{"id":"4b80155a-7760-4c0d-95bf-b41dbfc220b6","arxiv_id":"2602.13599","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Zero-energy momenta of arbitrarily stacked multilayer graphene are exactly the union of the zero-energy momenta of each embedded parallel (AA) stack, enabling design of flat bands.","lead":"This paper derives where zero-energy electronic states appear in multilayer graphene with any stacking sequence, showing that parallel (AA) stacking blocks determine these momenta. It also shows that inserting rhombohedral (ABC) blocks into other stackings can widen flat bands, a possible route to strongly-correlated and superconducting graphene devices.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The iff claim for arbitrary stacks is asserted, not proved: the derivation covers AA blocks at one end, but skips the induction over internal parallel runs.","rationale":"The reader's formal weakest_assumption is the nearest-neighbor-only ideal chiral limit; that is indeed a real scope limitation, explicitly acknowledged in Section III.C. However, I judge the more load-bearing concern to be the incompleteness of the proof of the if-and-only-if statement itself: the derivation only treats parallel blocks at an end of the stack, then asserts the general run-decomposition result. The reader did mention 'the proof of the if-and-only-if statement for arbitrary stacks is incomplete as written' in the rationale, so there is partial agreement, but the reader's stated weakest_assumption is the model limitation rather than the proof gap. I tested the structure of the recurrence and small example stacks; the union rule appears consistent, so I do not want to escalate the verdict. The correct outcome remains CONDITIONAL: the claim is plausible and well-supported by examples, but the main theorem needs a rigorous induction (or a counterexample) before full acceptance. The concrete symbolic-determinant test would settle the mathematical concern directly, independently of the separate question of whether real graphene's additional hoppings preserve the rings.","tokens_in":16648,"tokens_out":35490,"duration_ms":329471,"concrete_test":"For a curated set of stacks, including ABBA, AABAA, AABCC, and BABABABCABCBCBCB, construct the exact Hamiltonian H_S(p) of Eq. (10) with v=t⊥=1, symbolically compute det H_S(p) as a polynomial in x=|p|^2, and compare its positive roots with the predicted set { [2 cos(rπ/(N_i+1))]^2 } over all maximal parallel runs. If any positive root lies outside the predicted set, the iff claim is false; if all roots match, the missing induction should be supplied in a revision to make the paper's central theorem rigorous.","verdict_should_be":"UNCHANGED","load_bearing_attack":"In Section III.C, the proof establishes two ingredients: (i) a stack with no parallel connections has determinant proportional to |p|^{2N}, so only |p|=0 (Eq. 28); and (ii) a stack with a parallel block at one end has zero-energy momenta at the isolated AA-block radii, via the Schur-complement/Chebyshev recurrence (Eqs. 31-49). The central claim, Eq. (56)-(57), then asserts that for any stack decomposed into maximal parallel runs S_i of lengths N_i, the zero-energy condition is exactly |p| in the union of { (2t⊥/v) cos(rπ/(N_i+1)) }. But the transition is a leap: the text says 'It is understood by its derivation that Eq. (49) is true for all arbitrary stacks' without carrying out an induction over the run decomposition. In particular, a parallel run buried between two non-parallel blocks is never analyzed; it is not shown that such an interior run contributes exactly the same radii and no additional roots, nor that two separated runs cannot combine to create accidental zeros. The 'only if' direction is thereby unproved for the general case, which is the load-bearing part of the paper's headline result. The claim may well be true — small explicit checks such as ABBA and AABAA are consistent with the union rule — but as written the main theorem rests on an omitted derivation rather than a completed proof.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies multilayer graphene with arbitrary stacking sequences within a nearest-neighbor tight-binding model containing a single interlayer hopping t⊥ between vertically aligned atoms. It derives or recalls the dispersions for pure AA, AB, and ABC stacks, and then analyzes stacks with stacking faults. The central analytical result is stated in Eqs. (56)–(57): for any stack decomposed into maximal parallel runs S_i of lengths N_i, a momentum p is a zero-energy point if and only if |p| belongs to the union of the isolated AA-run zero sets { (2t⊥/v) cos(rπ/(N_i+1)) }. The paper also presents DOS/IDOS calculations for rhombohedral stacks with parallel faults and DFT band structures for selected stacks. The conclusion proposes flat-band