{"id":"60fa76c3-acf6-4005-b9b5-959d0210f3b5","arxiv_id":"2411.15911","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Simulations show a tilted silicon quadrumer metasurface produces a tunable collective EIT-like transparency window from coupling two quasi-bound states in the continuum, enabling slow light and refractive index sensing.","lead":"This paper uses computer simulations to show that a silicon surface made of tilted nanodisk quadrumers can produce a tunable transparency window, similar to electromagnetically induced transparency, by overlapping two bound states in the continuum. The effect slows light strongly and could improve refractive index sensing, with sensitivity rising then plateauing as the structure shrinks.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Central results lack a convergence study for the 21×21-order RCWA and 0.01 nm step; the reported q-BIC Q-factors, group index, and EIT window may shift under finer truncation.","rationale":"The reader's weakest-assumption analysis correctly identifies the home-developed RCWA truncation as the load-bearing risk. After reading the manuscript carefully, I find no internal inconsistency in the physical interpretation: the near-field maps and multipolar decompositions are consistent with ED and MQ q-BICs, the spectral-overlap mechanism is qualitatively plausible, and the group-index and sensitivity numbers follow from the stated formulas. The paper also cites prior agreement with experiment for the same code on a closely related structure, which is real supporting evidence. However, that citation does not validate convergence for the tilted geometry or for the ultra-narrow linewidths at small tilt angles, and no convergence study is provided. The central claims—especially the high Q-factors and the exact EIT-window parameters—depend on numerical resolution that has not been demonstrated. This is a concrete, testable gap rather than a speculative disagreement. The secondary 'exponential then plateau' sensitivity trend is based on few points and no error bars, but it is not the main load-bearing claim; even if that trend were overstated, the tunable EIT-like effect would still stand if convergence is confirmed. Therefore the appropriate verdict is unchanged: conditional acceptance pending a convergence study (and ideally code/data release). I agree with the reader that CONDITIONAL matches the current evidence, and the proposed test would settle whether the concern actually lands.","tokens_in":13305,"tokens_out":6947,"duration_ms":67865,"concrete_test":"Run a convergence series for the key configurations: (i) d=460 nm, α=4.2°; (ii) d=457 nm, α=6°; and (iii) the tilt-angle series in Fig. 2(e) for α=1°, 2°, 4.2°, 6°. Vary diffraction orders from 21×21 to 31×31 and 41×41, and wavelength steps from 0.01 nm to 0.002 nm and 0.0005 nm near the resonances. Extract the EIT-window transmittance, linewidth, group index, and q-BIC Q-factors, and require that all change by less than 5% between the coarsest and finest settings. Optionally cross-check one spectrum with an independent full-wave solver (FDTD or COMSOL) to rule out RCWA-specific truncation artifacts.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central quantitative claims—near-unitary transparency, group index 4.464×10^3, q-BIC Q-factors up to ~10^5, and the sensitivity plateau—are all produced by a home-developed RCWA code (Sec. 2) using 21×21 diffraction orders and a 0.01 nm wavelength step. No convergence study is reported; the only external check is 'good agreement' with experiment in the previous displaced-quadrumer geometry [41], which is not the same structure. For the smallest tilt angles in Fig. 2, fitted Q-factors exceed 10^4–10^5, corresponding to linewidths of ~0.1 nm down to ~0.01 nm. A 0.01 nm step then samples those lines with only a few points, making linewidth, Q-factor, and phase-slope (group delay) extraction numerically fragile. The EIT window itself has δλ ≈ 2.97 nm, so it is well resolved, but the individual q-BIC linewidths are used to identify the dual-band BICs and to establish the Q ∝ 1/sin²α scaling. If a finer truncation changes these linewidths or shifts the relative dispersion of the two modes, the spectral-overlap condition—and therefore the claimed EIT-like window—could move or disappear. The existence and tunability of the effect therefore rest on an unvalidated numerical truncation.