Formation of new stellar populations from gas accreted by massive young star clusters

A Corrigendum to this article was published on 07 September 2016


Stars in clusters are thought to form in a single burst from a common progenitor cloud of molecular gas. However, massive, old ‘globular’ clusters—those with ages greater than ten billion years and masses several hundred thousand times that of the Sun—often harbour multiple stellar populations1,2,3,4, indicating that more than one star-forming event occurred during their lifetimes. Colliding stellar winds from late-stage, asymptotic-giant-branch stars5,6,7 are often suggested to be triggers of second-generation star formation. For this to occur8, the initial cluster masses need to be greater than a few million solar masses. Here we report observations of three massive relatively young star clusters (1–2 billion years old) in the Magellanic Clouds that show clear evidence of burst-like star formation that occurred a few hundred million years after their initial formation era. We show that such clusters could have accreted sufficient gas to form new stars if they had orbited in their host galaxies’ gaseous disks throughout the period between their initial formation and the more recent bursts of star formation. This process may eventually give rise to the ubiquitous multiple stellar populations in globular clusters.

Access options

Rent or Buy article

Get time limited or full article access on ReadCube.


All prices are NET prices.

Figure 1: Colour–magnitude diagrams, including the best-fitting isochrones, and true-colour images for all three clusters.
Figure 2: Normalized radial distributions of the young sequences with respect to normal cluster stars of similar luminosity.


  1. 1

    Gratton, R. G., Carretta, E. & Bragalia, A. Multiple populations in globular clusters. Lessons learned from the Milky Way globular clusters. Astron. Astrophys. Rev. 20, 50 (2012)

    ADS  Article  Google Scholar 

  2. 2

    Piotto, G. et al. A triple main sequence in the globular cluster NGC 2808. Astrophys. J. 661, L53–L56 (2007)

    ADS  CAS  Article  Google Scholar 

  3. 3

    Milone, A. P. et al. Multiple stellar populations in 47 Tucanae. Astrophys. J. 744, 58–79 (2012)

    ADS  Article  Google Scholar 

  4. 4

    Bedin, L. R. et al. ω Centauri: the population puzzle goes deeper. Astrophys. J. 605, L125–L128 (2004)

    ADS  Article  Google Scholar 

  5. 5

    D’Ercole, A., Vesperini, E., D’Antona, F., McMillan, S. L. W. & Recchi, S. Formation and dynamical evolution of multiple stellar generations in globular clusters. Mon. Not. R. Astron. Soc. 391, 825–843 (2008)

    ADS  Article  Google Scholar 

  6. 6

    Bekki, K. Secondary star formation within massive star clusters: origin of multiple stellar populations in globular clusters. Mon. Not. R. Astron. Soc. 412, 2241–2259 (2011)

    ADS  CAS  Article  Google Scholar 

  7. 7

    Renzini, A. Origin of multiple stellar populations in globular clusters and their helium enrichment. Mon. Not. R. Astron. Soc. 391, 354–362 (2008)

    ADS  CAS  Article  Google Scholar 

  8. 8

    Bastian, N. & Lardo, C. Globular cluster mass loss in the context of multiple populations. Mon. Not. R. Astron. Soc. 453, 357–364 (2015)

    ADS  CAS  Article  Google Scholar 

  9. 9

    Marigo, P. et al. Evolution of asymptotic giant branch stars. II. Optical to far-infrared isochrones with improved TP-AGB models. Astron. Astrophys. 482, 883–905 (2008)

    ADS  CAS  Article  Google Scholar 

  10. 10

    Bastian, N., Cabrera-Ziri, I., Davies, B. & Larsen, S. S. Constraining globular cluster formation through studies of young massive clusters. I. A lack of ongoing star formation within young clusters. Mon. Not. R. Astron. Soc. 436, 2852–2863 (2013)

    ADS  Article  Google Scholar 

  11. 11

    Lu, P., Deng, L. & Zhang, X. Blue straggler formation via close binary mass transfer. Mon. Not. R. Astron. Soc. 409, 1013–1021 (2010)

    ADS  Article  Google Scholar 

  12. 12

    Hills, J. G. & Day, C. A. Stellar collisions in globular clusters. Astrophys. J. 17, L87–L93 (1976)

