Pith. sign in

REVIEW 3 major objections 5 minor 44 references

Integrated broadband optical isolator via dynamic rotating destructive interference

T0 review · 3 major / 5 minor · reviewed 2026-08-05 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read By driving four parallel optical channels with quarter-period-delayed RF waves, this paper demonstrates a traveling-wave optical isolator that continuously cancels backward light while leaving forward light unmodified, achieving roughly 30

desk verdict Clever new 4-channel DRDI isolator, real measured isolation, but the 30 nm bandwidth needs active heater retuning and the per-channel phase symmetry is asserted, not verified. read the letter →

arxiv 2509.02866 v1 pith:QVN2ZJXK submitted 2025-09-02 physics.optics

Kyunghun Han , Yiliang Bao , Junyeob Song , David Long , Sean Bresler , Daron Westly , Jason Gorman , Thomas LeBrun
show 2 more authors
Kartik Srinivasan Vladimir Aksyuk
This is my paper · ORCID
classification physics.optics PACS 42.25.Hz42.82.-m
keywords opticalisolatordynamicrotatingdestructiveinterferencetraveling-wavemodulationelectro-opticeffectthin-filmlithiumniobatesiliconnitridephotonicsbroadbandnon-reciprocityatomicspectroscopy
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 proposes and demonstrates an integrated optical isolator that needs no magnets and no resonant cavities. The trick is to phase-modulate four parallel waveguides with RF waves that are delayed by one quarter period relative to each other, so that at every moment two pairs of channels are exactly out of phase and cancel the backward-propagating light, while the forward light accumulates zero net phase and passes through unchanged. The authors measure about 30 dB isolation at 789.7 nm, more than 24 dB across 770–800 nm with active heater tuning, and simultaneous >20 dB isolation for two lasers within a 10 nm window, with no detectable optical sidebands or added forward loss. If correct, this removes the historical trade-off between isolation strength and bandwidth for on-chip isolators, and it opens a practical path to protecting integrated lasers across wavelength bands without exotic materials.

What carries the argument

The central object is the dynamic rotating destructive interference (DRDI) phasor pattern: four phase-modulated channels whose accumulated phases are arranged so that at any instant two channel pairs differ by exactly π (opposite phasors), and the pairing rotates each quarter RF period. A periodic zero-mean waveform—expressed as a triangular wave in principle but implemented as an optimized finite set of odd harmonics—guarantees forward light sees zero net phase while backward light is continuously cancelled, without needing waveform discontinuities or external optical filters.

What would settle it

Sweep the RF drive frequency around 6.5 GHz while monitoring both forward and backward optical spectra with the heaters fixed at their optimized values: the DRDI claim predicts a sharp backward null only at frequencies satisfying the integer-period condition of Eq. (3) with forward transmission remaining flat, so observing a broad backward minimum, a forward ripple at the null frequency, or failure of the null to regenerate after heater re-optimization would indicate that the suppression is not produced by the rotating interference mechanism.

Watch

Extended reading notes

Core claim

The central claim is that dynamic rotating destructive interference (DRDI) provides a complete, broadband, magnet-free and resonance-free mechanism for on-chip optical isolation. In a four-channel Mach-Zehnder modulator, two push-pull pairs are driven by RF waveforms separated by a quarter period. Over one RF cycle the pairs that cancel each other rotate every quarter period, so the backward-propagating light always experiences destructive interference, while the forward-propagating light, chosen so that its interaction time is an integer number of RF periods, sees zero net phase modulation. The paper shows experimentally that a 6.5 GHz optimized odd-harmonic waveform with rounded peaks and

Load-bearing premise

The four optical channels must stay matched in amplitude and phase across the entire 15 mm modulation region under RF drive, and the six microheaters plus the pre-distorted RF waveform have to hold that balance tightly enough to maintain the rotating destructive interference for the backward direction.

