhibit starkly different interlayer shear-mode Raman spectra: one domain shows only the S1 shear mode, the other only the S2 mode, while the polar ABC domain shares its shear signature with one of the neutral domains. The KPFM-Raman correspondence persists after annealing, is reproduced across multiple flakes, and survives encapsulation. Low-temperature PL shows distinct A-exciton lineshapes for the two neutral domains, tentatively assigning them to ABA and BAB. On this basis the authors conclude that the standard bond-polarizability model cannot account for the ABA/BAB shear-mode difference and that KPFM and low-frequency Raman probe complementary aspects of stacking order.","tokens_in":8954,"tokens_out":6610,"duration_ms":72162,"significance":"The experimental observation that two electrostatically identical domains differ in their vibrational fingerprint is valuable for the sliding-ferroelectric community, as it challenges the common assumption that KPFM and shear-mode Raman give redundant stacking information. The paper's strength is its multi-probe, multi-sample approach: direct spatial correlation between KPFM and Raman maps, reproducibility across flakes, and checks after annealing and encapsulation all support the robustness of the core observation. The low-temperature PL data add an independent, though model-dependent, corroboration. However, the central negative claim against the bond-polarizability model is not quantitatively substantiated: no model prediction is shown, and the ABA/BAB assignment is explicitly tentative. If the standard model actually predicts identical ABA/BAB shear activity, or if the S1-active neutral domain is inconsistent with expected selection rules, the headline conclusion would need substantial reframing.","major_comments":[{"comment":"The claim that the standard bond-polarizability model 'does not account for the difference in shear-mode activity between the ABA and BAB configurations' is made without showing any model prediction. The manuscript never specifies the predicted I(S1)/I(S2) ratio for ABA and BAB in the experimental cross-polarized backscattering geometry, nor does it compare those predictions with the measured spectra. If, as a symmetry argument suggests, the standard model predicts identical ABA/BAB shear activity with the antisymmetric shear mode forbidden in exact backscattering, then the observed S1-only neutral domain would be a selection-rule inconsistency that undermines the stacking assignment rather than a benign 'beyond the model' effect. The authors should either perform the model calculation explicitly (e.g., following Refs. 25 and 26) and display the predicted spectra, or clearly state which assumptions of the model are being relaxed.","section":"Results and discussion, 'Low-frequency Raman spectroscopy' (Fig. 2b)"},{"comment":"The ABA/BAB labels used to interpret the PL lineshapes and the shear-mode contrast are explicitly 'tentative', based on the hybridization model of Ref. 31 from the authors' own group. The core observation that two zero-polarization domains have different shear modes is independent of these labels, but the specific conclusion that a particular neutral domain (n1 or n2) is ABA or BAB, and hence the comparison with bond-polarizability predictions, depends on them. The manuscript should provide an independent check of the ABA/BAB assignment (e.g., from the absolute shear-mode frequencies or a first-principles calculation) or explicitly limit the conclusion to 'two unassigned neutral stackings differ in shear-mode activity.'","section":"Low-temperature photoluminescence, Fig. 3e"},{"comment":"The paper states that the observed shear-mode behavior is 'at variance' with previous studies (Refs. 27-29) that found 3R-MoS2 shear modes fully consistent with the bond-polarizability picture, but it never discusses what those works predict for ABA/BAB trilayers or why the present result differs. Given that the central claim is a contradiction of an established model, the authors should articulate the specific point of departure, ideally by reproducing the relevant prediction and showing where it fails, rather than only citing the earlier works.","section":"Results and discussion, 'Low-frequency Raman spectroscopy'"}],"minor_comments":[{"comment":"The phrase 'of a exfoliated trilayer' should read 'of an exfoliated trilayer.'","section":"Abstract"},{"comment":"The sentence 'It distinguishes only two potential domains, divided in the surface-potential map into three spatially distinct regions' is confusing; the map has three regions but only two potential levels, so a clearer phrasing would be 'the map contains three spatially separated regions but only two distinct potential levels.'","section":"Results and discussion, first paragraph"},{"comment":"The labels n1, n2, p+ appearing in panels (c) and (d) are not defined in the caption; they should be defined there or in the figure itself.","section":"Figure 2 caption"},{"comment":"The inset schematics labelled 'displacement associated with each region/shear mode' are difficult to interpret; explicit arrows showing the atomic displacement pattern for S1 and S2 in each stacking would improve clarity.","section":"Figure 2b inset"},{"comment":"The caption contains the typo 'of a the same flake' and should read 'of the same flake.'","section":"Figure 3 caption"}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Known you'd want my read on 2608.11988. Short version: the core observation is real and likely correct; the headline interpretation outruns the evidence in one specific place.\n\nWhat's new: two domains in a 3R-MoS2 trilayer that are identical in KPFM surface potential (both zero-net-polarization) produce completely different low-frequency shear-mode spectra — one shows only S1, the other only S2. That's reproduced across flakes, survives annealing, and persists after encapsulation. Earlier work (Liang et al., Liu et al.) reported shear modes consistent with the bond-polarizability model, so this is a genuine counter-example and corrects the working assumption that shear-mode Raman alone can assign stacking in 3R-MoS2.