{"id":"f0e4458a-e80e-45c0-ad1e-c77e86d50edf","arxiv_id":"2412.01895","paper_version":2,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":8,"one_line_summary":"Disc evolution models with sublimating ices predict a water-to-CO2 chemical transition in inner discs, with dust traps raising CO2/H2O and linking the ratio to trap location.","lead":"This paper models how inward-drifting dust grains carry water and carbon-dioxide ices into the inner regions of planet-forming discs, producing a water-rich phase early and a CO2-rich phase later. It shows that the CO2/H2O ratio in infrared spectra could reveal the presence and location of dust-trapping gaps.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The CO2/H2O vs. gap-location correlation rests on pure-CO2 ice desorption (§2.3.3); with CO2 co-desorbing from polar H2O ice at the H2O snowline, the H2O-rich/CO2-rich phase separation disappears. A mixed-ice model test would settle whether the central prediction survives.","rationale":"The reader identified the same weakest assumption: the separation of the H2O-rich and CO2-rich phases depends on CO2 behaving as a pure ice with a single desorption temperature. I agree that this is the most load-bearing point. The paper is transparent, the code is released, and the modelling is internally consistent, so this is not a reason to reject the work. However, the central observable prediction—NCO2/NH2O as a tracer of drift and trap location—would be qualitatively different if a large fraction of CO2 is released with H2O from polar ice. Because the authors themselves note that the pure-ice case is an extreme scenario, the acceptance should be conditional on either a quantitative mixed-ice sensitivity test or an explicit reframing of the headline prediction as applying only to pure CO2 ice. The concrete test above would settle the question directly.","tokens_in":44589,"tokens_out":5702,"duration_ms":69508,"concrete_test":"Rerun the fiducial Scenario 1 grid with a mixed-ice prescription: let a fraction f_polar (e.g. 0.5–0.8, guided by Pontoppidan et al. 2008 and Brunken et al. 2024) of CO2 desorb together with H2O at the H2O snowline, with the remainder desorbing as pure CO2 at its own snowline. Recompute the analogues of Fig. 9 and Fig. 10 and check whether the positive NCO2/NH2O versus Rgap correlation at 1–3 Myr is suppressed or reversed. If it persists, the concern is resolved; if it disappears or reverses, the abstract and conclusions should be reframed as conditional on pure CO2 ices.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"In §2.3.3 the authors state the implicit assumption that ices are pure, successively layered, and have a single desorption temperature, and then immediately note that ice observations imply pure CO2 is never more than about 20% of CO2 ice (usually <10%), with the polar, H2O-rich phase most abundant. The central prediction—a H2O-rich phase followed by a CO2-rich phase, and the positive NCO2/NH2O versus Rgap trend at 1–3 Myr—requires CO2 to be released at its own snowline outside the H2O snowline, then advected inward on the longer viscous timescale at roughly 2 au. If a substantial fraction of CO2 is trapped in polar H2O ice and released near the H2O snowline, both molecules enter the same inner reservoir and are advected on the same short timescale; the ratio then tends to track the initial ice abundance ratio and loses sensitivity to gap location. The authors plausibly argue that volcano desorption and segregation may preserve some separation, but they also state that the paper 'assumes the most extreme scenario,' and no model variant quantifies the mixed-ice case. Since the proposed observational diagnostic and its correlation with trap location are the central claim, this untested assumption is load-bearing. The concern is not an internal inconsistency; it is a boundary on the domain of validity of the headline prediction.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript uses a 1D protoplanetary disc evolution code with a two-population dust model and molecular ice tracers to model how radially drifting dust and dust traps in gas gaps shape the inner-disc H2O and CO2 vapour abundances, and it post-processes the resulting structures with 0D LTE slab spectral models to predict observable MIR column densities. The central claims are that discs evolve through an H2O-rich phase into a CO2-rich phase as H2O vapour is accreted and CO2 is advected inward from its snowline, that dust traps hasten this transition and raise CO2/H2O, and that the retrieved NCO2/NH2O ratio is a more robust tracer of drift/trapping than individual column densities and may correlate with gap location at 1–3 Myr. The model predictions are compared qualitatively to a small sample of JWST MIRI-MRS discs, and the code is released at a tagged version.","tokens_in":44924,"tokens_out":7042,"duration_ms":80330,"significance":"If the predicted relationship between inner-disc CO2/H2O and the presence/location of dust traps holds, it would offer a new population-level observational diagnostic for the role of radial drift and substructure