{"id":"136f31c4-b7b0-4039-8ada-2c00bec8c4da","arxiv_id":"2504.20450","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"Atmospheric C/O ratios of magma-ocean sub-Neptunes are set by equilibrium with the molten interior, not inherited from the protoplanetary disk, and can be depressed by carbon sequestration into metal.","lead":"This paper models how magma oceans beneath sub-Neptune atmospheres reset the atmospheric carbon-to-oxygen ratio through chemical equilibrium, making it very different from the disk value. The result matters because JWST now measures C/O in sub-Neptunes, and this could turn those measurements into probes of the planets' hidden interiors.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Deep C/O is not preserved at low Kzz: the paper's own Fig. 4 shows transmission-region C/O diverging from the magma-ocean value, so the interior-probe conclusion depends on an unverified assumption of strong vertical mixing.","rationale":"The reader's weakest assumption was global chemical equilibrium across metal, silicate, and gas phases. I partially agree, but the more load-bearing and internally testable tension is the paper's own atmospheric-transport step. Equilibration between magma ocean and deep atmosphere is the framework the paper adopts; if it fails, the 'homemade' label fails. However, the paper's new 1D modeling already shows that even with deep equilibrium, the observably accessed C/O need not equal the deep value. Since the conclusion that observed C/O ratios can be used to probe interiors is the headline application for JWST, the missing constraint on Kzz is the decisive condition. This concern does not move the verdict: the paper remains a conditional accept. It should either state the transport caveat explicitly in the conclusions or provide a Kzz threshold above which the interior probe works.","tokens_in":14082,"tokens_out":8536,"duration_ms":97018,"concrete_test":"Re-run the HELIOS/VULCAN iteration for the 6 M_Earth, AMOI=3000 K case with Kzz = 10^4, 10^5, 10^6, and 10^7 cm2/s, generate synthetic transmission spectra from the converged profiles, and retrieve C/O over 10^-3 to 10^-5 bar. If the Kzz=10^4 retrieved C/O differs from the deep equilibrium value by more than the typical retrieval uncertainty (about 0.1 to 0.2 dex), while Kzz=10^7 agrees, the 'largely preserved' conclusion requires an explicit constraint on Kzz before observed C/O can be used as an interior probe.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that observed sub-Neptune C/O probes the magma ocean has two links: deep equilibrium sets the C/O, and vertical transport preserves it in the observable atmosphere. Section 3.3 and Figure 4 only establish the second link for Kzz=10^7 cm2/s. At Kzz=10^4 cm2/s, gravitational settling and photochemistry deplete CO, CO2, CH4, and H2O above about 10^-2 bar, and the C/O ratio in the transmission region (10^-3 to 10^-5 bar) is shifted away from the deep equilibrium value. This matters because the TOI-270d C/O measurements cited in Section 1 are transmission measurements. Section 5 nevertheless states that 'the deep C/O ratio is largely preserved in the observable atmosphere' and that low observed C/O implies iron-metal carbon sequestration. That inference is not supported by the low-mixing case, and no constraint on Kzz is provided. Until the vertical-transport step is demonstrated at realistic Kzz values, a low observed C/O could reflect photochemical and dynamical fractionation rather than the interior.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper investigates whether the atmospheric C/O ratios of sub-Neptunes are set by formation (inherited from the protoplanetary disk) or by chemical equilibrium between the atmosphere and an underlying magma ocean. The authors use the global chemical equilibrium framework of Schlichting & Young (2022), extended with an explicit metal–silicate reaction for carbon partitioning (reaction 1), to compute bulk atmospheric C/O ratios across a grid of planetary masses, atmospheric mass fractions, and two thermal states. They find that including carbon in the metal phase strongly lowers the atmospheric C/O ratio at low atmospheric mass fractions, while excluding it yields nearly constant, often super-solar C/O. They then couple one representative 6 M_Earth case to 1D atmospheric models (HELIOS, FastChem, VULCAN, HELIOS-K) with photochemistry and two vertical mixing strengths, finding that the C/O ratio is altitude-independent for K_zz = 10^7 cm2/s but varies significantly for K_zz = 10^4 cm2/s. The paper concludes that atmospheric C/O is 'homemade rather than inherited' and that observed low C/O ratios imply iron metal in the magma ocean has sequestered carbon.","tokens_in":14328,"tokens_out":7313,"duration_ms":74836,"significance":"If the central claim holds, the paper would change how observed sub-Neptune C/O ratios are interpreted: they would report on interior conditions and magma-ocean chemistry rather than formation location. This is a timely and relevant contribution given JWST transmission spectra of sub-Neptunes such as TOI-270d. The paper's strengths are its internally consistent equilibrium network, the inclusion of a physically motivated carbon metal-silicate exchange reaction, and sensitivity tests over carbon partitioning fits and silicate-metal equilibrium temperature. The coupling to 1D atmospheric models with photochemistry