{"id":"989aab57-4731-4db7-adc4-0e764e024c4f","arxiv_id":"2506.05548","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The angular alignment between two encapsulating BN layers, not graphene's alignment with BN, controls anomalous gating, with gate ineffectiveness and hysteresis appearing only in a 15-45 degree BN-BN window.","lead":"Rotating the top boron nitride layer relative to the bottom one switches anomalous gating effects in bilayer graphene devices on and off. The effect appears only for a 15 to 45 degree alignment window between the two BN layers, not for graphene/BN alignment, giving experimenters a direct knob for reproducing or suppressing gate ineffectiveness and hysteresis.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Internal evidence of OFF zones at nominally identical BN-BN twist angles undermines the claim that the twist angle alone is the control parameter.","rationale":"The reader's conditional verdict identifies the angle calibration and possible translational sliding as the weakest assumptions. The paper's own discussion of OFF zones inside the active angular window is direct internal evidence that the reported ΘBN is not a sufficient control parameter, which is precisely the load-bearing concern. The 0°/60° ambiguity is real but secondary: it affects the absolute value of the claimed window, not the comparison between ΘBN=30° and ΘBN=90° that establishes the lack of 60° periodicity. The central empirical observation remains plausible and important, but the language of proof and the strong claim of angular control need to be qualified, ideally by correlating each measurement with the rotator's full translational state. Therefore the conditional verdict should stand without escalation.","tokens_in":9442,"tokens_out":8598,"duration_ms":94130,"concrete_test":"Re-examine the sample-1 measurement log behind Fig. 4f and, for every transport map, reconstruct the rotator's full rigid-body state (angle and lateral overlap with the bottom BN) from AFM push records and post-rotation images. Group the maps by nominal ΘBN within the 15°–45° window and compare the lateral-overlap/translation distributions of ON and OFF states. If OFF zones occur at the same inferred ΘBN and the same overlap geometry as ON zones, the angle-only claim is falsified; if OFF zones correspond to a different lateral offset, the conclusion must be revised to 'BN-BN twist plus translational stacking.' A complementary check is to translate the active rotator at fixed ΘBN using a two-point AFM push and test whether the anomalous gating toggles.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central claim is that the angular alignment between the two BN layers is the key parameter governing the anomalous gating, acting as an ON/OFF switch. The text itself, however, reports that while mapping the angular range, \"OFF zones\" (standard gating) appear \"inside this range\" of ΘBN≈30°±15°, and that different types of behavior are found at very similar angles, attributing this to a possible translational sliding mechanism. If the effect can be absent at the same inferred twist angle, then ΘBN is not sufficient to determine the ON/OFF state; a hidden degree of freedom, most plausibly the lateral stacking/translation of the rotator, also matters. The paper later states it has \"proved\" that the BN-BN alignment, not the graphene moiré, governs the effect, which overreaches given this internal admission of non-single-valued behavior. The unresolved 0°/60° convention ambiguity shifts the absolute window (15°–45° vs. 75°–105°) but does not by itself invalidate the non-60°-periodicity observation, since the ΘBN=30° vs. 90° comparison is internally consistent. The load-bearing weakness is sufficiency: the reported angle is a correlate, not a proven sole control parameter.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript reports charge transport measurements on dual-gated, dynamically rotatable bilayer-graphene/BN heterostructures. The authors demonstrate that the relative twist angle between the top and bottom BN flakes acts as an ON/OFF switch for anomalous gating effects (gate ineffectiveness and hysteresis). They classify the observed behaviors into three types, map an angular window in which the effect appears (reported as 15°-45° in an adopted convention), and argue that the effect is governed by BN-BN alignment rather than by the graphene-BN moiré, based on measurements on three samples. They also report a surprising absence of 60° periodicity and suggest a possible 