Editorial extensions

If this is right

  • On-chip isolators no longer need to trade isolation against bandwidth: the scheme avoids both magneto-optic materials and resonances, so isolation is set by channel balancing and waveform fidelity, not by cavity linewidth.
  • A single isolator can simultaneously protect multiple lasers at different wavelengths, as demonstrated for two lasers detuned up to 10 nm with >20 dB isolation for both.
  • The principle is wavelength-scalable: redesigning the passive splitters and combiners should extend the same cancellation mechanism to the telecommunication C-band or other visible-to-infrared bands.
  • The absence of measurable forward loss and the absence of generated optical sidebands mean the isolator can be inserted in front of integrated lasers without degrading their output spectrum or efficiency.
  • Integration with alkali atomic transitions (rubidium and potassium) enables chip-scale protection for atomic clocks, laser cooling, and precision spectroscopy systems that currently rely on bulky free-space isolators.

Reading between the lines

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

  • If the directional couplers are replaced with broadband or adiabatic designs, the isolation band could plausibly extend far beyond the demonstrated 30 nm, since the rotating interference condition itself imposes no wavelength limit—the paper shows the measured roll-off is dominated by splitter dispersion.
  • Because the cancellation is dynamic rather than resonant, the device may also work as an actively controllable 'optical gate' or isolating switch by simply turning the RF drive on and off, with no need to re-bias optical elements.
  • The same DRDI principle could be ported to other modulation platforms, such as acousto-optic or plasma-dispersion devices, provided four equal-amplitude channels can be driven with the required quarter-period delays and odd-harmonic waveform shape.
  • In a chip-scale spectroscopy system, the linear drift of the optimized heater powers with wavelength (reported in the supplemental information) could be exploited to implement fast feed-forward wavelength sweeps without re-optimizing each point.
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 / 5 minor

Summary. The paper proposes and experimentally demonstrates an integrated, magnet-free, resonance-free optical isolator based on dynamic rotating destructive interference (DRDI) in a four-channel traveling-wave electro-optic phase modulator. Four SiN/TFLN hybrid waveguides are phase-modulated by two push-pull RF drives delayed by one quarter period, so that, in the ideal limit, backward-propagating light experiences continuous destructive interference while forward-propagating light sees zero net phase accumulation. The device is fabricated on a 350 nm SiN platform with flip-chip bonded thin-film lithium niobate, and characterized at 770–800 nm. The central experimental claims are: about 30 dB isolation at 789.7 nm, >24 dB isolation over a 30 nm band with active heater tuning, >20 dB isolation over a 10 nm window for two simultaneous lasers, and no measurable forward-transmission penalty under RF drive. The paper includes static extinction-ratio calibration that bounds the EO performance, RF output spectra showing the predistorted multi-harmonic waveform, optical spectra showing sideband-free backward cancellation, and a dual-laser demonstration.

Significance. If the claims hold, this is a notable advance: it demonstrates a broadband, resonance-free, magnet-free integrated isolator in the visible/near-IR, with isolation comparable to prior resonant or magneto-optic devices but without their bandwidth limitations. The experimental methodology is largely internally consistent: static thermo-optic extinction ratios (≥30 dB) correctly bound the achievable EO isolation, the forward transmission with RF overlaps the no-RF baseline, and the wavelength-dependent isolation follows the splitter/combiner dispersion expected from the device design. The dual-laser experiment is a useful and nontrivial addition, showing simultaneous suppression for two wavelengths. The main weakness is that the DRDI mechanism is not directly verified at the per-channel level: the paper relies on aggregate optical measurements and RF output spectra, leaving open the question of whether the exact quarter-period/push-pull phase relationship in the optical phase domain is realized.

