Environ. Sci. Technol. 2006, 40, 3402-3408
Impact of Aging on the Formation of Bound Residues after Peroxidase-Mediated Treatment of 2,4-DCP Contaminated Soils M OÄ N I C A P A L O M O A N D A L O K B H A N D A R I * Department of Civil Engineering, Kansas State University, Manhattan, Kansas 66506-5000
This study evaluated the impact of solute-soil contact time on the formation of “bound” residue in two surface soils exposed to solutions containing 2,4-dichlorophenol (DCP) or DCP polymerization products (DPP). DPP was generated by horseradish peroxidase (HRP) mediated oxidative polymerization of 14C-labeled DCP in the soil slurry. Soils were preloaded with DCP or DPP for durations ranging from 2 h to 84 days. Bound residue was described as solute that was resistant to methanol extraction. Alkali extractions were conducted to estimate the 14C-activity associated with the humic acid, fulvic acid, and humin/mineral components of the soil. Changes in the distribution of the preloaded 14C-DCP and 14C-DPP were observed as a function of the solute-soil contact time. Results suggest that an assumption of sorption equilibrium based solely on the achievement of constant aqueous- or solid-phase solute concentrations can lead to erroneous conclusions about the establishment of true thermodynamic sorption equilibrium. This work also illustrated that (i) significant “irreversible” binding of phenolic contaminants to soils can be achieved during peroxidase-mediated treatment; and (ii) the “aging” process can lead to greater bound-residue formation over time.
Introduction Solute-soil contact time can have a significant impact on the transport, fate, and bioavailability of organic contaminants. Physicochemical phenomena such as the movement of solute molecules along tortuous pathways within the soil matrix, the retarded solute mass transfer associated with intraparticle environments, and the physical reorientation of sorption domains in soil organic matter (SOM) can produce contact-time-dependent changes in the phase-distribution behaviors of organic solutes. Several studies have documented that longer contact times between organic solutes and geosorbents lead to an “aging” phenomena characterized by higher solute retention (1-7), reduced desorption and extractability with organic solvents (7-10), and greater contaminant sequestration or binding to soil components (1, 6-8). The enhanced retention of solute has been shown to depend on the composition of the geosorbent, such as its mineral and organic characteristics and content (2, 10, 12, 13), solute properties (2, 10-12), and solution chemistry (2, 9, 11, 12). Soil organic matter is among the most important factors responsible for contact-time-dependent attenuation in solute * Corresponding author e-mail:
[email protected]; phone: (785) 532-1578; fax: (785) 532-7717. 3402
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 10, 2006
extractability perhaps due to its heterogeneous structure and the presence of reactive functional groups (14, 16). SOM broadly consists of three components: fulvic acids (FA), humic acids (HA), and humin. The operational definition of SOM fractions is based on their solubility under acidic or basic conditions. While FA is soluble at any pH, HA is soluble only under basic conditions and humin is insoluble at all pHs (16). While FA is composed of relatively small oligomeric molecules containing predominantly polar and ionizable groups, HA consists of larger and more hydrophobic macromolecules that are held together by weak interactions (16). An accurate correlation between the sorption behavior of organic solutes and the chemical or structural properties of HA or FA is often difficult due to the lack of a unique composition of these SOM components. In a recent study, the sorption behavior of phenanthrene was reported to be a function of the geosorbent’s oxygen/carbon (O/C) ratio (15). The O/C ratio of humic substances has been observed to decrease in the order FA > HA > humin (17), and several studies have reported that organic chemicals tend to associate preferentially with FA, followed by humin and HA (8, 1822). Soil solution chemistry, specifically pH, can control the reactivity of SOM functional groups as well as the speciation of the solute. For example, the retention of phenolic solutes on SOM is significantly enhanced at pHs that allow the solute to be protonated (2, 9-12). Phenols are less mobile in soils at pHs at least one unit below their pKa values. At