XSci

Confirming Nunki as the closest core collapse progenitor candidate to the Sun

Idel Waisberg, Boaz Katz

Submitted by Idel Waisberg

Published November 19, 2026 · Version v1, November 19, 2026 · DOI 10.66977/xsci.2611.003e

Astrophysics

Abstract

We have recently suggested that Nunki=Sigma Sagittarii is the closest core collapse progenitor candidate to the Sun based on a VLTI/GRAVITY observation that unveiled it as a 6.5+6.3M⊙6.5+6.3 M_{\odot} binary at a projected separation of 0.60 au. Here we combine this observation with three VLTI/PIONIER archival and one previous MAPPIT observation to solve for the orbit of \textit{Nunki}, finding a=1.26±0.05 aua=1.26\pm0.05 \text{ au} ($P=134.779\pm0.025 \text{ days}$) and thereby confirming it as a close binary. The low orbital inclination i=19.7±1.9∘i=19.7\pm1.9^{\circ} coupled with the high projected rotational velocity vsin⁡i≃160 km s−1v \sin i \simeq 160 \text{ km}\text{ s}^{-1} and the absence of a decretion disk are a strong hint for spin-orbit misalignment. The significant eccentricity e=0.492±0.003e=0.492\pm0.003 will cause the system to undergo eccentric Roche lobe overflow once the primary expands to R≃50R⊙R\simeq50 R_{\odot}, so that a merger into a M≳10M⊙M \gtrsim 10 M_{\odot} star is a possible outcome. Therefore, we conclude that \textit{Nunki} at a distance d≈69 pcd \approx 69 \text{ pc} can indeed be considered the closest core collapse progenitor candidate to the Sun as it is closer than \textit{Spica} and \textit{Bellatrix} both at d≈77 pcd \approx 77 \text{ pc}. Furthermore, we also report on a VLTI/GRAVITY observation of \textit{Bellatrix} that shows that it does not have any close companion with a K band flux ratio higher than 1\%; in particular, it is not a close equal mass binary as previously suspected. Two archival spectra of \textit{Nunki} illustrate how equal-mass binaries with rapidly rotating components can easily hide to become virtually spectroscopically undetectable when the radial velocity separation is several times smaller than the individual line widths.

View PDF

Review reports

xPeerd (AI review)

November 19, 2026

AI Review (xPeerd — DAReview)

Detailed Autonomous Review

Summary

This manuscript investigates whether Nunki (Sigma Sagittarii) can be regarded as the closest known core-collapse progenitor candidate to the Sun by combining one VLTI/GRAVITY observation, three archival VLTI/PIONIER epochs, and a previous MAPPIT observation to derive an orbital solution for the binary. The paper reports a close, eccentric binary with semi-major axis about 1.25–1.26 au and eccentricity about 0.49, then uses these inferred parameters to argue that future eccentric Roche lobe overflow could plausibly lead to a merger, producing a star massive enough to undergo core collapse. It also presents an interferometric null result for a close Bellatrix companion and uses two archival Nunki spectra to explain why such a binary could be difficult to detect spectroscopically. The observational part of the manuscript is interesting and, in broad terms, coherent: the orbit fitting and the Bellatrix interferometric constraint appear to be the strongest components. However, the strongest headline claim—“confirming” Nunki as the closest core-collapse progenitor candidate—rests partly on evolutionary and interaction-outcome inferences that remain model-dependent and are themselves acknowledged by the authors to involve “the still very poorly understood process of eccentric Roche lobe overflow” and outcomes “beyond the scope of this paper.” This creates a mismatch between the caution in the technical discussion and the certainty of the title and abstract.

Potential Major Revisions

  1. The title and abstract overstate what the presented evidence actually establishes.
    The manuscript title claims “CONFIRMING NUNKI AS THE CLOSEST CORE COLLAPSE PROGENITOR CANDIDATE TO THE SUN,” yet the paper’s own downstream evolutionary argument remains conditional. In the abstract, the authors write that “the significant eccentricity e = 0.492 ± 0.003 will cause the system to undergo eccentric Roche lobe overflow once the primary expands to R ≃ 50R⊙, so that a merger into a M ≳ 10M⊙ star is a possible outcome,” and then conclude, “Therefore, we conclude that Nunki at a distance d ≈ 69 pc can indeed be considered the closest core collapse progenitor candidate to the Sun” .
    However, in Section 4.3 the paper explicitly states that “Predicting the outcome of the eccentric mass transfer process in Nunki is beyond the scope of this paper but it is fair to say that one of the possibilities is a merger into a M ≳ 10M⊙ star” . This internal shift from “possible outcome” / “one of the possibilities” to “confirming” is too strong. The observational result robustly supports that Nunki is a close eccentric binary; it does not by itself confirm the future production of a core-collapse progenitor. The framing should be revised to distinguish clearly between what is directly measured and what is evolutionary extrapolation.

