{"id":"040530f6-df4a-415f-a01e-bd917b3c3709","arxiv_id":"2508.03303","paper_version":1,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A new coherent-control scheme stabilizes homodyne detection of two-color EPR light, yielding 9 dB of two-mode squeezing and verified entanglement at acoustic frequencies.","lead":"This paper demonstrates a stable way to control the phases of two-color entangled light beams from a nondegenerate optical parametric oscillator, using a single injected seed beam as a reference. The technique achieves 9 dB of two-mode squeezing, a strong level of entanglement useful for quantum networks and distributed sensing.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Eq. 6b's 1/(1−i∆′) factor is dropped in Eq. 7b; it shifts the CL phase sum by atan(Ω/γ)≈11°, so the claim that detuning phases don't affect the relative condition needs explicit calibration before the 9 dB result is fully accounted for.","rationale":"The strongest claim is the demonstration of a phase-control scheme for two-color EPR. The most load-bearing condition is that the bright lock fields carry the correct relative phase. The paper's own Eqs. (6a)-(6b) contain a detuning-dependent complex factor in ACLi whose phase is dropped when writing Eq. (7b). This is not a mere convention: the lock equations convert this phase into a relative offset between the two LO setpoints. At the quoted parameters, the offset is about 11°, which is far beyond the tolerance at 9 dB squeezing. The experiment may well have compensated this by optimizing the electronic phase references (the text mentions optimizing setpoints), but then the statement that detuning shifts don't affect the relative phase condition is misleading, and the condition θref_s=θref_i is not generally correct. This does not make the claimed entanglement impossible—an in-situ phase calibration would restore it—so I would not reject. It does mean the paper should be accepted only after the calibration is documented or the cancellation is derived. Hence CONDITIONAL. I partially agree with the reader, who identified the same weak spot more generally; my reading sharpens it to a concrete omitted phase and specifies the test.","tokens_in":9828,"tokens_out":19512,"duration_ms":246100,"concrete_test":"Recompute the CL phase sum from Eqs. (6a)-(6b) without the neglect in Sec. II.A and predict the offset δ=atan(Ω/γ). Then compare δ with the calibrated lock-point difference (θref_i−θref_s) recorded while optimizing the Fig. 5 data: if the optimum is shifted by ≈11° (or by the predicted δ), the 9 dB result relies on setpoint calibration and the paper must state and justify it; if no shift is found, the phase cancellation must be demonstrated analytically, since the present text does not provide it.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"From Eq. (6b), ACLi = ε e^{iϕp}/(1−i∆′) A*_CLs, so arg ACLi = ϕp − arg ACLs + atan∆′; hence ϕCLs + ϕCLi = ϕp + atan∆′, not Eq. (2). The text in Sec. II.A says constant detuning phase shifts 'do not affect the relative phase condition,' but with Ω/2π≈3 MHz and γ=15 MHz, ∆′≈0.2 and atan∆′≈11°. At 9 dB squeezing (V_min≈0.13, V_anti≈50), an 11° offset in the relative lock point raises the measured variance to ≈2, destroying the observed squeezing unless θref_s and θref_i are independently calibrated to cancel this offset. The paper neither derives this cancellation nor reports the calibrated setpoint difference; it states the optimal condition is θref_s=θref_i. This is a load-bearing gap in the theoretical description of how the bright locking fields transfer the phase reference to the quantum modes, exactly the assumption the experimental claim rests on.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents a coherent phase-control scheme for a nondegenerate optical parametric oscillator (NOPO) that generates two-color continuous-variable EPR-entangled light. A single seed beam injected at the signal wavelength is parametrically amplified to produce two bright locking fields at the signal and idler wavelengths, which serve as phase references for the homodyne detection of both entangled modes. The scheme is implemented on a 1064 nm/852 nm NOPO, and the authors report 9 dB of two-mode squeezing in the 5–15 kHz band and a Duan–Simon parameter ΔQ²₋ + Δ(Q^{π/2}₊)² = 0.26(1), demonstrating strong EPR entanglement. The paper also models the effect of residual phase noise, calibrates it from error signals, and fits the squeezing-vs-pump-power curves to extract detection efficiency η = 0.89 and phase noise σ̃Θ = 8 mrad.","tokens_in":10022,"tokens_out":17531,"duration_ms":180032,"significance":"If the claims hold, this is a significant technical advance: it provides a practical method for phase-stabilizing homodyne detection of highly nondegenerate two-color EPR sources, which is a known bottleneck for CV quantum networking and distributed sensing. The demonstrated 9 dB of two-mode squeezing at a wavelength separation of about 200 nm is a clear improvement over previous two-color sources. The paper gives explicit experimental attention to phase-noise characterization and to the optimization of the detection setpoints, which strengthens the empirical case. The Duan–Simon violation by a factor of roughly eight is compelling evidence of entanglement. However, two load-bearing theoretical points need to be corrected or clarified before the manuscript is fully convincing.","major_comments":[{"comment":"The statement that detuning-induced constant phase shifts 'do not affect the relative phase condition' is not justified. From Eq. (6b), ACLi = ε e^{iφp}/(1−i∆′) A*_CLs, so the phase sum of the locking fields is φCLs + φCLi = φp + atan(∆′), not φp as stated in Eq. (7b). With the experimental parameters Ω/2π ≈ 3 MHz and γ = 15 MHz, ∆′ ≈ 0.2 and atan(∆′) ≈ 11°, a substantial offset. Because the local oscillators are locked to these locking fields, this offset would shift the homodyne detection away from the optimal squeezing condition unless the electronic phase references θref_s and θref_i are calibrated to compensate. The paper neither derives such a cancellation nor reports a calibrated setpoint difference; it asserts that the optimal condition is θref_s = θref_i. This is a load-bearing gap in the theoretical description of how the bright locking fields transfer the phase reference to the quantum modes, and it directly affects the interpretation of the 9 dB squeezing result.","section":"Section II.A, Eqs. (6)–(7)"},{"comment":"Equation (14) gives the noise spectral densities as ΔQ²± = 1 ± η 4ε/(Ω²/γ² + (ε+1)²), with the same denominator for both the squeezed and anti-squeezed quadratures. For a nondegenerate OPA below threshold, the correct expressions are ΔQ²₋ = 1 − η 4ε/((1+ε)² + Ω²/γ²) and ΔQ²₊ = 1 + η 4ε/((1−ε)² + Ω²/γ²). At ε = 0.8 and Ω = 0, the paper's formula would give ΔQ²₊ ≈ 1.9 (about 2.8 dB of anti-squeezing), whereas the experimental data in Fig. 5 show anti-squeezing around 17 dB (variance ≈ 50) at the same pump power. The formula as written is therefore inconsistent with the reported measurements. Since this formula underpins the phase-noise model in Eq. (18) and the subsequent fit that yields η = 0.89 and σ̃Θ = 8 mrad, the fit results are called into question. The authors must correct the formula and refit the data, or explain the discrepancy.","section":"Section II.C, Eq. (14)"}],"minor_comments":[{"comment":"The caption lists ΔQ²₊ and Δ(Q^{π/2}₋)² as anti-squeezed and then ΔQ²₋ and Δ(Q^{π/2}₋)² as squeezed, repeating the same symbol for the second pair. Presumably one of the latter should be Δ(Q^{π/2}₊)²; please correct the typo.","section":"Fig. 6 caption"},{"comment":"The text states that 180 mW of pump corresponds to ε = 0.75, while the caption of Fig. 5 states that the maximum squeezing is achieved near ε = 0.8. These numbers should be reconciled.","section":"Section IV, experimental parameters"},{"comment":"The wavelengths 1064 nm and 852 nm are separated by 212 nm, not '200 nanometers' as stated in the abstract and introduction. Please use the accurate value or say 'approximately 200 nm'.","section":"Abstract and Introduction"},{"comment":"The definition of φLO_i in Eq. (9b), φLO_i = φp − φCLi + θref_i, is not symmetric with the signal expression and appears to be a source of confusion for the reader. The sign conventions for θref_s and θref_i and their connection to Eq. (8) should be stated more clearly.","section":"Section II.B, Eq. (9)"},{"comment":"The sentence 'The roughly constant σΘ = 10 ± 2 mrad highlights the stability and robustness of the locking scheme' is slightly ambiguous, as Fig. 4 suggests that σΘ may vary with ε; please clarify whether the quoted value is an average over all operating points.","section":"Section IV, phase noise estimation"}],"recommendation":"major_revision","confidential_remarks":"The experimental results appear to be solid and the 9 dB two-mode squeezing at a wavelength separation of ~200 nm is a valuable advance for the CV quantum information community. The two major issues are fixable in revision: the first requires deriving or explicitly calibrating the detuning