{"id":"babd1f73-2b56-4cda-92fa-44c33eb72a4e","arxiv_id":"2505.12438","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A MgO-doped thin-film lithium niobate waveguide pumped at 1550 nm produces a gap-free supercontinuum from 330 nm to 2250 nm, spanning 2.7 octaves.","lead":"Researchers generated a gap-free 2.7-octave supercontinuum spanning 330 nm to 2250 nm in a thin-film lithium niobate waveguide pumped by a femtosecond laser, without complex periodic poling. The result could make chip-scale optical frequency combs that reach into the ultraviolet cheaper and easier to build.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 2.7-octave claim rests on the 330–500 nm trace being genuine guided-wave signal above the noise floor, but no raw traces, error bars, or stitching/calibration details are provided, and this segment defines the bandwidth.","rationale":"The reader's conditional verdict and weakest assumption correctly identify the UV edge near the detection floor and the concatenation of spectra as the key risk. My read agrees: the 2.7-octave number is numerically dominated by the 330 nm edge, and the evidence for that edge is thinner than for the telecom/IR part of the spectrum. The simulation mismatch (0.81 nJ used, 500 nm discrete feature absent) is a secondary concern: it weakens the mechanistic interpretation but does not by itself falsify the measured spectrum. The missing raw data, noise floors, and stitching details are addressable, so the appropriate verdict remains CONDITIONAL rather than a rejection. I therefore recommend no change to the reader's verdict.","tokens_in":7220,"tokens_out":3552,"duration_ms":41104,"concrete_test":"Request that the authors release the raw spectral traces from each instrument (grating spectrometer, AQ6315A, AQ6375) together with separately recorded noise floors with the pump blocked, and describe the calibration and stitching procedure. Then perform a control measurement of the 330–500 nm band with a calibrated UV-enhanced power meter behind a 330–500 nm bandpass filter, with the waveguide pumped and with the pump blocked or waveguide removed. If the integrated UV power is below the noise-equivalent power or does not vanish in the blocked case, the 330 nm edge and hence the 2.7-octave claim are unsupported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The 2.7-octave bandwidth (330–2250 nm) is set by the short-wavelength edge, so the 330–500 nm portion is the load-bearing part of the central claim. In Section 2.3.1 and Fig. 2(b), this portion is measured only with a grating spectrometer whose stated lower detection limit is -90 dBm, and the red gap-free curve is described as taken 'without considering the coupled loss (~-10 dB)'. The black comparison curve at 0.625 nJ is discontinuous, while the red curve at 0.687 nJ is gap-free; a 10% increase in pump energy converting a discontinuous spectrum into the claimed gap-free UV edge is exactly the kind of transition that could be affected by the noise floor. The manuscript does not provide raw spectral traces, error bars, the measured noise spectrum with the pump blocked, the calibration factor relating the grating-spectrometer trace to the OSA traces, or the stitching procedure used to assemble Fig. 2(a). The discrete feature near 500 nm is acknowledged in Section 2.3.3 not to be reproduced by the GNLSE simulation, and the simulation used 0.81 nJ rather than the experimental 0.687 nJ, weakening the 'agrees well with numerical simulation' support. Thus the central claim is only as strong as the evidence that the UV trace is genuine guided supercontinuum signal above the detector noise, not an artifact of detector sensitivity, stray light, or spectral concatenation.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports a 2.7-octave supercontinuum spanning 330 nm to 2250 nm generated in a 6.5-mm-long MgO-doped thin-film lithium niobate (TFLN) ridge waveguide pumped by 81-fs, 1550-nm pulses with 0.687 nJ pulse energy. The claimed spectral coverage is achieved without periodic poling, using dispersion-engineered anomalous GVD, modal-matched SHG, THG, dispersive-wave emission, and MgO doping to suppress the photorefractive effect. The authors support the claim with power-dependent spectra (Fig. 3), a numerical GNLSE simulation (Fig. 4c), a comparison to undoped TFLN waveguides (Fig. 5), and a photograph of visible scattering along the waveguide. The central assertion is that the measured spectrum is gap-free across the full 330--2250 nm range, with the short-wavelength edge measured by a grating spectrometer with a stated lower detection limit of -90 dBm.","tokens_in":7530,"tokens_out":2987,"duration_ms":31428,"significance":"If the measurement is fully substantiated, this would be a record spectral coverage for unpoled TFLN waveguides and would represent a meaningful step toward chip-scale frequency combs spanning the UV to near-infrared. The paper has several strengths: the PLACE fabrication route is relevant to low-loss TFLN devices, the dispersion-engineering rationale is clearly presented, the pump-power evolution showing SHG, THG, and SPM is valuable, and the comparison with undoped TFLN helps isolate the role of MgO doping. However, the central 'gap-free 