Nitrogen rises to the top: evidence of enhanced mixing in very massive stars

Pablo Marchant; Tomer Shenar

REVIEW 3 major objections 5 minor 86 references

Nitrogen rises to the top: evidence of enhanced mixing in very massive stars

T0 review · 3 major / 5 minor · reviewed 2026-07-10 · grok-4.5

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Nearly all stars above 100 solar masses in the Tarantula are nitrogen-rich, a pattern that mass loss alone cannot produce and that requires early mixing beyond standard convective cores.

desk verdict Solid empirical result: near-100% N-enrichment above ~100 M☉ cannot be produced by published winds, SFHs or binary fractions; only high overshoot works, and that wrecks the HRD. read the letter →

arxiv 2607.06828 v1 pith:B6FN67KC submitted 2026-07-07 astro-ph.SR astro-ph.GA

Pablo Marchant , Tomer Shenar This is my paper · ORCID

classification astro-ph.SRastro-ph.GA

keywords massivestarsnitrogenenrichmentstellarmixingcoreovershootingwindsTarantulaNebulahigh-redshiftgalaxiesevolution

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Very massive stars are widely invoked to explain the strong nitrogen signatures seen in high-redshift galaxies, because their winds should eventually expose CNO-processed material. This paper tests that idea against a carefully assembled sample of 122 stars above 30 solar masses in the Tarantula Nebula. Almost every star above roughly 100 solar masses is strongly nitrogen-enriched. Standard single-star models with ordinary mass-loss recipes and ordinary core overshooting cannot reproduce so high an enriched fraction, no matter how the star-formation history or binary-product fraction is varied within plausible ranges. Only models with extreme overshooting succeed at the abundances, yet those same models force quasi-homogeneous evolution that places the stars too hot on the Hertzsprung–Russell diagram. The authors therefore conclude that an efficient mixing process must operate early—during or shortly after formation—rather than throughout the main sequence. That early mixing is missing from current evolutionary calculations, and its inclusion would reshape predictions for the structure, expansion, and final fates of the most massive stars while offering a local pathway to the rapid nitrogen enrichment observed at high redshift.

What carries the argument

Bayesian comparison of the luminosity-dependent nitrogen-rich fraction measured in the Tarantula sample against population-synthesis predictions that combine single-star tracks (varying mass-loss recipe and constant step-overshoot parameter alpha_ov) with a free binary-product fraction f_bin.prod.

What would settle it

A complete census of nitrogen abundances and effective temperatures for stars above 100 solar masses that either (a) reveals a substantial population of unevolved, nitrogen-normal objects at the highest luminosities or (b) shows that models with early-time-only mixing simultaneously match both the enrichment fraction and the observed cool locations on the Hertzsprung–Russell diagram.

Watch

Extended reading notes

Core claim

The observed fraction of nitrogen-rich stars with masses greater than or equal to about 100 solar masses cannot be explained by mass loss alone; it requires significantly more efficient mixing beyond the convective core than is present in current evolutionary models. Enhanced step overshooting of order unity can match the enrichment statistics, but produces chemically homogeneous tracks inconsistent with the observed Hertzsprung–Russell diagram, pointing instead to efficient early mixing that is not yet included in the models.

Load-bearing premise

That a single, luminosity-independent binary-product fraction plus a constant overshooting parameter capture all the non-single-star and mixing physics, so any residual enrichment must be attributed to early mixing.

