How black holes restrain old galaxies

Supermassive black holes are thought to keep star formation under control by ejecting or stirring gas in galaxies. Observations of an old galaxy reveal a potential mechanism for how this process occurs. See Letter p.504

When supermassive black holes at the centres of galaxies accrete matter, they turn into powerful engines that can potentially expel the gas of their host galaxies, thereby halting star formation1. Despite numerous efforts2,3 to observe supermassive black holes in the process of quenching star formation, conclusive evidence for such a process has remained elusive, particularly in the nearby Universe. On page 504 of this issue, Cheung et al.4 present state-of-the-art observations that might finally show how supermassive black holes can prevent galaxies that are already dominated by old, red stars from forming new ones.

According to our current view of their formation, galaxies grow by merging with other galaxies or by forming new stars, either from freshly acquired gas or from material that is lost by their old and dying stars. Merging events can further rearrange them into rounder shapes, whereas galaxies that benefit from a constant supply of external or recycled gas can form a stellar disk dominated by young, blue stars. Galaxies that have no external gas supply evolve passively into old, red stellar systems collectively called early-type galaxies (otherwise known as elliptical or lenticular galaxies).

And yet, up to 75% of early-type galaxies contain gas that could potentially fuel new bursts of star formation5. The fact that stars are observed to form in only 10–20% of such galaxies6,7 suggests that some kind of star-formation quenching is taking place. This is consistent with the finding that gas in early-type galaxies is usually in a warm, ionized state, rather than existing as cold clouds of gaseous molecules from which stars can form.

The radiation emitted from ionized gas is generally powered by hot but old stars6,8,9, rather than by massive, newly born stars such as those in the disk of the Milky Way and in other spiral galaxies. Also, unlike cold molecular gas, which always orbits in a thin disk at the circular velocity set by the local gravitational potential, the warm gas of early-type galaxies often shows sizeable, random motion — suggesting either that it is being stirred up somehow, or has yet to settle down10. However, the kinematics of the ionized gas in early-type galaxies has so far been considered to be consistent with coherent, although perhaps not completely ordered, rotation.

Enter Cheung and colleagues. The authors used spectroscopic observations that allowed them to map the motion of ionized gas across a galaxy and to infer what is powering the gas's emission. In this way, they conclusively show that ionized gas in a substantial fraction of early-type galaxies does not rotate in a coherent fashion. Instead, the authors propose a model whereby the approaching and receding material that is observed across the gas-velocity field of such objects is due to a bi-conical outflow of gas powered by a centrally active supermassive black hole (Fig. 1), rather than resulting from the circulation of gas in an inclined disk.

Figure 1: Bi-conical gas outflow from an early-type galaxy.

Cheung et al.4 propose a model of gas outflow from an old, red galaxy (an early-type galaxy) that explains their observations of an early-type galaxy nicknamed Akira. The authors suggest that, when the supermassive black hole at the centre of such a galaxy accretes matter, this can stimulate nuclear activity that drives a bi-conical outflow of gas — stirring up gas that is moving elsewhere in the galaxy, and possibly expelling part of it. This provides a possible mechanism by which star formation in early-type galaxies is quenched.

In maps of the intensity of ionized-gas emission, the class of early-type galaxy in which such bi-conical outflows occur is characterized by bisymmetrical, elongated features that align with the gradient in the gas-velocity field. Cheung and colleagues' model has the advantage of having a simple explanation for this defining characteristic: it reflects the accumulation of material on both the approaching and receding sides of the outflow.

The researchers present observations for one prototypical example of this class of object, in which the activity of the central supermassive black hole seems to have been triggered by interaction with a nearby companion galaxy. In this example, which Cheung et al. nickname Akira, such activity is sufficient to sustain the kinetic power of the outflowing wind, which in turn balances the cooling of the warm, ionized gas. Even if the central activity is not enough to rid Akira of its gas, it would provide sufficient energy to stir it by causing turbulence and shocks, and therefore still prevent the gas cooling that leads to star formation.

