Pith. sign in

REVIEW 3 major objections 3 minor 1 cited by

Hybrid integrated ultra-low linewidth coil stabilized isolator-free widely tunable external cavity laser

T0 review · 3 major / 3 minor · reviewed 2026-08-10 · deepseek-v4-flash

Pith's one-line read A hybrid-integrated external-cavity laser locked to an on-chip 10-meter coil resonator reports 3–7 Hz fundamental linewidth over 60 nm and 27–60 Hz integral linewidth over 40 nm, with no optical isolator.

desk verdict Genuine experimental advance with an addressable close-to-carrier measurement question; referees should engage. read the letter →

arxiv 2501.15010 v2 pith:TWOEJRYM submitted 2025-01-25 physics.optics quant-ph

classification physics.opticsquant-ph
keywords externalcavitylasersiliconnitridephotonicsintegratedcoilresonatorfrequencystabilizationlinewidthreductionopticalfeedbackresiliencePound-Drever-Halllocking
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper reports a chip-scale route to ultra-low-phase-noise, widely tunable lasers: a hybrid-integrated silicon nitride external-cavity tunable laser (ECTL) is Pound-Drever-Hall locked to an on-chip 10-meter coil resonator, with no optical isolator between them. The reported performance is a fundamental linewidth of $3$ to $7$ Hz across a 60 nm tuning range, an integral linewidth of $27$ to $60$ Hz across a 40 nm range, and an Allan deviation of $1.8\times10^{-13}$ at 6.4 ms. A sympathetic reader would care because precision applications in quantum computing, atomic clocks, metrology, and mmWave/RF generation currently use table-scale bulk-optics systems, and this design points to a fully integrated, portable source with comparable stability. If the numbers hold, it would be the lowest integral linewidth reported for a widely tunable integrated laser.

What carries the argument

The load-bearing pieces are the 10-meter coil resonator and the dual-ring ECTL. The coil, a bus-coupled spiral waveguide with an intrinsic Q around 200 million and a 20 MHz free-spectral range, provides a large mode volume that lowers the thermorefractive noise floor and offers lock frequencies almost everywhere across the tuning range. The ECTL's two high-Q Vernier rings (intrinsic Q 3.5 million) and its tunable Sagnac loop mirror extend the cavity photon lifetime, which narrows the fundamental linewidth and, through the feedback parameter, gives the roughly 30 dB feedback resilience. A Pound-Drever-Hall servo using electro-optic-modulator sidebands feeds back to the gain chip current, transferring the coil's low noise onto the laser.

What would settle it

Measure the locked laser's beat with an independent, sub-Hz-linewidth reference laser using a phase-noise analyzer from 1 Hz to 100 kHz, integrate the frequency-noise spectrum with the same $1/\pi$ method, and compare with the paper's 27 Hz at 1550 nm; an independently measured integral linewidth well above 27 Hz would contradict the central claim.

Watch

Extended reading notes

Core claim

The paper claims that a widely tunable integrated laser can be stabilized to a single on-chip reference cavity and still deliver the low phase noise previously reserved for table-top systems. The device is a silicon nitride external-cavity tunable laser (ECTL) with two high-Q Vernier ring resonators and a Sagnac loop mirror, hybrid-integrated with a reflective semiconductor optical amplifier, locked through a Pound-Drever-Hall servo to a 10-meter-long silicon nitride coil resonator fabricated in the same process. At 1550 nm the stabilized laser shows a 6.08 Hz fundamental linewidth (3–7 Hz across the 60 nm tuning range), a 27 Hz $1/\pi$ integral linewidth (27–60 Hz across 40 nm), frequency noise reduced by more than six orders of magnitude at low offsets, and an Allan deviation of $1.8\times10^{-13}$ at 6.4 ms. Because the ECTL's long cavity photon lifetime and high-Q rings make it resilient to optical feedback, the laser runs isolator-free: it tolerates up to $-10$ dB of feedback, about 30 dB more than a commercial DFB, and remains single-mode without frequency-noise degradation.

Load-bearing premise

The reported integral linewidths assume that the precision frequency counter used for offsets below 3 kHz captures all close-to-carrier frequency noise without adding or masking any; if slow jitter escapes the counter, the true 27–60 Hz numbers would be higher.

