Giant crystals from the young Earth's magma ocean

The early Earth was, according to the standard model, a molten world: a magma ocean hundreds of kilometres deep, left over from the giant impacts that assembled the planet, cooling over tens of millions of years as the first solid crust crystallised from its surface. The story has always had a weak point. The deep interior of that ocean, under pressures so high that the minerals there do not exist at the surface, was understood only through laboratory experiments done at those pressures, and experiments that run for hours or days cannot easily answer a question that played out over millions of years. A paper in the 21 January issue of Nature, from a team led by Northwestern Polytechnical University, closes part of that gap with a striking result: in the deep magma ocean, the mineral bridgmanite can grow into crystals centimetres to metres across, giants by any geological standard.

Bridgmanite is the most abundant mineral in the Earth. Below a depth of roughly 660 kilometres, the lower mantle is composed largely of it, and its behaviour sets the pace at which the whole mantle convects, cools and moves. How it crystallised from the primordial magma ocean determines the initial structure of the lower mantle, which in turn shapes everything that has happened since: the style of early mantle convection, the timing of when the deep interior began to mix, even the way the first reservoirs of geochemical heterogeneity were established. The size of the crystals matters because size controls settling. Small crystals sink slowly and stay mixed with the melt; giants settle fast, dragging the growing solid layer downward and separating it from the remaining liquid in ways that small-grain models never predicted.

The team’s central result is that under the conditions of a deep magma ocean, bridgmanite growth is not limited to the tiny grains that laboratory crystallisation produces. Their work, combining experiments and modelling, shows that the kinetics of growth in the deep, hot, slowly cooling melt allow individual crystals to reach enormous sizes, and that this growth regime changes the physical behaviour of the crystallising layer: it becomes gravitationally unstable in a different way, with large crystals foundering and reorganising the deep mantle as it formed. The implication is that the lower mantle was not simply a uniform layer that crystallised from the bottom up, but a dynamic, differentiated system shaped by the giants growing within it.

The result matters beyond the early Earth. The same physics should apply to any rocky planet that passed through a magma ocean phase, and the model planets are now numerous: Mars, Venus, the young Moon, and the growing catalogues of exoplanets whose interiors are inferred from their radii and masses. If giant crystals formed in all of them, then the long-term evolution of a rocky world, its magnetism, its heat budget, its volcanic history, carries the fingerprint of how its deep mantle first solidified. The standard picture of a smoothly crystallising magma ocean is being replaced by something rougher and more interesting.

As with most planetary-scale results, the work is a model grounded in experiments rather than a direct observation of the Archaean mantle, which no instrument will ever see. The team is careful about the assumptions, and the numbers will be tested and revised. But the direction is clear: the deep crystallisation of the early Earth was not a uniform settling of small grains, and the giants that grew then may still be echoing through the mantle today.