Investigating globular cluster abundance anomalies in different environments

Abstract

The origin of the star-to-star abundance variations for light elements shown in globular clusters (GCs) is not well understood, which represents a significant problem for stellar astrophysics. While the light element abundance variations are very common in globular clusters, they are comparatively rare in the galactic halo field population. However, relatively little is known regarding the occurrence of the abundance anomalies in other environments such as dwarf spheroidal galaxies and massive star clusters younger than the ancient globular clusters. Consequently, spectra obtained at Gemini-S with the GMOS spectrograph and at the AAT with the 2dF/AAOmega instrument have been used to investigate the anti-correlation and bimodality of CH and CN band strengths through the measurement of CH(lambda ~ 4300), S(3839) and S(4142) indices for 45 red giants stars in the Sculptor dwarf spheroidal galaxy. Using a calibration from five GCs (NGC 1851, NGC 288, NGC 6752, M 55 and M 30) we also have studied the CN/Na correlation in the Sculptor red giant sample. Our results indicate that variations analogous to those seen in GCs are not present in our Sculptor sample. Instead we find a positive correlation between CH and CN, and no correlation between Na and CN. Also, a deficiency of [Na/Fe] in Sculptor relative to GCs has been found that is consistent with previous work for dSph galaxies. The outcome reinforces the apparent need for a high stellar density environment to produce the light element abundance variations. We also report the discovery of a third carbon-enhanced metal-poor star in Sculptor that shows evidence of a strong s-process element enhancement. In addition, spectra obtained with the VLT/FORS and Gemini-S/GMOS instruments have been used to investigate CN and CH band strengths along with carbon and nitrogen abundances in a sample of massive star clusters in the Small Magellanic Cloud. The clusters have metallicity and luminosity comparable to GCs but are younger. The clusters investigated are Lindsay 1, Kron 3 and NGC 339, which have ages between 6 and 7.5 Gyr. Our results for Lindsay 1 and NGC 339 are not as clear-cut as they are for Kron 3, where the large spread in nitrogen (~ 1.2 dex), and a well-established bimodality of CH and CN band strengths with a CN/CH anti-correlation support the presence of multiple populations. Due to the possibility of evolutionary mixing on the red giant branch, the existence of CN-CH anti-correlation is not sufficient in itself to indicate the presence of abundance anomalies. Consequently, we have investigated the Na, CN correlation through the measurement of sodium D-line strengths and Na abundances, finding a spread in sodium of ~ 0.8 dex in Kron 3. For this cluster we therefore confirm a positive correlation between nitrogen and sodium abundances. These results add to existing photometric and spectroscopic indications of abundance anomalies in clusters of intermediate age, confirming that the mechanism responsible for the multiple populations cannot be an early cosmological effect applying only to old GGCs. This outcome is a key factor in gaining understanding on the origin of the abundance anomalies.