{"id":"db22e55d-1ce1-402d-92ad-05f2f7dbd8b1","arxiv_id":"2608.07426","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"An electrically tunable dark mode in a silicon-nitride microresonator is populated by four-wave mixing, extends the partner photon's coherence time to 2.1 ns, and adds 12 dB parametric gain.","lead":"Researchers made a tiny silicon-nitride light trap with one deliberately dark resonance that ordinary laser light cannot enter. They show this hidden mode can still be created inside by a nonlinear process, and it stretches the coherence of its visible partner photon and boosts optical gain by about 12 dB.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Dark-idler claim rests on an unquantified bound for residual MZI coupling; with vanishing extinction, Q_L/Q_int=1 cannot distinguish a true BIC from a quasi-BIC with γ_ext comparable to γ_int.","rationale":"The paper's central demonstration is that a mode made linearly inaccessible through a Friedrich-Wintgen BIC remains populated by SFWM and that its long, intrinsic-loss-limited lifetime is imprinted on the radiatively coupled signal photon. The reader's weakest-assumption analysis correctly locates the load-bearing premise: residual radiative coupling at the BIC must be negligible compared with the intrinsic loss rate for the dark-idler interpretation to hold. I agree with that assessment. The manuscript states the premise in Section III A but provides no numerical bound, and the Q_L/Q_int=1 evidence in Section III B is extracted from transmission spectra at the point where the resonance essentially disappears, which is a difficult measurement. My concrete test uses the nonlinear readout itself—the StFWM-generated idler line shape—to determine γ_i without relying on transmission extinction. If the test confirms γ_i≈γ_int, the central claim is on solid ground; if it reveals a broader linewidth, the coherence-transfer and gain-enhancement results remain real but should be attributed to a quasi-BIC with finite residual coupling. The reader's CONDITIONAL verdict is appropriate because the claim is plausible and directly measured in several independent ways (22 dB suppression, 12 dB gain advantage, 1.6 dB threshold reduction, 2.14 ns coherence time) but depends on an assumption that should be made quantitative. No change to the verdict is needed; the proposed check would either validate or refute the key premise.","tokens_in":20690,"tokens_out":4467,"duration_ms":48297,"concrete_test":"Use the nonlinear readout as a spectroscopic probe of the dark mode. With a narrow-linewidth CW pump and a weak seed scanned across the signal resonance in the BIC configuration, record the idler output spectrum with a heterodyne or high-resolution OSA (≤20 MHz). Fit the narrow feature to the convolution of the pump envelope, the signal Lorentzian, and an idler Lorentzian, treating γ_i as the sole free parameter. If the best-fit γ_i is within, say, 20% of γ_int=2π×74 MHz, the residual coupling is negligible; if it exceeds 1.2γ_int, the intrinsic-loss-limited interpretation and the 2.13 ns prediction must be revised. Additionally, verify Q_int independently by operating in a strongly undercoupled regime (γ_ext<<γ_int) where the resonance remains visible, rather than relying on the critical-coupling relation Q_int=2Q_L.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim—that the idler at the BIC condition is intrinsic-loss-limited—depends on the assertion in Section III A that 'the residual coupling plays a negligible role in determining the lifetime of the BIC,' which is never quantified. The supporting evidence in Section III B is the ratio Q_L/Q_int=1 extracted from transmission spectra as the resonance extinction vanishes. At the decoupling point the resonance is almost invisible, so fitting a Lorentzian to a disappearing dip is ill-conditioned; a residual coupling of, say, 0.2γ_int would make Q_L=0.83Q_int, shortening the predicted signal coherence time from 2.13 ns to about 1.8 ns and weakening the quasi-dark character of the mode. Because the same Q_int is inferred from a critical-coupling assumption (Q_int=2Q_L at Q_L=1.3e6), a small error in setting critical coupling shifts the predicted 2.13 ns. Thus the headline agreement between 2.14 ns and 2.13 ns is not by itself sufficient to establish that residual coupling is negligible; an independent, extinction-free measurement of the idler linewidth is needed.