We report the discovery of eight lithium-rich field giants found in a high resolution spectroscopic sample of over 700 metal-poor stars ([Fe/H]<-0.5) selected from the RAVE survey. The majority of the Li-rich giants in our sample are very metal-poor
([Fe/H]<-1.9), and have a Li abundance (in the form of 7Li), A(Li)=log(n(Li)/n(H))+12, between 2.30 and 3.63, well above the typical upper red giant branch limit, A(Li)<0.5, while two stars, with A(Li)~1.7-1.8, show similar lithium abundances to normal giants at the same gravity. We further included two metal-poor, Li-rich globular cluster giants in our sample, namely the previously discovered M3-IV101 and newly discovered (in this work) M68-A96. This comprises the largest sample of metal-poor Li-rich giants to date. We performed a detailed abundance analysis of all stars, finding that the majority our sample stars have elemental abundances similar to that of Li-normal halo giants. Although the evolutionary phase of each Li-rich giant cannot be definitively determined, the Li-rich phase is likely connected to extra mixing at the red giant branch bump or early asymptotic giant branch that triggers cool bottom processing in which the bottom of the outer convective envelope is connected to the H-burning shell in the star. The surface of a star becomes Li-enhanced as 7Be (which burns to 7Li) is transported to the stellar surface via the Cameron-Fowler mechanism. We discuss and discriminate among several models for the extra mixing that can cause Li-production, given the detailed abundances of the Li-rich giants in our sample.
We have undertaken the study of the elemental abundances and kinematic properties of a metal-poor sample of candidate thick-disk stars selected from the RAVE spectroscopic survey of bright stars to differentiate among the present scenarios of the for
mation of the thick disk. In this paper, we report on a sample of 214 red giant branch, 31 red clump/horizontal branch, and 74 main-sequence/sub-giant branch metal-poor stars, which serves to augment our previous sample of only giant stars. We find that the thick disk [alpha/Fe] ratios are enhanced, and have little variation (<0.1 dex), in agreement with our previous study. The augmented sample further allows, for the first time, investigation of the gradients in the metal-poor thick disk. For stars with [Fe/H] < -1.2, the thick disk shows very small gradients, <0.03 +/- 0.02 dex/kpc, in alpha-enhancement, while we find a +0.01 +/- 0.04 dex/kpc radial gradient and a -0.09 +/- 0.05 dex/kpc vertical gradient in iron abundance. In addition, we show that the peak of the distribution of orbital eccentricities for our sample agrees better with models in which the stars that comprise the thick disk were formed primarily in the Galaxy, with direct accretion of stars contributing little. Our results thus disfavor direct accretion of stars from dwarf galaxies into the thick disk as a major contributor to the thick disk population, but cannot discriminate between alternative models for the thick disk, such as those that invoke high-redshift (gas-rich) mergers, heating of a pre-existing thin stellar disk by a minor merger, or efficient radial migration of stars.
Very metal-poor stars are of obvious importance for many problems in chemical evolution, star formation, and galaxy evolution. Finding complete samples of such stars which are also bright enough to allow high-precision individual analyses is of consi
derable interest. We demonstrate here that stars with iron abundances [Fe/H] < -2 dex, and down to below -4 dex, can be efficiently identified within the Radial Velocity Experiment (RAVE) survey of bright stars, without requiring additional confirmatory observations. We determine a calibration of the equivalent width of the Calcium triplet lines measured from the RAVE spectra onto true [Fe/H], using high spectral resolution data for a subset of the stars. These RAVE iron abundances are accurate enough to obviate the need for confirmatory higher-resolution spectroscopy. Our initial study has identified 631 stars with [Fe/H] <= -2, from a RAVE database containing approximately 200,000 stars. This RAVE-based sample is complete for stars with [Fe/H] < -2.5, allowing statistical sample analysis. We identify three stars with [Fe/H] <= -4. Of these, one was already known to be `ultra metal-poor, one is a known carbon-enhanced metal-poor star, but we obtain [Fe/H]= -4.0, rather than the published [Fe/H]=-3.3, and derive [C/Fe] = +0.9, and [N/Fe] = +3.2, and the third is at the limit of our S/N. RAVE observations are on-going and should prove to be a rich source of bright, easily studied, very metal-poor stars.
