Oxide- and Silicate-Water Interfaces and Their Roles in Technology and the Environment.

Interfacial reactions drive all elemental cycling on Earth and play pivotal roles in human activities such as agriculture, water purification, energy production and storage, environmental contaminant remediation, and nuclear waste repository management. The onset of the 21st century marked the beginning of a more detailed understanding of mineral aqueous interfaces enabled by advances in techniques that use tunable high-flux focused ultrafast laser and X-ray sources to provide near-atomic measurement resolution, as well as by nanofabrication approaches that enable transmission electron microscopy in a liquid cell. This leap into atomic- and nanometer-scale measurements has uncovered scale-dependent phenomena whose reaction thermodynamics, kinetics, and pathways deviate from previous observations made on larger systems. A second key advance is new experimental evidence for what scientists hypothesized but could not test previously, namely, interfacial chemical reactions are frequently driven by "anomalies" or "non-idealities" such as defects, nanoconfinement, and other nontypical chemical structures. Third, progress in computational chemistry has yielded new insights that allow a move beyond simple schematics, leading to a molecular model of these complex interfaces. In combination with surface-sensitive measurements, we have gained knowledge of the interfacial structure and dynamics, including the underlying solid surface and the immediately adjacent water and aqueous ions, enabling a better definition of what constitutes the oxide- and silicate-water interfaces. This critical review discusses how science progresses from understanding ideal solid-water interfaces to more realistic systems, focusing on accomplishments in the last 20 years and identifying challenges and future opportunities for the community to address. We anticipate that the next 20 years will focus on understanding and predicting dynamic transient and reactive structures over greater spatial and temporal ranges as well as systems of greater structural and chemical complexity. Closer collaborations of theoretical and experimental experts across disciplines will continue to be critical to achieving this great aspiration.

Molecular models of cesium and rubidium adsorption on weathered micaceous minerals.

Understanding the adsorption mechanisms of metal cations onto soils and sediments is of critical importance in the protection of the environment, especially for the case of radioactive materials including the fission product (137)Cs. Mechanism-based adsorption models for the long-term interaction of chemical and radionuclide species with clay minerals are needed to improve the accuracy of groundwater reaction and flow models, as well as related simulations for performance assessment of waste sites and repositories. Toward this goal, molecular simulation using geometry optimization and molecular dynamics methods have been used to investigate the adsorption behavior of Cs(+) and Rb(+) cations at frayed edge wedges (a proxy for frayed edge sites, FES) and in the interlayer region formed as a result of the transformation of muscovite to Al-hydroxy interlayered vermiculite (HIV) during weathering and pedogenesis. Frayed edge wedges, formed both on individual smectite and illite phases and on the mica-HIV intergrade, have previously been recognized as significant sinks for the strong adsorption of Cs(+) and Rb(+). Atomic density profiles, interlayer adsorption site maps, radial distribution functions, and adsorption enthalpies derived from the equilibrated structural models are used to evaluate the optimal adsorption configurations and thermodynamics for Cs- and Rb-endmembers, a 50:50 Cs-Rb composition for the aqueous interlayer of vermiculite, and for the interlayer wedge zone as mica is transformed to HIV (i.e., HIV-mica wedge). Adsorption enthalpies for both cations are significantly larger for the frayed edge wedges (as represented by the HIV-mica wedge model) compared to values for the vermiculite and mica interlayers. Cesium cation binds more strongly than Rb(+) in the vermiculite interlayer, while Rb(+) binds more strongly than Cs(+) in the HIV-mica wedge. In all cases, the derived adsorption enthalpies for both cations indicate a preference for the wedge environment where electrostatic interaction is enhanced due to the presence of layer charge and the increased size of interlayer at the wedge accommodating cations larger than K(+).

Molecular simulation of carbon dioxide, brine, and clay mineral interactions and determination of contact angles.