Editorial extensions

If this is right

Standard evolutionary tracks under-predict surface nitrogen for the most massive stars and must incorporate an early-mixing channel not presently included.
Larger helium cores and restricted radial expansion would alter the production of Wolf–Rayet stars, the location of the Humphreys–Davidson limit, and the orbital separations at which Roche-lobe overflow occurs.
Very massive stars with early mixing become a stronger local analogue for the rapid nitrogen enrichment observed in high-redshift galaxies.
Final compact-object masses and the yields returned to the interstellar medium at the upper mass end would shift once early mixing is accounted for.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

If the required early mixing is metallicity-independent, the same process should operate in metal-poor high-redshift environments and could be tested with abundance patterns of local extremely metal-poor massive stars.
A time-dependent mixing efficiency that is strong only near the zero-age main sequence would simultaneously solve the abundance and Hertzsprung–Russell diagram tensions and could be constrained by asteroseismology of intermediate-mass analogues.
The same early-mixing physics would raise the minimum mass for pair-instability supernovae and change the expected black-hole mass spectrum from the most massive stars.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

The new thing here is clean and quantitative. Marchant & Shenar assemble a luminosity-complete sample of 122 stars in 30 Dor with N abundances and show that essentially every object above ~10^6 L☉ is strongly nitrogen-rich. Standard single-star tracks with three different wind recipes (Brott, Brands, Sabhahit), two SFHs, and a free binary-product fraction all under-predict that fraction. Only step overshoot α_ov ≳ 1 recovers the enrichment, but those tracks stay too hot and produce quasi-homogeneous evolution that is ruled out by the observed HRD. That tension is real and well documented.

What they do well: the Bayesian construction of P_N-rich(L), the exhaustive sensitivity tests (WR cut at X_H > 0.4, mass-loss boosts imes2 and imes3, both SFH priors), and the public Zenodo deposit of code, tracks and samples. The appendices make the result reproducible. They also correctly note that sustained high overshoot is inconsistent with the HRD and therefore float the idea of early, time-limited mixing. That is an honest framing rather than a forced solution.

The soft spot is exactly the one the reader flagged: the residual enrichment is absorbed into a single luminosity-independent f_bin.prod plus a constant α_ov. That is phenomenological. Extreme wind boosts can also match the enrichment numbers, but they contradict both theory and the HRD in the same way high overshoot does, so the authors rightly discard them. The central observational claim survives those tests; the physical agent remains unspecified. That is a limitation of the current models, not a flaw in the paper.

This is for anyone working on very massive stars, WR formation, the Humphreys-Davidson limit, or high-z N/O ratios. The math and data look solid; the citation pattern is appropriate. I would send it to referees without hesitation and would cite the enrichment discrepancy myself. The early-mixing suggestion is a useful prompt for the next generation of models, not a finished theory.

Referee Report

3 major / 5 minor

Summary. The paper compiles a sample of 122 massive stars (M ≳ 30 M⊙) in 30 Doradus with nitrogen abundance constraints and shows that the observed fraction of N-rich stars rises sharply to near 100% above ~10^6 L⊙ (M ≳ 100 M⊙). Using MESA tracks with three wind prescriptions (Brott et al. 2011; Brands et al. 2022; Sabhahit et al. 2022), two star-formation histories, and a Bayesian model that includes a luminosity-independent binary-product fraction f_bin.prod, the authors demonstrate that standard step overshooting (α_ov = 0.335) under-predicts the high-luminosity enrichment. Fitting α_ov and f_bin.prod yields α_ov ≳ 1, which reproduces the enrichment but produces quasi-chemically homogeneous tracks inconsistent with the observed HRD. Extreme mass-loss boosts can also match the enrichment but conflict with both the HRD and current wind constraints. The authors therefore argue for efficient early mixing (during or shortly after formation) that is not present in current evolutionary models, with implications for high-redshift N enrichment and the upper-mass end.

Significance. If correct, the result is a clear empirical requirement that present single-star models under-mix the envelopes of very massive stars. The discrepancy survives three wind recipes, two SFHs, a hydrogen-cut sample that removes possible classical WR contaminants (Appendix E), and mass-loss boosts of factors 2–3 (Appendix F). The public Zenodo deposit of code, tracks and MCMC samples makes the inference reproducible. The paper is careful to flag the HRD tension of sustained high overshooting and to propose time-limited early mixing as a possible resolution. The finding therefore supplies a concrete local-Universe anchor for interpreting rapid nitrogen enrichment at high redshift and motivates the development of physically motivated early-mixing prescriptions.