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports a reconfigurable silicon-nitride microresonator platform in which a Friedrich-Wintgen bound state in the continuum is realized by an interferometric coupler with an auxiliary ring, and it uses this device to study spontaneous four-wave mixing when the idler mode is tuned to the BIC. The authors report linear decoupling of the idler resonance, about 22 dB suppression of the output idler power under stimulated FWM, spontaneous-regime data on single counts, coincidences, cross-correlation and signal self-correlation, a signal coherence time of 2.14 +/- 0.06 ns at the BIC condition, and a BIC-assisted OPA gain advantage of about 12 dB with a 1.6 dB lower oscillation threshold. The interpretation is supported by a quantum Hamiltonian treatment that explicitly includes the BIC mode and by simulations of the photon-pair generation and OPA gain.","tokens_in":20864,"tokens_out":8283,"duration_ms":88929,"significance":"If the quantitative claims hold, this is a significant advance: it demonstrates that a mode linearly decoupled from the bus waveguide remains dynamically active in a chi^(3) process, transferring its long intrinsic lifetime to the radiatively coupled twin photon. The experimental work is direct, internally consistent, and contains a strong falsifiable prediction: the measured signal coherence time at the BIC, 2.14 +/- 0.06 ns, matches Q_int/omega = 2.13 ns. The imaging and 22 dB suppression data provide visual and spectral evidence of a dark idler, and the OPA/OPO results show practical utility. The principal weakness is that the central 'intrinsic-loss-limited' claim relies on an unquantified statement about residual coupling, and several headline numbers are presented without uncertainties. The theoretical framework is mostly parameter-free; the only fitted parameter in the supporting models is the thermo-optic coupling-rate coefficient xi in Appendix B, which affects the simulated gain curves but not the measured values.","major_comments":[{"comment":"The claim that the BIC idler is intrinsic-loss-limited is load-bearing and currently rests on the statement in Section III A that 'the residual coupling plays a negligible role in determining the lifetime of the BIC,' which is not quantified. The supporting evidence, Q_L/Q_int = 1 extracted from a transmission dip that vanishes at the decoupling point, is ill-conditioned: at vanishing extinction, a Lorentzian fit to the disappearing resonance cannot reliably distinguish Q_L = Q_int from, say, Q_L = 0.83 Q_int. The 2.14 ns versus 2.13 ns agreement is encouraging, but the predicted 2.13 ns inherits the systematic calibration of Q_int = 2 Q_L at critical coupling, for which no uncertainty is reported. Please provide a quantitative upper bound on gamma_ext/gamma_int; one constructive route is to convert the 22 dB output suppression into an escape-efficiency bound after characterizing the Raman/bus-waveguide background, and another is an independent, extinction-free measurement of the idler linewidth, such as nonlinear or pump-probe spectroscopy of the dark mode.","section":"III A and III B"},{"comment":"The quantitative reductions of idler singles by about 9 dB, coincidences by about 7 dB, and the doubling of signal counts are quoted without error bars or a background-subtraction protocol. The manuscript itself attributes residual counts to Raman scattering and SFWM in the access waveguide, and the Appendix A simulation predicts a much deeper coincidence dip than observed. Because these numbers quantify the central claim that the BIC suppresses extraction while enhancing internal generation, the paper should report uncertainties on all points in Fig. 5(a,b) and describe how the background is modeled or subtracted before the dB values are computed.","section":"III B, Figs. 5 and 6"},{"comment":"Two quantitative conclusions are based on fits whose systematic uncertainty is not addressed. First, at the BIC condition the cross-correlation has a double-exponential tail arising from the coexistence of resonant and continuum components, yet the signal self-correlation is fitted with a single exponential; a residual fast component could bias the extracted tau_c, which is a central number. Second, the 12 dB gain advantage and 1.6 dB threshold reduction are point estimates, and the corroborating simulation uses a fitted coefficient xi with no reported uncertainty or sensitivity analysis. Please justify the single-exponential model for tau_c and provide error bars or confidence intervals for the gain and threshold data, together with a brief analysis of how the simulated gain curves depend on xi.","section":"III B and III C"}],"minor_comments":[{"comment":"The phrase 'In particularly' should read 'In particular.'","section":"Abstract"},{"comment":"The sentence describing the broad background of the BIC StFWM lineshape—'which tracks the form of the field enhancement of only the seed wave, still close to the expected value of 1.5 GHz, following the same procedure'—is grammatically incomplete and should be rewritten for clarity.","section":"III A, fourth paragraph"},{"comment":"The bottom panel of Fig. 5(c) displays Q_L/Q_int values, but the caption does not define the symbol Q_L/Q_int or state how the loaded Q-factor was extracted; please add the definition and the extraction procedure.","section":"Fig. 5 caption"},{"comment":"The nonlinear coupling amplitude S(omega_1, omega_BIC, omega_3, omega_4) is used before its normalization and dependence are described; a brief definition would help the reader connect Eq. (4) to the asymptotic-field formalism of