{"id":"278dc1d8-1e2b-43a0-b87f-6ae5eee2c522","arxiv_id":"2501.07395","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"A direct TCSPC measurement gives the strontium 5s5p 1P1 lifetime as 5.216 ns, agreeing with the tune-out method and providing a new anchor for clock BBR shift corrections.","lead":"Researchers measured how long an excited strontium atom stays in a specific energy state: about 5.216 nanoseconds, with a total uncertainty near 0.25 percent. The number supports strontium atomic clocks, whose accuracy depends on exactly this type of atomic data.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Reported 5.216 ns is not corrected for two known sign-definite systematics (TAC nonlinearity and radiation trapping); applying both shifts the central value by ~0.015 ns, comparable to the quoted uncertainty.","rationale":"The reader's stated weakest assumption concerns the temporal shape of the background in Eq. (2). I see that risk, but it is substantially mitigated by the truncation procedure: the background tail is reported to last about 7.5 ns, and the fit starts at 7.6-10.6 ns, so a time-dependent background distortion in the first few nanoseconds is largely excluded from the fitted region. The scalar factor p then only needs to account for a slowly varying background, which is a reasonable first-order model supported by the before/after background measurements. I therefore do not regard the background-shape assumption as the single most load-bearing concern. Instead, the more concrete and internally verifiable problem is that two known, sign-definite systematic effects are quantified but not applied to the central value. The TAC nonlinearity section explicitly says the true lifetime is shorter, and the radiation-trapping section explicitly reports a zero-density lifetime of 5.211 ns, yet Eq. (3) reports the in-cell, TAC-uncorrected value 5.216 ns. Because both corrections are negative and together are comparable to the quoted systematic uncertainty, the central value as stated is not the best estimate of the atomic lifetime. This does not necessarily invalidate the experiment: the raw data are available, the eight datasets are independent, and the error budget is otherwise careful. But the paper should either correct the central value or unambiguously state that Eq. (3) is an uncorrected in-cell value with the corrections included only as uncertainties. Since the reader's verdict is already CONDITIONAL and this concern reinforces that condition without making the result unrecoverable, I recommend no change to the verdict.","tokens_in":12094,"tokens_out":7600,"duration_ms":77155,"concrete_test":"Using the published raw data, redo the weighted-mean analysis of the eight datasets in two passes: (1) apply the day-specific TAC time-base calibration as a multiplicative correction to each bin before fitting (or, if day-specific slopes were not recorded, use the maximal measured slope 1.00240 as a bounding case), and (2) replace the in-cell value with the zero-density extrapolation from the OD linear fit in Sec. V.E, i.e. subtract 0.096% (0.005 ns) and propagate the 0.002 ns fit uncertainty. Compare the resulting central value and total uncertainty with Eq. (3). If the shift exceeds ~0.013 ns, the paper must report the corrected value or explicitly label Eq. (3) as the uncorrected in-cell value; if the shift is smaller, the presentation is still ambiguous and should state which quantity is being reported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The most load-bearing problem is internal: the number quoted in Eq. (3), tau = 5.216 +/- 0.006 +/- 0.013 ns, is not the corrected lifetime the text describes. In Sec. V.B the authors find the TAC has a slope up to 1.00240 and state that 'the actual lifetime is shorter,' yet they only fold this into the uncertainty (0.2%) instead of applying the known-sign correction to the central value. In Sec. V.E the linear OD fit gives an extrapolated zero-density lifetime of 5.211 +/- 0.002 ns, a 0.096% correction relative to the 5.216 ns measured at OD=0.0011, but Eq. (3) retains 5.216 ns. Both corrections are negative, so the reported central value is biased high by roughly 0.010-0.015 ns, comparable to the total quoted uncertainty (~0.014 ns). Because the paper's stated purpose is to arbitrate the 7-sigma discrepancy between the tune-out value [23] and photoassociation [22], a central-value shift of this size is material: it can change whether the result is 'within 1 sigma' of [23] and how the discrepancy is interpreted. The issue is not an external disagreement; it is that the value advertised as the lifetime does not match the authors' own extrapolation and calibration.