Dry metastable olivine and slab deformation in a wet subducting slab

While the plates carry water to the Earth's interior, phase transitions of dry olivine, the main mineral in the plates, are thought to be responsible for deep-focus earthquakes and plate deformation. This study resolves the contradiction of the presence of dry olivine even in wet plates. Takayuki Ishii, a researcher at the Center for High Pressure Science & Technology Advanced Research (HPSTAR), China and the Bavarian Institute of Geosciences, University of Bayreuth, Germany, and Eiji Otani, a professor emeritus at Tohoku University, used high-pressure and high-temperature experiments to determine the water content of olivine under the conditions of a subducting plate containing water. The results show that the hydrous minerals absorb water, while the coexisting olivine contains no water at all. This experimental result overturned previous theories on the role of hydrous minerals, and revealed that deep-focus earthquakes can occur even in a wet plate, and that large plate deformation can also occur.The results of this research will be published Nature Geoscience.

The most widely accepted explanation for deep-focus earthquakes is that they are caused by the delayed transformation kinetics of dry olivine, which would seem to require that subducting slabs are dry. Metastable persistence of olivine to great depths (~630 km) leads to a wedge of olivine in the slab, and seismic observations of such wedges add weight to the proposed mechanism. However, in direct opposition to the dry wedge hypothesis, water is circulating not only at the Earth's surface but also in the interior through subducting oceanic lithosphere: many geochemical and geophysical observations and mineral physics data indicate that 'water', in the form of hydroxyl groups, is present within both hydrous and nominally anhydrous minerals, implying that subducting slabs are hydrated. The presence of dry metastable olivine in wet subducting slabs is therefore paradoxical, and the hydration state of the slabs remains a topic of debate.

Most previous studies focused on maximum solubility of water in olivine coexisting with hydrous melts outside the conditions where deep-focus earthquakes happen, under water-saturated conditions. However, natural wet subducting slabs may consist of olivine and hydrous minerals under water-undersaturated condition.

Here, they determined the water contents of olivine coexisting with hydrous phase A, a major dense hydrous magnesium silicate produced by dehydration of serpentine, under water-undersaturated conditions. They show that olivine is dry even under wet conditions. Alternately, a minor mineral of dense hydrous silicate preferentially crystallises as a host of water, which is the opposite conclusion to previous understanding of the water partitioning: olivine preferentially accommodates water and hydrous minerals forms when excess water exists. This finding changes our knowledge of the role of water in the deep Earth, implying that hydrous minerals play more important roles on water cycle in the Earth's interior.

"We thus solve the paradox: dry olivine experiences the delay transformation even in hydrated slabs, causing deep-focus earthquakes. Furthermore, this result suggests that dehydration of hydrous minerals, which is usually considered for an origin of intermediate-deep earthquakes, may also cause deep-focus earthquake," said Dr. Ishii.

Their finding also newly suggests a novel hypothesis that mysterious phenomena of slab bending and stagnation in the deep interior are caused jointly by dehydration of hydrous silicates and the subsequent rapid phase transformation of olivine enhanced by their released water, due to hydrolytic weakening of olivine and its high-pressure polymorphs.

"These results suggest that hydrous minerals not only play an important role in transporting water to the Earth's interior, but are also responsible for the occurrence of deep earthquakes and large plate deformations," Dr. Ishii added. "Dehydration of hydrous minerals has been thought to be one of the causes of shallower earthquakes than deep earthquakes, but our results suggest that it can also be a cause of deep-focus earthquakes deeper than 660 km, which cannot be explained by phase transitions in metastable olivine. The results are expected to provide an important clue to a full understanding of plate behaviour, including the occurrence of deep earthquakes and large deformations in the deep mantle."