Influence of Aeration on Hydrophobic Organic ... - ACS Publications

Jul 19, 2003 - (28) Lin, C.-H. M. Ph.D. Dissertation. University of California, Los. Angeles, CA, 2001. (29) Brannon, J. M.; Myers, T. E.; Gunnison, D...
1 downloads 0 Views 149KB Size
Download PDF
Environ. Sci. Technol. 2003, 37, 3547-3554

Influence of Aeration on Hydrophobic Organic Contaminant Distribution and Diffusive Flux in Estuarine Sediments

increase in diffusive flux was observed at the low ionic strength site (p < 0.10). This latter result represented approximately a doubling in diffusive flux. In the systems studied, facilitation of TeCB transport across the sedimentwater interface by organic colloids did not appear important.

C H E N - H U N G M I C H A E L L I N , †,‡ J O E L A . P E D E R S E N , §,| A N D I R W I N H . S U F F E T * ,†,§ Department of Environmental Health Sciences and Environmental Science and Engineering Program, University of California, Los Angeles, California 90095

Introduction

Dredging operations, resuspension events during storms, and bioturbation alter the oxic state of estuarine sediments and induce changes in the composition of dissolved and particle-associated natural organic matter. These changes may alter the distribution of hydrophobic organic chemicals (HOCs) in sediments and their diffusive flux across the sediment-water interface. In this study, the impact of aerating anoxic sediments on the distribution and diffusive flux of a model HOC, 2,2′,4,4′-tetrachlorobiphenyl (TeCB), was investigated. Anoxic estuarine sediments collected from three sites along a salinity gradient were used to determine site-specific apparent sorption coefficients for porewater dissolved organic carbon (Kpwdoc) and sediment organic carbon (Koc) under anoxic and oxic conditions. A two-compartment sediment flux model was employed to examine the diffusive flux of TeCB under both oxic states. Aeration of anoxic porewaters resulted in significant decreases in porewater dissolved organic matter (DOMpw) aromaticity as indicated by declines in molar absorptivity at 254 nm (p < 0.005). Aeration also resulted in a 9-13% decrease in DOMpw concentration (p < 0.005) at the two sites exhibiting lower ionic strengths; the high ionic strength site did not exhibit a significant change in DOMpw concentration (p > 0.10). The impact of aeration on TeCB distribution and diffusive flux appeared to be site-specific. Aeration of anoxic sediments induced a significant 1.4 log unit reduction in Kpwdoc at the lowest ionic strength site (p < 0.0005), while sediments from the intermediate ionic strength site exhibited a significant 0.6 log unit increase (p < 0.005). No significant change in sorption to DOMpw was observed for the high ionic strength site (p > 0.10). The sediment displaying the drop in Kpwdoc also exhibited a significant 0.4 log unit drop in Koc (p < 0.01), while the other two sites did not exhibit significant aeration-induced changes in sorption to particle-associated organic matter (p > 0.10). No significant change in diffusive flux was observed for two sites (p > 0.10), while a significant 89-110 mg m-2 yr-1 * Corresponding author phone: (310)206-8230; fax: (310)206-3358; e-mail: [email protected]. † Department of Environmental Health Sciences. ‡ Present address: Enviropro, Inc., Chastworth, CA 91311. § Environmental Science and Engineering Program. | Present address: Molecular and Environmental Toxicology Center and Department of Soil Science, University of Wisconsin, Madison, WI 53706. 10.1021/es026048p CCC: $25.00 Published on Web 07/19/2003

 2003 American Chemical Society

The distribution and mobility of hydrophobic organic compounds (HOCs) in estuarine sediment environments is determined primarily by sorption to natural organic matter associated with sediment particles and present in porewater. Disturbance of bed sediments due to dredging operations, resuspension events during storms, and bioturbation induce changes in porewater chemistry affecting the composition and possibly conformation of porewater dissolved organic matter (DOMpw) and particle-associated sediment organic matter (SOM). These changes in organic matter phases alter HOC association with DOMpw and SOM (1, 2) and thus the chemical concentration gradient across the sediment-water interface. The diffusive flux of HOCs from bed sediments may therefore be altered by changes in porewater chemistry. In estuarine sediment environments, HOCs distribute themselves between the aqueous, DOMpw, and SOM phases. The equilibrium distribution of HOCs between the porewater aqueous and DOM phases can be expressed by an equilibrium distribution coefficient (Kpwdoc) defined as

