{"id":"ee67ebff-455c-4d00-b991-466300276505","arxiv_id":"2506.08301","paper_version":2,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Activation measurements of muon-capture radionuclide production on Al, Si, Co, and Ta show the PHITS dataset is generally conservative for radiation-safety estimates but requires corrections for isomer production, multiple neutron emission, and proton-emission reactions.","lead":"This experiment measured how often negative muons captured by aluminum, silicon, cobalt, and tantalum nuclei produce specific radioactive isotopes. The results show the Monte Carlo dataset used for radiation-safety estimates is generally conservative, but it badly misses three classes of reactions.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Absolute scale set by authors' own 27Mg reference; 'generally safe side' claim is sensitive to a ~10% normalization shift, though the three specific correction cases are robust.","rationale":"We agree with the reader's weakest_assumption. The single most load-bearing condition is the absolute scale, because the headline claim has two parts: the three correction cases are secure, but the 'generally safe' statement is an aggregate over C/E ratios that are close to unity. The normalization ties all measured probabilities together through the 27Mg reference, and the reference is the authors' own, so the calibration is not externally anchored. The cross-check with Wyttenbach provides some assurance but not enough to rule out a 10% scale shift that could flip the safety classification for borderline channels. This does not invalidate the paper's main empirical findings or the large discrepancies (factors of 20-50), so the verdict should remain CONDITIONAL as the reader concluded, with a request to either provide an independent N_mu measurement or qualify the 'generally safe' wording to reflect the normalization sensitivity. We do not see an internal inconsistency that would justify REJECT.","tokens_in":18609,"tokens_out":17640,"duration_ms":194648,"concrete_test":"Perform a renormalization sensitivity test: treat the true 27Mg reference P_ref as 8.9% and 10.9% (i.e., ±10%, about 2 sigma of the 5% uncertainty), rescale every measured P in Table 2 by P_ref(9.9)/P_ref(new) (equivalently rescale f_mu in Eq. (8)), and recompute all C/E ratios against the PHITS values. If any of the currently 'safe' channels (e.g., 23Ne, 25Na on 27Al; 56Mn, 59Fe on 59Co; 178mHf, 181Hf on natTa) crosses below C/E = 1, then the 'generally on the safe side' conclusion is not robust to the normalization and should be stated with a caveat; if all remain above 1, the concern is refuted for the general claim.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Every measured probability is normalized through Eq. (8) to the 27Mg production probability P=9.9(5)% taken from the authors' companion paper (Ref. [41], Mizuno et al.) via the conversion factor f_mu. No independent determination of N_mu (e.g., muonic X-ray counting) is performed, and because Ref. [41] is from the same collaboration (and likely the same experiment), the calibration is not independent. This is load-bearing because the 'generally on the safe side' conclusion rests on C/E ratios that are only modestly above unity: on 27Al, C/E = 1.04 (23Ne), 1.00 (25Na), 1.44 (26Na), 1.60 (27Mg), and only 24Na (C/E=0.58) is clearly below; similar margins hold on 59Co and natTa. A 2-sigma shift of the 5% reference uncertainty (i.e., ±10% in f_mu) would rescale all measured P and could push several near-unity C/E values across unity, changing the 'safe side' classification. The three proposed corrections (isomer production, multiple-neutron emission, proton-involving channels) survive such a shift because they are based on factors of 2-50, but the general statement is scale-sensitive. Cross-checks with Wyttenbach et al. [21] (independent counter-telescope normalization) are consistent but have uncertainties too large to constrain a 10% scale change. The paper should either obtain an independent N_mu or soften the 'generally safe' wording.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper reports activation measurements of radionuclide production probabilities per negative-muon nuclear capture on 27Al, natSi, 59Co, and natTa targets, combining in-beam gamma-ray counting (half-lives from about 1 s to about 1 h) with off-line counting (half-lives up to about 70 d). The measured probabilities are compared with a PHITS-based dataset (JAEA-Data/Code 2024-008) that has been incorporated into the SARE-MLF sample-radioactivity estimation program. The central claims are: (i) the dataset is \"generally on the safe side\" for radioactivity estimation, overestimating most measured channels by less than a factor of two; and (ii) three classes of channels need correction—isomer production (177mHf and 180mHf overestimated; 58mMn absent from the calculation), production by multiple-neutron emission at high excitation energy (175Hf underestimated by roughly a factor of 40; 173Hf absent), and proton-involving channels of the (mu-,nu_mu p2n) type (178mLu overestimated by about a factor of 20; 24Na underestimated by a factor of 1.6). The paper also reports consistency with previous activation experiments and interprets the long-standing 24Na conflict