A Bayesian Analysis of Physical Parameters for 783 Kepler (Near-)Contact Binaries: Extreme-Mass-Ratio Systems and a New Mass Ratio versus Period Lower Limit

Abstract

Contact binary star systems represent the long-lived penultimate phase of binary evolution. Population statistics of their physical parameters inform understanding of binary evolutionary pathways and end products. We use light curves and new optical spectroscopy to conduct a pilot study of ten (near-)contact systems in the long-period ($P$>0.5 d) tail of close binaries in the Kepler field. We use PHOEBE light curve models to compute Bayesian probabilities on five principal system parameters. Mass ratios and third-light contributions measured from spectra agree well with those inferred from the light curves. Pilot study systems have extreme mass ratios $q$<0.32. Most are triples. Analysis of the unbiased sample of 783 0.15 d<$P$<2 d (near-)contact binaries results in 178 probable contact systems, 114 probable detached systems, and 491 ambiguous systems for which we report best-fitting and 16th/50th/84th percentile parameters. Contact systems are rare at periods $P$>0.5 d, as are systems with $q$>0.8. There exists an empirical mass ratio lower limit $q_{min}$($P$)$\approx$0.05--0.15 below which contact systems are absent, supporting a new set of theoretical predictions obtained by modeling the evolution of contact systems under the constraints of mass and angular momentum conservation. Pre-merger systems should lie at long periods and near this mass ratio lower limit, which rises from $q$=0.044 for $P$=0.74 d to $q$=0.15 at $P$=2.0 d. These findings support a scenario whereby nuclear evolution of the primary (more massive) star drives mass transfer to the primary, thus moving systems toward extreme $q$ and larger $P$ until the onset of the Darwin instability at $q_{min}$ precipitates a merger.