is also valuable. The main significance, however, is conditional on the vertical transport step, because the observable atmosphere may not preserve the deep equilibrium C/O ratio under realistic mixing conditions.","major_comments":[{"comment":"The conclusion in §5 that 'the deep C/O ratio is largely preserved in the observable atmosphere' is not supported by the low-Kzz case shown in Figure 4. In the low-mixing scenario, gravitational settling and photochemistry deplete CO, CO2, CH4, and H2O above about 10^-2 bar, and the C/O ratio in the transmission region (10^-3 to 10^-5 bar) deviates from the deep equilibrium value. Since the TOI-270d C/O measurement cited in §1 is from transmission spectroscopy and no constraint on Kzz is provided, the inference that a low observed C/O implies iron metal carbon sequestration is not supported. The authors should either demonstrate preservation at realistic Kzz values, provide constraints on Kzz from the models, or present the interior-probe conclusion as explicitly conditional on strong vertical mixing.","section":"§3.3 and Figure 4"},{"comment":"The conclusion that atmospheric C/O is 'not inherited from the protoplanetary disk' is not supported by the model as presented, because the initial envelope C/O is fixed to the solar value and not varied. In global chemical equilibrium, the final atmospheric C/O depends on the total planetary C/O inventory (core plus envelope), so a different disk C/O would shift the results. Moreover, Figure 1 shows that at high atmospheric mass fractions the atmospheric C/O approaches the initial disk value, which is the inherited regime. The authors should qualify the conclusion to state that magma ocean interactions can strongly modify a solar-composition envelope at low atmospheric mass fractions, rather than claiming inheritance is erased, and ideally test sensitivity to the initial envelope C/O.","section":"§2.5 and Figure 1"},{"comment":"The central claim assumes global chemical equilibrium across the metal, silicate, and gas phases on timescales shorter than atmospheric evolution, but no timescale or kinetic argument is presented. If the magma ocean solidifies or transport timescales prevent re-equilibration, the atmospheric C/O could retain a disk-inherited fingerprint. The paper's conditional 'if there is a magma ocean' is acknowledged in places, but §5 states the 'not inherited' conclusion without that caveat. The authors should add a discussion of the relevant equilibration timescales and clearly state the conditions under which the homemade scenario applies.","section":"Equations (2)–(4) and §5"}],"minor_comments":[{"comment":"The statement 'As CH4 overtakes CO as the main carbon carrier, the C/O ratio rises slightly at high altitudes' appears to contradict the earlier statement that the C/O ratio decreases above ~10^-2 bar in the low-mixing case; please clarify the vertical profile.","section":"§3.3"},{"comment":"There are minor grammar issues: 'a AMOI' should be 'an AMOI', and 'a atmosphere' should be 'an atmosphere'.","section":"§2.4 and Figure 1 caption"},{"comment":"Please specify that the pressure term is included only for gas-phase species, so that the summation notation is unambiguous.","section":"§2.1, Eq. (2)"},{"comment":"The absence of pressure corrections to the Gibbs free energies, especially for reaction (1), is a potential source of uncertainty in the carbon partitioning results; the sensitivity tests in Appendix A do not cover this assumption.","section":"§2.2"},{"comment":"The phrase 'SiH4 reduces the atmosphere' is unclear; please specify whether it reduces the atmospheric mass, metallicity, or something else.","section":"§4"},{"comment":"The atmospheric models do not include helium; this may affect the mean molecular weight and hence gravitational settling in the 1D calculations, and a brief justification or caveat would be helpful.","section":"§2.4"}],"recommendation":"major_revision","confidential_remarks":"The paper builds heavily on the authors' earlier framework (Schlichting & Young 2022; Young et al. 2023) and the incremental novelty is the carbon metal-silicate partitioning and the 1D coupling. The major issues are fixable within the manuscript's scope: the transport caveat can be addressed by softening the observable-probe conclusion, and the inheritance overclaim can be tempered by acknowledging the dependence of the equilibrium C/O on the total elemental inventory. I would suggest the authors consider a more nuanced title and abstract framing, since the phrase 'not inherited' is not supported for thick-envelope planets."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The new piece here is real: adding carbon partitioning into an iron metal phase inside the Schlichting-Young global equilibrium framework, then coupling that deep composition to 1D photochemical-transport models. That carbon sink can pull atmospheric C/O down by orders of magnitude at low atmospheric mass fractions, and it is a genuine extension over Tian & Heng (2024) and Seo et al. (2024), which both omit metal-silicate partitioning. The equilibrium network is internally consistent, the mass-balance equations are standard, and they do honest sensitivity tests on carbon partitioning fits and SME temperature. The comparison to prior work is fair, and they are open about not including SiH4 and about the CH4 dissolution caveat. The circularity burden is low: the C/O outputs are computed, not fitted to the values being explained.