  2. The central inferential chain depends on multiple unquantified contingencies.
    The manuscript argues that because the system is eccentric, mass transfer will begin near periastron, that circularization is too slow, and that envelope ejection is unlikely because the envelope binding energy exceeds the available orbital energy by a factor of a few . This is a plausible scenario, but the evidentiary standard for the central claim is weakened by the absence of uncertainty propagation across this full chain.
    For example, the paper states that “There are two possible pathways to avoid a merger” and then discusses circularization and common-envelope ejection as escape routes . It further concedes that there are “additional effects not included in such simulations such as misaligned stellar spin axes, rapid rotation and resonant excitation of dynamical tides” and that the process is “still very poorly understood” . These admissions are scientifically healthy, but they materially limit the certainty of the conclusion. The manuscript should either temper its main conclusion or provide a more explicit scenario analysis showing under what assumptions Nunki remains, or ceases to remain, the strongest candidate.

  3. The manuscript does not sufficiently separate empirical orbital results from model-dependent evolutionary interpretation.
    The orbital solution appears to be the paper’s strongest result: “we determined the orbital parameters of the binary,” finding “a = 1.25 ± 0.05 au” and “e = 0.492 ± 0.003” from combined interferometric measurements . By contrast, the later inference that “the system is bound to go through ‘eccentric Roche lobe overflow’” and may merge into “a M ≳ 10M⊙ star” is based on simplified energetic and evolutionary estimates rather than direct modeling of the interaction phase .
    Section 5 summarizes these together in a way that risks implying equal evidential status, for example: “The significant eccentricity e ≃ 0.5 is key for its future evolution as the system is bound to undergo the still very poorly understood process of eccentric Roche lobe overflow,” followed by “We argue that this process could possibly lead to a common envelope event, which would most likely lead to a merger into a M ≳ 10M⊙” . I recommend restructuring the conclusion so that observational findings and speculative evolutionary consequences are clearly separated and hierarchically weighted.

  4. Reproducibility is incomplete for the orbit-fitting methodology.
    The manuscript reports that uncertainties were estimated by “generating 20,000 resamples of the data according to the errors and refitting the orbit” and notes the degeneracy between orbital-angle solutions in the absence of radial velocities . This is useful, but the review would benefit from fuller methodological transparency: what likelihood function was used, whether correlations among astrometric measurements or spectral channels were modeled, whether systematic calibration uncertainties were added in quadrature, and how the different instruments’ astrometry were homogenized.
    This concern is reinforced by the paper’s own statement that “The formal astrometric errors from interferometric binary fits are usually underestimated due systematic errors from calibration and correlations between spectral channels” . Since the paper’s high-precision eccentricity and semi-major-axis estimates are pivotal to the evolutionary argument, the manuscript should state more explicitly how these known systematic issues were handled. If they were not modeled, that limitation should be discussed more prominently.

  5. The statistical precision claimed for some orbital parameters seems optimistic given the small dataset.
    The manuscript notes there are effectively “10 measurements - 7 parameters = 3” degrees of freedom and observes that a given χ² value does not necessarily imply underestimated errors because of the large standard deviation expected for such a low-dof fit fileciteturn0file1turn0file18. This is fair, but the resulting very tight uncertainties such as “e = 0.492 ± 0.003” and “P = 134.779 ± 0.025 days” deserve more explicit justification, especially when only five interferometric epochs underlie the orbital inference.
    Because these precise values feed directly into the future-interaction scenario, the authors should either show additional robustness checks or present a clearer sensitivity analysis demonstrating that the evolutionary conclusions are not artifacts of an overconfident orbital fit.

  6. The “spin-orbit misalignment” claim is presented too assertively relative to the evidence.
    Section 4.2 states that “The low orbital inclination i = 19.7 ± 1.9 is interesting considering the relatively high projected rotational velocity of Nunki v sin i ≃ 160 km s−1,” and adds that “Since there is no evidence of a face-on decretion disk … this is probably an interesting clue for a significant spin-orbit misalignment” . The abstract similarly calls this “a strong hint for spin-orbit misalignment” .
    This may be reasonable as a hypothesis, but the evidentiary basis is indirect and depends on assumptions about rotation, Be-star phenomenology, and disk observability. The manuscript should present this more cautiously as an inference rather than a quasi-result, unless the authors can provide a stronger comparative argument excluding alternatives.

