{"id":"8c7a85b5-646c-4d42-9c43-048d09cfa9ea","arxiv_id":"2607.24087","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Bilayer 1T-TaSe2 with A-C' stacking forms a charge-ordered stripe phase driven by interlayer Coulomb repulsion, producing an electronically generated out-of-plane ferroelectric polarization.","lead":"Bilayer 1T-TaSe2 in one stacking geometry develops an electric polarization from electron-electron repulsion alone, without atomic distortion. The result suggests a new family of electronic ferroelectrics in van der Waals materials.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Claimed zero-field ferroelectricity is contradicted by the paper's own phase classification: the observed A-C' stripe phase is antiferroelectric, and the ferroelectric phase is only field-induced.","rationale":"The reader's weakest assumption focused on whether the dI/dV stripe contrast and setpoint-induced change are electronic in origin rather than tip artifacts. That is a valid experimental concern, but the more fundamental issue is that the paper's own model and language place the zero-field A-C' ground state in an antiferroelectric striped phase, not a ferroelectric one. The abstract and title claim ferroelectricity as an emergent zero-field property, yet the only zero-field state observed is antiferroelectric, and the ferroelectric state appears only under an applied field. This is a load-bearing internal inconsistency: even if all experimental artifacts are excluded, the central claim as written is not established. The paper still has substantial value as evidence for an interaction-driven antiferroelectric CDW and a field-induced ferroelectric transition, so outright rejection is too strong; however, acceptance should be conditional on reframing the central claim or providing direct zero-field evidence of net polarization. The reader did not explicitly flag the antiferroelectric/ferroelectric mismatch, so agreement is only partial.","tokens_in":12612,"tokens_out":6929,"duration_ms":64400,"concrete_test":"Compute the net layer-resolved charge imbalance of the A-C' mean-field ground state at the experimental parameter point (t⊥=40 meV, V=60 meV, U=100 meV) by summing the staggered occupation difference over one full unit cell. If the sum vanishes, the zero-field ground state has zero net polarization, confirming that the observed stripe phase is antiferroelectric and the abstract's spontaneous ferroelectric polarization claim is not supported.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim—'spontaneous charge imbalance between layers that gives rise to an out-of-plane ferroelectric polarization' (Abstract)—is not supported by the paper's own results. In the Results, the stripe CDW phase observed in the A-C' bilayer is identified with the TB mean-field phase in which the 'charge transfer produces an out-of-plane polarization whose direction alternates between adjacent sites, indicating that the stripe CDW is an antiferroelectric phase.' Thus the zero-field ground state has no net polarization. The uniform ferroelectric phase is described as higher in energy than the striped antiferroelectric phase in the experimental regime; it is stabilized only by an applied electric field (computed switching at E_z = 0.058 mV/Å, Fig. 5a). The experimental 'switching' in Fig. 5e is a single high-setpoint STM image showing loss of stripe contrast; no hysteresis, reversibility, or opposite-polarity switching is shown. A ferroelectric requires a spontaneous, switchable polarization in zero field. At best, the data support an interaction-driven antiferroelectric ground state with a field-induced transition to a ferroelectric phase; they do not establish 'interaction-driven electronic ferroelectricity' as stated in the title and abstract. This is not an external-consensus dispute but an internal inconsistency with the paper's own phase classification.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports an STM/STS and DFT-based study of two stackings of bilayer 1T-TaSe2, A-C and A-C′. The authors argue that A-C stacking yields a dimerized band insulator, while A-C′ stacking is governed by interlayer Coulomb interactions, producing a striped CDW phase with charge imbalance between layers. They further claim that this charge imbalance gives rise to out-of-plane ferroelectric polarization and that a perpendicular electric field can switch between antiferroelectric and ferroelectric interchain configurations. The central claim is that this constitutes interaction-driven electronic ferroelectricity in a van der Waals bilayer.","tokens_in":13004,"tokens_out":4460,"duration_ms":39661,"significance":"If the central claim were correct, this would establish a new mechanism for ferroelectricity driven purely by electronic correlations, distinct from lattice-driven ferroelectricity. The paper combines careful MBE growth, atomically resolved STM/STS, and a tractable mean-field tight-binding model with fitted parameters. The identification of two distinct stacking-dependent electronic phases is interesting, and the field-induced transition between stripe and uniform charge orders is a plausible theoretical prediction. However, the central claim is undermined by the paper's