{"id":"b29d6e5f-b771-4c41-a6b3-2d5eb4ddd0b0","arxiv_id":"2412.05099","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"In self-gravitating discs with dust evolution, gravitational-instability and streaming-instability planet-forming regions have different C/O ratio ranges, and total C/O anticorrelates with dust-to-gas ratio.","lead":"This simulation study tracks how carbon and oxygen move through gas and ice in young, self-gravitating protoplanetary discs as dust drifts and grows. It finds that regions where different planet-forming instabilities occur have distinct C/O ratios, which may explain the different elemental compositions observed in two families of exoplanets.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The claimed GI/SI C/O dichotomy relies on different C/O definitions for the two mechanisms (total for GI, gas/ice separately for SI), so the agreement with observed exoplanet populations may reflect the phase choice rather than a robust formation-mechanism signature.","rationale":"The paper is a serious modelling study: FEOSAD is an established code, the two models bracket core masses, the snowline and dust-substructure results are internally consistent, and the authors explicitly list the chemistry and vertical-structure caveats. The central claim, however, rests on Figure 10, and that figure is where I find the weakest link. The reader's chemistry concern is real but secondary for the GI branch: total volatile C/O is conserved under gas-grain reactions, so simplified chemistry mainly affects the gas/ice split used for the SI branch, not the total used for GI. The more direct threat is that the GI and SI branches are not the same physical quantity. Because observed exoplanet C/O is a single gas-phase atmospheric value, one needs a common conversion from disc C/O to planet C/O. Without that conversion, the narrow-versus-wide match could be a selection effect introduced by choosing 'total' for one mechanism and 'gas' for the other. The proposed planet-formation pipeline test would settle this directly. In the meantime, the reader's CONDITIONAL verdict remains appropriate; I would not change it, but the condition should include this phase-definition and prediction test, not only code release and chemistry.","tokens_in":48379,"tokens_out":15041,"duration_ms":159901,"concrete_test":"Feed the FEOSAD time-sequence into a planet-formation and atmosphere model, e.g., the population-synthesis framework of Mordasini et al. (2016) or the retrieval-oriented pipeline of Mollière et al. (2022). For each GI clump and each SI region, compute the planet's atmospheric C/O using one common prescription: accrete gas at the local gas C/O and solids at the local ice C/O, with the same solid-accretion efficiency for both mechanisms. Then build predicted C/O histograms for GI and SI planets and compare them to the Hoch et al. (2023) samples with a formal test (e.g., Kolmogorov-Smirnov or Anderson-Darling). If, under a common definition, the SI distribution narrows and overlaps the GI distribution, the proposed C/O formation diagnostic does not survive.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section 4 and Figure 10 define the planet-forming C/O differently for the two mechanisms. For GI, the plotted C/O is the total volatile C/O (gas and ice together), justified by direct collapse of gas and dust. For SI, gas-phase and ice-phase C/O are shown as separate distributions. Observed exoplanet atmospheric C/O, however, is a single gas-phase quantity. A GI planet's atmosphere need not equal the local total volatile C/O unless all ices vaporize and mix into the envelope, and a core-accretion hot Jupiter's atmosphere need not equal the local gas C/O if solids contribute to the envelope; the paper itself notes that hot Jupiters may accrete hundreds of Earth masses of solids, which lowers atmospheric C/O. Thus the narrow GI peak and broad SI gas-phase tail that appear to match the two observed populations are not predictions of the same observable. The apparent separation could be produced by the phase choice itself: high dust-to-gas ratios in SI regions lower the total C/O (the anticorrelation in Figure 8), while the gas there is carbon-rich; the final planetary C/O then depends on the unspecified solid-to-gas accretion ratio. The abstract presents the diagnostic as established, but the comparison is not a single mechanism-to-atmosphere mapping.