major comments (3)
  1. [Section II, Fig. 3d] The central mechanism requires four optical channels with equal dynamic phase-modulation amplitudes and exact quarter-period-delayed, push-pull phase waveforms over the full 15 mm interaction length. The paper calibrates static power/phase with heaters (Extended Data Fig. 5) and matches the RF output spectrum (Fig. 3b), but does not report per-channel (or per-push-pull-pair) RF-to-optical phase transfer functions at 6.5/19.5 GHz. The measured 6.5 GHz operating frequency differs from the value predicted by Eq. (3) for L = 15 mm, and the authors attribute this to RF attenuation or bonding-induced phase-index changes (Extended Data Fig. 3), which means the interaction is not the ideal uniform, velocity-matched model. Static heaters cannot correct a dynamic amplitude/phase imbalance among channels. Please add direct per-channel phase-modulation characterization (e.g., optical sideband amplit
  2. [Fig. 3d and 'For a practical realization...' paragraph] The theoretical isolation-versus-N curves are computed with the same numerical optimizer that was used to design the RF waveform, so those curves are not independent predictions. The single measured point (N=3, black star) is compared to a curve that was optimized for exactly that N, making the agreement partly by construction. Please state this explicitly, and, if feasible, measure isolation for at least one other truncation (e.g., N=1 or N=5) to test whether the predicted trend is observed. Alternatively, compute the theoretical isolation using the actual measured RF output spectrum rather than the ideal optimized waveform, which would give a more meaningful comparison to the 30 dB experimental value.
  3. [Abstract, Section II, Extended Data Table 1] The abstract and introduction state 'over 24 dB isolation across a 30 nm bandwidth' and 'broadband optical isolator' without clearly qualifying that this requires active retuning of six microheaters at each wavelength. The fixed-heater result in Fig. 3e (upper) shows isolation varying by roughly 21 dB across the 30 nm span, with the >20 dB passive window being narrower (~10 nm, as reflected in Extended Data Table 1's 'without active tuning' footnote). Since the broadband claim is central, the distinction between actively tuned bandwidth and passive bandwidth must be stated in the abstract and in the main text near the claim, not only in a table footnote. Please revise the wording and the table header/notes accordingly.
minor comments (5)
  1. [Figure 3f] The statement 'no other sidebands exist in the spectra' would benefit from a stated OSA resolution bandwidth and measurement noise floor. Otherwise the reader cannot judge how small a sideband could have been hidden in the baseline.
  2. [Figure 3b] The RF power axis is 'normalized to the fundamental frequency power.' Please state the absolute measured RF power at the CPW output (or input) as well, so the 32 mW statement is traceable to the plotted spectrum.
  3. [Extended Data Table 1] The table footnote 'Bandwidth for isolation > 20 dB without active tuning' is useful but it appears only as a footnote. A short sentence in the main text, near Fig. 3e, would prevent readers from conflating the actively tuned 30 nm claim with the passive 10 nm result.
  4. [Supplementary Figure S1] The forward transmission frequency sweep (Fig. S1a) shows a strong dip at 0.5 GHz and saturation near 6.5 GHz, but the residual ripple above 2.5 GHz is not discussed. If this ripple reflects RF impedance or bonding nonuniformity, a brief note would help the reader understand the uncertainty in the 'zero net phase' condition.
  5. [General notation] The paper switches between 'Ch1' and 'Ch 1' and between 'H1' and 'Heater 1' inconsistently. A single consistent notation would improve readability.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the measured isolation is an experimental observable, and the theoretical curves are model-based design calculations rather than fitted outputs disguised as predictions.

full rationale

The paper's derivation chain is self-contained. The DRDI principle is introduced through explicit physical assumptions: Eqs. (1)-(2) give the accumulated phase from a distributed electro-optic interaction, and Eq. (3) gives the design frequency for zero forward net phase. The four-channel quarter-period-delayed push-pull scheme is then shown in the time domain to produce two π-opposed cancellation pairs at every instant. This is a stated mechanism, not a restatement of the measured isolation. The measured isolation (~30 dB, >24 dB over 30 nm, >20 dB over 10 nm) is a direct experimental observable obtained under an explicitly described optimization protocol for the six microheaters and the RF harmonic amplitudes/phases. Optimizing a device to maximize a figure of merit and then reporting the achieved value is not a circular prediction. The theoretical isolation-versus-harmonics curves in Fig. 3d are model calculations for waveforms generated by the same numerical optimizer used in the design; they are design estimates, not independent first-principles benchmarks, but they are not used to force the experimental result. No equation in the paper reduces to its own input by construction, and no fitted parameter is renamed as a prediction. Self-citations (e.g., refs. [1], [17]) appear only in background or application contexts and are not load-bearing for the central claim. The paper's own limitation comments about RF attenuation, velocity mismatch, and static-heater compensation further indicate that the experimental success is treated as a hardware achievement, not as a mathematical tautology. Therefore no circularity is present.

