{"id":"8fd06512-58fd-459d-8d71-5d1d6efcd686","arxiv_id":"2504.16811","paper_version":2,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"A variable-length optical cavity with modulated light measures attenuation length in transparent liquids to about 2 percent and group velocity to about 0.03 percent.","lead":"This paper introduces CELLPALS, a cavity-based method that measures how far light travels through transparent liquids before being absorbed or scattered, plus the speed of light pulses in the liquid. It matters for neutrino detectors like JUNO, which need very transparent liquid scintillators and precise light-timing information to reconstruct events.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 2% attenuation-length claim rests on Eq. 12 assuming only Beer's-law loss varies with L; the paper's qualitative collimation cut and acknowledged length-dependent residuals leave this assumption untested.","rationale":"The reader's CONDITIONAL verdict is already well calibrated: the strongest claim's accuracy rests on Eq. 12, and the paper's own data-handling and residual discussion make unmodeled length-dependent loss the key unverified link. I do not see an internal inconsistency that would justify REJECT: the derivation is standard, the air measurement (Fig. 5) validates the free-spectral-range route to v_g, the comparison with Ref. [30] supports the group-velocity values for LAB and LS, and the UV-Vis comparison is broadly consistent. Eq. 9's inverted printed form is a reproducibility issue rather than the core physical risk, and the suspicious EJ-309 group velocity in Table 2 is a red flag for a typo or an unmodeled phase effect, but it is secondary to the L_att bias question. The proposed empty-cavity slope test would directly turn the conditional into a verified claim, or expose the bias.","tokens_in":14265,"tokens_out":14202,"duration_ms":146794,"concrete_test":"Run an empty-cavity control over the full 0.3-0.9 m range (40+ lengths) and fit ln A_eff versus L using the same fitting procedure. Since air has effectively infinite attenuation length, Eq. 12 predicts a zero slope. If the fitted slope is inconsistent with zero at the level of about 0.003 m^-1 (the slope that would shift a 12 m L_att by 2%), the central assumption of no unmodeled length-dependent loss fails; if the slope is consistent with zero within that tolerance, the 2% attenuation-length claim is supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The determination of L_att uses Eq. 12, A_eff(L)=R^2 exp(-2L/L_att), and fits the slope of ln A_eff versus L. That slope is interpreted entirely as Beer's-law attenuation. The paper never quantifies L-dependent losses that would enter the same slope: changes in the Gaussian mode size on the concave mirrors and finite clipping by the 1-inch mirrors or the PTFE bore, diffraction at the edge of the stability range, scattering at the mirror/liquid interfaces, and alignment changes of the movable mirror. The discussion in Sec. 3.4 covers only incomplete collimation of the outgoing beam after the cavity and rejects data with a qualitative 'characteristic deviations' criterion; the conclusion itself admits that some resonator lengths show deviations larger than the corresponding fit errors. Since lengths are excluded post hoc and no loss budget is given, the quoted <2% uncertainty is a statement about fit precision, not accuracy. A length-dependent loss of about 0.2% over the 0.3-0.9 m scan would bias a 12 m attenuation length by roughly 2%, exactly the claimed precision.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents CELLPALS, a new method for measuring the attenuation length and group velocity of transparent liquids using a variable-length Fabry-Pérot cavity with an amplitude-modulated laser. The method fits the frequency-dependent amplitude ratio and phase shift of the cavity output to extract the free spectral range Δν and the attenuation parameter A_eff, from which the group velocity v_g = 2L Δν and the attenuation length L_att via A_eff = R^2 exp(-2L/L_att) are derived. The authors demonstrate the method on several liquids, report ~2% relative uncertainty for L_att and ~0.03% for v_g, validate v_g against the known group velocity of air, and compare L_att results with UV-Vis spectroscopy. The paper includes a derivation in Appendix A, an experimental setup description, and