phase offset, and the second is a clear formula error in Eq. (14) that should be corrected and the data refit. I do not see evidence that the central experimental claims are wrong, but the theoretical framework as written is not yet trustworthy enough for publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First, the thing you should know: this is a real experimental step forward. A single seed injected into a NOPO generates the two coherent lock fields for homodyne detection of EPR modes separated by 200 nm, with no seed-pump phase lock. They see 9 dB two-mode squeezing and a Duan-Simon sum of 0.26(1). That's a clear advance for two-color CV entanglement, and the technique is useful.\n\nWhat's good: the phase-noise calibration is done two ways—error-signal PSD and photocurrent fluctuation fits—and they agree. The in-situ optimization of the lock point is the right engineering instinct. The paper is honest about long-term drift and PZT bandwidth limits. The citation pattern is fair; it builds on Vahlbruch/Chelkowski and the group's own earlier work.\n\nNow the soft spots. First, the stress-test note is right: Eq. (6b) has a 1/(1−iΔ′) factor, and Eq. (7b) drops it. At Ω/γ≈0.2 that is an 11° phase offset in φ_CLi. The sentence 'constant phase shifts introduced by the detuning ... do not affect the relative phase condition' is asserted, not derived. With 17 dB anti-squeezing, an uncorrected 11° offset would push the squeezed variance up to about 2, so the fact that they see 9 dB means either the setpoints were calibrated to cancel the offset (not reported) or there's a cancellation in the locking equations the text doesn't show. The paper needs to close that gap.\n\nSecond, and I think more serious, Eq. (14) gives the same (ε+1)² denominator for both Q− and Q+, which caps the anti-squeezing at 1+η (about 2.7 dB). But Fig. 6 reports 17 dB of anti-squeezing. The standard NOPO formula has (1−ε)² in the anti-squeezed denominator; near threshold that produces large anti-squeezing. As written, Eq. (14) cannot fit the data, so the fitted η=0.89 and σΘ=8 mrad are on shaky ground. This doesn't invalidate the measured Duan-Simon number—that's direct—but it does mean the model needs correcting.\n\nWho should read it: people building hybrid quantum networks or distributed sensing with disparate wavelengths. The experimental result deserves a serious referee; the theory as written doesn't. I'd send it back for a revision that fixes the phase-transfer derivation and the anti-squeezing formula, and I'd ask for raw data or at least the calibrated setpoint values.","headline":"A real experimental advance in two-color CV entanglement, but the paper's phase-transfer argument drops an 11° detuning phase and its anti-squeezing formula looks wrong, so the theory needs correction before the parameter claims are taken at face value.","tokens_in":10614,"tokens_out":8964,"would_cite":true,"duration_ms":105196,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A single injected seed beam, amplified inside a nondegenerate optical parametric oscillator, produces two bright coherent locking fields that hold both homodyne detectors at the entanglement phase condition, yielding 9 dB of two-mode…","keywords":["continuous-variable entanglement","two-color EPR state","coherent control","phase stabilization","homodyne detection","optical parametric oscillator","two-mode squeezing","quantum networking"],"falsifier":"Record the local-oscillator setpoints that minimize the Duan–Simon variance sum while stepping the seed detuning $\\Omega_{\\mathrm{CL}}$ from near zero to several MHz across the cavity linewidth, keeping all other settings fixed. If the optimal setpoints shift measurably with detuning, or if the achieved variance sum degrades faster than the phase-noise model of Eq. (18) predicts, the assumption that detuning-induced phase shifts leave the relative phase condition untouched is falsified.","tokens_in":9584,"feed_emoji":"🔗","tokens_out":20138,"duration_ms":193287,"temperature":0.7,"pith_summary":"Quantum entanglement between two light beams whose wavelengths are separated by about 200 nm is a building block for hybrid quantum networks and distributed sensing, but it requires keeping the detection phases of both beams locked to the pump that created them. This paper presents a control scheme that does this with a single weak seed beam: the seed is injected into the same nondegenerate optical parametric oscillator, and parametric gain converts it into two bright coherent locking beams, one at each entangled wavelength, whose