{"id":"4890bf60-d98c-4bc9-914a-f80585d47807","arxiv_id":"2412.19604","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A monolithic all-glass spun tapered double-clad fiber amplifier directly amplifies picosecond pulses to 2 MW peak power and 645 W average power across 1 MHz to 1 GHz repetition rates.","lead":"This paper demonstrates a monolithic all-glass tapered fiber amplifier that directly boosts picosecond pulses to record power levels, up to 2 MW peak power at 1 MHz and over 600 W average power at higher repetition rates. It matters because it shows a single, alignment-free fiber stage can replace bulkier multi-stage systems for high-power pulsed lasers.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Peak-power claim rests on assumed 50/20 ps output duration; no autocorrelation at high power is reported.","rationale":"The reader's weakest-assumption analysis identifies exactly the same load-bearing concern: the output pulse duration is assumed, not measured, and the headline peak-power number depends on it. I agree with that assessment. The measured quantities that are reported—average power curves, M2 factors, degree of polarization, and optical spectra—are presented with concrete methods and are internally consistent. The missing temporal characterization is the soft spot, and it is addressable with one additional measurement. Because the reader's verdict is already CONDITIONAL, my stress-test pass does not move the verdict; it reinforces the condition. An additional secondary observation is that even if the pulse duration were preserved, the conversion from average power and FWHM to 'peak power' implicitly assumes a rectangular pulse shape; for a Gaussian pulse the true peak power is about 6% lower than E/(RR × FWHM). This is minor relative to the possible factor-of-two broadening, but it further supports treating the peak-power entries as estimates pending temporal data. The manuscript also contains a small ambiguity in the description of the pump fiber being spliced to the endcap facet, but that does not affect the central numerical claims as directly as the missing autocorrelation.","tokens_in":10483,"tokens_out":4413,"duration_ms":45017,"concrete_test":"Measure the output pulse duration of the 1 MHz channel at maximum power using a background-free intensity autocorrelator, with a 1 nm bandpass filter at 1040 nm to reject ASE before the nonlinear crystal. Record autocorrelation traces at output powers of 3 W, 50 W, 100 W, and 155 W. If the deconvolved FWHM at 155 W exceeds ~60 ps, the peak-power extraction must be revised downward and the 'over 2 MW' claim is unsupported. Repeating the same check at 20 MHz and 1 GHz would validate the pulse-duration labels in Table 1.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's headline 2 MW peak power is a derived quantity, not a measured one. In the 'Amplification of 1 MHz laser' section, the text states: 'The peak power was determined by considering only the signal region, excluding the amplified spontaneous emission contribution from the total output power.' No output pulse-duration measurement (autocorrelation, FROG, or streak camera) appears anywhere in the manuscript. The 2 MW value is therefore computed as P_signal / (RR × τ_seed), where τ_seed is the seed pulse duration. Table 1 lists peak powers of 2 MW, 625 kW, and 32 kW for the three channels, all implicitly assuming that the 50 ps or 20 ps seed duration survives amplification. This is load-bearing because SPM is demonstrably active at 1 MHz: the 3 dB spectral bandwidth grows from 0.45 nm at 3 W to 0.67 nm at 155 W, indicating nonlinear temporal reshaping is plausible. If the actual deconvolved output FWHM at 155 W were 70 ps instead of 50 ps, the true peak power would be about 1.4 MW rather than over 2 MW. The 625 W and 645 W average-power claims do not depend on pulse duration, but the '50 ps' and '20 ps' output labels, and every peak-power entry, do. This is not an internal inconsistency; it is a missing measurement on which a central advertised number depends.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports an experimental study of a monolithic all-glass spun tapered double-clad fiber (sT-DCF) amplifier in a master oscillator-power amplifier configuration. Three seed systems are amplified: 50 ps pulses at 1 MHz to 155 W average power with a claimed peak power above 2 MW, 50 ps pulses at 20 MHz to 625 W average power, and 20 ps pulses at 1 GHz to 645 W average power. For each case the paper reports output power scaling, optical spectra, ASE and Raman levels, beam quality M2, and degree of polarization, and it attributes the 1 MHz performance limits mainly to ASE and polarization modulation instability.","tokens_in":10762,"tokens_out":4339,"duration_ms":40736,"significance":"If the results hold, this is a significant engineering demonstration: a single-stage all-fiber amplifier with no free-space pump optics reaches 625 W and 645 W average power at high repetition rates and MW-level peak power at 1 MHz, with DOP between 70% and 88% and M2 between 1.28 and 2.0. The monolithic construction and the alignment-free kW-level pumping scheme are