[1] We studied in situ colloid mobilization under transient flow conditions using columns repacked with Hanford sediments. Rainfall infiltration was experimentally simulated using different flow rates and initial moisture conditions. Five series of column experiments were performed with initial infiltration rates of 0.018, 0.036, 0.072, 0.144, and 0.288 cm/min, and the columns reached water saturations in the range of 53 to 81%. The infiltration of water into the columns provided unfavorable conditions for colloid attachment to the sediments. Colloids were eluted by the infiltrating water with the peak colloid concentrations in the outflow coinciding with the arrival of the infiltration front. A larger flow rate led to a greater amount of colloids released from the column. The cumulative amount of colloids released was proportional to the column water content established after steady state flow rates were achieved. We used the advection-dispersion equation with a first-order colloid release reaction to analyze the experimental data. The colloid release rate coefficient increased with the increase of water content. We calculated forces exerted on colloids, and found that electrostatic and van der Waals interactions, calculated based on the DLVO (Derjaguin-Landau-Verwey-Overbeek) theory, and hydrodynamic forces, were all less important than capillary forces in controlling colloid release. In one experiment, the ionic strength of the infiltration solution was increased, such that colloid attachment was favorable. Nonetheless, colloids were mobilized and eluted with the infiltration front, implying that non-DLVO forces, such as capillary forces, played a prominent role in colloid mobilization.Citation: Shang, J., M. Flury, G. Chen, and J. Zhuang (2008), Impact of flow rate, water content, and capillary forces on in situ colloid mobilization during infiltration in unsaturated sediments, Water Resour. Res., 44, W06411,

No abstract

All natural surfaces exhibit nanoscale roughness (NR) and chemical heterogeneity (CH) to some extent. Expressions were developed to determine the mean interaction energy between a colloid and a solid-water interface, as well as for colloid-colloid interactions, when both surfaces contain binary NR and CH. The influence of heterogeneity type, roughness parameters, solution ionic strength (IS), mean zeta potential, and colloid size on predicted interaction energy profiles was then investigated. The role of CH was enhanced on smooth surfaces with larger amounts of CH, especially for smaller colloids and higher IS. However, predicted interaction energy profiles were mainly dominated by NR, which tended to lower the energy barrier height and the magnitudes of both the secondary and primary minima, especially when the roughness fraction was small. This dramatically increased the relative importance of primary to secondary minima interactions on net electrostatically unfavorable surfaces, especially when roughness occurred on both surfaces and for conditions that produced small energy barriers (e.g., higher IS, lower pH, lower magnitudes in the zeta potential, and for smaller colloid sizes) on smooth surfaces. The combined influence of roughness and Born repulsion frequently produced a shallow primary minimum that was susceptible to diffusive removal by random variations in kinetic energy, even under electrostatically favorable conditions. Calculations using measured zeta potentials and hypothetical roughness properties demonstrated that roughness provided a viable alternative explanation for many experimental deviations that have previously been attributed to electrosteric repulsion (e.g., a decrease in colloid retention with an increase in solution IS; reversible colloid retention under favorable conditions; and diminished colloid retention and enhanced colloid stability due to adsorbed surfactants, polymers, and/or humic materials).

[1] Capillary forces acting at the air-water interface play an important role in colloid fate and transport in subsurface porous media. We quantified capillary forces between different particles (sphere, cylinder, cube, disk, sheet, and natural mineral particles) and a moving air-water interface. The particles had different sizes and contact angles (ranging from 14°to 121°). Theoretical calculations using the Young-Laplace equation were used to support and generalize the experimental data. When the air-water interface moved over the particles, there were strong capillary forces acting on the particles in the direction of the moving interface. The measured maximum capillary forces were similar to those calculated by the Young-Laplace equation. The larger the contact angles and the larger the particle size, the stronger were the capillary forces. Particles with irregular shape and sharp edges experienced greater forces than smooth particles. Generalization of the results indicates that capillary forces exerted by a moving air-water interface can readily exceed attractive Derjaguin-Landau-Verwey-Overbeek (DLVO) and gravity forces for typical subsurface particles, and a moving air-water interface is therefore an effective mechanism for mobilization of particles in porous media. Particles in the colloidal size range are particularly susceptible for mobilization by a moving air-water interface.Citation: Shang, J., M. Flury, and Y. Deng (2009), Force measurements between particles and the air-water interface: Implications for particle mobilization in unsaturated porous media, Water Resour. Res., 45, W06420,

The contribution of variable grain sizes to uranium adsorption/desorption was studied using a sediment from the US DOE Hanford site. The sediment was wet sieved into four size fractions: coarse sand (1-2 mm), medium sand (0.2-1 mm), fine sand (0.053-0.2 mm), and clay/silt fraction (<0.053 mm). For each size fraction and their composite (sediment), batch and flow-cell experiments were performed to determine uranium adsorption isotherms and kinetic uranium adsorption and subsequent desorption. The results showed that uranium adsorption isotherms and adsorption/desorption kinetics were size specific, reflecting the effects of size-specific adsorption site concentration and kinetic rate constants. The larger-size fraction had a larger mass percentage in the sediment but with a smaller adsorption site concentration and generally a slower uranium adsorption/desorption rate. The same equilibrium surface complexation reaction and reaction constant could describe uranium adsorption isotherms for all size fractions and the composite after accounting for the effect of adsorption site concentration. Mass-weighted, linear additivity was observed for both uranium adsorption isotherms and adsorption/desorption kinetics in the composite. One important implication of this study is that grain-size distribution may be used to estimate uranium adsorption site and adsorption/desorption kinetic rates in heterogeneous sediments from a common location.