{"id":"5451ccf3-346e-46ab-81d2-37aadc3c83de","arxiv_id":"2505.09788","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"New effective collision strengths for Sr I, Sr III, Sr IV and Sr V are produced with R-matrix methods and used to propose Sr IV 1027.69 nm and Sr V 1203.35 nm kilonova diagnostics.","lead":"This paper computes new electron impact excitation data for four ionisation stages of strontium, Sr I, Sr III, Sr IV and Sr V, using established R-matrix atomic codes. The data are meant for modelling strontium emission in kilonova and white dwarf spectra, and the authors identify candidate Sr IV and Sr V lines in the near infrared.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 1203.35 nm Sr V kilonova diagnostic rests on inter-ion PEC comparisons the paper itself declares unphysical; no ionization balance supports the line's prominence.","rationale":"The reader's conditional verdict focuses on the lack of an independent benchmark for the collision strengths and on target-structure uncertainties, particularly the Sr V unshifted energy errors. Those are legitimate concerns, but they are somewhat generic for new R-matrix datasets and would require an expensive independent scattering or structure calculation to test. The concern I identify is sharper and internal: the paper's own PEC definition invalidates the central diagnostic claim as presented. If a single check coupling the ion stages shows Sr V is negligible, the paper's most prominent application claim is unsupported. I therefore agree partially with the reader: the dataset may still be conditionally acceptable, but the kilonova diagnostic framing must be fixed or removed. This does not change the verdict category, but it gives the condition a specific, testable target rather than leaving it as a general accuracy worry.","tokens_in":26077,"tokens_out":5081,"duration_ms":50703,"concrete_test":"Re-run the Section 4 colradpy spectrum calculation with all five Sr ion stages coupled through a Saha or full collisional-radiative ionization balance at T_e = 0.3-1.0 eV and n_e = 10^6-10^9 cm^-3, then weight each ion's PEC by its fractional abundance and compare the 1203.35 nm Sr V emissivity with the Sr I and Sr II lines. A minimal analytic version is sufficient: compute N(Sr V)/N(Sr I) from the Saha equation at 0.3 eV and 10^8 cm^-3. If this ratio is astronomically small, the 'stronger than Sr I/II' claim and the kilonova-diagnostic proposal do not survive, and the abstract and Section 4.1 must be revised.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's own Section 4.1 states that the ion populations are carried out independently for each ion stage, so the relative heights of spectral features between ion stages have no physical meaning; only PEC ratios within an ion stage are meaningful. Yet the same section presents the Sr V 1-2 line at 1203.35 nm as 'stronger than the other Sr I and Sr II lines presented in the full spectra' and the abstract offers it as a benchmark for the degree of Sr ionization in a kilonova. That comparison is precisely the inter-ion comparison the caveat forbids: each PEC is normalized to a unit ground-state population for its own ion, so a large Sr V PEC says nothing about whether Sr V exists in sufficient abundance to emit. The authors further concede that only trace Sr IV and Sr V are expected at KNe temperatures, given the ~3-4 Ryd ionization energies versus kT = 0.3 eV. Without a coupled ionization balance or an assumed abundance, the 'plausible kilonova ionization diagnostic' claim is unsupported as written. The underlying atomic data may still be useful, but the headline diagnostic inference should be retracted or explicitly recast as a conditional probe whose detectability depends on an unmodelled Sr V abundance.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents new atomic structure models and R-matrix electron-impact excitation calculations for Sr I, Sr III, Sr IV, and Sr V, together with energy levels, Einstein A coefficients, and Maxwellian-averaged effective collision strengths in adf04 format. The target structures are built with AUTOSTRUCTURE and GRASP0, and the computed energies and A-values are compared with NIST and with previous theoretical calculations. The new effective collision strengths are then used in the collisional-radiative code colradpy to generate synthetic spectra for the first five ionisation stages of Sr under kilonova and white-dwarf conditions. The authors identify candidate diagnostic lines, most prominently the Sr V 1-2 transition at 1203.35 nm, which they propose as a possible kilonova ionisation diagnostic.","tokens_in":26302,"tokens_out":3908,"duration_ms":38869,"significance":"If the collision data are reliable, this is a valuable contribution: it appears to be the first R-matrix treatment of