{"id":"477f15af-0889-4d5b-a8b9-4b42807a25e2","arxiv_id":"2506.23666","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Cascaded disordered gold nanoparticle metasurfaces are fabricated by thin-film deposition and dewetting, and modeled with a layer-resolved incoherent-scattering model that ties diffuse color to the coherent driving field.","lead":"This paper shows how to build stacks of randomly arranged gold nanoparticles on a wafer and predict their reflected colors with a simple model. The stacking makes the diffuse color change under direct versus room lighting, which can be used for visual encryption.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The diffuse model is tested only at 70–90 µm−2, far above its stated 10 µm−2 validity, with the correction factor set to 1 and arbitrary rescaling; the MS-3 large-angle discrepancy is blamed on the very neglected correlations, so the central color-tuning claim is not quantitatively established.","rationale":"The paper has real independent strengths: a scalable ALD/sputtering/dewetting fabrication route, TEM/SEM characterization, angle-resolved specular modeling that matches up to 60°, and a visually striking chromo-encryption demonstration. The concern is not about fabrication or the specular model; it is about the evidential basis of the diffuse model. The incoherent-sum structure of Eq. (1) requires both independent scattering and cross-layer positional independence. The authors explicitly state that the correction factor is needed at ρ≈10 µm−2 and then set it to 1, while applying the model at 70–90 µm−2. The only angular diffuse comparison shows a qualitative failure for MS-3, and the authors attribute it to the neglected multiple scattering and inter-layer conformity. Thus the central alpha mechanism is not separately confirmed: a low-density validation, or a direct estimate of the cross-layer interference term, would settle it. This is addressable and does not invalidate the fabrication or the empirical color-tuning observation, so the reader's CONDITIONAL verdict remains appropriate. My agreement is partial because I place more weight on the validity-regime and C=1 issues than on inter-layer independence alone, but both point to the same need: the diffuse model must be tested where its assumptions hold.","tokens_in":12309,"tokens_out":6418,"duration_ms":76880,"concrete_test":"Compute the cross-layer interference term in Suppl. Eqs. S2.4–S2.5 using measured particle positions (TEM/SEM) for MS-3, and compare its magnitude with the incoherent sum in Eq. (1). If the term is comparable to the incoherent contribution near the predicted large-angle lobe, the independence assumption is the cause of the discrepancy and Eq. (1) is not valid at ρ=90 µm−2; if it is negligible, the failure lies instead in the single-layer form factor or in alpha, and the paper must revise its diffuse model before claiming predictive color control.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Eqs. (1)–(3) rest on two assumptions: within-layer independent scattering (Eq. 2 with C=1, after the authors state the correction is needed even at ρ≈10 µm−2) and cross-layer positional independence (needed for the incoherent sum and for neglecting the interference term in Suppl. Eqs. S2.4–S2.5). All three experimental metasurfaces have ρ=70, 80, 90 µm−2, an order of magnitude above the regime where the model is claimed to apply (ρ<10 µm−2, with the underlying approximation stated as valid for 1–2 µm−2). At these densities the model is only qualitatively compared at normal incidence, with the MS-1 and MS-2 curves arbitrarily multiplied by ×3 and ×1.5, and it fails the one angular test available: it predicts a large-angle lobe for MS-3 that the data do not show. The authors attribute this to multiple scattering and residual inter-layer conformity, i.e., to the two effects the model neglects. Therefore the alpha mechanism—the central physical claim that stacking modifies the coherent exciting field and thereby tunes diffuse color—has not been isolated or quantitatively confirmed in the demonstrated regime. The manuscript itself limits the model's potential to ρ<10 µm−2 and admits the interference term 'might impact directionality,' so the abstract's claim of models 'accurate ... even at large incidence angles' is unsupported for the diffuse component.