Pith. sign in

REVIEW 4 major objections 4 minor 26 references

Low threshold integrated optical parametric oscillator with a compact Bragg resonator

T0 review · 4 major / 4 minor · reviewed 2026-08-10 · deepseek-v4-flash

Pith's one-line read A Fabry-Pérot Bragg cavity on lithium niobate reaches a 2.5 mW OPO threshold, the lowest reported for a double-resonant device.

desk verdict Solid LNOI Fabry-Pérot OPO with a plausible but not yet auditable record threshold; worth refereeing, with metrology fixes before the headline number is trusted. read the letter →

arxiv 2501.17050 v1 pith:ID3VKGTZ submitted 2025-01-28 physics.optics

classification physics.optics PACS 42.65.Yj42.60.Da42.82.-m
keywords opticalparametricoscillatorlithiumniobateoninsulatorintegratedphotonicsBraggresonatordouble-resonantOPOdegenerateoperationthermo-optictuningcomputing
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 an integrated optical parametric oscillator on thin-film lithium niobate that reaches oscillation with 2.5 mW of pump power, the lowest threshold reported so far for a double-resonant OPO. The key design change is a Fabry-Pérot cavity made from two Bragg reflectors instead of the usual ring resonator, which shrinks the active footprint to 0.45 mm² and cuts device area by about a factor of ten. The authors show that both pump wavelength tuning and a local thermo-optic heater can place the oscillator at degeneracy, where signal and idler share one frequency and the output phase is bistable. That combination of low threshold, compact footprint, and independent degeneracy tuning is what a spatially multiplexed network of OPOs for phase-encoded optical computing would need.

What carries the argument

The load-bearing component is the Fabry-Pérot cavity formed by two integrated Bragg reflectors in a lithium niobate on insulator waveguide, with a periodically poled section inside for parametric gain. The cavity is double-resonant for signal and idler, while the pump is injected through a directional-coupler wavelength demultiplexer so it does not pass through the Bragg gratings, which would deflect it out of plane. A second WDM is used only for linear characterization, and a thermo-optic electrode shifts one section of the cavity locally to tune the resonance frequencies. This combination does three jobs at once: it keeps the cavity short enough to cut device area by an order of magnitude compared with rings, it lowers the threshold by reducing escape of the signal from the cavity, and it gives an independent tuning knob per device that can bring every OPO to degeneracy at a shared pump frequency.

What would settle it

Measure the absolute on-chip pump power at the input grating and repeat the threshold extraction with a noise-modeled fit over the full wavelength sweep; if the resulting on-chip threshold is higher than one of the published comparison values, the record claim fails.

Watch

Extended reading notes

Core claim

The paper's central claim is that a double-resonant OPO can reach a record-low threshold in a small footprint if the cavity is a Fabry-Pérot resonator defined by two integrated Bragg reflectors rather than a ring. With this layout, the measured threshold is 2.5 mW and the active footprint is 0.45 mm², which the authors compare with published double-resonant OPO thresholds of 25–80 mW and footprints of roughly 8–10 mm². They also demonstrate that the oscillator can be placed at degeneracy by tuning either the pump wavelength or a local thermo-optic phase shifter, and that only every second cluster of supported modes contains the degenerate pair because energy conservation requires the half-pump to sit on a cavity resonance. The intended consequence is a scalable building block for spatially coupled OPO networks.

Load-bearing premise

The claim that 2.5 mW is a record assumes the reported power is measured at the same reference plane (for instance on-chip or in-fiber) as the literature values, and that the threshold-extraction rule of picking the brightest point in each wavelength sweep does not bias the number downward.

Editorial extensions

If this is right

  • If the 2.5 mW threshold transfers to a network setting, many OPO units could share one pump laser with sustainable total power.
  • The Fabry-Pérot layout makes spatial multiplexing practical because the transverse device size is an order of magnitude below ring-resonator OPOs.
  • Independent thermo-optic tuning per device allows multiple OPOs on one chip to operate at degeneracy at the same frequency despite fabrication variations.
  • Operating at degeneracy gives the phase bistability needed to represent Ising spins, so the device is a direct building block for optical Ising machines.
  • The reduced escape efficiency means output power is low, which the authors note is acceptable for phase readout but would limit applications requiring high generated signal power.

