{"id":"579b4b03-d492-4979-bc77-7cebb7939fe9","arxiv_id":"2507.16948","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"An integrated PZT actuator on a silicon nitride soliton microcomb performs optical frequency division, producing a 109.5 GHz signal with -114 dBc/Hz phase noise at 10 kHz offset.","lead":"This paper demonstrates a photonic chip that converts stable optical frequencies into low-noise microwave signals, using a piezoelectric actuator to tune a soliton microcomb. The work adds a fast, integrated control mechanism for chip-scale optical frequency division, relevant to compact radar, communications, and precision timing.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Unverified VCO phase-noise assumption underpins both the noise attribution and the 'full stabilization' claim; a direct VCO phase-noise measurement is needed to confirm the OFD result is optically-reference-limited.","rationale":"The paper's central demonstration is credible: PZT tuning coefficient of 43.7 MHz/V, 13 MHz modulation bandwidth, soliton generation, phase locking, two independent phase-noise measurement methods, and comparison to a previous non-PZT OFD oscillator. The raw result of -114 dBc/Hz at 10 kHz offset is a measurement and stands on its own. The issue is interpretational: the paper claims the output is optically-reference-limited and fully stabilized, using an unmeasured inequality about VCO noise. Because the OFD equation explicitly contains VCO terms, this inequality is load-bearing. If the VCOs have phase noise near -95 dBc/Hz or higher at 10 kHz at their operating frequencies, the divided contribution could be comparable to the measured trace, and the 'full stabilization' wording would be misleading. The requested measurement is straightforward with commercial equipment and would settle the point. No internal inconsistency beyond this was found; the N=-54/55 rounding discrepancy is within the precision of the stated laser frequencies and does not affect the conclusion. Therefore the reader's CONDITIONAL verdict is appropriate; this concern reinforces it rather than changing it.","tokens_in":7084,"tokens_out":10797,"duration_ms":109135,"concrete_test":"Measure the SSB phase noise of VCO1 and VCO2 at their operating frequencies with a calibrated phase-noise analyzer over 100 Hz to 1 MHz offset, with the OFD loop open but the VCOs running under nominal settings. Compute the VCO contribution to the repetition-rate phase noise as 10log10(10^(S1/10)+10^(S2/10)) - 20log10(N), where S1(f) and S2(f) are the measured VCO phase-noise spectra in dBc/Hz and N is the actual comb-line index used (near 54-55). Overlay this trace on Fig. 3c. If the VCO trace is >10 dB below the measured PZT-OFD phase noise at 10 kHz and across the plotted offsets, the assumption and the 'full stabilization' claim are supported. If it approaches or exceeds the measured trace, the phase noise is not optically-reference-limited and the OFD claim must be revised.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central OFD equation in Results, fr = (fB - fA)/N + (fVCO2 - fVCO1)/N, retains the two VCOs in the frequency chain, but the noise analysis immediately drops them: 'the phase noise of the two VCOs is much lower than that of the two optical reference lasers', giving Sr = (SA + SB)/N^2. No VCO phase-noise data or model are presented. The headline number, -114 dBc/Hz at 10 kHz offset at 109.5 GHz, is interpreted as transferring optical-reference stability to the mmWave, and the Summary further claims the comb is 'fully phase stabilized' with 'no free-running parameter', even though f0 = fA + fVCO1 and fN = fB + fVCO2 depend on two free-running VCOs. If the VCO phase noise is not far below the optical-reference contribution at the relevant offsets, the measured phase noise could be limited by the VCOs (or the servo) rather than by the divided optical references, and the 'full stabilization' statement would be overstated. This does not invalidate the raw mmWave measurement or the PZT tuning demonstration, but it is the least-secure assumption supporting the OFD interpretation.