2.7-octave' claim rests on the 330--500 nm portion of the spectrum, and the evidence provided for that portion is not yet sufficient to rule out detector-noise, stray-light, or spectral-stitching artifacts. The supporting simulation also uses a different pulse energy from the experiment and does not reproduce a prominent discrete feature near 500 nm.","major_comments":[{"comment":"The 2.7-octave bandwidth is set by the short-wavelength edge at 330 nm, so the 330--500 nm trace is load-bearing. The grating spectrometer's lower detection limit is stated as -90 dBm, and the red curve is described as recorded 'without considering the coupled loss (~-10 dB)'. However, the manuscript does not provide a raw spectral trace with the pump blocked, a calibration factor relating the grating-spectrometer trace to the OSA traces, or a description of how the spectra were stitched together in Fig. 2(a). The comparison between the discontinuous 0.625-nJ black curve and the supposedly gap-free 0.687-nJ red curve is a ~10% pump-energy change, and such a threshold-sensitive transition is exactly where the detection floor could affect the apparent envelope. Please provide the raw traces, a noise-floor measurement, detector-calibration details, and the stitching/concatenation procedure.","section":"Sec. 2.3.1, Fig. 2(b)"},{"comment":"The GNLSE simulation is run with a pump energy of 0.81 nJ, while the experiment is claimed at 0.687 nJ; no explanation for this mismatch is given. Additionally, the text acknowledges that the strong discrete spectral component near 500 nm is not reproduced by the simulation, yet the paper concludes that the simulation 'agrees well' with the experiment. This discrepancy weakens the numerical support for the UV extension. Please simulate at the experimental pulse energy, or explicitly justify the 0.81-nJ value, and quantify the agreement (e.g., spectral overlap or residual) in the region where the discrete feature appears.","section":"Sec. 2.3.3, Fig. 4(c)"},{"comment":"No error bars, repeated-measurement statistics, or spectrometer resolution/dynamic-range specifications are reported for the spectra in Fig. 2. The claim of a 'gap-free' envelope requires knowing that the absence of signal between the measured points is not due to the instrument floor or to stitching discontinuities. Please state the resolution and dynamic range of each spectrometer used for the concatenated spectrum, and report at least one repeated acquisition or an estimate of measurement uncertainty.","section":"Sec. 2.3.1 and Fig. 2"},{"comment":"The paper attributes the improved SCG in MgO-doped TFLN partly to suppression of the photorefractive effect and states that the doped waveguide maintains SCG 'for a long time of period', but no time-resolved spectral measurement is shown for either the doped or the undoped waveguide. The inset photograph and the qualitative statement are not sufficient to support this claim. Please provide a quantitative comparison of spectral stability over time under nominally identical pumping conditions.","section":"Sec. 3.1 and Fig. 5"}],"minor_comments":[{"comment":"The phrase 'spit-step Fourier method' should read 'split-step Fourier method'.","section":"Sec. 2.3.3"},{"comment":"The sentence beginning 'Here, Such ultrabroad-bandwidth SCG' contains an inappropriate capital 'S' after the comma; please revise for grammar.","section":"Abstract and Sec. 2.3.1"},{"comment":"The highest pump energy in Fig. 3 is listed as 0.675 nJ, whereas the main text and Fig. 2(a) quote 0.687 nJ; please reconcile these values.","section":"Fig. 3"},{"comment":"The phrase 'the lower detected limit of the spectrometer' should be 'the lower detection limit of the spectrometer'.","section":"Fig. 2(b) caption"}],"recommendation":"major_revision","confidential_remarks":"The central issue is verifiability of the UV-edge measurement. If the authors can provide raw spectra, a noise-floor trace, calibration/stitching details, and a simulation at the stated experimental energy, the claim could become publishable. The current manuscript is not suitable for acceptance without those additions."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe paper reports a real experimental advance: 2.7-octave SCG from 330 nm to 2250 nm in an unpoled MgO-doped TFLN ridge waveguide, using PLACE fabrication and dispersion engineering. That extends the unpoled record and gets close to the poled record (Wu 2024), with the added benefit of MgO doping against photorefractive damage. The power-dependent spectra, soliton-fission interpretation, and comparison with undoped TFLN all hang together; the physics is presented honestly.\n\nThe soft spots are concentrated on the UV edge, which is exactly what sets the bandwidth. The 330–500 nm trace is measured with a grating spectrometer whose -90 dBm floor is stated, but there are no raw traces, error bars, noise background, calibration against the OSA curves, or stitching procedure. The red gap-free curve is compared to a black discontinuous curve at only 0.625 nJ; a 10% pump change flipping a discontinuous spectrum into a gap-free UV edge is exactly the sort of thing that can be an artifact of the noise floor. The GNLSE simulation also uses 0.81 nJ rather than the experimental 0.687 nJ with no explanation, and the paper itself concedes the simulation misses the discrete ~500 nm feature that helps extend the spectrum into the UV. So the 'agrees well' claim is generous.