{"id":"fb3a4cc1-e479-4622-a87c-50dd3f8ae41d","arxiv_id":"2509.11041","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":7,"one_line_summary":"Hartree-Fock predicts that a twisted-hBN-spaced TMD bilayer at nu=1 can host a p-wave exciton condensate with coexisting quantum anomalous Hall and counterflow superfluid phases.","lead":"Using mean-field calculations, the authors predict that two transition-metal-dichalcogenide layers separated by twisted hBN can condense holes into a chiral p-wave exciton state with a quantum anomalous Hall effect. If realized, this device would be a clean platform for zero-field exciton condensation and lossless counterflow, addressing a long-standing question in semiconductor bilayers.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Single-band projection is uncontrolled at the p-EI parameters: the moiré miniband gap (~V_m/3 ≈ 2-3 meV) is far smaller than the interaction scale (~30 meV), so truncating each layer to its topmost miniband may eliminate the physics producing the p-EI.","rationale":"The reader's weakest assumption is the generic HF overestimation of symmetry-broken states, which the authors themselves acknowledge. I identified a more specific and arguably more fundamental weakness: the uncontrolled single-band projection. The paper states the assumption explicitly, but it is most fragile precisely in the V_m range where the p-EI appears (weak modulation → small miniband gap). Since the interaction energy far exceeds the interband gap, the two-band truncation is not a controlled approximation. This could invalidate the central claim even at the mean-field level, not just after including quantum fluctuations. The proposed test—a multi-band HF calculation at the p-EI parameters—would directly determine whether band mixing changes the phase. If the p-EI survives, the concern is resolved; if not, the paper's central prediction is an artifact of the projection. This is a concrete, computational check that is more decisive than a generic QMC call and directly targets the model's internal consistency.","tokens_in":13133,"tokens_out":11683,"duration_ms":134867,"concrete_test":"Repeat the self-consistent Hartree-Fock calculation for the p-EI parameters (e.g., V_m = 7 meV, V_D = 0, d = 4 nm, a_m = 8 nm, ε = 6, α = 1) keeping at least the two lowest minibands per layer instead of only the topmost. If the p-EI state (|C|=1, counterflow superfluid) no longer appears as the ground state, or its phase boundary shifts substantially, the single-band projection is the load-bearing limitation. If the p-EI persists with the same topology, the projection is validated.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper projects each layer's Hamiltonian onto the topmost miniband (SM Eq. S1–S2) with the assumption that interactions do not cause significant band mixing. This assumption is not justified in the parameter region where p-EI occurs: for V_m = 6–8 meV and a_m = 8 nm, the gap to the next moiré miniband is of order V_m/3 ≈ 2–3 meV, while the characteristic Coulomb energy e²/(ε a_m) ≈ 30 meV. Interactions are therefore not small compared to the band separation, and the truncation is uncontrolled. The p-EI phase in Fig. 3 sits precisely at these weak-modulation parameters, so the central prediction may be an artifact of the projection rather than a property of the full continuum model. This is distinct from the acknowledged HF overestimation of symmetry-broken states; even at the mean-field level, the two-band ansatz may be missing essential band-mixing effects.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a moiré device in which two TMD monolayers are separated by a twisted hBN multilayer that suppresses interlayer tunneling and imprints opposite triangular-lattice moiré potentials on the two layers. Using a continuum model plus dual-gate-screened Coulomb interactions and self-consistent Hartree-Fock in a single miniband per layer, the authors compute the ν=1 phase diagram as a function of modulation strength V_m and displacement field V_D. They identify layer-polarized, s-wave excitonic insulator, nematic excitonic insulator, metallic, and chiral p-wave excitonic insulator (p-EI) phases. The p-EI, found near V_D=0 and intermediate V_m, is characterized by a skyrmionic layer-pseudospin texture, Chern number |C|=1, and counterflow superfluidity. Phase diagrams for different α, d, ε, and a_m are collected in the Supplemental Material.","tokens_in":13358,"tokens_out":6469,"duration_ms":77912,"significance":"If the prediction is robust, the work would be a substantial advance: a zero-field equilibrium exciton condensate with spontaneous interlayer coherence, no single-particle tunneling, and coexisting quantum anomalous Hall and counterflow superfluid responses. The paper gives a concrete experimental signature (quantized Hall drag) and the mean-field calculations are self-consistent and internally consistent. The authors also honestly acknowledge that Hartree-Fock overestimates symmetry-broken states. However, the central p-EI claim rests on two approximations