A. Lynn Roberts

Longevity of granular iron in groundwater treatment processes: changes in solute transport properties over time.

Although progress has been made toward understanding the surface chemistry of granular iron and the mechanisms through which it attenuates groundwater contaminants, potential long-term changes in the solute transport properties of granular iron media have until now received relatively little attention. As part of column investigations of alterations in the reactivity of granular iron, studies using tritiated water (3H(2)O) as a conservative and non-partitioning tracer were periodically conducted to independently isolate transport-related effects on performance from those more directly related to surface reactivity. Hydraulic residence time distributions (HRTDs) within each of six 39-cm columns exposed to bicarbonate solutions were obtained over the course of 1100 days of operation. First moment analyses of the data revealed generally modest increases in mean pore water velocity (v) over time, indicative of decreasing water-filled porosity. Gravimetric measurements provided independent estimates of water-filled porosity that were initially consistent with those obtained from 3H(2)O tracer tests, although at later times, porosities derived from gravimetric measurements deviated from the tracer test results owing to mineral precipitation. The combination of gravimetric measurements and 3H(2)O tracer studies furnished estimates of precipitated mineral mass; depending on the assumed identity of the predominant mineral phase(s), the porosity decrease associated with solute precipitation amounted to 6-24% of the initial porosity. The accumulation of mineral and gas phases led to the formation of regions of immobile water and increased spreading of the tracer pulse. Application of a dual-region transport model to the 3H(2)O breakthrough curves revealed that the immobile water-filled region increased from initially negligible values to amounts ranging between 3% and 14% of the total porosity in later periods of operation. For the aged columns, mobile-immobile mass transfer coefficients (k(mt)) were generally in the range of 0.1-1.0 day(-1) and reflected a slow exchange of 3H(2)O between the two regions. Additional model calculations incorporating sorption and reaction suggest that although changes in HRTD can have an appreciable effect on trichloroethylene (TCE) transformation, the effect is likely to be minor relative to that stemming from passivation of the granular iron surface.

Trace determination of pharmaceuticals and other wastewater-derived micropollutants by solid phase extraction and gas chromatography/mass spectrometry.

The presence of pharmaceuticals and other wastewater-derived micropollutants in surface and groundwaters is receiving intense public and scientific attention. Yet simple GC/MS methods that would enable measurement of a wide range of such compounds are scarce. This paper describes a GC/MS method for the simultaneous determination of 13 pharmaceuticals (acetaminophen, albuterol, allopurinol, amitriptyline, brompheniramine, carbamazepine, carisoprodol, ciclopirox, diazepam, fenofibrate, metoprolol, primidone, and terbinafine) and 5 wastewater-derived contaminants (caffeine, diethyltoluamide, n-butylbenzene sulfonamide, n-nonylphenol, and n-octylphenol) by solid phase extraction (SPE) and derivatization with BSTFA. The method was applied to the analysis of raw and treated sewage samples obtained from a wastewater treatment plant located in the mid-Atlantic United States. All analytes were detected in untreated sewage, and 14 of the 18 analytes were detected in treated sewage.

Chlorine Monoxide (Cl2O) and Molecular Chlorine (Cl2) as Active Chlorinating Agents in Reaction of Dimethenamid with Aqueous Free Chlorine

HOCl is often assumed to represent the active oxidant in solutions of free available chlorine (FAC). We present evidence that Cl(2)O and Cl(2) can play a greater role than HOCl during chlorination of the herbicide dimethenamid. Reaction orders in [FAC] were determined at various solution conditions and ranged from 1.10 +/- 0.13 to 1.78 +/- 0.22, consistent with the concurrent existence of reactions that appear first-order and second-order in [FAC]. Solution pH, [Cl(-)], [FAC], and temperature were systematically varied so that the reactivity and activation parameters of each FAC species could be delineated. Modeling of kinetic data afforded calculation of second-order rate constants (units: M(-1) s(-1)) at 25 degrees C: k(Cl2O) = (1.37 +/- 0.17) x 10(6), k(Cl2) = (1.21 +/- 0.06) x 10(6), and k(HOCl) = 0.18 +/- 0.10. Under conditions typical of drinking water chlorination, Cl(2)O is the predominant chlorinating agent of dimethenamid. To the extent that Cl(2)O represents the active species in reactions with other disinfection byproduct (DBP) precursors, the influence of [FAC] and pH on DBP precursor reaction rates is different than if HOCl were the principal oxidant. Moreover, these findings call into question the validity of apparent rate constants (k(app)) commonly reported in the chlorination literature.