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports a direct lifetime measurement of the 5s5p 1P1 state of strontium using time-correlated single-photon counting of laser-induced fluorescence from a hot atomic beam excited by a femtosecond laser. Eight independent decay datasets are fitted with an exponential plus a scaled background, giving a weighted mean lifetime of tau = (5.216 +/- 0.006_stat +/- 0.013_sys) ns in Eq. (3). The authors evaluate systematic effects including TAC nonlinearity, magnetic field, radiation trapping, and laser power, and they claim that this direct measurement agrees with the tune-out-based value of Heinz et al. [23] within one sigma, thereby providing new information on the 7-sigma discrepancy with the photoassociation value of Yasuda et al. [22].","tokens_in":12360,"tokens_out":2114,"duration_ms":21091,"significance":"If the reported lifetime is accurate, this is a valuable contribution: it is the first direct measurement of this important state lifetime, it bears directly on the dynamic polarizability correction for strontium optical lattice clocks, and the raw data are made openly available. The experimental approach, including the use of a femtosecond laser for fast switch-off, the global truncation-time fitting procedure, and the explicit error budget, is mostly careful and reproducible. However, the central value as stated in Eq. (3) is not the corrected lifetime that the authors' own systematic analyses describe, and this affects the paper's main interpretive claim about agreement with the tune-out value.","major_comments":[{"comment":"The reported central value is not corrected for two known sign-definite systematic effects that the authors themselves quantify. In Sec. V.B the TAC calibration gives a slope of 1.00240(4), and the text states that 'the actual lifetime is shorter'; yet the authors only fold this into a 0.2% systematic uncertainty and do not shift the central value. In Sec. V.E the linear fit to the lifetime versus optical depth yields an extrapolated zero-density lifetime of 5.211 +/- 0.002 ns, a 0.096% correction relative to the 5.216 ns measured at OD=0.0011, but Eq. (3) retains 5.216 ns. Both corrections are negative, so the advertised value is biased high by roughly 0.010-0.015 ns, which is comparable to the quoted total uncertainty of about 0.014 ns. Because the paper's purpose is to arbitrate the 7-sigma discrepancy, this shift is material: it can change whether the result is 'within 1 sigma' of [23] and how the discrepancy is interpreted. The authors should correct the central value for both effects (or justify why they are not applied) and propagate the associated uncertainties accordingly.","section":"Sec. V.B, Sec. V.E, Eq. (3)"},{"comment":"The background model assumes that the background measured with the laser detuned to 457 nm has the same temporal shape as the background during on-resonance acquisition, up to a single scalar factor p. This is load-bearing because the fit includes p(B1(t)+B2(t)) as the only background term, and a time-dependent background error, especially in the first few nanoseconds after the pulse, would bias the fitted exponential time constant. The paper does not provide a direct check that resonant excitation does not change the temporal shape of the scattered-light background. I recommend an explicit test, for example comparing the background shape with the laser tuned to a nearby non-resonant transition under otherwise identical conditions, or allowing a time-dependent scaling in a control fit and checking whether tau shifts by more than the quoted uncertainty.","section":"Sec. III, Eq. (2)"}],"minor_comments":[{"comment":"The systematic uncertainty is given as 0.013 ns in the abstract and Eq. (3), but as 0.012 ns in Sec. VI; these should be consistent.","section":"Abstract and Sec. VI"},{"comment":"The text says 'TSCPC technique'; this should be 'TCSPC'.","section":"Sec. VI"},{"comment":"The branching ratio notation '1: 20 500' is unclear; please spell out that it means approximately one 1D2 decay per 20,500 decays from 1P1.","section":"Fig. 1 caption"},{"comment":"The time window is stated as 66.025 ns and the ADC bin width