Mass Loss from Red Giants in the Magellanic Clouds

Ecology We are made of 'stardust'. Without the creation of heavy elements inside the stellar furnaces, there would be no oxygen to breath or water to drink, no carbon upon which all life is based. But neither would there be, if it wasn't for dying stars to expel these products into interstellar space, so that new generations of stars and planets may form from this enriched material.

Stellar death is of paramount importance for the evolution of the Universe: the chemical enrichment of the interstellar medium in a galaxy modifies the thermal and density structure of the interstellar medium, with consequences for subsequent epochs of star formation. The formation and processing of dust depends on the chemical composition and abundance of its condensates, as does the formation of molecules such as carbon-monoxide (CO) and water. These are both important coolants, helping molecular clouds collapse and form stars, as well as important constituents of Earth-like planetary bodies. The presence of increasing amounts of interstellar dust and molecular gas, and the formation of stars with increasingly enhanced metal content, both have a tremendous impact on the appearance and evolution of planets, stars, galaxies and the Universe at large.

However, stellar death is also one of the least understood phases of stellar evolution. With birth-masses between a little less massive than the Sun (about 0.8 Msun) and several times as massive (about 8 Msun), intermediate-mass stars after having exhausted the hydrogen fuel in their cores will swell up to become cool, luminous red giants. Evolving, first along the Red Giant Branch (RGB) and later along the Asymptotic Giant Branch (AGB), their convective mantles become increasingly unstable and loosely bound. A combination of stellar pulsation (radial oscillations) and radiation pressure on molecules and/or dust grains that form in the cool dense atmospheres (stellar wind) are held responsible for driving an outflow through which these stars loose between at least half and up to 80% of their birth-mass. It is this material that enrichens the interstellar medium. The last ejected material is sometimes lit up by the barren, hot core of the dying star, shining as a brilliant Planetary Nebula, whilst the core itself gradually fades into oblivion. Because intermediate-mass stars have lifetimes ranging from about 30 Myr up to the age of the Universe (about 15 Gyr), they not only have nearly always played an important role in the history of the Universe, but at the same time they can also be used to trace this very history.

There are serious gaps in our understanding of these phenomena. To begin with, there is no adequate theory to describe stellar convection. It is thus difficult to predict the exact conditions for nuclear burning, the dredge-up of nuclear-processed material to the stellar surface, and the excitation of the radial oscillations and the effect this has on the stellar structure. The rarefied atmospheres of red giants are difficult to model due to complexities such as optical depth effects and asymmetric, non-planar geometries. The chemical path through which a variety of molecular species form and grow, and the process of their sublimation into dust grains, are largely unknown. As a consequence it has not been possible yet to predict, ab initio, the evolution of the mass-loss rate and its detailed chemical composition, as a function of stellar parameters (birth-mass, birth-metallicity, binarity, rotation, etc), nor do we know the details about when and how this material is mixed with the surrounding interstellar medium.

The Large Magellanic Cloud (LMC) provides us with a unique opportunity to study this problem. It contains millions of evolved stars covering the entire mass range from about a solar mass up to tens of solar masses. As these stars are all at approximately the same reasonably well known distance from us, their relative luminosities can be determined accurately. Estimates for the mass-loss rates also scale with the inferred distances and hence, unlike for most stars in our own Milky Way galaxy, mass-loss rates of LMC giants can be determined accurately as well. The LMC is near enough that the infrared (IR) emission from the circumstellar dust around cool evolved giants can be detected. From this, dust mass-loss rates can be estimated which may be scaled to estimate the total (gas+dust) mass-loss rates by adopting a certain dust-to-gas ratio. The dust-to-gas ratio is almost invariably obtained by scaling dust-to-gas ratios determined for galactic objects, where molecular emission, e.g. CO, can be detected to yield the gas mass-loss rate (after scaling by the, again to a certain extent uncertain, CO abundance relative to hydrogen). The Spitzer Space Telescope (SST) can detect photospheres of RGB stars in the LMC, and it will thus be able to measure excess IR emission above these photospheres from very small amounts of circumstellar dust.

The Small Magellanic Cloud (SMC) provides similar advantages as the LMC. It is (only) slightly more distant, but significantly smaller, which limits the number of stars that can be studied. This is especially problematic for rare objects that represent short-lived phases in stellar evolution. However, the metallicity of the SMC is a few times lower than that of the LMC, and a comparative study of stellar populations in both Clouds is highly interesting as it may reveal dependences on metallicity. Due to the lower metallicity the dust content of the mass loss is also diminished. although this does not a priori mean that the mass-loss rates are lower, it does render the IR excess emission fainter and the obscuration at shorter wavelengths less severe which in turn makes it harder to recognise objects with moderate mass-loss rates as well as to estimate these rates. The Spitzer Space Telescope, however, can detect very low mass-loss rates even at the small dust-to-gas ratios encountered in the SMC.

For the first time we will be able to map the mass-loss rate across the entire RGB and AGB. Photometry with the Spitzer Space Telescope at mid to far-IR wavelengths (with the MIPS instrument) is complemented by low-resolution spectroscopy (with the IRS instrument) to determine the detailed shape of the spectral energy distribution (SED) at wavelengths between 5 and 38 micrometre. The spectra yield unambiguous chemical types (oxygen- or carbon-dominated) and degree of crystallinity of the dust grains, which is important when measuring mass-loss rates from modeling of the SED because the results depend on the optical properties of the dust species. Besides obtaining reliable mass-loss rates, the spectra also help us gauge the type of matter with which these stars enrich the interstellar medium.

Stars in star clusters in the Magellanic Clouds are included, as for cluster members individual metallicities and ages (hence progenitor masses for the giants) can be determined. We are already in the process of obtaining ancilliary (groundbased) observations at optical and near-IR wavelengths in order to determine the stellar photospheric temperatures and chemical composition that are needed to derive mass-loss rates from the shape of the IR spectral energy distribution. It will thus be possible to reconstruct the relation between the mass-loss rate and the birth-mass, birth-metallicity and time. This will then form the empirical basis against which to test our understanding of stellar evolution, and serve as input for models of galacto-chemical evolution.