Chondrules, the stony, seed-like grains in meteorites, were formed when some event melted rock in the solar nebula. The latest analyses narrow the possible ‘when’, ‘where’ and ‘how’ of that process.
Open up almost any stony meteorite, as scientists have been doing for more than 200 years1, and you will find hundreds of millimetre-sized bits of rock. These ‘chondrules’ (named after the Greek for seeds) were formed at the birth of the Solar System, and as such potentially bear witness to conditions — pressures, temperatures, chemical composition and so on — in the solar nebula. But obtaining that information depends on identifying how they formed, a topic tackled by Cuzzi and Alexander on page 483 of this issue2.
With the invention of the petrographic microscope in the late nineteenth century, it was recognized that chondrules are silicate rocks with igneous textures, and that these “drops of fiery rain”3 individually crystallized from a molten state while floating freely. Clearly, at the birth of the Solar System there were events hot enough to melt rock, and frequent enough to melt most of the mass of the asteroids — asteroids, which reside between Mars and Jupiter today, are the source of most meteorites. Identifying what melted the chondrules has long been a central theme of meteoritics. Dozens of explanations have been advanced, but detailed, quantitative tests of these models have come of age only recently.
Cuzzi and Alexander's work2 represents a major advance in constraining the circumstances of chondrule formation. It shows that not only were the chondrule-melting events very energetic, but they were also very large — thousands of kilometres — in extent. This finding limits the number of possible mechanisms by which chondrules formed, and goes some way to demystifying their origins.
Chondrules are known to have reached peak temperatures in excess of 1,800 K, and to have remained partially molten for a matter of hours4. This degree of heating not only melts the silicate rock, but can also cause relatively volatile elements such as potassium, iron, silicon and magnesium to evaporate. Indeed, these elements are relatively depleted in chondrules compared with the average in meteorites. Because light isotopes evaporate more readily, one would expect the Si in chondrules, for example, to be relatively depleted in light isotopes such as 28Si compared with heavy isotopes such as 30Si. In fact, no such isotopic fractionation is observed5.
As Cuzzi and Alexander2 discuss, the most plausible explanation for this discrepancy is that the K, Fe, Si and Mg vapour never left the region in which the chondrules formed. The chondrules would then reequilibrate with the vapour of nearby chondrules, and the light isotopes that most readily evaporate from chondrules would just as quickly recondense onto them. Based on the upper limit on the isotopic fractionation, and using an elegantly simple analytic expression confirmed by numerical chemical kinetic modelling, Cuzzi and Alexander estimate that the vapour pressures of the volatile elements must have reached at least 95% of their saturation pressures for the duration of the chondrule-melting event. This constraint sets a lower limit on the density of chondrules floating in the chondrule-formation region of about 10 per cubic metre.
An additional constraint arises from the need to keep the rock vapour from diffusing away from the chondrules, which sets a lower limit on the size of the chondrule-forming region. In the region where meteorites are thought to have formed (the location of the present-day asteroid belt), gas pressures probably were about 10−4 bar (ref. 6). In that case, a chondrule would have to be surrounded by other evaporating chondrules for about 3 km in each direction to keep the rock vapour pressure close to saturation. For the majority (99%) of chondrules to escape isotopic fractionation, the chondrule-forming region would have to be about 300 times larger still — that is, about 1,000 km in extent.
These new estimates severely restrict the ‘when’, ‘where’ and ‘how’ of chondrule melting. Previous studies of how often chondrules collided with and stuck to each other led to estimates of 1–10 chondrules per cubic metre in the chondrule-forming region7, and it is significant that these new constraints from isotopic fractionation confirm this high density. Chondrule densities of as much as 10 per cubic metre are unexpectedly high and can exist only where the gas densities are highest, near the midplane of the Solar System's proto-planetary disk. Even so, chondrules must be concentrated by factors of several hundred compared with their average throughout the disk, so that they outweigh the gas locally. This argues against heating mechanisms acting far from the midplane, such as shocks driven by X-ray flares8 or clumpy accretion9. The constraint on the size of the chondrule-formation region, 1,000 km or more, is fundamentally new: it constitutes evidence against such proposed heating mechanisms as nebular lightning10, and perhaps even shock waves driven by large asteroids, several hundred kilometres in diameter, orbiting through the solar nebula gas11.
All in all, Cuzzi and Alexander2 have notably advanced our understanding of chondrule formation. On the evidence of their analyses, a favoured explanation is that the melting mechanism for chondrule creation was shock waves, on a scale of the whole solar nebula12,13,14. These shock waves were probably driven by gravitational instabilities that arose when the gravitational forces of the solar nebula gas on itself were comparable to the gravitational pull of the Sun. These instabilities probably manifested themselves as spiral density waves akin to the spiral arms in our Galaxy. Further modelling, in concert with petrological and isotopic analyses of chondrules, will test this idea. Despite their size, smaller than seeds of grain, chondrules may bear witness to events that occurred on a stage as large as the Solar System itself.