{"id":"29047188-81d0-45e0-b1d2-3ecab370750f","arxiv_id":"1908.06988","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"Ingestion of a 5.2-Earth-mass rocky planet on a grazing orbit can explain the 0.128 dex metal enhancement of turn-off star M67 Y2235.","lead":"This paper shows that a star in the cluster M67 could have swallowed a super-Earth planet, which would explain why its surface contains more heavy elements than its twin stars. The finding offers a concrete mechanism for abundance anomalies in stars and suggests a way to measure how often planetary systems become unstable.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Dissolution criterion rests on CH, which is varied by only a factor of two in Sec. 3.3.5 despite the order-of-magnitude uncertainty admitted in Sec. 3.1; a smaller CH shifts the v_tilde<0.4 boundary and could let the planet sink below the convective zone before depositing its metals.","rationale":"The paper is best read as arguing that super-Earth ingestion can explain the observed enhancement, not as a measured rate: the N-body experiments are explicitly representative and the authors decline to compute an ingestion rate. Under that reading, the decisive burden is to show that a planet on a grazing orbit dissolves before sinking below the convective zone. That burden is carried by Equation 7, whose coefficient CH is not calibrated for planetary-scale bodies and is only varied by a factor of two in the robustness tests. The reader identified this same weak point, and the concern is concrete: because the deposited mass fraction in Figure 8 changes visibly between CH=0.005 and CH=0.02 in the transition region, an order-of-magnitude error in CH could move the v_tilde boundary substantially. The proposed test would settle this by spanning the plausible range and applying the actual impactor distributions from the N-body simulations. Other acknowledged limitations, such as differential settling (Sec. 3.3.1), impact time (Sec. 3.3.3), and the absence of a rate calculation (Sec. 4.1), are less likely to invalidate a single-object can-explain claim, especially because the observed abundance pattern is a constant offset and the timing can be adjusted. The conditional verdict is therefore appropriate and does not need to be changed.","tokens_in":16633,"tokens_out":14897,"duration_ms":174975,"concrete_test":"Re-run the fiducial dissolution integration (Equations 5-7) for CH = 1e-4, 1e-3, 3e-3, 0.01, 0.03, 0.1 and for planet masses 1-30 Mearth, recording for each CH the critical v_tilde below which at least 5.2 Mearth is deposited in the convective zone. Then convolve those thresholds with the v_tilde distributions in Figure 9 (Kozai 2 au, Kozai 5 au, and scattering). If at CH=1e-3 the fraction of impactors satisfying the criterion drops below roughly half, the CH uncertainty is decisive; if the fraction remains above 90 per cent, the conclusion is robust despite the coefficient uncertainty.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that planets with v_tilde<0.4 dissolve entirely in the convective envelope is carried by Equations 5-7, in which the mass-loss rate is linear in CH, the fraction of frictional energy deposited in the planet. Section 3.1 concedes that obtaining a good estimate of CH is challenging and that the meteor-ablation analogy gives values between roughly 3 and 30 per cent; the authors set CH=0.01, below the lower end of that range, calling it conservative. Section 3.3.5 then varies CH only by a factor of two (0.005 and 0.02), which does not bracket the order-of-magnitude uncertainty. If the true CH were, for example, 0.001, the ablation rate would be ten times slower and a grazing planet could survive long enough for its decaying orbit to carry it below the base of the convective zone before it is fully consumed. The authors do not map the critical v_tilde threshold as a function of CH over the plausible range, so the clean separation shown in Figure 7 is not yet robust. The delivery simulations in Figure 9 produce very grazing orbits, but the scattering sample already contains a tail of impactors near v_tilde~0.4; a lower threshold would move part of that tail into the regime where metals are deposited where they cannot be observed. Because the mechanism requires essentially all of the 5.2 Mearth of metals to remain in the small convective envelope, this uncalibrated coefficient is the load-bearing uncertainty in the argument.