{"id":"3633fab8-0b58-4aab-9890-5bb879a70bb0","arxiv_id":"1909.02207","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":3.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"A liquid argon detector yields 7.28 and 7.66 photoelectrons per keV for 41.5 keV krypton and 511 keV sodium events, with light output falling by 8.4% and 21.4% at 200 V/cm drift field.","lead":"This paper injects a short-lived krypton isotope into a liquid argon detector and measures how much light the detector produces per unit energy, both at zero drift field and under electric fields up to 200 V/cm. The numbers are calibration anchors for argon-based dark matter detectors.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Position-dependent light collection, evidenced by the 17.5% 83mKr peak width versus 5.8% photostatistics, likely biases the volume-averaged 83mKr yield and confounds the 5% 83mKr/22Na comparison.","rationale":"The reader's verdict is already CONDITIONAL, and this concern reinforces it rather than overturning the paper. The specific condition should be: demonstrate spatial uniformity of light collection (or correct for it) using the PMT-asymmetry test, and then reassess the absolute 83mKr yield and its comparison with 22Na. The LED-SPE concern flagged by the reader is also valid but is secondary here, because a common SPE error would shift both absolute yields without changing the 5% comparison or the field-quenching ratios; the position-dependence issue directly attacks the comparison and the meaning of the distributed-source calibration. I therefore recommend keeping the CONDITIONAL verdict, with this position-uniformity check added to the required revisions.","tokens_in":8085,"tokens_out":16122,"duration_ms":180775,"concrete_test":"Reanalyze the recorded two-PMT waveforms from the 83mKr and 22Na runs: compute the charge asymmetry A=(Q_top-Q_bottom)/(Q_top+Q_bottom) for each event and bin the full-absorption peak centroid in A bins. If the 83mKr peak centroid for events with A≈0 (central axial position) is more than ~3% above the all-event centroid and moves toward the 22Na value, while the 22Na centroid under the same collimated geometry is flat in A, then the reported 83mKr light yield and the 5% 83mKr-22Na difference are biased by position-dependent light collection; if the A≈0 83mKr centroid remains near 7.28 pe/keV, the difference is physical.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The 83mKr source is introduced through the recirculation system and therefore populates the full active volume, whereas the 22Na calibration uses a 10-mm collimated beam aimed at the centre (Sec. 4.1). Section 4.1 reports for 83mKr at 0 V/cm a fitted peak at 301.9 p.e. with sigma 52.88 p.e. (17.5% resolution); for the same peak, photostatistics alone give sqrt(301.9)=17.4 p.e. (5.8%), so ~16.5% of the width is unaccounted. For the central-collimated 22Na 511-keV peak, the measured width is 3.4% against a photostatistics contribution of 1.6%, leaving only ~3.0% excess. The much larger unexplained fractional width for the only volume-distributed source is strong evidence of position-dependent light collection, likely from the PTFE/ESR/TPB geometry and the two-PMT readout. The paper makes no position correction and reports no spatial-uniformity check. Consequently the comparison 'the light of 511 keV is about 5% higher than 41.5 keV... because the greater stopping power [27]' is confounded: the 22Na value samples a narrow central volume while the 83mKr value is averaged over the whole active volume. The headline absolute yields and the 8.4% versus 21.4% field-quenching comparison therefore rest on an unverified and likely false assumption that the light response is spatially uniform.