Kpwdoc )

Cpwdoc

(1)

Cpwaq[DOC]pw

where Cpwdoc and Cpwaq represent the DOMpw-associated and truly dissolved HOC concentrations (µg L-1), and [DOC]pw is the concentration of dissolved organic carbon in the sediment porewater (kgC L-1). In this paper, dissolved organic carbon refers to organic matter passing a 0.7-µm filter and includes both truly dissolved and colloidal organic carbon. Over limited concentration ranges, the equilibrium distribution of HOCs between sediment particles and the interstitial water aqueous phase can be described as a linear process (3). When the mass fraction of organic carbon in sediment particles ( foc) exceeds approximately 0.001, association with organic matter controls HOC sorption to sediment solids (4). The sediment porewater distribution coefficient (Kd) is expressed as

Kd )

Csed Coc foc ) ) focKoc Cpwaq Cpwaq

(2)

and carries units of liters per kilograms. In this expression, Csed and Coc are the concentration of HOC sorbed to sediment solids (µg kgsolid-1) and to sediment organic carbon (µg kgC-1), and Koc is the sediment organic carbon-porewater distribution coefficient. With respect to diffusive flux, chemicals dissolved in the porewater and associated with DOMpw can be considered mobile (5, 6). The mobile phase concentration in the porewater can be defined as

Cpw ) Cpwaq + Cpwdoc ) Cpwaq(1 + Kpwdoc[DOC]pw) (3) Dredging operations, resuspension events during storms, bioturbation, and seasonal migration of the redoxcline can alter the oxic state of estuarine bed sediments. Aeration of VOL. 37, NO. 16, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3547

anoxic sediments results in complex changes to porewater chemistry, shifts in microbial respiration, and alteration of sediment and porewater organic matter. Organic matter may be released into interstitial waters when anoxic sediments are oxidized, but coagulation with precipitating iron (III) hydr(oxides) may reduce DOMpw concentrations (1, 7, 8). Upon aeration, molecular oxygen oxidizes dissolved Fe(II) to insoluble Fe(OH)3 (9, 10). Complexation of Fe(III) with relatively strong organic ligands present in the DOMpw pool enhances the rate of Fe(II) oxidation (10). Oxidation of anoxic sediments enhances the rate of refractory organic matter degradation (11-13). Decreased organic matter mineralization under anoxic conditions leads to a buildup of relatively low molecular weight refractory DOMpw (12) and a concomitant increase in dissolved organic matter flux across the sediment-water interface (14). Dissolved organic matter concentrations in anoxic interstitial waters tend to be higher than those in mixed anoxic/oxic or oxic porewaters. Investigations of Chesapeake Bay interstitial waters revealed distinct compositional differences between DOMpw from anoxic and mixed anoxic/oxic sediments (13). Changes in oxic state also appear to induce conformational changes in humic substances (15). Our laboratory has conducted a series of studies examining changes in HOC distribution brought about by alterations in sediment oxic state (1, 2, 7). Association of a polychlorinated biphenyl (PCB) with sediment porewater DOM was altered as a result of aeration-induced changes in the structure and composition of DOMpw in freshwater (7) and estuarine sediment systems (1, 2). Hunchak-Kariouk et al. (7) observed that aeration of anoxic freshwater sediment porewater effected a 0.7 log unit increase in Kpwdoc. In an independent study conducted in low ionic strength solution, Coates et al. (15) observed a 20% increase in naphthalene binding when reduced humic acids were exposed to O2. Subsequent investigations in our laboratory using estuarine sediments (ionic strength, I ) 0.3-1.2) demonstrated negligible changes to order of magnitude decreases in Kpwdoc upon aeration of anoxic porewaters (1, 2). Only one prior study examined changes in HOC sorption to sediment organic matter induced by altered apparent sediment reduction potential (EH). Pedersen et al. (2) observed decreased PCB association with SOM (0.5 log unit reduction in Koc) when anoxic estuarine sediments were disturbed and aerated (measured apparent EH increased from -9 to +14 mV). These investigators also observed a 0.3 log unit reduction in Kpwdoc upon aeration of anoxic porewater. However, the potential effect of such changes in sorption on the diffusive flux of HOCs across the sediment-water interface has not been extensively investigated. The objectives of this study were (i) to experimentally examine the effect of change in oxic state on the distribution of a model HOC, 2,2′,4,4′-tetrachlorobiphenyl (TeCB), in estuarine sediments and (ii) to model the effect of such change on TeCB diffusive flux across the sediment-water interface. The sediments used in this study were collected along a transect through an estuary representing the salinity gradient encountered by sediments during transport from riverine to marine environments in successive resuspension and deposition events. A two-compartment sediment flux model developed by Chen (16) was applied to examine the diffusive flux of TeCB under anoxic and oxic conditions and to examine the importance of porewater colloids in diffusive flux.

