{"id":"43209d08-a994-4a12-865e-c2065a26646a","arxiv_id":"2411.08880","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Helical trilayer graphene with momentum-dependent tunneling is predicted to host a fractional Chern mosaic at filling 3+1/3: local fractional Chern insulators whose fractionalization pattern varies on the supermoiré scale.","lead":"This paper predicts a new electronic state, a fractional Chern mosaic, in a three-layer graphene stack with successively rotated layers. In this state, electrons split into fractional quasiparticles whose pattern changes across a large-scale moiré pattern, creating a patchwork of different topological orders.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The fractional mosaic claim rests on single-domain FCI robustness; multi-band ED at the headline angle θ=1.8° has not yet demonstrated a robust FCI, so band-mixing effects could destroy the fractional state before the mosaic extrapolation is made.","rationale":"The reader's CONDITIONAL verdict is appropriate. I read the paper in good faith: the MDT derivation in App. A is parameter-free up to the Slater-Koster fit, the HF phase diagram reproduces the observed particle-hole asymmetry, the ν=+3 |C|=1 insulator is strong, and the 1-band ED evidence (N=27 and N=30, flux threading, CDW exclusion) is solid for a frozen-band model. The paper also discloses the multi-band limitation explicitly in App. D4, which counts as honest support rather than a hidden flaw. I do not select the supermoiré separation as the single most load-bearing assumption, even though the reader's 'weakest_assumption' does. The supermoiré concern is real but downstream: the authors only need a local FCI in each domain and C2z to relate domains, and they explicitly argue that the domain-wall network prevents global transport signatures while local probes can test the domains. The more fundamental gap is whether the local FCI survives when the lower HF band is allowed to fluctuate. The main text's central FCI claim is at θ=1.8°, but the only multi-band FCI spectrum is at θ=1.76° with N=18 and Nmax≤2, below the size where 1-band FCI becomes robust. Thus the exact two-band ground state at the headline parameters is unverified. This does not warrant rejection: the bandmax convergence in Fig. S13 is encouraging, and the B-band occupancy stays above 90% in smaller 2-band ED at 1.8°. It does justify keeping the verdict CONDITIONAL, with the multi-band N=18/N=21 calculation as the explicit acceptance criterion.","tokens_in":22051,"tokens_out":10770,"duration_ms":104844,"concrete_test":"At θ=1.8°, ϵr=8, U=0, run two-band ED at ν=3+1/3 on tilted N=18 and, if feasible, N=21 clusters, computing the momentum-resolved many-body spectrum and flux-threading evolution for bandmax Nmax=0, 2, 4, ... up to the full two-band limit (Nmax=N). The decisive check is whether the three-fold quasi-degenerate FCI ground states at the predicted momenta, with a neutral gap that remains open under flux threading, persist as Nmax increases and at the largest accessible N near the 1-band robustness threshold. If the gap closes or the degeneracy splits before full band mixing is allowed, the frozen-HF-band FCI is not the exact two-band ground state; if it survives, the multi-band mixing concern is resolved.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim requires that at ν=3+1/3, the |C|=1 conduction band of the ν=+3 HF insulator hosts an FCI in the full two-band model, not only when the lower HF band is frozen. The strongest evidence is 1-band ED on N=30 at θ=1.8°, which shows three-fold degeneracy and flux-threading gaps. But the paper's own multi-band ED at θ=1.8° consists of N=12, 15, 16 clusters and only reports sublattice occupations, not FCI degeneracy or spectral flow; the FCI spectra in App. D4 are on a tilted N=18 cluster at θ=1.76°, with bandmax Nmax≤2. The text states that \"direct signatures of FCIs in our multi-band calculations are present but not fully robust,\" and attributes this to finite size because N=18 is below the threshold where 1-band ED at 1.8° becomes robust. This is an admission that the exact two-band FCI is not established at the parameter set promoted in the main text. If virtual B-band holes renormalize the A band so that the neutral gap closes or the three-fold degeneracy splits at larger N, the 1-band FCI is an artifact of the frozen-band approximation, and the fractional Chern mosaic claim would fail even if the supermoiré separation-of-scales assumption is granted. The supermoiré and domain-wall extrapolation is a legitimate second concern, but it is logically downstream: local fractionalization is the necessary first step.