{"id":"c7bd56c3-b155-4fe6-81c2-4cda3b41f388","arxiv_id":"2507.00745","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Graphene/hexagonal-boron-nitride moiré stacks show ferroelectric polarization that magnetic fields enhance up to room temperature, with purely electronic origin, and this polarization can suppress Shubnikov-de Haas oscillations.","lead":"Researchers found that stacking graphene and hexagonal boron nitride in twisted layers creates a ferroelectric that gets stronger when a magnetic field is applied, even at room temperature and without any magnetic atoms. The effect could lead to two-dimensional memory and logic devices whose state is controlled by both electric and magnetic fields.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Polarization is inferred from the same transport hysteresis that magnetic field is claimed to modulate; without a no-moiré control or direct probe, the B-enhancement of P2D is not distinguished from a magnetoresistance or charge-trapping artifact.","rationale":"The reader's weakest_assumption is exactly the point on which the central claim rests: the conversion of a sweep-direction-dependent CNP shift into a polarization, with no independent verification. I read the paper in good faith: it reports coherent, reproducible transport data across several device geometries, including room-temperature persistence, and it cites prior ferroelectric moiré literature consistent with its zero-field interpretation. Those strengthen the case that some hysteresis is present. However, the magnetic-field enhancement—the paper's distinctive claim—is specifically vulnerable because the same Rxx maps that define the polarization also show the SdH suppression and quantum-Hall changes attributed to that polarization. This creates a circularity: the 'polarization' is not measured by a thermodynamic or structural probe, but by the very transport features it is invoked to explain. The only way to break the circle is a negative control without a moiré superlattice, or a direct polarization probe (e.g., PFM, SHG, or displacement-current measurement) under magnetic field. The DFT support is also under-specified—'applying B by including SOC' is not a standard magnetic-field treatment and would itself need scrutiny—but the experimental extraction is more load-bearing because it is the sole quantitative basis for the headline effect. The recommended verdict is unchanged from the reader's CONDITIONAL: the paper is promising but should not be accepted as established until the artifact alternative is falsified. I agree with the reader's identification of the weakest assumption.","tokens_in":21352,"tokens_out":3740,"duration_ms":51451,"concrete_test":"Fabricate a control device with the same graphite/hBN gate stack and same graphene layer(s) but with the graphene/hBN interface deliberately misaligned by about 30° so no moiré superlattice forms. Run the identical forward/backward V_b sweep protocol at B = 0, 0.5, 5.6, and 13.5 T, and extract Δn_H from the CNP line positions using the same peak-finding and capacitance-calibration algorithm. If Δn_H remains zero or B-independent in the misaligned control while it grows with B in the aligned devices, the ferroelectric interpretation survives the artifact check; if the control reproduces the B-dependent Δn_H, the claimed B-enhanced polarization is an artifact of the transport-based extraction.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central observable is the sweep-direction-dependent charge-neutrality-point shift, converted to P2D = eΔn_H d in Fig. 2c,d and Methods ('Determination of gate capacitance'). At B = 13.5 T, the Rxx maps used to extract the CNP line are themselves dominated by SdH oscillations and quantum Hall transport, and the paper later attributes suppression of SdH oscillations and changes in Gxy to the same polarization. The quantity used to define the polarization and the effect used to demonstrate its consequence are therefore not independent: the 'B-enhanced polarization' is read off the same resistance features it is claimed to explain. The gate-density conversion n = (C_b V_b + C_t V_t)/e assumes a rigid parallel-plate capacitor with no B-dependent trapping or contact effects. Reference 24 documents an anomalous hysteresis in graphite/hBN transistors without a moiré superlattice, explicitly offering a non-ferroelectric origin for similar hysteresis. The paper's multi-device reproducibility is not a negative control, because all devices share the moiré superlattice that also changes the band structure and magnetoresistance; no misaligned (non-moiré) stack or direct polarization probe is presented. The self-stated limitation that 'the precise link between magneto-transport and ferroelectric polarization remains unclear' further underscores this gap. If the B-dependent CNP shift is a charge-trapping or magnetoresistance artifact, the headline claim of magnetic-field-enhanced ferroelectricity collapses, while the room-temperature persistence and the SdH suppression would be artifacts of the same extraction.