{"id":"61deba20-8890-4461-8c3a-c83a6cac3639","arxiv_id":"2505.05414","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"Electron-electron interactions in slightly doped rhombohedral multilayer graphene are predicted to create a time-reversal-breaking Fermi surface shaped like a crescent (the Fermi lune), giving large non-reciprocal resistance and in-plane orbital magnetization.","lead":"A theory and transport study reports a new interaction-driven Fermi surface shape, a crescent-like Fermi lune, in five-to-nine-layer rhombohedral graphene, together with giant current-direction-dependent resistance and an anomalous Hall effect. The paper argues this is a form of orbital magnetism that appears only when the sample thickness lies between two and three dimensions.","discovery_kind":"first_principles","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 22.9% non-reciprocal signal is not shown to be an intrinsic single-domain Fermi-lune signature: six degenerate TOM domains and no quantitative δRxx/Rxx calculation leave contact, domain, and heating artifacts uncontrolled.","rationale":"The reader's weakest assumption is the same as the load-bearing concern I identify: the non-reciprocal resistance and the transdimensional AHE are assumed to be intrinsic single-domain signatures of the Fermi-lune Hartree-Fock state rather than artifacts. The paper's own theory predicts six degenerate TOM domains, so the measured 22.9% asymmetry requires either a domain imbalance or an external bias that is not described. The absence of any quantitative calculation of δRxx/Rxx from the TOM band structure is a concrete gap: Table I establishes only symmetry permission, not magnitude or sign. I did not find an obvious algebraic error in the Hartree-Fock or renormalization-group construction, and the phase diagrams are internally consistent, so the theoretical Fermi-lune state may well exist. The issue is the evidential link to the single-device transport data. The proposed field-poling test would settle whether the non-reciprocal signal is domain-magnetization driven; if it fails, the conditional verdict should be lowered, because the only in-paper experimental support for the Fermi-lune state would then be absent. Since the reader already captured this as the weakest assumption and assigned CONDITIONAL, no verdict change is needed.","tokens_in":34588,"tokens_out":6290,"duration_ms":82950,"concrete_test":"Cool the same 9-layer device to base temperature in an in-plane field By = +0.5 T to pole the TOM_y domains, measure Rxx(+I) and Rxx(-I) at D = 0.9 V/nm and n = 1.5×10^12 cm^-2; then warm above the ordering temperature and cool in By = -0.5 T and repeat. An intrinsic single-domain Fermi-lune signal should reverse the sign of δRxx = Rxx(+I) - Rxx(-I) between the two poling states and should be suppressed after zero-field cooling; a heating, contact, or Hall-mixing artifact will not show this controlled sign reversal.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The experimental pillar of the central claim is the giant non-reciprocal longitudinal transport in Sec. IV: δRxx/Rxx = 22.9% at D = 0.9 V/nm and n = 1.5×10^12 cm^-2 (Fig. 3(c)). For this to count as evidence for a Fermi-lune state, the signal must be intrinsic to a single TOM domain and odd in current with a sign set by the domain's orbital magnetization. However, the theory in Sec. III gives six energetically equivalent TOM_y domains connected by C3, T, and My operations (Fig. 2(f)). In a multi-domain sample, the odd-in-I longitudinal asymmetry of different domains cancels, and no poling protocol or domain-imbalance estimate is described for the measurement in Fig. 3. The paper also provides no quantitative transport calculation of δRxx/Rxx from the TOM_y band structure; Table I only lists that δRxx is symmetry-allowed. Without such a calculation, the measured magnitude and sign are not compared with any theoretical prediction, so current-induced heating, contact rectification, or admixture of the companion paper's large Hall voltage [17] into Rxx cannot be excluded. This concern is load-bearing because, if the non-reciprocal signal is not intrinsic, the only direct transport experiment in this paper supporting the Fermi-lune state is weakened; the AHE evidence is deferred to [17]. The mismatch between the calculated phase diagram (D ≤ 0.7 V/nm, Fig. 4) and the measured TDAHE region (D ≈ 0.75-0.95 V/nm, Fig. 3(a)) compounds the gap.