major comments (3)

Appendix C and Eq. (C.2): the residual enrichment is absorbed into a single luminosity-independent f_bin.prod. Binary products (mergers, mass gainers) are expected to be more common at the highest luminosities; a luminosity-dependent f_bin.prod could therefore reduce the required α_ov. The manuscript should either demonstrate that a mass- or luminosity-dependent binary fraction cannot erase the need for α_ov ≳ 1, or quantify how large such a dependence would have to be.
Section 4 and Figs. 3–4: the central claim is that efficient early (time-limited) mixing is required. The paper shows that constant α_ov ≳ 1 produces quasi-homogeneous tracks that are too hot, but does not present any evolutionary calculation in which mixing is strong only near the ZAMS and then declines. Without at least a schematic demonstration that such a time-dependent prescription can simultaneously match both the N-enrichment fraction and the observed HRD, the proposed resolution remains an untested conjecture.
Appendix F: mass-loss rates boosted by factors of 2–3 relative to Sabhahit et al. (2022) can formally reproduce the enrichment trend with low α_ov. While the authors correctly note that such boosts conflict with recent theoretical and observational wind constraints, the argument would be stronger if they quantified how far the boosted rates lie outside the allowed range of Björklund et al. (2023) and Verhamme et al. (2026) at the relevant luminosities and temperatures, rather than relying on a qualitative statement.

minor comments (5)

Figure 1 caption and top axis: the conversion from luminosity to ZAMS mass is model-dependent; state which wind and overshoot prescription is used for that axis.
Appendix A: three objects (R136a8, H31, H47) lack reliable N lines; the decision to exclude only R136a8 while retaining the other two as 'pessimistic' should be stated more explicitly in the main text.
Table 1 and Fig. 2: the 90% HDI for α_ov under the Brands winds is systematically higher than under Brott or Sabhahit; a one-sentence physical explanation would help the reader.
Section 2: the statement that the sample is 'largely complete at high luminosities' would benefit from a brief quantitative completeness estimate (e.g., relative to Schneider et al. 2018b).
Typographical: 'Fibre Large Array Multi Element Spectrograph' should be 'Fibre Large Array Multi-Element Spectrograph'; 'Very Large Telescpoe' → 'Telescope'.

Circularity Check

1 steps flagged · score 1.0 of 10

No significant circularity: Bayesian fit of α_ov and f_bin.prod is used comparatively against independent mass-loss/SFH priors and HRD constraints, not as a tautological prediction.

fitted input called prediction [§3 / Appendix C, Eq. (C.2) and Table 1]
"we perform a Bayesian analysis to constrain the possible values of α_ov and f_bin.prod ... PN-rich(log L/L☉)=f_bin.prod+(1-f_bin.prod)×PN-rich,single(log L/L☉,α_ov). ... the results for α_ov support a value in excess of unity"

α_ov and f_bin.prod are free parameters fitted directly to the observed enrichment fractions. The high-α_ov solution is therefore a best-fit description of the same data, not an independent prediction. The paper treats it as such (a phenomenological requirement that then conflicts with the HRD), so the circularity is mild and acknowledged rather than hidden.

full rationale

The paper's central claim is that the observed near-100% nitrogen-enrichment fraction above ~10^6 L☉ cannot be reproduced by standard single-star models (fixed α_ov=0.335 from Brott et al. 2011, three independent wind recipes, two SFH priors) even after allowing a free luminosity-independent binary-product fraction f_bin.prod. The Bayesian model of Appendix C then fits α_ov and f_bin.prod to the same enrichment data; the resulting high α_ov ≳ 1 is reported as a phenomenological requirement, not as a first-principles derivation. The paper itself immediately notes the tension with the observed HRD (Figs. 3–4) and explores alternatives (early-only mixing, mergers, extreme winds in Appendix F). No equation reduces the enrichment fraction to a definitional identity of the fitted parameters, no uniqueness theorem is imported from the authors' prior work, and the mass-loss and SFH inputs are external. The single mild self-referential element is the use of Brott et al. (2011) both for the baseline α_ov and for the LMC abundance mixture; this is ordinary model inheritance, not load-bearing circularity. Score 1 reflects that minor self-citation without elevating it to a circular derivation.