Ref. [32].","section":"II, Eq. (4)"},{"comment":"The pump and seed input terms contain a factor FSR that is not explicitly defined in the equation; please state its value and units, and clarify whether it is the same for all three modes.","section":"Appendix B, Eq. (B1)"},{"comment":"The text refers to 'CAR = g_si^(2) - 1' as a coincidence-to-accidental ratio; please define CAR explicitly and use consistent nomenclature, since the usual CAR definition differs from this expression by a factor involving the coincidence window and singles rates.","section":"III B"}],"recommendation":"major_revision","confidential_remarks":"For the editor: the experimental effort is substantial and the central idea is interesting and well within the journal's scope. I do not see a fundamental flaw that would require rejection; the revision should focus on quantifying the residual-coupling bound and the uncertainties of the headline numbers. If the authors can provide a clean bound on gamma_ext/gamma_int and report error bars on the count-rate and gain data, the paper would be a strong candidate for acceptance."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a real advance, and the central claim survives the obvious objections. The paper demonstrates an electrically tunable Friedrich-Wintgen BIC in a SiN microresonator, populates the dark mode through SFWM, and shows that the signal photon inherits the idler's long coherence time (2.14 ns vs 2.13 ns expected). The 12 dB OPA gain advantage and 1.6 dB threshold reduction are direct comparisons between BIC and equal-Q configurations, so the main results don't lean on fitted parameters.\n\nThe strongest part is the coherence-time transfer measurement. The g^(2)_ss fit gives τ_c = 2.14±0.06 ns at the BIC condition, matching Q_int/ω = 2.13 ns. This is better evidence that residual coupling is negligible than the vanishing-extinction fit, because it's an independent, extinction-free probe of the idler linewidth: if residual coupling were 0.2γ_int, τ_c would be ~1.8 ns, which is many sigma away. The stress-test note worries about this, but I think the data already answer it, provided the Q_int assignment from critical coupling is credible. The critical-coupling inference is standard, and the internal consistency of the coherence match supports it.\n\nWhere the paper is soft: headline numbers like 22 dB idler suppression, 12 dB gain, and 1.6 dB threshold lack error bars. The OPA simulation uses a fitted thermo-optic coupling coefficient ξ, so the 'quantitative agreement' there is weaker than it looks. No data/code released. And there's a small rhetorical tension: the text says absolute calibration is prevented by Raman and waveguide SFWM, but then says the signal doubling 'demonstrates' enhanced internal generation. It's plausible, but it's an inference, not a calibrated measurement.\n\nNone of these sink the paper. The residual-coupling premise is the only load-bearing assumption, and the coherence-time data effectively bound it. I'd send this to peer review with requests for error bars, a quantitative residual-coupling bound (which they can extract from the coherence measurement), and a clear statement of which simulation parameters are fitted. It's a solid contribution for quantum and nonlinear integrated photonics.","headline":"Dark idler claim holds up: the coherence-time measurement already bounds residual coupling, so the paper's main results are direct and credible.","tokens_in":21455,"tokens_out":3724,"would_cite":true,"duration_ms":35273,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper demonstrates that a bound state in the continuum, a cavity mode that cannot be excited or extracted through the bus waveguide, remains fully active in four-wave mixing and can be read out through its radiative twin.","keywords":["bound states in the continuum","silicon nitride microresonator","spontaneous four-wave mixing","Friedrich-Wintgen BIC","optical parametric amplification","photon-pair correlations","coherence time","mode-selective lifetime engineering"],"falsifier":"At the BIC heater setting, measure the spectral width of the idler generated by stimulated four-wave mixing after removing pump and seed broadening; an intrinsic-loss-limited dark idler should show a linewidth near $\\gamma_{\\rm int}/2\\pi = 74$ MHz, while a broader idler or a signal coherence time that saturates below $Q_{\\rm int}/\\omega$ would show that residual radiative coupling is not negligible and the BIC is quasi-dark rather than dark.","tokens_in":20445,"feed_emoji":"🔬","tokens_out":6704,"duration_ms":63033,"temperature":0.7,"pith_summary":"The paper shows that a resonant mode which is optically dark—a bound state in the continuum (BIC)—can still drive nonlinear light generation when it is created inside the cavity by four-wave mixing. In a silicon nitride microresonator with an interferometric coupler, the authors tune only the idler mode into the BIC condition and read the dark mode through its radiative partner, the signal photon. The signal photon's coherence time