The Oldest Crust and the Earliest Water on Earth

I’ve written a few posts now on the distribution of water on the Moon, the formation of the Earth-Moon system, and the origin of its volatiles. I've been wading through the literature for a few months now and I'm only now starting to pull it together. It's worthwhile, though, because one of the things that makes Earth special is that it's wet.

Water is intrinsic to all parts of the plate tectonic cycle: when the oceanic crust forms at mid-ocean ridges it incorporates water; when this same crust is recycled into the mantle in subduction zones, metamorphic reactions at depth cause the slab to give up water, and this causes the volcanism that creates continental crust. A trace of water deep in the Earth lowers the mantle’s viscosity enough that it can flow and convect, and this drives plate movement. Plate tectonics regulates Earth’s climate, replenishes its surface nutrient supply, and is probably responsible for Earth’s long-lived magnetic field, all of which are conducive to, if not necessary for, the evolution of complex life. Clearly, understanding how water gets into a planetary system is important in determining how Earth-like planets, and complex life, come to exist.

The oldest known zircon. The spot of the ion microprobe analysis is circled. From John Valley's website.

About eight years ago, there was a surprising discovery of just how early large amounts of liquid water appeared on Earth. The proto-Earth accreted by 4.55 billion years ago and the giant impact formed the moon about 4.5 billion years ago. Until recently the early stage of Earth’s history was thought to have been terribly inhospitable: largely molten and dry. It was named the Hadean eon for its presumed hell-like conditions. There is now very good evidence for surface water by 4.4 billion years ago. Although we don’t have rocks that old, we do have individual crystals of the mineral zircon (ZrSiO₄).

Zircon is a very robust mineral; it can survive in sedimentary, metamorphic, or igneous environments and remain relatively unchanged. The ancient zircons in question were found in metamorphic rocks of the Jack Hills of Western Australia. They originally crystallized in magmatic rocks that were later eroded, and many of the grains were deposited in river deltas. These sedimentary deposits turned to sedimentary rocks (conglomerates) and were later metamorphosed to become metaconglomerates. The zircons have ages as old as 4.4 billion years and are the oldest continental material yet found. Moreover, their oxygen isotope signatures indicate that their magmas formed by melting of material that had interacted with low-temperature water. Two seminal papers on these zircons are Mojzsis et al. (2001) and Peck et al. (2001).

Outcrop in the Jack Hills, western Australia, from which the oldest known zircons were collected. These minerals provide evidence for continental crust and liquid water by 4.4 billion years ago. From John Valley's website.

This means that several oceans worth of water accumulated on the Earth within 50-100 million years after the giant impact (if you agree with Albarède and his sources that support the late veneer model). The impact flux was much higher then – about one million to one billion times higher than now. The Moon is much smaller than the Earth and has less gravity, so it would have experienced fewer impacts and more of an impactor’s volatiles would escape upon collision with the Moon. It’s possible that one test of the late veneer model may be whether the Moon’s volatile budget meets these expectations; data from LCROSS, LRO, and future missions could help evaluate this. Of course, this is complicated by the later impacts which overprint the earliest history. (Post-late veneer impacts will be a topic of a future post). It’s not clear if it will be possible to unravel the earliest portions.