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper predicts that helical trilayer graphene (HTG) at twist angle θ≈1.8° hosts a fractional Chern mosaic at filling ν=3+1/3. The authors introduce a momentum-dependent tunneling (MDT) correction to the Bistritzer-MacDonald model, which breaks particle-hole symmetry and explains the observed electron-hole asymmetry in transport. Hartree-Fock calculations show a strong-coupling |C|=1 insulator at ν=+3 and a sublattice transition at νc≈3.5, so that at ν=3+1/3 the low-energy physics is an effectively isolated |C|=1 conduction band of predominantly A character. One-band exact diagonalization on a 30-site cluster gives a three-fold quasi-degenerate FCI ground state with correct momenta and flux-threading gaps for εr=8 and 15. Two-band ED with a bandmax constraint on an 18-site tilted cluster at θ=1.76° suggests the FCI survives band mixing, but direct two-band FCI signatures at 1.8° are acknowledged in the text as not fully robust. The mosaic claim follows from the supermoiré structure: h-HTG and \\bar{h}-HTG domains are related by C2z and have opposite local topology, yielding domain-wall chiral edge modes and quasiparticles with opposite braiding phases. The paper is a substantial proposal with strong local-domain numerical evidence but an extrapolative central claim.","tokens_in":22379,"tokens_out":6658,"duration_ms":61739,"significance":"If the local FCI is confirmed in the full two-band model at the headline parameters, the paper would establish a concrete material platform for fractional Chern mosaics, a novel class of states with spatially varying topological order. The work combines a parameter-free derivation of the MDT term from a Slater-Koster model, systematic Hartree-Fock phase diagrams, ED evidence with correct quantum numbers and spectral flow, and explicit falsifiable predictions for SET, STM, and transport experiments. The authors are transparent about the limitations of their multi-band ED, which is a strength. The significance for the correlated moiré electron community is high, provided the multi-band extrapolation is substantiated or the claim is appropriately softened.","major_comments":[{"comment":"The central claim that the ν=3+1/3 fractional Chern insulator appears in the full two-band model at the headline parameters (θ=1.8°, εr=8,15) is not yet established. The 1-band ED freezes the B (|C|=-2) band; the only multi-band ED FCI spectra are the bandmax-restricted tilted N=18 calculation at θ=1.76° (App. D4, Fig. S13), and the text states that 'direct signatures of FCIs in our multi-band calculations are present but not fully robust.' The N=12,15,16 two-band calculations at 1.8° report only sublattice occupations (Fig. 3b), not degeneracy or spectral flow. Because virtual B-band holes could renormalize the A band enough to close the neutral gap or split the three-fold degeneracy at larger N, the frozen-band approximation could be invalid. The paper should either present full two-band ED or bandmax ED at θ=1.8° at a size comparable to the 1-band threshold (e.g., N=18 with Nmax=2 or 3) showing three-fold degeneracy with flux-threading gaps, or explicitly temper the mosaic prediction to the frozen-band regime. As written, the mosaic rests on an extrapolation that the authors themselves flag.","section":"Exact diagonalization; App. D4"},{"comment":"The fractional Chern mosaic is inferred from local-domain calculations under the assumption that the supermoiré modulation and the gapless domain walls do not substantially hybridize the local Chern bands: 'Due to the separation of lengthscales asm ≫ am, it suffices to model the system as locally periodic within each domain.' No calculation or estimate is given for the strength of inter-domain coupling relative to the FCI neutral gap, and the authors acknowledge that the favored mosaic scenario depends on fine details of domain-wall energetics. To make the mosaic claim quantitative, the paper should provide an estimate (or a minimal coupled-domain model) showing that the domain-wall bandwidth is small compared to the FCI gap; otherwise the conclusion that a fractional Chern mosaic 'can be realized' is a plausible scenario rather than a demonstrated result.","section":"Introduction; Discussion (mosaic)"}],"minor_comments":[{"comment":"The definition δρ_q = ρ_q - 4ρ_q uses the same symbol for the projected density operator and for the mean density of the central bands; please disambiguate the notation (e.g., δρ_q = ρ_q - 4⟨ρ_q⟩) so that the 'average' density subtraction is unambiguous.","section":"Interactions, Eq. (5)"},{"comment":"Typo: the sentence 'the strong-coupling phase only exists for for relatively strong interactions' contains a duplicated 'for'.","section":"Hartree-Fock phase diagram"},{"comment":"The text 'Lattice relaxation creates large triangular moiré-periodic domains, referred to as h-HTG and h-HTG' should read 'h-HTG and \\bar{h}-HTG'; the overbar is