engineering by combining parallel and rhombohedral stackings.","tokens_in":17014,"tokens_out":12754,"duration_ms":119085,"significance":"If fully established, the result is an elegant and useful design rule: in the chiral nearest-neighbor limit, the zero-energy manifold of an arbitrary stack is completely determined by its maximal AA blocks, and the quantitative ring positions are parameter-free predictions. The paper is self-contained, uses no fitted parameters (only t, t⊥, and a DOS smearing width), and the DFT/DOS calculations are independent checks rather than inputs. The determinant recurrence for a single AA block at the edge and the Chebyshev solution are genuine analytic contributions. However, the generalization from one edge AA block to arbitrary run decompositions is asserted rather than proved, and the real-graphene robustness of the rings is not quantitatively tested. Both issues are repairable but are load-bearing for the paper's headline claim.","major_comments":[{"comment":"The 'if and only if' statement for arbitrary stacks is not proved. The derivation establishes two ingredients: (i) stacks with no parallel connections have only the |p|=0 zero (Eq. (28)); (ii) a stack with a single parallel block at one end has zero-energy momenta given by the AA-block radii (Eqs. (29)–(49)). The step to arbitrary stacks is the sentence 'It is understood by its derivation that Eq. (49) is true for all arbitrary stacks,' followed by the decomposition into maximal runs. No induction over multiple runs or interior runs is given. In particular, the Schur-complement argument in Eq. (31) is applied from one end with a monolayer top block; for a run buried between non-parallel layers the same argument must be repeated from the other side or replaced by a different factorization, and the text does not show that an interior run contributes exactly its isolated radii and no additi","section":"Section III.C, Eqs. (49)–(57)"},{"comment":"The conclusion states that the formula is 'quantitatively accurate for finding the rings of zero energy,' but the analytical result is obtained in a nearest-neighbor chiral model with a single t⊥. Actual graphene has Slonczewski-Weiss-McClure hoppings and trigonal warping that can gap or distort these rings. The paper acknowledges this limitation in Section III.C but does not test it quantitatively. The DFT data in Fig. 7 are band dispersions along Γ→K→M; they give only a visual, rough match to the predicted crossings and do not confirm the closed-ring structure implied by Eq. (56). I recommend either (a) restating the quantitative claim as a theorem about the nearest-neighbor model, or (b) adding a full SWMcC calculation for at least one representative stack (e.g., AAABCCC) and comparing the actual zero-energy contours with Eq. (56).","section":"Section IV / Fig. 7"}],"minor_comments":[{"comment":"The definition of p_{r_i}^i in Eq. (56) lacks an absolute value, although radii must be non-negative. The text later says one may take absolute values; the definition should match Eq. (17).","section":"Eq. (56)"},{"comment":"The notation S = P_{i=1}^M S_i for string concatenation is unusual and undefined; use e.g. S = S_1 S_2 ⋯ S_M to avoid confusion with a sum.","section":"Eq. (55)"},{"comment":"Typo: 'psuedopotentials' should be 'pseudopotentials'. Also, 'for |p > Δ|p|' is missing a norm bar; it should read 'for |p| > Δ|p|'.","section":"Section II / after Eq. (23)"},{"comment":"Reference [23] (Bohr) appears unrelated to the statement about electrons filling zero-energy regions; please check whether a different citation is intended.","section":"Reference [23]"},{"comment":"The abstract and introduction state that low-energy dispersions may be deduced from substacks in isolation. In the body this is demonstrated mainly for zero-energy momenta and flat-band presence, not for the full dispersion. Please qualify the wording.","section":"Abstract / Introduction"}],"recommendation":"major_revision","confidential_remarks":"The manuscript shows good physical intuition, and the local derivations are mostly sound. My main concern is the missing induction for the arbitrary-stack claim; I would not accept until that is supplied or the claim is explicitly downgraded to a conjecture supported by numerical checks. The DFT comparison is too coarse to validate the quantitative ring positions in real graphene, so the real-material claim should also be softened or tested with a fuller hopping model."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: the union rule for zero-energy momenta (Eqs. 56-57) is a genuine contribution, not just repackaged Koshino-McCann. It explicitly predicts that only AA runs matter, which is new. The paper does a lot right: the derivation for a single AA block is coherent, uses no fitted parameters, and the DFT comparisons are honest independent checks that support the qualitative picture. The DOS/IDOS flat-band enhancement story is credible and useful for the rhombohedral-graphene program.