pith. sign in

arxiv: 2604.02111 · v2 · submitted 2026-04-02 · 🌌 astro-ph.SR · astro-ph.GA

Multiplicity of Massive stars in the Milky Way (M3W). I. Project description, UNWIND, application to GLS 11 448, and DIB catalog

J. Ma\'iz Apell\'aniz , R. C. Gamen , G. Holgado , S. Rosu , J. I. Arias , S. Sim\'on-D\'iaz , A. Pellerin , M. Abdul-Masih
show 3 more authors
E. Madero Fuentes J. A. Molina-Calzada R. H. Barb\'a
This is my paper

Pith reviewed 2026-05-13 21:03 UTC · model grok-4.3

classification 🌌 astro-ph.SR astro-ph.GA
keywords massive starsbinary starsO starsspectral disentanglingstellar massesdiffuse interstellar bandsMilky Way
0
0 comments X

The pith

The binary GLS 11 448 contains two O3.5 II(f*) stars with evolutionary masses of 70 and 76 solar masses, the highest measured for O stars.

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

The M3W project works to enlarge the sample of massive star binaries that have well-determined orbits and masses so that multiplicity statistics, mass functions, and related questions can be addressed with better data. In the first paper the UNWIND algorithm is applied to GLS 11 448 to produce separate spectra for each component across the full 3820-11000 Å range. Both stars are classified O3.5 II(f*), a new orbital solution is obtained, and evolutionary models give masses of 70±10 and 76±11 solar masses. The same analysis supplies a new detection of the interstellar He I 10830 triplet and a catalog of 631 diffuse interstellar bands.

Core claim

Application of the UNWIND spectral disentangling algorithm to GLS 11 448 yields individual spectra for the Aa and Ab components over the entire 3820-11000 Å interval for the first time in an OB binary. Both components receive the classification O3.5 II(f*), a new orbit is derived, and evolutionary models assign masses of 70±10 M⊙ to Aa and 76±11 M⊙ to Ab. The work also reports the first absorption detection of the interstellar He I 10830 triplet along an OB sightline and produces a DIB library containing 631 features between 4000 and 17100 Å.

What carries the argument

The UNWIND algorithm for separating the composite spectrum of a spectroscopic binary into clean individual stellar spectra across a wide wavelength range.

If this is right

  • New stellar lines become available for the study of O stars.
  • The system supplies data on very massive stars that may be progenitors of gravitational-wave events.
  • A DIB catalog with 631 entries, 119 of them previously unidentified, is now available for ISM studies.
  • The analysis framework supports later papers in the series on multiplicity statistics and mass discrepancies.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • If the high masses hold, current models may need revision for the upper end of the O-star mass range.
  • Wider application of the disentangling technique could help reconcile spectroscopic, evolutionary, and Keplerian mass estimates in other binaries.
  • The same sightline data could be used to test predictions about wind properties or runaways among very massive stars.

Load-bearing premise

Evolutionary masses are calculated under the assumption that standard stellar evolution tracks apply without major revision to these very massive and rapidly rotating O stars.

What would settle it

An independent dynamical mass determination from the binary orbit that yields values significantly different from 70 and 76 solar masses.

Figures

Figures reproduced from arXiv: 2604.02111 by A. Pellerin, E. Madero Fuentes, G. Holgado, J. A. Molina-Calzada, J. I. Arias, J. Ma\'iz Apell\'aniz, M. Abdul-Masih, R. C. Gamen, R. H. Barb\'a, S. Rosu, S. Sim\'on-D\'iaz.