    ADS  Google Scholar 

  13. 13

    Li, C., de Grijs, R. & Deng, L. The binary fractions in the massive young Large Magellanic Cloud star clusters NGC 1805 and NGC 1818. Mon. Not. R. Astron. Soc. 436, 1497–1512 (2013)

    ADS  Article  Google Scholar 

  14. 14

    Hunter, D. A., Elmegreen, B. G., Dupuy, T. J. & Mortonson, M. Cluster mass function in the Large and Small Magellanic Clouds: fading and size-of-sample effects. Astron. J. 126, 1836–1848 (2003)

    ADS  Article  Google Scholar 

  15. 15

    McLaughlin, D. E. & van der Marel, R. P. Resolved massive star clusters in the Milky Way and its satellites: brightness profiles and a catalog of fundamental parameters. Astrophys. J. 161 (Suppl.), 304–360 (2005)

    ADS  CAS  Article  Google Scholar 

  16. 16

    Recchi, S. & Danziger, I. J. Self-enrichment in globular clusters. I. An analytic approach. Astron. Astrophys. 436, 145–154 (2005)

    ADS  CAS  Article  Google Scholar 

  17. 17

    Kroupa, P. On the variation of the initial mass function. Mon. Not. R. Astron. Soc. 322, 231–246 (2001)

    ADS  Article  Google Scholar 

  18. 18

    Fukui, Y. & Kawamura, A. Molecular clouds in nearby galaxies. Annu. Rev. Astron. Astrophys. 48, 547–580 (2010)

    ADS  CAS  Article  Google Scholar 

  19. 19

    Hopkins, P. F., Quataert, E. & Murray, N. The structure of the interstellar medium of star-forming galaxies. Mon. Not. R. Astron. Soc. 421, 3488–3521 (2012)

    ADS  CAS  Article  Google Scholar 

  20. 20

    de Grijs, R. & Parmentier, G. The long-term survival chances of young massive star clusters. Chin. J. Astron. Astrophys. 7, 155–186 (2007)

    ADS  Article  Google Scholar 

  21. 21

    de Grijs, R. A revolution in star cluster research: setting the scene. Phil. Trans. R. Soc. A 368, 693–711 (2010)

    ADS  Article  Google Scholar 

  22. 22

    Shustov, B. M. & Wiebe, D. S. First supernovae in Galactic globular clusters. Mon. Not. R. Astron. Soc. 319, 1047–1055 (2000)

    ADS  CAS  Article  Google Scholar 

  23. 23

    Conroy, C. & Spergel, D. N. On the formation of multiple stellar populations in globular clusters. Astrophys. J. 726, 36–48 (2011)

    ADS  Article  Google Scholar 

  24. 24

    Pflamm-Altenburg, J. & Kroupa, P. Recurrent gas accretion by massive star clusters, multiple stellar populations and mass thresholds for spheroidal stellar systems. Mon. Not. R. Astron. Soc. 397, 488–494 (2009)

    ADS  Article  Google Scholar 

  25. 25

    Naiman, J. P., Ramirez-Ruiz, E. & Lin, D. N. C. External mass accumulation onto core potentials: implications for star clusters, galaxies, and galaxy clusters. Astrophys. J. 735, 25–37 (2011)

    ADS  Article  Google Scholar 

  26. 26

    Milone, A. P., Bedin, L. R., Piotto, G. & Anderson, J. Multiple stellar populations in Magellanic Cloud clusters. I. An ordinary feature for intermediate age globulars in the LMC? Astron. Astrophys. 497, 755–771 (2009)

    ADS  Article  Google Scholar 

  27. 27

    Rich, R. M., Shara, M. M. & Zurek, D. New photometry for the intermediate-age Large Magellanic Cloud globular cluster NGC 2121 and the nature of the LMC age gap. Astron. J. 122, 842–848 (2001)

    ADS  Article  Google Scholar 

  28. 28

    Glatt, K. et al. Age determination of six intermediate-age Small Magellanic Cloud star clusters with HST/ACS. Astron. J. 136, 1703–1727 (2008)

    ADS  CAS  Article  Google Scholar 

  29. 29

    Davis, L. E. A reference guide to the IRAF/DAOPHOT package. (1994)

  30. 30

    Dolphin, A. DOLPHOT user’s guide, version 2.0. (2013)

  31. 31

    Dolphin, A. DOLPHOT/ACS user’s guide, version 2.0. (2011)

