{"id":"8d55bb52-233a-4efa-89b1-13d594d545e8","arxiv_id":"2508.16513","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"Six new solar twins with lithium below about 0.5 dex at ages under 5 Gyr form an anomalous population; early episodic accretion is favored over planet engulfment or blue straggler formation.","lead":"Six Sun-like stars with extremely low lithium were found, far below what stellar mixing models predict for their ages. The authors argue that bursty accretion in the first million years is the most plausible cause, but only two of the stars have beryllium data to test it.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Six new 'solar twins' have Teff outside the paper's own 5777±100 K definition (e.g., HIP 91700 at 5533 K); Li depletion is strongly mass/Teff dependent, so low Li may reflect lower-mass stars, not anomalous depletion.","rationale":"The reader's weakest assumption (comparability of the 7/150 incidence to Baraffe et al. 2017a) concerns the preferred scenario, but a more fundamental issue undermines the central observational claim itself. The paper defines solar twins with Teff = 5777 ± 100 K, yet all six newly reported stars fall outside this interval, with Teff differences up to −244 K. Since lithium depletion scales strongly with mass/Teff, these cooler, lower-mass stars may naturally have low A(Li) without any anomalous mechanism. The paper does not compare against standard models for the individual stellar parameters; it only compares to the solar-twin Li-age relation, which is inappropriate for non-solar twins. This is an internal inconsistency with the paper's own definition, not a matter of outside consensus. If the sample is not actually solar twins, the title, abstract, and core conclusion 'cannot be explained by standard post-ZAMS lithium depletion' lack support. The paper could be revised by recomputing expected Li for the actual masses/Teff and reframing the sample as solar analogs, but as written the central claim is unsupported. The reader's condition on Be data and model fractions is secondary; the Teff/mass issue must be addressed first.","tokens_in":41296,"tokens_out":9782,"duration_ms":111674,"concrete_test":"Recompute the expected lithium depletion using standard stellar evolution models (e.g., MESA or YREC) for the exact masses, metallicities, and isochronal ages of each star from Table 2 (masses 0.93–1.08 M_sun, [Fe/H] ~0 to +0.14, ages 2.7–5.3 Gyr), with only standard mixing (convection, possibly atomic diffusion, no extra mixing). For each star, compare the predicted A(Li) with the observed values (Table 4). If any standard-model prediction is below the observed A(Li), the star is not anomalously Li-poor at its mass/Teff; if all predictions are well above 0.5 dex, the anomaly is real despite the non-solar-twin Teff. This directly tests whether the central claim survives the sample definition issue.","verdict_should_be":"REJECT","load_bearing_attack":"The paper's central claim depends on the sample being solar twins. Section 1 explicitly adopts the Ramírez et al. (2014) definition: Teff = 5777 ± 100 K, log g = 4.44 ± 0.10, [Fe/H] = 0.00 ± 0.10. Yet Table 2 lists Teff for all six newly reported stars outside that range: HIP 53087 (5633 K), HIP 91700 (5533 K), HIP 93858 (5671 K), HIP 116937 (5659 K), HD 221103 (5884 K), HD 236254 (5632 K). Only the previously known HIP 8522 (5729 K) qualifies. The corresponding isochrone masses are lower for the cooler stars (e.g., 0.93 M_sun for HIP 91700). Li depletion during the pre-main sequence is highly sensitive to mass and Teff: a 0.93 M_sun star has a deeper convective envelope and burns lithium more efficiently than a 1 M_sun star. The paper compares the targets only to the solar-twin Li-age relation (Fig. 3) and not to standard models computed at the individual masses and Teff. Therefore, the observed A(Li) ~0.4–0.6 dex may be perfectly compatible with standard (or mildly non-standard) models for these evolved late-G/early-K stars. This does not require episodic accretion or mergers. The reader's identified weakness (the 7/150 versus 3% incidence) is secondary because it only affects the plausibility of the preferred scenario, not the existence of the anomaly. Our concern attacks the anomaly itself.