{"id":"39fa1d98-8b0e-43e0-94cd-12cd0094c492","arxiv_id":"2412.06397","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Adding cosmic-ray radiolysis and sputtering to a multi-phase Monte Carlo model reproduces observed TMC-1 abundances of five oxygen-bearing complex organic molecules within a factor of 3.","lead":"Astrochemists modeled the cold cloud TMC-1 with cosmic-ray driven reactions inside ice mantles and sputtering that ejects molecules into gas. Their models match observed abundances of five complex organic molecules within a factor of 3, supporting a non-thermal route to these molecules at 10 K.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Suprathermal reaction set (Table 10) is unvalidated and drives the claimed factor-of-3 COM agreement; Table 7 itself shows MC3 C2H5OH and MC4 CH3CHO outside factor 3.","rationale":"The reader's weakest assumption correctly identifies the unvalidated suprathermal chemistry as a key risk. My stress-test agrees but sharpens the concern: the agreement is not merely uncertain, it is driven almost entirely by a speculative network of 343 reactions with no barriers, no branching ratios, and no validation. In addition, the paper's own Table 7 shows that two of the four models (MC3 C2H5OH and MC4 CH3CHO) fall outside the stated factor of 3 at the local best-fitting time, so the headline quantitative claim is internally inconsistent. These issues do not necessarily invalidate the entire modeling approach, which may still be a useful proof of concept, but they do mean the paper should not be accepted without correction of the factor-of-3 claim and a sensitivity analysis of the assumed suprathermal channels. The reader's CONDITIONAL verdict already captures the need for revisions, so I leave the verdict unchanged.","tokens_in":47258,"tokens_out":7488,"duration_ms":68332,"concrete_test":"Re-run MC2 with the suprathermal reaction set restricted to experimentally characterized channels only (e.g., Shingledecker & Herbst 2018; Paulive et al. 2021) and with radiolysis attenuated by ice depth (e.g., limited to the top ~10 monolayers). If the abundances of CH3OH, HCOOCH3, CH3OCH3, CH3CHO, C2H5OH at t=2.93e5 yr change by more than an order of magnitude, the claimed agreement is contingent on the unvalidated 343-channel network. Also recompute the Table 7 ratios for MC3 C2H5OH and MC4 CH3CHO to verify the factor-of-3 statement.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that MC1-MC4 reproduce TMC-1 O-bearing COMs within a factor of 3 rests on the 343 suprathermal reactions in Table 10. The text states these reactions proceed with 'equal probability' based only on neighbor abundance, with no activation barriers, branching ratios, or experimental/theoretical validation (Section 2.3; Appendix A). More than 99% of JHCOOCH3 and the dominant channels for CH3OCH3 and C2H5OH go through these assumed channels. The radiolysis rates themselves are taken from Shingledecker et al. (2018) and applied uniformly through the ice mantle with no depth attenuation, an assumption the authors concede is made 'due to limited knowledge.' If these channels are absent, barrier-suppressed, or have different products, the agreement is not robust. Independently, the factor-of-3 claim is internally contradicted: at the local best-fitting times in Table 7, MC3 C2H5OH is 4.46e-10 vs observed 1.1e-10 (factor 4.05) and MC4 CH3CHO is 9.4e-11 vs observed 3.5e-10 (factor 3.72).","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper presents a 0D multiphase Monte Carlo gas-grain astrochemical model of TMC-1 that adds cosmic-ray-induced radiolysis and sputtering desorption to the authors' earlier Lu et al. (2018) model. Six models (MC1-MC6) are run, differing in the assumed sputtering ice composition (mixed, CO, CO2, H2O) and in whether radiolysis and sputtering are included; a reactive-desorption variant of MC2 is also studied. The manuscript's central claims are (i) that models MC1-MC4 reproduce the observed gas-phase abundances of CH3OH, HCOOCH3, CH3OCH3, CH3CHO, and C2H5OH toward TMC-1 within a factor of 3 at a local best-fitting time, and (ii) that at a global best-fitting time 63 of 94 compared species (67.02%) agree within an order of magnitude for MC2/MC3. The paper also analyzes carbon-chain species and the effect of varying C/O ratios.","tokens_in":47519,"tokens_out":6531,"duration_ms":60921,"significance":"The paper is potentially significant because it applies a