{"id":"fd59c106-bd5e-4c63-b6a9-af3a824f9f6c","arxiv_id":"1908.08047","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Nine metal-polluted white dwarfs show rocky, volatile-poor accreted compositions with diversity that includes a mantle-like planetesimal with a small Fe-Ni core and a post-accretion system.","lead":"This paper analyzes nine white dwarfs polluted by rocky debris and finds they are accreting volatile-poor, differentiated asteroidal material. It introduces a logit-based comparison method and identifies unusual compositions, including a Ca-poor star, two Na-rich stars, and one star whose accretion stopped about 5 million years ago.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Post-accretion identification of 2216-657 rests on Mg/Ca and Mg/Fe sinking-time ratios that are not tabulated and carry inter-model scatter; independent diffusion tables would settle whether the 5 Myr interpretation is robust.","rationale":"The reader's weakest-assumption identification is correct: the three-phase diffusion model, and especially the relative sinking times of Mg versus Ca and Fe, is the least secure input to the most novel part of the paper. I agree that this supports a conditional rather than unconditional verdict. The broad rocky-accretion conclusion is robust, and the paper deserves credit for disclosing the inter-model disagreement in Appendix A3 and for performing the random-abundance sanity check at 2216-657. The residual concern is specific: the Mg/Ca and Mg/Fe sinking-time ratios underpin the 5 Myr post-accretion age, yet they are not tabulated, and the factor-1.2 element-to-element scatter in independent models is large enough to warrant a direct recalculation. Such a check would likely confirm the qualitative conclusion, but it should be done before the post-accretion interpretation is treated as settled. Therefore the reader's CONDITIONAL verdict should stand unchanged.","tokens_in":26330,"tokens_out":5515,"duration_ms":64418,"concrete_test":"Recompute the Section 4.2 decreasing-phase chi-squared grid for 2216-657 using the public Montreal diffusion-timescale tables for He-dominated atmospheres at Teff about 9190 K and log g about 8.05, and also perturb the Mg/Ca and Mg/Fe sinking-time ratios by the factor-1.2 inter-model spread quoted in Appendix A3. If the best-fit post-accretion time shifts outside the 2-10 Myr window, or if ongoing accretion becomes consistent with any rocky or solar-system composition, the post-accretion claim is model-dependent; if the minimum remains near 5 Myr with Mg still strongly enriched relative to Ca and Fe, the concern is resolved.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The broad claim that the sample accretes rocky, volatile-poor asteroidal material is well supported by multiple refractory element ratios and is consistent with the field consensus. The load-bearing soft spot is the novel post-accretion interpretation of 2216-657. That interpretation depends on converting photospheric abundances to accreted compositions via the three-phase diffusion model of Section 3.1, specifically on the statement that Mg sinks about 2.5 times more slowly than Ca and Fe at Teff about 9200 K and log g about 8.05 (Section 4.3). If the true Mg/Ca and Mg/Fe sinking-time ratios were near unity, the Mg-dominated photosphere could be a direct compositional signature of the accreted body rather than evidence that accretion ceased about 5 Myr ago. The paper itself reports in Appendix A3 that absolute sinking times from the Montreal models differ from the Koester models by factors up to 85 for H-dominated and six for He-dominated stars, while element-to-element ratios usually agree within a factor of 1.2. A factor of 1.2 would shift the recovered post-accretion time by only about 20 percent, so this is not by itself fatal, but the Mg/Ca and Mg/Fe ratios used for 2216-657 are not tabulated and are the least directly checkable input to the headline claim. The random-abundance sanity check in Section 4.3 mitigates the possibility of a non-rocky mimic, but it is computed with the same diffusion model and does not remove the need for an independent ratio check.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"Drawing on medium-resolution X-shooter spectroscopy of nine metal-polluted white dwarfs (with supplementary Keck/HIRES and HST/COS data for some stars), this paper derives photospheric abundances from model atmospheres and converts them to accreted compositions under the standard constant-accretion, three-phase diffusion model. A logit-transformed reduced chi-squared test is introduced to compare the inferred compositions with solar system materials. The authors conclude that all nine stars are accreting rocky, volatile-poor asteroidal material; that 0446-255 best matches a mantle-like planetesimal with a small Fe-Ni core; that 0449-259, 1350-162 and 2105-820 show notable Na excesses or Ca depletion; and that 2216-657 is observed in the post-accretion (decreasing) phase, with accretion having ceased about 5 Myr ago. Extensive attention is given to error propagation, upper limits, and systematic uncertainties.","tokens_in":26559,"tokens_out":8746,"duration_ms":85642,"significance":"The paper's broad conclusion is in line with the field consensus and is supported by multiple refractory-element ratios. Its main new contributions are the logit-based comparison method, the well-characterized 0446-255 composition with 14 detected metals, and the identification of 