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports six newly identified low-Li stars from the Inti solar-twin survey, plus the previously known HIP 8522, giving a total of seven extremely Li-poor objects. Using high-resolution spectra from HARPS, HDS, TS23, and MIKE, the authors derive stellar parameters, ages, masses, chemical abundances, and 3D NLTE Li abundances. They then test three Li-depletion scenarios: planet engulfment with MESA simulations, blue-straggler formation via radial-velocity/SED companion searches and chemical-clock age discrepancies, and early episodic accretion using the Baraffe et al. (2017a) model grid. They conclude that early episodic accretion is the most likely explanation, with HIP 91700 possibly explained by planet engulfment, and report two new exoplanet candidates around HIP 53087 and HIP 93858.","tokens_in":41827,"tokens_out":9125,"duration_ms":96441,"significance":"If the Li anomalies are genuine, this is a valuable sample that challenges standard and non-standard Li-depletion models for Sun-like stars and provides a population-level test of episodic accretion. The observational analysis is careful: differential abundance analysis relative to the Sun, NLTE corrections, independent checks of stellar parameters, a thorough SED/RV companion search, and falsifiable Be predictions for the remaining stars. However, the interpretation currently rests on several load-bearing assumptions that are not fully supported: most targets do not satisfy the paper's own solar-twin definition, the incidence comparison with episodic-accretion models lacks statistical weight, and the Be discriminator is available for only two of seven stars. The significance of the paper is therefore conditional on a reanalysis that establishes the anomaly at the individual stellar parameters.","major_comments":[{"comment":"The six newly reported stars do not satisfy the paper's own solar-twin Teff criterion (5777±100 K): HIP 53087 (5633 K), HIP 91700 (5533 K), HIP 93858 (5671 K), HIP 116937 (5659 K), HD 221103 (5884 K), and HD 236254 (5632 K). Only HIP 8522 (5729 K) lies inside the window. Because PMS Li depletion is strongly mass- and Teff-dependent, the Fig. 3 comparison against the Inti solar-twin Li-age relation is not an adequate null model: a 0.93 M_sun star at 5533 K is expected to burn more Li during the PMS than a 1.0 M_sun twin. The central claim of an unexplained anomaly requires computing standard-model Li expectations at the individual masses and effective temperatures, or at least for the most deviant cases. Without this, the statement that these stars challenge non-standard mixing models is not demonstrated for the six new objects.","section":"§1 (footnote 18), Table 2, §6/Fig. 3"},{"comment":"The inference that 7/150 ≈ 5% is consistent with the episodic-accretion prediction of 2/60 ≈ 3% has no statistical support. A binomial 95% confidence interval for 7/150 is roughly [1.3%, 8.0%] (or similar width), so the 3% prediction is not meaningfully distinguished from a much wider range of rates. Moreover, one of the two Baraffe et al. (2017a) models with strong Li depletion has a final mass of 0.735 M_sun (Table A2 here), below the inferred mass range of this sample (0.93–1.08 M_sun, Table 2); the only in-range model (0.926 M_sun) yields partial depletion. The paper should quote the effective predicted rate for solar-mass models and give a proper uncertainty interval for the observed 7/150 fraction before using this as quantitative support.","section":"§8.3, §9"},{"comment":"The Be test is the main discriminator offered in favor of episodic accretion over FBSS/merger scenarios, but it is based on only two of the seven stars (HIP 116937 and HIP 93858), and the model extension in Table A2 predicts essentially no Be depletion for the two matched cases (Be/Be0 = 0.85–0.99). With n = 2 and five stars unmeasured, the Be data are at most a consistency check, not a confirmation. The conclusion that episodic accretion is 'the most likely scenario' therefore goes beyond the evidence currently presented. I recommend either rephrasing the conclusion as provisional pending Be measurements for the remaining stars, or substantially strengthening the quantitative model-vs-observation comparison.","section":"§8.3, Fig. 8, Table A2, §9"}],"minor_comments":[{"comment":"The second exoplanet is called HIP 116937 b in the Summary but HIP 93858 b in Table 5 and in §7.2. Please