genuinely stochastic multiphase treatment to non-thermal COM formation, using experimental sputtering fits and an explicitly expanded suprathermal reaction network. The model description is unusually complete: Tables 3, 5, 7, 8, and 10 document the radiolysis reactions, sputtering parameters, abundance comparisons, and the 343 suprathermal channels; the authors also include a reactive-desorption sensitivity test and comparisons with seven previous models. These strengths make the manuscript a useful reference point for future work on cold-core COM chemistry. However, the central factor-of-3 claim is contradicted by the authors' own Table 7 for two model-species combinations, and the 343 suprathermal reactions that dominate COM production are assumed without experimental or theoretical validation. As it stands, the paper demonstrates that a specific assumed network can be tuned to match TMC-1, but it does not yet establish that the proposed mechanisms are responsible for the observed COM abundances.","major_comments":[{"comment":"The statement in §3.3 that 'the simulation results for COMs in our models, MC1 to MC4, all deviated from the observed values by less than a factor of 3' is not supported by Table 7. At the local best-fitting time, MC3 gives C2H5OH = 4.46e-10 versus observed (1.1 ± 0.3)e-10, a ratio of about 4.1 even using the upper 1σ bound of 1.4e-10, and MC4 gives CH3CHO = 9.4e-11 versus observed (3.5 ± 0.2)e-10, a ratio of about 3.7 even against the lower 1σ bound of 3.3e-10. The factor-of-3 claim must be revised, or the definition of 'agreement' and the treatment of observational errors must be stated precisely.","section":"§3.3, Table 7"},{"comment":"The dominant formation channels for the claimed COM detections are inserted as assumptions. The text states in §3.1.2 that about 99% of JHCOOCH3 is formed through radiative mechanisms, and Table 4 lists the product-forming reactions (e.g., JCH3O* + JHCO → HCOOCH3); §3.1.5 and §3.1.6 show that the main channels for CH3OCH3 and C2H5OH are likewise suprathermal reactions with specified COM products. The 343 reactions in Table 10 are taken to be barrierless and to proceed with equal probability based on neighbor abundance, with no branching ratios, experimental measurements, or theoretical calculations cited, and §2.3 concedes that radiolysis is assumed uniform in all ice layers 'due to limited knowledge.' Since the product COMs are inputs to the network rather than emergent consequences of the model, the reported factor-of-3 agreement is a test of the assumed channels' rates, not a validation of the mechanism. The authors should either supply external constraints for the suprathermal reactions or explicitly reframe the results as an exploration conditional on the network.","section":"§2.3 and Appendix A, Table 10"},{"comment":"The stochastic Monte Carlo results are presented without error bars or multiple realizations, yet the comparisons use abundance values at single times. Table 8 contains numerous 0.00(0) entries and the text describes 'notably large' abundance fluctuations in MC1 and MC4 (§3.1.3), so run-to-run scatter could plausibly affect the factor-of-3 and 67% statistics. In addition, the global best-fitting time is defined as the time that maximizes the number of matches, so the 67.02% figure is a best-case, in-sample statistic rather than a predictive accuracy measure. At minimum, the authors should report the stochastic uncertainty (e.g., multiple seeds or a Poisson error estimate) and state how the time-selection procedure affects the reported agreement.","section":"§2.2, §3.3, Table 8"},{"comment":"The local best-fitting time is selected by minimizing D(t) computed from the very five COMs that are then used to claim factor-of-3 agreement. This in-sample optimization inflates the apparent agreement; a fixed epoch (e.g., 1e6 years) or an out-of-sample comparison would be a more meaningful test. The authors should provide the abundance ratios at a common time or quantify the sensitivity of the conclusion to the chosen epoch.","section":"§3.3, Eq. (15)"}],"minor_comments":[{"comment":"The title contains a typo ('F ormation'), and Section 4 uses 'oxygen-obeying COMs' instead of 'oxygen-bearing COMs'.","section":"Title and §4"},{"comment":"The formula for the layer distribution probability is garbled as 'Pji = SU M Kji/P jSU M Kji'; please reformat the equation and define the summation index.","section":"§2.3"},{"comment":"The