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports an on-chip optical frequency division (OFD) demonstration in which a PZT-integrated SiN soliton microcomb is phase-locked to two reference lasers that are themselves stabilized to an integrated 4-m coil cavity. The integrated PZT actuator tunes the soliton repetition rate, with a measured tuning coefficient of 43.7 MHz/V and a small-signal 3-dB bandwidth around 13 MHz. Photodetection of the stabilized comb on a MUTC-PD produces a 109.5 GHz millimeter-wave signal with phase noise of -114 dBc/Hz at 10 kHz offset, equivalent to -135 dBc/Hz when rescaled to 10 GHz. The authors claim that the phase noise is dominated by the divided optical references, and that the comb is fully phase-stabilized with no free-running parameter in its frequency.","tokens_in":7343,"tokens_out":3522,"duration_ms":39560,"significance":"This work is significant because it adds a fast integrated tuning actuator to OFD, which has previously relied on pump-laser tuning or passive injection locking in on-chip demonstrations. The PZT tuning speed (13 MHz bandwidth) far exceeds typical OFD servo bandwidths, and the demonstration uses only the two reference lasers for both soliton generation and OFD locking, simplifying the architecture. The measured phase noise is competitive with prior integrated OFD results. Strengths of the manuscript include direct characterization of the PZT actuator, clear presentation of the frequency-division relation, and phase-noise data under both free-running and locked conditions. No fitted parameters or circular derivations are used in the analysis. The main weakness is an unverified assumption about VCO phase noise that underpins the optical-reference-limited interpretation.","major_comments":[{"comment":"The assumption that 'the phase noise of the two VCOs is much lower than that of the two optical reference lasers' is stated without measurement, model, or upper bound. The equations f0 = fA + fVCO1 and fN = fB + fVCO2 retain the VCOs in the frequency chain, and the noise expression Sr = (SA + SB)/N^2 drops them by assumption. Because the central claim is that the measured -114 dBc/Hz at 10 kHz reflects the divided optical references, the VCO contribution must be verified. I recommend adding direct VCO phase-noise measurements at the relevant offset frequencies, or at least a conservative estimate of the VCO phase noise referenced to the comb line. Without this, the measured phase noise could be VCO- or servo-limited rather than optically-reference-limited.","section":"Results, second paragraph (frequency-division equation)"},{"comment":"The claim that the comb is 'fully phase stabilized' and that this 'eliminates any free-running parameter in the frequency of the microcomb' is overstated. The zero-th comb line f0 = fA + fVCO1 and the locked line fN = fB + fVCO2 both depend on free-running VCOs, so the comb frequencies are not determined solely by the optical references and the microresonator. While the comb is phase-locked relative to the references, the VCOs still contribute phase noise and their absolute frequencies are not stabilized. Please qualify the statement to acknowledge the residual VCO dependence, e.g., by saying the comb lines are stabilized to the references with a residual VCO noise contribution.","section":"Summary, last sentence"},{"comment":"The resonance dip around 20 MHz is attributed to the electronic circuits 'likely' without supporting evidence. Since the paper uses this measurement to claim a 3-dB bandwidth of about 13 MHz and to argue that the actuator will not limit OFD bandwidth, the origin of the dip matters. If the dip is an electrical artifact of the probe or cabling, a control measurement (e.g., with different cables, a 50-ohm through, or a network analyzer calibration) should be reported; if it is a real feature of the PZT response, the claim that the actuator bandwidth exceeds tens of MHz needs to be qualified. As presented, this is a speculation that could be resolved with a simple measurement.","section":"Results, S21 characterization (Figure 1e, text around 'resonance dip')"}],"minor_comments":[{"comment":"The carrier frequency is given as 109.5 GHz in the Results and Abstract, but Figure 2 and Figure 3 label it as 110 GHz. Please unify the value.","section":"Throughout"},{"comment":"The phrase '10s MHz' should be written as 'tens of MHz' or '10s of MHz' for clarity.","section":"Abstract and Introduction"},{"comment":"Panel (c) is marked 'not drawn to scale' and the layer labels are small; please enlarge the