120° periodicity.","tokens_in":9681,"tokens_out":7853,"duration_ms":81357,"significance":"If the central claim is correct, this is an important advance: it identifies a new control parameter for ferroelectric-like anomalous gating in van der Waals heterostructures and provides an experimental protocol for switching these phenomena on and off. The rotatable device with a reference rotator is a significant technical achievement enabling systematic angular studies on a single sample. The observation of three reproducible regimes and the apparent dependence on BN-BN twist rather than graphene-BN alignment are novel and will stimulate further theoretical work. However, the paper's own evidence of OFF zones at nominally similar angles and the unresolved 0°/60° convention ambiguity mean that the strongest claims in the abstract and conclusions overreach the data as they stand.","major_comments":[{"comment":"The conclusion that the angular alignment between the two BN layers is the key parameter governing these effects, and that the effect is \"proved\" to be governed by this alignment, is weakened by the authors' own observation of OFF zones inside the nominal ON window. The text states that \"different types ... can be found at very similar angles and particularly 'OFF zones' ... can be found inside this range. This points to a possible role of translational movement, suggesting a sliding mechanism.\" If the same (or nearly the same) ΘBN can yield either standard or anomalous gating, then ΘBN is not sufficient to determine the ON/OFF state; an additional degree of freedom (e.g., lateral stacking) must be involved. Please revise the abstract and conclusions to state that the twist angle is a necessary but not sufficient control parameter, or provide a quantitative characterization of the OFF zones showing that they occur at measurably different angles rather than at identical nominal angles.","section":"Conclusions and paragraph after Fig. 4f"},{"comment":"The absolute angular window \"between 15° and 45°\" is convention-dependent because the reference rotator lock cannot distinguish AA (0°) from AA' (60°), as acknowledged in the text: \"this technique does not allow us to distinguish parallel (AA - ΘBN = 0°) from antiparallel (AA' - ΘBN = 60°) alignment.\" With the opposite convention, the window becomes 75° to 105°. Since the abstract reports a definite numerical window without this caveat, it is not a convention-independent physical result as written. Please state the convention explicitly in the abstract or main text and discuss explicitly how the non-60°-periodicity conclusion is, as noted in the text, insensitive to this 0°/60° ambiguity.","section":"Paragraph describing the reference rotator (angle calibration)"},{"comment":"The evidence that graphene-BN alignment plays no role is based on only three samples, with the full angular window mapped on one sample at room temperature; samples 2 and 3 are measured at 6 K at only two fixed values of ΘBN (30° and 90°). This supports the claim but does not, by itself, \"prove\" the independence from the graphene-BN angle. The manuscript should provide explicit error estimates for the angular calibration (precision of the reference lock, repeatability of manual AFM rotation) and ideally show a partial angular map on a second sample to demonstrate that the 15°-45° window is not specific to sample 1.","section":"Section \"Anomalous gating type III\" through \"Angular dependence of the anomalous gating\" (Fig. 4a-d)"}],"minor_comments":[{"comment":"The abstract reports the window as \"between 15 deg and 45 deg\" while the main text describes it as \"close to ΘBN ≈ 30° ± 15°\" and marks a shaded region that ends at 45°; please harmonize these statements and clarify whether the boundaries are measured or interpolated, since no data points are shown exactly at 15° or 45°.","section":"Abstract and Fig. 4e"},{"comment":"The subtraction maps (forward minus backward traces) combine information on hysteresis and gate ineffectiveness; the color scale may obscure the direction of the hysteresis. Consider showing the forward and backward maps separately or adding directional arrows in the color scale legend.","section":"Fig. 2b, d, f"},{"comment":"The sentence \"Nonetheless, this cannot explain the lack of 60° periodicity in the anomalous gating effect\" would benefit from a brief justification, because the relationship between the proposed long-range commensurate angles and the 60° rotational symmetry of the BN lattice is not immediately obvious to the reader.","section":"Discussion of possible