formation in setting the volatile chemistry available to planet formation. The paper’s strengths are its clear model specification, the explicit exploration of three dust-evolution scenarios, the careful treatment of continuum obscuration in the synthetic spectra, and the release of the model code. The authors are also candid about limitations that they do not model, such as vertical thermal structure and chemical reprocessing. The main risk to the central claim is the assumption of pure, successively layered CO2 ice, which is acknowledged in the text but not subject to a sensitivity test.","major_comments":[{"comment":"The central prediction—an H2O-rich phase followed by a CO2-rich phase, and the positive NCO2/NH2O–Rgap trend at 1–3 Myr—relies on the assumption stated in §2.3.3 that CO2 ice is pure and has a single desorption temperature at its own snowline (about 2.2 au in the model). The text immediately notes that protostellar ice observations imply pure CO2 is never more than about 20% of CO2 ice (usually <10%), with the polar, H2O-rich phase most abundant, and that the paper assumes the most extreme scenario. If a substantial fraction of CO2 is trapped in polar H2O ice and released near the H2O snowline at about 0.47 au, both molecules would enter the same inner reservoir and be advected on the same short viscous timescale (Eq. 19), so the CO2/H2O ratio would tend to track the initial ice abundance ratio and lose the dependence on Rgap shown in Figs. 9 and 10. The qualitative arguments for volcano desorption and segregation are plausible but do not provide a quantitative bound on the polar-phase fraction needed to preserve the correlation. I request a sensitivity test with a mixed-ice prescription—for example, a parametrized fraction of CO2 that co-desorbs with H2O—or an equivalent demonstration that the predicted ratio and its trend with gap location survive for observationally motivated ice-phase distributions.","section":"§2.3.3 and Figs. 9–10"}],"minor_comments":[{"comment":"The statement that 'the trends with respect to H2O are lost' is too absolute, since Fig. 9 shows a weak late-time negative trend in the H2O column density for the most distant gaps; please qualify the sentence.","section":"§4.3.3"},{"comment":"Equation (12) contains a factor (1−b) in the denominator, so the special case b = 1 is singular; please either exclude it explicitly or comment on the limiting behaviour.","section":"§2.2.2"},{"comment":"The markers labeled 'RM24 Compact' and 'RM24 Extended' are not described in the caption; please add one sentence explaining the observational estimates they represent.","section":"Fig. 3 caption"},{"comment":"The manuscript uses both 'snowline' and 'snow line'; please unify the spelling for consistency.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"The main technical issue is the pure-ice assumption for CO2 desorption. The authors acknowledge it explicitly, but because the CO2/H2O–gap-location correlation is the headline result, a sensitivity test or a quantitative robustness argument is needed. If the authors can show that the correlation survives a mixed-ice treatment (or at least clarify the domain of validity in terms of polar-phase fraction), I would support publication. The manuscript is otherwise well within the scope of A&A and the code release is a valuable asset."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line: this is a careful, useful modeling study. It gives JWST observers a new diagnostic—the CO2/H2O column density ratio—that may connect inner-disc chemistry to outer-disc dust traps. The central claim is well argued within the model's assumptions, and the paper is honest about the main caveat, but that caveat is more load-bearing than the authors let on.\n\nWhat's new: Booth, Kalyaan, and Mah did H2O and C/O effects of traps; this paper adds explicit CO2, a systematic grid of gap radii and formation times, and forward modeling into synthetic MIRI-MRS spectra with slab-model retrieval. That last step is valuable: it tests what an observer would actually measure, and the finding that the ratio survives dust-obscuration effects is the strongest part. The code is released at a tagged version, the initial ice abundances come from independent protostellar ice observations, and the binding energies from lab experiments; the comparison to observed discs is made after the models are run. No circularity problem.\n\nThe weakest spot is exactly what the authors flag in §2.3.3: they assume pure, layered ices with single desorption temperatures, and then note that pure CO2 ice is never more than ~20% of the CO2 ice budget, usually <10%. If most CO2 is in polar H2O ice and is released near the H2O snowline, the H2O-rich then CO2-rich phase separation weakens, and the NCO2/NH2O vs Rgap correlation loses its diagnostic power. The authors argue for volcano desorption and segregation, but they also call their choice the 'most extreme scenario.' A mixed-ice model variant would have settled this; its absence leaves the domain of validity unclear. A second soft spot is scenario dependence: the positive correlation at 1–3 Myr appears in Scenarios 1 and 3, but in Scenario 2 the ratio behaves differently and column densities overshoot observations by orders of magnitude. The paper acknowledges this, but it means the diagnostic is only as good as the assumption that the dust dynamics are in the Scenario 1/3 regime.