Where, oh where has the r-process gone?

We present a review of the possible sources for r-process nuclei (r-nuclei). It is known that there is as yet no self-consistent mechanism to provide abundant neutrons for a robust r-process in the neutrino-driven winds from nascent neutron stars. We consider that the heavy r-nuclei with mass numbers A > 130 (Ba and above) cannot be produced in the neutrino-driven winds. Nonetheless, the r-process and the neutrino-driven winds may be directly or indirectly related by some unknown additional mechanism, which, for example, could provide ejecta with very short dynamic timescales of less than or similar to 0.004 s. This undetermined mechanism must supply a neutron source within the same general stellar sites that undergo core collapse to produce the neutron star. Observational data on low-metallicity stars in the Galactic halo show that sites producing the heavy r-nuclei do not produce Fe or any other elements between N and Ge. Insofar as a forming neutron star is key to producing the heavy r-nuclei, then the only possible sources are supernovae resulting from collapse of O-Ne-Mg cores or accretion-induced collapse of white dwarfs, neither of which produce the elements of the Fe group or those of intermediate mass (above C and N). Observational evidence on s and r-nuclei in low-metallicity stars with high C and N abundances shows that the r-process is also active in binary systems. The nuclei with A similar to 90-110 produced by charged-particle reactions (CPR) in the neutrino-driven winds are in general present in metal-poor stars with high or low abundances of heavy r-nuclei. The CPR nuclei and the heavy r-nuclei are not strongly coupled. Some metal-poor stars show extremely high enrichments of heavy r-nuclei and have established that the abundance patterns of these nuclei are universally close to the solar abundance pattern of heavy r-nuclei. Using a template star with high enrichments of heavy r-nuclei and another with low enrichments we develop a two-component model based on the abundances of Eu (from sources for heavy r-nuclei) and Fe (from Fe core-collapse supernovae). This model gives very good quantitative predictions for the abundances of all the other elements in those metal-poor stars with [Fe/H]less than or similar to - 1.5 for which the Eu and Fe abundances are known. We attribute the CPR elements such as Sr, Y, and Zr to reactions in the neutrino-driven winds from a nascent neutron star and the heavy r-nuclei to the hypothecated true r-process. The CPR nuclei should be produced whenever a neutron star is formed regardless of whether heavy r-nuclei are produced or not. Using the two-component model we estimate the yield of the CPR element Sr to be similar to 10(-6) M-circle dot for a single neutron star formation event. Self-consistent astrophysical models are needed to establish that the CPR nuclei are common to the neutron stars produced in both sources for the heavy r-nuclei and those for Fe. We show that the observational data appear fully consistent with the two-component model. The specific mechanism and site for the production of heavy r-nuclei remains to be found. (c) 2007 Elsevier B.V. All rights reserved.