The Moon Was Formed in a Smashup Between Earth and a Near Twin

But solving one puzzle of lunar origins has raised another linked to the abundances of tungsten in the primordial bodies

When young planets collide. Hagai Perets

The moon was born in the collision of a Mars-sized body and the early Earth, but beyond that, much about the world we see in our skies every night is still a mystery. After 61 missions, including six astronaut visits that collected samples of moon rocks, many questions remain, including how much of the moon is made from that lost planet's leftovers, and how much was stolen from Earth? Answering these questions could offer fresh insights into the evolution of both celestial bodies.

Now, scientists in France and Israel have found evidence that the smaller body that smashed into the proto-Earth was likely made of similar stuff to our home world. Also, according to their computer models, the current composition of lunar material is best explained if whatever hit early Earth formed nearby. Two additional studies suggest that both bodies then built up a veneer of extra material as smaller protoplanets continued to bombard the young system, but Earth picked up much more of this later coating.

According to the "giant impact hypothesis," the moon formed about 4.5 billion years ago, when a planet-like object about a tenth of Earth's current mass slammed into our planet. Simulations and recent studies of moon rocks suggest that the moon should be mostly made from the remains of the impactor, nicknamed Theia. This would explain why the moon seems to be made of material that looks a lot like Earth's mantle, as seen in rock samples and mineral maps.

The problem is that planets tend to have distinct compositions. Mars, Mercury and big asteroids such as Vesta all have somewhat different ratios of various elements. If Theia was formed someplace else in the solar system, its makeup should have been rather different from Earth's, and the bulk composition of the moon shouldn't look so similar to Earth's mantle.

To try and solve the conundrum, Alessandra Mastrobuono-Battisti and Hagai Perets at the Israel Institute of Technology analyzed data from simulations of 40 artificial solar systems, applying more computer power than has been used in previous work. The model grew the known planets and a hypothetical number of planetesimals and then let them loose in a game of cosmic billiards.

The simulations assume that planets born farther from the sun tend to have higher relative abundances of oxygen isotopes, based on the observed chemical mix in Earth, the moon and Mars. That means any planetesimals that spawned close to Earth should have similar chemical traces. "If they are living in the same neighborhood, they will be made of roughly the same material," says Perets.

The team found that a lot of the time—20 to 40 percent—big impacts involved collisions between bodies that formed at similar distances from the sun and so had similar makeup. Described this week in Nature, the work backs up the intuitive idea that it's less likely something will sail in and hit you from afar, and it goes a long way toward explaining the moon's bulk composition.  

So far so good, but that doesn't explain everything. There's still a lingering puzzle linked to abundances of the element tungsten. This siderophile, or iron-loving, element should sink towards the cores of planets over time, making its abundance much more variable in different bodies even if they formed close together. That's because bodies of different sizes will form cores at different rates. While there would be a little mixing from the impact, most of Theia's tungsten-rich mantle material would have been flung into orbit and incorporated into the moon, so the amount of tungsten in Earth and the moon should be very different.

In two independent studies also appearing in Nature, Thomas Kruijer at the University of Münster in Germany and Mathieu Touboul at the University of Lyon in France examined the ratio of two tungsten isotopes—tungsten-184 and tungsten-182—in moon rocks and in Earth as a whole. The moon rocks have slightly more tungsten-182 than Earth, the teams report.

This is intriguing, because that particular isotope of tungsten comes from the radioactive decay of an isotope of the element hafnium. Its half-life is short, only about 9 million years. So while iron-loving tungsten tends to sink towards the core, the hafnium isotope stays closer to the surface and, over time, turns into tungsten-182. That leaves an excess of tungsten-182 in a planet's mantle versus the amount of tungsten-184 and other natural isotopes.

The difference between Earth and the moon is relatively small: the two studies find it at the level of 20 to 27 parts per million. But even that tiny shift would require a lot of chemical fine-tuning, says Kruijer, which makes it unlikely that it was just chance. "Varying the tungsten by only a percent or so has a dramatic effect," he says. "The only solution is if the mantle of proto-Earth had similar tungsten-182 content to Theia, and the core of the impactor directly merged with Earth's."

That's not likely, though. While much of Theia's core, being heavier than its mantle, will remain as part of the Earth, the mantle will mix with Earth's as it gets flung into orbit. More mixing happens as the moon accretes. The proportion of Theia's core and mantle material that gets turned into the moon is random chance, but there had to have been at least some core material, says Kruijer. Touboul's team came to a similar conclusion: If the differences in tungsten abundance were due to random mixing as Theia's innards were sloshing around with Earth's, the planet and the moon should be even more different than they are.

The simplest solution, the authors say, seems to be the "late veneer" hypothesis, which suggests that Earth and the proto-moon started with similar tungsten isotope ratios. Earth, being larger and more massive, would continue to attract more planetesimals after the impact, adding new material to the mantle. The veneer from those planetesimals would have had more tungsten-184 relative to tungsten-182, while the moon would have kept the ratio that dated from the impact.

"This looks like solid data," Fréderic Moynier, a cosmochemist and astrophysicist at the Institut de Physique du Globe de Paris, says via email. "It fits with the present theory of late veneer, which is simply based on the elemental abundance of the siderophile elements (among them tungsten): there is simply too many siderophile elements in the present Earth's mantle (they should all be in the core) and therefore they must have been brought to Earth after core formation via meteorite impacts."

One mystery remains: For the proto-moon to match Earth's tungsten ratio, Theia and Earth must have started with very similar tungsten abundances. Solving that puzzle will be the work of future planetary studies, but at least for now, the lunar origin story is starting to look a little clearer. 

Get the latest Science stories in your inbox.