Assumptions & free parameters 3 free parameters · 4 assumptions · 1 invented entities

The central claim rests on a modest set of free parameters (α_ov, f_bin.prod) fitted to the enrichment data, standard domain assumptions about IMF, convection, and mass-loss recipes drawn from the literature, and no newly invented physical entities beyond the phenomenological suggestion of early mixing.

free parameters (3)

α_ov (step overshooting) = 1.4–2.6 (90 % HDI depending on wind/SFH)
Fitted via Bayesian comparison to the observed N-rich fraction; values ≳1 are required to match data but produce homogeneous evolution inconsistent with the HRD.
f_bin.prod = 0.14–0.33 (90 % HDI)
Constant luminosity-independent fraction of binary products assumed nitrogen-rich; anti-correlated with α_ov and constrained to ≲0.5.
mass-loss boost factors = 2 and 3
Ad-hoc multiplicative factors (×2, ×3) applied to the Sabhahit et al. (2022) recipe to test whether extreme winds can remove the need for high α_ov.

assumptions (4)

domain assumption Salpeter IMF (dN/dm ∝ m^−2.35) for population synthesis
Used throughout Appendix C to convert evolutionary tracks into luminosity-dependent enrichment fractions.
ad hoc to paper Step overshooting with a single mass- and age-independent α_ov is an adequate proxy for any envelope mixing
Explicitly adopted in §3; the authors note that real mixing may be mass- or age-dependent or driven by rotation/gravity waves.
domain assumption Nitrogen-rich defined as surface 12+log(N/H) exceeding the LMC carbon baseline 7.75
Adopted from Brott et al. (2011) and used to convert abundance measurements or qualitative flags into p_N-rich probabilities (Appendix A–B).
domain assumption Mass-loss rates of Brott, Brands or Sabhahit recipes (or modest boosts thereof) bound the plausible wind physics
Central to the claim that winds alone cannot explain the data; tested in §3 and Appendix F.

invented entities (1)

efficient early mixing (time-limited, operating during/shortly after formation)
purpose: To reconcile high surface N enrichment with an HRD that is inconsistent with lifelong high overshooting
Postulated in the Discussion as the remaining viable explanation once winds, binaries and constant high α_ov are ruled out; no microphysical model is supplied.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Nitrogen rises to the top: evidence of enhanced mixing in very massive stars." pith.science (2026). https://pith.science/paper/B6FN67KC

@misc{pith2026260706828,
  author       = {Pith},
  title        = {Pith review of: Nitrogen rises to the top: evidence of enhanced mixing in very massive stars},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/B6FN67KC}},
  note         = {Machine review of arXiv:2607.06828}
}

abstract

Recent observations of young galaxies in the high-redshift Universe reveal signs of early enrichment of nitrogen. Extremely massive stars ($M \gtrsim 10^{2}-10^3\,M_\odot$) with strong stellar winds have been proposed as a potential driver of this phenomenon. Here, we show that the observed fraction of nitrogen rich stars with masses $\gtrsim 100M_\odot$ cannot be explained solely by mass loss, requiring significantly more efficient mixing beyond their convective cores than accounted in present evolutionary models. We compile a representative sample of 122 stars in the Tarantula Nebula of the Large Magellanic Cloud (LMC) with masses $M \gtrsim 30\,M_\odot$. Nearly all stars with masses $M \gtrsim 100\,M_\odot$ exhibit strong nitrogen enrichment, by factors $\gtrsim 5-10$. We demonstrate that this trend cannot be reproduced by varying assumptions on binary fraction, star formation history, or mass-loss rates within ranges predicted by current empirical and theoretical models. In contrast, enhanced core overshooting of $\alpha_\text{ov}\gtrsim 1$ can account for the observed enrichment, but leads to quasi-chemically homogeneous evolution that is inconsistent with the observed Hertzsprung-Russell diagram. While the origin of this discrepancy remains unclear, our results, in combination with observational and theoretical constraints on mass-loss rates, suggest the presence of efficient early mixing operating during or shortly after the formation of very massive stars. Such mixing models are currently not included in stellar evolution models. These findings have immediate implications for the formation, radial expansion, evolution, and final fates of stars at the upper mass end, and provide a potential pathway to explaining the rapid nitrogen enrichment observed in the high-redshift Universe.