\n\nThe soft spot is exactly what the stress-test says. In Section III.C the proof nails down (i) stacks with no parallel connections, determinant ~ |p|^{2N}, and (ii) a single parallel block at one end, via Schur complement and Chebyshev. The jump to arbitrary stacks rests on 'It is understood by its derivation that Eq. (49) is true for all arbitrary stacks' — no explicit induction over interior runs, no demonstration that two separated runs cannot produce accidental zeros, no treatment of a run buried between non-parallel blocks. That's the load-bearing 'if and only if.' It may be true; ABBA and AABAA checks work; but the paper doesn't prove it. This is fixable — write out the induction or a matrix factorization — but as submitted the gap is real.\n\nOther soft spots are minor. The novelty split with Sarsfield et al. and Koshino-McCann is not clearly drawn; Section III.A-B is mostly textbook review. The nearest-neighbor-only idealization is acknowledged in III.C, and the DFT alignment is rough, so the quantitative ring positions are approximate in real graphene. No code or data is provided, which makes the explicit stack checks hard to reproduce.\n\nBottom line: this deserves a serious referee. The central idea is plausible, analytically coherent in the single-block case, and likely correct in general, but the main theorem needs a complete proof. I'd send it out and ask for the induction plus a clearer statement of what is new vs. prior work.","headline":"Genuine analytical result with a real proof gap in the general 'iff' claim; worth refereeing but needs an explicit induction over interior runs.","tokens_in":17438,"tokens_out":1608,"would_cite":false,"duration_ms":16005,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["73.22.Pr","71.15.Mb"],"model":"deepseek-v4-flash","headline":"A multilayer graphene stack of any stacking order has its zero-energy momenta fully determined by its maximal same-letter (parallel) runs: those momenta are exactly the AA-stack roots, |p| = (2t⊥/v) cos(rπ/(N+1)).","keywords":["multilayer graphene","stacking order","tight-binding model","zero-energy states","flat bands","rhombohedral stacking","AA/parallel stacking","band structure engineering"],"falsifier":"A full tight-binding or first-principles calculation for a stack with a long same-letter run, such as AAAABBB or AAABCCC, including the standard extra interlayer hoppings, should be compared with Eq. (56): if any of the predicted zero-energy rings splits or gaps out at momenta |p|=2t⊥/v cos(rπ/(N+1)), the 'if and only if' statement fails quantitatively; if the rings survive with energies below measurement resolution, the decomposition rule holds.","tokens_in":16579,"feed_emoji":"⚛️","tokens_out":8207,"duration_ms":67265,"temperature":0.7,"pith_summary":"Using a nearest-neighbor tight-binding model, the paper shows that the low-energy electronic structure of multilayer graphene with any stacking order can be deduced from its substacks in isolation. Its central claim is that the momenta at which the Hamiltonian has a zero eigenvalue are exactly the union of the zero-energy momenta of its maximal same-letter (parallel) runs: |p| must equal 2t⊥/v times cos(rπ/(N+1)) for some run length N and index r. Bernal and rhombohedral connections by themselves contribute nothing away from |p|=0, while rhombohedral (ABC) substacks reproduce their characteristic flat band when embedded in any other stacking, and adding parallel runs extends the flat-band radius to at least twice the pure-ABC value. If correct, this turns stacking sequence into a quantitative design rule for placing zero-energy crossings and flat bands, and it explains why multilayer graphene without twisting can host correlated-electron physics.","feed_headline":"One formula predicts zero-energy bands of any graphene stack","feed_subtitle":"Stacking order becomes a design knob: embed a rhombohedral run to keep its flat band; add parallel runs to double the flat-band radius.","key_machinery":"The central object is the 2N×2N tight-binding Hamiltonian H_S(p) built from monolayer Dirac blocks vσ·p and interlayer hopping t⊥ between vertically aligned atoms, with stacking encoded by a letter A, B, or C. The proof rides on a determinant recurrence: peeling off the top layer expresses det H in terms of the determinant of the remaining substack; for non-AA links this reduces to det H ∝ |p|^{2N}, zero only at |p|=0. For an AA link, a block-determinant formula replaces the low-energy block by an effective momentum factor (1−Q_n)vπ, and the recurrence Q_{n+1}=t⊥²/((1−Q_n)v²|p|²) is solved by recognizing x f_n(x)=U_{n+1}(x/2)/U_n(x/2) in terms of Chebyshev polynomials