that are not fully controlled—the single-miniband projection at weak modulation and the neglect of spin/valley fluctuations—and the paper's abstract/conclusion are considerably stronger than these caveats.","major_comments":[{"comment":"The single-band projection is uncontrolled in the parameter region of the p-EI. The paper keeps only the topmost miniband per layer (SM Eq. S1–S2) assuming interactions do not cause significant band mixing. In Fig. 3 the p-EI appears for V_m ≈ 5–8 meV, a_m = 8 nm, ε = 6. The separation to the next miniband is O(V_m/3) ≈ 2–3 meV, while e²/(ε a_m) ≈ 30 meV. Since the interaction exceeds this gap by an order of magnitude, the truncation is not justified at these parameters. A two-band or full continuum HF calculation is needed to establish that the p-EI is not a projection artifact.","section":"Model / SM Sec. I.A"},{"comment":"The authors correctly state that Hartree-Fock overestimates symmetry-broken states and that quantum fluctuations will enlarge the metallic region and move the p-EI/NEI boundary to larger V_m. However, the abstract and introduction present the p-EI as a definitive prediction, despite this acknowledged caveat. The parameters highlighted in Fig. 4 (V_m=7 meV, V_D=0) lie precisely in the most fragile regime. Please either provide a beyond-mean-field estimate (e.g., RPA or QMC) for the p-EI region, or temper the central claim accordingly.","section":"Discussion, third paragraph"},{"comment":"The choice of pairing momentum Q=κ is load-bearing for the p-EI state, but the paper only states that 'the ground state energy is minimized when Q=±κ' without showing the energy comparison. Please provide the energy versus Q scan or a symmetry argument that rules out other momenta (e.g., Q=0). Without this evidence, the phase diagram is restricted to an assumed pairing channel.","section":"Mean-field theory, paragraph on pairing momentum Q"},{"comment":"The spin (valley) degree of freedom is neglected with the argument that spin-order energy scales are smaller than layer-order scales (footnote [49]). At ν=1 in a spinful system, the spinless p-EI state requires full spin polarization. The cited Ref. [56] supports kinetic ferromagnetism in a tunneling-coupled MoTe2/WSe2 model; its applicability to the hBN-separated bilayer is not demonstrated. A calculation, or at least a more detailed symmetry argument, is needed to show that spin-unpolarized competing states do not preempt the QAH phase.","section":"Model, spin neglect"}],"minor_comments":[{"comment":"Typo: 'transiton' should be 'transition'. In the Model section, 'Hamiltonain' should be 'Hamiltonian'.","section":"Introduction, first paragraph"},{"comment":"The band-structure plot would benefit from labeling the topmost miniband and the gap to the next band. The y-axis range (0–80 meV) obscures the relevant miniband gap, which is central to the projection issue.","section":"Fig. 2(b)"},{"comment":"The statement that a 15×15 k-grid and reciprocal-lattice cutoff at 4|g_1| 'ensures convergence' is not supported by a convergence test. Provide at least one check of the ground-state energy or phase boundary versus k-grid size.","section":"SM Sec. I.B"},{"comment":"The pseudospin texture plots are dense; please define the normalization of n_z and clarify the color scale. A zoomed inset near one vortex would improve readability.","section":"Fig. 4"}],"recommendation":"major_revision","confidential_remarks":"The manuscript reports a thought-provoking mean-field prediction, but the central p-EI phase sits in a regime where the single-band projection is questionable. I would push for a multi-band check before acceptance. The abstract's wording is also stronger than the caveats in the Discussion. The paper is within scope for the journal, but the load-bearing approximations need to be addressed or explicitly downgraded."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The genuinely new thing here is the device geometry: two TMD monolayers separated by twisted hBN, whose ferroelectric domains imprint opposite triangular moiré potentials on the two layers. That gives automatic lateral alignment, suppressed tunneling, and a clean platform for interlayer coherence. The p-wave excitonic insulator and its QAH signature are not new — they are carried over from the MoTe2/WSe2 heterobilayer proposals, Refs. [56,57], and the authors say so themselves. What they add is a concrete, experimentally testable stack where the coherent state is not contaminated by interlayer tunneling, plus a quantized Hall drag prediction that would be an unambiguous signature.\n\nThe mean-field calculation is honest and internally consistent. The p-EI phase emerges self-consistently, Q is chosen by energy minimization rather than fit, and the parameter scans in Fig. S1 show the phase is robust across a reasonable range of d, epsilon, alpha, and a_m. The Hartree-Fock ansatz, the Haldane-model analogy, and the Berry curvature / pseudospin diagnostics all hang together. The authors also flag the standard HF overestimation of symmetry-broken states, which is the right caveat to emphasize. I do not see circular fitting.