correct this inconsistency and ensure the period/mass assignments match throughout.","section":"Summary, §7.2, Table 5"},{"comment":"The labels 'HD 10725' and 'HIP 38909' should be 'HIP 10725' and 'HIP 38908'.","section":"§9 and Fig. 8 caption"},{"comment":"HD 221103 is listed as 'No refractory enrichment' in Table 7, but §8.1 and Table 6 report a significant positive TC slope (10.77 ± 2.46 × 10^-5 dex/K, p = 0.010). This internal contradiction should be reconciled.","section":"Table 7 vs §8.1/Table 6"},{"comment":"The MESA engulfment masses (30, 265, 1000 M_Earth) are fit parameters tuned to reproduce the observed A(Li) and A(Fe). The text should state explicitly that these simulations are illustrative rather than predictive; the current wording in places implies the engulfment scenario is constrained by the data.","section":"§8.1, Fig. 7"}],"recommendation":"major_revision","confidential_remarks":"The dataset and abundance analysis are of high quality, and the companion search is thorough. However, the central anomaly claim needs to be re-established for stars that are cooler than the adopted solar-twin window. A revision that compares the stars to standard/non-standard models at their individual Teff and mass, replaces the incidence comparison with proper binomial uncertainties, and reconciles the internal inconsistencies could make the paper publishable. If the reanalysis shows that the low Li is compatible with standard PMS depletion at these lower masses, the central claim would not survive."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line: the measurements are careful and the six stars are likely genuinely low in Li, but the paper overstates the challenge by calling these stars solar twins. Five of the six new stars fall outside the paper's own adopted Teff window (5777±100 K), and HIP 91700 is at 5533 K with a corresponding isochrone mass of 0.93 M_sun. Since PMS Li depletion is steeply mass- and Teff-dependent, the correct test is against standard (or non-standard) models at each star's actual mass and Teff, not just against the solar-twin Li-age relation. The paper doesn't do that, so the central claim — that these require exotic mixing or episodic accretion — isn't yet established.\n\nWhat is genuinely new: six previously unreported very Li-poor stars (A(Li) 0.4–0.6 dex, some upper limits) with high-S/N, high-resolution spectra and careful differential abundances against the Sun. The parameter checks (photometric Teff, trigonometric log g) are reassuring. The exploration of planet engulfment with MESA is a reasonable attempt, though the engulfed masses are fitted, not predicted, and only HIP 91700 gives an internally consistent match. The RV work is solid enough to claim two new planet candidates, and the SED fitting rules out stellar companions.\n\nThe soft spots, in order of size:\n1. Teff/mass mismatch (above).\n2. The 3% vs 5% episodic accretion comparison: Baraffe et al. (2017) produced only 2/60 models with strong Li depletion, one at 0.735 M_sun, below the sample range. Comparing that to 7/150 without stating uncertainties on either fraction is not a quantitative test.\n3. Internal inconsistency: the Summary assigns the 766-day planet to HIP 116937, while Table 5, the RV section, and Fig. A2 are for HIP 93858. Table 7 calls HIP 93858 a confirmed planet and HIP 116937 a candidate. That needs fixing.\n4. Be is only measured for two stars, so the \"no Be depletion\" support for episodic accretion rests on two points.\n\nThe paper deserves a serious referee, but it needs major revision: recompute or at least display standard-model Li at the actual masses, fix the planet ID, and soften the incidence comparison. The six stars are worth following up with Be spectroscopy.","headline":"Careful new data, but five of the six new 'solar twins' fall outside the paper's own Teff criterion, so the claimed challenge to mixing models is not yet established.","tokens_in":42293,"tokens_out":3183,"would_cite":false,"duration_ms":35932,"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":"The paper reports six new solar twins with A(Li) below about 0.5 dex at ages below about 5 Gyr, and argues that early episodic accretion is the most likely explanation.","keywords":["lithium depletion","solar twins","episodic accretion","blue straggler stars","stellar abundances","stellar