Stellar astrophysics

The mystery of globular clusters

The discovery of multiple stellar populations — formed at different times — in several young star clusters adds to the debate on the nature and origin of such populations in globular clusters from the early Universe. See Letter p.502

Since the discovery of the first globular cluster in 1665, these large, ancient agglomerates of stars — which can host up to a million suns — have fascinated both astronomers and the public. They are visible through small telescopes, and their exquisite spherical symmetry singles them out in the sky and makes them easy to classify. However, their formation and evolution history is unclear. On page 502 of this issue, Li et al.1 report observations of young star clusters (Fig. 1) that may help to crack the mystery of the oldest star clusters.

Figure 1

ESA/Hubble & NASA

Hubble Space Telescope image of the young star cluster NGC 411.

Globular clusters remain gravitationally bound as they orbit their host galaxy, on a timescale comparable to the lifetime of the low-mass stars they host. They are 10 billion to 13 billion years old — their age defines the boundaries of the age of the Universe. These well-studied systems were long thought to be simple and to host a single population of stars that all formed at the same time.

But in 2004, everything we knew about globular clusters changed radically. Using accurate Hubble Space Telescope photometry2,3, astronomers detected not one, but multiple stellar populations in ω Centauri, one of the most massive globular clusters in the Milky Way. Subsequent studies (see ref. 4, for example) of other globular clusters confirmed that this was not an isolated finding but the discovery of a general feature that revolutionized our understanding of such objects. These stellar populations were shown to exhibit unique chemical properties that are not found in any other stellar environment5. This means that these clusters are not simple at all, and have experienced more than one star-forming event during their lifetime.

The exciting news2,3 inspired different star-formation models to account for the photometric and spectroscopic properties of the different populations hosted by single clusters. For example, colliding winds from late-stage, medium-mass stars or ejecta from fast-rotating massive stars were invoked as a trigger and/or an origin of the processed material that fuelled the second generation of star formation. Astronomers also proposed an explanation based on a single generation of stars. They suggested that very-low-mass stars with protoplanetary disks — gas rotating around newly formed stars — could acquire the observed properties by sweeping up material shed from interacting binary or rapidly rotating massive stars in a continual growth process.

However, most of these theories suffer from major issues6. Some of them imply a more massive original cluster than observed, with a substantial fraction of the original stars lost to the galactic halo. But this enhanced mass is in contrast to theoretical expectations for the dynamical evolution of these systems7 and to observations of dwarf galaxies and of young massive clusters in our Local Group of galaxies8. The consensus in the community is that we urgently need alternative, innovative ideas to overcome the impasse.

Enter Li and colleagues1. The authors challenge the status quo by presenting observations of three massive clusters that are 1 billion to 2 billion years old in the Magellanic Clouds, members of our Local Group. The authors show clear evidence of a late burst of star formation that occurred a few hundred million (up to one billion, within the errors) years after the clusters' initial formation epoch. Multiple stellar population sequences are visible even in the authors' raw diagrams. The colours of the younger stellar sequences are consistent with an enhanced abundance of helium. This would be expected as a result of the chemical anomalies in globular clusters that are due to hydrogen burning at high temperature (helium is the main yield of hydrogen burning).

To explain the data, Li et al. propose that such clusters orbiting within the gaseous disks of their host galaxies could accrete sufficient gas reservoirs to form the next generation of stars. They suggest that this mechanism may account for the ubiquitous multiple stellar populations in globular clusters, assuming that the observed massive young star clusters in the Magellanic Clouds are the modern counterparts of the old globular clusters. It is a plausible working hypothesis, but not a consensus view: some astronomers think that the formation mechanism of young massive star clusters in our neighbouring galaxies might be different from the globular-cluster formation that occurred in the early Universe. The link between young and old clusters has yet to be fully established. The determination of whether the chemical properties of the youngest stars in the Magellanic clusters are similar to those of globular-cluster stars will eventually address the issue.

Li and colleagues may not yet have all the answers needed to put the proposed general model on a firm quantitative footing. For example, the mass in the younger observed population sequence is much smaller than that typically observed in globular clusters. There is no explanation given for the origin of the additional helium that is claimed to be present in the youngest stars, nor for the fact that the accreted material must have the same metal content as the original protocluster, even though the host galaxies might have evolved chemically between the distinct star-formation events. (By 'metal', astronomers mean any element after helium in the periodic table.)

Nevertheless, the findings present an innovative approach that deserves further attention. It will certainly advance the ongoing debate, as well as trigger original thoughts, future observations and corresponding interpretations. And it could lead to a final, robust explanation in the not too distant future — an example of how scientific debate works at its best.Footnote 1


  1. 1.

    See all news & views


  1. 1

    Li, C. et al. Nature 529, 502–504 (2016).

    ADS  CAS  Article  Google Scholar 

  2. 2

    Bedin, L. R. et al. Astrophys. J. 605, L125–L128 (2004).

    ADS  Article  Google Scholar 

  3. 3

    Piotto, G. et al. Astrophys. J. 621, 777–784 (2005).

    ADS  CAS  Article  Google Scholar 

  4. 4

    Milone, A. P. et al. Astrophys. J. 808, 51 (2015).

    ADS  Article  Google Scholar 

  5. 5

    Gratton, R. G., Carretta, E. & Bragaglia, A. Astron. Astrophys. Rev. 20, 50 (2012).

    ADS  Article  Google Scholar 

  6. 6

    Bastian, N. Preprint at (2015).

  7. 7

    Krause, M. G. H., Charbonnel, C., Bastian, N. & Diehl, R. Preprint at (2015).

  8. 8

    Larsen, S. S., Brodie, J. P., Grundahl, F. & Strader, J. Astrophys. J. 797, 15 (2014).

    ADS  CAS  Article  Google Scholar 

Download references

Author information



Corresponding authors

Correspondence to Antonella Nota or Corinne Charbonnel.

Related links

Related links

Related links in Nature Research

Astronomy: A new spin on the first stars

Astrophysics: Two black holes found in a star cluster

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Nota, A., Charbonnel, C. The mystery of globular clusters. Nature 529, 473–474 (2016).

Download citation


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing