{"id":"594104fc-6093-4174-804f-1b67f43040a9","arxiv_id":"2505.21470","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"With internal photoevaporation, inner discs around M dwarfs stay carbon-poor until the evaporation rate is reduced by a factor of 3-4, which reproduces observed high C/O ratios.","lead":"This paper simulates how X-rays from a small, cool star evaporate the gas disc around it and change the disc's carbon-to-oxygen balance. It finds young inner discs stay oxygen-rich with strong evaporation, while weaker evaporation lets carbon build up, matching recent space-telescope observations.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The low-C/O result rests on the fixed mid-plane temperature: late-time equilibrium cycles require ice lines near the gap's inner edge, and M-dwarf luminosity evolution shifts them. Relaxing this makes the claimed factor 3-4 photoevaporation overestimate non-unique.","rationale":"Central claim: with nominal photoevaporation rates the inner C/O stays low, contradicting observed high C/O in M-dwarf discs; reducing rates by 3-4 resolves the tension. This holds only if the gap-ice-line geometry is right. The model is internally consistent and the limitation is disclosed, but the fixed-temperature assumption is not neutral: it is set at 2.5 Myr and never evolved, even though the late-time equilibrium cycles that set the low C/O operate for millions of years while the star dims. The paper's justification in Section 2.1 covers early pebble delivery, not the late-time recycling that determines C/O after gap opening. The 0.3 M_sun result is especially fragile because the authors show that a modest viscosity change destroys the CO2 cycle; a luminosity-driven ice-line shift is a similar perturbation. My concern is not that the model is wrong but that the nominal-vs-reduced comparison conflates two unknowns: photoevaporation strength and thermal evolution. The proposed simulation separates them. Since the reader already judged the paper CONDITIONAL with moderate confidence, I do not change the verdict; I would add this time-dependent-temperature test as an explicit condition.","tokens_in":34691,"tokens_out":12569,"duration_ms":137388,"concrete_test":"Rerun the nominal and reduced photoevaporation cases with a time-dependent mid-plane temperature, e.g. updating the stellar luminosity from Baraffe et al. (2015) tracks at each output time, and recompute the ice-line radii as a function of time. For each stellar mass, overlay the time-dependent CO2 and H2O ice lines on the photoevaporative gap inner edge. If any ice line crosses the gap edge between 1 and 8 Myr, recompute the C/O trajectory; if the nominal-rate model yields C/O >1 at any age or the reduced-rate prediction changes sign, the claimed contradiction with observations and the factor 3-4 reduction do not follow uniquely.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's quantitative conclusion, that nominal Picogna et al. (2021) photoevaporation rates are too high by factor 3-4 because they force low inner C/O, depends on the disc's thermal structure as much as on the photoevaporation rate. Under nominal rates, the low C/O is produced by an equilibrium cycle operating at the outer edge of the inner disc: a water cycle for the 0.5 and late-time 0.1 M_sun models, and a CO2 cycle for the 0.3 M_sun model (Sections 3.1.2 and 3.2). The identity and survival of these cycles depend on whether the relevant ice line lies inside, at, or outside the inner edge of the photoevaporative gap. The code fixes the mid-plane temperature in time and adopts the stellar luminosity at 2.5 Myr (Sections 2.1, 2.4), while the paper itself notes that low-mass stars dim substantially with age. For a 0.1 M_sun star, the luminosity change between 1 and 10 Myr can shift irradiation-dominated ice lines inward by factors of about 2, potentially moving the CO2 or H2O ice line from the inner disc into the gap or vice versa. That would change which molecule is recycled and could raise or lower the late-time C/O. The sensitivity is already visible in Appendix C.2, where increasing alpha alone moves the CO2 line into the gap and destroys the 0.3 M_sun cycle; a time-dependent temperature is a comparable knob. Relaxing the fixed-temperature assumption may therefore remove the discrepancy with observed high C/O without lowering photoevaporation at all, making the claimed factor 3-4 overestimate non-unique.