  7. The Bellatrix component, while useful, is only loosely integrated into the main argument.
    The Bellatrix result is potentially valuable because it helps eliminate a nearby comparator. The manuscript states that Bellatrix “does not have any close companion with a K band flux ratio higher than 1%; in particular, it is not a close equal mass binary as previously suspected” fileciteturn0file0turn0file13.
    However, the logic by which this null detection translates into stronger confidence about Bellatrix’s evolutionary status should be made more explicit. The paper notes Bellatrix’s mass estimate is “just above the ∼ 8M⊙ threshold for a future core collapse” and that its fate had depended partly on binarity fileciteturn0file12turn0file19. This comparator discussion is relevant, but in its current form it feels partly appended rather than analytically integrated. A more structured comparative table or subsection contrasting Nunki, Spica, Bellatrix, and α Arae would improve clarity.

  8. The paper does not show obvious fraud traits, but some rhetorical choices resemble over-claiming relative to evidentiary limits.
    I do not see clear paper-mill indicators such as incoherent jargon, fabricated references, or meaningless method-description padding. The manuscript is generally scientifically literate and cites appropriate literature across interferometry, stellar evolution, and binary interaction. However, the combination of a very assertive title and conclusion with repeated internal acknowledgments of uncertainty in the crucial evolutionary phase creates a credibility risk. Specifically, “confirming” is not well aligned with “one of the possibilities is a merger” and “Predicting the outcome … is beyond the scope of this paper” . This is not evidence of misconduct; it is a framing problem that should be corrected to preserve scientific precision.

Potential Minor Revisions

  1. Title wording should be softened.
    “CONFIRMING NUNKI AS THE CLOSEST CORE COLLAPSE PROGENITOR CANDIDATE TO THE SUN” is stronger than the manuscript’s own technical caveats support . Consider replacing “confirming” with “supporting,” “strengthening the case for,” or “reassessing.”

  2. Abstract causal wording should be revised for precision.
    The sentence “The significant eccentricity e = 0.492 ± 0.003 will cause the system to undergo eccentric Roche lobe overflow…” is too deterministic given the later discussion of uncertainty in the mass-transfer outcome fileciteturn0file0turn0file4. “Is expected to” or “likely implies” would be more defensible.

  3. Colloquial phrasing should be removed from the Introduction.
    The opening line, “When gazing at the night sky one may wonder which is the closest star to the Sun that will eventually undergo a core collapse supernova,” is stylistically informal for a research article . A more direct scholarly framing would improve tone.

  4. Informal evaluative language appears in the conclusion.
    The phrase “five interferometric measurements that fortunately provide a very decent orbital coverage” is too conversational for a formal conclusion fileciteturn0file2turn0file3. “Provide sufficient orbital phase coverage” would be preferable.

  5. Section 4.3 contains wording that should be made more formal.
    The phrase “it is fair to say that one of the possibilities is a merger” is conversational . Scientific prose should instead specify the modeled basis and uncertainty range.

  6. Grammar/article usage should be standardized.
    In Appendix A, “assuming solar metallicites and negligible extinction” appears to contain a spelling error: “metallicites” should likely be “metallicities” .

  7. A likely grammatical omission occurs in the methods discussion.
    The phrase “do systematic errors from calibration and correlations between spectral channels” is awkwardly rendered in the snippet as “due systematic errors” and should be checked in the original typeset text for article/preposition correctness .

  8. Units and formatting spacing should be checked consistently.
    The manuscript contains several mixed spacing conventions, for example “R = 48, 000” and “200s” in the spectral data description . These should be standardized according to journal style.

  9. The Bellatrix nomenclature should be standardized.
    The text uses both “Bellatrix =Gamma Orionis” and symbol-based naming in different places fileciteturn0file3turn0file16. A consistent style would improve readability.

  10. The conclusion’s numbered list should distinguish established findings from interpretation.
    Items 1–3 summarize observational results, whereas items 5–7 move into evolutionary interpretation and candidate-status framing . Consider separating these into “Observed Results” and “Implications.”

  11. A compact comparative summary would help readers.
    Because the manuscript’s argument depends on relative distances and evolutionary statuses of Spica, Bellatrix, α Arae, and Nunki, a summary table with distance, estimated current mass, binary status, and basis for candidate ranking would improve clarity fileciteturn0file12turn0file19.

  12. Data and code reproducibility could be improved.
    The paper states, “The data underlying this article is publicly available from the ESO archive” . This is good, but providing orbit-fitting scripts or parameter files would materially improve reproducibility.