practical strengths. However, the headline peak-power value is not directly measured but computed from average power and an assumed pulse duration, and none of the quantitative results are accompanied by measurement uncertainties; these points weaken confidence in the exact advertised numbers.","major_comments":[{"comment":"The headline claim of over 2 MW peak power is not based on a direct measurement of the output pulse duration. The text states that the peak power was determined by considering only the signal region and excluding the ASE contribution from the total output power, but no autocorrelation, FROG, or streak-camera measurement is reported anywhere in the manuscript. The value is therefore computed as P_signal/(RR × τ) with τ taken as the 50 ps seed duration. Since Fig. 4c shows spectral broadening from 0.45 nm at 3 W to 0.67 nm at 155 W, nonlinear temporal reshaping cannot be ruled out; if the output FWHM were 70 ps rather than 50 ps, the peak power would drop to about 1.4 MW. The same issue applies to the 20 ps output label for the 1 GHz channel, although the peak power there is much lower. Please add a measured output pulse duration at maximum power, or explicitly relabel the peak-power and pulse-duration claims as estimates that assume unchanged seed duration.","section":"Amplification of 1 MHz laser; Table 1"},{"comment":"No measurement uncertainties or repeated-measurement statistics are provided for any of the reported quantities: average output power, slope efficiency, DOP, M2, spectral bandwidth, or ASE/Raman suppression ratios. DOP and M2 are also single-point characterizations taken at the maximum power only. Because the paper makes quantitative comparisons with previous demonstrations (for example, 1.26 MW peak power and 573 W average power) and reports fine distinctions such as 88.3% versus 87.6% DOP, error bars or multiple measurements are needed to support the precision implied by the claims.","section":"Figures 4-6; Table 1"}],"minor_comments":[{"comment":"The sentence 'In the amplification of 20 ps pulses with 1 MHz RR' should read '50 ps pulses'; the 1 MHz channel is described elsewhere as 50 ps, and Table 1 consistently lists 50 ps for 1 MHz.","section":"Discussion, first paragraph"},{"comment":"The contributions state 'A.G. and E.G. made the taper,' but E.G. is not an author of the manuscript; this is likely a typo for A.G. (Andrey Grishchenko) and should be corrected.","section":"Author contributions"},{"comment":"The claim that 'fast growth of ASE and degradation of DOP limit further power scaling' is not directly documented: DOP is shown only at maximum output power, and no DOP-versus-power curve is presented. Either provide that curve or soften the claim to say that ASE growth was observed.","section":"Amplification of 1 MHz laser; Fig. 4"},{"comment":"The phrase 'at different output power' should be 'at different output powers' in the captions of Figs. 4(e), 5(e), and 6(e).","section":"Captions of Figs. 4, 5, and 6"}],"recommendation":"major_revision","confidential_remarks":"The paper is a useful engineering demonstration and the central average-power results appear credible, but the 2 MW peak-power headline rests on an unmeasured pulse duration. I do not see grounds for rejection; a revision that adds output temporal characterization (or clearly re-labels the estimate) plus uncertainty statements would make the claims solid. The missing autocorrelation is a standard and fixable measurement for a fiber amplifier group."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The main thing you should know: this is a solid experimental report of a genuinely new amplifier configuration. The authors take the spun tapered double-clad fiber that previously required free-space pump injection and make it fully monolithic with dual-sided fused pumping. That alone is a practical step forward, and the numbers are strong: 645 W average power at 1 GHz, 625 W at 20 MHz, and 155 W at 1 MHz with near-diffraction-limited beam quality and decent polarization. The average power figures are direct measurements; the spectra, M2 curves, and DOP traces are internally consistent, and the absence of SRS is believable given the taper geometry. I have no reason to doubt the main engineering claim.\n\nThe soft spot is the one the stress-test flags: the 2 MW peak power is not measured. It is computed from average power, repetition rate, and the seed pulse duration, with no output autocorrelation or streak measurement anywhere in the text. The authors say only that they excluded ASE from the signal power. Given that the spectral bandwidth grows from 0.45 nm to 0.67 nm at 1 MHz, some nonlinear temporal reshaping is plausible, so the true peak could be meaningfully lower. This is not a fatal flaw, but it is load-bearing for the headline number. The fix is straightforward: one autocorrelation trace at high power, or a clear statement that the pulse duration is assumed unchanged.\n\nMinor issues: no error bars on any power, DOP, or M2 measurement; the comparison to prior monolithic side-pumped designs is qualitative only; and the M2 and DOP values are single-point characterizations rather than stability studies. None of these change the core result, but they weaken the quantitative claims more than necessary.