Figure 1
Figure 1. Figure 1: Example of UNWIND output for the Hα region of a GLS 11 448 Aa,Ab spectroscopic epoch and the result of combining 77 epochs. The top spectra (green and blue) show the normalized output for the two components, shifted horizontally to the velocities of the epoch and upwards 0.1 continuum units, and diluted by their flux fractions. The bottom (orange and purple) spectra show the fitted telluric lines, specific… view at source ↗
Figure 2
Figure 2. Figure 2: Phased radial velocity curves for GLS 11 448 Aa and Ab. 4.1.2. The systemic velocities In Sect. 4.2 we fit FASTWIND models to the dis￾entangled spectra to derive synthetic H+He spectra for GLS 11 448 Aa and Ab that include the effect of unresolved multiplets and lines from different species and of infilling from stellar winds. Here we use them to adjust the systemic velocities, for which we analyse the dif… view at source ↗
Figure 3
Figure 3. Figure 3: Disentangled Brackett (left) and Pfund (right) He ii series of GLS 11 448 Aa (black) and Ab (red). The spectra are nor￾malised, displaced in continuum units, and placed in the rest ve￾locity frame of each component systemic velocity without taking into account wind infilling effects (hence, the small displacement in centroids towards the blue). Differences in S/N are caused by extinction increasing towards… view at source ↗
Figure 4
Figure 4. Figure 4: Disentangled blue-violet spectra of GLS 11 448 Aa,Ab shifted to the stellar reference frame and degraded to the spectral classification resolution R = 2500. The two spectra are almost identical and there is no sign of the subtracted ISM. This is an update of [PITH_FULL_IMAGE:figures/full_fig_p009_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Section of the UNWIND extraction for GLS 11 448 Aa (in the rest frame of the star determined from [PITH_FULL_IMAGE:figures/full_fig_p010_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: A comparison of the profiles of the interstellar He i λ10 830 triplet and K i λ7698.974 line in the GLS 11 448 sightline, with the latter placed in the velocity reference frame of the first component of the He i triplet (10 829.0911 Å), see also Fig. B.1. The dashed blue vertical lines mark the tentative He i velocity of −23.5 km/s in our barycentric reference frame, which is blueshifted with respect to ei… view at source ↗
read the original abstract

(ABRIDGED BUT NOT TOO FAR) Multiplicity is ubiquitous among massive stars and its understanding is constrained by the sample of well-determined orbits. The immediate goal of M3W is to significantly increase the number of massive multiple systems with well-determined orbits and masses. We will address issues such as multiplicity statistics, the mass function in clusters and the field, the properties of binaries with compact companions and gravitational-wave progenitors, the origin and characteristics of runaways and their 3-D motions, the use of apsidal motion as a probe of stellar interiors, and the mass discrepancy between different methods (evolutionary, spectroscopic, and Keplerian). In this first paper, we present the project; describe the data and tools that will be used, including the disentangling UNWIND tool; analyse the very massive twin binary system GLS 11 448; and briefly introduce some of the following papers of the series. We present a new orbit for GLS 11 448, using UNWIND to obtain for the first time disentangled spectra for the full 3820-11 000 $\mathring{A}$ range for an OB spectroscopic binary. We derive the stellar parameters, making new stellar lines available for the study of O stars. The Aa and Ab components of GLS 11 448, both classified as O3.5 II(f*), are the two most massive O stars ever detected according to the evolutionary masses of 70$\pm$10 M$_\odot$ and 76$\pm$11 M$_\odot$ determined in this paper. We also report the first-ever detection of the interstellar He I 10 830 triplet in absorption in an OB-star sightline. As a by-product of the ISM model derived for UNWIND using GLS 11 448 and five other standard stars, we present the most detailed diffuse-interstellar-band (DIB) library ever built, with a total of 631 DIBs in the 4000-17 100 $\mathring{A}$ range, of which 37 are fitted with multiple-Gaussian profiles and 119 had never been identified before.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The paper introduces the M3W project to increase the sample of massive-star binaries with well-determined orbits and masses, describes the UNWIND spectral-disentangling algorithm, applies it to the O+O binary GLS 11 448 to derive a new spectroscopic orbit and full 3820–11 000 Å disentangled spectra, reports stellar parameters and evolutionary masses of 70±10 M⊙ and 76±11 M⊙ for the two O3.5 II(f*) components (claimed to be the most massive O stars known), detects the interstellar He I 10 830 triplet, and releases a catalog of 631 DIBs (37 fitted with multiple Gaussians, 119 previously unidentified).