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper proposes that the 0.128 dex enhancement in the photospheric metal abundances of the M67 turn-off star Y2235, relative to two sibling turn-off stars, is due to the late ingestion of a ~5.2 M_earth rocky or icy planet. The authors construct a stellar model of Y2235 and find a surface convective envelope of mass 3.45e-3 M_sun at 4 Gyr, so that Eq. (2) yields 5.2 M_earth of accreted metals for the observed offset. They then develop a simple drag-plus-vaporization model (Eqs. 3-7) in which a planet dissolves in the convective envelope if its radial velocity at first contact is less than about 40% of the total velocity (impact parameter b > 0.9). N-body simulations of planet-planet scattering and Lidov-Kozai cycles show that most delivered planets arrive on grazing orbits. The paper concludes that super-Earth ingestion is a good explanation and suggests a high-resolution spectroscopic survey of M67 to test the frequency of such events.","tokens_in":17016,"tokens_out":13140,"duration_ms":128372,"significance":"If the result holds, the paper provides a concrete, physically motivated explanation for a striking abundance anomaly in a cluster member, with a testable observational prediction. The calculation of the required planet mass is transparent and parameter-free once the convective envelope mass is accepted, and the delivery simulations are a useful first exploration of the orbital phase space of ingested planets. The paper's main strengths are its clarity, the explicit treatment of the dissolution criterion, and the fact that the proposed survey is falsifiable. However, the central quantitative claim rests on the value of the heat-transfer coefficient C_H, which is acknowledged to be highly uncertain and is tested over only a factor of two; this weakens the robustness of the v_tilde < 0.4 dissolution boundary.","major_comments":[{"comment":"The mass-loss rate in Eq. (7) is proportional to C_H, the fraction of frictional energy deposited in the planet, yet the paper sets C_H = 0.01 after noting that meteor-ablation formulae give values between about 3% and 30%, and then varies C_H only by a factor of two (0.005 and 0.02). This does not bracket the order-of-magnitude uncertainty the authors themselves identify. If C_H were as low as 0.001, the ablation rate would be ten times slower and a grazing impactor could survive long enough to sink below the base of the convective zone before fully dissolving, moving the critical v_tilde threshold to lower values. Because the scenario requires essentially all 5.2 M_earth of metals to remain in the thin convective envelope, the paper should either map the dissolution boundary as a function of C_H over the plausible range (e.g., 0.001-0.3) or provide a first-principles argument that C_H cannot be much smaller than 0.01. As written, the clean separation in Fig. 7 and the 40% threshold are not yet robust.","section":"Sec. 3.1, Eq. (7); Sec. 3.3.5, Fig. 8"},{"comment":"The treatment of differential settling and of ingestion time is a hand-wave. The present-day abundance is compared with the stellar model at 4 Gyr, but ingestion could occur at earlier times when the convective envelope was smaller (Fig. 3), and subsequent settling/levitation could alter the surface pattern. The manuscript states that the 'only likely effect is to slightly increase the mass of planet' (Sec. 3.3.1), but no calculation is shown. A simple time-dependent model--add the planet's metals at time t_ing, then evolve the convective mass and surface abundances using the same diffusion treatment that fits the comparison stars--is needed to demonstrate that the 5.2 M_earth estimate and the flat abundance pattern survive. This is load-bearing because the quantitative claim is the 0.128 dex offset at the current age.","section":"Sec. 3.3.1 and Sec. 3.3.3"},{"comment":"The N-body delivery statistics are reported without uncertainties. The scattering sample yields 49 impactors from 100 systems, with a median tilde-v_perp of 0.12 and 'only three' impactors that would penetrate below the convective zone; the Poisson uncertainty on that number is large, and the fraction of impactors above the critical threshold (which itself depends on C_H, see comment 1) is not quantified. Since the paper's conclusion is that 'almost all planets that are ingested arrive at the star on grazing orbits,' the authors should report the full distribution or at least a confidence interval on the fraction of grazing impacts. This is especially important because the scattering tail lies near tilde-v_perp ~ 0.4 and could cross the threshold if C_H is lower.","section":"Sec. 4.3, Fig. 9"}],"minor_comments":[{"comment":"The units 'g cm-2' for density should be 'g cm-3'; this appears for rho_p = 7.9, 5.5, and 10 g cm-2.","section":"Sec. 3.3.4"},{"comment":"The phrase 'between 0.02 and 0.035 M_sun from the stellar surface' should presumably read 'between 0.02 and 0.035 R_sun from the stellar surface' (or a mass coordinate), as the units are inconsistent with the context of Fig. 3.","section":"Sec. 3.3.3"},{"comment":"The term 'entropy