{"id":"e78ba0ac-9dc0-4ee2-b224-dd0f4e79299a","arxiv_id":"2505.09908","paper_version":1,"verdict":"REJECT","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"A modified RCWA with non-zero-flux evanescent bases computes twisted-bilayer photonic-slab eigenmodes, and a five-layer uniform-slab model with fitted coupled-mode theory explains the resonance spectra and the 0.7 c/a transition frequency.","lead":"The authors modify a standard electromagnetic simulation method so it can compute the eigenmodes of two twisted photonic crystal slabs, and they approximate the resonances with a five-layer uniform slab model. The work aims to give optical engineers a simpler design tool for twisted bilayer photonic devices such as beam steering and lasers.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The modified evanescent basis underlying the claimed eigenmode calculation is not flux-normalized, not shown unitary, and is never benchmarked; until it is, the central claim is unsupported.","rationale":"The abstract's strongest claim is that eigenmode calculation 'has not been realized before'; the five-layer approximation and the CMT transition are secondary and, in the paper, are validated against the modified RCWA itself rather than against independent data. Thus the modified basis is the keystone. My reading of Table S1 and Supplement Sec. 1 found no proof that the new modes are flux-normalized, orthogonal, or that the resulting S-matrix is unitary; the heuristic divergence argument appears to omit the decay e^{-|kz|d} across the air gap, which is exactly the term that would make the standard series converge. If the basis is merely a change of variables, transmission observables must be basis-independent, so a benchmark against RCWA4D would settle the issue. The mismatch between S^(1) and S^(3) in the eigenmode operator is an additional internal inconsistency. For these reasons I agree with the reader's REJECT; no verdict change is needed.","tokens_in":14964,"tokens_out":10290,"duration_ms":111217,"concrete_test":"Implement the modified RCWA of Sec. 2.1 and Supplemental Secs. 2–3 for the Fig. 1(b) geometry at theta=30 deg, truncating at increasing M (3x3, 5x5, 7x7, 9x9 reciprocal vectors); compute both the transmission series Eq. (2) and the eigenvalues of the round-trip matrix S^(2)_21 S^(3)_21 over f=0.5–1.0 c/a. Compare against an independent RCWA4D or FDTD transmission calculation with the same parameters, and, if possible, against scattering-matrix poles from RCWA4D. Also verify numerically that the Table S1 modes, after flux normalization, give S^dagger S = I for each single-slab S^(i). If the spectra or the eigenmode frequencies/Q factors do not match the independent solver within numerical tolerance, or if unitarity fails, the central eigenmode claim is not established.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central novelty—eigenmode calculation via modified RCWA—rests entirely on the Table S1 evanescent basis and on the claim that this basis carries non-zero flux so that the round-trip operator S^(2)_21 S^(3)_21 has physical eigenvalues. This premise is unproven. The modes in Table S1 are not flux-normalized: for S1 TE+ the flux scales as e^{2|kz|h}, while for S2 TE+ it is independent of h, so a common normalization is missing; without it, energy conservation does not imply unitarity of S^(i), and the condition |lambda|~1 is not a well-defined resonance criterion. The Supplemental Sec. 1 divergence argument also does not establish the need for a new basis: it considers reflection at a single interface with |r|=1 but omits the finite-gap propagation factor e^{-|kz|d}, which normally makes the round-trip amplitude vanish and the multiple-scattering series converge. No truncation-convergence study or independent benchmark (RCWA4D, FDTD) is provided, so the computed eigenmodes may be numerical artifacts of the chosen normalization. In addition, Eq. (2) defines the round-trip operator as S^(2)_21 S^(3)_21, while Sec. 3.2 refers to S^(2)_21 S^(1)_21; this ambiguity affects the reproducibility of the headline result.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript proposes a modified rigorous coupled-wave analysis in which the evanescent basis functions in the air gap are replaced by hybrid modes carrying non-zero Poynting flux, claiming that this enables, for the first time, eigenmode computation for twisted bilayer photonic-crystal slabs. The authors also propose a five-layer uniform-slab approximation to predict twist-angle-dependent and twist-angle-independent guided resonances, use D4 group theory to explain resonance splitting, and develop a coupled-mode theory with expanded channels to identify f_c=0.7c/a as the boundary between angle-dependent and Fabry-Perot transmission regimes. Experimental transmission spectra from 75 GHz to 110 GHz are presented.","tokens_in":15271,"tokens_out":4088,"duration_ms":41899,"significance":"If substantiated, the eigenmode calculation would be a genuinely useful tool for incommensurate twisted bilayer photonics, and the