{"id":"abf98d4c-a0a3-458a-9377-79d76d7e2770","arxiv_id":"2507.22667","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"First demonstration of second harmonic generation in polycrystalline ZnS nanowaveguides, converting 1474 nm pump light to 737 nm with an estimated efficiency of 0.4 %/W/cm².","lead":"Researchers made zinc sulfide (ZnS) nanowaveguides on a silicon chip and detected red light at 737 nm when pumping with 1474 nm infrared light, the first reported second harmonic generation in this platform. If confirmed, ZnS offers a cheap, non-critical-material route to compact photonic frequency converters.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No power-scaling or background check for the 737 nm signal leaves the central SHG claim relying on a single spectrum and build-up image.","rationale":"The reader's verdict is CONDITIONAL, and the weakest assumption identified is exactly the one I would press hardest: the 737 nm signal has not been shown to be quadratic in pump power, nor is there an on/off or background measurement. All other concerns—the missing SI for the efficiency calculation, the 50x gap explained by an unmeasured 60% orientation fraction, and the citation mismatch—are secondary to this. If the 737 nm light is not genuine SHG from the waveguide, the paper's headline claim fails regardless of how well the tensor theory is constructed. The paper does provide real supportive evidence: the spectrum at 737 nm with the pump at 1474 nm and the scattering build-up image are consistent with guided SHG, and the fabricated nanowaveguides are characterized in detail. However, consistency is not enough. A single spectral peak could be produced by stray light or by nonlinearities in the 20 m fiber link, especially because the pump is pulsed and the paper does not state that background subtraction was performed. The efficiency estimate being placed in unavailable supporting information means the most quantitative claim cannot be audited from the manuscript alone. I therefore agree with the reader's conditional posture and would not change the verdict on the basis of the present analysis. The proposed power-dependence and background check is the minimal experiment that would elevate or falsify the central claim.","tokens_in":10681,"tokens_out":3108,"duration_ms":38251,"concrete_test":"Measure the 737 nm signal power while attenuating the 1474 nm pump in calibrated steps (e.g., 3, 6, 9, 12 dB) with the OPO locked at 1474 nm and the waveguide aligned. Plot log10(P_SH) vs log10(P_pump); genuine SHG must give slope 2.0 ± 0.1. Repeat with the waveguide removed (or with pump blocked) to establish the background, and with the pump detuned 5 nm away from the phase-match peak to show the signal disappears. If the slope is not quadratic or the background is comparable, the central SHG claim fails.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that phase-matched SHG occurs in the fabricated ZnS nanowaveguides, evidenced by a sharp peak at 1474 nm producing 737 nm light and by a scattering build-up image. For that claim to hold, the 737 nm light must be second harmonic generated inside the ZnS guide, not stray pump leakage, fluorescence, or nonlinearity in the 20 m fiber link. Section 5 reports the spectrum and top-view image but provides no power-dependence curve, no on/off control (e.g., waveguide removed or pump blocked), and no background measurement. Without a quadratic power dependence, a single spectral peak at exactly half the pump wavelength is suggestive but not conclusive: the same signature could arise from pump harmonics generated in the fiber or from detector/OSA artifacts, especially since the pump is delivered through a 20 m fiber at 8.2 mW peak power. The paper explicitly states the conversion-efficiency estimate is in supporting information, which is not included here, so the 0.4 %/W/cm² number and the 50x discrepancy attributed to a 60% orientation fraction cannot be independently checked. The internally consistent tensor calculation in Section 3 is not at issue; rather, the experimental link between the observed 737 nm light and the ZnS waveguide is undersupported.