From numerical expectations to experimental observations: Unraveling the complexities in CO2-basalt interactions and mineral growth in porous media

Subsurface fluid flow and solute transport are pivotal in addressing pressing energy, environmental, and societal challenges, such as the geological storage of carbon dioxide (CO2). Basaltic rocks have emerged as highly suitable geological substrates for injecting large volumes of CO2 with emission reduction and carbon mineralization purposes. This preference is attributed to their widespread occurrence at Earth's surface, high concentrations of cation-rich silicate minerals, reported fast mineralization rate, and often favorable characteristics such as porosity, permeability, and injectivity. The mineralization process within basaltic rocks is intricately linked, involving the dissolution of silicate minerals and the subsequent precipitation of carbonate minerals. During this chemical interplay, silicates play a crucial role by contributing vital calcium (Ca), magnesium (Mg), and iron (Fe) ions essential for the precipitation of carbonate minerals, including Ca-, Mg-, and Fe-carbonates. Understanding the consequences of mineral nucleation and growth in porous media and the fate of subsurface flow and transport necessitates spatial and temporal knowledge of solid precipitation locations and amounts. Only then can the reactive transport models provide precise and realistic predictions on the intricate interplay between transport mechanisms and reaction kinetics and, therefore, advection-diffusion-reaction (ADR). However, accurately representing the dynamics and dimensionality of mineral nucleation and growth in porous media is still challenging. There is a continued need for theoretical development to precisely predict the occurrences where ADR coupling occurs in the space and time domains. We conducted an integrated investigation involving experimental and numerical approaches to gain deeper insights into the spatial distribution of secondary mineral growth. Laboratory experiments under static (batch reactors) and dynamic (columnar flow cells) conditions at elevated pressure and temperature explored variations in aqueous solutes, pH, thermodynamic conditions, and residence time. Despite numerical predictions suggesting the formation of MgFeCa-carbonates in CO2-basalt interactions at higher temperatures, our laboratory findings primarily indicated the growth of calcium carbonates, namely calcite and aragonite. The lack of MgFeCa-carbonates, such as ankerite, siderite, and magnesite, remains elusive but is presumed to be associated with the concurrent occurrence of clay fractions, particularly smectites, consistently observed in batch-type experiments. Columnar flow experiments revealed the spontaneous formation of a limited number of large crystals at various locations, rationalized by the overarching influence of probabilistic mineral nucleation. This underscores the need for a new probabilistic approach to accurately model kinetics and crystal growth distribution in numerical simulations, where dynamic ADR may steer geochemical reactions towards favorable or unfavorable regions in terms of carbon mineralization efficiency.