{"id":"6363ea5e-27d1-435d-958c-aeb7a5445e24","arxiv_id":"2607.08710","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"At low temperature, the ν=2/3 state in moiré rhombohedral graphene is an integer Chern insulator (charge e), not a fractional Chern insulator masked by edge effects.","lead":"A gate-defined antidot reveals that the ν=2/3 state in moiré rhombohedral graphene carries whole-electron quasiparticles at low temperature. This rules out an edge-equilibration explanation and points to a genuine bulk transition into an integer Chern crystal.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Central claim depends on doping-independent antidot capacitance; if C_CG shifts by ~3x between ν=1 and 2/3, identical periods do not prove e*=e.","rationale":"The reader's weakest_assumption correctly identifies the doping-independent C_CG assumption as the key vulnerability. The paper's central claim—that ν=2/3 becomes a C=1 integer Chern insulator below ~100 mK—rests on the approximate equality of ΔV_CG at ν=1 and ν=2/3, combined with the assumption that C_CG is unchanged. The manuscript explicitly states this assumption and says it is 'often an excellent approximation,' but provides no direct experimental calibration at ν=2/3. The COMSOL simulation, while useful for the geometric capacitance, treats the mRG as a metallic sheet and therefore cannot address state-dependent screening; a compressible low-temperature state such as the proposed generalized anomalous Hall crystal could screen differently from an incompressible Chern insulator. A factor-of-three change in C_CG would fully mask a fractional quasiparticle charge. The Laughlin-pumping data support a similar antidot area but not a similar capacitance. Because this is a genuine but not necessarily fatal gap, and the paper also presents supporting evidence from bulk resistance and Chern-number dispersion, the appropriate verdict remains CONDITIONAL, matching the reader's assessment. No verdict change is needed.","tokens_in":16190,"tokens_out":6376,"duration_ms":73914,"concrete_test":"Repeat the COMSOL simulation of Extended Data Fig. 7 with the mRG layer represented as a 2D screening layer with a Thomas-Fermi screening length (or quantum capacitance) appropriate for (i) the incompressible ν=1 Chern insulator and (ii) the low-temperature ν=2/3 state, e.g., a compressible generalized anomalous Hall crystal. Compute ΔV_CG required to add one electron to the antidot in each case. If the two values differ by <10%, the equal-period inference is safe; if they differ by ~30% or more, the data are consistent with e*=e/3 and the central claim is not established.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The inference that equal voltage periods at ν=1 and ν=2/3 imply equal quasiparticle charge is the single load-bearing step. It rests on the stated assumption of a doping-independent control-gate capacitance C_CG. This assumption is not directly calibrated at ν=2/3. The COMSOL simulation in Extended Data Fig. 7 models all conducting layers as metallic sheets at V_TG=V_BG=0; it captures the geometric capacitance but not state-dependent screening by the mRG electron system. A low-temperature state such as the proposed generalized anomalous Hall crystal could be more compressible than an incompressible Chern insulator, reducing C_CG. A threefold reduction in C_CG at ν=2/3 would make a fractional e*=e/3 period equal to the observed integer period. The similar Laughlin-pumping period ΔB≈52 mT (vs 55 mT at ν=1) constrains the antidot area, not C_CG. Thus the charging data alone cannot exclude an FCI with e*=e/3 and altered capacitance.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports bulk-sensitive antidot charging measurements in moiré rhombohedral graphene at fillings ν=1 and ν=2/3. At T=12 mK, Coulomb oscillations in the diagonal resistance give nearly identical control-gate voltage periods (ΔV_CG≈9.5 mV and 9.7 mV), which the authors interpret as identical quasiparticle charge e*=e; they therefore conclude that the ν=2/3 state at low temperature is a C=1 integer Chern insulator, not an FCI with unequilibrated edges. Laughlin-pumping stripes give magnetic-field periods ΔB≈55 mT and 52 mT, and the bulk resistance shows a peak-to-dip evolution with temperature near ν=2/3, with a non-monotonic R_bulk minimum near 150 mK. The authors argue for a thermodynamic phase transition from an FCI at high T to a generalized anomalous Hall crystal at low T.","tokens_in":16453,"tokens_out":11172,"duration_ms":119227,"significance":"If correct, the work resolves an important debate about the low-temperature fate of FCIs in mRG: the