Formation of Interstellar C60 from Silicon Carbide Circumstellar Grains
J. J. Bernal1, P. Haenecour2, J. Howe3, T. J. Zega2,4, S. Amari5, and L. M. Ziurys1,6,7
The Astrophysical Journal Letters, Volume 883, Number 2
A long-standing mystery in astronomical spectroscopy concerns diffuse interstellar bands, a family of absorption features seen in the spectra of the interstellar medium of the Milky Way and of other galaxies. First observed almost 100 years ago, the origin of any of the bands was unknown until 2015, when four of them were assigned1 to the cation of buckministerfullerene (C60+; the uncharged molecule is often referred to simply as fullerene, or colloquially as a buckyball). Fullerene and its analogue, C70, are by far the biggest molecules detected in space, raising the question of how such large species can form in those rarified conditions. Researchers have suggested that fullerene forms in the outflows of old, carbon-rich stars known as asymptotic giant branch stars2 — the temperatures and densities of these outflows promote chemistry similar to that of combustion. This could lead to the formation of soot, which can contain fullerene-like structures. Writing in Astrophysical Journal Letters, Bernal et al.3 propose a very different formation route for fullerene.
The carbon atoms in fullerene are arranged in the shape of a football, a molecular structure that is remarkably stable but also difficult to construct. Fullerene has been made in the laboratory in experiments designed to probe the chemistry that occurs in carbon-rich stars: carbon in the form of graphite was vaporized into a high-density helium flow, producing carbon clusters4. The discovery that fullerene was among the reaction products led to the award of the Nobel Prize in Chemistry to Harry Kroto, Richard Smalley and Robert Curl in 1996.
However, the range of temperatures required to create fullerene in this way is quite specific2; outside that range, molecules known as polycyclic aromatic hydrocarbons (PAHs) are produced instead. These molecules are 2D sections of a single layer of graphite (a graphene sheet), decorated with hydrogen atoms. Subsequent experiments5,6 have shown that PAHs that contain more than 60 carbon atoms are converted into fullerenes when exposed to sufficient ultraviolet irradiation.
The first astronomical source in which fullerene was detected was the star Tc 1 (ref. 7). Puzzlingly, however, the emission associated with fullerene came from a location far away from the star and its ultraviolet photons, whereas the PAH emissions were closer to the star. On the basis of the previously reported laboratory experiments, this is the opposite of what should happen if fullerene forms from PAHs in this source8. So how can the locations of the emissions be explained?
Bernal and co-workers now report that fullerene also forms readily from silicon carbide (SiC), which has been proposed to be the first carbonaceous material to condense out of old, carbon-rich stars9. The authors rapidly heated grains of the crystalline form of SiC that is found in highest abundance in meteorites10, and irradiated them with xenon ions, mimicking the heating caused by shock waves around old stars.
Using a transmission electron microscope to image the surfaces of the samples down to the subnanometre scale, the scientists observed that the grain material had altered notably as a result of its treatment (Fig. 1). Silicon atoms had percolated to the outer layers of the grains, leaving behind what looked like sheets of carbon atoms in a hexagonal ‘chicken-wire’ arrangement — that is, graphene sheets.
The transformation of the outer layers of SiC into graphene sheets at high temperatures had been reported11 previously for a different form of SiC from that studied by Bernal and colleagues. However, Bernal et al. also observed the formation of hemispherical structures with diameters similar to that of fullerene. Their work thus provides a convincing new mechanism for the formation of fullerene in evolved stars.
Bernal et al. report another piece of evidence supporting the idea that SiC grains are rapidly heated and bombarded with ions in evolved stars. They have identified a fragment of the Murchison meteorite — a highly studied meteorite that is rich in organic compounds — in which the ratio of carbon-12 to carbon-13 isotopes is typical of material from an old, carbon- rich star. This indicates that the fragment was not produced during or after the formation of the meteorite, but instead is stardust that originated in an old star. The fragment has a core of SiC surrounded by graphene sheets. However, previous analyses12 of graphite-containing stardust found evidence only of titanium carbide cores, rather than SiC cores. This raises the question of how common SiC cores are in graphite-containing stardust.
The rapid heating of SiC grains in the presence of hydrogen can lead to the formation of PAHs13. Bernal and colleagues’ findings therefore suggest that the thermal conversion of SiC to graphene sheets in evolved stars could be the first step in the formation of large carbon-containing molecules in general: subsequent (or simultaneous) exposure of the graphene to atomic hydrogen produces PAHs, whereas ion bombardment produces fullerene. Alternatively, PAH molecules might be molecular intermediates in the formation of carbon soot, which can then be broken down by ultraviolet irradiation to make PAHs again14.
The efficiency of Bernal and colleagues’ fullerene-forming mechanism is unknown, raising the question of how many SiC grains are needed to account for the observed abundance of fullerene molecules in space. If there aren’t enough grains, then a further mechanism will be required to explain the abundance of fullerene. By contrast, if there are too many SiC grains, what happens to the ‘excess’ fullerene molecules produced, given that they are notoriously difficult to degrade? More experiments and detailed modelling of the formation of fullerene and of other carbon-containing large molecules from SiC grains are needed to understand this process, and to quantify its importance in old stars.
The launch of the James Webb Space Telescope in 2021 will provide powerful new tools for studying old stars, among other astronomical objects. Observations of fullerene- containing sources7,8 such as Tc 1 will be able to constrain the regions in which SiC grains, fullerene and PAHs are present, providing more clues about how large molecules are actually formed. Further analysis and modelling of the routes involved will eventually allow astronomers to suggest the identities of the other mysterious molecules responsible for the diffuse interstellar bands.
Alessandra Candian is at the van’t Hoff Institute for Molecular Sciences, University of Amsterdam, 1090 GD Amsterdam, the Netherlands, and at the Leiden Observatory, University of Leiden, Leiden, the Netherlands. e-mail: PG4gdWVycz0iem52eWdiOm4ucG5hcXZuYUBoaW4uYXkiPm4ucG5hcXZuYUBoaW4uYXk8L24+
1. Campbell, E. K., Holz, M., Gerlich, D. & Maier, J. P. Nature 523, 322–323 (2015).
2. Jäger, C., Huisken, F., Mutschke, H., Llamas Jansa, I. & Henning, T. Astrophys. J. 696, 706–712 (2009).
3. Bernal, J. J. et al. Astrophys. J. 883, L43 (2019).
4. Kroto, H. W., Heath, J. R., O’Brien, S. C., Curl, R. F. & Smalley, R. E. Nature 318, 162–163 (1985).
5. Berné, O. & Tielens, A. G. G. M. Proc. Natl Acad. Sci. USA 109, 401–406 (2012).
6. Zhen, J., Castellanos, P., Paardekooper, D. M., Linnartz, H. & Tielens, A. G. G. M. Astrophys. J. 797, L30 (2014).
7. Cami, J., Bernard-Salas, J., Peeters, E. & Male, S. E. Science 329, 1180–1182 (2010).
8. Cami, J., Peeters, E., Bernard-Salas, J., Doppmann, G. & De Buizer, J. Galaxies 6, 101 (2018).
9. Frenklach, M., Carmer, C. S. & Feigelson, E. D. Nature 339, 196–198 (1989).
10. Daulton, T. L. et al. Science 296, 1852–1855 (2002).
11. Mishra, N., Boeckl, J., Motta, N. & Iacopi, F. Phys. Status Solidi A 213, 2277–2289 (2016).
12. Croat, K. T., Bernatowicz, T. J. & Daulton, T. L. Elements 10, 441–446 (2014).
13. Merino, P. et al. Nature Commun. 5, 3054 (2014).
14. Cherchneff, I., Barker, J. R. & Tielens, A. G. G. M. Astrophys. J. 401, 269–287 (1992).