REVIEW 3 major objections 4 minor 36 references
On-chip Quantum Measurement of Squeezing Generated from a Silicon Nitride Micro-ring Resonator
T0 review · 3 major / 4 minor · reviewed 2026-08-15 · deepseek-v4-flash
Pith's one-line read By placing a second silicon nitride micro-ring amplifier before the lossy chip-to-fiber interface, this paper demonstrates on-chip measurement of 4.6 dB of two-mode squeezing even when downstream losses exceed 7 dB.
desk verdict Nice device concept, but the 4.6 dB claim fails a reciprocity check, so the loss-tolerance result is not established as written. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The load-bearing element is the second ring operated as a phase-sensitive, high-gain non-degenerate optical parametric amplifier whose pump acts as the equivalent local oscillator. In the high-gain limit, the quadrature output follows $\hat{X}_s^{\rm out}(\theta)\approx G\hat{X}_+(\theta,\phi_p)$, where $\hat{X}_+$ is the sum of the signal quadrature and the corresponding idler quadrature; comparing squeezed-input to vacuum-input noise gives $R=\langle\Delta^2\hat{X}_+\rangle_{\rm sq}/\langle\Delta^2\hat{X}_+\rangle_{\rm vac}$. Because the amplified noise is about $2G^2R$ before any loss, the vacuum entering through loss $L$ is negligible, and the detected ratio $R_{\rm det}\approx R$. This noiseless amplification before detection is what carries the loss tolerance; the cascaded filters and heaters merely prepare matched operation of the two rings.
What would settle it
Turn on R1 below threshold while keeping Pump 2 fixed and monitor the R2 gain at 100 MHz with a weak seed; if the R2 gain or its phase response shifts by more than the quoted uncertainties when Pump 1 is switched on, the 4.6 dB value does not isolate R1's squeezing and the reference must be remeasured under identical on-chip conditions. A simpler check is to record the R2-alone noise trace before and after a period of R1 pumping and see whether the 10 dB-above-shot-noise level remains unchanged.
Extended reading notes
Core claim
The central claim is that a high-gain optical parametric amplifier can perform the quantum measurement of squeezing on-chip, making the result insensitive to downstream loss. For two-mode squeezed light generated by ring R1, ring R2 acts as a non-degenerate amplifier with gain $G$; in the large-gain limit the detected noise reduction $R_{\rm det}$ becomes equal to the input squeezing $R$ and independent of the loss $L$ (Eq. 6). Using this architecture, the paper reports $4.6\pm 0.4$ dB of measured squeezing and $13.2\pm 0.4$ dB of anti-squeezing at 100 MHz, with only about 70% escape efficiency in the squeezer, and estimates a 5 dB quantum SNR enhancement of the resulting SU(1,1) interferometer. The experiment also shows the measured noise level of R2 alone sits about 10 dB above shot noise, consistent with roughly 17 dB on-chip gain, confirming the amplifier operates in the required high-gain regime.
Load-bearing premise
The measurement is only as good as the assumption that the noise of the second ring with Pump 1 off is the right empty-input reference for the output of the first ring, meaning that switching on the first ring does not change the second ring's gain or phase through thermal or refractive crosstalk.
Editorial extensions
If this is right
- On-chip squeezing can be characterized without monolithically integrated detectors, because the measurement happens in the same nonlinear platform as the source.
- Chip-to-fiber coupling loss of 5 dB or more no longer caps the observed squeezing; the same device should preserve a fixed $R_{\rm det}$ as long as the amplifier gain stays high.
- The cascaded-ring device is a functional SU(1,1) interferometer on a chip, so phase sensing with quantum-enhanced SNR could be done in a compact CMOS-compatible circuit.
- The 4.6 dB level is bounded by the squeezer's roughly 70% escape efficiency, so improving the ring coupling ratio should directly raise the measured squeezing toward the theoretical limit.
- The measurement establishes a way to verify non-classical correlations at the point of generation, which is relevant for chip-based quantum sensors and continuous-variable quantum computing.
Reading between the lines
- A natural extension the paper does not explore is using the same amplifier-assisted readout to characterize other two-mode states, such as EPR-entangled or frequency-bin entangled light, on the same chip; the derivation only assumes a two-mode input and high gain.
- A testable prediction is that as Pump 2 gain $G$ is increased, $R_{\rm det}$ should stay approximately constant while the absolute noise floor rises by $20\log_{10}G$; measuring that plateau would confirm the loss-tolerance mechanism.
- The 5 dB SNR estimate relies on comparing signal gain with output noise, so a chip with on-chip phase modulation and a linear interferometer reference would turn this estimate into a direct measurement, which the authors say is planned.
