{"id":"757a87ad-e559-4470-a6fd-ca45dda178ce","arxiv_id":"2505.01852","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"A record 2 kW laser delivery over 2.45 km through a hollow-core anti-resonant fiber with 0.175 dB/km loss and no silica Raman scattering was demonstrated using all-fiber splicing and Raman noise suppression.","lead":"This paper reports an all-fiber system that delivers a 2 kilowatt laser through 2.45 kilometers of hollow-core fiber with 85.3 percent transmission efficiency. It sets records for transmission loss at 1 micron and identifies stimulated Raman scattering inside the fiber's silica tubes as the key nonlinear limit.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"SRS amplification ratio reversal (2.98→1.28) contradicts the constant-gain model used to derive γ_R and to construct Fig. 6, leaving the no-SRS margin unquantified.","rationale":"The central experimental demonstration—2 kW delivery over 2.45 km at 85.3% efficiency with no observed SRS after CTFBG filtering—is credible and directly supported by the reported output spectra, power measurements, and cutback loss. I do not see an internal inconsistency that overturns the headline result. However, the conditionality of the paper is warranted. The reader focused on the fitted 3 μm fusion offset; my concern is more fundamental. The measured SRS amplification ratio falls from 2.98 at 2259 W to 1.28 at 2400 W, which contradicts the constant-γ exponential model used to extract γ_R and to generate the scalability predictions. The stated explanation via differential attenuation is insufficient because that attenuation is power-independent and tiny over 0.2 km. If the ratio reversal is real, then γ_R derived from the 2259 W point is not a material constant, the inferred 3 μm offset is not trustworthy, and the Fig. 6 limits are unreliable. This does not change the verdict of CONDITIONAL, but it sharpens the condition: the authors should either explain the reversal with a quantitative model or present measurements that resolve it. Because the no-SRS observation at 2.45 km is direct spectral evidence, the central claim remains intact; hence UNCHANGED.","tokens_in":11747,"tokens_out":25619,"duration_ms":253507,"concrete_test":"Re-measure the 2 m vs 200 m SRS ratio at 2259, 2315, and 2400 W input power using a calibrated, unsaturated OSA (or spectrometer with verified neutral-density linearity), and independently record the Raman seed power launched into the fiber. If the ratio still falls with increasing pump power, the constant-γ model must be rejected and Fig. 6 rebuilt without it. If the ratio rises to ~3, the reported reversal is a measurement artifact and the derived γ is reliable.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper derives the effective Raman gain coefficient γ_R^HCF = 2.4 km^-1·kW^-1 from Eq. (4) using a measured 200 m/2 m SRS amplification of 2.98 at 2259 W (Methods 3a). The same measurements at 2315 W and 2400 W give amplification ratios of only 1.33 and 1.28. This is not a small perturbation: the constant-γ exponential model predicts the ratio should increase modestly with pump power (≈2.9, 3.0, 3.1 at these powers), not drop by more than half over a 56 W increase. The paper attributes the decrease to differential attenuation (0.175 vs 0.557 dB/km), but attenuation is power-independent and over 0.2 km accounts for only ~2.5% Raman loss, far too small to explain the observed reversal. Unless the spectra saturated or the seed power changed between points, the measured γ is not a constant, and the model underlying the fitted 3 μm offset and the Fig. 6 scalability curves is invalid. Because the paper uses the same model to argue that the 2.45 km, 2 kW point is safely below the SRS threshold, the quantitative content of the 'without observable SRS' claim rests on a model that fails at the upper end of the same power range.