{"id":"0e3739c0-d559-40fe-b192-5bfb02e75640","arxiv_id":"2502.01765","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"In chemical models of inner protoplanetary disk gas, X-ray-driven chemistry stores carbon in long-chain hydrocarbons and CO, while UV-driven chemistry stores it in atomic carbon and CO, and water strongly enhances C2H2.","lead":"Protoplanetary disks with bright hydrocarbon emission might be explained by chemical models where the dominant radiation source, X-rays versus ultraviolet light, decides which carbon molecules survive. The models show that water strongly boosts acetylene production, and that methane is a short-lived carbon donor, which can help interpret JWST spectra of planet-forming disks.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The claimed water enhancement of C2H2 is confounded: the 'water-rich' comparison also changes the carbon carrier and CO abundance, and the C/O grid is degenerate with initial water.","rationale":"I read the paper as a deliberately simplified single-cell survey whose headline is that the chemistry driver and water abundance control carbon partitioning and C2H2 production. The driver contrast and network limitations are discussed candidly, and the UMIST/three-body concern identified by the reader is real but applies to any gas-phase network study; it mostly affects quantitative yields and is already flagged by the authors. The most load-bearing defect is internal to the presented grid: Table 3 conflates initial water abundance with C/O and with the initial carbon carrier. Because the water claim is in the abstract, Fig. 5, and Section 4.5, this confound touches a primary conclusion. A single controlled rerun with oxygen supplied as atomic O rather than water, holding C/O and the CH4 carbon donor fixed, would settle whether water itself is responsible. I therefore keep the reader's conditional verdict, but the justification shifts from network completeness to an experimental-design confound that is directly testable. The model-count inconsistency (42 vs 80 vs 48) and CO-photodissociation ambiguity also need fixing, but neither is as directly tied to the central water claim as the confounding in Table 3.","tokens_in":19378,"tokens_out":8473,"duration_ms":75973,"concrete_test":"Run a controlled X-ray fiducial pair (ζ15, 400 K, n_H = 1e8 cm^-3, C/O = 1) with CH4 = 1.5e-4 and CO = 1e-14 fixed, while exchanging oxygen between H2O and atomic O: (A) H2O = 1.5e-4, O = 1e-14; (B) H2O = 1e-8, O = 1.5e-4. If equilibrium C2H2 is similar in A and B, the water-dependence claim fails; if B is suppressed by roughly an order of magnitude, the claim survives. As a secondary check, repeat Models 3-5 with H2O held at 1.5e-4 and oxygen compensated by lowering CO or adding atomic O to see whether the C/O = 1 optimum persists when water is constant.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section 4.5 and Fig. 5 base the claim that C2H2 is enhanced up to an order of magnitude by initial water abundance on a comparison of Models 2 and 6. Table 3, however, changes three initial abundances simultaneously: H2O drops from 1.5e-4 to 1e-8, CO rises from 1e-14 to 1.5e-4, and CH4 drops from 1.5e-4 to 1e-14. The high-C2H2 model thus starts with carbon in reactive CH4 and oxygen in water, while the low-C2H2 model starts with carbon locked in stable CO and almost no CH4. The lower C2H2 could be caused by the absence of the CH4 carbon donor or by the presence of CO as a carbon sink, not by the absence of water. The same degeneracy affects the optimum at C/O=1: Models 2-5 keep CH4 and CO fixed and raise C/O by decreasing H2O, so the reported decline of C2H2 with increasing C/O is perfectly anti-correlated with initial H2O. Reaction 8 (H2O + C2H3+ -> C2H2 + H3O+) makes it natural that C2H2 tracks H2O, but the grid cannot separate water abundance from C/O or from the chemical form of the initial oxygen and carbon reservoirs.