Assumptions & free parameters 3 free parameters · 6 assumptions · 0 invented entities

The paper introduces no new physical entities, forces, particles, or dimensions. The central claim rests on standard electro-optic traveling-wave physics plus a new four-channel interference scheme. The main free inputs are empirically tuned parameters: RF frequency, RF harmonic waveform, and six heater powers. The static extinction ratio and half-wave voltage are measured device properties, not free parameters in the physical model.

free parameters (3)
  • Base RF frequency f_RF = 6.5 GHz
    Chosen empirically from the measured forward transmission response, where net phase shift saturates above about 6.5 GHz (Fig. S1a). The theoretical value from Eq. (3) is lower, so the actual RF phase velocity is inferred from data rather than independently measured.
  • RF harmonic amplitudes and phases = Fundamental and third harmonic; about 32 mW at CPW output
    Optimized iteratively by monitoring the CPW output spectrum and minimizing backward optical transmission (Fig. S5). These are device-specific tuning values, not derived from first principles.
  • Microheater powers H1-H6 = Wavelength-dependent, roughly 0.12 to 0.28 W (Fig. S4)
    Tuned at each wavelength to balance channel powers and set static phase offsets. Active heater tuning is required to achieve more than 24 dB isolation across 30 nm.
assumptions (6)
  • domain assumption Time-periodic spatiotemporal modulation of the refractive index breaks time-reversal symmetry and Lorentz reciprocity (Eqs. 1-3, Section II).
    The entire non-reciprocity relies on this standard traveling-wave isolator physics; the paper cites it but does not re-derive the underlying electro-optic response.
  • domain assumption Electro-optic phase modulation is linear, instantaneous, and lossless in the hybrid SiN/TFLN waveguide (Methods, PIC mode management).
    The analysis assumes the phase shift is proportional to the applied RF voltage with negligible optical absorption and no nonlinearity; no nonlinearity measurements are presented.
  • standard math Half-wave symmetry f(t+T0/2) = -f(t) with quarter-period-delayed drives generates pairwise pi phase differences at all times (Supplementary Section C, Eqs. S1-S2).
    This is the mathematical construction of the DRDI scheme; it is a symmetry argument and is essentially self-contained.
  • standard math Forward propagation accumulates zero net phase when the total interaction time L/v_RF + L/v_opt is an integer multiple of the RF period (Eq. 3).
    The derivation assumes negligible RF attenuation, alpha approximately 0, which the paper acknowledges is not exactly true experimentally; the operating frequency is adjusted empirically.
  • domain assumption Static thermo-optic extinction ratio is an upper bound for electro-optic isolation (Section II, static extinction paragraph).
    The paper assumes both effects interfere through the same linear optical paths, so a static extinction ratio bounds the dynamic case. This is plausible but not proven over all RF frequencies and wavelengths.
  • domain assumption Directional couplers and combiners are reciprocal and become balanced after heater compensation (Section II, heater tuning protocol).
    The heaters compensate for fabrication asymmetries, but the wavelength dependence of the couplers is the main reason the passive bandwidth is limited to about 10 nm.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Integrated broadband optical isolator via dynamic rotating destructive interference." pith.science (2026). https://pith.science/paper/QVN2ZJXK

@misc{pith2026250902866,
  author       = {Pith},
  title        = {Pith review of: Integrated broadband optical isolator via dynamic rotating destructive interference},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/QVN2ZJXK}},
  note         = {Machine review of arXiv:2509.02866}
}
read the original abstract