Capture and subsequent geologic storage of CO2 in deep brine reservoirs plays a significant role in plans to reduce atmospheric carbon emission and resulting global climate change. The interaction of CO2 and brine species with mineral surfaces controls the ultimate fate of injected CO2 at the nanoscale via geochemistry, at the pore-scale via capillary trapping, and at the field-scale via relative permeability. We used large-scale molecular dynamics simulations to study the behavior of supercritical CO2 and aqueous fluids on both the hydrophilic and hydrophobic basal surfaces of kaolinite, a common clay mineral. In the presence of a bulk aqueous phase, supercritical CO2 forms a nonwetting droplet above the hydrophilic surface of kaolinite. This CO2 droplet is separated from the mineral surface by distinct layers of water, which prevent the CO2 droplet from interacting directly with the mineral surface. Conversely, both CO2 and H2O molecules interact directly with the hydrophobic surface of kaolinite. In the presence of bulk supercritical CO2, nonwetting aqueous droplets interact with the hydrophobic surface of kaolinite via a mixture of adsorbed CO2 and H2O molecules. Because nucleation and precipitation of minerals should depend strongly on the local distribution of CO2, H2O, and ion species, these nanoscale surface interactions are expected to influence long-term mineralization of injected carbon dioxide.

Molecular simulations of carbon dioxide and water: cation solvation.

Proposed carbon dioxide sequestration scenarios in sedimentary reservoirs require investigation into the interactions between supercritical carbon dioxide, brines, and the mineral phases found in the basin and overlying caprock. Molecular simulations can help to understand the partitioning of metal cations between aqueous solutions and supercritical carbon dioxide where limited experimental data exist. In this effort, we used classical molecular dynamics simulations to compare the solvation of alkali and alkaline-earth metal cations in water and liquid CO(2) at 300 K by combining a flexible simple point charge model for water and an accurate flexible force field for CO(2). Solvation energies for these cations are larger in water than in carbon dioxide, suggesting that they will partition preferentially into water. In both aqueous and CO(2) solutions, the solvation energies decrease with cation size and increase with cation charge. However, changes in solvation energy with ionic radii are smaller in CO(2) than in water suggesting that the partitioning of cations into CO(2) will increase with ion size. Simulations of the interface between aqueous solution and supercritical CO(2) support this suggestion in that some large cations (e.g., Cs(+) and K(+)) partition into the CO(2) phase, often with a partial solvation sphere of water molecules.

Effects of thermodynamic ensembles and mineral surfaces on interfacial water structure.

While performing molecular dynamics simulations of water or aqueous solutions in a slab geometry, such as at mineral surfaces, it is important to match bulk water density in the diffuse region of the model system with that expected for the appropriate experimental conditions. Typically, a slab geometry represents parallel surfaces with a variable region of confined water (this region can range in size from a few Ångstroms to many tens of Ångstroms). While constant-pressure simulations usually result in appropriate density values in the bulk diffuse region removed from either surface, constant-volume simulations have also been widely used, sometimes without careful consideration of the method for determining water content. Simulations using two thermodynamic ensembles as well as two methods for calculating the water-accessible volume have been investigated for two distinct silicate surfaces-hydrophilic cristobalite (111) and hydrophobic pyrophyllite (001). In cases where NPT simulations are not feasible, a simple geometry-based treatment of the accessible volume can be sufficient to replicate bulk water density far from the surface. However, the use of the Connolly method can be more appropriate in cases where a surface is less well-defined. Specific water-surface interactions (e.g., hydrophobic repulsion) also play a role in determining water content in a confined water simulation. While reported here for planar surfaces, these results can be extended to an interface with any solvent, or to other types of surfaces and geometries.

Nanoconfined water in magnesium-rich 2:1 phyllosilicates.

