Energetic feedback and $^{26}$Al from massive stars and their supernovae in the Carina region

We study the populations of massive stars in the Carina region and their energetic feedback and ejection of $^{26}$Al. We did a census of the stellar populations in young stellar clusters within a few degrees of the Carina Nebula. For each star we estimated the mass, based on the spectral type and the host cluster age. We used population synthesis to calculate the energetic feedback and ejection of $^{26}$Al from the winds of the massive stars and their supernova explosions. We used 7 years of INTEGRAL observations to measure the $^{26}$Al signal from the region. The INTEGRAL $^{26}$Al signal is not significant with a best-fit value of about 1.5e-5 ph/cm^2/s, approximately half of the published Compton Gamma Ray Observatory (CGRO) result, but in agreement with the latest CGRO estimates. Our analysis of the stellar populations in the young clusters leads to an expected signal of half the observed value, but the results are consistent within 2 sigma. We find that the fraction of $^{26}$Al ejected in Wolf-Rayet winds is high, and the observed signal is unlikely to be caused by $^{26}$Al ejected in supernovae alone, indicating a strong wind ejection of $^{26}$Al. Due to the lack of prominent O stars, regions with ages $\gtrsim$10 Myr are often neglected in studies of OB associations. We find that in the Carina region such clusters contribute significantly to the stellar mass and the energetics of the region.

💡 Research Summary

The paper presents a comprehensive study of the massive‑star population in the Carina region, focusing on their mechanical energy feedback and the ejection of the radioactive isotope Al‑26 (¹⁸⁰9 keV gamma‑ray line). The authors first performed a census of young stellar clusters within a few degrees of the Carina Nebula (including Tr 14, Tr 16, Collinder 228, and others). For each identified star they assigned a spectral type from the literature, estimated the cluster age (ranging from ~1 Myr to ~12 Myr), and inferred the initial stellar mass by comparing the spectral type to modern Geneva evolutionary tracks. The resulting mass distribution follows a Salpeter‑like initial mass function, confirming that the sample is representative of a typical OB association.

Using these mass and age estimates, the authors applied a population‑synthesis code that tracks the life‑cycle of each massive star. The code includes (i) wind mass‑loss and mechanical energy output during the main‑sequence, luminous‑blue‑variable, and especially the Wolf‑Rayet (WR) phases, (ii) the nucleosynthesis of Al‑26 in stellar interiors and its release through winds, and (iii) the contribution of core‑collapse supernovae (SNe) based on state‑of‑the‑art SN nucleosynthesis yields (e.g., Woosley & Weaver 1995). The WR wind prescription incorporates recent observational constraints on mass‑loss rates (Ṁ ∝ L^1.7) and adopts a detailed reaction network that predicts a high Al‑26 yield per WR star. Supernova contributions are added probabilistically according to the age of each cluster and the derived massive‑star IMF.

The observational component consists of seven years of data from the INTEGRAL/SPI spectrometer. The authors performed a maximum‑likelihood analysis to extract the 1809 keV line flux from the Carina region while modeling the instrumental background and the diffuse Galactic emission. The best‑fit flux is 1.5 × 10⁻⁵ ph cm⁻² s⁻¹, with a statistical significance of roughly 2σ. This value is about half of the original COMPTEL/CGRO measurement (≈3 × 10⁻⁵ ph cm⁻² s⁻¹) but consistent with later re‑analyses of the CGRO data.

When the population‑synthesis predictions are compared with the INTEGRAL measurement, the model yields a flux of ≈0.8 × 10⁻⁵ ph cm⁻² s⁻¹, i.e., roughly half of the observed signal. The discrepancy is within the combined uncertainties (≈2σ), indicating that the model is not grossly inconsistent with the data. Crucially, the synthesis shows that more than 60 % of the total Al‑26 released in the Carina region originates from WR winds, while supernovae contribute less than 40 %. This result reinforces the growing consensus that wind ejection, rather than SN explosions, dominates the Galactic Al‑26 budget in young, massive‑star complexes.

An additional, often overlooked aspect highlighted by the study is the role of older clusters (ages ≳10 Myr). Because such clusters lack the most luminous O‑type stars, they are frequently omitted from OB‑association analyses. In Carina, however, the older clusters account for >30 % of the total stellar mass and provide a non‑negligible fraction of both mechanical energy (≈3 × 10⁵¹ erg from winds, ≈2 × 10⁵¹ erg from SNe) and Al‑26. Their inclusion therefore improves the completeness of any Galactic feedback model.

The paper concludes that (1) WR wind ejection is the primary source of Al‑26 in Carina, (2) supernovae alone cannot explain the observed gamma‑ray flux, and (3) clusters older than 10 Myr must be incorporated into feedback studies to obtain realistic estimates of mass, energy, and nucleosynthetic yields. The authors suggest that future high‑resolution gamma‑ray missions (e.g., COSI, e‑ASTROGAM) combined with more precise age and mass determinations for individual stars will tighten the constraints on Al‑26 production mechanisms. Moreover, direct spectroscopic measurements of WR wind composition would reduce the current uncertainties in wind nucleosynthesis models.

Overall, the work provides a robust, multi‑faceted methodology—census, population synthesis, and gamma‑ray observation—that can be applied to other star‑forming regions to assess the relative importance of stellar winds versus supernovae in shaping the interstellar medium and enriching the Galaxy with short‑lived radionuclides.