2025A&A...701A.126S

Very massive stars (VMS) play a fundamental role in astrophysics. Their powerful stellar winds, which dictate their evolution, supernovae, and fate as black holes (BHs), are a key uncertainty, as evidence suggests their mass-loss rates may exceed standard predictions. To address this, we investigated the effect of enhanced winds on the single and binary VMS evolution by implementing new stellar wind prescriptions in the stellar evolution code PARSEC v2.0 and in the binary population synthesis code SEVN. Our updated models are sensitive to the Eddington parameter (Γ_e) and the luminosity-to-mass ratio. We used them to simulate the VMS evolution from 100-600 M_⊙ at the metallicity of the Large Magellanic Cloud (LMC) to model the VMS population in the Tarantula Nebula. Our results show that Γ_e-enhanced single-star tracks agree better with the observed VMS properties in the Tarantula Nebula than the standard wind models. When the most massive star in the region, R136a1, is explained via a single-star evolution, a lower limit on the initial mass of ≳300 M_⊙ is required, regardless of the wind recipe used. We also show that binary stellar mergers offer another suitable formation channel that might lower the required initial mass limit by ∼100 M_⊙. The choice of the wind treatment profoundly impacts the BH populations. Stronger winds yield smaller BHs, which inhibits the formation of objects above the lower edge of the pair-instability mass gap (∼50 M_⊙). For merging binary BHs, enhanced-wind models predict more primary BHs above 30 M_⊙ and enable secondary BHs between 30-40 solar masses, which is a range not found with standard stellar winds at the metallicity of the LMC. This study highlights the crucial role of stellar wind physics and binary interactions in the evolution of VMS and resulting BH populations. It offers predictions that are relevant for interpreting VMS observations and gravitational-wave sources.