Inelastic neutron scattering, density functional theory, ab initio molecular dynamics, and classical molecular dynamics were used to examine the behavior of nanoconfined water in palygorskite and sepiolite. These complementary methods provide a strong basis to illustrate and correlate the significant differences observed in the spectroscopic signatures of water in two unique clay minerals. Distortions of silicate tetrahedra in the smaller-pore palygorskite exhibit a limited number of hydrogen bonds having relatively short bond lengths. However, without the distorted silicate tetrahedra, an increased number of hydrogen bonds are observed in the larger-pore sepiolite with corresponding longer bond distances. Because there is more hydrogen bonding at the pore interface in sepiolite than in palygorskite, we expect librational modes to have higher overall frequencies (i.e., more restricted rotational motions); experimental neutron scattering data clearly illustrates this shift in spectroscopic signatures. It follows that distortions of the silicate tetrahedra in these minerals effectively disrupt hydrogen-bonding patterns at the silicate-water interface, and this has a greater impact on the dynamical behavior of nanoconfined water than the actual size of the pore or the presence of coordinatively unsaturated magnesium edge sites.

Implementation of a Morse potential to model hydroxyl behavior in phyllosilicates.

The accurate molecular simulation of many hydrated chemical systems, including clay minerals and other phyllosilicates and their interfaces with aqueous solutions, requires improved classical force field potentials to better describe structure and vibrational behavior. Classical and ab initio molecular dynamics simulations of the bulk structure of pyrophyllite, talc, and Na-montmorillonite clay phases exhibit dissimilar behavior in the hydroxyl stretch region of power spectra derived from atomic trajectories. The classical simulations, using the CLAYFF force field, include either a standard harmonic potential or a new Morse potential parametrized for both dioctahedral and trioctahedral phases for the O-H bond stretch. Comparisons of classical results with experimental values and with ab initio molecular dynamics simulations indicate improvements in the simulation of hydroxyl orientation relative to the clay octahedral sheet and in the O-H bond stretch in the high frequency region of the power spectrum.

Vibrational spectra of methane clathrate hydrates from molecular dynamics simulation.

Molecular dynamics simulations were performed on methane clathrate hydrates at ambient conditions. Thermal expansion results over the temperature range 60-300 K show that the unit cell volume increases with temperature in agreement with experiment. Power spectra were obtained at 273 K from velocity autocorrelation functions for selected atoms, and normal modes were assigned. The spectra were further classified according to individual atom types, allowing the assignment of contributions from methane molecules located in small and large cages within the structure I unit cell. The symmetric C-H stretch of methane in the small cages occurs at a higher frequency than for methane located in the large cages, with a peak separation of 14 cm(-1). Additionally, we determined that the symmetric C-H stretch in methane gas occurs at the same frequency as methane in the large cages. Results of molecular dynamics simulations indicate the use of power spectra obtained from the velocity autocorrelation function is a reliable method to investigate the vibrational behavior of guest molecules in clathrate hydrates.

Structure, energetics, and dynamics of water adsorbed on the muscovite (001) surface: a molecular dynamics simulation.

Molecular dynamics (MD) computer simulations of liquid water adsorbed on the muscovite (001) surface provide a greatly increased, atomistically detailed understanding of surface-related effects on the spatial variation in the structural and orientational ordering, hydrogen bond (H-bond) organization, and local density of H2O molecules at this important model phyllosilicate surface. MD simulations at constant temperature and volume (statistical NVT ensemble) were performed for a series of model systems consisting of a two-layer muscovite slab (representing 8 crystallographic surface unit cells of the substrate) and 0 to 319 adsorbed H2O molecules, probing the atomistic structure and dynamics of surface aqueous films up to 3 nm in thickness. The results do not demonstrate a completely liquid-like behavior, as otherwise suggested from the interpretation of X-ray reflectivity measurements and earlier Monte Carlo simulations. Instead, a more structurally and orientationally restricted behavior of surface H2O molecules is observed, and this structural ordering extends to larger distances from the surface than previously expected. Even at the largest surface water coverage studied, over 20% of H2O molecules are associated with specific adsorption sites, and another 50% maintain strongly preferred orientations relative to the surface. This partially ordered structure is also different from the well-ordered 2-dimensional ice-like structure predicted by ab initio MD simulations for a system with a complete monolayer water coverage. However, consistent with these ab initio results, our simulations do predict that a full molecular monolayer surface water coverage represents a relatively stable surface structure in terms of the lowest diffusional mobility of H2O molecules along the surface. Calculated energies of water adsorption are in good agreement with available experimental data.