Editorial extensions

If this is right

  • Precision systems that currently rely on table-scale stabilized lasers could be served by a chip-scale source with comparable noise performance.
  • The isolator-free lock means the laser and reference cavity can be integrated onto a single silicon nitride chip in a common CMOS-compatible process.
  • The 20 MHz free-spectral range of the 10-meter coil allows the lock to be acquired at almost any wavelength across the 60 nm tuning range, avoiding the single-resonance limitation of bulk reference cavities.
  • The paper predicts that higher loaded ring Q and longer coil resonators will push fundamental and integral linewidths still lower, because the Schawlow-Townes linewidth scales as $1/Q^2$ and the thermorefractive noise floor falls with mode volume.
  • The common fabrication platform is claimed to extend from visible to shortwave-infrared wavelengths, covering atomic transitions and fiber communication bands.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The feedback-resilience mechanism is not specific to the Vernier geometry: any external-cavity laser with a long cavity photon lifetime and high effective mirror reflectivity should show similar isolator-free tolerance, so the design principle could be ported to other wavelengths and gain chips.
  • The observed drop in frequency noise as feedback increased suggests controlled, phase-coherent feedback can act like an extension of the laser cavity; this could be studied deliberately as a linewidth-narrowing tool rather than avoided as a hazard.
  • The 30 dB figure is relative to one commercial DFB class; a standardized 'feedback level at coherence collapse' metric across integrated lasers would make such resilience claims comparable between platforms.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 3 minor

Summary. The manuscript reports a hybrid-integrated Si3N4 external cavity tunable laser (ECTL) stabilized to a 10-meter-long coil resonator. The authors claim a 60 nm tuning range, 3–7 Hz fundamental linewidth, 27–60 Hz integral linewidth, an Allan deviation of 1.8e-13 at 6.4 ms, and isolator-free operation with roughly 30 dB improved feedback resilience compared to a commercial DFB laser. The laser and reference cavity are co-fabricated in an 80 nm Si3N4 platform. The central experimental results are direct frequency-noise, linewidth, and Allan-deviation measurements using a ULE-cavity-stabilized frequency comb and a fiber-MZI discriminator.

Significance. If the reported close-to-carrier noise numbers are accurate, this is a significant advance in integrated widely tunable lasers: it would be the lowest integral linewidth reported for such a laser across a wide tuning range, and the demonstration of isolator-free stabilization to an integrated reference cavity is an important step toward fully integrated frequency-stabilized sources. The use of direct measurements rather than fitting, the co-fabrication claim, and the explicit reporting of both fundamental and integral linewidths across a tuning range are strengths. However, the headline ILW and ADEV values depend on a frequency-counter-based close-to-carrier measurement whose stitching with the OFD data is not fully validated, and the paper reports inconsistent ADEV values without measurement details.

major comments (3)
  1. [Supplementary Note 4; Fig. 3a] The close-to-carrier frequency noise measurement, which dominates the claimed 27 Hz integral linewidth, is stitched from a frequency-counter measurement below 3 kHz and an OFD measurement above 3 kHz. However, Supplementary Note 4 explicitly states that the counter measurement 'becomes limited by the speed of the frequency counter at frequency offsets above ~1 kHz.' If the counter response rolls off between 1 kHz and 3 kHz, the stitched spectrum in that band could be attenuated, and the integrated noise—and therefore the ILW—would be understated. Please justify the 3 kHz stitch frequency, provide a calibration of the counter frequency response, or re-analyze the ILW using a conservative stitch frequency (e.g., 1 kHz) and show the resulting ILW sensitivity.
  2. [Abstract, Results, Discussion, Table 1] The Allan deviation is reported inconsistently: the abstract and main text state 1.8e-13 at 6.4 ms, while the Discussion and Table 1 state 1.6e-13 at the same averaging time. Since ADEV is a headline claim, this discrepancy must be resolved. In addition, the ADEV computation lacks essential metadata: no gate time, dead time, number of samples, or drift-removal procedure are reported. Please provide the measurement details and a single consistent value with uncertainty.
  3. [Results, Fig. 3] No error bars or uncertainty estimates are given for the fundamental linewidths, integral linewidths, Allan deviation, or the 30 dB feedback-resilience comparison. Given that the record claims rest on quantitative comparisons with prior work, the paper should report measurement uncertainties for these quantities. At minimum, specify the statistical and systematic uncertainties in the FN spectrum and the ILW integration.
minor comments (3)
  1. [Throughout] The symbol '1/p' appears in several places (abstract, Table 1) but should be '1/π' for the integral linewidth definition; the text uses both forms and should be made consistent.
  2. [Optical feedback measurements, Fig. 4] The 30 dB feedback-resilience comparison relative to a commercial DFB relies on a coherence-collapse threshold of -40 dB taken from prior literature (refs 16, 43–45) rather than measured in the same setup; please state this explicitly in the text and clarify that the comparison is to a representative published DFB value.
  3. [Table 1] For the comparison entries, the ILW values are described as calculated from published FN data, but no description of the calculation method or the frequency range used is provided. A brief note on how these calculations were performed would strengthen the comparison.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the headline linewidth, stability, and feedback-resilience claims are direct measurements with standard parameter-free processing; self-citations are not load-bearing.

