{"id":"41f7b269-5b1d-4963-85c5-47aaade288f9","arxiv_id":"2607.21470","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"In 4Hb-TaS2, the incommensurate CDW on 1H layers adopts two discrete elastic states (−2.3% compressed or +3.2% stretched) selected by the relative rotation of the surrounding 1T CDWs.","lead":"Scanning tunneling measurements on 4Hb-TaS2 show that the charge-density wave on the buried 1H layers changes its spacing depending on how the surrounding 1T layers' CDWs are rotated. This identifies interlayer registry as a controllable source of the wave-vector variability long seen in layered CDW materials.","discovery_kind":"unification","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Surface 1H CDW is an unverified zero-strain reference; burial by a second 1T layer could shift qH as much as the registry splitting, so the compressive/tensile labels are not secure.","rationale":"The reader's weakest assumption correctly identifies the surface 1H-terminated CDW as an unverified zero-strain reference. The core claim that the 1H CDW q-vector depends on the registry of the surrounding 1T CDWs is supported by the D1–D2 difference, which is large compared to the quoted resolution and based on direct moiré observations. However, the specific headline numbers (−2.3% compressive, +3.2% tensile) and the sign of the strain relative to 'unstrained' are only meaningful if the surface qH00 is the equilibrium q for a buried 1H layer. Since a buried layer has two 1T neighbors instead of one, charge transfer and dielectric environment differ, and the paper provides no measurement of an unstrained buried 1H CDW. This is a genuine soft spot because the title and abstract emphasize the absolute compressive/tensile states and the elastic model uses qH00 as the zero-force reference. The paper remains conditionally acceptable: the authors should either justify the surface baseline (e.g., by showing qH00 is insensitive to the subsurface registry) or reframe the result as registry-induced relative shifts. The circular stiffness estimate is a secondary issue; the surface-reference problem is the most load-bearing because it affects the quantitative central claim, not just the interpretation.","tokens_in":18823,"tokens_out":16261,"duration_ms":146290,"concrete_test":"Perform DFT total-energy calculations of the 1H CDW wave vector in three geometries: a freestanding 1H monolayer, a 1H monolayer between two rigid 1T CDW layers with the aligned registry, and the same trilayer with the 27.8°-rotated registry. If the computed surface-to-buried shift (freestanding vs trilayer) is ≥ 0.01 Å^-1 (comparable to the 0.043 Å^-1 D1–D2 splitting), the surface reference is not a valid zero-strain state; the paper should re-report strains relative to a buried reference or as registry-induced differences. If the burial shift is negligible, the original compressive/tensile labels are supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The quantitative strain labels are anchored to qH00 = 0.779 Å^-1 measured on a 1H-terminated surface, where the 1H layer has only one 1T neighbor and a vacuum interface. The buried 1H layers in regions D1 and D2 have two 1T neighbors. The D1–D2 splitting (q0H0 = 0.797 vs 0.754 Å^-1, about 5.4%) is internally robust because it rests on directly observed moiré peaks, not on the surface baseline. However, labeling D1 as compressive (−2.3%) and D2 as tensile (+3.2%) assumes that the surface qH00 equals the zero-force equilibrium for a buried 1H layer. In 4Hb-TaS2, interlayer charge transfer between 1T and 1H layers is known and burial itself could shift the CDW q by an amount comparable to the registry splitting (0.043 Å^-1). If the true buried reference were 0.797 Å^-1, D1 would be unstrained and D2 tensile at −5.4%; if it were lower, both could be compressive. Thus the headline numbers and the sign of the registry-induced strain are not uniquely determined by the data. The elastic model F(q,θ) also uses qH00 as the zero-force minimum, so the extracted λ and K inherit this reference uncertainty. This does not invalidate the central observation that registry changes q, but it means the paper's quantitative claim of a compressive −2.3% and tensile +3.2% pair is conditional on an unverified baseline.