{"id":"9592e253-98b9-49b6-8a44-4a20c7e1dc3e","arxiv_id":"2507.16578","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"A free-space Sagnac polarization encoder for quantum-dot single photons achieves a 0.69% quantum bit error rate at 152 MHz, the lowest reported for high-speed single-photon encoding, with 60-hour stability.","lead":"Researchers built a free-space Sagnac interferometer that encodes quantum information in the polarization of individual telecom photons from a quantum dot, and measured a record-low error rate of 0.69%. The design is inherently phase-stable and stayed stable for 60 hours, which matters for quantum communication systems that need fast, low-error single-photon encoding.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 'lowest QBER to date' claim is unsubstantiated because Table 2 omits the only prior Sagnac-based SPS encoding experiment (Ref. [45]), which the paper itself cites.","rationale":"The reader's verdict is CONDITIONAL, primarily because the 0.69% QBER may not reflect the encoder's intrinsic error floor given the temporal post-selection and one-minute acquisition. I agree that concern is real. However, I identified a more directly checkable threat: the superlative claim 'lowest error rate reported to date' depends on a complete literature comparison, and the paper's own Table 2 omits the only prior Sagnac-based SPS encoding experiment ([45]). This is not a minor citation gap; it is a missing control for the central comparative statement. The paper has strong points: it provides a detailed Methods section, transparently reports the temporal filter and the one-minute acquisition, acknowledges the temporal side-channel risk, and gives a careful stability analysis for the passive path. Those features support the reliability of the experimental work. But the headline claim of a record-low QBER cannot be accepted without verifying that no existing result, especially [45], is lower. Since we do not know the outcome, the existing CONDITIONAL verdict remains appropriate: the paper should be accepted only after the comparison is completed and, if needed, the claim is revised. My concern does not move the verdict; it strengthens the condition already placed on acceptance.","tokens_in":14607,"tokens_out":13567,"duration_ms":140496,"concrete_test":"Retrieve the data of Ref. [45] and any other SPS polarization-encoding experiments published before 22 July 2025. Construct a table with QBER, repetition rate, encoding method, and source type. If any entry reports a QBER equal to or lower than 0.69% at a comparable rate, the paper's 'lowest error rate' claim is false. If all comparable reports have higher QBER, the claim is verified. The comparison should be added to the manuscript regardless of outcome.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that a QBER of 0.69% is the lowest error rate reported to date for high-speed information encoding on single photons. A comparative claim of this kind requires a complete and fair comparison set. The paper explicitly cites Ref. [45] (Xingjian et al., Natl. Sci. Rev. 2025) as the first Sagnac-based SPS encoding experiment, yet Table 2 contains no entry for it. No justification is given for this exclusion, and the table includes a less relevant entry ([53], with g(2)(0) > 0.5) while marking it as incomplete. Unless [45] reports a strictly higher QBER at a comparable repetition rate, the 'lowest to date' claim is not established. The issue is directly checkable: the omitted paper is in the reference list, so its QBER and rate can be extracted and compared. This is more load-bearing than the well-described temporal-filter caveat, because even a perfectly valid 0.69% measurement does not support a superlative claim without a complete comparison. The stability measurement (Sec. 4.2) also used a CW laser with the PM inactive, so it does not directly bound the long-term dynamic single-photon QBER, but the comparison gap is the most immediate threat to the headline claim.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports a free-space Sagnac interferometer (FSSI) based BB84 polarization encoder for single photons from an InGaAs quantum dot emitting at 1560.4 nm. The encoder operates at 151.894 MHz and is claimed to achieve a quantum bit error rate (QBER) of 0.69(2)% and an encoding agreement of 96.5(2)% from one minute of accumulated data with a 2.65 ns temporal filter. The paper also reports a 60-hour polarization stability measurement performed with a CW laser and a polarimeter, and presents an asymptotic secret-key-rate simulation using the measured QBER as an input. The authors explicitly acknowledge the one-minute accumulation, the manual fiber polarization controller optimization, the temporal filtering, and the potential for temporal side-channel leakage.","tokens_in":14831,"tokens_out":6159,"duration_ms":60054,"significance":"If substantiated, a 0.69% QBER at 152 MHz would be a notable result for high-speed single-photon encoding, and the FSSI approach is an interesting way to obtain phase-stable polarization