Main

Since August 2014, the ESA Comet Rendezvous Mission, Rosetta16,17, has been in orbit around the Jupiter-family comet 67P/Churyumov-Gerasimenko, monitoring the evolution of the comet’s nucleus, near-nucleus region, and inner coma as a function of increasing solar flux input, as the comet moves towards the Sun. As part of these studies, the COmetary Secondary Ion Mass Analyser (COSIMA)18 onboard Rosetta is collecting comet grains from the near-nucleus region and the inner coma onto special target plates19, which are subsequently imaged and compositionally investigated by time-of-flight secondary ion mass spectrometry using an indium ion source. The grain collection commenced at a heliocentric distance of 3.57 astronomical units (where 1 au is the average Sun–Earth distance), when the comet was still at low activity. The optical analysis of the grains captured on the target plates at distances beyond 3 au shows that most have fragmented upon capture and a large fraction of grains more than 50 µm across have shattered. Figure 1a shows a typical example of a dust particle that has crumbled into a rubble pile upon collection, while Fig. 1b shows an example of a dust particle that has shattered into a loosely connected cluster with a wide range of sub-component sizes. These two types of feature are representative of most large particles collected at less than 30 km from the nucleus during the first three months of the orbital phase. Given that the dust particles hit the target with a relatively low velocity (1–10 m s−1)19, their tensile strength must be very low. From the inertial deceleration forces upon grain capture the strength of the material can be approximated, and a first rough estimate relevant for the present fragmentation process is on the order of 1,000 Pa.

Figure 1: Dust particles.
figure 1

a, An example of a dust particle that crumbled into a rubble pile when collected. The particle was collected at a nucleus distance of 10–20 km, between 25 and 31 October 2014, with corresponding heliocentric distance range 3.11–3.07 au. The image was obtained with two different grazing illumination conditions (top image illuminated from the right, bottom image from the left). The brightness is presented in logarithmic scale to emphasize the shadows, which indicate that the altitude above the target reaches about 100 µm. As the particle lies 4.2 mm below the centre of the collecting target, the shadows are tilted with regard to the horizontal direction. b, An example of a dust particle that shattered when collected. The distance, time of collection, illumination conditions, and logarithmic scale are the same as for a. The shadows indicate that the altitude above the target reaches about 60 µm. The two grains visible on the right are not part of the shattered cluster.

PowerPoint slide

Full size image

The disintegration of cometary grains in the coma is often described as resulting from an icy grain component that evaporates when exposed to solar radiation, producing a secondary source for comet gaseous material20,21. A dusty secondary source can, however, also be attributed to certain organic grains that are not mantled by water ice22. The coma dust returned by Stardust23 featured various types of grain, including specimens that had disintegrated along the deceleration tracks when entering the aerogel (the ultralight porous gel in which the grains were captured) at velocities of the order of 6 km s−1, and hence were composed of very fine or thermally unstable components24,25. The morphology of the grains collected by COSIMA supports the presence of solely refractory material. A grain composed of an ice–mineral mixture would not shatter at low-velocity collection; instead, the icy part of such a grain would evaporate very shortly after collection, leaving one or more voids in the particle that remains on the target plate. Grains composed of (nearly) pure water-ice would evaporate at or shortly after collection and create a dark signature on the target plate. At the scale of the COSIMA image resolution (pixel size is 14 µm), there is no hint of volatiles having left the grains after collection. In other words, there is no indication of an ice–mineral mixture, or of pure icy grains hitting the target. This is in contrast to cometary grains remotely observed, or collected before the Rosetta mission.

The most important difference between the Stardust and COSIMA grains is the heliocentric distance at which they were captured. The Stardust samples were collected during a comet fly-by at 1.85 au, whereas the grains collected by COSIMA were dragged off the nucleus of a re-approaching comet at heliocentric distances greater than 3 au (as 67P/Churyumov-Gerasimenko returned from its aphelion passage at 5.68 au having spent about four years at a distance beyond 4 au). These COSIMA grains therefore come from a dusty layer that has built up over those four years, when the comet was so far from the Sun that the solar radiation was no longer able to create a gas drag that could efficiently remove the dust. The dust therefore remained on the surface, building up an ice-free, fluffy layer, below which lies an ice–dust mixture. When the comet returned to regions of higher solar irradiation the evaporation rate of the volatile gases underneath the dust layer increased again, lifting the particles from the dry upper dust layer into the inner coma, and leaving their original dusty cohabitants (dust frozen together with the gas) behind. This left-behind dust replenishes the existing dusty layer from below, thereby maintaining its thickness in a quasi-steady state until the solar radiation is high enough that the amount of dust removed from the upper layer is larger than the new volatile-free dust produced underneath. As a consequence, the dusty layer will disappear over time and fresh material will come to the surface. The transition may be gradual but could be violent if there is a hard zone under the dusty layer (as may be indicated by the re-bounce of the Philae lander) below which high gas pressures are building up. From the increase in dust production rate observed telescopically in 2008 (ref. 14) we infer that the dusty layer was lost at some stage between 2.7 au and 2.5 au. That orbital section will be reached again during the present recurrence of the comet between 24 December 2014 and 20 January 2015, so the loss of the dusty layer has probably already occurred.

The mass spectra of the surface of the COSIMA grains collected beyond 3 au show a high abundance of sodium. Preliminary values obtained after calibration26 are as high as 0.8, normalized to Mg = 1. For comparison, the Na abundances (Mg = 1) for comet 81P/Wild-2 are 0.13 (collected in aerogel) and 0.2 (collected on aluminium foil)27, 0.1 ± 0.06 for comet 1P/Halley28, and 0.055 for CI chondrites29. The Na abundance observed in Perseid and Leonid meteoroids is a factor of 1.5 higher than the chondritic value15, which fits very well with the value measured by COSIMA. Furthermore, the fluffiness of the COSIMA grains suggests that they would fragment with time after release into the coma. From remote observations, such fragmentation of coma grains has regularly been proposed30. Therefore we conclude that the high Na abundance measured by COSIMA, combined with the fluffiness of the grains, supports the hypothesis that these grains represent the parent population of interplanetary dust particles in meteor streams of cometary origin.

Beyond 3 au, COSIMA has not collected any of the dust that is mixed with sublimating ice, but rather the dust that is present in the upper ice-free dust layer. When the comet loses its fluffy mantle, it is expected that the properties of the grains collected will be very different from those of the grains currently under analysis, which show the properties of ‘space-weathered’ comet refractory material. The fresh material is likely to be a mixture of ice and dust, and its analysis should provide the detailed structure of this mixture. However, when the comet returns to the outer Solar System, a new dusty mantle will form as the upper layer again becomes free of ice. The formation of such a mantle was considered for re-occurring comets8 and detailed models exist for short-period comet nuclei9,10. The physical processes and timescales of these models are consistent with assumptions made about the nucleus size, orbit and so on for 67P/Churyumov-Gerasimenko. Therefore, the grains collected from this comet provide direct evidence for the existence of its dusty mantle and also an indication of the structure of dust mantles in short-period comets.