Vibrational Analysis of Brucite Surfaces and the Development of an Improved Force Field for Molecular Simulation of Interfaces.

We introduce a nonbonded three-body harmonic potential energy term for Mg-O-H interactions for improved edge surface stability in molecular simulations. The new potential term is compatible with the Clayff force field and is applied here to brucite, a layered magnesium hydroxide mineral. Comparisons of normal mode frequencies from classical and density functional theory calculations are used to verify a suitable spring constant (k parameter) for the Mg-O-H bending motion. Vibrational analysis of hydroxyl librations at two brucite surfaces indicates that surface Mg-O-H modes are shifted to frequencies lower than the corresponding bulk modes. A comparison of DFT and classical normal modes validates this new potential term. The methodology for parameter development can be applied to other clay mineral components (e.g., Al, Si) to improve the modeling of edge surface stability, resulting in expanded applicability to clay mineral applications.

Molecular dynamics studies of nanoconfined water in clinoptilolite and heulandite zeolites.

The complete periodic series of alkali and alkaline earth cation variants (Li(+), Na(+), K(+), Rb(+), Cs(+), Mg(2+), Ca(2+), Sr(2+), and Ba(2+)) of clinoptilolite (Si : Al=5) and heulandite (Si : Al=3.5) aluminosilicate zeolites are examined by large-scale molecular dynamics utilizing a flexible SPC water and aluminosilicate force field. Calculated hydration enthalpies, radial distribution functions, and ion coordination environments are used to describe the energetic and structural components of extra-framework species while power spectra are used to examine the intermolecular dynamics. These data are correlated to evaluate the impact of ion-zeolite, ion-water, and water-zeolite interactions on the behavior of nanoconfined water. Analysis of the correlated data clearly indicates that the charge density of extra-framework cations appears to have the greatest influence on librational motions, while the anionic charge of the framework (i.e. Si:Al ratios) has a lesser impact.

Water structure and aqueous uranyl(VI) adsorption equilibria onto external surfaces of beidellite, montmorillonite, and pyrophyllite: results from molecular simulations.

Molecular dynamics simulations were performed to provide a systematic study of aqueous uranyl adsorption onto the external surface of 2:1 dioctahedral clays. Our understanding of this key process is critical in predicting the fate of radioactive contaminants in natural groundwaters. These simulations provide atomistic detail to help explain experimental trends in uranyl adsorption onto natural media containing smectite clays. Aqueous uranyl concentrations ranged from 0.027 to 0.162 M. Sodium ions and carbonate ions (0.027-0.243 M) were also present in the aqueous regions to more faithfully model a stream of uranyl-containing groundwater contacting a mineral system comprised of Na-smectite. No adsorption occurred near the pyrophyllite surface, and there was little difference in uranyl adsorption onto the beidellite and montmorillonite, despite the difference in location of clay layer charge between the two. At low uranyl concentration, the pentaaquouranyl complex dominates in solution and readily adsorbs to the clay basal plane. At higher uranyl (and carbonate) concentrations, the mono(carbonato) complex forms in solution, and uranyl adsorption decreases. Sodium adsorption onto beidellite occurred both as inner- and outer-sphere surface complexes, again with little effect on uranyl adsorption. Uranyl surface complexes consisted primarily of the pentaaquo cation (85%) and to a lesser extent the mono(carbonato) species (15%). Speciation diagrams of the aqueous region indicate that the mono(carbonato)uranyl complex is abundant at high ionic strength. Oligomeric uranyl complexes are observed at high ionic strength, particularly near the pyrophyllite and montmorillonite surfaces. Atomic density profiles of water oxygen and hydrogen atoms are nearly identical near the beidellite and montmorillonite surfaces. Water structure therefore appears to be governed by the presence of adsorbed ions and not by the location of layer charge associated with the substrate. The water oxygen density near the pyrophyllite surface is similar to the other cases, but the hydrogen density profile indicates reduced hydrogen bonding between adsorbed water molecules and the surface.