five-layer approximation could serve as a convenient design rule. The group-theory classification of the split modes is standard but clearly presented, and the five-layer model is a plausible simplification. However, the paper's central claims are not currently supported: the modified basis is not normalized or benchmarked, the key eigenvalue figure is missing, and the CMT explanation fits its parameters to the same data it is meant to explain. These issues affect the main novelty claims, not merely the presentation. No code or data repository is provided to allow independent verification.","major_comments":[{"comment":"The modified evanescent basis in Table S1 is not flux-normalized, and no common normalization that would make the scattering matrices unitary is given. For example, the S1 TE+ mode contains a factor a=e^{|kz|h} multiplying E+ while the S2 TE+ mode contains 1/a multiplying E+, so the z-flux of these two representations scales differently with h. Without a well-defined normalization, energy conservation does not imply unitarity of S^(i), and the criterion |lambda|≈1 for identifying resonances is not a well-defined condition. The central eigenmode claim therefore rests on an unproven premise.","section":"Supplemental Sec. 2; Table S1"},{"comment":"The divergence argument used to justify the new basis is incomplete. The reflection coefficient at a single interface has |r|=1 for a purely evanescent wave, but the round-trip amplitude in the air gap includes the propagation factor e^{-|kz|d}, where d is the finite gap thickness. For d>0 this factor makes the multiple-scattering series converge in the standard evanescent basis, so the stated divergence does not establish the necessity of the modified basis. The manuscript does not provide a truncation-convergence study or an independent benchmark (e.g., RCWA4D or FDTD) for the computed eigenmodes.","section":"Supplemental Sec. 1"},{"comment":"The eigenmode analysis that is claimed as the central result is not actually shown. The text states that 'The eigenvalue in Fig.S3(a) demonstrates a similar angle dependence as the transmission spectra,' but the supplementary material contains no such figure; Fig.S3 in the supplemental is the scattering-matrix synthesis diagram. Without a plot of the eigenvalues versus angle or frequency, and without a comparison to an independent solver, the claim that eigenmodes have been calculated for the first time is unsupported.","section":"Sec. 3.2"},{"comment":"The coupled-mode theory explanation of the critical frequency f_c=0.7c/a is circular. The decay rates tau0 and tau1 are fitted to the same RCWA transmission data that are used to locate the resonance and its lower boundary; Eq. (6) and Eq. (7) are then used to identify f_c with the half-maximum of the fitted resonance. This reduces the 'transition mechanism' to the fitted lineshape rather than providing a predictive derivation. To support the claim, tau0 and tau1 should be computed from the mode profile or an independent first-principles method, not fitted to the target data.","section":"Sec. 3.3; Eqs. (6)-(7)"},{"comment":"There is an inconsistency in the definition of the round-trip operator. Equation (2) expands the transmission in powers of S^(2)_21 S^(3)_21, but the following sentence states that the eigenfunction of the matrix S^(2)_21 S^(1)_21 dictates the stationary field. Since S^(1) and S^(3) are described in the text as different physical processes (slab 1 in reflection vs. slab 1 in the multi-scattering region), the choice of operator affects the eigenvalue spectrum and the reproducibility of the headline result. This ambiguity should be resolved explicitly.","section":"Eq. (2)"},{"comment":"The claimed accuracy 'around 0.04a/c' for the five-layer uniform-slab approximation is not accompanied by any data table or plot. Supplemental Sec. 5 states that three parameter sweeps were performed, but no numerical comparison between the approximation and the RCWA results is presented. A quantitative accuracy claim with no supporting data cannot be evaluated.","section":"Sec. 3.1; Supplemental Sec. 5"}],"minor_comments":[{"comment":"The manuscript contains numerous grammatical and typographical errors, including 'To counter for the transmission property,' 'have never reveled the transition mechanism,' 'pave the wave,' and inconsistent capitalization. These should be corrected in a thorough language edit.","section":"Throughout"},{"comment":"The text refers to 'the crossing exemplified by the intersection of the blue circle and the gray parallelogram in Fig.3(c),' but the figure has no legend identifying these markers, making the statement difficult to verify.","section":"Fig. 3(c)"},{"comment":"The CMT equations use tau, tau0, and tau1 without stating the required relationship among them beyond the text near Eq. (5); the denominator in Eqs. (6)-(7) should be written with explicit parentheses, e.g., j(omega-omega0)+1/(2tau)+1/tau0+2/tau1, to avoid ambiguity.","section":"Eqs. (4)-(7)"},{"comment":"Reference [45] is cited for the energy-conservation