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports the fabrication of ZnS nanowaveguides by sputtering and electron-beam lithography, their linear characterization, and the observation of second harmonic generation (SHG) in the TM→TM configuration. A tensor rotation calculation for (111)-oriented zincblende ZnS predicts a d33 coefficient of 1.15 d14, and COMSOL simulations predict modal phase matching for TM00→TM02 at 1480 nm. Experiments using an OPO pump at 1474 nm show a spectral peak at 737 nm and a top-view scattering image consistent with signal build-up; the instantaneous conversion efficiency is estimated at 0.4 %/W/cm². The authors claim this is the first demonstration of SHG in ZnS nanowaveguides.","tokens_in":10919,"tokens_out":2895,"duration_ms":37255,"significance":"If confirmed, this would establish ZnS as a new integrated nonlinear photonic platform, with the advantages of a wide transparency range, high nonlinear coefficients, and non-critical raw materials. The paper includes several strengths: the fabrication process is described in detail, the tensor rotation is benchmarked to the known 43m zincblende tensor and gives the correct effective coefficient for [111] propagation, the modal phase-matching prediction is concrete, and the observation of a spectral peak at exactly half the pump wavelength is a necessary signature. However, the central experimental claim currently lacks a standard control (quadratic power dependence) and the efficiency estimate is not independently verifiable, so the significance is contingent on additional evidence.","major_comments":[{"comment":"The central claim that the 737 nm signal is second harmonic generated inside the ZnS waveguide rests on a single spectrum and a scattering image. The paper reports no power-dependence measurement, no on/off control (e.g., pump blocked or light injected into a region without the waveguide), and no background measurement. Without a quadratic power-dependence curve, the spectral peak at exactly half the pump wavelength does not exclude stray light, detector artifacts, or nonlinearity in the 20 m fiber link. This is a load-bearing gap for the paper's central claim.","section":"Section 5"},{"comment":"The reported 0.4 %/W/cm² instantaneous conversion efficiency is stated to be derived in the supporting information, which is not included in the manuscript, and no error bars or systematic uncertainties are given for pump power, coupling efficiency, collection efficiency, or propagation losses. Furthermore, the statement that the discrepancy between theory and experiment is due to '60% of the crystallites' sharing a similar [111] orientation is inferred from the very same efficiency discrepancy; this is circular unless an independent structural measurement (e.g., EBSD or pole figures) supports the 60% fraction.","section":"Section 5, conversion efficiency"},{"comment":"The rotation description is internally inconsistent: the text first gives θ0 = 45° around [001] and φ0 = 54.73° around [1̄10], then later gives φ0 = −45° and θ0 = −54.74°, and Eqs. (2) and (3) differ in form (one uses inverse rotation with Kronecker products, the other does not). While the final tensor in Eq. (4) appears consistent with the known [111] effective coefficient d33 = 1.15 d14, the presentation must be corrected so the derivation is traceable.","section":"Section 3, Eqs. (2)–(4)"}],"minor_comments":[{"comment":"The abstract says 'scanning electron microscopy (SEM)' while Section 2 says 'scanning electron beam microscopy'; please unify the terminology.","section":"Abstract and Section 2"},{"comment":"The sentence 'The OPO's wavelength was scanned from ݉ to ݉' contains missing numerical values; the scanned wavelength range should be stated explicitly.","section":"Section 5, paragraph 2"},{"comment":"The axes of the dispersion-curve plot are not described in the caption or text; please specify the plotted quantities and units.","section":"Section 5, Figure 5a"},{"comment":"The claim that the observation of a phase-matching peak at the expected spectral position indicates 'the absence of random quasi-phase matching' needs quantitative support, since a narrow spectral peak alone does not rule out RQPM; the spectral width and the expected RQPM bandwidth should be compared.","section":"Section 5, discussion of RQPM"},{"comment":"Propagation losses are reported as averages without uncertainties; at least the spread or standard deviation should be given, and the statement 'competitive with the ones of nanowaveguides from other more mature platforms' should be supported by a direct comparison with specific references.","section":"Section 4, loss measurements"},{"comment":"Reference [35] concerns SHG in gallium phosphide microdisks; its direct relevance to the ZnS crystallite orientation fraction should be clarified or replaced with a source specific to ZnS.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The manuscript cannot be fully evaluated without the supporting information, which contains the conversion-efficiency measurement details and the loss-origin discussion. Please ensure the SI is available to reviewers. The novelty of 'first ZnS nanowaveguides' appears defensible against refs. [14] and [15], which concern non-nanoscale or different geometries, but check that the claim is consistent with the literature. The main risk to the central claim is the lack of a power-dependence control; this should be added or the claim softened."