Methods Sample Collection. Anoxic sediments were collected from three sites in the Mugu Lagoon estuary in southern California. The estuary receives freshwater inputs from Calleguas Creek, Revelon Slough, and several agricultural drainage ditches. Sites were selected to represent the salinity gradient that 3548

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 16, 2003

particle-associated HOCs might encounter as they are transported from the river mouth to the marine environment in successive resuspension and deposition events. Site A was located in the tidal portion of the mouth of Calleguas Creek. Site B was located in a depositional area where the zone of maximum turbidity often occurred. Site C was a nearly fully marine site near the point where the estuary enters the Southern California Bight. Samples were collected by filling submerged 1-L glass jars with anoxic sediment and sealing while submerged near the sediment-water interface. While sample collection may have resulted in the introduction of small amounts of dissolved oxygen, microbial activity was sufficiently high to maintain anoxic conditions as evidenced by no visible oxidation of ferrous sulfide. Samples were stored headspacefree at 4 °C in N2-filled glovebags until processed. Preparation of Anoxic and Oxic Samples. With the exception of aeration, all sample manipulations were performed under N2 in a controlled atmosphere glovebox (LabCono, Inc., Kansas City, MO) to prevent oxygen contamination. Prior to use, the glovebox was repeatedly purged and filled with N2 to produce an oxygen-free environment. The nitrogen gas used was passed through a molecular sieve gas purifier, activated carbon, and an oxygen trap. Anaerobic indicator strips (GasPak Strips, Becton Dickson and Co., Sparks, MD) were used to monitor oxygen levels in the glovebox. Anoxic porewater was extracted from sediment samples by pressure filtration through 0.45-µm mixed cellulose ester filters (Millipore HA, Millipore Corp., Bedford, MA) under a maximum pressure of 35 psi N2 (oxygen-free grade) in a Teflon-lined hazardous waste filtration unit (Millipore Corp., Bedford, MA). Anoxic porewater samples were split; half the volume was maintained under anoxic conditions, and half was aerated to produce oxic conditions. Porewater samples were aerated by bubbling zero-grade air (first passed through a moisture trap, granular activated carbon (GAC), and organic carbonfree water) through anoxic porewater for 7 d. The aeration treatment was used to achieve a level of oxidation that would produce measurable changes in the association of the model HOC with DOMpw. These samples are referred to as “oxic” porewater samples. Oxic samples were filtered through a 0.45-µm filter prior to use to remove precipitated iron and coagulated DOM. Reference porewater was prepared for each site by passing anoxic and oxic porewater through a GAC column to remove DOM. Sediment solids were split, and half the mass was maintained under anoxic conditions. The second half was exposed to atmospheric oxygen until the oxidation of ferrous sulfide was evidenced by change in sediment color (24-48 h). Sediment-Porewater and Solids Characterization. Sediment porewater was characterized before and after aeration. Conductivity was measured with a model 1484 conductivity meter (Chemtrix, Inc., Hillsboro, OR), and pH was measured with an Accumet 950 pH/ion meter (Fisher Scientific, Pittsburgh, PA) equipped with Accumet combination pH electrode. Total alkalinity was determined by acidimetric titration (17). Metal concentrations were determined by inductively coupled plasma/atomic emission spectrometry (ICP/AES) using a Perkin-Elmer Optima 3000 DA ICP-AES spectrometer following standard method 3120B (17). DOMpw for each anoxic and oxic sample was measured as dissolved organic carbon (DOC) concentration using a Dohrman DC80 TOC analyzer (Xertex Corp., Santa Clara, CA) by the persulfate-ultraviolet oxidation method (17). A HewlettPackard 8451A diode array spectrophotometer was used to measure ultraviolet absorbance at 254 nm (A254). Absorptivity on a mole carbon basis (254) was calculated from A254 and DOC concentration. Molar absorptivity at 250-280 nm has