of vaporisation of 5 kJ g-1' should be 'latent heat of vaporisation' or 'specific energy of vaporisation,' since entropy has units of energy per temperature, not per mass.","section":"Sec. 3.3.4"},{"comment":"The text gives 'ten Earth masses (3e-5 M_sun)'; using 3.003e-5 M_sun would be more precise.","section":"Sec. 3.2"},{"comment":"The results of the C_H sensitivity runs appear only in the caption of Fig. 8; the text should explicitly state that changing C_H from 0.01 to 0.005 and 0.02 has a small effect on the mass deposited, and should comment on the direction of the effect.","section":"Sec. 3.3.5 and Fig. 8"},{"comment":"The statement that the flat abundance pattern 'likely represents the ingestion of an icy/water-rich super-Earth' is a strong inference; the paper should acknowledge that a planet with exactly stellar metal ratios (except H/He) is a non-trivial compositional requirement and discuss its plausibility in more detail, or soften the claim.","section":"Sec. 5"}],"recommendation":"major_revision","confidential_remarks":"The paper is interesting and likely within scope for MNRAS. The main technical weakness is the uncalibrated C_H coefficient; I would be willing to accept after the authors either map the dissolution boundary over a wider C_H range or justify why C_H cannot be far below 0.01. The settling/timing issue also deserves a quantitative treatment. The N-body statistics are of secondary importance."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear [colleague],\n\nThe short version: this paper gives the cleanest quantitative statement I've seen of when an ingested rocky planet actually gets mixed into a thin surface convection zone. The criterion — radial velocity fraction below about 0.4, i.e. impact parameter above ~0.9 — comes out of a simple drag/ablation model, and the dynamical part shows that both Lidov-Kozai and scattering deliver planets on very grazing orbits. So the mechanism is not just hand-waved; it connects a specific observed anomaly (the 0.128 dex offset in M67 Y2235) to a plausible late dynamical event.\n\nWhat it does well: the required metal mass (5.2 Mearth) is derived directly from the model convective envelope mass, not fitted. The authors are explicit about the simplifying assumptions, and they test several: overshooting, time of impact, planet density. The N-body experiments are representative rather than exhaustive, but the median impact parameters are so grazing (0.03–0.29) that the conclusion seems robust to reasonable variations.\n\nSoft spots: the biggest is CH, the fraction of frictional energy deposited in the planet. They set it to 0.01, citing meteor ablation work that gives 3–30%, so it's below the lower end. They call it conservative, which is right in direction — lower CH means harder to dissolve — but they only vary it by a factor of two (0.005–0.02), not over the range implied by the uncertainty. The stress-test worry is that if CH were 0.001, a grazing planet would survive longer and its decaying orbit could take it below the convective zone before it is fully consumed. That hypothetical is well below any published estimate, so it isn't disqualifying, but the authors should map the dissolution threshold over a wider CH range and show how the delivery tails in Figure 9 compare. The settling argument (Sec 3.3.1) is admittedly a hand-wave; it's probably fine for a 0.128 dex difference, but it deserves a more serious treatment if the mechanism is to be considered established. The N-body results also lack uncertainties; with only 49 scattering impactors, the median has a noticeable Poisson spread.\n\nOverall, the central argument holds up as a plausible explanation. It's not a detection of the event, but it's a well-constructed plausibility model with a testable prediction (a spectroscopic survey of M67 turn-off stars). This is the kind of paper I'd want a referee to spend time on rather than desk-reject. I would recommend sending it to review, with the requirement that the authors widen the CH exploration and add error estimates to their simulation statistics.\n\nWho it's for: anyone working on planet-star interactions, chemical tagging, or cluster abundance anomalies. I'd probably cite it if I were writing about pollution signatures in evolved stars.