{"id":"180af6c2-b7c5-4188-bc41-76a5c72566dd","arxiv_id":"2505.02993","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"Very massive stars produce detectable HeII 1640 emission in integrated starburst light down to 0.1 solar metallicity, with detectability at 0.01 solar metallicity depending on the assumed wind mass loss.","lead":"What happens to the heaviest stars, over 150 times the Sun's mass, when they form in gas with very few heavy elements? This paper models their evolution and spectra and shows they leave a specific ultraviolet helium line in the combined light of young starbursts down to very low metallicity.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"VMS HeII 1640 claim rests on an unbracketed mass-loss transition scenario; a steeper effective Z-scaling or a metallicity-dependent thin/thick wind transition could erase the feature at 0.1 Zsun.","rationale":"The reader correctly identified the mass-loss prescriptions and the fixed thin/thick transition as the weakest assumption. My stress-test agrees and sharpens it: the two adopted x values do not bracket the physically motivated scenario in which the transition itself shifts with metallicity, as predicted by Sabhahit et al. (2023). This is load-bearing because the HeII 1640 visibility at 0.1 Zsun in the scaled case is already marginal in the paper's own models (EW near zero at late ages), so a modest additional wind reduction could push the integrated feature below detectability. The concern is not an internal inconsistency; the paper's models are internally coherent and the comparison to II Zw 40-A, MrK71-A, and SB 126 is a genuine strength. It is also not an ad hominem or a disagreement with consensus: the paper itself states that no empirical constraint exists below LMC metallicity and that it cannot determine which framework is favoured. The appropriate verdict remains CONDITIONAL, exactly as the reader concluded: the scientific core is defensible, but the quantitative and general claims should be used with caution until the mass-loss transition behaviour at low Z is constrained. My read does not change the reader's verdict, so I recommend UNCHANGED.","tokens_in":38540,"tokens_out":8688,"duration_ms":87452,"concrete_test":"Recompute the 150-300 Msun evolutionary tracks at Z=0.1 and 0.01 Zsun using the Sabhahit et al. (2023) prescription for the metallicity-dependent thin/thick wind transition (or, as a simpler proxy, adopt x=1.5 and x=2.0 in Eq. 2), then recompute CMFGEN spectra and population-synthesis EW(HeII 1640) for burst and CSF models. If EW(HeII 1640) at Z=0.1 Zsun falls below a detection threshold (e.g., <0.5 Å) for more than a negligible fraction of the SFH parameter space, the headline claim 'down to 0.1 Zsun whatever the star formation history' fails.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that HeII 1640 remains visible in integrated light down to 0.1 Zsun 'whatever the star formation history' is only tested for two mass-loss prescriptions: no Z-scaling (x=0) and a linear Z-scaling (x=1) of the optically thick wind rates, with the thin/thick wind transition fixed at the LMC calibration (Eq. 2 and Sect. 2.1). The paper explicitly notes (Sect. 5.3) that Sabhahit et al. (2023) predict the transition to thick winds moves to higher luminosity at lower metallicity, which would reduce VMS winds more strongly than a linear scaling alone. This scenario is not bracketed by the two adopted cases: x=1 at Z=0.1 gives a wind reduction factor of about 0.25 (if Z_LMC≈0.4 Zsun), whereas a shifted transition could suppress thick winds entirely for much of the main sequence, causing an even earlier redward evolution and a weaker or absent HeII 1640. Since the visibility of HeII 1640 at 0.1 Zsun in the scaled case already relies on the hot early phases (ages ≲1 Myr; EW drops to near zero by 2.5 Myr in Fig. 7), any additional wind reduction would shorten that hot phase and could make the integrated feature undetectable. The paper's honest admission that it cannot identify the favoured mass-loss framework does not remove the risk: the claim's generality depends on the true behaviour lying within the x=0 to x=1 envelope, which is not guaranteed by any empirical or theoretical constraint below LMC metallicity.