Significance. If the evolutionary masses prove robust, the identification of the two highest-mass O stars would directly constrain the upper initial-mass function, binary-interaction channels, and the mass discrepancy between evolutionary, spectroscopic, and dynamical methods; the extensive DIB library would also constitute a major reference resource for ISM studies across the optical–near-IR.

major comments (2)
  1. [§ on application to GLS 11 448 and evolutionary-mass determination] § on application to GLS 11 448 and evolutionary-mass determination: the claim that the components are the most massive O stars rests on placing UNWIND-derived T_eff, log L, and abundances onto standard evolutionary tracks, yet the text provides no quantitative assessment of how enhanced rotational mixing, wind mass loss, or possible binary-interaction effects at >70 M⊙ would shift the mass–luminosity relation; without this, the 70±10 and 76±11 M⊙ values cannot be shown to exceed all prior records.
  2. [§ describing the UNWIND algorithm and its application] § describing the UNWIND algorithm and its application: validation tests for residual wind or circumstellar contamination in the separated spectra are not presented; such residuals would systematically bias the line profiles used to derive the atmospheric parameters that feed the evolutionary-mass calculation, directly affecting the central claim.
minor comments (2)
  1. [Abstract] The abstract states that 'new stellar lines' become available but neither lists them nor indicates where in the disentangled spectra they appear; a short table or explicit wavelength list would improve traceability.
  2. [DIB catalog section] The DIB catalog is presented as a by-product; the fitting procedure for the 37 multi-Gaussian profiles and the criteria used to claim 119 new identifications should be stated more explicitly to allow independent verification.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their constructive and detailed report. We address each major comment below and outline the revisions we will make to strengthen the manuscript.

read point-by-point responses

  1. Referee: the claim that the components are the most massive O stars rests on placing UNWIND-derived T_eff, log L, and abundances onto standard evolutionary tracks, yet the text provides no quantitative assessment of how enhanced rotational mixing, wind mass loss, or possible binary-interaction effects at >70 M⊙ would shift the mass–luminosity relation; without this, the 70±10 and 76±11 M⊙ values cannot be shown to exceed all prior records.

    Authors: We agree that the manuscript would benefit from an explicit discussion of how non-standard effects at these masses could alter the mass-luminosity relation. While our evolutionary masses follow the standard procedure used in the literature for O-star comparisons (including the quoted uncertainties), we will add a new subsection in the revised version that reviews available models for enhanced rotational mixing, wind mass loss, and binary-interaction channels above 70 M⊙. This will contextualize our results, quantify the possible systematic shifts where literature allows, and moderate the claim accordingly while preserving the comparison to prior records obtained with similar methods. revision: yes

  2. Referee: validation tests for residual wind or circumstellar contamination in the separated spectra are not presented; such residuals would systematically bias the line profiles used to derive the atmospheric parameters that feed the evolutionary-mass calculation, directly affecting the central claim.

    Authors: We acknowledge that explicit validation tests for residual contamination were not included in the original text. The UNWIND disentangling incorporates an ISM absorption model fitted simultaneously, and the separated spectra were verified for consistency with the observed composite and with the O3.5 II(f*) classification. In the revised manuscript we will add a dedicated validation subsection that presents quantitative tests, including reconstruction of the composite spectrum from the disentangled components, comparison against synthetic spectra computed from the derived parameters, and checks for systematic residuals in wind-sensitive lines across the 3820–11000 Å range. revision: yes

Circularity Check

0 steps flagged

No circularity: evolutionary masses derived from external tracks applied to observationally disentangled parameters

full rationale

The paper derives stellar parameters (T_eff, log L, abundances) for GLS 11 448 Aa/Ab from UNWIND-disentangled spectra over 3820-11000 Å, then places those parameters onto standard stellar evolution tracks to obtain the reported evolutionary masses of 70±10 M⊙ and 76±11 M⊙. These tracks are external inputs, not defined in terms of the paper's fitted values or results. The new orbital solution is presented separately and does not enter the mass calculation. No self-definitional steps, fitted-input predictions, or load-bearing self-citations appear in the derivation chain; the central claim rests on the applicability of standard tracks rather than on any internal redefinition or renaming of quantities.