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper reports calibration measurements of a small two-phase liquid argon TPC using 83mKr introduced via the gas recirculation system and 22Na as an external collimated source. The authors measure the scintillation light yield at 41.5 keV and 511 keV for drift fields of 0, 50, 100, 150, and 200 V/cm, report the field-induced quenching (8.4% for 83mKr and 21.4% for 22Na at 200 V/cm), and verify the 83mKr half-life (1.83±0.11 h) after halting source injection. The paper's central claims are that 83mKr provides a distributed low-energy calibration source for LAr detectors and that the absolute yields and field dependence are measured to ~0.01–0.02 pe/keV precision.","tokens_in":8403,"tokens_out":7055,"duration_ms":64063,"significance":"If the quantitative results are correct, the paper offers a simple and useful validation of 83mKr-based distributed calibration for liquid argon detectors, along with a concrete dataset on electric-field quenching of the scintillation yield. The half-life check is a nice internal consistency test, and the use of two independent sources with different energies is a reasonable approach. However, the paper as presented does not yet support the stated absolute precision because it quotes only statistical uncertainties and because the volume-distributed 83mKr sample may be subject to position-dependent light collection that is not characterized.","major_comments":[{"comment":"The 83mKr full-absorption peak at 0 V/cm has µ=301.9 pe and σ=52.88 pe, giving a resolution of 17.5%, whereas photostatistics alone contributes 1/sqrt(301.9)≈5.8%. In contrast, the 22Na 511-keV peak has ~3.4% resolution versus a 1.6% photostatistics contribution. This large unexplained broadening for the volume-distributed 83mKr source, compared with the centrally-collimated 22Na source, is strong evidence of position-dependent light collection that is not addressed anywhere in the paper. Because the 83mKr events populate the entire active volume while 22Na samples a small central region, the quoted 5% difference between the 41.5-keV and 511-keV light yields is confounded by spatial geometry. The authors should either provide a position-uniformity map or position correction based on the PMT charge ratio, or restrict both calibrations to a matching fiducial volume; without this, the absolute 83mKr yields and the field-quenching comparison between the two sources in Section 5 are not reliable.","section":"Sec. 4.1, Figs. 6–7, Tables 1–2"},{"comment":"The uncertainties reported for the light yields (e.g., 7.28±0.02 pe/keV) are only statistical fit errors. The absolute scale relies on the LED-based single-photoelectron calibration of Section 3.2, which yields mean SPE charges of 0.53 pC and 0.44 pC with a monitored variation below 2%. The paper does not estimate the systematic error in transferring this LED calibration to 128-nm scintillation light shifted by TPB to 420 nm, including possible differences in photoelectron collection, pulse shape, or gain. Therefore the claims of 0.01–0.02 pe/keV precision are not supported. A systematic error budget, or at least a conservative estimate based on the SPE calibration uncertainty and optical collection variations, is needed.","section":"Sec. 3.2 and Sec. 4.1"}],"minor_comments":[{"comment":"The abstract states 7.26±0.02 pe/keV for 83mKr, while Section 4.1 and Table 1 report 7.28±0.02; the value should be reconciled.","section":"Abstract vs. Sec. 4.1"},{"comment":"σp is listed as 1475.3 pe, which is far outside the range of the other rows and inconsistent with Fig. 10; it is presumably a typo for 147.5 pe.","section":"Table 2, 200 V/cm row"},{"comment":"The Section 4.2 title says 'from 0 to 200V/cm' but the abstract says 'from 50 to 200V/cm'; the stated range should be consistent.","section":"Sec. 4.2 vs. Abstract"},{"comment":"The claim that the PARIS model 'provides a good description' of the data is not supported by any quantitative comparison in the paper; the authors should add a model curve or residuals, or qualify the statement.","section":"Sec. 5"},{"comment":"There are minor typos: 'E_drfift' in the captions of Figs. 9–10, 'after stop filling 83mKr atom' in Section 4.1, and non-standard capitalization in several references (e.g., refs. [20] and [26]).","section":"Figures and references"}],"recommendation":"major_revision","confidential_remarks":"The spatial-uniformity issue is the central technical concern and should be the main request in revision. The authors should be asked to provide a position-dependent analysis or a clear demonstration that the 22Na and 83mKr samples see the same light response. They should also correct the 7.26/7.28 inconsistency and the Table 2 typo. The paper is within the journal's scope, but the main quantitative claims need stronger systematic support."