- Actively locking Pump 2's phase and the ring temperatures would likely reduce the $\pm 0.4$ dB uncertainty in the reported squeezing and improve the stability of the monolithic interferometer.
Signed reviews
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper proposes and implements an amplifier-assisted on-chip quantum measurement scheme using two cascaded silicon nitride micro-ring resonators. A first ring (R1) generates a two-mode squeezed state via four-wave mixing, and a second ring (R2) acts as a high-gain parametric amplifier that measures the squeezing before off-chip loss. The authors report 4.6 dB of detected squeezing with an estimated 7 dB downstream loss, and claim the first monolithic SU(1,1) interferometer with an estimated 5 dB signal-to-noise enhancement. The theoretical basis (Eqs. 1-6) is standard, and the device fabrication and linear characterization are described in detail.
Significance. If the 4.6 dB squeezing claim is correct, the work demonstrates a practical route to loss-resilient quantum measurement in integrated photonics, which would be a valuable advance. The paper includes a clean derivation of the loss-tolerant detection principle, a plausible device architecture, and a direct measurement of the noise reduction. However, the central experimental claim currently rests on an unverified assumption about the stability of the amplifier gain, and the data themselves show an internal inconsistency that must be resolved before the claim can be accepted.
major comments (3)
- [Quantum noise measurement and observation of squeezing, Fig. 6(B), Eq. (6)] Equation (6) assumes that the gain G of R2 is identical in the squeezed-input measurement (both pumps on) and in the vacuum-reference measurement (Pump 1 off). The reported data are inconsistent with this assumption. For a pure two-mode squeezed state measured with the same amplifier gain, the maxima and minima of the phase-dependent noise in dB must sum to 0 dB relative to the same vacuum reference. The reported values of +13.2 dB and -4.6 dB sum to +8.6 dB. This common-mode offset implies that the reference level (the orange trace in Fig. 6(B)) does not correspond to the same amplifier response as the red trace, or that an additional phase-insensitive noise contribution is present. If the offset is interpreted as a gain increase of R2 when Pump 1 is turned on, the inferred squeezing from R1 would be -13.2 dB, exceeding the paper's own escape-efficiency bound of -5.2 dB; if interpreted as added noise, it would raise the minimum above the observed value. The paper provides no measurement of the R2 gain stability between the two configurations, despite the acknowledged sensitivity of the gain near the OPO threshold. The authors must provide a direct calibration of the R2 gain (or an equivalent reference with both pumps on) to validate Eq. (6).
- [Discussion, second paragraph] The paper attributes the discrepancy between the observed -4.6 dB and the -5.2 dB escape-efficiency upper bound to coupling loss between the two rings and detuning in the squeezer ring. This explanation is not consistent with the observed anti-squeezing of +13.2 dB. Loss between the rings would reduce both the squeezing and the anti-squeezing relative to the ideal values, but the measured anti-squeezing is 8.6 dB higher than the reciprocal of the measured squeezing. Therefore, the stated loss mechanism cannot account for the data. The manuscript needs to explicitly address the origin of the common-mode offset and to show that it does not affect the squeezing estimate.
- [On-chip SU(1,1) Interferometer with Injected Seed] The claimed 5 dB SNR enhancement of the SU(1,1) interferometer is an indirect estimate that depends on the same unverified parameters. The estimate uses an R2 classical gain of 17 dB and an output noise level '5 dB above shot noise' from Fig. 6(B). If the R2 gain changes when Pump 1 is on, or if the noise reference is invalid, both the phase-signal gain (10 dB) and the noise penalty (5 dB) are incorrect. The paper also acknowledges that the quantum enhancement is not directly measured. To support this secondary claim, the authors should either measure the SNR enhancement directly or clearly state that the value is conditional on the validity of the calibration.
minor comments (4)
- [Gain characterization of OPAs] The text states that the R2 working pump power is 16.3 mW, equal to the measured OPO threshold of 16.3 mW, while also claiming the pumps remain below threshold; clarify how detuning changes the threshold and whether the working point is indeed below the detuned threshold.
- [Quantum noise measurement and observation of squeezing, Fig. 6(B)] The origin of the ±0.4 dB uncertainties is not described; specify whether these are standard deviations over repeated phase scans, and the averaging procedure.
- [Principle of loss-tolerant, amplifier-assisted measurement, Eq. (5)] Equation (5) introduces L as the overall loss but does not explicitly define it as an intensity loss between 0 and 1; please define the variable and state its relationship to the dB losses quoted in the text.
