{"id":"043ff9a5-3de7-4f1f-9c4c-1d4f86481fc6","arxiv_id":"2412.05535","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":11,"one_line_summary":"The ~34 Myr old M4.5 star J0446B hosts a long-lived, hydrocarbon-dominated primordial disk with a gas-phase C/O ratio of at least 2.","lead":"JWST's mid-infrared spectrum of a 34-million-year-old star reveals a gas disk still rich in hydrocarbons, with more than a dozen carbon molecules and almost no water. This is the first detailed look at disk gas at an age where disks are usually long gone, and it suggests very old, low-mass stars can host carbon-rich planet-forming gas.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The quantitative 'C/O ≳2' claim rests on an optically thick C2H2 slab component whose column density scales inversely with the assumed 2 km/s line width; without propagating that systematic, the late carbon-rich phase result is not yet robust.","rationale":"The reader's conditional verdict already flags age and the unresolved-wind caveat as the weakest assumptions, and lists propagated C/O uncertainties as a condition. I agree that those are genuine concerns, but I see the most load-bearing issue as the unpropagated systematic in the C/O estimate: the abstract's 'gas-phase C/O ratio ≳2' is a headline quantity, and it depends on a compact optically thick C2H2 component whose column is tied to an assumed 2 km/s line width and to continuum placement in wavelength regions dominated by overlapping hydrocarbon emission. This is not an internal inconsistency, and the qualitative C-rich nature of the spectrum is credible; the concern is quantitative robustness. The concrete σ-variation test is cheap and would settle whether the headline C/O value survives a plausible systematic. Since the reader already reached CONDITIONAL and asked for propagated C/O uncertainties, my read does not move the verdict; it strengthens the motivation for that specific condition and provides an operational check.","tokens_in":26464,"tokens_out":7831,"duration_ms":87053,"concrete_test":"Re-run the iris/dynesty slab fits of §3.2 with the turbulent line width set to σ = 1 and 5 km/s (and, ideally, with σ as a free parameter), recomputing N(C2H2,thick)/N(CO2) and the resulting C/O from the Najita et al. grid. Because optically thick columns scale roughly as 1/σ, a factor-2.5 change in σ changes N(C2H2,thick) by about 0.4 dex; if the inferred C/O drops below ~1.5, or if the Bayesian evidence no longer requires the separate thick C2H2 component, the quantitative 'C/O ≳ 2' claim and the carbon-rich-phase framing would need to be correspondingly softened.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central quantitative claim, gas-phase C/O ≳ 2, is not a direct observable. It is obtained by mapping the slab-model column-density ratio N(C2H2)/N(CO2) through the Najita et al. (2011) chemical grid. The numerator is dominated by a separate optically thick C2H2 component with log N = 22.54 ± 0.26 cm^-2 and Rslab = 0.019 AU (§3.2, Table 1). The paper itself notes in §3.2 that, for optically thick lines, column densities scale roughly inversely with the assumed Gaussian line width σ = 2 km/s. No uncertainty is propagated for σ, for the continuum placement in the 7 and 14 μm pseudo-continuum regions, or for the plane-parallel geometry. If the true line width is larger, or if part of the 'thick' C2H2 pseudo-continuum is actually produced by overlapping lines of the many other hydrocarbons detected here, the inferred C2H2 column can decrease substantially. Because the same pipeline gives near-solar C/O for the water-rich disk Sz 114, the O-rich/C-rich dichotomy could be exaggerated if the thick component is partly an artifact. Since 'C/O ≳ 2' is a headline result and underpins the late carbon-rich evolutionary narrative, its current model-dependent status is the most load-bearing weakness.