pHs where phenolic solutes can exist in both the ionized and neutral forms, determination of interactions between the solute molecules and the soil/sediment particles is more difficult to predict. Among chlorinated phenols, 2,4-dichlorophenol (DCP) falls into this category because it has a pKa of 7.69. At near-neutral pHs (pH 6-8), between 33% and 98% of DCP may be ionized in solution. In its ionized form, DCP is significantly more water-soluble and has the potential to migrate rapidly in soils and groundwater contaminating pristine aquifer zones and expanding the risk to various ecoreceptors. A large number of phenolic contaminants are structural analogues of naturally occurring precursors of soil organic matter. The transformation of such phenolic precursors into SOM components is often mediated by extracellular enzymes such as peroxidases and phenoloxidases. Several of these enzymes, including horseradish peroxidase (HRP), have been proposed for use in the in-situ stabilization of phenolic pollutants at contaminated sites (20, 21, 24-29). HRPmediated oxidative-coupling reactions have the potential to transform toxic phenols in groundwater into sparingly soluble polymeric species, thereby arresting their subsurface mobility and reducing the potential for the offsite migration of contaminant plumes. The enzyme-mediated process results in bound-residue formation due to precipitation of the polymers, sorption of polymerization products to soil components, and cross-coupling between the solute or its polymerization products and reactive SOM moieties (20, 21, 28-30). Although the protonated form of DCP is considered a better substrate for HRP, ionized phenols also readily undergo HRP-mediated oxidative polymerization, (31). Several studies have explored bound-residue formation in geosorbents from a perspective of physical processes such as sequestration or aging (4, 14) or from a standpoint of biochemical processes such as enzyme-mediated oxidative coupling (25, 26, 28-30, 32, 33). Ahn et al. (34) recently reported the impact of contact times ranging from 0 to 14 days on laccase-mediated contaminant immobilization in a 10.1021/es052265p CCC: $33.50
2006 American Chemical Society Published on Web 05/09/2006
DCP contaminated soil. Our study explores a more likely scenario where enzymatic transformation of phenols in the soil solution is followed by long aging periods of up to 3 months. Results from this study demonstrate the impact of aging on bound-residue formation via stabilization processes that are often considered analogous to natural humus formation (25, 26).
Experimental Section Soils. The two soils used in this study were collected from adjacent agricultural and wooded areas (7). Soils were sterilized using a conventional multiple autoclaving procedure. Selected characteristics of the soils are summarized in Table 1 of the Supporting Information. A detailed soil characterization is available elsewhere (7, 30). Chemicals. The target chemical used in this study was 2,4-dichlorophenol. Unlabeled and U-ring 14C-DCP (specific activity 20.9 mCi/mmol) were purchased from Sigma Aldrich (St. Louis, MO.) DCP solutions were prepared in pH 7 buffer consisting of 2.82 mM K2HPO4, 1.8 mM KH2PO4, and 500 mg/L of sodium azide (NaN3). NaN3 was added to prevent biological growth over long contact periods and did not appear to impact HRP activity. Precise volumes of 14C-DCP were added to nonlabeled DCP solutions to obtain the working solutions. Radioactivity was enumerated by transferring 250 µL aliquots of the solution into scintillation vials containing 5 mL of scintillation cocktail (Fisher Scientific Scinti-Safe 50%) and counting in a liquid scintillation counter (LSC, 6500 Beckman). Solid samples were combusted at 905 °C in an OX-500 biological oxidizer (R. J. Harvey Instruments, Hillsdale, NJ). The resulting 14CO2 was trapped in a 14C scintillation cocktail (R. J. Harvey Instruments, Hillsdale, NJ) and counted using LSC. Quench and luminescence in samples were corrected using the LSC’s H-number and LumEX features and the 14C-activity in solution was expressed as disintegrations per minute (dpm). HRP (Type II, RZ 2:2) and hydrogen peroxide (30% w/w, 8.2 M) were purchased from Sigma Aldrich and used without further purification. One activity unit (AU) of HRP was defined by Sigma Aldrich as the amount of enzyme that forms 1.0 mg of purpurogallin from pyrogallol in 20 s at pH 6.0 at 20 °C. HPLC grade methanol was purchased from