Elemental abundance ratios in stars of the outer Galactic disk. I. Open clusters

We summarize radial velocity studies of selected stars in the old, distant clusters Berkeley 20, Berkeley 21, NGC 2141, Berkeley 29, and Berkeley 31. Cluster members are identified using optical and infrared color-magnitude diagrams, as well as radial velocities derived from high-resolution echelle spectra. Three members of M67 were observed similarly, and those velocities compare extremely well with prior measures. Mean cluster radial velocities are determined. We also employ the highest quality spectra to analyze the chemical compositions of all six clusters for [Fe/H], as well as abundances of alpha-elements, iron-peak elements, and those synthesized in either the s-process or the r-process. In Be 21 our observed star is found to be rotating rapidly and overabundant in lithium, the second Li-rich star found in this sparse cluster. We confirm the lack of correlation between abundance and age. For the outer disk, the abundance gradient for [Fe/H] deviates from the trend defined near the solar neighborhood. Rather than declining with increasing galactocentric distance, [Fe/H] appears to reach a basement'' at [Fe/H] approximate to -0.5 beyond R-GC approximate to 10-12 kpc. Our radial abundance distribution for [Fe/H] is not inconsistent with the radial abundance discontinuity exhibited by Cepheids. We find enhanced [O/Fe], [alpha/Fe], and [Eu/Fe] in the outer disk, revealing a rapid star formation history. The outer disk also exhibits enhancements for s-process elements. We compare the open cluster compositions with those of the thin disk, thick disk, halo, bulge, and dwarf spheroidal galaxies. None of these stellar populations perfectly matches the abundance ratios of the outer disk open clusters. Several key points arise from these comparisons: ( 1) [O/Fe] and [alpha/Fe] resemble those of the thick disk. ( 2) [Na/Fe] and [Al/Fe] are enhanced relative to those of the thin disk. ( 3) [Ni/ Fe] and [Mn/Fe] are in accord with those of the thin disk, while [Co/Fe] may be slightly enhanced. ( 4) The neutron-capture elements indicate different ratios of s-process to r-process material, with no cluster showing a pure r-process distribution. ( 5) An unusual pattern exists among the alpha-elements, with [[Mg + Ti]/Fe] enhanced while [[Si + Ca]/Fe] is normal. Similar abundance ratios have been reported for Galactic bulge giants and indicate a common but not necessarily shared nucleosynthetic history between the bulge and the outer disk. Enhanced ratios of [Al/Fe] and [Co/Fe] offer another possible similarity between the bulge and the outer disk. An intriguing but tentative conclusion is that the outer disk open cluster abundance ratios are consistent with the outer disk being formed via a merger event or series of merger events. The basement in [Fe/H] and enhanced [alpha/Fe] suggest that the outer disk formed from a reservoir of gas with a star formation history distinct from the solar neighborhood. That the open clusters may be associated with an accreted dwarf galaxy or galaxies is appealing, since the clusters are young and have [alpha/Fe] ratios indicating a rapid star formation history. However, the high [alpha/Fe] ratios are unlike those seen in any current dwarf galaxies at the same [Fe/H]. Therefore, the open clusters may have formed as a result of star formation triggered by a merger event or series of mergers in the outer disk. The ages of the outer disk open clusters would then be a measure of when the merger(s) occurred. However, Be 29 is a candidate merger member, while Be 31 is not. One problem with the merger scenario is that open clusters with presumably very different origins have similar and unusual compositions.