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports a low-temperature STM study of 4Hb-TaS2, focusing on the incommensurate CDW on 1H layers that are sandwiched between commensurate 1T-CDW layers. By analyzing large-field-of-view dI/dV maps and their Fourier transforms, the authors resolve two distinct buried-1H-CDW states, D1 and D2, with wave vectors q0H0 = 0.797 Å-1 and 0.754 Å-1. Taking qH00 = 0.779 Å-1 measured on a 1H-terminated surface as the zero-strain reference, they label these compressive (−2.3%) and tensile (+3.2%) elastic states. The two states are correlated with the relative rotational alignment of the 1T CDWs on the two surrounding 1T layers. QPI measurements show flat-band features shifted by a few meV between D1 and D2. A minimal elastic free energy F(q,θ) = (K/2)(q−qH00)2 + λ(θ)(q−qH00) is used to relate the observed q shifts to an interlayer registry force and to estimate K ≈ 10 eVÅ2 and an elastic energy cost of a few meV.","tokens_in":19140,"tokens_out":5514,"duration_ms":55722,"significance":"If the quantitative strain interpretation is accepted, the paper establishes a conceptually important mechanism: the ordering wave vector of an incommensurate CDW behaves as an elastic degree of freedom controlled by interlayer registry. This could explain sample-to-sample variability of CDW wave vectors in layered TMDs and connects stacking to low-energy electronic structure. The central observation—that the buried 1H CDW q-vector changes systematically with the stacking of the adjacent 1T CDWs—is plausible and supported by careful Fourier analysis, moiré simulations, and the use of the lattice-pinned 1T CDW as an internal calibration. The paper also demonstrates a technically strong spectroscopic mapping approach. However, the absolute signs and magnitudes of the strain labels, and the causal interpretation of the flat-band shifts, are less secure than the abstract suggests.","major_comments":[{"comment":"The D1–D2 splitting is internally robust because it is based on directly observed moiré peaks (Δq ≈ 0.043 Å−1, about 5.4%). However, the compressive (−2.3%) and tensile (+3.2%) labels are anchored to qH00 measured on a 1H-terminated surface, where the 1H layer has one 1T neighbor and a vacuum interface. The buried 1H layers in D1 and D2 have two 1T neighbors and experience charge transfer; no unstrained buried reference is measured. If the true zero-force q for a buried 1H layer were 0.797 Å−1, D1 would be unstrained and D2 would be tensile at +5.4%; if it were lower than 0.754 Å−1, both states would be compressive. Thus the sign and magnitude of the reported strains are conditional on an unverified baseline. The paper should either supply a buried reference (e.g., via DFT or a different experimental geometry) or reframe the quantitative claim as a registry-induced q-splitting without ab","section":"Results, Fig. 2 and Table S2"},{"comment":"The estimate of the elastic parameters is not an independent test of the model. In the Discussion, λ is approximated as ΔE/Δq using the measured flat-band shift ΔE ≈ 5 meV and the measured wave-vector change Δq ≈ 0.02 Å−1; K is then obtained from the minimization equation using the same Δq. The resulting elastic energy cost ΔF ∼ K(Δq)2/2 ≈ few meV is therefore a restatement of the two input measurements, not a verification of the elastic free-energy form. The claim that the energy cost is 'few meV' is thus not independently established. Please label this as a consistency estimate and provide, if possible, an independent bound on K (e.g., from DFT or from the curvature of the CDW free energy).","section":"Discussion, elastic model"},{"comment":"The abstract states that the few-meV flat-band shifts demonstrate that interlayer interactions reshape the electronic structure 'through the intrinsic elasticity of the incommensurate CDW.' However, the measurements show only that D1 and D2 differ in both the CDW q-vector and the flat-band position. The Discussion itself attributes the flat-band energy to interlayer charge transfer, which also differs between the stacking configurations. The flat-band shift is therefore not uniquely attributable to the CDW strain; it could be a direct consequence of the different 1T/1H/1T stacking and charge transfer. To support the causal claim, one would need to vary q while holding the stacking/charge environment fixed, or show DFT results with the measured q values.","section":"Abstract and Discussion, Fig. 3"}],"minor_comments":[{"comment":"The phrase 'with a mean wave vector close to 2.8×2.8' lacks units; presumably Å−1 or dimensionless q/q_at. Please clarify.","section":"Introduction"},{"comment":"The super-moiré vector qTH0-T0T and the inset in panel (f) are described only in the caption; a short explanation in the main text would aid readability.","section":"Results, Fig. 2 caption"},{"comment":"The real-space lattice expression in the Methods uses f_α = 1/9 + 8/9 ∏ cos(q_α,i r), while Section S5 writes f_i = 1/9 + 8/9 ∏ cos(1/2 k_i r). These are equivalent only if q = k/2; please make the notation consistent.","section":"Methods and Section S5"},{"comment":"Typo: 'a board peak' should be 'a broad peak'.","section":"Fig. 3(h)"},{"comment":"References 45 and 48 appear to duplicate the same arXiv preprint; consider consolidating. Several 2026 in-press references (e.g., refs 27, 44) would benefit from DOI/arXiv identifiers.