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents JWST MIRI/MRS 4.9–28.6 μm spectroscopy of J0446B, an M4.5 star in the ~34 Myr-old χ1 For association with weak accretion signatures. It reports the detection of 14 molecular species, dominated by hydrocarbons (CH4, C2H2, C2H4, C2H6, C6H6, etc.), plus H2 pure-rotational lines, [Ne II], and [Ar II], with only a marginal H2O detection. Slab-model fits yield a very high column density optically thick C2H2 component and a high N(C2H2)/N(CO2) ratio. Using the Najita et al. (2011) chemical grid, the authors infer a gas-phase C/O ≳2. They argue that the spatially unresolved H2 and [Ne II] emission indicates a long-lived primordial disk, and chemcomp models suggest that maintaining a carbon-rich inner disk for tens of Myr requires α-viscosity ≲1e-4. The paper interprets J0446B as the first detailed characterization of disk gas at ~30 Myr and as evidence for a late-stage carbon-rich phase in disk evolution.","tokens_in":26825,"tokens_out":7458,"duration_ms":75830,"significance":"If the claims hold, this is a significant observational advance: it extends the well-characterized gas-disk phase from ~10 Myr to ~30 Myr, presents the richest hydrocarbon inventory yet seen in an old disk, and connects the observed chemistry to pebble-drift/volatile-transport disk evolution. The strengths of the paper include the use of a published reduction pipeline, public JWST data, careful line identification with HITRAN/iSLAT, and explicit acknowledgment of several caveats (e.g., line-width scaling, model-grid limitations, unresolved wind alternative). The qualitative result that J0446B is molecule-rich and hydrocarbon-dominated appears credible and is well documented. However, the headline quantitative claim (C/O ≳2) and the primordial-disk classification are not yet as secure as the line detections themselves: systematic uncertainties in the slab-model columns and in the chemical-grid mapping are not propagated into the reported C/O, and the H2/[Ne II] unresolved detections do not uniquely exclude a disk wind. These issues are addressable with additional analysis and do not undermine the value of the observations themselves.","major_comments":[{"comment":"The central claim of gas-phase C/O ≳2 (abstract; §4.3.1) is derived from N(C2H2)/N(CO2), where the numerator is dominated by the optically thick C2H2 component with log N = 22.54 ± 0.26 cm^-2. The text in §3.2 states that for optically thick lines the column density scales roughly inversely with the assumed Gaussian width σ = 2 km/s, and the table caption notes that the quoted uncertainties are statistical and likely underestimated. No systematic uncertainty is propagated for σ, for the continuum/pseudo-continuum placement (including the excluded 7.1–8.5 and 12.0–16.5 μm regions), or for the plane-parallel slab geometry. Since the ratio in Figure 6 is what is mapped to C/O, an unquantified factor of a few in the thick C2H2 column can shift the inferred C/O from ≳2 toward ~1. The authors should refit with σ = 3 and 4 km/s, vary the continuum level, and report the resulting range in N(C2H2)/N(CO2) and in C/O. In addition, part of the 'thick' pseudo-continuum could in principle be produced by the many blended hydrocarbon lines detected here; this degeneracy should be explicitly discussed and, if possible, tested with the full molecular inventory included in the fit.","section":"§3.2, Table 1"},{"comment":"The statement in §4.1 that 'the detection of spatially unresolved H2 and [Ne II] lines strongly suggests that J0446B hosts a long-lived primordial gas disk' goes beyond what §3.3 establishes. The same section concedes that 'the possibility of tracing a small-scale unresolved disk wind cannot be completely ruled out, given the moderate spatial resolution of MIRI/MRS (0.2–0.3).' Since [Ne II] is routinely identified as a disk-wind/jet tracer (as the paper itself notes), the unresolved detection does not uniquely require a quasi-static primordial disk. The authors should provide a quantitative discriminator — for example, line-centroid shifts, flux ratios among the H2 transitions, or an upper limit on the emitting radius from the point-spread function — or should temper the 'first confirmed case' claim to say that the emission