\n\nBest,\n[Your name]","headline":"A credible quantitative case that a ~5 Mearth super-Earth ingested by M67 Y2235 on a grazing orbit explains its 0.128 dex metal offset; the main weakness is the undertested CH coefficient, but the argument is worth engaging.","tokens_in":17547,"tokens_out":5838,"would_cite":true,"duration_ms":57779,"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":"A single ingested super-Earth of roughly five Earth masses explains the anomalously high metal abundances of the turn-off star M67 Y2235.","keywords":["super-Earth ingestion","M67 Y2235","stellar abundance enhancement","convective envelope","planet-planet scattering","Lidov-Kozai oscillations","differential stellar spectroscopy","open clusters"],"falsifier":"Measure the heat-transfer coefficient $C_H$ for iron-rich bodies at Mach numbers near 50 in tenuous gas through meteor-ablation observations or shock-tube experiments: if $C_H$ is much smaller than 0.005, the dissolution boundary shifts from $\\tilde v_\\perp\\lesssim0.4$ toward far more grazing orbits, so typical scattering and Kozai impacts would no longer dissolve in the convective zone and the explanation for Y2235 would fail.","tokens_in":16440,"feed_emoji":"🪐","tokens_out":11339,"duration_ms":106847,"temperature":0.7,"pith_summary":"This paper argues that the unexplained 0.128 dex metal excess of the turn-off star M67 Y2235 can be produced by the late ingestion of a single rocky super-Earth. The mechanism is economical: the star's surface convection zone holds only $3.45\\times10^{-3}\\,M_\\odot$, so about $5.2\\,M_\\oplus$ of metals is enough to raise every measured abundance by a constant factor. The paper's key result is a geometric dissolution criterion: a planet dissolves completely in the convective envelope if it hits the star with radial speed less than 40% of its total speed at the surface, i.e. impact parameter above about 0.9. It then shows that both planet-planet scattering and Lidov-Kozai oscillations deliver planets almost always on such grazing orbits. If correct, the result turns one anomalous star into evidence that late dynamical instability of planetary systems can be detected through cluster abundance surveys.","feed_headline":"A swallowed super-Earth explains a star's metal excess","feed_subtitle":"Grazing impacts dissolve the planet in the thin surface convection zone, so all its metals reach the photosphere.","key_machinery":"The central object is the fractional radial velocity at first stellar contact, $\\tilde v_\\perp = v_\\perp/v_0$, equivalently the impact parameter $b = v_\\parallel/v_0$; the key identity is the dissolution boundary $\\tilde v_\\perp \\lesssim 0.4$ ($b\\gtrsim 0.9$). It is obtained from a numerical model in which a spherical iron-density planet loses mass at the rate $\\dot M_p = C_H F_D v /(\\epsilon_{\\rm bind,p}+L_{\\rm vap})$ under gas drag $F_D=\\frac12 C_D\\pi R_p^2 \\rho_\\star v^2$ while its orbit decays. This boundary does the central work of the paper: it turns the abundance anomaly into an orbital-geometry question, and the paper's N-body simulations then show that both delivery channels produce the required grazing geometry.","core_discovery":"The paper's central claim is that the anomalously high metal abundances of the M67 turn-off star Y2235—a constant 0.128 dex offset in every measured species relative to two sibling turn-off stars—can be explained by the late ingestion of a single super-Earth. For its fiducial 4 Gyr, 1.18 solar-mass model of Y2235, the surface convective envelope contains only $3.45\\times10^{-3}\\,M_\\odot$, so the required metal mass is about $5.2\\,M_\\oplus$. The paper's dissolution model then shows that a rocky planet of up to about $30\\,M_\\oplus$ is fully vaporized inside this convective layer provided it enters on a grazing orbit, $\\tilde v_\\perp = v_\\perp/v_0 \\lesssim 0.4$, i.e. impact parameter $b\\gtrsim0.9$. Planets that plunge more steeply deposit most of their mass below the convective zone and leave no photospheric signature. Since N-body simulations of both planet-planet scattering and Lidov-Kozai oscillations deliver planets with median $\\tilde v_\\perp$ of 0.12 and 0.05–0.29 respectively, essentially all ingested planets dissolve where their metals are observable. The paper therefore concludes that super-Earth ingestion is a viable explanation and suggests M67-like clusters as testbeds for late planetary-system instability.","pith_inferences":["If ingestion is common, abundance scatter among coeval turn-off stars should grow with stellar mass, because the surface convective envelope shrinks for higher-mass stars; a cluster survey could test this gradient.","The dissolution criterion implies that stars with essentially radiative envelopes would show no photospheric pollution even after ingesting planets, so the absence of metal offsets in warmer stars would not rule out ingestion.","The $\\tilde v_\\perp\\lesssim0.4$ boundary is a sharp prediction: any polluted star should show evidence of a dynamical delivery channel, such as an eccentric outer giant planet or a binary companion, which radial-velocity monitoring could check."],"forward_implications":["A total of about 5 Earth masses of rock, delivered as one super-Earth, is sufficient to raise the photospheric