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper uses the chemcomp 1D semi-analytical disc model to study how internal photoevaporation shapes the C/O ratio in the inner regions of protoplanetary discs around low-mass stars (0.1-0.5 Msun). The authors compare simulations with nominal photoevaporation rates from Picogna et al. (2021) against simulations with X-ray-luminosity-reduced rates (by factors of 2.2-5.0), which lower the mass-loss rates by a factor of 3-4. They find that nominal rates open photoevaporative gaps early, block inward-drifting pebbles, and remove carbon-rich gas from the outer disc, leaving the inner disc with persistently low subsolar C/O ratios. Reduced rates delay gap opening, allowing carbon-rich vapour to raise the C/O ratio to supersolar values in some cases, which the authors argue agrees with JWST observations such as the high C/O of Sz 28. The paper concludes that the nominal photoevaporation rates overestimate the mass-loss efficiency for M dwarfs and that a factor 3-4 reduction reconciles the models with observations.","tokens_in":34987,"tokens_out":6103,"duration_ms":68980,"significance":"The paper addresses an observationally important tension: many inner discs around low-mass stars show high C/O ratios, whereas standard photoevaporation models are thought to produce low C/O. The study provides a concrete mechanistic explanation (equilibrium cycles at the photoevaporative gap edge) and a falsifiable prediction that young (<2 Myr) inner discs should be oxygen-rich and old (>2 Myr) discs carbon-rich. The inclusion of comparison simulations without photoevaporation, with a planet, and with a different viscosity strengthens the interpretion of the mechanism. The authors also openly state several model simplifications, including time-independent temperature and uniform photoevaporation of all molecular species. However, as detailed in the major comments, the central quantitative conclusion that nominal rates overestimate by factor 3-4 rests on a thermal-structure assumption that is not tested, and the reduced rates are hand-calibrated to the very observations they are meant to reproduce. If the thermal assumption is relaxed, the discrepancy with observations may not uniquely require lower photoevaporation rates.","major_comments":[{"comment":"The claim that nominal photoevaporation rates produce persistently low inner-disc C/O ratios depends directly on the time-independent mid-plane temperature, which is calculated once using the stellar luminosity at 2.5 Myr. The equilibrium cycles that set the late-time C/O (water cycle for 0.5 Msun and late-time 0.1 Msun; CO2 cycle for 0.3 Msun) require the relevant ice line to sit at the inner edge of the photoevaporative gap. The authors themselves note that low-mass stars dim substantially with age; for a 0.1 Msun star, the irradiation-dominated ice lines can move inward by factors of about 2 between 1 and 10 Myr. A time-dependent temperature could shift the CO2 or H2O ice line into the gap or into the inner disc, destroying or creating the equilibrium cycle and thereby changing the C/O trajectory. The sensitivity of the mechanism is already visible in Appendix C.2, where increasing alpha alone moves the CO2 line into the gap and removes the 0.3 Msun cycle. I ask the authors to test the fixed-temperature assumption by rerunning the nominal-rate simulations with time-evolving stellar luminosity (or at least with luminosities appropriate for 1 Myr and 5 Myr) and to show whether the low-C/O result persists. Without such a test, the conclusion that the nominal rates overestimate photoevaporation by factor 3-4 is not robust and the discrepancy with observations is not uniquely attributable to the photoevaporation rate.","section":"§2.1, §2.4, §3.1.2"},{"comment":"The X-ray reduction factors [5.0, 3.5, 2.2] for stellar masses [0.5, 0.3, 0.1] Msun are chosen, in the words of the authors, so that the reduced rates bring the calculated C/O ratios into better agreement with observations. The same observed C/O values are then used as evidence that the nominal rates overestimate mass loss by factor 3-4. This is close to circular: the data are used both to set the reduction factors and to validate them. The factors are within the observed spread of X-ray luminosities, but the spread is large and a range of factors would be consistent with it; the paper does not show that the required factor is independently predicted. I recommend that the authors present a forward-model comparison over the observed X-ray luminosity distribution, or adopt a statistical measure (e.g., likelihood or chi-square against the observed C/O constraints), and explicitly state that the reduction factors were calibrated to match the C/O observations. This would clarify the logical status of the conclusion.","section":"§2.2, §3.2"},{"comment":"The comparison to observations is qualitative and rests primarily on a single system, Sz 28, with a lower-limit C/O>1. The model C/O ratio varies strongly with radius and time, while the MIRI observations probe a range of radii weighted by the disc emission and a specific disc age; the paper does not quantify the radius range or the age interval over which the match is claimed to hold. A more quantitative comparison, accounting for the observed emission region and the stellar age, would materially strengthen the claim that the reduced rates 'fit' the observations and that the nominal rates are excluded.","section":"§4.2"}],"minor_comments":[{"comment":"The