The incidence of binaries in Globular Cluster stellar populations

TL;DR

This study measures binary star fractions in ten globular clusters, revealing a significantly lower binary fraction in second-generation stars, indicating denser formation environments for these stars compared to first-generation stars.

Contribution

It provides the first detailed binary fraction measurements for multiple populations in globular clusters, supporting models of dense environment formation for second-generation stars.

Findings

Binary fraction in clusters is about 2.2%.

Second-generation stars have a lower binary fraction (~1.2%) than first-generation stars (~4.9%).

Results suggest second-generation stars form in denser environments.

Abstract

Binary fraction and orbital characteristics provide indications on the conditions of star formation, as they shed light on the environment they were born in. Multiple systems are more common in low density environments rather than in higher density ones. In the current debate about the formation of Globular Clusters and their multiple populations, studying the binary incidence in the populations they host offers a crucial piece of information on the environment of their birth and their subsequent dynamical evolution. Through a multi-year observational campaign using FLAMES at VLT, we monitored the radial velocity of 968 Red-Giant Branch stars located around the half-light radii in a sample of 10 Galactic Globular Clusters. We found a total of 21 radial velocity variables identified as {\it bona fide} binary stars, for a binary fraction of 2.2%$\pm$0.5%. When separating the sample into first generation and second generation stars, we find a binary fraction of 4.9%$\pm$1.3% and 1.2%$\pm$0.4% respectively. Through simulations that take into account possible sources of bias in detecting radial velocity variations in the two populations, we show that the difference is significant and only marginally affected by such effects. Such a different binary fraction strongly suggests different conditions in the environment of formation and evolution of first and second generations stars, with the latter being born in a much denser environment. Our result hence strongly supports the idea that the second generation forms in a dense subsystem at the center of the loosely distributed first generation, where (loose) binaries are efficiently destroyed.