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents single-cell gas-phase chemical models of the inner regions of protoplanetary disks, integrating the UMIST 2012 network with a stiff ODE solver. It explores how the carbon reservoir is partitioned as a function of C/O ratio, ionization rate, and the dominant ionization driver (X-ray/cosmic-ray versus UV). The authors report that X-ray-driven chemistry favors CO and long-chain hydrocarbons as carbon sinks, whereas UV-driven chemistry favors atomic carbon and CO, and they claim that C2H2 production is optimized at C/O=1 and enhanced by up to an order of magnitude by the initial water abundance through reactions such as H2O + C2H3+ -> C2H2 + H3O+. The paper is framed as a pathfinder for interpreting JWST hydrocarbon spectra of planet-forming disks.","tokens_in":19655,"tokens_out":9353,"duration_ms":85417,"significance":"The methodological core is conventional but sound: the chemical network is an externally published standard, the integration uses a public stiff solver, and the Jacobian-based pathway analysis is a clear and useful tool. The abundance maps for C2H2, CH4, CO2, and other hydrocarbons across ionization conditions and C/O ratios are a useful reference for interpreting mid-infrared disk observations. There is no circularity: no quantity is fitted to data, and the results are emergent outputs of the assumed network. The authors also deserve credit for explicitly flagging the omission of three-body reactions and PAH formation in Section 4.3. However, the two headline conclusions currently sit on fragile supports: the water-enhancement claim is confounded by simultaneous changes in the initial carbon and oxygen carriers, and the long-chain hydrocarbon carbon-sink claim has not been tested against a network that includes three-body chemistry. If the authors add the requested control runs and reconcile the model-count accounting, the paper will be a useful contribution to the field.","major_comments":[{"comment":"The central claim in the abstract and §4.5 that C2H2 production is enhanced by up to an order of magnitude by the initial water abundance is not supported by the model grid as presented. Figure 5 compares Models 2 and 6, but Table 3 changes three abundances at once: H2O is lowered from 1.5e-4 to 1e-8, CO is raised from 1e-14 to 1.5e-4, and CH4 is lowered from 1.5e-4 to 1e-14. The lower C2H2 could therefore be caused by carbon being locked in CO or by the absence of a reactive CH4 donor, rather than by the absence of H2O. In addition, the C/O sequence Models 2-5 is constructed by decreasing H2O while keeping CH4 and CO fixed, so the reported decline of C2H2 with increasing C/O is degenerate with decreasing initial water. Reaction (8), H2O + C2H3+ -> C2H2 + H3O+, makes a real water dependence plausible, but the model grid alone cannot separate water abundance from C/O or from the chemical state of the initial oxygen and carbon reservoirs. I request at least one control model in which only H2O is varied at fixed C/O, fixed CO, and fixed CH4, or an explicit sensitivity analysis that isolates the contribution of Reaction (8).","section":"§4.5, Fig. 5, Table 3, Reactions (8) and (24)"},{"comment":"The abstract's statement that CO and long-chain hydrocarbons act as carbon sinks in the cosmic/X-ray-driven limit is conditional on the UMIST 2012 gas-phase network. Section 4.3 itself notes that Kanwar et al. (2023) included three-body reactions and PAH formation, and that this 'can lead to vastly different reaction pathways than the simple ones we explore.' At the adopted n_H = 1e8 cm^-3 and T = 400 K, three-body association is expected to be competitive, so the carbon-sink partitioning could change qualitatively. Since Figure 4 and §4.1 use the long-chain species (C5, C6, C8) to contrast the X-ray and UV carbon-sink behavior, this omission is load-bearing. I recommend either adding three-body reactions and PAH formation in at least the fiducial X-ray and UV models, or explicitly framing the long-chain result as a two-body-only network result and qualifying the abstract accordingly.","section":"§4.3, §4.1, Fig. 4"}],"minor_comments":[{"comment":"The number of models is inconsistent: the abstract states 42 models, the introduction states 80 individual models, and Tables 2 and 3 imply 48 if every initial-condition model is run under every physical condition. Please reconcile the number and state the exact grid in one place.","section":"Abstract, §1, Tables 2-3"},{"comment":"The '...' notation in Table 3 is ambiguous for Model 6, because the row implies solar CH4 (1e-14) and solar CO (1.5e-4); please give explicit values and state which models correspond to the 'reduced water' case.","section":"Table 3, Model 6"},{"comment":"The Fig. 5 caption says 'with and without the presence of water (Models #2 vs. 6)', but these models also differ in CH4 and CO; rephrase to describe the two initial-condition prescriptions in Table 3 or add the requested control model.","section":"Fig. 5 caption"},{"comment":"The phrase 'we proved one simulation' should read 'we