{"id":"e1f73540-535a-43c5-a197-f9661e455972","arxiv_id":"2505.13705","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"A merged exoplanet-disk chemical network shows that thermodynamic equilibrium is not established in planet-forming disks, with the closest approach in a small shielded region behind the inner rim.","lead":"Planet-forming disks are almost nowhere in thermodynamic equilibrium, except a tiny warm dense pocket behind the inner rim, according to a new chemical network that merges exoplanet and disk chemistry. The network, ChaiTea, changes predicted inner-disk abundances of C2H2, HCN, and CO2 that JWST can observe.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"N2 network incompleteness may inflate the slowest relaxation mode; if missing neutral-neutral routes shorten the 1200 K timescale, the 'no TE anywhere' result weakens.","rationale":"The central claim is negative and universal: no disk region reaches thermodynamic equilibrium. The only identified reason the warm inner disk fails is the long N2 formation timescale, and the paper explicitly flags that the network may miss neutral-neutral N2 routes (Sect. 4.3). This is a load-bearing internal limitation, not a mere caveat, because a faster N2 channel would shorten the relaxation timescale and could put the inner disk into TE over a disk lifetime. The grid_0 agreement with GGchem is reassuring but does not resolve the issue: time-independent solves find the equilibrium composition, not the rate of approach, so a missing fast pathway is invisible in that benchmark. The authors deserve credit for the detailed balance verification and for expressly stating the N2 problem, but the strongest claim is conditional on network completeness. The proposed test adds the missing neutral-neutral channels and re-measures the relaxation timescale; this directly determines whether the slow mode is physical or a network artifact. The reader's weakest_assumption is the same, so the conditional verdict stands unchanged pending this check.","tokens_in":33306,"tokens_out":5939,"duration_ms":61759,"concrete_test":"Repeat the time-dependent relaxation run shown in Fig. 2 at T=1200 K and nH=7.56e13 cm-3, augmenting ChaiTea with every exothermic neutral-neutral reaction from KIDA and NIST that forms or breaks an N≡N bond (at least N+NH2→N2+H2, N+NH→N2+H, NH+NO→N2+OH, N+N2H→N2+NH, NCN+H→N2+CH), plus one benchmark grain-surface N2 formation rate, keeping all other settings unchanged; then re-measure tau_chem,3 and tau_chem,2. If either drops below ~1e7 yr, the 'TE cannot be established anywhere' conclusion must be revised for the inner disk.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The 'no thermodynamic equilibrium anywhere' claim hinges on the slowest chemical mode being gas-phase N≡N bond formation (Sect. 4.3, Table 7). At 1200 K, nH=7.56e13 cm-3, the visual relaxation time is ~1e8 yr and the Jacobian mode is ~4e7 yr, both above any plausible disk age. But the authors concede in Sect. 4.3 that 'we may miss some other neutral-neutral reaction pathways to form or destroy N≡N bonds'; adding N2H, NCN, HNCO changes the timescale by only a factor of 2-3. The grid_0 benchmark does not test this because it solves time-independently: a missing fast pathway leaves the steady-state equilibrium composition unchanged but can shorten the relaxation timescale by orders of magnitude. If an unlisted route (e.g., N+NH2, NH+NO, N+N2H, or grain-surface formation) is fast at 1200 K, the warm inner disk could relax to TE on disk-evolutionary timescales, directly contradicting the abstract. The 2D sigma maps inherit the same network, so they cannot rule this out.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops a new chemical network, ChaiTea, for planet-forming disk models by merging the STAND exoplanet network, UMIST 2022, and the DIANA disk network, and by adding reverse reactions derived from Gibbs free energies and termolecular reactions via a Lindemann-Hinshelwood treatment. The authors verify that, in the absence of photorates, cosmic rays, and X-rays, the network relaxes to thermodynamic equilibrium by comparing single-point ProDiMo results with the equilibrium code GGchem. They then quantify deviations from equilibrium induced by a Planck radiation field and by cosmic-ray ionization, and compute chemical relaxation timescales. Finally, they compare 2D disk abundance maps from the standard DIANA network, the DIANA network with reverse reactions, and the full ChaiTea network. The headline result is that thermodynamic equilibrium cannot be established anywhere in a typical T Tauri disk, although a small, warm, high-density region behind the inner rim, shielded from cosmic rays, approaches equilibrium; a secondary result is that the slowest relaxation mode is N2 formation, with timescales as long as ~1e8 yr at 1200 K.","tokens_in":33588,"tokens_out":6100,"duration_ms":59626,"significance":"If