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports magneto-transport experiments on graphene/hBN moiré devices, including an ABA-trilayer-graphene/hBN moiré superlattice and an MLG-intercalated “across-layer” version. At zero magnetic field, the authors observe sweep-direction-dependent resistance hysteresis, which they attribute to interfacial ferroelectricity. They then claim that a perpendicular magnetic field enhances the ferroelectric polarization up to room temperature, that this enhancement is purely electronic, and that the polarized state suppresses Shubnikov–de Haas oscillations and modifies quantum Hall conductance. The central observable throughout is the sweep-direction-dependent shift of the charge neutrality point, converted to a polarization through P2D = eΔnH d.","tokens_in":21610,"tokens_out":6395,"duration_ms":75980,"significance":"If correct, the result would be a striking example of magnetoelectric coupling in non-magnetic moiré materials and would establish ferroelectric polarization as a switch for quantum transport. The manuscript has genuine strengths: the zero-field hysteresis is observed in several devices with different layer configurations, the temperature dependence in Fig. 3b extends to 275 K, and the authors attempt to separate ionic and electronic contributions with DFT. However, the key quantity P2D is not measured directly, and the paper does not exclude an artifact of the same transport hysteresis used to define the polarization. The significance is therefore conditional on new control experiments.","major_comments":[{"comment":"The central quantity P2D is obtained from the difference in charge-neutrality-point (CNP) positions between forward and backward sweeps under magnetic field via P2D = eΔnH d, assuming a rigid parallel-plate capacitor with d = 0.34 nm. At B = 13.5 T, the Rxx maps in Fig. 2a,b that are used to locate the CNPs are themselves dominated by Shubnikov–de Haas oscillations and quantum Hall features, so the extracted shift may reflect magnetoresistance or field-dependent charge trapping rather than a change in electric polarization. No control device without a moiré superlattice is presented; all devices share the moiré pattern, and multi-device reproducibility therefore does not rule out a common artifact. Reference 24 (Waters et al.) documents exactly such anomalous hysteresis in graphite/hBN transistors without a moiré lattice and is cited but not addressed. The authors should provide a non-moiré control, a direct polarization probe, or quantitative estimates of the proposed artifact; without this, the B-enhancement claim is not established.","section":"Magnetic field-enhanced ferroelectricity / Methods: Determination of gate capacitance"},{"comment":"The polarized (P) and nonpolarized (N) phases in Fig. 4a-d are identified from the same Rxx(Vb/db, Vt/dt) maps in which SdH suppression is later claimed. The phase label and the effect are therefore not independent: the quantity used to define the polarization is read off the same resistance features it is said to explain. The authors themselves state in the Results that “the precise link between magneto-transport and ferroelectric polarization remains unclear and demands further theoretical and microscopic investigation.” An independent determination of the polarization state (for example, from a simultaneous capacitance, optical, or scanning-probe measurement) is needed before concluding that polarization gates quantum oscillations.","section":"Effects of ferroelectric polarization on magneto-transport"},{"comment":"The DFT description says that “The magnetic field was applied self-consistently by including SOC in the calculations.” This is not a transparent representation of an external magnetic field: spin-orbit coupling produces spin splitting but not the orbital effects of a perpendicular magnetic field. Please specify whether a Zeeman term, an orbital Peierls phase, or a spin-dependent potential was used, and justify why this captures the experimental field range. The theoretical claim of a purely electronic B-enhanced polarization (Supplementary Fig. S29) depends on this point.","section":"Methods: Computational methods"},{"comment":"The Methods state that the device is “assumed to be a series of parallel-plate capacitors” and that the graphene multilayers are “treated as a single charge layer.” In the intercalated devices, however, the MLG is described as decoupled from the TLG (Supplementary Fig. S10e), so a single-layer capacitance model may misassign gate-voltage changes between two independent carrier systems. This affects the quantitative P2D values in the across-layer devices and should be justified or replaced by a two-carrier analysis.","section":"Methods: Determination of gate capacitance"}],"minor_comments":[{"comment":"The abbreviations “P phase” and “N phase” are used in the text and figure before being explicitly defined; please define them (polarized and nonpolarized) in the main text or caption at first use.","section":"Fig. 4"},{"comment":"The inset showing dP2D/dB as a function of temperature would benefit from a clear statement of the fitting range and from error bars on the slopes, given that the P2D(B) curves are presented without uncertainties.","section":"Fig. 3b"},{"comment":"Several figure captions contain garbled unicode placeholders (e.g., strings starting with “/uni000000…”) that should be cleaned before publication.","section":"Figure captions"},{"comment":"The sentence in the Results that “the precise link between magneto-transport and ferroelectric polarization remains unclear” is an important caveat; please expand it in the Conclusions to specify which missing measurement or theory would establish the proposed link.","section":"Results / Conclusion"}],"recommendation":"major_revision","confidential_remarks":"The main technical concern is the artifact alternative for the B-enhanced polarization: the paper cites Waters et al. (ref. 24) but does not address its non-ferroelectric hysteresis mechanism. The zero-field ferroelectricity observations are plausible and consistent with the literature, but the headline claim of magnetic-field-enhanced polarization rests on one indirect observable with no negative control. I would support publication after the requested control experiments or an independent polarization probe are added; in its current form the central claim is not fully supported."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The headline result — magnetic-field growth of switchable polarization up to 275 K in a non-magnetic moiré stack — is unusual enough to demand a careful look, and the intercalated across-layer TLG/hBN device is a genuinely new configuration. The multi-device reproducibility and the clean hysteresis at zero field are real strengths. But I side with the stress-test note: the B-enhancement is read off the same CNP shift that later is said to be a consequence of the polarization, and there is no control without a moiré pattern. Ref. 24 (Waters et al.) documents similar-looking hysteresis in graphite/hBN without any moiré, and the authors cite it without refuting it for their geometry. Their own sentence that 'the precise link between magneto-transport and ferroelectric polarization remains unclear' is an honest flag, but it also means the SdH suppression story is not yet tied to a verified polarization state.\n\nThe DFT section does not help as written: 'magnetic field applied self-consistently by including SOC' is not a standard way to include a B field, and without details on how orbital or Zeeman terms were treated, the claimed electronic ME mechanism is not established. The room-temperature persistence is the most interesting observation, and if it survives a non-moiré control and an independent polarization probe (e.g., via capacitance or local probe), it would be a real step. As it stands, the paper is a well-executed transport study with an interpretation that outruns the evidence.\n\nI would send it to referees, but I would tell the authors that the central claim needs a negative control and a direct check that the hysteresis is not B-dependent charge trapping. The across-layer ferroelectricity and the temperature dependence can be salvaged; the B enhancement needs more support.","headline":"Striking B-enhanced ferroelectricity claim in non-magnetic moiré stacks, but the evidence is not yet independent of the transport artifact it is meant to explain.","tokens_in":22269,"tokens_out":2442,"would_cite":false,"duration_ms":28181,"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":"This paper claims that a perpendicular magnetic field increases the switchable electric polarization of non-magnetic graphene/hBN moir\\'e superlattices through a purely electronic mechanism, and