Although the bi-conical-outflow model presented by Cheung and co-workers is only qualitative, the authors' results might aid our understanding of the role of supermassive black holes in galaxy evolution. Akira is just one of the 10,000 objects that will eventually be targeted by the ongoing campaign (the MaNGA survey11) from which the authors' data are drawn. From the 700 MaNGA galaxies presently being surveyed, Akira-like objects occur as a small (5%), yet non-negligible fraction of early-type galaxies4. This could be just the tip of the iceberg — central-black-hole activity can be triggered several times by different accretion episodes, and thus many galaxies that do not presently show outflows could have been stirred, or their gas expelled, by a previous episode.

The observed outflows may also help to solve another riddle: the origin of gas in early-type galaxies. One way to tell whether an early-type galaxy acquired gas from other galaxies or from recycled material lost from its stars is to compare the gas's angular momentum with that of the stars'. If the gas was internally produced, it should follow the motions of the stars, whereas if it was externally acquired it could just as well move in the opposite direction. Observations12 of the stellar and gaseous kinematics of early-type galaxies show that the gas comes from mixed sources. This is puzzling, but not really problematic, given that, for instance, galaxies in crowded environments such as galaxy clusters do not interact easily with each other and therefore find it hard to steal gas from smaller companions12.

More troubling is the fact that 25% of early-type galaxies have little or no gas at all5. Because early-type galaxies have similar, old stellar populations that would also return gas to their hosts over time, one would expect all early-type galaxies to retain at least some of this recycled material in the absence of an external gas source. By providing evidence for a mechanism capable of removing at least part of the gas, Cheung and colleagues' work might bring us a step closer to explaining why some early-type galaxies seem to be devoid of gas.

Footnote 1


  1. 1.

    See all news & views


  1. 1

    Fabian, A. C. Annu. Rev. Astron. Astrophys. 50, 455–489 (2012).

    ADS  CAS  Article  Google Scholar 

  2. 2

    Schawinski, K. et al. Mon. Not. R. Astron. Soc. 382, 1415–1431 (2007).

    ADS  CAS  Article  Google Scholar 

  3. 3

    Cicone, C. et al. Astron. Astrophys. 562, A21 (2014).

    Article  Google Scholar 

  4. 4

    Cheung, E. et al. Nature 533, 504–508 (2016).

    ADS  CAS  Article  Google Scholar 

  5. 5

    Sarzi, M. et al. Mon. Not. R. Astron. Soc. 366, 1151–1200 (2006).

    ADS  CAS  Article  Google Scholar 

  6. 6

    Sarzi, M. et al. Mon. Not. R. Astron. Soc. 402, 2187–2210 (2010).

    ADS  CAS  Article  Google Scholar 

  7. 7

    Young, L. M. et al. Mon. Not. R. Astron. Soc. 414, 940–967 (2011).

    ADS  CAS  Article  Google Scholar 

  8. 8

    Binette, L., Magris, C. G., Stasińska, G. & Bruzual, A. G. Astron. Astrophys. 292, 13–19 (1994).

    ADS  CAS  Google Scholar 

  9. 9

    Stasińska, G. et al. Mon. Not. R. Astron. Soc. 391, L29–L33 (2008).

    ADS  Google Scholar 

  10. 10

    Young, L. M., Bureau, M. & Cappellari, M. Astrophys. J. 676, 317–334 (2008).

    ADS  CAS  Article  Google Scholar 

  11. 11

    Bundy, K. et al. Astrophys. J. 798, 7 (2015).

    ADS  Article  Google Scholar 

  12. 12

    Davis, T. A. et al. Mon. Not. R. Astron. Soc. 417, 882–899 (2011).

    ADS  CAS  Article  Google Scholar 

Download references

Author information



Corresponding author

Correspondence to Marc Sarzi.

Related links

Related links

Related links in Nature Research

Astrophysics: The slow death of red galaxies

Galaxy formation: When the wind blows

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Sarzi, M. How black holes restrain old galaxies. Nature 533, 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.