\n\nThe observational comparison is appropriately cautious—small, heterogeneous sample, different fitting methods—and they don't overclaim. The discussion of what could produce CO2-dominated spectra is balanced.\n\nWho this is for: anyone interpreting MIRI-MRS spectra of T Tauri discs, or trying to connect ALMA substructure to inner-disc chemistry. It deserves a serious referee; I'd send it out. I'd also suggest the authors add a mixed-ice run in revision, because it directly tests the headline prediction.","headline":"Solid modeling paper with a useful new tracer, but the headline correlation rests on an explicitly flagged pure-CO2 ice assumption that the authors never test.","tokens_in":45525,"tokens_out":2626,"would_cite":true,"duration_ms":29141,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Drifting ice grains make planet-forming discs turn CO2-rich on million-year timescales, and the CO2/H2O ratio marks where dust traps sit.","keywords":["protoplanetary discs","dust radial drift","dust trapping","snow lines","CO2/H2O ratio","mid-infrared spectroscopy","ice sublimation","disc chemical evolution"],"falsifier":"A population survey combining ALMA gap radii with JWST MIRI-MRS measurements of CO2/H2O column-density ratios could settle the claim: if discs at 1–3 Myr show no positive correlation between the ratio and gap radius, the predicted dependence fails. The mechanism would also collapse if experiments or observations show that the majority of CO2 desorbs at the H2O sublimation temperature rather than at its pure-phase temperature.","tokens_in":44386,"feed_emoji":"🪐","tokens_out":8176,"duration_ms":77904,"temperature":0.7,"pith_summary":"This paper argues that the chemical diversity seen in the inner regions of planet-forming discs—some dominated by water vapour, others by carbon dioxide—can arise from the inward drift of dust grains carrying ices, and that the ratio of CO2 to H2O is a practical tracer of where dust trapping occurs. Using a 1D disc evolution model, the authors show that discs first become water-rich as water ice sublimates at its snow line, then turn CO2-rich as the water vapour drains onto the star and CO2 ice drifts inward from its more distant snow line. Dust traps—pressure bumps created by gaps in the gas disc—interrupt this delivery, speed up the transition, and raise CO2/H2O, with the size of the effect depending on the gap's radius and formation time. Synthetic mid-infrared spectra analysed with standard slab models show that the CO2/H2O column-density ratio is far less affected by dust obscuration than either molecule's absolute column density, making it the more reliable observable. If the picture is right, the ratio offers a way to read both the age of a disc and the location of its dust traps from JWST spectra.","feed_headline":"Dust traps leave a CO2/H2O fingerprint","feed_subtitle":"A disc-evolution model predicts water-rich discs age into CO2-rich ones, with the ratio revealing gap location.","key_machinery":"The load-bearing mechanism is the snow-line sequence, in which each ice sublimates at a characteristic radius set by its binding energy—H2O closest to the star, CO2 further out—and the released vapour is then drained onto the star on a viscous timescale proportional to that radius. The models implement this with a 1D viscous evolution code using the two-population dust model, and represent dust traps as Gaussian perturbations to the effective viscosity that create pressure bumps. The paper's central diagnostic is the CO2/H2O column-density ratio $N_{\\mathrm{CO_2}}/N_{\\mathrm{H_2O}}$: because the dust that delivers H2O also adds continuum opacity that hides the water, absolute columns are unreliable, whereas the ratio largely cancels this obscuration and tracks the underlying chemistry.","core_discovery":"The central claim is that inward-drifting pebbles deliver H2O and CO2 ices to the inner disc in a volatility-ordered sequence: H2O sublimates closest to the star and briefly dominates, then its vapour is accreted onto the star on a viscous timescale of roughly 0.25 Myr, while CO2 continues to arrive from its snow line further out, leaving the inner disc CO2-rich on Myr timescales. Introducing a dust trap cuts off the pebble flux, and because H2O is drained faster than CO2, traps raise the CO2/H2O vapour ratio; close-in or early-opening gaps block more of the CO2 ice and therefore produce smaller increases. The paper further claims this behaviour survives in observable form: the CO2/H2O column-density ratio retrieved from 0D LTE slab fits to synthetic MIR spectra mirrors the underlying vapour-mass ratio and depends only weakly on dust continuum obscuration, so it can serve as a tracer of radial drift and trapping in real discs.","pith_inferences":["A null result in the predicted correlation