is consistent with primordial disk gas but an unresolved wind is not excluded.","section":"§3.3 and §4.1"},{"comment":"The conversion from N(C2H2)/N(CO2) to C/O relies on the Najita et al. (2011) chemical grid, which the text itself notes was developed for warmer T Tauri disks and does not include pathways to the complex hydrocarbons (C2H4, C2H6, C6H6, etc.) that dominate the observed spectrum. These two effects may act in opposite directions on the derived C/O, but they are not quantified, so the resulting 'C/O ≳2' is a model-dependent inference without a stated systematic error. The authors should test the mapping with a network that includes these species at intermediate C/O values (e.g., a Kanwar et al. 2024b-type grid with C/O = 1–4 for J0446B-like parameters) or otherwise provide a quantitative uncertainty on C/O from the grid choice. Without this, the abstract's headline number is not yet robust, even if the qualitative carbon-rich conclusion survives.","section":"§4.3.1, Figure 6"}],"minor_comments":[{"comment":"The note that the uncertainties are 'statistical and likely underestimated' is important for the C/O derivation and should be repeated in the main text where the column densities are used, not only in the table footnote.","section":"Table 1"},{"comment":"The statement that 'the resulting C/O ratio is also close to 2 based on the column density ratio of C2H2 and H2O' is based on an upper limit for H2O; it should be phrased as a lower limit on C/O from water, not as a measurement.","section":"§4.3.1"},{"comment":"The observed values/ranges plotted in the left panel lack visible error bars; since the x-axis is a model-derived quantity, the figure should include propagated uncertainties or state explicitly that they are omitted.","section":"Figure 6"},{"comment":"The text first quotes the M6 spectral type from Silverberg et al. (2020) and then derives M4.5; please state explicitly that the adopted spectral type is M4.5 (Teff ≈ 3100 K) throughout the remainder of the paper to avoid confusion.","section":"§2.1"},{"comment":"The phrase 'first confirmed case of disk gas surviving for more than 30 Myr' is stronger than the abstract's 'first detailed characterization of disk gas at ~30 Myr'; consider harmonizing the wording given the unresolved-wind degeneracy.","section":"§4.1 and Abstract"},{"comment":"The conclusion that maintaining C/O > 1 at ~30 Myr requires α ≲ 1e-4 depends on the assumed initial volatile partitioning (60% refractory carbon, 20% CO, 10% CO2, 10% CH4) and v_frag = 5 m/s; the paper should state how sensitive Figure 8 is to these choices.","section":"§4.3.2, Figure 8"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is a strong observational paper whose raw detections and qualitative carbon-rich characterization are likely solid. The main risk is that the quantitative C/O ≳2 and the primordial-disk classification are presented as more secure than the current analysis warrants. The paper also leans on coauthored references (Najita et al. 2011; Mah et al. 2023) for the central interpretation; this is not a problem in itself, but the referee should be aware that the C/O inference is not independent of the authors' own theoretical framework. If the authors can quantify the systematic uncertainties or soften the claims accordingly, the paper would be suitable for publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Feng and colleagues have done something genuinely new: they obtained a MIRI/MRS spectrum of the inner disk of J0446B, a ~34 Myr M4.5 star with hints of accretion, and it is molecule-rich, hydrocarbon-dominated. The detections of H2, [Ne II], [Ar II], and 14 molecular species at this age are firsts, and the line identifications look solid; the reduction follows the published JDISCS pipeline and the data are in MAST. The placement of J0446B in the M-dwarf disk sample (Table 3, Figure 5) is a useful synthesis, showing it as the most extreme in C2H2/H2O ratio while sharing excitation conditions with younger carbon-rich disks like ChaI-147.