mixing","planet engulfment","beryllium abundances"],"falsifier":"Measure beryllium in the remaining five low-Li stars (HIP 53087, HIP 91700, HD 221103, HD 236254, HIP 8522). Episodic accretion predicts A(Be) near 0.85–0.99 dex, following the solar-twin trend; a clear beryllium depletion in any of these stars would falsify the episodic-accretion explanation for that star and push it toward a merger or mass-transfer origin.","tokens_in":41276,"feed_emoji":"☀️","tokens_out":10634,"duration_ms":112332,"temperature":0.7,"pith_summary":"Six newly found solar twins have almost no lithium left in their atmospheres (A(Li) ≲ 0.5 dex) even though they are only 3–5 Gyr old, far younger than the age at which standard and current non-standard mixing models would remove lithium from Sun-like stars. The paper argues these stars, together with one previously reported anomalous twin, form a distinct population: about 5% of the 150 solar twins surveyed show this extreme depletion. After simulating planet engulfment, inspecting radial velocities for hidden companions, and weighing blue straggler scenarios, the authors conclude that early episodic accretion—violent bursts of mass falling onto the star during its first few million years—is the most likely cause. The reason this matters is that lithium-poor solar twins could become an observable fingerprint of a rare, violent accretion history that leaves no other trace in an otherwise ordinary Sun-like star.","feed_headline":"Episodic accretion strips lithium from young solar twins","feed_subtitle":"Seven of 150 Sun-like stars ran out of lithium early; the count matches bursty star-formation models, not later mixing.","key_machinery":"The load-bearing comparison is an incidence ratio: in the episodic-accretion grid, two of sixty models (about 3%) show strong lithium depletion by a few Myr, and seven of the roughly 150 solar twins in the parent sample (about 5%) show A(Li) below ~0.5 dex. The second tracer is beryllium: lithium burns near 2.5 million K while beryllium survives to about 3.5 million K, so the episodic-accretion models predict strong Li loss with Be intact, matching the two stars with measured Be. Lithium itself is measured by spectral synthesis of the 6707.8 Å Li I feature with 3D non-LTE corrections; ages come from isochrones and are cross-checked with chemical-clock and activity relations. The planet-engul","core_discovery":"The authors identify six previously unknown solar twins—stars with Sun-like temperature, gravity, and metallicity—whose lithium abundance A(Li) is at or below roughly 0.5 dex at ages between roughly 3 and 5 Gyr. Standard stellar models predict these stars should still hold near-primordial lithium, around 3 dex, and known non-standard mixing models do not remove enough lithium this early. Together with the previously reported anomalous twin HIP 8522, the six make seven such objects in a parent sample of 150 solar twins. After ruling out unresolved stellar companions via spectral energy distributions and radial velocities, the authors test three depletion channels. Planet engulfment simulation","pith_inferences":["A natural next step would be to compare the full predicted lithium-abundance distribution from the episodic-accretion models against the survey, rather than only the strong-depletion tail; the paper's incidence match uses two small counts (2/60 models vs 7/150 stars).","The same lithium–beryllium diagnostic could be applied to lithium-depleted halo and thick-disk stars below the Spite plateau; any with intact Be would suggest episodic accretion operates beyond solar twins and is not limited to blue-straggler formation.","Because the two strongly depleted models in the grid include a 0.735-solar-mass case below the solar-twin mass range, extending the grid to 0.9–1.1 solar masses at solar metallicity would directly test whether the predicted ~3% incidence holds for the stars actually being compared.","If confirmed, these stars would transform lithium measurements into a field-star census of protostellar episodic accretion, complementing the luminosity-based statistics from embedded young stellar objects."],"forward_implications":["Current non-standard mixing models that act only after the zero-age main sequence cannot account for A(Li) below 0.5 dex in stars younger than 5 Gyr; the low-Li twins