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports an all-fiber delivery system for a 1080 nm continuous-wave laser through a self-fabricated anti-resonant hollow-core fiber (AR-HCF). The headline results are 2 kW delivered over 2.45 km with 85.3% transmission efficiency, a cut-back loss of 0.175 dB/km at 1080 nm, and fusion splicing of AR-HCF to anti-reflection-coated solid-core fiber with splice loss below 0.2 dB. The authors also report the first observation of stimulated Raman scattering (SRS) amplified within the silica nested tubes of the AR-HCF, attribute this SRS to Raman noise from the laser seed, and show that a chirped and tilted Bragg grating suppresses the Raman component. A scalability model based on an effective Raman gain coefficient is used to predict power-distance limits for larger-core fibers. The central experimental demonstration is internally consistent: cut-back loss, delivered power, and output spectra agree with each other, and the beam quality M² ≈ 1.3 is preserved after both 1 km and 2.45 km.","tokens_in":12002,"tokens_out":8448,"duration_ms":80939,"significance":"If the results hold, this is a substantial advance in high-power laser delivery: a factor of roughly 500 improvement in power-distance product over prior all-fiber AR-HCF demonstrations, a record low loss at 1 μm, and a practical all-fiber architecture with a protected output end cap. The paper's strengths include the cut-back loss measurement over an 8.65 km length, the directly measured power-delivery values, a two-hour stability test, and the systematic use of a CTFBG to separate source-generated Raman noise from intrinsic fiber nonlinearity. These are concrete, checkable experimental results. The main weakness is the modeling of the effective Raman gain coefficient: the extraction uses a single measurement point and is inconsistent with the other two measurements at nearby powers. This modeling underpins the Fig. 6 scalability projections and the quantitative 'no SRS' margin, so those parts need substantial revision. The core delivery demonstration, however, is not dependent on the Raman model.","major_comments":[{"comment":"The effective Raman gain coefficient γ_R^HCF ≈ 2.4 km⁻¹·kW⁻¹ is derived from a single measured amplification ratio G = 2.98 at 2259 W, but the same experiment gives G = 1.33 at 2315 W and G = 1.28 at 2400 W. Under the constant-γ exponential model used, the ratio R(200 m/2 m) = exp[γ_R P (L_eff(200 m) − L_eff(2 m)) − α_R(198 m)] is predicted to increase modestly with P (from ≈2.6 at 2259 W to ≈2.8 at 2400 W for γ_R = 2.4 km⁻¹·kW⁻¹), not to drop by more than half. The paper attributes the reversal to differential attenuation (α_R = 0.557 dB/km vs α_S = 0.175 dB/km), but over 198 m this contributes only a factor exp(−0.110) ≈ 0.90, far too small to reverse the trend. The measured γ_R is therefore not a constant, and the model used to infer the 3 μm offset and to construct Fig. 6 is not validated at the upper end of the power range. Please provide a self-consistent extraction with error bars, or explicitly discuss saturation, pump-depletion, or detector nonlinearity effects. Without this, the no-SRS margin and the quantitative scalability claims are unsupported.","section":"§4, Methods 3a, Fig. 4b, Eqs. (4)–(7)"},{"comment":"The 3 μm fusion offset is obtained by matching the simulated effective Raman coefficient (2.55 km⁻¹·kW⁻¹) to the measured value (2.4 km⁻¹·kW⁻¹) at a single power. This is effectively a one-parameter calibration with no stated uncertainty, and the same fitted model is then used to generate the power-distance curves in Fig. 6. The sensitivity of the predicted limits (2.8 km for a 25 μm core and 6.82 km for a 30 μm core at 2 kW) to the assumed offset should be quantified, for example by showing curves for offsets of 2, 3, and 4 μm. If possible, the offset should also be verified independently from microscope images of the splice or from mode-field measurements, rather than inferred solely from the Raman gain fit.","section":"§4, Fig. 4c, Discussion"},{"comment":"The text states that the 1 km (2160 W) and 2.45 km (2050 W) power-delivery results imply a fiber transmission loss of 0.172 dB/km. However, assuming identical input coupling loss in the two builds, the ratio gives 10·log10(2160/2050)/1.45 ≈ 0.157 dB/km, not 0.172 dB/km. Please reconcile this discrepancy or state explicitly which additional loss terms (for example, different splice losses in the two systems) are included in the 0.172 dB/km value. The