caption says 'Model 1 and Model 6' but the figure compares MC2 and MC6; make the model labels consistent.","section":"Figure 7 caption"},{"comment":"There are notation inconsistencies, e.g., the CH2NH upper limit is given as '<3.6(-0.9)' and several entries use '0.00(0)' that is not defined; also Table 9 lists Carder et al. twice (rows 25 and 28) with the same reference.","section":"Table 8 and Table 9"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is within the journal's scope and the authors are transparent about the network and mechanisms. However, the overstatement in §3.3 relative to Table 7 and the conditional nature of the suprathermal network need to be addressed before publication. The referee report highlights the contradiction in Table 7 and the need for uncertainty quantification."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThis is a serious modeling paper, not a crackpot claim. The genuinely new part is implementing cosmic-ray radiolysis and sputtering in the authors' layer-resolved Monte Carlo multiphase code, with a 343-reaction suprathermal network, and applying it to TMC-1. That is a real step beyond the rate-equation treatments of Wakelam et al. and Paulive et al., and the paper is readable and complete: physical parameters, reaction tables, and model variants are all spelled out in the text and appendix.\n\nIt also does one thing well that often gets skipped: it checks the impact of the new mechanisms on carbon-chain species. The finding that the carbon-chain families are not wrecked by the new chemistry, and in fact are boosted in the late stages, is useful and plausible.\n\nThe soft spots are real and go to the headline. The factor-of-3 agreement is not actually met in Table 7 at the local best-fitting times: MC3's C2H5OH is 4.05 times the observed value, and MC4's CH3CHO is 3.72 times. That is a direct contradiction of the abstract's claim. The 67% figure is computed at a time chosen to maximize the number of within-order-of-magnitude matches, so it is a fit, not a prediction. And the larger issue is that the suprathermal reactions in Table 10, which produce essentially all of the COMs, are assumed with equal probability and no barriers, with no experimental or theoretical validation. The authors are honest about this and about ignoring depth attenuation, but that does not change the fact that the COM agreement is a test of abundances given an assumed chemistry, not a validation of the mechanism. The Monte Carlo runs also have no error bars, so we do not know the run-to-run scatter.\n\nNone of this makes the paper worthless. The mechanism is plausible, previous models did underproduce HCOOCH3 by orders of magnitude, and this model shows what a layer-resolved treatment can do. But the claims need to be softened and the sensitivity to the assumed channels needs to be examined before the factor-of-3 language is justified. I would send this to a serious astrochemistry referee. The referee should ask for error bars, a branching-ratio sensitivity study, and a correction of the internal factor-of-3 discrepancy. Worth one reading group session to discuss how much circularity is acceptable in grain-surface models.","headline":"A plausible and well-documented port of radiolysis plus sputtering into a Monte Carlo multiphase code, but the factor-of-3 claim is contradicted by the paper's own Table 7.","tokens_in":48094,"tokens_out":2681,"would_cite":false,"duration_ms":28573,"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":"Cosmic-ray radiolysis and sputtering together can explain the five oxygen-bearing complex organic molecules observed in TMC-1 to within a factor of 3.","keywords":["astrochemistry","complex organic molecules","cold dark clouds","cosmic-ray radiolysis","suprathermal reactions","sputtering desorption","grain-surface chemistry","TMC-1"],"falsifier":"Irradiate mixed $\\mathrm{H_2O}$-CO-$\\mathrm{CO_2}$-$\\mathrm{CH_3OH}$ ices with cosmic-ray-like particles at 10 K and measure whether excited methoxy plus formyl actually produces methyl formate with appreciable branching; alternatively, rerun the same model with depth-dependent radiolysis attenuation and only experimentally confirmed suprathermal channels and check whether the factor-of-3 agreement with TMC-1 survives.","tokens_in":47004,"feed_emoji":"🧊","tokens_out":11966,"duration_ms":109411,"temperature":0.7,"pith_summary":"The paper claims that the oxygen-bearing complex