Detection of single-degenerate massive binaries with Gaia: The impact of blue supergiants, triples, mass precision, and high-precision parallax requirements

Abstract

Context. X-ray-quiet single-degenerate massive binaries are notoriously difficult to detect, and only a few have been identified so far. Yet, recent investigations have shown that hundreds of black holes (BHs) with massive main-sequence (MS) companions (OB+BHs) might be identifiable from Gaia astrometry by using the astrometric mass-ratio function (AMRF). Aims. We aim to investigate a number of biases that can have an impact on the astrometric identification of OB+BH binaries, namely the presence of blue supergiant (BSG) companions instead of dwarfs and the presence of additional companions in the system that are not resolved by Gaia. We also explore the accuracy with which the primary mass needs to be constrained. Moreover, we assess the impact of high-precision constraints on the detection of binaries by using the conservative constraints imposed to release astrometric orbits in the latest Gaia data release, DR3. We then investigate how much less stringent these constraints need to be in order to obtain information on the BH-formation scenario. Methods. We established a mass-magnitude relation of BSGs and from this computed BSG AMRF curves. A mock population of OB/BSG+BH binaries, non-degenerate binaries with an OB or BSG primary (OB/BSG+MS), and triples with OB and BSG primaries is used to determine the fraction of false-positive identifications and the effect of the BSG mass-magnitude relation. We compare the number of sources with astrometric DR3 orbits in the second Alma Luminous Star catalogue with new predictions regarding the detection of OB+BHs using the conservative selection criterion used for publishing astrometric orbits in DR3. Results. We show that the addition of systems with BSG primaries does not significantly impact the fraction of false-positive identifications. Only for triple systems where the outer star is more luminous and a BSG will the usage of the previously established MS curves potentially result in a high amount of false-positive identifications. However, such systems are expected to be rare. We also demonstrate that the mass of the primary does not need to be accurately known to benefit from both the high identification fraction of OB/BSG+BHs and the low fraction of false positives. We find that 11 sources have an astrometric binary orbit available in DR3. None of these sources are OB/BSG+BH candidates. This is in line with the new predictions that use the Gaia DR3 selection criterion. Conclusions. If the evolutionary stages of the primary stars are unknown, the usage of the BSG curves is recommended over the MS curves to avoid high contamination from BSG+MS systems or triples with a more luminous outer star. This way, the false-positive fractions are decreased by an order of magnitude, reaching values ≪1%. However, the fraction of identifiable OB+BHs is also significantly reduced (from 68% to 29%). If the mass of the primary star is not known, it is possible to use a fixed estimate. The non-detection of astrometric OB/BSG+BH systems cannot be attributed to the underlying BH-formation scenario, but rather to the stringent selection criterion imposed on the parallax relative uncertainty in DR3. A relaxation of this condition would be needed to find the bulk of the OB+BH population with Gaia. If possible, we propose that the constraint on the relative parallax precision in DR4 be improved to n × (ϖ/σϖ)DR3,single >  n × 1000/Pday, with (ϖ/σϖ)DR3,single the relative parallax precision for the single source solution in DR3 and n the DR4 improvement of the relative precision compared to the DR3 single star solutions. In DR3, this means a 95% decrease in the original criterion.