Photonic integrated circuits route and shape light on a chip, but back-reflections feed back into coherent on-chip lasers, destabilizing operation and corrupting signals. Robust operation requires an integrated optical isolator that strongly suppresses backward propagation while maintaining low-loss, broadband forward transmission. However, prior on-chip isolators rely on magneto-optic materials or resonance-based filters, which respectively demand non-standard processes or inherently constrain bandwidth. Here, we propose and experimentally demonstrate a traveling-wave optical isolator without magnetic materials or resonant elements. By driving four parallel optical channels with periodic RF waves, we realize dynamic rotating destructive interference that continuously cancels backward-propagating light while leaving forward-propagating light unaffected. We achieve about 30 dB isolation at a wavelength of 789.7 nm and maintain over 24 dB isolation across an approximately 30 nm bandwidth (770 nm to 800 nm), including >20 dB isolation for two simultaneous lasers within an approximately 10 nm wavelength window. This wavelength span covers key alkali atomic transitions, enabling strong suppression of feedback-induced frequency noise and laser instability in atomic spectroscopy, laser cooling, and locking applications. We demonstrate a practical, broadband on-chip isolator applicable from the visible to the near-infrared, which is a crucial step toward fully integrated photonic platforms.

Figures

Figures reproduced from arXiv: 2509.02866 by the authors.

Figure 1
Figure 1. Integrated broadband traveling-wave optical isolator. [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Dynamic rotating destructive interference with 4-channel phase modulation. [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Broadband optical isolation demonstration with an optimized multi-harmonic RF waveform. [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Double-laser experiment for broadband isolation performance. a. Experimental setup shows two external-cavity [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

44 extracted references · 44 canonical work pages

  1. [1]

    A., Stone, J

    Long, D. A., Stone, J. R., Sun, Y., Westly, D. & Srinivasan, K. Sub-Doppler spectroscopy of quantum systems through nanophotonic spectral translation of electro-optic light. Nature Photonics 18, 1285–1292 (2024)

    work page 2024

  2. [2]

    Isichenko, A. et al. Photonic integrated beam delivery for a rubidium 3D magneto-optical trap.Nature Com- munications 14, 3080 (2023)

    work page 2023

  3. [3]

    Martin, K. W. et al. Compact Optical Atomic Clock Based on a Two-Photon Transition in Rubidium.Phys- ical Review Applied 9, 014019 (2018)

    work page 2018

  4. [4]

    Yao, C. et al. Integrated reconstructive spectrometer with programmable photonic circuits.Nature Commu- nications 14, 1–10 (2023)

    work page 2023

  5. [5]

    Optical atomic clock interrogation using an integrated spiral cavity laser.Nature Photonics 19, 277–283 (2025)

    Loh, W.et al. Optical atomic clock interrogation using an integrated spiral cavity laser.Nature Photonics 19, 277–283 (2025)

    work page 2025

  6. [6]

    Rizzo, A. et al. Massively scalable Kerr comb-driven silicon photonic link. Nature Photonics 17, 781–790 (2023)

    work page 2023

  7. [7]

    Ultrafast tunable lasers using lithium niobate integrated photonics

    Snigirev, V.et al. Ultrafast tunable lasers using lithium niobate integrated photonics. Nature 615, 411–417 (2023)

    work page 2023

  8. [8]

    Siddharth, A. et al. Ultrafast tunable photonic- integrated extended-DBR Pockels laser. Nature Pho- tonics 19, 709–717 (2025)

    work page 2025

Show all 44 references
  1. [9]

    Jalas, D. et al. What is-and what is not-an optical isolator. Nature Photonics 7, 579–582 (2013)

    2013

  2. [10]

    & Fan, S

    Shi, Y., Yu, Z. & Fan, S. Limitations of nonlinear op- tical isolators due to dynamic reciprocity.Nature Pho- tonics 9, 388–392 (2015)

    2015

  3. [11]

    Huang, D. et al. Integrated broadband Ce:YIG/Si Mach–Zehnder optical isolators with over 100 nm tun- ing range. Optics Letters 42, 4901 (2017)

    2017

  4. [12]

