2008AJ....135..722A

In the past there have been many discrepant studies about whether metal-poor (Population II) stars were deficient in binaries or had approximately the same binary frequency as metal-rich stars. We consider here the distributions of periods in five samples of binaries: (1) 138 field halo binaries with low metal abundances ([Fe/H] < -0.30) observed by Latham et al. and Goldberg et al., (2) 42 blue metal-poor binaries discovered by Preston & Sneden, (3) all 145 known FG dwarf binaries with [Fe/H] > -0.30 as listed on-line by Pourbaix, (4) 31 binaries among the well-studied FG dwarfs within 25 pc of the Sun, and (5) binaries in three open clusters with ages of 10⁸-10⁹ yr. The main result is that the metal-poor stars lack short-period binaries relative to the metal-rich stars; the period distributions are very different. This tells us two things. First, the determination of the frequency of discovered binaries for metal-poor and metal-rich stars will depend upon the equipment used: if one uses low-resolution spectrographs, one will detect many of the short-period binaries but not many of the long-period binaries and conclude that the metal-poor stars are deficient in binaries. If one uses higher resolution, as did Latham et al., one will detect many more binaries and conclude that there is no difference in binary frequency between metal-poor and metal-rich stars. Second, if binaries are formed in three-body interactions in clusters and, as n-body simulations show, the longer they remain in dense cluster environments, the harder (shorter periods) they become.