\n\nThe paper is honest about its main caveats. Table 1 flags that the quoted uncertainties are statistical and likely underestimated. Section 3.3 admits the unresolved H2 and [Ne II] could trace a small-scale disk wind rather than quiescent disk gas. The age rests on chi1 For membership, which is reasonable but not independent.\n\nThe soft spot is the headline 'gas-phase C/O ratio ≳2.' That number is not a direct observable. It comes from mapping the slab-model column-density ratio N(C2H2)/N(CO2) through the Najita et al. grid. The C2H2 column is dominated by an optically thick component with log N = 22.54, and the paper itself notes in §3.2 that optically thick columns scale roughly inversely with the assumed Gaussian width σ = 2 km/s. No uncertainty is propagated for σ, for the continuum placement, or for the plane-parallel geometry. If σ is larger or part of the 12–16 µm pseudo-continuum is actually overlapping lines of the many other hydrocarbons detected here, the C2H2 column drops and the C/O estimate weakens. The stress-test note is right to call this the most load-bearing weakness. It does not kill the paper: even the optically thin C2H2 component gives a super-solar ratio, and the qualitative conclusion of a very carbon-rich inner disk is well supported. But '≳2' is a first-order estimate, and the abstract overstates its firmness.\n\nSimilarly, the α < 10^-4 conclusion comes from forward chemcomp runs, not a fit, and is presented as a suggestion. Fine. The ALMA 0.9 mm flux is a private communication and should be public.\n\nMy take: this is a solid observational discovery paper that deserves a serious referee. The central detection—a long-lived, gas-rich, hydrocarbon-dominated disk at ~30 Myr—will hold. I would condition acceptance on propagating the line-width systematic into the C/O estimate, softening 'first confirmed case' to account for the wind caveat, and making the ALMA flux available. Those are revisions, not reasons to desk reject.","headline":"First JWST spectrum of a ~30 Myr disk is a real observational step forward; the carbon-rich conclusion is credible but the C/O>2 number is softer than the abstract implies.","tokens_in":27513,"tokens_out":3807,"would_cite":true,"duration_ms":34042,"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":"JWST spectroscopy shows a 34-million-year-old low-mass star still surrounded by a carbon-rich planet-forming gas disk.","keywords":["protoplanetary disks","JWST MIRI/MRS","disk chemistry","carbon-rich disk","hydrocarbons","disk lifetime","low-mass stars","C/O ratio"],"falsifier":"A decisive check would be to resolve the H$_2$ and [Ne II] emission spatially or spectrally at higher angular and spectral resolution: if the lines are extended or systematically blueshifted, they trace a disk wind rather than a primordial disk. Independently, a lithium detection or a revised moving-group membership that puts J0446B at an age under $\\sim$10 Myr would remove the '30-Myr-old disk' framing, and a direct measurement of the gas-phase $\\mathrm{C/O}$ from another tracer (for example, ALMA observations of HCN or C$_2$H) that gives $\\mathrm{C/O} < 1$ would undercut the carbon-rich claim.","tokens_in":26255,"feed_emoji":"🔭","tokens_out":13383,"duration_ms":82175,"temperature":0.7,"pith_summary":"This paper uses JWST mid-infrared spectroscopy to show that the inner disk around J0446B, an M4.5 star thought to be $\\sim$34 Myr old, is filled with hydrocarbons—acetylene, benzene, methane, and many others—and almost no water. The detection of spatially unresolved molecular hydrogen and neon emission lines is read as evidence that this is a long-lived primordial gas disk, not a debris disk, making it the first confirmed case of disk gas surviving past 30 Myr. From the ratio of acetylene to carbon dioxide, the authors estimate a gas-phase carbon-to-oxygen ratio $\\mathrm{C/O} \\gtrsim 2$, so the inner disk is very carbon-rich. They argue this fits a picture in which low-mass stars deplete their icy pebbles early, pass through a water-rich phase, and end up carbon-rich while the disk slowly drains inward. If