{"id":"6547feb0-1c1c-4f4e-847b-b9a58bb8f371","arxiv_id":"2502.07180","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"Soliton microcombs at 25 GHz are demonstrated in X-cut thin-film lithium niobate by orienting racetrack waveguides to suppress Raman scattering.","lead":"Researchers generated soliton microcombs, stable trains of evenly spaced light pulses, in X-cut lithium niobate microchips, a platform previously blocked by unwanted Raman scattering. The key was to orient the chip's racetrack waveguides so the light polarization runs perpendicular to the crystal axis, and the combs were then produced with both continuous-wave and pulsed lasers.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Raman-suppression mechanism is not isolated from Q differences or validated in waveguide geometry; matched-Q angular control is needed.","rationale":"The reader identified exactly the load-bearing concern: the two-device comparison conflates orientation with Q, and the bulk Raman angular data are not shown to transfer to the waveguide mode. I agree that this is the weakest point in the argument. The direct experimental demonstration of soliton microcombs in X-cut TFLN—with sech2 fits, soliton steps, and a narrow 25.037 GHz beatnote—is credible and stands independently. However, the paper's central design rule for suppressing Raman is not yet isolated, so a conditional verdict remains appropriate. My concrete test would settle the causal claim by controlling Q across orientation angles and by computing the waveguide-mode Raman gain from the bulk tensor; until such a test is performed, the mechanism should be treated as a plausible hypothesis rather than an established result. The paper's strengths include clean device characterization and convincing soliton evidence, but the mechanism claim requires stronger experimental isolation.","tokens_in":10383,"tokens_out":3920,"duration_ms":40271,"concrete_test":"Fabricate a series of racetrack resonators with identical cross-section, bending radius, coupling gap, and target Q, but with straight-section angles theta = 0, 30, 60, and 90 degrees relative to the optical axis. For each device, measure the on-chip Raman lasing threshold and the soliton existence range under identical pump and auxiliary-laser conditions. Also simulate the effective stimulated Raman gain for the TE mode using the measured bulk Raman tensor and the actual mode profile (including the 62-degree sidewall angle), and compare the predicted Raman threshold to the measured value. If the Raman threshold scales with orientation as predicted and solitons appear only when the Raman threshold exceeds the available pump power, the mechanism is validated; if Q or fabrication variations dominate, the causal claim must be weakened.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central causal claim—that orienting the racetrack straight sections perpendicular to the optical axis suppresses Raman and thereby enables soliton formation—rests on a two-device comparison that does not control for resonator Q. Device (i) (strong-Raman orientation) is reported to have a loaded Q of ~1e6, while device (ii) (weak-Raman orientation) has an intrinsic Q up to 3e6 (Fig. 3b). Since the soliton threshold and thermal stability depend strongly on Q, the observed failure in device (i) could be due to its lower Q rather than to the orientation-dependent Raman response. Additionally, the polarization-dependent Raman spectra in Fig. 2a were measured on a planar X-cut film with free-space polarization rotation, and there is no demonstration that the Raman gain coefficient for the confined TE waveguide mode in the etched racetrack follows the same angular scaling. The Raman tensor components and the mode's electric-field distribution (including sidewall-angle-induced components) can alter the effective Raman gain; a 40% residual spontaneous A(TO)1 peak at 90 degrees (Fig. 2b) does not by itself establish that the stimulated Raman threshold exceeds the soliton threshold. Thus the design rule 'straight section parallel to the optical axis suppresses Raman' is plausible but not yet established.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports the generation of TE-polarized soliton microcombs in X-cut thin-film lithium niobate (TFLN) racetrack microresonators, with a repetition rate of about 25 GHz and spectral spans beyond 200 nm under continuous-wave pumping and about 350 nm under pulsed pumping. The proposed enabling mechanism is the orientation of the racetrack straight sections relative to the optical axis so that the TE mode is polarized perpendicular to the optical axis, thereby reducing the Raman response that has previously hindered soliton formation on this platform. The manuscript presents Raman spectroscopy of an X-cut film, device characterization (Q factor, coupling, dispersion), and demonstrations of soliton steps, sech2-shaped spectra, and narrow beatnotes for both CW and pulse-pumped operation.","tokens_in":10575,"tokens_out":3514,"duration_ms":33539,"significance":"If the orientation-based Raman suppression is validated, this result is significant because it opens a route toward monolithically integrated X-cut TFLN photonics combining soliton microcombs with high-speed electro-optic modulators and efficient frequency doublers. The direct evidence for soliton microcombs is solid: discrete soliton steps in the transmitted power, sech2-shaped spectral envelopes, and a narrow 25.037 GHz beatnote. The reproducibility of the core observation across CW and pulse pumping strengthens the demonstration. However, the causal claim that the improved soliton generation is due specifically to the polarization-orientation-dependent Raman response is not yet established with the two-device comparison presented.","major_comments":[{"comment":"The central claim that device (ii) supports soliton microcombs because of reduced Raman nonlinearity is not isolated from a difference in resonator Q. Device (i) is reported to have a loaded Q of approximately 1e6, while device (ii) is reported to have an intrinsic Q up to 3e6. Soliton threshold and thermal stability depend strongly on Q, so the failure of device (i) to reach the soliton state could be due to its lower Q rather than to the orientation-dependent Raman gain. The authors should present a matched-Q comparison, or at minimum a quantitative estimate showing that the orientation-induced Raman gain reduction exceeds the Q-related change in soliton threshold, before claiming the design rule is validated.","section":"Results, Fig. 2c-f; Fig. 3b"},{"comment":"The polarization-dependent Raman spectra are measured on the planar X-cut TFLN film in a free-space backscattering geometry, not on the confined TE waveguide mode inside the etched racetrack. The effective Raman gain for a waveguide mode depends on the mode's electric-field polarization components, including any sidewall-angle-induced components, so the free-space angular scaling may not transfer directly to the waveguide geometry. Furthermore, the A(TO)1 mode retains about 40% of its peak intensity at 90 degrees (Fig. 2b), and no measurement or estimate is given for the stimulated Raman threshold relative to the soliton formation threshold in the actual device. The authors should provide waveguide-level Raman gain calculations or a direct measurement of Raman suppression in the etched waveguide to support the causal mechanism.","section":"Raman response, Fig. 2a,b; Methods"},{"comment":"The spectral-span claims of 'more than 200 nm' (CW) and 'about 350 nm' (pulsed) are based on measured spectra that show significant deviations from the sech2 envelope due to mode crossings and birefringence-induced mode mixing, particularly below 1470 nm. This does not undermine the existence of solitons, but it makes the practical usable bandwidth unclear. The authors should specify the wavelength range over which the comb lines follow the soliton envelope and are coherent, rather than quoting the raw spectral extent.","section":"Fig. 4c; Fig. 5c; Extended Data Fig. 1"}],"minor_comments":[{"comment":"In the introduction, 'remarkedly enhanced' should be 'remarkably enhanced'.","section":"Introduction"},{"comment":"The caption contains a typo: 'T echnologies' should be 'Technologies'.","section":"Fig. 1 caption"},{"comment":"The right inset caption reads 'electrical beanote' and should read 'electrical beatnote'.","section":"Fig. 5 caption"},{"comment":"The y-axis label 'Raman scattering spectrum (a.u.)' appears to be shared by several panels; please clarify the scaling between the angle-dependent spectra.","section":"Fig. 2a"},{"comment":"The fit range for the parabolic dispersion (D2/2pi = 24.4 kHz) is not specified; the mode-crossing deviations near 1560 nm and 1610 nm are visible, and stating the exact fitting window would make the D2 value reproducible.","section":"Fig. 3d"},{"comment":"The data and code availability statements say that materials are available 'upon reasonable request'; depositing the soliton spectra, beatnote recordings, and dispersion data in a public repository would improve reproducibility.","section":"Data availability"}],"recommendation":"major_revision","confidential_remarks":"The soliton microcomb demonstration itself is convincing and likely of high interest to the journal's readership. The main risk is the uncontrolled two-device comparison underpinning the Raman-suppression mechanism; a matched-Q control or a quantitative waveguide Raman-gain analysis would considerably strengthen the paper. The authors should also be asked to provide the fitting parameters for the sech2 envelopes and a clearer definition of the usable comb bandwidth."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The headline is simple: this is the first TE-polarized soliton microcomb on X-cut TFLN, and the direct evidence for solitons is clean. Sech2-shaped spectra, discrete soliton steps, and a narrow 25.037 GHz beatnote are all there. That is a real advance, and it matters because X-cut TFLN is the platform that also hosts high-speed modulators and efficient doublers.\n\nWhat is genuinely new is the design rule: orient the racetrack straight sections so the TE mode is polarized perpendicular to the optical axis, reducing Raman gain. The polarization-dependent Raman spectra in Fig. 2 are measured and fitted, and the contrast between the strong-Raman and weak-Raman orientations is visible in the raw data. The paper also does something unusual that deserves credit: it reports the failure case (device (i), strong Raman orientation) rather than only the success. That makes the comparison more informative.\n\nThe soft spot is the causal claim. Device (i) has a loaded Q around 1e6, while device (ii) has intrinsic Q up to 3e6. Soliton threshold and thermal stability depend strongly on Q, so the failure in device (i) could be due to lower Q rather than to the orientation-dependent Raman response. The Raman angular data come from a planar film with free-space polarization rotation; the effective Raman gain in the confined waveguide mode, with sidewall angles and field components, may not follow the same scaling. A 40% residual A(TO)1 peak at 90 degrees does not by itself prove that the stimulated Raman threshold exceeds the soliton threshold. These are real limitations, but they are not fatal. The soliton demonstration does not depend on the mechanism claim, and the mechanism is plausible even if not fully established.\n\nA few smaller issues: no error bars on the Q, coupling, or dispersion values, and data/code are only available on request. That is mostly a completeness concern, not a correctness one.\n\nThis paper deserves a serious referee. The result is important, the evidence for solitons is solid, and the mechanism question can be addressed in revision (matched-Q controls, or direct in-resonator Raman measurement). I would cite this for the platform advance even while treating the Raman-suppression rule as promising rather than proven.","headline":"First TE soliton microcombs on X-cut TFLN, with clean soliton evidence; the Raman-suppression mechanism is plausible but not fully isolated from device-Q differences.","tokens_in":11234,"tokens_out":1194,"would_cite":true,"duration_ms":12375,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"TE-polarized soliton microcombs at 25 GHz are generated on X-cut thin-film lithium niobate by rotating the racetrack's straight sections so the mode's polarization lies perpendicular to the optical axis, suppressing the Raman gain that…","keywords":["soliton microcomb","thin-film lithium niobate","X-cut TFLN","Raman nonlinearity","optical frequency comb","microresonator","racetrack resonator","electro-optic integration"],"falsifier":"Fabricate two racetrack resonators with identical geometry, etch depth, and loaded Q near 3 million, differing only in straight-section orientation, and compare the threshold for Raman lasing and the existence of soliton steps; if the perpendicular-oriented device still shows strong Raman lasing or no soliton access, the orientation rule is not the enabling mechanism.","tokens_in":10129,"feed_emoji":"💡","tokens_out":4753,"duration_ms":41317,"temperature":0.7,"pith_summary":"This paper reports the first soliton microcombs on X-cut thin-film lithium niobate (TFLN), a platform prized for co-integrating high-speed modulators and frequency doublers but previously blocked from soliton formation by strong Raman gain. The authors show that the block is orientation-dependent: if the straight sections of a racetrack resonator hold the TE mode polarized perpendicular to the crystal's optical axis, the Raman response drops, and a continuous-wave pump produces 25 GHz soliton microcombs spanning over 200 nm. With a pulsed pump, the span extends beyond 350 nm. If the claim holds, it removes the last major obstacle to fully monolithic, fast-tunable and self-referenced microcombs on lithium niobate.","feed_headline":"Racetrack orientation unlocks soliton microcombs on X-cut LiNbO3","feed_subtitle":"TE-mode combs at 25 GHz span 200 nm cw and 350 nm pulsed, clearing the X-cut TFLN barrier.","key_machinery":"The load-bearing element is the racetrack geometry itself. Because most of the mode energy lives in the straight waveguides, rotating the straight sections by 90 degrees relative to the optical axis switches the dominant contribution from extraordinary (strong Raman) to ordinary (weak Raman) light behavior, without changing material or etch process. This polarization-orientation design rule is what suppresses stimulated Raman scattering enough for soliton formation; auxiliary-laser thermal control and wavelength-dependent coupling losses are supporting tools used to reach and stabilize the soliton state.","core_discovery":"The central discovery is that TE-polarized soliton microcombs can be generated in X-cut TFLN microresonators by orienting the racetrack so that the polarization of the TE mode in the long straight waveguides is perpendicular to the optical axis. In this orientation the Raman peaks A(TO)1 and A(TO)4 fall to roughly 40% and 20% of their parallel-polarization intensity, suppressing Raman lasing and letting Kerr soliton formation win. The resulting single-soliton state has an $sech^{2}$ spectrum, a 25.037 GHz repetition-rate beat note, and more than 200 nm span under CW pumping; under pulsed pumping the span reaches about 350 nm, with deviations below 1470 nm attributed to birefringence-induced TE-TM mixing. The paper claims this removes the previously identified Raman barrier specific to X-cut TFLN while retaining the platform's electro-optic and nonlinear advantages.","pith_inferences":["The paper's two-point comparison (parallel vs. perpendicular) leaves the threshold angle untested; a natural extension would be to map soliton existence across intermediate orientation angles (15°, 30°, 45°, 60°, 75°) to find where Raman lasing yields to Kerr solitons.","One could also rotate only the pump-coupling section rather than the whole racetrack, trading Raman suppression against modal confinement—an optimization the paper does not explicitly explore.","The 200 nm CW span at 25 GHz repetition implies tens of thousands of comb