- [Abstract and Introduction] The phrase 'the first monolithic SU(1,1) interferometer' should be supported with a comparison to prior integrated implementations, or softened to avoid a potentially contested novelty claim.
Circularity Check
No significant circularity: the 4.6 dB squeezing claim follows from measured noise levels and independent vacuum references, not from fitted inputs or self-citational load-bearing premises.
full rationale
The paper's central derivation (Eqs. (4)-(6)) is self-contained. It defines R as the ratio of the phase-sensitive amplifier output noise for squeezed input to that for vacuum input, and shows that this ratio is preserved after downstream loss because the same large gain G appears in both numerator and denominator. The experimental implementation measures the denominator directly from the R2-only trace (Pump 1 off) and the numerator from the dual-pump trace, so the ratio is a measurement rather than a fit. The claim that the R2-only trace provides the vacuum reference is a legitimate calibration step: with Pump 1 off, the first ring outputs vacuum, so the second ring amplifies vacuum noise. The gain-matching assumption between the Pump1-off and dual-pump configurations is an experimental validity condition, not a definitional circularity; if violated, the data would be unreliable, but the derivation itself does not presuppose the result. The estimated 5 dB SNR enhancement also uses measured gain and noise values (17 dB classical gain minus 7 dB loss, and about 5 dB output noise above shot noise) rather than fitted parameters, and the escape-efficiency bound of -5.2 dB is obtained from independently measured transmission spectra. Self-citations to prior SU(1,1) work provide background and estimation formulas but are not load-bearing for the observed 4.6 dB noise reduction. The skeptic's internal-consistency concern about a possible common-mode gain shift is a legitimate experimental critique, but it is not a circularity of the derivation chain.
Assumptions & free parameters
assumptions (5)
- domain assumption Non-degenerate OPA input-output relation, a_out = G a_in + g e^(2i phi_p) a_in^dagger, with G^2 - g^2 = 1 (Eq. 1).
- domain assumption In the high-gain limit G >> 1, the output quadrature is approximately G times the correlated combination X_s + X_i, so lambda -> 1 (Eq. 2).
- domain assumption The two-mode squeezed state from the first ring has reduced variance in the summed quadrature X+ compared to vacuum (Eqs. 3-4).
- domain assumption Loss before an ideal detector is modeled as (1-L) variance plus L vacuum contributions (Eq. 5).
- domain assumption The LO at 1542 nm and detection of only the signal field suffices because the idler correlation is mapped into the signal quadrature by the non-degenerate amplifier (Eq. 2).
Cite this review
Pith. "Pith review of On-chip Quantum Measurement of Squeezing Generated from a Silicon Nitride Micro-ring Resonator." pith.science (2026). https://pith.science/paper/UIC2FWRS
@misc{pith2026260803402,
author = {Pith},
title = {Pith review of: On-chip Quantum Measurement of Squeezing Generated from a Silicon Nitride Micro-ring Resonator},
year = {2026},
howpublished = {\url{https://pith.science/paper/UIC2FWRS}},
note = {Machine review of arXiv:2608.03402}
}
read the original abstract
Integration of quantum optical technique on-chip is crucial for large scale applications of quantum technology, which were proven in a free space environment to be superior to the corresponding classical technology. Squeezed states of light can be used for enhancing the sensitivity of quantum sensors and for fault-tolerant quantum computing. Although chip-based squeezed light generation has advanced significantly, practical impact remains limited because coupling losses between the chip and off-chip detectors destroy delicate quantum correlations, restricting the amount of observed squeezing. Here, we overcome this limitation by implementing the idea of on-chip quantum measurement with the aid of a parametric amplifier and applying it to the squeezed state generated by a silicon nitride (SiN) microring resonator. In our scheme, two matched SiN micro-rings are sequentially constructed. The first ring generates a squeezed state, whereas the second ring acts as a high-gain parametric amplifier (PA) that measures the squeezed state before the light experiences significant off-chip loss. This architecture is inherently loss-tolerant: the amplifier elevates the quantum noise well above the vacuum level, making the measurement insensitive to downstream losses. We directly observe a quantum noise reduction of 4.6 dB from the first ring, despite a chip-to-fiber coupling loss exceeding 5 dB. This work also demonstrates the first monolithic SU(1,1) interferometer with an estimated 5 dB signal-to-noise enhancement compared to traditional linear interferometers, and thus establishes a practical pathway for chip-based quantum sensors.