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper uses the thin-disc hydrodynamic code FEOSAD, including self-gravity, thermal balance, dust growth and drift, and turbulent diffusion, to follow the phase transitions (freeze-out, thermal desorption, photodesorption) of H2O, CO2, CH4, and CO in two self-gravitating protoplanetary disc models over 0.5 Myr. The authors map the resulting C/O ratio in the gas, in the ice, and in total, and connect these distributions to potential planet formation sites. They identify regions where gravitational instability (GI) and streaming instability (SI) criteria are satisfied, and claim that GI regions have a narrow C/O range (≈0.3–0.6, total) matching directly imaged exoplanets, while SI regions have a wide C/O range (≈0–1.4, gas phase) matching transiting hot Jupiters. They also report an anticorrelation between dust-to-gas ratio and total C/O, and an ice-phase C/O of ≈0.2–0.3 between the CO2, CH4, and CO snowlines consistent with Solar System comets.","tokens_in":48663,"tokens_out":4179,"duration_ms":44677,"significance":"If the GI/SI C/O dichotomy survives scrutiny, it would provide a useful observational diagnostic for distinguishing planet formation mechanisms from atmospheric compositions. The paper's strengths include a global 2D model with self-consistent dust dynamics and phase transitions, explicit consideration of disc substructures and multiple snowline geometries, two different core masses, and quantitative comparisons with exoplanet population data. The authors are also transparent about the model's simplifications, such as the absence of gas-phase and surface chemistry and the thin-disc approximation. However, the central claim is currently built on an inconsistent definition of the observable C/O between the two formation channels, and on a simplified chemical scheme whose quantitative ranges are not shown to be robust. These issues need to be addressed before the main diagnostic can be accepted.","major_comments":[{"comment":"The comparison between GI and SI regions and observed exoplanet C/O ratios uses different C/O definitions: for GI the total volatile C/O is plotted, while for SI the gas-phase and ice-phase C/O are plotted separately. Observed atmospheric C/O is a single gas-phase quantity, so this is not a like-for-like comparison. A GI planet's atmospheric C/O equals the local total C/O only if all accreted solids are vaporized and mixed into the envelope; otherwise it is closer to the gas-phase value. Similarly, an SI-formed planet's atmospheric C/O depends on its solid-to-gas accretion ratio, which is not specified. The apparent narrow GI peak and broad SI gas-phase tail could therefore arise in part from the phase choice rather than from a genuine formation-mechanism difference. I request that Figure 10 be recomputed using a common observable (e.g., gas-phase C/O for both mechanisms, or a simple envelope-formation model that mixes gas and solids with an explicit accretion ratio), and that the conclusions be re-evaluated accordingly.","section":"Section 4, Figure 10"},{"comment":"The chemical model includes only freeze-out, thermal desorption, and photodesorption of four species, with no gas-phase or surface reactions. The authors correctly note in the Discussion that CO-to-CO2 conversion, methanol formation, and volatile trapping in ice mantles can substantially change C/O ratios. Since the paper makes a quantitative prediction (GI total C/O ≈0.3–0.6, SI gas-phase C/O ≈0–1.4), it is important to quantify how sensitive these ranges are to the missing chemistry. A simple post-processing test, such as converting a fraction of CO ice to CO2 in the relevant regions or adopting a methanol abundance, would show whether the GI/SI dichotomy survives. Without such a test, the claimed C/O fingerprints are statements about the simplified chemistry, not robust predictions of the physical disc.","section":"Section 2.3 and Discussion"},{"comment":"The GI C/O distribution in Figure 10 is derived from all regions with Q_Toomre ≤ 1, but the actual GI clumps are under-resolved: the authors state that the clump lifetime is too short for differentiation because of insufficient numerical resolution, and that focused higher-resolution studies are needed. If GI planets form preferentially inside clumps, the material C/O in those clumps may differ from the surrounding Q ≤ 1 regions. The paper should either provide a resolution/convergence test demonstrating that the clump C/O is adequately captured, or justify why the disc-averaged total C/O in GI-unstable regions is representative