Southern massive stars at high angular resolution: Physical separations and mass ratios

Context. A key property of massive stars is their high degree of multiplicity, which can impact their evolution and end-of-life products. The Southern Massive Stars at High angular resolution ( SMASH +) survey uses interferometric and high-angular-resolution techniques to detect companions at intermediate separations, from about 1 mas to 8″, a domain that so far has remained largely unexplored. Aims. For this study, we converted the angular separations and magnitude contrasts into physical units, i.e., projected physical separations and mass ratios. We also derived the sensitivity of the survey for various physical and orbital parameters and we investigated the orbital separation and mass ratio distributions. Methods. We developed a spectral-type and luminosity class – H -band luminosity – mass calibration based on existing grids of physical parameters for O stars, and we used it to obtain the photometric distance to each system, correcting for all known companions within the 2MASS point-spread function. We also derived the individual masses of the primaries and of each detected companion. Results. The probability of detecting companions is very uniform within the sensitivity limits of the SMASH + survey, which can be considered almost complete for binaries with 1 < log( a / AU )< 4 and q = M 2 / M 1 > 0.2. The projected separations follow a uniform distribution in log separation. The mass ratios follow a power-law distribution f q ∝ q κ with κ values that seem to decrease toward larger separations. For resolved companions within 100 AU, we find κ <100 = −0.6 +0.9 −0.7 which is compatible with both the power-law distributions derived for spectroscopic binaries ( κ ≈ 0) and with what was proposed in an earlier study ( κ = −1.4 ± 0.4) based on about half the SMASH + sample in the same range. Beyond ≈1000 AU, we observe a clear lack of (near-)equal-mass companions, with an upper mass-ratio limit declining toward larger separations.