on 27Al as resolved in favor of the lower value of Heisinger et al.","tokens_in":18878,"tokens_out":18379,"duration_ms":199722,"significance":"If the results hold, they provide a useful expansion of a sparse experimental database, especially for high-Z targets, and the paper's mechanism-level analysis is a genuine strength. Each of the three identified deficiencies is traced to a specific model component (EBITEM level-structure data for isomers; the Singer excitation-energy distribution's high-energy tail for multiple-neutron emission; JQMD preequilibrium proton emission for the (mu-,nu_mu p2n) channels), and the Coulomb-barrier and threshold-energy analyses in Figs. 11-13 yield concrete, falsifiable predictions, such as the MEC-modified excitation distribution and the V-dependence of the p2n production probability. The 24Na measurement provides independent support for Heisinger et al. over Heusser et al. The experimental uncertainties are presented with statistical and systematic components separated, and the three correction cases (i)-(iii) are robust to the normalization issue because they rest on factors of 2-50.","major_comments":[{"comment":"The absolute scale of every measured probability in Table 2 is fixed by the conversion factor f_mu in Eq. (8), calibrated against P(27Mg)=9.9(5)% taken from the companion paper by the same collaboration (Ref. [41]); no independent determination of N_mu (e.g., by muonic X-ray counting, as used by Heisinger et al.) is reported. Because the calibration reference and the validated dataset (Ref. [12]) are both products of the same group, the 27Mg entry provides no independent anchor, and a bias in the reference value propagates coherently to all reported probabilities and all C/E ratios. This scale sensitivity is numerically relevant for the central claim: with C/E = 1.04 for 23Ne, 1.00 for 25Na, 1.21 for 59Fe, and 1.25 for 56Mn, a 5% (1-sigma) downward shift of the 27Mg reference moves the first two channels below unity, and a 10% (2-sigma) shift creates further exceptions to the \"generally on the safe side\" statement. The three proposed corrections, being based on factors of 2-50, survive such shifts. The paper should either add an independent N_mu determination or explicitly qualify the headline claim as holding modulo the common 5% reference scale and identify which channels would change classification under a +/-1-sigma scale shift; the text should also state that the f_mu uncertainty is 100% correlated across all rows of Table 2.","section":"§3 (Eq. (8)); §5; Abstract"},{"comment":"The 27Mg row on the 27Al target is not an independent experimental validation point. The measured value P(27Mg)=9.90(12)(79)% is the calibration input for N_mu via Eq. (8), so the displayed C/E = 1.60 merely compares the PHITS calculation with the assumed reference value from Ref. [41]. Including this point in the Section 5 statements that the calculations \"follow the general trend\" (and showing it without distinction in Fig. 8) overstates the number of independent model-data comparisons. The row should be explicitly labelled as the calibration anchor and either removed from the validation comparisons or clearly marked as non-independent.","section":"Table 2, Fig. 8, §5"},{"comment":"The natSi results are never compared with the PHITS dataset: the PHITS column in Table 2 is empty for all six natSi nuclides, and the C/E comparisons in Figs. 8-10 cover only 27Al, 59Co, and natTa. Since the natSi target was selected in part as a practical MLF sample element and includes the largest measured channel (28Al at 22.3(2)(18)%), the absence of a PHITS comparison leaves the validation claim, and the practical usefulness of the dataset for silicon sample estimation, unquantified for this target. The authors should add the calculated values and C/E ratios for the natSi channels or justify the omission in the text.","section":"Table 2 (natSi rows); §4-5"}],"minor_comments":[{"comment":"The validated dataset (Ref. [12], JAEA-Data/Code 2024-008) was produced by two of the present authors; the text should state this explicitly so that \"validation of the dataset\" is not misread as an independent model assessment.","section":"§1, §5 (Refs. [12], [41])"},{"comment":"The text quotes \"a factor of twenty\" for 178mLu (Table 2 gives 0.675(8)/0.030(4)(4) ≈ 22.5) and \"a factor of fifty\" for 175Hf (1.3(1)(2)/0.0313(8) ≈ 40); the quoted ratios should match the table values.","section":"§5"},{"comment":"The convention for off-scale C/E values is described only for large values (\"plotted at C/E = 2 by multiplying the factor in the parenthesis\"); the treatment of the strongly underestimated 175Hf point (C/E ≈ 0.024) should be specified in the caption as well.","section":"Fig. 10 caption"},{"comment":"Equation (3) uses a fixed dead time Td = 0.5 ms with no stated uncertainty or rate dependence; since this correction affects all in-beam results (including the short-lived 26Na, 30Al, and 177mHf channels), a sentence on its estimated uncertainty, or on why it is negligible, should be added to the systematic-error budget.","section":"§3, Eq. (3)"},{"comment":"The statement that the present experiment \"resolved the conflict\" for 24Na on 27Al is stronger than the evidence warrants: the new result is consistent