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports a wafer-scale, lithography-free fabrication route to stacked disordered plasmonic metasurfaces based on ALD-deposited Al2O3 spacers, sputtered Au films, and solid-state dewetting. It presents a semi-analytical transfer-matrix model for the specular (coherent) reflectance of these stacks, validated against angle-resolved measurements up to 60 degrees, and a layer-resolved diffuse BRDF model (Eqs. 1-3) in which each layer's diffuse contribution is weighted by the ratio of the local coherent excitation intensity to the background field intensity. The paper applies this framework to a chromo-encryption demonstration in which the diffuse color changes with the number of layers and with direct versus diffuse illumination. The central claim is that stacking modifies the coherent field exciting each layer and that this mechanism quantitatively controls the diffuse color of the stack.","tokens_in":12583,"tokens_out":7620,"duration_ms":77131,"significance":"The specular model is a genuine strength: it uses only SEM-extracted geometry and tabulated optical constants, is computationally fast, and reproduces angle-resolved measurements up to 60 degrees for three stacks with different layer counts. The diffuse model is physically appealing and, if validated, would provide a design rule for multilayer disordered metasurfaces for which no comparable simple model exists. The chromo-encryption device, though not quantitatively modeled, is an attractive visual demonstration with potential applications in authentication. However, the quantitative validation of the diffuse model is currently restricted to normal incidence, high-density samples with ad hoc intensity scaling, and its only angular test (MS-3) fails. The central claim that the coherent-excitation mechanism quantitatively tunes diffuse color is therefore plausible but not established by the data presented.","major_comments":[{"comment":"The diffuse model is compared to experiment in Fig. 4 only for samples with densities rho = 70, 80, 90 um^-2 (MS-1, MS-2, MS-3), which are one to two orders of magnitude above the stated validity range of the independent scattering approximation (rho < 1-2 um^-2, and rho = 10 um^-2 even with the correction factor). The correction factor C^(m) is nevertheless set to 1. At these densities, quantitative comparison is impossible because the measured BRDFs of MS-1 and MS-2 are arbitrarily multiplied by x3 and x1.5 before plotting. The authors should either (i) validate the model on samples with densities within the claimed validity range and report absolute (unscaled) BRDF values, or (ii) implement and calibrate the correction factor C^(m) and demonstrate that it accounts for the high-density data.","section":"Model for the diffuse light component; Eq. (2)"},{"comment":"The model predicts a strong large-angle peak around 650 nm for MS-3 that is not observed in the experimental data, which instead show an intense peak at small scattering angles. The authors attribute this discrepancy to multiple scattering and residual inter-layer conformity, i.e., precisely the mechanisms neglected in Eqs. (1)-(3). This is not a minor deviation: it is a failure of the model's angular prediction for the triple-layer stack at the demonstrated density. At minimum, the abstract's statement that the models 'remain valid even at large incidence angles' must be qualified to the specular component, and the diffuse model's angular predictions need to be tested at oblique incidence and at densities within its validity range.","section":"Comparison model vs experiment; Fig. 4g-i"},{"comment":"The incoherent sum in Eq. (1) relies on the assumption that meta-atom positions on different layers are statistically independent. The authors state that 'some residual conformity likely exists due to the fabrication process and the interference term highlighted in Suppl. Note S2.1 (Eqs. S2.4-S2.5) might impact the directionality of the radiation.' Since the central physical mechanism of the paper is that the coherent excitation field (alpha) tunes the diffuse color, the model cannot isolate this mechanism if inter-layer correlations contribute to the measured directionality. The authors should either quantify the expected magnitude of the interference term for their fabricated stacks (e.g., via electron microscopy pair-correlation analysis between layers or by simulating correlated stacks) or restrict the model's claims to the uncorrelated regime.","section":"Model for the diffuse light component; Eq. (1) and Suppl. Eqs. S2.4-S2.5"},{"comment":"The attribution of the diffuse color changes to the coherent-excitation mechanism is confounded in the main BRDF comparison: MS-1, MS-2, and MS-3 differ not only in layer number and spacer thickness but also in nanodisc diameter (60, 52, 42 nm) and density (70, 80, 90 um^-2). The observed spectral shifts could therefore reflect particle size or density effects rather than the alpha mechanism. The chromo-encryption samples do control for size in Set 1, but no model comparison is shown for those samples. I request a controlled experiment in which single-layer parameters (particle size, density, morphology) are held constant while only the layer number and spacer thickness are varied, together with model predictions for those samples.","section":"Fig. 4 and Fig. 5; Application to chromo-encryption"}],"minor_comments":[{"comment":"In the text describing the chromo-encryption samples, the references to 'Fig. 4b' and 'Fig. 4d' (e.g., 'Fig. 4b, left' and 'Fig. 4b, right') should be to Fig. 5b and Fig. 5d.","section":"Application to chromo-encryption"},{"comment":"The sentence 'the ratio C^(m)/(cos theta_i cos theta_s) in Eq. (2) is set to 1' is ambiguous; setting the whole ratio to 1 would remove the angle-dependent cosines, which is unlikely intended. Please clarify that only C^(m) is taken as unity.","section":"Model for the diffuse light component"},{"comment":"The caption phrase 'a. laying of the layered substrate' is ungrammatical; it should read 'a metasurface placed on a layered substrate' or similar.","section":"Figure 3 caption"},{"comment":"The nanodisc height of 35 nm is introduced without stating its source; please state whether it is obtained from TEM/SEM cross-sections or is a model parameter.","section":"Model for the diffuse light component"}],"recommendation":"major_revision","confidential_remarks":"The supporting information was not part of the review package, so I could not verify the derivation in Suppl. Note S2 or the pair-correlation/interference term; the editorial process should ensure these are checked. The main text's claims about the diffuse model should be softened or backed by additional low-density experiments. I see no reason to question the integrity of the work; the issues are technical validation rather than misconduct."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nRead this because the fabrication route is genuinely nice: ALD-grown Al2O3 spacers plus sputtered Au and solid-state dewetting in one reactor gives centimeter-scale, lithography-free stacked disordered metasurfaces with sub-micron thickness. Up to three layers, with careful TEM/EDS characterization. That alone is worth knowing.\n\nThe specular model—single-particle scattering plus transfer matrices—is solid. It matches angle-resolved reflectance up to 60° for all three stacks, with particle sizes and densities taken from SEM, not fitted. That part is in good shape.\n\nThe soft spot is the diffuse model, the actual claim to fame. Equations (1)–(3) extend the group's earlier single-layer BRDF framework to stacks via a layer-dependent coherent excitation field alpha^(m). The idea is clean, and qualitative features like the blue peak in MS-1 and the red shift in MS-3 are captured. But validation is thin. All three samples have densities 70–90 µm^-2, an order of magnitude above the regime the paper itself says the model should work in (ρ < 10 µm^-2). The comparison is only at normal incidence, the experimental MS-1 and MS-2 curves are arbitrarily multiplied by ×3 and ×1.5, and the model predicts a large-angle lobe for MS-3 that the data do not show. The authors blame multiple scattering and inter-layer conformity—exactly the effects the model neglects. So the alpha mechanism, the claim that stacking tunes diffuse color via the coherent field, is not quantitatively isolated. The abstract's \"accurate ... even at large incidence angles\" applies to the specular model, not the diffuse one.\n\nTo their credit, the authors are upfront in the text about these limitations; the problem is the abstract doesn't match. The paper would be stronger if diffuse validation were done near the model's stated range, or if the claims were dialed back to \"qualitative agreement.\"\n\nWho should read this: people in structural color, appearance engineering, and disordered photonics. The fabrication method is immediately useful, and the diffuse model is a reasonable starting point for stacks. I'd cite it for fabrication and the specular model, but not for the diffuse predictions. It deserves peer review, though I'd ask for a revised framing of the diffuse claims before publication.\n\nRecommendation: send it to review, with the expectation of major revisions on the diffuse-model framing.","headline":"Fabrication and specular model are solid; the diffuse-model claim overreaches and the abstract overstates its accuracy.","tokens_in":13146,"tokens_out":3199,"would_cite":true,"duration_ms":35003,"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 stack of disordered metasurfaces scatters diffuse light layer by layer, with each layer driven by the coherent field the whole stack creates.","keywords":["disordered metasurfaces","metasurface stacks","diffuse reflectance","BRDF","chromo-encryption","plasmonic nanoparticles","atomic layer deposition","structural color"],"falsifier":"Measure the diffuse BRDF of two three-layer stacks made from the same nominal layers but with one middle layer deliberately shifted or rotated in plane; if the diffuse pattern changes with that shift, the layer positions are not statistically independent and Eq. (1)'s incoherent sum is incomplete. A more direct check is a high-resolution angular scan of MS-3 near 650 nm, where the model predicts a pronounced large-angle peak while the measured data show a broad small-angle peak.","tokens_in":12090,"feed_emoji":"🎨","tokens_out":6657,"duration_ms":68451,"temperature":0.7,"pith_summary":"This paper argues that the diffuse, angle-resolved light scattered by a stack of disordered metasurfaces can be predicted by treating each layer separately, as long as each layer is driven by the