{"id":"734b7d72-d257-4e31-9e4c-ee0ad056b8e5","arxiv_id":"2507.23003","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":10,"one_line_summary":"Spurious single- and few-electron events in the DarkSide-50 argon TPC are mostly delayed electron releases from impurities (5-50 ms), and multi-electron events may come from grid photo-ionization.","lead":"This paper reports the first systematic study of spurious electron signals in an argon dark matter detector, using data from DarkSide-50. It finds that most of these background events are delayed releases of electrons from impurities, and that multiple-electron events may come from photo-ionization of the detector's steel grid.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Main causal claim (30-70% of SEs from impurity capture/release) is an inference from correlations; the paper itself defers confirmation to spiked-impurity studies, so acceptance should remain conditional.","rationale":"The paper's phenomenological findings are strong: the SE rate is temporally, spatially, and energetically correlated with preceding events; the multiplicity is Poisson; the getter-off and radon-trap correlations are suggestive; and the work is the first systematic argon-TPC characterization of this background. None of those findings are in question here. The load-bearing weakness is the causal step from 'correlated delayed electron emission' to 'electrons captured by impurities and later released.' That step is not directly tested, and the paper explicitly calls for spiked-impurity studies to confirm it. The reader's weakest-assumption assessment identifies exactly this gap, and the manuscript itself contains the relevant limitation statement in Sec. 7.1, which should be weighed in the verdict. Because the authors are appropriately cautious and the requested confirmation is feasible, the correct disposition is not rejection but a conditional acceptance contingent on the mechanism test or a clearly weakened central claim. No internal inconsistency was found; the issue is underdetermination between impurity release and other metastable-reservoir mechanisms. The concrete spiking test would settle the ambiguity by checking whether added impurities generate the predicted dT components with linear concentration scaling and species-specific time constants.","tokens_in":21258,"tokens_out":11110,"duration_ms":153057,"concrete_test":"Run a controlled spiking study in a small liquid-argon TPC (or a dedicated DS-50-like detector) with purified baseline and then fixed, known concentrations of candidate impurities (N2, O2, H2O, and TPB), holding electric fields, trigger, and circulation conditions fixed. Measure the SE-parent dT spectrum and Ne-distribution at each concentration. If the impurity-capture/release hypothesis is correct, each spiked species should introduce a new exponential dT component whose amplitude scales linearly with concentration and whose time constant is reproducible and species-specific; toggling the getter and radon-trap conditions should correspondingly modulate the associated component. If the dT spectrum is unchanged by spiking, the correlated SE population is not impurity-driven and the 30-70% attribution must be revised.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that a significant fraction of spurious-electron events are 'caused by electrons captured from impurities and later released' rests on interpreting the temporally correlated SE component as impurity anion capture/autodetachment. The supporting evidence is correlational: exponential dT components (Fig. 6, Eq. 2), parent S2 and drift-time scaling (Figs. 12, 14), the getter-off spike (Fig. 2), and the radon-trap temperature correlation (Fig. 16). These data establish that a large fraction of SEs have delayed, volume-distributed, ionization-fed parentage, but they do not directly identify an impurity species or demonstrate electron capture followed by delayed release. The same dT signatures could in principle be produced by any metastable electron reservoir that is populated by ionization and drains exponentially, including surface/interface traps, grid or electrode charge states, or long-lived molecular anions that are not gas-phase impurities. The paper itself acknowledges this in Sec. 7.1, stating that to 'confirm the hypothesis, identify impurities, and develop means to reduce SE events in future experiments, the SE rate needs to be studied with spiked impurity concentrations in a dedicated system.' The photo-ionization interpretation for multi-electron SEs is similarly left provisional in Sec. 6, where the authors note that 'further investigations of the pulse finder algorithm's behavior at low PE are needed to fully test this hypothesis.' Because the headline quantitative statement (30-70%) inherits this unconfirmed mechanism assignment, the causal attribution is the least secure link in the paper, even though the phenomenological correlations themselves appear robust.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper presents a systematic study of spurious-electron (SE) signals in the DarkSide-50 liquid argon TPC. Events with a single S2-like pulse and reconstructed charge below 4 e− are classified as SEs, and their rates, time correlations with preceding high-energy 'parent' events, purity metrics, and electron multiplicities are analyzed over roughly 1000 days of operation. The main results are: (i) a significant fraction of SEs, quoted as 30–70% across the run, is temporally correlated with parent events and is described by two exponential delay components with time constants of about 5 ms and 40–80 ms, plus an additional ~16 ms component during a getter-off period; (ii) the correlated SE probability scales with parent S2 size and drift time, giving a trapping probability of 1.74e-8 per electron per mm; (iii) the slow component's rate correlates with radon-trap temperature; and (iv) the SE multiplicity is Poisson-like with mean µSE = 0.062±0.001, consistent, within a factor, with photo-ionization of the steel grid by electroluminescence photons. The paper interprets the correlated population as delayed electron release from impurity molecules and discusses implications for S2-only dark matter searches.","tokens_in":21596,"tokens_out":9085,"duration_ms":102027,"significance":"If correct, this is the first comprehensive characterization of spurious electrons in argon and provides a quantitative basis for modeling the dominant S2-only background in low-mass dark matter searches. The main strength is the combination of several independent correlational handles: the getter-off response, the exponential dT distributions with time-ordered/random pair subtraction, parent-energy and drift-time scaling, spatial proximity, radon-trap temperature, and stability checks across the run and across Ne− bins. The paper is careful to define event classes and to verify that µSE does not depend on dT, tdrift, or S2. The 30–70% correlated fraction and the 1.74e-8 per mm trapping probability are concrete, falsifiable inputs for future argon TPCs and for spiked-impurity experiments. The grid photo-ionization hypothesis, while provisional, is useful because it yields a numerical prediction from external photoelectric data. The main limitations are that no impurity species is directly identified and that no systematic uncertainty budget is provided; these limitations are acknowledged in part by the authors in Sec.","major_comments":[{"comment":"The abstract states that 'a significant fraction of spurious-electron events ... are caused by electrons captured from impurities and later released', and the Summary (§8) repeats the mechanism as 'likely electrons released from impurities'. The evidence in §§3.3.1, 5.1, 5.3, 5.4 and 5.6 is correlational: it demonstrates a delayed, spatially correlated, ionization-fed population, but no impurity species is identified and no direct observation of capture followed by delayed release is made. The manuscript itself concedes in §7.1 that 'to confirm the hypothesis, identify impurities, and develop means to reduce SE events in future experiments, the SE rate needs to be studied with spiked impurity concentrations in a dedicated system.' Because the 30–70% fraction quoted in the abstract is a statement about a population attributed to this mechanism, the wording should be revised to present impurity capture/release as a well-motivated hypothesis consistent with the data, or supplemented with decisive spiking evidence.","section":"Abstract; §7.1"},{"comment":"All quoted uncertainties are stated in §3 to be purely statistical, and no systematic uncertainty analysis appears anywhere in the paper. The central quantitative claims—the correlated fraction (R1+R2)/R_SE, the time constants τ1 and τ2, and the 30–70% range in the abstract—depend on several modeling choices: the parent selection threshold S1>1000 PE, the 10 s pairing window, the two-exponential form of Eq. (2), the logarithmic binning, and the treatment of random coincidences. The uncertainty on g2=(23±1) PE/e− is also not propagated into Ne− boundaries or into the fitted rates. A systematic-error section that varies these choices (e.g., parent threshold, number of exponentials, window length, fitting range) and reports the resulting ranges is needed before the 30–70% fraction can be taken as a robust quantitative result.","section":"§3; §5.1; Fig. 8"},{"comment":"Section 5.6 states that deviations of R2 and the radon-trap temperature 'correlate with each other, offset by approximately 30 days', but no correlation coefficient, significance, or statistical test is reported. This correlation is one of the main supports for the impurity mechanism, and the 30-day offset is unexplained; a quantitative analysis (e.g., a time-shift scan with a reported p-value and a trials factor) is needed. In addition, the uncertainty on R2 shown as the residual comes from fits in §5.1 that carry no systematic component, so the visual agreement in Fig. 16 may overstate the significance of the correlation.","section":"§5.6; Fig. 16"}],"minor_comments":[{"comment":"The phrase 'gas pocket produciung S2 via electroluminescence' contains a typo: 'producung' should be 'producing'.","section":"§2"},{"comment":"The caption for Fig. 3 contains garbled fit text ('Const 0.05 ± 27.93 EL ... 