ran one simulation'.","section":"§4.6"},{"comment":"The text says the evolution is run 'with logarithmic steps assuming a ratio of 1.1'; please define whether 1.1 is the ratio of successive time steps and state how many time steps were used.","section":"§2.2"},{"comment":"The claim that the UV-driven limit places the majority of carbon in atomic carbon and CO is based on Figure 4 for a single representative C/O ratio (Model #3); please state more explicitly how the conclusion extends across the full C/O grid, or add a supporting figure.","section":"§4.1"}],"recommendation":"major_revision","confidential_remarks":"The water-dependence conclusion appears overstated in the abstract relative to what the current grid can demonstrate, but it is repairable with control models. The model-count discrepancy between the abstract and the text suggests the manuscript needs a careful consistency pass. The authors' explicit caveats about the network are a positive sign, but those caveats must be propagated into the abstract so that the long-chain hydrocarbon sink claim is not read as a network-independent result."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's my take on Raul et al. The useful core is the systematic grid: six initial chemical conditions, C/O from 0.4 to 100, four ionization rates, four UV fields, run to 3 Myr with UMIST 2012. The contrast between X-ray-driven chemistry (carbon ending in CO and long chains like C5, C6, C8) and UV-driven chemistry (carbon ending mostly in C and CO) is clean and potentially useful for interpreting JWST spectra of inner disks. The C2H2 peak at C/O=1, rather than at extreme C/O, is also worth knowing. The paper is honest about its own limitations, including the lack of CO photodissociation, the single-point geometry, and the absence of three-body reactions and PAHs in UMIST 2012.\n\nThe big problem is the water–C2H2 claim in §4.5 and Fig. 5. Models 2 and 6 differ in three initial abundances at once: H2O (1.5e-4 vs 1e-8), CO (1e-14 vs 1.5e-4), and CH4 (1.5e-4 vs 1e-14). The water-rich model starts with carbon in reactive CH4 and almost no CO; the water-poor model starts with carbon locked in stable CO. The reported order-of-magnitude water enhancement cannot be separated from the carbon carrier or the CO reservoir. The same degeneracy runs through the C/O grid: Models 2–5 raise C/O by lowering H2O while keeping CH4 fixed, so the C2H2 decline with C/O is also a decline with initial water. Reaction 8 (H2O + C2H3+ -> C2H2 + H3O+) makes a water link plausible, but the grid does not demonstrate it. This needs a cleaner test, e.g., varying H2O at fixed CO and CH4, before the claim goes out.\n\nThe model-count inconsistency (42 vs 80 vs 48) is minor but should be fixed. No code or data are provided; that is not fatal given the standard network, but a table of equilibrium abundances would help reproduction.\n\nThe X-ray/UV dichotomy and the C/O trends are likely qualitatively robust. The water claim is the one that matters for observers, and it is the one that needs rework. This paper deserves a serious referee — the grid is new and the interpretive tool is useful — but I would send it back for major revision, not reject it.","headline":"Useful hydrocarbon grid for JWST interpreters, but the headline water–C2H2 claim is confounded by the model setup.","tokens_in":20240,"tokens_out":4389,"would_cite":false,"duration_ms":35322,"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":"The chemistry driver decides where carbon ends up in planet-forming disk gas.","keywords":["astrochemistry","protoplanetary disks","hydrocarbon chemistry","C/O ratio","cosmic-ray ionization","UV photochemistry","C2H2","carbon grain destruction"],"falsifier":"Rerun the same single-cell grid at $n_{\\mathrm{H}} = 10^8$ cm$^{-3}$ and 400 K with the three-body and PAH-forming reactions that the paper cites in Section 4.3 included; the central claim fails if long-chain hydrocarbons no longer dominate the carbon budget in the X-ray limit, or if the order-of-magnitude water enhancement of C2H2 disappears. A complementary check: find a water-poor inner disk with strong persistent C2H2 emission, which the models say should not occur.","tokens_in":19138,"feed_emoji":"🪐","tokens_out":7337,"duration_ms":61629,"temperature":0.7,"pith_summary":"This paper asks where carbon goes when carbon-rich grains release their carbon into the hot inner gas of planet-forming disks, the region where terrestrial planets assemble. Using a grid of single-cell