During their formation, many rocky planets go through a phase known as a magma ocean, during which they are mostly or completely molten. Many researchers thought that the solidification of Mars’s magma ocean was protracted, perhaps lasting for up to 100 million years (Myr) after the ocean’s formation1–3. But in a paper in Nature, Bouvier et al.4 show that this process was completed in less than 10 Myr. The finding suggests that habitable conditions existed on Mars up to 100 Myr before they did on Earth.
In the contest between scientific models, empirical evidence is the arbiter. More than 100 meteorites that originated on Mars have been identified on Earth, providing samples of the Martian crust. And advances in the sensitivity of instruments that measure the concentrations of individual isotopes allow the ages of these materials to be determined with high precision.
Bouvier and colleagues looked for minerals known as zircons in Martian meteoritic material. When a zircon crystallizes from its parent magma, its crystal structure allows stray uranium atoms to be trapped in the growing crystal, but rejects lead atoms. Consequently, when researchers study these minerals billions of years later, they can be confident that any lead in the crystals was produced by uranium decay and that no other sources of lead need to be considered. Furthermore, two uranium–lead decay processes (235U to 207Pb and 238U to 206Pb) can be used simultaneously to improve the precision of the results. Uranium–lead geochronology using zircons therefore yields the most precise ages of ancient geological materials that are currently possible.
The authors analysed seven hard-earned zircons and obtained ages ranging from 4,476 to 4,430 Myr. For comparison, the first solids in the gas disk around the growing young Sun, known as calcium–aluminium-rich inclusions (CAIs), formed 4,567.3 Myr ago5. Therefore, in the astonishingly short interval of 90 Myr, Mars grew from dust to a planet, solidified from its initial magma-ocean state and formed a crust containing zircons.
This result already shows that models predicting a protracted magma-ocean stage on Mars1–3 cannot be correct, but Bouvier and co-workers’ study yielded even finer constraints. The lutetium–hafnium decay process, 176Lu to 176Hf, can be used to constrain the melting history of the zircons’ parent magmas, because the two isotopes behave differently during melting. The authors found that the zircons have unusually low concentrations of 176Hf. This indicates that the parent magmas had lower amounts of 176Lu than would be expected if they originated from the solidified products of Mars’s magma ocean. To form these parent magmas, the planet must have partially melted after it had solidified.
Bouvier and colleagues’ findings provide a revised timeline for the early stages of Mars’s history (Fig. 1). The planet grew to approximately its current size within less than 10 Myr (and probably less than 5 Myr) of the formation of CAIs6,7. It then took less than 10 Myr to solidify from its initial magma-ocean phase. To put these timescales into perspective, if the Solar System were one day old, Mars would have fully formed in the first 6 minutes. About 20 Myr after the formation of CAIs, the planet partially melted to produce magmas that rose to the planet’s surface; 70 Myr later, these magmas had solidified to form a zircon-containing crust.
The rapid solidification of Mars’s magma ocean has important implications for our understanding of both Mars and the planet’s formation of rocky planets in general. The speediness suggests that heat was easily lost from Mars, which implies that the planet’s atmosphere was relatively thin. Two processes could have produced such an atmosphere: a low release of volatile gases from the magma ocean; and a stripping of the atmosphere by the young Sun8,9. Researchers can now constrain the extent to which such processes occur much more closely, and can apply the results to the young Earth.
The early growth and magma-ocean phase of Mars, and, by extension, of other planetary embryos, means that at least some of the planet’s formation probably happened while the gas disk was still present around the young Sun — on average, such disks exist for only a few million years10. Therefore, there is strong reason to think that gas in the disk would have diffused into the magma oceans on these embryos.
This diffusion process could help to answer some long-standing questions about, for example, the noble-gas content of Earth. Today, Earth releases noble gases that must have been implanted in the mantle at the time of the planet’s formation. The origin of these gases has been unclear because the rocky material that built Earth contained only a small quantity of noble gases. The diffusion of noble gases from the gas disk directly into the magma ocean might solve the mystery.
Finally, Bouvier and co-workers’ timeline allows the early histories of Earth and Mars to be compared directly. About 100 Myr after the formation of CAIs, Earth went through a magma-ocean phase that is thought to have been initiated by the collision of the planet with a Mars-sized body — a collision that led to the formation of the Moon11. Consequently, the authors’ results suggest that Mars had clement conditions, and was possibly even hospitable to the formation of life, for as long as 100 Myr before such conditions existed on Earth. Mars had a head start on Earth in the planetary-evolution game.
Nature 558, 522-523 (2018)