Looking at the chemical composition of stars that host planets,  astronomers have found that while dwarf stars often show iron  enrichment on their surface, giant stars do not. The astronomers  think that the planetary debris falling onto the outer layer of the  star produces a detectable effect in a dwarf star, but this pollution  is diluted by the giant star and mixed into its interior.

"It is a little bit like a Tiramisu or a Capuccino," says Luca  Pasquini from ESO, lead-author of the paper reporting the results.  "There is cocoa powder only on the top!'

Just a few years after the discovery of the first exoplanet it became  evident that planets are preferentially found around stars that are  enriched in iron. Planet-hosting stars are on average almost twice as  rich in metals than their counterparts with no planetary system.

The immediate question is whether this richness in metals enhances  planet formation, or whether it is caused by the presence of planets.  The classic chicken and egg problem. In the first case, the stars  would be metal-rich down to their centre. In the second case, debris  from the planetary system would have polluted the star and only the  external layers would be affected by this pollution.

When observing stars and taking spectra, astronomers indeed only see  the outer layers and can't make sure the whole star has the same  composition. When planetary debris fall onto a star, the material  will stay in the outer parts, polluting it and leaving traces in the  spectra taken.

A team of astronomers has decided to tackle this question by looking  at a different kind of stars: red giants. These are stars that, as  will the Sun in several billion years, have exhausted the hydrogen in  their core. As a result, they have puffed up, becoming much larger  and cooler.

Looking at the distribution of metals in fourteen planet-hosting  giants, the astronomers found that their distribution was rather  different from normal planet-hosting stars.

"We find that evolved stars are not enriched in metals, even when  hosting planets," says Pasquini. "Thus, the anomalies found in planet- hosting stars seem to disappear when they get older and puff up!"

Looking at the various options, the astronomers conclude that the  most likely explanation lies in the difference in the structure  between red giants and solar-like stars: the size of the convective  zone, the region where all the gas is completely mixed. In the Sun,  this convective zone comprises only 2% of the star's mass. But in red  giants, the convective zone is huge, encompassing 35 times more mass.  The polluting material would thus be 35 times more diluted in a red  giant than in a solar-like star.

"Although the interpretation of the data is not straightforward, the  simplest explanation is that solar-like stars appear metal-rich  because of the pollution of their atmospheres," says co-author Artie  Hatzes, Director of the Thüringer Landessternwarte Tautenburg  (Germany) where some of the data were obtained.

When the star was still surrounded by a proto-planetary disc,  material enriched in more heavy elements would fall onto the star,  thereby polluting its surface. The metal excess produced by this  pollution, while visible in the thin atmospheres of solar-like stars,  is completely diluted in the extended, massive atmospheres of the  giants.

More Information

"Evolved stars hint to an external origin of enhanced metallicity in  planet-hosting stars", by L. Pasquini et al. To appear in Astronomy  and Astrophysics. The paper is available on astro-ph at http:// arxiv.org/abs/0707.0788

The team is composed of L. Pasquini and M.P. Doellinger (ESO), A.  Weiss (Max-Planck-Institut fuer Astrophysik, Garching, Germany), L.  Girardi (INAF-Osservatorio Astronomico di Padova, Italy), C. Chavero  (Instituto de Astrofisica de Canarias, Tenerife, Spain, and  Observatorio Nacional/MCT, Rio de Janeiro, Brasil), A. P. Hatzes  (Thüringer Landessternwarte Tautenburg, Germany), L. da Silva  (Observatorio Nacional/MCT, Rio de Janeiro, Brasil), and J. Setiawan  (Max Planck Institute fuer Astronomie, Heidelberg, Germany). The data have been partially collected at ESO, and partially at the 2- m telescope of the Thueringer Landessternwarte Tautenburg (TLS).

A high-resolution image is available at http://www.eso.org/public/outreach/press-rel/pr-2007/pr-29-07.html

Science Contacts:

Luca Pasquini
ESO Garching, Germany
Phone: +49 89 3200 6792
Email: lpasquin@eso.org

Artie Hatzes
Thueringer Landessternwarte Tautenburg, Germany
Phone: +49-36427-863-51
Email: artie@tls-tautenburg.de