Figures

Figures reproduced from arXiv: 2607.06828 by the authors.

**Figure 1.** Observed fraction of nitrogen rich stars in 30 Doradus (gray band showing 50% and 90% highest density intervals) compared against theoretical predictions from single star models using different assumptions of wind physics with a value of overshoot of αov = 0.335 (Brott et al. 2011). Solid lines indicate a constant star formation prior, while the dashed line lines uses a star formation history based on Schneider et … view at source ↗

**Figure 2.** Posterior distributions for αov and fbin. prod. inferred by comparison to our 30 Doradus sample. Results for two star formation models (constant and following Schneider et al. 2018a) and three wind prescriptions (Brott et al. 2011; Brands et al. 2022; Sabhahit et al. 2022) are shown. Panel for the 2D distribution of shows a 90% highest density interval [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗

**Figure 3.** Location of stars in our sample indicating the probability that they are nitrogen enriched, plotted together with main-sequence evolutionary tracks with different assumptions on wind physics and a step overshoot parameter of αov = 0.335. Numbers indicate initial masses of the evolutionary tracks in M⊙, while think lines indicate the phases where the nitrogen abundance exceeds the carbon baseline of the LMC. Teff [k… view at source ↗

Figures from the paper (1 more)

**Figure 4.** Figure 4: Same as [PITH_FULL_IMAGE:figures/full_fig_p004_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

86 extracted references · 86 canonical work pages

[1]

A., Sana , H., Taylor , W., et al

Almeida , L. A., Sana , H., Taylor , W., et al. 2017, , 598, A84

work page 2017
[2]

2023, , 519, 5333

Baraffe , I., Clarke , J., Morison , A., et al. 2023, , 519, 5333

work page 2023
[3]

M., Crowther , P

Bestenlehner , J. M., Crowther , P. A., Broos , P. S., Pollock , A. M. T., & Townsley , L. K. 2022, , 510, 6133

work page 2022
[4]

M., Crowther , P

Bestenlehner , J. M., Crowther , P. A., Caballero-Nieves , S. M., et al. 2020 a , , 499, 1918

work page 2020
[5]

M., Crowther , P

Bestenlehner , J. M., Crowther , P. A., Caballero-Nieves , S. M., et al. 2020 b , , 499, 1918

work page 2020
[6]

M., Gr \"a fener , G., Vink , J

Bestenlehner , J. M., Gr \"a fener , G., Vink , J. S., et al. 2014, , 570, A38

work page 2014
[7]

O., Singh , S

Bj \"o rklund , R., Sundqvist , J. O., Singh , S. M., Puls , J., & Najarro , F. 2023, , 676, A109

work page 2023
[8]

1958, , 46, 108

B \"o hm-Vitense , E. 1958, , 46, 108

work page 1958

Show all 86 references

[9]

A., de Koter , A., Bestenlehner , J

Brands , S. A., de Koter , A., Bestenlehner , J. M., et al. 2022, , 663, A36

2022
[10]

E., Cantiello , M., et al

Brott , I., de Mink , S. E., Cantiello , M., et al. 2011, , 530, A115

2011
[11]

J., Saxena , A., Cameron , A

Bunker , A. J., Saxena , A., Cameron , A. J., et al. 2023, , 677, A88

2023
[12]