Figures
Reference graph
Works this paper leans on
-
[1]
Dutt,et al., On-chip optical squeezing.Physical Review Applied3(4), 044005 (2015)
A. Dutt,et al., On-chip optical squeezing.Physical Review Applied3(4), 044005 (2015)
work page 2015
-
[2]
P.-K. Chen, I. Briggs, S. Hou, L. Fan, Ultra-broadband quadrature squeezing with thin-film lithium niobate nanophotonics.Optics Letters47(6), 1506–1509 (2022)
work page 2022
-
[3]
Y. Zhao,et al., Near-degenerate quadrature-squeezed vacuum generation on a silicon-nitride chip.Physical Review Letters124(19), 193601 (2020)
work page 2020
-
[4]
V. D. Vaidya,et al., Broadband quadrature-squeezed vacuum and nonclassical photon number correlations from a nanophotonic device.Science advances6(39), eaba9186 (2020)
work page 2020
-
[5]
Zhang,et al., Squeezed light from a nanophotonic molecule.Nature communications12(1), 2233 (2021)
Y. Zhang,et al., Squeezed light from a nanophotonic molecule.Nature communications12(1), 2233 (2021)
work page 2021
-
[6]
Yang,et al., A squeezed quantum microcomb on a chip.Nature Communications12(1), 4781 (2021)
Z. Yang,et al., A squeezed quantum microcomb on a chip.Nature Communications12(1), 4781 (2021)
work page 2021
-
[7]
Liu,et al., Wafer-scale squeezed-light chips.arXiv preprint arXiv:2509.10445(2025)
S. Liu,et al., Wafer-scale squeezed-light chips.arXiv preprint arXiv:2509.10445(2025)
arXiv 2025
-
[8]
F. Najafi,et al., On-chip detection of non-classical light by scalable integration of single-photon detectors.Nature communications6(1), 5873 (2015)
work page 2015
Show all 36 references
-
[9]
Shaked,et al., Lifting the bandwidth limit of optical homodyne measurement with broadband parametric amplification.Nature communications9(1), 609 (2018)
Y. Shaked,et al., Lifting the bandwidth limit of optical homodyne measurement with broadband parametric amplification.Nature communications9(1), 609 (2018)
2018
-
[10]
Li,et al., Pulsed entanglement measured by parametric amplifier assisted homodyne detec- tion.Optics Express27(21), 30552–30562 (2019)
J. Li,et al., Pulsed entanglement measured by parametric amplifier assisted homodyne detec- tion.Optics Express27(21), 30552–30562 (2019)
2019
-
[11]
Li,et al., Measuring continuous-variable quantum entanglement with parametric-amplifier- assisted homodyne detection.Physical Review A101(5), 053801 (2020)
J. Li,et al., Measuring continuous-variable quantum entanglement with parametric-amplifier- assisted homodyne detection.Physical Review A101(5), 053801 (2020)
2020
-
[12]
T. Kashiwazaki,et al., Fabrication of low-loss quasi-single-mode PPLN waveguide and its application to a modularized broadband high-level squeezer.Applied Physics Letters119(25), 251104 (2021). 17
2021
-
[13]
Nehra,et al., Few-cycle vacuum squeezing in nanophotonics.Science377(6612), 1333– 1337 (2022)
R. Nehra,et al., Few-cycle vacuum squeezing in nanophotonics.Science377(6612), 1333– 1337 (2022)
2022
-
[14]
Yurke, S
B. Yurke, S. L. McCall, J. R. Klauder, SU(2) and SU(1,1) interferometers.Physical Review A 33(6), 4033 (1986)
1986
-
[15]
J. Jing, C. Liu, Z. Zhou, Z. Ou, W. Zhang, Realization of a nonlinear interferometer with parametric amplifiers.Applied Physics Letters99(1) (2011)
2011
-
[16]
Hudelist,et al., Quantum metrology with parametric amplifier-based photon correlation interferometers.Nature communications5(1), 3049 (2014)
F. Hudelist,et al., Quantum metrology with parametric amplifier-based photon correlation interferometers.Nature communications5(1), 3049 (2014)
2014
-
[17]
Z. Y. Ou, X. Li, Quantum SU (1, 1) interferometers: Basic principles and applications.APL Photonics5(8) (2020)
2020
-
[18]
Z. Y. Ou, Enhancement of the phase-measurement sensitivity beyond the standard quantum limit by a nonlinear interferometer.Physical Review A85(2), 023815 (2012)
2012
-
[19]
Manceau, G