chemical models run at 400 K with gas density $10^8$ cm$^{-3}$, the authors vary the C/O ratio from 0.4 to 100 and compare two chemistry drivers: ionization by cosmic rays/X-rays versus ultraviolet light. They find the driver, not the C/O ratio, mostly sets the final carbon reservoirs: X-ray-driven chemistry parks carbon in CO and long-chain hydrocarbons (C5, C6, C8), while UV-driven chemistry parks it in atomic carbon and CO. They also find that C2H2, a key JWST tracer, is produced most efficiently at C/O = 1 and is boosted up to an order of magnitude by water, through the reaction $\\mathrm{H_2O + C_2H_3^+ \\rightarrow C_2H_2 + H_3O^+}$. The work matters because JWST now sees hydrocarbon-rich inner disks, and knowing which conditions produce those molecules constrains how refractory carbon is destroyed and whether forming planets inherit carbon-poor solids.","feed_headline":"X-rays send disk carbon to long chains; UV to atoms","feed_subtitle":"Single-cell chemistry models show where carbon from destroyed grains lands, and why C2H2 peaks at C/O=1 with water.","key_machinery":"The engine of the analysis is a single-cell chemical kinetics model built on the UMIST 2012 gas-phase network, with 536 species and 7349 reactions, integrated to 3 Myr at a fixed 400 K and $n_{\\mathrm{H}} = 10^8$ cm$^{-3}$ with $10^{-2}$ cm grains. Carbon supply is injected by raising initial CH4, C, H2O, or CO abundances to set C/O ratios from 0.4 to 100, and the same grid is solved twice: once with only cosmic-ray/X-ray ionization ($\\zeta = 5\\times10^{-17}$ to $10^{-12}$ s$^{-1}$) and once with only UV photochemistry (0.007 to 45 $G_0$, i.e., multiples of the standard interstellar UV field). Dominant production and destruction pathways are extracted from the Jacobian of the network, which identifies which reaction carries the flux for each species at each time. The specific identity carrying the water dependence of C2H2 is the neutralization reaction $\\mathrm{H_2O + C_2H_3^+ \\rightarrow C_2H_2 + H_3O^+}$, which ties acetylene production to the presence of water rather than to free oxygen.","core_discovery":"The paper's central claim is that the carbon chemistry of inner disk gas is governed by what powers the chemistry, not simply by how much carbon is added. In the cosmic-ray/X-ray-driven limit, the excess carbon released from grains is funneled into CO and long-chain hydrocarbons C5, C6, C8, with C2H2, C2H4, C4H2, and C6H6 appearing transiently, while in the UV-driven limit the vast majority of carbon ends in atomic carbon and CO, with hydrocarbon enrichment short-lived. A second load-bearing claim is that C2H2 is optimally produced at C/O = 1, not at higher C/O, and that its abundance depends strongly, up to an order of magnitude, on initial water abundance via $\\mathrm{H_2O + C_2H_3^+ \\rightarrow C_2H_2 + H_3O^+}$. The authors further find that CH4 is a transient carbon donor that hands its carbon to CO, C2H2, and HCN, and that the choice of methane versus atomic carbon as the initial carrier of excess carbon hardly affects the final equilibrium, with both converging on similar timescales set by the ionization rate.","pith_inferences":["The authors do not pursue it, but their C/O = 1 optimum for C2H2 implies that high observed C2H2/CO2 ratios need not indicate extremely carbon-rich gas; they could mark moderate C/O combined with water-rich, X-ray-irradiated gas.","An untested consequence of the UV-limit result is that atomic carbon (C I) should be a bright reservoir in UV-driven inner disks, so C I line observations could directly test where carbon grain destruction products land.","Given the authors' flag that three-body reactions can change the pathways, rerunning this grid with the UMIST 2022 network or with three-body and PAH chemistry included would show whether the long-chain hydrocarbon sink is robust or an artifact of the 2012 network.","The convergent behavior of CH4 versus C as initial carriers suggests that in real disks the identity of the grain-destruction product matters mainly for early-time chemistry, not for the equilibrium composition; this could simplify how carbon-supply terms are parameterized in disk models."],"forward_implications":["In X-ray/cosmic-ray-dominated inner disk gas, most excess carbon from grain destruction will be hidden in CO and in long-chain hydrocarbons (C5, C6, C8), so those species, not just C2H2, are the reservoirs to observe.","In UV-dominated gas, atomic carbon and CO hold the vast majority of the