cut-back value of 0.175 dB/km is the primary loss claim, but the text's use of the power-delivery data to derive loss and coupling loss should be internally consistent.","section":"§5, Fig. 5a"},{"comment":"Key quantitative claims are reported without measurement uncertainty or repeated-measurement information: the 0.175 dB/km cut-back loss, the 85.3% transmission efficiency, the Raman amplification ratios (2.98, 1.33, 1.28), and the derived γ_R = 2.4 km⁻¹·kW⁻¹. The factor-of-two variation in the Raman ratios at 2315 and 2400 W relative to 2259 W may be partly a measurement artifact, but without error bars this cannot be assessed. Please provide estimates of measurement precision for these central numbers, either from repeated measurements or from instrument specifications.","section":"§5 and Methods 3a"}],"minor_comments":[{"comment":"The phrase 'Raman amplification of 200 m AR-HCF relative to 2 m AR-HCF' should be defined explicitly as the ratio of integrated Raman spectral intensity, since Fig. 4b shows normalized spectra; this will help readers connect the measurement to Eq. (4).","section":"§4, Fig. 4b"},{"comment":"The 'minor spectral redshift' attributed to the Raman response of atmospheric air within the core is implausible as stated: the vibrational Raman shifts of N₂ (~2331 cm⁻¹) and O₂ (~1556 cm⁻¹) would produce discrete Stokes lines well separated from 1080 nm, not a small continuous redshift. Please clarify what is meant or find another explanation.","section":"§5, Fig. 5b"},{"comment":"Eq. (7) appears to duplicate Eq. (4) with an explicit attenuation term, but it is unclear whether the α in Eq. (7) is α_S or α_R and how L_eff is used. Please state the conventions explicitly and check the units consistently.","section":"Methods 3a, Eqs. (4)–(7)"},{"comment":"The data availability statement says the data are not publicly available. Given that the central quantitative claims are a record loss and a record power-distance product, depositing the raw cut-back, power-delivery, and spectral data in a public repository would greatly strengthen the paper's verifiability.","section":"Data Availability"},{"comment":"Minor presentation issues: 'the practical maps' should be 'photographs'; in Methods 3b, 'According to the analysis.' is an incomplete sentence; and References 15 and 32 are the same paper and should be unified.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"The experimental demonstration is strong and likely to be of high interest to the optics community. My main concern is that the Raman-gain modeling, which supports both the 'without SRS' statement and the Fig. 6 scalability curves, is internally inconsistent at the powers adjacent to the single calibration point. This is fixable: the authors could re-analyze the data with a self-consistent model, add uncertainty estimates, or soften the scalability claims to an empirical observation. I therefore recommend major revision rather than rejection; the central delivery result itself appears sound and should survive revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The core result is solid and significant: all-fiber delivery of 2 kW at 85.3% efficiency over 2.45 km of self-fabricated hollow-core fiber, with a record low loss of 0.175 dB/km at 1080 nm. That is a genuine leap over prior all-fiber systems, and the transmission numbers are internally consistent—0.172 dB/km derived from delivery versus 0.175 dB/km by cutback is believable. The fusion splicing of AR-HCF to anti-reflection-coated SCF with <0.2 dB loss and −28.7 dB return loss is also a real engineering advance. And the first observation of SRS amplified inside the silica nested tubes, plus its suppression with a CTFBG, is a new and useful piece of physics, even if the underlying SRS mechanism is familiar. Credit where due: the loss measurement, the delivery demonstration, and the stability data (2-hour 2.3% fluctuation) all look credible.