own phase classification, which identifies the zero-field stripe CDW as antiferroelectric, not ferroelectric. The presented experimental switching evidence is also limited to a single high-current image, without hysteresis or polarization reversal. Thus, while the work presents promising data and modeling, the headline claim of ferroelectricity is not supported as stated.","major_comments":[{"comment":"The abstract and title claim that the A-C′ ground state has a spontaneous charge imbalance giving rise to an out-of-plane ferroelectric polarization. This is contradicted by the authors' own results: in the Results section, they state that the stripe CDW has a polarization 'whose direction alternates between adjacent sites, indicating that the stripe CDW is an antiferroelectric phase.' They also state that the striped antiferroelectric phase is 'narrowly favored by the model' over the uniform ferroelectric phase at zero field. Thus the zero-field ground state has no net polarization, and the ferroelectric phase only appears under an applied field. The phrasing in the abstract and conclusions should be corrected to antiferroelectricity with field-induced ferroelectricity, or the authors must provide direct evidence of a spontaneous net out-of-plane polarization in zero field.","section":"Results, 'Ferroelectric CDW in the A-C′ bilayer' and Fig. 5"},{"comment":"The experimental claim of electrical switching is based on a single STM image at a high current setpoint (Fig. 5e) that shows loss of stripe contrast. No hysteresis loop, no switching back to the stripe phase by reducing the field, and no reversal to the opposite ferroelectric polarization are shown. Such behavior is essential to establish field-induced ferroelectric switching. The observed contrast change could result from a tip-induced structural modification, electronic effect, or a change in tunneling conditions rather than a true polarization switch. Reversible and repeatable switching with opposite bias polarities should be demonstrated.","section":"Methods and Fig. 2f-h"},{"comment":"The interlayer Coulomb interaction V is set to 60 meV in the 'experimental region' without an independent determination from DFT or from the experimentally measured gap. Since the ferroelectric CDW phase appears only for sufficiently large V relative to the hopping t⊥, choosing V=60 meV to place the model in the experimental region makes the theoretical prediction partially circular. An estimate of V from first principles (e.g., constrained RPA or DFT-based screening) would be needed to validate that the observed stripe CDW is genuinely interaction-driven rather than an artifact of the chosen parameter.","section":"Results, 'Ferroelectric CDW in the A-C′ bilayer'"}],"minor_comments":[{"comment":"The schematic labels the A-C′ configuration as 'ferroelectric CDW phase' while the text later identifies the actual zero-field ground state as antiferroelectric. The label should be changed to 'antiferroelectric stripe CDW' or 'ferroelectric/antiferroelectric CDW' and should distinguish the zero-field and field-induced states.","section":"Conclusions"},{"comment":"The phrase 'ferroelectric alternating polarization' is internally contradictory; the authors likely mean 'antiferroelectric' or 'alternating polarization with no net polarization.' Please choose terminology consistently.","section":"Methods, modeling of external field"},{"comment":"The field strength at the STM tip is not quantified. The statement that changing the current setpoint from 5 pA to 1 nA changes the electric field should be supported by estimates of the tip-sample distance change or by a separate calibration; otherwise the interpretation of Fig. 5e as field-induced switching is not unique.","section":"Methods, mean-field calculations"},{"comment":"The mean-field decoupling includes Hartree and Fock terms, but the notation is compact. It would help to specify which terms are kept and how the self-consistency is initialized (e.g., random vs. ordered), because the coexistence of ferroelectric and antiferroelectric solutions suggests a nontrivial energy landscape.","section":"DFT methods"},{"comment":"The DFT calculations use PBE without spin-orbit coupling. For 1T-TaSe2, spin-orbit coupling can affect the band structure near the Fermi level. A statement justifying this choice or showing its effect would strengthen the model fitting.","section":"Fig. 2"},{"comment":"In the phase diagrams (panels e-f), the experimental region is marked but the boundaries between phases are not explicitly defined (e.g., order parameter values). Adding the order parameter plots from the SI would make the phase assignment more transparent.","section":"Data availability"},{"comment":"The data availability statement says data are available 'upon request.' In light of the strong claims of ferroelectricity, the authors should consider depositing the STM data and the tight-binding model in a public repository to enable independent verification.","section":"Introduction"}],"recommendation":"major_revision","confidential_remarks":"The paper's central claim, as stated in the title and abstract, is not supported by the authors' own phase classification, which identifies the zero-field stripe state as antiferroelectric. The field-induced uniform state is interesting but does not constitute spontaneous ferroelectricity. The V=60 meV parameter is not independently justified, and the experimental switching evidence is minimal. These issues are fixable: the authors should reframe the claim to antiferroelectricity plus field-induced ferroelectricity, provide direct evidence of reversible switching, and justify the interaction parameter. The underlying data and modeling are of good quality, but the current framing overstates the result."