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Let me give you the short version: this is a competent but incremental calibration study of a small liquid argon TPC using 83mKr and 22Na. The 83mKr circulation technique is established, and the new content is a set of absolute light yields and drift-field dependence for this particular detector. That is legitimate progress, but the paper has a real soft spot: the 83mKr peak is 17.5% wide while photostatistics alone predict 5.8%, which is strong evidence of position-dependent light collection. Because 83mKr is distributed throughout the active volume and 22Na is collimated to the center, the reported comparison between the two yields is confounded.\n\nLet me give credit first. The detector is described clearly, the data acquisition and SPE calibration are standard, and the half-life check (1.83 ± 0.11 h vs 1.83 ± 0.02 h) is a nice cross-check. The field-dependence measurements from 0 to 200 V/cm are smoothly behaved and plausible. The paper does not oversell its novelty.\n\nNow the soft spots, in order of severity. The absence of systematic uncertainties is a problem. The abstract reports 7.26 pe/keV while Table 1 says 7.28; that inconsistency should never survive copyediting. Table 2 has a sigma of 1475.3 at 200 V/cm that is clearly a typo for ~147.5. These are fixable, but they make you distrust the precision claims. More fundamental: the paper quotes only statistical errors, so the 0.01–0.02 pe/keV precision is not supported. The SPE calibration is the backbone of every absolute yield, and the paper does not estimate how the LED-based calibration transfers to TPB-shifted scintillation pulses.\n\nThe position-dependent response is the most serious issue. The 83mKr peak width is three times the photostatistics limit, while the central 22Na peak has a much smaller unexplained excess. That asymmetry strongly suggests the light collection varies with position, and the paper makes no position correction and shows no uniformity map. Consequently, the 5% difference between the 511 keV and 41.5 keV light yields may reflect geometry rather than physics. The comparison to the PARIS model is qualitative; a quantitative fit would be more convincing.\n\nFor whom is this paper? Anyone building a small LAr TPC and wanting a reference for achievable light yields and field quenching will find it useful, but it should be read as a prototype study, not a high-precision calibration. It deserves a serious referee — the measurement is honest and reproducible in principle — but it needs revision before acceptance. I'd want to see a treatment of systematics, an estimate of the spatial uniformity bias, and corrected typographic inconsistencies.\n\nRecommendation: send it out for peer review, with a referee who understands position-dependent light collection in gas/argon TPCs. If I were writing my own paper, I'd cite it for the field-dependence trends, but not for absolute yields until the systematics are addressed.","headline":"Useful incremental LAr calibration data, but the volume-averaged 83mKr light yield is likely biased by position-dependent response and the paper quotes only statistical errors.","tokens_in":8978,"tokens_out":2625,"would_cite":true,"duration_ms":30819,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["85.60.Ha","14.60.Pq"],"model":"deepseek-v4-flash","headline":"Krypton-83m, carried in by recirculating argon, calibrates the low-energy light response of a liquid argon TPC.","keywords":["liquid argon","time projection chamber","83mKr calibration","light yield","drift electric field","electron recoil","noble-liquid detector"],"falsifier":"One concrete check: recalibrate the same detector's light yield using an independent method, such as a collimated gamma source with a known emission rate and a Monte Carlo of the geometry, or a precision source of single photons with known quantum efficiency. If the resulting photoelectrons-per-keV differs from 7.26 at 41.5 keV by more than the quoted statistical uncertainty, the LED-based single-photoelectron scale is wrong. Alternatively, using a second PMT with a different photocathode or TPB configuration would reveal whether the field-quenching trend is a detector artifact or a genuine argon property.","tokens_in":7870,"feed_emoji":"⚛️","tokens_out":6894,"duration_ms":61043,"temperature":0.7,"pith_summary":"This paper demonstrates that the radioactive isotope $^{83m}$Kr can be swept into a liquid argon (LAr) detector through the existing argon recirculation system and