{"id":"050172c7-0b89-4383-9d2d-1899a57d472d","arxiv_id":"2607.22820","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"In over 150,000 equilibrium crust-atmosphere models, the thermal stability of 23 minerals is tied to atmospheric type, independent of the six tested refractory-element abundance sets.","lead":"This paper uses equilibrium chemistry models to map how the mineral makeup of a rocky exoplanet's surface changes with the composition of its atmosphere. It identifies 23 minerals whose presence in the model crust is tied to a handful of observable atmospheric types, offering a way to infer surface minerals from atmospheric spectra.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 'independent of refractory element ratios' claim is based on only two correlated compositional families; BC2 minerals are untested for Ca-rich/S-poor compositions.","rationale":"The reader's CONDITIONAL verdict is directionally correct. The single most load-bearing issue is not the equilibrium assumption, which the authors openly discuss and scope in Sect. 4.1 and which is partially supported by Earth/Venus examples and by Charnoz et al. (2026) for T>=800 K. More load-bearing is the abstract's headline independence claim, which is the paper's main novelty and the basis for the suggested observational application. The grid design deliberately keeps refractory ratios fixed within each grid and varies only C/H/O; independence is inferred across grids. But the six grids collapse into two correlated families for the refractory elements that control the key minerals (Ca/Si, S/Fe, Fe/Si). The BC2 type is only sampled by CI/Ryugu. Thus there is no evidence that the mineral-atmosphere link is independent of refractory ratios over the range of compositions expected for exoplanets. This is a within-model concern, so it can be settled by additional GGchem calculations without needing a kinetic model. The equilibrium concern remains valid but is already acknowledged and affects applicability at T<800 K; if the independence test fails, the headline claim itself needs revision. We therefore keep the CONDITIONAL verdict unchanged.","tokens_in":25111,"tokens_out":7535,"duration_ms":81367,"concrete_test":"Run GGchem grids with synthetic refractory compositions that break the existing correlations: (i) Earth-like Ca/Si and Mg/Si with CI-like S/Fe; (ii) CI-like Ca/Si with Earth-like S/Fe; (iii) a S/Fe sequence from 0.01 to 100 at fixed Ca/Si and Mg/Si; (iv) a Ca/Si sequence from 0.1 to 10 at fixed S/Fe. Repeat at 700 K and 1000 K over the same CHO grid at 1 bar. Check whether each of the 23 claimed minerals remains in the same class for the same atmospheric type. If any Class I mineral becomes Class II/III or absent, the independence claim must be restricted to solar-system-like refractory ratios.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim in the abstract is that a link between 23 minerals and atmospheric types is 'independent of the ratio of the refractory elements present.' That claim is underdetermined by the grids. Only six input compositions are used, and they are not independent samples: MORB and BSE cannot form type-B atmospheres at all (Sect. 2.2), so the type-B mineral set is tested only on Earth/CC (Ca-rich, Si-rich, S-poor) and CI/Ryugu (Ca-poor, Fe-rich, S-rich). The two families are correlated along exactly the axes that matter for the minerals (Ca/Si, S/Fe, Fe/Si). The sulphur-rich type BC2 appears only in the CI and Ryugu grids (Sect. 3.2, Fig. 7), so the claim that CaSO4, Na2SO4, Fe2O3 etc. are linked to BC2 independently of refractory ratios is untested for Ca-rich, S-poor refractory compositions. Because class assignments (Class I/II/III) are defined by which grids are included, adding a composition that breaks the Earth/CC vs CI/Ryugu correlation could reclassify minerals and change the 23-mineral table. Thus the practical inference—observe atmospheric type and p,T, infer crustal mineralogy regardless of refractory composition—is not yet supported.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents grid calculations with the equilibrium chemistry solver GGchem for 1-bar crust–atmosphere systems at 500–1000 K, using six solar-system-inspired bulk compositions (Earth, CC, BSE, MORB, CI, Ryugu) and systematically varying C, H, and O abundances to span the CHO atmospheric-parameter space. The authors classify the thermal stability of condensates into four classes and report 23 minerals whose stability they find linked to atmospheric type, including the FeS → FeS2 → CaSO4 sequence, iron oxides, and iron-bearing silicates. They further find that the average iron oxidation state tracks atmospheric type. The central claim, stated in the abstract, is that the mineral–atmosphere link is independent of the ratio of refractory elements present.","tokens_in":25335,"tokens_out":4286,"duration_ms":46453,"significance":"If the central claim held, the paper would provide a practically valuable shortcut: an observed atmospheric type, together with surface temperature and pressure, could constrain crustal mineralogy and iron oxidation state. The manuscript has clear strengths: it is built on a large, internally consistent grid (over 150,000 individual models), the