CAPOS: The bulge Cluster APOgee Survey XI. Unraveling the chemical composition of the bulge globular cluster NGC 6304

Abstract

Context. With CAPOS, we can mitigate the observational difficulties limiting access to bulge globular clusters in the optical and investigate them in more detail in the near-IR. Aims. To perform a rigorous abundance analysis of the metal-rich bulge globular cluster NGC 6304, in order to determine its detailed chemical composition and identify multiple populations. Methods. We analyzed APOGEE-2 near-IR spectra of 27 giant members. The abundances of 17 elements (C, N, O, Na, Mg, Al, Si, S, K, Ca, Ti, V, Cr, Mn, Fe, Ni, and Ce) were derived using the BACCHUS code, using atmospheric parameters from both ASPCAP and photometry (Gaia and 2MASS). Results. We derived $[{\rm Fe/H}] = -0.45\pm0.05$ using the ASPCAP parameters, and $[{\rm Fe/H}] = -0.45\pm0.08$ when using photometric parameters, with no evidence of an intrinsic metallicity spread. NGC 6304 shows $[α/{\rm Fe}]_{\rm spec} = 0.24\pm0.07$ and $[α/{\rm Fe}]_{\rm phot} = 0.23\pm0.08$. We find a significant spread in $[{\rm N/Fe}]$, with $σ_{\rm spec} = 0.54$ and $σ_{\rm phot} = 0.46$, along with a C-N anticorrelation. Furthermore, we detect a correlation of Ce with both N and Al, consistent with patterns observed in some metal-rich bulge globular clusters. Conclusions. We find a significant star-to-star variation in Na, but a minimal variation in O. The absence of the Mg-Al anticorrelation supports the evidence that the MgAl cycle is not active in globular clusters at high metallicity. The observed correlation between Ce and both N and Al suggests that the enrichment of these elements may be driven by asymptotic giant branch stars, positioning Ce as an element involved in multiple populations in metal-rich globular clusters. We find that abundances are consistent with those of bulge field stars of similar metallicity, suggesting a similar origin and chemical evolution.

Figures (9)

• Figure 1: Multiband (JHKs-combined color) infrared view of NGC 6304 obtained with the Gemini South Adaptive Optics Imager+Gemini Multi-Conjugate Adaptive Optics System (SAOI+GeMS; Haro et al. in preparation)
• Figure 2: Selection of NGC 6304 membership. Top left: Spatial distribution of observed stars within the APOGEE-2 survey area. The dashed black circumference indicates the tidal radius of the cluster ($t_{\mathrm{r}}$=15.25). NGC 6316 can be seen to the NE. Top Right: PM distribution of the stars within the tidal radius obtained from Gaia DR3. The dashed circumference has a radius of 0.5 [mas/yr]. Bottom left: Selection in the RV vs. [Fe/H] plane from ASPCAP for the candidate members. The point with error bars shows the mean RV and its standard deviation from the 2023MNRAS.521.3991B database. Bottom right: G vs. ($G_{\mathrm{bp}}$ - $G_{\mathrm{rp}}$) CMD. Final selected members are shown in all plots as points color-coded according to their S/N, while the field stars are represented as gray dots. The plotted isochrone of 12 Gyr represents the best PARSEC fit. Black squares in the plots are those stars with S/N > 70.
• Figure 3: Differences in atmospheric parameters, as well as the differences in [Fe/H] resulting from two BACCHUS runs, each adopting different values of $\rm T_{\rm eff}$ and log(g) derived from spectroscopic and photometric parameters. The vertical axis is the difference of each atmospheric parameter and [Fe/H], while the horizontal axis is the $\rm T_{\rm eff}^{\rm phot.}$. Points are color-coded according to S/N. The mean and standard deviation of the differences is represented in each panel by a black line and a gray shadow. These values for each parameter difference are specified at the lower left part of each panel.
• Figure 4: Quality of the model fit obtained with BACCHUS around the molecular lines ($^{12}$C$^{16}$O, $^{12}$C$^{14}$N, and $^{16}$OH) and atomic lines (Mg, Al, Si, Ca, Ti, Mn, Fe, Ni, and Ce) for the star 2M17143117-2930149 (S/N = 88). The dotted black line represents the observed spectrum, while the solid colored lines correspond to synthetic spectra with abundances varying by $\pm$ 0.3 dex, except for Mg, Al, Si, Ti, and Fe, whose abundances vary by $\pm$ 0.5 dex. The best-fitting synthetic spectrum is shown in red. Each panel is centered on the selected lines with the dashed black lines indicating the positions of the air wavelength lines.
• Figure 5: Comparison between $\sigma_{\rm tot}$ and $\sigma_{\rm obs}$ as a function of the atomic number (Z). Total error of our measurements and the observed error derived from BACCHUS using spectroscopic stellar parameters are represented by yellow and pink circles, respectively, while those obtained using photometric parameters are shown as blue and green triangles. Error bars indicate the error of the mean. Element abbreviations are shown for clarity.