lines, so a practical test would be to demonstrate error-free wavelength-division multiplexed transmission on a subset of those lines.","The observed TE-TM mixing below 1470 nm suggests that birefringence engineering, such as waveguide width tuning, could flatten and extend the spectrum without changing the orientation design rule."],"forward_implications":["Monolithic X-cut TFLN chips can now combine soliton microcombs with high-speed electro-optic modulators and PPLN frequency doublers on one platform, enabling on-chip feedback control of both repetition rate and carrier-envelope offset.","The orientation rule applies to smaller racetracks with larger free-spectral ranges, so repetition rates in the hundreds of gigahertz, suited to wavelength-division multiplexing, become accessible on the same platform.","With pulsed pumping, the 350 nm span brings octave-spanning combs within reach once dispersion engineering and power budgets are improved, allowing on-chip f-2f self-referencing.","The same ordinary-versus-extraordinary polarization lever helps suppress Raman parasitics in other anisotropic photonic platforms, notably lithium tantalate, as the paper notes from prior work."],"supporting_citations":[{"why":"Supplies the Raman mode assignment and polarization dependence used to identify A(TO)1 and A(TO)4 in lithium niobate.","marker":"[22]"},{"why":"Provides reference Raman spectra for lithium niobate and lithium tantalate with force-field calculations, used in the spectroscopic comparison.","marker":"[30]"},{"why":"Establishes the precedent that incorporating more ordinary light in lithium tantalate mitigates the Raman effect, motivating the same approach here.","marker":"[31]"},{"why":"Prior lithium niobate microcomb work from the group that the present racetrack design extends.","marker":"[16]"},{"why":"Documents Raman lasing and soliton mode-locking in lithium niobate microresonators, defining the barrier the new orientation rule addresses.","marker":"[25]"},{"why":"Introduces photonic dissipation control for Kerr soliton generation in strongly Raman-active media, informing the wavelength-dependent coupling design.","marker":"[26]"},{"why":"Provides the standard continuous-wave laser tuning protocol used to access soliton states in microresonators.","marker":"[34]"},{"why":"Demonstrates temporal solitons in microresonators driven by optical pulses, the method used here to extend the comb span.","marker":"[42]"}],"fun_headline_variants":["Racetrack rotation silences Raman to seed soliton microcombs on X-cut LiNbO3","Rotate the racetrack to beat Raman and get soliton combs on X-cut LiNbO3","Raman no match for rotated racetrack: soliton microcombs on X-cut LiNbO3","X-cut LiNbO3 soliton combs: simply rotate the racetrack"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The claim rests on the assumption that the difference between the two devices comes from the orientation-dependent Raman response, not from the different Q factors or other fabrication variations, and that the polarization-dependent Raman spectra measured on the planar film transfer to the etched waveguide mode.","fun_headline_variants_meta":{"raw":{"variants":["Racetrack rotation silences Raman to seed soliton microcombs on X-cut LiNbO3","Rotate the racetrack to beat Raman and get soliton combs on X-cut LiNbO3","Raman no match for rotated racetrack: soliton microcombs on X-cut LiNbO3","X-cut LiNbO3 soliton combs: simply rotate the racetrack"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000943,"raw_usage":{"total_tokens":4040,"prompt_tokens":966,"completion_tokens":3074,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":582,"completion_tokens_details":{"reasoning_tokens":2969}},"tokens_in":582,"tokens_out":3074,"duration_ms":21409,"temperature":1.0,"reasoning_tokens":2969,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-08T13:32:27.836868+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Fabricate two racetrack resonators with identical geometry, etch depth, and loaded Q near 3 million, differing only in straight-section orientation, and compare the threshold for Raman lasing and the existence of soliton steps; if the perpendicular-oriented device still shows strong Raman lasing or no soliton access, the orientation rule is not the enabling mechanism.","supporting_citations":[{"cited_title":"& Weber, M","cited_arxiv_id":null,"evidence_quote":"Supplies the Raman mode assignment and polarization dependence used to identify A(TO)1 and A(TO)4 in lithium niobate."},{"cited_title":"& Proust, C","cited_arxiv_id":null,"evidence_quote":"Provides reference Raman spectra for lithium niobate and lithium tantalate with force-field calculations, used in the spectroscopic comparison."},{"cited_title":"Broadband microwave-rate dark pulse microcombs in dissipation-engineered LiNbO$_3$ microresonators","cited_arxiv_id":"2404.19584","evidence_quote":"Prior lithium niobate microcomb work from the group that the present racetrack design extends."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Introduces photonic dissipation control for Kerr soliton generation in strongly Raman-active media, informing the wavelength-dependent coupling design."},{"cited_title":"& Herr, T","cited_arxiv_id":null,"evidence_quote":"Demonstrates temporal solitons in microresonators driven by optical pulses, the method used here to extend the comb span."}],"review_version":1}