photoevaporation rates in Table 2 are quoted without uncertainties; given the spread in the Güdel et al. (2007) relation and the adopted scaling, a statement of the expected uncertainty in the reduced mass-loss rates would help the reader judge the significance of the factor 3-4.","section":"§2.2"},{"comment":"In the 0.1 Msun case, the text says the C/O ratio increases 'up to super-solar values' after the initial drop; please specify the numerical value and time at which this maximum occurs, since the later evolution is driven by the switch from a CO2 cycle to a water cycle.","section":"§3.1.2"},{"comment":"The authors correctly note that photoevaporation is implemented as acting uniformly on all molecular species, whereas in reality lighter species are removed more efficiently. This assumption could affect the C/O ratio in either direction; I suggest adding a brief quantitative estimate or citing a specific result from the literature that bounds this effect, rather than leaving it as an open qualitative caveat.","section":"§4.1.2"},{"comment":"No data availability statement is provided. Since the paper is based on a semi-analytical code with multiple parameters and appendices, a statement on code/plotting scripts availability would aid reproducibility.","section":"General"}],"recommendation":"major_revision","confidential_remarks":"The paper is within the scope of A&A and the authors have presented an internally consistent modelling study with clearly stated limitations. The main scientific concern is that the central conclusion, the factor 3-4 overestimate of photoevaporation rates, depends on the time-independent temperature assumption in a way that is not tested; this is a load-bearing point that can be addressed with additional simulations. The second concern is the hand-calibration of reduction factors to the observed C/O values. I believe these are fixable within a revision and do not require rejection, but the current version overstates the uniqueness of the reconciliation. The authors should also consider whether the quantitative comparison to Sz 28 can be made more rigorous."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This is a careful M-dwarf extension of the photoevaporation-chemistry story, and the nominal-rate result deserves to be taken seriously. The mechanism from Paper I carries over cleanly: the gap blocks pebbles, the wind removes carbon-rich gas, and the inner-disc C/O stays low. The authors do a solid job mapping how gap position and gap-opening time vary with stellar mass, and the mass-dependent CO2 equilibrium cycle for 0.3 Msun is a genuinely new piece of behavior. The comparison with pure viscous discs and planet-gap discs helps isolate what photoevaporation actually does, and the paper is honest about its simplifications: uniform molecular loss, no refractory carbon, fixed mid-plane temperature.\n\nThe soft spot is the reconciliation step. The nominal model gives low inner C/O, which does contradict observations like Sz 28. But the fix is to reduce the X-ray luminosity by hand-chosen factors and then read off agreement with the same observations. This is not fundamentally circular, because the low-C/O prediction is a real model output, but the factor 3-4 is a fitted knob, not a derived quantity. The stress-test point about the fixed temperature has real teeth here. The late-time C/O is controlled by whether a water or CO2 equilibrium cycle operates at the inner edge of the gap, which depends on the ice lines sitting in the right place. The paper assumes a constant temperature set by the 2.5 Myr luminosity, while low-mass stars dim substantially. A time-dependent temperature could move the CO2 ice line into or out of the gap, changing the cycle and possibly producing high C/O without lowering photoevaporation at all. The authors mention this sensitivity in section 3.1.2, but they do not test it. Given that Appendix C.2 shows the CO2 cycle dies when alpha is raised, this is a legitimate sensitivity, not a nitpick.\n\nThe observational side is also thin: essentially one strong M-dwarf inner-disc C/O measurement, plus a few indirect cases. No code or data is released, so reproducibility is limited. These are fixable issues, not fatal ones.