thermodynamic sequence from FeS to FeS2 to CaSO4 across redox regimes is physically sensible, and the classification tables in Appendix C are a useful reference resource. However, the headline generality is not established by the evidence presented. The claim of independence from refractory element ratios rests on only two correlated compositional families, and the paper's own equilibrium assumptions are acknowledged to break down at the low end of the investigated temperature range. The thermodynamic consistency of the results is credible, but the observational, generalizable inference is substantially stronger than the current grid supports.","major_comments":[{"comment":"The claim of independence from refractory element ratios is underdetermined by the grid design. Only four compositions (Earth, CC, CI, Ryugu) form full parameter-space grids; MORB and BSE cannot produce type-B atmospheres (Sect. 2.2), and type-BC2 atmospheres occur only in the CI and Ryugu grids (Sect. 3.2). The Earth/CC family and the CI/Ryugu family are correlated along exactly the axes relevant to the reported minerals (Ca/Si, Fe/Si, S/Fe). Consequently, the Class I/II/III assignments and the list of 23 minerals are conditional on the specific composition families included. Adding a composition that breaks this correlation could reclassify minerals. The abstract's statement 'independent of the ratio of the refractory elements present' is therefore too strong. Either additional grids spanning the refractory-element space independently (e.g., varying Ca/Si, Fe/Si, S/Fe separately) must","section":"Abstract; Sect. 2.2; Sect. 3.2; Tables C.1–C.2"},{"comment":"The equilibrium assumption is load-bearing for the practical inference, and the manuscript itself provides reasons to doubt it at the lower temperatures. Sect. 4.1 states that 'if a planet experiences moderate but consistent volcanism and has a relatively cool atmosphere (T<700 K), it may never reach equilibrium' (citing Liggins et al. 2023) and that Charnoz et al. (2026) suggest good kinetic–equilibrium agreement only for T≥800 K. Yet Sect. 5 presents detailed 500 K and 600 K mineral lists as predictions, and Sect. 4.2 tells the reader that an observed atmospheric type plus surface temperature and pressure can constrain the surface mineralogy without restricting to the regime where equilibrium is plausible. The temperature-dependent validity should be stated as a central caveat in the abstract and conclusions, not only in the limitations section.","section":"Sect. 4.1 vs. Sect. 5 and Sect. 4.2"},{"comment":"The definition of sulphur-poor versus sulphur-rich atmospheres uses temperature-dependent thresholds chosen from the data (e.g., S/(C+H+O+S) ≤ 0.005 at 1000 K, ≤ 0.002 at 900/800 K, ≤ 0.001 for T ≤ 700 K). Because the atmospheric types themselves are defined from the same GGchem gas-phase output, and the mineralogical classifications are then made after applying these thresholds, the reported 'link' is to some extent a statement about the internal consistency of one equilibrium solver rather than a hypothesis test against independent data. The paper would be strengthened by a sensitivity analysis showing how the Class I/II/III assignments change with reasonable variations of these thresholds, or by anchoring one boundary to an external benchmark (e.g., laboratory data or an independent equilibrium code). Without this, the 23-mineral list should be presented as a model-dependent mapping,","section":"Sect. 3.1; Fig. 3; Fig. 5"}],"minor_comments":[{"comment":"The abstract uses 'oxidisation' while the rest of the paper uses 'oxidation'; please standardize the spelling.","section":"Abstract"},{"comment":"The caption begins 'Figure ??:' with an unresolved placeholder. This needs to be fixed before publication.","section":"Appendix B"},{"comment":"Fayalite is Fe2SiO4, not FeSiO4 as written. The formula is correct elsewhere (e.g., Fe2SiO4 entries in Fig. B.1 and Tables C.1/C.2), but the text in §3.1.2 should be corrected.","section":"Sect. 3.1.2"},{"comment":"The stated grid size is inconsistent: Sect. 2.2 says 'about 5000 crust-atmosphere models per temperature' for each of four full grids, which at six temperatures gives about 120,000 models, while Sect. 4 states 'over 150000'. Please clarify the total count, including any partial BSE/MORB grids.","section":"Sect. 2.2 and Sect. 4"},{"comment":"There are several typographical slips, e.g., 'Flouride' in §3.1.4. A careful proofread is recommended.","section":"General"}],"recommendation":"major_revision","confidential_remarks":"The paper's core thermodynamic grid work is solid and the classification tables are a useful contribution. The main issue is that the headline claim of 'independence of refractory element ratios' oversells the evidence: only two correlated compositional families are available, and the BC2 mineral set in particular is untested for Ca-rich, S-poor refractory compositions. If the authors cannot add new grids that break the correlation, they should substantially temper the abstract and conclusion. The equilibrium-timescale caveat is acknowledged but not elevated to the level the practical inference requires. I would be comfortable accepting after the claim is either backed by broader compositional sampling or properly qualified."