John T. Coates

Desorption kinetics for selected PCB congeners from river sediments

Abstract Desorption of PCBs from sediment can significantly affect the ultimate fate and effects of PCBs in aquatic systems. Using a gas purging technique to strip soluble and sorbed polychlorinated biphenyls (PCBs) from solutions and sediment suspensions, Henrys law constants, approach to equilibrium, and desorption rate constants for four PCB congeners were measured. Henrys law constants were on the order of 10 −4 m 3 atm mole −1 . Desorption rate constants measured for a predominantly kaolinitic, low-organic carbon sediment were on the order of 0.03–0.1 days −1 . In contrast, desorption rate constants measured for a sediment composed of montmorillonite with a 3% organic carbon content were on the order of 0.009–0.04 days −1 . Desorption data suggest that equilibration times for PCBs with low chlorine content are on the order of six weeks, and months to years for PCBs with a significantly higher chlorine content.

Field test of high molecular weight alcohol flushing for subsurface nonaqueous phase liquid remediation

A pilot scale field test of non-aqueous phase liquid (NAPL) removal using high molecular weight alcohols was conducted at Operable Unit 1, Hill Air Force Base, Utah. Petroleum hydrocarbons and spent solvents were disposed of in chemical disposal pits at this site, and these materials are now present in the subsurface in the form of a light non-aqueous phase liquid (LNAPL). This LNAPL is a complex mixture of aromatic and aliphatic hydrocarbons, chlorinated solvents, and other compounds. The field experiment was performed in a 5 m by 3 m confined test cell, formed by driving interlocking sheet pile walls through the contaminated zone into an underlying clay. The test involved the injection and extraction of about four pore volumes (1 pore volume=7000 L) of a mixture of 80% tert-butanol and 15% n-hexanol. The contaminants were removed by a combination of NAPL mobilization and enhanced dissolution, and the results of postflood soil coring indicate better than 90% removal of the more soluble contaminants (trichloroethane, toluene, ethylbenzene, xylenes, trimethylbenzene, naphthalene) and 70–80% removal of less soluble compounds (decane and undecane). The results of preflood and postflood NAPL partitioning tracer tests show nearly 80% removal of the total NAPL content from the test cell. The field data suggest that a somewhat higher level of removal could be achieved with a longer alcohol injection.

Cosolvent effects of alcohols on the Henry's law constant and aqueous solubility of tetrachloroethylene (PCE).

The effects of selected cosolvents ethyl alcohol (EtOH), isopropyl alcohol (IPA), and tertbutyl alcohol (TBA) on the Henrys law constant (H) of tetrachloroethylene (PCE) in aqueous solutions were investigated using the static headspace method. Alcohols in solution at a concentration around 20% and above acted as cosolvents increasing the aqueous solubility of PCE, which resulted in lower H values for PCE as compared to the value of H in deionized water. TBA, the most hydrophobic of the three alcohols, exhibited the strongest cosolvent effects, while EtOH had the weakest effects. A ln-linear relationship was observed between H and the volumetric fraction of alcohol added. Investigation of the solubilization of PCE in alcohol solutions confirmed the cosolvent trend observed for the three alcohols. A ln-ln relationship was observed between H and the enhanced solubility of PCE at a particular alcohol concentration. It was also observed that the value of H is a function of the enhanced solubility regardless of the type of cosolvent used. The results from this research further define the behavior of PCE in alcohol flooding solutions used in the remediation of PCE contaminated media.

Sorption of humic acids and α-endosulfan by clay minerals

Sorption of alpha-endosulfan by kaolinite andmontmorillonite alone and in the presence of sorbed and dissolved humicacid (HA) was investigated (pH 8 and 25oC). Three types of HA, Elliotsoil HA (EHA), Peat HA (PHA), and Summit Hill HA (SHHA), were used torepresent typical humic substances found in soils. For sorption of HA byeither mineral, Freundlich sorption coefficient (Kf) values appeared todecrease in the order of EHA>PHA>SHHA, which followedincreasing polarity (expressed as the O/C atomic ratio) and decreasingpercent-carbon content. For both clays, sorption of alpha-endosulfan bythe HA mineral complex was greater than for sorption by the clay alone.Sorption of alpha-endosulfan by the HA mineral complexes followed thesame order as the Kf of the HAs (EHA>PHA>SHHA). Based on theamount of HA adsorbed by each mineral, organic carbon partitioncoefficients (KOC) were determined for sorption of alpha-endosulfan bytwo of the HA mineral complexes. The value of KOC for alpha-endosulfansorption was greater for kaolinite EHA than kaolinite SHHA. However, theopposite trend was found with the montmorillonite HA complexes.Montmorillonite appeared to sorb alpha-endosulfan and/or HA with higheraffinity than kaolinite, which likely is due to its 2:1 layer structureand higher surface area. Sorption of endosulfan diol, a hydrolysisproduct, by the minerals was much less than the parentpesticide.

