{"id":"7289c963-0167-4165-bca1-120d57ad7c48","arxiv_id":"2412.16688","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A complete HFR opacity dataset for Z=20-103 is published, and for the modeled merger the abundant d-shell elements, not lanthanides, dominate the average kilonova ejecta opacity.","lead":"This paper releases a full set of atomic opacity tables for every element from calcium to lawrencium, computed for kilonova conditions one to seven days after a neutron star merger. For one merger model, abundant lighter elements such as chromium and zirconium block as much light as the famous lanthanides, so kilonova interpretations may need to change.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Composition weighting in Eq. (9) scales single-element LTE level densities by abundance without recomputing Saha with the mixture electron density; the 1.13 cm2/g result and the 'a posteriori averaging fails' claim rest on this untested mixing approximation.","rationale":"The paper is a serious and valuable extension of the authors' HFR program: it provides a public database, compares with Tanaka et al. and Fontes et al., and documents convergence checks and calibration studies. My disagreement with the reader's identification of the weakest assumption is deliberate. The reader highlighted the single abundance pattern and atomic-data uncertainties; these are real scope caveats, but the paper already largely scopes the lanthanide result to one model and to the 'average' mixture. The more fundamental issue is internal to the mixing method used to produce the very numbers supporting the headline. Equation (9) applies a linear abundance scaling to pure-element level populations while the underlying Saha system, Eq. (5), is nonlinear in electron density. A trace species in the mixture sees the mixture electron density, not the electron density of its own pure plasma; the paper does not show that freezing pure-element ionization fractions is accurate. This is not an external-consensus disagreement but a potential correctness gap in the central demonstration. The proposed multi-element Saha computation is a direct, computationally feasible check. If it confirms Eq. (9), the reader's conditional acceptance remains appropriate; if it does not, the numerical conclusions and the 'cannot be combined a posteriori' claim need reworking.","tokens_in":90307,"tokens_out":14282,"duration_ms":141407,"concrete_test":"Using the publicly archived HFR line lists and the sym-n1-a6 abundances, solve the coupled Saha equations for all elements simultaneously with one electron density fixed by charge neutrality at T = 6000 K, rho = 1e-13 g/cm3, t = 3.5 d, and recompute the Planck-mean expansion opacity from the combined line list. Compare the total and the per-element contributions (especially lanthanides/actinides versus Z = 24 and Z = 40) with the Eq. (9) values. If the total shifts by more than about 30% or the lanthanide+actinide partial contribution becomes comparable to or larger than the d-shell contribution, the headline claim must be revised; a small shift would validate the present approximation quantitatively.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central quantitative claim—that the composition-weighted HFR Planck expansion opacity for sym-n1-a6 is 1.13 cm2/g while abundance-weighted single-element opacities give 0.11 cm2/g—is derived from Eq. (9), not from a true multi-element LTE mixture calculation. Section 6 defines n_l,Z = y_Z n_l, where n_l is the lower-level number density from a pure-element calculation at the same T and bulk density. But the Saha equation, Eq. (5), depends on the electron density n_e, and the electron density of a pure lanthanide plasma at rho = 1e-13 g/cm3 is not the same as the electron density of the full mixture in which lanthanides are trace species (8.4e-5 molar fraction). Multiplying pure-element level densities by y_Z rescales the element's own electron contribution but leaves the ionization balance frozen at the pure-element value; it does not solve Eq. (5) with a common n_e and charge conservation. The method therefore assumes, without demonstration, that ion fractions of every element in the mixture equal those in its own pure plasma at the same bulk density. If this assumption fails, the partial contributions shown in Figs. 9-10, the total opacity 1.13 cm2/g, and the comparison with 0.11 cm2/g are not established. Since the claim that single-element opacities cannot be combined a posteriori is specifically about the difference between these two mixing procedures, this untested step is load-bearing.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper presents new pseudo-relativistic Hartree-Fock (HFR) atomic structure calculations for all elements from Ca (Z = 20) to Lr (Z = 103) in ionization stages I-IV, together with expansion and line-binned opacities for a kilonova-relevant grid (T = 1000-10000 K, rho = 1e-17 to 1e-13 g/cm^3, t = 1-7 days). The data are made publicly available on Zenodo. The authors compare their Planck mean opacities with Tanaka et al. (2020) and Fontes et al. (2020, 2023), and apply the opacities to the neutron star merger model sym-n1-a6 from Just et al. (2023). Their central result is that, for this model, the composition-weighted HFR expansion opacity is 1.13 cm^2/g, about 25% lower than the parametric prescription of Just et al. (2022), and that lanthanides are not the dominant contributors 'at least on average'; instead, 3d-shell (Z ~ 24) and 4d-shell (Z ~ 40) elements contribute substantially. They also find that abundance-weighted single-element Planck opacities (0.11 cm^2/g) underestimate the mixture opacity by an order of magnitude, and that HFR opacities shift the modeled bolometric peak from roughly 1-2 days to roughly 7-8 days.","tokens_in":90488,"tokens_out":6234,"duration_ms":53905,"significance":"The database is a major community resource: consistent HFR oscillator strengths and energy levels for 84 elements in four charge states are not currently available elsewhere with this coverage. The comparison with existing data sets and the public release are valuable. If the central claim survives scrutiny, it challenges the common practice of treating lanthanide opacity as the sole determinant of kilonova red/infrared emission and the practice of building mixture opacities by abundance-weighting single-element tables. The result is, however, conditional on the adopted nucleosynthesis composition and on the composition-mixing approximation discussed below. The manuscript is generally careful in hedging ('for a given model', 'at least on average'), but the headline result deserves stronger validation.","major_comments":[{"comment":"The composition-weighted opacity is constructed by scaling pure-element lower-level densities n_l by the elemental molar fraction y_Z, i.e., n_{l,Z} = y_Z n_l. This does not solve the Saha equation (Eq. 5) with a common electron density n_e and charge conservation for the mixture; it implicitly assumes that each element's ionization balance in the mixture is identical to that in a pure plasma of that element at the same bulk density. Because n_e in a mixture is dominated by the abundant lighter elements, while the trace lanthanides (molar fraction 8.4e-5 in the full ejecta) have their own pure-plasma n_e, the ion fractions of the important species are not established. The claimed totals (1.13 cm^2/g for the full ejecta, 5.42 cm^2/g for the dynamical component) and the factor-10 discrepancy with the abundance-weighted single-element value (0.11 cm^2/g) therefore rest on an untested approximation. Please either implement a self-consistent multi-element LTE calculation for the sym-n1-a6 composition, or provide a quantitative demonstration that the pure-element Saha fractions are insensitive to n_e over the relevant range.","section":"§6, Eq. (9)"},{"comment":"The paper states that HFR energy-level discrepancies of 18-30% (49% for Nd III) and inverted ground states in 23 of the 120 lanthanide/actinide ions leave expansion opacities essentially unaffected, but this claim is demonstrated only for Nd II/III, U II/III, and Er III. The extrapolation to all 429 ions is a load-bearing step because the main astrophysical conclusion depends on the relative opacities of d-shell, lanthanide, and actinide species. A sensitivity test covering representative ions from each group (e.g., Cr, Mo, Sm, U) with and without level-energy calibration, or a propagated uncertainty estimate on the total mixture opacity, would substantially strengthen the robustness of the