missing on the second occurrence.","section":"Introduction"},{"comment":"The momentum-resolved spectra in Fig. 3a would be easier to interpret if the high-symmetry momenta (γ, κ, κ′) were labeled directly on the momentum axis.","section":"Fig. 3a caption"},{"comment":"In the second mosaic scenario (opposite valley polarizations), the local Hall conductivity is identical in both domains; the sense in which the topological order varies (valley quantum numbers of the e/3 quasiparticles) should be stated more explicitly, since the first scenario (opposite σxy) is the more conventional Chern mosaic.","section":"Discussion"}],"recommendation":"major_revision","confidential_remarks":"This is a strong and timely proposal, and the 1-band ED evidence is convincing. The main gap is the direct two-band FCI evidence at the headline angle; the authors themselves acknowledge this in App. D4. If the authors can add bandmax ED at θ=1.8° at a size comparable to N=18 or larger and show the three-fold degeneracy with flux-threading gaps, the paper would be acceptable. Otherwise, the central claim should be softened to a frozen-band fractional Chern insulator that suggests, but does not fully establish, a fractional Chern mosaic. The paper's fit to the journal's scope is good."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Punchline: this is a genuinely new proposal—fractional Chern mosaic—and the single-domain FCI evidence is solid at the frozen-band level. But the multi-band FCI at the headline angle is not yet established, so the mosaic claim is a prediction rather than a demonstrated result.\n\nWhat's new: the fractional Chern mosaic concept, where the pattern of fractionalization varies across the supermoiré lattice, goes beyond the integer Chern mosaic. The paper also identifies a momentum-dependent tunneling term that breaks particle-hole symmetry and drives a sublattice transition at νc≈3.5, matching the experimental anomalous Hall window in HTG. The MDT parameter is derived from a Slater-Koster model, not fitted to the FCI, so there is no circularity here. The HF phase diagrams are thorough, and the 1-band ED at ν=3+1/3 on N=30 shows three-fold degeneracy and flux-threading gaps; N=27 rules out CDW. The authors are transparent about their limitations.\n\nThe soft spots are real but disclosed. First, the full two-band FCI at θ=1.8° is not demonstrated. The 2-band ED results at that angle only show sublattice occupations; the spectral evidence with bandmax is on a tilted N=18 cluster at θ=1.76°, and the authors themselves say the multi-band FCI signatures are not fully robust. They attribute this to finite size, and that is plausible, but band mixing could still spoil the FCI at larger N. This is the load-bearing gap for the fractional mosaic claim, because local fractionalization must survive in the full band model.\n\nSecond, the mosaic itself is inferred from symmetry and separation of length scales, not simulated with the supermoiré potential or domain walls. That is a softer concern and logically downstream: if the local FCI holds, the mosaic argument is reasonable, though the gapless domain walls could complicate the picture.\n\nOverall, this paper is clear, honest, and worth engaging with. It should go to peer review. I would ask the authors to either push the multi-band ED at 1.8° to larger Nmax, or temper the central claim to 'a proposal consistent with available evidence.' The right reader is anyone working on moiré topology or fractional Chern insulators.","headline":"A solid proposal for a fractional Chern mosaic in helical trilayer graphene, with a genuine gap between the strong 1-band FCI evidence and the untested multi-band and supermoiré extrapolations.","tokens_in":22961,"tokens_out":2854,"would_cite":true,"duration_ms":26279,"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 paper predicts that helical trilayer graphene at twist angle 1.8° hosts a fractional Chern mosaic at filling ν=3+1/3, with alternating domains of opposite fractional Hall conductivity and opposite anyonic braiding phases.","keywords":["fractional Chern insulator","helical trilayer graphene","supermoiré","fractional Chern mosaic","topological order","momentum-dependent tunneling","exact diagonalization","anyon braiding"],"falsifier":"A scanning single-electron transistor measurement on a θ≈1.8° helical trilayer graphene device at ν=3+1/3 should see alternating supermoiré-scale regions that are incompressible and disperse with perpendicular magnetic field according to the Streda formula, with local Hall conductivities of opposite sign; their absence would rule out the mosaic. Alternatively, an exact diagonalization calculation on a supercell that includes the full supermoiré potential in the active bands, and that shows the three-fold degenerate fractional