The IACOB project: CVIII. Hunting for spectroscopic binaries in the O and B supergiant domain.The threat of pulsational variability

TL;DR

This study estimates the binary fraction among Galactic OB supergiants using high-resolution spectra, revealing intrinsic variability can mimic binarity and highlighting the importance of careful analysis to understand massive star evolution.

Contribution

It provides the first empirical estimate of the spectroscopic binary fraction in OB supergiants and establishes thresholds to distinguish pulsational variability from true binarity.

Findings

Approximately 10% of OB supergiants are spectroscopic binaries.

Intrinsic variability can cause RV amplitudes up to 25 km/s.

Binary fraction decreases significantly from O stars to B supergiants.

Abstract

Observations have definitively strengthened the long-standing assertion that binaries are crucial in massive star evolution. While the percentage of spectroscopic binary systems among main-sequence O stars is well-studied, other phases of massive star evolution remain less explored. We aim to estimate the spectroscopic binary fraction in Galactic late O- and B-type supergiants (OB-Sgs) and set empirical thresholds in radial velocity (RV) to avoid misidentifying pulsating stars as single-line spectroscopic binaries. Using over 4500 high-resolution spectra of 56 Galactic OB-Sgs (plus 13 O dwarfs/subgiants and 5 early-B giants) from the IACOB project (2008-2020), we apply Gaussian fitting and centroid computation techniques to measure RV for each spectrum. Our findings reveal that intrinsic variability in OB-Sgs can result in peak-to-peak RV amplitudes (RVpp) of up to 20-25 km/s, notably in late-O and early-B Sgs, and decreases to typical values of RVpp in the range of 1-5 km/s for O dwarfs and 2-15 km/s for late B-Sgs. Considering these results and evaluating line-profile variability in each star, we find that 10$\pm$4% of OB-Sgs are clearly single-line spectroscopic binaries. In addition, we find that the percentage of double-line spectroscopic binaries in late O- and early B-Sgs is ~6%, much lower than the ~30% in O-type dwarfs and giants. This study, along with prior research on B-Sgs in the 30 Doradus region of the LMC, indicates that the spectroscopic binary percentage decreases by a factor of 4-5 from O stars to B-Sgs. Our study underscores the need for a thorough characterization of spectroscopic variability due to intrinsic sources to reliably determine the spectroscopic binary fraction among OB-Sgs and O-type stars in general, offering valuable insights into the impact of binaries on massive star evolution.