zoomed-in cross section and label all layers (Au, Si3N4, SiO2, Si, Pt) more legibly.","section":"Figure 1"},{"comment":"This is a useful detail, but the power of the comb line used for OFD locking is not stated. Reporting the comb-line power and the beatnote SNR would help readers assess the locking robustness.","section":"Results, 'No optical amplifier is used'"},{"comment":"The comparison with the previous non-PZT integrated OFD oscillator (Ref. 19) should state the offset frequency and measurement conditions for both data sets to make the comparison meaningful.","section":"Results, '2 dB better'"},{"comment":"The data and code are available only 'upon reasonable request.' For reproducibility, consider depositing the phase-noise traces and analysis scripts in a public repository.","section":"Data and code availability"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is within the scope of physics.optics and reports a genuinely useful advance in integrated OFD actuation. The main technical concern is the unverified VCO phase-noise assumption; if the authors can add a direct measurement or a conservative bound, the central interpretation would be secure. The 'full stabilization' wording should be softened regardless. I saw no sign of circularity, fabricated data, or unsupported claims beyond the ones noted in the major comments."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick read: the paper does what it says. It puts an integrated PZT stress-optic actuator inside the OFD feedback loop for a SiN soliton microcomb, locks the comb to two reference lasers, and generates a 109.5 GHz tone with -114 dBc/Hz at 10 kHz. To my knowledge the combination is new; prior chip OFD work tuned via pump laser or injection locking. The PZT actuator bandwidth (~13 MHz) and tuning coefficient (43.7 MHz/V) are cleanly characterized, and the phase-noise data show clear suppression versus free-running soliton. The mmWave-to-microwave division and dual-tone self-heterodyne methods are established, and the agreement between them is decent. This is a solid engineering advance, not a paradigm shift.\n\nWhere I would push back: the noise attribution rests on one sentence. The OFD equation explicitly keeps the two VCOs, f_r = (f_B - f_A)/N + (f_VCO2 - f_VCO1)/N, and then they are dropped with 'the phase noise of the two VCOs is much lower than that of the two optical reference lasers.' No VCO phase-noise data or model appears anywhere. That matters because the headline -114 dBc/Hz is interpreted as transferring optical-reference stability to the mmWave, and the summary claims the comb is 'fully phase stabilized' with 'no free-running parameter' while the two VCOs are free-running in the frequency chain. If the VCO noise is not far below the optical-reference contribution at relevant offsets, the measured floor could be VCO- or servo-limited. The raw mmWave measurement and the PZT tuning demonstration survive regardless, but the 'full stabilization' sentence should be softened and the VCO noise measured.\n\nMinor issues: the 20 MHz S21 dip is attributed to electronics without direct evidence; the servo bandwidth being 200 kHz while the actuator is 13 MHz is left as 'we believe limited by the loop'—probably true, but a measurement would help. Data and code are request-only, which is normal for this group but keeps reproducibility at 'ask and see.'\n\nCitation pattern looks fine; self-citations to prior OFD and PZT work are used for methods, and the central claim is independent. The paper is concise, honest about what was measured, and does not oversell the actuator speed. The only real fix needed is the VCO noise check plus a revised stabilization claim. I would send it to review; a referee should ask for the VCO phase-noise data before acceptance.","headline":"Solid demonstration of PZT-based OFD with a real but narrow advance; the main caveat is the unmeasured VCO phase-noise floor.","tokens_in":7907,"tokens_out":1655,"would_cite":true,"duration_ms":18027,"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":"This paper demonstrates optical frequency division on a chip using an integrated piezoelectric actuator that rapidly tunes a soliton microcomb's repetition rate, yielding a 109.5 GHz signal with -114 dBc/Hz phase noise at 10 kHz offset.","keywords":["optical frequency division","soliton