\n\nThe soft spot the stress-test note flags is real. The paper derives γ_R^HCF = 2.4 km⁻¹·kW⁻¹ from a measured SRS amplification of 2.98 at 2259 W, but the same experiment at 2315 W and 2400 W gives ratios of only 1.33 and 1.28. That drop is far too large to be explained by differential attenuation (0.175 vs 0.557 dB/km over 0.2 km is ~2.5% effect). The paper's explanation in terms of attenuation does not work. Either the seed changed, the spectra saturated, or the constant-gain model is not valid at these powers. This matters for the scalability projections in Fig. 6, which are built on that fitted γ_R and on the inferred 3 μm fusion offset. Those projections are speculative anyway, and the paper would be better if it flagged that.\n\nThat said, the central claim—2 kW over 2.45 km without observable SRS in the final system—does not rest on the γ_R model. The output spectra shown in Fig. 5b directly display no SRS peak. The problematic measurements are from the 200 m SRS characterization without the CTFBG, a separate diagnostic. So the stress-test concern undercuts the model, not the headline result.\n\nOther soft spots: no error bars on loss or efficiency, and the data are not public. The paper also states the coupling loss rose from 0.2 to 0.27–0.29 dB at high power without explaining why. These should be addressed, but they are revision-level issues.\n\nWho is this for: anyone working on high-power laser delivery, hollow-core fiber fabrication, or nonlinear effects in anti-resonant fibers. It deserves a serious referee—the result is important enough that the modeling inconsistency should be fixed in review, not used to desk-reject. My recommendation: send it to peer review, with the SRS reversal question going to the authors.","headline":"A credible experimental milestone—2 kW all-fiber delivery over 2.45 km with record-low 0.175 dB/km loss—but the SRS model used for scalability has a real unexplained inconsistency.","tokens_in":12653,"tokens_out":1489,"would_cite":true,"duration_ms":16105,"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":"An all-fiber hollow-core link delivers 2 kW over 2.45 km at 85.3% efficiency, a nearly 500-fold jump in power-distance product over prior all-fiber systems.","keywords":["hollow-core fiber","anti-resonant fiber","high-power laser delivery","fusion splicing","stimulated Raman scattering","2 kW laser","low transmission loss","all-fiber laser system"],"falsifier":"Run the 2.45 km delivery at 2400 W with the chirped-and-tilted Bragg grating removed and monitor the output at 1140 nm; the paper's mechanism predicts the Stokes peak returns. Separately, image or tomographically measure the lateral offset at the fusion splice: if it is away from roughly 3 µm, the simulated effective Raman gain of $2.55\\,\\mathrm{km}^{-1}\\mathrm{kW}^{-1}$ can no longer be reconciled with the measured $2.4\\,\\mathrm{km}^{-1}\\mathrm{kW}^{-1}$, and the Fig. 6 projections lose their basis.","tokens_in":11534,"feed_emoji":"🔦","tokens_out":10284,"duration_ms":96990,"temperature":0.7,"pith_summary":"An all-fiber line built from a self-fabricated anti-resonant hollow-core fiber carried 2 kW of 1080 nm laser light over 2.45 km with 85.3% transmission efficiency, about 500 times the power–distance product of the previous all-fiber hollow-core delivery system. The fiber achieves a loss of 0.175 dB/km at 1080 nm, the lowest reported in this band, and the whole chain—solid-core launch fiber, fusion splice, hollow-core span, output end cap—is spliced rather than free-space coupled, which is what makes it stable in moving or vibrating environments. The paper also reports the first observation of stimulated Raman scattering amplified inside the silica nested tubes of an anti-resonant fiber; filtering the laser source's Raman noise with a chirped-and-tilted Bragg grating suppresses this effect, letting the system reach 2 kW without silica Raman scattering. If these results hold, multi-kilometer, multi-kilowatt laser delivery becomes practical for factory floors, nuclear decommissioning, and deep drilling.","feed_headline":"2 kW laser travels 2.45 km in hollow-core fiber at 85.3% efficiency","feed_subtitle":"All-spliced chain with record 0.175 dB/km loss; filtering source Raman noise unlocks kilometer-scale 2 kW delivery.","key_machinery":"The central object is a five-element triple-nested anti-resonant hollow-core fiber: a 28 µm air core surrounded by five capillary elements, each