Minerals are widely assumed to protect organic matter (OM) from degradation in the environment, promoting the persistence of carbon in soil and sediments. In this Review, we describe the mechanisms and processes operating at the mineral-organic interface as they relate to OM transformation dynamics. A broad set of interactions occur, with minerals adsorbing organic compounds to their surfaces and/or acting as catalysts for organic reactions. Minerals can serve as redox partners for OM through direct electron transfer or by generating reactive oxygen species, which then oxidize OM. Finally, the compartmentalization of soil and sediment by minerals creates unique microsites that host diverse microbial communities. Acknowledgement of this multiplicity of interactions suggests the general assumption that the mineral matrix provides a protective function for organic matter is overly simplistic. Future work must recognize adsorption as a condition for further reactions instead of as a final destination for organic adsorbates, and should consider the spatial and functional complexity that is characteristic of the environments where mineral-OM interactions are observed.

We present molecular dynamics (MD) simulations of water-filled silica nanopores such as those that occur in ordered oxide ceramics (MCM-41, SBA-15), controlled pore glasses (such as Vycor glass), mesoporous silica, bioglasses, and hydrous silica gel coatings of weathered minerals and glasses. Our simulations overlap the range of pore diameters (1 to 4 nm) where confinement causes the disappearance of bulk-liquid-like water. In ≥ 2 nm diameter pores, the silica surface carries three statistical monolayers of density-layered water, interfacial water structure is independent of confinement or surface curvature, and bulk-liquid-like water exists at the center of the pore (this last finding contradicts assumptions used in most previous neutron diffraction studies and in several MD simulation studies of silica nanopores). In 1 nm diameter pores, bulk-liquidlike water does not exist and the structural properties of interfacial water are influenced by confinement. Predicted water diffusion coefficients in 1 to 4 nm diameter pores agree with quasi-elastic neutron scattering (QENS) data and are roughly consistent with a very simple "core-shell" conceptual model whereupon the first statistical water monolayer is immobile and the rest of the pore water diffuses as rapidly as bulk liquid water.

We report new molecular dynamics results elucidating the structure of the electrical double layer (EDL) on smectite surfaces contacting mixed NaCl-CaCl 2 electrolyte solutions in the range of concentrations relevant to pore waters in geologic repositories for CO 2 or high-level radioactive waste (0.34 to 1.83 mol c dm -3 ). Our results confirm the existence of three distinct ion adsorption planes (0-, β-, and d-planes), often assumed in EDL models, but with two important qualifications: (1) the location of the β-and dplanes are independent of ionic strength or ion type and (2) "indifferent electrolyte" ions can occupy all three planes. Charge inversion occurred in the diffuse ion swarm because of the affinity of the clay surface for CaCl + ion pairs. Therefore, at concentrations ≥ 0.34 mol c dm -3 , properties arising from long-range electrostatics at interfaces (electrophoresis, electro-osmosis, co-ion exclusion, colloidal aggregation) will not be correctly predicted by most EDL models. Co-ion exclusion, typically neglected by surface speciation models, balanced a large part of the clay mineral structural charge in the more concentrated solutions. Water molecules and ions diffused relatively rapidly even in the first statistical water monolayer, contradicting reports of rigid "ice-like" structures for water on clay mineral surfaces.

The diffusion coefficients (D) of water and solutes in nanoporous Na-smectite clay barriers have been widely studied because of their importance in high-level radioactive waste (HLRW) management and in the isolation of contaminated sites. However, few measurements have been carried out at the high temperatures that are expected to occur in HLRW repositories. We address this knowledge gap by using molecular dynamics (MD) simulations to predict the temperature dependence of diffusion in clay interlayer nanopores, expressed as a pore scale activation energy of diffusion (E a ). Our sensitivity analysis shows that accurate prediction of pore scale D and E a values requires careful consideration of the influence of pore size, simulation cell size, and clay structure flexibility on MD simulation results. We find that predicted D values in clay interlayer nanopores are insensitive to the size of the simulation cell (contrary to the behavior observed in simulation of bulk liquid water) but sensitive to the vibrational motions of clay atoms (particularly in the smallest pores investigated here, the one-, two-, and three-layer hydrates). Our predicted D and E a values are consistent with experimental data. They reveal, for both water and Na + , that E a increases by ∼6 kJ mol −1 with increasing confinement, when going from bulk liquid water to the one-layer hydrate of Na-montmorillonite.

International audienceThe screening of surface charge by dissolved ions at solid–liquid interfaces—in the region of interfacial fluid known as the electrical double layer (EDL)—plays a recurrent role in surface science, from ion adsorption to colloidal mechanics to the transport properties of nanoporous media. A persistent unknown in theories of EDL-related phenomena is the structure of the Stern layer, the near-surface portion of the EDL where water molecules and adsorbed ions form specific short-range interactions with surface atoms. Here, we describe a set of synchrotron X-ray reflectivity (XRR) experiments and molecular dynamics (MD) simulations carried out under identical conditions for a range of 0.1 M alkali chloride (Li-, Na-, K-, Rb-, or CsCl) solutions on the basal surface of muscovite mica, a mineral isostructural to phyllosilicate clay minerals and one of the most widely studied reference surfaces in interfacial science. Our XRR and MD simulation results provide a remarkably consistent view of the structure and energetics of the Stern layer, with some discrepancy on the fraction of the minor outer-sphere component of Rb and on the adsorption energetics of Li. The results of both techniques, along with surface complexation model calculations, provide insight into the sensitivity of water structure and ion adsorption to surface topography and the type of adsorbed counterion