microcomb","piezoelectric actuator","PZT","millimeter-wave generation","phase noise","integrated photonics","frequency stabilization"],"falsifier":"Directly measure the phase noise of the two VCOs at 10 kHz offset; if either VCO's noise is comparable to or larger than the divided optical reference noise, the observed -114 dBc/Hz floor cannot be attributed to optical frequency division.","tokens_in":6943,"feed_emoji":"📡","tokens_out":10639,"duration_ms":96939,"temperature":0.7,"pith_summary":"Optical frequency division transfers the stability of a laser reference to a microwave or millimeter-wave carrier by locking a frequency comb to that reference. This paper shows that a piezoelectric (PZT) actuator integrated directly onto a silicon-nitride soliton microcomb cavity can provide the fast, direct tuning of the comb's repetition rate that this locking requires, with bandwidth exceeding tens of megahertz. Using two reference lasers locked to a 4-meter coil cavity, the authors lock the $N=-54$ comb line via the PZT, achieving a division ratio of 54 and a 109.5 GHz millimeter-wave output. The measured phase noise is -114 dBc/Hz at 10 kHz offset, equivalent to -135 dBc/Hz when scaled to 10 GHz. The significance is that the entire frequency-comb control chain is integrated and requires no separate pump laser, simplifying chip-scale low-noise oscillators for communications, sensing, and timing.","feed_headline":"PZT-tuned chip comb makes 110 GHz with low noise","feed_subtitle":"A fast piezoelectric actuator locks the microcomb to two references, reaching -114 dBc/Hz at 10 kHz offset.","key_machinery":"The load-bearing component is the integrated PZT stress-optic actuator: a thin lead zirconate titanate film deposited on the oxide cladding of the SiN racetrack microresonator. An applied voltage strains the SiN waveguide and changes its refractive index, shifting the microresonator resonance (43.7 MHz/V) and, with it, the soliton repetition rate, while leaving the comb offset frequency unchanged. The actuator's small-signal 3-dB bandwidth is about 13 MHz, well above the 200 kHz servo bandwidth used in the OFD loop, so the loop electronics—not the actuator—limit the locking speed. This actuator carries the entire feedback signal that phase-locks the $N$-th comb line to the reference laser, which is what transfers optical reference stability to the millimeter-wave carrier.","core_discovery":"The paper reports that an integrated PZT stress-optic actuator on a SiN soliton microcomb resonator can tune the resonance frequency by 43.7 MHz/V with a 3-dB bandwidth around 13 MHz, and that this tuning shifts the soliton repetition rate while leaving the carrier-envelope offset fixed. Using this actuator as the feedback element, the authors lock the $N=-54$ comb line to a 1600 nm reference laser while the $0$-th comb line is derived from a 1550 nm reference laser, both references being PDH-locked to an integrated 4-meter coil cavity. The frequency division then maps a 6 THz optical spacing to the 109.5 GHz repetition rate, a division ratio of 54. The photodetected millimeter-wave output shows a phase noise of -114 dBc/Hz at 10 kHz offset, equivalent to -135 dBc/Hz at 10 GHz, which is 2 dB better than the authors' earlier non-PZT integrated OFD oscillator. The paper interprets this as establishing the PZT actuator as a fast, simple, integrated control mechanism for on-chip optical frequency division.","pith_inferences":["If the offset frequency is truly untouched by the PZT, the same actuator could in principle lock the carrier-envelope offset directly, opening a route to fully self-referenced chip-scale frequency combs without an octave-spanning spectrum.","The independent control of repetition rate and offset suggests a straightforward extension to dual-comb spectroscopy on a chip, where one PZT tunes the line spacing while a second controls the offset for heterodyne detection.","Other integrated stress-optic or electro-optic platforms (for example lithium niobate) could adopt the same OFD feedback architecture, potentially raising the actuator bandwidth from tens of megahertz to gigahertz and further reducing noise at high offset frequencies.","A clean test of the division mechanism would be to repeat the measurement with a different comb line index (say $N=-27$) and verify that the phase noise at 10 