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The core observation—registry-dependent q-splitting of the buried 1H CDW—is solid and publishable. The paper's current framing overstates certainty in the absolute strain signs/magnitudes and the causal role of CDW elasticity in the flat-band shift. These are fixable with careful reframing and additional analysis, so I recommend major revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The useful core: this paper shows, with high-quality large-field STM and moiré simulations, that the incommensurate CDW on the buried 1H layers in 4Hb-TaS2 has two distinct wave vectors, q = 0.797 Å^-1 and 0.754 Å^-1, depending on whether the surrounding 1T CDWs are aligned or rotated by 27.8°. The D1–D2 difference (~5%) is read off directly from different moiré peaks, not from any model, and the QPI data add correlated flat-band shifts of a few meV. That is a real, reproducible observation and a plausible resolution of why CDW q-vectors vary between nominally identical samples.\n\nWhere I'd push back: the paper calls D1 'compressive' (−2.3%) and D2 'tensile' (+3.2%) by taking the q-vector on a 1H-terminated surface as the zero-strain reference. But a surface 1H layer has one 1T neighbor and vacuum; the buried 1H layers in D1/D2 have two 1T neighbors. Burial itself, via charge transfer or the extra interlayer interaction, could shift q by an amount comparable to the D1–D2 splitting. The paper never measures an unstrained buried 1H CDW. So the sign and magnitude of the registry-induced strain are conditional. This doesn't kill the central claim—registry changes q—but it does mean the headline compressive/tensile numbers are not secure. The elastic model compounds this: qH00 is the zero-force minimum, and the stiffness K and force λ are estimated from the same Δq and ΔE they are supposed to explain. That's a minor-to-moderate circularity, not a fatal one. The flat-band shifts also come without error bars, which is a small gap given they carry the energy scale.\n\nWhat's solid: the internal assignment of the two moiré patterns, the use of the commensurate 1T CDW as a geometric reference, the simulations reproducing higher-order peaks, and the transverse electron diffraction showing the two 1T CDW orientations. The paper is honest about the phenomenology and doesn't oversell the model's predictive power.\n\nWho this is for: anyone working on TaS2 polytypes, CDW elasticity, or stacking-dependent electronic order in van der Waals heterostructures. It deserves a serious referee: the observation is novel and likely important, but the reference-state question needs to be addressed directly, either by measuring an unstrained buried 1H layer or by reframing the claims as registry-induced differences without absolute strain labels.\n\nMy recommendation: send it to peer review, with the baseline issue as the main request for revision.","headline":"Careful STM work showing the buried 1H CDW in 4Hb-TaS2 takes two discrete q-vectors that track the stacking of the adjacent 1T CDWs; the absolute 'compressive/tensile' labels rest on a surface reference the paper never validates.","tokens_in":19763,"tokens_out":2584,"would_cite":true,"duration_ms":23448,"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":"In 4Hb-TaS2, the incommensurate charge-density wave on 1H layers acts as an elastic degree of freedom: its wave vector compresses by 2.3% or expands by 3.2% depending on the stacking registry of the surrounding 1T CDW layers.","keywords":["charge density wave","4Hb-TaS2","electronic elasticity","interlayer registry","scanning tunneling microscopy","moiré pattern","flat band","van der Waals heterostructures"],"falsifier":"Grow or pattern 1T/1H/1T stacks where the relative orientation of the two 1T CDWs is controlled (aligned vs. 27.8°-rotated) while keeping all other conditions identical, and measure the 1H CDW q-vector with the same STM method. If the q-vector does not switch between the two elastic states, or if the same q difference appears in regions with unchanged registry, the central claim fails. A simpler check: compare a buried 1H layer in a region where the two 1T CDWs are aligned but the interlayer distance is modified; the flat-band shift should track the CDW strain if the proposed elastic mechanism","tokens_in":18651,"feed_emoji":"📏","tokens_out":6928,"duration_ms":63577,"temperature":0.7,"pith_summary":"This paper argues that the ordering wave vector of an incommensurate charge-density wave is not a fixed material constant but an elastic, soft degree of freedom that responds to the stacking arrangement of neighboring layers. Using 4Hb-TaS2, where 1H layers with an incommensurate CDW sit between 1T layers with a lattice-locked CDW, the authors find that the 1H CDW compresses by 2.3% when the surrounding CDWs are aligned and expands by 3.2% when they are rotated by 27.8°. These elastic states come with few-meV shifts of a flat band, showing that weak interlayer interactions can reshape electronic structure through CDW elasticity. If right, this explains why nominally identical samples report different CDW wave vectors and points to a general mechanism by which stacking controls correlated electronic phases in layered materials.","feed_headline":"Stacking registry stretches and compresses a CDW by 2-3 