When and why does CO2 mineralization in basalts stall? A critical review of microscale passivation, pore-clogging mechanisms and the multiscale imaging needed to resolve them

Tagir Karamov , Alina Karamova

In-situ CO2 mineralization in basalts is among the most secure forms of geological carbon storage. Field pilots report rapid mineralization and, where monitored, stable injectivity, yet many laboratory experiments show that carbonation slows or stops long before the reactive cation inventory of the rock is consumed. We synthesise four field pilots (CarbFix1, CarbFix2, Wallula, Jizan) and 37 laboratory studies (36 experimental, one modelling) spanning 5–200 °C and 0.1–30 MPa to identify why, and to ask whether these limits apply at reservoir scale. Three microscale mechanisms are distinguished: (i) chemical isolation of olivine, glass and pyroxene by nanometre-thick Si-rich amorphous layers, which lower dissolution rates by two to more than four orders of magnitude; (ii) precipitation of smectites, zeolites and serpentine-type phases that coat pore walls, bridge micro-cracks and compete with carbonates for divalent cations; and (iii) clogging of pore throats and micro-fractures by carbonates; if all products are retained in place, carbonation increases the solid volume by 35–90%. Where both were measured, permeability and porosity were decoupled: in five of seven experiments permeability fell by up to one to two orders of magnitude for porosity losses of less than 0.2–4 percentage points — 3–60 times more than Kozeny–Carman predicts for an initial porosity of 10% — whereas in the other two permeability increased despite secondary precipitation. A Pilling–Bedworth-type volume criterion predicts that leached Si residues and calcite crusts on plagioclase are porous, so that the dense passivating layers observed on olivine require interfacial silica reprecipitation; pyrite armour proved temporary. Transport regime (Péclet and Damköhler numbers) and pore connectivity, rather than bulk porosity, decide whether reaction proceeds, passivates or clogs. Only 3 of the 37 studies imaged the reacting interface at the nanometre scale. We therefore propose a correlative workflow in which time-lapse micro-CT locates clogging and site-specific FIB-SEM/TEM resolves passivating interfaces, and formulate a research agenda for parameterising reactive-transport models of CO2 mineralization.