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports a combined theory-and-transport study of slightly electron-doped rhombohedral multilayer graphene (RMG) with 4 to 9 layers. The central claim is that electron-electron interactions drive a spontaneous symmetry-broken metallic state, the 'Fermi lune,' in which the Fermi surface has a crescent shape and breaks time-reversal, C3, and mirror symmetries. This state is argued to produce large in-plane orbital magnetization ('transdimensional orbital magnetism'), a transdimensional anomalous Hall effect, and giant intrinsic non-reciprocal longitudinal transport. Unrestricted self-consistent Hartree-Fock calculations are presented for several layer numbers, showing TOM_y, TOM_x, and SVP phases as a function of density and displacement field. The experimental part reports a 22.9% non-reciprocal longitudinal resistance signal in a 9-layer device at D = 0.9 V/nm and n = 1.5e12 cm^-2, together with the absence of quantum oscillations in the claimed Fermi-lune regime. A superlattice-potential version of the state is also shown to produce a Chern band with quantized anomalous Hall response to in-plane fields.","tokens_in":34941,"tokens_out":3037,"duration_ms":36954,"significance":"If the central claims hold, the paper identifies a genuinely new class of interaction-driven Fermi-surface topology in a widely studied material platform, with an unusual combination of broken symmetries and orbital magnetism. The theoretical work is ambitious and largely transparent: the unrestricted Hartree-Fock search over 128 order parameters, the explicit phase diagrams for N = 4 through 9 layers, and the numerical evaluation of in-plane orbital magnetization via coupling to a small in-plane field are concrete and useful. The model parameters are taken from prior literature rather than fitted to the transport data, and the measured non-reciprocal signal is not used as an input, so the core theoretical prediction is not circular. However, the experimental identification of the Fermi-lune state rests on transport signatures whose intrinsic, single-domain origin is not established by the present manuscript, and the calculated phase diagram does not extend into the experimentally claimed region. These gaps are substantive because the paper's headline experimental number, 22.9%, is not accompanied by a quantitative transport calculation.","major_comments":[{"comment":"The claim that the measured 22.9% non-reciprocal resistance is an intrinsic signature of the Fermi-lune state is not supported by any quantitative transport calculation. Table I only establishes that delta_Rxx is symmetry-allowed, while the text states that the asymmetry of forward and backward Fermi velocities produces the effect; no Boltzmann or Kubo calculation of delta_Rxx/Rxx from the TOM_y band structure is given, so neither the magnitude, sign, nor density dependence of the measured signal is compared with theory. Without such a calculation, current-induced heating, contact rectification, or admixture of the Hall voltage into Rxx cannot be excluded, and the central experimental claim remains unquantified.","section":"Sec. IV, Fig. 3(c)"},{"comment":"The six degenerate TOM_y domains shown in Fig. 2(f) would have opposite non-reciprocal responses for opposite in-plane orbital magnetizations, so a multi-domain sample would see cancellation of the odd-in-current longitudinal asymmetry. The measurement in Fig. 3 is not accompanied by any poling protocol, domain-imbalance estimate, or single-domain detection, and the paper itself acknowledges in Sec. VI that six domains will form domain walls and Z6 vortices. The authors need to explain how the observed large odd-in-I signal survives in a multi-domain sample, or demonstrate single-domain behavior, before the signal can be attributed to the intrinsic Fermi-lune state.","section":"Sec. IV, Fig. 2(f) and Sec. VI"},{"comment":"The calculated Hartree-Fock phase diagram in Fig. 4(a) for N = 9 covers displacement fields only up to D = 0.7 V/nm, while the measured transdimensional AHE and non-reciprocal transport region in Fig. 3(a) is D ≈ 0.75–0.95 V/nm. The paper describes the agreement as 'perfect,' but no calculation is shown in the measured field range. The authors should extend the HF calculations to the measured range, or present a quantitative argument (for example, using renormalized parameters) for why the TOM_y phase should persist to D ≈ 0.9 V/nm.","section":"Sec. V, Fig. 4(a) and Fig. 3(a)"},{"comment":"The theoretical phase diagram depends on the low-energy window E*_C ≈ 0.3 eV, the cutoff ratio L_s/(n_cut a0), and the RG treatment that assumes approximate particle-hole symmetry, as stated in Supp. S4. No sensitivity analysis is provided for these choices, nor is the dependence on the 46x46 k-mesh or the n_cut = 1 band truncation discussed. Since the Fermi-lune state is the central prediction, the authors should show that the TOM_y phase and its orbital magnetization are robust to reasonable variations of these numerical and RG parameters, and should quantify the effect of the particle-hole breaking terms that are neglected in the RG flow.","section":"Supp. S4, S5"}],"minor_comments":[{"comment":"The header for the last column reads 'δRxx = σ+xx−R−xx', which mixes conductivity and resistance notation; this should be corrected to R+xx − R−xx or defined consistently.","section":"Table I"},{"comment":"There are several typographical errors, including 'transidimensional' for 'transdimensional,' 'foward-moving' for 'forward-moving,' 'ennealayer' for 