Reading between the lines

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

  • A natural extension would be to measure the pump power at the waveguide input directly, so the 2.5 mW claim can be compared with literature values on a common power reference; until then the record depends on an unstated coupling-loss assumption.
  • The same cavity geometry could be tested with resonant pumping instead of a nonresonant pump, which would trade a higher threshold for a different tuning landscape.
  • Coupling two such OPOs through a shared waveguide could test injection locking or phase coupling, which is the step between a single oscillator and a functional Ising network.
  • Repeating the threshold measurement with a noise-aware fit across the whole wavelength sweep, rather than selected maximum points, would show whether 2.5 mW is a stable operating point or an optimistic selection.
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

4 major / 4 minor

Summary. The paper demonstrates a double-resonant integrated optical parametric oscillator in thin-film lithium niobate, using a Fabry-Pérot cavity formed by two Bragg reflectors instead of the usual ring resonator. The authors report a 2.5 mW oscillation threshold, which they claim is the lowest reported for double-resonant OPOs, and a device footprint of 0.45 mm2, about ten times smaller than ring-based devices. They also study the tuning behavior via pump wavelength and a thermo-optic phase shifter, and show that the OPO can be operated at degeneracy, which is relevant for phase-encoded Ising machines. The paper includes linear characterization, dispersion measurements, threshold measurements, and a comparison with previous double-resonant OPOs.

Significance. If the central claim survives scrutiny, this is a meaningful advance for integrated nonlinear photonics: the Fabry-Pérot geometry provides a compact alternative to ring cavities, and the 2.5 mW threshold would substantially lower the power budget for spatially multiplexed OPO networks. The manuscript is also valuable for its detailed characterization of the tuning physics, including the role of dispersion in discrete mode selection and the demonstration of thermo-optic tuning to reach degeneracy. The experimental work is described in considerable detail, and the direct threshold measurement is a clean experimental approach rather than an indirect estimate. However, the headline record-threshold claim currently lacks the calibration and statistical support needed to make it fully auditable against the cited literature.

major comments (4)
  1. [II.D Methods and Fig. 5] The power reference plane for the threshold value is not defined. The Methods state that input power is monitored by a 1% tap before coupling, but the paper does not state whether the 2.5 mW value refers to the fiber-tip power, the power at the grating coupler, or the on-chip waveguide power, and no fiber-to-chip insertion loss is reported. Since the record claim in Table 1 depends on comparing powers at the same reference plane, this omission makes the comparison ambiguous. Please specify the reference plane and provide a coupling-loss calibration or an on-chip power estimate.
  2. [II.B, Fig. 5(a) and 5(b)] The threshold extraction procedure is non-standard and potentially biased. For each pump power, only the highest-output points near the transition between degenerate and non-degenerate operation are collected to build the threshold curve, with no detector noise model and no error bars. This selected-envelope approach can systematically shift the inferred threshold relative to a full-data power-law fit. The authors should justify this procedure quantitatively or provide an alternative analysis using all data points, including an estimate of the uncertainty on the 2.5 mW value.
  3. [II.C, Table 1] The comparison with literature values in Table 1 is not fully apples-to-apples. Entries [14] and [24] are marked as peak power in the pulsed regime, while the present work appears to be continuous-wave; comparing pulsed peak power with CW average power can be misleading. Additionally, the paper does not state the pump repetition rate or pulse duration for those references, nor the exact extraction method used for the threshold in each cited work. Please clarify the operating regime of each comparison and state explicitly whether the threshold values are all on-chip powers measured with comparable methods.
  4. [II.A and II.C, footprint definition] The reported footprint of 0.45 mm2 is not clearly defined. The text says that an additional WDM used only for linear characterization does not contribute to the device footprint 'in the perspective of integrating coupled devices,' but the actual area used for the 0.45 mm2 number is not specified. If the footprint excludes part of the fabricated circuit, this should be stated explicitly so that the factor-of-ten comparison with ring resonators in Table 1 is meaningful.
minor comments (4)
  1. [Abstract] There is a formatting error in the author list ('A LESSANDRA SABATTI') and a typo in the first sentence of the abstract ('T uning' should be 'Tuning').
  2. [II.B] In the paragraph on thermo-optic tuning, the text reads 'for a a pump wavelength' — there is a duplicated 'a'.
  3. [II.B, Fig. 2(c)] The polynomial fit used for the dispersion curve is not described; specifying the polynomial order and the fitted coefficients would improve reproducibility of the tuning simulation.
  4. [Data availability] The data availability statement indicates that data are not public but may be obtained upon request. Given that the record-threshold claim is central, making the threshold dataset and the linear characterization data publicly available would strengthen the paper's auditability.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the central threshold and tuning claims are experimental measurements, with simulations used only as consistency checks.