measurements of LAB at different purification stages.","tokens_in":14467,"tokens_out":10136,"duration_ms":87975,"significance":"If the accuracy claims hold, CELLPALS is a significant practical advance for characterizing liquid scintillators and water in neutrino detectors: it requires only about a liter of sample, measures both L_att and v_g simultaneously, and reaches uncertainties competitive with long-tube methods while being far more compact. The manuscript is thorough in its empirical characterization: 15,675 measurements at 51 cavity lengths, a dedicated test of the group-velocity determination against the literature value for air, and a comparison with UV-Vis spectroscopy. The derivation in Appendix A is standard and correctly gives the resonance structure. The main weakness is that the central ~2% accuracy claim for L_att rests on an implicit assumption that no length-dependent loss other than Beer's-law attenuation contributes to the fitted slope, and the manuscript does not provide a quantitative loss budget for this assumption. A clear typographical error in Eq. (9) also needs correction. Overall, the method is promising and the results are broadly supportive, but the accuracy claim requires additional systematic analysis.","major_comments":[{"comment":"Equation (9) is printed as A(ν) = b1 √(A_eff² − 2A_eff cos(2πν/Δν) + 1), i.e., the square root is multiplied by b1. The derivation in Appendix A.4, however, gives the amplitude as IT² exp(−L/L_att) divided by this same square-root factor, so the amplitude ratio should be b1 / √(A_eff² − 2A_eff cos(2πν/Δν) + 1). With the printed multiplicative form, A(ν) would be minimal at the resonance ν = Δν, inverting the peak structure shown in Fig. 3. Since the fits evidently use the denominator form (Fig. 3 shows resonance peaks), this is a severe typographical error in a central equation that must be corrected; the fitting code should also be cross-checked against the published formula.","section":"§2.2, Eq. (9)"},{"comment":"The claim that L_att is determined with ~2% relative uncertainty rests on Eq. (12), which assumes that the only length-dependent loss is Beer's-law attenuation, so that ln A_eff(L) = ln R² − 2L/L_att is a straight line in L. The manuscript does not provide a quantitative budget for other losses that can scale with cavity length: changes in the Gaussian mode size on the 1-inch mirrors and clipping by the PTFE bore, diffraction at the edges of the stability range, scattering at the liquid–mirror interfaces, and alignment changes as the movable mirror is translated. The stability criterion (Eq. 16) and the qualitative 'characteristic deviations' collimation test of §3.4 do not bound these effects at the 0.2% level in A_eff over the 0.3–0.9 m scan. A length-dependent loss of ~0.2% over the scan would bias a 12 m attenuation length by ~2%, exactly the claimed precision. Please provide a quantitative estimate of each length-dependent loss, or an experimental test (e.g., measuring A_eff(L) with different mirror apertures or beam sizes, or measuring a sample with an independently known L_att), to support the accuracy rather than only the precision.","section":"§3.3, Eq. (12); §3.4; §7"},{"comment":"The conclusion states that 'the data for some resonator lengths show deviations that are larger than the corresponding fit errors' and attributes these to collimation effects, while §3.4 says that measurements showing 'characteristic deviations' are dismissed. This indicates that data points are excluded from the L_att fit after inspection. The manuscript does not state how many lengths were excluded, the exact selection criterion, or the resulting change in the fitted L_att. Without this information, the least-squares uncertainty does not include model-selection uncertainty, and the fitted slope could be biased by the post hoc removal of points. Please quantify the exclusions, report a fit that includes all measured lengths, and discuss whether L_att changes by more than the quoted 2%.","section":"§7, Fig. 11"}],"minor_comments":[{"comment":"The stability parameter is printed as g = 1 + L/r. For concave mirrors with the usual sign convention (r > 0 for the radius of curvature), the g-factor is g = 1 − L/r. The stated maximum length L_max = 2r corresponds to the minus sign, so Eq. (16) appears to have a sign error unless a nonstandard sign convention is being