mechanisms"},{"comment":"The Methods section is very brief. For reproducibility, please specify the AFM rotation speed, the step size of the rotations (the text mentions 0.05° steps in some cases), and how the locked position of the reference rotator is identified, and indicate which of these details are provided in the supplementary notes.","section":"Methods and supplementary notes"}],"recommendation":"major_revision","confidential_remarks":"The experimental work is technically solid and the rotatable-device platform with a reference rotator is a genuine methodological asset. The main problem is overclaiming: the abstract and conclusions state that the BN-BN twist angle is the key parameter and that the phenomena are \"proved\" to be governed by it, while the text's own admission of OFF zones at very similar angles and a possible sliding mechanism indicate that the angle is not sufficient. This is fixable with careful revision and possibly a small amount of additional analysis. The 0°/60° ambiguity should be disclosed in the abstract if the numerical window is retained. I would support publication after a major revision that tempers the central claims and quantifies the angular uncertainties."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"I read the twist-angle gating paper. What you should know first: the data are real and the central claim is mostly right, but the conclusion overstates the control parameter. The rotatable platform is a genuine step up from static devices. Same sample, same BN flakes, only the BN-BN angle changes, and the anomalous gating switches ON and OFF. That directly addresses the reproducibility debate in the prior literature. The three-type classification (I, II, III) with careful gate-asymmetry maps is systematic and holds together. The non-60-degree periodicity, checked at 6 K on two other samples, is a clean new observation. Credit where it is due: the geometry does rule out the graphene-BN moiré as the trigger. That part of the conclusion is solid.\n\nThe soft spots, in proportion. First, the \"15 to 45 degree window\" is not a clean window. The paper itself reports OFF zones inside that range at nominally the same Theta_BN, and different types at nearly the same angle, and explains this by possible translational sliding. That means the twist angle alone is not sufficient to set the ON state. It may be necessary, but the failure of single-valued behavior is load-bearing. The conclusion says \"proved\" - that is overreach. Proved that the graphene moiré is not the cause, yes. Proved that the BN-BN angle alone governs, no.\n\nSecond, the angle calibration: they admit they cannot distinguish Theta_BN = 0 from 60 degrees (AA vs AA'), so the reported window is convention-dependent. That does not kill the non-60-degree observation, since the 30 vs 90 degree comparisons are internally consistent, but it does affect the absolute window quoted in the abstract. Third, no error bars, and the room-temperature map is essentially one sample. The 6 K checks on two other samples at two angles each are useful but thin. All of this is addressable.\n\nThe microscopic mechanism is untouched, and they say so themselves. I prefer that to a hand-waving model. Who gets value: anyone working on ferroelectric-like gating in graphene/BN, and the rotator community. It deserves a serious referee, and the referee should push for raw data and for an explicit statement that sliding/translation is a candidate second control parameter. Send it to review.","headline":"A genuinely useful rotatable-platform experiment that pins the effect to the BN-BN interface, but the paper overclaims a twist-angle-only switch when its own OFF zones at the same nominal angle show another degree of freedom is in play.","tokens_in":10207,"tokens_out":2068,"would_cite":true,"duration_ms":21056,"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":"The angular alignment between the two boron nitride layers—not the graphene/BN moiré—controls whether electrostatic gating in encapsulated bilayer graphene works normally or shows ineffectiveness and hysteresis, with the effect switchable…","keywords":["anomalous gating","gate ineffectiveness","hysteresis","twist angle","bilayer graphene","hexagonal boron nitride","ferroelectric-like behavior","van der Waals heterostructure"],"falsifier":"Fix the BN-BN twist at angles from $0^\\circ$ through $60^\\circ$ in devices whose crystallographic orientation is verified by diffraction or atomic imaging, and measure top- and bottom-gate response; if gate ineffectiveness or hysteresis