FIGURE 1. Analytical scheme for determining the apparent equilibrium distribution coefficients Kd, Koc, and Kpwdoc. Mspike and CT are the mass of TeCB added and the total TeCB concentration. Measured quantities are in bold. The superscript “obs” refers to “observed” Kd and Koc values obtained without taking into account DOMpw as a sorptive phase. been shown to correlate with the aromaticity of dissolved humic substances and their affinity for HOCs (18-22). Sediment porosity and water content were measured using ASTM methods (23). The solid density of sediment was analyzed using a water displacement technique (24). Sediment organic carbon content was determined by the hightemperature oxidation method using the sludge/sediment combustion unit on the DC-80 TOC analyzer (25). HOC Distribution Coefficients. The batch equilibrium sorption method (26) was modified to allow determination of apparent equilibrium distribution coefficients for sediment solids (Kd) and porewater dissolved organic matter (Kpwdoc). The experimental scheme depicted in Figure 1 was used to determine values for Kd, Koc, and Kpwdoc. In each case, a 20mL aliquot of porewater was spiked with 14C-labeled 2,2′,4,4′tetrachlorobiphenyl (TeCB) to yield an initial concentration 8.9 +873 -0.1 -61*

a Abbreviations: [Alk], alkalinity; DO, dissolved oxygen concentration; E , measured apparent reduction potential; DOC, dissolved organic H carbon; 254, molar absorptivity; I, ionic strength; κ, electrical conductivity. b Duplicate measurements. Mean ( standard deviation reported. Significance of differences: *, p < 0.005; **, p < 0.0005.

exists between the porewater and DOMpwaq phases; (iv) the chemical concentration profile is invariant with time; and (v) both the diffusivity and the distribution coefficient for DOMpw and DOM in the diffusive boundary layer are equal. These assumptions allow diffusive flux (FD) to be expressed as

FD )

(CL - Cw) rs + rw

(4)

where CL and Cw are the truly dissolved chemical concentrations at the depth of the active sediment layer (L) and in the bulk water; and rs and rw are the sediment- and water-side mass transfer resistances. The sediment-side mass transfer resistance is a function of the sediment porosity (φ), desorption enhancement factor (Ψ), and the effective bioturbation (D h b) and molecular diffusion (D h m) coefficients for the mobile species (i.e., truly dissolved and DOMpw-associated):

rs )

L φΨ(D hb + D h m)

(5)

The effective bioturbation and molecular diffusion coefficients are defined as D h b ) Db(1 + Kpwdoc[DOC]pw) and D hm ) D′m + D′pwdocKpwdoc[DOC]pw, where D′m and D′pwdoc are the aqueous molecular diffusivities for the HOC and porewater DOC corrected for sediment porosity and tortuosity. The desorption enhancement factor quantifies the enhanced flux of a sorptive compound due to desorption and is dependent in a complex manner on D h b, D h m, Koc, and the desorption rate constant (k1) (16). The water-side mass transfer resistance is defined as

rw )

Zw Dw

(6)

where Zw represents the thickness of the diffusive boundary layer; and the diffusivity in the bulk water phase (Dw) equals Dm + DdomKdoc[DOC]w. The total mass transfer resistance across the sediment-water interface (rtotal) consists of the sum of the two individual resistances (rtotal ) rs + rw). Assuming Cw ) 0 and equilibrium between the porewater and sediment organic matter at the depth of the active layer (i.e., CL ) Coc,L/Kd), eq 4 can be expressed as

FD )

Coc,L 1 foc Kocrtotal

(7)

Because Coc,L/foc was held constant in model simulations, the overall effects of Koc and total mass transfer resistance controlled HOC diffusive flux.