the results hold, this is a timely and useful contribution. It brings pressure-dependent and reverse reactions, standard in exoplanet chemistry, into disk modeling, and it directly addresses a question of broad interest: when can equilibrium chemistry be used for inner disk observations with JWST and ALMA. The paper is strong on transparency: it defines a quantitative deviation metric σ, computes detailed balance tests (Table 6), reports multiple relaxation-timescale definitions (Table 7), and explicitly identifies spurious endothermic reactions in existing networks. The authors also acknowledge the main weakness of their own analysis, namely the possible incompleteness of the network for N≡N bond formation. For these reasons the work is a valuable reference point even if its strongest conclusion needs qualification.","major_comments":[{"comment":"The central claim that thermodynamic equilibrium is never established anywhere in the disk rests on the identification of N2 formation as the slowest chemical mode, with relaxation timescales of ~1e8 yr at 1200 K and >1e10 yr below ~1100 K. However, the authors themselves state in §4.3 that 'we may miss some other neutral-neutral reaction pathways to form or destroy N≡N bonds.' The grid_0 benchmark in §4.1 cannot detect such a missing pathway because it solves the time-independent chemical equilibrium problem: a missing fast reaction leaves the converged equilibrium composition unchanged but can shorten the relaxation timescale by orders of magnitude. If an unlisted route (e.g., N+NH2, NH+NO, or grain-surface N2 formation) is fast at 1200 K, the warm inner-disk region behind the inner rim could relax to thermodynamic equilibrium on disk-evolutionary timescales, weakening the abstract's 'cannot be established anywhere' statement. The authors should either conduct a targeted literature search for N-N bond forming/destroying reactions and demonstrate that they do not change the conclusions, or consistently qualify all headline statements as 'within the current ChaiTea network.'","section":"§4.3, Table 7, Fig. 2"},{"comment":"There is a direct numerical inconsistency in the reported deviation in the warm intermittent molecular layer. The Abstract states 'In the warm intermittent molecular layer, which is observable, σ≥10,' whereas §4.7 states 'there is ... the warm molecular intermittent surface layer where σ again reaches values as low as 5,' and the bullet list in §6 states 'There is a warm intermittent molecular layer where σ∼5.' Since this quantity is a headline result about the observability of disequilibrium, the authors must decide on the correct value and use it consistently.","section":"Abstract vs §4.7 and §6"},{"comment":"The verification against GGchem is not fully independent: §4.1 states that GGchem was run 'with a new option to use the same Burcat thermochemical data as used in this work.' Because both codes share the same Gibbs free energy data, the <1% agreement in grid_0 is primarily a detailed-balance self-consistency check of the kinetic network, not a validation of the thermochemical data themselves. The authors do compare formation enthalpies with NIST and ATcT for some species, which partially mitigates this, but the manuscript should explicitly frame the grid_0 benchmark as an internal consistency check and should justify that the shared thermochemical data are not the source of the reported disk conclusions.","section":"§4.1, §2.3"},{"comment":"The three relaxation timescale definitions in Table 7 disagree by three to four orders of magnitude at 1200 K (τchem,1 = 1.68×10^4 yr, τchem,2 = 4.03×10^7 yr, visual τchem,3 ~1e8 yr). The paper explains that τchem,1 is misleading because it tracks individual species rather than collective modes, but it then admits that even the Jacobian-based τchem,2 is 'not fully satisfactory' due to numerical ill-conditioning of the Jacobian. Given that these timescales are load-bearing for the conclusion that TE cannot be established, the manuscript should report the condition numbers or another convergence diagnostic for the eigen-decomposition and should present the resulting uncertainty on the derived temperature below which relaxation exceeds the disk age.","section":"§4.2, Table 7"}],"minor_comments":[{"comment":"The word 'studed' is a typo and should read 'studied.'","section":"§4.1"},{"comment":"The caption gives the density as 7.65×10^13 cm^-3, while the text and Table 7 use 7.56×10^13 cm^-3; the values should be harmonized.","section":"Fig. 2 caption"},{"comment":"The text says the grid_0 deviations are '<1%' but the color scale in Fig. 3 spans 10^-4 to 10^2; a zoomed panel or an explicit color-bar range for the converged models would make the claim easier to verify.","section":"§4.4, Fig. 3"},{"comment":"The sentence 'the CRIs scale with n⟨H⟩' is imprecise; the ionization rate per unit volume