Abstract :
[en] Massive stars, the most luminous stars, are the true “cosmic engines” of our Universe. They eject large quantity of material throughout their life, which strongly influences their evolutionary path as well as their environment. An important feature of massive stars is their high rotational velocities that are either acquired at birth or due to the influence of a companion. Rotation is believed to transport nitrogen-rich and carbon/oxygen-poor material generated in the stellar core through the CNO cycle, to the surface. A way to test the e ciency of rotational mixing is to study the chemical composition at the surface of stars, in particular the fastest rotators.
The incentive for this study was the discovery, in the context of the VLT- FLAMES Survey of Massive Stars, of fast rotators exhibiting an unenriched nitrogen composition at their surface, contrary to predictions from single-star evolutionary models including rotation. However, their multiplicity may a ect this conclusion, since both rotation and abundances can change as a result of binary interactions. In this work, we combined, for the first time, a detailed surface abundance analysis with a radial-velocity study to quantify the importance of binary e ects. This work was conducted for a sample of 40 bright, OB fast rotators in our Galaxy. Statistical tests and period-search techniques revealed that ≥ 40% of our targets whose multiplicity status can be probed, are binary or binary candidates. We derived the projected rotational velocity of our targets and model atmosphere codes were then used to derive stellar parameters and surface abundances of all sample stars. This abundance study revealed a correlation between the helium and nitrogen abundances of our targets, which is predicted by the rotational mixing theory. Finally, we compared our results to predictions of single-star evolutionary models. We found that 10 – 20% of our 40 targets exhibit no enhancement of the [N/O] abundance ratio, in line with results of the VLT-FLAMES Survey of Massive Stars. The properties of only half of our sample are explained by such models, and surprisingly we also uncovered a quite common large abundance of helium at the surface of our targets. Modifying the di usion coe cient in single-star models and models of non-rotating mergers did not reproduce simultaneously both the observed helium abundances and the [N/O] abundance ratios. Binary models considering a mass-transfer episode can, however, reproduce the [N/O] values of the majority of our targets and even the helium abundances of some of the most helium-enriched targets, but they cannot explain stars displaying little helium enrichment but high [N/O] values. In conclusion, we found that not every feature of massive stars can be explained by models, suggesting that they lack a physical ingredient and thus require further improvements.
The second part of this thesis aimed at improving our knowledge of the X-ray emission of early B-type stars. We studied 11 such stars at high resolution thanks to two X-ray facilities, XMM-Newton and Chandra, doubling the number of B-stars analysed at high resolution. In many aspects, our study confirmed previous ones: early B-stars display rather narrow and unshifted lines arising from a warm (of typically 0.2 – 0.6 keV) plasma located at a few stellar radii over the stellar surface. We also found that abundances derived in the X-ray domain are in fair agreement with photospheric ones derived in the optical domain. Furthermore, most early B-stars are moderately bright X-ray emitters – though we also unexpectedly found that this X-ray emission varies, on short and/or long timescales, for half of our sample. A few stars display peculiar features: the presence of a very hot (1.6 – 4.4 keV) component and strong variations. These features suggest that the recorded X-ray emission may not be entirely linked to the B-stars, but could be contaminated by emission from a companion or an interaction with it. Indeed, in one case, HD 79351, a flare was detected, of a luminosity compatible with those from PMS stars, and which could be associated to its companion. Finally, the data used also led to the discovery of the second case of X-ray pulsations associated to beta Cephei activity.
