Sorption of Sr and U[VI] on Natural Mixtures of Hydrous Ferric and Aluminum Oxides

Evaluating and predicting vadose and groundwater reactive transport at field-scales requires an understanding of the abundance and nature of sorptive materials along subsurface flow paths. As part of our ongoing research focused on evaluating in situ reactive surface area, we have investigated the sorption of Sr and U[VI] using batch experiments and sands coated with naturally occurring mixtures of hydrous ferric and aluminum oxides over a range of environmentally relevant pH and sorbate concentrations. The goal of our research is to develop field applicable methodologies for predicting reactive transport in variably saturated media of radionuclide contaminants (e.g., U[VI]) from the behavior of simple benign tracers (e.g., Sr). Our tracer approach represents an effective method to integrate abundant and inexpensive laboratory sorption measurements across multiple scales with field-scale tests and observations. Materials used in the study were sands naturally coated with hydrous ferric and aluminum oxides from a field site near Oyster, Virginia and had hydrous metal oxide contents that ranged from 1 to 70 μmole g^-1 and 8 to 50 μmole g^-1 for iron and aluminum, respectively. Surface areas ranged from 0.1 to 3.2 m² g^-1 and are closely correlated with the abundance of the hydrous metal oxide coatings. After accounting for variations in surface area, Sr sorption was linear, suggesting that it can be used as a surrogate for reactive surface area. The adsorption of U[VI] was more complex. Uranium adsorption can be described by the empirical relationship: C_s = K [C_l x A]ⁿ where C_s and C_l are the concentration of sorbed (mass/mass sand) and dissolved (mass/volume solution) uranium, respectively. A is the surface area (area/mass) of the sand, and K and n are the Freundlich constant and exponent, both of which exhibit pH dependency. The more complex sorption behavior of U is attributed to multiple sorption sites and changes in speciation as a function of pH. Preliminary results suggest that surface areas estimated from Sr sorption can be used to predict U sorption.