Antimony (Sb) is a co-contaminant with lead (Pb) in shooting range soils at Department of Defense (DoD) installations throughout the United States. The in situ immobilization of Pb in shooting range soils can be accomplished through the application of phosphate (PO₄); however, the impact of this treatment on the mobility and bioaccessibility of Sb is unknown. Further, the ability to predict Sb fate and behavior in contaminated soils, or as influenced by treatment technologies, has not been suitably developed. In soil, Sb commonly exists in the Sb(V) oxidation state as the hydroxyanion (Sb(OH)₆^–). This anionic species is derived through the hydrolysis of antimonic acid (Sb(OH)₅ ), a weak acid). As such, the principal mechanisms of retention in soils are anion exchange (weak outer-sphere electrostatic adsorption) and ligand exchange (strong inner-sphere covalent adsorption) by variable-charge soil minerals, such as iron (Fe), aluminum (Al), and manganese (Mn) oxyhydroxides. Available research findings suggest that Sb(V) is associated with Fe oxyhydroxides in soils and that PO₄ amendments can enhance Sb(V) mobility and bioaccessibility.

The objectives of this project were to: (1) determine the mechanisms and thermodynamics of Sb(V) adsorption by hydrous Fe, Al, and Mn oxyhydroxides (goethite, gibbsite, kaolinite, and birnessite) as a function of ionic environment, pH, temperature, and Sb(V) concentrations; (2) quantify the competitive effects of PO₄ and sulfate (SO₄) on Sb(V) adsorption; and (3) develop and evaluate the capability of chemical models to predict Sb(V) adsorption within the holistic framework of a complex chemical environment.

A series of laboratory-based experiments were performed to determine the Sb(V) adsorption mechanisms, and the tenacity and reversibility of the adsorption processes. Adsorption edge studies were used to assess the mechanisms of Sb(V) retention by reactive soil minerals as a function of several environmental variables, including pH, ionic strength (I_s, controlled by KNO₃, potassium nitrate), and the presence of competing ligands (PO₄ and SO₄). Adsorption isotherms were developed as a function of Sb(V) concentration, pH, and temperature to assess the thermodynamics of Sb(V) adsorption. The data accumulated from these experimental activities, including the identified adsorption mechanisms, were then used to develop mechanistic predictive models that combine aqueous speciation and surface complexation (adsorption) phenomenon. The chemical modeling activity resulted in mechanistic parameters that described Sb(V) retention and that are transferable; they can be used to predict Sb(V) fate and behavior in any soil environment (given that soil chemical information are available).

The aluminol group (≡AlOH) is the reactive surface functional group on kaolinite and gibbsite. The aluminol group has relatively low affinity for Sb(V), and retention is both pH- and ionic strength-dependent. Kaolinite exhibits the lowest capacity to retain Sb(V) (1.48 mmol kg^‒1 adsorbed Sb(V) at pH 5.5), with minimal adsorption (~0% of added Sb(V)) in pH > 7 suspensions. In pH < 4 suspensions, adsorption increases to approximately 50% of the added Sb(V) in 0.1 I_s and to 80% in 0.01 I_s. Similarly, Sb(V) retention by gibbsite is pH- and I_s-dependent (4.32 mmol kg^‒1 adsorbed Sb(V) at pH 5.5), with between 0% and 10% of the added Sb(V) retained in pH > 9 suspensions. In pH < 4 suspensions, retention increases to approximately 80% of the added Sb(V) in 0.1 I_s and to > 90% in 0.01 I_s. The ionic strength-dependency of Sb(V) adsorption by kaolinite and gibbsite indicates that the weak, electrostatic retention of Sb(V) is an important mechanism. However, in strongly acidic suspensions (pH < 5 to 6), Sb(V) adsorption is irreversible, suggesting strong covalent bonding. The mechanistic interpretation of the adsorption edge results are supported by the adsorption isotherm results and surface electrostatics. In pH 8 suspensions, Sb(V) adsorption is exothermic, indicating that the predominant retention mechanism is anion exchange. In pH 5.5 suspensions, there is an endothermic component to the adsorption process, indicating covalent bonding by the aluminol functional group. Antimonate adsorption generates a negative shift in surface charge and an increase in proton adsorption, both of which are consistent with covalent bonding. Both sulfate and phosphate interfere with Sb(V) retention on kaolinite and gibbsite.

The ferrol group (≡FeOH) on goethite has a high capacity to retain Sb(V) (88.5 mmol kg^‒1 adsorbed Sb(V) at pH 5.5). Adsorption by goethite is pH-dependent, independent of ionic strength, and generally irreversible. Approximately 40% of the added Sb(V) is retained by goethite in pH 10 suspensions, increasing to 100% when pH < 6. Antimonate adsorption is endothermic in both pH 5.5 and 8 suspensions and adsorption generates a negative shift in goethite surface charge, indicating covalent bonding by the ferrol functional group. Antimonate adsorption by goethite is not impacted by sulfate. However, phosphate strongly inhibits Sb(V) retention.

Like the ferrol group on goethite, the manganol group (≡MnOH) on birnessite is a scavenger for Sb(V) (14.8 mmol kg^‒1 adsorbed Sb(V) at pH 5.5). Adsorption by birnessite is pH- and ionic strength-dependent. Approximately 10% (low I_s) to 20% (high I_s) of the added Sb(V) was retained by birnessite in pH > 9 suspensions, increasing to 100% when pH < 5. Antimonate adsorption generates a negative shift in birnessite surface charge, indicating covalent bonding by the manganol functional group. Antimonate adsorption by birnessite is not impacted by either sulfate or phosphate.

The experimental findings suggest that the retention of Sb(V) by kaolinite and gibbsite occurs via a combination of electrostatic and covalent bonding mechanisms; whereas, retention by goethite and birnessite occurs predominately by covalent mechanisms. Antimonate adsorption by all surfaces, as a function of pH and ionic strength, was successfully predicted by employing the triple-layer surface complexation model that considered both outer-sphere [≡SOH₂^+‒Sb(OH)₆^‒] and inner-sphere [≡SOSb(OH)₅^‒] adsorption mechanisms. In general, however, the models generated for single ligand systems required reoptimization to successfully predict adsorption in the competitive systems (Sb(OH)₆ and SO₄ or Sb(OH)₆ and PO₄).

This research describes the adsorption of Sb(V) by the surface-reactive minerals that are common to soils and sediments. The results indicate that Sb(V) retention is strongly dependent on pH. Depending on the adsorbent, Sb(V) adsorption is also influenced by the ionic strength (salinity) and the presence of ligands (SO₄ and PO₄) that compete for adsorption sites. In general, Sb(V) is immobilized in strongly acidic environments and by Fe- and Mn-rich phases (but not by Al-rich phases). The research findings also indicate that the addition of PO₄-based fertilizer amendments to immobilize lead in shooting range soils will potentially enhance Sb(V) mobility and bioaccessibility. Geochemical models that predict the distribution of Sb(V) between soluble and adsorbed phases as a function of pH and ionic environment were successfully developed. However, the application of these models to predict behavior in Sb(V)-affected environments will require site-specific chemical information and calibration. The results of this project will help establish technically defensible cleanup goals and priorities at DoD facilities and will improve public and DoD site manager confidence in the management of contaminated environments.