CAPOS: The bulge Cluster APOgee Survey VIII. Final ASPCAP results for all clusters

TL;DR

CAPOS provides final ASPCAP-derived metallicities, alpha abundances, and radial velocities for 18 bulge/disk globular clusters using APOGEE-2S DR17, under a stringent membership framework that yields 303 high-SNR members and 125 lower-SNR members. A key methodological advance is the correction for 2P star metallicities, revealing precise mean [Fe/H] values to 0.05 dex and [Si/Fe] as a robust alpha proxy, with radial velocities accurate to about 3.4 km s$^{-1}$. The study finds no strong evidence for internal metallicity spreads in most clusters, presents a new chemo-dynamical classification distinguishing in situ vs ex situ GCs, and shows the bulge globular cluster metallicity distribution to be bimodal with peaks near $[Fe/H]\approx -1.08$ and $-0.45$ (metal-poor peak dominating); ex situ GCs are largely unimodal around $[Fe/H]\approx -1.6$. Comparisons with literature and BACCHUS highlight both general agreement and systematic differences in alpha- and metallicity analyses, motivating further homogeneous, near-IR abundance work and future SDSS-V Open Fiber observations to extend CAPOS.

Abstract

Bulge globular clusters(BGCs) are exceptional tracers of the formation and chemodynamical evolution of this oldest Galactic component. However, until now, observational difficulties have prevented us from taking full advantage of these powerful Galactic archeological tools. CAPOS, the bulge Cluster APOgee Survey, addresses this key topic by observing a large number of BGCs, most of which have been poorly studied. We aim to obtain accurate mean values for metallicity,[alpha/Fe],and radial velocity, as well as abundances for 11 other elements. We present final parameters based on ASPCAP for all 18 CAPOS BGCs. We carry out a stringent membership selection, finding 303 with SNR>70 and 125 with lower SNR. We reinforced the finding that stars with high [N/Fe] abundances show higher [Fe/H] than their lower [N/Fe] counterparts. Mg,Ca and global alpha abundances show similar trends, while Si is well-behaved. The [Fe/H] value of these 2nd population stars is corrected to derive the mean metallicity. Mean metallicities are determined to a precision of 0.05 dex,[alpha/Fe] to 0.06 dex, and radial velocity to 3.4 km/s. No clusters show strong evidence for internal metallicity variation, including M22. Abundances for 11 other elements using only 1st population stars are calculated and are generally in good agreement with the literature. We develope a new chemodynamical GC classification scheme, synthesizing several recent studies. We also compile up-to-date metallicities. The BGC metallicity distribution is bimodal, with peaks at [Fe/H]=-0.45 and -1.1, with the metal-poor peak strongly dominant, while exsitu GCs are unimodal, with a peak at -1.6. Surprisingly, we find only a small, statistically insignificant difference in the mean [Si/Fe] of in and exsitu GCs. The 4 GCs with the lowest [Si/Fe] values are all exsitu, relatively young, and 3 belong to Sagittarius, but no other correlations are evident.

Figures (12)

• Figure 1: Membership determination in NGC 6380, including location with respect to tidal radius (upper left), PM space (upper right), metallicity:RV space (lower left) and CMD (lower right). Shown are CAPOS final members in blue, with boxes around those with SNR$<$70, non-CAPOS members in yellow, and other stars observed by APOGEE in grey. The Vasiliev & Baumgardt (2021) mean RV and error are shown in cyan. The clustering seen in the grey points are due to two other GCs observed in the same CAPOS field.
• Figure 2: Comparing our membership determination in NGC 6380 vs. that of S24. Diagrams and point descriptions as in Figure \ref{['N6380members']}, but with yellow points indicating members selected by S24, and green pentagons showing duplications in their selection. The typical error in PM is about 0.01 mas $yr^{-1}$ for both pmRa and pmDec, and in the RV vs metallicity plot, typical errors are shown in red. S24 have included a number of stars which do not satisfy our more stringent membership requirements.
• Figure 3: a). $\Delta([Fe/H])$ (difference in [Fe/H] from the mean based on only 1P stars = those with ([N/Fe]$<$0.7) as a function of [N/Fe], A clear trend is found for 2P (higher [N/Fe]) stars to have a higher [Fe/H] than 1P stars in the same cluster, with the difference in metallicity increasing with increasing [N/Fe] for 2P stars. The best-fit line is shown for 2P stars. b). As in a) but for [Mg/Fe]. c). As in a) but for [Si/Fe]. d). As in a) but for [Ca/Fe]. The stars are color-coded according to metallicity in each panel.
• Figure 4: Mean abundance ratio of the Fe-peak elements Mn (top) and Ni (bottom) for each of our clusters (filled blue circles with error bars), compared with APOGEE bulge stars (asterisks - Rojas-Arriagada et al. 2020) and the general trend of disk stars (circles - Hayden et al. 2015). Our BGCs generally follow the bulge field-star trend but extend to lower metallicity.
• Figure 5: Mean [Mg/Mn] vs. [Al/Fe] for CAPOS clusters. We include the purported division between various origins from Horta et al. (2021).