result.","section":"§2"},{"comment":"The 'lanthanides are not dominant' conclusion is drawn from a single r-process abundance pattern, model sym-n1-a6 with the BSkG3 mass model. The lanthanide-plus-actinide molar fraction in this model is 8.4e-5 in the full ejecta and 8.7e-4 in the dynamical component. Because single-element lanthanide/actinide opacities exceed those of the abundant d-shell elements by several orders of magnitude (Section 3), the ranking is a cancellation between per-element opacity and abundance. Published r-process calculations with other mass models or fission treatments show lanthanide fractions varying by orders of magnitude. The paper should either test the sensitivity of the ranking and of the 1.13 cm^2/g total to variations in the lanthanide/actinide fraction within the range of current models, or explicitly narrow the claim to this specific nucleosynthesis model in the abstract.","section":"§6"}],"minor_comments":[{"comment":"The density unit in the figure captions is given as '10^-13 cm^-1'; it should be '10^-13 g cm^-3' as in the text.","section":"Figure captions 2-5"},{"comment":"Typos: 'The purpose of this work is to is to build' and 'impoortance' should be corrected; 'Eriii' should be 'Er III'; in the reference list, 'Astronomy & Astroohysics' should be 'Astronomy & Astrophysics'.","section":"Section 7"},{"comment":"The reference to 'KILONOV A' should be 'KILONOVA'.","section":"Acknowledgements"},{"comment":"The Planck mean opacity definition uses B_lambda(T) in both numerator and denominator; it may be worth noting explicitly that the integration is over the full wavelength range, as stated in the text, to avoid confusion with the line-binned version.","section":"Eq. (7)"}],"recommendation":"major_revision","confidential_remarks":"The paper is potentially important and the database is a useful contribution. The main reservations are the unvalidated composition-mixing approximation in Eq. (9) and the extrapolation of the level-error insensitivity to all ions. If the authors can provide a true multi-element LTE test for one or two compositions and a sensitivity analysis on the abundance pattern, the paper would be suitable for publication. The current version's central claim is conditional, but the abstract may overstate the generality; the authors should ensure the hedging is visible in the abstract and conclusions."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"I have two takeaways for you. The paper delivers a genuinely useful resource: consistent HFR atomic data and opacity tables for all elements Z=20-103 in ionization stages I-IV, published on Zenodo, with full configuration lists in the appendix. And its headline claim—lanthanides are not the dominant opacity source, 'at least on average'—is honest but rests on one r-process abundance pattern, which the authors acknowledge.\n\nWhat is actually new: the complete periodic-table coverage with one consistent model set; the composition-weighted treatment of expansion opacity; and the demonstration that abundance-weighting single-element tables (0.11 cm2/g) is an order of magnitude below their recomputed mixture value (1.13 cm2/g) for the sym-n1-a6 model. The comparisons against Tanaka et al. (HULLAC) and Fontes et al. (line-binned) are fair, and the explanations offered for the lanthanide discrepancy—restricted configuration sets and approximated partition functions in the earlier work—are credible. Nothing here is fitted to a target; the opacities are derived from structure calculations. That should be credited.\n\nThe soft spots, in order of weight. First, the lanthanide conclusion depends on one abundance pattern: sym-n1-a6 with the BSkG3 mass model, where the lanthanide-plus-actinide fraction is 8.4e-5 in the full ejecta. With other mass models or fission prescriptions that number moves by orders of magnitude, and the ranking could flip. Second, the stress-test concern about Eq. (9) has real content: multiplying pure-element level densities by y_Z keeps the Saha ionization balance frozen at pure-element electron densities instead of solving the mixture with a common n_e. I think this is a genuine gap, but it does not overturn the qualitative conclusion—single-element tables still cannot be combined a posteriori, because the expansion opacity is nonlinear in the level densities and saturation physics alone produces an order-of-magnitude effect. The factor of ten is just not pinned down precisely until someone does the full multi-element solve. Third, there is no uncertainty propagation despite documented 18-30% level errors (49% for Nd III) and inverted ground states in 23 of 120 ions, with the claimed robustness verified for only Nd, U, and Er. Fourth, the computed light curve peaks at 7-8 days versus the observed AT2017gfo peak near 1-2 days; the paper shows the shift but does not reconcile it.