Superbubble dynamics in globular cluster infancy

II. Consequences for secondary star formation in the context of self-enrichment via fast-rotating massive stars

¹ Geneva Observatory, University of Geneva, 51 chemin des Maillettes, 1290 Versoix, Switzerland
e-mail: Martin.Krause@unige.ch
² Max-Planck-Institut für extraterrestrische Physik, Postfach 1312, Giessenbachstr., 85741 Garching, Germany
³ Excellence Cluster Universe, Technische Universität München, Boltzmannstrasse 2, 85748 Garching, Germany
⁴ IRAP, UMR 5277 CNRS and Université de Toulouse, 14 Av. E. Belin, 31400 Toulouse, France
⁵ Institut d’Astrophysique de Paris, UMR 7095 CNRS, Univ. P. & M. Curie, 98bis Bd. Arago, 75104 Paris, France

Received: 5 November 2012
Accepted: 5 February 2013

Abstract

Context. The self-enrichment scenario for globular clusters (GC) requires large amounts of residual gas after the initial formation of the first stellar generation. Recently, we found that supernovae may not be able to expel that gas, as required to explain their present-day gas-free state, and suggested that a sudden accretion onto the dark remnants at a stage when type II supernovae have ceased may plausibly lead to fast gas expulsion.

Aims. Here, we explore the consequences of these results for the self-enrichment scenario via fast-rotating massive stars (FRMS).

Methods. We analysed the interaction of FRMS with the intra-cluster medium (ICM), in particular where, when, and how the second generation of stars may form. From the results, we developed a timeline of the first  ≈ 40 Myr of GC evolution.

Results. Our previous results imply three phases during which the ICM is in a fundamentally different state, namely the wind bubble phase (lasting 3.5 to 8.8 Myr), the supernova phase (lasting 26.2 to 31.5 Myr), and the dark remnant accretion phase (lasting 0.1 to 4 Myr): (i) Quickly after the first-generation massive stars have formed, stellar wind bubbles compress the ICM into thin filaments. No stars may form in the normal way during this phase because of the high Lyman-Werner flux density. If the first-generation massive stars have equatorial ejections however, as we proposed in the FRMS scenario, accretion may resume in the shadow of the equatorial ejecta. The second-generation stars may then form due to gravitational instability in these disc, which are fed by both the FRMS ejecta and pristine gas. (ii) In the supernova phase the ICM develops strong turbulence, with characteristic velocities below the escape velocity. The gas does not accrete either onto the stars or onto the dark remnants in this phase because of the high gas velocities. The strong mass loss associated with the transformation of the FRMS into dark remnants then leads to the removal of the second-generation stars from the immediate vicinity of the dark remnants. (iii) When the supernovae have ceased, turbulence quickly decays, and the gas can once more accrete, now onto the dark remnants. As discussed previously, this may release sufficient energy to unbind the gas, and may happen fast enough so that a large fraction of less tightly bound first-generation stars are lost.

Conclusions. Studying the FRMS scenario for the self-enrichment of GCs in detail reveals the important role of the physics of the ICM for our understanding of the formation and early evolution of GCs. Depending on the level of mass segregation, this sets constraints on the orbital properties of the stars, in particular high orbital eccentricities, which likely has implications on the GC formation scenario.