Probing the low-mass end of the companion mass function for O-type stars

Abstract

Context. Past observations of O-type stars in the Galaxy have shown that almost all massive stars are part of a binary or higher-order multiple system. Given the wide range of separations at which these companions are found, several observational techniques have been adopted to characterize them. Despite the recent advancements in interferometric and adaptive optics observations, contrasts greater than 4 in the H band have never been reached between 100 and 1000 mas. Aims. Using new adaptive optics (AO) assisted coronagraphic observations, we aim to study the multiplicity properties of a sample of 18 dwarf (or sub-giant) O stars in the galactic field and in OB associations to probe the existence of stellar companions in the angular separation range from 0.″15 to 6″ down to very low mass ratios. Methods. We used VLT/SPHERE to observe simultaneously with the IRDIS and IFS sub-systems 18 O-type stars within 6 kpc and ages between 1 and 5 Myr. The IFS YJH band observations have allowed us to probe the presence of sub-solar companions in a 1.7″ × 1.7″ field-of-view down to magnitude limits of ΔH = 10 at 0.″4. In the wider 12″ × 12″ IRDIS field-of-view, we reached contrasts of ΔK = 12 at 1″, enabling us to look for even fainter companions at larger angular separations and to probe the source density of the surrounding portion of the sky. Results. This paper presents five newly discovered intermediate (< 1″) separation companions, three of which are smaller than 0.2 M⊙. If confirmed by future analyses of proper motions, these new detections represent the lowest-mass companions ever found around O-type stars. Additionally, 29 other sources are found in the IRDIS field-of-view with spurious association probabilities smaller than 5%. Assuming that all sources detected within 1″ are physically bound companions, the observed (uncorrected for bias) fraction of companions for O-type stars between 150 and 900 mas is 0.39 ± 0.15, whereas it increases to 1.6 ± 0.3 in the separation range from 0.″9 to 6″. Conclusions. These findings clearly support the notion that massive stars form almost exclusively in multiple systems, serving as proof of concept that supports the application of larger AO-assisted coronagraphic surveys as a crucial step in placing constraints on the multiplicity properties of massive star companions in regions of the parameter space that have previously gone unexplored. These results also demonstrate that the companion mass function is populated down to the lowest stellar masses.