full rationale

The paper's central claim of a 2.5 mW threshold is a direct experimental measurement, not the output of a model fitted to the same data. The threshold is extracted from selected maximum OPO-power points at each pump power, which is a non-standard procedure that may bias the value, but it is not circular: the reported threshold is not derived from an input that presupposes it. The tuning simulation in Fig. 3 uses the measured resonance dispersion to model which mode pairs satisfy energy conservation, and is explicitly presented as a comparison with measured tuning, not as an independent prediction; the wavelength offset is attributed to temperature drift. The dispersion is measured and then fit, but the fit does not feed into the headline performance number. Literature comparisons in Table 1 use external benchmarks, and the ambiguity about the power reference plane (fiber versus waveguide) affects auditability, not circularity. Self-citations (e.g., refs. 17, 18, 22, 25) concern Bragg reflector and phase-shifter fabrication details and are not load-bearing for the record-threshold claim, which is an externally compared measurement. No step in the derivation chain reduces by construction to its own input.

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

The threshold claim itself is directly measured and does not rely on fitted parameters. The simulation of tuning uses a polynomial fit to the measured dispersion and a wavelength offset, so those are the only fitted inputs, and they do not enter the threshold value. No invented entities are introduced.

free parameters (3)
  • Dispersion polynomial coefficients (D2 and higher) fitted to measured cavity resonances = Not specified numerically in text
    The measured dispersion curve in Fig. 2(c) is fitted with a polynomial and then used to simulate OPO tuning in Fig. 3(b). This is a model fitted to the same device, not a first-principles input.
  • Wavelength offset between linear and OPO spectra = Not quantified
    The paper explains the wavelength shift between experimental and simulated tuning plots as due to temperature differences, effectively a free offset that aligns simulation with data.
  • Thermo-optic resonance shift coefficient = Inset Fig. 4(b), not quantified in text
    The linear fit of resonance position versus applied thermo-optic power is used to interpret tuning; it is measured for this device.
assumptions (4)
  • standard math Energy conservation for parametric oscillation: pump frequency equals sum of signal and idler frequencies.
    Used throughout tuning analysis, e.g., 'because of energy conservation between pump, signal and idler'.
  • domain assumption The PPLN waveguide provides effective chi-2 nonlinearity via quasi-phase matching.
    Assumed for parametric gain; SHG calibration supports it, but the nonlinear coefficient value is not independently measured here.
  • domain assumption Bragg reflectors act as frequency-selective mirrors for signal and idler and do not introduce significant loss at operating powers.
    Stop-band characterization shows high extinction, but the cavity loss and its effect on threshold are not separately quantified.
  • domain assumption Pump does not resonate in the cavity; the WDM separates pump and signal paths.
    Design choice stated in Section A; pumping through the WDM instead of the Bragg reflector because the BR deflects the pump out of plane.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Low threshold integrated optical parametric oscillator with a compact Bragg resonator." pith.science (2026). https://pith.science/paper/ID3VKGTZ

@misc{pith2026250117050,
  author       = {Pith},
  title        = {Pith review of: Low threshold integrated optical parametric oscillator with a compact Bragg resonator},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ID3VKGTZ}},
  note         = {Machine review of arXiv:2501.17050}
}
read the original abstract

Optical parametric oscillators (OPOs) have been studied as basic components for optical computing with phase encoding and Ising machines. Integrated photonics offers a scalable solution to incorporate a progressively larger number of devices towards a functional computing module. Among the available platforms, lithium niobate on insulator is an excellent candidate for this goal thanks to its large second order nonlinearity, which can be leveraged via periodic poling of the thin film. In this work, we show a device with a 2.5 mW threshold for parametric oscillation, which is the lowest reported to date among double-resonant OPOs. We use a novel configuration with a Fabry-P\'erot cavity, which reduces the footprint compared to a typical ring resonator by a factor 10. Tuning our devices using pump wavelength and local heating, we can operate the oscillators at degeneracy, which is crucial for logical operations requiring phase bistability. Our results showcase the device as an ideal building block for phase-encoded integrated optical computing, enabling spatial multiplexing with reduced footprint and power consumption.

Figures

Figures reproduced from arXiv: 2501.17050 by the authors.