\n\nThe soft spots, in order of seriousness. First, the single-band projection is uncontrolled precisely where the p-EI lives: for V_m ~ 6–8 meV and a_m ~ 8 nm, the miniband gap is a few meV while the Coulomb scale is ~30 meV. The claim that interactions do not cause significant band mixing is asserted, not checked. This is independent of the HF caveat and could kill the p-EI even if HF were exact in the two-band subspace. A simple check would be to repeat the self-consistent calculation keeping the next miniband and see whether the topological state survives. Second, the calculation is spinless and single-valley; the authors argue spin order is a smaller scale, and Ref. [56] backs kinetic ferromagnetism, but the intervalley competition is waved off rather than resolved. Third, no code or data file is provided, which slows verification of a 15x15-grid HF result. Minor points: the introduction says zero-field exciton condensation is experimentally unconfirmed while citing two 2025 Science papers on perfect Coulomb drag in dipolar excitonic insulators — that is a citation-pattern inconsistency worth fixing.\n\nWho is this for? Theorists working on moiré exciton condensates and experimental groups chasing zero-field interlayer coherence. It deserves a serious referee even if the final verdict is skeptical, because the geometry is realistic, the prediction is falsifiable, and the uncontrolled-projection issue is fixable in revision. I would send it to peer review, with the recommendation that the referee press for a two-band projection check.","headline":"A credible mean-field proposal for a zero-field chiral p-wave excitonic insulator in TMD/hBN/TMD stacks; the physics is borrowed from MoTe2/WSe2 theory, the geometry is new, and the main risk is the uncontrolled single-band truncation.","tokens_in":13941,"tokens_out":719,"would_cite":true,"duration_ms":10470,"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":"Twisted hBN between two TMD layers can produce a chiral p-wave exciton condensate that is both a Chern-number-one anomalous Hall insulator and a counterflow superfluid, at one hole per moiré cell, with no magnetic field.","keywords":["exciton condensate","quantum anomalous Hall effect","twisted hBN","transition metal dichalcogenide bilayers","moiré potential","chiral p-wave pairing","counterflow superfluidity","mean-field phase diagram"],"falsifier":"Measure Hall resistance and counterflow conductance in a device tuned to one hole per moiré cell, V_D=0, and V_m around 7 meV; if there is no quantized Hall plateau and no superfluid counterflow response, the predicted p-wave exciton condensate is absent. Equivalently, an unbiased numerical solution of the projected two-band model at those parameters that finds a metal or a topologically trivial insulator would falsify the mean-field prediction.","tokens_in":12940,"feed_emoji":"🌀","tokens_out":10803,"duration_ms":106198,"temperature":0.7,"pith_summary":"The paper predicts that two TMD monolayers separated by a twisted hBN spacer can form an equilibrium exciton condensate at zero magnetic field: at one hole per moiré unit cell, interlayer Coulomb interactions spontaneously establish phase coherence between the electrically isolated layers. The central result is a mean-field phase diagram showing that near zero displacement field and intermediate moiré modulation strength, the ground state is a chiral p-wave exciton condensate—a quantum anomalous Hall insulator with Chern number |C|=1 that also carries counterflow superfluidity. This matters because spontaneous interlayer coherence at zero field has been predicted but never confirmed in semiconductor bilayers, and the proposed geometry avoids the usual requirement of lattice alignment: the twisted hBN's ferroelectric moiré imprints opposite potentials on the two layers automatically. If the prediction holds, the p-wave state would be a concrete realization of a simultaneous topological insulator and exciton superfluid in a two-dimensional bilayer.","feed_headline":"Twisted hBN spacer predicted to make a zero-field Chern superfluid","feed_subtitle":"At one hole per moiré cell, two TMD layers could form a chiral exciton superfluid with quantized Hall response.","key_machinery":"The load-bearing object is the momentum-space layer pseudospin texture n_k = (sin θ_k cos φ_k, sin θ_k sin φ_k, cos θ_k) defined by the interlayer-coherent mean-field state. The paper shows that when pairing occurs at momentum shift Q = ±κ, the maximum of one band aligns with the minimum of the other, and the resulting mean-field Hamiltonian reduces to nearest-neighbor interlayer coherence amplitudes on a honeycomb lattice—formally the graphene tight-binding model with complex hopping. The winding of the pseudospin texture then determines the pairing symmetry: p_x ± i p_y vortices at κ and κ′ with opposite layer polarizations give a skyrmion texture and Chern number |C|=1 (the p-EI), while v","core_discovery":"On the paper's own terms, the discovery is that a bilayer