Massive Star Clusters in Non- Interacting Galaxies Dynamical Mass Estimates and the (I)MF Søren S. Larsen ESO / ST-ECF, Garching Tom Richtler, Concepcion Jean P. Brodie, UCO / Lick Deidre A. Hunter, Lowell See also: Larsen & Richtler, A&A, in press (astro-ph/ ) Larsen, Brodie & Hunter, AJ, in press (astro-ph/ )
Motivation “Massive” ( M ) young star clusters have been found in several nearby galaxies Appear similar to old globular clusters in terms of sizes and masses, but will they really evolve into bona-fide old (~1 Hubble time) GCs? One key question is the (I)MF shape - if deficient in low-mass stars, clusters might disrupt prematurely Is IMF universal, or are there variations? Direct observations of low-mass stars generally unfeasible beyond Local Group => dynamical M/L ratios of YMCs represent one way to constrain IMF
M/L ratios and IMFs for YMCs Mengel et al. 2002, A&A 383, 137 SSP models from Leitherer et al (Starburst99; M ) M/L ratios from high-dispersion spectroscopy and HST imaging M vir ≈ 10 R hlr v x 2 / G R hlr = half-light radius v x = line-of-sight velocity dispersion Degeneracy: IMF slope / lower mass limit Top-heavy IMFs Bottom-heavy IMFs
Caveats Hard to find good targets (spatial resolution, bright enough for high-dispersion spectroscopy, isolated, uniform background) Youngest clusters relaxed? Mass segregation (primordial or dynamical) Macroturbulence in red supergiants ~ 10 km/s => dominates over velocity dispersions for masses < ~10 5 M Statistical fluctuations (~20 RSGs in 10 5 M cluster at 10 7 years)
Our Observations 7 YMCs in 4 nearby (3-6 Mpc) galaxies: NGC 4214 (irr), NGC 4449 (irr), NGC 5236 (sp) and NGC 6946 (sp) Masses > 10 5 M , ages 15 Myrs Myrs (from broad-band colours) HST imaging: cluster profiles well resolved (1 WFPC2 pixel ~ 1.5 pc at 3 Mpc) VLT/UVES and Keck/HIRES/NIRSPEC echelle spectroscopy => velocity dispersions through cross- correlation analysis (Tonry & Davis 1979)
Target galaxies NGC 6946 NGC 5236 NGC 4449 NGC 4214
Cluster sizes: EFF Model Fits Sizes determined with baolab/ishape software (Larsen 1999). Convolves TinyTim PSF with Elson, Fall & Freeman (EFF) models of the form P(r) ~ [1-(r/r c ) 2 ] - F555W Residuals
Velocity dispersions Keck/HIRES spectra Velocity dispersion: v x 2 = TC 2 - TT 2 where TC and TT are the dispersions of the cluster-template and template-template CCFs (Tonry & Davis 1979) Notes: 1) No individual strong lines are required for this technique to work. 2) Intrinsic broadening of lines (macroturbulence etc.) “cancels out”. Cross-Correlation Functions (CCFs)
Cluster Properties R hlr [pc] V x [km/s] Log(age) [yr] M vir [10 5 M ] 0 [M pc -3 ] N ± ± ± ± 1.0 (2.5±1.0) 10 3 N ± ± ± ± 2.4 (1.9±0.6) 10 5 N ± ± ± ± 0.9 (1.9±0.8) 10 3 N ± ± ± ± 1.6 (6.8±2.4) 10 3 N ± ± ± ± 0.8 (2.8±1.0) 10 3 N ± ± ± ± 0.7 (1.6±1.1) 10 4 N ± ± ± ±5 (2.3±0.8) 10 4
M/L ratios and the IMF UVES data (M83) HIRES data (Dwarfs) NIRSPEC (N6946) All 7 clusters consistent with Kroupa-type or Salpeter (M min =0.1 M ) IMF. Solid black curve: Bruzual+Charlot SSP models. Others: Basic SSP models based on Padua isochrones No top-heavy IMFs (Models for Z=0.008)
Summary Clusters with masses in the range 10 4 M M can form in disks of “normal” spirals and in dwarf galaxies, in addition to starbursts and mergers The clusters analyzed here have M/L ratios consistent with “normal” IMFs (usual disclaimers apply..) Such objects may provide direct insight into processes related to the formation of globular clusters in the early Universe
Velocity dispersions - MK type
M/L: Reddening effects