appears at $0^\\circ$ or $60^\\circ$, or is absent between $15^\\circ$ and $45^\\circ$, the central claim is wrong.","tokens_in":1834,"feed_emoji":"🔄","tokens_out":1935,"duration_ms":107298,"temperature":0.7,"pith_summary":"This paper argues that the anomalous gating effects seen in bilayer graphene encapsulated in boron nitride—where a gate voltage stops changing the resistance and sweeping it produces hysteresis—are controlled by the twist angle between the top and bottom BN layers, not by the moiré superlattice formed with graphene. Using devices whose top BN flake can be rotated in situ, the authors show that rotation acts as an ON/OFF switch for the anomalous response in the same sample. At room temperature the effect appears only when the BN-BN angle lies roughly between 15° and 45°, and it shows no 60° periodicity, which they interpret as evidence for a 120° periodicity. They also sort the response into three distinct regimes that differ in which gate is ineffective and how the hysteresis behaves. If the claim is right, the effect is an interface property of the two BN flakes, switchable by rotation and largely independent of the graphene orientation.","feed_headline":"BN-BN twist angle, not graphene moiré, rules anomalous gating","feed_subtitle":"In rotatable bilayer graphene devices, gate ineffectiveness and hysteresis appear only for BN-BN twists between 15° and 45°.","key_machinery":"The central piece is a dual-gated, dynamically rotatable van der Waals heterostructure: the top BN flake (the active rotator) can be pushed with an atomic force microscope to change its twist relative to the bottom BN flake, while a reference rotator cut from the same BN material locks at a crystallographically aligned position and calibrates the reported angle $\\Theta_{\\mathrm{BN}}$. This design allows the same device to be mapped over roughly $140^\\circ$ of twist, so the BN-BN angle is effectively the only variable changed between measurements. A separate top gate and bottom gate then reveal which gate is ineffective and which shows hysteresis, which is what separates the response into the three types.","core_discovery":"The core discovery is that $\\Theta_{\\mathrm{BN}}$, the relative twist angle between the two encapsulating BN flakes, is the key control parameter for anomalous gating in bilayer graphene/BN heterostructures. In a single device, moving $\\Theta_{\\mathrm{BN}}$ from about $-27^\\circ$ (standard gating) to about $+25^\\circ$ turns on gate ineffectiveness and hysteresis, and the effect disappears again when the angle is moved out of the active window. The active window at room temperature is between roughly $15^\\circ$ and $45^\\circ$, with no sign of the $60^\\circ$ periodicity that the individual crystal lattices would suggest; the authors expect a $120^\\circ$ periodicity instead. Three types of behavior are identified—type I, with strong gate asymmetry, top-gate ineffectiveness, and large hysteresis when the top gate is swept; type II, with hysteresis on both gates in opposite senses; and type III, where the bottom gate is standard and the top gate is almost entirely ineffective. The graphene/BN alignment is reported not to matter: active angles were observed for graphene-BN alignments of about $10^\\circ$, $50^\\circ$, and $60^\\circ$ across three samples.","pith_inferences":["A direct extension the paper leaves implicit: if the BN-BN interface is what matters, the same anomalous gating should appear in a device with no graphene at all, for example a BN/BN capacitor whose capacitance or current shows the same angle-dependent hysteresis.","Because 'OFF zones' inside the active window hint at translational sliding, the real control parameter may be the full stacking configuration (twist plus in-plane shift), not the twist angle alone; imaging the flake position before and after each rotation would test this.","The predicted $120^\\circ$ periodicity could be checked by mapping the hysteresis-loop area continuously over a full rotation; a clean threefold repeat would point to a specific stacking motif at the BN-BN interface."],"forward_implications":["If correct, anomalous gating is a controllable interface property: rotating the top BN flake can switch gate ineffectiveness and hysteresis on and off at room temperature.","The graphene/BN moiré and the graphene layer's own orientation are not responsible, so the microscopic mechanism should be sought at the BN/BN interface rather than in graphene band structure or correlated-electron states.","Reproducing the effect in future devices