electron-impact excitation for these four Sr ionisation stages, and the adf04 delivery format makes the data directly usable in NLTE modelling codes. The paper contains careful structure comparisons with NIST and with earlier calculations, and the A-value comparisons cover multiple independent theoretical approaches. However, the central deliverable—the effective collision strengths—has no independent benchmark in the paper, and the headline kilonova diagnostic claim is contradicted by the paper's own caveat about inter-ion PEC comparisons. The data may still be useful for modellers, but the diagnostic inference as currently stated is unsupported and needs to be substantially qualified or removed.","major_comments":[{"comment":"The paper's own caveat that 'the relative heights of spectral features between ion stages have no physical meaning' is violated by the central diagnostic claim. The abstract and Section 4.1 identify the Sr V 1-2 line at 1203.35 nm as a benchmark for the degree of Sr ionisation in a kilonova, and Section 4.1 states it is 'stronger than the other Sr I and Sr II lines presented in the full spectra.' Because each PEC is normalised to its own ion's ground-state population, this inter-ion comparison is exactly the one the caveat forbids: a large Sr V PEC says nothing about whether Sr V is sufficiently abundant to emit. The claim should be retracted or explicitly recast as conditional on an assumed Sr V abundance or on a coupled ionisation balance.","section":"§4.1 and Abstract"},{"comment":"The central deliverable, the Maxwellian-averaged effective collision strengths, has no independent benchmark. For every ion the authors state that no other collision data exist in the literature with which to compare. Given that the accuracy of the collision strengths is inherited from the target structures, and given the Sr V energy errors noted in Table 13, the accuracy of the collision strengths is unquantified. I request a sensitivity test (for example, comparing collision strengths from the unshifted versus shifted target, or a distorted-wave cross-check) or an explicit uncertainty statement, before these data are used as benchmarks for astrophysical diagnostics.","section":"§3.2–3.5"},{"comment":"The Sr V unshifted energy levels 2–5 differ from NIST by 19.9–24.4%. Since the scattering calculation uses levels shifted to their NIST positions, the target Hamiltonian and the shifted thresholds are inconsistent for these low-lying levels. For the 1203.35 nm (1–2) line, which is the headline diagnostic, this is particularly relevant because the transition lies within the ground configuration and the 24.4% error in level 2 affects threshold and resonance positions. The paper should discuss the possible impact of these large shifts on the effective collision strengths and, ideally, quantify it.","section":"§2.5, Table 13"}],"minor_comments":[{"comment":"The text says the first sample Sr V transition is at 92.2397 nm, but the Figure 9 caption lists 101.371 nm for the same transition (4s24p34d 3F4 -> 4s24p35p 3D3). Table 14 contains both wavelengths. Please align these values and ensure the figure caption describes the correct transitions.","section":"§3.5, Fig. 9"},{"comment":"The caption states that 'the average percentage between the shifted and unshifted levels was 1.250%,' but the table lists only unshifted AS energies and NIST energies. Please clarify what is being averaged and, if shifted energies are used, show them explicitly.","section":"Table 13 caption"},{"comment":"The text refers to 'bprmcodes'; this should likely be 'rmbp codes', matching the terminology used elsewhere in the paper.","section":"§3.5"},{"comment":"Several table captions refer to the wrong table number for the level indices: Table 5 refers to 'Table 2' instead of Table 4, Table 8 refers to Table 7, Table 11 refers to Table 10, and Table 14 refers to Table 13. Please correct these cross-references.","section":"Table captions 5, 8, 11, 14"},{"comment":"Equation (6) rescales A-values using NIST wavelengths, but it is not stated explicitly whether the A-values stored in the adf04 files are the rescaled values. Please state this clearly, since the collisional-radiative modelling in Section 4 uses these A-values.","section":"§2.2, Eq. (6)"}],"recommendation":"major_revision","confidential_remarks":"The paper's atomic data are likely useful, and the extensive structure comparisons are a strength. The main issue is that the headline kilonova diagnostic claim is not supported by the PEC modelling as presented, and the central collision strengths lack any validation. A major revision that removes or carefully qualifies the diagnostic claim and adds an uncertainty discussion for the collision strengths would make the contribution solid. I would also encourage the authors to consult whether the adf04 files are made available in a public repository at the time of revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The real deliverable here is a set of effective collision strengths for four Sr ions that did not have them before. That is genuinely useful for NLTE modelling of kilonovae and hot white dwarfs, and the paper does the work properly: target structures are described in detail, energy levels are matched to NIST, A-values are compared against every available calculation, and the R-matrix parameters (partial waves, mesh, top-up) are specified well enough to reproduce. The adf04 output is the right format for the community.