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: first ZnS nanowaveguides patterned by e-beam and dry etch, with a phase-matched SHG peak at 1474 nm producing 737 nm light and a clear build-up image. The materials characterization and the tensor-rotation analysis are careful, and the comparison with prior planar ZnS waveguides and GaP is honest. I believe the core observation is very likely real, but the paper as written does not fully prove it.\n\nWhat is new: no prior ZnS integrated waveguide via nanolithography; refs 14 and 15 are planar films. Fabrication details are solid: sputtering, XRD, ellipsometry, Fabry-Perot loss measurements. The theoretical section is standard 43m tensor rotation, but it is done cleanly and gives a useful prediction of the phase-matching wavelength and the relevant d' coefficient.\n\nSoft spots, in order. (1) The 737 nm signal lacks a power-dependence curve, an on/off control, or a background estimate. A sharp peak at exactly half the pump wavelength with a 20 m fiber link and 8.2 mW pulses is suggestive but not conclusive; pump harmonics in fiber or detector artifacts can mimic that. This is the load-bearing gap. (2) The conversion efficiency of 0.4 %/W/cm² is stated to be in the supporting information, which is not included, so the number and the 50x discrepancy cannot be checked. (3) The 60% crystallite orientation fraction is inferred from that same discrepancy, making it effectively a fitting parameter rather than a measured quantity. (4) No error bars on losses or efficiency, which is a minor issue for a first demonstration but matters for the comparison with GaP.\n\nThe stress-test note is right about the missing quadratic power check; that is the main fix. I do not think the tensor calculation is circular, and the paper's own statements honestly flag the twist/tilt complexity. The central claim—phase-matched SHG in ZnS nanowaveguides—is probably correct but under-supported as submitted.\n\nWho this is for: the integrated nonlinear photonics community, especially people looking for non-critical, low-cost platforms. It deserves a serious referee; the fix is a simple power-scaling measurement plus moving the efficiency derivation into the main text or a complete SI.\n\nRecommendation: accept for peer review, with the clear expectation of a power-dependence measurement before publication.","headline":"Credible first SHG in ZnS nanowaveguides, but missing a power-scaling check leaves the central claim one control short.","tokens_in":11505,"tokens_out":2341,"would_cite":true,"duration_ms":27324,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["42.65.Ky","42.82.-m","78.20.-e"],"model":"deepseek-v4-flash","headline":"Zinc sulfide nanowaveguides, fabricated for the first time, produce phase-matched second harmonic generation at 737 nm when pumped at 1474 nm, supporting ZnS as an integrated nonlinear photonics platform.","keywords":["zinc sulfide","second harmonic generation","nanowaveguides","integrated nonlinear photonics","phase matching","polycrystalline thin films","RF magnetron sputtering","wide bandgap semiconductors"],"falsifier":"Measure the 737 nm signal power as the 1474 nm pump power is varied over at least a decade on a log-log plot: genuine SHG must fall on a straight line of slope 2, and the signal must disappear when the pump is blocked or detuned far from 1474 nm. A slope near 1, or a signal that survives with the waveguide removed while pumping the bare fiber, would refute the central claim.","tokens_in":10503,"feed_emoji":"✨","tokens_out":8086,"duration_ms":83237,"temperature":0.7,"pith_summary":"The paper reports the first zinc sulfide (ZnS) nanowaveguides and the first observation of phase-matched second harmonic generation (SHG) in them: light at 1474 nm injected into a sputtered polycrystalline waveguide emerges at 737 nm, with a sharp resonance at the wavelength predicted by modal phase matching. The authors argue this establishes ZnS as a practical integrated nonlinear photonics platform, using a high-index wide-bandgap material with transparency from 400 nm to 10 µm and only one independent second-order nonlinear coefficient, rather than the costly single-crystal or poled materials used today. They support the claim with a tensor analysis showing that a specifically chosen TM→TM polarization configuration keeps its nonlinear coefficient constant under random crystallite twist, and with imaging that shows the second-harmonic signal building up along the guide. The measured instantaneous conversion efficiency is estimated at 0.4 %/W/cm², to be compared with a theoretical upper value of 21 %/W/cm² for a perfect monocrystalline device, and the gap is interpreted as partial crystallite misorientation.","feed_headline":"First