A Detailed Spectroscopic Study of the WN6ha + O5 Colliding-wind Binary WR 25

Massive stars have a major dynamical as well as chemical impact on their environment in their host galaxy. Despite their importance, many details of their life and evolution remain poorly constrained. Observational determinations of their physical parameters are therefore required to better constrain evolutionary models. In particular, the mass and the mass-loss rate—two key parameters—are known to be more reliably determined through the study of binary stars. In this context, the binary system WR 25 (≡ HD 93162) whose dominant spectral type is that of a nitrogen sequence Wolf–Rayet star, is of particular interest. It exhibits the highest known hydrogen content among the Milky Way Wolf–Rayet stars, as well as additional absorption lines. It is therefore most probably an object still on the main-sequence or close to leaving it. We acquired new high-resolution spectra of this extremely interesting object. Following a detailed analysis, we present an improved version of the SB1 orbital solution for the motion of the WN component. We confirm the detection of the previously reported probable O companion and definitively prove that it is indeed gravitationally bound to the WN star. A first SB2 solution is presented, and a mass-ratio WN/O of 2.02 ± 0.36 (1σ) is derived. Following a spectral disentangling procedure, we identify the companion as an O5(f+) star belonging to either the main-sequence (V) or giant (III) luminosity classes. The spectral type of the companion, together with the above-mentioned mass ratio, points to a very massive primary WN star of at least 55 M⊙. This confirms that WR 25 belongs to the family of very massive WNLh objects, a very poorly populated class in the Milky Way. We demonstrate that the absorption component in the P-Cygni profile of the He ii λ4859+Hβ line of WR 25 is mainly due to the O-type companion, and we propose to adopt a WN6ha spectral type for the WN component of the system. On the basis of the extracted disentangled spectra, we derive approximate physical parameters through a detailed CMFGEN analysis, given the difficulty in constraining both the extinction towards WR 25 and the definitive brightness ratio between the components. Finally, we propose the presence of a third component contributing to the spectrum of WR 25.