percent","feed_subtitle":"In 4Hb-TaS2, the ordering wave vector of an incommensurate CDW flexes between two elastic states and shifts a flat band by a few meV.","key_machinery":"The argument rests on three tools: the lattice-locked √13×√13 CDW of the 1T layers, used as a rigid in-situ reference for measuring the buried 1H CDW q-vector; Fourier analysis of large-field STM conductance maps, with moiré-pattern simulations that reproduce the observed peaks and fix the strained q-values; and a minimal free-energy model F(q,θ) = (K/2)(q−qH00)² + λ(θ)(q−qH00), whose minimization gives (q−qH00)/qH00 = −λ/(K qH00). The model turns the registry-dependent force λ(θ) into a measurable strain and yields an estimated stiffness K ≈ 10 eVŲ, making a 2–3% deformation cost only a few meV.","core_discovery":"The paper's central claim is that the ordering wave vector of an incommensurate CDW is an intrinsic elastic variable, not a fixed material constant. In 4Hb-TaS2, where 1H layers host an incommensurate CDW and 1T layers host a commensurate √13×√13 CDW, the authors use the 1T CDW as a rigid internal reference and measure the q-vector of the 1H CDW in two buried configurations. When the two surrounding 1T CDWs are aligned, the 1H CDW is compressed by 2.3% relative to a 1H-terminated surface; when rotated by 27.8°, it is expanded by 3.2%. Corresponding flat dispersions sit at 14 mV and 5 mV in the compressed region and at 4 mV and 0 mV in the tensile region, i.e., a few-meV shift toward the Ferm","pith_inferences":["If the elasticity picture holds, controlled twist-angle experiments in artificial 1T/1H/1T stacks should reveal intermediate elastic states and map the registry force λ(θ) as a continuous function, not just two discrete points.","The registry-dependent elastic energy is a plausible microscopic channel connecting stacking to the pressure dependence of superconductivity in 4Hb-TaS2; the paper suggests but does not prove this link.","The same analysis could be applied to the scattered q-values reported in 4Hb-TaSe2 and other natural heterostructures, predicting that registry variations account for at least part of that scatter."],"forward_implications":["Sample-to-sample scatter in incommensurate CDW wave vectors can be intrinsic, set by stacking registry, rather than only by disorder or external strain.","The incommensurate CDW q-vector can serve as a local, surface-sensitive probe of interlayer registry in van der Waals stacks.","Registry-controlled CDW elasticity directly shifts low-energy electronic states, demonstrated here by the few-meV flat-band shift between the two elastic states.","The estimated stiffness K ≈ 10 eVŲ means 2–3% CDW deformations cost only a few meV, so comparable registry-induced q-shifts should be expected in other layered CDW systems.","The framework unifies previously reported thickness-, pressure-, and strain-dependent CDW periodicities in NbSe2 and related dichalcogenides."],"fun_headline_variants":["CDW's wave vector bends with stacking registry","Interlayer registry elastically alters CDW and flat bands","In TaS2, stacking alignment twists CDW's q-vector","Elastic CDW: few-meV band shifts from layer registry","Stacking registry flexes CDW order by up to 3.2%"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The load-bearing premise is that the CDW q-vector on a 1H-terminated surface is the unstrained reference; if simply sandwiching a 1H layer between two 1T layers shifts its q-vector through charge transfer or burial effects, the quoted −2.3% and +3.2% strains would need renormalizing.","fun_headline_variants_meta":{"raw":{"variants":["CDW's wave vector bends with stacking registry","Interlayer registry elastically alters CDW and flat bands","In TaS2, stacking alignment twists CDW's q-vector","Elastic CDW: few-meV band shifts from layer registry","Stacking registry flexes CDW order by up to 3.2%"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000796,"raw_usage":{"total_tokens":3379,"prompt_tokens":821,"completion_tokens":2558,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":565,"completion_tokens_details":{"reasoning_tokens":2469}},"tokens_in":565,"tokens_out":2558,"duration_ms":18311,"temperature":1.0,"reasoning_tokens":2469,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T07:19:05.736690+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Grow or pattern 1T/1H/1T stacks where the relative orientation of the two 1T CDWs is controlled (aligned vs. 27.8°-rotated) while keeping all other conditions identical, and measure the 1H CDW q-vector with the same STM method. If the q-vector does not switch between the two elastic states, or if the same q difference appears in regions with unchanged registry, the central claim fails. A simpler check: compare a buried 1H layer in a region where the two 1T CDWs are aligned but the interlayer distance is modified; the flat-band shift should track the CDW strain if the proposed elastic mechanism","supporting_citations":[],"review_version":1}