Dr. Andrea Stolte


Image: Hubble Space Telescope/WFC3,
Credits: A. M. Ghez, M. R. Morris, A. Stolte

The present-day stellar mass function in Milky Way starburst clusters

One of the major questions investigated in the starburst cluster project was:

  • Is the initial stellar mass function universal?
  • Do extreme environments, such as starburst clusters, produce a larger fraction of high-mass stars?
  • If we only consider the central regions of starburst clusters, we clearly observe an overdensity of high-mass stars: the present-day mass function is flat. However, when accounting for the dynamical evolution, the results on the Arches cluster suggest that no over-formation of exceptionally massive stars is needed to explain the mass distribution in these clusters. In other words, a normal initial mass function is sufficient to explain the observed mass distribution and its spatial variation.
    The stellar mass function was investigated - with and without the additional aid of proper motion membership - in all PhD theses related to the cluster survey. The major results are illustrated in the figures extracted from the respective publications below.


The slope of the present-day mass function of the Arches cluster, located near the center of our Galaxy, was investigated in Habibi (2014) and Habibi et al. (2013). In this figure, the observed mass function slopes at each radius (red and blue points) are compared to dynamical simulations of the cluster during its short 2.5 million year lifetime. All black models start with a normal, Salpeter-like slope. Note how the cluster has evolved such that the mass function appears flat in the inner parts at small radii, where mass segregation has caused an overdensity of high-mass stars. In the outskirts of the cluster, the present-day MF appears much steeper than the initial MF at birth, because low-mass stars have migrated to larger radii. Although we cannot exclude a deviation in the initial mass function at the time the cluster stars emerged from their native cloud from these observations alone, the models suggest that the dynamical evolution of the cluster is sufficient to explain the observed radial distribution of cluster stars today.


The radial trend in the slope of the stellar mass function of the Quintuplet cluster was investigated in Hußmann (2014) and Hußmann et al. (2012). The figure shows how the present-day mass function steepens slightly, but visibly, with increasing cluster center distance. Still, the outer mass function is by far less steep then the outer Arches MF, and the total mass function from all stars included with membership probabilities in the survey remains still slightly shallower than the standard Salpeter IMF. Whether or not the Quintuplet was born with an overdensity of high-mass stars thus remains even more a question than in the case of the Arches cluster. A set of dynamical models covering the full cluster area is needed to decide whether a normal IMF would lead to the stellar radial mass distribution as we observe it today.

Westerlund 1

Radial variation in the elongated cluster Westerlund 1 illustrated by colour gradients. A detailed analysis of various slopes of the mass function and the radial distribution of stars can be found in Gennaro et al. (2011). Although no full dynamical model of the cluster is available at the present time, mass segregation is a likely scenario to explain the increasing number of low-mass stars towards the cluster outskirts (blue colours in the cluster map). High-mass stars are concentrated in the center, as discussed in detail in the paper.

Conclusion: Even though we observe local variations in the slope of the present-day mass function in Milky Way starburst clusters, the evidence that the clusters were born with deviating initial mass functions is weak. The variations can be quite severe, as seen particularly well in the Arches cluster, yet dynamical evolution and the segregation of masses are sufficient to explain the observed radial trend in the MF slope. During mass segregation, the high-mass stars sink to the center, while the low-mass stars absorb energy in swing-by events and are flung outwards, to orbits with larger radii. In the strong tidal field of the Galactic center, stars can be tidally dispersed from the cluster and end up in the cluster's tidal tail. These stars are lost from the cluster area and are not counted in the present-day mass function. In fact, they cannot even be distinguished from stars in the dense Galactic center stellar field.

Spiral arm vs. Galactic center starbursts

  • Are starburst clusters in the Galactic center different from spiral arm clusters?
  • Can we see any evidence that the Galactic center - the most extreme star-forming environment in the Milky Way - shapes starburst clusters differently than the moderate environment of the spiral arms?
  • The present-day mass functions in the spiral-arm starbursts NGC 3603 YC and Westerlund 1 were scrutinised in two PhD theses at MPIA.

    NGC 3603 YC: Using proper motion membership to discern cluster members from field stars, Boyke Rochau could show that the present-day mass function in the inner parts of the cluster is flat, and hence biased to high-mass stars ( Rochau 2011).

    Westerlund 1: With a thorough colour analysis, yet without proper motion membership, Mario Gennaro found the present-day MF in the Wd 1 full cluster area to be normal ( Gennaro 2011, Gennaro et al. 2011). For both clusters, no dynamical model exists so far. A full dynamical analysis of these clusters would be very valueable to conclude whether their mass distributions were normal at the time of their birth.
  • Conclusion: From the present evidence, Galactic center clusters and spiral arm clusters show very similar radial variations in their present-day mass distributions. Although the tidal evolution of Galactic center clusters along their orbits is dynamically much faster than the dispersal of stars from spiral arm clusters, it is too early to conclude whether any of these starbursts formed with a deviating initial mass function compared to the solar neighbourhood.

Gennaro, M., Brandner, W., Stolte, A., Henning, Th. 2011:
Mass segregation and elongation of the starburst cluster Westerlund 1
Monthly Notices of the Royal Astronomical Society, 412, 2469

Gennaro, M. 2011:
Massive clusters revealed in the near infrared: Constraining the early stages of stellar evolution
PhD thesis, University of Heidelberg, Germany, Novembre 2011

Habibi, M., Stolte, A., Brandner, W., Hußmann, B., Motohara, K. 2013:
The Arches cluster out to its tidal radius: dynamical mass segregation and the effect of the extinction law on the stellar mass function
Astronomy & Astrophysics, 556, A26

Hußmann, B., Stolte, A., Brandner, W., Gennaro, M., Liermann, A. 2012:
The present-day mass function of the Quintuplet cluster based on proper motion membership
Astronomy & Astrophysics, 540, A57

Hußmann, B. 2014:
The Quintuplet cluster: A young massive cluster study based on proper motion membership
PhD thesis, University of Bonn, Germany, January 2014

Rochau, B. 2011:
Young massive star clusters as probes for stellar evolution, cluster dynamics and long term survival
PhD thesis, University of Heidelberg, Germany, May 2011

© Andrea Stolte -- April 2015