0Rs µ 550 ± 8322 eτ'); please clean up the automated fit-parameter string.","section":"Fig. 3 caption"},{"comment":"The description of random pairs ('dT for parents that follow SEs in the same time span') is clear, but the relationship between the constant C in Eq. (2) and the explicit random-pair subtraction used in Fig. 9 should be stated explicitly so the reader can compare the two fitting procedures.","section":"§5.1"},{"comment":"The predicted grid photo-ionization probability of 2–4% is compared with µSE = 0.062 ± 0.001 with no error on the 'half of the UV photons directed towards the grid' assumption; please give this at least as a bracketing estimate (e.g., 0.25–0.5) and quote the resulting range.","section":"§6"},{"comment":"The statement that varying the fit range from 5 µs to 15 µs increases τℓ by up to 3% is useful; please also state whether this variation is included in the quoted τℓ values or only as a cross-check.","section":"§4.2"}],"recommendation":"major_revision","confidential_remarks":"This is a solid phenomenological background study, and the correlational evidence is strong enough to justify publication after revision. The main issue is that the abstract overstates the causal mechanism relative to what the data demonstrate; this is fixable by language changes and by adding the systematic-error analysis requested in the major comments. The radon-trap correlation needs a significance statement. I do not see grounds for rejection; the authors' own Sec. 7.1 already frames the mechanism as a hypothesis to be tested with spiked impurities."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First thing to know: this is the first systematic study of spurious electrons in an argon TPC, and the core phenomenology is solid. The group shows that SEs come with exponentially-distributed delays (about 5 ms and 40-80 ms, plus a 16 ms component during the getter-off run), that the delayed population scales with parent S2 and drift length, that the multiplicity is Poisson-like with a stable mean, and that the slow SE rate tracks the radon trap temperature. That is new and should become a reference for DarkSide-20k, ARGO, and anyone doing S2-only analyses. The contrast with xenon — power-law delays in XENON1T/LUX, exponentials here — is useful and appropriately cautious.\n\nThe paper earns credit for not overselling. Section 6 flags that the photo-ionization explanation for multi-electron SEs needs more pulse-finder work; Section 7.1 explicitly defers impurity identification to spiked-impurity studies in a dedicated system. Those are real limitations, not afterthoughts.\n\nThe soft spots are proportionate. The headline number — 30 to 70% of SEs caused by electrons captured from impurities — is an interpretation from correlations, not a direct measurement. The getter-off spike, radon-trap temperature correlation, dT shapes, and energy scaling make the impurity-capture story the natural reading, but no impurity species is identified and the abstract states the causal claim more firmly than the evidence allows. Also, only statistical uncertainties are quoted for the important ratios; the correlated fraction and the radon-trap correlation need systematic errors before the abstract number is taken at face value. The photo-ionization estimate leans on an assumed 0.5 photon fraction toward the grid and an external photoelectric yield; plausible but not decisive. No data or code release, which limits independent checking.\n\nThese issues are addressable and do not threaten the descriptive backbone. The timescales, rates, scaling laws, and multiplicity will stay. The causal hypothesis is clearly labeled as a hypothesis inside the paper.\n\nMy recommendation: send it to a serious referee. A good referee should push for systematic uncertainties on the central fractions, softer wording of the causal claim in the abstract, and a clearer separation between measured correlations and inferred mechanisms. I would cite this in future work and bring it to a reading group.","headline":"First systematic argon TPC spurious-electron study with solid phenomenology; the 30-70% impurity claim in the abstract goes beyond what the data can directly show, and the paper itself says so.","tokens_in":23024,"tokens_out":2456,"would_cite":true,"duration_ms":30846,"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":"The DarkSide-50 liquid-argon detector's spurious-electron background is largely caused by impurity molecules capturing drifting electrons and releasing them a few to tens of milliseconds later, with multi-electron events traced to grid…","keywords":["spurious electrons","dual-phase argon TPC","S2-only analysis","low-mass dark matter","impurity capture","delayed electron release","grid photo-ionization","DarkSide-50"],"falsifier":"A small argon TPC could spike the argon with one candidate impurity at a time—for example molecular nitrogen, water, or dissolved TPB—and check whether a specific exponential component (around 5 ms, 40–80 ms, or 16 ms) appears or grows; if the components do not respond to the added species, the impurity-capture explanation is wrong. Independently, replacing the steel grid with a material of much lower UV photoelectric yield should make the Poisson multi-electron peak with $\\mu_{\\mathrm{SE}} \\approx 0.06$ disappear if the grid photo-ionization mechanism is correct.","tokens_in":21046,"feed_emoji":"⚛️","tokens_out":10772,"duration_ms":115019,"temperature":0.7,"pith_summary":"The paper sets out to identify the origin of the small ionization-only signals that limit low-mass dark-matter searches in dual-phase