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The new physics here is the A-C' bilayer stripe phase: a clear, reproducible STM/STS pattern with alternating CDW-row intensity and reversed contrast at opposite bias, backed by a plausible charge-transfer explanation. The A-C bilayer as a dimerized band insulator is also a nice, clean result. The DFT-fitted tight-binding model with mean-field order parameters gives a coherent account of both stackings, and the theoretical prediction of a field-driven transition from an alternating to a uniform charge imbalance is genuinely interesting. The experimental work is careful, the figures are convincing, and the authors know their materials.\n\nThe problem is the headline. The paper's own text says the stripe CDW is an antiferroelectric phase: the polarization direction alternates between adjacent sites, so there is no net zero-field polarization. The uniform ferroelectric phase is higher in energy and is stabilized only by a perpendicular electric field. That makes the title and abstract claim of 'spontaneous charge imbalance that gives rise to an out-of-plane ferroelectric polarization' inaccurate. What is actually demonstrated is interaction-driven antiferroelectricity plus a field-induced ferroelectric state. That is still a worthwhile result, but it should be named correctly.\n\nThe switching evidence is also thin. Figure 5e is one STM image at high setpoint showing loss of stripe contrast. There is no hysteresis loop, no reverse switching, no measurement of polarization, and the possibility of a tip-induced electronic or mechanical artifact is not excluded. The model's V=60 meV is placed in the 'experimental region' rather than independently determined, so the phase diagram is consistent with the data but not a sharp prediction. No code or raw data are deposited, which makes it harder to check the model details.\n\nThese are correctable issues. The experimental observation of the stripe phase and the band-insulator versus charge-transfer distinction are solid. The paper deserves a serious referee, but the authors should be pushed to reframe the central claim, strengthen the switching evidence, and justify V more carefully. I would probably cite the stripe-phase observation and the AFE/FE competition if I worked in this area, but not the 'electronic ferroelectricity' claim as it currently stands.","headline":"The stripe-CDW observation and the A-C band-insulator result are likely real, but the 'ferroelectricity' in the title and abstract overstates what the paper actually shows: the zero-field ground state is antiferroelectric, and the field-induced ferroelectric phase is demonstrated only once, without switching back.","tokens_in":13455,"tokens_out":2820,"would_cite":true,"duration_ms":27449,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["71.27.+a","77.80.-e","68.37.Ef"],"model":"deepseek-v4-flash","headline":"This paper claims that in the A-C′ stacking of bilayer 1T-TaSe2, interlayer Coulomb interactions drive a spontaneous charge imbalance between layers and thereby an out-of-plane electronic ferroelectric polarization, distinct from lattice-dr","keywords":["electronic ferroelectricity","charge density wave","1T-TaSe2","van der Waals heterostructures","scanning tunneling microscopy","interlayer Coulomb interaction","mean-field tight-binding model","ferroelectric switching"],"falsifier":"Measure the same A-C′ bilayer with a probe that directly senses interlayer charge, such as out-of-plane piezoresponse or core-level spectroscopy, at the stripe periodicity; or sweep the tip field continuously while tracking stripe contrast. If stripe contrast persists unchanged, or if the supposedly switched state shows a lattice reconstruction rather than an electronic polarization reversal, the central claim fails.","tokens_in":12529,"feed_emoji":"⚡","tokens_out":3631,"duration_ms":31507,"temperature":0.7,"pith_summary":"The paper tries to establish that ferroelectric order can emerge purely from electron-electron interactions, without structural inversion-symmetry breaking, in a van der Waals bilayer. It grows bilayer 1T-TaSe2 in two stackings, A-C and A-C′, which both form quasi-1D interacting chains from coupled Star-of-David charge-density-wave units. Model and experiment show A-C is a dimerized band insulator, while A-C′ is dominated by interlayer Coulomb interaction V, producing a nearly complete interlayer charge transfer and an out-of-plane polarization. The paper further shows that antiferroelectric striped and ferroelectric uniform configurations can be switched by an electric field, making the polar state electrically tunable. If right, this adds