    & Mizumoto, T

    Yamaguchi, R., Shoji, Y. & Mizumoto, T. Low-loss waveguide optical isolator with tapered mode converter and magneto-optical phase shifter for TE mode input. Optics Express 26, 21271 (2018)

    2018

  5. [13]

    Microring-Based Optical Isolator and Circulator with Integrated Electromagnet for Silicon Photonics

    Pintus, P.et al. Microring-Based Optical Isolator and Circulator with Integrated Electromagnet for Silicon Photonics. Journal of Lightwave Technology 35, 1429– 1437 (2017)

    2017

  6. [14]

    Zhang, Y. et al. Monolithic integration of broadband optical isolators for polarization-diverse silicon photon- ics. Optica 6, 473 (2019)

    2019

  7. [15]

    Bi, L. et al. On-chip optical isolation in monolithi- cally integrated non-reciprocal optical resonators.Na- ture Photonics 5, 758–762 (2011)

    2011

  8. [16]

    Ultra-broadband magneto-optical isola- tors and circulators on a silicon nitride photonics plat- form

    Yan, W.et al. Ultra-broadband magneto-optical isola- tors and circulators on a silicon nitride photonics plat- form. Optica 11, 376 (2024)

    2024

  9. [17]

    & Stadler, B

    Srinivasan, K. & Stadler, B. J. H. Review of integrated magneto-optical isolators with rare-earth iron garnets 10 for polarization diverse and magnet-free isolation in silicon photonics. Optical Materials Express 12, 697 (2022)

    2022

  10. [18]

    & Lipson, M

    Lira, H., Yu, Z., Fan, S. & Lipson, M. Electrically driven nonreciprocity induced by interband photonic transition on a silicon chip. Physical Review Letters 109, 033901 (2012)

    2012

  11. [19]

    D., Fang, K., Nussenzveig, P., Fan, S

    Tzuang, L. D., Fang, K., Nussenzveig, P., Fan, S. & Lipson, M. Non-reciprocal phase shift induced by an effective magnetic flux for light. Nature Photonics 8, 701–705 (2014)

    2014

  12. [20]

    & Fan, S

    Fang, K., Yu, Z. & Fan, S. Photonic Aharonov-Bohm effect based on dynamic modulation. Physical Review Letters 108, 153901 (2012)

    2012

  13. [21]

    Novel nonmagnetic 30-dB traveling- wavesingle-sidebandopticalisolatorintegratedinIII/V material

    Bhandare, S.et al. Novel nonmagnetic 30-dB traveling- wavesingle-sidebandopticalisolatorintegratedinIII/V material. IEEE Journal on Selected Topics in Quantum Electronics 11, 417–421 (2005)

    2005

  14. [22]

    R., Dupuis, N

    Doerr, C. R., Dupuis, N. & Zhang, L. Optical isolator using two tandem phase modulators.Optics Letters 36, 4293–4295 (2011)

    2011

  15. [23]

    & Popović, M

    Dostart, N., Gevorgyan, H., Onural, D. & Popović, M. A. Optical isolation using microring modulators. Optics Letters 46, 460 (2021)

    2021

  16. [24]

    R., Chen, L

    Doerr, C. R., Chen, L. & Vermeulen, D. Silicon pho- tonicsbroadbandmodulation-basedisolator. Optics Ex- press 22, 4493 (2014)

    2014

  17. [25]

    Gao, L. et al. Thin-film lithium niobate electro-optic isolator fabricated by photolithography assisted chemo- mechanical etching. Optics Letters 49, 614 (2024)

    2024

  18. [26]

    Travelling-wave Mach-Zehnder modulators functioning as optical isolators

    Dong, P. Travelling-wave Mach-Zehnder modulators functioning as optical isolators. Optics Express 23, 10498 (2015)

    2015

  19. [27]

    & Fan, L

    Shah, M., Briggs, I., Chen, P.-K., Hou, S. & Fan, L. Visible-telecom tunable dual-band optical isolator based on dynamic modulation in thin-film lithium nio- bate. Optics Letters 48, 1978 (2023)

    1978

  20. [28]