relative phase automatically obeys the entanglement condition. Locking each detector's phase reference to its corresponding locking beam holds both detection angles at the optimum. With that control, the paper stabilizes a two-color Einstein-Podolsky-Rosen (EPR) state at 1064 nm and 852 nm, observes up to 9 dB of two-mode squeezing, and verifies entanglement through the Duan–Simon criterion, whose variance sum reaches $0.26(1)$, far below the separability bound of $2$. The scheme matters because it removes the need to phase-lock the seed to the pump and eases the phase-noise constraint that previously limited such sources to weaker correlations.","feed_headline":"One seed beam locks two-color entanglement at 9 dB","feed_subtitle":"A single injected seed creates two pilot beams that keep the detectors at the optimal phase, giving 9 dB of squeezing.","key_machinery":"The central machinery is the pair of coherent locking fields generated from a single injected seed by the same parametric process that produces the entangled state. Solving the classical intracavity dynamics of the nondegenerate optical parametric amplifier in steady state gives $A_{\\mathrm{CL}_i} = \\epsilon e^{i\\phi_p}(1-i\\Delta')^{-1} A_{\\mathrm{CL}_s}^*$, so the locking fields automatically satisfy $\\phi_{\\mathrm{CL}_i} = \\phi_p - \\phi_{\\mathrm{CL}_s}$. These fields sit at $\\pm\\Omega_{\\mathrm{CL}}$ from the cavity resonances, within the linewidth but outside the analysis band, so they experience the nonlinear interaction without contaminating the quantum sidebands. Homodyne beat notes at $\\pm\\Omega_{\\mathrm{CL}}$ are demodulated to produce error signals whose setpoints $\\theta_{\\mathrm{ref}}^s$ and $\\theta_{\\mathrm{ref}}^i$ directly choose the measured quadrature angles, forcing the detection to follow the entanglement phase condition.","core_discovery":"On the paper's own terms, the central discovery is that the same nonlinear interaction that entangles the quantum modes also imprints the entanglement phase relation onto two bright classical fields, and these bright fields can serve as error-signal references for homodyne detection. A single seed at $\\omega_s+\\Omega_{\\mathrm{CL}}$ injected into the NOPO is parametrically amplified into CL$_s$ and, by difference-frequency generation, produces CL$_i$ whose phase obeys $\\phi_{\\mathrm{CL}_i} = \\phi_p - \\phi_{\\mathrm{CL}_s}$, exactly the condition $\\phi_i = \\phi_p - \\phi_s$ required for maximal EPR correlations. Demodulated beat notes between each locking field and its local oscillator provide sine-shaped error signals whose setpoints set the local-oscillator phases to enforce the entanglement condition. With this locking, the measured joint quadrature variances follow the phase-noise-degraded theoretical model, giving 9 dB of two-mode squeezing and a Duan–Simon sum $\\Delta Q_-^2 + \\Delta(Q^{\\pi/2}_+)^2 = 0.26(1)$. The scheme is presented as valid in the highly nondegenerate regime and as removing the pump-seed phase-locking requirement of earlier coherent-control approaches.","pith_inferences":["The paper leaves implicit that the same coherent locking fields could also be used to stabilize the cavity length, since their phases carry cavity-error information, potentially consolidating two locking tasks into one setup.","A direct test of the core assumption would be to step the seed detuning across the cavity linewidth and see whether the local-oscillator setpoint that maximizes squeezing shifts; a measurable shift would mean the bright and quantum fields do not share the same frequency-dependent phase response.","The absence of a seed-pump phase lock suggests the scheme could be portable to field or network settings where the pump and seed sources are not mutually coherent, as long as the relative phase condition is still enforced inside the OPO."],"forward_implications":["Two-color CV entanglement can be phase-stabilized with one seed and no seed-pump phase lock, so the technique applies to wavelength pairs that are hard to lock by conventional means.","Because the locking fields and the entanglement are generated by the same nonlinear process, the local-oscillator phases automatically track the pump phase, allowing both quadratures to be measured at their optimal angles.","With common-mode phase noise of about 10 mrad, the source reaches its optimum near $\\epsilon \\approx 0.8$, yielding up to 9 dB of two-mode squeezing and narrowing the gap to single-mode squeezed sources.","The Duan–Simon variance sum of $0.26(1)$ confirms genuine entanglement over