Neutral degradates of chloroacetamide herbicides: Occurrence in drinking water and removal during conventional water treatment

Treated drinking water samples from 12 water utilities in the Midwestern United States were collected during Fall 2003 and Spring 2004 and were analyzed for selected neutral degradates of chloroacetamide herbicides, along with related compounds. Target analytes included 20 neutral chloroacetamide degradates, six ionic chloroacetamide degradates, four parent chloroacetamide herbicides, three triazine herbicides, and two neutral triazine degradates. In the fall samples, 17 of 20 neutral chloroacetamide degradates were detected in the finished drinking water, while 19 of 20 neutral chloroacetamide degradates were detected in the spring. Median concentrations for the neutral chloroacetamide degradates were approximately 2-60ng/L during both sampling periods. Concentrations measured in the fall samples of treated water were nearly the same as those measured in source waters, despite the variety of treatment trains employed. Significant removals (average of 40% for all compounds) were only found in the spring samples at those utilities that employed activated carbon.

Reactivity of BrCl, Br2, BrOCl, Br2O, and HOBr Toward Dimethenamid in Solutions of Bromide + Aqueous Free Chlorine

HOBr, formed via oxidation of bromide by free available chlorine (FAC), is frequently assumed to be the sole species responsible for generating brominated disinfection byproducts (DBPs). Our studies reveal that BrCl, Br(2), BrOCl, and Br(2)O can also serve as brominating agents of the herbicide dimethenamid in solutions of bromide to which FAC was added. Conditions affecting bromine speciation (pH, total free bromine concentration ([HOBr](T)), [Cl(-)], and [FAC](o)) were systematically varied, and rates of dimethenamid bromination were measured. Reaction orders in [HOBr](T) ranged from 1.09 (±0.17) to 1.67 (±0.16), reaching a maximum near the pK(a) of HOBr. This complex dependence on [HOBr](T) implicates Br(2)O as an active brominating agent. That bromination rates increased with increasing [Cl(-)], [FAC](o) (at constant [HOBr](T)), and excess bromide (where [Br(-)](o)>[FAC](o)) implicate BrCl, BrOCl, and Br(2), respectively, as brominating agents. As equilibrium constants for the formation of Br(2)O and BrOCl (aq) have not been previously reported, we have calculated these values (and their gas-phase analogues) using benchmark-quality quantum chemical methods [CCSD(T) up to CCSDTQ calculations plus solvation effects]. The results allow us to compute bromine speciation and hence second-order rate constants. Intrinsic brominating reactivity increased in the order: HOBr ≪ Br(2)O < BrOCl ≈ Br(2) < BrCl. Our results indicate that species other than HOBr can influence bromination rates under conditions typical of drinking water and wastewater chlorination.

Calculated One- and Two-Electron Reduction Potentials and Related Molecular Descriptors for Reduction of Alkyl and Vinyl Halides in Water

One- and two-electron reduction potentials (E1 and E2 values) were calculated from published thermodynamic data for 39 halogenated C1 and C2 compounds, including many commonly encountered groundwater contaminants. Because reductive dehalogenation is an important pathway for their destruction under anaerobic conditions, information concerning the relevant reduction potentials may be useful for assessing the thermodynamic feasibility of a particular reaction, as well as in developing linear free energy relationships (LFERs) or other quantitative structure-activity relationships (QS ARs) that may enable prediction of rates of transformation. E1 values were calculated assuming a stoichiometry corresponding to dissociative electron transfer, which produces a carbon-centered radical and a halide ion. E2 values were calculated for both hydrogenolysis and reductive β-elimination reactions. Uncertainties in the thermodynamic data for the organohalides under consideration may introduce substantial uncertainty in the resulting E1 values. Hence, relationships between calculated E1 values and various surrogate parameters were also investigated. E1 values were correlated with lowest unoccupied molecular orbital (LUMO) energies and carbon-halogen homolytic bond dissociation energies (BDE values), which were computed via density functional theory. Correlations were also attempted between E1 values and vertical attachment energies (VAE values), the latter representing experimental measures of the ease of reduction of a molecule to a radical anion in the gas phase. These alternate descriptors may provide a means for estimating E1. Additional studies will need to be undertaken to establish which descriptor best correlates with reactivity in environmental reductive dehalogenation.