Molecular dynamics simulation of uranyl(VI) adsorption equilibria onto an external montmorillonite surface.

We used molecular dynamics simulations to study the adsorption of aqueous uranyl species (UO(2)(2+)) onto clay mineral surfaces in the presence of sodium counterions and carbonato ligands. The large system size (10,000 atoms) and long simulation times (10 ns) allowed us to investigate the thermodynamics of ion adsorption, and the atomistic detail provided clues for the observed adsorption behavior. The model system consisted of the basal surface of a low-charge Na-montmorillonite clay in contact with aqueous uranyl carbonate solutions with concentrations of 0.027 M, 0.081 M, and 0.162 M. Periodic boundary conditions were used in the simulations to better represent an aqueous solution interacting with an external clay surface. Uranyl adsorption tendency was found to decrease as the aqueous uranyl carbonate concentration was increased, while sodium adsorption remained constant. The observed behavior is explained by physical and chemical effects. As the ionic strength of the aqueous solution was increased, electrostatic factors prevented further uranyl adsorption once the surface charge had been neutralized. Additionally, the formation of aqueous uranyl carbonate complexes, including uranyl carbonato oligomers, contributed to the decreased uranyl adsorption tendency.

Characterization of adsorption sites on aggregate soil samples using synchrotron X-ray computerized microtomography.

Synchrotron-source X-ray computerized microtomography (CMT) was used to evaluate the adsorptive properties of aggregate soil samples. A linear relationship between measured mean mass attenuation coefficient (sigma) and mass fraction iron was generated by imaging mineral standards with known iron contents. On the basis of reported stoichiometries of the clay minerals and identifications of iron oxyhydroxides (1), we calculated the mass fraction iron and iron oxyhydroxide in the intergranular material. The mass fractions of iron were estimated to range from 0.17 to 0.22 for measurements made at 18 keV and from 0.18 to 0.21 for measurements made at 26 keV. One aggregate sample also contained regions within the intergranular material with mass fraction iron ranging from 0.29 to 0.31 and from 0.33 to 0.36 for the 18 and 26 keV measurements, respectively. The mass fraction iron oxyhydroxide ranged from 0.18 to 0.35 for the low-iron intergranular material and from 0.40 to 0.59 for the high-iron intergranular material. Using absorption edge difference imaging with CMT, we visualized cesium on the intergranular material, presumably because of adsorption and possible exchange reactions. By characterizing the mass fraction iron, the mass fraction iron oxyhydroxide, and the adsorptive capacity of these soil mineral aggregates, we provide information useful for conceptualization, development, and parametrization of transport models.

Linear free energy relationships between dissolution rates and molecular modeling energies of rhombohedral carbonates.

Bulk and surface energies are calculated for endmembers of the isostructural rhombohedral carbonate mineral family, including Ca, Cd, Co, Fe, Mg, Mn, Ni, and Zn compositions. The calculations for the bulk agree with the densities, bond distances, bond angles, and lattice enthalpies reported in the literature. The calculated energies also correlate with measured dissolution rates: the lattice energies show a log-linear relationship to the macroscopic dissolution rates at circumneutral pH. Moreover, the energies of ion pairs translated along surface steps are calculated and found to predict experimentally observed microscopic step retreat velocities. Finally, pit formation excess energies decrease with increasing pit size, which is consistent with the nonlinear dissolution kinetics hypothesized for the initial stages of pit formation.