used as an internal, low-energy calibration source. The authors measure the detector's light yield at the 41.5 keV sum of the two $^{83m}$Kr transitions and compare it with the 511 keV line from $^{22}$Na. They report absolute light yields of 7.26$\\pm$0.02 photoelectrons per keV for $^{83m}$Kr and 7.66$\\pm$0.01 photoelectrons per keV for $^{22}$Na at zero drift field, and they track how both yields fall as the drift field rises to 200 V/cm. If the result stands, large LAr detectors can calibrate their central volume without deployment hardware, and field-dependent quenching must be folded into the energy response.","feed_headline":"Krypton-83m calibrates liquid argon from within","feed_subtitle":"A 41.5 keV source spreads through the whole liquid volume, mapping absolute yield and field effects.","key_machinery":"The mechanism that carries the argument is the transport of $^{83m}$Kr atoms from a $^{83}$Rb-doped zeolite trap through the closed argon circulation loop, so the isotope is distributed throughout the active volume rather than collimated from outside. Inside the liquid, $^{83m}$Kr decays via two conversion-electron/x-ray transitions summing to 41.5 keV; because the 154 ns separation is much shorter than the argon triplet scintillation time, the two interactions merge into a single light pulse. Light yield is then inferred from Gaussian fits to the full-absorption peak, converted from photoelectrons to keV using a single-photoelectron calibration of the two immersed PMTs based on a PMT response function fitted to LED data (mean charges 0.53 pC and 0.44 pC, monitored to less than 2% variation). Field scans from 0 to 200 V/cm expose the recombination-driven decrease in scintillation yield, which the paper interprets with the PARIS recombination model.","core_discovery":"The paper's central finding is that $^{83m}$Kr introduced through the gas recirculation system behaves like a well-understood point-like source of 41.5 keV electronic recoils inside a liquid argon TPC: the two cascaded transitions at 32.1 keV and 9.4 keV fire within about 154 ns, well inside the 1.6 $\\mu$s argon triplet time, so they appear as a single peak. Fitting that peak gives a light yield of 7.28$\\pm$0.02 photoelectrons per keV at zero field, with 17.6% energy resolution, versus 7.66$\\pm$0.01 photoelectrons per keV and 3.6% resolution for the 511 keV line from $^{22}$Na. Raising the drift field from 0 to 200 V/cm suppresses recombination and quenches the light yield by 8.4% for $^{83m}$Kr and 21.4% for $^{22}$Na; the weaker quenching at 41.5 keV indicates that recombination is more complete for the denser ionization column of the lower-energy electrons. The fitted half-life of 1.83$\\pm$0.11 h after stopping the fill matches the known 1.83$\\pm$0.02 h, confirming that the source decays cleanly away with no permanent contamination.","pith_inferences":["The reported light-yield ratio between 511 keV and 41.5 keV, with the higher-energy line brighter by about 5%, could be tested against models that treat recombination as a function of linear energy transfer; if the trend extends to tens of keV, calibration sources at multiple energies would be needed to map the nonlinearity.","Because the single-photoelectron calibration relies on LED pulses, a subtle systematic would arise if the LED optical pulse shape or PMT gain differs from that for TPB-shifted 128 nm scintillation events; comparing the 511 keV peak position with an independent, absolutely calibrated source would settle the scale.","If $^{83m}$Kr can be circulated through a full-scale detector without degrading argon purity, as this small detector suggests, it may also serve for continuous stability monitoring during long physics runs, not just calibration campaigns."],"forward_implications":["Large liquid argon TPCs can use $^{83m}$Kr as a low-energy, volume-filling calibration source without any insertion mechanism, since it enters with the recirculating gas and leaves with the known 1.83 h half-life.","The measured absolute light yields give a direct energy-scale conversion for electronic recoils at 41.5 keV and 511 keV in a TPB-coated LAr detector with two immersed PMTs.","The field-induced quenching curves (8.4% at 41.5 keV, 21.4% at 511 keV, at 200 V/cm) provide data that recombination models such as PARIS must reproduce, and imply that the energy scale shifts with