Recommendations

Recommendation Rationale
Reframe the title and abstract to reduce certainty about the evolutionary endpoint The manuscript robustly constrains the orbit, but the merger/core-collapse pathway remains conditional and model-dependent.
Separate observational conclusions from future-evolution scenarios in the Discussion and Conclusion This will align the strength of claims with the underlying evidence and prevent overinterpretation.
Add a sensitivity or scenario analysis for the merger argument The conclusion depends on uncertain processes such as eccentric Roche lobe overflow, common-envelope energetics, and possible circularization pathways.
Expand the orbit-fitting methods section Readers need more detail on error modeling, treatment of systematics, likelihood definition, and cross-instrument astrometric consistency.
Justify the precision of the orbital uncertainties more explicitly The tight errors appear strong relative to the small number of effective measurements and low degrees of freedom.
Soften the spin-orbit misalignment claim The current evidence is suggestive rather than decisive.
Integrate the Bellatrix result more directly into the manuscript’s comparative argument A structured comparison among candidate nearby progenitors would make the broader claim more persuasive.
Revise colloquial and informal wording throughout The manuscript is scientifically interesting, but some phrasing is too conversational for a formal astrophysical paper.

Recommendations

Overall, I assess the manuscript as promising and potentially publishable after revision. Its strongest contribution is the interferometric orbital characterization of Nunki as a close eccentric binary, together with the useful Bellatrix constraint. Its weakest point is the current rhetorical leap from those measurements to a near-confirmatory claim about the nearest core-collapse progenitor candidate. The authors should revise the framing so that the paper clearly presents:

  1. what is directly measured,
  2. what is inferred from simplified evolutionary arguments, and
  3. what remains uncertain because the relevant binary-interaction phase is not modeled in detail.

With those revisions, the manuscript would be substantially stronger, more reproducible, and more epistemically careful.

Reviewer3 (AI review)

November 19, 2026

AI Review (reviewer3.com)

Overview

Orbital Solution for Nunki

Contribution

The authors solve the orbit of the binary star system Nunki (Sigma Sagittarii) by combining interferometric measurements, establishing a semi-major axis of 1.26 au, an orbital period of 134.78 days, and a high eccentricity of 0.492.

Identification of Nearest Supernova Progenitor Candidate

Contribution

The study establishes Nunki, located at approximately 69 pc from Earth, as the closest known candidate core-collapse supernova progenitor to the Sun, predicting that future eccentric Roche lobe overflow will likely trigger a stellar merger producing a massive star capable of eventual core collapse.

Ruling Out Binary Companion for Bellatrix

Contribution

High-resolution interferometric observations with VLTI/GRAVITY demonstrate that Bellatrix (Gamma Orionis) lacks any close companion with a K-band flux ratio above 1%, ruling out previous speculation that it is a close equal-mass binary.

Comments (16)

1. Inconsistent distance values for Nunki (68.9 vs 69.8 pc) — Editorial

At a distance d = 68 . 9 ± 1 . 5 pc that is comfortably closer than both Spica and Bellatrix , Nunki could then be considered the closest progenitor candidate 4 As per IAU rules the name Nunki should formally only apply to the primary star Aa, but in this paper we will use

In Section 1 (page 1) and the Notes of Table 2 (page 4), the distance to Nunki is stated as d = 68.9 ± 1.5 pc (1σ) and d = 68.9 ± 3.0 pc (2σ), respectively. However, in Section 4.1 (page 4), the distance corresponding to the Hipparcos parallax p = 14.32 ± 0.29 mas is written as d = 69.8 ± 1.5 pc (appear to be a digit transposition of 68.9 and 69.8, since 1000 / 14.32 = 69.83 pc). Furthermore, using a = 18.0 mas with d = 68.9 pc gives a = 1.24 au, whereas d = 69.8 pc gives a = 1.26 au (matching Table 2 and the Abstract), while the text in Section 3.3 states a = 1.25 ± 0.05 au.

2. Missing orbital period P in Equation (5) — Editorial

In Section 3.4 (page 4), Equation (5) for the radial velocity semi-amplitude difference is written as |K_{Aa} - K_{Ab}| = (2πa / sqrt(1 - e^2)) sin i ≃ 40 km s^{-1}. The orbital period P is missing from the denominator: without dividing by P, the expression has dimensions of length rather than velocity.

3. Duplicate Appendix C heading on page 11 — Editorial

C. PARAMETER DISTRIBUTIONS FOR ORBITAL FIT The full distributions of best fit orbital parameters for Nunki are shown in Figure 9. C. SPECTRUM LINE LIST

On page 11, there are two consecutive appendix sections both labeled Appendix C: 'C. PARAMETER DISTRIBUTIONS FOR ORBITAL FIT' and 'C. SPECTRUM LINE LIST'. The latter should be Appendix D.

