Whether supernovae create most of the dust in the cosmos is a controversial question. Observations of a distant supernova have revealed signs of freshly formed dust, but the properties of the dust are unexpected. See Letter p.326
Dust grains play a crucial part in galaxy evolution. They aid in the formation of stars and provide the building blocks of rocky planets and life itself. However, the origin of dust is a contentious topic: it remains unclear whether dust is formed in the violent deaths of massive stars. Supernova explosions are often portrayed as the villains in the life cycle of dust in galaxies, with the harsh million-kelvin gas of the debris thought to efficiently destroy dust grains — produced by the supernova and in the surrounding material — through high-speed collisions with atoms and other grains1,2. But indirect observations of considerable quantities of dust in galaxies at low and high redshifts suggest either that supernovae are producing lots of dust3,4,5,6 or that dust destruction by the supernovae is inefficient. In this issue, Gall et al.7 (page 326) describe observations of telltale signatures from dust in an extragalactic supernova. The results reveal, for the first time, that both of these scenarios are likely to be true.
Over the past few years, thanks in part to far-infrared, millimetre and submillimetre telescopes such as the European Space Agency's Herschel Space Observatory8,9 and the Atacama Large Millimeter/submillimeter Array (ALMA)10,11, evidence has slowly mounted that dust formation in the aftermath of a supernova may in fact be ubiquitous12. In their study, Gall and colleagues investigated whether dust grains were present in the distant supernova 2010jl (Fig. 1). They did this by checking for signs of absorption of light — owing to dust within the supernova — from debris moving towards and away from us, and by searching for thermal emission from dust in the near-infrared (NIR) part of the electromagnetic spectrum.
Using the Very Large Telescope in Chile, the team observed the supernova over 10 epochs starting 26 days after the initial explosion, and found clear evidence that dust grains were formed in the dense shell that lies just behind the expanding supernova shock. They found that, by day 868 after the explosion, the amount of dust in the supernova had grown considerably compared with their observations at earlier epochs. From the NIR emission, they derived a mass of dust equivalent to 830 Earth masses, which is 40 times lower than observed in the ancient Crab Nebula supernova remnant9. Such a small dust mass is unsurprising, given the relative youthfulness of SN 2010jl.
As well as measuring the quantity of freshly formed dust, Gall et al. used their data to graphically show the extent of the absorption of light by the dust grains as a function of wavelength — the extinction curve. This curve provides information on the dust composition (carbon-rich in this case) and size distribution, and reveals perhaps the most significant result from this work: newly formed supernova dust grains are gigantic compared with dust typically found in our Galaxy. The same type of analysis for the Milky Way requires dust grains with a maximum size of 0.25 micrometres to reproduce the observed extinction curve, but in SN 2010jl, the grains need to be greater than 1 μm with a maximum grain radius of 4.2 μm.
The presence of such large grains in a distant supernova is at odds with the size distribution assumed in theoretical dust models used in the literature13. However, this is not the first time that astronomers have observed large grains. The Ulysses robotic spacecraft mission14 recorded substantial emission from grains larger than 2 μm entering our Solar System, and grains as large as 6 μm were detected hitting our planet's atmosphere15. Similarly large dust grains have also been seen in distant γ-ray bursts16.
These large grains seen in our Solar System, and now in an extragalactic supernova, imply not only that is dust created directly as a result of the explosion, but also that supernova dust might be hardy enough to survive the explosion's harsh environment. Owing to their size, larger grains will be more resilient to high-speed collisions compared with smaller grains, and could well survive the explosion in the long term, albeit chipped into smaller pieces as they make their way into the surrounding gas.
Another supernova (SN 1987A) in the nearby Large Magellanic Cloud, a satellite galaxy of the Milky Way, perhaps provides researchers with an ideal laboratory to directly measure the efficiency of dust destruction in supernova shocks. The debris of SN 1987A10,11 is currently moving at 2,000 kilometres per second, and will soon collide with a ring of material left over from the progenitor star before the explosion. Astronomers will be able to observe with ALMA the thermal emission from the dust as the supernova ejecta and the ring collide in real time. Such observations will detect an evolution in dust formation and destruction even at a distance of 50,000 parsecs (the distance from Earth at which the debris of SN 1987A is located). If collisions do prove to be less destructive than theoretical models currently suggest, this will be comforting news to astronomers trying to explain the large dust masses observed in galaxies6,7,17. It seems that supernovae may not be the bad guys after all.
Jones, A. P., Tielens, A. G. G. M., Hollenbach, D. J. & McKee, C. F. Astrophys. J. 433, 797–810 (1994).
Barlow, M. & Silk, J. Astrophys. J. 211, L83–L87 (1977).
Gall, C., Hjorth, J. & Andersen, A. C. Astron. Astrophys. Rev. 19, 43 (2011).
Morgan, H. L. & Edmunds, M. G. Mon. Not. R. Astron. Soc. 343, 427–442 (2003).
Matsuura, M. et al. Mon. Not. R. Astron. Soc. 396, 918–934 (2009).
Dunne, L. et al. Mon. Not. R. Astron. Soc. 417, 1510–1533 (2011).
Gall, C. et al. Nature 511, 326–329 (2014).
Pilbratt, G. L. et al. Astron. Astrophys. 518, L1 (2010).
Gomez, H. L. et al. Astrophys. J. 760, 96–108 (2012).
Matsuura, M. et al. Science 333, 1258–1261 (2011).
Indebetouw, R. et al. Astrophys. J. Lett. 782, L2 (2014).
Gomez, H. L. Proc. Sci. http://pos.sissa.it/archive/conferences/207/146/LCDU2013_146.pdf (2014).
Zubko, V., Dwek, E. & Arendt, R. G. Astrophys. J. Suppl. 152, 211–249 (2004).
Grün, E. et al. Nature 362, 428–430 (1993).
Meisel, D. D., Janches, D. & Mathews, J. D. Astrophys. J. 579, 895–904 (2002).
Li, Y., Li, A. & Wei, D. M. Astrophys. J. 678, 1136–1141 (2008).
Rowlands, K. et al. Mon. Not. R. Astron. Soc. 441, 1040–1058 (2014).