    Kittlaus, E. A.et al. Electrically driven acousto-optics and broadband non-reciprocity in silicon photonics.Na- ture Photonics 15, 43–52 (2021)

    2021

  21. [29]

    B., Örsel, O

    Sohn, D. B., Örsel, O. E. & Bahl, G. Electrically driven optical isolation through phonon-mediated pho- tonic Autler–Townes splitting. Nature Photonics 15, 822–827 (2021)

    2021

  22. [30]

    Tian, H. et al. Magnetic-free silicon nitride integrated optical isolator. Nature Photonics 15, 828–836 (2021)

    2021

  23. [31]

    A., Otterstrom, N

    Kittlaus, E. A., Otterstrom, N. T., Kharel, P., Gertler, S. & Rakich, P. T. Non-reciprocal interband Brillouin modulation. Nature Photonics 12, 613–619 (2018)

    2018

  24. [32]

    B., Kim, S

    Sohn, D. B., Kim, S. & Bahl, G. Time-reversal sym- metry breaking with acoustic pumping of nanophotonic circuits. Nature Photonics 12, 91–97 (2018)

    2018

  25. [33]

    Integrated electro-optic isolator on thin- film lithium niobate

    Yu, M.et al. Integrated electro-optic isolator on thin- film lithium niobate. Nature Photonics 17, 666–671 (2023)

    2023

  26. [34]

    Jin, M., Chen, J.-Y., Sua, Y. M. & Huang, Y.-P. High- extinction electro-optic modulation on lithium niobate thin film. Optics Letters 44, 1265 (2019)

    2019

  27. [35]

    Atomic spectroscopy on a chip.Nature Photonics 1, 331–335 (2007)

    Yang, W.et al. Atomic spectroscopy on a chip.Nature Photonics 1, 331–335 (2007)

    2007

  28. [36]

    Schwindt, P. D.et al. Chip-scale atomic magnetometer. Applied Physics Letters 85, 6409–6411 (2004)

    2004

  29. [37]

    Hummon, M. T.et al. Photonic chip for laser stabiliza- tion to an atomic vapor with 10 -11 instability.Optica 5, 443 (2018)

    2018

  30. [38]

    Parabolic MMI Coupler for 2× 2 Silicon Optical Switch With Robustly High Extinction Ratio for Four Paths.IEEE Photonics Technology Letters 35, 737–740 (2023)

    Jin, M.et al. Parabolic MMI Coupler for 2× 2 Silicon Optical Switch With Robustly High Extinction Ratio for Four Paths.IEEE Photonics Technology Letters 35, 737–740 (2023)

    2023

  31. [39]

    & Dagenais, M

    Xie, S., Veilleux, S. & Dagenais, M. On-Chip High Ex- tinction Ratio Single-Stage Mach-Zehnder Interferome- ter Based on Multimode Interferometer.IEEE Photon- ics Journal 14, 2237906 (2022)

    2022

  32. [40]

    & Jaeger, N

    Yun, H., Chrostowski, L. & Jaeger, N. A. F. Ultra-broadband 2 × 2 adiabatic 3 dB coupler us- ing subwavelength-grating-assisted silicon-on-insulator strip waveguides. Optics Letters 43, 1935 (2018)

    1935

  33. [41]

    Fan, L. et al. An All-Silicon Passive Optical Diode. Science 335, 447–450 (2012)

    2012

  34. [42]

    Abdelsalam, K., Li, T., Khurgin, J. B. & Fathpour, S. Linear isolators using wavelength conversion.Optica 7, 209 (2020)

    2020

  35. [43]

    White, A. D.et al. Integrated passive nonlinear optical isolators. Nature Photonics 17, 143–149 (2023)

    2023

  36. [44]

    White, A. D.et al. Unified laser stabilization and isola- tion on a silicon chip.Nature Photonics 18, 1305–1311 (2024). 11 METHODS Device fabrication The device was fabricated on a 100 mm diameter sil- icon wafer with a 3 µm wet thermal oxide layer and 350 nm stoichiometric si...

    2024

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.