coherent field that the whole stack produces at that layer's depth. The authors build a lithography-free fabrication route, using atomic layer deposition, metal sputtering, and thermal dewetting, to make one-, two-, and three-layer gold nanoparticle stacks with sub-micron total thickness, and they measure both specular and diffuse reflection with a BRDF setup. Their model sums the diffuse BRDFs of the individual layers, with each contribution weighted by the ratio of the local coherent field intensity to the background field intensity. Using this picture, they show that stacking shifts the spectral weight of the diffuse scattering, and they exploit the effect in a centimeter-scale chromo-encryption surface that looks similar under room light but displays different diffuse colors under a collimated beam.","feed_headline":"Stacked metasurfaces put hidden colors in diffuse light","feed_subtitle":"The diffuse color of a stack is set by the coherent field each layer sees, enabling illumination-dependent encryption.","key_machinery":"Equation (1), $f_{\\mathrm{diff}} = \\sum_m f_{\\mathrm{diff}}^{(m)}$, is the central identity: it asserts that inter-layer diffuse interference is absent because the random particle positions in different layers are statistically independent. Each $f_{\\mathrm{diff}}^{(m)}$ follows the single-layer independent-scattering form, density times form factor times structure factor, with the form factor built from the Jones matrix of one meta-atom, $\\mathbf{J}^{(m)}$, and the excitation correction $\\alpha^{(m)}$ in front. The coherent excitation field $\\mathbf{E}_{\\mathrm{coh}}$ comes from a standard $2\\times2$ transfer-matrix computation of the stack, and $\\alpha^{(m)}$ measures how much that field is enhanced or suppressed relative to the layer's isolated background field. This construction disentangles the intrinsic scattering pattern of a nanoparticle, the form factor, from the layered environment that drives it, the coherent field.","core_discovery":"The paper's central claim is that the diffuse reflection of a stacked disordered metasurface is the incoherent sum of layer-resolved diffuse reflections, $f_{\\mathrm{diff}} = \\sum_m f_{\\mathrm{diff}}^{(m)}$, provided that for each layer $m$ the single-particle scattering form factor is multiplied by $\\alpha^{(m)}=|\\mathbf{E}_{\\mathrm{coh}}(\\mathbf{k}_i,z^{(m)})|^2/|\\mathbf{E}_b(\\mathbf{k}_i,z^{(m)})|^2$, the ratio of the coherent intensity at that layer to the background intensity in the absence of the particles. The coherent field is obtained from a transfer-matrix model of the whole stack, so it already contains the modifications caused by all other layers. The authors validate this model against measured BRDF maps for one-, two-, and three-layer samples at normal incidence and against specular reflection maps up to 60 degrees. The main consequence is that the diffuse color of a stack is set not only by the nanoparticle resonances but by where the layers sit in the stack's standing-wave coherent field, giving a design lever that single-layer metasurfaces do not have.","pith_inferences":["Beyond the paper: because $\\alpha^{(m)}$ varies with depth, a laterally patterned spacer thickness could encode a diffuse-color image readable only under direct illumination, without lithographic patterning of the nanoparticles themselves.","Beyond the paper: the large-angle discrepancy seen in the three-layer sample points to the next corrections a fuller model should include, namely inter-particle multiple scattering and any residual layer-to-layer positional correlation, with an explicit inter-layer interference term replacing the zero that Eq. (1) assumes.","Beyond the paper: the same layer-resolved excitation logic should transfer to diffuse transmittance and to random multilayer coatings for lighting and display optics, where the coherent field inside the stack is likewise depth-dependent."],"forward_implications":["The diffuse color of a stack can be tuned by changing the number of layers or the spacer thickness, because these shift the coherent standing-wave field at each layer's position.","Specular and diffuse responses can be designed together: the same transfer-matrix field that predicts the specular reflection also determines the weighting factors in the diffuse model.","Centimeter-scale chromo-encryption is realizable: stacks can be made that look nearly identical under ambient diffuse light but show distinct diffuse colors under direct collimated illumination.","Because the model is semi-analytical, design scans over layer number, spacer thickness, particle size, and density can be computed in seconds rather than by full-wave simulation.","The framework is not tied to gold or to visible wavelengths; any material and spectral range with a known single-particle Jones matrix can be inserted."],"supporting_citations":[{"why":"Supplies the single-layer diffuse BRDF model and the independent-scattering approximation that the stack model extends.","marker":"[9]"},{"why":"Provides