\n\nNone of this overturns the core observation, and the missing details are standard referee requests: raw data, calibration, and a simulation at the true pump energy. The citation pattern is fine; self-citations are to the group's own fabrication methods, which is appropriate.\n\nThis paper is for people working on integrated frequency combs and TFLN nonlinear photonics. It deserves a serious referee: the fabrication and measurement are likely sound, but the record claim needs the UV evidence strengthened before it becomes a citable number. Send it to review; require raw data and a simulation at the correct pump energy.","headline":"A credible record-claim for unpoled TFLN supercontinuum, but the UV-edge evidence needs to be shown above the noise floor before it fully lands.","tokens_in":8108,"tokens_out":1737,"would_cite":false,"duration_ms":17423,"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":"An unpoled MgO-doped lithium niobate waveguide converts 0.687-nJ, 81-fs pulses at 1550 nm into one gap-free supercontinuum spanning 330–2250 nm.","keywords":["supercontinuum generation","thin-film lithium niobate","dispersion engineering","MgO doping","photorefractive effect","second-harmonic generation","third-harmonic generation","integrated photonics"],"falsifier":"Measure the waveguide output at 330–500 nm with a calibrated photon-counting detector through narrow bandpass filters while blocking the pump; if the apparent UV signal is absent, matches stray-light background, or does not rise steeply with pump energy through the same threshold as the harmonics, then the 2.7-octave envelope is not genuine.","tokens_in":7029,"feed_emoji":"🌈","tokens_out":7867,"duration_ms":72531,"temperature":0.7,"pith_summary":"This paper reports that a 6.5-mm-long ridge waveguide in magnesium-oxide-doped thin-film lithium niobate, pumped by 0.687-nJ, 81-fs pulses at 1550 nm, produces a gap-free supercontinuum from 330 nm to 2250 nm, covering 2.7 octaves without any periodic poling. The authors attribute the breadth to dispersion engineering that makes the pump band anomalously dispersive, so soliton fission, dispersive-wave emission, and modal-matched second- and third-harmonic generation all contribute to the spectrum. They also argue that 5% MgO doping suppresses lithium niobate's photorefractive drift, so the broad spectrum persists rather than fading, and that their chemo-mechanical fabrication keeps waveguide losses low enough for the effect to develop. If correct, this would mean chip-scale supercontinuum sources covering the visible and ultraviolet can be made in simple, unpoled waveguides with modest pulse energies, a step toward integrated optical clocks and self-referenced frequency combs.","feed_headline":"Unpoled lithium niobate waveguide spans 2.7 octaves, 330–2250 nm","feed_subtitle":"A 6.5-mm MgO-doped waveguide turns 0.687 nJ laser pulses into light spanning 330 to 2250 nm.","key_machinery":"The load-bearing object is a ridge waveguide etched from a 900-nm Z-cut MgO:LiNbO3 film into a 2-µm-wide mesa with a wedge angle of about 8.36°, fabricated by photolithography-assisted chemo-mechanical etching (PLACE) and designed so that the second-order dispersion β2 at 1550 nm is anomalous. Three mechanisms in that geometry do the work: the anomalous dispersion lets soliton fission and dispersive-wave emission widen the spectrum toward both shorter and longer wavelengths; careful mode matching makes the second harmonic at 775 nm and third harmonic at 517 nm strong enough to seed new spectral regions; and the 5% magnesium-oxide doping suppresses photorefractive index drift so the waveguide output stays stable under sustained pumping. The paper uses a generalized nonlinear Schrödinger equation solved by the split-step Fourier method to connect the waveguide cross-section and pump conditions to the observed spectrum.","core_discovery":"The central discovery is that a single unpoled MgO-doped thin-film lithium niobate ridge waveguide, with a 2-µm top width and an etched depth chosen so that the telecom band is anomalously dispersive, converts 1550-nm femtosecond pulses at 0.687 nJ into one continuous spectrum spanning 330–2250 nm. The envelope is built by several processes acting at once: self-phase modulation and soliton fission in the anomalous-dispersion region, a short-wavelength dispersive wave near 1100 nm, modal-matched second-harmonic generation at 775 nm, third-harmonic generation at 517 nm, and sum-frequency mixing that fills the gaps between harmonics. The authors show that at lower pump energies the harmonic and supercontinuum components broaden and overlap, reaching 2.5 octaves by 0.54 nJ, and that the full 2.7-octave envelope at 0.687 nJ matches generalized nonlinear-Schrödinger-equation simulations. They further report that an undoped control waveguide gives only 1.636 octaves and unstable output, which they attribute to higher scattering loss and the photorefractive effect.","pith_inferences":["If the UV edge is confirmed by a calibrated photon-counting