Revisiting the Evolutionary Status of Massive Stars at the central parsec of the Milky Way

Massive stars and their winds have a large influence in their environment, e.g, determining the accretion rate on to the Galactic Centre (GC) super-massive black hole Sgr A*. The winds of those stars collide and are accreted, at a rate that depends on their chemical composition. Here we aim to revisit the evolutionary status of the evolved massive stars at the GC, by means of new tracks based on updated mass-loss rate recipes for the earlier stages of massive stars. We use the Geneva-evolution-code for initial stellar masses ranging from 20 to 60 $M_\odot$, for metallicity $Z=0.020$. We adopt a new mass-loss rate recipe for the line-driven winds of O-type stars and B-supergiants, plus a new recipe for the dust-driven winds of red supergiants (RSG). Additionally, we set up initial rotation $Ω/Ω_\text{crit}=0.4$, and we adopt the Ledoux criterion for the treatment of convection in inner layers. We found that evolution models adopting new mass-loss rate prescriptions predict that stars will lose less of their outer layers during their initial phases, while a big reduction of mass happens at the RSG phase. As a consequence, the resulting Wolf-Rayet (WR) stars are less radially homogeneous in their inner structure from the core to the surface. Also, these new evolution models predict the absence of hydrogen-free WN stars. These evolutionary predictions agree better with the observed chemical abundances of the WR stars at the GC. We provide a table with the chemical H, He, and CNO abundances calculated for the different subtypes of WR stars. We propose a different re-arrangement of the WR subtypes to be used for the modelling of the collision of their winds. We discuss the potential implications of these changes for the colliding winds generated from the massive stars at the GC, which are accreting onto the supermassive black hole Sgr A*.

💡 Research Summary

The authors present a comprehensive re‑evaluation of the evolutionary status of massive stars within the central parsec of the Milky Way, employing updated stellar‑wind mass‑loss prescriptions that are significantly lower than the classic Vink et al. (2001) rates for O‑type and B‑supergiant phases and higher for the red‑supergiant (RSG) phase. Using the Geneva‑evol code, they compute evolutionary tracks for initial masses of 20–60 M☉ at a super‑solar metallicity Z = 0.020, with an initial rotation of Ω/Ω_crit = 0.4 and the Ledoux criterion for convection.

Key methodological advances include: (1) adoption of the Krtička et al. (2024) mass‑loss formula (K24) for T_eff > 10 kK, which provides a smooth temperature dependence and avoids the controversial bi‑stability jump; (2) implementation of the Yang et al. (2023) RSG prescription (Y23), which is essentially metallicity‑independent and yields very high mass‑loss rates (up to 3 × 10⁻³ M☉ yr⁻¹) during the cool supergiant phase; (3) retention of Nugis & Lamers (2000) and Gräfener & Hamann (2008) prescriptions for Wolf‑Rayet (WR) winds, appropriate for the mass range considered.

The resulting tracks differ markedly from earlier models that used Vink et al. (2001) rates. In the early main‑sequence phase, reduced mass loss allows stars to retain more mass, stay larger and more luminous, and preserve angular momentum for longer, enhancing rotational mixing and surface enrichment of helium and nitrogen. When the stars reach the RSG stage, the dramatically increased mass loss strips almost the entire hydrogen envelope, leading to a rapid transition to the WR phase. Consequently, the new models predict that hydrogen‑free WN stars (WN‑h = 0) do not appear at all; instead, the sequence Ofpe/WN9 → WNL → WN/C → WC matches the observed chemical abundances of Galactic‑center WR stars.

Quantitatively, the authors provide tables of surface mass fractions (H, He, C, N, O) for each WR subtype, showing higher nitrogen and lower carbon/oxygen in the WNL phase than in previous models, in agreement with spectroscopic analyses (e.g., Martins et al. 2007). The lifetimes of the WR sub‑phases are also revised: the WNL stage is longer, while the WC stage is slightly shortened, reflecting the altered envelope stripping history.

These chemical and structural differences have direct implications for the colliding‑wind environment around Sgr A*. WR winds are fast (v ≳ 1000 km s⁻¹), but their cooling efficiency after shock depends strongly on hydrogen content. Winds with residual hydrogen cool more efficiently, fostering the formation of dense clumps that can be accreted episodically, potentially feeding Sgr A* in bursts. In contrast, hydrogen‑poor WR winds remain hot, producing a smoother, Bondi‑like inflow. The revised abundances therefore necessitate a re‑calibration of the wind‑collision simulations that have been used to estimate the accretion rate onto the super‑massive black hole.

Finally, the paper proposes a re‑arranged classification scheme for the WR subtypes based on the new abundance patterns, and supplies a ready‑to‑use table for modelers of the Galactic‑center gas dynamics. By integrating modern mass‑loss recipes, rotation, and Ledoux convection, the study offers a more realistic picture of massive‑star evolution in the extreme environment of the Galactic nucleus, with consequential impacts on our understanding of the feeding and growth of Sgr A*.