used; please clarify.","section":"§3.3, Eq. (16)"},{"comment":"Cross-references to sections appear as 'In 2', 'cf. 2.1', 'in 3', 'in 6', etc. Please replace these with 'Section 2', 'Section 2.1', 'Section 3', 'Section 6', and similar.","section":"Throughout"},{"comment":"The phrase 'Beers’ law' is used consistently; the standard name is 'Beer’s law' (after August Beer). Please correct this throughout.","section":"Abstract and §1"},{"comment":"The transmission coefficient is defined as T = 1 − R, which assumes lossless mirrors. Since the mirror reflectivity is specified as R = 0.95, a small absorption/scattering loss in the mirrors is possible. This does not affect the slope used for L_att, but a sentence stating this assumption would improve clarity.","section":"§2.1, Eq. (4)"},{"comment":"The text states that Asub = Afull − (1/2)Aempty is ≥ 0, but immediately afterward the plot shows Asub < 0 for some wavelengths. This is discussed as a puzzle, but the inequality in the text is incorrect as stated; it should say that the theoretical model predicts nonnegative values, while the data show negative values, which is the observed discrepancy.","section":"§5, Eq. (20)"}],"recommendation":"major_revision","confidential_remarks":"The paper is within the scope of physics.ins-det and presents a genuinely useful instrument concept. The main load-bearing issues are the missing loss budget for the L_att accuracy claim, the post-hoc exclusion of data points, and the typographical error in Eq. (9); all are fixable within the manuscript's scope. I do not see a fundamental flaw in the method itself, and the group-velocity determination is well validated by the air measurement. No concerns about novelty disclosure or citation practices."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a legitimate new instrument paper. The variable-length cavity trick to cancel mirror reflectivity in the attenuation fit is real, and the group velocity measurement via free spectral range is a nice bonus. The air calibration checks out, and the reported ~2% attenuation length uncertainty with a ~1 L sample volume would be useful for JUNO-type purification monitoring.\n\nThe math in Appendix A is standard phase-shift cavity ring-down, and they cite Reid et al. honestly. The amplitude and phase formulas are correct, with one important exception: Eq. 9 as printed has the square root in the numerator instead of the denominator (compare App. A.4). That's clearly a typo, but it needs fixing because it changes the resonance shape.\n\nThe bigger question is the attenuation length accuracy, not precision. Eq. 12 assumes that the only length-dependent loss is Beer's law. If beam clipping, diffraction at the stability edge, or mirror-liquid interface scattering varies with L, that goes directly into the slope. The paper's collimation cut is qualitative—\"characteristic deviations\"—and the conclusion admits some lengths show deviations larger than fit errors. The stress-test estimate that a 0.2% length-dependent loss would bias a 12 m attenuation length by ~2% is plausible, and it means the quoted uncertainty is currently a precision statement, not an accuracy budget. I'd want to see either a quantitative loss model or a measurement that varies something other than L (e.g., different mirror sets) to confirm the slope is really Beer's law.\n\nAlso check Table 2: EJ-309 at 56.55% of c is far below what any xylene-based scintillator should give (vg/c ~0.65-0.68). That looks like an error in the analysis or the table, not a real measurement.\n\nThe comparison with UV-Vis is fine and appropriately critical. The JUNO purification samples show the method works in practice.\n\nBottom line: the paper deserves a serious referee and likely publication after the typo fix and a real loss budget. For a neutrino detector community, this could be a useful tool. I would bring it to a reading group interested in instrumentation, but not to a general audience.","headline":"CELLPALS is a genuinely new variable-length cavity instrument for simultaneous attenuation length and group velocity in liquids; the method looks sound but the paper needs a corrected Eq. 9, a quantitative loss budget, and a check of the EJ-309 group velocity.","tokens_in":15021,"tokens_out":1780,"would_cite":false,"duration_ms":16884,"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":"Amplitude-modulated