Pu(V) transport through Savannah River Site soils - an evaluation of a conceptual model of surface- mediated reduction to Pu (IV).

Over the last fifteen years the Savannah River Site (SRS) in South Carolina, USA, was selected as the site of three new plutonium facilities: the Mixed Oxide Fuel Fabrication Facility, Pit Disassembly and Conversion Facility, and the Pu Immobilization Plant. In order to assess the potential human and environmental risk associated with these recent initiatives, improved understanding of the fate and transport of Pu in the SRS subsurface environment is necessary. The hypothesis of this study was that the more mobile forms of Pu, Pu(V) and Pu(VI), would be reduced to the less mobile Pu(III/IV) oxidation states under ambient SRS subsurface conditions. Laboratory-scale dynamic flow experiments (i.e., column studies) indicated that Pu(V) was very mobile in SRS sediments. At higher pH values the mobility of Pu decreased and the fraction of Pu that became irreversibly sorbed to the sediment increased, albeit, only slightly. Conversely, these column experiments showed that Pu(IV) was essentially immobile and was largely irreversibly sorbed to the sediment. More than 100 batch sorption experiments were also conducted with four end-member sediments, i.e., sediments that include the chemical, textural, and mineralogical properties likely to exist in the SRS. These tests were conducted as a function of initial Pu oxidation state, pH, and contact time and consistently demonstrated that although Pu(V) sorbed initially quite weakly to sediments, it slowly, over the course of <33 days, sorbed very strongly to sediments, to approximately the same degree as Pu(IV). This is consistent with our hypothesis that Pu(V) is reduced to the more strongly sorbing form of Pu, Pu(IV). These studies provide important experimental support for a conceptual geochemical model for dissolved Pu in a highly weathered subsurface environment. That is that, irrespective of the initial oxidation state of the dissolved Pu introduced into a SRS sediment system, Pu(IV) controls the environmental transport within a couple weeks and Pu strongly binds to the sediment, limiting its mobility.

Enantiomeric composition of chiral polychlorinated biphenyl atropisomers in dated sediment cores

Enantiomer fractions (EFs) of seven chiral polychlorinated biphenyls (PCBs) were measured in dated sediment cores of Lake Hartwell (SC, USA) and Lake Ontario (USA) to detect, quantify, and gain insight regarding microbial reductive dechlorination of PCBs in lake sediments with high and low concentrations, respectively. Lake Hartwell sediments had high total PCBs (5-60 microg/g), with significantly nonracemic EFs that generally were consistent with those from previous laboratory microcosm reductive dechlorination experiments using sediments from these sites. Thus, stereoselective reductive dechlorination had occurred in situ, including at total PCB concentrations of less than the threshold of approximately 30 to 80 microg/g suggested as being necessary for reductive dechlorination. Enantiomer fractions of PCBs 91, 95, 132, and 136 in Lake Hartwell cores were significantly correlated both with concentrations of those individual congeners and with total PCB concentration for some sites. This result indicates that enantioselective microbial dechlorination activity increases with higher concentrations within sediments for these congeners. Enantiomer composition reversed with depth for PCBs 91, 132, and 176, suggesting that multiple microbial populations may be present within the same core that are enantioselectively dechlorinating PCBs. Such observations indicate that concentration and time are not the only factors affecting biotransformation, complicating prediction of enantioselectivity. Comparison of EFs with dates suggested biotransformation half-lives of approximately 30 years, which is on the same time scale as sequestration by burial. In contrast, Lake Ontario sediments (maximum total PCBs, 400 ng/g) had racemic or near-racemic amounts of most congeners throughout the core profile, which is consistent with achiral indicators suggesting no microbial biotransformation within Lake Ontario sediments. Thresholds for reductive dechlorination may exist, but they would be at concentrations of less than 30 to 80 microg/g.