full rationale

All headline quantities (3–7 Hz fundamental linewidth, 27–60 Hz integral linewidth, 1.8e-13 Allan deviation, 60 nm tuning, and ~30 dB feedback resilience) are experimental measurements rather than outputs of a fitted model or of a self-referential derivation. The integral linewidth is computed by integrating the measured frequency-noise spectrum using the standard 1/π reverse-integration convention; the cited prior work (ref. 40) supplies only that integration convention and is parameter-free, so it is not load-bearing circular evidence under Hard Rule 4. The fundamental linewidth is read from the measured white-frequency-noise floor of the OFD spectrum. The TRN and PTN dashed curves in Fig. 2e and Supplementary Fig. 3 are model estimates used for comparison, not parameters fitted to the laser's headline linewidths. The feedback-resilience claim rests on a direct measurement of frequency noise under controlled feedback levels (-32 to -10 dB) compared with a commercial DFB coherence-collapse threshold quoted from the literature; no parameter is fitted to the claimed 30 dB improvement. The self-citations that appear are to the authors' earlier coil-resonator, Brillouin-laser, and frequency-noise-measurement papers, and none is used to forbid alternatives or to define the target quantity in terms of itself. The close-to-carrier stitching concern in Supplementary Note 4 (the frequency counter becoming limited above ~1 kHz while the composite spectrum is stitched at 3 kHz) is a real measurement-validity question about possible understatement of low-offset noise, but it is not circularity: the ILW is still obtained from measured data rather than from an input that already contains the claimed result. Likewise, the abstract/Discussion discrepancy in Allan deviation (1.8e-13 vs 1.6e-13 at 6.4 ms) is an internal-consistency issue, not a circular-reasoning issue. I find no step in which an equation reduces to its own inputs, no fitted parameter renamed as a prediction, and no load-bearing self-citation chain; the derivation is self-contained against external measurement benchmarks.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The central claims are experimental, so the ledger contains no fitted parameters that produce the headline numbers. The free parameters listed are used only in supplementary noise-floor calculations. The main assumptions are about the reliability of the close-to-carrier measurement and the representativeness of the DFB baseline.

free parameters (3)
  • Photothermal absorption loss fraction xi = not stated (estimated experimentally)
    Used in Eq. (1) of Supplementary Note 4 to estimate the photothermal noise floor; does not feed the headline linewidth.
  • Estimated on-chip optical power P_int = not stated
    Used in the photothermal frequency noise model; estimated rather than directly measured.
  • Linewidth enhancement factor alpha = not stated
    Used in the feedback C-parameter estimate in Supplementary Note 5; taken from literature for a semiconductor gain chip.
assumptions (4)
  • standard math Integral linewidth is obtained by 1/pi reverse integration of the measured frequency noise spectrum.
    Standard definition used in the field; cited in Methods.
  • domain assumption The frequency counter with the ULE-comb reference measures the laser's close-to-carrier frequency noise without significant added noise.
    The 27-60 Hz integral linewidth values depend on this measurement; the paper states the counter measurement becomes limited by counter speed above roughly 1 kHz.
  • domain assumption The 10-meter coil resonator's thermorefractive noise limit, calculated from its measured quality factor, is the actual floor reached by the stabilized laser.
    The stabilized frequency noise spectrum is compared with this calculated limit; errors in the TRN model would affect interpretation, though not the measured linewidth.
  • domain assumption A commercial III-V DFB laser collapses at -40 dB feedback, as reported in prior literature.
    The paper's 30 dB improvement claim uses this literature baseline rather than a same-setup measurement.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Hybrid integrated ultra-low linewidth coil stabilized isolator-free widely tunable external cavity laser." pith.science (2026). https://pith.science/paper/TWOEJRYM