of two TMD monolayers separated by a twisted hBN spacer, with interlayer tunneling suppressed and opposite triangular-lattice moiré modulations imposed on the two layers, spontaneously develops interlayer coherence at total filling ν=1. In the mean-field ground state, the layer pseudospin winds in momentum space in a chiral pattern: vortices at the κ and κ′ points carry opposite layer polarizations, producing a skyrmion texture with Chern number |C|=1 and therefore a quantum anomalous Hall effect without any Landau levels or magnetic field. The same state is an exciton condensate with counterflow superfluidity, since electrons and hol","pith_inferences":["If the p-wave condensate is confirmed, the same geometry is a natural platform to search for fractional Chern states at other fillings, since the spontaneous coherence generates topologically nontrivial bands without band-structure engineering.","The honeycomb-lattice analogy suggests that the p-EI is equivalent to a spontaneously generated complex nearest-neighbor hopping; a concrete extension would be to measure chiral edge conduction directly and test whether the edge current direction is set by the sign of V_D or by sample-specific disorder.","The authors' mean-field caveat points to a direct theoretical test: run an unbiased many-body simulation of the same single-band model at the predicted p-EI parameters (e.g., V_m≈7 meV, V_D=0, d=4 nm, ε=6) to see whether the Chern-number-one interlayer coherent state survives beyond the mean-field approximation.","Because the hBN spacer guarantees independent electrical contacts, counterflow measurements in this geometry could cleanly separate spontaneous interlayer coherence from single-particle tunneling artifacts, making it a sharper test of exciton condensation than tunnel-coupled heterobilayers."],"forward_implications":["At total filling ν=1, the TMD/hBN/TMD stack is predicted to be an insulator with spontaneous interlayer coherence over a wide range of displacement fields and modulation strengths, with no single-particle tunneling and no applied magnetic field.","Near V_D=0 and V_m around 6–8 meV (with d=4 nm, ε=6, a_m=8 nm), the ground state is the p-EI: a quantum anomalous Hall insulator with Chern number |C|=1 and counterflow superfluidity.","Because the twisted hBN spacer imprints the moiré potential remotely, no lateral alignment between TMD layers or between TMD and hBN is required—removing the main experimental obstacle to zero-field exciton condensation in semiconductor bilayers.","The p-EI should show a quantized Hall drag effect, analogous to bilayer quantum Hall exciton condensates, providing a clear experimental signature that would also support the excitonic scenario proposed for MoTe2/WSe2 heterobilayers.","The qualitative structure of the phase diagram—layer-polarized, s-wave, nematic, p-wave, and metallic regions—persists across changes in layer asymmetry, dielectric constant, interlayer distance, and moiré length, with the p-EI favored at weaker interactions and near V_D=0."],"fun_headline_variants":["Twisted hBN spacer enables zero-field Chern superfluid","Zero-field Chern superfluid from twisted hBN bilayer","Chiral exciton condensate with quantized Hall effect","Chern superfluid without magnetic field via twisted hBN","Topological excitonic insulator: zero-field Chern superfluid"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The prediction depends on the mean-field approximation being reliable in the region where the p-wave state appears; the authors themselves note that quantum fluctuations are likely to enlarge the metallic region and push the p-EI/nematic boundary to stronger interactions, so if fluctuations wipe out the p-wave state near zero displacement field, the central claim fails.","fun_headline_variants_meta":{"raw":{"variants":["Twisted hBN spacer enables zero-field Chern superfluid","Zero-field Chern superfluid from twisted hBN bilayer","Chiral exciton condensate with quantized Hall effect","Chern superfluid without magnetic field via twisted hBN","Topological excitonic insulator: zero-field Chern superfluid"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000948,"raw_usage":{"total_tokens":3882,"prompt_tokens":745,"completion_tokens":3137,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":489,"completion_tokens_details":{"reasoning_tokens":3067}},"tokens_in":489,"tokens_out":3137,"duration_ms":28712,"temperature":1.0,"reasoning_tokens":3067,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-04T17:12:05.875661+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure Hall resistance and counterflow conductance in a device tuned to one hole per moiré cell, V_D=0, and V_m around 7 meV; if there is no quantized Hall plateau and no superfluid counterflow response, the predicted p-wave exciton condensate is absent. Equivalently, an unbiased numerical solution of the projected two-band model at those parameters that finds a metal or a topologically trivial insulator would falsify the mean-field prediction.","supporting_citations":[],"review_version":1}