as 16.12 ps with 4096 bins; the product is about 66.0 ns, but the small discrepancy (0.025 ns) is not explained.","section":"Sec. II"},{"comment":"The statement that 'A was confirmed to correlate with the concentration of strontium atoms' is vague; please provide the linear relationship or a reference, since this correlation is used to estimate OD for low currents.","section":"Sec. V.E"}],"recommendation":"major_revision","confidential_remarks":"The central-value issue is real and fixable, but it is not a purely presentational point: the advertised lifetime is not the authors' own corrected value, and the discrepancy with the tune-out result could change sign of agreement after correction. The data availability statement is a strength, and the experimental methodology is solid apart from the correction issue. I would encourage the editor to request a reanalysis rather than a rejection, as the raw data appear sufficient to apply the corrections."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First thing you should know: this is the first direct measurement of the Sr 5s5p 1P1 lifetime, done with TCSPC on a hot beam excited by a femtosecond laser at 461 nm. That alone makes it useful. The measurement itself looks careful: eight independent runs, a global fit that explicitly handles truncation-time dependence, a background model with a scaling parameter, and a genuine effort on systematics like quantum beats, laser power, and radiation trapping. The data and analysis code are available, which helps.\n\nThe soft spot is internal and it is real. The paper reports tau = (5.216 +/- 0.006_stat +/- 0.013_sys) ns, but that number is not the lifetime the text describes. The TAC calibration in Sec. V.B found a slope up to 1.00240, and the authors state this means the actual lifetime is shorter. They then fold this into a 0.2% uncertainty instead of applying the correction. The radiation-trapping extrapolation in Sec. V.E gives 5.211 +/- 0.002 ns at zero density, a 0.096% correction relative to 5.216 ns measured at OD = 0.0011. Both corrections are negative. Added together, the true central value should be about 0.015 ns lower, which is larger than the total quoted uncertainty. That matters because the paper's purpose is to weigh in on the 7-sigma discrepancy between [22] and [23]. A shift of that size could move the result from 'within 1 sigma of [23]' to something less comfortable, and it changes how the discrepancy reads.\n\nThis is not a fatal flaw; it is a reporting error that a referee should catch. The right fix is to apply the corrections and quote the corrected value, or to state clearly that Eq. (3) is the in-cell value with corrections included only as uncertainties, and then give the zero-density value as the main result. The background-scaling assumption in Eq. (2) is worth a closer look too - the model assumes the detuned background has the same temporal shape as the on-resonance background up to a scalar. If resonant excitation changes the scattered-light shape in the first few nanoseconds, the fitted tau would be biased. The authors do estimate a background error by comparing B1 and B2 fits, but that only captures drift, not a shape change from resonance. I would call that a minor concern rather than a demonstrated flaw.\n\nOverall: the experiment is solid, the paper is worth a serious referee, and the atomic-clock community will want this value. My recommendation is to send it to review, and make sure the referee asks for the corrected central value before acceptance.","headline":"Direct measurement of the Sr 1P1 lifetime is a useful new data point, but the reported central value doesn't include two known-sign corrections that shift it by about a sigma.","tokens_in":12938,"tokens_out":2130,"would_cite":true,"duration_ms":18992,"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":"The paper reports the first direct measurement of the strontium 5s5p 1P1 lifetime, giving 5.216 ns with 0.006 ns statistical and 0.013 ns systematic uncertainty.","keywords":["strontium","5s5p 1P1 state","lifetime measurement","time-correlated single-photon counting","femtosecond laser excitation","blackbody radiation shift","optical lattice clock","atomic spectroscopy"],"falsifier":"One decisive check is to record the background with the femtosecond laser on resonance but with the strontium dispenser cold, so no atoms are excited, and compare its temporal shape with the 457 nm