grows to 2.14 ns, matching the intrinsic-loss-limited 2.13 ns lifetime of the dark idler, even though the idler output is suppressed by roughly 22 dB. In the stimulated regime, the same dark mode raises parametric gain by about 12 dB and lowers the oscillation threshold by 1.6 dB. The intended consequence is that BICs enable mode-selective dissipation engineering, letting a device control biphoton wavepackets and amplifier thresholds separately from photon extraction.","feed_headline":"Hidden mode boosts photon-pair coherence and amplifier gain","feed_subtitle":"Suppressing the idler's escape by 22 dB still lengthens signal coherence to 2.14 ns and raises parametric gain by 12 dB.","key_machinery":"The central object is the Friedrich-Wintgen bound state in the continuum: a square-integrable cavity mode at a frequency inside the radiation continuum whose net radiative amplitude vanishes by destructive interference of multiple leakage paths, leaving only intrinsic loss to limit its lifetime. In the device, the leakage paths are the two effective couplings of the main ring to the bus through a Mach-Zehnder interferometer, and an auxiliary ring provides a resonant phase shift that makes the cancellation wavelength-selective. The theoretical machinery is a quantization of the displacement field that explicitly includes the discrete BIC operator alongside the continuum asymptotic-in operators; the four-wave-mixing Hamiltonian then contains a term that creates one BIC photon and one continuum photon from two pump photons. Because the BIC mode is orthogonal to the asymptotic fields even at the same frequency, it can be populated only internally by the nonlinearity and read out through its twin. The measured effects—idler escape suppression, signal coherence transfer, and gain enhancement—follow from the same rate asymmetry $\\gamma_i \\ll \\gamma_s$ between the dark idler and the radiative signal.","core_discovery":"The central claim is that suppressing linear access to one participating mode does not remove that mode from the nonlinear dynamics, and the dark mode's long lifetime can be transferred to a radiatively coupled partner. The authors realize electrically tunable Friedrich-Wintgen BICs with quality factors above $10^6$ in a silicon nitride microresonator, and place the idler of spontaneous four-wave mixing at the BIC. The idler generated inside the resonator cannot escape to the output waveguide (suppression of about 22 dB), yet the pair-generation process is not suppressed: signal singles roughly double, and the extracted signal photon inherits the dark idler's lifetime, reaching 2.14 ns against the expected $Q_{\\rm int}/\\omega = 2.13$ ns. In the stimulated regime, optical parametric amplification with the idler as a BIC gives about 12 dB more signal gain at 13.5 dBm on-chip pump power and a 1.6 dB lower oscillation threshold than the equal-Q configuration. The paper's central statement is that wavelength-selective suppression of radiative loss controls nonlinear interaction, spectral bandwidth, and photon extraction as distinct degrees of freedom within a single integrated device.","pith_inferences":["Because the joint spectral amplitude is set by the dark idler's linewidth, BIC-assisted spontaneous four-wave mixing should generate two-mode squeezed states with strongly asymmetric marginal bandwidths, which could be matched to narrow-band quantum memory interfaces.","A direct consequence not tested here is that the dark-mode population should ring up on the intrinsic-loss timescale; time-resolving the radiative signal after a pump pulse would confirm that the idler really lives for $Q_{\\rm int}/\\omega$ and not a shorter quasi-BIC lifetime.","The residual 22 dB idler suppression is attributed to point-coupler imbalance; if fabrication symmetry improves, the suppression should grow and the signal coherence time should approach the full intrinsic limit, a testable scaling relation between escape suppression and coherence transfer.","The same selective-dark-mode architecture could be applied to frequency comb generation, where a single dark comb line may suppress mode competition or reshape comb bandwidth by the same loss-asymmetry mechanism used here for optical parametric amplification."],"forward_implications":["The internal pair-generation rate is not suppressed when the idler becomes dark: signal singles roughly double at the BIC condition while extracted idler singles drop by about 9 dB and signal-idler coincidences by about 7 dB.","The signal photon inherits the spectral bandwidth of its dark partner, so the emitted photon's coherence time can be much longer than the loaded lifetime of the emitting resonance, reaching 2.14 ns against the 2.13 ns intrinsic-loss limit.","With the idler as a BIC, the same device gives roughly 12 dB higher optical parametric amplification at 13.5 dBm on-chip pump power and a 1.6 dB lower oscillation threshold than the equal-Q configuration.","The BIC condition is electrically tunable and wavelength-selective: the auxiliary