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"Martins et al. compute evolutionary tracks for very massive stars (VMS) of 150-300 Msun at Z=0.2, 0.1 and 0.01 Zsun with the STAREVOL code, using two ad hoc metallicity scalings for optically thick VMS winds (no Z dependence, x=0, and linear Z dependence, x=1, in Eq. 2). They then compute CMFGEN non-LTE atmosphere models at selected ages along the tracks and include the resulting spectra in BPASS-based population synthesis models. The paper finds that VMS with Z-independent mass loss remain hot and near the ZAMS throughout their evolution, whereas VMS with linearly scaled mass loss evolve redward. VMS show HeII 1640 emission in most individual phases except when they become too cool, and this feature appears in integrated light down to 0.1 Zsun under most of the investigated configurations, though it is weaker or absent in the scaled mass-loss case at late ages. The models are compared to HST UV spectra of II Zw 40-A, MrK71-A and SB126, yielding qualitative and sometimes quantitative matches without clearly favouring either mass-loss prescription.","tokens_in":38860,"tokens_out":6903,"duration_ms":71576,"significance":"The paper provides a consistent and useful extension of VMS evolution, atmosphere and population-synthesis modelling to sub-LMC metallicities. Its central predictions, especially the behaviour of HeII 1640 and the increase in ionizing photons below about 45 eV when VMS are included, are falsifiable and can be confronted with existing and future UV spectroscopy of low-metallicity starbursts. A clear strength is the comparison with independent HST observations of three star-forming regions rather than a fit to the data used to calibrate the models. The authors are also transparent about the exploratory nature of the mass-loss assumptions and about the fact that their data cannot identify the preferred framework. The synthetic spectra are made publicly available on the POLLUX database, which aids reproducibility.","major_comments":[{"comment":"The Abstract's claim that HeII 1640 is present in integrated light down to 0.1 Zsun 'whatever the star formation history' is not established for both adopted mass-loss prescriptions simultaneously. In the models with Z-scaled VMS mass loss, Sect. 4.2 states that HeII 1640 disappears at 2.5 Myr, and Sect. 4.3 reports that scaled-mass-loss burst models have almost no EW(HeII 1640) below Z=0.2 Zsun. The conclusion should be rephrased to make explicit which mass-loss framework, age range and detection criterion support the 'presence' claim, or the analysis should be repeated with a quantitative line-detection threshold rather than visual inspection.","section":"Abstract; Sect. 2.1, Eq. (2); Sect. 5.3"},{"comment":"The two adopted wind prescriptions do not bracket the scenario in which the optically thin/thick wind transition itself shifts to higher luminosity at lower metallicity, as predicted by Sabhahit et al. (2023) and acknowledged by the authors in Sect. 5.3. If the transition moves in that direction, optically thick winds can be suppressed over much of the main sequence even for a linear Mdot(Z) scaling, leading to an earlier redward evolution than in the x=1 tracks. Because the HeII 1640 visibility of the integrated x=1 models already relies on the hottest early phases (the EW drops to near zero at ages near 2.5 Myr in Fig. 7, and to small values in CSF models after roughly 3 Myr), this unbracketed scenario could remove the feature at 0.1 Zsun. Including a third set of models with a metallicity-dependent thin/thick transition, or explicitly limiting the headline claim to the x=0 and x=1 frameworks, is necessary for the stated generality.","section":"Sect. 2.1, Eq. (2); Sect. 5.3"},{"comment":"The paper uses the wording 'a weak emission is detected' for models where EW(HeII 1640) is close to zero after absorption lines over the 1625-1655 A window compensate the emission. This makes the central 'presence' claim difficult to evaluate quantitatively. I recommend defining a detection threshold based on line flux or on an EW significance criterion, so that statements such as 'present' or 'vanishes' are unambiguous across the different models and ages.","section":"Sect. 4.3; Fig. 8"}],"minor_comments":[{"comment":"The caption gives 'Z=0.001 Zsun' for the orange lines, while the main text and Fig. 20 description refer to Z=0.01 Zsun; this appears to be a typo.","section":"Fig. 20 caption"},{"comment":"In the Z=0.01 Zsun scaled-mass-loss table, the 150 Msun model row at 2.50 Myr appears to have a malformed hydrogen abundance entry; the H and He columns should be checked.","section":"Table A.6"},{"comment":"For ages 0 and 0.5 Myr the BPASS 1 Myr model is used for the normal-star population; this approximation is acknowledged in Sect. 4.6.2 for MrK71-A, but its systematic effect on the youngest burst comparisons in Figs. 15 and 16 is not quantified.","section":"Sect. 4.1"},{"comment":"The phrase 'whatever the star formation history' is stronger than what is computed; the authors consider one burst and one constant star-formation history built from discrete age bins, so a formulation such as 'for both the burst and constant star-formation histories considered here' would be more precise.","section":"Abstract"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is within the scope of A&A and reports a careful, internally consistent modelling effort. The main concern is not the quality of the calculations but the robustness of the headline claim: the two mass-loss prescriptions