2216-657 as a rare post-accretion system. The authors are unusually explicit about limitations: Appendix A3 documents model-dependent diffusion timescales and shows that independent re-analyses of the same stars can change the chi-squared consistency conclusions. The logit method is validated on simulated data and on G29-38, and the comparisons use external benchmark compositions rather than being fit to the model, which reduces circularity concerns. If the post-accretion timing holds, 2216-657 will be a useful anchor for accretion-history modelling.","major_comments":[{"comment":"The post-accretion identification and the quoted 5 Myr timing rest on the relative sinking times of Mg versus Ca and Fe, but the paper never tabulates tau_Mg/tau_Ca or tau_Mg/tau_Fe at the star's parameters (Teff about 9190 K, log g about 8.05). Table 1 lists only tau_Ca, and the statement that Mg sinks around 2.5 times more slowly than the other detected elements appears in prose without an associated uncertainty. Please provide these ratios, with their Monte Carlo or formal errors, and show how the recovered post-accretion time changes when independent diffusion timescales (e.g., the Montreal tables discussed in Appendix A3) are used. A factor of about 1.2 in the ratio would alter the inferred time by roughly 20 per cent, which would be acceptable, but the current text does not allow the reader to verify that this is the relevant uncertainty.","section":"§4.3, 2216-657; Table 1"},{"comment":"The random-abundance sanity check is computed with the same diffusion model that produces the presumed Mg/Fe enhancement, so it cannot independently test whether an unusual accreted composition could mimic the observed ratios. Given the inter-model discrepancies quoted in Appendix A3 (up to a factor of six in absolute sinking times for He-atmosphere stars, with element-to-element ratios usually within about 1.2), the authors should re-run the 2216-657 analysis with an independent diffusion-time table, or at minimum perturb the Mg/Ca and Mg/Fe ratios to the extremes of the model scatter. This would convert the post-accretion identification from a model-contingent inference into a quantitatively robust one.","section":"§4.3, 2216-657 sanity check; §Appendix A3"}],"minor_comments":[{"comment":"Please clarify exactly which elements enter the mass-fraction normalization for each star-comparison pair. The text says fractions are measured with respect to only the observed metals, but for comparisons such as 67P/C-G (no Ti or Ni) and pallasites (no Na or Ti) the effective denominator changes; a sentence stating that only the common detected elements are used, and whether the missing-element treatment was tested, would remove ambiguity.","section":"§3.3, Eq. (1)"},{"comment":"There is a typo, 'thisisduetothetheunusualNaabundance', with a duplicated 'the'; please correct it.","section":"§4.3, 0449-259"},{"comment":"The sentence containing 'abundances have have diverged' has a duplicated 'have'; please correct it.","section":"§5.2"},{"comment":"The caption ends with '(Section 4.2.' and is missing the closing parenthesis; please fix the incomplete reference.","section":"Figure 5 caption"},{"comment":"The phrase 'uncertain (1)' appears to be an orphan footnote marker; if no footnote is intended, please remove the '(1)' or convert it to a proper citation.","section":"§4.3, 0122-227"}],"recommendation":"major_revision","confidential_remarks":"The paper is a careful, well-scoped observational study appropriate for MNRAS. The requested additions are modest: tabulated diffusion-time ratios and an independent-model sensitivity check for the 2216-657 decreasing-phase claim. I do not see a need for a deeper reanalysis or additional data, and I found no evidence of circular reasoning or unsupported novelty claims."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nRead this one if you work on white dwarf pollution or exoplanet bulk compositions. The paper reports photospheric abundances for nine metal-polluted white dwarfs from X-shooter and Keck/HIRES, with a couple supplemented by HST/COS. The new data are useful: nine more systems with multi-element detections, including the most heavily polluted star with 14 metals. The analysis is unusually careful about uncertainties—Monte Carlo propagation, correlated errors, comparisons with independent models. The logit-based chi-squared comparison is a genuinely sensible methodological improvement over raw mass-fraction chi-squared, and they validate it with simulations and a re-analysis of G29-38. I was also impressed that they test their decreasing-phase interpretation for 2216-657 with random abundance sets; that is the kind of sanity check that should be standard.\n\nThe central conclusion—all nine stars are accreting rocky, volatile-poor asteroidal material, mostly consistent with differentiated bodies—is well supported and matches the consensus. The interesting outliers are the Ca-depleted 2105-820, the Na-rich 0449-259 and 1350-162, and especially 2216-657, which they argue is observed a few sinking times after accretion ceased, around 5 Myr ago.