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe headline: this is the most systematic equilibrium mapping of crust mineralogy to atmospheric types I've seen. The FeS→FeS2→CaSO4 sequence across redox regimes, and the iron oxidation state tracking atmospheric type, are thermodynamically sensible and likely hold within the model domain. But the paper's central claim—that the link is 'independent of the ratio of the refractory elements present'—is stronger than the evidence supports.\n\nWhat's new and good: The paper runs ~150,000 GGchem equilibrium models, covering the full CHO parameter space at 500–1000 K with four complete composition grids (Earth, CC, CI, Ryugu), plus partial MORB/BSE grids. It resolves the sparse sampling of Herbort & Sereinig (2025) and extends to sulphur-rich BC2 atmospheres. The result is a clean atlas: 71 condensates, 23 minerals assigned to atmospheric types, with temperature-dependent class definitions. The limitations section is honest—equilibrium timescales, kinetic condensation, pure-mineral approximation, solid solutions, p-T profile effects, and solar system bias are all acknowledged.\n\nSoft spots, in order of importance. The independence claim is underdetermined. The four full grids are really two compositional families: Earth/CC (Ca-rich, Si-rich, S-poor) and CI/Ryugu (Ca-poor, Fe-rich, S-rich). MORB and BSE cannot form type B atmospheres at all, so type B minerals are tested on only four correlated compositions. The sulphur-rich type BC2 appears only in CI/Ryugu grids, so the BC2 mineral set (CaSO4, Na2SO4, Fe2O3) is untested for Ca-rich, S-poor refractory compositions. Because class assignments depend on which grids are included, a composition that breaks that correlation could change the 23-mineral table. The claim should be scoped to 'the explored compositions' or tested with a deliberately orthogonal composition.\n\nSecond, the sulphur-poor/sulphur-rich cutoffs are hand-set, temperature-dependent values. They influence which minerals fall into which class, and no sensitivity analysis shows how robust the classification is to those thresholds. Third, no data or scripts are released, so the 23-mineral table cannot be independently checked. Fourth, the equilibrium assumption is the load-bearing premise; the authors argue it well for warm planets, but the paper doesn't quantify how far disequilibrium could move the boundaries.\n\nThe central tendency—mineralogy and iron oxidation state track atmospheric type in equilibrium—is probably correct. The overreach is the word 'independent.' This paper deserves a serious referee: it's a useful interpretive tool for JWST/Ariel/ELT observations, provided the claims are scoped and the data released. I'd recommend sending it out with a request for softening the independence language, sensitivity tests on the sulfur thresholds, and a data release.","headline":"A genuinely useful equilibrium grid atlas, but the 'independent of refractory element ratios' claim is overreached—two correlated compositional families and hand-set sulphur thresholds carry more weight than the headline admits.","tokens_in":25884,"tokens_out":3446,"would_cite":true,"duration_ms":35445,"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":"A rocky exoplanet's atmospheric type can constrain which minerals are stable on its surface, regardless of the planet's refractory element abundances.","keywords":["rocky exoplanets","crust-atmosphere equilibrium","atmospheric types","surface mineralogy","iron oxidation state","sulfur chemistry","thermo-chemical equilibrium","CHO parameter space"],"falsifier":"Observe a warm rocky exoplanet with a confidently classified type B atmosphere whose surface shows spectral evidence of ferrous minerals (e.g. FeS or Fe2SiO4) or lacks the predicted CaSO4/Fe2O3 assemblage; conversely, a type alpha planet with ferric oxides would falsify the mapping. A more direct test: measured reaction kinetics showing that the FeS-to-CaSO4 transition does not occur on geologic timescales at 700-1000 K would undercut the equilibrium basis.","tokens_in":24910,"feed_emoji":"🪨","tokens_out":4220,"duration_ms":44986,"temperature":0.7,"pith_summary":"This paper tries to establish a direct, abundance-independent link between the observable type of a rocky exoplanet's atmosphere and the minerals that can coexist with it on the ground. Using equilibrium chemistry over a grid of about 5000 crust-atmosphere models per temperature and composition, the authors find 23 minerals whose thermal stability tracks the atmospheric type across four different starting rock compositions. The link is strongest for sulfur-bearing minerals and for the average oxidation state of iron: oxidising atmospheres select sulfates and ferric iron, reducing atmospheres select sulfides and ferrous iron. If the claim holds, an observed atmospheric type plus surface temperature and pressure is enough to start mapping the unseen surface of a terrestrial exoplanet.","feed_headline":"Atmospheric type predicts a rocky planet's surface minerals","feed_subtitle":"Grids of 150,000 equilibrium models tie 23 crust minerals to gas chemistry, independent of rock