\n\nWho is this for? People modeling disc chemistry and planet formation around low-mass stars. It deserves a serious referee. My recommendation: send it to review, but the referee should push for a time-dependent temperature run or a clear argument for why it does not matter, and for a more systematic treatment of the reduction factor, ideally tied to a specific lower-photoevaporation model rather than an a-posteriori luminosity tweak.","headline":"Solid M-dwarf extension of the photoevaporation-chemistry model, but the claimed reconciliation with observed C/O relies on hand-tuned X-ray reduction factors and a fixed temperature profile, so the factor 3-4 overestimate is not yet uniquely established.","tokens_in":35569,"tokens_out":2715,"would_cite":false,"duration_ms":33152,"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":"Internal photoevaporation sets the C/O ratio of inner M-dwarf discs, and a 3–4 times weaker wind is needed to match carbon-rich observations.","keywords":["protoplanetary discs","M dwarf stars","internal photoevaporation","C/O ratio","pebble drift","disc chemistry","planet formation","X-ray luminosity"],"falsifier":"Measure the inner-disc C/O ratio of a sample of M dwarfs spanning ages of roughly 1 to 6 Myr, with their X-ray luminosities recorded; the model is falsified if old ($>2$ Myr) low-X-ray discs are not carbon-rich, or if young ($<2$ Myr) discs already show C/O above solar, because the timing of the water-rich phase and the gap opening is the causal hinge.","tokens_in":34414,"feed_emoji":"🪐","tokens_out":13359,"duration_ms":121922,"temperature":0.7,"pith_summary":"This paper tries to establish that internal photoevaporation, not just viscous transport and pebble drift, controls the carbon-to-oxygen ratio (C/O) of the inner regions of protoplanetary discs around low-mass stars of roughly $0.1$ to $0.5\\,M_\\odot$. In the model, the photoevaporative wind opens a gap that blocks inflowing pebbles and carries away carbon-rich vapour from the outer disc, so the inner disc stays oxygen-rich and C/O-poor when the nominal wind rate is used. Since observed M-dwarf discs show high C/O, the paper argues that the nominal X-ray photoevaporation rates are too strong, and that reducing the mass-loss rate by a factor of 3 to 4 delays gap opening, lets carbon-rich gas enter, and reproduces the observed C/O values. This matters because the C/O of the disc gas sets the compositions of planets that form there, and M dwarfs are the most common planet hosts.","feed_headline":"M-dwarf discs need a weaker X-ray wind to stay carbon-rich","feed_subtitle":"The model says a 3 to 4 times weaker wind lets carbon-rich gas reach the inner disc, explaining observed high C/O ratios.","key_machinery":"The load-bearing mechanism is the soft X-ray photoevaporation prescription combined with equilibrium cycles at evaporation fronts. The prescription fixes the radial profile and total rate of gas loss, which decides where the gap opens (roughly 0.7 to 3 AU) and when (1 to 2 Myr for nominal rates, 5 to 6 Myr for reduced rates). At the gap's inner edge a strong pressure gradient pushes vapour outwards; if that edge coincides with a particular ice line, the vapour recondenses into pebbles that drift back inwards and evaporate again, forming an equilibrium cycle. In these simulations the recycled volatile is water for some stellar masses, keeping oxygen in the inner disc, and CO2 for the $0.3\\,M_\\odot$ case, keeping carbon in; the evaporation-front positions are set by a time-independent mid-plane temperature. This contrast between the water cycle and the CO2 cycle is what produces the different C/O trajectories across stellar mass.","core_discovery":"For discs around stars of $0.1$ to $0.5\\,M_\\odot$, the central claim is that internal photoevaporation sets the inner-disc C/O ratio through the timing and location of the photoevaporative gap. With the nominal soft X-ray mass-loss rates, the gap opens at 1 to 2 Myr, separating the inner disc from carbon-rich gas evaporated from CO, CO2 and CH4 ice; a water or CO2 equilibrium cycle can then recycle one volatile at the gap's inner edge, leaving the inner disc oxygen-rich with subsolar C/O. Observations of low-mass-star discs instead show C/O above solar in systems such as Sz 28, so the paper concludes that nominal photoevaporation rates overestimate mass loss for M dwarfs. A factor 3 to 4 reduction delays gap opening to 5 to 6 Myr, allowing carbon-rich gas to enrich the inner disc first and yielding C/O values in line with observations. The model therefore predicts a two-phase composition history: young inner discs, younger than about 2 Myr, should be oxygen-rich and carbon-poor, while older discs should be carbon-rich if the wind is weak enough.","pith_inferences":["Beyond the paper: the water-versus-CO2 cycle switch with stellar mass implies a sharp transition in the carbon-to-oxygen ratio available to planet formation around M dwarfs, so planets forming around stars near $0.3\\,M_\\odot$ may show a different atmospheric C/O than those around slightly more or less massive stars.","Beyond the paper: if the nominal photoevaporation rates are too high because the model omits cooling, then standard estimates of M-dwarf disc lifetimes are too short, which would push the window for forming planets around the most common stars later than usually assumed.","Beyond the paper: the NH3 equilibrium cycle visible in the C/H and N/H results suggests that measuring nitrogen abundances in inner discs would provide an independent observational test of the gap-edge recycling picture."],"forward_implications":["Inner discs around M