metal abundance of a turn-off star such as Y2235 by 0.128 dex.","Most planets ingested after dynamical instability arrive with $\\tilde v_\\perp$ well below 0.4 (medians 0.05–0.29 in the paper's simulations), so full dissolution in the convective envelope is the typical outcome.","The species-independent enhancement observed in Y2235 points to an icy/water-rich super-Earth rather than an Earth-like dry body, because carbon and oxygen are not depleted relative to other metals.","Late ingestion introduces abundance scatter among coeval stars, limiting the precision of chemical tagging and potentially biasing abundance-based inferences about exoplanet compositions.","A high-resolution spectroscopic survey of M67 turn-off and upper-main-sequence stars could measure how often late dynamical instability destroys planetary systems."],"supporting_citations":[{"why":"Provides the differential M67 spectra showing Y2235 is enhanced by a constant 0.128 dex; this is the observation the paper sets out to explain.","marker":"Liu et al. (2019)"},{"why":"Identifies planet accretion as the cause of abundance differences in the HD 80606/80607 binary, the immediate precedent for the ingestion scenario.","marker":"Liu et al. (2018)"},{"why":"Gives meteor-ablation formulae used to choose the heat-transfer coefficient C_H = 0.01 in the planet dissolution model.","marker":"Brykina (2018)"},{"why":"Shows that in systems destabilized by outer giant planets, 40 percent of lost super-Earths collide with the star, motivating the scattering delivery channel.","marker":"Mustill et al. (2017)"},{"why":"Provides the mercury N-body code used to simulate Lidov-Kozai and planet-planet scattering delivery to the stellar surface.","marker":"Chambers (1999)"},{"why":"Describes the stellar evolution code used to compute the 1.18 solar-mass model and its 3.45e-3 solar-mass convective envelope.","marker":"Pols et al. (1995)"},{"why":"Supplies the solar metal mass fraction Z_0 = 0.013 used to convert the 0.128 dex offset into the required 5.2 Earth masses of accreted metals.","marker":"Asplund et al. (2009)"},{"why":"Simulates stellar encounters in birth clusters and informs the claim that 2 au binary perturbers are rare while 5 au Kozai forcing is plausible.","marker":"Malmberg et al. (2011)"}],"fun_headline_variants":["Super-Earth meal leaves star metal-rich","Star's metal excess traced to super-Earth snack","Grazing super-Earth impact explains stellar anomaly","Swallowed super-Earth boosts star's metals"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole dissolution criterion assumes that one percent of the drag energy goes into vaporizing the planet (C_H = 0.01); if the true fraction is much smaller, a grazing planet could survive passage through the convective zone and deposit its metals deeper, where they would not be observed.","fun_headline_variants_meta":{"raw":{"variants":["Super-Earth meal leaves star metal-rich","Star's metal excess traced to super-Earth snack","Grazing super-Earth impact explains stellar anomaly","Swallowed super-Earth boosts star's metals"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000257,"raw_usage":{"total_tokens":1662,"prompt_tokens":1111,"completion_tokens":551,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":727,"completion_tokens_details":{"reasoning_tokens":491}},"tokens_in":727,"tokens_out":551,"duration_ms":5577,"temperature":1.0,"reasoning_tokens":491,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T12:29:31.390935+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the heat-transfer coefficient $C_H$ for iron-rich bodies at Mach numbers near 50 in tenuous gas through meteor-ablation observations or shock-tube experiments: if $C_H$ is much smaller than 0.005, the dissolution boundary shifts from $\\tilde v_\\perp\\lesssim0.4$ toward far more grazing orbits, so typical scattering and Kozai impacts would no longer dissolve in the convective zone and the explanation for Y2235 would fail.","supporting_citations":[{"cited_title":"J., Mel \\'e ndez J., Ram \\' rez I., Lin J., 2018, @doi [ ] 10.1051/0004-6361/201832701 , https://ui.adsabs.harvard.edu/abs/2018A&A...614A.138L 614, A138","cited_arxiv_id":null,"evidence_quote":"Identifies planet accretion as the cause of abundance differences in the HD 80606/80607 binary, the immediate precedent for the ingestion scenario."},{"cited_title":"R., Tout C","cited_arxiv_id":null,"evidence_quote":"Describes the stellar evolution code used to compute the 1.18 solar-mass model and its 3.45e-3 solar-mass convective envelope."},{"cited_title":"B., Heggie D","cited_arxiv_id":null,"evidence_quote":"Simulates stellar encounters in birth clusters and informs the claim that 2 au binary perturbers are rare while 5 au Kozai forcing is plausible."}],"review_version":1}