Chern insulator ground states collapse, would falsify the single-domain separation-of-scales assumption.","tokens_in":21800,"feed_emoji":"🧩","tokens_out":10524,"duration_ms":83286,"temperature":0.7,"pith_summary":"This paper predicts that helical trilayer graphene, three graphene sheets stacked with alternating small twist angles, can realize a fractional Chern mosaic: a state in which electrons fractionalize into charge e/3 quasiparticles in some regions while the pattern of fractionalization reverses on a much larger supermoiré length scale. The specific claim is that at twist angle θ≈1.8° and filling ν=3+1/3, the conduction band of the ν=+3 Chern insulator supports a fractional Chern insulator, and that the two types of relaxation domains, h-HTG and hbar-HTG, are related by a two-fold rotation so they host opposite local Hall conductivities and opposite braiding phases. This combination of topology and supermoiré periodicity would be a new kind of quantum state: topological order that varies from place to place. The paper argues the ingredients are realistic because the momentum-dependent tunneling correction explains the observed particle-hole asymmetry and stabilizes the relevant Chern band, and exact diagonalization finds the characteristic three-fold degenerate ground states of a fractional Chern insulator.","feed_headline":"Fractional Chern mosaic predicted in twisted trilayer graphene","feed_subtitle":"The predicted e/3 quasiparticles would braid with opposite phases in neighboring supermoiré domains","key_machinery":"The machinery has two components. The first is the generalized Bistritzer-MacDonald continuum model (the standard continuum description of moiré graphene bands) of a single h-HTG domain, augmented by a momentum-dependent tunneling (MDT) term: a leading correction arising from the decay of interlayer tunneling with in-plane distance between atomic pz orbitals. MDT breaks particle-hole symmetry, lowers the mean energy of the B Chern-sublattice band relative to the A band, stabilizes the strong-coupling |C|=1 insulator at ν=+3, and pushes the first-order sublattice transition filling to νc≈3.5, above 3+1/3; the unoccupied |C|=1 A band then has favorable quantum geometry. The second component is the supermoiré lattice of alternating h-HTG and hbar-HTG domains: the paper invokes the separation of length scales (supermoiré period much larger than moiré period) to treat each domain as locally periodic, and uses the C2z rotation relating the two domain types to combine the single-domain fractional Chern insulator into a spatially varying topological order.","core_discovery":"On the paper's own terms, the central discovery is that fractionalization can occur locally in each moiré-periodic domain of helical trilayer graphene, and that the global C2z symmetry of the supermoiré lattice turns this into a fractional Chern mosaic. In the h-HTG domain, the |C|=1 conduction band of the ν=+3 Hartree-Fock insulator at θ=1.8° hosts a fractional Chern insulator at ν=3+1/3, with local Hall conductivity σxy=−(2/3)τz e²/h, where τz=±1 labels the valley. The hbar-HTG domains are C2z-related, so their partially filled band has opposite topology and local σxy=+(2/3)τz e²/h; electrons in both domains fractionalize into charge e/3 quasiparticles, but the anyonic exchange phase is $e^{{+iπ/3}}$ in one domain and $e^{{−iπ/3}}$ in the other. If the valley polarizations are instead opposite in the two domains, the local Hall conductivities are equal while the e/3 quasiparticles carry opposite valley quantum numbers.","pith_inferences":["Beyond the paper: the same sublattice-transition cascade suggests fractional Chern insulator plateaus at other fractional fillings of the form ν=3+k/(2k+1) below νc≈3.5, and possibly near other integer fillings where the flavor and sublattice physics resets.","Beyond the paper: if the supermoiré potential or domain-wall coupling hybridizes the local Chern bands enough to destroy the FCI, the system would instead show an integer Chern mosaic or reentrant metallic regions; local Hall imaging as a function of filling and displacement field would map how robust the mosaic is.","Beyond the paper: aligning the device with hBN breaks the C2z relation between domains and could confine fractionalization to one domain type, effectively turning the mosaic into a single uniform fractional Chern insulator that would be much easier to see in transport."],"forward_implications":["At ν=3+1/3 the full HTG crystal would consist of alternating domains with local Hall conductivities ±(2/3)e²/h, separated by domain walls that carry fractionally charged chiral edge modes.","Charge e/3 quasiparticles in one domain type would acquire an anyonic phase of e^{+iπ/3} upon exchange, while those in the other domain type acquire e^{−iπ/3}; the