background; if the shapes differ in the first 10 ns by more than the quoted background error, the single-scale-factor model fails and the lifetime would shift.","tokens_in":11889,"feed_emoji":"⏱️","tokens_out":7154,"duration_ms":65571,"temperature":0.7,"pith_summary":"The paper reports the first direct measurement of the lifetime of the strontium 5s5p ^1P_1 state, the dominant contributor to the dynamic polarizability that limits blackbody-radiation shift calculations in optical lattice clocks. Using femtosecond 461 nm pulses in a hot atomic beam and time-correlated single-photon counting, the authors record exponential fluorescence decays and obtain $\\tau = (5.216 \\pm 0.006_{\\mathrm{stat}} \\pm 0.013_{\\mathrm{sys}})$ ns as a weighted mean of eight independent runs. They argue this result agrees with the lifetime inferred from a tune-out wavelength measurement and therefore brings new evidence to bear on a known 7-$\\sigma$ disagreement with the photoassociation-derived value. If correct, the measurement provides an independent anchor for the Einstein A coefficient of the ^1P_1 \\to ^1S_0 transition, which is needed to lower the uncertainty of Sr lattice clocks.","feed_headline":"Strontium 1P1 lifetime measured directly at 5.216 ns","feed_subtitle":"The first direct TCSPC value agrees with the tune-out route and sharpens the 7-sigma clash with photoassociation data.","key_machinery":"The carrying mechanism is the time-correlated single-photon counting histogram of laser-induced fluorescence, fit by $N(t) = A\\exp(-t/\\tau) + p(B_1(t)+B_2(t))$, where $B_1$ and $B_2$ are background signals measured before and after the decay run and $p$ is a fitted scale factor accounting for background drift. The femtosecond pulse provides switch-off faster than the lifetime without an electro-optic modulator, and the global fit across multiple truncation times (7.6 to 10.6 ns) plus extrapolation of $\\tau$ versus optical depth to zero density are what turn the raw histogram into a lifetime with controlled systematics.","core_discovery":"On the paper's own terms, the central discovery is that a femtosecond-pulsed excitation at approximately 461 nm followed by hybrid-detector time-correlated single-photon counting can cleanly resolve the roughly 5.2 ns decay of the 5s5p ^1P_1 state in a hot atomic beam, yielding the first direct lifetime $\\tau(^1P_1) = (5.216 \\pm 0.006_{\\mathrm{stat}} \\pm 0.013_{\\mathrm{sys}})$ ns. The value is reported as the weighted mean of eight independent measurements with different accumulation times, after correcting for pulse pile-up, fitting with a global truncation-time procedure, and applying systematic corrections for TAC nonlinearity, magnetic-field quantum beats, and radiation trapping extrapolated to zero optical depth. The authors find that their lifetime agrees within one $\\sigma$ with the tune-out-based value and frame the result as new evidence on the discrepancy with photoassociation spectroscopy.","pith_inferences":["If the quoted systematic budget is complete, the dominant TAC nonlinearity term could be reduced by an FPGA-based time-to-digital converter, and a future measurement might reach below 0.1% total uncertainty and tighten the comparison with theory.","The same femtosecond-excitation and TCSPC scheme could be applied to other alkaline-earth or alkaline-earth-like atoms with blue or ultraviolet transitions, giving direct lifetimes that feed polarizability estimates in other optical-clock candidates.","A simultaneous in-situ measurement or modeling of the scattered-light background during resonance, rather than the detuned-laser proxy, would directly test the single-scale-factor background assumption and would also benefit any reanalysis of the present data."],"forward_implications":["The measured lifetime fixes the Einstein A coefficient for the 5s5p ^1P_1 \\to 5s^2 ^1S_0 transition, the dominant term in the dynamic polarizability of the clock ground state, so blackbody-radiation shift calculations for Sr lattice clocks can be anchored to a direct value.","Within one sigma, the result supports the tune-out-based lifetime of Heinz et al. over the photoassociation value, meaning the 7-sigma discrepancy is likely due to an error in one of the two older routes rather than in both.","The agreement gives an independent check on the ground-state polarizability used in Sr clock uncertainty budgets, which is relevant to the redefinition of the second