  32. 32

    Dolphin, A. DOLPHOT/WFC3 user’s guide, version 2.0. (2011)

  33. 33

    Werchan, F. & Zaritsky, D. The star clusters of the Large Magellanic Cloud: structural parameters. Astron. J. 142, 48–58 (2011)

    ADS  Article  Google Scholar 

  34. 34

    Rafelski, M. & Zaritsky, D. The star clusters of the Small Magellanic Cloud: age distribution. Astron. J. 129, 2701–2713 (2005)

    ADS  Article  Google Scholar 

  35. 35

    Niederhofer, F. et al. Controversial age spreads from the main sequence turn-off and red clump in intermediate-age clusters in the LMC. Astron. Astrophys. (in the press); preprint at (2015)

  36. 36

    Girardi, L. et al. An extended main-sequence turn-off in the Small Magellanic Cloud star cluster NGC 411. Mon. Not. R. Astron. Soc. 431, 3501–3509 (2013)

    ADS  Article  Google Scholar 

  37. 37

    Li, C., de Grijs, R. & Deng, L. The exclusion of a significant range of ages in a massive star cluster. Nature 516, 367–369 (2014)

    ADS  CAS  Article  Google Scholar 

  38. 38

    Bastian, N. & Niederhofer, F. The morphology of the subgiant branch and red clump reveal no sign of age spreads in intermediate-age clusters. Mon. Not. R. Astron. Soc. 448, 1863–1873 (2015)

    ADS  Article  Google Scholar 

  39. 39

    Goudfrooij, P. et al. On the interpretation of subgiant branch morphologies of intermediate-age star clusters with extended main sequence turnoffs. Mon. Not. R. Astron. Soc. 450, 1693–1704 (2015)

    ADS  Article  Google Scholar 

  40. 40

    Dotter, A. et al. The Dartmouth stellar evolutionary database. Astrophys. J. 178 (Suppl.), 89–101 (2008)

    CAS  Article  Google Scholar 

  41. 41

    Binney, J. & Tremaine, S. Galactic Dynamics (Princeton Univ. Press, 1998)

  42. 42

    Ashman, K. M. & Zepf, S. E. Globular Cluster Systems (Cambridge Astrophys. Ser. Vol. 30, Cambridge Univ. Press, 1998)

  43. 43

    Goudfrooij, P., Girardi, L. & Kozhurina-Platais, V. Extended main sequence turnoffs in intermediate-age star clusters: a correlation between turnoff width and early escape velocity. Astrophys. J. 797, 35–56 (2014)

    ADS  Article  Google Scholar 

  44. 44

    van der Marel, R. P., Alves, D. R., Hardy, E. & Suntzeff, N. B. New understanding of Large Magellanic Cloud structure, dynamics, and orbit from carbon star kinematics. Astron. J. 124, 2639–2663 (2002)

    ADS  Article  Google Scholar 

Download references


We thank F. Ferraro, P. Kroupa and X. K. Liu for discussions. Partial financial support for this work was provided by the National Natural Science Foundation of China through grants 11073001, 11373010 and 11473037. C.L. was also partially supported by the Chinese Academy of Sciences (grant XDB09000000) and the 973 Program (grant 2014CBB45700). A.M.G. was funded by a National Science Foundation Astronomy and Astrophysics Postdoctoral Fellowship under award no. AST-1302765. C.-A.F.-G was supported by a National Science Foundation grant AST-1412836.

Author information




C.L., R.d.G. and L.D. jointly designed and coordinated this study. C.L. performed the data reduction. C.L., R.d.G., Y.X. and Y.H. collaborated on the detailed analysis. L.D. provided ideas that improved the study’s robustness. A.M.G. and C.-A.F.-G. led the theoretical analysis of the gas-accretion physics. All authors read, commented on and jointly approved submission of this article.

Corresponding author

Correspondence to Chengyuan Li.

Ethics declarations

Competing interests

The authors declare no competing financial interests.

Extended data figures and tables

Extended Data Figure 1 Stellar number-density profiles.

Here ρ is the number density of stars at a given radius R. a, NGC 1783. The vertical solid line indicates the cluster’s core radius, that is, the position where the number density decreases to half the central value. The ±1σ uncertainties shown are due to Poisson noise. b, c, As a, but for NGC 1696 (b) and NGC 411 (c).