of the material that ends up in a GI-formed planet.","section":"Section 3.5 and Figure 9"}],"minor_comments":[{"comment":"There are several typos and grammatical slips: \"the disc matter can be roughly divided into three component\" should be \"three components\"; \"annual structures\" in Section 3.1 should be \"annular structures\"; \"circumcise\" in Section 3.1 should be \"circumscribe\"; \"protorstar\" in Section 3.4 should be \"protostar\"; \"he most important\" in Section 2.3 should be \"the most important\"; \"drown dust\" in Section 4 should be \"grown dust\"; \"boader\" in Section 3.3 should be \"border\"; \"V ariations\" in Section 3.3 should be \"Variations\"; \"preset\" in Section 3.3 should be \"present\"; \"two-dimentional\" in Section 3.5 should be \"two-dimensional\"; and \"F ormation\" in the References should be \"Formation\".","section":"Introduction, Section 1"},{"comment":"The caption of Figure 3 reads \"Same as Figure 3 but for model M2\" but it should refer to Figure 2; please correct the cross-reference.","section":"Figure 3 caption"},{"comment":"In the caption of Figure 4, \"By the age of 490 yr\" in the text should read \"490 kyr\"; the time unit is inconsistent with the rest of the paper.","section":"Section 3.3, Figure 4 caption"},{"comment":"The sentence \"The most prominent dust rings are located in the vicinity of the water snowline: the ring outside the primary snowline at 5 – 8 au (depending on the time) and the ring at 1 – 2 au, inside the primary snowline, which at later times also contains water ice and additional snowlines\" is somewhat confusing because the second ring is first described as inside the primary snowline and then as containing water ice. Please rephrase for clarity.","section":"Section 3.1"},{"comment":"The observed exoplanet C/O data from Hoch et al. (2023) are plotted at arbitrary y-axis positions; the figure would be clearer if the data points were shown in a separate panel or with a legend explaining the arbitrary offset.","section":"Section 4, Figure 10"}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper extends the FEOSAD model to track C/O ratios of four volatiles through freeze-out, sublimation, and advection in self-gravitating discs with dust evolution. What is genuinely new: the anticorrelation between total volatile C/O and dust-to-gas ratio, and the maps of where GI and SI conditions overlap with distinct C/O ranges. The model work is careful and the substructure effects—multiple snowlines, ring-driven C/O variations—are described concretely. The comet comparison is a nice touch.\n\nThe main soft spot is the central claim in the abstract and Section 4/Figure 10. For GI regions they plot total C/O (gas plus ice); for SI they plot gas-phase and ice-phase separately. Observed atmospheric C/O is a single gas-phase quantity. The match between the narrow GI total C/O peak and directly imaged planets assumes that GI planets fully vaporize and mix their accreted solids into the envelope, and that the local total C/O survives collapse. That is not argued in the paper. Meanwhile the broad SI gas-phase tail matches transiting planets, but those planets are hot Jupiters that likely accreted solids too, which the paper itself notes lowers C/O. So the apparent dichotomy may be partly an artifact of choosing different definitions. The paper softens this in the discussion, but the abstract states it as a result. A referee should ask for a consistent observable mapping, or at least a careful statement of the assumption.\n\nOther soft spots: the chemistry is limited to four species with zero reactions; CO-to-CO2 conversion and methanol are named as omissions in the discussion, but they can shift C/O at snowlines. The thin-disc geometry excludes vertical C/O gradients, which the authors acknowledge. No convergence tests are shown, and the code is not public—\"data on reasonable request\" only. One key mechanism is attributed to \"Molyarova et al., in prep\", which is hard to check.\n\nThe reader's conditional verdict seems right to me. The stress-test note is valid; I don't think it is a takedown, because the paper's own discussion is hedged, but the abstract overreaches.\n\nOverall: worth sending to peer review. The modeling is credible and the anticorrelation is a genuinely testable prediction. The referee should focus on the phase definition issue and the chemistry simplifications. I'd assign to a competent referee, not desk-reject.