kHz scales as $1/N^2$ after correcting for the reference noise."],"forward_implications":["A division ratio of 54 maps 6 THz of optical spacing to a 109.5 GHz carrier, with the two reference lasers' phase noise divided down by $54^2 \\approx 35$ dB.","The entire soliton microcomb is phase-stabilized with no free-running parameters, because both the 0-th comb line and the $N=-54$ comb line are locked to reference lasers.","The PZT actuator's tens-of-megahertz bandwidth leaves headroom beyond the demonstrated 200 kHz servo, so faster electronics could push the lock bandwidth and improve phase noise at larger offsets.","Because the 0-th comb line comes from a modulated sideband of reference laser A, no separate pump laser is needed to generate and tune the comb, simplifying the OFD architecture."],"supporting_citations":[{"why":"Supplies the integrated 4-meter SiN coil reference cavity to which both reference lasers are locked.","marker":"[12]"},{"why":"Provides the prior integrated OFD demonstration without PZT that this work compares against and improves by 2 dB.","marker":"[19]"},{"why":"Demonstrates monolithic piezoelectric control of soliton microcombs, the actuation principle the paper builds on.","marker":"[28]"},{"why":"Describes the SiN stress-optic PZT modulator whose fabrication and tuning are used for the actuator.","marker":"[30]"},{"why":"Provides the MUTC photodiode used to photodetect the soliton comb and generate the 110 GHz mmWave signal.","marker":"[36]"},{"why":"Supplies the dual-tone delayed self-heterodyne method used for optical-domain phase noise measurement above 30 kHz.","marker":"[38]"}],"fun_headline_variants":["Chip-scale PZT-tuned comb yields low-noise mmWave output","PZT actuator on chip microcomb achieves -135 dBc/Hz at 10 GHz","Fast PZT tuning locks soliton comb for low-noise frequency division","Chip-scale comb generates 110 GHz with low noise via PZT","Chip-scale frequency division with PZT-integrated microcomb hits low noise"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The results assume that the two voltage-controlled oscillators used for offset and reference frequencies add noise far below the two optical reference lasers; the paper states this inequality but does not directly measure either VCO's phase noise.","fun_headline_variants_meta":{"raw":{"variants":["Chip-scale PZT-tuned comb yields low-noise mmWave output","PZT actuator on chip microcomb achieves -135 dBc/Hz at 10 GHz","Fast PZT tuning locks soliton comb for low-noise frequency division","Chip-scale comb generates 110 GHz with low noise via PZT","Chip-scale frequency division with PZT-integrated microcomb hits low noise"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00114,"raw_usage":{"total_tokens":4760,"prompt_tokens":1004,"completion_tokens":3756,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":620,"completion_tokens_details":{"reasoning_tokens":3670}},"tokens_in":620,"tokens_out":3756,"duration_ms":24969,"temperature":1.0,"reasoning_tokens":3670,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T14:59:39.263959+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Directly measure the phase noise of the two VCOs at 10 kHz offset; if either VCO's noise is comparable to or larger than the divided optical reference noise, the observed -114 dBc/Hz floor cannot be attributed to optical frequency division.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the integrated 4-meter SiN coil reference cavity to which both reference lasers are locked."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the prior integrated OFD demonstration without PZT that this work compares against and improves by 2 dB."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Demonstrates monolithic piezoelectric control of soliton microcombs, the actuation principle the paper builds on."},{"cited_title":", author Liu, K","cited_arxiv_id":null,"evidence_quote":"Describes the SiN stress-optic PZT modulator whose fabrication and tuning are used for the actuator."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the MUTC photodiode used to photodetect the soliton comb and generate the 110 GHz mmWave signal."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the dual-tone delayed self-heterodyne method used for optical-domain phase noise measurement above 30 kHz."}],"review_version":1}