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a silicon quadrumer metasurface in which C4v symmetry is broken by tilting the top and bottom nanodisk rows, and it uses a home-developed rigorous coupled-wave analysis (RCWA) code to study the resulting dual-band q-BICs. The central claims are: (i) under y-polarized illumination the two q-BICs can be tuned by nanodisk diameter or tilt angle until they spectrally overlap, producing a collective EIT-like transparency window with near-unitary transmittance; (ii) the transparency window can be tuned by nanodisk size (height and diameter), with group index up to 4.464×10^3; and (iii) refractometric bulk sensitivity first increases exponentially and then reaches a plateau around 367 nm/RIU as the nanodisk size decreases, with FOM up to 333 RIU^-1. The paper is purely numerical: all results are produced by RCWA simulations at 21×21 diffraction orders and a 0.01 nm wavelength step, with no experimental measurement and no convergence study reported.","tokens_in":13587,"tokens_out":4085,"duration_ms":40608,"significance":"If the numerical results are reliable, the paper would make a useful contribution by extending collective EIT-like effects to the coupling of electric-dipolar and magnetic-quadrupolar q-BICs in a tunable all-dielectric metasurface, and by connecting slow light to an exponential-then-plateau sensitivity trend. The geometry is clearly described, the multipolar decomposition is informative, and the reported values of group index, sensitivity, and FOM are concrete and falsifiable. On the other hand, the entire quantitative edifice rests on a home-developed RCWA code, and the acknowledged validation (agreement with a different quadrumer geometry in Ref. [41]) does not substitute for a convergence study of the present structure, where the ultra-narrow linewidths make the simulations especially delicate. The exponential-plateau claim is also presented without fit parameters, residuals, or a mechanistic derivation.","major_comments":[{"comment":"The statement that a 21×21 diffraction order and a 0.01 nm wavelength step are 'large enough' and 'small' is not supported by any convergence study for this structure. The reported Q-factors in Fig. 2(b) extend to the 10^4-10^5 range; at λ≈1.3 μm, Q=10^5 implies a linewidth of about 0.013 nm, which is only slightly larger than the 0.01 nm sampling step. This makes the extracted linewidths, the Q∝1/sin^2α scaling, the phase derivative, and hence the group index of order 4.464×10^3 all numerically fragile. Please add convergence tests (e.g., 31×31 and 41×41 orders with 0.005 and 0.002 nm steps) for the q-BIC linewidths, the EIT transparency window, the group delay, and the bulk sensitivity, and quantify how the reported values change.","section":"Sec. 2 and Figs. 2-3"},{"comment":"The claim that the bulk sensitivity 'first increases exponentially and then reaches a plateau' is empirical and is not backed by a specified functional form, fit parameters, or residuals. Moreover, the horizontal axes in Fig. 6(e)-(f) are not independent: height and diameter are simultaneously changed along the path that maintains the EIT overlap, so the trend may be a property of this particular parameter path rather than a general scaling. Equation (3) explains monotonic improvement with the fraction of energy in the sensing volume, but it does not by itself predict either an exponential regime or a plateau. Please state the fitted function with uncertainties and test whether the plateau persists when the parameters are varied independently.","section":"Sec. 3.3, Fig. 6(e)-(f)"},{"comment":"The central EIT-like effect depends on the numerical positions and widths of two ultra-narrow modes. The spectral overlap is demonstrated by a selected parameter pair (d=460 nm, α=4.2°), but no tolerance is given for how far the two q-BIC lines can be separated before the near-unitary window and the group-delay peak disappear. Because the mode dispersions are calculated numerically, a truncation-induced shift in the relative dispersion could move or destroy the overlap. I request a sensitivity analysis around the overlap condition, including how the transparency contrast and group index vary with small deviations in d and α around the claimed EIT parameters.","section":"Sec. 3.1 and Fig. 3(b)-(c)"}],"minor_comments":[{"comment":"The sentence 'the Q-factor of the q-BIC I-y remains almost constant, whereas that of the q-BIC I-y increases dramatically' should refer to q-BIC II-y in the second clause.","section":"Sec. 3.2"},{"comment":"There is a typo: 'the gaps between the left two nanodiks' should read 'nanodisks'.","section":"Sec. 3.1"},{"comment":"'Refratrometric' should be 'Refractometric'.","section":"Sec. 3.3 heading"},{"comment":"Equation (3) is cited from the plasmonic sensing literature [53]; its applicability to dielectric q-BIC metasurfaces is plausible but should be stated explicitly, and the notation for the material derivative should be defined.","section":"Eq. (3)"},{"comment":"The data availability statement says data are available on request; providing the RCWA input files and parameter sets for the EIT parameters would materially improve reproducibility.","section":"Data availability"}],"recommendation":"major_revision","confidential_remarks":"This is a numerical-only design study. I do not consider the lack of experiment itself disqualifying, but the missing convergence