organic molecules detected in cold dark clouds such as TMC-1 can be produced in place by two non-thermal processes acting together: cosmic-ray-induced radiolysis, which excites molecules locked inside dust ice mantles so they react with adjacent molecules without diffusing, and cosmic-ray sputtering, which lifts the products from the ice into the gas phase. At 10 K, ordinary diffusive grain-surface chemistry is too slow, and gas-phase reactions alone underpredict these species, so the mechanism matters if models are to explain the observations. The authors add 46 radiolysis reactions, 343 suprathermal (excited-state) reaction channels, and 197 sputtering channels to a multiphase Monte Carlo gas-grain model. In four model variants with different ice-composition sputtering rates, the predicted abundances of $\\mathrm{CH_3OH}$, $\\mathrm{HCOOCH_3}$, $\\mathrm{CH_3OCH_3}$, $\\mathrm{CH_3CHO}$, and $\\mathrm{C_2H_5OH}$ agree with TMC-1 observations within a factor of 3 at the best-fitting time, and 63 of 94 compared species agree within one order of magnitude.","feed_headline":"Cosmic-ray ice chemistry hits cold-core molecules within factor of 3","feed_subtitle":"Adding sputtering and radiolysis to ice models brings five key molecules in line with TMC-1 observations.","key_machinery":"The load-bearing objects are the suprathermal reaction network and the sputtering rate law. In the network, 25 ice species can be excited by cosmic rays; an excited species reacts immediately with one neighbouring molecule selected by abundance-weighted probability, with 343 product channels (for example, $\\mathrm{CH_3O^{*}} + \\mathrm{HCO} \\rightarrow \\mathrm{HCOOCH_3}$). The multiphase model resolves the ice into a diffusive active surface layer, immobile normal sites, and mobile interstitial sites, and it is simulated with an accelerated Gillespie stochastic algorithm. Sputtering enters through $k_{scr} = (\\zeta / 3 \\times 10^{-17})\\, Y_{eff}\\, \\pi r_d^2 / N_s$ with $Y_{eff} = \\alpha(1-e^{-(n_{layers}/\\beta)^{\\gamma}})$, fitted to experimental yields for CO, $\\mathrm{CO_2}$, and $\\mathrm{H_2O}$ ices; the mixed-ice rate is composition-weighted in real time.","core_discovery":"The central claim is that cosmic-ray radiolysis paired with sputtering closes the cold-core COM problem. When a cosmic ray strikes an ice molecule, the molecule enters a short-lived electronically excited state and reacts immediately with a neighbouring species instead of waiting for diffusion, which makes the reaction possible at 10 K. The resulting COMs accumulate in the bulk ice, where radicals are protected from rapid hydrogenation, and later sputtering desorbs them into the gas phase. The paper reports factor-of-3 agreement for the five key COMs ($\\mathrm{CH_3OH}$, $\\mathrm{HCOOCH_3}$, $\\mathrm{CH_3OCH_3}$, $\\mathrm{CH_3CHO}$, $\\mathrm{C_2H_5OH}$), and shows that without sputtering the molecules stay trapped in the ice, while without both mechanisms most COMs are not formed at all.","pith_inferences":["Beyond the paper, the same radiolysis-plus-sputtering mechanism should be testable in other cold cores such as L1544 and L1689B, where the same COMs are observed; the model's age at best fit (about 2-7 x 10^5 years) gives a concrete prediction for how COM abundances should track core age.","Beyond the paper, if the 343 suprathermal channels are later found to have different branching ratios, the factor-of-3 claim would likely shift by orders of magnitude, so the network presently functions as a parameterization of unknown ice chemistry rather than a measured mechanism.","Beyond the paper, the model's insensitivity to C/O suggests that in cold cores the ice mantle, not the initial gas composition, controls the carbon budget available to COM formation; direct ice observations toward TMC-1 could check whether mantle composition tracks the assumed model layers."],"forward_implications":["Without sputtering, COMs formed inside the ice mantle cannot reach the gas phase; adding sputtering makes it the dominant desorption channel after about $10^5$ years, accounting for 58-89% of desorption depending on the model.","Methyl formate and dimethyl ether are built almost entirely inside the bulk ice by suprathermal reactions, while acetaldehyde is built mostly in the gas phase from sputtered precursors such as $\\mathrm{C_2H_5}$ and $\\mathrm{C_2H_5OH}$.","The new mechanisms widen the epoch over which carbon-chain molecules