scales linearly with n⟨H⟩, while the underlying rate coefficient is independent of density.","section":"§4.6"},{"comment":"The reaction count summary '5539 (2462 +375+1155+1547)' is mathematically correct but could be clearer if the subtotals (2837+1155+1547 = 5539) were stated explicitly in the text.","section":"§3.2"},{"comment":"The sentence after Eq. (9) says the auto-cleaning multiplies both Kf and Kr by a common factor when ΔRG°/R > γf; it would help to state explicitly that this preserves the ratio Kf/Kr and therefore detailed balance at equilibrium, since that property is central to the paper's verification.","section":"§3.3, Eq. (9)"},{"comment":"The erroneous reaction is said to have a reaction enthalpy of '+6.5 eV' while the correct reaction is '+0.43 eV'; the second value lacks explicit units and should read '+0.43 eV'.","section":"§5.1, Eq. (22)-(23)"},{"comment":"The radiative association H + e− → H− + hν is listed as a source of deviation from detailed balance, but the sentence would be clearer if it stated that this reaction is treated as non-invertible, which is why it breaks detailed balance.","section":"§4.5, Eq. (20)"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is within the scope of A&A and addresses a question with strong current interest. The main issue is not the quality of the implementation but the strength of the headline claim: the authors themselves identify the N2 formation network as possibly incomplete, and that incompleteness is the hinge for the 'no thermodynamic equilibrium anywhere' conclusion. A major revision that either secures this point with a dedicated analysis of N-N bond formation routes or consistently limits the claim to the current network would make the paper suitable for publication. The abstract-vs-section inconsistency on σ in the warm layer should also be fixed before acceptance; this is a minor textual fix but it is a visible contradiction."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Jayatee et al. have built something the field can actually use: ChaiTea, a 239-species network that folds STAND's pressure-dependent termolecular reactions and Gibbs-derived reverse rates into the UMIST 2022 and DIANA disk chemistry, implemented in ProDiMo. The benchmark against GGchem is clean — <1% deviations down to 600 K when photorates and cosmic rays are off — and the auto-detection of 45–68 spurious endothermic reactions in standard disk networks is a tangible service to the community. The 2D sigma maps are a nice way to see where TE fails, and the changes they find for C2H2, CO2, and HCN in the warm inner disk plausibly matter for JWST interpretation. This deserves serious referee time.\n\nThe soft spots are just as real. First, the grid_0 agreement is partly built in: both ProDiMo and GGchem use the same Burcat thermodynamic polynomials, so the check demonstrates detailed-balance self-consistency, not independent validation. Second, the headline that \"thermodynamic equilibrium cannot be established anywhere in the disk\" rests on the slowest chemical mode being gas-phase N≡N bond formation. At 1200 K the relaxation time is ~10^8 yr by their own visual estimate, but they concede in Sect. 4.3 that missing neutral-neutral N2 formation routes could shorten this substantially — they tried N2H, NCN, HNCO and got only a factor 2–3, but that does not rule out faster pathways. Because the benchmark is time-independent, it cannot catch this; a missing fast pathway changes the steady-state composition not at all but can crash the relaxation timescale down to disk ages. If such routes exist, the warm inner rim region could relax toward TE, directly softening the abstract's claim. The paper itself is honest about this, which I credit, but the claim in the abstract goes beyond what the network can currently establish.\n\nThird, the full network and data are not released with the paper, which is a real problem for a paper whose main deliverable is a network. The species tables and reaction lists are in the text, but not machine-readable.\n\nMy advice: send it to review, but tell the authors to soften the \"anywhere\" claim to \"in the regions sampled by this network,\" release the network and input files, and add a dedicated test of N2 relaxation with an independent route, even a toy one. The paper is useful and the thinking is clear; it just overstates the certainty of its central negative result.","headline":"Useful new network and analysis, but the 'no TE anywhere' headline leans on the slowest N2 formation routes being complete, which the authors themselves doubt.","tokens_in":34042,"tokens_out":2438,"would_cite":true,"duration_ms":24027,"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":"No region of a typical planet-forming disk reaches thermodynamic equilibrium; only a tiny, warm, dense pocket shielded from cosmic rays behind the inner rim comes