This post was written by Prof Jamie Gilmour
Around 4.5 billion years ago the sun formed when part of a huge cloud of gas and dust began to collapse. If you could go back in time at look at what was happening, you’d see a disk of gas and dust – the solar nebula – feeding material inwards to the centre where the sun was growing. Within the disk, dusk grains accumulated to form larger bodies. There came a time when there were millions of asteroid-sized (30-50 km diameter) bodies that we call “planetesimals” in the region where the Earth and other terrestrial planets would eventually grow. These would mostly go on to build even larger bodies, finally resulting in the terrestrial planets we are familiar with. A small proportion of planetesimals avoided this fate and became the ancestors of present-day asteroids.
The material that arrives at Earth bears witness to this planetesimal stage of solar system formation. Meteorites mostly come from asteroids, the descendants of planetesimals. Some of these, chondrites, are assembled mixtures of silicate and iron-nickel grains that used to exist independently. Others are massive chunks of solid iron-nickel, or resemble volcanic rocks that crystallized from a molten magma. We can roughly put these in a sequence of planetesimal evolution – the first planetesimals consisted of assembled grains. Some of these got so hot that they melted, at which point the dense molten mixture of iron and nickel sank to the centre of the planetesimal to form a core similar to the Earth’s, and the less dense rocky silicate melt floated to the top to form a mantle. Planetesimals were later disrupted by collisions or ripped apart by planetary tides, so today’s asteroid belt contains fragments of planetesimal cores, fragments of mantles, and parts of other planetesimals that never got hot enough to melt. All of these are contributing to the population of meteorites.
When I first started working on meteorites in the 1980s, one big debate centred on why planetesimals had got hot (even the chondrites show evidence of heating, though they didn’t get hot enough to melt). One option was heat from decay of radioactive aluminium. The problem was whether there was enough. The earliest grains that actually formed in our solar system contain enough radioactive aluminium to heat the material they are contained in through 5000 °C, but this only works if the heat can be trapped while the aluminium decays over around 2 million years. Trapping the heat over this timescale requires a body of 10 km or more – a planetesimal. So for this idea to work, planetesimals had to grow in the first 2 million years after the formation of the first solid grains in our solar system, which seemed a tall order. Most of the chondrites record too little radioactive aluminium to lead to melting.
The solution to this conundrum became apparent once new techniques developed that allowed us to date the melting process experienced by the planetesimal sources of iron meteorites. Although they represent a later stage in planetesimal evolution than chondritic material, it turns out that they were processed earlier in solar system history. They represent what happens to a planetesimal when it forms early on, before the radioactive aluminium has decayed away. This changed our ideas about how long it takes planetesimals to form, but the heat source had been identified. A form of selection bias is at work – planetesimals that form with a lot of radioactive aluminium melt, those with less remain chondritic; this is why the chondritic samples record too little radioactive aluminium to power melting.
But maybe another form of selection bias is at work. In 2009 Ceri Middleton and I wrote a paper suggesting that the heating of planetesimals fundamentally changed the composition of the material that went on to form the planets; even material that started out with a moderate amount of water would “dry out” as it was heated by the decay of radioactive aluminium. We argued that the presence of this radioactive aluminium had dried out the planetesimals and led to our planet having much less water than it would have had otherwise. The Earth seems to be finely balanced, with many factors combining in a configuration that allows a habitable environment to persist over geological time, a balance that arguably depends on the right abundance of water. In addition, building a technological civilization, we thought, would be much harder under several kilometres of ocean. In effect, our presence on the planet selects an initial abundance of water, which in turn selects a planetary system that formed with at least as much 26Al as ours had.
In this scenario, we can understand why there was a debate about the heat source for processing of planetesimals. Models suggest that most solar systems form with no radioactive aluminium – it has to be made in a massive star near enough that it can get into our forming solar system before it decays. Of those that do incorporate radioactive aluminium, larger amounts are less likely than smaller amounts. Now factor in a requirement that there be enough radioactive aluminium present to dry out planetesimals and allow a planet like ours to form. The most likely observation for a civilization like ours is a solar system with just enough to do the job; more radioactive solar systems are even more rare. In a sense we can “explain” the initial controversy about whether there was enough radioactive aluminium as reflecting the most likely situation for a civilization like ours
Since we published our paper, the field has developed. Detailed models suggest that most planetary systems form with much less
26Al than ours – around 85% get none at all. Now, Tim Lichtenberg has looked in detail at how the compositions of terrestrial planets are controlled by the presence of 26Al in their parent system. It turns out that we should expect most terrestrial planets like the Earth to be ocean worlds, with only a small proportion ending up relatively water poor like the Earth itself. Even more exciting, it is predicted that we might be able to test this idea by measuring the masses and radii of planets orbiting other stars – water rich planets can be as much as 10% larger than water-poor planets of the same mass.
Lichtenberg et al. (2019) A water budget dichotomy of rocky protoplanets from
26Al-heating, Nature Astronomy, 2, pg 307-313. doi: