Post-AGB evolution (ROB and KU Leuven)

       The final stages of evolution of low- and intermediate mass stars start when they have reached the -asymptotic giant branch (AGB), and continues through the so-called post-AGB phase and the planetary-nebula stage. This evolution is dominated by mass-loss processes which are still poorly understood. Moreover, the interplay among various physical processes between the star and its circumstellar nebula complicates the interpretation of the observations, and conversely increases their diagnostic value. Finally, the mass-loss and dredge-up processes reveal products of internal nucleosynthesis, hence providing important clues about internal astrophysical processes.

       The ROB group continued to study the circumstellar shells of post-AGB stars. To model them Ueta's radiative transfer code for an axisymmetric system, 2-Dust, was used in order to understand the energetics of thermal emission within the shells and the properties of dust grains and the central star. The 2-Dust code has been continuously upgraded so that the execution time is reduced about 50% and that radiative transfer is done following dust temperature for each grain size and for each dust species. The latter modification is required to incorporate non-equilibrium heating of very small dust grains. This work has been done in collaboration with R. Szczerba (Torun, Poland).

       The 2-Dust code was used to compute a radiative transfer model of the post-AGB star, HD 56126. This model was contributed to a study of the object's molecular and dust shells by Meixner (STScI, USA). 2-Dust was also used by Mieke De Vlieger, graduate student at the K.U. Leuven, to construct a model for the post-AGB star IRAS 16594-4656. The modelling effort resulted in an undergraduate thesis which is now reformulated for a refereed publication. We would also like to gain insight in the winds structuring this object and the shocks they produce. To this aim we are carefully analyzing the lines in the high-resolution near infrared spectra obtained with Gemini-South, in collaborations with P. van Hoof, who will join the ROB team early in 2005. To better understand the spatial distribution of molecular hydrogen, images in molecular hydrogen were retrieved from the HST archive and analyzed.

       Detecting structure developing shells would provide crucial clues to our understanding how the circumstellar shells are shaped. To investigate the critical phases of structure formation in the circumstellar shells of evolved stars, we analyzed imaging polarimetry data obtained with NICMOS polarizers on board the Hubble Space Telescope. It successfully revealed the 3-dimensional structure of the circumstellar shells of post-AGB stars (especially optically thin shells), confirming their hollow and spheroidal nature with equatorial enhancement. The results have been presented as a contributed talk at the "Astronomical Polarimetry - Current Status and Future Directions" conference in March 2004 and a publication has been accepted by Astronomical Journal.

       Based on the success in the previous research projects new observing proposals using NACO/VLT and TIMMI2/ESO3.6 have been submitted and accepted. A total of 4 nights of observing time was obtained. Near-IR coronographic imaging polarimetry with NACO has been performed in December 2004 while mid-IR imaging with TIMMI2 will be performed in January 2005. For the NACO observations we selected objects which are likely developing the axisymmetric structure for the first time in their mass loss history. Coronographic imaging polarimetry with NACO allows (1) high spatial resolution probing of circumstellar environment with significantly reduced PSF effects and (2) separation of dust-scattered light off the shell (polarized component) from the direct star light (unpolarized component). The observations with TIMMI2 are obtained in a collaboration ROB-KULeuven. The objective is to understand how circumstellar shells of evolved stars assume the axisymmetric structure during the late AGB phase by directly probing the dust distribution via thermal mid-IR data. Well-resolved mid-IR images show structures that provide direct clues to trace the mass loss history during the late AGB Phase. Even marginally resolved images yield an upper limit for the dust-shell inner radius, which is critical for constraining dust radiative transfer models. Critical constraints such as dust temperature and optical depth maps will also be obtained.

       A chemical analysis of the metal-poor post-AGB star V453 Ophiuchi shows that this object is enriched in s-process elements, but not so in carbon. This circumstance is most unusual, since it is generally thought (and confirmed by observation) that s-process enhancement and carbon enrichment go hand in hand. Two solutions are proposed: (1) extrinsic s-process enrichment by a more massive companion, and (2) extreme parental cloud enrichment by a previous generation. Models involving standard AGB evolution fail to reproduce the observations within the acknowledged uncertainties of the parameters and theoretical recipes involved.

       The main focus of the post-AGB research of the KULeuven group has shifted more and more to post-AGB binaries. The conjecture that post-AGB stars with near-IR excesses are binaries, has indeed proven true, and the orbital motion could be determined from long-term campaigns with the Coralie spectrograph at the Swiss Telescope at La Silla. Typical orbital periods are of the order of a few years, which implies that binarity has had a substantial influence on evolution, hence its interest. Two peculiar aspects are: (1) the non-zero and sometimes large eccentricities of several systems, despite the short orbits, which according to first-order evolutionary scenarios should have implied circularisation, and (2) the peculiar chemical abundances of many primary components, which are very deficient in refractory elements. The clue to both probably is the presence of a circumsystem disk, which enhances the eccentricity through tidal interaction, and is responsible for element separation through dust formation processes and subsequent reaccretion of 'cleaned' gas.

       A next step in the study of these binary post-AGB stars is made possible by the substantial Belgian involvement in VLTI. Our first results with VLTI, equipped with MIDI, show that it the circumsystem disks can be resolved. Moreover, the visibilities depend on wavelength, in the sense that it becomes possible to spatially locate the different dust features, hence opening new constraints on the study of their formation histories.

Halo stars (ULg)

       The ULg Group has proceeded with the chemical analysis of metal-poor stars in order to test in more detail the EASE scenario which had been proposed by our team a few years ago on the basis of the abundance correlations for heavy elements in 21 mildly metal-poor stars. They have examined the behaviour of lighter elements, from carbon to calcium. These lighter elements are important tracers of the nucleosynthetic processes at work in the stellar interiors.

       They have found that, in stars that are thought to have accreted matter expelled by AGB (Asymptotic Giant Branch) stars, i.e. those exhibiting excesses of neutron-capture (s-process) elements, the carbon abundance is anticorrelated with the s-process abundances; they interpret this as the effect of the so-called 'hot bottom burning', which would bring to the surface of the AGB stars the products of hydrogen-burning by the CNO-cycle.

       On the other hand, the Liège group found that, although all the -elements abundances are correlated with each other, the slopes of the correlations differ from unity, in the sense that the lighter elements show a larger range of variation than the heavier ones. This does not fit well with the original EASE scenario, which postulates that stars showing different -elements abundances (relative to iron) have formed and escaped the proto-globular cloud at different stages during the explosions of SNeII. The stars having the lowest /Fe ratios would have formed earlier, when only the most massive of the SNeII have exploded, while stars with higher /Fe would have formed at the end of the SNeII phase. As the most massive stars are expected to overproduce the lighter -elements, the observed trends cannot be explained by this original EASE scenario.

       They propose a modification of this scenario, in which all observed field halo stars have formed and escaped proto-globular clouds after the end of the phase of SNeII explosions. However, we postulate that the proto-globular clouds exhibit a large mass range, and that the less massive of these clouds are unable (or less able) to give birth to very massive stars. Thus, they would only be enriched by the nucleosynthetic products of moderate-mass stars (e.g., M < 20 M) and the second generation stars that we observe would show lower abundances of the lighter -elements, as well as lower values of /Fe. We point out that the inability of low-mass clouds to give birth to massive stars does not require a variation of the initial mass function for stars with the mass of the cloud. Indeed, simple statistical arguments show that many low mass clouds will lack the most massive (and rarest stars). This new scenario also neatly explains why none of these stars with lower /Fe show signs of accretion: indeed, these low-mass clouds are weakly bound, unable to retain the matter expelled by the AGB stars and promptly disrupted by the tidal field of the galactic disk (De Cauwer et al., in press).

       Moreover, they have completed the reduction and analysis of VLT spectra for an additional sample of metal-poor stars, selected on a kinematic basis. While the stars in the original sample mostly belong to the thick disk and inner halo components of our Galaxy, the new sample consists of stars with extreme kinematics, belonging to the outer halo. According to the generally accepted view, these stars might originate from small satellite galaxies captured and disrupted by the gravitational field of the Milky Way. They shall compare them with the stars of the first sample, which are believed to have formed inside the Milky Way. Such a comparison is expected to shed more light on the environments in which these stars formed.

Web pages:Sandrine Sohy
Last modification: April 27, 2005