\n\nWho should read it: kilonova light-curve modelers and anyone comparing opacity methods. It deserves a serious referee. I would ask the authors for uncertainty quantification on the tables, a test of Eq. (9) against a true mixture Saha solve, an explicit abstract caveat on the single abundance pattern, and a statement on what the 7-8 day peak implies for the absolute opacity scale. Send it to review.","headline":"A solid, citable full-periodic-table opacity resource whose main claims hold up; the lanthanide conclusion depends on one abundance model, and the composition-mixing procedure deserves a proper multi-element Saha test.","tokens_in":91223,"tokens_out":14830,"would_cite":true,"duration_ms":121932,"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":"Lanthanides do not dominate kilonova opacity, new atomic data show.","keywords":["kilonova","neutron star merger","opacity","atomic data","lanthanides","actinides","r-process","expansion opacity"],"falsifier":"Recompute the full-ejecta Planck opacity at $t=3.5$ d, $T=6000$ K, $\\rho=10^{-13}\\ \\mathrm{g\\,cm^{-3}}$ with the paper's HFR tables but with r-process abundances from a different nuclear mass model or fission prescription; if the lanthanide and actinide contribution then exceeds the d-shell contribution, the average non-dominance claim fails. A second check is to compare the predicted $\\sim$7-8 day bolometric peak with the observed AT2017gfo light curve; if the HFR-opacity model is conclusively excluded by the timing, the opacity shift is not the right description.","tokens_in":89996,"feed_emoji":"💥","tokens_out":11790,"duration_ms":92550,"temperature":0.7,"pith_summary":"The paper builds a complete set of atomic data and opacities for every element from calcium ($Z=20$) to lawrencium ($Z=103$), in the first four ionization stages, using the pseudo-relativistic Hartree-Fock (HFR) method, for the temperature and density conditions of a kilonova photosphere one day to one week after a neutron-star merger. It then folds these per-element opacities into the r-process abundance pattern of a realistic merger simulation. The central claim is that, at least on average, lanthanides are not the dominant opacity sources: their single-element opacities are orders of magnitude larger than those of other elements, but their abundances in this model are so low that abundant 3d- and 4d-shell elements around $Z\\simeq 24$ and $Z\\simeq 40$ contribute just as much. The paper further argues that single-element expansion opacities cannot be averaged with abundances afterward, because that shortcut gives $0.11\\ \\mathrm{cm^2\\,g^{-1}}$ instead of the correct $1.13\\ \\mathrm{cm^2\\,g^{-1}}$. These opacities shift the predicted bolometric light-curve peak from about 1-2 days to 7-8 days after merger.","feed_headline":"Lanthanides do not dominate kilonova opacity, new atomic data show","feed_subtitle":"In a realistic merger ejecta, abundant elements near atomic numbers 24 and 40 matter as much as lanthanides, moving the peak to 7-8 days.","key_machinery":"The machinery is the composition-weighted expansion opacity, built on the standard Sobolev expansion formula $\\kappa_{\\mathrm{exp}}(\\lambda) = (1/ct\\rho)\\sum_l (\\lambda_l/\\Delta\\lambda)(1-e^{-\\tau_l})$ and its Planck mean. The defining move is to scale each element's lower-level number density inside the optical depth by its molar abundance, $n_{l,Z} = y_Z n_l$, so the mixture opacity is computed before summing over elements rather than by averaging completed single-element opacities. The atomic input comes from large multiconfiguration pseudo-relativistic Hartree-Fock calculations for ions I-IV of all elements from calcium to lawrencium; the paper argues that the level errors and