Models indicate that there are strong gradients in element concentrations and in the pH of fluids at the slab–mantle interface — a major discontinuity deep within Earth. This transforms our view of global geochemical transport. See Letter p.420
The chemical and isotopic composition of Earth's rocks and minerals carries a fingerprint of how and where they formed. This information provides a unique record of the spatial and time scales of past planetary processes and the fluxes of the chemical elements involved. Interpreting this record requires knowledge of the chemical conditions generated when aqueous fluids and minerals interact at extreme depths, but such knowledge has been hard to acquire. On page 420, Galvez et al.1 report a method for computing the composition and chemical speciation of aqueous fluids in equilibrium with different rock types, and a predictive model for mass transport across the top of Earth's subduction zones — the regions at which portions of a tectonic plate are being forced underneath another plate.
Mass transport on our planet is driven by the motions of plates in Earth's outer shell (the lithosphere), and rocks formed at Earth's surface can be recycled to depths of up to 200 kilometres. As temperature and pressure rise with depth, the rocks undergo complex mineral transformations, including the release of volatile constituents such as water and carbon dioxide. These fluids are highly mobile in rock environments because of their low density and high buoyancy (Fig. 1), and act as transport agents for diverse chemical elements.
When the fluids penetrate major discontinuities in the Earth, such as the boundary between subducting slabs of lithosphere and the upper mantle, they act as a medium for complex dissolution–precipitation reactions, which allow subducted minerals to attain chemical equilibrium in their new rock environment. These reactions strongly alter element and isotope concentrations in the fluid and surrounding rocks, and thus chemically differentiate Earth's materials, dictate element mobility and define global element fluxes.
Understanding the reactivity of aqueous and carbon-containing fluids with minerals and rocks in Earth's deep interior has been a long-standing challenge for laboratory research and dynamic simulations. However, experimental advances2,3 in the past few years have allowed mineral solubilities and the stabilities of chemical species at extreme temperatures and pressures to be measured with unprecedented accuracy. In parallel, atomistic simulations4,5 have delivered new calibrations of the electrostatic properties of water at such temperatures and pressures. These emerging data are essential for constructing sets of thermodynamic data for aqueous chemical species, which are used in simulations to predict fluid–mineral reactions and element transfer in natural systems. Such predictive simulations typically involve more than six chemical components, multiple mineral phases and tens of aqueous species in the fluid6,7.
“The authors realized that equilibria between mineral components and ionic species can be expressed using simple hydrolysis reactions.”
Nevertheless, obtaining robust information about the simultaneous exchange of multiple elements between minerals and fluids at Earth's extreme depths has not been possible — until now. Galvez and colleagues' computational method is based on the idea that the chemical potentials of major components of Earth's mineral assemblages (such as silica, alumina and sodium oxide) are the same regardless of whether they exist as aqueous species or as solids in rocks. Chemical potentials can be determined for all components within stable mineral assemblages, and dictate the chemical activity of each species and overall element concentrations in coexisting fluids at any pressure and temperature. The authors realized that equilibria between mineral components (oxides) and ionic species can be expressed using simple, independent hydrolysis reactions, and used this as the basis of their computational approach.
Galvez and co-workers' simulations reveal that, during subduction of Earth's oceanic crust, minerals containing hydroxyl or carbonate groups break down as the pressure and temperature increase, forming volatile-free solid phases and a free fluid. The fluid progressively dissolves the mineral phases, and thus acquires diverse cations, which make it moderately alkaline. Minerals in peridotite rock in the upper mantle surrounding the subducting slab also initially produce fluid that is alkaline, underpinned by the release of calcium cations from the mineral assemblage to the fluid. But when the temperature exceeds about 500 °C, the calcium cations become compatible with high-temperature minerals and therefore partition into the minerals from the fluid. The acidity of the aqueous fluid in the mantle thus increases, reaching weakly alkaline or near-neutral conditions. This phenomenon does not occur in the rocks of the subducting slab, and so a sharp acidity gradient is generated at the slab–mantle boundary.
But how does the difference in acidity affect fluid reactivity and chemical modification across this major geological interface? Galvez and colleagues' simulations show that infiltration of slab-derived fluid into the overlying mantle causes highly nonlinear effects: a rapid increase in fluid pH, followed by a sharp decrease. The infiltrated rocks respond in a highly nonlinear way, and may reach the limit of their ability to buffer or moderate the pH changes after coming into contact with a certain amount of fluid. This finding transforms our view of when the chemical composition of Earth's major geological domains is modified and where these processes are localized.
The authors extended their predictive approach to fluids of even greater complexity that contain appreciable concentrations of ligand species, such as carbonates or chlorine. This allowed the chemical cycles of volatiles (including CO2) and feedback relationships between ligand behaviour and metal mobility to be addressed.
The principal sources of carbonates at subduction zones are oceanic sediments and variably altered oceanic crust, whereas chlorine is derived from ocean water in the rock's pore and fracture space. At temperatures of between 200 and 700 °C, these ligands gradually form negatively charged species in fluids, and so concentrations of cations — including hydrogen ions, H+, which determine pH — in the fluid rise to balance the negative charge. Aqueous fluids associated with any slab rock type will therefore probably be weakly alkaline to almost neutral. Many dissolution–precipitation reactions are acid–base exchange processes (they involve the exchange of metal ions by H+), and so this relative increase in fluid acidity will probably promote the solubility and mobility of metals.
Earth's redox state changes between the upper continental crust and mantle by more than eight orders of magnitude8,9, and exerts additional controls on the distribution of species in aqueous fluids. For example, iron(III) species are much less stable and are less soluble in water than their iron(II) counterparts, and CO2 is more acidic than methane (CH4). The effects of these characteristics act in conjunction to lower the net charge in the deep fluid and promote an increase in acidity. This perturbation propagates through different chemical species, resulting in new shifts of chemical equilibria and feedback relationships. Incorporation of redox equilibria into Galvez and colleagues' model will therefore probably allow previously unknown chemical gradients or interfaces to be identified in Earth's dynamic subduction systems.Footnote 1
Galvez, M. E., Connolly, J. A. D. & Manning, C. E. Nature 539, 420–424 (2016).
Wilke, M. et al. Earth Planet. Sci. Lett. 349–350, 15–25 (2012).
Bernini, D., Wiedenbeck, M., Dolejš, D. & Keppler, H. Contrib. Mineral. Petrol. 165, 117–128 (2013).
Pan, D., Spanu, L., Harrison, B., Sverjensky, D. A. & Galli, G. Proc. Natl Acad. Sci. USA 110, 6646–6650 (2013).
Sverjensky, D. A., Harrison, B. & Azzolini, D. Geochim. Cosmochim. Acta 129, 125–145 (2014).
Dolejš, D. & Wagner, T. Geochim. Cosmochim. Acta 72, 526–553 (2008).
Galvez, M. E., Manning, C. E., Connolly, J. A. D. & Rumble, D. Earth Planet. Sci. Lett. 430, 486–498 (2015).
Frost, D. J. & McCammon, C. A. Annu. Rev. Earth Planet. Sci. 36, 389–420 (2008).
Evans, K. A., Elburg, M. A. & Kamenetsky, V. S. Geology 40, 783–786 (2012).