study is load-bearing because the main quantitative outputs (Q, group index, and sensitivity) are extracted from ultra-narrow resonances sampled at the edge of resolution. The self-citations to Refs. [28] and [41] are contextually relevant and do not appear to force the results. If the authors add a genuine convergence study, quantify the exponential-plateau fit, and include a robustness check around the EIT overlap, the paper would be acceptable for publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This is a competent, honest simulation paper, and the new bit is real: by breaking C4v symmetry with a tilt rather than a displacement, the authors get two q-BICs under y-polarization whose spectral distance shrinks as the tilt or diameter changes, so they can be pushed into overlap. That gives a near-unitary EIT-like transparency window with group index up to 4.46e3, tunable by nanodisk size. The multipolar decomposition and near-field plots back up the ED/MQ assignment, and the Q ∝ 1/sin²α scaling is a clean signature of symmetry-protected q-BICs. The sensing analysis is also fairly done, with FOMs of 111–333 RIU⁻¹ and the sensitivity plateau around 367 nm/RIU. Compared to the authors' own prior work in refs. [28] and [41], the tilt geometry is a genuine variant, and the spectral-overlap mechanism is new enough to be worth reporting.\n\nSoft spots, in order of weight. First, the entire quantitative edifice rests on one home-built RCWA code with 21×21 orders and a 0.01 nm step, and there is no convergence study. For the smallest tilt angles, fitted Q-factors are 1e4–1e5, meaning linewidths of 0.01–0.1 nm. A 0.01 nm step samples those lines with only a few points, so the extracted linewidths, Q-factors, and phase slope used for group delay could shift under finer sampling or higher truncation. The EIT transparency window itself (δλ ≈ 2.97 nm) is well resolved, but the dual-band identification and the Q-scaling claim depend on the narrow linewidths. This is the most serious issue, and it is fixable with a convergence check.\n\nSecond, the 'exponential increase then plateau' sensitivity claim is based on five simulated height/diameter pairs and an empirical fit. It may well be right, and the field arguments via Eq. (3) are plausible, but exponential is a strong word for so few points. A toned-down 'monotonic increase saturating around 367 nm/RIU' would be more defensible.\n\nThird, no data or code are released, and there is no experiment. For a simulation-only paper in a crowded field, that limits reproducibility, though not fatally.\n\nThe citation pattern is fair; the authors cite their own prior work appropriately and the broader BIC-EIT literature is represented. I would send this to peer review: the geometry is novel enough, the claims are concrete, and a good referee can push for the convergence study and language fixes. If the convergence check holds, this is a publishable incremental advance.","headline":"Solid incremental simulation study: tilted quadrumers let dual-band q-BICs spectrally overlap into a tunable EIT-like window with high group index; the physics is plausible but the paper needs a convergence study and language toning down.","tokens_in":43,"tokens_out":1484,"would_cite":false,"duration_ms":51409,"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":"Tilting silicon nanodisk quadrumers couples two dark resonances into a tunable transparency window that slows light by a factor of about 4464 and pushes refractive-index sensitivity toward a plateau of 367 nm/RIU.","keywords":["bound states in the continuum","quasi-BIC coupling","electromagnetically induced transparency analogue","all-dielectric metasurface","slow light","group index","refractive index sensing","silicon quadrumer"],"falsifier":"Fabricate a tilted quadrumer array with approximately $d=457$ nm, $h=200$ nm, $\\Lambda=1.1$ $\\mu$m, and $\\alpha=6^\\circ$ in an index-matched environment near $n=1.45$, and measure the zero-order transmittance and phase delay; if no near-unity transparency window appears near 1.385 $\\mu$m with a group delay of order 1.5 ps, or if an RCWA convergence sweep with $31\\times 31$ orders and 0.001 nm steps moves the window by more than the quoted linewidth, the central numerical claim fails.","tokens_in":62,"feed_emoji":"🔬","tokens_out":8273,"duration_ms":133007,"temperature":0.7,"pith_summary":"This paper sets out to show that two high-quality quasi-bound states in the continuum (q-BICs) in a silicon nanodisk metasurface can be tuned into spectral overlap, creating a collective electromagnetically induced transparency-like (EIT-like) window with strong slow light. The authors report that tilting the disk quadrumers provides one dark mode dominated by a collective electric dipole and another by a magnetic quadrupole, and that shrinking or enlarging the disks brings the two resonances together. At overlap, the structure transmits