\n\nThe soft spot is exactly the one you flagged. The 2216-657 timing claim rests on the ratio of Mg sinking time to Ca and Fe sinking times, which they say is about 2.5 at that star's parameters. That ratio is the difference between 'Mg-rich accreted body' and 'post-accretion enhancement,' and I couldn't find the ratio tabulated anywhere. The paper itself reports that independent diffusion models (Montreal vs Koester) disagree on element-to-element ratios by typically a factor of 1.2, which would shift the recovered time by roughly 20 percent—not fatal, but an independent check would make me much more comfortable. Also, the chi-squared machinery is explicitly indicative, and for stars with only three detected elements (e.g., 2105-820) the pallasite match is suggestive rather than conclusive.\n\nMinor gripes: the data and model grids aren't public, which limits reproducibility, and some of the speculative discussion about Na origins is necessarily hand-wavy. Neither undermines the paper.\n\nBottom line: this deserves a serious referee and, with the sinking-time ratios tabulated and the model dependence acknowledged more explicitly, it is publishable as is.","headline":"A careful nine-star abundance study that confirms the rocky-accretion consensus and offers one credible post-accretion system; the timing claim is model-dependent but not fragile enough to sink the paper.","tokens_in":27170,"tokens_out":3366,"would_cite":true,"duration_ms":31271,"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":"Nine polluted white dwarfs are accreting rocky, volatile-poor asteroidal debris from differentiated bodies, including a mantle-like planetesimal with a small iron core and one system where accretion ceased about five million years ago.","keywords":["white dwarfs","metal pollution","exoplanetary material","asteroid accretion","photospheric abundances","planetesimal differentiation","diffusion timescales","circumstellar debris"],"falsifier":"Compute the diffusion (sinking) timescales for magnesium, calcium, and iron with an independent public model for a helium-atmosphere white dwarf at around 9190 K and $\\log g \\sim 8.05$, the parameters of 2216−657; if the Mg/Ca and Mg/Fe sinking-time ratios differ substantially from the values used here, the post-accretion interpretation and the inferred age of about 5 Myr are not settled. A long-baseline search for changes in the Mg/Fe ratio of 2216−657 could test the predicted exponential divergence, though the relevant timescales are millions of years.","tokens_in":26069,"feed_emoji":"🪐","tokens_out":11063,"duration_ms":100553,"temperature":0.7,"pith_summary":"The paper argues that nine metal-polluted white dwarfs are all currently or recently accreting rocky, volatile-poor asteroidal material, and that the compositions trace the interiors of differentiated planetesimals. It presents a new way to compare photospheric abundances with solar-system materials—a statistical test on log-odds scales that is less dominated by abundant elements—and applies it to medium-resolution optical spectra. The payoff is diversity: one star appears to be accreting a mantle-like planetesimal with a small iron–nickel core, two show sodium-rich material, one is strongly calcium-poor, and one appears to have stopped accreting about five million years ago, leaving magnesium to dominate because it sinks slowest. If the interpretation is right, white dwarf spectroscopy is a direct way to measure the bulk composition of exoplanetary bodies.","feed_headline":"Nine white dwarfs are accreting rocky, differentiated asteroid debris","feed_subtitle":"One is a mantle-like planetesimal with a small iron core; another stopped accreting about 5 Myr ago.","key_machinery":"The argument rests on the three-phase model of photospheric metal abundance evolution driven by element-dependent diffusion (sinking) timescales. During constant accretion, abundances rise and approach steady-state values; once accretion stops, element ratios diverge exponentially because elements leave the photosphere at different rates. Coupled to this is a logit-transformed reduced chi-squared comparison between the observed abundances and a database of solar-system materials (Earth components, meteorite groups, comets), which prevents a dominant element from manufacturing false consistency. The paper uses the differential sinking of magnesium versus calcium and iron to read the timing of the 2216−657 system: since Mg lingers longest, its dominance marks a post-accretion phase about five calcium sinking timescales after the event.","core_discovery":"The central claim is that the photospheric metals in all nine white dwarfs were delivered by accretion of rocky, volatile-poor asteroidal debris whose parent bodies were differentiated, in line with the consensus model for white dwarf pollution. The most metal-rich object, WD 0446−255, shows fourteen detected elements whose relative abundances match a mixture of mantle rock with a small fraction of iron–nickel core material, implying the accretion of a differentiated minor planet with a small core. At WD 2216−657 the element ratios are best explained by accretion that ceased about 5 Myr ago: magnesium, which sinks about 2.5 times more slowly than calcium and iron, has come to dominate the photosphere. The paper also reports unusual sodium enhancements at two stars and a severe calcium deficiency at another, and argues these reflect genuine diversity in the source planetesimals rather than observational artifacts.","pith_inferences":["Editorial extension: if the Mg/Fe diagnostic is as reliable as this analysis suggests, large spectroscopic surveys of polluted white dwarfs could identify many more post-accretion systems from spectra with only a few detected metals, turning rare objects into a population sample.","Editorial extension: the sodium-rich stars