Figure 1
Figure 1. Device concept and overview. (a) Sketch of the device showing the Fabry-Pérot cavity, the PPLN waveguide and the thermo￾optic phase shifter. (B) SEM image of a Bragg reflector. (c) SEM image of a PPLN waveguide, with the domain pattern revealed by an intensity contrast in the electronic signal. (d) Mechanisms for tuning the OPO through the pump wavelength (top) and through the thermo-optic shift of the signal resona… view at source ↗
Figure 2
Figure 2. Characterisation of nonlinear devices. (a) Experimen￾tal setup. See the text for more details. (b) Measured Bragg grating stop band (blue curve) and second-harmonic signal measured from a calibration waveguide in the same poling region. (c) Measured and simulated dispersion. FPC fiber po￾larization controller, PD photodiode, OSA optical spectrum analyzer, WDM wavelength demultiplexer. connected to the BR and measuri… view at source ↗
Figure 3
Figure 3. Tuning between degenerate and non-degenerate optical parametric oscillation through pump wavelength tuning. (a) Mea￾sured signal and idler wavelength as a function of pump detuning. The tuning curve has a periodicity of one quarter of the res￾onator free spectral range and presents the degenerate mode once every two periods. (b) Simulation of the tuning behaviour based on energy conservation for the measured dispers… view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Tuning between degenerate and non-degenerate optical parametric oscillation through sweeping the power applied to the thermo-optic electrode. (a) OPO spectra as a function of power applied to the electrode. (b) Resonance plot for differ￾ent thermo-optic powers and reso…
Figure 5
Figure 5. Figure 5: Threshold study for the OPO in the degenerate mode. (a) Pump wavelength sweeps for increasing pump power. The maximum power of each sweep is indicated in orange and the non-degenerate OPO excitations are plotted in light blue. (b) OPO output power as a function of pump…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

26 extracted references · 24 canonical work pages

  1. [14]

    Octave-spanning tunable infrared parametric oscillators in nanophotonics,

    L. Ledezma, A. Roy, L. Costa, et al., “Octave-spanning tunable infrared parametric oscillators in nanophotonics,” Sci. Adv. 9, eadf9711 (2023)

  2. [24]

    Visible-to-mid-ir tunable frequency comb in nanophotonics,

    A. Roy, L. Ledezma, L. Costa,et al., “Visible-to-mid-ir tunable frequency comb in nanophotonics,” Nat. Commun. 14, 6549 (2023)

  3. [1]

    The physics of optical computing,

    P . L. McMahon, “The physics of optical computing,” Nat. Rev. Phys. 5, 717–734 (2023)

  4. [2]

    Integrated photonic neu- romorphic computing: opportunities and challenges,

    N. Farmakidis, B. Dong, and H. Bhaskaran, “Integrated photonic neu- romorphic computing: opportunities and challenges,” Nat. Rev. Electr. Eng. pp. 1–16 (2024)

  5. [3]

    On the computational complexity of ising spin glass models,

    F . Barahona, “On the computational complexity of ising spin glass models,” J. Phys. A: Math. Gen. 15, 3241 (1982)

  6. [4]

    Coherent ising machine based on degenerate optical parametric oscillators,

    Z. Wang, A. Marandi, K. Wen, et al., “Coherent ising machine based on degenerate optical parametric oscillators,” Phys. Rev. A88, 063853 (2013)

  7. [5]

    Coherence properties of a doubly resonant monolithic optical parametric oscillator,

    C. D. Nabors, S. T. Y ang, T. Day, and R. L. Byer, “Coherence properties of a doubly resonant monolithic optical parametric oscillator,” J. Opt. Soc. Am. B 7, 815–820 (1990)

  8. [6]

    A fully programmable 100-spin coherent ising machine with all-to-all connections,

    P . L. McMahon, A. Marandi, Y . Haribara,et al., “A fully programmable 100-spin coherent ising machine with all-to-all connections,” Science 354, 614–617 (2016)

Show all 26 references
  1. [7]

    Spectral phase transitions in optical parametric oscillators,

    A. Roy, S. Jahani, C. Langrock, et al., “Spectral phase transitions in optical parametric oscillators,” Nat. communications 12, 835 (2021)

  2. [8]

    100,000-spin coherent ising machine,

    T. Honjo, T. Sonobe, K. Inaba, et al. , “100,000-spin coherent ising machine,” Sci. Adv. 7, eabh0952 (2021)

  3. [9]

    Network of time-multiplexed optical parametric oscillators as a coherent ising machine,

    M. Alireza, W. Zhe, T. Kenta,et al., “Network of time-multiplexed optical parametric oscillators as a coherent ising machine,” Nat. Photonics 8, 937–942 (2014)

  4. [10]

    Demonstration of chip- based coupled degenerate optical parametric oscillators for realizing a nanophotonic spin-glass,