a mechanism—interaction-driven electronic ferroelectricity—to the known lattice- and stacking-driven ferroelectric phases.","feed_headline":"Electrons alone make bilayer TaSe2 ferroelectric","feed_subtitle":"A stacking change makes interlayer Coulomb repulsion shift charge between layers, creating a switchable out-of-plane polarization.","key_machinery":"A minimal mean-field tight-binding model of coupled quasi-1D chains, with one orbital per Star-of-David cluster. Intralayer hopping t≈3 meV, interlayer hopping t⊥≈35–40 meV, with dimerization asymmetry δt⊥=t⊥/2 in A-C and δt⊥=0 in A-C′, on-site U=100 meV, and interlayer nearest-neighbor Coulomb V≈60 meV in the experimental regime. The competition between interlayer hybridization, which dimerizes and opens a band gap, and interlayer Coulomb interaction V, which induces interlayer charge transfer, is what separates the two phases; V is the term that carries the ferroelectric order.","core_discovery":"The central claim, argued jointly from STM/STS and ab initio plus mean-field tight-binding calculations, is that the A-C′ stacking of bilayer 1T-TaSe2 hosts a correlated electronic ground state in which interlayer Coulomb repulsion V transfers close to one electron between Star-of-David units in adjacent layers. The resulting charge imbalance between layers is an antiferroelectric striped CDW that, under an out-of-plane field, converts to a uniform ferroelectric state with net polarization; the estimated switching field is 0.058 mV/Å. This is electronic ferroelectricity: the polarization comes from a charge-ordered electronic state, not from ionic displacements, interlayer sliding, or moiré","pith_inferences":["If the mechanism generalizes, other 1T TMD bilayers with appropriate stacking—such as TaS2 or NbSe2—could show interaction-driven ferroelectricity, and multilayer stacking sequences might yield layered polar textures.","Because the polarization is electronic rather than ionic, switching could in principle be faster or lower-energy than in displacive ferroelectrics, though this is not tested in the paper.","The nearly full charge transfer suggests a description as a charge-transfer ferroelectric; a testable extension is to measure the interlayer electric field directly via core-level shifts or optical second-harmonic generation while sweeping the tip field.","The model's predicted switching field could be compared with a measured coercive field from transport or piezoresponse experiments; any large discrepancy would tighten or refute the model parameters."],"forward_implications":["Bilayer 1T-TaSe2 in the A-C′ stacking becomes a platform for electrically switchable correlated polar states.","Ferroelectricity can be stabilized without lattice distortion or sliding, purely by electronic correlations in a flat-band system.","Stacking engineering selects between two distinct ground states: a dimerized band insulator (A-C) and a ferroelectric CDW (A-C′).","Antiferroelectric striped and ferroelectric uniform configurations have close energies, with a computable switching field of about 0.058 mV/Å.","The discovered phase extends the phase diagram of layered 1T transition-metal dichalcogenides, implying similar phases in related Star-of-David multilayers."],"fun_headline_variants":["TaSe2 bilayer turns ferroelectric purely from electron interactions","Interlayer electron repulsion gives TaSe2 a switchable polarization","Electronic ferroelectricity without lattice distortion in TaSe2 bilayer","Charge repulsion alone makes bilayer TaSe2 ferroelectric","Stacking A-C' turns TaSe2 into an electronic ferroelectric"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The claim stands on the interpretation that the striped STM contrast reflects interlayer charge transfer (modeled with V=60 meV) rather than a structural reconstruction or tip artifact, and that the setpoint-induced contrast change is true electric-field switching.","fun_headline_variants_meta":{"raw":{"variants":["TaSe2 bilayer turns ferroelectric purely from electron interactions","Interlayer electron repulsion gives TaSe2 a switchable polarization","Electronic ferroelectricity without lattice distortion in TaSe2 bilayer","Charge repulsion alone makes bilayer TaSe2 ferroelectric","Stacking A-C' turns TaSe2 into an electronic ferroelectric"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00073,"raw_usage":{"total_tokens":3127,"prompt_tokens":788,"completion_tokens":2339,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":532,"completion_tokens_details":{"reasoning_tokens":2248}},"tokens_in":532,"tokens_out":2339,"duration_ms":16394,"temperature":1.0,"reasoning_tokens":2248,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-31T23:02:58.052880+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the same A-C′ bilayer with a probe that directly senses interlayer charge, such as out-of-plane piezoresponse or core-level spectroscopy, at the stripe periodicity; or sweep the tip field continuously while tracking stripe contrast. If stripe contrast persists unchanged, or if the supposedly switched state shows a lattice reconstruction rather than an electronic polarization reversal, the central claim fails.","supporting_citations":[],"review_version":1}