can form: after carbon is locked into CO, sputtering of $\\mathrm{C_3}$ and other precursors from ice keeps hydrocarbons and cyanopolyynes abundant at late times.","Changing the elemental C/O ratio has little effect on the fraction of reproduced species in the new models, unlike earlier rate-equation models where C/O changed the fit quality.","A test model with 1% reactive desorption added changes mainly methanol, because the other COMs form below the surface where reactive desorption does not act."],"supporting_citations":[{"why":"Provides the base multiphase gas-grain model and reaction network that the new radiolysis and sputtering channels are added to.","marker":"Lu et al. (2018)"},{"why":"Supplies the accelerated Gillespie algorithm used to simulate the stochastic gas-grain network in feasible wall-clock time.","marker":"Chang et al. (2017)"},{"why":"Supplies the radiolysis rate coefficients and decomposition pathways for the excited species used in Table 3.","marker":"Shingledecker et al. (2018)"},{"why":"Provides the experimental sputtering-yield fit parameters (alpha, beta, gamma) that set the cosmic-ray sputtering rates in MC2-MC4.","marker":"Dartois et al. (2021)"},{"why":"Establishes the previous rate-equation baseline where HCOOCH3 is underproduced and raising sputtering overproduces CH3OCH3, the discrepancy this work targets.","marker":"Paulive et al. (2022)"},{"why":"First applied experimental sputtering rates to cold-core models; its low gas-phase COM results are the comparison baseline.","marker":"Wakelam et al. (2021)"},{"why":"Tested non-diffusive mechanisms with chemical desorption and still underestimated CH3OCH3 by two orders of magnitude, the gap this paper addresses.","marker":"Jin & Garrod (2020)"},{"why":"Provides the observed TMC-1 abundances of CH3OH and CH3CHO used in Table 7.","marker":"Cernicharo et al. (2020b)"},{"why":"Provides the observed TMC-1 abundances of HCOOCH3 and CH3OCH3 used in Table 7.","marker":"Agúndez et al. (2021b)"}],"fun_headline_variants":["Radiolysis and sputtering reproduce cold-core COMs within factor 3","Non-thermal ice chemistry aligns TMC-1 organics within factor 3","Cosmic-ray sputtering and radiolysis explain cold-core COMs","Multiphase models with sputtering match TMC-1 five COMs","Cold-core complex organics formed by non-thermal ice processes"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that an excited ice molecule reacts with whichever neighbour it happens to touch and forms exactly the products listed in the 343-channel table, and that cosmic-ray radiolysis works uniformly through all ice layers with no depth attenuation; neither part is backed by experiment or theory.","fun_headline_variants_meta":{"raw":{"variants":["Radiolysis and sputtering reproduce cold-core COMs within factor 3","Non-thermal ice chemistry aligns TMC-1 organics within factor 3","Cosmic-ray sputtering and radiolysis explain cold-core COMs","Multiphase models with sputtering match TMC-1 five COMs","Cold-core complex organics formed by non-thermal ice processes"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000799,"raw_usage":{"total_tokens":3518,"prompt_tokens":954,"completion_tokens":2564,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":570,"completion_tokens_details":{"reasoning_tokens":2466}},"tokens_in":570,"tokens_out":2564,"duration_ms":20937,"temperature":1.0,"reasoning_tokens":2466,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T19:42:13.075786+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Irradiate mixed $\\mathrm{H_2O}$-CO-$\\mathrm{CO_2}$-$\\mathrm{CH_3OH}$ ices with cosmic-ray-like particles at 10 K and measure whether excited methoxy plus formyl actually produces methyl formate with appreciable branching; alternatively, rerun the same model with depth-dependent radiolysis attenuation and only experimentally confirmed suprathermal channels and check whether the factor-of-3 agreement with TMC-1 survives.","supporting_citations":[{"cited_title":"2017, ApJ, 851, 68, doi: 10.3847/1538-4357/aa99d9","cited_arxiv_id":null,"evidence_quote":"Supplies the accelerated Gillespie algorithm used to simulate the stochastic gas-grain network in feasible wall-clock time."},{"cited_title":"T., & Herbst, E","cited_arxiv_id":null,"evidence_quote":"Establishes the previous rate-equation baseline where HCOOCH3 is underproduced and raising sputtering overproduces CH3OCH3, the discrepancy this work targets."}],"review_version":1}