of the second kind. The","core_discovery":"On the paper's own terms, the discovery is a determinant identity and its consequence: for an arbitrary N-layer stacking sequence S decomposed into maximal parallel runs S_i of lengths N_i, the condition det H_S(p)=0 is equivalent to |p| ∈ union over i of { (2t⊥/v) cos(rπ/(N_i+1)) : r=1,...,N_i }. Bernal and rhombohedral links only pass the determinant condition down the stack and force |p|=0 on their own; the sole sources of nonzero-momentum zero modes are AA blocks. The same machinery shows that an embedded ABC substack pins a flat band to the full stack—localizing at the substack's edges when embedded in Bernal material—and that parallel runs move additional zero crossings outward, giving","pith_inferences":["A directly testable prediction follows from the radius bound: scanning tunnelling spectroscopy on AA-plus-ABC stacks should show a wider low-energy DOS plateau than pure ABC of comparable thickness, with the plateau width scaling as 2t⊥/v rather than t⊥/v.","Because the zero-energy roots are set only by run lengths, stacking sequence becomes a discrete design parameter: one could engineer stacks with several zero-energy rings at chosen radii simply by choosing parallel-run lengths—an untwisted route to flat-band platforms.","The determinant retention mechanism is not obviously specific to graphene: any bipartite layered lattice with AA-style vertical registry should show the same cosine-root spectrum, so the design rule may transfer to other van der Waals materials.","The paper leaves open whether the zero-energy rings carry nontrivial topology; if the rings are protected by the chiral (sublattice) symmetry of the nearest-neighbor model, realistic symmetry-breaking terms would gap them, which is the main quantitative risk to the rule."],"forward_implications":["Any stack containing a same-letter run of length N acquires zero-energy band crossings on rings of radius 2t⊥/v cos(rπ/(N+1)) around each Dirac point; the ring radii are determined by that run alone.","Stacks with no same-letter runs of length ≥2 (pure Bernal/rhombohedral alternation) have zero-energy states only at |p|=0; any flat bands there come from embedded ABC substacks, not from the alternating parts.","Embedding an ABC substack yields the ABC flat band in the full stack; in a Bernal host the low-energy wavefunction localizes at the edges of the ABC block, while in an AA host it does not.","Parallel runs combined with rhombohedral substacks extend the flat-band radius to at least 2t⊥/v in the large-N limit, twice the pure-ABC radius of t⊥/v, with a correspondingly higher density of states near the Fermi level.","Density-functional calculations for stacks such as AAABCCC and AAAAAABCABCCCCCC show flat bands extending beyond the pure-ABC limit, with the parallel-fault crossings roughly matching Eq. (56) along Γ–K–M."],"fun_headline_variants":["Stacking sequence spells out graphene's zero-energy bands","One rule predicts flat bands in any graphene stack","Graphene's flat bands: a stack-by-stack formula","Zero-energy bands from arbitrary stacking orders","Stacking runs pin flat bands in graphene multilayers"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The whole result assumes that only nearest-neighbor intralayer hopping and a single vertical interlayer hopping t⊥ are present; real graphene has additional interlayer couplings and trigonal warping that can open gaps or distort the predicted zero-energy rings, so the exact radii in Eq. (56) are only guaranteed in that ideal model.","fun_headline_variants_meta":{"raw":{"variants":["Stacking sequence spells out graphene's zero-energy bands","One rule predicts flat bands in any graphene stack","Graphene's flat bands: a stack-by-stack formula","Zero-energy bands from arbitrary stacking orders","Stacking runs pin flat bands in graphene multilayers"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000157,"raw_usage":{"total_tokens":1041,"prompt_tokens":707,"completion_tokens":334,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":451,"completion_tokens_details":{"reasoning_tokens":259}},"tokens_in":451,"tokens_out":334,"duration_ms":3915,"temperature":1.0,"reasoning_tokens":259,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-02T23:30:57.354722+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A full tight-binding or first-principles calculation for a stack with a long same-letter run, such as AAAABBB or AAABCCC, including the standard extra interlayer hoppings, should be compared with Eq. (56): if any of the predicted zero-energy rings splits or gaps out at momenta |p|=2t⊥/v cos(rπ/(N+1)), the 'if and only if' statement fails quantitatively; if the rings survive with energies below measurement resolution, the decomposition rule holds.","supporting_citations":[],"review_version":1}