extended QAH background is a genuine bulk competing phase, not an edge-equilibration artifact. The experimental strengths are substantial: the measurement is bulk-sensitive by design, the Coulomb oscillations and Laughlin pumping are seen in multiple devices, the B-field dispersion independently yields C≈2/3 at 400 mK and C≈1 at 12 mK (Extended Data Figs. 5–6), and the electrostatic simulation reproduces the geometric periodicity (Extended Data Fig. 7). These features make the central claim credible. The main weakness is that the equality-of-periods argument for e*=e relies on an uncalibrated doping-independent capacitance assumption; the B-field dispersion partially mitigates this, but the charging inference itself needs additional support or a clear caveat.","major_comments":[{"comment":"The inference that identical ΔV_CG at ν=1 and ν=2/3 implies e*=e rests entirely on the stated assumption of a doping-independent gate capacitance C_CG. This assumption is not directly calibrated at ν=2/3. The COMSOL model (Extended Data Fig. 7) treats all conducting layers as metallic sheets at V_TG=V_BG=0 and therefore captures only geometric capacitance; it cannot exclude a state-dependent change of C_CG by a factor of ~3 due to different bulk compressibility or screening in the proposed low-temperature state. The similar Laughlin-pumping period ΔB≈52 mT (vs 55 mT) fixes the antidot area, not C_CG. I recommend either a direct lever-arm calibration at both fillings (e.g., Coulomb-diamond slope at ν=2/3) or a clear downgrading of this inference in favor of the independent B-field dispersion evidence.","section":"Main text, 'Quasiparticle charging near ν=2/3'"},{"comment":"The 'thermodynamic phase transition' at T≈150 mK is inferred from the temperature dependence of a two-terminal quasi-bulk resistance R_bulk. This transport proxy can be non-monotonic for reasons unrelated to a bulk thermodynamic transition, such as contact effects, edge-state equilibration, or percolation. The claim would be strengthened by a direct thermodynamic probe (e.g., compressibility or heat capacity) or by showing a scaling collapse of R_bulk around the transition. As written, the abstract's 'suggest' is appropriate, but the later text—'the results suggest the presence of a thermodynamic phase transition'—goes somewhat beyond the transport evidence.","section":"Main text, 'Competing ground states probed by bulk resistance' (Fig. 4e)"}],"minor_comments":[{"comment":"The notation '𝝂𝝂 = 𝟏𝟏 and 2/3' appears with bold/math artifacts; please use standard symbols (ν=1 and ν=2/3).","section":"Abstract"},{"comment":"'above 2000C' should read 'above 200 °C'.","section":"Methods, 'Device fabrication'"},{"comment":"The text refers to dashed lines i, ii, and iii in Fig. 1g; these labels are not visible in the figure and should be added or described more explicitly.","section":"Figure 1 and main text"},{"comment":"The description of the electrostatics simulation should state explicitly how ΔV_CG≈9.0 mV is extracted from the simulated charge distribution; the current text says only that the distribution was used.","section":"Extended Data Fig. 7"},{"comment":"The claim that a doping-independent C_CG is 'often an excellent approximation' cites mainly GaAs/AlGaAs quantum Hall antidot experiments; the applicability of this assumption to a strongly correlated moiré system should be discussed or qualified.","section":"Main text, 'Quasiparticle charging near ν=2/3'"}],"recommendation":"major_revision","confidential_remarks":"The referee stress-test concern about C_CG is real and should be addressed, but it is not necessarily fatal because Extended Data Figs. 5–6 provide an independent, capacitance-free indication (C≈2/3 at 400 mK, C≈1 at 12 mK) that supports the transition. The paper would be strengthened by a quantitative calibration or a more careful wording of the charging inference. I see no grounds for rejection; the physics is important and the experimental approach is novel."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The core result here is clean and important: quasiparticle charging into an antidot in pentalayer mRG gives e* = e at both ν=1 and ν=2/3 at low temperature, and the B-field dispersion of the ν=2/3 state changes from C≈2/3 at 400 mK to C≈1 at 12 mK. Those two independent probes together strongly argue that the low-T state is an integer Chern insulator, and the extended QAH background is a bulk phase transition, not an edge-equilibration artefact. This directly settles a debate that has been running in the moiré graphene community, so it is a genuinely new and useful result. The antidot fabrication is careful, the Fourier analysis is clean, and the COMSOL simulation reproduces the absolute charging period. The Laughlin pumping period is a nice consistency check, and the quasi-bulk resistance measurement is a reasonable attempt to get away from edge artefacts. Having three devices helps.