carbon; hydrocarbon molecules are transient, surviving only briefly or when carbon is continuously resupplied.","C2H2 is predicted to be brightest at C/O = 1 and in water-rich gas; boosting C/O beyond unity lowers its equilibrium abundance because carbon is parsed into chains in the X-ray case or into atomic carbon in the UV case.","Water presence enhances C2H2 by up to an order of magnitude through $\\mathrm{H_2O + C_2H_3^+ \\rightarrow C_2H_2 + H_3O^+}$, while water-poor gas shifts carbon toward CH3, CH4, and CH3+.","CH4 is a short-lived carbon donor in this chemistry, so its detection implies either recent or ongoing methane supply or a weakly ionizing environment, and whether the initial carrier is CH4 or atomic C barely changes the final equilibrium."],"supporting_citations":[{"why":"Provides the UMIST 2012 gas-phase reaction network that defines the chemistry solved in every model.","marker":"McElroy et al. 2013"},{"why":"Provides the X-ray ionization-rate estimates used to choose the ionization grid for inner disk surfaces.","marker":"Glassgold et al. 1997"},{"why":"Supplies the X-ray absorption cross-sections that justify the ionization-rate range as corresponding to molecular emission layers.","marker":"Bethell & Bergin 2011"},{"why":"Is the comparison point for C2H2 formation at C/O = 0.4 and for water-assisted inner-disk chemistry.","marker":"Duval et al. 2022"},{"why":"Is the prior inner-disk chemistry model whose long-chain carbon reservoirs are compared with the present long-chain sinks.","marker":"Wei et al. 2019"},{"why":"Is cited for the three-body and PAH-forming reactions that the paper flags as a possible alternative pathway absent from its network.","marker":"Kanwar et al. 2023"},{"why":"Provides the JWST observations of hydrocarbon-rich disk spectra that motivate the elevated-C/O scenario.","marker":"Tabone et al. 2023"},{"why":"Supplies the carbon-grain-destruction scenario that is the proposed source of excess gaseous carbon.","marker":"Li et al. 2021"}],"fun_headline_variants":["X-rays build long chain hydrocarbons; UV breaks to atoms","Disk carbon fate: X-rays make chains, UV makes atoms","C2H2 peaks at C/O=1, water boosts it tenfold","Radiation type decides disk carbon: chains or atoms","Methane hands off carbon; X-rays and UV steer the path"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing assumption is that the UMIST 2012 gas-phase reaction set, without three-body reactions or polycyclic aromatic hydrocarbon formation, captures the dominant hydrocarbon chemistry at 400 K and $n_{\\mathrm{H}} = 10^8$ cm$^{-3}$; the paper itself flags in Section 4.3 that adding three-body pathways can lead to vastly different reaction pathways.","fun_headline_variants_meta":{"raw":{"variants":["X-rays build long chain hydrocarbons; UV breaks to atoms","Disk carbon fate: X-rays make chains, UV makes atoms","C2H2 peaks at C/O=1, water boosts it tenfold","Radiation type decides disk carbon: chains or atoms","Methane hands off carbon; X-rays and UV steer the path"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000465,"raw_usage":{"total_tokens":2390,"prompt_tokens":1081,"completion_tokens":1309,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":697,"completion_tokens_details":{"reasoning_tokens":1221}},"tokens_in":697,"tokens_out":1309,"duration_ms":8830,"temperature":1.0,"reasoning_tokens":1221,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-09T14:32:07.707424+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Rerun the same single-cell grid at $n_{\\mathrm{H}} = 10^8$ cm$^{-3}$ and 400 K with the three-body and PAH-forming reactions that the paper cites in Section 4.3 included; the central claim fails if long-chain hydrocarbons no longer dominate the carbon budget in the X-ray limit, or if the order-of-magnitude water enhancement of C2H2 disappears. A complementary check: find a water-poor inner disk with strong persistent C2H2 emission, which the models say should not occur.","supporting_citations":[{"cited_title":"2019, ApJ, 870, 129, doi: 10.3847/1538-4357/aaf390","cited_arxiv_id":null,"evidence_quote":"Is the prior inner-disk chemistry model whose long-chain carbon reservoirs are compared with the present long-chain sinks."},{"cited_title":"Hydrocarbon chemistry in inner regions of planet forming disks","cited_arxiv_id":"2310.04505","evidence_quote":"Is cited for the three-body and PAH-forming reactions that the paper flags as a possible alternative pathway absent from its network."}],"review_version":1}