Billions of years ago, Earth was a red-hot sphere covered by an ocean of liquid magma. Over time it cooled and crystallized, forming the mantle, the layer between the crust and the core. Today, deep flows of material, mantle plumes, rise to the surface and cause eruptions. Hydrogen from the planet's deepest interior is sometimes found in such lava.
Hydrogen exists in two stable forms: ordinary light hydrogen and heavier deuterium. In most volcanic rocks their ratio is roughly the same. But in lavas associated with deep plumes, an unusually low share of deuterium is sometimes found, for example on Baffin Island in the Canadian Arctic and in Iceland.
This rare composition has sparked a scientific debate. One hypothesis explains it by the freezing of the ancient magma ocean. Another suggests that low-deuterium hydrogen seeps from Earth's metallic core.
To test the versions, a group led by Wenzhong Wang of the University of Science and Technology of China carried out computer modeling of conditions at depths of thousands of kilometers. The scientists studied how the two forms of hydrogen behave during cooling and crystallization of melt into different minerals, and then modeled the distribution of isotopes in the early mantle.
The results, published in the journal Science Advances, showed that as the magma ocean solidified, the two forms of hydrogen were distributed through the mantle relatively evenly. There were small differences at different depths, but they did not reach the anomalously low deuterium levels observed in volcanic rocks. "Our results demonstrate that crystallization of the magma ocean does not lead to significant fractionation of hydrogen isotopes in Earth's mantle," the authors write.
This means the solidification of the magma ocean cannot explain the unusual signatures. The researchers therefore lean toward the core-leak version: low-deuterium hydrogen probably seeped from the metallic core into the mantle over billions of years.
There is another clue. Some of these rocks have an unusually high ratio of helium-3 to helium-4 isotopes. Helium-3 is a rare form of helium that could have been preserved in the core since the earliest times.
Although the data strengthen the arguments for hydrogen leaking from the core, the exact mechanism for transferring material from the core to the mantle remains unclear. "Elucidating the mechanisms by which core signatures are transferred to the mantle remains an important task for future research," the authors note.
It is important to emphasize: the study is based on computer modeling, not on direct observations. It does not prove a causal link, but only shows that the observed isotope ratios are better consistent with the core-leak hypothesis than with the alternative.