{"id":"6c9449a8-fdc8-4f3d-9726-b2352932d40b","arxiv_id":"2607.11768","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"FDTD simulations of size-tuned Au nanocube arrays on Al2O3 or h-BN spacers predict up to ~4× excitation and >80× radiative-rate gains in monolayer MoS2, implying ~350× photoluminescence enhancement without strict A/B exciton selectivity.","lead":"A numerical study shows gold nanocube arrays on thin dielectric spacers can boost monolayer MoS2 exciton generation and emission by concentrating light at cube edges. The work maps how cube size and spacer choice trade off field strength against metal losses for 2D optoelectronics.","discovery_kind":"extension","skeptic_critique":{"model":"grok-4.5","headline":"Headline 350× PL rests on PLE ≈ (G/G0)×Frad while QE maps already show metal quenching that the product formula does not fold back into the emission factor.","rationale":"The reader correctly isolates Eq. 2 and the continuum dipole model as the load-bearing soft spot for the 350-fold headline. The paper itself supplies the missing piece: Figure 5 and Eq. 3 already quantify non-radiative loss, yet those numbers are never folded back into Table 2. That internal inconsistency is more concrete than a generic “no experiment” complaint; fixing it is a pure post-processing check on existing simulation outputs. Because the rest of the design map (size/spacer trends, non-selectivity language, field maps) remains useful once the PLE numbers are corrected, the verdict stays CONDITIONAL rather than REJECT. Agreement with the reader is therefore full on the weakest assumption; the stress-test only sharpens the same point into a falsifiable recalculation.","tokens_in":16046,"tokens_out":673,"duration_ms":6155,"concrete_test":"Recompute every Table 2 PLE entry as (G/G0)×[Frad/(Frad+Fnonrad)] (or the exact low-QY form (G/G0)×Frad/(1+η0(Ftot-1)) with η0=10^-3 and 10^-4) using the same dipole powers already extracted for Figure 5. If the 2 nm Al2O3 / 130 nm AuNC B-exciton entry falls below ~100 (or changes by >30 %), the Abstract’s 350-fold claim must be revised.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central numerical claim (Abstract; Table 2) is ~350-fold photoluminescence enhancement from excitation factors ≤4.35 times Frad >80. That product is justified only by Eq. 2 under the low-QY approximation η0~10^-3–10^-4 (citing [30,38]). Methodology separately defines QE = Frad/(Frad+Fnonrad) (Eq. 3) and Figure 5 maps η/η0 that already encode distance-dependent quenching. Yet Table 2 PLE entries are still pure products of the two columns; they never re-weight Frad by the computed QE (or by 1/(1+η0(Ftot-1))). At ts=2 nm the same geometry that maximizes Frad also maximizes non-radiative transfer to Au, so the product systematically overstates net PL. The continuum FDTD sheet-conductivity + single in-plane dipole model (Eqs. 7–13) further omits exciton diffusion, saturation, and array-period effects that would reduce the effective enhancement. Without that correction the headline 350-fold number is not a self-consistent prediction of the paper’s own decay-rate framework.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.5","summary":"The manuscript presents a numerical FDTD study of monolayer MoS2 coupled to periodic Au nanocube arrays (side lengths 105 and 130 nm, period 250 nm) separated by Al2O3 or h-BN spacers of thickness 2–10 nm. Size- and spacer-tuned LSPR is used to modulate the A- and B-excitonic windows. From extinction, near-field maps, absorption, carrier-generation rate, and dipole-based radiative/non-radiative rates the authors report excitation enhancements up to 4.35 (B, 605 nm) and 3.66 (A, 650 nm), radiative-rate factors exceeding 80, and photoluminescence enhancements up to ~350-fold via the low-QY product PLE ≈ (G/G0) × Frad (Eq. 2; Table 2). They conclude that the platform yields wavelength-dependent rather than strictly exciton-selective enhancement and is a simple, scalable design for excitonic light–matter engineering in ML MoS2.","tokens_in":16370,"tokens_out":1086,"duration_ms":8335,"significance":"If the quantitative claims hold under a self-consistent treatment of quantum yield, the work supplies a compact, fabrication-compatible design map (cube size + spacer material/thickness) for plasmon-enhanced excitonic absorption and emission in monolayer TMDs. The systematic comparison of Al2O3 versus h-BN, the sensing-volume interpretation of spacer-induced LSPR shifts (Eq. 1), and the multi-metric reporting (CGR, Frad, QE maps) go beyond single-geometry field-enhancement papers and would be useful for photodetectors, emitters, and sensors. Strengths include a transparent, literature-standard FDTD workflow (Johnson–Christy Au, Palik dielectrics, experimental MoS2 sheet conductivity, TFSF extinction, power-dissipation CGR, in-plane dipole decay rates) and clear tables that make the parameter trends reproducible. The absence of experiment is acceptable for a pure design study provided the headline numbers are internally consistent with the paper’s own decay-rate framework.","major_comments":[{"comment":"Abstract and Table 2 claim ~350-fold PL enhancement from the product PLE ≈ (G/G0) × Frad (Eq. 2) under the low-QY approximation η0 ~ 10^{-3}–10^{-4}. Methodology simultaneously defines QE = Frad/(Frad + Fnonrad) (Eq. 3) and Figure 5 maps η/η0 that already encode strong