Hidden Binaries in Star-forming Regions

Abstract A significant fraction of, and possibly all, stars form in binary or multiple systems. For solar-mass stars in the Galactic field, the distribution of orbital separations is log-normal over 7 orders of magnitude, from 10 −2 to 10 5 au. In contrast, the separation distributions of systems in nearby star-forming regions paint a much more confusing picture. There appears to be an excess of systems in the separation range 10–1000 au, and recent high-resolution spectroscopic observations of close (<10 au) systems suggest a field-like distribution in some star-forming regions, but a possible excess with respect to the field in other regions. Furthermore, the resolution limit of numerical simulations of binary star formation is ∼1 au, and consequently, comparisons with the binary distributions in star-forming regions and in the field are restricted. In this paper, we demonstrate that these observational uncertainties and limitations in the simulations are potentially a much bigger problem than previously realised. We show that the log-normal separation distribution in the field can be reproduced by combining constituent binary populations whose initial separation distributions have a very different form from a log-normal. We also argue that the observed excess of binaries in the range 10–62 au in the Orion Nebula Cluster (ONC) compared to the Galactic field is not necessarily inconsistent with the field population, because the ONC is only one of many star-forming regions that populate the field. We propose that further observations of spectroscopic binaries in star-forming regions to probe and complete the <10 au parameter space are urgently needed.