inverted ground states these models leave (18-30% for most ions, 49% for Nd III) have little effect on expansion opacities, based on tests for Nd, U, and Er. The line-binned opacity $\\kappa^{\\mathrm{bin}}_\\nu\\propto \\sum_l N_l f_l$ scales linearly with abundance and therefore does not suffer from the a posteriori averaging problem.","core_discovery":"The discovery is a change in who controls kilonova opacity when the ejecta composition is taken seriously. For the sym-n1-a6 merger model at 3.5 days ($T=6000\\ \\mathrm{K}$, $\\rho=10^{-13}\\ \\mathrm{g\\,cm^{-3}}$), the composition-weighted HFR Planck opacity is $1.13\\ \\mathrm{cm^2\\,g^{-1}}$, about 25% below the $1.43\\ \\mathrm{cm^2\\,g^{-1}}$ obtained from a parametric formula that links opacity to the lanthanide-plus-actinide molar fraction. The opacity budget is shared among lanthanides and actinides and the much more abundant elements near $Z\\simeq 24$ and $Z\\simeq 40$. The paper's second result is methodological: because the expansion opacity is a nonlinear Sobolev sum over lower-level populations, replacing those populations by abundance-scaled ones inside the formula is not the same as averaging single-element opacities; the a posteriori average gives $0.11\\ \\mathrm{cm^2\\,g^{-1}}$, an order of magnitude too low. When the HFR opacities replace the parametric formula in light-curve simulations, the bolometric peak moves from roughly 1-2 days to about 7-8 days, with a slightly lower peak luminosity for the full ejecta.","pith_inferences":["The paper's lanthanide-non-dominance conclusion should be read as abundance-pattern dependent: replacing the BSkG3-based abundances with another r-process mass model or fission prescription, which can change lanthanide fractions by orders of magnitude, could restore lanthanide dominance, and this test is not performed in the paper.","The same composition-weighting failure likely affects any kilonova or transient model that stores Planck mean opacities per element and later mixes them with abundances; line-binned opacities, being linear in abundance, are the safer stored quantity for such post-processing.","Recomputing the light curve with per-trajectory compositions instead of the uniform composition assumed here could sharpen or soften the 7-8 day peak shift; the paper identifies this as future work.","If the non-dominance holds broadly, the atomic-data bottleneck shifts from all lanthanides to a short list of species (Cr, Fe, Zr, Mo, Sm, Nd, Dy, Er, U, Np, Pu), and improving those atoms experimentally and theoretically would be the highest-value next step."],"forward_implications":["Kilonova light curves computed with these opacities peak at roughly 7-8 days rather than 1-2 days, a shift that is directly testable against the observed evolution of AT2017gfo.","Lanthanide and actinide molar fraction alone is not a reliable opacity proxy; the parametric formula built on it is 25% high for the full ejecta and a factor of two low for the dynamical ejecta.","Single-element expansion opacity tables should not be combined with arbitrary abundance patterns a posteriori; the correct composition-weighted calculation changes the answer by an order of magnitude here.","The published grid of HFR atomic data and expansion/line-binned opacity tables for $Z=20$-103, ions I-IV, over $t=1$-7 days, $\\rho=10^{-17}$-$10^{-13}\\ \\mathrm{g\\,cm^{-3}}$, and $T=1000$-10000 K provides a common baseline for kilonova modeling.","When only the fast dynamical ejecta is considered, the HFR opacities give a brighter peak than the parametric opacity, so atomic data matter even in the lanthanide-dominated component."],"supporting_citations":[{"why":"Supplies the heuristic opacity formula and the ALCAR-kilonova light-curve code that the HFR opacities are compared against.","marker":"Just et al. (2022)"},{"why":"Provides the sym-n1-a6 neutron-star merger model, its ejecta components, and the r-process composition used for the average-opacity result.","marker":"Just et al. (2023)"},{"why":"Establishes the HFR multiconfiguration modelling strategy and the expansion-opacity convergence checks for Nd and U that the paper extends to all elements.","marker":"Flörs et al. (2023)"},{"why":"Supplies the HULLAC single-element Planck opacities for Z=26-88 used as the main