Young Massive Clusters in Non-Interacting Galaxies

Abstract

Young star clusters with masses well in excess of 100.000 Msun have been observed not only in merger galaxies and large-scale starbursts, but also in fairly normal, undisturbed spiral and irregular galaxies. Here we present virial mass estimates for a sample of 7 such clusters and show that the derived mass-to-light ratios are consistent with "normal" Kroupa-type stellar mass distributions.

Full text

Page 1

arXiv:astro-ph/0412512v1 20 Dec 2004
YOUNGMASSIVECLUSTERSINNON-INTERACTING
GALAXIES
Søren S. Larsen,1Jean P. Brodie2, Deidre A. Hunter3and Tom Richtler4
1ESO / ST-ECF,2UCO / Lick Obs.,3Lowell Obs.,4Universidad de Concepción
Abstract
Young star clusters with masses well in excess of 105M⊙ have been observed
not only in merger galaxies and large-scale starbursts, but also in fairly normal,
undisturbed spiral and irregular galaxies. Here we present virial mass estimates
for a sample of 7 such clusters and show that the derived mass-to-light ratios are
consistent with “normal” Kroupa-type stellar mass distributions.
Keywords:
galaxies: star clusters — galaxies: spiral — galaxies:irregular.
1. Introduction
Young star clusters with masses in the range 104M⊙–106M⊙(“young mas-
sive clusters”; YMCs) are frequently observed in a wide variety of external
galaxies (see review in Larsen 2004). Determining the mass-to-light ratios of
YMCs is a matter of considerable interest, because this may lead to constraints
on the stellar mass function (MF). The MF, in turn, is of interest not only for
the general question whether or not there is a universal MF, but also for the
long-term survival of YMCs. Here we discuss new results for a sample of
7 YMCs, for which we have attempted to constrain the MF using dynamical
mass-to-light ratios.
2. Data
Three of our target clusters are located in spiral galaxies (one in NGC 6946
and two in NGC 5236) and were selected from the sample of Larsen & Richtler
(1999). The other four are in dwarf irregulars (2 in each of NGC 4214 and
NGC4449) and were selected from Billett et al. (2002) and Gelatt et al. (2001).
Archive HST/WFPC2 imaging is available for all our targets, and the clus-
ters are all free of crowding and appear superimposed on a reasonably smooth
background. The host galaxies have estimated distances in the range 2–6 Mpc,
making the clusters appear well resolved on HST images although a correction
for instrumental resolution is still necessary when measuring the sizes. Based

Page 2

2
Table 1.
line-of-sight velocity dispersion and ρ0is the estimated central density.
Data for the YMCs. Rhlris the half-light radius, Mvir is the virial mass, vx is the
ID
Rhlr
[pc]
vx
Log(age)
[yr]
Mvir
ρ0
[km/s][105M⊙][M⊙pc−3]
N4214-10
N4214-13
N4449-27
N4449-47
N5236-502
N5236-805
N6946-1447
4.33±0.14
3.01±0.26
3.72±0.32
5.24±0.76
7.6±1.1
2.8±0.4
10.2±1.6
5.1±1.0
14.8±1.0
5.0±1.0
6.2±1.0
5.5±1.0
8.1±1.0
8.8±1.0
8.3±0.1
8.3±0.1
8.9±0.3
8.5±0.1
8.0±0.1
7.1±0.2
7.05±0.1
2.6±1.0
14.8±2.4
2.1±0.9
4.6±1.6
5.2±0.8
4.2±0.7
17.6±5
(2.5 ± 1.0) × 103
(1.9 ± 0.6) × 105
(1.9 ± 0.8) × 103
(6.8 ± 2.4) × 103
(2.8 ± 1.0) × 103
(1.6 ± 1.1) × 104
(2.3 ± 0.8) × 104
on photometry, ages were estimated to be in the range 11 Myr – 800 Myr
and all clusters have photometric mass estimates greater than 105M⊙. The
expected velocity dispersions are of the order 5–10 km/s.
The two clusters in NGC 5236 were observed with the UVES echelle spec-
trograph on the ESO Very Large Telescope, while the remaining clusters were
observed with HIRES and NIRSPEC on the Keck I telescope. The spectral
resolution was between λ/∆λ = 25000 and λ/∆λ = 60000 and the S/N
typically about 20–30 or better per resolution element. Velocity dispersions
were measured using the cross-correlation technique of Tonry & Davis (1979).
In brief, the cluster spectra were first cross-correlated with the spectrum of a
suitable (red supergiant) template star. The template star spectrum was then
cross-correlated with the spectrum of another template star. The velocity dis-
persion of the cluster spectrum was then essentially given as the quadrature
difference between the gaussian dispersions of the peaks of the two cross-
correlation functions. The cluster sizes were derived from the HST/WFPC2
images by convolving EFF models (Elson, Fall & Freeman 1987) with the
WFPC2 point-spread function and solving for the best fit to the observed im-
ages (Larsen 1999). Cluster ages and reddenings were estimated by comparing
multi-colour photometry with Bruzual & Charlot (2003) simple stellar popula-
tion (SSP) models. For further details concerning the data reduction and anal-
ysis we refer to Larsen et al. (2005; for NGC 4214, NGC 4449 and NGC 6946)
and Larsen & Richtler (2005; for NGC 5236).
3. Virial mass-to-light ratios and the MF
Our results for the 7 YMCs are summarised in Table 1. As expected from
the photometry, all clusters have virial masses in excess of 105M⊙, and two
have masses > 106M⊙. Interestingly, there is little if any correlation between
cluster mass and half-light radius. Central densities, estimated from the EFF