dwarfs younger than about 2 Myr should be oxygen-rich and carbon-poor, because water-ice pebbles evaporate early and dominate the inner gas before carbon-rich vapour arrives.","Discs older than about 2 Myr should be carbon-rich, with C/O above solar, only if the photoevaporation rate is low enough to delay gap opening until that carbon-rich vapour reaches the inner disc.","Nominal soft X-ray photoevaporation rates overestimate mass loss for M dwarfs; reducing them by a factor of 3 to 4 brings both the inner-disc C/O ratio and the disc lifetime into better agreement with observations.","The stellar mass determines which evaporation front sits at the gap's inner edge, so the equilibrium cycle that preserves carbon or oxygen changes with stellar mass, giving different C/O evolutions for each mass.","Observed high inner-disc C/O values such as that of Sz 28 can be reproduced without changing the pebble or chemistry model, purely by lowering the wind strength."],"supporting_citations":[{"why":"supplies the soft X-ray photoevaporation prescription (equations 5-6) and nominal mass-loss rates whose strength the paper tests against observations.","marker":"Picogna et al. (2021)"},{"why":"provides the stellar-mass-to-X-ray-luminosity relation (equation 7) that sets both nominal and reduced photoevaporation rates.","marker":"Güdel et al. (2007)"},{"why":"provides the soft X-ray luminosity scaling and X-ray-dependent mass-loss relation used to compute the wind rates.","marker":"Ercolano et al. (2021)"},{"why":"describes the chemcomp disc model, including viscous evolution, pebble growth and drift, and chemical partitioning, which carries the simulations.","marker":"Schneider & Bitsch (2021a)"},{"why":"is the Paper I solar-mass companion study whose photoevaporation-and-C/O method and conclusions are extended here to low-mass stars.","marker":"Lienert et al. (2024)"},{"why":"reports C/O greater than 1 for the M dwarf system Sz 28, the key observed anchor for the high-C/O prediction.","marker":"Kanwar et al. (2024a)"},{"why":"provides the low-mass-star viscous C/O evolution and the initial disc radius scaling used for initial conditions.","marker":"Mah et al. (2023)"},{"why":"models photoevaporation with additional cooling and finds lower mass-loss rates, supporting the factor 3-4 reduction.","marker":"Sellek et al. (2024)"}],"fun_headline_variants":["M-dwarf discs need weaker winds to stay carbon-rich","Carbon-rich M-dwarf discs imply gentler X-ray winds","Weak photoevaporation lets carbon reach inner disc","M-dwarf disc C/O ratio hinges on wind strength","Disc carbon wealth tied to subtle winds around M dwarfs"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument rests on the assumed strength and position of the photoevaporation wind, together with evaporation-front locations set by a time-independent mid-plane temperature; if these do not hold for M dwarfs, the predicted C/O trajectories and the claimed mismatch with observations do not follow.","fun_headline_variants_meta":{"raw":{"variants":["M-dwarf discs need weaker winds to stay carbon-rich","Carbon-rich M-dwarf discs imply gentler X-ray winds","Weak photoevaporation lets carbon reach inner disc","M-dwarf disc C/O ratio hinges on wind strength","Disc carbon wealth tied to subtle winds around M dwarfs"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000191,"raw_usage":{"total_tokens":1455,"prompt_tokens":1167,"completion_tokens":288,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":783,"completion_tokens_details":{"reasoning_tokens":207}},"tokens_in":783,"tokens_out":288,"duration_ms":3897,"temperature":1.0,"reasoning_tokens":207,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T13:27:17.275417+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the inner-disc C/O ratio of a sample of M dwarfs spanning ages of roughly 1 to 6 Myr, with their X-ray luminosities recorded; the model is falsified if old ($>2$ Myr) low-X-ray discs are not carbon-rich, or if young ($<2$ Myr) discs already show C/O above solar, because the timing of the water-rich phase and the gap opening is the causal hinge.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"supplies the soft X-ray photoevaporation prescription (equations 5-6) and nominal mass-loss rates whose strength the paper tests against observations."},{"cited_title":"J., & Preibisch, T","cited_arxiv_id":null,"evidence_quote":"provides the soft X-ray luminosity scaling and X-ray-dependent mass-loss relation used to compute the wind rates."},{"cited_title":"L., Bitsch, B., & Henning, Th","cited_arxiv_id":null,"evidence_quote":"is the Paper I solar-mass companion study whose photoevaporation-and-C/O method and conclusions are extended here to low-mass stars."},{"cited_title":"D., Grassi, T., Picogna, G., et al","cited_arxiv_id":null,"evidence_quote":"models photoevaporation with additional cooling and finds lower mass-loss rates, supporting the factor 3-4 reduction."}],"review_version":1}