modulation. The manuscript is careful in several respects: it provides a detailed loss budget, openly describes the measurement limitations, discloses the temporal side-channel risk, and presents the SKR as a simulation rather than as a measured rate. However, the headline comparative claims are not fully supported in the current version: Table 2 omits Ref. [45], which the text identifies as the first Sagnac-based SPS encoding experiment, and the encoding-agreement calculation appears to use an unnormalized theoretical state vector. Both issues bear directly on the two headline quantitative claims and require revision.","major_comments":[{"comment":"The claim that 0.69% is the lowest error rate reported to date for high-speed information encoding on single photons is not established, because Table 2 omits Ref. [45], which Sec. 1 identifies as the first Sagnac-based SPS encoding experiment. Since Ref. [45] is a directly relevant SPS polarization-encoding demonstration, its QBER and repetition rate must be included in the comparison or the exclusion must be explicitly justified; without that, the superlative claim is unsupported.","section":"Table 2 / Abstract"},{"comment":"The theoretical state vector in Eq. (8), |D⟩_theo = √(3/2)·(1, 0, 1/2, 1/2), is not normalized (its norm is 1.5), despite the text stating that normalized theoretically expected state vectors are used. Since the encoding agreement in Eq. (9) is computed from the Frobenius distance between M_exp and M_theo, the normalization of the theoretical matrix directly affects the reported 96.5(2)% value. The authors should correct the formula or clearly define the intended normalization and recompute the agreement accordingly.","section":"Methods, Eq. (8)-(9)"},{"comment":"The 60-hour stability measurement is performed with a CW laser, with the phase modulator inactive, and with a polarimeter placed directly after the Sagnac loop, rather than with encoded single photons through the full encoder-fiber-decoder chain. As the authors acknowledge, this isolates encoder stability from fiber-induced fluctuations, but it does not directly demonstrate that the dynamic single-photon QBER of 0.69% persists over 60 hours. The 'ultrastable' title claim should be qualified accordingly, or a long-term single-photon QBER measurement should be provided.","section":"Sec. 4.2 / Fig. 4"}],"minor_comments":[{"comment":"The one-minute accumulation and manual FPC optimization mean the reported QBER is an instantaneous, optimized value; adding a footnote to Table 2 indicating this would help readers compare this work with other experiments on equal footing.","section":"Sec. 4.1"},{"comment":"Equation (13) appears to have a missing closing parenthesis in the denominator; please check and correct the formula.","section":"Methods, Eq. (13)"},{"comment":"The phrase 'best results reported at present regarding dynamic information inscription onto single photons' is broader than what is currently demonstrated; consider restricting the claim to the specific encoder type or completing the comparison table first.","section":"Conclusion"}],"recommendation":"major_revision","confidential_remarks":"The two major issues are both fixable within the scope of the manuscript: adding the omitted comparison entry (Ref. [45]) and correcting the normalization in the encoding-agreement definition. The stability measurement is honestly described but the title overstates what is shown. Overall, the experimental work appears carefully done and the limitations are transparent, so I would encourage revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's the short version: this is a solid, carefully written experimental paper. The authors build a free-space Sagnac polarization encoder for quantum-dot single photons at telecom wavelength, drive it at 152 MHz, and measure a BB84 QBER of 0.69(2)% from direct polarization projection. That's a real advance over the 2.5%-ish numbers that dominate the SPS-encoding literature, and the 60-hour baseline with no active stabilization is genuinely useful for practical QKD. The methods are transparent enough to reproduce: they give the sequence, the temporal filtering procedure, the loss budget, and the SKR simulation is explicitly a simulation using the measured QBER as input.\n\nWhat's actually new is the free-space Sagnac implementation with an asymmetric delay for deterministic single photons, not the Sagnac concept itself. The paper is honest about that and cites the prior WCP Sagnac work and the one prior SPS Sagnac experiment [45]. That's where the biggest problem sits. The abstract claims 'the lowest error rate reported to date for high-speed information encoding on single photons.' But Table 2, which is supposed to justify that, omits [45] — the only prior Sagnac-based SPS encoding experiment, which the paper itself cites. No reason for the exclusion is given, and the table includes a less relevant entry with g(2)>0.5. Unless [45] reports a strictly higher QBER at a comparable rate, the superlative claim is not established. This is directly checkable, and it's the most load-bearing issue in the paper.\n\nThe secondary caveats are real but smaller. The 0.69% number comes from one minute of data after manual fiber polarization control, with a 2.65 ns temporal filter that discards ~10% of events. The paper states this clearly, so it's not hidden, but it means the quoted QBER is a best case, not a guaranteed operating point. The 60-hour stability measurement was done with a CW laser and a polarimeter after the Sagnac loop, not with single photons through the full encoder-fiber-decoder chain. Again, the paper says this explicitly, and it's a reasonable way to isolate encoder stability from fiber drift. So those are minor-to-moderate concerns, not fatal.\n\nOverall, the central result — that a free-space Sagnac encoder can encode QD single photons with sub-1% QBER at 152 MHz — looks correct. The math and data are consistent, the citation pattern is generally good except for the Table 2 gap, and the authors are upfront about side channels and repetition-rate limits. The paper deserves a serious referee. I'd send it out, and I'd make the referee check [45]'s QBER and force the authors to either include it in the comparison or soften the 'lowest to date' phrasing. The temporal-filter and one-minute-data concerns should be discussed but shouldn't block publication if the comparison issue is resolved.","headline":"A clean Sagnac encoder demo for QD single photons with a record-low QBER claim that is not yet substantiated because the closest prior Sagnac-SPS experiment is missing from the comparison table.","tokens_in":15467,"tokens_out":2793,"would_cite":true,"duration_ms":24940,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["42.50.Ex","03.67.Dd"],"model":"deepseek-v4-flash","headline":"The paper demonstrates a free-space Sagnac interferometer that encodes BB84 polarization states onto single photons at 151.894 MHz with a 0.69(2)% error rate, the lowest reported for high-speed single-photon encoding.","keywords":["single-photon source","quantum dots","Sagnac interferometer","polarization encoding","BB84 protocol","quantum key distribution","quantum bit error rate","telecom C-band"],"falsifier":"Leave the full chain (quantum-dot source, encoder, connection fiber, decoder) running for many hours without touching the fiber polarization controller and compute the single-photon QBER minute by minute: if the QBER drifts substantially above 0.69% (for example above 1%) over the course of a day while the loop-only polarization stays at the reported $10^{-4}$ level, the record-low value reflects the controlled short-term measurement rather than long-term operational encoding. As a second test, widen or remove the 2.65 ns temporal filter: if the QBER degrades markedly when the rejected ~10% of events are included, the error floor depends on discarding mistimed photons, which a real protocol could only do at the cost of efficiency or security.","tokens_in":14395,"feed_emoji":"🔐","tokens_out":17063,"duration_ms":147508,"temperature":0.7,"pith_summary":"The paper reports the first polarization encoder for single-photon qubits built from a free-space Sagnac interferometer, and claims it reaches the lowest error rate reported to date for high-speed information encoding on single photons. Using a quantum dot emitting at 1560.4 nm, deterministically generated photons are modulated into the four BB84 polarization states (the two-basis set used in quantum key distribution) at 151.894 MHz with a quantum bit error rate of 0.69(2)% and an encoding agreement of 96.5(2)% with the ideal states. The central claim is that the Sagnac geometry, in which the two interferometer paths share the same physical optics, removes the environmental phase drift that limits single-pass interferometric encoders, so low error and long-term stability come from the architecture itself rather than from active stabilization. The module runs for 60 hours with output polarization deviations at the $10^{-4}$ level and no temperature or mechanical control. If true, this removes a practical bottleneck for quantum key distribution and other single-photon protocols that need fast, robust state preparation.","feed_headline":"Single-photon encoder hits 0.69% error at 152 MHz","feed_subtitle":"A free-space Sagnac loop writes BB84 states onto quantum-dot photons with the lowest error yet reported.","key_machinery":"The free-space Sagnac interferometer (FSSI): a loop in which both interferometer arms are the same physical path, so clockwise and counterclockwise light accumulate identical phase and the interferometer is intrinsically stable. Its essential elements are a Wollaston polarizer with $10^{6}$ extinction that both creates the two arms and filters the recombined output, a low-$V_\\pi$ phase modulator ($V_\\pi = 4.2$ V) driven by an arbitrary waveform generator synchronized to the excitation laser, and a 3.3 ns delay between the arms so the modulator imprints a different phase on the early (clockwise) and late (counterclockwise) photon components. The combination converts applied voltages into