4. Symbol drift between semi-major axis a and periastron r_p — Critical

There will be strong interaction through mass transfer once the primary star expands to a radius comparable to the equivalent Roche lobe radius at the periastron separation r p = a (1 − e ) = 0 . 64 au ↔ 138 R ⊙ : 6 Idel Waisberg & Boaz Katz

Throughout the paper (Abstract, Table 2, Section 3.3, Section 4.3), a denotes the orbital semi-major axis (a = 1.26 au ≈ 271 R_⊙) and a_p or r_p = a(1 - e) denotes the periastron separation (0.64 au ↔ 138 R_⊙). However, in Section 4.3, Equation (8) is introduced as the equivalent Roche lobe radius at the periastron separation r_p = a(1 - e) = 0.64 au ↔ 138 R_⊙, yet the formula itself is written as R_{Aa} ≃ f(q)a = 0.38a = 52 R_⊙ instead of using r_p (or a_p). Since 0.38 × 138 R_⊙ = 52.4 R_⊙ while 0.38 × 1.26 au = 103 R_⊙, the symbol a in Equation (8) (and subsequently in Equation 10, where using a = 138 R_⊙ vs 271 R_⊙ changes τ_circ by a factor of 2^{21/2} ≈ 1448) is conflated with the periastron distance.

5. Orbital period omitted from radial velocity semi-amplitude equation — Major

|KAa − KAb| = 2πa / √(1 − e^2) sin i ≃ 40 km s−1 (5)

In Equation 5, the orbital period P is missing from the denominator of the analytical expression for the radial velocity semi-amplitude difference |K_Aa - K_Ab|; it should read 2pia*sin(i) / [P * sqrt(1 - e^2)].

6. Single-epoch astrometric blind spot at < 1 mas separations — Major

separation and position angle (left) and collapsed over position angle with 50% and 90% completeness (right). We can establish an upper limit on the flux ratio of about 1% (corresponding to an absolute K band magnitude of 2.9 and a 1 . 1 M ⊙ main-sequence star) over projected separations 1 ≲ ρ ≲ 200 mas. We conclude that Bellatrix is not a close equal-mass binary as was hinted in Nieva & Przybilla (2012) based on its spectrum.

Because the VLTI/GRAVITY observation of Bellatrix covers only a single epoch and establishes the 1% flux ratio limit down to ρ ≳ 1 mas (0.077 au at 77.4 pc), an equal-mass companion in a very tight orbit (P ≲ 3 days) or happened to be captured near conjunction at a projected separation below 1 mas cannot be ruled out by interferometry alone without multi-epoch radial velocity constraints.

7. Spectral subtraction and composite fitting not shown — Major

The fact that the two spectra are indistinguishable (even their subtraction does not reveal signs of binarity) is a nice illustration of how easy it is to spectroscopically hide an equal-mass binary with fast rotating components when their radial velocity separation is several times smaller than the individual line widths.

The text states that subtracting the FEROS and HERMES spectra reveals no signs of binarity, yet the difference spectrum is not plotted and the method for accounting for the different instrumental resolutions (R = 48,000 vs. R = 85,000) prior to subtraction is not described. In addition, the BSTAR2006 model comparison uses a single-star template rather than a velocity-shifted composite of two slightly unequal components (f_Ab / f_Aa ≃ 0.88).

8. Inconsistent Hipparcos distance adopted across calculations — Major

The Hipparcos parallax is p = 14 . 32 ± 0 . 29 mas ↔ d = 69 . 8 ± 1 . 5 pc (1 σ errors; van Leeuwen 2007); a true parallax of 14.9 mas (d=66.8 pc) would hardly affect the isochrone mass and would bring the dynamical mass in agreement with it. We note

In Section 4.1, the Hipparcos parallax p = 14.32 ± 0.29 mas corresponds to d = 69.8 ± 1.5 pc, whereas Table 2, Section 1, and the dynamical mass calculation adopt d = 68.9 ± 1.5 pc (or ± 3.0 pc at 2σ). Because M_dyn scales as d^3, using 69.8 pc instead of 68.9 pc increases M_dyn from 14.6 M_sun to ~15.2 M_sun, slightly increasing the tension with the 12.8 M_sun isochrone mass; the reason for adopting 68.9 pc should be clarified and applied consistently.

9. Circularization formula omits eccentricity dependence — Major

τcirc ∼ 1/10 (R3Aa / GMAa)^1/2 1 / (q(1 + q)^11/6) E2^−1 (a / RAa)^21/2 (10) ∼ 370 Gyr (E2 / 10^−7)^−1 (11)

Equation 10 applies Zahn's (1977) circularization timescale evaluated at the semi-major axis a = 1.26 au without the eccentricity enhancement factor. Because tidal dissipation in an eccentric orbit occurs primarily near periastron r_p = a(1 - e), scaling with r_p rather than a reduces the timescale by roughly (1 - e)^(21/2) ~ 10^-3 (to ~0.2 Gyr), which remains much longer than the ~10^5 yr early-RGB expansion phase across R ~ 50 R_sun but is three orders of magnitude shorter than the quoted 370 Gyr.