J., Katz , H., Rey , M

Cameron , A. J., Katz , H., Rey , M. P., & Saxena , A. 2023, , 523, 3516

2023
[13]

2014, , 570, L13

Castro , N., Fossati , L., Langer , N., et al. 2014, , 570, L13

2014
[14]

& Torres , G

Claret , A. & Torres , G. 2019, , 876, 134

2019
[15]

Cox , J. P. & Giuli , R. T. 1968, Principles of stellar structure

1968
[16]

A., Caballero-Nieves , S

Crowther , P. A., Caballero-Nieves , S. M., Bostroem , K. A., et al. 2016, , 458, 624

2016
[17]

A., Schnurr , O., Hirschi , R., et al

Crowther , P. A., Schnurr , O., Hirschi , R., et al. 2010, , 408, 731

2010
[18]

A., & Beasor , E

Davies , B., Crowther , P. A., & Beasor , E. R. 2018, , 478, 3138

2018
[19]

A., et al

de Burgos , A., Sim \'o n-D \' az , S., Urbaneja , M. A., et al. 2025, , 695, A87

2025
[20]

R., & Hubeny , I

de Koter , A., Heap , S. R., & Hubeny , I. 1997, , 477, 792

1997
[21]

2016, , 222, 8

Dotter , A. 2016, , 222, 8

2016
[22]

L., Thompson , A., Crowther , P

Dufton , P. L., Thompson , A., Crowther , P. A., et al. 2018, , 615, A101

2018
[23]

J., Taylor , W

Evans , C. J., Taylor , W. D., H \'e nault-Brunet , V., et al. 2011, , 530, A108

2011
[24]

Fitzpatrick , E. L. & Garmany , C. D. 1990, , 363, 119

1990
[25]

E., Xu, K., Widmann, D., et al

Fjelde, T. E., Xu, K., Widmann, D., et al. 2025, ACM Trans. Probab. Mach. Learn., just Accepted

2025
[26]

2018, in Proceedings of Machine Learning Research, Vol

Ge, H., Xu, K., & Ghahramani, Z. 2018, in Proceedings of Machine Learning Research, Vol. 84, Proceedings of the Twenty-First International Conference on Artificial Intelligence and Statistics, ed. A. Storkey & F. Perez-Cruz (PMLR), 1682--1690

2018
[27]

2021, , 503, 1884

Gilkis , A., Shenar , T., Ramachandran , V., et al. 2021, , 503, 1884

2021
[28]

J., Ram \' rez-Agudelo , O

Grin , N. J., Ram \' rez-Agudelo , O. H., de Koter , A., et al. 2017, , 600, A82

2017
[29]

2014, , 565, A27

Hainich , R., R \"u hling , U., Todt , H., et al. 2014, , 565, A27

2014
[30]

Higgins , E. R. & Vink , J. S. 2019, , 622, A50

2019
[31]

Hoffman , M. D. & Gelman , A. 2011, arXiv e-prints, arXiv:1111.4246

2011 arXiv
[32]

Holgado , G., Sim \'o n-D \' az , S., Herrero , A., & Barb \'a , R. H. 2022, , 665, A150

2022
[33]

Iglesias , C. A. & Rogers , F. J. 1996, , 464, 943

1996
[34]

Irwin , A. W. 2004, The FreeEOS Code for Calculating the Equation of State for Stellar Interiors

2004
[35]

S., Bauer , E

Jermyn , A. S., Bauer , E. B., Schwab , J., et al. 2023, , 265, 15

2023
[36]

2024, , 535, 881

Ji , X., \"U bler , H., Maiolino , R., et al. 2024, , 535, 881

2024
[37]

H., et al

Johnston , C., Michielsen , M., Anders , E. H., et al. 2024, , 964, 170

2024
[38]

1980, , 91, 175

Kippenhahn , R., Ruschenplatt , G., & Thomas , H.-C. 1980, , 91, 175

1980
[39]

2012, , 50, 107

Langer , N. 2012, , 50, 107

2012
[40]

J., & Sugimoto , D

Langer , N., Fricke , K. J., & Sugimoto , D. 1983, , 126, 207

1983
[41]

1947, , 105, 305

Ledoux , P. 1947, , 105, 305

1947
[42]

J., Berlanas , S

Lennon , D. J., Berlanas , S. R., Herrero , A., et al. 2026, , 707, A204

2026
[43]

1987, , 178, 159

Maeder , A. 1987, , 178, 159

1987
[44]

2020, , 634, A118

Mahy , L., Sana , H., Abdul-Masih , M., et al. 2020, , 634, A118

2020
[45]

J., Barb \'a , R

Ma \' z Apell \'a niz , J., Evans , C. J., Barb \'a , R. H., et al. 2014, , 564, A63

2014
[46]

& de Mink , S

Mandel , I. & de Mink , S. E. 2016, , 458, 2634

2016
[47]

& Bodensteiner , J

Marchant , P. & Bodensteiner , J. 2024, , 62, 21

2024
[48]

M., & Moriya , T

Marchant , P., Langer , N., Podsiadlowski , P., Tauris , T. M., & Moriya , T. J. 2016, , 588, A50

2016
[49]

F., & Morrell , N

Massey , P., Neugent , K. F., & Morrell , N. I. 2023, , 947, 77

2023
[50]

M., Dufton , P

McEvoy , C. M., Dufton , P. L., Evans , C. J., et al. 2015, , 575, A70

2015
[51]

2025, , 704, A121

Moens , N., Debnath , D., Verhamme , O., et al. 2025, , 704, A121

2025
[52]

2011, , 192, 3

Paxton , B., Bildsten , L., Dotter , A., et al. 2011, , 192, 3

2011
[53]

2013, , 208, 4

Paxton , B., Cantiello , M., Arras , P., et al. 2013, , 208, 4

2013
[54]

2015, , 220, 15

Paxton , B., Marchant , P., Schwab , J., et al. 2015, , 220, 15

2015
[55]

B., et al

Paxton , B., Schwab , J., Bauer , E. B., et al. 2018, , 234, 34

2018
[56]

2019, , 243, 10

Paxton , B., Smolec , R., Schwab , J., et al. 2019, , 243, 10

2019
[57]

2013, , 495, 76

Pietrzy \'n ski , G., Graczyk , D., Gieren , W., et al. 2013, , 495, 76

2013
[58]

G., Sundqvist , J

Poniatowski , L. G., Sundqvist , J. O., Kee , N. D., et al. 2021, , 647, A151

2021
[59]

M., et al

Ramachandran , V., Hamann , W.-R., Oskinova , L. M., et al. 2019, , 625, A104

2019
[60]

H., Sana , H., de Mink , S

Ram \' rez-Agudelo , O. H., Sana , H., de Mink , S. E., et al. 2015, , 580, A92

2015
[61]

Rogers , T. M. & McElwaine , J. N. 2017, , 848, L1

2017
[62]

2023, , 525, 706

Romagnolo , A., Belczynski , K., Klencki , J., et al. 2023, , 525, 706

2023
[63]

N., Vink , J

Sabhahit , G. N., Vink , J. S., Higgins , E. R., & Sander , A. A. C. 2021, , 506, 4473

2021
[64]

N., Vink , J

Sabhahit , G. N., Vink , J. S., Higgins , E. R., & Sander , A. A. C. 2022, , 514, 3736

2022
[65]

Salpeter , E. E. 1955, , 121, 161

1955
[66]

E., de Koter , A., et al

Sana , H., de Mink , S. E., de Koter , A., et al. 2012, Science, 337, 444

2012
[67]

1993, , 279, 431

Schatzman , E. 1993, , 279, 431

1993
[68]

Schneider , F. R. N., Ram \' rez-Agudelo , O. H., Tramper , F., et al. 2018 a , , 618, A73

2018
[69]

Schneider , F. R. N., Sana , H., Evans , C. J., et al. 2018 b , Science, 359, 69

2018
[70]