the four BB84 states, and the shared-path geometry means any environmental disturbance affects both components identically, converting potential phase error into loss rather than bit error.","core_discovery":"The central claim is that a free-space Sagnac interferometer, not a single-pass Mach-Zehnder or Michelson setup, is the right carrier for dynamic polarization encoding of deterministic single photons. In this design an input $|D\\rangle$ state is split by a Wollaston polarizer into clockwise and counterclockwise components that traverse the same phase modulator, fibers, and waveplates; a 3.3 ns path delay lets the modulator apply a relative phase $\\phi \\in \\{0, \\pi/2, \\pi, -\\pi/2\\}$ between the early and late components, producing the BB84 states $|D\\rangle$, $|R\\rangle$, $|A\\rangle$, $|L\\rangle$. Because both components see the same optical path in the same polarization, environmental phase fluctuations cancel, and the Wollaston polarizer doubles as a final reject filter with $10^{6}$ extinction, so polarization misalignment inside the loop costs transmission rather than error. The authors demonstrate a mean QBER of 0.69(2)% over a pseudorandom 16-bit sequence (0.96(1)% in the X-basis, 0.43(2)% in the Y-basis) at a system rate of 151.894 MHz, an encoding agreement of 96.5(2)% with the theoretical states, and a 60-hour output stability with average polarization error $4\\times10^{-5}$ without active stabilization, and they state this is the lowest QBER reported to date for high-speed information encoding on single photons.","pith_inferences":["The 60-hour stability test measured the loop alone with a laser and polarimeter, while the 0.69% QBER was a one-minute single-photon run through the connection fiber and decoder after manual optimization of a fiber polarization controller; the natural confirmation test, which the paper suggests but does not perform, is to run the full chain for hours with automated feedback and watch the QBER cont","Because the temporal filter keeps only photons inside a 2.65 ns window and rejects roughly 10% of events, the demonstrated error rate is tied to the fast ~1 ns decay of this particular quantum dot; a slower or jitterier emitter would blur the boundary between early and late components and raise the error rate, so part of the record is a source property, not purely an encoder property.","The paper acknowledges, citing the temporal side-channel study of Sagnac QKD encoders, that timing information can leak information; turning the 0.69% encoding error into a certified secure key rate would require quantifying that leakage for this specific module, which the paper does not attempt."],"forward_implications":["A real BB84 link built on this module could run at roughly a quarter of the error rate used as the lower bound of published single-photon encoding experiments (2.50%), and the paper's simulation shows that improving QBER from 2.50% to 0.69% raises the asymptotic secure key rate by more than 19% at every distance and adds 0.63 dB of tolerable channel loss.","The encoder operates at 151.894 MHz with a total module loss of 5.17(5) dB, and the paper estimates that replacing the fiber-based circulator with a free-space version would recover about 2 dB of that loss.","The architecture is reversible: operating it as an active single-photon decoder would enable QKD protocols with asymmetric basis choice, as the paper notes.","With minor modifications the same Sagnac design can implement phase and time-bin encoding, and all components have integrated-photonics counterparts, making the demonstrated performance a step toward on-chip encoders.","The 60-hour stability record shows flicker-dominated noise with no strong spectral lines, so the module needs no active thermal or mechanical stabilization under ordinary laboratory conditions."],"supporting_citations":[{"why":"Defines the BB84 protocol whose four states in two bases the encoder must produce; the paper's QBER and encoding-agreement figures are measured against this protocol.","marker":"[3]"},{"why":"A recent high-speed single-photon QKD experiment at 2.54% QBER, one of the results the paper compares against in its argument that 0.69% is the lowest reported value.","marker":"[29]"},{"why":"The bright Purcell-enhanced quantum-dot source with a circular Bragg grating cavity that supplies the telecom C-band single photons used in the experiment.","marker":"[32]"},{"why":"Provides the decoder setup that the authors refined and the asymptotic secure-key-rate formula used to project the benefit of the low QBER.","marker":"[36]"},{"why":"A high-speed Sagnac-based polarization modulation scheme for QKD that the paper extends from weak-coherent-pulse sources to deterministic single photons.","marker":"[38]"},{"why":"A stable, low-error, calibration-free polarization encoder in the Sagnac lineage; the paper positions its free-space Sagnac module as the next step of this line.","marker":"[40]"},{"why":"The first application of Sagnac encoding in a single-photon-source QKD experiment, cited as the only prior SPS use of the technique.","marker":"[45]"},{"why":"The study of time-dependent side channels in QKD that the paper cites when acknowledging potential temporal side-channel vulnerabilities in its own encoder.","marker":"[49]"},{"why":"A 2.5% QBER single-pass polarization encoding experiment used as the comparison baseline in the paper's SKR simulation.","marker":"[52]"}],"fun_headline_variants":["Sagnac loop polarizes single photons with record-low error","Ultralow-error photon encoding via Sagnac interferometer","0.69% QBER: Sagnac-based photon encoder sets record","Free-space Sagnac beats MZI for photon qubit encoding","Stable Sagnac encoder lowest-error single-photon encoding"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the reported 0.69% QBER reflects the encoder's intrinsic error floor rather than favorable measurement conditions, since the value was computed from just one minute of data after manually optimizing a fiber polarization controller and applying a 2.65 ns temporal filter that discards about 10% of events, while the 60-hour stability demonstration used a laser and polarimeter directly after the loop rather than single photons through the full encoder-fiber-decoder chain.","fun_headline_variants_meta":{"raw":{"variants":["Sagnac loop polarizes single photons with record-low error","Ultralow-error photon encoding via Sagnac interferometer","0.69% QBER: Sagnac-based photon encoder sets record","Free-space Sagnac beats MZI for photon qubit encoding","Stable Sagnac encoder lowest-error single-photon encoding"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000246,"raw_usage":{"total_tokens":1586,"prompt_tokens":1038,"completion_tokens":548,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":654,"completion_tokens_details":{"reasoning_tokens":458}},"tokens_in":654,"tokens_out":548,"duration_ms":5578,"temperature":1.0,"reasoning_tokens":458,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T15:06:52.145059+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Leave the full chain (quantum-dot source, encoder, connection fiber, decoder) running for many hours without touching the fiber polarization controller and compute the single-photon QBER minute by minute: if the QBER drifts substantially above 0.69% (for example above 1%) over the course of a day while the loop-only polarization stays at the reported $10^{-4}$ level, the record-low value reflects the controlled short-term measurement rather than long-term operational encoding. As a second test, widen or remove the 2.65 ns temporal filter: if the QBER degrades markedly when the rejected ~10% of events are included, the error floor depends on discarding mistimed photons, which a real protocol could only do at the cost of efficiency or security.","supporting_citations":[{"cited_title":"Theoretical Computer Science 560, 7–11 (2014)","cited_arxiv_id":null,"evidence_quote":"Defines the BB84 protocol whose four states in two bases the encoder must produce; the paper's QBER and encoding-agreement figures are measured against this protocol."},{"cited_title":"Physical Review Letters 134(21), 210801 (2025)","cited_arxiv_id":null,"evidence_quote":"A recent high-speed single-photon QKD experiment at 2.54% QBER, one of the results the paper compares against in its argument that 0.69% is the lowest reported value."},{"cited_title":"Advanced Quantum Technologies 6(11), 2300111 (2023)","cited_arxiv_id":null,"evidence_quote":"The bright Purcell-enhanced quantum-dot source with a circular Bragg grating cavity that supplies the telecom C-band single photons used in the experiment."},{"cited_title":"Light: Science & Applications13(1), 150 (2024)","cited_arxiv_id":null,"evidence_quote":"Provides the decoder setup that the authors refined and the asymptotic secure-key-rate formula used to project the benefit of the low QBER."},{"cited_title":"Optics Letters 44(21), 5262–5265 (2019)","cited_arxiv_id":null,"evidence_quote":"A high-speed Sagnac-based polarization modulation scheme for QKD that the paper extends from weak-coherent-pulse sources to deterministic single photons."},{"cited_title":"Optics Letters 45(17), 4706– 4709 (2020)","cited_arxiv_id":null,"evidence_quote":"A stable, low-error, calibration-free polarization encoder in the Sagnac lineage; the paper positions its free-space Sagnac module as the next step of this line."},{"cited_title":"National Science Review 12(8), 147 (2025)","cited_arxiv_id":null,"evidence_quote":"The first application of Sagnac encoding in a single-photon-source QKD experiment, cited as the only prior SPS use of the technique."},{"cited_title":"Physical Review A 111(4), 042622 (2025)","cited_arxiv_id":null,"evidence_quote":"The study of time-dependent side channels in QKD that the paper cites when acknowledging potential temporal side-channel vulnerabilities in its own encoder."},{"cited_title":"Nature 420(6917), 762–762 (2002)","cited_arxiv_id":null,"evidence_quote":"A 2.5% QBER single-pass polarization encoding experiment used as the comparison baseline in the paper's SKR simulation."}],"review_version":1}