No Arabic abstract
Globular clusters (GCs) are known to host multiple stellar populations showing chemical anomalies in the content of light elements. The origin of such anomalies observed in Galactic GCs is still debated. Here we analyse data compiled from the Hubble Space Telescope, ground-based surveys and Gaia DR2 and explore relationships between the structural properties of GCs and the fraction of second population (2P) stars. Given the correlations we find, we conclude that the main factor driving the formation/evolution of 2P stars is the cluster mass. The existing strong correlations between the 2P fraction and the rotational velocity and concentration parameter could derive from their correlation with the cluster mass. Furthermore, we observe that increasing cluster escape velocity corresponds to higher 2P fractions. Each of the correlations found is bimodal, with a different behaviour detected for low and high mass (or escape velocity) clusters. These correlations could be consistent with an initial formation of more centrally concentrated 2P stars in deeper cluster potentials, followed by a long-term tidal stripping of stars from clusters outskirts. The latter are dominated by the more extended distributed first population (1P) stars, and therefore stronger tidal stripping would preferentially deplete the 1P population, raising the cluster 2P fraction. This also suggests a tighter distribution of initial 2P fractions than observed today. In addition, higher escape velocities allow better retention of low-velocity material ejected from 1P stars, providing an alternative/additional origin for the observed differences and the distributions of 2P fractions amongst GCs.
Evidence that the multiple populations (MPs) are common properties of globular clusters (GCs) is accumulated over the past decades from clusters in the Milky Way and in its satellites. This finding has revived GC research, and suggested that their formation at high redshift must have been a much-more complex phenomenon than imagined before. However, most information on MPs is limited to nearby GCs. The main limitation is that most studies on MPs rely on resolved stars, facing a major challenge to investigate the MP phenomenon in distant galaxies. Here we search for integrated colors of old GCs that are sensitive to the multiple-population phenomenon. To do this, we exploit integrated magnitudes of simulated GCs with MPs, and multi-band Hubble Space Telescope photometry of 56 Galactic GCs, where MPs are widely studied, and characterized as part of the UV Legacy Survey of Galactic GCs. We find that both integrated $C_{rm F275W,F336W,F438W}$ and $m_{rm F275W}-m_{rm F814W}$ colors strongly correlate with the iron abundance of the host GC. In second order, the pseudo two-color diagram built with these integrated colors is sensitive to the MP phenomenon. In particular, once removed the dependence from cluster metallicity, the color residuals depend on the maximum internal helium variation within GCs and on the fraction of second-generation stars. This diagram, which we define here for Galactic GCs, has the potential of detecting and characterizing MPs from integrated photometry of old GCs, thus providing the possibility to extend their investigation outside the Local Group.
Our FLAMES survey of Na-O anticorrelation in globular clusters (GCs) is extended to NGC 4833, a metal-poor GC with a long blue tail on the horizontal branch (HB). We present the abundance analysis for a large sample of 78 red giants based on UVES and GIRAFFE spectra acquired at the ESO-VLT. We derived abundances of Na, O, Mg, Al, Si, Ca, Sc, Ti, V, Cr, Mn, Fe, Co, Ni, Cu, Zn, Y, Ba, La, Nd. This is the first extensive study of this cluster from high resolution spectroscopy. On the scale of our survey, the metallicity of NGC 4833 is [Fe/H]=-2.015+/-0.004+/-0.084 dex (rms=0.014 dex) from 12 stars observed with UVES, where the first error is from statistics and the second one refers to the systematic effects. The iron abundance in NGC 4833 is homogeneous at better than 6%. On the other hand, the light elements involved in proton-capture reactions at high temperature show the large star-to-star variations observed in almost all GCs studied so far. The Na-O anticorrelation in NGC 4833 is quite extended, as expected from the high temperatures reached by stars on the HB, and NGC 4833 contains a conspicuous fraction of stars with extreme [O/Na] ratios. More striking is the finding that large star-to-star variations are seen also for Mg, which spans a range of more than 0.5 dex in this GC. Depletions in Mg are correlated to the abundances of O and anti-correlated with Na, Al, and Si abundances. This pattern suggests the action of nuclear processing at unusually high temperatures, producing the extreme chemistry observed in the stellar generations of NGC 4833. This extreme changes are also seen in giants of the much more massive GCs M 54 and omega Cen, and our conclusion is that NGC 4833 has probably lost a conpicuous fraction of its original mass due to bulge shocking, as also indicated by its orbit.
