Abstract

Selenium (Se) contamination poses significant risks to aquatic ecosystems and human health, particularly in waters affected by mining operations, coal-fired power plants, and agricultural drainage. While Se is an essential trace element, elevated concentrations can cause reproductive failure in fish and birds and neurological abnormalities in humans. Se predominantly exists in four oxidation states (+VI, +IV, 0, and -II) in the environment, with Se(VI) (i.e., SeO42-) and Se(IV) (i.e., HSeO3- and SeO32-) being the dominant species in most contaminated waters. Se(VI) is highly soluble and exhibits limited adsorption affinity to mineral surfaces, making it challenging to remove in water treatment. Se(IV) is of particular environmental concern due to its greater toxicity compared to Se(VI). Despite its high reactivity and stronger affinity for adsorbents such as iron and manganese oxyhydroxides, Se(IV) still poses significant challenges in water treatment, especially in achieving concentrations below regulatory limits. Iron electrocoagulation (EC) is a promising treatment technique as it generates reactive Fe(II)/Fe(III) solids that serve as both adsorbents and reductants for Se removal. This work investigates the mechanisms governing Se(VI) and Se(IV) removal during iron EC under environmentally relevant conditions. Specifically, a series of batch experiments were conducted to determine Se removal rates and extents as a function of water chemistry and operational conditions. In addition, the temporal dynamics of dissolved Se during and after EC treatment were investigated, and the comparative performance of EC and chemical coagulation was examined. In simple aqueous systems, Se(VI) removal exhibited distinct pathways under oxic and anoxic conditions, with the removal extents being significantly affected by pH. Under oxic conditions, Se(VI) removal was dominated by adsorption onto Fe(III) (oxy)hydroxides, with enhanced performance at acidic pH. Under anoxic conditions, Se(VI) removal proceeded by chemical reduction on mixed-valence iron-containing solids, achieving better removal at neutral to alkaline pH. Higher ionic strength decreased Se(VI) removal rates and extents by altering surface potential profiles and decreasing the activity of SeO42- ion. Se(VI) removal rates and extents increased with increasing charge loading rates due to faster Fe(II) dosage rate. A reaction-based model was developed for Se(VI) removal under oxic conditions, enabling performance prediction and system optimization. In complex aqueous matrices, the effects of co-occurring constituents such as sulfate, nitrate, bicarbonate, and natural organic matter (NOM) on Se(VI) removal were investigated under both oxic and anoxic conditions. These constituents affect not only the availability of adsorption sites but also the mineralogy and reactivity of iron-containing solids essential for Se removal from water. Under oxic conditions, sulfate decreased Se(VI) removal rates and extents through competitive adsorption. Under anoxic conditions, sulfate and bicarbonate suppressed Se(VI) removal by favoring the formation of less reactive forms of green rust (GR-SO4 and GR-CO3) over chloride green rust (GR-Cl) and magnetite. Nitrate had minimal effects under oxic conditions but inhibited Se(VI) removal under anoxic conditions by acting as a competing electron acceptor. NOM had negligible effects under oxic conditions but decreased Se(VI) removal rates under anoxic conditions due to competitive adsorption. Iron EC performance for Se removal was further evaluated in two synthetic challenge waters (representing mining and agricultural sectors) and one real flue gas desulfurization (FGD). Se(VI) removal was limited in all three challenge waters (<15%), confirming that the presence of co-occurring solutes substantially limited treatment efficiency. A pretreatment strategy for sulfate removal substantially improved Se(VI) removal from mining water (achieving 80% removal) but offered limited benefit for the agricultural and FGD waters, where additional factors such as high ionic strength and bicarbonate concentration likely limited performance. The temporal dynamics of Se following EC treatment demonstrated distinct behaviors under oxic and anoxic conditions, driven by differences in removal mechanisms and solid-phase transformations. Under oxic conditions, Se removal occurred through instantaneous adsorption onto Fe(III) (oxy)hydroxides, with removal extent determined entirely by iron dose. Under anoxic conditions, Se removal proceeded through rapid adsorption followed by slow reduction on mixed-valence solids such as chloride green rust and magnetite. A transient release was observed at lower iron dose (≤ 31 mg/L) due to the transformation of chloride green rust to magnetite and the limited capacity of the resulting solids to retain Se. In contrast, higher iron doses (≥ 78 mg/L) enabled sustained Se immobilization after treatment due to the enhanced magnetite formation, which facilitated both adsorption and reduction of Se species. While Se(IV) and Se(VI) exhibited similar removal mechanisms via iron EC, Se(IV) achieved faster kinetics and greater removal extents than Se(VI) at the same pH values under both oxic and anoxic conditions. Under anoxic conditions, both Se(VI) and Se(IV) removal proceeded through chemical reduction on mixed-valence iron-containing solids. The faster removal of Se(IV) suggested that the reduction of Se(VI) to Se(IV) was the rate-limiting step in the EC treatment process for Se(VI) removal. Under oxic conditions, both Se(VI) and Se(IV) removal were governed by adsorption onto Fe(III)(oxy)hydroxides, with Se(IV) achieving greater removal for a given amount of iron solids. This was due to the stronger adsorption affinity of Se(IV) compared to Se(VI). Sulfate had only minor effects on Se(IV) removal even at concentrations up to 500 mg/L, while Se(VI) removal was substantially inhibited under the same conditions, indicating the stronger binding affinity of Se(IV) compared to Se(VI) to iron-containing solids. Toxicity characteristic leaching procedure tests confirmed that Se remained largely immobilized in the solid phase, with aqueous concentrations well below regulatory limits. Compared to chemical coagulation, EC provides several mechanistic and operational advantages for Se (both Se(VI) and Se(IV)) removal. These include higher localized pH near the cathode, in situ coagulant generation, and precise control of iron dosage through current application. EC outperformed Fe(II) chemical coagulation for Se removal at pH 4 under oxic conditions, while chemical Fe(III) addition resulted in faster Se(IV) removal than EC due to more immediate Fe(III) solids formation. Under anoxic conditions at pH 8, Se removal performance was similar between EC and Fe(II) chemical coagulation due to similar iron dosage rates. Collectively, these findings advance mechanistic understanding of Se behavior during and after EC and demonstrate its potential as a flexible and effective treatment technology for Se-laden waters across diverse environmental contexts.

Committee Chair

Daniel Giammar

Committee Members

Jason He; Jeffrey Catalano; Xinhua Liang; Young-Shin Jun

Degree

Doctor of Philosophy (PhD)

Author's Department

Energy, Environmental & Chemical Engineering

Author's School

McKelvey School of Engineering

Document Type

Dissertation

Date of Award

12-19-2025

Language

English (en)

Download

Available for download on Friday, December 18, 2026

Share

COinS