\n\nWho is this for? Anyone working on high-power pulsed fiber amplifiers, especially tapered-fiber and all-fiber architectures. The paper is worth a serious referee: the engineering advance is real, the data volume is decent, and the missing temporal measurement is eminently addressable. I would send it to peer review with a request for the autocorrelation and error bars before publication.","headline":"A real engineering advance in monolithic tapered-fiber amplifiers, but the 2 MW peak-power headline rests on an assumed pulse duration that is never measured.","tokens_in":11289,"tokens_out":793,"would_cite":true,"duration_ms":9493,"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":"A single monolithic all-glass fiber amplifier delivers 2 MW peak power at 1 MHz, 625 W at 20 MHz, and 645 W at 1 GHz.","keywords":["fiber amplifier","tapered double-clad fiber","spun fiber","monolithic all-glass","picosecond pulses","high peak power","high average power","polarization maintenance"],"falsifier":"Measure the output pulse duration at the maximum power of each system using a background-free autocorrelator or a fast streak camera; if the pulses broaden substantially (e.g., from 50 ps to 100 ps), the true peak power would be a factor of two or more below the stated 2 MW, 625 kW, and 32 kW values.","tokens_in":10327,"feed_emoji":"🔬","tokens_out":2889,"duration_ms":35881,"temperature":0.7,"pith_summary":"The paper reports a single-stage fiber amplifier built from a spun tapered double-clad fiber that directly amplifies narrow-linewidth picosecond pulses from a few tens of milliwatts to hundreds of watts of average power and megawatt-level peak power, without any free-space optics. Using the same amplifier module, the authors demonstrate 50 ps pulses reaching over 2 MW peak power at 1 MHz, 625 W average power at 20 MHz, and 20 ps pulses reaching 645 W average power at 1 GHz, with near-diffraction-limited beam quality at high average power and a high degree of linear polarization throughout. The significance is that one alignment-free, monolithic design can cover a wide range of repetition rates and power levels, overcoming the usual trade-off between nonlinear effects and transverse mode instabilities.","feed_headline":"One all-glass fiber hits 2 MW peak and 645 W average","feed_subtitle":"Spun tapered fiber amplifies picosecond pulses from 1 MHz to 1 GHz without free-space optics.","key_machinery":"The central object is the spun tapered double-clad fiber (sT-DCF), whose core and cladding diameters grow along its length from 8.3/75/90 µm to 90/814/977 µm over about 6.7 m. Tapering enlarges the mode area gradually, keeping the mode content stable and raising the thresholds for stimulated Raman scattering and other nonlinear effects, while spinning the preform during drawing reduces intrinsic birefringence to roughly $10^{-8}$, making the polarization state insensitive to heating. The fiber is pumped from both ends in a monolithic configuration, with a backward 976 nm pump of up to 840 W and a forward 915 nm pump of 18 W, enabling alignment-free high-power operation.","core_discovery":"A monolithic all-glass spun tapered double-clad fiber (sT-DCF) amplifier can amplify narrow-linewidth picosecond pulses in a single stage across three distinct repetition-rate regimes while maintaining high beam quality and polarization stability. The 1 MHz system produced 50 ps pulses with over 2 MW peak power (155 W average power) at a slope efficiency of 59% and a degree of polarization of 70%; the 20 MHz system produced 50 ps pulses with 625 W average power at 76.6% slope efficiency and 88.3% DOP; the 1 GHz system produced 20 ps pulses with 645 W average power at 78.6% slope efficiency and 87.6% DOP. The design eliminates free-space pump coupling by splicing a pump combiner directly to the fiber, removing alignment dependence and internal back reflections.","pith_inferences":["The technique could be combined with post-compression or coherent beam combining to further boost peak power, since the amplifier already delivers a clean, near-diffraction-limited beam at high average power.","The polarization instability at high peak power (attributed to polarization modulation instability) may be mitigated by slightly increasing fiber birefringence, which could extend the peak-power ceiling at low repetition rates.","If the pumped power is scaled further, the same fiber design may reach kilowatt-class average power in the 20 MHz and 1 GHz regimes, with the main remaining questions being thermal handling and SRS onset.","The monolithic all-glass approach could transfer to other rare-earth-doped fibers or wavelengths, potentially enabling high-power pulsed amplifiers beyond the 1 µm ytterbium band."],"forward_implications":["The same monolithic sT-DCF amplifier can serve applications across a wide range of repetition rates, from high-peak-power regimes at 1 MHz to high-average-power regimes at 1 GHz, without reconfiguration.","Because the 20 MHz and 1 GHz systems were limited only by available