nearly all light in a narrow window, with a group index of $4.464\\times 10^{3}$. They further report that reducing the nanodisk size strengthens the slow light and raises the refractive-index sensitivity, which first grows exponentially and then plateaus near 367 nm/RIU. If these numerical results hold, the design gives a tunable slow-light platform whose sensing performance is limited by one of the two coupled resonances.","feed_headline":"Metasurface couples two dark resonances into a 4464x slow-light window","feed_subtitle":"Tilted silicon nanodisks bring two ultra-sharp resonances together for tunable slow light and higher sensing sensitivity.","key_machinery":"The central object is the all-dielectric metasurface unit cell: four silicon nanodisks arranged as a $C_{4v}$ quadrumer whose top and bottom dimers are tilted by angle $\\alpha$, breaking symmetry to $C_s$ so that dark modes become weakly radiating q-BICs. Bound states in the continuum are modes that remain confined even though they sit inside the radiation continuum; quasi-BICs are their slightly leaky counterparts with very high quality factors. The carrying identity is $Q \\propto 1/\\sin^2 \\alpha$ for both q-BICs, which confirms their symmetry-protected character and shows why reducing tilt narrows the resonances. The specific pair producing the effect under $y$-polarization is a collective electric dipole (q-BIC I-$y$) and a magnetic quadrupole (q-BIC II-$y$); their spectral distance falls with tilt and rises with diameter, allowing an overlap point where Fano interference creates the transparency window. Rigorous coupled-wave analysis with $21\\times 21$ diffraction orders and 0.01 nm wavelength steps supplies the simulated spectra, and multipolar decomposition identifies which moments dominate each resonance.","core_discovery":"Under $y$-polarized illumination, the tilted silicon quadrumer supports two symmetry-protected q-BICs: q-BIC I-$y$, a collective electric dipolar mode, and q-BIC II-$y$, a magnetic quadrupole with antitoroidic character. Both $Q$-factors follow $Q \\propto 1/\\sin^2 \\alpha$, identifying them as q-BICs that vanish at zero tilt. Unlike the earlier displacement-broken quadrumer, the tilt configuration lets the two resonances approach each other as $\\alpha$ grows, and increasing the nanodisk diameter pushes them into spectral overlap. At overlap (for example $d=460$ nm, $\\alpha=4.2^\\circ$, or $d=457$ nm, $\\alpha=6^\\circ$), the two modes couple into a near-unitary transparency window; multipolar decomposition shows the window is a Fano interference between the electric-dipole and magnetic-quadrupole q-BICs. The window is highly dispersive, giving a group delay of 2.972 ps and a group index of $4.464\\times 10^{3}$ at the stronger overlap. Tuning the nanodisk height from 200 nm down to 160 nm shifts the window, raises its $Q$, and increases the group index; correspondingly, the bulk refractive-index sensitivity rises from 330 nm/RIU to a plateau around 367 nm/RIU while the figure of merit climbs to 333 RIU$^{-1}$. The authors identify the narrow magnetic-quadrupole q-BIC II-$y$ as the dominant contributor to the EIT window's $Q$, sensitivity, and figure of merit.","pith_inferences":["The plateau in sensitivity at a fixed field-confinement fraction suggests a general design rule: once most electric energy sits in the sensing volume, further miniaturization buys linewidth and figure of merit but not wavelength shift; the same scaling may appear in other slow-light refractometric sensors.","The contrast with the authors' previous displacement-broken quadrumer implies that the symmetry-breaking mechanism itself, not just the mode ordering, controls whether dual q-BICs can be brought into overlap; other breaking geometries might yield even closer spectral spacing.","A direct experimental test would be to fabricate several arrays with diameters from 440 to 460 nm and measure the group delay spectroscopically; a measured transparency window with group index within a factor of two of the prediction would confirm the numerical picture."],"forward_implications":["The transparency-window wavelength, its quality factor, and the group index can all be tuned by changing the nanodisk diameter, height, or tilt angle.","Shrinking the nanodisks produces slower light and larger sensing figures of merit: group index up to about $4.464\\times 10^{3}$, bulk sensitivity up to about 367 nm/RIU, and figure of merit up to 333 RIU$^{-1}$.","Because the window's sensing performance tracks the narrower magnetic-quadrupole q-BIC II-$y$, that dark mode sets the practical limit on sensitivity and figure of merit.","The exponential-then-plateau behavior means that once the sensitivity saturates, further miniaturization improves quality factor and figure of merit but not the wavelength shift per refractive index unit.","Dynamic tuning of the collective EIT-like resonance should be