that this polarization can suppress…","keywords":["moiré ferroelectricity","graphene/hBN","sliding ferroelectricity","magnetoelectric effect","Shubnikov-de Haas oscillations","quantum Hall effect","two-dimensional ferroelectric","van der Waals heterostructures"],"falsifier":"A graphene/hBN device with no moir\\'e superlattice, measured under the same forward/backward gate sweeps in a perpendicular field, would reveal whether the field-dependent neutrality-point shift is intrinsic to the moir\\'e domains; a direct electrostatic-force-microscopy measurement of $P_{2D}$ at fixed field would settle whether the extracted polarization is real.","tokens_in":21116,"feed_emoji":"🧲","tokens_out":4084,"duration_ms":42833,"temperature":0.7,"pith_summary":"This paper reports that a perpendicular magnetic field strengthens the switchable electric polarization of graphene/hBN moir\\'e superlattices, even though these devices contain no magnetic elements, and that the strengthened polarization can in turn switch off Shubnikov-de Haas oscillations and reshape quantum Hall plateaus. The authors show the effect in several layer configurations, including an across-layer stack where a rotated monolayer graphene is intercalated between the trilayer graphene and hBN, and they report that it persists up to room temperature. The claimed mechanism is purely electronic: density functional theory calculations with fixed ionic positions reproduce a field-dependent polarization, ruling out lattice or magnetostrictive contributions. If correct, this gives a non-magnetic platform for magnetoelectric coupling and a ferroelectric knob for quantum transport.","feed_headline":"Magnetic fields amplify ferroelectricity in non-magnetic stacks","feed_subtitle":"A field-boosted polarization in graphene/hBN moiré devices survives to room temperature and can switch quantum oscillations off.","key_machinery":"The carrier is sliding ferroelectricity in the moir\\'e superlattice: a network of non-centrosymmetric ABCA and ABAC stacking domains in TLG/hBN carries opposite out-of-plane polarization, and an electric field switches between them by interlayer sliding. The observable is the hysteresis of the charge-neutrality point, converted into a two-dimensional polarization $P_{2D}=e\\Delta n_H d$. For the magnetic enhancement, the load-bearing calculation is a first-principles Berry-phase evaluation of electronic polarization with ionic positions frozen at finite magnetic field, which isolates the electronic contribution and shows a much larger $B$-response in the ABAC domain.","core_discovery":"The central claim is that a perpendicular magnetic field enhances the remnant ferroelectric polarization of graphene/hBN moir\\'e superlattices through an electronic, not ionic, response, and that the resulting polarized phase modifies the magneto-transport of the device. In the experiments the charge-neutrality-point shift between forward and backward gate sweeps, converted to $P_{2D}=e\\Delta n_H d$, grows from $-0.05$ to $0.08$ pC/m at fixed displacement field as $B$ rises to 13.5 T, and field-enhanced hysteresis persists up to 275 K. The polarization is found to suppress the Shubnikov-de Haas oscillations of the trilayer graphene and to change the Hall conductance filling factors, with local and nonlocal measurements suggesting unidirectional current propagation along one edge. The authors attribute the effect to the combined breaking of spatial-inversion symmetry by the ferroelectric domains and of time-reversal symmetry by the field, with first-principles calculations showing a linear $B$-dependence of electronic polarization that is an order of magnitude stronger in one stacking domain than the other.","pith_inferences":["Because the effect is claimed to be purely electronic, the field-enhancement rate $dP_{2D}/dB$ should track the moir\\'e band structure; a device series varying the twist angle or interlayer distance could test that scaling.","The edge-asymmetric and $B$-asymmetric SdH suppression resembles chiral edge transport; reversing the current direction in the nonlocal geometry would test whether the unidirectionality follows the current or the field.","No control device without a moir\\'e superlattice is reported, so a key test is to measure an aligned but non-superlattice stack and check that its neutrality-point hysteresis stays $B$-independent."],"forward_implications":["Any graphene/hBN moir\\'e stack with non-centrosymmetric domains should show $B$-enhanced polarization, independent of layer count or intercalation.","The polarized phase acts as a voltage-tunable switch for Shubnikov-de