measurement, the practical span may extend beyond 2250 nm too, since the stated limit is the optical spectrum analyzer's cutoff rather than the waveguide.","The same dispersion-engineering recipe without poling should transfer to other pump wavelengths, such as 1030 nm or 2 µm, by rescaling the cross-section, opening UV-visible combs for telecom- and mid-IR-pumped chips.","MgO doping could be applied to other lithium niobate nonlinear devices where photorefractive drift limits long-term operation, such as electro-optic modulators, microring combs, and quantum sources.","A direct test of the mechanism would be to measure the 330–500 nm output with a silicon photon counter and bandpass filters; if the counts track the pump pulse energy through the same nonlinear threshold as the visible harmonics, the guided-wave origin is settled."],"forward_implications":["A 2.7-octave, gap-free supercontinuum from an unpoled waveguide means chirped periodic poling is not needed for full-visible-to-UV coverage on thin-film lithium niobate, relaxing fabrication demands and cost.","At pump energies around 0.405 nJ the second-harmonic light and the supercontinuum already overlap spectrally, which is the condition required for f-2f self-referencing of the comb on a chip.","MgO doping makes the broad spectrum persist over time, directly addressing the photorefractive instability that limited earlier TFLN supercontinuum demonstrations.","Numerical simulations reproduce the measured profile and soliton-fission position, giving a design loop for moving the spectrum edge by changing the waveguide cross-section."],"supporting_citations":[{"why":"Supplies the prior TFLN supercontinuum record spanning 330–2400 nm achieved with chirped periodic poling, the baseline this work extends by removing the poling.","marker":"[23]"},{"why":"Provides the photolithography-assisted chemo-mechanical etching method used to fabricate the low-loss MgO-doped waveguides.","marker":"[31]"},{"why":"Cited with [33] as the numerical-simulation and waveguide-model basis for predicting the spectrum and soliton-fission position.","marker":"[32]"},{"why":"Provides the generalized nonlinear Schrödinger equation and split-step Fourier method used to simulate the measured supercontinuum.","marker":"[33]"},{"why":"Documents earlier TFLN supercontinuum work with photorefractive limitations and f-2f referencing context that motivated the MgO doping.","marker":"[20]"}],"fun_headline_variants":["Unpoled lithium niobate waveguide spans 330–2250 nm in one shot","2.7-octave supercontinuum from a single unpoled TFLN waveguide","Dispersion-managed waveguide delivers 2.7-octave UV-to-IR spectrum","MgO-doped TFLN waveguide: 2.7 octaves from 330 nm to 2250 nm","Unpoled TFLN waveguide rivals poled record for supercontinuum"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The result collapses if the 330–500 nm light measured by the grating spectrometer is not genuine guided supercontinuum above the noise floor, or if stitching three detectors' traces creates a continuous envelope that no single detector would actually see.","fun_headline_variants_meta":{"raw":{"variants":["Unpoled lithium niobate waveguide spans 330–2250 nm in one shot","2.7-octave supercontinuum from a single unpoled TFLN waveguide","Dispersion-managed waveguide delivers 2.7-octave UV-to-IR spectrum","MgO-doped TFLN waveguide: 2.7 octaves from 330 nm to 2250 nm","Unpoled TFLN waveguide rivals poled record for supercontinuum"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001038,"raw_usage":{"total_tokens":4443,"prompt_tokens":1094,"completion_tokens":3349,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":710,"completion_tokens_details":{"reasoning_tokens":3230}},"tokens_in":710,"tokens_out":3349,"duration_ms":23627,"temperature":1.0,"reasoning_tokens":3230,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T20:33:28.491998+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the waveguide output at 330–500 nm with a calibrated photon-counting detector through narrow bandpass filters while blocking the pump; if the apparent UV signal is absent, matches stray-light background, or does not rise steeply with pump energy through the same threshold as the harmonics, then the 2.7-octave envelope is not genuine.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the prior TFLN supercontinuum record spanning 330–2400 nm achieved with chirped periodic poling, the baseline this work extends by removing the poling."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the photolithography-assisted chemo-mechanical etching method used to fabricate the low-loss MgO-doped waveguides."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Cited with [33] as the numerical-simulation and waveguide-model basis for predicting the spectrum and soliton-fission position."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the generalized nonlinear Schrödinger equation and split-step Fourier method used to simulate the measured supercontinuum."},{"cited_title":"Okawachi, M","cited_arxiv_id":null,"evidence_quote":"Documents earlier TFLN supercontinuum work with photorefractive limitations and f-2f referencing context that motivated the MgO doping."}],"review_version":1}