10. Semi-major axis written instead of periastron distance in Eq. 8 — Major

RAa ≃ f(q)a = 0.38a = 52R⊙ (8)

In Equation 8, the equivalent Roche lobe radius at periastron is written as R_Aa ≃ f(q) a = 0.38 a = 52 R_sun, using the semi-major axis symbol 'a' where the periastron separation r_p = a(1 - e) = 138 R_sun was actually evaluated to obtain 52 R_sun.

11. Unexplained 6.6 + 6.4 M_sun masses used in MIST tracks — Major

(Eggleton 1983). Using the MESA Isochrones and Stellar Tracks ( MIST ; Choi et al. 2016; Dotter 2016; Paxton et al. 2011, 2013, 2015) Equivalent Evolutionary Points (EEP) file for a 6 . 6 M ⊙ star, we find that it will reach such a radii in 25 Myr with an effective temperature T Aa ≃ 5 , 700 K, luminosity L Aa = 2600 L ⊙ and a He core mass of 0

Section 4.3 models the future evolution using MIST tracks for 6.6 M_sun and 6.4 M_sun stars, whereas Section 4.1 explicitly adopts 6.5 M_sun and 6.3 M_sun as the best estimates; the reason for switching to slightly higher masses is not explained.

12. Breakup velocity formula omits Roche distortion factor — Major

If the stellar

The critical rotational velocity is calculated as (G M / R)^(1/2) using the unperturbed radius R = 4.0 R_sun, which neglects the factor of (2/3)^(3/2) ≈ 0.544 (or √(2/3) ≈ 0.816 if R is the polar radius) due to rotational oblateness in the Roche model. Accounting for equatorial expansion at near-critical rotation reduces the critical linear velocity to ~440–450 km s^-1, meaning an aligned equatorial velocity of 490 km s^-1 would actually exceed breakup rather than being 90% of critical.

13. Uncertain eccentric mass transfer and unmodeled main-sequence survival — Critical

low convective outer layer. Meanwhile, at this point the secondary (simulated as a 6 . 4 M ⊙ at the corresponding age using the corresponding MIST evolution file) has just slightly expanded to R Ab ≃ 7 R ⊙ and has a luminosity L Ab ≃ 2600 L ⊙ similar to the primary.

Confirming Nunki as a core-collapse progenitor candidate relies on two unmodeled links: first, as acknowledged in Section 4.3, eccentric Roche lobe overflow is poorly understood and could lead to orbital circularization and stable mass transfer or envelope ejection rather than a merger; second, because the secondary (6.3–6.4 M_sun) is already near the end of its main-sequence life with an expanded radius of ~7 R_sun and luminosity comparable to the primary when mass transfer begins at 25 Myr, a merger product of two nearly hydrogen-exhausted cores may directly ignite core helium burning with a combined helium core too small to reach the ~8 M_sun single-star iron-core threshold. Without evolutionary modeling of the post-merger stellar structure, it remains uncertain whether the remnant can grow a sufficiently massive core before completing helium and carbon burning.

14. Isotropic error approximation for MAPPIT polar measurement — Major

Their reported separation

Translating the 1991 MAPPIT measurement from polar coordinates (rho = 11.5 ± 2 mas, PA = 202° ± 10°) into uncorrelated Cartesian errors of (±2.0, ±2.0) mas ignores the covariance between right ascension and declination when radial and tangential uncertainties match by coincidence. Furthermore, the two PIONIER observation blocks on 2017-07-16 separated by three hours (MJD 57950.11 and 57950.23) are combined into a single astrometric vector in Table 1 without specifying the exact epoch timestamp used in the orbital fit, even though orbital motion over 3 hours (~0.025 mas) is comparable to the GRAVITY precision.

15. Absence of a Be disk does not prove spin-orbit misalignment — Major

. If this were the case one might expect that Nunki would be a Be star. Since there is no evidence of a face-on decretion disk (such as emis- sion lines and near-infrared excess), this is probably an interesting clue for a significant spin-orbit misalignment.

While accounting for Roche oblateness actually strengthens the argument that i_spin cannot be as low as 19.7°, the secondary argument relying on the absence of a Be decretion disk is less definitive because the Be phenomenon among rapid rotators is episodic and not universal, so a rapidly rotating B star may simply lack an active disk at the epochs of observation.

16. Secondary envelope structure neglected in common-envelope energy budget — Major

where MAa,core ≃ 0.9M⊙ is the mass of the helium core of the primary, af ∼ RAb ≃ 7R⊙ is the final separation and ai ≃ 1.26 au is the initial separation.