may be fossil signatures of earlier accretion events of crust-like material, because sodium sinks slowly; the paper raises this possibility but leaves quantitative modelling for future work, and detecting phosphorus, sulfur, or other moderately volatile elements would test it.","Editorial extension: if the strongly calcium-poor star really accreted pallasite-like core–mantle boundary material, white dwarf spectra would be probing the internal layering of exoplanetesimals, not just their bulk composition—a scenario that predicts distinctive co-variation of siderophile and lithophile elements in other such stars."],"forward_implications":["If the claims hold, white dwarf pollution spectroscopy is a reliable route to the bulk compositions of exoplanetary bodies, and most such bodies in these systems are rocky and volatile-poor.","The results for 0446−255 imply that differentiated rocky planetesimals with small iron–nickel cores exist around white dwarfs, and that the accreted mass of about $10^{23}\\,\\mathrm{g}$ is comparable to small asteroids or icy moons like Enceladus.","A decreasing-phase system such as 2216−657 rules out exponentially decaying accretion histories, because those would keep abundance ratios near steady-state values rather than letting Mg diverge, and it places a lower limit near $10^{23}\\,\\mathrm{g}$ on the parent body mass in the paper's accretion-model context.","The Mg/Fe ratio, when Ca/Fe looks normal, can serve as a preliminary diagnostic for post-accretion systems, allowing more such systems to be identified from fewer detected elements.","The logit-transformed comparison should be used for abundance consistency tests; without it, false consistencies with meteorite compositions extend too far into the post-accretion phase, as the re-analysis of a known polluted white dwarf illustrates."],"supporting_citations":[{"why":"Supplies the diffusion and accretion model that converts photospheric abundances into accreted compositions and defines the three-phase framework.","marker":"Koester 2009"},{"why":"Establishes the metal sinking timescales that make photospheric metals evidence of recent or ongoing accretion.","marker":"Paquette et al. 1986"},{"why":"Proposes the tidal disruption of asteroids as the source of white dwarf pollution, the framework within which the sample is interpreted.","marker":"Jura 2003"},{"why":"Provides the earlier chi-squared comparison of accreted material with solar system objects that this paper improves with the logit transform.","marker":"Xu et al. 2013"},{"why":"Supplies the meteorite abundance database used as the main comparison set for the accreted compositions.","marker":"Nittler et al. 2004"},{"why":"Gives the large sample of polluted white dwarfs against which the calcium-poor star 2105−820 is identified as an outlier.","marker":"Hollands et al. 2018"},{"why":"Reports the earlier ultraviolet detections of carbon and silicon at 2216−657 that support the post-accretion interpretation.","marker":"Wolff et al. 2002"},{"why":"Models planetesimal accretion events and is used to interpret 2216−657 as a single-event system with a lower limit on parent body mass.","marker":"Wyatt et al. 2014"}],"fun_headline_variants":["White dwarf with 14 metals reveals a mantle-like rock and small iron core","One white dwarf stopped accreting 5 Myr ago, leaving Mg to dominate","Diverse planetesimals: a Ca-poor star and two Na-rich polluters","Nine polluted white dwarfs trace rocky, differentiated asteroid debris","WD 0446-255's spectrum hints at a mantle plus a Fe-Ni core"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that magnesium really does sink out of these white-dwarf atmospheres about 2.5 times more slowly than calcium and iron; if independent diffusion models disagree with that relative ratio, the post-accretion identification and 5 Myr timing for 2216−657 would collapse.","fun_headline_variants_meta":{"raw":{"variants":["White dwarf with 14 metals reveals a mantle-like rock and small iron core","One white dwarf stopped accreting 5 Myr ago, leaving Mg to dominate","Diverse planetesimals: a Ca-poor star and two Na-rich polluters","Nine polluted white dwarfs trace rocky, differentiated asteroid debris","WD 0446-255's spectrum hints at a mantle plus a Fe-Ni core"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000706,"raw_usage":{"total_tokens":3174,"prompt_tokens":932,"completion_tokens":2242,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":548,"completion_tokens_details":{"reasoning_tokens":2138}},"tokens_in":548,"tokens_out":2242,"duration_ms":19263,"temperature":1.0,"reasoning_tokens":2138,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:51:23.930221+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the diffusion (sinking) timescales for magnesium, calcium, and iron with an independent public model for a helium-atmosphere white dwarf at around 9190 K and $\\log g \\sim 8.05$, the parameters of 2216−657; if the Mg/Ca and Mg/Fe sinking-time ratios differ substantially from the values used here, the post-accretion interpretation and the inferred age of about 5 Myr are not settled. A long-baseline search for changes in the Mg/Fe ratio of 2216−657 could test the predicted exponential divergence, though the relevant timescales are millions of years.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Reports the earlier ultraviolet detections of carbon and silicon at 2216−657 that support the post-accretion interpretation."}],"review_version":1}