@misc{pith2026250115010,
  author       = {Pith},
  title        = {Pith review of: Hybrid integrated ultra-low linewidth coil stabilized isolator-free widely tunable external cavity laser},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TWOEJRYM}},
  note         = {Machine review of arXiv:2501.15010}
}
read the original abstract

Precision applications including quantum computing and sensing, mmWave/RF generation, and metrology, demand widely tunable, ultra-low phase noise lasers. Today, these experiments employ table-scale systems with bulk-optics and isolators to achieve requisite noise, stability, and tunability. Photonic integration will enable scalable, reliable and portable solutions. Here we report a hybrid-integrated external cavity widely tunable laser stabilized to a 10-meter-long integrated coil-resonator, achieving record-low 3 - 7 Hz fundamental linewidth across a 60 nm tuning range and 27 - 60 Hz integral linewidth with 1.8E-13 Allan deviation at 6.4 ms across 40 nm, delivering orders of magnitude frequency noise and integral linewidth reduction over state of the art. Stabilization is achieved without an optical isolator, leveraging resilience to optical feedback of 30 dB beyond that of commercial DFB lasers. The laser and reference cavity are fabricated in the same Si3N4 CMOS-compatible process, unlocking a path towards fully integrated visible to ShortWave-IR frequency stabilized lasers.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Kerr nonlinearity, self-injection locking and correlation in a microresonator

    physics.optics 2025-06 conditional novelty 6.0 of 10

    Two self-injection-locked, counter-propagating lasers in a microresonator produce non-degenerate four-wave-mixing sidebands, and the pump lasers become correlated when the sidebands are generated.

Reference graph

Works this paper leans on

11 extracted references · 9 canonical work pages · cited by 1 Pith paper

  1. [1]

    1 Hybrid integrated ultra-low linewidth coil stabilized isolator-free widely tunable external cavity laser David A. S. Heim1, Debapam Bose1, Kaikai Liu1, Andrei Isichenko1, and Daniel J. Blumenthal1* 1Department of Electrical and Computer Engineering, University of California Santa Barbara, Santa Barbara, CA 93106 USA. *Corresponding author (danb@ucsb.edu...

  2. [2]

    b Fundamental vs

    a Frequency noise plots comparing hybrid-integrated low noise lasers. b Fundamental vs. 1/p integral linewidths. The bubble size represents the tuning range and color heat map represents output power. Increased Tuning Range a b 11 Since both the ECTL and coil resonator are fabricated in the same 80 nm thick, CMOS-compatible Si3N4 platform, these results p...

  3. [3]

    Liu, K. et al. Photonic circuits for laser stabilization with integrated ultra-high Q and Brillouin laser resonators. APL Photonics 7, 096104 (2022)

  4. [4]

    Liu, K. et al. 36 Hz integral linewidth laser based on a photonic integrated 4.0 m coil resonator. Optica 9, 770 (2022)

  5. [5]

    Liu, K. et al. Common cavity waveguide coil-resonator stabilized hybrid integrated WDM laser with 89 Hz integral linewidth. in 2024 Optical Fiber Communications Conference and Exhibition (OFC) 1–3 (2024)

  6. [6]

    Gundavarapu, S. et al. Sub-hertz fundamental linewidth photonic integrated Brillouin laser. Nature Photon 13, 60–67 (2019)

  7. [7]

    Chauhan, N. et al. Visible light photonic integrated Brillouin laser. Nat Commun 12, 4685 (2021)

  8. [8]

    Zhang, Z. et al. High-Speed Coherent Optical Communication With Isolator-Free Heterogeneous Si/III-V Lasers. Journal of Lightwave Technology 38, 6584–6590 (2020)

Show all 11 references
  1. [9]

    Gomez, S. et al. High coherence collapse of a hybrid III–V/Si semiconductor laser with a large quality factor. J. Phys. Photonics 2, 025005 (2020)

  2. [10]

    Harfouche, M. et al. Kicking the habit/semiconductor lasers without isolators. Opt. Express 28, 36466 (2020)

  3. [11]

    Oldenbeuving, R. M. et al. 25 kHz narrow spectral bandwidth of a wavelength tunable diode laser with a short waveguide-based external cavity. Laser Phys. Lett. 10, 015804 (2012)

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.