the visual-appearance framework and BRDF measurement methodology for single disordered metasurfaces used as the experimental baseline.","marker":"[22]"},{"why":"Shows how nanoparticle form factors near a substrate produce angular radiation features, including the large-angle peak that appears in the model.","marker":"[23]"},{"why":"Establishes dewetted metal nanoparticles as structural-color elements, underpinning the fabrication route and particle-size control.","marker":"[39]"},{"why":"Supplies the standard transfer-matrix method used to compute the coherent field and therefore the alpha correction factors.","marker":"[42]"},{"why":"Provides the near-to-far-field transformation used to compute each meta-atom's Jones matrix and form factor.","marker":"[43]"},{"why":"Provides the tabulated gold refractive-index data used in all modeled spectra.","marker":"[44]"},{"why":"Documents the multiple-scattering correction at high particle densities, which the paper discusses but does not include in the simplified diffuse model.","marker":"[47]"}],"fun_headline_variants":["Stacked metasurfaces change diffuse color with lighting","Coherent field sets diffuse hue in stacked metasurfaces","Layer position in field tunes diffuse color of stacks","Illumination-dependent encryption from stacked metasurfaces","Standing wave governs diffuse color in cascaded metasurfaces"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The calculation assumes that the random positions of nanoparticles in different layers are statistically independent, so the diffuse contributions of different layers add without interference; if layer-to-layer correlation is significant, the predicted angular pattern fails.","fun_headline_variants_meta":{"raw":{"variants":["Stacked metasurfaces change diffuse color with lighting","Coherent field sets diffuse hue in stacked metasurfaces","Layer position in field tunes diffuse color of stacks","Illumination-dependent encryption from stacked metasurfaces","Standing wave governs diffuse color in cascaded metasurfaces"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000297,"raw_usage":{"total_tokens":1729,"prompt_tokens":961,"completion_tokens":768,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":577,"completion_tokens_details":{"reasoning_tokens":690}},"tokens_in":577,"tokens_out":768,"duration_ms":7999,"temperature":1.0,"reasoning_tokens":690,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T21:34:45.270101+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the diffuse BRDF of two three-layer stacks made from the same nominal layers but with one middle layer deliberately shifted or rotated in plane; if the diffuse pattern changes with that shift, the layer positions are not statistically independent and Eq. (1)'s incoherent sum is incomplete. A more direct check is a high-resolution angular scan of MS-3 near 650 nm, where the model predicts a pronounced large-angle peak while the measured data show a broad small-angle peak.","supporting_citations":[{"cited_title":"Disordered optical metasurfaces: basics, design and applications","cited_arxiv_id":null,"evidence_quote":"Supplies the single-layer diffuse BRDF model and the independent-scattering approximation that the stack model extends."},{"cited_title":"The visual appearances of disordered optical metasurfaces","cited_arxiv_id":null,"evidence_quote":"Provides the visual-appearance framework and BRDF measurement methodology for single disordered metasurfaces used as the experimental baseline."},{"cited_title":"Tailoring iridescent visual appearance with disordered resonant metasurfaces","cited_arxiv_id":null,"evidence_quote":"Shows how nanoparticle form factors near a substrate produce angular radiation features, including the large-angle peak that appears in the model."},{"cited_title":"Structural coloring of glass using dewetted nanoparticles and ultrathin films of metals","cited_arxiv_id":null,"evidence_quote":"Establishes dewetted metal nanoparticles as structural-color elements, underpinning the fabrication route and particle-size control."},{"cited_title":"Yeh, Optical waves in layered media, J","cited_arxiv_id":null,"evidence_quote":"Supplies the standard transfer-matrix method used to compute the coherent field and therefore the alpha correction factors."},{"cited_title":"Near-to-far field transformations for radiative and guided waves","cited_arxiv_id":null,"evidence_quote":"Provides the near-to-far-field transformation used to compute each meta-atom's Jones matrix and form factor."},{"cited_title":"Optical constants of the noble metals","cited_arxiv_id":null,"evidence_quote":"Provides the tabulated gold refractive-index data used in all modeled spectra."},{"cited_title":"Emergent scattering regimes in disordered metasurfaces near critical packing","cited_arxiv_id":null,"evidence_quote":"Documents the multiple-scattering correction at high particle densities, which the paper discusses but does not include in the simplified diffuse model."}],"review_version":1}