Chemical composition of old, metal-poor stars poses new questions

New measurements of the element abundances of strontium and barium in the Galactic disk and halo have led to intriguing results. Scientists at the Max Planck Institute for Astrophysics determined the element abundances in a large sample of old metal-poor stars and found that they cannot be explained by any of conventional nucleosynthesis scenarios predicted by models of stellar evolution. The data indicate that some - still unknown - process has to be mainly responsible for the production of these elements in the early Galaxy.

Fig. 1: The Milky Way contains about 300 billion stars. The galactic disk, which is partly hidden behind gas and dust clouds, is surrounded by a spherical halo of old stars and stellar clusters.
Image: Wei-Hao Wang

Fig. 2: The new barium (Ba) and strontium (Sr) abundances for the halo and disk stars as a function of metallicity. The measurements show the difference between the ratio of each element to iron in a star and that of the Sun (i.e., a value of zero means that the star has the same abundance ratio as the Sun). Disc stars with low metallicity (-1.5 < [Fe/H] -0.5) show solar barium abundances, whereas for halo stars the barium abundance is consistent with that predicted from the r-process in supernova explosions. The strontium abundances, however, do not agree with predictions from standard nucleosynthesis models.

Fig. 3: Relative abundance of strontium and barium, where disk and halo stars are indicated by red and black circles. Conventional nucleosynthesis scenarios would give a slightly negative value for [Sr/Ba], whereas the observed value is clearly positive, especially for the stars with very low metallicity, [Fe/H] < -1.5.

In theory, the nucleosynthesis of elements heavier than iron is thought to be well understood. Most of them are formed in the so-called slow (s) or rapid (r) neutron capture processes on lighter seed nuclei. The s-process elements are mainly produced during the late stages of evolution of low-mass stars; a small fraction of these elements originates from massive stars, which explode as core-collapse supernovae.

However, predictions of the stellar models are not fully supported by observations. The observed s-process abundances in the Sun and metal-poor stars, tracing the composition of the interstellar matter in different stages of Galactic evolution, need to be explained by alternative nucleosynthesis scenarios.

Strontium (Sr) and barium (Ba) with magic neutron numbers of 50 and 82, respectively, are the most abundant s-process elements. This is a direct consequence of their nuclear properties, namely very small neutron capture cross-sections. In addition, their strongest absorption lines are in the optical part of a spectrum, accessible with ground-based telescopes, and can be measured even in most metal-poor and distant Galactic halo stars.

The most difficult part, in fact, is to compute the abundance from the observed data points: line profiles are very sensitive to the treatment of radiative transfer in stellar atmosphere models. So far, most observational studies in the literature relied on highly-simplified models, neglecting the important influence of non-local thermodynamic equilibrium (non-LTE) effects on element abundances.

Scientists at MPA have now performed, for the first time, a consistent non-LTE spectroscopic analysis of a large sample of metal-poor stars, which belong to various Galactic populations (halo, thick and thin disks). Non-LTE effects were taken into account both in the determination of the basic stellar parameters and the element abundances (see Fig. 2), using new quantum-mechanical atomic data.

The disk stars show barium abundances in agreement with scaled solar abundances. This can be fully explained by models of nucleosynthesis in low-mass stars. The halo stars exhibit a strong deficiency of barium compared to iron. This is again consistent with stellar models, which predict that only a small range of massive stars contribute to producing barium via the r-process, whereas iron is made in supernovae explosions of stars with different initial masses.

The strontium abundances, however, increase with decreasing metallicity of a star eventually flattening off at the lowest metallicities. This conflicts with predictions from stellar models. What process could have produced strontium in the early Galaxy? Furthermore, what process over-produced strontium compared to barium at the time when the oldest and very metal-poor stars were born? Summing up all known contributions from both r-process and s-process gives a strontium-to-barium ratio, which is much lower than the one observed for stars with low metallicity (see Fig. 3).

A few non-standard, exotic nucleosynthesis scenarios, such as the so-called Light Element Primary Process or low-mass electron-capture supernovae, might have the potential to explain these observations.


Gregory Ruchti, Maria Bergemann


References

Travaglio, C., Gallino, R., Arnone, E., et al., "Galactic Evolution of Sr, Y, and Zr: A Multiplicity of Nucleosynthetic Processes", 2004, linkPfeilExtern.gifApJ, 601, 864

Wanajo, S., Janka, H.-T., Mueller, B., "Electron-capture Supernovae as The Origin of Elements Beyond Iron", 2011, linkPfeilExtern.gifApJL, 726, L15

Ruchti, G. and Bergemann, M., "New NLTE Results for Neutron-Capture Elements in Metal-Poor Milky Way Field Stars", 2012, MNRAS, in prep.