require a pre-main-sequence depletion episode.","If episodic accretion is the mechanism, the remaining unmeasured twins should have normal beryllium, matching the two stars measured; a Be-depleted outlier would point back to a merger or mass-transfer origin.","The match between the model-predicted ~3% and observed ~5% incidence means strongly Li-depleting episodic accretion is rare but not exotic, occurring in a few percent of solar-mass stars.","The radial-velocity campaign rules out stellar companions and instead finds planetary-mass companions around two low-Li stars, so these systems retain planets despite the history that destroyed their lithium."],"supporting_citations":[{"why":"Supplies the episodic-accretion models whose ~3% strong-lithium-depletion rate is compared with the observed 7/150 incidence.","marker":"I. Baraffe et al. (2017a)"},{"why":"Reported the first anomalous low-Li solar twin HIP 8522 and the lithium synthesis methodology that this sample extends.","marker":"J. Yana Galarza et al. (2025)"},{"why":"Defines the solar-twin catalog and parent sample of roughly 150 stars from which the six new stars were selected.","marker":"J. Yana Galarza et al. (2021c)"},{"why":"Supplies the beryllium abundances of two sample stars and the solar-twin Be-age trend used to test episodic accretion.","marker":"H. Reggiani et al. (2025)"},{"why":"Provides the planet-engulfment simulation prescription with thermohaline mixing used for the four engulfment candidates.","marker":"J. Sevilla et al. (2022b)"},{"why":"Provides the 3D non-LTE corrections applied to the measured lithium abundances.","marker":"E. X. Wang et al. (2021)"},{"why":"Reports the field blue straggler HIP 38908 and the merger-based interpretation against which the new stars are compared.","marker":"A. Rathsam et al. (2025)"},{"why":"Reports the low-lithium solar analog HIP 10725, the standard example of a field blue straggler formed by mass transfer.","marker":"L. Schirbel et al. (2015)"}],"fun_headline_variants":["Seven solar twins defy lithium depletion models","Episodic accretion explains lithium-poor solar twins","Seven Sun-like stars ran out of lithium early","Lithium mystery: solar twins lost more than models allow","Early accretion, not later mixing, stripped these solar twins"],"cache_read_input_tokens":2688,"weakest_assumption_plain":"The episodic-accretion conclusion assumes the 150-star survey is an unbiased census of solar twins and that the small grid of episodic-accretion models, with only two strongly depleted cases (one below the solar-twin mass range), can be compared directly as a predicted ~3% incidence; if either assumption fails, the quantitative support for episodic accretion over mergers loses its footing.","fun_headline_variants_meta":{"raw":{"variants":["Seven solar twins defy lithium depletion models","Episodic accretion explains lithium-poor solar twins","Seven Sun-like stars ran out of lithium early","Lithium mystery: solar twins lost more than models allow","Early accretion, not later mixing, stripped these solar twins"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000801,"raw_usage":{"total_tokens":3392,"prompt_tokens":812,"completion_tokens":2580,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":556,"completion_tokens_details":{"reasoning_tokens":2506}},"tokens_in":556,"tokens_out":2580,"duration_ms":21597,"temperature":1.0,"reasoning_tokens":2506,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-05T17:15:10.287453+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure beryllium in the remaining five low-Li stars (HIP 53087, HIP 91700, HD 221103, HD 236254, HIP 8522). Episodic accretion predicts A(Be) near 0.85–0.99 dex, following the solar-twin trend; a clear beryllium depletion in any of these stars would falsify the episodic-accretion explanation for that star and push it toward a merger or mass-transfer origin.","supporting_citations":[{"cited_title":"2025, ApJ, 984, 108, doi: 10.3847/1538-4357/adc43e","cited_arxiv_id":null,"evidence_quote":"Supplies the beryllium abundances of two sample stars and the solar-twin Be-age trend used to test episodic accretion."},{"cited_title":"I., et al","cited_arxiv_id":null,"evidence_quote":"Reports the low-lithium solar analog HIP 10725, the standard example of a field blue straggler formed by mass transfer."}],"review_version":1}