    Y . Okawachi, M. Yu, J. K. Jang, et al. , “Demonstration of chip- based coupled degenerate optical parametric oscillators for realizing a nanophotonic spin-glass,” Nat. communications 11, 4119 (2020)

  5. [11]

    Large-scale time- multiplexed nanophotonic parametric oscillators,

    R. M. Gray1, R. Sekine1, L. Ledezma, et al. , “Large-scale time- multiplexed nanophotonic parametric oscillators,” arXiv preprint arXiv:2405.17355v1 (2024)

  6. [12]

    Ultra-low-power second-order nonlinear optics on a chip,

    T. P . McKenna, H. S. Stokowski, V. Ansari, et al. , “Ultra-low-power second-order nonlinear optics on a chip,” Nat. Commun. 13, 4532 (2022)

  7. [13]

    Ultralow-threshold thin-film lithium niobate optical parametric oscillator,

    J. Lu, A. A. Sayem, Z. Gong, et al., “Ultralow-threshold thin-film lithium niobate optical parametric oscillator,” Optica 8, 539–544 (2021)

  8. [15]

    Monolithically integrated femtosecond optical parametric oscillators,

    N. Jornod, M. Jankowski, L. M. Krüger, et al., “Monolithically integrated femtosecond optical parametric oscillators,” Optica10, 826–831 (2023)

  9. [16]

    Tunable ultranarrowband grating filters in thin-film lithium niobate,

    A. Prencipe, M. A. Baghban, and K. Gallo, “Tunable ultranarrowband grating filters in thin-film lithium niobate,” ACS Photonics 8, 2923–2930 (2021)

  10. [17]

    Extreme electro- optic tuning of bragg mirrors integrated in lithium niobate nanowaveg- uides,

    M. R. Escalé, D. Pohl, A. Sergeyev, and R. Grange, “Extreme electro- optic tuning of bragg mirrors integrated in lithium niobate nanowaveg- uides,” Opt. Lett. 43, 1515–1518 (2018)

  11. [18]

    100-gbd waveguide bragg grating modulator in thin-film lithium niobate,

    D. Pohl, A. Messner, F . Kaufmann,et al., “100-gbd waveguide bragg grating modulator in thin-film lithium niobate,” IEEE Photonics Technol. Lett. 33, 85–88 (2021)

  12. [19]

    Single-mode squeezed-light generation and tomography with an integrated optical parametric oscil- lator,

    T. Park, H. Stokowski, V. Ansari,et al., “Single-mode squeezed-light generation and tomography with an integrated optical parametric oscil- lator,” Sci. Adv. 10, eadl1814 (2024)

  13. [20]

    Integrated frequency- modulated optical parametric oscillator,

    H. S. Stokowski, D. J. Dean, A. Y . Hwang,et al., “Integrated frequency- modulated optical parametric oscillator,” Nature 627, 95–100 (2024)

  14. [21]

    Mid-infrared spectroscopy with a broadly tunable thin-film lithium niobate optical parametric oscil- lator,

    A. Y . Hwang, H. S. Stokowski, T. Park,et al., “Mid-infrared spectroscopy with a broadly tunable thin-film lithium niobate optical parametric oscil- lator,” arXiv preprint arXiv:2307.04199 (2023)

  15. [22]

    High-bandwidth thermo- optic phase shifters for lithium niobate-on-insulator photonic integrated circuits,

    A. Maeder, F . Kaufmann, D. Pohl, et al. , “High-bandwidth thermo- optic phase shifters for lithium niobate-on-insulator photonic integrated circuits,” Opt. Lett. 47, 4375–4378 (2022)

  16. [23]

    High-fidelity and polarization-insensitive universal photonic processors fabricated by femtosecond laser writing,

    C. Pentangelo, N. D. Giano, S. Piacentini, et al. , “High-fidelity and polarization-insensitive universal photonic processors fabricated by femtosecond laser writing,” Nanophotonics 13, 2259–2270 (2024)

  17. [25]

    Redeposition-free inductively-coupled plasma etching of lithium niobate for integrated photonics,

    F . Kaufmann, G. Finco, A. Maeder, and R. Grange, “Redeposition-free inductively-coupled plasma etching of lithium niobate for integrated photonics,” Nanophotonics 12, 1601–1611 (2023)

  18. [26]

    Wavelength-accurate and wafer-scale process for nonlinear frequency mixers in thin-film lithium niobate,

    C. J. Xin, S. Lu, J. Y ang,et al., “Wavelength-accurate and wafer-scale process for nonlinear frequency mixers in thin-film lithium niobate,” arXiv preprint arXiv:2404.12381 (2024)

Pith tools

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