pump power, adding more pump diodes should scale average power beyond 625 W and 645 W while retaining beam quality and polarization.","Eliminating free-space pump alignment improves reliability and compactness, making the amplifier attractive for industrial and scientific laser systems where alignment drift causes downtime.","A single amplifier stage can replace multi-stage or free-space amplifier chains for picosecond pulses, reducing system complexity and footprint.","The demonstrated combination of high average power, high peak power, and polarization maintenance is directly relevant to material processing, high-harmonic generation, and burst-mode laser systems."],"supporting_citations":[{"why":"Introduces the tapered double-clad fiber concept that underlies the amplifier's nonlinearity suppression and mode-content preservation.","marker":"[18]"},{"why":"Establishes the spun tapered double-clad fiber design and its low-birefringence polarization maintenance mechanism used here.","marker":"[21]"},{"why":"Shows an alternative polarization-maintaining approach (stress rods) in tapered fibers, providing the comparison baseline for the spun design.","marker":"[19]"},{"why":"Demonstrates active tapered double-clad fiber with low birefringence, supporting the spun fiber's polarization stability claims.","marker":"[20]"},{"why":"Reports a previous 1.26 MW peak power result using a PT-DCF, setting the peak-power benchmark this work surpasses with a monolithic design.","marker":"[24]"},{"why":"Reports a previous 573 W average power result in sT-DCF, the baseline for the average-power scaling achieved here.","marker":"[25]"},{"why":"Demonstrates a monolithic side-pumped tapered fiber amplifier, the direct precursor to the all-glass monolithic pumping scheme used in this paper.","marker":"[27]"},{"why":"Explains polarization modulation instability in weakly birefringent fibers, which the authors invoke to account for the spectral sidebands and beam-degradation at high peak power.","marker":"[29]"}],"fun_headline_variants":["All-glass fiber amplifies pulses to 2 MW peak, 645 W avg","Single all-glass amplifier delivers MW peak and hundreds of watts","No free-space optics: fiber achieves 2 MW peak, 645 W average","Fiber amplifier spans 1 MHz to 1 GHz with MW peak power","Monolithic all-glass amplifier: 2 MW peak, 645 W average"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The claimed peak powers assume that the seed pulse duration (50 ps or 20 ps) stays unchanged through high-power amplification, because only average power and spectra were measured at maximum output, with no direct autocorrelation or temporal characterization.","fun_headline_variants_meta":{"raw":{"variants":["All-glass fiber amplifies pulses to 2 MW peak, 645 W avg","Single all-glass amplifier delivers MW peak and hundreds of watts","No free-space optics: fiber achieves 2 MW peak, 645 W average","Fiber amplifier spans 1 MHz to 1 GHz with MW peak power","Monolithic all-glass amplifier: 2 MW peak, 645 W average"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000186,"raw_usage":{"total_tokens":1301,"prompt_tokens":896,"completion_tokens":405,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":512,"completion_tokens_details":{"reasoning_tokens":306}},"tokens_in":512,"tokens_out":405,"duration_ms":20984,"temperature":1.0,"reasoning_tokens":306,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T00:09:56.365203+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the output pulse duration at the maximum power of each system using a background-free autocorrelator or a fast streak camera; if the pulses broaden substantially (e.g., from 50 ps to 100 ps), the true peak power would be a factor of two or more below the stated 2 MW, 625 kW, and 32 kW values.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes the spun tapered double-clad fiber design and its low-birefringence polarization maintenance mechanism used here."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Shows an alternative polarization-maintaining approach (stress rods) in tapered fibers, providing the comparison baseline for the spun design."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Demonstrates active tapered double-clad fiber with low birefringence, supporting the spun fiber's polarization stability claims."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Reports a previous 1.26 MW peak power result using a PT-DCF, setting the peak-power benchmark this work surpasses with a monolithic design."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Reports a previous 573 W average power result in sT-DCF, the baseline for the average-power scaling achieved here."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Demonstrates a monolithic side-pumped tapered fiber amplifier, the direct precursor to the all-glass monolithic pumping scheme used in this paper."},{"cited_title":"G., Leonhardt, R","cited_arxiv_id":null,"evidence_quote":"Explains polarization modulation instability in weakly birefringent fibers, which the authors invoke to account for the spectral sidebands and beam-degradation at high peak power."}],"review_version":1}