\n\nRecommendation: accept with major revision if the authors clarify the observable mapping and add convergence or resolution checks.","headline":"Solid modeling study of C/O in self-gravitating discs, but the GI/SI dichotomy in Figure 10 is softer than the abstract claims because it compares different phase quantities to observations.","tokens_in":49163,"tokens_out":2608,"would_cite":false,"duration_ms":27389,"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":"This paper shows that the carbon-to-oxygen ratio of the gas and ices in self-gravitating protoplanetary discs evolves with dust dynamics, producing a narrow C/O fingerprint for gravitational instability and a wide one for streaming…","keywords":["carbon-to-oxygen ratio","protoplanetary discs","dust evolution","snowlines","gravitational instability","streaming instability","exoplanet atmospheres","volatile transport"],"falsifier":"Observe the $C/O$ ratios of a dozen directly imaged giant planets at separations beyond 10 au: if their $C/O$ ratios scatter as widely as those of transiting hot Jupiters instead of clustering near 0.3–0.6, the gravitational-instability fingerprint proposed here would be falsified.","tokens_in":48205,"feed_emoji":"🪐","tokens_out":7160,"duration_ms":66279,"temperature":0.7,"pith_summary":"The paper argues that the carbon-to-oxygen ratio ($C/O$) of planet-forming material is not just a chemical snapshot but a diagnostic of how planets form. In two self-gravitating disc models evolved for 0.5 Myr with dust growth, drift, and freeze-out/sublimation of H$_2$O, CO$_2$, CO, and CH$_4$, the $C/O$ ratio of gas and ices becomes strongly non-uniform. Regions where gravitational instability (GI) can operate have a narrow total $C/O$ range of about 0.3–0.6, centred near the initial value of 0.34, while regions where streaming instability (SI) can operate show a wide range from 0 to 1.4. This split matches the observed two populations of exoplanets: directly imaged planets cluster around 0.5–0.8, while transiting hot Jupiters scatter widely. If correct, atmospheric $C/O$ becomes a practical way to tell whether a giant planet formed by disc fragmentation or by core accretion with migration.","feed_headline":"C/O ratio fingerprints each planet-formation route","feed_subtitle":"GI regions hold C/O 0.3–0.6; SI regions span 0–1.4, matching the two observed exoplanet populations.","key_machinery":"The central object is the coupled hydrodynamics–dust–ice model FEOSAD, treated as a thin disc with self-gravity, temperature-dependent viscosity, dust growth and drift, and a phase-transition network for four volatiles. The mechanism carrying the argument is the combination of radial drift of icy grains, which transports oxygen-rich ices inward and concentrates them at snowlines, and the criterion-based identification of GI ($Q\\le1$) versus SI regions (the Li–Youdin threshold). The paper then compares the $C/O$ distributions in those regions with observed exoplanet populations and cometary compositions.","core_discovery":"Using the FEOSAD thin-disc hydrodynamics code with self-gravity, two-temperature dust, and a four-species volatile model (H$_2$O, CO$_2$, CO, CH$_4$) that allows freeze-out, thermal desorption, and photodesorption, the authors simulate disc formation and evolution to 0.5 Myr for two core masses. They show that dust drift and the feedback of icy mantles on fragmentation create rings, spirals, and multiple snowlines, and that the total $C/O$ ratio anticorrelates with the dust-to-gas ratio because oxygen-rich ices dominate in dust-rich regions. Weighting all disc regions where the Toomre criterion ($Q \\le 1$) is met, they find that the $C/O$ distribution for gravitational instability has a narrow peak near 0.5, while the regions meeting the Li–Youdin criterion for streaming instability give a broad gas-phase $C/O$ range of roughly 0.2–1.4 and low ice-phase values. They argue this dichotomy is consistent with the observed separation between directly imaged exoplanets (narrow, lower $C/O$) and transiting hot Jupiters (wide $C/O$), and that the ice-phase $C/O\\approx 0.2$–0.3 between the CO, CO$_2$, and CH$_4$ snowlines matches Solar System comets.","pith_inferences":["If gas-phase and surface reactions (CO-to-CO$_2$ conversion, methanol formation, volatile trapping) were added, the boundaries of the $C/O$ zones would likely shift, but the qualitative anticorrelation and the GI/SI separation may survive; rerunning the same model with a reduced chemical network is a testable extension.","The narrow GI