Page 3

Young Massive Clusters in Non-interacting Galaxies
3
Figure 1.
Comparison of observed mass-to-light ratios with SSP models
fits, are listed in the last column and range from ∼ 1000M⊙pc−3to greater
than 105M⊙pc−3.
In Fig. 1 we compare the observed mass-to-light ratios with predictions
by SSP models for various MFs. The solid line shows the MV magnitude
per solar mass for solar metallicity and a Salpeter (1955) MF with a lower
mass limit of 0.1M⊙from Bruzual & Charlot (2003). The other curves show
our calculations for Salpeter MFs with lower mass limits of 0.01M⊙, 0.1M⊙
and 1.0M⊙(long-dashed, dotted-dashed and triple dotted-dashed lines) and a
Kroupa (2002) MF(short-dashed line), obtained bypopulating solar-metallicity
isochrones from Girardi (2000) according to the various MFs. It is not strictly
correct to put all YMCs on the same plot, since the clusters in NGC 4214 and
NGC 4449 may have metallicities of only 1/4-1/3 solar (Larsen et al. 2005).
However, the V -band M/L ratios are predicted to change by less than 0.2 mag
for models of 1/5 solar metallicity (shifting the curves in Fig. 1 upwards).
The comparison in Fig. 1 suggests that our data are mostly consistent with a
Kroupa-type MF or a Salpeter MF extending down to 0.1 M⊙. At the present
level of accuracy, we cannot distinguish between these possibilities. It should
also be noted that there is a degeneracy between the MF slope and the lower
mass limit. However, there is no suggestion that any of the YMCs studied
here have significantly top-heavy MFs. It should be kept in mind that the virial
mass estimates are subject to a number of uncertainties which are not easily
quantified. In order to assess the role of macroturbulence in the template stars

Page 4

4
we used several stars in the cross-correlation analysis and found no strong de-
pendence on the choice of template star. However, it is difficult to find local
red supergiants which are as luminous as those expected to be present in the
youngest clusters, and if the macroturbulence varies significantly with lumi-
nosity then this may lead to systematic errors. Mass segregation (primordial
or dynamical) can also lead to erroneous results if the virial theorem is applied
blindly. Recent calculations by A. Lançon (this meeting) suggest that mass
segregation will typically lead to the cluster masses being underestimated, i.e.
the data points would shift downwards in Fig. 1 and the case for top-heavy
MFs would then seem even weaker.
4. Summary
Usingacombination ofground-based high-dispersion spectroscopy and HST
imaging, we have derived virial mass-to-light ratios for a sample of 7 YMCs
with masses in the range 105M⊙–106M⊙in nearby spiral and irregular galax-
ies. By comparing the mass-to-light ratios with predictions by SSP models, we
conclude that our data are consistent with “normal” (e.g. Kroupa 2002-type)
stellar mass functions, suggesting that the clusters may eventually evolve into
objects which will be very similar to the globular clusters commonly observed
around galaxies.
Acknowledgments
JPB and SSL acknowledge support by National Science Foundation grant
AST-0206139 and HST archival grant AR-09523-01-A. DAH acknowledges
National Science Foundation grant AST-0204922. TR gratefully acknowl-
edges support from the Chilean Centerfor Astrophysics FONDAPNo.15010003.
References
Billett, O. H., Hunter, D. A., & Elmegreen, B. G., 2002, AJ, 123, 1454
Bruzual, G., & Charlot, S., 2003, MNRAS, 344, 1000
Elson, R. A. W., Fall, S. M., & Freeman, K., 1987, 323, 54
Gelatt, A. E., Hunter, D. A., & Gallagher, J. S., 2001, PASP, 113, 142
Girardi, L., Bressan, A., Bertelli, G., & Chiosi, C., 2000, A&A Suppl., 141, 371
Kroupa, P., 2002, Science, 295, 82
Larsen, S. S., 1999, A&A Suppl., 139, 393
Larsen, S. S., 2004, in: “Planets to Cosmology: Essential Science in Hubble’s Final Years”, ed.
M. Livio, STScI, May 2004”
Larsen, S. S., Brodie, J. P., & Hunter, D. A., 2005, AJ, in press
Larsen, S. S., Richtler, T., 1999, A&A, 345, 59
Larsen, S. S., Richtler, T., 2005, A&A, in press
Salpeter, E. E., 1955, ApJ, 121, 161
Tonry, J., & Davis, M., 1979, AJ, 84, 1511