composition.","key_machinery":"The central object is the atmospheric type, a coarse classification of gas-phase equilibrium composition in the C-H-O (plus S) parameter space, paired with the equilibrium condensate set returned by Gibbs free-energy minimisation. The paper sorts the 71 thermally stable condensates into four classes (omnipresent, strongly linked, partially linked, unconstrained) according to whether their stability is tied to an atmospheric type independent of element-abundance input. This machinery shows that sharp crustal transitions fall on the same borders as atmospheric-type transitions, producing the sequence FeS -> FeS2 -> CaSO4+Fe2O3 across the redox gradient from type alpha through the transitional","core_discovery":"Across roughly 150,000 equilibrium crust-atmosphere models at 1 bar and 500-1000 K, the paper claims that the stable mineral assemblage of a rocky surface is set by the atmosphere's chemical type, not by the relative abundances of refractory elements. Twenty-three minerals show this type-independent link: oxidising type B atmospheres select sulfates and ferric iron (CaSO4, Fe2O3), reducing type alpha atmospheres select sulfides and ferrous iron (FeS, Fe2SiO4), and a narrow transitional regime is marked by FeS2 and Fe3O4. The crust's average iron oxidation state tracks the atmosphere — iron(III) under type B, iron(II) under type alpha — so an observed atmospheric type plus surface temperature","pith_inferences":["Because the equilibrium assumption weakens below roughly 800 K, the practical reach of these constraints is likely limited to moderately warm planets; cooler worlds may preserve kinetic, volcanic, or photochemical disequilibrium mineralogies that blur the type-mineral link.","The same grid logic could be extended to other surface pressures or to include solid solutions and partial melts; the sharpness of the atmospheric-type transitions suggests the mineral-type mapping may persist but with shifted boundaries.","A testable extension: for a tidally locked planet whose day side and night side straddle the 500-700 K range, the two hemispheres should host different stable mineral assemblages, and thermal phase curves might reveal surface-driven signatures.","The strong CaSO4/FeS dichotomy implies that gas-phase sulfur abundance in a secondary atmosphere could be buffered by the crust, a feedback that interior-atmosphere models could incorporate to predict observable SO2 levels."],"forward_implications":["An observed atmospheric type plus a surface temperature and pressure measurement directly constrains which minerals can be stable on the crust, including the iron oxidation state.","Rocky-exoplanet spectroscopy can target warm planets (roughly 700 K and above) where the equilibrium assumption is strongest and the type-alpha/type-B distinction is sharp.","Sulfur-bearing minerals act as a redox indicator: oxidising atmospheres should yield sulfate crusts like CaSO4, reducing atmospheres sulfide crusts like FeS, with pyrite confined to a narrow transition.","Precise C/H/O abundances are not needed; only the atmospheric type, so modest-quality spectra can still constrain surface composition.","Sharp mineralogical transitions in the parameter space imply that a single planet with day-night temperature contrasts could harbour different stable mineral assemblages across its surface."],"fun_headline_variants":["Atmosphere reveals a rocky exoplanet's hidden crust","What a rocky planet's air says about its rocks","23 minerals tied to exoplanet atmospheric type","Iron oxidation state reads from a planet's atmosphere","Sky chemistry fingerprints ground minerals on rocky worlds"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The load-bearing premise is that a 1-bar crust-atmosphere interface reaches full thermo-chemical and phase equilibrium at 500-1000 K, so that the computed equilibrium minerals are the ones actually present on real exoplanet surfaces.","fun_headline_variants_meta":{"raw":{"variants":["Atmosphere reveals a rocky exoplanet's hidden crust","What a rocky planet's air says about its rocks","23 minerals tied to exoplanet atmospheric type","Iron oxidation state reads from a planet's atmosphere","Sky chemistry fingerprints ground minerals on rocky worlds"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000244,"raw_usage":{"total_tokens":1397,"prompt_tokens":803,"completion_tokens":594,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":547,"completion_tokens_details":{"reasoning_tokens":531}},"tokens_in":547,"tokens_out":594,"duration_ms":5538,"temperature":1.0,"reasoning_tokens":531,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T04:22:42.310781+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Observe a warm rocky exoplanet with a confidently classified type B atmosphere whose surface shows spectral evidence of ferrous minerals (e.g. FeS or Fe2SiO4) or lacks the predicted CaSO4/Fe2O3 assemblage; conversely, a type alpha planet with ferric oxides would falsify the mapping. A more direct test: measured reaction kinetics showing that the FeS-to-CaSO4 transition does not occur on geologic timescales at 700-1000 K would undercut the equilibrium basis.","supporting_citations":[],"review_version":1}