\n\nThe soft spot is the vertical-transport link, and it is load-bearing. Their own Figure 4 shows that at Kzz = 10^4 cm2/s, the C/O ratio in the transmission region (10^-3 to 10^-5 bar) diverges from the deep equilibrium value, while it remains constant only for Kzz = 10^7 cm2/s. The TOI-270d C/O measurements that motivate the paper are transmission measurements. So the conclusion that “the deep C/O ratio is largely preserved in the observable atmosphere,” and that a low observed C/O implies iron-metal carbon sequestration, is only supported if you assume strong vertical mixing. They provide no constraint on Kzz, and that is not a minor omission—it is the bridge between their deep equilibrium result and the JWST data they want to interpret.\n\nTwo smaller concerns: the global chemical equilibrium assumption itself means the atmosphere is fully re-equilibrated with the magma ocean on timescales shorter than atmospheric evolution, which is plausible but not demonstrated. And there is no public code or data release—they point to Grimm et al. in prep for the computational speedup, which makes the quantitative results hard to reproduce independently.\n\nWho this is for: sub-Neptune atmosphere modelers and JWST observers who want a physically motivated framework for interpreting C/O measurements. It deserves a serious referee, but the referee should push hard on the Kzz dependence and on whether the equilibrium assumption holds over the relevant timescales. I would not desk reject this, and I would want to see the vertical-transport conclusion reworked or heavily caveated before publication.","headline":"A genuinely new metal-carbon sink in a magma-ocean equilibrium model that can make sub-Neptune C/O ratios very low, but the vertical-transport step needed to connect deep C/O to transmission spectra is only shown to work under strong mixing.","tokens_in":14847,"tokens_out":1598,"would_cite":true,"duration_ms":18439,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Magma-ocean chemistry, not the protoplanetary disk, sets sub-Neptune C/O ratios.","keywords":["sub-Neptunes","C/O ratio","magma ocean","chemical equilibrium","carbon sequestration","exoplanet atmospheres","metal-silicate partitioning","JWST"],"falsifier":"Measure atmospheric C/O in several sub-Neptunes spanning at least a factor of a few in inferred atmospheric mass fraction from bulk density and compare with the predicted monotonic rise of C/O with envelope mass fraction; seeing no such correlation, or finding C/O values that track the planet's likely formation radius instead of its interior state, would falsify the equilibrium-driven picture.","tokens_in":13918,"feed_emoji":"🌋","tokens_out":6007,"duration_ms":51119,"temperature":0.7,"pith_summary":"This paper argues that the atmospheric carbon-to-oxygen ratio (C/O) of a sub-Neptune is not a fossil of where the planet formed in the protoplanetary disk. Instead, it is set by chemical equilibrium between the hydrogen-dominated atmosphere and the molten rock and metal deep inside the planet. Using a global chemical equilibrium calculation across gas, silicate, and metal phases, the authors show that the resulting C/O can range from orders of magnitude below solar to a few times solar depending on planet mass, atmospheric mass fraction, and temperature. A key new ingredient is carbon partitioning into iron metal, which can sequester much of the carbon and drive atmospheric C/O far below the disk value, especially for small envelopes. If right, this means JWST C/O measurements can probe the interior chemistry of sub-Neptunes rather than their formation location.","feed_headline":"Magma oceans set sub-Neptune C/O ratios","feed_subtitle":"Low measured C/O would reveal an iron core that locked away carbon, opening a window into exoplanet interiors.","key_machinery":"The central object is the global chemical equilibrium framework: a set of linearly independent chemical reactions among 26 species distributed across three phases, iron metal, silicate melt, and gas, solved with Gibbs free-energy equilibrium conditions, phase mass-balance, and elemental conservation. The paper's extension adds a carbon partitioning reaction between metal and silicate, $\\mathrm{C_{metal}} + \\mathrm{O_{metal}} \\rightleftharpoons \\mathrm{CO_{silicate}}$, letting carbon be drawn out of the atmosphere into the metal phase. The same equilibrium framework is then coupled to one-dimensional atmosphere models with photochemistry and vertical mixing to check how the deep composition is transmitted to observable pressures.","core_discovery":"On its own terms, the paper's central discovery is that the atmospheric C/O ratio in sub-Neptunes emerges from the chemical exchange between the atmosphere and the underlying magma ocean rather than being inherited from the protoplanetary disk. The authors compute global chemical equilibrium among gas, silicate melt, and iron metal, including a new reaction that lets carbon partition from silicate into metal. Across a grid of 2 to 10 Earth-mass planets with atmospheric mass fractions of 0.1 to 10 percent, they find atmospheric C/O spans several orders of magnitude below solar to a few times solar, with carbon sequestration into the metal phase producing large depletions at low atmospheric mass fractions. They further show, with coupled