ZnS nanowaveguides double a 1474 nm laser to 737 nm","feed_subtitle":"Sputtered polycrystalline zinc sulfide delivers phase-matched second harmonic generation, opening a new integrated nonlinear photonics…","key_machinery":"The argument is carried by the second-order susceptibility tensor $\\chi^{(2)}$ of zinc-blende ZnS expressed in the crystallographic basis aligned with the [111] growth direction. Rotating the standard tensor, whose only nonzero components in the cubic frame are $d_{14}=d_{25}=d_{36}$, into the $[1\\bar{1}0]$, $[11\\bar{2}]$, $[111]$ basis yields effective nonlinear coefficients, and one specific process, $E_{y''} E_{y''} \\to P_{z''}$ (involving $d_{33}''$), has a coefficient independent of rotation around [111], so uncontrolled crystallite twist does not destroy the nonlinear response. The second essential piece is modal phase matching: finite-element dispersion calculations for an 810 nm wide, 500 nm high guide show that only the TM00 pump can phase match around 1480 nm, to the TM02 mode, locating the observable SHG peak and tying the spectrum to the crystal orientation.","core_discovery":"On the paper's own terms, the central claim is that 737 nm light with a sharp spectral peak at a 1474 nm pump wavelength, together with a scattered-light image showing signal growth along the propagation direction, demonstrates phase-matched TM00→TM02 second harmonic generation inside a polycrystalline ZnS nanowaveguide. Because the phase-matching wavelength coincides with the calculated position for a zinc-blende crystal oriented along [111], the authors conclude that random quasi-phase-matching is absent and that most crystallites share a preferential [111] orientation; the previously noted crystallite twist around that axis should not suppress the chosen TM→TM process, since its effective nonlinear coefficient is twist-invariant. This makes ZnS, previously only a bulk or thin-film nonlinear material, a viable integrated waveguide platform whose SHG efficiency is claimed to be close to that of GaP nanowaveguides under the same type of modal phase matching.","pith_inferences":["A power-dependence measurement, which the paper does not report, would directly test the SHG attribution: genuine second-harmonic power should grow as the square of the pump power and vanish when the pump is blocked or the guide is removed.","The twist-invariance argument likely generalizes to other polycrystalline zinc-blende materials such as ZnSe and ZnTe, suggesting that sputtered wide-bandgap films could form a family of integrated nonlinear platforms without epitaxial growth.","If crystallite size can be engineered independently of orientation, the same platform could be steered between the two regimes discussed in the paper: narrowband oriented-crystal phase matching or broadband random quasi-phase-matching conversion.","The inferred 60% crystallite alignment could be checked directly by electron-backscatter-diffraction or pole-figure measurements on the same films, offering an independent validation of the efficiency interpretation."],"forward_implications":["Phase-matched SHG in polycrystalline ZnS waveguides implies that integrated frequency conversion does not require single-crystal epitaxy or periodic poling for twist-insensitive polarization configurations.","Reducing propagation losses from the measured 40–55 dB/cm toward 5 dB/cm and increasing guide length toward 1.2 cm is forecast by the authors to raise conversion efficiency by two orders of magnitude.","The sharp phase-matching peak at the monocrystal position shows that the polycrystalline film retains a dominant [111] crystallite orientation, and the efficiency shortfall allows an estimate that about 60% of crystallites are aligned.","The platform's transparency from 400 nm to 10 µm opens a route to on-chip frequency conversion into the visible and mid-infrared using the same fabrication flow.","Because ZnS uses elements not on the European critical raw materials list, the platform could provide a resource-supply advantage over lithium niobate and III-V waveguides."],"supporting_citations":[{"why":"Ito et al. 1974, phase-matched guided SHG in oriented ZnS polycrystalline thin-film waveguides; establishes the prior art the present work extends to nanowaveguides.","marker":"[15]"},{"why":"Saitoh et al. 1991, single-crystalline epitaxial ZnS waveguides for phase-matched SHG; provides the earlier epitaxial approach this work replaces with sputtered polycrystalline devices.","marker":"[14]"},{"why":"Pantzas et al. 2022, continuous-wave SHG in orientation-patterned GaP waveguides; supplies the comparison platform for conversion efficiency and phase-matching approach.","marker":"[5]"},{"why":"Boyd, Nonlinear Optics; source of the $\\chi^{(2)}$ tensor form with $d_{14}$ as the only independent component of zinc-blende ZnS.","marker":"[28]"},{"why":"Wagner et al. 1998, dispersion of the second-order nonlinear susceptibility in ZnS, ZnSe and ZnTe; supplies the material's nonlinear coefficient values used in the predictions.","marker":"[19]"},{"why":"Combrié et al. 2006, Fabry-Perot fringe method for propagation loss extraction; the measurement technique used to characterize the ZnS waveguides.","marker":"[34]"},{"why":"Anthur et al. 2021, second harmonic generation in gallium phosphide nanowaveguides; the GaP nano-waveguide result the authors compare their efficiency against.","marker":"[36]"},{"why":"Guillemé et al. 2017, SHG in GaP microdisks from strict orientation to random quasi-phase matching; gives the framework for interpreting the crystallite orientation distribution from efficiency.","marker":"[35]"},{"why":"Baudrier-Raybaut et al. 2004, random quasi-phase-matching in polycrystalline materials; the concept the authors rule out based on the sharp phase-matching peak.","marker":"[32]"}],"fun_headline_variants":["ZnS nanowaveguides double 1474 nm light to 737 nm","First phase-matched SHG in polycrystalline ZnS waveguides","ZnS waveguide doubles 1474 nm laser to 737 nm","Phase-matched doubling in ZnS nanowaveguides at 1474 nm"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The detected 737 nm light is genuinely second harmonic light generated inside the ZnS waveguide, because the paper reports no power-dependence curve, no on/off control, and no background estimate to rule out stray light, fluorescence, or fiber-link nonlinearity.","fun_headline_variants_meta":{"raw":{"variants":["ZnS nanowaveguides double 1474 nm light to 737 nm","First phase-matched SHG in polycrystalline ZnS waveguides","ZnS waveguide doubles 1474 nm laser to 737 nm","Phase-matched doubling in ZnS nanowaveguides at 1474 nm"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00077,"raw_usage":{"total_tokens":3373,"prompt_tokens":871,"completion_tokens":2502,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":487,"completion_tokens_details":{"reasoning_tokens":2437}},"tokens_in":487,"tokens_out":2502,"duration_ms":19354,"temperature":1.0,"reasoning_tokens":2437,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T11:25:30.284206+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the 737 nm signal power as the 1474 nm pump power is varied over at least a decade on a log-log plot: genuine SHG must fall on a straight line of slope 2, and the signal must disappear when the pump is blocked or detuned far from 1474 nm. A slope near 1, or a signal that survives with the waveguide removed while pumping the bare fiber, would refute the central claim.","supporting_citations":[{"cited_title":"Ru et al","cited_arxiv_id":null,"evidence_quote":"Ito et al. 1974, phase-matched guided SHG in oriented ZnS polycrystalline thin-film waveguides; establishes the prior art the present work extends to nanowaveguides."},{"cited_title":"Saitoh, T","cited_arxiv_id":null,"evidence_quote":"Saitoh et al. 1991, single-crystalline epitaxial ZnS waveguides for phase-matched SHG; provides the earlier epitaxial approach this work replaces with sputtered polycrystalline devices."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Pantzas et al. 2022, continuous-wave SHG in orientation-patterned GaP waveguides; supplies the comparison platform for conversion efficiency and phase-matching approach."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Boyd, Nonlinear Optics; source of the $\\chi^{(2)}$ tensor form with $d_{14}$ as the only independent component of zinc-blende ZnS."},{"cited_title":"Fang et al","cited_arxiv_id":null,"evidence_quote":"Wagner et al. 1998, dispersion of the second-order nonlinear susceptibility in ZnS, ZnSe and ZnTe; supplies the material's nonlinear coefficient values used in the predictions."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Combrié et al. 2006, Fabry-Perot fringe method for propagation loss extraction; the measurement technique used to characterize the ZnS waveguides."},{"cited_title":"Singh, R","cited_arxiv_id":null,"evidence_quote":"Anthur et al. 2021, second harmonic generation in gallium phosphide nanowaveguides; the GaP nano-waveguide result the authors compare their efficiency against."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Guillemé et al. 2017, SHG in GaP microdisks from strict orientation to random quasi-phase matching; gives the framework for interpreting the crystallite orientation distribution from efficiency."},{"cited_title":"ܧ ௬ \" ܲ ௬","cited_arxiv_id":null,"evidence_quote":"Baudrier-Raybaut et al. 2004, random quasi-phase-matching in polycrystalline materials; the concept the authors rule out based on the sharp phase-matching peak."}],"review_version":1}