My research uses star clusters as fossils that provide information about galaxy histories. Stars have a chemical imprint of the gas conditions of the galaxy at the time when they formed. I measure the abundances of different chemical elements in stars to infer how the gas content and details of star formation evolved over time.

The M31 Project

We have observed a sample of 31 globular clusters in M31 using HIRES on the Keck I telescope.

M31 is the closest “normal” sized galaxy to the Milky Way. At a distance of ~780 Mpc, stars in M31 are too faint for detailed chemical abundance analysis from high resolution (R>20,000) spectra. However, globular clusters in M31 are perfect targets for high resolution spectrographs on 6-10 m class telescopes. Out of the hundreds of confirmed globular clusters in M31, we picked ~30 from the Revised Bologna Catalog, with the goal of sampling the range in properties available in the cluster system of M31. The clusters have V band magnitudes of 14 to 17, M31 galactocentric radii of 4 – 117 kpc, and metallicities of [Fe/H] = -2 to -0.2.

This slideshow requires JavaScript.

We measured abundances of Fe I, Fe II, Ca I, Si I, Ti I , Ti II, Mg I, Na I and Al I in this sample in Colucci, Bernstein & Cohen (2014). This is the first time that the abundances of these elements have been measured in any other normal mass galaxy!  We determined that the M31 clusters have a very similar alpha-element (Ca, Si, Ti) abundance pattern to Milky Way clusters. We found a different alpha-element abundance pattern for one cluster (G002), which may indicate it was accreted by M31.

We are currently in the process of measuring many more Fe-peak and neutron-capture element abundances in this exciting sample of clusters. All figures are reproduced from Colucci, Bernstein & Cohen (2014).

The Cen A Project

We are building a sample of globular clusters with detailed chemical abundances in the nearest early type galaxy Cen A (NGC 5128).


At a distance of ~4 Mpc, Cen A is the nearest early type galaxy. Until now the detailed abundance pattern of any early type galaxy was an unknown because it is impossible to obtain high resolution spectra of individual stars at these distances. Using our original technique for analysis of integrated light high resolution spectra of globular clusters, we have been able to get a first peek at the chemical evolution of an early type galaxy, Cen A. The first results for Fe I and Ca I for a sample of 10 clusters were published in Colucci et al. (2013). The clusters have [Fe/H] from -1.6 to -0.2 and [Ca/Fe] indicative of enrichment dominated by core collapse supernovae. The [Ca/Fe] of these clusters remains enhanced at high [Fe/H], which is similar to what is seen for stars in the bulge of the Milky Way. This may indicate a similar formation pattern for spheroidal galaxy components.

This slideshow requires JavaScript.

We are currently using the MIKE spectrograph on the Magellan Clay telescope in Las Campanas, Chile to increase the sample of clusters observed from 10 to ~25. We are also measuring abundances of additional alpha, Fe-peak, and heavy elements.

Ph. D. Dissertation: Detailed Chemical Abundances in Clusters in the Large Magellanic Cloud

I used a sample of star clusters in the LMC to test a new method for detailed abundance analysis of integrated light spectra.

As part of my dissertation I further developed a new method for obtaining the first detailed chemical abundances for unresolved clusters, which makes it possible to determine detailed abundances beyond the Milky Way and its closest satellites for the first time. Detailed abundances in extragalactic clusters can help constrain galaxy formation theories by providing information on star formation rates and durations, chemical enrichment history, and supernovae yields for galaxies beyond the Milky Way.

I used a sample of 8 star clusters in the Large Magellanic Cloud to develop techniques for analyzing high signal-to-noise ratio, high resolution spectra. These clusters have ages of 0.05 to 12 Gyr, and they are a particulary important sample, as they can be used to demonstrate the accuracy (+/- 0.1 dex in [X/Fe]) of our new method on clusters with ages younger than ~10 Gyr, which is an age range that cannot be probed in the Milky Way. The large age range of the sample also provides a comprehensive picture of the chemical enrichment and star formation history of the LMC.

The high signal-to-noise ratio, high resolution spectra we obtain for abundance analyses also contains a wealth of other information about globular cluster stellar populations. We are able to measure internal velocity dispersions of extragalactic globular clusters to high precision, providing constraints on their total masses and mass-to-light ratios. Additionally, we are investigating methods for constraining horizontal branch morphology of extragalactic clusters using their integrated Balmer line profiles. In conjunction with the ages and metallicities we determine with our abundance analysis, we can also use the observed and predicted integrated colors of globular clusters to probe the reddening in galaxies.

This slideshow requires JavaScript.