\n\nThe soft spots are real but mostly fixable. The Sr V target has unshifted energy errors of 20–24% for the lowest excited levels; the scattering runs shift to NIST positions, which is standard practice, but it means the collision data inherit whatever the shifted structure does to the coupling. There is no independent collision strength benchmark for any of these ions, so accuracy rests entirely on the target models. That is not a fatal flaw, but it should temper the claims.\n\nThe bigger problem is the kilonova diagnostic. Section 4.1 explicitly says the ion populations are computed independently for each stage and therefore relative heights between ion stages have no physical meaning — only ratios within an ion stage count. A few paragraphs later the same section calls the Sr V 1–2 line at 1203.35 nm stronger than the Sr I and Sr II lines in the full spectrum, and the abstract offers it as a benchmark of the degree of Sr ionisation. Those are inter-ion comparisons. Each PEC is normalised to its own ground state, so a large Sr V PEC says nothing about how much Sr V exists. The authors also concede that only trace Sr IV and Sr V are expected at kilonova temperatures. Without a coupled ionisation balance or an assumed abundance, the 1203.35 nm line as an ionisation diagnostic is unsupported as written. That claim should be pulled back or explicitly recast as a conditional probe.\n\nThe circularity the reader flags — fitting orbital scaling to NIST energies, shifting levels, rescaling A-values from NIST wavelengths — is worth noting but is common practice in this field; it limits independent validation, it does not invalidate the data. The data availability is also thin: no hashes, no permanent archive, adf04 files not yet deposited. That should be fixed before the dataset becomes a reference.\n\nThis paper deserves a serious referee. The collision data are new and likely to be used. The referee should push for a corrected discussion of the PEC comparison, an explicit statement of what the 1203.35 nm line can and cannot diagnose, and proper archival of the data files. I would send it to review rather than desk reject, with revisions expected.","headline":"Useful new R-matrix effective collision strengths for Sr I, III, IV and V, but the 1203.35 nm kilonova diagnostic claim leans on inter-ion PEC comparisons the paper itself says are not physical, so that inference should be recast.","tokens_in":26849,"tokens_out":1308,"would_cite":true,"duration_ms":14684,"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":"This paper delivers the first Maxwellian-averaged electron-impact excitation data for Sr I, III, IV, and V, and identifies a Sr V line at 1203.35 nm as a candidate kilonova ionisation diagnostic.","keywords":["atomic data","electron-impact excitation","effective collision strengths","strontium","kilonovae","white dwarfs","R-matrix method","non-LTE modelling"],"falsifier":"A direct measurement of the electron-impact excitation cross-section for the Sr V ground-term forbidden transition at 1203.35 nm in a merged electron-ion beam or storage-ring experiment at energies around 0.1–1 Ryd; if the resulting Maxwellian-averaged effective collision strength differs from the paper's value significantly, the kilonova diagnostic prediction loses its quantitative basis. Alternatively, a high-resolution infrared spectrum of a kilonova at 1.15–1.25 µm across several epochs that shows no 1203.35 nm feature while Sr II lines evolve normally would call the modelled Sr V ionisation or excitation rates into question.","tokens_in":25868,"feed_emoji":"⚛️","tokens_out":11378,"duration_ms":99030,"temperature":0.7,"pith_summary":"Strontium lines show up in kilonova spectra and in the atmospheres of hot white dwarfs, but modelling those lines under non-LTE conditions requires electron-impact excitation rates that simply did not exist for most strontium ions. This paper fills that gap for four species: neutral Sr and the doubly, triply, and quadruply ionised stages Sr III, Sr IV, and Sr V. The authors build new target structures, check their energies and Einstein A-values against laboratory data and earlier calculations, and then compute R-matrix collision strengths, Maxwellian-averaged into effective collision strengths. When those data are fed into a collisional-radiative solver, a Sr V forbidden line at 1203.35 nm stands out as strong and density-sensitive within kilonova conditions, making it a plausible benchmark for how far strontium ionisation can go in a neutron-star merger. The paper also reports that most other potential diagnostic ratios across these ion stages fail, because the levels sit in either coronal or local-thermal-equilibrium conditions.","feed_headline":"New strontium collision data flag 1203 nm kilonova probe","feed_subtitle":"The 1203.35 nm Sr V line is a strong, density-sensitive candidate for tracking strontium ionisation in a kilonova.","key_machinery":"The machinery is the close-coupling R-matrix method for electron-impact excitation, in two implementations: the semi-relativistic Breit-Pauli package RMBP for Sr I and Sr IV, and the fully relativistic Dirac R-matrix package DARC for Sr III and Sr V. The output is the dimensionless collision strength $\\Omega_{i\\to j}$, which is then converted to the Maxwellian-averaged effective collision strength $$\\Upsilon_{i\\to j}(T_e) = \\int_0^\\infty \\Omega_{i\\to j} $e^{{-\\epsilon_j/kT_e}}$ d\\left(\\frac{\\epsilon_j}{kT_e}\\right).$$ The scattering runs use targets truncated to the lowest 57 (Sr I), 65 (Sr III), 70 (Sr IV), and 95 (Sr V) levels, all shifted to the experimental energies, with partial-wave top-up and Burgess–Tully infinite-energy points to handle the higher partial waves. A final, equally important component is the collisional-radiative solver that turns the atomic data into level populations and photon emissivity coefficients, which is what lets the authors scan for density- and temperature-sensitive lines.","core_discovery":"The central deliverable is a self-consistent set of atomic data — fine-structure energy levels, Einstein A coefficients, and Maxwellian-averaged effective collision strengths $\\Upsilon_{i\\to j}$ — for Sr I, Sr III, Sr IV, and Sr V, presented in adf04 format for use in NLTE codes. Each target structure was built with either AUTOSTRUCTURE or GRASP0, with orbital scaling parameters tuned to reproduce the measured energy levels; for the scattering calculations the levels were shifted to their spectroscopic positions. The authors argue the new data are urgently needed because previously only Sr II had full R-matrix excitation data, while other ions had to rely on approximate van Regemorter or Axelrod rates. Running the data through a collisional-radiative solver, they find that in kilonova conditions the spectrum is dominated by Sr I and Sr II, but six higher-ionisation lines appear between 400 and 1300 nm, including a single Sr IV line at 1027.69 nm and five Sr V lines, the strongest being the 1-2 forbidden transition at 1203.35 nm. Because the Sr V line's photon emissivity varies with electron density in the $10^{6}$–$10^{9}$ cm$^{-3}$ range, the authors propose it as a candidate diagnostic for the ionisation stage reached in an evolving kilonova. They also identify one Sr III line ratio (1-4/1-3, 55.49/56.28 nm) as a narrow-range density and temperature diagnostic for kilonova plasmas, and one Sr V ratio (1-34/2-6) for white dwarf conditions; most other ratios were found to be poor because the level populations follow coronal or LTE behaviour.","pith_inferences":["If the Sr V 1-2 line is confirmed observationally, its density sensitivity could be exploited as a time-resolved probe of the ionisation structure of a kilonova, something the paper does not itself demonstrate.","Because the collision data are the first of their kind, an independent calculation using a different structure basis (e.g., a B-spline or convergent close-coupling approach) or a beam experiment at low energies would be the natural test of the Sr V target, where the unshifted levels 2–5 deviate by 20–24% before shifting.","The near-infrared window around 1.2 µm where the Sr V line falls is also the region of the disputed Sr II P Cygni line in AT2017gfo; the new data could help future analyses separate Sr II from Sr V contributions.","The paper's finding that most levels are in coronal or LTE conditions implies that, for many lines, simpler equilibrium modelling suffices, so the main observational payoff of the new data may concentrate in a handful of transitions — exactly the Sr IV/V lines the paper highlights."],"forward_implications":["Kilonova NLTE models can now replace approximate excitation rates for Sr I, Sr III, Sr IV, and Sr V with R-matrix effective collision strengths, which is expected to change predicted line strengths, especially for forbidden transitions.","The 1203.35 nm Sr V line becomes a concrete prediction: if strontium is ionised that far in a neutron-star merger, this near-infrared line should appear and vary with ejecta density as the remnant evolves.","White dwarf models of hot stars can test their Sr V line identifications against the new A-values and collision data rather than relying on structure calculations alone.","The Sr III 1-4/1-3 ratio (55.49/56.28 nm) offers a narrow but usable density and temperature diagnostic for kilonova plasma, and the Sr V 1-34/2-6 ratio does the same for white dwarf conditions.","The data provide a benchmark for evaluating the older van Regemorter and Axelrod approximations, quantifying where those approximations fail for forbidden transitions."],"supporting_citations":[{"why":"Provides the experimental energy levels against which the target structures are tuned and to which the scattering levels are shifted.","marker":"Kramida et al. (2023)"},{"why":"Supplies the Sr II R-matrix data used to complete the first five ionisation stages and demonstrates the inaccuracy of approximate excitation rates for forbidden transitions.","marker":"Mulholland et al. (2024)"},{"why":"The collisional-radiative solver that converts the new atomic data into level populations, photon emissivity coefficients, and synthetic spectra.","marker":"Johnson et al. (2019)"},{"why":"Identified Sr V (and proposed Sr IV) lines in the hot white dwarf RE 0503-289; its A-values are used for comparison and its line list motivates the Sr V diagnostic search.","marker":"Rauch et al. (2017)"},{"why":"Prior GRASP0 structure calculation for Sr IV whose A-values are compared against the new target.","marker":"Aggarwal & Keenan (2015)"},{"why":"Prior Sr V structure calculation used for A-value comparison and for Stark-width work in hot white dwarfs.","marker":"Aloui et al. (2022)"},{"why":"Theoretical Sr III A-values (MCHF) used to validate the Sr III target.","marker":"Sureau et al. (1984)"},{"why":"Observed the Sr II P Cygni line in the kilonova AT2017gfo, establishing the observational context for the kilonova diagnostics.","marker":"Watson et al. (2019)"},{"why":"Identifies Sr I lines useful for stellar abundances, providing the transitions highlighted in the Sr I sample collision strengths.","marker":"Bergemann et al. (2012)"}],"fun_headline_variants":["Sr V 1203 nm line: new kilonova ionisation probe","New atomic data expose 1203 nm kilonova Sr tracer","Computed Sr collision data flag 1203 nm kilonova line","Strontium excitation data reveal 1203 nm kilonova tracer","Fresh Sr data point to 1203 nm kilonova diagnostic"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The collision strengths are only as good as the target structures, and for Sr V the fitted model's unshifted levels 2–5 are 20–24% off the measured values; everything then relies on shifting those levels to the measured positions, with no independent collision data anywhere to check the outcome.","fun_headline_variants_meta":{"raw":{"variants":["Sr V 1203 nm line: new kilonova ionisation probe","New atomic data expose 1203 nm kilonova Sr tracer","Computed Sr collision data flag 1203 nm kilonova line","Strontium excitation data reveal 1203 nm kilonova tracer","Fresh Sr data point to 1203 nm kilonova diagnostic"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000768,"raw_usage":{"total_tokens":3538,"prompt_tokens":1217,"completion_tokens":2321,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":833,"completion_tokens_details":{"reasoning_tokens":2226}},"tokens_in":833,"tokens_out":2321,"duration_ms":15819,"temperature":1.0,"reasoning_tokens":2226,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T21:23:52.908778+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A direct measurement of the electron-impact excitation cross-section for the Sr V ground-term forbidden transition at 1203.35 nm in a merged electron-ion beam or storage-ring experiment at energies around 0.1–1 Ryd; if the resulting Maxwellian-averaged effective collision strength differs from the paper's value significantly, the kilonova diagnostic prediction loses its quantitative basis. Alternatively, a high-resolution infrared spectrum of a kilonova at 1.15–1.25 µm across several epochs that shows no 1203.35 nm feature while Sr II lines evolve normally would call the modelled Sr V ionisation or excitation rates into question.","supporting_citations":[{"cited_title":"W., Demleitner M., 2017, @doi [Astronomy & Astrophysics] 10.1051/0004-6361/201730383 , 606, A105","cited_arxiv_id":null,"evidence_quote":"Identified Sr V (and proposed Sr IV) lines in the hot white dwarf RE 0503-289; its A-values are used for comparison and its line list motivates the Sr V diagnostic search."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Theoretical Sr III A-values (MCHF) used to validate the Sr III target."}],"review_version":1}