Evaluation of a rapid technique for measuring actinide oxidation states in a ground water simulant

A system using an ion chromatograph coupled to a flow-cell scintillation detector for rapidly measuring the oxidation states of actinides at low concentrations (<10–6M) in aqueous solutions was evaluated. The key components of the system are a cation–anion separation column (Dionex, CS5) and a flow cell detector with scintillating cerium activated glass beads. The typical procedure was to introduce a 0.5 ml aliquot of sample spiked with actinides in the +III to +VI oxidation states into a 5 ml sample loop followed by 4 ml of synthetic groundwater simulant. Separation was achieved at a flow rate of 1 ml/min using an isocratic elution with oxalic, diglycolic, and nitric acids followed by distilled water. Tests were first conducted to determine elution times and recoveries for an acidic solution (pH ≈ 2) and a ground water simulant (pH ≈ 8) containing Am(III), Pu(IV), Th(IV), Pu(V), and U(VI). Then, an analysis was performed using a mixture of Pu(IV), Pu(V), and Pu(VI) in the ground water simulant and compared to results using the DBM extraction technique. Approximate elution times were the same for both the acidic solution and the ground water simulant. These were as follows: Pu(V) at 10 min, Am(III) at 15 min, Pu(IV) at 25 min, Th (IV) at 28 min and U(VI) at 36 min. Recoveries for the acidic solution were quantitative for U(VI) and Th(IV) and exceeded 80% for Am(III). Recoveries for the ground water simulant were quantitative for U(VI), but they were generally not quantitative for Th(IV), Pu(IV), and Am(III). For Th(IV) and Pu(IV), less than quantitative recoveries were attributed to the formation of neutral hydroxides and colloids; for Am(III) they were attributed to insoluble carbonates and/or hydroxycarbonates. When applied to the measurement of plutonium in the ground water simulant, the technique provided showed good agreement with the dibenzoylmethane (DBM) extraction technique, but it could not distinguish between Pu(V) and Pu(VI). This was likely due to the reduction of Pu(VI) to Pu(V) in the sample by the oxalic acid eluent. However, in spite of this limitation, the technique can be used to distinguish between Pu(IV) and Pu(V) in aqueous environmental samples within a pH range of 4 to 8 and an EH range of -0.2 to 0.6 V, the predominance region for Pu(III), (IV), and (V). In addition, this technique can be used to corroborate oxidation state analysis from the dibenzoylmethane (DBM) extraction method for environmental samples.

Distributions of radionuclide sorption coefficients (Kd) in sub-surface sediments and the implications for transport calculations

The effect of the spatial variability of K(d) on calculations of contaminant travel time in the vadose zone was determined. Depth discrete measurements of K(d) were made for a suite of radionuclides ((109)Cd, (57)Co, (60)Co, (85)Sr, (137)Cs, and (88)Y) utilizing a sediment core from the E-Area at the Savannah River Site. The K(d)s were ordered as (85)Sr(2+) < (137)Cs(+) < (109)Cd(2+) < (57)Co(2+) = (60)Co(2+) << (88)Y(3+) and the values generally fell below or near the lowest quartile of values reported in the literature. Correlations were generally weak between soil properties and K(d) values. Most importantly, all of the K(d) distributions could be reasonably approximated as log-normal. Deterministic and stochastic calculations of contaminant travel time to the water table were made. The deterministic calculations were based on each of three conceptual models of the vadose zone: complete stratification (17 strata, each with a different K(d)), two strata (two sections of the vadose zone, each characterized by a single, average K(d)), and unstratified (a single zone with an average K(d)). Stochastic calculations were based on log-normal fits to the K(d) data. The two strata model generally yielded travel times 2x greater than those in the completely stratified model. The unstratified model yielded travel times that were between 3 and 5 times greater than the completely stratified model. The stochastic mean travel times were comparable to those of the two strata model.

Mobilities of uranium species in sedimentary interbed from the Snake River Plain

Laboratory column tests were conducted to characterize the mobility of uranium in a composite of sedimentary interbed samples from the Snake River Plain. The objective of the studies was to investigate the influence on uranium mobility of carbonate, fluoride, sulfate, and hydroxide complexes at pH 8, which is representative of ground water from the region. It was determined that carbonate species were the most dominant followed by hydroxide species. The observations are consistent with the results of speciation modeling. The carbonate species, which had retardation factors between 5 and 10, were considerably more mobile than the hydroxide species, which had retardation factors in excess of 300.