Deep mantle water cycle

An ocean-floor rock can sink with a plate, carrying water bound inside mineral crystals. How much stays in the sinking rock, passes into other minerals, or is released before reaching the deeper mantle?[1] [2]

A sinking rock carries hydrogen within mineral structures, rather than transporting a buried liquid ocean. How much remains bound as pressure, temperature and mineral composition change?

In 2026 experiments, warming mineral mixtures transferred water from hydrous minerals into two neighboring high-pressure forms of olivine, wadsleyite and ringwoodite. That transfer need not create a free fluid. Other experiments found composition-dependent silica-bearing assemblages that could store water deeper in subducted crust. The examples establish possible carriers and exchanges, rather than a measured global inventory.[1] [2]

Experiments reveal possible carriers and transfers. Their water storage capacity does not tell us how much water real slabs contain or transport worldwide.[1] [2]

Real slabs follow different temperature and composition histories. Knowing a phase can store water does not establish how abundant it is, how full its storage sites are, or how much rock crosses a depth over time. Those missing quantities separate laboratory water storage capacity from the actual water flux and accumulated deep inventory.[1] [2]

It helps explain how plate recycling redistributes Earth’s water.[1]

Mineral-bound water can influence melting and the ease of rock deformation. Following its redistribution links the surface history of sinking plates to the physical behavior of the mantle over geological time. A carrier found in a laboratory matters because it provides a route for that connection; its capacity alone does not establish the route’s global importance.[1]