with Heisinger et al. and inconsistent with Heusser et al., but the explanation offered for Heusser's discrepancy is interpretive; \"supports the lower value\" would be more precise.","section":"§5, §6"},{"comment":"The statement that Fig. 11(a) shows probabilities \"per threshold energy gap\" would benefit from one sentence describing the binning procedure (threshold energies from AME2020 and the definition of the gaps) to make the comparison reproducible.","section":"§5, Fig. 11"},{"comment":"The natSi data offer an internal cross-check opportunity, since 24Na and 27Mg are each measured on more than one target; a brief comment on whether these cross-target values are mutually consistent would strengthen confidence in the common normalization.","section":"§4, Table 2"}],"recommendation":"major_revision","confidential_remarks":"The recommendation is driven mainly by the normalization issue: the absolute scale of all measured probabilities is anchored to the authors' own 27Mg value (Ref. [41], same collaboration) and the validated dataset is the authors' own (Ref. [12]), making the validation partially self-referential. The three identified correction cases are solid and would survive an independent normalization, but the \"generally safe side\" claim needs either an independent N_mu determination (e.g., muonic X-ray counting) or a clearly stated scale-sensitivity caveat. The journal fit is good for NIM B, and the companion paper Ref. [41] is cited as an arXiv preprint whose publication status should be resolved before acceptance."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This is a careful activation measurement worth refereeing. The genuinely new part is the 59Co and natTa data, which extend muon-capture production probabilities to high Z and give the PHITS dataset something it didn't have: a set of C/E comparisons that point to specific model defects (EBITEM isomer handling, Singer neutron distribution, JQMD preequilibrium proton emission). The three identified deficiencies are supported by factors of 2–50 and survive any plausible normalization shift.\n\nWhat the paper does well: the experimental methods are standard and clearly described, systematic and statistical uncertainties are separated, and the authors are transparent about cumulative yields and the pion-contamination upper limit for 173Hf. The comparison with Wyttenbach et al. provides a useful independent cross-check. The resolution of the 24Na conflict on Al is credible: the Heusser value is shown to be suspect due to its theoretical normalization.\n\nThe soft spots are real but manageable. First, the absolute scale is set by the authors' own P(27Mg)=9.9(5)% from a companion paper, so the 27Mg comparison to PHITS (C/E=1.6) is not an independent validation. A 10% scale shift would move several near-unity C/E values across 1, so the 'generally safe side' claim is scale-sensitive. However, the claim still holds within uncertainties for most channels, and the three specific defects are not affected. Second, the 27Al and natSi data appear to overlap with the companion paper (Ref. [41]); the authors should state explicitly what is new versus previously reported. Third, the 'safe side' wording is a bit loose—the paper itself lists channels where the dataset underestimates (24Na, 175Hf, 173Hf), so 'generally safe' is doing a lot of work. That is not a fatal flaw, but the abstract could be more precise.\n\nFor whom: experimentalists using muon beams for activation or radiation safety, and anyone maintaining PHITS or similar Monte Carlo for muon capture. The paper deserves a serious referee; the issues are about bookkeeping and wording, not about the core data. I would recommend sending it to review with a request for clarification on the normalization and data overlap.","headline":"Solid activation measurements and a genuinely useful PHITS validation; the normalization dependence and data overlap need clarifying but don't undermine the main conclusions.","tokens_in":19487,"tokens_out":3264,"would_cite":true,"duration_ms":35974,"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":"By measuring radionuclide production probabilities in negative muon capture on aluminum, silicon, cobalt, and tantalum, this paper finds the PHITS-based dataset used for sample radioactivity estimation is generally conservative, but must…","keywords":["negative muon","nuclear capture","radionuclide production probability","activation experiment","PHITS","sample radioactivity","isomer production","multiple neutron emission"],"falsifier":"Count the stopped muons in the same irradiation independently by detecting muonic X-rays, as earlier activation experiments did, and recompute all probabilities and C/E ratios; if the muonic-X-ray count disagrees with the 27Mg-calibrated count by more than the quoted uncertainties, the absolute scale is biased and the 'safe side' conclusion for channels with C/E ratios near unity would need re-examination.","tokens_in":18375,"feed_emoji":"☢️","tokens_out":12650,"duration_ms":119317,"temperature":0.7,"pith_summary":"To keep sample-activation estimates honest at high-intensity muon facilities, this paper measures how often negative muons captured by a nucleus end up in specific radioactive isotopes, using targets of aluminum, natural