\n\nThe main weak spot is the doping-independent capacitance assumption. If C_CG at ν=2/3 were a third of its ν=1 value, then equal voltage periods would not imply e* = e. But there is no physical reason to expect a factor-of-three jump: the gate geometry is identical, and both the integer Chern insulator and the proposed low-T state are gapped and incompressible in the relevant sense. The simulation based on the known structure matches the absolute period, so the burden is on the skeptic to propose a specific screening mechanism that changes C_CG by that much. More importantly, the B-field dispersion in Extended Data Fig. 5 is independent of the capacitance assumption and already gives C≈1 at 12 mK. The charging data and the dispersion triangulate the same conclusion, so the capacitance caveat is not load-bearing.\n\nTwo other caveats, both minor: the claim of a thermodynamic phase transition at ~150 mK is inferred from transport, which is suggestive but not thermodynamic proof; and the identification of the low-T state as a generalized anomalous Hall crystal is speculative—reasonable, but not proven. The raw data are not included, so error bars cannot be independently verified. These are revision-level issues.\n\nThis paper deserves a serious referee. The measurement is hard, the comparison against the edge-equilibration scenario is sharp, and it resolves a live debate. I would ask the authors to show the raw oscillations at both fillings and to address the capacitance assumption directly, but the central argument holds. Send it to review.","headline":"Solid experiment; the e* and C=1 data make the case, and the capacitance caveat is real but not fatal.","tokens_in":16981,"tokens_out":3199,"would_cite":true,"duration_ms":35110,"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":"Quasiparticle charging shows ν=2/3 in moiré rhombohedral graphene is an integer Chern state at low temperature, not a fractional one.","keywords":["moiré rhombohedral graphene","fractional Chern insulator","quantum anomalous Hall effect","Chern number","antidot","quasiparticle charging","Laughlin charge pumping","competing ground states"],"falsifier":"Measure the antidot's control-gate capacitance independently at both fillings (for example, from the Coulomb diamond area together with a simultaneous density calibration, or via a nearby charge sensor). If C_CG at ν=2/3 turns out to be roughly one-third of its value at ν=1, then the measured ΔV_CG would correspond to e*/3, overturning the integer-charge conclusion. Alternatively, a shot-noise measurement through the antidot near 100–150 mK should reveal fractional charge e/3 if any fractionally charged quasiparticles are present in the transition region.","tokens_in":16100,"feed_emoji":"🧲","tokens_out":4122,"duration_ms":41932,"temperature":0.7,"pith_summary":"The paper aims to settle why the fractional Chern insulator at ν=2/3 in moiré rhombohedral graphene gives way to an extended quantum anomalous Hall background when the temperature is lowered. By measuring the voltage needed to add one quasiparticle to a gate-defined antidot, the authors find nearly identical charging periods at ν=2/3 and ν=1, implying the quasiparticle charge is one electron at both fillings. Combined with bulk resistance measurements and magnetic-field dispersion showing the Chern number changing from 2/3 to 1, they conclude the low-temperature state is a C=1 integer Chern insulator and that the transition is a genuine bulk phase transition, not an edge-state equilibration effect. A sympathetic reader should care because this identifies a competing ground state—likely a generalized anomalous Hall crystal—that must be suppressed to stabilize and probe fractional excitations in these materials.","feed_headline":"Charging probe shows ν=2/3 is a C=1 Chern insulator","feed_subtitle":"Equal gate-voltage periods at ν=1 and ν=2/3 give e*=e, making the low-temperature QAH plateau a bulk phase transition.","key_machinery":"The central probe is Coulomb charging of a gate-defined antidot in the weak-tunneling regime. The control-gate voltage period ΔV_CG = e*/C_CG directly reads out the quasiparticle charge, while the magnetic-field period of Laughlin charge pumping, ΔB = h/(eA), confirms the antidot area and the pumping mechanism. The crucial comparison is the near-identity of ΔV_CG for ν=1 and ν=2/3, which converts