drift field.","The small 5% light-yield difference between 41.5 keV and 511 keV indicates a mild non-linearity in LAr scintillation response over this range, relevant for low-energy rare-event searches."],"supporting_citations":[{"why":"Provides the $^{83}$Rb to $^{83m}$Kr decay properties and the 1.83 h half-life against which the measured decay is compared.","marker":"[16]"},{"why":"Established the use of $^{83m}$Kr as a calibration source in noble-liquid detectors, the method this paper adapts to liquid argon.","marker":"[20]"},{"why":"Supplies the PMT response function used to fit the single-photoelectron spectra.","marker":"[25]"},{"why":"Gives the calibrated low-energy response of liquid argon to $^{83m}$Kr, providing a reference for the observed light yield.","marker":"[26]"},{"why":"Defines the PARIS recombination model used to interpret the field-induced yield quenching.","marker":"[29]"},{"why":"Provides the measured LAr energy response to electronic recoils used to explain the energy-dependent quenching.","marker":"[30]"}],"fun_headline_variants":["83mKr: internal calibrator for liquid argon","Internal 83mKr calibrates LAr detectors","83mKr and 22Na probe LAr field effects","Liquid argon self-calibrates with 83mKr","83mKr: the built-in yardstick for LAr detectors"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The result assumes that the PMT single-photoelectron calibration, established with an LED, accurately gives the charge for the TPB-shifted 420 nm scintillation pulses recorded in the physics runs; if it does not, every reported light yield scales by a common factor that the paper does not estimate.","fun_headline_variants_meta":{"raw":{"variants":["83mKr: internal calibrator for liquid argon","Internal 83mKr calibrates LAr detectors","83mKr and 22Na probe LAr field effects","Liquid argon self-calibrates with 83mKr","83mKr: the built-in yardstick for LAr detectors"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001047,"raw_usage":{"total_tokens":4435,"prompt_tokens":1017,"completion_tokens":3418,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":633,"completion_tokens_details":{"reasoning_tokens":3335}},"tokens_in":633,"tokens_out":3418,"duration_ms":25110,"temperature":1.0,"reasoning_tokens":3335,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T04:57:15.900672+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"One concrete check: recalibrate the same detector's light yield using an independent method, such as a collimated gamma source with a known emission rate and a Monte Carlo of the geometry, or a precision source of single photons with known quantum efficiency. If the resulting photoelectrons-per-keV differs from 7.26 at 41.5 keV by more than the quoted statistical uncertainty, the LED-based single-photoelectron scale is wrong. Alternatively, using a second PMT with a different photocathode or TPB configuration would reveal whether the field-quenching trend is a detector artifact or a genuine argon property.","supporting_citations":[{"cited_title":"Venos et al., 83Krm radioactive source based on 83Rb trapped in cation-exchange paper or in zeolite, Appl","cited_arxiv_id":null,"evidence_quote":"Provides the $^{83}$Rb to $^{83m}$Kr decay properties and the 1.83 h half-life against which the measured decay is compared."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Established the use of $^{83m}$Kr as a calibration source in noble-liquid detectors, the method this paper adapts to liquid argon."},{"cited_title":"Bellamy et al, 1994 Absolute calibration and monitoring of a spectrometric channel using a photomultiplier, Nucl","cited_arxiv_id":null,"evidence_quote":"Supplies the PMT response function used to fit the single-photoelectron spectra."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the calibrated low-energy response of liquid argon to $^{83m}$Kr, providing a reference for the observed light yield."},{"cited_title":"Agnes et al","cited_arxiv_id":null,"evidence_quote":"Defines the PARIS recombination model used to interpret the field-induced yield quenching."},{"cited_title":"Agnes et al, Measurement of the liquid argon energy re- sponse to nuclear and electronic recoils, Phys","cited_arxiv_id":null,"evidence_quote":"Provides the measured LAr energy response to electronic recoils used to explain the energy-dependent quenching."}],"review_version":1}