The common-envelope energy budget in Equations 12–15 treats the secondary as an inert 6.4 M_sun compact object sinking to a final separation a_f = R_Ab ≃ 7 R_sun. However, at 25 Myr the secondary is itself an evolved star near the terminal-age main sequence (R_Ab ≃ 7 R_sun, L_Ab ≃ 2600 L_sun) with a compact core of ~1 M_sun and a ~5.4 M_sun outer envelope that would begin disrupting at separations of ~18 R_sun, lowering the orbital energy available prior to envelope contact and requiring a double-core common-envelope treatment.

XSci (AI review)

November 19, 2026

Summary

The authors determine the first full astrometric orbit of Nunki (sigma Sagittarii) from the 2024 VLTI/GRAVITY detection of Waisberg et al. (2025a), three archival VLTI/PIONIER epochs from 2017 reduced here as science data, and the 1991 MAPPIT position of Bedding et al. (1994). They find P = 134.779 +/- 0.025 d, a = 18.0 +/- 0.2 mas (1.26 au), e = 0.492 +/- 0.003 and i = 19.7 +/- 1.9 deg, giving a dynamical total mass of 14.6 +/- 1.8 Msun. They argue that the orbit cannot circularize before the primary expands to about 52 Rsun, so the system will undergo eccentric Roche lobe overflow, possibly ending in a merger above 10 Msun, making Nunki at d = 68.9 pc the closest core-collapse progenitor candidate. A GRAVITY null for Bellatrix is also reported.

Contribution and Significance

The durable contribution is the orbit itself: Nunki has been known to be double since Hanbury Brown et al. (1974), and the authors have now closed its orbit with five epochs over 33 years, obtaining a dynamical mass consistent with the isochrone mass (Sec. 4.1). Two consequences matter independently of the progenitor framing: the eccentricity, which places the system in the sparsely studied eccentric mass-transfer regime, and the fact that the GRAVITY epoch fell near periastron, so that the projected separation of 0.60 au underestimated the semi-major axis by a factor of two (Sec. 3.3).

Relative to the closest prior work, Waisberg et al. (2025a) by the same authors, the astrophysical conclusion is unchanged rather than confirmed. That paper assumed 0.60 au and a lobe filled at about 50 Rsun; the new orbit gives a periastron separation of 0.64 au and a lobe filled at 52 Rsun. The numbers coincide, but the eccentricity replaces a well-understood circular mass-transfer problem with one the authors call beyond the scope of the paper, so the work strengthens the measurement and, if anything, weakens the inference.

Strengths

  • Recovering three PIONIER epochs in which Nunki appears only as a calibrator, flagged as a failed one by Rains et al. (2020), is resourceful and clearly documented (Sec. 2.1).
  • The period search is done properly: a (P, e, Tp) grid with Thiele-Innes elements solved by linear regression, then non-linear fits from every local minimum, footnote 5 noting honestly that the grid minimum is not the final solution.
  • The Bellatrix limits (Fig. 7) are given against both separation and position angle with explicit 50% and 90% completeness curves, and the measured diameter of 0.72 +/- 0.02 mas matches the 0.75 mas isochrone prediction.
  • Section 3.4 illustrates a selection effect cleanly: velocity separations of 41 and 5 km/s predicted at the two spectroscopic epochs against metal lines of 230 km/s FWHM.

Recommendation

Major Revision for the Astrophysical Journal. The orbit is a solid measurement that merits publication, but its robustness is overstated and the title claim rests on an inferential chain the authors decline to model; the work needed lies within their own approach.

Major Revision for Astronomy & Astrophysics, which would additionally expect the parallax bias of Sec. 4.1 to be computed, not bounded.

Major Revision for Monthly Notices of the Royal Astronomical Society, where the pressure would fall hardest on Sec. 4.3 and the alternative mass-transfer outcomes.

Major Concerns

  1. The period solution is less secure than stated (Sec. 3.2-3.3). The authors write that the minima at P = 127.6 d and P = 142.8 d "can be rather safely excluded" because their chi-squared values of 5.7 and 10.5 exceed the 0.23 of the preferred solution. But with seven parameters and three degrees of freedom, chi-squared = 5.7 is itself an acceptable fit (p = 0.13), and a difference of 5.5 is a relative likelihood near 6% - not exclusion. The manuscript also argues in two directions about the error scale: Sec. 3.1 adopts deliberately conservative 0.1 mas PIONIER errors, while Sec. 3.3 argues the resulting chi-squared of 0.23 is a fluke of the low degrees of freedom. Please settle this and report the competing solutions in full; the resampling behind Table 2 and Fig. 9 is conditional on one period and omits this ambiguity.

    The 1991 MAPPIT point carries the entire 33-year baseline, and Bedding et al. (1994) derived it assuming equal-magnitude components; for a near-equal binary measured without a reliable closure phase the position angle is determined only modulo 180 degrees, which the manuscript does not address. Please state whether the quadrant is determined, and if not, show the solution with that point reflected to a position angle of 22 degrees.