metal quenching at ts = 2 nm—the same geometry that maximizes Frad. Table 2 PLE entries never re-weight Frad by the computed QE (or by the standard factor 1/(1 + η0(Ftot – 1))). Consequently the headline 350× figure systematically overstates net emission relative to the paper’s own decay-rate framework and must be recomputed or clearly caveated before the central claim can stand.","section":null},{"comment":"Methodology (Eqs. 7–13) models MoS2 as a continuum sheet conductivity and the exciton as a single in-plane dipole. At the reported Frad > 80 and sub-10 nm metal–semiconductor separations this continuum treatment omits exciton diffusion, saturation, and possible array-period collective effects that would reduce the effective enhancement. Without a short discussion or estimate of these corrections, the absolute 350-fold number remains an upper-bound continuum prediction rather than a robust device-level forecast.","section":null}],"minor_comments":[{"comment":"Figure 1e,f and Table 1: the LSPR peak shifts for Al2O3 and h-BN are reported to 0.1 nm precision; a brief statement of mesh convergence or wavelength sampling would strengthen confidence in the sub-nanometer differences.","section":null},{"comment":"Figure 3 absorption spectra are normalized; absolute absorption or a side-by-side un-normalized panel would better support the claim that the AuNC only moderately reshapes the excitonic profile.","section":null},{"comment":"Equation 1 is introduced as a sensing-volume model; the fitted or estimated values of m and ld used for the present AuNC geometry are not stated, making the quantitative link to Table 1 incomplete.","section":null},{"comment":"Scattered typographical issues (e.g., “10 5” for 105 nm, missing spaces around units, “t s” vs ts) should be cleaned for readability.","section":null},{"comment":"The self-citation to arXiv:2603.07732 is appropriate for related geometry work but should be clearly distinguished as prior numerical work by the same group rather than independent experimental validation.","section":null}],"recommendation":"major_revision","confidential_remarks":"The core FDTD workflow is sound and the design-parameter map is useful; the only load-bearing defect is the inconsistent use of the low-QY product versus the paper’s own QE maps. Once Table 2 is recomputed (or the 350× claim is explicitly labeled an upper-bound continuum estimate) the manuscript should be publishable after minor polishing. Scope fits a solid optics/nanophotonics journal; novelty is incremental but well-executed."},"author_rebuttal":null,"desk_editor":{"model":"grok-4.5","letter":"This is a competent classical FDTD parameter study, not new physics. What is actually new is the quantitative side-by-side map for 105/130 nm Au nanocubes on Al2O3 vs h-BN spacers over ML MoS2: extinction shifts (Table 1), in-plane |E|^{2}, carrier-generation spectra, Frad, and tabulated excitation/Frad/PLE factors (Table 2). The workflow is standard and defensible—Johnson–Christy Au, Palik dielectrics, experimental MoS2 sheet conductivity, TFSF extinction, power-dissipation CGR, in-plane dipole decay rates. Trends are physically right: larger cubes redshift LSPR, thinner spacers give stronger near-field and generation, both A and B channels rise with wavelength-dependent weights rather than strict selectivity. The authors are careful on that last point, which is good.\n\nThe soft spot is real and load-bearing for the abstract claim. They define QE = Frad/(Frad+Fnonrad) and plot η/η0 maps that already show metal quenching at small ts, yet Table 2 PLE is still pure (G/G0)×Frad under the low-η0 product approximation (Eq. 2). At 2 nm the same geometry that maximizes Frad also maximizes non-radiative transfer, so the ~350× number systematically overstates net PL relative to their own decay framework. Continuum sheet + single dipole also omits diffusion, saturation, and period effects. That is a framing and consistency problem, not a fabrication problem; the underlying field and rate trends remain usable if claims stay labeled as simulation predictions.\n\nWho it is for: people designing plasmon-enhanced 2D emitters/detectors who want a practical cube-size and spacer map. Citation pattern is appropriate; circularity is low. I would send it to referees with the instruction to force the PLE formula into consistency with the QE maps and to tone the headline. Worth engaging as a design reference once that is fixed; not a device result.","headline":"Solid FDTD design map for Au nanocubes on spacer/MoS2; the ~350\times PL number is an overstated product that ignores the paper’s own quenching maps.","tokens_in":17025,"tokens_out":513,"would_cite":false,"duration_ms":4845,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.5","headline":"Size-tuned gold nanocube arrays on thin dielectric spacers give monolayer MoS2 up to ~350-fold photoluminescence enhancement via edge-localized plasmons.","keywords":["Monolayer MoS2","Exciton-plasmon coupling","Purcell enhancement","Charge generation rate","Gold nanocubes","Localized surface plasmon resonance","Dielectric spacer"],"falsifier":"Fabricate 130 nm Au nanocube arrays on 2 nm Al2O3 over monolayer MoS2, measure absolute photoluminescence enhancement at 605 nm and 650 nm under the same excitation conditions used in the model, and check whether the observed factors approach the predicted ~350 and ~180; a large shortfall would falsify the product approximation or the continuum dipole treatment.","tokens_in":16920,"feed_emoji":"🔆","tokens_out":973,"duration_ms":14849,"temperature":0.7,"pith_summary":"Monolayer MoS2 is only about a nanometer thick, so it absorbs and emits light poorly even though its A and B excitons are strong. This numerical study shows that periodic gold nanocubes, separated from the sheet by a few nanometers of Al2O3 or h-BN, concentrate light at the cube edges and can be size-tuned so their plasmon resonance overlaps those excitons. Excitation-rate gains of a few times multiply with radiative-decay-rate gains above 80 to produce estimated photoluminescence enhancements of order 350, largest for the thinnest spacers. Both excitonic channels brighten at once; relative strength shifts with cube size, spacer material, and thickness rather than switching one exciton fully off. The design is presented as a simple, scalable route to stronger light-matter interaction in 2D semiconductors without external fields or doping.","feed_headline":"Gold nanocubes give MoS2 a 350-fold light boost","feed_subtitle":"Edge plasmons tuned by cube size and spacer thickness brighten both A and B excitons in a one-atom sheet.","key_machinery":"Spacer-mediated gold nanocube arrays: flat-faceted Au cubes whose corner hot spots and size-dependent LSPR are further redshifted and attenuated by the dielectric spacer thickness and refractive index, so near-field intensity, carrier generation, and Purcell-type radiative enhancement at the MoS2 plane can be tuned together.","core_discovery":"Localized surface plasmons of size-tuned gold nanocube arrays, mediated by Al2O3 or h-BN spacers of controlled thickness, produce simultaneous excitation-rate enhancements up to 4.35 (B exciton, 605 nm) and 3.66 (A exciton, 650 nm) and radiative decay-rate enhancements exceeding 80 in monolayer MoS2, yielding estimated photoluminescence enhancements of roughly 350-fold while preserving the intrinsic excitonic spectrum and enabling wavelength-dependent rather than strictly selective excitonic modulation.","pith_inferences":["Because the geometry is compatible with colloidal or lithographic nanocube placement, the predicted gains are in principle transferable to large-area devices once experimental quantum-yield maps are measured.","The same spacer-and-size tuning should apply, with only modest re-optimization, to other group-VI TMD monolayers whose A/B excitons lie nearby in the visible.","Edge localization implies that sparse or non-periodic cube placements could still deliver local enhancement, potentially relaxing array-period constraints for integration."],"forward_implications":["Cube side length plus spacer thickness and material become practical knobs for placing the plasmon resonance across the A/B excitonic window of MoS2.","Thinnest spacers maximize both carrier generation and radiative decay enhancement, giving a clear design rule for maximum PL gain.","The same platform can serve enhanced 2D photodetectors, nanoscale emitters, and excitonic sensors without external fields or chemical doping.","Both A and B channels remain active, so the structure supplies broadband wavelength-dependent brightening rather than single-exciton switching."],"fun_headline_variants":["Spacer-tuned Au nanocubes lift MoS2 excitons up to 350-fold","Size-tuned gold nanocube arrays boost MoS2 PL by ~350x","Edge plasmons in Au nanocubes enhance A/B excitons of ML MoS2","Al2O3/hBN spacers mediate 80x radiative gains in MoS2 via cubes","Nanocube LSPR shifts yield 4x excitation boost for MoS2 excitons"],"cache_read_input_tokens":128,"weakest_assumption_plain":"The headline 350-fold photoluminescence boost is obtained by simply multiplying excitation and radiative-rate factors because the monolayer’s intrinsic quantum yield is taken to be extremely small, so non-radiative metal losses and any saturation under strong near-field coupling are assumed not to cancel the product.","fun_headline_variants_meta":{"raw":{"variants":["Spacer-tuned Au nanocubes lift MoS2 excitons up to 350-fold","Size-tuned gold nanocube arrays boost MoS2 PL by ~350x","Edge plasmons in Au nanocubes enhance A/B excitons of ML MoS2","Al2O3/hBN spacers mediate 80x radiative gains in MoS2 via cubes","Nanocube LSPR shifts yield 4x excitation boost for MoS2 excitons"]},"model":"grok-4.5","effort":"low","cost_usd":0.003854,"raw_usage":{"total_tokens":1303,"prompt_tokens":895,"num_sources_used":0,"completion_tokens":122,"cost_in_usd_ticks":38540000,"prompt_tokens_details":{"text_tokens":895,"audio_tokens":0,"image_tokens":0,"cached_tokens":256},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":286,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":895,"tokens_out":122,"duration_ms":4728,"temperature":1.0,"reasoning_tokens":286,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-14T03:20:05.891341+00:00","model_set":{"reader":"grok-4.5"},"falsifier":"Fabricate 130 nm Au nanocube arrays on 2 nm Al2O3 over monolayer MoS2, measure absolute photoluminescence enhancement at 605 nm and 650 nm under the same excitation conditions used in the model, and check whether the observed factors approach the predicted ~350 and ~180; a large shortfall would falsify the product approximation or the continuum dipole treatment.","supporting_citations":[],"review_version":1}