argon time projection chambers. Using about 1000 days of data from the DarkSide-50 detector, it argues that 30–70% of these spurious-electron (SE) events are electrons that impurity molecules capture and release after a delay, with characteristic exponential time constants near $\\tau_1 \\approx 5$ ms and $\\tau_2 \\approx 40$–$80$ ms. It also argues that multi-electron SE events are consistent with electroluminescence photons photo-ionizing the stainless steel extraction grid. If both claims are right, the dominant background can be attacked directly by improving argon purity and by choosing a grid material that does not emit electrons under 128 nm light, which would let future experiments search for dark matter at lower energies.","feed_headline":"30-70% of spurious argon signals are trapped impurity electrons","feed_subtitle":"If right, purifying the argon and suppressing grid photo-emission could lower dark-matter search thresholds.","key_machinery":"The load-bearing tool is the time-difference distribution $dT$ between each SE and all preceding parent events, binned logarithmically and fit as a sum of exponential decay components $R_i/\\tau_i \\, e^{-dT/\\tau_i}$ plus a constant for random coincidences. The decay constants are the paper's handle on the delayed-release mechanism: a stable $\\tau_1 \\approx 5$ ms, a drifting $\\tau_2 \\approx 40$–$80$ ms, and a getter-off $\\tau \\approx 16$ ms component. This machinery separates temporally correlated SEs from accidental pairs, tracks the correlated rate against purification-system conditions and parent energy, and yields the per-mm trapping probability. A second mechanism is identified through a Poisson fit to the SE multiplicity, interpreted with the measured photoelectric yield of stainless steel to attribute multi-electron SEs to grid photo-ionization.","core_discovery":"Spurious-electron events—single S2-like pulses with fewer than four extracted electrons and no S1 pulse—are not a uniform detector artifact. The paper shows that the time separating an SE from the preceding high-energy 'parent' event follows a sum of exponentials: a fast component with $\\tau_1 \\approx 5$ ms, a slower component with $\\tau_2 \\approx 40$–$80$ ms that changes over the experiment's lifetime, and an additional $\\approx 16$ ms component that appears only while the hot getter is bypassed. The correlated component reconstructs within about 5 cm of its parent, grows linearly with the parent's S2 charge and drift distance (a trapping probability of $1.74 \\times 10^{-8}$ electrons per electron per mm of drift), and increases sixfold during getter-off operation. From these correlations the paper concludes that electrons are captured by impurity species and released milliseconds later, with evidence for at least three impurity-related components. The remaining 'uncorrelated' SE rate is proportional to the total ionization rate, which the paper reads as evidence that most of those, too, have parents on timescales beyond the 1 s coincidence window. The electron multiplicity is Poisson with mean $\\mu_{\\mathrm{SE}} = 0.062$, and the paper argues this matches the 2–4% per-electron probability that 128 nm electroluminescence photons ionize the steel grid, making grid photo-ionization the likely source of multi-electron SEs.","pith_inferences":["A testable extension the paper leaves implicit: spiking a small argon TPC with one candidate impurity at a time should reproduce a specific exponential time constant, and identifying which molecule gives the 5 ms and which gives the 40–80 ms components would turn the correlation into a direct measurement.","If delayed release from impurities is generic, the same mechanism should appear in xenon TPCs, but the paper's comparison suggests argon shows exponential delays while xenon shows power-law delays; a cross-detector study using identical analysis could separate universal trapping physics from species-specific chemistry.","The grid photo-ionization estimate implies a simple engineering rule for next-generation detectors: measure the UV photoelectric yield of the actual grid surface at 128 nm and keep it well below 1% to keep the multi-electron SE rate negligible.","Because the apparently uncorrelated SEs still scale with ionization, a low-background detector with a lower trigger rate might reveal parents at delays of seconds, making the entire SE population modelable as delayed ionization rather than intrinsic noise."],"forward_implications":["If 30–70% of SEs come from impurity capture, then keeping the purification system running and avoiding getter bypasses directly suppresses a large fraction of the background that currently forces S2-only analyses to discard events below about four electrons.","The exponential time constants provide a template: future argon TPCs can model the correlated SE population as a sum of exponentials and subtract it statistically, improving sensitivity to genuine low-energy ionization events.","If multi-electron SEs are grid photo-ionization, then engineering the extraction grid—material, coating, or field configuration—should reduce the multi-electron SE rate and allow the S2-only analysis threshold to drop.","Because even the 'uncorrelated' SE rate is proportional to total ionization, most SEs have parents on timescales