bracket a plausible range of VMS wind scalings, but they do not bracket the physically motivated scenario in which the thin/thick wind transition itself depends on metallicity. The authors already cite this scenario in Sect. 5.3, so adding a test model or tightening the conclusion should be feasible within a revision. I therefore recommend major revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a workmanlike extension of Martins & Palacios's VMS modeling program to 0.2, 0.1, and 0.01 Zsun. The main result is that HeII 1640 emission from very massive stars remains visible in integrated starburst light down to 0.1 Zsun under both of the paper's mass-loss prescriptions, and the models actually do a decent job matching the UV spectra of II Zw 40-A, MrK71-A, and SB 126. That is a genuinely useful result for anyone interpreting JWST-era rest-frame UV spectra.\n\nWhat is new: new evolutionary tracks for 150-300 Msun at three low metallicities with two mass-loss scalings, corresponding CMFGEN synthetic spectra, and population synthesis including VMS. The ionizing photon comparison below 45 eV is a solid quantitative addition over the earlier Z=0.4 Zsun work. The pipeline is described clearly and the internal consistency between evolution and atmospheres is a real strength. The paper is also honest that it cannot decide between the two mass-loss frameworks.\n\nThe soft spot is the mass-loss envelope. The two recipes bracket the optically thick wind scaling with x=0 and x=1, but they keep the thin-to-thick transition fixed at the LMC calibration. If, as Sabhahit et al. (2023) predict, the transition moves to higher luminosity at lower metallicity, the true wind reduction is steeper than the linear case, and the hot main-sequence phase that powers HeII 1640 in the scaled models would shorten. That scenario is not bracketed by the x=0–1 envelope. So the paper's broad statement that HeII 1640 survives down to 0.1 Zsun 'whatever the star formation history' is only as secure as the assumption that the real behavior lies between these two prescriptions. The authors acknowledge the framework uncertainty, but the conclusion is worded more strongly than the bracket justifies.\n\nAlso minor: no uncertainty estimates on the predicted EWs and ionizing fluxes, and only the synthetic spectra are public; evolutionary tracks and population models come 'on reasonable request,' which is a step short of reproducible. One of the three observed targets, MrK71-A, is not quantitatively matched, so the observational support is partly qualitative.\n\nBottom line: this is a serious paper that deserves a proper referee. I would send it out with the expectation of a revision that softens the generalization and ideally releases the population models. I'd cite it if I worked on starbursts.","headline":"The paper convincingly shows HeII 1640 from VMS is a plausible low-metallicity tracer, but the 'survives down to 0.1 Zsun whatever the SFH' claim is bracketed by two mass-loss recipes that may not cover the real metallicity behavior.","tokens_in":39511,"tokens_out":3832,"would_cite":true,"duration_ms":35253,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper argues that very massive stars at low metallicity retain a distinctive He II 1640 emission line in the integrated light of starbursts down to 0.1 solar metallicity, and that adding them hardens the ionizing spectrum below about…","keywords":["very massive stars","mass loss","stellar evolution","synthetic spectra","population synthesis","He II 1640 emission","starbursts","low metallicity"],"falsifier":"Observe a very young (≲1 Myr) massive cluster at Z ≈ 0.01 solar metallicity, for example a lensed low-metallicity star-forming galaxy that can be resolved into a compact cluster, and measure whether He II 1640 appears in emission in its integrated UV spectrum. A strong detection would rule out a linear or steeper metallicity scaling of VMS winds, because in those models the stars are too cool and their winds too weak to produce the line; a clean upper limit would rule out metallicity-independent winds, because those models keep the line strong at all ages.","tokens_in":38303,"feed_emoji":"🔭","tokens_out":10781,"duration_ms":106121,"temperature":0.7,"pith_summary":"Very massive stars (VMS, stars with initial masses above roughly 100 solar masses) are securely known only in the Milky Way and the Large Magellanic Cloud, and nothing is empirically known about their winds at lower metallicity. This paper predicts their evolution and spectra at 0.2, 0.1 and 0.01 solar metallicity under two bracketing assumptions for wind mass loss: no metallicity dependence, or a linear scaling with metallicity. The central finding is that the helium line He II 1640 stays in emission through most of a VMS lifetime and survives in the integrated light of population synthesis models at least down to 0.1 solar metallicity under