holding three nested silica tubes, with walls 1.3 µm thick, so that 1080 nm light is confined by anti-resonant reflection and overlaps silica only weakly. The argument is carried by that fiber plus two enabling components: an anti-reflection-coated fusion splice between the 20/250 µm solid-core launch fiber and the AR-HCF (splice loss below 0.2 dB, return loss below −28.7 dB), and a chirped-and-tilted Bragg grating after the laser source that removes the seed Raman noise. The load-bearing numbers are the 0.175 dB/km cutback loss at 1080 nm, the measured effective Raman gain coefficient $2.4\\,\\mathrm{km}^{-1}\\mathrm{kW}^{-1}$, and the SRS critical-power relation used to project how far the approach can scale.","core_discovery":"On its own terms, the paper establishes that an all-fiber chain built around a five-element triple-nested anti-resonant hollow-core fiber transmits a 1080 nm laser at 2400 W input over 2.45 km with 2050 W output (85.3% efficiency), and over 1 km with 89.9% efficiency. The authors attribute the result to four ingredients: a fiber loss of 0.175 dB/km at 1080 nm, the lowest reported in the 1 µm band; fusion splices to anti-reflection-coated solid-core fiber with loss below 0.2 dB and return loss below −28.7 dB; suppression of source Raman noise by a chirped-and-tilted Bragg grating so that the silica nested tubes no longer amplify Stokes light; and a protective output end cap. From the observed 2.98 times Raman growth over 200 m, they extract an effective Raman gain of $2.4\\,\\mathrm{km}^{-1}\\mathrm{kW}^{-1}$ for the AR-HCF, match it with simulations at a roughly 3 µm splice offset, and use the SRS critical power to project transmission distances of 2.8 km (25 µm core) and 6.82 km (30 µm core) at 2 kW output, with 10 kW requiring core diameters above 35 µm.","pith_inferences":["Editorial inference: because the Raman Stokes line is attenuated faster than the signal, the all-fiber design may be most vulnerable to SRS at intermediate lengths rather than at the longest lengths, so splice-offset control and Raman filtering should be optimized jointly with target distance.","Editorial inference: the same spliced hollow-core platform should extend to single-frequency delivery limited by stimulated Brillouin scattering, since air-core guidance sharply reduces the acousto-optic overlap that creates Brillouin gain.","Editorial inference: directly measuring the fusion-splice offset, rather than inferring it from Raman gain, would turn the Fig. 6 scalability curves into a reliable design tool; offsets below 3 µm would push the predicted maximum distances upward.","Editorial inference: replacing the two-mode 20/250 µm solid-core launch fiber with a fiber that excites mainly LP01 would reduce the field overlap with the nested tubes and could raise the power ceiling independently of the grating."],"forward_implications":["After 2.45 km, 2050 W reaches the output with a beam quality factor of 1.29, so kilometer-scale multi-kilowatt delivery preserves beam quality.","All-fiber construction removes the free-space coupling optics that drift and heat in industrial settings; the 1 km end-cap system held 2.3% power fluctuation over a 2-hour test.","Filtering the source's Raman noise with the chirped-and-tilted Bragg grating is sufficient to avoid silica SRS in AR-HCF; without it, 2400 W produces a Stokes peak near 1140 nm.","At 2 kW output, the scaling simulations give roughly 2.8 km for a 25 µm core and 6.82 km for a 30 µm core; reaching 10 kW over 1 km requires core diameters above 35 µm.","Because the 1140 nm Stokes light attenuates at 0.557 dB/km while the 1080 nm signal attenuates at 0.175 dB/km, long fibers lose Raman light faster than signal light, an asymmetry favorable to long-distance delivery."],"supporting_citations":[{"why":"Provides the previous best AR-HCF loss in the 1 µm band (0.3 dB/km with five nested tubes) that the new fiber's design extends to 0.175 dB/km.","marker":"24"},{"why":"Demonstrates 1 kW over 1 km in hollow-core fiber with spatial coupling, the kilometer-scale high-power baseline this all-fiber system surpasses.","marker":"25"},{"why":"Reports 2.2 kW narrow-linewidth delivery over 100 m, one