possible by embedding active materials in the metasurface, as noted in the paper's concluding remarks."],"supporting_citations":[{"why":"Provides the prior all-dielectric quadrumer with dual-band q-BICs whose simulated spectra were experimentally validated, grounding the numerical setup and the comparison structure.","marker":"[41]"},{"why":"Establishes the ultrahigh-Q collective EIT-like effect in all-dielectric metasurfaces that this work extends and compares against.","marker":"[28]"},{"why":"Introduces the tilted-bar design and the inverse-quadratic $Q$-scaling used to identify the symmetry-protected q-BICs.","marker":"[42]"},{"why":"Supplies the stable rigorous coupled-wave analysis formulations on which all simulated transmittance spectra are based.","marker":"[43-45]"},{"why":"Justifies the $21\\times 21$ diffraction-order truncation used to reach the convergence region in the RCWA simulations.","marker":"[46]"},{"why":"Provides the wavelength-dependent silicon refractive index used in every simulation.","marker":"[47]"},{"why":"Gives the Cartesian multipole-moment expressions used to attribute each q-BIC to an electric dipole or a magnetic quadrupole.","marker":"[51,52]"},{"why":"States the electric-energy-fraction formula used to explain the exponential-then-plateau sensitivity behavior.","marker":"[53]"}],"fun_headline_variants":["Dual bound states couple into tunable slow-light window","Two dark modes merge into EIT window with high sensitivity","Tunable transparency from coupled BICs boosts sensing","Silicon quadrumer merges dark modes for slow-light sensing"],"cache_read_input_tokens":16256,"weakest_assumption_plain":"The load-bearing premise is that the home-built rigorous coupled-wave analysis, truncated at $21\\times 21$ diffraction orders with a 0.01 nm wavelength step, resolves the ultra-narrow q-BIC linewidths and the near-unity transparency window without appreciable error; no convergence study for this geometry appears in the paper.","fun_headline_variants_meta":{"raw":{"variants":["Dual bound states couple into tunable slow-light window","Two dark modes merge into EIT window with high sensitivity","Tunable transparency from coupled BICs boosts sensing","Silicon quadrumer merges dark modes for slow-light sensing"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000887,"raw_usage":{"total_tokens":3916,"prompt_tokens":1118,"completion_tokens":2798,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":734,"completion_tokens_details":{"reasoning_tokens":2740}},"tokens_in":734,"tokens_out":2798,"duration_ms":17587,"temperature":1.0,"reasoning_tokens":2740,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T13:44:05.896831+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Fabricate a tilted quadrumer array with approximately $d=457$ nm, $h=200$ nm, $\\Lambda=1.1$ $\\mu$m, and $\\alpha=6^\\circ$ in an index-matched environment near $n=1.45$, and measure the zero-order transmittance and phase delay; if no near-unity transparency window appears near 1.385 $\\mu$m with a group delay of order 1.5 ps, or if an RCWA convergence sweep with $31\\times 31$ orders and 0.001 nm steps moves the window by more than the quoted linewidth, the central numerical claim fails.","supporting_citations":[{"cited_title":"Silicon (Si),","cited_arxiv_id":null,"evidence_quote":"Provides the wavelength-dependent silicon refractive index used in every simulation."},{"cited_title":"Theoreticallimitoflocalizedsurfaceplasmonresonancesensitivitytolocalrefractiveindexchangeanditscomparison to conventional surface plasmon resonance sensor,","cited_arxiv_id":null,"evidence_quote":"States the electric-energy-fraction formula used to explain the exponential-then-plateau sensitivity behavior."},{"cited_title":"Homogeneous and significant near-field enhancement in all-dielectric metasurfaces for sensing applications,","cited_arxiv_id":null,"evidence_quote":"Provides the prior all-dielectric quadrumer with dual-band q-BICs whose simulated spectra were experimentally validated, grounding the numerical setup and the comparison structure."},{"cited_title":"Ultrahigh-𝑄 metasurface transparency band induced by collective–collective coupling,","cited_arxiv_id":null,"evidence_quote":"Establishes the ultrahigh-Q collective EIT-like effect in all-dielectric metasurfaces that this work extends and compares against."},{"cited_title":"Asymmetric metasurfaces and high-𝑄 resonances governed by bound states in the continuum,","cited_arxiv_id":null,"evidence_quote":"Introduces the tilted-bar design and the inverse-quadratic $Q$-scaling used to identify the symmetry-protected q-BICs."},{"cited_title":"High-q quadrupolar plasmonic lattice resonances in horizontal metal-insulator- metal gratings,","cited_arxiv_id":null,"evidence_quote":"Justifies the $21\\times 21$ diffraction-order truncation used to reach the convergence region in the RCWA simulations."}],"review_version":1}