the 5–17 kHz acoustic band, a range suited to quantum metrology and sensing.","The error signals themselves provide real-time phase-noise estimates, so the same locks can be used to monitor and optimize the system during operation."],"supporting_citations":[{"why":"Introduces the single-beam coherent control of vacuum squeezing that this paper generalizes to two-mode EPR states.","marker":"[4]"},{"why":"Supplies the generalized-quadrature phase space decoupled from the pump and the entanglement phase condition used throughout.","marker":"[7]"},{"why":"Earlier coherent control of broadband vacuum squeezing in which pump-seed phase locking was required; this scheme removes that requirement.","marker":"[10]"},{"why":"Demonstrates phase control of EPR-based squeezing for near-degenerate modes, the regime this work extends to strongly nondegenerate two-color fields.","marker":"[14]"},{"why":"Duan inseparability criterion that the paper uses to certify entanglement.","marker":"[15]"},{"why":"Simon version of the separability criterion used together with the Duan criterion.","marker":"[16]"},{"why":"Previous two-colour EPR source from the same setup that reported a lower optimum pump amplitude, providing the baseline this work improves on.","marker":"[17]"}],"fun_headline_variants":["One seed beam locks two-color EPR at 9 dB","Single seed creates phase references for 9 dB squeezing","Self-referencing two-color EPR locked by one seed","One injected beam yields two-lock fields for EPR"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument assumes that detuning-induced phase shifts do not change the relative phase between the bright locking fields and the quantum modes, so locking the detectors to the bright fields automatically puts the measurement at the entanglement optimum; if the two kinds of fields pick up different frequency-dependent phase shifts inside the cavity, the locked angles would be offset and the reported correlations could be a lock-point artifact.","fun_headline_variants_meta":{"raw":{"variants":["One seed beam locks two-color EPR at 9 dB","Single seed creates phase references for 9 dB squeezing","Self-referencing two-color EPR locked by one seed","One injected beam yields two-lock fields for EPR"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000435,"raw_usage":{"total_tokens":2192,"prompt_tokens":901,"completion_tokens":1291,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":517,"completion_tokens_details":{"reasoning_tokens":1223}},"tokens_in":517,"tokens_out":1291,"duration_ms":13497,"temperature":1.0,"reasoning_tokens":1223,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T04:31:41.004120+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Record the local-oscillator setpoints that minimize the Duan–Simon variance sum while stepping the seed detuning $\\Omega_{\\mathrm{CL}}$ from near zero to several MHz across the cavity linewidth, keeping all other settings fixed. If the optimal setpoints shift measurably with detuning, or if the achieved variance sum degrades faster than the phase-noise model of Eq. (18) predicts, the assumption that detuning-induced phase shifts leave the relative phase condition untouched is falsified.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Introduces the single-beam coherent control of vacuum squeezing that this paper generalizes to two-mode EPR states."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the generalized-quadrature phase space decoupled from the pump and the entanglement phase condition used throughout."},{"cited_title":"& Schnabel, R","cited_arxiv_id":null,"evidence_quote":"Earlier coherent control of broadband vacuum squeezing in which pump-seed phase locking was required; this scheme removes that requirement."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Demonstrates phase control of EPR-based squeezing for near-degenerate modes, the regime this work extends to strongly nondegenerate two-color fields."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Duan inseparability criterion that the paper uses to certify entanglement."},{"cited_title":"Peres-Horodecki separability criterion for continu- ous variable systems","cited_arxiv_id":null,"evidence_quote":"Simon version of the separability criterion used together with the Duan criterion."},{"cited_title":"B., Novikov, V ., Kerdoncuff, H., Lassen, M","cited_arxiv_id":null,"evidence_quote":"Previous two-colour EPR source from the same setup that reported a lower optimum pump amplitude, providing the baseline this work improves on."}],"review_version":1}