longer than 1 s; a longer coincidence window or a lower event rate would expose more of that hidden parent population.","The getter-off spike with no measured electron-lifetime degradation shows that the impurity responsible for the $\\sim 16$ ms component is not the same as the electron-lifetime-limiting impurity, so purity metrics alone do not monitor this background."],"supporting_citations":[{"why":"Supplies the previously proposed delayed electron-extraction mechanism that this study tests in argon.","marker":"[20]"},{"why":"Reports a xenon experiment's observation of drifting electrons trapped by impurities, a comparative basis for the impurity-release claim.","marker":"[40]"},{"why":"Shows SEs in a xenon TPC are temporally and spatially correlated with preceding events, and the contrast clarifies the exponential behavior found here.","marker":"[57]"},{"why":"Measured the photoelectric yield of stainless steel at 124 and 130 nm, the quantitative input for the grid photo-ionization estimate.","marker":"[43]"},{"why":"Provides the electroluminescence yield and grid geometry used to compute the 2–4% per-electron photo-ionization probability.","marker":"[26]"},{"why":"Established the N$_2$ dependence of argon scintillation pulse shape, used to rule out >1 ppm N$_2$ as the SE impurity.","marker":"[36]"},{"why":"Gives the O$_2$ electron-attachment rate constant used to convert electron lifetime into equivalent impurity concentration.","marker":"[32]"},{"why":"The DarkSide-50 S2-only analysis that defines the single-electron gain $g_2$, the calibration at the heart of the SE classification.","marker":"[5]"},{"why":"The low-mass dark-matter search that restricted its energy range because of SE background, motivating the present study.","marker":"[15]"},{"why":"Provides drift speeds in the extraction field, used to estimate the ~1.4 $\\mu$s delay of secondary electrons from the grid.","marker":"[44]"}],"fun_headline_variants":["Impurity electrons: 30-70% of argon spurious signals","Argon spurious signals: impurity electron trapping explains 30-70%","Delayed impurity electrons: main source of argon spurious events","Grid photo-ionization linked to multi-electron argon spurious signals","Argon spurious events: impurity capture and delayed release"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The paper infers that impurity molecules capture and later release the electrons from timing, energy, position, and purification-system correlations, but it never directly detects the impurity species, so the mechanism stands on the assumption that no other correlated process can reproduce the same exponential delay pattern.","fun_headline_variants_meta":{"raw":{"variants":["Impurity electrons: 30-70% of argon spurious signals","Argon spurious signals: impurity electron trapping explains 30-70%","Delayed impurity electrons: main source of argon spurious events","Grid photo-ionization linked to multi-electron argon spurious signals","Argon spurious events: impurity capture and delayed release"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000822,"raw_usage":{"total_tokens":3654,"prompt_tokens":1057,"completion_tokens":2597,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":673,"completion_tokens_details":{"reasoning_tokens":2507}},"tokens_in":673,"tokens_out":2597,"duration_ms":25796,"temperature":1.0,"reasoning_tokens":2507,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T11:09:00.011414+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A small argon TPC could spike the argon with one candidate impurity at a time—for example molecular nitrogen, water, or dissolved TPB—and check whether a specific exponential component (around 5 ms, 40–80 ms, or 16 ms) appears or grows; if the components do not respond to the added species, the impurity-capture explanation is wrong. Independently, replacing the steel grid with a material of much lower UV photoelectric yield should make the Poisson multi-electron peak with $\\mu_{\\mathrm{SE}} \\approx 0.06$ disappear if the grid photo-ionization mechanism is correct.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Reports a xenon experiment's observation of drifting electrons trapped by impurities, a comparative basis for the impurity-release claim."},{"cited_title":"Aprile et al","cited_arxiv_id":null,"evidence_quote":"Shows SEs in a xenon TPC are temporally and spatially correlated with preceding events, and the contrast clarifies the exponential behavior found here."},{"cited_title":"Laulainen et al., AIP Conf","cited_arxiv_id":null,"evidence_quote":"Measured the photoelectric yield of stainless steel at 124 and 130 nm, the quantitative input for the grid photo-ionization estimate."},{"cited_title":"Acciarri et al","cited_arxiv_id":null,"evidence_quote":"Established the N$_2$ dependence of argon scintillation pulse shape, used to rule out >1 ppm N$_2$ as the SE impurity."},{"cited_title":"Bakale, U","cited_arxiv_id":null,"evidence_quote":"Gives the O$_2$ electron-attachment rate constant used to convert electron lifetime into equivalent impurity concentration."},{"cited_title":"Li et al., Nucl","cited_arxiv_id":null,"evidence_quote":"Provides drift speeds in the extraction field, used to estimate the ~1.4 $\\mu$s delay of secondary electrons from the grid."}],"review_version":1}