2012ApJ...754...76D

We present a quantitative study on the properties at death of fast-rotating massive stars evolved at low-metallicity–objects that are proposed as likely progenitors of long-duration γ-ray bursts (LGRBs). We perform one-dimensional+rotation stellar-collapse simulations on the progenitor models of Woosley and Heger, and critically assess their potential for the formation of a black hole and a Keplerian disk (namely, a collapsar) or a proto-magnetar. We note that theoretical uncertainties in the treatment of magnetic fields and the approximate handling of rotation compromise the accuracy of stellar-evolution models. We find that only the fastest rotating progenitors achieve sufficient compactness for black hole formation while the bulk of models possess a core density structure typical of garden-variety core-collapse supernova (SN) progenitors evolved without rotation and at solar metallicity. Of the models that do have sufficient compactness for black hole formation, most of them also retain a large amount of angular momentum in the core, making them prone to a magneto-rotational explosion, therefore preferentially leaving behind a proto-magnetar. A large progenitor angular-momentum budget is often the sole criterion invoked in the community today to assess the suitability for producing a collapsar. This simplification ignores equally important considerations such as the core compactness, which conditions black hole formation, the core angular momentum, which may foster a magneto-rotational explosion preventing black hole formation, or the metallicity and the residual envelope mass which must be compatible with inferences from observed LGRB/SNe. Our study suggests that black hole formation is non-trivial, that there is room for accommodating both collapsars and proto-magnetars as LGRB progenitors, although proto-magnetars seem much more easily produced by current stellar-evolutionary models.