The hydrostatic alphaelements oxygen and magnesium

We first look at the two alpha-elements primarily made in hydrostatic equilibrium in massive stars: oxygen and magnesium. Oxygen is primarily produced during the He-burning phases, whereas magnesium is made during the C-burning phases. The yields of both are proportional to the mass of the respective burning shell (Woosley & Weaver 1995, Timmes et al. 1995). This means that, if there is no mass loss, more massive stars will have larger shells, and therefore produce more O and Mg than do lower-mass stars. Models of Type-Ia SNe do not produce significant amounts of either O or Mg.

We plot the observed bulge [O/Fe] ratios in Figure 12.1. The top panel shows only the bulge field stars and in the lower panel we add in the bulge globular clusters.

[Fe/H]

Figure 12.1. Top panel: [O/Fe] versus [Fe/H] for bulge field stars from three surveys in Baade's Window. The solid line gives a rough indication of where disk and halo field stars lie. Field stars in the bulge lie along, or slightly above, the disk-halo relation at the metal-poor end, but have higher [O/Fe] values for the most metal-rich stars. Dashed lines connect multiple observations of the same star, as discussed in the text. Two metal-poor Baade's Window stars have lower [O/Fe] and higher [Na/Fe] and [Al/Fe] ratios than other metal-poor bulge stars. They lie along the Na-O anti-correlation, as if they were globular-cluster stars like those in M4. Bottom panel: the same plot as the top panel, but here all the field stars are marked by crosses and the large symbols reflect the mean values from studies of bulge globular clusters. The triangles mark data from Cohen et al. and Carretta et al. (see Table 12.1), the square points and HP-1 data from Barbuy et al. and Zoccali et al., and the pentagons and stars data from Origlia, Rich, and collaborators (the order of clusters given on the star-symbol label is that of increasing metallicity). The bulge cluster stars lie in the same region as the field stars.

Figure 12.1. Top panel: [O/Fe] versus [Fe/H] for bulge field stars from three surveys in Baade's Window. The solid line gives a rough indication of where disk and halo field stars lie. Field stars in the bulge lie along, or slightly above, the disk-halo relation at the metal-poor end, but have higher [O/Fe] values for the most metal-rich stars. Dashed lines connect multiple observations of the same star, as discussed in the text. Two metal-poor Baade's Window stars have lower [O/Fe] and higher [Na/Fe] and [Al/Fe] ratios than other metal-poor bulge stars. They lie along the Na-O anti-correlation, as if they were globular-cluster stars like those in M4. Bottom panel: the same plot as the top panel, but here all the field stars are marked by crosses and the large symbols reflect the mean values from studies of bulge globular clusters. The triangles mark data from Cohen et al. and Carretta et al. (see Table 12.1), the square points and HP-1 data from Barbuy et al. and Zoccali et al., and the pentagons and stars data from Origlia, Rich, and collaborators (the order of clusters given on the star-symbol label is that of increasing metallicity). The bulge cluster stars lie in the same region as the field stars.

Oxygen has been observed in all three of our bulge field-star surveys. Fulbright et al. (2006b) used the [O i] 6,300 A line, whereas the near-IR groups used the OH molecular absorption lines available in that wavelength region.

There are a few cases in which the same star has been analyzed by two different groups. The multiple observations of these stars (three in common between Fulbright et al. (2006b) and Cunha & Smith (2006), two in common between Fulbright et al. (2006b) and Rich & Origlia (2005)) are connected by dotted lines. In the former case the disagreement can be large - but in two of the three cases the oxygen abundances are in good agreement, and it is the Fe abundances that differ by sizable amounts. In the latter case, the agreement is very good and the points are very close to each other in Figure 12.1.

Also plotted is a solid line representing the rough trend of values seen in disk and halo field stars. The bulge [O/Fe] ratios lie at or above the disk/halo relations at the highest metallicities both for the field and for the cluster stars. At the highest metallicities, the [O/Fe] ratio is about +0.3 dex higher. The basic interpretation is the ratio of Type-II to Type-Ia supernovae contributions in the bulge has to be higher than that in the local disk.

In the panels of Figure 12.2, we see a similar result - the [Mg/Fe] values of the metal-rich bulge stars and clusters lie well above that of the metal-rich disk. However, the [Mg/Fe] ratios undergo only a mild decline at the highest metallicities: [Mg/Fe] « + 0.3 at [Fe/H] = +0.5. The [Mg/O] ratio rises from roughly Solar at [Fe/H] = -0.5 to + 0.3 at [Fe/H] = +0.5. This effect cannot be reproduced by the inclusion of Type-Ia ejecta because these supernovae do not produce significant amounts of either element. If O and Mg are produced in the shells of massive stars, then how can the yields change with increasing metallicity?

One possible answer is that mass loss by the Type-II progenitor could possibly decrease the oxygen yield by lowering the mass of the He-burning shell without significantly affecting the mass of the C-burning shell. The details of this scenario are given in Fulbright (2006b), but there is only mixed observational and theoretical support for this type of Wolf-Rayet mass loss altering oxygen yields at high metallicity.

The abundances of the light odd-Z elements sodium and aluminum lend more support for the scenario that the high [Mg/Fe] values at high metallicity are the true measure of the Type-II contribution of the bulge. Both [Na/Fe] and [Al/Fe] ratios for bulge stars stay far above Solar at all metallicities, like [Mg/Fe] ratios. Neither Na nor Al is predicted to be made in Type-Ia supernovae, so, again, the only option is that the chemical evolution of the bulge is dominated by Type-II ejecta.

Exceptions to the high [Mg/Fe] ratios in the bulge are the cluster HP-1 and possibly NGC 6528. The [Mg/Fe] ratio for the metal-poor cluster HP-1 lies well below what is expected for metal-poor systems. One solution is that this cluster

Fulbright et al. 2006 • J Rich & Origlia 2005 # Halo/Disk Stars —

5s • ;

i ,

Figure 12.2. Top panel: [Mg/Fe] versus [Fe/H] for bulge field stars from two surveys in Baade's Window. The solid line gives a rough indication of where disk and halo field stars lie. Field stars in the bulge lie above the disk-halo relation at all metallicities, with only slightly lower [Mg/Fe] values for the most metal-rich stars. This is among the strongest evidence that the chemical evolution of the bulge was dominated by Type-II supernovae. Bottom panel: the same plot as the top panel, but here all the field stars are marked by crosses and the large symbols (the same ones as in Figure 12.1) reflect the mean values from studies of bulge globular clusters. As with [O/Fe], the bulge globular clusters lie in similar regions, although two of the three observations of NGC 6528 lie on the lower envelope of the field-star distribution. Data for the cluster HP-1 lie well below both the bulge and the disk-halo relations.

Figure 12.2. Top panel: [Mg/Fe] versus [Fe/H] for bulge field stars from two surveys in Baade's Window. The solid line gives a rough indication of where disk and halo field stars lie. Field stars in the bulge lie above the disk-halo relation at all metallicities, with only slightly lower [Mg/Fe] values for the most metal-rich stars. This is among the strongest evidence that the chemical evolution of the bulge was dominated by Type-II supernovae. Bottom panel: the same plot as the top panel, but here all the field stars are marked by crosses and the large symbols (the same ones as in Figure 12.1) reflect the mean values from studies of bulge globular clusters. As with [O/Fe], the bulge globular clusters lie in similar regions, although two of the three observations of NGC 6528 lie on the lower envelope of the field-star distribution. Data for the cluster HP-1 lie well below both the bulge and the disk-halo relations.

is truly alpha-poor, like the relatively young clusters Ruprecht 106 and Pal 12 (Brown et al. 1997). The younger age allows for more Type-Ia enrichment to dilute the [a/Fe] ratios. The other solution is mentioned in Barbuy et al. (2006). They found that the spectroscopic [Fe/H] solution was about 0.5 dex higher than what one would assume from the shape of the red giant branch and the cluster's blue horizontal branch. Lowering the [Fe/H] value by itself would raise the various [element/Fe] ratios.

Finally, two of the Baade's Window giants examined by Fulbright et al. (2006b) have low [O/Fe] ratios yet high [Na/Fe] and [Al/Fe] values. This is reminiscent of what is seen in many globular clusters. In fact, these two stars have abundances that place them on the same region of the [Na/Fe] versus [O/Fe] and [Al/Fe] versus [O/Fe] diagrams as the stars of the similar-metallicity globular cluster M4 (Ivans et al. 1999). These two stars are the first non-cluster stars in the Milky Way found to exhibit this effect. The origin of the so-called Na-O and Al-O anti-correlations in globular cluster stars is unknown, but two possible theories for the existence of bulge field stars exhibiting this pattern are that these stars were stripped from metal-poor globular clusters like M4 or that the conditions that caused the abundance trends in globular-cluster stars were present in the bulge when these stars formed.

Was this article helpful?

0 0

Post a comment