silicon, cobalt, and natural tantalum. The measured probabilities are compared channel by channel with the Monte Carlo dataset that the facility's radioactivity-estimation program relies on. The comparison shows the dataset generally overestimates production, which is the safe direction for radiation-safety estimates, but it fails in three identifiable ways: isomer states are mispredicted, multiple-neutron-emission channels are badly underestimated at high excitation energies, and proton-emitting channels show almost no Coulomb-barrier suppression, producing a twenty-fold overestimate on tantalum. The measurements also settle an old disagreement over the 24Na production probability on aluminum, agreeing with the lower of the two earlier values. If these findings hold, the radioactivity code can be corrected in targeted ways rather than replaced.","feed_headline":"Radioactivity code passes muon-capture test with 3 exceptions","feed_subtitle":"On Al, Si, Co, and Ta, the PHITS dataset usually errs high; three channels need correction.","key_machinery":"The central object is the per-capture production probability $P$, obtained from activation counting with $P = N_\\gamma / (I_\\gamma \\varepsilon_\\gamma \\varepsilon_{\\mathrm{LT}} P_{\\mathrm{cap}} P_d N_\\mu)$, where $P_{\\mathrm{cap}}$ is the nuclear capture probability deduced from the total capture rate and $N_\\mu$ is the number of stopped muons. The absolute scale of every reported probability is fixed by calibrating the muon counter against the known $^{27}$Mg production probability $P = 9.9(5)\\%$ from the companion measurement, via the conversion factor $f_\\mu$ of Eq. (8). On the calculation side, the validated dataset is produced by a Monte Carlo chain in which the excitation energy imparted to the capturing nucleus is sampled from the Singer distribution with the Amado momentum distribution, the nucleon system evolves with the JQMD quantum molecular dynamics model, particle evaporation is described by the GEM model, and gamma-ray deexcitation including isomers is handled by EBITEM. The comparison is quantified as C/E ratios for each channel, and the paper isolates the failing physics by examining the threshold-energy dependence of multiple-neutron-emission probabilities and the Coulomb-barrier dependence of proton-emission probabilities.","core_discovery":"The central claim is that the PHITS-based dataset of radionuclide production probabilities from negative muon nuclear capture is, for the four tested targets, generally on the safe side for radioactivity estimation—most calculated probabilities exceed the measured ones by less than a factor of two—but that the dataset needs correction in three cases: isomer production (for example, 177mHf and 180mHf are overestimated by more than a factor of two, and 58mMn is not produced at all in the calculation), radionuclide production by multiple neutron emission (175Hf from high-threshold channels is underestimated by a factor of fifty, and 173Hf is missing entirely from the calculation), and radionuclide production by particle emissions involving a proton (178mLu is overestimated by a factor of twenty on tantalum, while 24Na on aluminum is underestimated by a factor of 1.6). The paper attributes the isomer problems to incomplete level-structure data in the gamma-deexcitation model, the multi-neutron problems to the too-sharp high-energy tail of the excitation-energy distribution, and the proton problems to the preequilibrium emission model, which yields calculated proton-emission probabilities that are almost independent of the Coulomb barrier. The measured 24Na probability of 1.93% agrees with the earlier activation value of 2.1% and resolves the conflict with the higher 3.5% value.","pith_inferences":["If the 27Mg reference value is ever revised, the scale of all probabilities reported here would shift, but the largest discrepancies (factors of 20 and 50) would likely survive, so the qualitative correction cases are robust to that systematic.","The Coulomb-barrier independence of the calculated proton-emission probabilities implies the preequilibrium model is missing a surface or barrier-suppression effect; a direct test would be to measure proton-emitting channels on intermediate-mass targets where evaporation and preequilibrium contributions are comparable.","Because the validation now covers light through heavy nuclei, the same activation approach could be applied to the actual sample materials used in muon spin rotation experiments, such as magnesium and titanium, to validate the radioactivity estimates for those routine irradiations."],"forward_implications":["For the four tested targets, the existing Monte Carlo dataset used in sample radioactivity estimation gives conservative (overestimated) activities in most channels, so no wholesale replacement is needed before use.","The 24Na production probability on aluminum is about 1.9%, resolving a previous discrepancy between two earlier activation measurements and supporting the lower of the two values.","The three documented failure modes—isomer production, multiple-neutron emission at high threshold energy, and proton-involving channels—require targeted corrections to the gamma-deexcitation model, the excitation-energy distribution, and the