of the high-power AR-HCF demonstrations that motivate and bracket the new result.","marker":"26"},{"why":"Reports 3 kW multi-mode delivery over 100 m and the spatial-coupling thermal issues that this work's fusion splicing removes.","marker":"27"},{"why":"Previous all-fiber AR-HCF delivery, 100 W over 100 m, the direct benchmark for the nearly 500-fold power-distance improvement.","marker":"31"},{"why":"Establishes the nested antiresonant nodeless fiber design principle used to choose the triple-nested structure.","marker":"32"},{"why":"Supplies the anti-reflection coating fusion-splicing method that enables the low-loss, low-return splice.","marker":"35"},{"why":"Gives the SRS gain and critical-power formulas used for the effective Raman gain calculation and the scalability projections.","marker":"36"}],"fun_headline_variants":["All-fiber 2 kW laser: 2.45 km delivery at 85.3% efficiency","2 kW laser delivered 2.45 km via all-fiber hollow-core link","Record all-fiber laser delivery: 2 kW over 2.45 km at 85.3%","Hollow-core fiber sends 2 kW laser 2.45 km with 85.3% efficiency"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The scalability curves rest on the assumption that the production splice has about a 3 µm lateral offset between the solid-core and hollow-core fibers, an offset inferred by matching theory to a single measured Raman-gain point rather than measured directly.","fun_headline_variants_meta":{"raw":{"variants":["All-fiber 2 kW laser: 2.45 km delivery at 85.3% efficiency","2 kW laser delivered 2.45 km via all-fiber hollow-core link","Record all-fiber laser delivery: 2 kW over 2.45 km at 85.3%","Hollow-core fiber sends 2 kW laser 2.45 km with 85.3% efficiency"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000703,"raw_usage":{"total_tokens":3231,"prompt_tokens":1066,"completion_tokens":2165,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":682,"completion_tokens_details":{"reasoning_tokens":2062}},"tokens_in":682,"tokens_out":2165,"duration_ms":16050,"temperature":1.0,"reasoning_tokens":2062,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T04:09:21.284398+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the 2.45 km delivery at 2400 W with the chirped-and-tilted Bragg grating removed and monitor the output at 1140 nm; the paper's mechanism predicts the Stokes peak returns. Separately, image or tomographically measure the lateral offset at the fusion splice: if it is away from roughly 3 µm, the simulated effective Raman gain of $2.55\\,\\mathrm{km}^{-1}\\mathrm{kW}^{-1}$ can no longer be reconciled with the measured $2.4\\,\\mathrm{km}^{-1}\\mathrm{kW}^{-1}$, and the Fig. 6 projections lose their basis.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the previous best AR-HCF loss in the 1 µm band (0.3 dB/km with five nested tubes) that the new fiber's design extends to 0.175 dB/km."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Demonstrates 1 kW over 1 km in hollow-core fiber with spatial coupling, the kilometer-scale high-power baseline this all-fiber system surpasses."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Reports 2.2 kW narrow-linewidth delivery over 100 m, one of the high-power AR-HCF demonstrations that motivate and bracket the new result."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Reports 3 kW multi-mode delivery over 100 m and the spatial-coupling thermal issues that this work's fusion splicing removes."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Previous all-fiber AR-HCF delivery, 100 W over 100 m, the direct benchmark for the nearly 500-fold power-distance improvement."},{"cited_title":"Nested antiresonant nodeless hollow core fiber","cited_arxiv_id":null,"evidence_quote":"Establishes the nested antiresonant nodeless fiber design principle used to choose the triple-nested structure."},{"cited_title":"Y., Yu, R","cited_arxiv_id":null,"evidence_quote":"Supplies the anti-reflection coating fusion-splicing method that enables the low-loss, low-return splice."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the SRS gain and critical-power formulas used for the effective Raman gain calculation and the scalability projections."}],"review_version":1}