A1260686720xcRMZ

1
Star clusters in the Antennae galaxies
Sabine Mengel, ESO, Garching
Ages
Extinction
Narrow band diagnostics (equivalent widths of
Brg, Ha and CO, images from SOFI, HST and ISAAC,
respectively) were used together with a
simultaneous fit of individual cluster broadband
SEDs for age and extinction (in comparison
to Starburst99 (Leitherer et al.) models). Our
available broadband magnitudes span the
wavelength range from U to Ks, which turns out to
be crucial if reliable cluster properties are to
be determined. Only the Ks-detected clusters
(1072) were taken into account. Knowledge of age
and extinction together with absolute magnitudes
allowed us to determine photometric cluster
masses (assuming a Salpeter IMF).
Narrow band line ratios (Brg and Ha) were used
together with broadband fits to determine
extinction values towards single star clusters.
Note the excellent agreement between regions of
high extinction and high CO density (in the CO
contours taken from Wilson et al. 01). From the
distribution of high extinction Clusters and the
evolution of extinction with age we deduce that a
typical cluster has blown free of its natal dust
cocoon by the age of 8-9 Myrs.
VLT-ISAAC Ks-band image (red) with HST (Whitmore
et al. 99) UB (blue) and VI (green) images.
Results from the imaging data will be included in
Mengel et al. (2004, in preparation)
Star Formation History
There has been a debate about the nature of the
brightest 15mm peak in this merger, which
contributes 15 of the flux at that wavelength
range. We converted our photometric mass
estimates into expected bolometric luminosities
according to Starburst99 models, and converted
this into an image shown here as contours. We
expect the 15mm cluster (arrow at the bottom)
to contribute 3.6 of the flux. The colour image
is the ISOCAM 15mm image from Mirabel et al (98).
We believe one of the suggestions by Wilson et
al. (01) is quite likely, that the region of the
clusters is swept by small dust particles from
moving molecular clouds.
These two figures show the age distribution of
the Ks-band detected star clusters. Even though
the Ks-band completion limit for a cluster
with more than 105 solar masses is reached only
at around 30 Myrs, a significant drop in number
counts takes place before then. This means that
either the recent interaction-induced starburst
has been much stronger than any other over the
last 30 (up to 200) Myrs, or that a large
fraction of the clusters gets destroyed on
timescales of 10 Myrs. This is in general
agreement with what Fall et al. find from HST
data only.
IMF
Stellar velocity dispersions measured on some of
the clusters (using ISAAC, and UVES on the VLT),
together with size estimates (from HST) allowed
us to determine dynamical masses. Comparing them
to the photometric cluster masses and to
evolutionary synthesis models potentially allows
conclusions about the IMF, which seems to vary in
slope or low mass cutoff. Another possibility is
that these variations reflect the disruption
experienced by some clusters, since our measured
concentrations and IMF slopes would suggest that
at least these clusters are stable for several
Gyrs (they end up in the green box inserted into
a figure by Takahashi Portegies Zwart (2000),
where the numbers inside the box are log(survival
time in Grys)). Yet another possibility is that
mass segregation plays a role, compromizing both
the size and the velocity dispersion estimate
(see Mengel et al., 2002).