Determination of pharmaceuticals and antiseptics in water by solid-phase extraction and gas chromatography/mass spectrometry: analysis via pentafluorobenzylation and stable isotope dilution

A sensitive yet robust analytical method is presented for the simultaneous determination of 12 human pharmaceuticals (valproic acid, phenytoin, ibuprofen, gabapentin, acetaminophen, gemfibrozil, naproxen, ketoprofen, secobarbital, phenobarbital, 5-fluorouracil, and diclofenac) and 6 antiseptics (biosol, biphenylol, p-chloro-m-cresol, p-chloro-m-xylenol, chlorophene, and triclosan). The method employs solid-phase extraction (SPE) followed by a novel pentafluorobenzylation using a mixture of acetontrile/water (1/1, v/v). The method is simple to perform (derivatization can be completed in a single test tube) and eliminates the need for any solvent/SPE cartridge drying or blow-down. It affords excellent resolution, high sensitivity and reproducibility, and freedom from interference even for matrices as complex as untreated sewage. The method was applied to the analysis of sewage samples using 15 isotopically labeled surrogates, which resulted in the detection of 10 of the 12 pharmaceuticals and all of the antiseptics sought. Ten of 15 surrogates were synthesized from pure analytes by a simple H-D exchange reaction employing D2O and D2SO4. Measured recoveries were sensitive to matrix effects and varied substantially among analytes, indicative of the limitations associated with using a single surrogate standard.

Reactivity of Alkyl Polyhalides toward Granular Iron: Development of QSARs and Reactivity Cross Correlations for Reductive Dehalogenation

Attempts to develop quantitative structure-activity relationships (QSARs) for reductive dehalogenation by granular iron have been hindered by the unavailability of high quality predictor variables, have included relatively few compounds, and on occasion have relied on data lacking internal consistency. We herein investigate the reduction of 24 alkyl polyhalides by granular iron and the better-defined, homogeneous reductants Cr(H(2)O)(6)(2+) and an Fe(II) porphyrin. QSARs were constructed with a new set of computationally derived gas phase homolytic carbon-halogen bond dissociation energies and solvated one-electron reduction potentials determined using a quantum chemistry composite method (G3MP2). Reactivity cross correlations between reductant systems were also developed. Reactivity trends were generally consistent among all reductants and revealed pronounced structural influences. Compounds reduced at C-Br were orders of magnitude more reactive than analogues reduced at C-Cl; the number and identity of α- (Br ∼ Cl > CH(3) > F > H) and β-substituents (Br > Cl) also influenced reactivity. Nonlinearities encountered during QSAR and cross correlation development suggest that reactions of highly halogenated alkyl polyhalides with granular iron are limited by mass transfer, as supported by estimates of mass transfer coefficients. For species not suspected to exhibit mass transfer limitations, reasonably strong cross correlations and comparable substituent effects are consistent with dissociative electron transfer as the rate-determining step.

Applications of surface analysis in the environmental sciences: dehalogenation of chlorocarbons with zero-valent iron and iron-containing mineral surfaces

Halogenated organic compounds are common pollutants in groundwater. Consequently, there is widespread interest in understanding the reactions of these compounds in the environment and developing remediation strategies. One area of ongoing research involves the reductive dechlorination of organohalides with zero-valent metals or metal sulfide minerals. These processes have been studied almost exclusively from the perspective of the aqueous phase. In this paper we illustrate the utility of surface analysis techniques, including electron spectroscopies, vibrational spectroscopies, and atomic force microscopy in elucidating the roles played by the surface. A dual analysis approach to the study of reductive dechlorination, combining traditional solution phase analysis with surface analytical techniques, also is demonstrated using a liquid cell coupled to an ultrahigh vacuum surface analysis chamber.

Interaction of abiotic and microbial processes in hexachloroethane reduction in groundwater

In order to gain insight into mechanisms of hexachloroethane reduction, hexa- and pentachloroethane transformation rates were measured in anaerobic groundwater samples. For samples spiked with pentachloroethane, disappearance of pentachloroethane was accompanied by tetrachloroethylene production. Transformation rates were similar in unpoisoned and in HgCl2-poisoned samples, and rates were within ±20% of predictions based on measured pH and second-order dehydrochlorination rate constants determined in clean laboratory systems, indicating that the fate of pentachloroethane in this system is dominated by abiotic reactions. No hexachloroethane transformation was observed in HgCl2-poisoned samples, whereas in unpoisoned samples, hexachloroethane disappearance was accompanied by production of tetrachloroethylene as well as traces of pentachloroethane. Although only minor amounts of pentachloroethane accumulated, as much as 30% of the hexachloroethane transformation pathway proceeds via a pentachloroethane intermediate. This suggests that the microbial reduction of hexachloroethane proceeds at least in part through a free-radical mechanism. To the extent that hexachloroethane reduction to tetrachloroethylene occurs through a pentachloroethane intermediate, the first step in the sequence, the microbially-mediated step, is the slow step; the subsequent abiotic dehydrohalogenation step occurs much more rapidly.