Understanding the carbon cycle via molecular dynamics simulations of carbonate solutions under extreme conditions in Earth's upper mantle.

Abstract

The deep Earth hosts the majority of Earth's carbon in geological reservoirs, while the deep carbon cycle governs the long-term exchange and transport of carbon between the surface and Earth's interior, thereby influencing long-term climate. In subduction zones, carbon is transported into the mantle as carbonate minerals and organic matter. As these materials descend into the transition zone, water acts as the dominant solvent, driving complex geochemical reactions under extreme conditions. Understanding carbon speciation under these conditions is essential for quantifying deep carbon reservoirs, yet direct experimental observations at mantle conditions remain extremely challenging. While ab initio molecular dynamics (AIMD) provides accurate predictions, it is computationally intensive for complex geological fluids. Here, we employ classical molecular dynamics with a reactive force field (CMD-ReaxFF), whose performance is validated against AIMD calculations to enable efficient simulations of bond-breaking and bond-forming processes in aqueous systems under high-pressure and high-temperature conditions relevant to the deep Earth. Simulations were conducted at 1000 K and 10 GPa to evaluate the influence of alkali and alkaline earth metal cations - specifically Na+, Mg2+, and Ca2+ - on carbonate speciation in aqueous solutions, representative of upper mantle fluids. The results show that aqueous CO2 is only a minor species under these conditions, challenging traditional thermodynamic models. Instead, bicarbonate ion (HCO3-) dominates, particularly in the presence of divalent cations. Notably, Mg2+ and Ca2+ promote the formation and stabilization of neutral carbonic acid (H2CO3), which may serve as a transient carbon carrier and facilitate vertical transport and degassing. These findings enhance our understanding of carbon speciation and transport in deep Earth systems and demonstrate CMD-ReaxFF to be an efficient and powerful tool for modeling geochemical processes relevant to the global carbon cycle.

Abstract as supplied by the indexing source. © the authors / publisher.