would independently point to CO2 being largely locked in polar water ice, or to surface-layer chemistry resetting delivered abundances; 2D thermochemical models could distinguish these.","If the correlation is confirmed, it would favour early-forming, non-planetary dust traps such as MHD zonal flows, because planets struggle to open gaps before ~0.1 Myr.","The time at which a disc switches from water-dominated to CO2-dominated encodes the viscous timescale at the water snow line, so the ratio could serve as a chemical age indicator for individual discs.","The paper's 'traffic jam' scenario implies water column densities are fixed by the ice-to-dust ratio rather than by delivered water mass; unusually high observed water columns would therefore signal dust loss or decoupling inside the snow line."],"forward_implications":["CO2-dominated inner-disc spectra should be a common late stage of disc evolution, appearing on ~1–3 Myr timescales even without gaps, as H2O drains onto the star faster than CO2 arrives.","Dust traps should raise the CO2/H2O ratio, and in discs with early-formed traps the ratio should increase with gap radius during the 1–3 Myr window, offering a direct observational test with ALMA gap catalogues.","Traps opened close to the star or very early block more CO2 ice and produce smaller ratio increases, so the ratio can constrain both the location and formation time of substructure.","Because the ratio is only weakly sensitive to dust continuum obscuration, retrieving it from weak features such as CO2 hot bands or 13CO2 should trace the bulk delivered chemistry better than absolute water or CO2 column densities.","Trap formation delayed beyond roughly 0.1 Myr weakens or reverses the ratio–gap-location trend, so the observed strength of the correlation constrains when substructures formed."],"supporting_citations":[{"why":"Supplies the 1D disc evolution code with gas viscous evolution, dust radial drift, and ice tracer transport that underlies all the models.","marker":"Booth et al. 2017"},{"why":"Provides the two-population dust model that sets the small and large grain fractions and the fragmentation and drift limits on grain growth.","marker":"Birnstiel et al. 2012"},{"why":"Previous modelling of how dust traps at different radii affect inner-disc H2O, which this paper extends to CO2 and to observable spectral signatures.","marker":"Kalyaan et al. 2021"},{"why":"Establishes that transport outpaces chemical processing for H2O and CO2, justifying the paper's neglect of chemical reactions.","marker":"Booth & Ilee 2019"},{"why":"Supplies the experimental binding energies and desorption prefactors that fix the H2O and CO2 snow-line locations.","marker":"Minissale et al. 2022"},{"why":"Provides the observed CO2-dominated spectrum of GW Lup and the 13CO2-based column density estimate used for comparison.","marker":"Grant et al. 2023"},{"why":"Provides the CO2-dominated CX Tau spectrum and 13CO2-based estimate used as an additional observational comparison.","marker":"Vlasblom et al. 2024a"},{"why":"Supplies the slab-model fitting grid and retrieval approach applied to the synthetic spectra.","marker":"Gasman et al. 2023"}],"fun_headline_variants":["Water-rich discs turn CO2-rich as pebbles drift","Dust traps raise CO2/H2O and hint at gap location","CO2/H2O ratio traces dust drift and trapping","Discs evolve from H2O-rich to CO2-rich over time","Dust traps etch CO2/H2O signature in discs"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The predicted separation of the water-rich and CO2-rich phases—and therefore the diagnostic power of the CO2/H2O ratio—rests on the assumption that CO2 ice is pure and sublimates at its own snow line; if most CO2 is trapped in polar water ice and released together with water, the two delivery episodes would merge and the ratio would lose its sensitivity to gap location.","fun_headline_variants_meta":{"raw":{"variants":["Water-rich discs turn CO2-rich as pebbles drift","Dust traps raise CO2/H2O and hint at gap location","CO2/H2O ratio traces dust drift and trapping","Discs evolve from H2O-rich to CO2-rich over time","Dust traps etch CO2/H2O signature in discs"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000796,"raw_usage":{"total_tokens":3591,"prompt_tokens":1121,"completion_tokens":2470,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":737,"completion_tokens_details":{"reasoning_tokens":2381}},"tokens_in":737,"tokens_out":2470,"duration_ms":18391,"temperature":1.0,"reasoning_tokens":2381,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T00:52:18.981796+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A population survey combining ALMA gap radii with JWST MIRI-MRS measurements of CO2/H2O column-density ratios could settle the claim: if discs at 1–3 Myr show no positive correlation between the ratio and gap radius, the predicted dependence fails. The mechanism would also collapse if experiments or observations show that the majority of CO2 desorbs at the H2O sublimation temperature rather than at its pure-phase temperature.","supporting_citations":[],"review_version":1}