cavity measures liquid transparency to 2 percent","keywords":["attenuation length","group velocity","Fabry-Pérot cavity","liquid scintillator","linear alkylbenzene","optical cavity resonator","frequency-domain measurement","variable-length cavity"],"falsifier":"Take a liquid whose attenuation length has been established with an independent long-path method and fit the CELLPALS data in two separate length ranges, say 30–60 cm and 60–90 cm; if the two fitted values of $L_\\mathrm{att}$ differ by more than the quoted 2 percent, unmodeled length-dependent losses are biasing the result.","tokens_in":14055,"feed_emoji":"💧","tokens_out":9904,"duration_ms":87722,"temperature":0.7,"pith_summary":"The paper introduces CELLPALS, a frequency-domain method that measures the attenuation length of highly transparent liquids to about 2 percent relative uncertainty and their group velocity to about 0.03 percent. It uses an amplitude-modulated laser and a variable-length Fabry-Pérot cavity so that light passes through the sample many times, letting a compact, roughly one-litre cell probe attenuation lengths above ten metres. The key step is to fit the measured amplitude ratio and phase shift between a reference beam and the beam leaving the cavity as the modulation frequency is swept, which yields the free spectral range and an effective attenuation parameter. Those two quantities give the group velocity and the attenuation length directly. The paper demonstrates the method on several organic liquids and on purified linear alkylbenzene samples, and shows it outperforms UV-Vis spectroscopy for very transparent media.","feed_headline":"Amplitude-modulated cavity measures liquid transparency to 2 percent","feed_subtitle":"CELLPALS also returns the liquid's group velocity to 0.03 percent, sharpening neutrino-detector design.","key_machinery":"The central object is a variable-length Fabry-Pérot cavity formed by two concave mirrors of reflectivity $R$ and radius of curvature $r$, filled with the liquid under test, illuminated by a sinusoidally modulated laser whose modulation frequency is swept up to about 350 MHz. Each ray that leaves the cavity has been reflected $n$ times and has traversed the liquid a distance $(2n+1)L$; summing the geometric series of these rays produces the measured amplitude ratio and phase shift. The two fit parameters are the free spectral range $\\Delta\\nu=v_g/(2L)$, which gives the group velocity, and the attenuation parameter $A_\\mathrm{eff}=R^2\\exp(-2L/L_\\mathrm{att})$, which gives the attenuation length once $R$ has been removed by a multi-length fit. A stability criterion and a collimation-deviation check are used to identify data affected by incomplete detection of the outgoing beam.","core_discovery":"CELLPALS determines the attenuation length $L_\\mathrm{att}$ and the group velocity $v_g$ of a transparent liquid inside an optical cavity by measuring, as a function of modulation frequency $\\nu$, the amplitude ratio $A(\\nu)$ and the phase shift $\\Delta\\phi(\\nu)$ between a reference beam and light transmitted through the cavity. The measured curves follow closed-form expressions in which the free spectral range $\\Delta\\nu=v_g/(2L)$ fixes the resonance positions and the parameter $A_\\mathrm{eff}=R^2\\exp(-2L/L_\\mathrm{att})$ sets the resonance contrast. Because $A_\\mathrm{eff}$ is recorded at many cavity lengths $L$, the mirror reflectivity $R$ cancels when the exponential decay is fitted, removing the dominant systematic error of a single-cavity measurement. The reported uncertainties are about $0.03\\%$ for $v_g$ and below $2\\%$ for $L_\\mathrm{att}$ when more than fifty cavity lengths are used, and the group-velocity results agree with literature values in air and with refractive-index-based values in a liquid scintillator.","pith_inferences":["A natural extension, not discussed in the paper, would be to sweep the laser wavelength as well as the modulation frequency, turning the same cavity into a dispersion-measuring instrument that maps both $L_\\mathrm{att}(\\lambda)$ and $v_g(\\lambda)$.","A testable implication is that selecting the detection geometry could separate absorption from scattering: rays that scatter out of the cavity mode contribute to the fitted loss, while small-angle scattered rays that remain in the cavity do not, so a variant with a movable iris might distinguish