'nine-layer,' 'stongally' for 'trigonally,' and 'vanishment' for 'vanishing.' These should be corrected.","section":"Throughout"},{"comment":"The text states that the non-reciprocal signal is 'in perfect agreement with theoretical expectation,' but no theoretical curve or quantitative prediction is shown; the wording should be softened until such a comparison is provided.","section":"Sec. IV"},{"comment":"The definition of the measured Rxx should state explicitly whether the four-terminal longitudinal voltage contacts exclude the Hall contribution, and how the AC lock-in current sign is reversed to define Rxx(+I) and Rxx(−I).","section":"Supp. S1"},{"comment":"Reference [33] is cited as 'Manuscript in preparation (2025)' for a more comprehensive experimental study; if possible, the authors should provide additional details of the planned study or remove the reliance on unpublished work.","section":"Sec. VI / Ref. [33]"}],"recommendation":"major_revision","confidential_remarks":"The paper's experimental pillar is partly deferred to the companion manuscript [17], which supplies the transdimensional AHE data; the present manuscript's self-contained experimental evidence is the non-reciprocal transport measurement. Given that the central identification of the Fermi-lune state depends on both experiments, the editor may wish to ensure that Ref. [17] is available for cross-checking before final acceptance. The lack of a quantitative transport calculation and the D-range mismatch are the main technical obstacles."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Fermi lune is a real addition to the vocabulary. The core idea: in slightly doped RMG with 5-9 layers, the non-interacting double-ring Fermi surface is unstable to an interaction-driven state that breaks T, C3, and mirror symmetry and supports a crescent-shaped FS. This state has not been discussed in prior RMG literature, which focused on SVP and superconductivity. The paper does the symmetry analysis properly — Table I is clear, the six TOM_y domains and their C3/T/My relations are spelled out, and the phase diagram vs NL, n, D is a useful map. The prediction of in-plane orbital magnetization 1-10 μB per electron in the transdimensional regime is concrete and testable, and the idea that orbital currents circulate both in-plane and out-of-plane is worth taking seriously.\n\nThe soft spots are where theory meets experiment. The transport pillar is a single 9-layer device. The headline number δRxx/Rxx=22.9% is not compared quantitatively to any band-structure calculation — Table I only says δRxx is symmetry-allowed. With six degenerate TOM domains, the odd-in-I non-reciprocal response should cancel in a multi-domain sample unless there is domain imbalance or poling, and the paper supplies neither. I don't think that kills the theory — the AHE in the companion paper may well be the stronger evidence — but the claim of 'unambiguous evidence' goes beyond what the paper shows. Also, the calculated phase diagram stops at D=0.7 V/nm while the measured effect sits at 0.75-0.95 V/nm. That mismatch could be a parameter-renormalization artifact, but it is not discussed. The Hartree-Fock scheme has two tuning knobs (E_C, L_s/(n_cut a0)) and no robustness scan, so I would not treat the phase boundaries as quantitative.\n\nNet: this is a theory paper with an experimental garnish. The theory is novel and plausible; the experiment is suggestive but incomplete. A good referee should ask for a transport calculation, a domain analysis, and a phase-diagram extension before publication. The paper deserves peer review, not a desk reject. I'd bring it to a reading group if the group likes RMG; otherwise read it when the next version lands.","headline":"A genuinely new mean-field metallic state ('Fermi lune') in rhombohedral multilayer graphene, with a solid symmetry analysis and a plausible phase diagram — but the transport pillar is a single device with no quantitative theory comparison, so the paper needs serious referee revision, not a desk reject.","tokens_in":35482,"tokens_out":3256,"would_cite":true,"duration_ms":33728,"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":"Electron interactions in five-to-nine-layer rhombohedral graphene reshape the Fermi surface into a crescent that breaks time-reversal symmetry, generating in-plane orbital magnetism and current-direction-dependent resistance as large as…","keywords":["Fermi lune","rhombohedral multilayer graphene","transdimensional orbital magnetism","non-reciprocal transport","Hartree-Fock theory","spontaneous symmetry breaking","anomalous Hall effect","Chern insulator"],"falsifier":"Cool a 9-layer device into the TOM phase under in-plane magnetic fields of opposite directions and measure $\\delta R_{xx}/R_{xx}$ at fixed carrier density and displacement field; the Fermi-lune mechanism predicts that the sign of the non-reciprocity should follow the selected magnetic domain, whereas a heating or contact artifact would be insensitive to the