Fisher Scientific. DPP were generated “in situ” by adding known activities of HRP and H2O2 to soil slurries containing DCP. DPP was quantified as molar equivalents of the parent DCP. Sodium hydroxide was used to extract humic and fulvic acids. Preloading. DCP and DPP were allowed to sorb to the two soils for contact times ranging from 0.083 days (2 h) to 84 days. One and half grams of soil was placed in triplicate 10-mL glass centrifuge tube reactors and contacted with approximately 8.6 mL of DCP solution. The completely mixed batch reactors (CMBRs) were maintained at room temperature (22 ( 1 °C) and contained the following: (i) soil + DCP solution, (ii) soil + DCP solution + HRP + H2O2, and (iii) DCP solution only. The H2O2 concentrations used in this and previous studies had no impact on solute sorption or binding in the absence of HRP (28). The last set of CMBRs was included to monitor solute losses. Preloading was conducted with five initial aqueous phase DCP concentrations (5, 10, 50, 100, and 500 µM) and was allowed to progress for solute-soil contact periods of 0.083, 1, 7, 14, 28, 56, and 84 days. Enzyme-amended soil slurries contained 1.2 AU/ mL of HRP. DCP polymerization in these CMBRs was initiated by adding peroxide at a H2O2/DCP molar ratio of 1.2. The H2O2 dose was based on results of preliminary experiments that revealed an optimum H2O2/DCP molar ratio of 1.2:1 for >98% DCP polymerization in aqueous solutions. The reactors were continuously mixed in an end-over-end tumbler. At the end of the preloading period, the contents of the CMBRs
were centrifuged at 550g, and the residual aqueous phase 14 C-activity in the supernatant was measured. A simple mass balance of the 14C-activity in each reactor was used to determine the solid-phase solute concentration. Extractions. The residual supernatant in the CMBRs was discarded and the soil was subjected to a single extraction with the pH 7 buffer followed by three sequential extractions with methanol. These extractions were sufficient to reduce the radioactivity in the extract to a background level of 50 dpm. Extracted soils were air-dried inside a fume hood. Thereafter, the contents of three replicate CMBRs were combusted at 905 °C. The 14CO2 recovered was quantified on the LSC and was attributed to preloaded solute that was resistant to solvent extractions (14CNX). The methanolextracted soil in the remaining triplicate CMBRs was subjected to a series of sequential 24-hr extractions with deoxygenated 0.1 N NaOH until the supernatant became visibly clear (22). The contents of each CMBR were centrifuged at 550g and the supernatant containing HA and FA was transferred into a 50 mL glass centrifuge tube for storage. The alkali-extracted soil was air-dried inside the fume hood and combusted to determine the radioactivity associated with the HM component (14CHM). Supernatant stored in the 50 mL centrifuge tubes was acidified to pH 2 with H2SO4 to precipitate HA. A period of 24 h was allowed for precipitate formation after which the suspension was centrifuged and the supernatant containing FA was removed and discarded. Direct measurements of the 14C-activity associated with FA (14CFA) were not performed due to interference caused by color in the solution. The precipitated HA was air-dried and combusted to estimate 14C-activity associated with HA (14CHA). 14C FA was determined by the simple mass balance described in eq 1. 14
CFA ) 14CNX - 14CHM - 14CHA
(1)
Statistical Analysis. Data were analyzed for statistical significance using analysis of variance (SAS version 6.1). A GLM procedure with a confidence level of R ) 0.05 was employed. The statistical analysis was performed using a completely randomized design and a factorial treatment with 4 factors: (i) solute type (DCP and DPP); (ii) soil type (woodland and agricultural soil); (iii) solute concentrations (5, 10, 50, 100, and 500 µM); and (iv) soil-solute contact times (0.083, 1, 7, 14, 28, and 84 days).
Results and Discussion The main differences between the two soils used in this study were their overall SOM content and SOM characteristics including HA, FA, and humin content. At 6.2%, the SOM content of woodland soil was nearly twice that of the agriculture soil (3.4%). While woodland soil contained significantly more HA and humin (0.73% and 4.35%, respectively) than the agricultural soil (0.33% and 1.95%), the mass fraction of FA (1.12%) was identical in both soils. The aromatic carbon and aromatic hydrogen contents were