one-dimensional radiative-convective and photochemical models, that the deep-equilibrium C/O is largely preserved in the observable atmosphere under strong vertical mixing, so a measured C/O far below the host star's value would indicate that iron metal in the magma ocean has locked away a large fraction of the carbon.","pith_inferences":["If the equilibrium assumption holds, the scatter in observed sub-Neptune C/O ratios could be mapped onto a scatter in interior metal content or thermal state, turning atmospheric retrievals into a crude tomographic probe of the planet's deep interior.","The metal-phase carbon sink suggests that a sub-Neptune with a larger iron core fraction would show a lower atmospheric C/O for the same envelope mass, a correlation that could be tested with planets of known mass-radius composition.","The strong sensitivity to silicate-metal equilibrium temperature implies that C/O measurements might also constrain the depth at which metal-silicate equilibration occurred, placing a constraint on the thermal state of the magma ocean during the era when the atmosphere formed.","A direct test for a sample of sub-Neptunes spanning a range of masses and atmospheric mass fractions inferred from bulk density is that the model predicts a monotonic increase of C/O with atmospheric mass fraction; if observations instead show no such trend, the assumption of global equilibrium would need to be revisited."],"forward_implications":["Observed sub-Neptune C/O ratios, especially values far below the host star's, should be read as diagnostics of interior conditions rather than formation location.","Planets with small atmospheric mass fractions and low masses are the most likely to show strongly sub-stellar C/O, because most of the carbon is sequestered into the metal phase.","Under strong vertical mixing the deep C/O is preserved at observable pressures, so transmission and emission spectroscopy can both probe the interior, whereas under weak mixing the C/O varies with altitude and transmission measurements would sample a different value than the deep one.","If carbon is excluded from the metal phase, atmospheric C/O stays close to the initial disk value, so the metal phase is the main agent that decouples atmospheric composition from the disk.","Atmospheric retrievals that impose disk-based C/O priors may be biased for sub-Neptunes with magma oceans."],"supporting_citations":[{"why":"Supplies the global chemical equilibrium framework and the original reaction network that this paper extends with carbon partitioning.","marker":"Schlichting & Young (2022)"},{"why":"Provides the adopted experimental metal-silicate carbon partitioning data used to set the free energy of the new carbon partitioning reaction.","marker":"Blanchard et al. (2022)"},{"why":"Provides the initial planetary compositions and the internally consistent thermodynamic treatment of Si and O in the metal phase.","marker":"Young et al. (2023)"},{"why":"Reports the first JWST measurement of C/O in a sub-Neptune, which motivates the question addressed here.","marker":"Benneke et al. (2024)"},{"why":"A comparative magma-atmosphere equilibrium study without metal-silicate partitioning, used to contrast the present results.","marker":"Tian & Heng (2024)"},{"why":"A comparative study of magma-gas equilibrium whose O/H versus C/O trend the paper reproduces and builds upon with a metal phase.","marker":"Seo et al. (2024)"},{"why":"Supports the interpretation of altitude-dependent C/O in terms of gravitational settling and photochemical depletion of carbon and oxygen species.","marker":"Tsai et al. (2021)"}],"fun_headline_variants":["Sub-Neptune C/O: homemade in magma oceans, not inherited","Low sub-Neptune C/O means an iron core ate the carbon","Magma oceans, not protoplanetary disks, set sub-Neptune C/O","C/O ratio in sub-Neptune: made by magma, not by disk"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole result rests on the assumption that the atmosphere, magma ocean, and metal phase stay in full chemical equilibrium long enough to re-set the envelope's composition; if gas transport or a solidifying interior shut that exchange down, the atmosphere could keep a disk-inherited C/O fingerprint.","fun_headline_variants_meta":{"raw":{"variants":["Sub-Neptune C/O: homemade in magma oceans, not inherited","Low sub-Neptune C/O means an iron core ate the carbon","Magma oceans, not protoplanetary disks, set sub-Neptune C/O","C/O ratio in sub-Neptune: made by magma, not by disk"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000862,"raw_usage":{"total_tokens":3815,"prompt_tokens":1095,"completion_tokens":2720,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":711,"completion_tokens_details":{"reasoning_tokens":2648}},"tokens_in":711,"tokens_out":2720,"duration_ms":20528,"temperature":1.0,"reasoning_tokens":2648,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T05:28:20.574874+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure atmospheric C/O in several sub-Neptunes spanning at least a factor of a few in inferred atmospheric mass fraction from bulk density and compare with the predicted monotonic rise of C/O with envelope mass fraction; seeing no such correlation, or finding C/O values that track the planet's likely formation radius instead of its interior state, would falsify the equilibrium-driven picture.","supporting_citations":[],"review_version":1}