\n\nAlso good: the KPFM+Raman logic forcing the two low-potential domains to be ABA and BAB is internally consistent. Even if the PL-based ABA/BAB labeling (which the paper honestly calls tentative) were swapped, the main observation stands: two electrostatically degenerate stackings differ vibrationally. The annealing and encapsulation controls are the right ones.\n\nSoft spots. The claim that the bond-polarizability model fails is asserted, not demonstrated. No predicted intensities are shown. The stress-test note worried that the standard model would predict S1 forbidden in backscattering; I think that specific symmetry argument is wrong — E'' shear modes are allowed in cross-polarized backscattering, and a simple two-interface bond-polarizability sum gives both S1 and S2 active in both ABA and BAB with equal magnitudes. So the model mismatch, if anything, is even more striking: the model predicts identical spectra, experiment shows one mode absent per domain. But the authors need to show that calculation. Without it, the 'beyond the model' statement is unanchored. They also omit error bars on the mode frequencies, which is minor but fixable.\n\nOne more caveat: the PL-based assignment of which neutral domain is which depends on a same-group model (Ref. 31) and the asymmetric dielectric environment. They flag this. It's a real weakness, but not load-bearing for the central empirical claim.\n\nWho should read it: anyone using shear-mode Raman or KPFM to assign stacking in sliding ferroelectrics. It's a strong multi-technique case study and deserves a serious referee. My recommendation: send to peer review; require the authors to add the bond-polarizability prediction for their exact geometry. That's a revision, not a rejection.","headline":"Solid multi-probe result — the KPFM-degenerate neutral domains genuinely differ in shear modes — but the bond-polarizability failure claim needs the model calculation shown.","tokens_in":9560,"tokens_out":3707,"would_cite":true,"duration_ms":36676,"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":"In trilayer 3R-MoS2, low-frequency shear-mode Raman distinguishes two mirror-symmetric, zero-polarization stackings that Kelvin-probe force microscopy cannot tell apart, a contrast the standard bond-polarizability model fails to explain.","keywords":["sliding ferroelectricity","3R-MoS2","shear modes","Raman spectroscopy","Kelvin probe force microscopy","stacking order","bond polarizability","photoluminescence"],"falsifier":"Direct atomic-resolution imaging of the two low-potential domains, for example cross-sectional scanning transmission electron microscopy, would settle the claim: if the domains prove to have identical stacking, or any stacking outside the assumed ABC/CBA/ABA/BAB set, the central conclusion collapses. Alternatively, a first-principles Raman calculation that reproduces the S1/S2 contrast within the bond-polarizability framework would falsify the claim that the model fails.","tokens_in":8476,"feed_emoji":"🔬","tokens_out":5856,"duration_ms":53922,"temperature":0.7,"pith_summary":"The paper sets out to test whether non-destructive optical probes can assign the stacking order of a trilayer 3R-MoS2 flake whose ferroelectric polarization is already mapped by Kelvin-probe force microscopy. Combining KPFM, low-frequency Raman, and low-temperature photoluminescence, it finds that the two mirror-symmetric stackings ABA and BAB, both of which have zero net polarization and are therefore indistinguishable in KPFM, have drastically different shear-mode Raman spectra and different excitonic line shapes. The authors argue that this difference is intrinsic: it survives annealing, domain-wall rearrangement, and encapsulation, and appears on multiple flakes. They further show that the standard interlayer bond-polarizability model cannot account for the shear-mode contrast, and conclude that no single technique used here is sufficient to assign sliding-ferroelectric stacking order. If correct, low-frequency shear-mode Raman becomes a necessary complement to electrostatic probes for stacking-domain identification.","feed_headline":"Shear modes distinguish 3R-MoS2 stackings invisible to KPFM","feed_subtitle":"In trilayer 3R-MoS2, Raman shear modes separate the two zero-polarization stackings that KPFM cannot resolve.","key_machinery":"The central objects are the two rigid-layer interlayer shear modes, S1 and S2, of the trilayer, measured in a cross-polarized low-frequency Raman configuration that suppresses layer-breathing modes, together with the KPFM surface-potential map and its roughly 60 mV-per-interface potential ladder. The argument works by comparing two fingerprints of the same domains: electrostatics, which groups ABA and BAB together as neutral, and vibrational selection, which separates them. The mechanism that supposedly fails is the interlayer bond-polarizability model, which links Raman intensities to stacking configuration and predicts equivalent shear activity for the mirror stackings; the data contradict that prediction.","core_discovery":"The central discovery is that in a trilayer 3R-MoS2 stack the two KPFM-degenerate zero-polarization domains, assigned to ABA and BAB stacking, activate different interlayer shear modes: one region shows only the lower-energy S1 mode, the other only the higher-energy S2 mode. Since KPFM distinguishes only high- and low-potential regions with a roughly 110 mV offset, the shear-mode maps lift the degeneracy of the two neutral domains and force the assignment that the high-potential domain is ABC while the two low-potential