equal charging periods into equal quasiparticle charges and rules out fractional e*/3 at low temperature.","core_discovery":"At temperatures below about 100 mK, the ν=2/3 state in moiré rhombohedral graphene is a C=1 integer Chern insulator, so the extended quantum anomalous Hall background seen in transport is a real bulk phase transition rather than a failure of edge states to equilibrate. The key evidence is that the control-gate voltage period of Coulomb oscillations in an antidot is nearly identical at ν=2/3 (9.7 mV) and ν=1 (9.5 mV), giving the same quasiparticle charge e* = e. Temperature-dependent bulk resistance shows a non-monotonic behavior near ν=2/3 with a resistance minimum near 150 mK, suggesting a thermodynamic phase transition, and the magnetic-field dispersion of the insulating state changes from","pith_inferences":["A testable extension is to track the antidot charging period while tuning the gate-to-sample distance: if reducing Coulomb screening stabilizes the fractional Chern insulator, the voltage period should jump to the fractional value e*/3 as the low-temperature integer state is suppressed.","If the low-temperature ground state is a generalized Wigner crystal, scanning tunneling spectroscopy should reveal a charge-ordered honeycomb lattice with a single-particle gap in the bulk—an observation that would independently confirm the proposed picture.","The same antidot charging technique could be applied to other moiré fractional Chern insulators to test whether extended integer quantum anomalous Hall states are a generic low-temperature competitor across different material platforms."],"forward_implications":["The extended quantum anomalous Hall plateau at low temperature is a bulk property of the ground state, not an artifact of edge-state non-equilibration.","The fractional Chern insulator at ν=2/3 is stable only above roughly 150 mK; below that it loses to a generalized anomalous Hall crystal, so experiments seeking fractional excitations must either stay above this transition or suppress the competing crystal.","The magnetic-field dispersion shifting from C≈2/3 at 400 mK to C≈1 at 12 mK is consistent with a thermodynamic phase transition, strengthening the bulk-transition interpretation.","The non-monotonic temperature dependence of the bulk resistance near ν=2/3 suggests the topological transition may be continuous, with gap closure near the critical temperature.","Mesoscopic antidot charging is a bulk-sensitive technique that can distinguish competing ground states in moiré materials, a capability demonstrated here for the first time in this context."],"fun_headline_variants":["ν=2/3 in moiré graphene is a C=1 Chern insulator","Quasiparticle charging proves ν=2/3 is a bulk phase transition","mRG ν=2/3: not an FCI but an integer Chern insulator","Bulk transition, not edge failure, explains ν=2/3 QAH state","Same charging period at ν=1 and ν=2/3 reveals e* = e"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The whole argument rests on assuming that the antidot's control-gate capacitance is the same at ν=1 and ν=2/3; if the capacitance differed by a factor of about three between the two fillings, then equal voltage periods would not imply equal quasiparticle charge.","fun_headline_variants_meta":{"raw":{"variants":["ν=2/3 in moiré graphene is a C=1 Chern insulator","Quasiparticle charging proves ν=2/3 is a bulk phase transition","mRG ν=2/3: not an FCI but an integer Chern insulator","Bulk transition, not edge failure, explains ν=2/3 QAH state","Same charging period at ν=1 and ν=2/3 reveals e* = e"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000231,"raw_usage":{"total_tokens":1352,"prompt_tokens":803,"completion_tokens":549,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":547,"completion_tokens_details":{"reasoning_tokens":448}},"tokens_in":547,"tokens_out":549,"duration_ms":5734,"temperature":1.0,"reasoning_tokens":448,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-02T07:46:54.708912+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the antidot's control-gate capacitance independently at both fillings (for example, from the Coulomb diamond area together with a simultaneous density calibration, or via a nearby charge sensor). If C_CG at ν=2/3 turns out to be roughly one-third of its value at ν=1, then the measured ΔV_CG would correspond to e*/3, overturning the integer-charge conclusion. Alternatively, a shot-noise measurement through the antidot near 100–150 mK should reveal fractional charge e/3 if any fractionally charged quasiparticles are present in the transition region.","supporting_citations":[],"review_version":2}