NGC 4833 is a metal-poor Galactic globular cluster (GC) whose multiple stellar populations present an extreme chemical composition. The Na-O anti-correlation is quite extended, which is in agreement with the long tail on the blue horizontal branch, and the large star-to-star variations in the [Mg/Fe] ratio span more than 0.5 dex. Recently, significant excesses of Ca and Sc with respect to field stars of a similar metallicity were also found, signaling the production of species forged in H-burning at a very high temperature in the polluters of the first generation in this cluster. Since an enhancement of potassium is also expected under these conditions, we tested this scenario by analysing intermediate resolution spectra of 59 cluster stars including the K I resonance line at 7698.98 A. We found a wide spread of K abundances, anti-correlated to Mg and O abundances, as previously also observed in NGC 2808. The abundances of K are found to be correlated to those of Na, Ca, and Sc. Overall, this chemical pattern confirms that NGC 4833 is one of the relatively few GCs where the self-enrichment from first generation polluters occurred at such high temperatures that proton-capture reactions were able to proceed up to heavier species such as K and possibly Ca. The spread in K observed in GCs appears to be a function of a linear combination of cluster total luminosity and metallicity, as other chemical signatures of multiple stellar populations in GCs.
We present the abundance analysis of 82 red giant branch stars in the dense, metal-poor globular cluster NGC 6093 (M 80), the largest sample of stars analyzed in this way for this cluster. From high resolution UVES spectra of 14 stars and intermediate resolution GIRAFFE spectra for the other stars we derived abundances of O, Na, Mg, Al, Si, Ca, Sc, Ti, V, Cr, Mn, Fe, Co, Ni, Cu, Zn, Y, Zr, Ba, La, Ce, Pr, Nd, Sm, Eu. On our UVES metallicity scale the mean metal abundance of M 80 is [Fe/H]=-1.791+/-0.006+/-0.076 (+/-statistical +/-systematic error) with rms=0.023 (14 stars). M 80 shows star to star variations in proton-capture elements, and the extension of the Na-O anticorrelation perfectly fit the relations with (i) total cluster mass, (ii) horizontal branch morphology, and (iii) cluster concentration previously found by our group. The chemistry of multiple stellar populations in M 80 does not look extreme. The cluster is also a typical representative of halo globular clusters for what concerns the pattern of alpha-capture and Fe-group elements. However we found that a significant contribution from the s-process is required to account for the distribution of neutron-capture elements. A minority of stars in M 80 seem to exhibit slightly enhanced abundances of s-process species, compatible with those observed in M 22 and NGC 1851, although further confirmation from larger samples is required.
It has become clear in recent years that globular clusters are not simple stellar populations, but may host chemically distinct sub-populations, typically with an enhanced helium abundance. These helium-rich populations can make up a substantial fraction of all cluster stars. One of the proposed formation channels for blue straggler stars is the physical collision and merger of two stars. In the context of multiple populations, collisions between stars with different helium abundances should occur and contribute to the observed blue straggler population. This will affect the predicted blue straggler colour and luminosity function. We quantify this effect by calculating models of mergers resulting from collisions between stars with different helium abundances and using these models to model a merger population. We then compare these results to four observed clusters, NGC 1851, NGC 2808, NGC 5634 and NGC 6093. As in previous studies our models deviate from the observations, particularly in the colour distributions. However, our results are consistent with observations of multiple populations in these clusters. In NGC 2808, our best fitting models include normal and helium enhanced populations, in agreement with helium enhancement inferred in this cluster. The other three clusters show better agreement with models that do not include helium enhancement. We discuss future prospects to improve the modelling of blue straggler populations and the role that the models we present here can play in such a study.