close.","keywords":["thermodynamic equilibrium","protoplanetary disks","chemical networks","ChaiTea","inner disk chemistry","detailed balance","cosmic rays","JWST"],"falsifier":"Measure or compute the rate of the slowest N2-forming neutral-neutral reactions near 1200 K; a fast channel that shortens the $\\sim 10^8$-year relaxation timescale would weaken the central 'no equilibrium' claim. Observationally, JWST mid-infrared spectra of a disk whose inner rim is directly viewed should show the ChaiTea prediction of reduced C2H2, reduced HCN, and increased CO2 inside 1 au; a spectrum matching the old standard network would refute the predicted kinetic structure.","tokens_in":33164,"feed_emoji":"🌌","tokens_out":9828,"duration_ms":81306,"temperature":0.7,"pith_summary":"The paper asks whether thermodynamic equilibrium can ever be established in planet-forming disks, and its answer is no: in a typical T Tauri disk model no region attains equilibrium, with one near-exception in a tiny, warm, high-density pocket just behind the inner rim where cosmic rays are shielded. To make this test the authors build a new chemical network, ChaiTea, that merges disk chemistry with planetary-atmosphere chemistry and enforces detailed balance by pairing every gas-phase reaction with a reverse rate derived from Gibbs free energies. They verify that the network relaxes to equilibrium when radiation is removed, then show how photochemistry and cosmic rays push abundances away from it. This matters because the same inner-disk molecules probed by JWST, such as C2H2, HCN, CH4, CO2 and N2H+, change by orders of magnitude when the new network replaces the standard disk network.","feed_headline":"No disk region reaches thermodynamic equilibrium except one shielded pocket","feed_subtitle":"A new network coupling disk and planet chemistry finds JWST-visible gas deviates by 10-30 orders of magnitude.","key_machinery":"The load-bearing object is the ChaiTea chemical network, a merged 239-species network that combines interstellar and disk reaction databases with the STAND planetary network. Its defining mechanism is reaction reversal: every invertible gas-phase reaction is paired with its reverse, and the reverse rate coefficient is computed from the forward rate and the reaction's Gibbs free energy, enforcing detailed balance at equilibrium. This is supplemented by a Lindemann-Hinshelwood treatment of termolecular reactions that captures pressure-dependent stabilisation in the dense inner disk. The network is embedded in a thermochemical disk model, benchmarked against an independent equilibrium chemistry code, and diagnosed with two tools: a deviation measure $\\sigma$ that averages logarithmic abundance differences between kinetic and equilibrium solutions, and eigenvalue-based chemical relaxation timescales obtained from the Jacobian of the rate equations.","core_discovery":"The central claim is that thermodynamic equilibrium cannot be established anywhere in a planet-forming disk. In the full 2D disk model, kinetic chemistry and equilibrium abundances never fall below about one order of magnitude apart everywhere ($\\sigma \\gtrsim 1$), and they differ by tens of orders of magnitude in the observable warm molecular layer. Only a small, warm, dense region in the midplane directly behind the inner rim ($r \\lesssim 0.1$ au), shielded from cosmic rays, approaches equilibrium. At a density of $7.6\\times 10^{13}$ cm$^{-3}$, the chemical relaxation timescale rises from about a month at 2000 K to roughly $10^8$ years at 1200 K, and exceeds the age of the universe below about 1100 K, with the slowest mode being the formation of N2 from other nitrogen carriers. Because kinetic equilibrium distributes carbon into CO and CO2, nitrogen into N2, and oxygen into H2O and CO, whereas thermodynamic equilibrium favours CH4, NH3 and H2O, switching to the new network changes the inner-disk abundances of JWST-visible molecules.","pith_inferences":["If faster neutral-neutral routes to N2 exist than the network currently contains, the $10^8$-year relaxation at 1200 K would shorten, and the warm inner-disk region approaching equilibrium could be larger than the paper's fiducial answer allows.","The same Gibbs-free-energy reversal and $\\sigma$ diagnostic could be carried back to exoplanet atmosphere chemistry to map, in a common metric, where equilibrium assumptions for hot Jupiter spectra break down.","Because the auto-cleaning step found dozens of spurious endothermic reactions in an established disk network, applying the same audit to other astrochemical networks could change predicted outer-disk abundances of H, N2H+, and related tracers.","Running the same 2D disk model across a range of carbon-to-oxygen ratios would test whether the size and location of the tiny equilibrium pocket track the thermochemical shifts, giving observers a concrete target to confirm or refute the