2026, , 707, A116

Schootemeijer , A., G \"o tberg , Y., Langer , N., et al. 2026, , 707, A116

2026
[71]

2024, , 689, A157

Schootemeijer , A., Shenar , T., Langer , N., et al. 2024, , 689, A157

2024
[72]

P., et al

Senchyna , P., Plat , A., Stark , D. P., et al. 2024, , 966, 92

2024
[73]

S., Sana , H., & Sander , A

Shenar , T., Gilkis , A., Vink , J. S., Sana , H., & Sander , A. A. C. 2020, , 634, A79

2020
[74]

D., Sablowski , D

Shenar , T., Richardson , N. D., Sablowski , D. P., et al. 2017, , 598, A85

2017
[75]

P., Hainich , R., et al

Shenar , T., Sablowski , D. P., Hainich , R., et al. 2019, , 627, A151

2019
[76]

2022, , 665, A148

Shenar , T., Sana , H., Mahy , L., et al. 2022, , 665, A148

2022
[77]

2021, , 650, A147

Shenar , T., Sana , H., Marchant , P., et al. 2021, , 650, A147

2021
[78]

A., Crowther , P

Tehrani , K. A., Crowther , P. A., Bestenlehner , J. M., et al. 2019, , 484, 2692

2019
[79]

Temaj , D., Schneider , F. R. N., Laplace , E., Wei , D., & Podsiadlowski , P. 2024, , 682, A123

2024
[80]

W., Stark , D

Topping , M. W., Stark , D. P., Senchyna , P., et al. 2024, , 529, 3301

2024
[81]

O., de Koter , A., et al

Verhamme , O., Sundqvist , J. O., de Koter , A., et al. 2026, , 710, A166

2026
[82]

Vink , J. S. 2023, , 679, L9

2023
[83]

S., Brott , I., Gr \"a fener , G., et al

Vink , J. S., Brott , I., Gr \"a fener , G., et al. 2010, , 512, L7

2010
[84]

S., de Koter , A., & Lamers , H

Vink , J. S., de Koter , A., & Lamers , H. J. G. L. M. 2001, , 369, 574

2001
[85]

Walborn , N. R. & Blades , J. C. 1997, , 112, 457

1997
[86]

R., Sana , H., Sim \'o n-D \' az , S., et al

Walborn , N. R., Sana , H., Sim \'o n-D \' az , S., et al. 2014, , 564, A40

2014

Pith tools

Reviewed July 10, 2026 · model on record in the stance chip above.

[1] [1]

A., Sana , H., Taylor , W., et al

Almeida , L. A., Sana , H., Taylor , W., et al. 2017, , 598, A84

work page 2017

[2] [2]

2023, , 519, 5333

Baraffe , I., Clarke , J., Morison , A., et al. 2023, , 519, 5333

work page 2023

[3] [3]

M., Crowther , P

Bestenlehner , J. M., Crowther , P. A., Broos , P. S., Pollock , A. M. T., & Townsley , L. K. 2022, , 510, 6133

work page 2022

[4] [4]

M., Crowther , P

Bestenlehner , J. M., Crowther , P. A., Caballero-Nieves , S. M., et al. 2020 a , , 499, 1918

work page 2020

[5] [5]

M., Crowther , P

Bestenlehner , J. M., Crowther , P. A., Caballero-Nieves , S. M., et al. 2020 b , , 499, 1918

work page 2020

[6] [6]

M., Gr \"a fener , G., Vink , J

Bestenlehner , J. M., Gr \"a fener , G., Vink , J. S., et al. 2014, , 570, A38

work page 2014

[7] [7]

O., Singh , S

Bj \"o rklund , R., Sundqvist , J. O., Singh , S. M., Puls , J., & Najarro , F. 2023, , 676, A109

work page 2023

[8] [8]

1958, , 46, 108

B \"o hm-Vitense , E. 1958, , 46, 108

work page 1958