2010ApJ...724..975R

To better characterize the abundance patterns produced by the r-process, we have derived new abundances or upper limits for the heavy elements zinc (Zn, Z= 30), yttrium (Y, Z= 39), lanthanum (La, Z= 57), europium (Eu, Z= 63), and lead (Pb, Z= 82). Our sample of 161 metal-poor stars includes new measurements from 88 high-resolution and high signal-to-noise spectra obtained with the Tull Spectrograph on the 2.7 m Smith Telescope at the McDonald Observatory, and other abundances are adopted from the literature. We use models of the s-process in asymptotic giant branch stars to characterize the high Pb/Eu ratios produced in the s-process at low metallicity, and our new observations then allow us to identify a sample of stars with no detectable s-process material. In these stars, we find no significant increase in the Pb/Eu ratios with increasing metallicity. This suggests that s-process material was not widely dispersed until the overall Galactic metallicity grew considerably, perhaps even as high as [Fe/H] =-1.4, in contrast with earlier studies that suggested a much lower mean metallicity. We identify a dispersion of at least 0.5 dex in [La/Eu] in metal-poor stars with [Eu/Fe] <+0.6 attributable to the r-process, suggesting that there is no unique "pure" r-process elemental ratio among pairs of rare earth elements. We confirm earlier detections of an anti-correlation between Y/Eu and Eu/Fe bookended by stars strongly enriched in the r-process (e.g., CS 22892-052) and those with deficiencies of the heavy elements (e.g., HD 122563). We can reproduce the range of Y/Eu ratios using simulations of high-entropy neutrino winds of core-collapse supernovae that include charged-particle and neutron-capture components of r-process nucleosynthesis. The heavy element abundance patterns in most metal-poor stars do not resemble that of CS 22892-052, but the presence of heavy elements such as Ba in nearly all metal-poor stars without s-process enrichment suggests that the r-process is a common phenomenon.