prediction."],"forward_implications":["Thermodynamic equilibrium cannot be assumed when modelling or interpreting inner-disk chemistry; kinetic chemistry with pressure-dependent and reverse reactions is required.","JWST-visible abundances change by orders of magnitude with the new network: C2H2 drops in the inner midplane, CO2 increases within 1 au, and N2H+ and CH+ become more abundant in the warm molecular layer and outer disk.","Spurious endothermic reactions with zero or tiny activation energies are automatically corrected by Gibbs-free-energy pairing; the paper identifies 45 such reactions in the large DIANA-standard network and 68 once STAND reactions are added, eliminating artefacts such as excess atomic hydrogen in the outer midplane.","Below roughly 1100 K, chemical relaxation toward equilibrium takes longer than the age of the universe, so cold disk regions cannot be interpreted as equilibrium mixtures even if cosmic rays and photochemistry were absent."],"supporting_citations":[{"why":"Supplies the STAND planetary network with termolecular reactions and Gibbs-free-energy-based reverse rates, the core concept adopted here.","marker":"Rimmer & Helling 2016"},{"why":"Provides the large DIANA disk network and baseline species selection that ChaiTea extends.","marker":"Kamp et al. 2017"},{"why":"Provides the UMIST 2022 database, the primary source of gas-phase kinetic reaction data.","marker":"Millar et al. 2024"},{"why":"Supplies the GGchem equilibrium code used to benchmark the ChaiTea network against thermodynamic equilibrium abundances.","marker":"Woitke et al. 2018"},{"why":"Supplies the NASA seven-term polynomial thermochemical data used to compute Gibbs free energies for reaction reversal.","marker":"Burcat & Ruscic 2005"},{"why":"Establishes the thermochemical disk model framework within which the network is implemented and 2D disk structures are computed.","marker":"Woitke et al. 2009"},{"why":"Identifies spurious endothermic reactions with zero activation energies, motivating the auto-cleaning applied to the merged network.","marker":"Tinacci et al. 2023"},{"why":"Provides the previous extended hydrocarbon network results used to compare inner-disk C2H2 abundances and past ProDiMo chemistry.","marker":"Kanwar et al. 2024b"}],"fun_headline_variants":["Only a tiny shielded pocket near the star reaches equilibrium","Disk never reaches chemical equilibrium except a warm dense spot","Equilibrium impossible in disks except near the inner rim","Chemical relaxation in disk can exceed universe's age","Disk equilibrium timescale exceeds universe age below 1100 K"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The conclusion that thermodynamic equilibrium is nowhere established rests on the completeness of the ChaiTea network's slowest relaxation pathways, especially the neutral-neutral reactions that form N2 from other nitrogen-bearing molecules; if faster routes exist, more of the warm inner disk could relax toward equilibrium.","fun_headline_variants_meta":{"raw":{"variants":["Only a tiny shielded pocket near the star reaches equilibrium","Disk never reaches chemical equilibrium except a warm dense spot","Equilibrium impossible in disks except near the inner rim","Chemical relaxation in disk can exceed universe's age","Disk equilibrium timescale exceeds universe age below 1100 K"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000901,"raw_usage":{"total_tokens":3965,"prompt_tokens":1120,"completion_tokens":2845,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":736,"completion_tokens_details":{"reasoning_tokens":2768}},"tokens_in":736,"tokens_out":2845,"duration_ms":18827,"temperature":1.0,"reasoning_tokens":2768,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T20:12:07.330640+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure or compute the rate of the slowest N2-forming neutral-neutral reactions near 1200 K; a fast channel that shortens the $\\sim 10^8$-year relaxation timescale would weaken the central 'no equilibrium' claim. Observationally, JWST mid-infrared spectra of a disk whose inner rim is directly viewed should show the ChaiTea prediction of reduced C2H2, reduced HCN, and increased CO2 inside 1 au; a spectrum matching the old standard network would refute the predicted kinetic structure.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the STAND planetary network with termolecular reactions and Gibbs-free-energy-based reverse rates, the core concept adopted here."},{"cited_title":"& Ruscic, B","cited_arxiv_id":null,"evidence_quote":"Supplies the NASA seven-term polynomial thermochemical data used to compute Gibbs free energies for reaction reversal."},{"cited_title":"2023, , 266, 38","cited_arxiv_id":null,"evidence_quote":"Identifies spurious endothermic reactions with zero activation energies, motivating the auto-cleaning applied to the merged network."}],"review_version":1}