M. Manceau, G. Leuchs, F. Khalili, M. Chekhova, Detection loss tolerant supersensitive phase measurement with an SU (1, 1) interferometer.Physical Review Letters119(22), 223604 (2017)
2017
-
[20]
C. M. Caves, Quantum-mechanical noise in an interferometer.Physical Review D23(8), 1693–1708 (1981)
1981
-
[21]
Mallet,et al., Quantum State Tomography of an Itinerant Squeezed Microwave Field
F. Mallet,et al., Quantum State Tomography of an Itinerant Squeezed Microwave Field. Physical Review Letters106(22), 220502 (2011)
2011
-
[22]
Flurin, N
E. Flurin, N. Roch, F. Mallet, M. H. Devoret, B. Huard, Generating Entangled Microwave Radiation Over Two Transmission Lines.Physical Review Letters109(18), 183901 (2012)
2012
-
[23]
Frascella, S
G. Frascella, S. Agne, F. Y. Khalili, M. V. Chekhova, Overcoming Detection Loss and Noise in Squeezing-Based Optical Sensing.npj Quantum Information7(1), 72 (2021)
2021
-
[24]
Cui,et al., Compact microring resonator based on ultralow-loss multimode silicon nitride waveguide.Advanced Photonics Nexus2(4), 046007–046007 (2023)
S. Cui,et al., Compact microring resonator based on ultralow-loss multimode silicon nitride waveguide.Advanced Photonics Nexus2(4), 046007–046007 (2023). 18
2023
-
[25]
D. J. Moss, R. Morandotti, A. L. Gaeta, M. Lipson, New CMOS-compatible platforms based on silicon nitride and Hydex for nonlinear optics.Nature Photonics7(8), 597–607 (2013)
2013
-
[26]
D. J. Blumenthal, R. G. Heideman, D. H. Geuzebroek, A. Leinse, C. G. H. Roeloffzen, Silicon Nitride in Silicon Photonics.Proceedings of the IEEE106(12), 2209–2231 (2018)
2018
-
[27]
A. L. Gaeta, M. Lipson, T. J. Kippenberg, Photonic-chip-based frequency combs.nature photonics13(3), 158–169 (2019)
2019
-
[28]
J. S. Levy,et al., CMOS-compatible multiple-wavelength oscillator for on-chip optical inter- connects.Nature Photonics4(1), 37–40 (2010)
2010
-
[29]
Okawachi,et al., Octave-spanning frequency comb generation in a silicon nitride chip
Y. Okawachi,et al., Octave-spanning frequency comb generation in a silicon nitride chip. Optics Letters36(17), 3398–3400 (2011)
2011
-
[30]
Kues,et al., Quantum optical microcombs.Nature Photonics13(3), 170–179 (2019)
M. Kues,et al., Quantum optical microcombs.Nature Photonics13(3), 170–179 (2019)
2019
-
[31]
Samara,et al., High-rate photon pairs and sequential Time-Bin entanglement with Si 3N4 microring resonators.Optics Express27(14), 19309–19318 (2019)
F. Samara,et al., High-rate photon pairs and sequential Time-Bin entanglement with Si 3N4 microring resonators.Optics Express27(14), 19309–19318 (2019)
2019
-
[32]
Imany,et al., 50-GHz-spaced comb of high-dimensional frequency-bin entangled photons from an on-chip silicon nitride microresonator.Optics Express26(2), 1825–1840 (2018)
P. Imany,et al., 50-GHz-spaced comb of high-dimensional frequency-bin entangled photons from an on-chip silicon nitride microresonator.Optics Express26(2), 1825–1840 (2018)
2018
-
[33]
R. N. Clark,et al., Integrated Photonics for Continuous-Variable Quantum Optics.Nature Photonics20(5), 489–503 (2026)
2026
-
[34]
Gonz ´alez-Arciniegas, P
C. Gonz ´alez-Arciniegas, P. Nussenzveig, M. Martinelli, O. Pfister, Cluster States from Gaussian States: Essential Diagnostic Tools for Continuous-Variable One-Way Quantum Computing. PRX Quantum2(3), 30343 (2021)
2021
-
[35]
Aasi,et al., Enhanced Sensitivity of the LIGO Gravitational Wave Detector by Using Squeezed States of Light.Nature Photonics7(8), 613–619 (2013)
J. Aasi,et al., Enhanced Sensitivity of the LIGO Gravitational Wave Detector by Using Squeezed States of Light.Nature Photonics7(8), 613–619 (2013)
2013
-
[36]
D. I. Herman,et al., Squeezed Dual-Comb Spectroscopy.Science387(6734), 653–658 (2025). 19 Acknowledgments Funding:This work is supported by City University of Hong Kong (Project No. 9610522), the General Research Fund of the Hong Kong Research Grants Council (No. 11307823), an...
2025
Reviewed August 15, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.