requires controlling the BN-BN twist angle; samples outside the roughly $15^\\circ$–$45^\\circ$ window should show normal gating.","The three types (I, II, III) give a classification scheme for the phenomenon that other groups can use to compare observations across different graphene-based systems.","The absence of $60^\\circ$ periodicity, with an expected $120^\\circ$ periodicity, provides a sharp constraint that any theory of the effect has to satisfy."],"supporting_citations":[{"why":"Supplies the rotatable device and reference-rotator method used to measure the BN-BN angle in a single sample.","marker":"[12]"},{"why":"Documents the anomalous gating phenomenon in graphene/BN systems that the paper seeks to explain and reproduce.","marker":"[1]"},{"why":"Represents the moiré band-structure mechanism that the angle-independence results rule out.","marker":"[3]"},{"why":"Reports anomalous gating in bilayer graphene/BN and concludes correlated electrons are not needed, a conclusion this paper's data support.","marker":"[4]"},{"why":"Proposes a disorder or stack-fault origin for the effect; the ON/OFF switching by rotation directly contradicts that proposal.","marker":"[9]"},{"why":"Shows polarization switching in parallel-aligned BN systems and is used to explain the resistance jumps seen in type I behavior.","marker":"[15]"},{"why":"Provides a charge-localization-in-moiré-potential mechanism that the paper argues cannot explain the observed angular dependence.","marker":"[17]"}],"fun_headline_variants":["BN-BN twist, not graphene moiré, controls anomalous gating","Twist angle between BN layers toggles anomalous gating on and off","Anomalous gating requires BN-BN twists between 15° and 45°","Bilayer graphene gating anomalies switch with BN layer twist","BN alignment, not moiré, drives gate ineffectiveness and hysteresis"],"cache_read_input_tokens":12416,"weakest_assumption_plain":"The angle measurement assumes the reference rotator settles at a known crystallographic alignment and that rotating the top flake changes only the twist angle, not the in-plane position; because the setup cannot distinguish $0^\\circ$ from $60^\\circ$ and sliding is invoked to explain off zones, a wrong convention or hidden sliding would shift or erase the claimed $15^\\circ$–$45^\\circ$ window.","fun_headline_variants_meta":{"raw":{"variants":["BN-BN twist, not graphene moiré, controls anomalous gating","Twist angle between BN layers toggles anomalous gating on and off","Anomalous gating requires BN-BN twists between 15° and 45°","Bilayer graphene gating anomalies switch with BN layer twist","BN alignment, not moiré, drives gate ineffectiveness and hysteresis"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000295,"raw_usage":{"total_tokens":1752,"prompt_tokens":1019,"completion_tokens":733,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":635,"completion_tokens_details":{"reasoning_tokens":642}},"tokens_in":635,"tokens_out":733,"duration_ms":7807,"temperature":1.0,"reasoning_tokens":642,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T10:17:12.411601+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Fix the BN-BN twist at angles from $0^\\circ$ through $60^\\circ$ in devices whose crystallographic orientation is verified by diffraction or atomic imaging, and measure top- and bottom-gate response; if gate ineffectiveness or hysteresis appears at $0^\\circ$ or $60^\\circ$, or is absent between $15^\\circ$ and $45^\\circ$, the central claim is wrong.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the rotatable device and reference-rotator method used to measure the BN-BN angle in a single sample."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Documents the anomalous gating phenomenon in graphene/BN systems that the paper seeks to explain and reproduce."},{"cited_title":"Zheng, Q","cited_arxiv_id":null,"evidence_quote":"Represents the moiré band-structure mechanism that the angle-independence results rule out."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Reports anomalous gating in bilayer graphene/BN and concludes correlated electrons are not needed, a conclusion this paper's data support."},{"cited_title":"Icking, L","cited_arxiv_id":null,"evidence_quote":"Shows polarization switching in parallel-aligned BN systems and is used to explain the resistance jumps seen in type I behavior."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides a charge-localization-in-moiré-potential mechanism that the paper argues cannot explain the observed angular dependence."}],"review_version":1}