the two components.","The frequency-domain readout is the Fourier analogue of time-domain cavity ring-down, so a direct comparison of the two on the same liquid should expose unmodeled mode-dependent or polarization-dependent losses if they exist."],"forward_implications":["A single compact cell holding roughly one litre can monitor attenuation length with about 2 percent uncertainty, making inline purification monitoring practical for large liquid-scintillator detectors.","The same measurement returns the group velocity to about 0.03 percent, which improves timing and position reconstruction in large detectors and is not accessible with standard attenuation-length instruments.","For liquids with attenuation lengths above about ten metres, CELLPALS has substantially smaller uncertainties than UV-Vis cuvette spectroscopy, where the sample attenuation is below half a percent and reflection corrections dominate.","Measuring $A_\\mathrm{eff}(L)$ at many lengths also extracts the mirror reflectivity $R$, potentially with better precision than the manufacturer's specification."],"supporting_citations":[{"why":"Introduces the cavity-enhancement factor and the summing of multiple passes that the method builds on.","marker":"[16]"},{"why":"Provides the phase-shift formalism used to derive the frequency-dependent amplitude ratio and phase shift.","marker":"[17]"},{"why":"Supplies the literature value of the group velocity in air used to calibrate the free spectral range.","marker":"[18]"},{"why":"Gives the Gaussian-beam and ray-transfer optics used for the stability criterion and the collimation-deviation analysis.","marker":"[19]"},{"why":"Supplies the purified linear alkylbenzene and liquid-scintillator samples on which the method is demonstrated.","marker":"[7]"},{"why":"Provides the refractive-index-based group velocity used to independently check the CELLPALS group-velocity results.","marker":"[30]"}],"fun_headline_variants":["CELLPALS: 2% attenuation, 0.03% group velocity","Cavity method measures liquid clarity and light speed together","CELLPALS: one cavity, both attenuation length and group velocity","New optical cavity yields two key liquid properties at once","Variable-length cavity yields attenuation and group velocity"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The inference from the measured $A_\\mathrm{eff}(L)$ to $L_\\mathrm{att}$ assumes that the only loss growing with cavity length is Beer's-law attenuation in the liquid, so $A_\\mathrm{eff}=R^2\\exp(-2L/L_\\mathrm{att})$ with a constant mirror reflectivity; any other length-dependent loss, such as clipping, diffraction, or scattering at the liquid-mirror surfaces, would be misread as a shorter attenuation length.","fun_headline_variants_meta":{"raw":{"variants":["CELLPALS: 2% attenuation, 0.03% group velocity","Cavity method measures liquid clarity and light speed together","CELLPALS: one cavity, both attenuation length and group velocity","New optical cavity yields two key liquid properties at once","Variable-length cavity yields attenuation and group velocity"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00152,"raw_usage":{"total_tokens":6149,"prompt_tokens":1061,"completion_tokens":5088,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":677,"completion_tokens_details":{"reasoning_tokens":5004}},"tokens_in":677,"tokens_out":5088,"duration_ms":37970,"temperature":1.0,"reasoning_tokens":5004,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T10:56:00.171359+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a liquid whose attenuation length has been established with an independent long-path method and fit the CELLPALS data in two separate length ranges, say 30–60 cm and 60–90 cm; if the two fitted values of $L_\\mathrm{att}$ differ by more than the quoted 2 percent, unmodeled length-dependent losses are biasing the result.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the phase-shift formalism used to derive the frequency-dependent amplitude ratio and phase shift."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the literature value of the group velocity in air used to calibrate the free spectral range."},{"cited_title":"Hecht, Optics, Vol","cited_arxiv_id":null,"evidence_quote":"Gives the Gaussian-beam and ray-transfer optics used for the stability criterion and the collimation-deviation analysis."}],"review_version":1}