ring affects only one out of every four main-ring resonances, so pump and signal resonances remain radiatively accessible while the idler is dark.","These results establish mode-selective lifetime engineering as a way to control photon extraction, biphoton wavepacket shape, and parametric threshold independently within one integrated device."],"supporting_citations":[{"why":"Defines the Friedrich-Wintgen interference mechanism that the realized BIC relies on.","marker":"[3]"},{"why":"Prior integrated BIC photon-pair source; the paper's quantum application is contrasted with this down-conversion demonstration.","marker":"[11]"},{"why":"Earlier use of BICs for hyperparametric oscillation; the paper extends this to quantum and below-threshold regimes.","marker":"[13]"},{"why":"Supplies the resonant interferometric coupler architecture and selective linewidth control that the device is built from.","marker":"[29]"},{"why":"Establishes time-resolved coincidence measurements on the same silicon nitride platform used for the spontaneous four-wave-mixing experiments.","marker":"[30]"},{"why":"Provides the asymptotic-field Hamiltonian formalism that the quantization of the BIC mode builds on.","marker":"[32]"},{"why":"Extends the formalism to lossy structures with phantom waveguides, used for the numerical simulations.","marker":"[33]"},{"why":"Gives the temporal coupled-mode equations adapted to asymmetric signal-idler escape for the optical parametric amplification simulation.","marker":"[37]"}],"fun_headline_variants":["Dark idler BIC boosts photon coherence to 2.14 ns","Suppressing idler loss yields 12 dB gain boost in microresonator","BIC trick: hide idler, sharpen photon pairs, lower threshold","Electrically tunable BIC gives 2.14 ns photon coherence","Mode-selective BIC enhances parametric gain by 12 dB"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The central claim that the dark idler's lifetime is set purely by intrinsic loss depends on the residual radiative coupling at the BIC being negligible; the paper states this but gives no quantitative bound, so if residual coupling were comparable to intrinsic loss, the loaded Q at the BIC would drop and all the quoted effects would shrink.","fun_headline_variants_meta":{"raw":{"variants":["Dark idler BIC boosts photon coherence to 2.14 ns","Suppressing idler loss yields 12 dB gain boost in microresonator","BIC trick: hide idler, sharpen photon pairs, lower threshold","Electrically tunable BIC gives 2.14 ns photon coherence","Mode-selective BIC enhances parametric gain by 12 dB"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000385,"raw_usage":{"total_tokens":2078,"prompt_tokens":1028,"completion_tokens":1050,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":644,"completion_tokens_details":{"reasoning_tokens":954}},"tokens_in":644,"tokens_out":1050,"duration_ms":9807,"temperature":1.0,"reasoning_tokens":954,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T04:48:46.080204+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"At the BIC heater setting, measure the spectral width of the idler generated by stimulated four-wave mixing after removing pump and seed broadening; an intrinsic-loss-limited dark idler should show a linewidth near $\\gamma_{\\rm int}/2\\pi = 74$ MHz, while a broader idler or a signal coherence time that saturates below $Q_{\\rm int}/\\omega$ would show that residual radiative coupling is not negligible and the BIC is quasi-dark rather than dark.","supporting_citations":[{"cited_title":"Friedrich and D","cited_arxiv_id":null,"evidence_quote":"Defines the Friedrich-Wintgen interference mechanism that the realized BIC relies on."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Prior integrated BIC photon-pair source; the paper's quantum application is contrasted with this down-conversion demonstration."},{"cited_title":"Harnessing bound states in the continuum for quantum and nonlinear photonics in silicon nitride microresonators","cited_arxiv_id":"2608.07426","evidence_quote":"Earlier use of BICs for hyperparametric oscillation; the paper extends this to quantum and below-threshold regimes."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the resonant interferometric coupler architecture and selective linewidth control that the device is built from."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes time-resolved coincidence measurements on the same silicon nitride platform used for the spontaneous four-wave-mixing experiments."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the asymptotic-field Hamiltonian formalism that the quantization of the BIC mode builds on."},{"cited_title":"Liscidini, L","cited_arxiv_id":null,"evidence_quote":"Extends the formalism to lossy structures with phantom waveguides, used for the numerical simulations."},{"cited_title":"Christ, K","cited_arxiv_id":null,"evidence_quote":"Gives the temporal coupled-mode equations adapted to asymmetric signal-idler escape for the optical parametric amplification simulation."}],"review_version":1}