topological order itself varies across the supermoiré scale.","The particle-hole asymmetry seen in HTG transport, with correlated states only on the electron-doped side and anomalous Hall behavior persisting to ν≈3.5, is explained by MDT and the sublattice transition at νc≈3.5.","Because the conducting domain walls short global transport, the mosaic is best detected with local probes: scanning single-electron transistors should see incompressible features that disperse with magnetic field according to the Streda formula, and STM should image the sublattice cascade.","If a small magnetic field makes the valley polarizations opposite in the two domains, both domains have the same local σxy=−(2/3)e²/h but the e/3 quasiparticles carry opposite valley quantum numbers."],"supporting_citations":[{"why":"Provides the transport observations that MDT must explain: correlated states only on the electron-doped side and anomalous Hall effect persisting to about ν=3.5.","marker":"[51]"},{"why":"Supplies the strong-coupling Hartree-Fock framework, the average-interaction scheme, and the reference phase diagram that this work modifies with MDT.","marker":"[52]"},{"why":"Introduces the magic-angle helical trilayer graphene continuum model and the h-HTG versus hbar-HTG domain structure with valley-contrasting Chern bands.","marker":"[36]"},{"why":"Establishes the multiscale lattice relaxation that creates large triangular h-HTG and hbar-HTG domains on the supermoiré scale.","marker":"[37]"},{"why":"Reports scanning single-electron-transistor imaging of supermoiré relaxation and conductive domain walls, the experimental basis for the length-scale separation and for local-probe detection.","marker":"[67]"},{"why":"Derives the momentum-dependent interlayer tunneling that produces the MDT term central to breaking particle-hole symmetry.","marker":"[68]"},{"why":"Provides the Bistritzer-MacDonald continuum model generalized in Eq. (1) for the HTG band structure.","marker":"[69]"},{"why":"Defines the fractional Chern insulator diagnostic used in the exact diagonalization: three-fold quasi-degenerate ground states at expected momenta that stay gapped under flux threading.","marker":"[65]"}],"fun_headline_variants":["Supermoiré graphene hosts e/3 anyons with opposite braiding phases","Fractional Chern mosaic: topology varies across supermoiré domains","e/3 quasiparticles braid with opposite phases in supermoiré graphene","Supermoiré mosaic: fractional Chern order with paired anyonic phases"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the system can be modeled as locally moiré-periodic within each h-HTG or hbar-HTG domain, ignoring the supermoiré modulation and the gapless domain walls; if that coupling is strong enough to spoil the local fractional Chern insulator, the mosaic does not form.","fun_headline_variants_meta":{"raw":{"variants":["Supermoiré graphene hosts e/3 anyons with opposite braiding phases","Fractional Chern mosaic: topology varies across supermoiré domains","e/3 quasiparticles braid with opposite phases in supermoiré graphene","Supermoiré mosaic: fractional Chern order with paired anyonic phases"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000449,"raw_usage":{"total_tokens":2223,"prompt_tokens":860,"completion_tokens":1363,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":476,"completion_tokens_details":{"reasoning_tokens":1280}},"tokens_in":476,"tokens_out":1363,"duration_ms":10179,"temperature":1.0,"reasoning_tokens":1280,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T21:13:48.297613+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A scanning single-electron transistor measurement on a θ≈1.8° helical trilayer graphene device at ν=3+1/3 should see alternating supermoiré-scale regions that are incompressible and disperse with perpendicular magnetic field according to the Streda formula, with local Hall conductivities of opposite sign; their absence would rule out the mosaic. Alternatively, an exact diagonalization calculation on a supercell that includes the full supermoiré potential in the active bands, and that shows the three-fold degenerate fractional Chern insulator ground states collapse, would falsify the single-domain separation-of-scales assumption.","supporting_citations":[{"cited_title":"Multi-scale lattice relaxation in general twisted trilayer graphenes","cited_arxiv_id":"2305.13155","evidence_quote":"Establishes the multiscale lattice relaxation that creates large triangular h-HTG and hbar-HTG domains on the supermoiré scale."},{"cited_title":"Regnault and B","cited_arxiv_id":null,"evidence_quote":"Defines the fractional Chern insulator diagnostic used in the exact diagonalization: three-fold quasi-degenerate ground states at expected momenta that stay gapped under flux threading."}],"review_version":1}