either wind assumption, making it a usable signature of very massive stars in unresolved low-metallicity starbursts. Adding VMS to the models also raises the number of ionizing photons below about 45 eV and increases the ionizing photon efficiency, which eases the difficulty standard models have in reaching the high efficiencies measured in some star-forming galaxies. The paper is explicit that the available ultraviolet spectra do not yet identify which wind scaling is correct.","feed_headline":"He II 1640 emission survives to 0.1 solar metallicity","feed_subtitle":"Even at one-tenth solar metallicity, this helium line keeps identifying very massive stars in unresolved starbursts","key_machinery":"The machinery is a two-branch wind mass-loss recipe anchored on the empirical VMS calibration of Gräfener (2021): an optically thin branch and an optically thick branch, with the thick branch scaled by $(Z/Z_{\\rm LMC})^x$ for $x=0$ or $x=1$. Stellar evolution tracks for initial masses 150, 200, 250 and 300 $M_\\odot$ at $Z=0.2,0.1,0.01\\,Z_\\odot$ are computed with these recipes, and CMFGEN non-local thermodynamic equilibrium atmosphere models are made at selected ages using the predicted surface abundances, so the synthetic spectra and the population synthesis built from them are internally consistent. The argument-carrying observable is the He II 1640 emission line, formed in the dense winds of hot, helium-enriched VMS; its strength and profile track the stellar temperature, wind density, and surface composition across the model grid.","core_discovery":"On the paper's own terms, the discovery is that the fate of a 150 to 300 solar mass star at low metallicity is decided by the wind mass-loss recipe. If VMS winds do not weaken with metallicity, the stars keep losing mass, stay near the zero-age main sequence for about 2.5 million years, and then evolve blueward; they remain hot, and their synthetic spectra show strong He II 1640 emission throughout. If the mass-loss rates scale linearly with metallicity, the weaker winds allow a helium II opacity bump to appear below the surface, the star inflates, and the track moves toward the red part of the Hertzsprung-Russell diagram; He II 1640 weakens and can disappear at 0.01 solar metallicity. In population synthesis, He II 1640 emission is present in integrated light at 0.2 and 0.1 solar metallicity for both burst and constant star formation histories, and also appears at 0.01 solar metallicity when winds are metallicity-independent. Adding VMS raises the number of ionizing photons below about 45 eV and boosts the ionizing photon efficiency, while the flux above 45 eV depends on age, metallicity, star formation history, and the shortest final phases of VMS evolution. The models reproduce the ultraviolet spectra of the low-metallicity starbursts II Zw 40-A, MrK71-A and SB 126 qualitatively and sometimes quantitatively, but the authors conclude that no clear choice between the two mass-loss frameworks emerges.","pith_inferences":["If the true metallicity scaling is shallower than linear but not zero, VMS at 0.01 solar metallicity should show weak but detectable He II 1640 only at the youngest ages; stacking ultraviolet spectra of many low-metallicity star-forming regions could reveal this population-average line even where individual clusters are too faint.","The two wind scenarios predict different surface nitrogen enhancement and different amounts of nitrogen ejected into the surroundings, so the nitrogen abundances of young low-metallicity systems, including the high-redshift N-emitters, may discriminate between them even when ultraviolet line ratios cannot.","The models imply that the profile of He II 1640 becomes more purely emissive, with less P-Cygni absorption, as metallicity drops; equivalent-width thresholds for VMS selection should therefore be metallicity-dependent rather than fixed.","The predictions above 45 eV rest on only a few sampled points in the final phases of VMS evolution, so the claimed nebular He II 4686 contribution is the least secure element of the hard-UV part of the model grid; a finer time sampling of the last 0.2 Myr would settle it."],"forward_implications":["He II 1640 emission in an unresolved young starburst becomes a usable VMS tracer at metallicities down to at least 0.1 solar, for both burst and constant star formation histories; no resolved spectroscopy of individual stars is needed.","At 0.01 solar metallicity, only the metallicity-independent wind family keeps the line strong, so a detection there would indicate that VMS winds scale with metallicity more weakly than linearly.","Including VMS raises the number of ionizing photons below about 45 eV and raises the ionizing photon efficiency $\\xi_{\\rm ion}$, so high efficiencies measured in starburst galaxies no longer require a top-heavy initial mass function.","The optical Wolf-Rayet bumps discriminate