Axiom & Free-Parameter Ledger

1 free parameters · 2 axioms · 0 invented entities

The central claims rest on standard assumptions of stellar spectroscopy and evolutionary modeling plus the internal assumptions of the new UNWIND algorithm; no new physical entities are introduced.

free parameters (1)
  • evolutionary track parameters
    Masses are obtained from evolutionary models whose internal parameters (mass-loss rates, rotation, mixing) are calibrated to data rather than derived from first principles.
axioms (2)
  • domain assumption Spectral disentangling can be performed accurately for OB binaries across 3820-11000 Å
    UNWIND relies on this assumption to produce the separated spectra used for parameter derivation.
  • domain assumption Standard O-star classification criteria and line identifications remain valid for these extreme objects
    The O3.5 II(f*) types and new line identifications depend on this background assumption.

pith-pipeline@v0.9.0 · 5788 in / 1602 out tokens · 59160 ms · 2026-05-13T21:03:01.826553+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

6 extracted references · 6 canonical work pages

  1. [1]

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

    Almeida, L. A., Sana, H., Taylor, W., et al. 2017, A&A, 598, A84 Ansín, T., Gamen, R., Morrell, N. I., et al. 2023, MNRAS, 525, 4566 Arias, J. I., Barbá, R. H., Gamen, R. C., et al. 2010, ApJL, 710, L30 Banyard, G., Sana, H., Mahy, L., et al. 2022, A&A, 658, A69 Barbá, R. H., Gamen, R. C., Arias, J. I., et al. 2010, in RMxAC, V ol. 38, 30–32 Barbá, R. H.,...

    work page arXiv 2017

  2. [2]

    Stars used

    ′′428 and with Aa being itself an SB2 system (Maíz Apellániz et al. 2021b). Our data includes the combined light of the three stars. We used the multi-epoch high-resolution LiLiMaRlin spectroscopy to produce high-S/N spectra of the six standards. Before analysing the DIBs, we studied different ISM lines with well known wavelengths (Naiλλ5889.951,5895.924;...

    work page 2002

  3. [3]

    Of the 631 DIBs in Table B.2, 37 are fitted with multiple Gaussians (examples in Fig

    We also give the number of Gaussians used to fit each DIB (with a colour code to identify cases with multiple Gaussians, see below) and the stars used to build each profile (with a colour code to identify new DIBs, see below). Of the 631 DIBs in Table B.2, 37 are fitted with multiple Gaussians (examples in Fig. B.3) and their information is given in Table...

    work page 2013

  4. [4]

    but such an endeavour is not practical for our primary purpose of analysing the stellar spectra. Therefore, we set the rectification points around 4400 Å and 4460 Å (the pre- cise values depending on the star) and we find, after subtracting the stellar features, that the profile is slightly asymmetric and requires two Gaussians. DIBB4884 (a.k.a. the HβDIB...

    work page 1922

  5. [5]

    Most such studies con- centrate on the strongest DIB, DIBN8621 (e.g

    has become pop- ular for DIB studies thanks to the data availability inGaiaDR3, soon to increase significantly in DR4. Most such studies con- centrate on the strongest DIB, DIBN8621 (e.g. Schultheis et al. 2023), with some including the second one in EW for most sightlines, DIBI8646 (e.g. Zhao et al. 2024). A third previously known DIB, DIBN8530 (Jenniske...

    work page 2023

  6. [6]

    The DIB colour code is the same as in Table B.2

    The hori- zontal scale is in Å and the non-linear vertical scale is used to emphasise weak DIBs. The DIB colour code is the same as in Table B.2. Light gray areas indicate where telluric absorp- tion severely hampers the detection of DIBs and dark gray areas where it makes it impossible. Light orange areas are gaps (de- tectors or orders) in the CARMENES ...

    work page 2015