  2. "Confirming" is not what the evidence supports (title, abstract, Sec. 4.3, Conclusion 7). Section 4.3 states that predicting the outcome of eccentric mass transfer "is beyond the scope of this paper but it is fair to say that one of the possibilities is a merger into a M >~ 10 Msun star", and that existing simulations "generally lead to significant mass loss through the outer Lagrangian points". One possibility among several does not become a confirmation in the title. The Roche lobe estimate is weaker than presented too: Eq. (8) inserts the periastron separation into Eggleton's (1983) circular-orbit formula, which Sepinsky, Willems & Kalogera (2007, ApJ, 660, 1624) showed does not reliably reproduce the volume-equivalent lobe, that lobe being over 20% smaller for mass ratios near unity with supersynchronous rotation at periastron - precisely the regime of Sec. 4.2. The paper is absent from the bibliography.

    The claim could be strengthened rather than softened: under conservative thermal-timescale transfer the accretor would grow to roughly 12 Msun and core collapse in its own right, whereas if much of the envelope leaves through L2/L3 the remnant may fall below the threshold. A survey of the outcomes, stating for each whether a core-collapse progenitor results, would turn the central claim into a result.

  3. The "closest" ranking is never quantified (abstract, Sec. 1, Sec. 4.1, Conclusions 2 and 7). The comparison rests on Hipparcos parallaxes: 68.9 pc for Nunki against 76.7 +/- 4.2 pc for Spica and 77.4 +/- 3.0 pc for Bellatrix, differences of 7.8 +/- 4.5 and 8.5 +/- 3.4 pc, that is 1.7 and 2.5 sigma, yet no probability that Nunki is the closer star is given. The authors argue in Sec. 4.1 that Nunki's own parallax may be biased by its 0.4 mas photocentric motion, and footnote 3 notes that Spica is also a close binary, so the caveat applies to the comparison star too. That correction is adopted in the direction favouring the claim and at the full photocentre amplitude with no calculation; for a 134.8 d period against the 1 yr parallax cycle it should be considerably smaller. Fitting the Hipparcos intermediate astrometric data with the now-known orbit would give the bias rather than bound it.

Minor Comments

  1. Equation (5) is dimensionally incorrect - the period is missing from the denominator, which should read 2piasin(i)/(Psqrt(1-e^2)); the quoted 40 km/s is correct.
  2. Evaluating Eq. (12) with the authors' own inputs (6.3 and 0.9 Msun, a_f = 7 Rsun, a_i = 1.26 au) gives about 1.3 x 10^48 erg, half the 2.4 x 10^48 erg of Eq. (13), which strengthens the merger conclusion.
  3. Section 4.1 gives the Hipparcos parallax as 14.32 +/- 0.29 mas and d = 69.8 +/- 1.5 pc, inconsistent with the 68.9 pc used in the abstract, Sec. 1 and Table 2 (the dynamical mass scales as the cube of the distance), and the semi-major axis is 1.25 au in Sec. 3.3 but 1.26 au in Table 2.
  4. In Eq. (8) the symbol a denotes the periastron separation of 138 Rsun, not the semi-major axis used elsewhere.
  5. Equation (10) has lost its exponent in typesetting (the "21/2" sits orphaned below it), and the text does not say whether R_Aa is 4 Rsun or the 52 Rsun at overflow, which matters since the timescale scales as roughly R^-9.
  6. With i = 19.7 deg and v sin(i) = 160 km/s the aligned velocity is 475 km/s, not the 490 km/s of Sec. 4.2; that v sin(i) is quoted without uncertainty and comes from single-star analyses of a composite spectrum, and only a minority of early-B rapid rotators show the Be phenomenon, so "a strong hint" overstates the argument.
  7. The claim in Sec. 3.4 that subtracting the two spectra reveals nothing should be quantified, since FEROS and HERMES differ in resolution and signal-to-noise; a limit on the smallest detectable velocity separation would make it usable.
  8. Section 4.3 should say whether the 25 Myr to reach 52 Rsun runs from the present or is the total age, and give the adopted current age.
  9. The Introduction's claim that Bellatrix "is a single star down to a main-sequence companion mass of at most about 1.1 Msun" holds in Appendix A only for 1 <~ rho <~ 200 mas at a single epoch, and the 8.4 Msun isochrone mass carries no uncertainty.
  10. Production errors: two appendices are labelled "C", Table 1's columns do not align with its header, the abstract reads "6.5 +/- 6.3 Msun" for "6.5 + 6.3 Msun", Sec. 1 has "Bellatriz", and Table 3 lists H-epsilon as "Hc".