populations: the blue bump shows He II 4686 without N III 4634-42, and the red bump is a narrow C IV 5802-12 doublet, when VMS dominate instead of classical Wolf-Rayet stars.","The short final hot phases of VMS barely affect the ultraviolet and optical integrated spectra, but they can raise $Q({\\rm He\\,II})/Q({\\rm H\\,I})$ to a few times $10^{-3}$, enough for nebular He II 4686 at roughly one percent of H$\\beta$, yet not enough to explain the strongest observed He II emitters."],"supporting_citations":[{"why":"Supplies the empirical VMS mass-loss recipe, calibrated at LMC metallicity, whose optically thin and thick branches are the baseline for both wind frameworks.","marker":"Gräfener (2021)"},{"why":"Provides the UV spectrum of MrK71-A and the argument that VMS winds show no metallicity dependence, motivating the x=0 framework.","marker":"Smith et al. (2023)"},{"why":"Documents the roughly Z^1.2 metallicity dependence of classical WR star winds that motivates the linear Z-scaling framework.","marker":"Hainich et al. (2015)"},{"why":"Presents a theoretical framework for how the transition between normal and boosted VMS winds changes with metallicity, against which the ad hoc recipes are compared.","marker":"Sabhahit et al. (2023)"},{"why":"Describes the CMFGEN code used to compute all non-LTE expanding-atmosphere models and synthetic spectra.","marker":"Hillier & Miller (1998)"},{"why":"Supplies the BPASS population synthesis models for stars below 100 solar masses to which the VMS contributions are added.","marker":"Eldridge et al. (2017)"},{"why":"Established that VMS dominate the integrated ultraviolet light of clusters and that He II 1640 emission in R136 is produced only by VMS, making the line the key tracer.","marker":"Crowther et al. (2016)"},{"why":"Provides the method of coupling evolutionary and atmosphere models for VMS and the Z=0.4 solar models extended in this paper.","marker":"Martins & Palacios (2022)"},{"why":"Supplies the HST ultraviolet spectrum of II Zw 40-A, one of the three observed starbursts used to test the models.","marker":"Leitherer et al. (2018)"},{"why":"Supplies the HST ultraviolet spectrum of SB 126 and the claim that VMS are needed to reproduce its stellar and nebular features.","marker":"Senchyna et al. (2021)"}],"fun_headline_variants":["He II 1640 persists down to 0.1 solar metallicity","Wind scaling determines if massive stars stay hot at low Z","VMS fate splits: hot or red depending on wind loss recipe","He II 1640 emission flags VMS even at tenth solar Z","Low-metal starbursts reveal VMS via He II 1640 line"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the two adopted wind prescriptions bracket the true mass-loss behaviour of very massive stars below Large Magellanic Cloud metallicity, with the optically thin/thick wind transition left unchanged from the LMC calibration; no empirical constraint exists in that regime, and if the real metallicity dependence differs from zero or linear scaling, the evolutionary paths, the redward evolution, and the visibility of He II 1640 at 0.01 solar metallicity would change.","fun_headline_variants_meta":{"raw":{"variants":["He II 1640 persists down to 0.1 solar metallicity","Wind scaling determines if massive stars stay hot at low Z","VMS fate splits: hot or red depending on wind loss recipe","He II 1640 emission flags VMS even at tenth solar Z","Low-metal starbursts reveal VMS via He II 1640 line"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000529,"raw_usage":{"total_tokens":2705,"prompt_tokens":1256,"completion_tokens":1449,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":872,"completion_tokens_details":{"reasoning_tokens":1354}},"tokens_in":872,"tokens_out":1449,"duration_ms":12664,"temperature":1.0,"reasoning_tokens":1354,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T00:38:24.099250+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Observe a very young (≲1 Myr) massive cluster at Z ≈ 0.01 solar metallicity, for example a lensed low-metallicity star-forming galaxy that can be resolved into a compact cluster, and measure whether He II 1640 appears in emission in its integrated UV spectrum. A strong detection would rule out a linear or steeper metallicity scaling of VMS winds, because in those models the stars are too cool and their winds too weak to produce the line; a clean upper limit would rule out metallicity-independent winds, because those models keep the line strong at all ages.","supporting_citations":[{"cited_title":"& Palacios , A","cited_arxiv_id":null,"evidence_quote":"Provides the method of coupling evolutionary and atmosphere models for VMS and the Z=0.4 solar models extended in this paper."},{"cited_title":"C., & Levesque , E","cited_arxiv_id":null,"evidence_quote":"Supplies the HST ultraviolet spectrum of II Zw 40-A, one of the three observed starbursts used to test the models."}],"review_version":1}