Haas oscillations: oscillations are suppressed in the polarized phase and restored in the non-polarized phase.","Quantum Hall plateaus in the polarized phase can shift from integer to random fractional filling factors, offering a ferroelectric control of quantum Hall states.","The enhancement is not a low-temperature curiosity: it persists up to 275 K, so room-temperature operation is in reach."],"supporting_citations":[{"why":"Supplies the baseline finding of unconventional ferroelectricity in moir\\'e heterostructures that this work extends to intercalated stacks and magnetic fields.","marker":"[6]"},{"why":"Provides the $P_{2D}=e\\Delta n d$ method and the reference magnitude for bilayer graphene moir\\'e polarization.","marker":"[10]"},{"why":"Raises the alternative explanation that gate-sweep hysteresis in graphite/hBN may not be true polarization, which this work must exclude.","marker":"[24]"},{"why":"Gives the two-dimensional ferroelectricity model for binary compound bilayers that underlies the sliding mechanism.","marker":"[35]"},{"why":"Supplies the across-layer sliding ferroelectricity theory used to interpret the intercalated device.","marker":"[36]"},{"why":"Provides the theory that breaking spatial-inversion and time-reversal symmetry together enables a magnetic response of polarization in non-magnetic multilayers.","marker":"[23]"},{"why":"Supplies the concept that the magnetoelectric response can be dominated by electronic rather than ionic contributions.","marker":"[43]"},{"why":"Provides the Berry-phase method used in the DFT calculation of electronic polarization.","marker":"[59]"}],"fun_headline_variants":["Magnetic field boosts ferroelectricity in non-magnetic moiré stacks","Magnetic field amplifies ferroelectricity in non-magnetic moiré","Field-enhanced ferroelectricity in non-magnetic moiré superlattices","Non-magnetic moiré: field boosts polarization, quenches oscillations","A magnetic field turns up ferroelectricity in graphene/hBN stacks"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole claim rests on interpreting the gate-sweep hysteresis of the charge-neutrality point as a true electric polarization of stacking domains, rather than a magnetic-field-dependent charge-trapping or magnetoresistance artifact.","fun_headline_variants_meta":{"raw":{"variants":["Magnetic field boosts ferroelectricity in non-magnetic moiré stacks","Magnetic field amplifies ferroelectricity in non-magnetic moiré","Field-enhanced ferroelectricity in non-magnetic moiré superlattices","Non-magnetic moiré: field boosts polarization, quenches oscillations","A magnetic field turns up ferroelectricity in graphene/hBN stacks"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000474,"raw_usage":{"total_tokens":2367,"prompt_tokens":972,"completion_tokens":1395,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":588,"completion_tokens_details":{"reasoning_tokens":1306}},"tokens_in":588,"tokens_out":1395,"duration_ms":12307,"temperature":1.0,"reasoning_tokens":1306,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T21:08:34.484824+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A graphene/hBN device with no moir\\'e superlattice, measured under the same forward/backward gate sweeps in a perpendicular field, would reveal whether the field-dependent neutrality-point shift is intrinsic to the moir\\'e domains; a direct electrostatic-force-microscopy measurement of $P_{2D}$ at fixed field would settle whether the extracted polarization is real.","supporting_citations":[{"cited_title":"Zheng, Q","cited_arxiv_id":null,"evidence_quote":"Supplies the baseline finding of unconventional ferroelectricity in moir\\'e heterostructures that this work extends to intercalated stacks and magnetic fields."},{"cited_title":"On the origin of anomalous hysteresis in graphite/boron nitride transistors","cited_arxiv_id":"2410.02699","evidence_quote":"Raises the alternative explanation that gate-sweep hysteresis in graphite/hBN may not be true polarization, which this work must exclude."},{"cited_title":"Yang and M","cited_arxiv_id":null,"evidence_quote":"Supplies the across-layer sliding ferroelectricity theory used to interpret the intercalated device."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the theory that breaking spatial-inversion and time-reversal symmetry together enables a magnetic response of polarization in non-magnetic multilayers."},{"cited_title":"Bousquet, N","cited_arxiv_id":null,"evidence_quote":"Supplies the concept that the magnetoelectric response can be dominated by electronic rather than ionic contributions."}],"review_version":1}