domains are ABA and BAB. The contrast persists after thermal annealing and encapsulation, and the photoluminescence line shapes of the two neutral domains also differ, with a hybridization model assigning the split spectrum to ABA and the single-peaked spectrum to BAB. The bond-polarizability model, which predicts identical shear activity for mirror-symmetric stackings, fails to describe the observed difference.","pith_inferences":["The ABA/BAB shear-mode contrast may be a general feature of parallel-stacked rhombohedral TMDCs, meaning shear-mode Raman could become a standard label for neutral mirror domains in multi-stack devices.","If the bond-polarizability model fails because of interlayer wavefunction hybridization, shear-mode intensities might be usable as a quantitative probe of interlayer electronic coupling, testable by comparing measured intensities with first-principles calculations.","The finding implies that any purely electrostatic domain-mapping scheme, including KPFM-based ones, systematically misses a class of stacking distinctions, so polarization-only assignments of stacking order should be treated as incomplete.","A decisive test of the tentative ABA/BAB labels could come from an optical probe sensitive to the mirror symmetry of the stack, such as circular dichroism or second-harmonic generation, which has not been applied here."],"forward_implications":["Shear-mode Raman can distinguish zero-polarization mirror stackings that KPFM cannot, providing a non-destructive, spatially resolved probe of stacking order in sliding ferroelectrics.","No single technique among KPFM, low-frequency Raman, and low-temperature photoluminescence is sufficient for stacking assignment; reliable domain maps require combining an electrostatic probe with a vibrational one.","The failure of the bond-polarizability model for ABA versus BAB indicates that interlayer Raman intensities carry information beyond mechanical registry, likely tied to the electronic structure of the stack.","The contrast is robust against annealing, domain-wall rearrangement, and encapsulation, so it can serve as an intrinsic fingerprint rather than a strain artifact.","The same combined approach should extend to other rhombohedral transition-metal dichalcogenides and to thicker stacks where multiple neutral mirror domains appear."],"supporting_citations":[{"why":"Defines the interlayer bond-polarizability model whose prediction of equivalent shear activity for ABA and BAB is contradicted by the data.","marker":"[25]"},{"why":"Companion formulation of stacking-dependent low-frequency Raman intensities within the bond-polarizability picture.","marker":"[26]"},{"why":"Establishes the cumulative polarization and roughly 60 mV per-interface potential ladder used to index the KPFM domains.","marker":"[11]"},{"why":"Provides the low-frequency Raman fingerprints of layer stacking in TMDCs that ground the shear-mode interpretation.","marker":"[20]"},{"why":"Supplies the hybridization model used to tentatively assign which neutral domain is ABA and which is BAB via photoluminescence line shapes.","marker":"[31]"},{"why":"Earlier shear-mode study of trilayer 3R-MoS2 whose consistency with the bond-polarizability model the present findings contradict.","marker":"[28]"},{"why":"Prior shear-mode Raman imaging of multilayer 3R-MoS2, another previous study that the present ABA/BAB contrast goes beyond.","marker":"[29]"}],"fun_headline_variants":["Raman shear modes expose stacking hidden from KPFM","Shear modes break KPFM tie in 3R-MoS2 trilayers","Zero-polarization stackings diverge in Raman shear","3R-MoS2: Raman lifts KPFM stacking degeneracy","Shear spectra reveal what KPFM misses in 3R-MoS2"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The conclusion that the two KPFM-degenerate neutral domains are specifically ABA and BAB depends on assuming both that the four stackings ABC, CBA, ABA, and BAB are the only ones present in the flake and that the photoluminescence-based identification of which neutral domain is ABA and which is BAB is correct.","fun_headline_variants_meta":{"raw":{"variants":["Raman shear modes expose stacking hidden from KPFM","Shear modes break KPFM tie in 3R-MoS2 trilayers","Zero-polarization stackings diverge in Raman shear","3R-MoS2: Raman lifts KPFM stacking degeneracy","Shear spectra reveal what KPFM misses in 3R-MoS2"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00018,"raw_usage":{"total_tokens":1304,"prompt_tokens":945,"completion_tokens":359,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":561,"completion_tokens_details":{"reasoning_tokens":265}},"tokens_in":561,"tokens_out":359,"duration_ms":3754,"temperature":1.0,"reasoning_tokens":265,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T00:19:35.212228+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Direct atomic-resolution imaging of the two low-potential domains, for example cross-sectional scanning transmission electron microscopy, would settle the claim: if the domains prove to have identical stacking, or any stacking outside the assumed ABC/CBA/ABA/BAB set, the central conclusion collapses. Alternatively, a first-principles Raman calculation that reproduces the S1/S2 contrast within the bond-polarizability framework would falsify the claim that the model fails.","supporting_citations":[{"cited_title":"Broken mirror symmetry in excitonic response of reconstructed domains in twisted","cited_arxiv_id":null,"evidence_quote":"Defines the interlayer bond-polarizability model whose prediction of equivalent shear activity for ABA and BAB is contradicted by the data."},{"cited_title":"Nature Communications , year =","cited_arxiv_id":null,"evidence_quote":"Supplies the hybridization model used to tentatively assign which neutral domain is ABA and which is BAB via photoluminescence line shapes."}],"review_version":1}