cooling-field direction.","tokens_in":1917,"feed_emoji":"🌙","tokens_out":2382,"duration_ms":98297,"temperature":0.7,"pith_summary":"The paper claims that in slightly electron-doped rhombohedral multilayer graphene with roughly five to nine layers, electron-electron interactions spontaneously reorganize the Fermi surface into a crescent shape, named the Fermi lune, that breaks time-reversal, threefold-rotation, and mirror symmetries. In this transdimensional regime, where the film is thick enough to support vertical orbital motion yet thin enough to remain quantum coherent, the lune hosts circulating currents that produce magnetic moments both within and perpendicular to the layers, a state the authors call transdimensional orbital magnetism. The same broken symmetries make forward and backward electrical currents experience very different Fermi velocities, yielding large intrinsic non-reciprocal longitudinal resistance; the measured value reaches 22.9% in a nine-layer device. If correct, the work establishes a new symmetry-broken metallic state whose transport and magnetic responses are controlled by Fermi-surface geometry rather than by spin.","feed_headline":"Crescent Fermi surface drives 22.9% one-way resistance","feed_subtitle":"Interactions in five-to-nine-layer rhombohedral graphene make a crescent Fermi surface and an orbital magnet.","key_machinery":"The Fermi lune is the central object: a crescent-shaped Fermi-energy contour that emerges from the double-ring non-interacting Fermi surface once long-range, double-gate-screened Coulomb interactions are treated with unrestricted Hartree-Fock, a self-consistent mean-field treatment of the electron-electron interaction. The argument's load-bearing steps are a perturbative renormalization-group procedure that enhances the low-energy continuum model parameters through remote-band screening, and projection of the Coulomb interaction onto a low-energy window ($E_C^*\\sim 0.3$ eV) with one conduction and one valence band per spin and valley. The lune's geometry, a strongly asymmetric distribution of occupied states, directly carries the non-reciprocity and the orbital magnetism; the in-plane orbital magnetization $M_y \\sim \\langle \\hat{z}\\hat{v}_x\\rangle$ scales with sample thickness, which is why the effect grows with layer number.","core_discovery":"The central discovery is a new interaction-driven ground state in rhombohedral multilayer graphene: the Fermi lune. Starting from a non-interacting double-ring Fermi surface in the presence of a displacement field, long-range intra-valley Coulomb interactions drive a spontaneously symmetry-broken metal whose Fermi contour is a crescent that breaks time-reversal ($T$), threefold rotation ($C_3$), and mirror ($M_y$) symmetries while preserving the combined $M_yT$ symmetry (the TOM-$y$ state) or, in a second variant, breaks $C_3$ and $M_yT$ while preserving $M_y$ (the TOM-$x$ state). The lune produces strongly asymmetric Fermi velocities for forward- and backward-moving carriers, identified as the microscopic origin of giant non-reciprocal longitudinal transport. It also supports coherent orbital current loops both in-plane and out-of-plane, giving orbital magnetizations of order $1$-$10\\,\\mu_B$ per electron whose in-plane component grows nearly linearly with layer number. Hartree-Fock phase diagrams for four to nine layers show the TOM states occupying most of the parameter space for seven and nine layers, and a superlattice-potential calculation shows that the lune can fold into an isolated Chern-number-1 conduction band, a transdimensional Chern insulator whose quantized anomalous Hall effect responds hysteretically to an in-plane magnetic field.","pith_inferences":["A direct test the paper does not perform: if a single Fermi-lune domain can be prepared, for instance by cooling in a weak in-plane magnetic field, the non-reciprocal resistance should become antisymmetric under current reversal and its sign should flip when the domain is switched, whereas a heating artifact would not track the domain orientation.","The paper stops short of computing $\\delta R_{xx}/R_{xx}$ from the Hartree-Fock band structure; doing so would let theory predict the size, sign, and density dependence of the 22.9% signal and help separate the intrinsic lune mechanism from contact or mixing artifacts.","The same lune mechanism may appear in other layered metals once the out-of-plane mean free path is comparable to the film thickness; the paper names multilayer transition metal dichalcogenides as a natural venue, so checking for current-direction-dependent resistance there is a testable extension."],"forward_implications":["The Fermi-lune state should show no Shubnikov-de Haas oscillations in out-of-plane magnetic fields, because the lune's asymmetric velocities suppress closed cyclotron orbits; the paper reports this absence up to $12$ T.","The non-reciprocal longitudinal resistance and the transdimensional anomalous Hall effect should appear and vanish