and to clock-based searches for dark matter and constant variation.","The experiment demonstrates that femtosecond-laser excitation plus time-correlated single-photon counting is viable for short lifetimes in the blue spectral region where fast electro-optic switch-off is unavailable."],"supporting_citations":[{"why":"Photoassociation-derived lifetime that the new measurement is compared against; the paper's result is part of resolving a 7-sigma discrepancy.","marker":"[22]"},{"why":"Tune-out-based lifetime that the new result agrees with within one sigma; this comparison supports the paper's main interpretive claim.","marker":"[23]"},{"why":"Reference for the TCSPC method and reverse start-stop timing used to record decay histograms.","marker":"[25]"},{"why":"Supplies the ^1P_1-to-^1D_2 branching ratio (1 in 20500) used to justify neglecting the cascade loss channel.","marker":"[27]"},{"why":"Gives the pulse pile-up correction formula applied to each dataset before fitting.","marker":"[30]"},{"why":"Source of the systematic truncation-time procedure used to choose the fit start bin.","marker":"[31]"},{"why":"Earlier lifetime-measurement work whose truncation-time analysis informs the global fitting approach.","marker":"[32]"}],"fun_headline_variants":["Direct TCSPC measures Sr 1P1 at 5.216 ns","Femtosecond TCSPC pins Sr 1P1 lifetime at 5.216 ns","First direct TCSPC Sr 1P1 lifetime: 5.216 ns","Sr 1P1 direct measurement: 5.216 ns, agrees with tune-out","Direct Sr 1P1 lifetime 5.216 ns: matches tune-out, challenges photoassociation"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The fit assumes the background recorded with the laser detuned to 457 nm is the same function of time as the background during resonant excitation except for one scale factor; if resonant light changes the scattered-light background's time shape during the first nanoseconds, the fitted lifetime is biased.","fun_headline_variants_meta":{"raw":{"variants":["Direct TCSPC measures Sr 1P1 at 5.216 ns","Femtosecond TCSPC pins Sr 1P1 lifetime at 5.216 ns","First direct TCSPC Sr 1P1 lifetime: 5.216 ns","Sr 1P1 direct measurement: 5.216 ns, agrees with tune-out","Direct Sr 1P1 lifetime 5.216 ns: matches tune-out, challenges photoassociation"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001086,"raw_usage":{"total_tokens":4502,"prompt_tokens":870,"completion_tokens":3632,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":486,"completion_tokens_details":{"reasoning_tokens":3516}},"tokens_in":486,"tokens_out":3632,"duration_ms":26304,"temperature":1.0,"reasoning_tokens":3516,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T20:43:32.568620+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"One decisive check is to record the background with the femtosecond laser on resonance but with the strontium dispenser cold, so no atoms are excited, and compare its temporal shape with the 457 nm background; if the shapes differ in the first 10 ns by more than the quoted background error, the single-scale-factor model fails and the lifetime would shift.","supporting_citations":[{"cited_title":"Yasuda, T","cited_arxiv_id":null,"evidence_quote":"Photoassociation-derived lifetime that the new measurement is compared against; the paper's result is part of resolving a 7-sigma discrepancy."},{"cited_title":"Heinz, A","cited_arxiv_id":null,"evidence_quote":"Tune-out-based lifetime that the new result agrees with within one sigma; this comparison supports the paper's main interpretive claim."},{"cited_title":"Becker, The Bh TCSPC Handbook, 10th edition (Becker & Hickl, 2023)","cited_arxiv_id":null,"evidence_quote":"Reference for the TCSPC method and reverse start-stop timing used to record decay histograms."},{"cited_title":"Lifetime measurement of the 5s5p 1P1 state in strontium","cited_arxiv_id":"2501.07395","evidence_quote":"Supplies the ^1P_1-to-^1D_2 branching ratio (1 in 20500) used to justify neglecting the cascade loss channel."},{"cited_title":"Felinto, C","cited_arxiv_id":null,"evidence_quote":"Gives the pulse pile-up correction formula applied to each dataset before fitting."},{"cited_title":"Gomez, F","cited_arxiv_id":null,"evidence_quote":"Source of the systematic truncation-time procedure used to choose the fit start bin."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Earlier lifetime-measurement work whose truncation-time analysis informs the global fitting approach."}],"review_version":1}