Journal of Hazardous, Toxic, and Radioactive Waste

This study presents the results of an extensive beneficial-use evaluation of $3/8$-in. minus steel slag fines (SSF) to immobilize arsenic. Two primary sets of experiments were undertaken to assess (1) the ability of SSF to immobilize $100\mathrm{mg}/\mathrm{kg}$ arsenite (${\mathrm{As}}^{3+}$) and arsenate (${\mathrm{As}}^{5+}$) in dredged material when blended with SSF, including slag cement doses (up to 2%) to determine if additional environmental polishing was necessary; and (2) the ability of SSF alone to immobilize each $\mathrm{As}$ species. Visually, the SSF materials resemble an AASHTO No. 9 (fine) aggregate, with a small fraction passing the No. 200 (0.075 mm) sieve. In order to establish the design parameters for deploying the slag media in geoenvironmental applications (soil blending, drainage, reactive trenches, and filters), the soil classification and grain-size distribution, specific gravity, loss on ignition (ash content), standard and modified Proctor compaction behavior, direct shear strength, and swell behavior of the SSF media were evaluated. Additionally, the following geochemical attributes of the SSF media were evaluated: bulk chemistry, mineralogy, pH, anion scan, total priority pollutant list (PPL) metals, toxicity characteristic leaching procedure (TCLP), and synthetic precipitation leaching procedure (SPLP) leaching behavior for PPL metals. Arsenic thresholding studies were performed, in which the uptake of each $\mathrm{As}$ source on the SSF materials was evaluated. The SSF materials immobilized approximately $7,900\mathrm{mg}/\mathrm{kg}$ ${\mathrm{As}}^{3+}$ and $8,800\mathrm{mg}/\mathrm{kg}$ ${\mathrm{As}}^{5+}$, producing TCLP and SPLP concentrations less than $0.010\mathrm{mg}/\mathrm{L}$ in three of four cases. X-ray diffraction (XRD) and scanning electron microscopy (SEM) studies were used in combination with MINTEQ modeling to isolate the mechanisms responsible for the $\mathrm{As}$ immobilization in the SSF materials.

Batch and column leaching tests were conducted to assess the simultaneous stabilization of copper (Cu), lead (Pb), and tungsten (W) in eight representative contaminated firing range soils in the United States using various amendments. The amendments included granulated ferric oxide (GFO), granulated titanium dioxide (GTD), Pahokee peat soil (PPS), Gascoyne leonardite soil (GLS), Elliot silty loam soil (ESLS), calcium phosphate monobasic (CPM), and apatite II. The metal oxides and the organic soil amendments were applied at a dosage of 10%, and phosphates were applied at phosphorus to lead (P/Pb) molar ratio of 1.8. The experimental results indicated that GFO was superior to all materials tested for simultaneously stabilizing Cu, Pb, and W during the batch leaching tests. Flow-through column tests were conducted for one of the soil samples to test the effectiveness of GFO to immobilize Cu, Pb, and W. The concentrations of Cu, Pb, and W were significantly reduced in the effluent of the amended soil columns as compared with the control soil columns.

Together with industrialization, more sites have become contaminated with mixed wastes of organics. Contamination of groundwater with benzene, toluene, ethylbenzene, and three isomers of xylene (BTEX) and chlorinated aliphatic hydrocarbons (CAHs) allows for the application of aerobic bioremediation to achieve mineralization of both types of compounds. In this study, the aerobic bioremoval of mixtures of these compounds [BTEX, BTEX/trichloroethylene (TCE), BTEX/cis-1,1-dichloroethylene (cis-DCE), BTEX/total petroleum hydrocarbons (TPH)/TCE, and toluene and ortho-xylene (ToX/TCE] was assessed under different environmental conditions (pH and temperature) by using indigenous microorganisms isolated from potentially contaminated regional sites. The highest TPH bioremoval efficiencies ($<50\%$) occurred at 25°C and under neutral or alkaline conditions. TCE was cometabolized with toluene and $o$-xylene provided as growth substrates, and the highest bioremoval efficiency occurred at $\mathrm{T}o\mathrm{X}/\mathrm{TCE}$, and pH and temperature impacted the mineralization of compounds. This study would help enhance the applicability of bioremediation technology to the mixed-wastes contaminated sites.

Polycyclic aromatic hydrocarbons (PAHs) and heavy metals are environmental concerns and must be removed to acceptable levels. This paper evaluates different flushing agents to enhance the remediation of soil contaminated with PAHs and heavy metals at a former manufactured gas plant site. Four flushing column tests at a constant hydraulic gradient of 1.2 were conducted using four different flushing agents, which included deionized water, chelant (0.2 M EDTA), surfactant (5% Igepal CA-720), and cyclodextrin (10% hydroxypropyl-$\beta$-cyclodextrin or HPCD). Additional column tests using Igepal and HPCD at a lower hydraulic gradient of 0.2 were conducted to investigate the effects of rate-limited desorption or solubilization of PAHs. The results showed that the EDTA produced the maximum metal removal from the soil compared with deionized water, Igepal, and HPCD under different hydraulic gradient conditions. The 0.2 M EDTA flushing solution removed approximately 25–75% of the toxic heavy metals found in the soil. None of the PAHs were removed from the soil when deionized water and EDTA were the flushing solutions. The PAHs removal efficiencies in the Igepal and HPCD systems decreased as the hydraulic gradient decreased. However, the surfactant-enhanced systems were more efficient in removing PAHs from the soil than the HPCD systems under high- and low-hydraulic gradients. The results also demonstrated that the removal of PAHs in surfactant-enhanced systems depended upon the micelles formation, whereas in the HPCD-enhanced systems, it depended upon the sterioselective diffusion of the PAHs to the nonpolar cavity of the HPCD. Overall, this study showed that the contaminant removal in soil flushing systems depends on the flushing solution affinity and selectivity toward the target contaminant and the existing hydraulic gradient condition.

This study deals with the remediation of clayed soils contaminated with hydrophobic organic compounds (HOCs) by using electrokinetics (EK) coupled with a permeable reactive barrier (PRB) technique. Hexachlorobenzene (HCB) and kaolin were selected as the typical HOCs and clay, respectively, and microscale $\mathrm{Cu}/\mathrm{Fe}$ particles were synthesized as the functional materials of PRB. Furthermore, Triton X-100 was used as the solubility-enhanced agent to promote the movement and removal of contaminants. Four bench-scale EK tests were conducted with or without $\mathrm{Cu}/\mathrm{Fe}$ PRB. The results reveal that coupling EK with $\mathrm{Cu}/\mathrm{Fe}$ PRB greatly promoted the overall removal of HCB from soil, compared with EK alone. Other than the electroosmotic flow, the adsorption/reductive degradation by $\mathrm{Cu}/\mathrm{Fe}$ PRB may also be responsible for the HCB removal (over 82% of HCB passing by the PRB was removed). Our investigation suggests that the integration of EK and $\mathrm{Cu}/\mathrm{Fe}$ PRB is of great promise to promote the application of EK technique in the remediation of HOC-contaminated soils.

Determination of contaminant interaction with soil is essential for predicting its fate in geoenvironments. Several studies have appraised contaminant fate prediction based on the results of single contaminant-soil interaction. Although a few studies deal with multiple contaminant-soil interaction, its influence on contaminant fate prediction is not discussed in detail. Therefore, the present study will investigate the influence of multiple contaminant-soil interaction results about contaminant fate prediction. The influence of different ranges of contaminant concentration on fate prediction has also been studied. Results indicate that contaminant fate prediction based on multiple contaminant-soil interaction is higher than single contaminant interaction. Contaminant fate prediction based on linear isotherms is influenced by the range of concentration for single and multiple contaminants. For a lower range of concentration, contaminant fate prediction based on the Freundlich isotherm is greater than that of Langmuir, and vice versa for higher concentration ranges. Furthermore, the difference in contaminant fate prediction based on Freundlich and Langmuir nonlinear isotherms decreases with an increase in concentration.

The sorption of 12 neutral organic compounds (NOCs) contained in mixtures to both hydrophilic mineral surfaces [uncoated, iron (hydr)oxide-coated, and aluminum (hydr)oxide-coated sands] and humic acid-mineral complexes with different fractions of organic carbon, $f_{oc}$ ($=0.051$, 0.119, and 0.221%) was evaluated using batch equilibrium sorption tests. The 12 compounds included six nonpolar NOCs (1,2,4-trichlorobenzene; 1,4-dichlorobenzene; chlorobenzene; $m$-xylene; toluene; and benzene) and six polar NOCs (2,4-dimethyl phenol; $p$-cresol; phenol; 2-hexanone; 2-butanone; and acetone). For sorption of the nonpolar NOCs, mixture effects (e.g., competitive or cooperative effects) were not statistically significant (95% confidence level) regardless of changes in mixture composition or type of sorbent. In contrast, mixture effects for some of the polar NOCs (2,4-dimethyl phenol; $p$-cresol; phenol; and 2-hexanone) were statistically significant and depended on both the mixture compositions (concentration and polarity of the NOC) and the type of sorbent (availability of hydrophilic interaction sites). These results suggest that direct experimental measurement of sorption coefficients for polar NOCs contained in mixtures may be required to reduce errors in predicting the fate and transport of the polar NOCs and the risk posed to potential receptors.

A media developed in the laboratory by applying a coating of iron and manganese to a quartz sand surface, known as mixed-oxide-coated sand (MOCS) is tested in this paper for arsenic(III) adsorption from water. The media has shown alkali resistance. Scanning electron microscope (SEM) images of MOCS have shown uneven and chapped morphology throughout the coated sand surfaces, whereas images of plane quartz sand are flat. The pattern of X-ray diffractograms of plane sand, MOCS, and arsenic-loaded MOCS are similar. The MOCS has an XRD pattern like typical crystalline material. The Langmuir and Freundlich isotherm equations could be used to describe the partitioning behavior of systems at different pH and media doses. Studies on pH effect have shown maximum As(III) removal near neutral pH. The batch kinetic studies data were tested using active available site (AAS) and chemical reaction rate models. The rate constants, equilibrium sorption capacity, and normalized standard deviations were calculated for all models. The tested models almost accurately predict the sorption capacity with respect to time for the whole range of data points. However, sorption kinetic data were better correlated using an AAS equation model based on normalized standard deviation. The results of desorption studies using different regenerants show that 0.2 M NaOH has high desorption efficiency compared with other regenerants for desorption of As(III) from MOCS. The impact of pH on desorption of arsenic(III) was also studied, and results have shown that high pH values show a significant reduction in quantity of arsenic(III) as compared with lower pH values. At pH 11.1, the percentage of arsenic extraction was highest from MOCS media.

One of the important issues in simulation of contaminant transport in the subsurface is how to quantify the hydraulic properties of soil that are randomly variable in space because of soil heterogeneity. Stochastic approaches have the potential to represent spatially variable parameters, making them an appropriate tool to incorporate the effects of the spatial variability of soil hydraulic properties on contaminant fate. This paper presents development and application of a numerical model for simulation of advective and diffusive-dispersive contaminant transport using a stochastic finite-element approach. Employing the stochastic finite-element method proposed in this study, the response variability is reproduced with a high accuracy. Comparison of the results of the proposed method with the results obtained using the Monte Carlo approach yields a pronounced reduction in the computation cost while resulting in virtually the same response variability as the Monte Carlo technique.