2012A&A...544A..12G

It has been found that globular clusters host multiple stellar populations. In particular, in NGC 1851 the subgiant branch (SGB) can be divided into two components and the distribution of stars along the horizontal branch (HB) is multimodal. Various authors have found that NGC 1851 possibly has a spread in [Fe/H], but the relation between this spread and the division in the SGB is unknown. We obtained blue (3950-4600Å) intermediate resolution (R∼8000) spectra for 47 stars on the bright and 30 on the faint SGB of NGC 1851 (b-SGB and f-SGB, respectively). The spectra were analysed by comparing with synthetic spectra. The determination of the atmospheric parameters to extremely high internal accuracy allows small errors to be recovered when comparing different stars in the cluster, in spite of their faintness (V∼19). Abundances were obtained for Fe, C, Ca, Cr, Sr, and Ba. We found that the b-SGB is slightly more metal-poor than the f-SGB, with [Fe/H]=-1.227±0.009 and [Fe/H]=-1.162±0.012, respectively. This implies that the f-SGB is only slightly older by ∼0.6Gyr than the b-SGB if the total CNO abundance is constant. There are more C-normal stars in the b-SGB than in the f-SGB. This is consistent with what is found for HB stars, if b-SGB are the progenitors of red HB stars, and f-SGB those of blue HB ones. As previously found, the abundances of the n-capture elements Sr and Ba have a bimodal distribution, reflecting the separation between f-SGB (Sr and Ba-rich) and b-SGB stars (Sr and Ba-poor). In both groups, there is a clear correlation between [Sr/Fe] and [Ba/Fe], suggesting that there is a real spread in the abundances of n-capture elements. By looking at the distribution of SGB stars in the [C/H] versus (vs.) T_eff diagram and in the [Ba/H] vs. [Sr/H] diagram, not a one-to-one relation is found among these quantities. There is some correlation between C and Ba abundances, while the same correlation for Sr is much more dubious. We identified six C-rich stars, which have a moderate overabundance of Sr and Ba and rather low N abundances. This group of stars might be the progenitors of those on the anomalous RGB in the (v, v-y) diagram. These results are discussed within different scenarios for the formation of NGC 1851. It is possible that the two populations originated in different regions of a inhomogeneous parent object. However, the striking similarity with M 22 calls for a similar evolution for these two clusters. Deriving reliable CNO abundances for the two sequences would be crucial.