correct, the result extends the known lifetime of planet-forming gas disks by an order of magnitude and links late disk chemistry to the carbon content of planets forming around the smallest stars.","feed_headline":"A 34-million-year-old star still has a carbon-rich planet-forming disk","feed_subtitle":"JWST's first look at disk gas at 30 Myr finds hydrocarbons and almost no water.","key_machinery":"The central object is the MIRI/MRS spectrum itself: continuum-subtracted mid-infrared data over 4.9–28.6 $\\mu$m, fitted with LTE plane-parallel slab models that include optical-depth and line-overlap effects, producing column density, temperature, and emitting area for each molecule. Two derived ratios carry the chemical argument: the C$_2$H$_2$/CO$_2$ column-density ratio, converted to gas-phase $\\mathrm{C/O}$ through thermo-chemical model grids, and the H$_2$ S(1)/S(3) line ratio, which yields the warm gas temperature. The time-evolution claim is carried by a 1D $\\alpha$-disk model that tracks the inward drift and sublimation of icy pebbles and the accretion of outer carbon-rich gas, showing that a high inner $\\mathrm{C/O}$ can be sustained for tens of Myr at $\\alpha \\lesssim 10^{-4}$.","core_discovery":"The paper's central claim is that J0446B is the first confirmed example of a primordial protoplanetary disk with gas surviving past $\\sim$30 Myr, and that its inner disk has entered a late, hydrocarbon-dominated, carbon-rich phase. The evidence is a MIRI/MRS spectrum from 4.9 to 28.6 $\\mu$m showing 14 molecular species, including nine hydrocarbons with a very optically thick C$_2$H$_2$ component, a marginal water detection, five pure-rotational H$_2$ lines, and spatially unresolved [Ne II] and [Ar II] lines. Slab-model column densities, mapped through thermo-chemical model grids, imply a gas-phase $\\mathrm{C/O} \\gtrsim 2$. The authors place J0446B at the oldest end of the JWST M-star disk sample, with hydrocarbon excitation conditions (T $\\sim$ 250–300 K, emitting radii 0.05–0.1 AU) similar to younger carbon-rich disks, and use a 1D viscous disk model with pebble drift and volatile transport to argue that such a carbon-rich state can persist for tens of Myr only if the disk viscosity is low, $\\alpha \\lesssim 10^{-4}$.","pith_inferences":["The paper does not forecast how common long-lived carbon-rich disks are; a natural next step is a JWST survey of the known accreting M-dwarf disks at 20–50 Myr, measuring the fraction that are hydrocarbon-dominated. The pebble-drift picture predicts that fraction rises with age and anticorrelates with millimeter flux.","The carbon-rich inner disk implies carbon is being transferred from solids to gas; if this is general, terrestrial planets assembled late in such systems could be carbon-poor even while the gas they accrete is carbon-rich, a bias worth folding into interpretations of exoplanet atmospheric C/O for M-dwarf systems like TRAPPIST-1.","A testable extension of the low-viscosity requirement: high-spectral-resolution observations of CO or H$_2$ line profiles in J0446B could measure turbulence and constrain $\\alpha$ directly, providing an independent check on the $\\alpha \\lesssim 10^{-4}$ limit inferred from chemistry.","The unresolved [Ne II] and [Ar II] lines could be used as a diagnostic in other old disks to tell whether stellar high-energy radiation or an unseen wind controls the final disk dispersal."],"forward_implications":["If J0446B is truly a 34-Myr-old primordial disk, disk gas can survive roughly ten times longer than the canonical few-Myr disk lifetime, at least around very low-mass stars.","Planets forming in such a disk would accrete gas with $\\mathrm{C/O} \\gtrsim 2$, producing carbon-rich atmospheres with different chemistry and haze formation than solar-composition planets.","The millimeter-faint, carbon-rich disks around late M stars are expected to be the norm at ages beyond $\\sim$10 Myr, while water-rich disks like Sz 114 are the exception, tied to bright millimeter emission and dust substructures that trap icy pebbles.","Maintaining