An updated stellar census of the Quintuplet cluster

Context. Found within the central molecular zone, the Quintuplet is one of the most massive young clusters in the Galaxy. As a consequence it offers the prospect of constraining stellar formation and evolution in extreme environments. However, current observations suggest that it comprises a remarkably diverse stellar population that is difficult to reconcile with an instantaneous formation event. Aims. To better understand the nature of the cluster our aim is to improve observational constraints on the constituent stars. Methods. In order to accomplish this goal we present Hubble Space Telescope/NICMOS+WFC3 photometry and Very Large Telescope/SINFONI+KMOS spectroscopy for ∼100 and 71 cluster members, respectively. Results. Spectroscopy of the cluster members reveals the Quintuplet to be far more homogeneous than previously expected. All supergiants are classified as either O7–8 Ia or O9–B0 Ia, with only one object of earlier (O5 I–III) spectral type. These stars form a smooth morphological sequence with a cohort of seven early-B hypergiants and six luminous blue variables and WN9-11h stars, which comprise the richest population of such stars of any stellar aggregate known. In parallel, we identify a smaller population of late-O hypergiants and spectroscopically similar WN8–9ha stars. No further H-free Wolf–Rayet (WR) stars are identified, leaving an unexpectedly extreme ratio of 13:1 for WC/WN stars. A subset of the O9–B0 supergiants are unexpectedly faint, suggesting they are both less massive and older than the greater cluster population. Finally, no main sequence objects were identifiable. Conclusions. Due to uncertainties over which extinction law to apply, it was not possible to quantitatively determine a cluster age via isochrone fitting. Nevertheless, we find an impressive coincidence between the properties of cluster members preceding the H-free WR phase and the evolutionary predictions for a single, non-rotating 60 M⊙ star; in turn this implies an age of ∼3.0–3.6 Myr for the Quintuplet. Neither the late O-hypergiants nor the low luminosity supergiants are predicted by such a path; we suggest that the former either result from rapid rotators or are the products of binary driven mass-stripping, while the latter may be interlopers. The H-free WRs must evolve from stars with an initial mass in excess of 60 M⊙ but it appears difficult to reconcile their observational properties with theoretical expectations. This is important since one would expect th