expansion-opacity comparison.","marker":"Tanaka et al. (2020)"},{"why":"Supplies lanthanide line-binned Planck opacities used to validate the new HFR line-binned results.","marker":"Fontes et al. (2020)"},{"why":"Supplies actinide line-binned Planck opacities used to validate the new HFR line-binned results.","marker":"Fontes et al. (2023)"},{"why":"Provides the BSkG3 nuclear mass model that sets the r-process abundance pattern underlying the composition-weighted opacity.","marker":"Grams et al. (2023)"},{"why":"Provides the HFB-14 fission rates entering the nucleosynthesis calculation for the merger model.","marker":"Goriely et al. (2010)"},{"why":"Shows how approximate partition functions change expansion opacities, explaining differences from earlier work.","marker":"Carvajal Gallego et al. (2023b)"},{"why":"Shows that calibrating HFR levels to experiment or leaving inverted ground states uncorrected has minor impact on opacities for Er III.","marker":"Deprince et al. (2024)"}],"fun_headline_variants":["Lanthanides dethroned: abundant elements steer kilonova opacity","Kilonova opacity: lanthanides not dominant, new HFR data show","New atomic data move kilonova peak to 7-8 days post-merger","Composition-weighted opacity reshapes kilonova light curves","HFR calculations: abundant elements matter as much as lanthanides"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the single abundance pattern from the sym-n1-a6 merger simulation with the BSkG3 mass model is representative enough that the cancellation between low lanthanide abundance and high lanthanide opacity holds in real kilonova ejecta, because other r-process models vary that abundance by orders of magnitude.","fun_headline_variants_meta":{"raw":{"variants":["Lanthanides dethroned: abundant elements steer kilonova opacity","Kilonova opacity: lanthanides not dominant, new HFR data show","New atomic data move kilonova peak to 7-8 days post-merger","Composition-weighted opacity reshapes kilonova light curves","HFR calculations: abundant elements matter as much as lanthanides"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000299,"raw_usage":{"total_tokens":1833,"prompt_tokens":1151,"completion_tokens":682,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":767,"completion_tokens_details":{"reasoning_tokens":581}},"tokens_in":767,"tokens_out":682,"duration_ms":27063,"temperature":1.0,"reasoning_tokens":581,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T10:22:42.798233+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Recompute the full-ejecta Planck opacity at $t=3.5$ d, $T=6000$ K, $\\rho=10^{-13}\\ \\mathrm{g\\,cm^{-3}}$ with the paper's HFR tables but with r-process abundances from a different nuclear mass model or fission prescription; if the lanthanide and actinide contribution then exceeds the d-shell contribution, the average non-dominance claim fails. A second check is to compare the predicted $\\sim$7-8 day bolometric peak with the observed AT2017gfo light curve; if the HFR-opacity model is conclusively excluded by the timing, the opacity shift is not the right description.","supporting_citations":[{"cited_title":"2023, The Astrophysical Journal Letters, 951, L12","cited_arxiv_id":null,"evidence_quote":"Provides the sym-n1-a6 neutron-star merger model, its ejecta components, and the r-process composition used for the average-opacity result."},{"cited_title":"J., Fryer, C","cited_arxiv_id":null,"evidence_quote":"Supplies lanthanide line-binned Planck opacities used to validate the new HFR line-binned results."},{"cited_title":"J., Fryer, C","cited_arxiv_id":null,"evidence_quote":"Supplies actinide line-binned Planck opacities used to validate the new HFR line-binned results."},{"cited_title":"2023, Eur","cited_arxiv_id":null,"evidence_quote":"Provides the BSkG3 nuclear mass model that sets the r-process abundance pattern underlying the composition-weighted opacity."},{"cited_title":"2010, Phys","cited_arxiv_id":null,"evidence_quote":"Provides the HFB-14 fission rates entering the nucleosynthesis calculation for the merger model."},{"cited_title":"2024, Eur","cited_arxiv_id":null,"evidence_quote":"Shows that calibrating HFR levels to experiment or leaving inverted ground states uncorrected has minor impact on opacities for Er III."}],"review_version":1}