peak near 0.5 depends on excluding refractory carbon and on the assumed initial volatile abundances; including carbon-rich rock cores would raise the baseline $C/O$ and could move the GI peak closer to the observed directly imaged value.","This claim implies that atmospheric $C/O$ measurements, combined with orbital separation and age, could identify individual planets formed by disc fragmentation, a prediction testable with current or near-future JWST spectroscopy of wide-orbit planets.","The ice-phase match to comets suggests that most Solar System comets formed between snowlines, while carbon-rich comets like C/2016 R2 require formation at a carbon-rich snowline; linking specific comets to specific zones is a testable prediction."],"forward_implications":["Planets formed by gravitational instability should have atmospheres with $C/O$ clustered near 0.3–0.6, while planets formed by streaming instability and migration can show $C/O$ anywhere from 0 to 1.4.","Directly imaged exoplanets likely formed via GI in the outer disc, whereas transiting hot Jupiters' wide $C/O$ spread reflects migration through varied disc conditions.","Comets with ice-phase $C/O$ 0.2–0.3 formed between the CO and CO$_2$ snowlines, whereas carbon-rich comets likely originated at the snowlines themselves.","Dust-to-gas ratio and total $C/O$ are anticorrelated, so planetesimals assembled in dust-rich rings should start with low $C/O$ unless they later accrete carbon-rich gas."],"supporting_citations":[{"why":"Supplies the observational two-population $C/O$ data (directly imaged vs transiting exoplanets) that the simulations are compared against.","marker":"Hoch et al. (2023)"},{"why":"Provides the snowline-step framework for $C/O$ in planetary atmospheres that this work extends with dust dynamics.","marker":"Öberg, Murray-Clay, and Bergin (2011)"},{"why":"Previous model of volatile freeze-out and sublimation in FEOSAD discs, from which the phase-transition treatment and snowline accumulation results are taken.","marker":"Molyarova et al. (2021)"},{"why":"The FEOSAD code itself, providing self-gravity, dust evolution, and thermal balance as the base for the simulations.","marker":"Vorobyov et al. (2018)"},{"why":"Criterion for streaming instability used to identify SI-active regions in the simulation.","marker":"Li and Youdin (2021)"},{"why":"Companion study of pebble ice composition in the same model, used to discuss what pebble accretion would deliver to planets.","marker":"Topchieva et al. (2024)"},{"why":"Binding-energy data for H$_2$O, CO$_2$, and CO desorption adopted in the chemical model.","marker":"Cuppen et al. (2017)"},{"why":"Supplies the adopted initial ice abundances of H$_2$O, CO$_2$, CO, and CH$_4$ used in the simulations.","marker":"Karin I. Öberg et al. (2011)"}],"fun_headline_variants":["C/O ratio separates gravitational and streaming instabilities","C/O fingerprints split planet formation routes","Icy C/O points to planet formation mechanism","C/O ratio reveals birth zones of exoplanets","Two C/O ranges mark GI and SI planet formation"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The modelling restricts chemistry to freezing and desorption of four ices; if gas-phase or surface reactions (e.g. CO converting to CO$_2$, methanol forming, volatile trapping in mantles) materially alter the midplane composition, the assigned $C/O$ fingerprints would shift.","fun_headline_variants_meta":{"raw":{"variants":["C/O ratio separates gravitational and streaming instabilities","C/O fingerprints split planet formation routes","Icy C/O points to planet formation mechanism","C/O ratio reveals birth zones of exoplanets","Two C/O ranges mark GI and SI planet formation"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000519,"raw_usage":{"total_tokens":2650,"prompt_tokens":1214,"completion_tokens":1436,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":830,"completion_tokens_details":{"reasoning_tokens":1365}},"tokens_in":830,"tokens_out":1436,"duration_ms":10500,"temperature":1.0,"reasoning_tokens":1365,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T20:53:49.888667+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Observe the $C/O$ ratios of a dozen directly imaged giant planets at separations beyond 10 au: if their $C/O$ ratios scatter as widely as those of transiting hot Jupiters instead of clustering near 0.3–0.6, the gravitational-instability fingerprint proposed here would be falsified.","supporting_citations":[],"review_version":1}