the carbon-rich phase for tens of Myr requires a slowly evolving disk with $\\alpha \\lesssim 10^{-4}$, linking the observed chemistry to the disk's viscosity and lifetime.","The comparable fluxes of [Ne II] and [Ar II] imply soft X-ray/EUV ionization dominates, so stellar high-energy radiation alone may not disperse such disks."],"supporting_citations":[{"why":"Identified J0446 as a strong infrared-excess source with accretion hints and suggested membership in the Columba association, making it the target of this study.","marker":"Silverberg et al. 2020"},{"why":"Reassigned the system to a nearby moving group with an age of 33.7 Myr based on lithium depletion; this adopted age anchors the '30-Myr-old disk' claim.","marker":"Luhman 2024"},{"why":"Provides the thermo-chemical model grids that convert the measured C$_2$H$_2$/CO$_2$ column-density ratio into a gas-phase C/O estimate.","marker":"Najita et al. 2011"},{"why":"Proposed the pebble-drift plus gas-accretion scenario for inner-disk C/O evolution that this paper extends to 40 Myr.","marker":"Mah et al. 2023"},{"why":"Established the JWST MIRI/MRS methodology for optically thick C$_2$H$_2$ emission and a carbon-rich comparison disk in Upper Sco.","marker":"Tabone et al. 2023"},{"why":"Provides the JWST spectrum of the carbon-rich disk ChaI-147 used for species identification and sample comparison.","marker":"Arabhavi et al. 2024"},{"why":"Supplies the JWST spectrum of the water-rich M-dwarf disk Sz 114, the key opposite chemical state and bright-millimeter comparison.","marker":"Xie et al. 2023"},{"why":"Provides the 1D viscous disk evolution code used to model pebble drift and volatile transport for the time-evolution calculations.","marker":"Schneider & Bitsch 2021"},{"why":"Spitzer spectra of late-M disks showing weak water and enhanced C$_2$H$_2$, the earlier observational baseline this work extends.","marker":"Pascucci et al. 2013"},{"why":"Supplies the two-population dust-evolution algorithm used in the model to compute pebble drift and grain growth.","marker":"Birnstiel et al. 2012"}],"fun_headline_variants":["JWST unveils carbon-rich gas in a 34-Myr-old planet-forming disk","Oldest known disk gas: hydrocarbon-rich, almost no water","First JWST view of a 30-Myr disk reveals carbon-rich chemistry","A 34-Myr-old star still feeds a carbon-rich disk"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument assumes both that J0446B is truly $\\sim$34 Myr old (with membership and lithium age taken from the literature rather than measured independently for this star) and that the unresolved H$_2$ and neon lines trace disk gas rather than a small-scale disk wind.","fun_headline_variants_meta":{"raw":{"variants":["JWST unveils carbon-rich gas in a 34-Myr-old planet-forming disk","Oldest known disk gas: hydrocarbon-rich, almost no water","First JWST view of a 30-Myr disk reveals carbon-rich chemistry","A 34-Myr-old star still feeds a carbon-rich disk"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000378,"raw_usage":{"total_tokens":2155,"prompt_tokens":1235,"completion_tokens":920,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":851,"completion_tokens_details":{"reasoning_tokens":840}},"tokens_in":851,"tokens_out":920,"duration_ms":9278,"temperature":1.0,"reasoning_tokens":840,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T20:38:22.190888+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A decisive check would be to resolve the H$_2$ and [Ne II] emission spatially or spectrally at higher angular and spectral resolution: if the lines are extended or systematically blueshifted, they trace a disk wind rather than a primordial disk. Independently, a lithium detection or a revised moving-group membership that puts J0446B at an age under $\\sim$10 Myr would remove the '30-Myr-old disk' framing, and a direct measurement of the gas-phase $\\mathrm{C/O}$ from another tracer (for example, ALMA observations of HCN or C$_2$H) that gives $\\mathrm{C/O} < 1$ would undercut the carbon-rich claim.","supporting_citations":[],"review_version":1}