Earth was formed by numerous collisions with smaller and greater objects. Calculations from ETH researcher Alexander Halliday now show that the consequence of these collisions was not what has been assumed up until now.

Exactly how the Earth came into being is something that we will never be able to reconstruct via direct observation. The oldest stones found on our planet today are at least 500 million years younger than the Earth itself. Cosmologists measure their hypotheses against computer models when they want to simulate the genesis of Earth. Alexander Halliday, Professor at the ETH Institute of Isotope Geochemistry and Mineral Resources (1), recently published his deliberations in "Nature" that throw new light on Earth's early years (2). According to his calculations, current models – at least in certain points – have to be reconsidered.

Traces of the past

Generally speaking, one assumes nowadays that the inner planets – i.e., Mercury, Venus, Earth and Mars – created themselves out of the surplus material of the newly evolved Sun. In a spinning disc of dust and gas, material congregated to form large lumps, which, in their turn collided with each other to build even larger bodies. Little by little, the new planets came into being. It is, however, not clear how long this process took. Most current models indicate that Earth took between 10 and 100 million years to form.

Isotope dating of the rocks now makes is possible to pinpoint this process more accurately. If one knows the underlying mechanism, one can calculate, for example using the composition of the lead isotopes in rocks, how quickly Earth was formed. The composition of the lead that can be measured today was considerably influenced by processes happening at the time. Similarly, analyses of tungsten isotopes can help to determine the length of time it took Earth to evolve.

An inhomogeneous mass

However, the respective "clocks" of lead and tungsten give different ages for the beginnings of Earth. Evidently, says Halliday, a basic presumption in the models is wrong. Until now it has been presumed that objects that collided with Earth first bound themselves to its mantle and that the metal core of the Earth was formed in a second phase. Both clocks were set running at the occurrence of the event and should, therefore, deliver identical results regarding the age of the Earth.

Halliday now postulates that the consequences of the collisions were otherwise. There was no complete union with the Earth's mantle; rather the metal cores of the colliding objects smelted with the Earth's own metal core more or less straight away, owing to the higher density of metal. If one assumes that this hypothesis of the collision mechanism is correct, then the clocks agree with one another. According to this theory the Earth created itself over a period of 50 million years.

Our moon came into being following a collision between the planet Theia and the Earth. Isotope dating now shows that the composition of Theia was similar to that of Mars. (Picture Nasa) large

Costly collision

Halliday presents new results in his work of regarding another aspect. In the end phase of its formation, Earth probably collided with a big object called Theia. It is generally acknowledged today that this gigantic collision led to the formation of our moon (3). Earlier simulations indicate that this body is mainly composed of material from the planet Theia. The question now is, was the composition of Theia the same as that of the moon?

Halliday is convinced that it wasn't. If one compares rocks from the Earth, the moon and Mars one finds some quite striking differences. Samples from the moon exhibit far fewer volatile elements – such as rubidium and potassium – than rocks on Earth. Conversely, rocks from Mars contain precisely these elements far more often than those of Earth.

Mars as a prototype

Isotope data now reveal that Theia had a similar composition to that of Mars. This would indicate that Earth and the moon underwent very radical chemical changes on collision; the impact was so strong that both bodies lost a major proportion of their volatile elements.

However, this puts the spotlight on another pressing question: why is there so much water on Earth? In principle, one would expect to find that Earth not only lost rubidium and potassium when it collided with Theia, but also water. "The really exciting question is not whether there's water on Mars or not," says Halliday, "but why there is so much of it on Earth" (4).