Extended Data Figure 2 Spatial distributions of cluster and field stars.

Red and blue points represent cluster stars and the adopted field stars, respectively. The black dots are all observed stars except for the cluster and field stars. RA, right ascension; dec., declination. a, NGC 1783; b, NGC 1696; and c, NGC 411.

Extended Data Figure 3 Field-star-decontaminated colour–magnitude diagrams for samples of stars at different radii in NGC 1783.

a, R ≥ 30″; b, R ≥ 60″; and c, R ≥ 90″.

Extended Data Figure 4 Field-star-decontaminated colour–magnitude diagrams of NGC 1783 for three different adopted reference fields.

a, b, Resulting colour–magnitude diagram (a) based on the field-star sample drawn from the image containing the cluster (b). c, d, As a and b, but for a representative field region taken from a separate image. e, f, As c and d, but for a different field region taken from the same, separate image. The black points are stars which are located in the cluster region, whereas the red points are the adopted field stars.

Extended Data Figure 5 Field-star-decontaminated colour–magnitude diagrams of NGC 1783 for three different grid sizes.

a, 0.30 mag × 0.15 mag; b, 0.40 mag × 0.20 mag; and c, 0.50 mag × 0.25 mag.

Extended Data Figure 6 Colour–magnitude analysis.

ad, NGC 1783; eh, NGC 1696; il, NGC411. First column (a, e, i), raw colour–magnitude diagrams; second column (b, f, j), field-decontaminated colour–magnitude diagrams (also shown in Fig. 1); third column (c, g, k), colour–magnitude diagrams of the representative field regions; and fourth column (d, h, l), stellar colour–magnitude distributions that were removed from the raw catalogues.

Extended Data Figure 7 Colour–magnitude diagrams highlighting specific features.

a, Purple and dark green squares, stars in NGC 1783 sequences A and B, respectively. Dark green circles, corresponding red-giant-branch and red-clump stars, used for comparison with sequence B. The combination of dark green and purple circles represents the sample used for comparison with sequence A. b, Purple squares, NGC 1696 young-sequence stars; red circles, corresponding red-giant-branch and red-clump stars used for comparison. c, As b, but for NGC 411.

Extended Data Figure 8 Gas-accretion diagnostic diagram.

V is the relative velocity of the cluster with respect to the gas, and n represents the gas density. The shaded regions indicate the parameter space where an NGC 1783-like cluster can accrete the required mass to form the two additional generations of stars, namely one of 250M over 520 Myr (blue, corresponding to sequence B in Fig. 1), and a second of 370M over 440 Myr (green, corresponding to sequence A in Fig. 1). We have assumed a star-formation efficiency of 10% for all calculations in this figure. The regions to the right of the ‘Gravity’ curves correspond to where Bondi accretion can accumulate at least the required mass, and the regions to the right of the ‘Sweep’ curves correspond to where accretion by collisional sweeping up of ambient interstellar gas by seed intracluster gas can accumulate at least the required mass. The parameter space above the ‘Ram’ curves are excluded because ram pressure strips clusters of their gas in those regions. See Methods for details.

Extended Data Figure 9 Gas mass, M, accreted from the interstellar medium as a function of time, t.

We have adopted a star-formation efficiency of 10% and calculated representative interstellar gas-accretion frameworks that can explain the stellar masses in the secondary sequences A and B in NGC 1783. For the Bondi regime (solid lines), we used a relative velocity of 4 km s–1 and a density of 0.3 cm–3. For the sweeping regime (dashed lines), we used a velocity of 50 km s–1 and a density of 0.05 cm–3. The blue and green filled regions indicate the stellar masses and age offsets of NGC1783 sequences A and B, respectively. The vertical extent of each region provides an estimate of the range in allowed masses. Specifically, for each sequence we plot a region centred on the mass derived using the Kroupa initial mass function, as given in the main text; the range to higher and lower masses is equal to the mean of the differences between the total masses derived using a Salpeter and Kroupa initial mass function of 180M (multiplied by the assumed 10% star formation efficiency).

PowerPoint slides

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Li, C., de Grijs, R., Deng, L. et al. Formation of new stellar populations from gas accreted by massive young star clusters. Nature 529, 502–504 (2016).

Download citation

Further reading


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