and time-reversal constraints of CMT, but the listed source, K. J. Dean, 'Waves and Fields in Optoelectronics,' appears to be misattributed; the standard citation for this text is H. A. Haus. This should be checked.","section":"References"},{"comment":"In the sentence 'Denoting L1 ... L1 ...', the second occurrence of L1 should be L2, and the symbols for the two reciprocal lattices are otherwise easy to confuse.","section":"Supplemental Sec. 3"}],"recommendation":"reject","confidential_remarks":"The paper addresses a timely topic and contains a plausible five-layer approximation, but the central methodological claims are not supported by the evidence presented. The missing eigenvalue figure, the lack of normalization and benchmarking of the modified basis, the incomplete divergence argument, and the circularity of the CMT fitting are all load-bearing issues. In my view these cannot be resolved within a routine revision and would require a substantially reworked manuscript with independent validation."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things to know about arXiv:2505.09908. First, the five-layer uniform slab approximation is a genuinely useful design heuristic: the claim that twisted-bilayer resonances track the modes of an effective five-layer slab with averaged permittivity, within about 0.04a/c over parameter sweeps, is plausible and, if it holds up, gives optical engineers a simple route to predict angle-dependent resonances without full RCWA. Second, the paper's central methodological claim—that a modified RCWA with non-zero-flux evanescent bases computes eigenmodes that were previously inaccessible—is not backed by the evidence presented. The Table S1 basis is not flux-normalized across the two slab interfaces (the h-dependence differs between the S1 and S2 forms), so the scattering matrices are not manifestly unitary and the eigenvalue criterion |λ|≈1 is not well-defined. The Supplemental divergence argument for the new basis omits the finite-gap propagation factor e^{-|kz|d} that normally makes the multiple-scattering series converge; the single-interface |r|=1 example does not demonstrate a divergence in a real finite gap. And there is no independent benchmark against RCWA4D, FDTD, or any alternative solver. The eigenvalue plot referenced as Fig. S3(a) is absent (the supplement's Fig. S3 is the synthesis schematic), and the round-trip operator is written as S(2)_21 S(3)_21 in Eq. (2) but later referred to as S(2)_21 S(1)_21. These may be typos, but they hurt reproducibility.\n\nThe CMT section is also weaker than it looks. The decay rates τ0 and τ1 are fitted to the same RCWA transmission data that defines the resonance boundary, so the identification of f_c = 0.7c/a as the lower boundary of the 'expanded-channel' resonance is, by the paper's own equations, a statement about the fitted Lorentzian, not an independent prediction. The group-theory split analysis is fine and standard, but it is a small part of the paper.\n\nNet: the five-layer approximation and the phase-regime observation are worth a serious look, and the authors have clearly engaged with the Lou et al. and RCWA4D literature. But the eigenmode calculation—the paper's headline novelty—is under-built: unnormalized basis, no benchmark, missing figure, and an internal inconsistency. I would send this to a qualified referee, not desk-reject it, because the approximate model and the phase-transition question are timely and a referee could push for the missing comparisons. But as it stands, I would not cite it or build on the eigenmode results until the basis is fixed and benchmarked.\n\nBest,\n[Your name]","headline":"A useful five-layer design heuristic sits atop an unbenchmarked, not-flux-normalized eigenmode method that the paper presents as its headline result.","tokens_in":15842,"tokens_out":4068,"would_cite":false,"duration_ms":38541,"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":"A modified RCWA that replaces evanescent bases with flux-carrying superpositions computes the eigenmodes of twisted bilayer photonic slabs and identifies f_c = 0.7c/a as the boundary of the angle-dependent regime.","keywords":["twisted bilayer photonic crystal","modified rigorous coupled-wave analysis","guided resonance","multiple scattering","coupled-mode theory","moiré photonic structures","Fabry-Pérot resonance"],"falsifier":"Run an independent full-wave simulation of the same twisted-bilayer geometry, extract the transmission spectra from $0.5c/a$ to $1.0c/a$ for twist angles from 0 to 45 degrees, and compare: the paper's claim fails if angle-dependent resonances persist below $0.7c/a$, if the eigenmode frequencies differ from the five-layer prediction by much more than $0.04a/c$, or if no eigenvalue of $S^{(2)}_{21}S^{(1)}_{21}$ near unity can be associated with the observed resonance peaks. The experimental counterpart is to measure the fabricated PEEK slabs in the 75-110 GHz band and check the same boundary frequency.","tokens_in":14748,"feed_emoji":"💡","tokens_out":11249,"duration_ms":96534,"temperature":0.7,"pith_summary":"This paper tries to establish that the optical response of a twisted bilayer photonic crystal slab can be computed with a modified rigorous coupled-wave analysis (RCWA) in which the evanescent waves that shuttle between the two slabs are replaced by paired superpositions that carry real energy flux. With that replacement, the infinite multiple-scattering series between the slabs converges and the eigenmodes of the whole bilayer become accessible as eigenvectors of a product of single-slab scattering matrices, something the authors say standard RCWA could not provide. If the claim is right, the same modified solver explains two things at once: it predicts the angle-dependent guided resonances of moiré photonic slabs, and it locates the critical frequency, $f_c = 0.7c/a$, below which transmission is only Fabry-Pérot-like. It also supports a five-layer uniform-slab approximation, accurate to about $0.04a/c$, that gives optical engineers a quick design rule for twisted-bilayer devices such as beam-steering modules, tunable filters, and lasers.","feed_headline":"Modified RCWA computes twisted-bilayer photonic eigenmodes","feed_subtitle":"A scattering-matrix eigenmode solver explains angle-dependent resonances and puts the design boundary at 0.7c/a.","key_machinery":"The load-bearing object is the modified free-space eigenbasis used only in the air gap between the two slabs. In standard RCWA the gap modes are evanescent waves whose time-averaged Poynting vector along z is zero, which makes repeated reflections between the slabs unphysical and divergent. Each new mode is instead a normalized superposition of counter-propagating evanescent TE or TM waves, with factors $a^{\\pm 1/2}$ and $a=e^{|k_z|h}$, chosen so the composite carries non-zero energy flux along z and respects time-reversal symmetry; traveling modes are left unchanged, and the far-field regions keep the original eigenmodes. This modified basis does two jobs: it makes the scattering-matrix series converge, and it expands the channel space used by the coupled-mode theory; the expanded channels are what place the lower boundary of the angle-dependent regime at $f_c=0.7c/a$.","core_discovery":"The central discovery is that the transmission of a twisted bilayer photonic slab is governed by the multiple-scattering series $\\boldsymbol{u}^{(3)} = \\sum_{i=0}^{\\infty} S^{(1)}_{11} (S^{(2)}_{21} S^{(3)}_{21})^i S^{(2)}_{21} \\boldsymbol{u}^{(1)}$, and that the stationary fields localized between the slabs are the eigenfields of the product matrix $S^{(2)}_{21} S^{(1)}_{21}$; an eigenvalue near one marks a guided resonance. Standard RCWA cannot supply these matrices because its evanescent modes have zero Poynting flux along the decay direction, so a reflected evanescent wave does not attenuate and the scattering series diverges. The paper's modification replaces the evanescent basis in the air gap by superpositions of counter-propagating evanescent waves with amplitude factors $a^{\\pm 1/2}$ and $a=e^{|k_z|h}$, giving each basis mode a non-zero z-component of energy flux while preserving time-reversal symmetry. With this basis the authors compute bilayer eigenmodes, reproduce the transmission spectra, explain the splitting of the degenerate E resonance by the moiré potential, and identify the lowest single-slab resonance's lower boundary, $f_c=0.7c/a$, as the transition between angle-dependent resonance and Fabry-Pérot regimes.","pith_inferences":["Because the construction of the modified basis is general, the same product-of-scattering-matrices eigenmode idea should carry over to twisted multilayers and to slabs with different lattice symmetries, although the paper does not demonstrate those cases.","The critical frequency $f_c$ should depend on the hole radius, slab thickness, and dielectric contrast, since those parameters control the lowest even E-mode's channel opening; tracking that dependence would be a sharper test of the coupled-mode explanation than the single value reported.","The choice of the length scale $h$ entering $a=e^{|k_z|h}$ is not uniquely fixed by the physics; if the computed eigenmodes remain stable as $h$ is varied, the method is robust, and if not, that sensitivity would reveal where the modified-basis approximation breaks down."],"forward_implications":["Resonance lines of a twisted bilayer slab can be computed directly from the eigenvalues of $S^{(2)}_{21} S^{(1)}_{21}$, bypassing the need for large supercell simulations of the moiré pattern.","The five-layer uniform-slab dispersion curves $\\omega_i(\\mathbf{k}_{\\mathrm{inc}}+G^{(1)}_{m_1,n_1}+G^{(2)}_{m_2,n_2})$ predict the twist-angle-dependent resonances to about $0.04a/c$, giving engineers a closed-form design guide for choosing twist angle and lattice parameters.","Below $f_c=0.7c/a$ the bilayer behaves as a passive Fabry-Pérot cavity with no angle-dependent guided resonance, which sets a lower frequency bound for moiré-based beam steering and tunable filtering.","The moiré-induced splitting of the doubly degenerate