| Description |
Despite several decades of research there currently exists no mechanistic theory to predict deposition in porous media in the presence of colloid-collector repulsion (unfavorable conditions). Recently, mechanistic models have been developed that incorporate nanoscale surface heterogeneity in colloid-collector interactions. Comparison of simulations to experimental data allows backing out a representative heterogeneity for the surface, which to date has been reported only for silica. Colloid deposition onto a variety of representative aquifer materials expected to be unfavorable under environmental conditions (amorphous silica, muscovite and albite) was observed for 0.25, 1.1 and 2.0 m carboxylate-modified latex microspheres using an impinging jet system. Deposition efficiencies varied in response to changes in collector mineralogy, ionic strength, electrolyte valence and pH. Characteristics (size and spatial distribution) of surficial charge-heterogeneity on the collector were backed out from these experiments via comparison to particle trajectory simulations incorporating discrete nanoscale attractive domains (heterodomains). A bimodal distribution (1:4, ratio of large to small heterodomains) of 120nm and 60nm diameter heterodomains was found to predict attachment onto silica across the range of colloid sizes, solution chemistries and fluid velocities, whereas a uniform distribution of 120nm heterodomains was found to predictive of retention on muscovite. Varying the surface coverage of heterodomains allowed a characteristic coverage for each observed collector surface to be determined. This developing catalog of surface characteristics will aid prediction of colloid transport and deposition under environmentally relevant conditions (aquifer material composition and groundwater chemistry). This study also highlighted the fact that heterogeneity on the collector surface alone is not enough to capture trends in deposition across the range of minerals and solution chemistries examined and more mechanisms of attachment need to be examined. |