preequilibrium proton-emission model, respectively.","Additional activation measurements on other targets, especially those probing high threshold energies and isomer channels, are needed to turn the identified corrections into an updated dataset.","The method of plotting production probabilities per threshold-energy gap can be applied to future data to map where the calculated excitation distribution departs from reality."],"supporting_citations":[{"why":"supplies the reference 27Mg production probability used to calibrate the stopped-muon count and set the absolute scale","marker":"[41]"},{"why":"is the PHITS-based dataset of nuclide production probabilities whose validation is the paper's purpose","marker":"[12]"},{"why":"describes the muon interaction models in PHITS, including nuclear capture, on which the calculated dataset relies","marker":"[14]"},{"why":"gives the Singer excitation-energy distribution whose high-energy tail is implicated in the multiple-neutron-emission underestimation","marker":"[43]"},{"why":"is the JQMD quantum molecular dynamics model used for the preequilibrium stage, implicated in the proton-emission Coulomb-barrier problem","marker":"[45]"},{"why":"is the EBITEM gamma-deexcitation model whose level-structure data limit is implicated in the isomer production discrepancies","marker":"[47]"},{"why":"provides the previous activation result for 24Na on aluminum that agrees with the present measurement","marker":"[19]"},{"why":"supplies the total nuclear capture rates and Huff factors used to compute the capture probability Pcap","marker":"[40]"},{"why":"provides the in-beam activation method and the threshold-energy-dependence comparison for multiple-neutron-emission channels","marker":"[23]"}],"fun_headline_variants":["Muon capture: PHITS data mostly safe, needs 3 corrections","Negative muon capture: 3 flaws in Monte Carlo radioactivity model","PHITS muon-capture dataset: safe side but 3 exceptions","Muon capture radioactivity: model passes, but 3 channels off","Muon-capture simulation: 3 cases need correction"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The absolute scale of all measured probabilities rests on a single reference: the companion paper's 9.9% production probability for 27Mg from muon capture on aluminum, which is used to convert the muon-counter charge into the number of stopped muons; a bias in that reference would shift every reported probability and every calculated-to-experimental ratio together.","fun_headline_variants_meta":{"raw":{"variants":["Muon capture: PHITS data mostly safe, needs 3 corrections","Negative muon capture: 3 flaws in Monte Carlo radioactivity model","PHITS muon-capture dataset: safe side but 3 exceptions","Muon capture radioactivity: model passes, but 3 channels off","Muon-capture simulation: 3 cases need correction"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000205,"raw_usage":{"total_tokens":1431,"prompt_tokens":1023,"completion_tokens":408,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":639,"completion_tokens_details":{"reasoning_tokens":317}},"tokens_in":639,"tokens_out":408,"duration_ms":5411,"temperature":1.0,"reasoning_tokens":317,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T05:14:48.096924+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Count the stopped muons in the same irradiation independently by detecting muonic X-rays, as earlier activation experiments did, and recompute all probabilities and C/E ratios; if the muonic-X-ray count disagrees with the 27Mg-calibrated count by more than the quoted uncertainties, the absolute scale is biased and the 'safe side' conclusion for channels with C/E ratios near unity would need re-examination.","supporting_citations":[{"cited_title":"Measurement of production branching ratio after muon nuclear capture reaction of Al and Si isotopes","cited_arxiv_id":"2507.19753","evidence_quote":"supplies the reference 27Mg production probability used to calibrate the stopped-muon count and set the absolute scale"},{"cited_title":"Ikeda, J-parc status update, Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment 600 (1) (2009) 1–4","cited_arxiv_id":null,"evidence_quote":"is the PHITS-based dataset of nuclide production probabilities whose validation is the paper's purpose"},{"cited_title":"Niita, S","cited_arxiv_id":null,"evidence_quote":"is the JQMD quantum molecular dynamics model used for the preequilibrium stage, implicated in the proton-emission Coulomb-barrier problem"},{"cited_title":"Ogawa, S","cited_arxiv_id":null,"evidence_quote":"is the EBITEM gamma-deexcitation model whose level-structure data limit is implicated in the isomer production discrepancies"},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"provides the previous activation result for 24Na on aluminum that agrees with the present measurement"},{"cited_title":"Wu, Nuclear data sheets for a = 181, Nuclear Data Sheets 106 (3) (2005) 367–600","cited_arxiv_id":null,"evidence_quote":"supplies the total nuclear capture rates and Huff factors used to compute the capture probability Pcap"},{"cited_title":"Niikura, T","cited_arxiv_id":null,"evidence_quote":"provides the in-beam activation method and the threshold-energy-dependence comparison for multiple-neutron-emission channels"}],"review_version":1}