E modes, around $0.005c/a$ for the parameters studied, provides a knob for engineering $\\Gamma$-point band structure in photonic-crystal surface-emitting lasers."],"supporting_citations":[{"why":"Supplies the Fourier-modal/RCWA method whose evanescent basis the paper modifies.","marker":"[37]"},{"why":"Gives the prior theory of twisted-bilayer photonic slabs that the five-layer approximation and matrix-synthesis procedure build on and compare against.","marker":"[33]"},{"why":"Provides the method for synthesizing scattering matrices that cover the combined reciprocal lattice of both slabs, used in the expanded basis.","marker":"[35]"},{"why":"Establishes the guided-resonance condition and the coupled-mode framework that the paper adapts with expanded channels to locate $f_c$.","marker":"[36]"},{"why":"Supplies the coupling-of-modes formalism used for the resonance channel analysis.","marker":"[38]"},{"why":"Supplies the temporal coupled-mode theory equations for Fano resonances that underlie equations (4)-(7).","marker":"[39]"},{"why":"Provides the scattering-matrix formulation used to extract $S^{(i)}$ from the boundary conditions.","marker":"[41]"},{"why":"Gives the master equation for photonic bands used in the five-layer perturbation argument.","marker":"[42]"},{"why":"Provides the energy-conservation and time-reversal relations that fix the coupled-mode-theory coupling constants.","marker":"[45]"}],"fun_headline_variants":["Modified RCWA unlocks twisted-bilayer eigenmode calculations","Scattering-matrix eigenmodes explain twist-slab resonances","Angle-dependent resonances in twisted photonic slabs decoded","Design boundary at 0.7c/a for twist photonic slabs"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the modified evanescent basis inside the air gap, pairs of counter-propagating evanescent waves combined to carry real energy flux with $a=e^{|k_z|h}$, completely and correctly describes the multiple-reflection physics, so that the scattering-matrix series and the eigenmodes it yields are the physical ones, a claim supported by an energy-conservation argument but not by an independent convergence proof or benchmark.","fun_headline_variants_meta":{"raw":{"variants":["Modified RCWA unlocks twisted-bilayer eigenmode calculations","Scattering-matrix eigenmodes explain twist-slab resonances","Angle-dependent resonances in twisted photonic slabs decoded","Design boundary at 0.7c/a for twist photonic slabs"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000558,"raw_usage":{"total_tokens":2696,"prompt_tokens":1028,"completion_tokens":1668,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":644,"completion_tokens_details":{"reasoning_tokens":1597}},"tokens_in":644,"tokens_out":1668,"duration_ms":13375,"temperature":1.0,"reasoning_tokens":1597,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T21:23:07.397641+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run an independent full-wave simulation of the same twisted-bilayer geometry, extract the transmission spectra from $0.5c/a$ to $1.0c/a$ for twist angles from 0 to 45 degrees, and compare: the paper's claim fails if angle-dependent resonances persist below $0.7c/a$, if the eigenmode frequencies differ from the five-layer prediction by much more than $0.04a/c$, or if no eigenvalue of $S^{(2)}_{21}S^{(1)}_{21}$ near unity can be associated with the observed resonance peaks. The experimental counterpart is to measure the fabricated PEEK slabs in the 75-110 GHz band and check the same boundary frequency.","supporting_citations":[{"cited_title":"Li, New formulation of the Fourier modal method for crossed surface-relief gratings, Journal of the Optical Society of America A 14, 2758 (1997)","cited_arxiv_id":null,"evidence_quote":"Supplies the Fourier-modal/RCWA method whose evanescent basis the paper modifies."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the prior theory of twisted-bilayer photonic slabs that the five-layer approximation and matrix-synthesis procedure build on and compare against."},{"cited_title":"Lou and S","cited_arxiv_id":null,"evidence_quote":"Provides the method for synthesizing scattering matrices that cover the combined reciprocal lattice of both slabs, used in the expanded basis."},{"cited_title":"Fan and J","cited_arxiv_id":null,"evidence_quote":"Establishes the guided-resonance condition and the coupled-mode framework that the paper adapts with expanded channels to locate $f_c$."},{"cited_title":"Manolatou, M","cited_arxiv_id":null,"evidence_quote":"Supplies the coupling-of-modes formalism used for the resonance channel analysis."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the scattering-matrix formulation used to extract $S^{(i)}$ from the boundary conditions."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the master equation for photonic bands used in the five-layer perturbation argument."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the energy-conservation and time-reversal relations that fix the coupled-mode-theory coupling constants."}],"review_version":1}