We present the abundances of Li, Na, Mg, Al, Si, Ca, Sc, Ti, V, Cr, Mn, Co, Ni, Zn, Sr, Y, Zr, and Ba in a wide metallicity range ($-2.8 \le$ [Fe/H] $\le -1.5$). Using the differential technique allowed us to obtain an unprecedented low level of scatter in our analysis, with standard deviations as low as 0.05 dex, and mean errors as low as 0.05 dex for [X/Fe]. By expanding our metallicity range with precise abundances from other works, we were able to precisely constrain Galactic chemical evolution models in a wide metallicity range ($-3.6 \le$ [Fe/H] $\le -0.4$). The agreements and discrepancies found are key for further improvement of both models and observations. We also show that the LTE analysis of Cr II is a much more reliable source of abundance for chromium, as Cr I has important NLTE effects. These effects can be clearly seen when we compare the observed abundances of Cr I and Cr II with GCE models. While Cr I has a clear disagreement between model and observations, Cr II is very well modeled. We confirm tight increasing trends of Co and Zn toward lower metallicities, and a tight flat evolution of Ni relative to Fe. Our results strongly suggest inhomogeneous enrichment from hypernovae. Our precise stellar parameters results in a low star-to-star scatter ($0.04$ dex) in the Li abundances of our sample, with a mean value about $0.4$ dex lower than the prediction from standard Big Bang Nucleosynthesis; we also study the relation between lithium depletion and stellar mass, but it is difficult to assess a correlation due to the limited mass range. We find two blue straggler stars, based on their very depleted Li abundances. One of them shows intriguing abundance anomalies, including a possible zinc enhancement, suggesting that zinc may have been also produced by a former AGB companion.
H. Reggiani, J. Melendez, C. Kobayashi, et. al.
Wed, 13 Sep 17
Comments: Accepted for publication in A&A. 35 pages, 7 tables, 9 figures