together as density or displacement field crosses the TOM phase boundary; the measured phase boundary in the nine-layer device is consistent with this expectation.","In seven- and nine-layer samples the TOM states dominate the mean-field phase diagram, whereas four-layer systems are mostly spin/valley-polarized, predicting a sharp layer-number threshold for the lune.","Coupling a lune state to a superlattice potential of the right period should produce a correlated Chern insulator with Chern number $1$ and in-plane orbital magnetization on the order of $3\\,\\mu_B$ per electron, with a quantized anomalous Hall effect that is hysteretic in an in-plane magnetic field.","The six degenerate magnetic domains of the TOM-$y$ state imply domain walls and $\\mathbb{Z}_6$ vortices at their intersections, so the thermal ordering transition should be governed by proliferation of those defects."],"supporting_citations":[{"why":"Supplies the experimental transdimensional anomalous Hall effect data that indicate spontaneous in-plane orbital magnetization in the same 9-layer rhombohedral graphene device.","marker":"[17]"},{"why":"Provides the bulk graphite conductivity data used to estimate the vertical mean free path $l_\\perp\\sim 2$ nm, defining the transdimensional regime the lune requires.","marker":"[18]"},{"why":"Provides the perturbative renormalization-group treatment of remote-band electrons that renormalizes the low-energy continuum model parameters used as input to the Hartree-Fock calculation.","marker":"[30, 31]"},{"why":"Gives the comparable 22.9%-scale non-reciprocal transport signal in a quantum anomalous Hall insulator, establishing the benchmark against which the present signal is measured.","marker":"[5]"},{"why":"Reports the absence of quantum oscillations in the TOM phase up to $12$ T, evidence that electrons on the Fermi lune do not complete closed cyclotron orbits.","marker":"[32]"},{"why":"Documents the flavor-polarized metallic phases that compete with and contrast against the TOM states in the Hartree-Fock phase diagram.","marker":"[19–23]"}],"fun_headline_variants":["Crescent Fermi surface unlocks new orbital magnetism","Fermi lune drives giant non-reciprocal transport in graphene","Transdimensional orbital magnetism from a Fermi lune","Graphene's crescent Fermi surface spawns Chern insulator","Symmetry-broken Fermi lune: orbital currents and non-reciprocity"],"cache_read_input_tokens":37504,"weakest_assumption_plain":"The paper treats the measured non-reciprocal resistance and the companion transdimensional anomalous Hall effect as intrinsic, single-domain signatures of the Fermi-lune Hartree-Fock state rather than artifacts of current-induced heating, multi-domain averaging, contact effects, or mixing of longitudinal and Hall voltages.","fun_headline_variants_meta":{"raw":{"variants":["Crescent Fermi surface unlocks new orbital magnetism","Fermi lune drives giant non-reciprocal transport in graphene","Transdimensional orbital magnetism from a Fermi lune","Graphene's crescent Fermi surface spawns Chern insulator","Symmetry-broken Fermi lune: orbital currents and non-reciprocity"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000281,"raw_usage":{"total_tokens":1698,"prompt_tokens":1012,"completion_tokens":686,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":628,"completion_tokens_details":{"reasoning_tokens":600}},"tokens_in":628,"tokens_out":686,"duration_ms":6167,"temperature":1.0,"reasoning_tokens":600,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T23:05:19.394456+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Cool a 9-layer device into the TOM phase under in-plane magnetic fields of opposite directions and measure $\\delta R_{xx}/R_{xx}$ at fixed carrier density and displacement field; the Fermi-lune mechanism predicts that the sign of the non-reciprocity should follow the selected magnetic domain, whereas a heating or contact artifact would be insensitive to the cooling-field direction.","supporting_citations":[{"cited_title":"Transdimensional anomalous Hall effect in rhombohedral thin graphite","cited_arxiv_id":"2505.03891","evidence_quote":"Supplies the experimental transdimensional anomalous Hall effect data that indicate spontaneous in-plane orbital magnetization in the same 9-layer rhombohedral graphene device."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the bulk graphite conductivity data used to estimate the vertical mean free path $l_\\perp\\sim 2$ nm, defining the transdimensional regime the lune requires."},{"cited_title":"Yasuda, T","cited_arxiv_id":null,"evidence_quote":"Gives the comparable 22.9%-scale non-reciprocal transport signal in a quantum anomalous Hall insulator, establishing the benchmark against which the present signal is measured."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Reports the absence of quantum oscillations in the TOM phase up to $12$ T, evidence that electrons on the Fermi lune do not complete closed cyclotron orbits."}],"review_version":1}