Results and Discussion Effect of Aeration on Inorganic and Organic Porewater Constituents. Aeration of anoxic porewaters resulted in a 3550

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 16, 2003

TABLE 2. Sediment Characteristicsa characteristic Fs Fbb φb pH focb b

unit cm-3

g g cm-3

site A

site B

site C

1.27 ( 0.01 1.20 ( 0.01 1.95 ( 0.07 0.56 ( 0.00 0.63 ( 0.02 2.00 ( 0.01 0.69 ( 0.00 0.66 ( 0.05 0.49 ( 0.00 7.67 7.65 8.12 0.0059 ( 0.0000 0.0094 ( 0.0017 0.0064 ( 0.0005

a Abbreviations: F , solid density; F , bulk density; φ, porosity [φ ) s b 1-(Fb/Fs)]; foc, mass fraction of organic carbon. b Duplicate measurements. Mean and standard deviation reported.

0.4-0.6 unit increase in pH and a 0.38-1.78 mM decrease in alkalinity (Table 1). Aeration caused a loss of carbon dioxide, hydrogen sulfide, and other volatile organic compounds present in the porewaters. Hunchak-Kariouk et al. (7) and Pedersen et al. (1) observed comparable pH increases when anoxic porewaters were aerated. Aeration of anoxic porewater induced the precipitation of metal oxides and phosphates (cf. Supporting Information, Table S1) but had little effect on porewater ionic strength (Table 1). Porewater DOC concentrations were highest for site A under both anoxic and oxic conditions. Aeration resulted in modest but significant decreases in DOC concentration at sites A (13.1%; p < 0.0005) and B (8.9%; p < 0.005) but had negligible impact on the dissolved organic matter concentration at site C (p > 0.05). The decline in porewater DOC concentrations at sites A and B was likely a consequence of coagulation with precipitating iron(III) hydroxides (1, 7). Under both anoxic and oxic conditions, values for 254 decreased with increasing ionic strength, suggesting that DOMpw aromaticity declined from site A to site B to site C. Aeration of anoxic porewaters produced significant drops in 254 (p < 0.005) for all DOMpw samples. Pedersen et al. (1) observed similar decreases in absorptivity and attributed them to both selective coagulation of more absorptive DOM macromolecules and the oxidative cleavage of aromatic groups. Sediment Solid Characteristics. Values for sediment foc, solid density, bulk density, and porosity are presented in Table 2. The physical characteristics of sediment solids (i.e., solid density, bulk density, porosity) were assumed unaffected by changes in oxic state and therefore measured for oxic samples only. Sediment physical characteristics were similar at sites A and B. The mass fraction of sediment organic carbon varied slightly among the samples (Table 2). Association of TeCB with Porewater Dissolved Organic Matter. Figure 3a presents log Kpwdoc values for DOMpw from the three sites under anoxic and oxic conditions. Kpwdoc varied considerably among sites under both anoxic and oxic conditions. Anoxic Kpwdoc values varied over an order of magnitude (1.2 log units) among the samples, with site A DOMpw exhibiting the highest value (105.7 L kg-1). Oxic Kpwdoc

FIGURE 4. Effect of change in oxic state on the apparent equilibrium distribution of 2,2′,4,4′-tetrachlorobiphenyl in sediment porewater. Significance of differences: *, p < 0.001. Asterisks inside bars signify significance of difference for Cpwaq and Cpwdoc; asterisks above bars indicate significance of difference for Cpw ) Cpwaq + Cpwdoc upon aeration.

FIGURE 3. Effect of change in oxic state on apparent (a) porewater DOC-water (Kpwdoc) and (b) particle-associated organic carbon (Koc) distribution coefficients (L kg-1). Values given represent mean ( standard deviation (n ) 3); significance of increase or decrease: *, p < 0.01; **, p < 0.005; ***, p < 0.0005. values displayed somewhat less variation (0.7 log unit) among sites. Anoxic DOMpw from site A displayed the highest sorption capacity for TeCB, while oxic porewater DOM from the same site had the lowest sorption capacity. No trend was apparent in the magnitude of anoxic or oxic Kpwdoc values with respect to solution ionic strength. Parallel studies in our laboratory examining the influence of ionic strength on TeCB association with DOM indicated that Kdoc declines with increasing ionic strength for a given DOC isolate (28). The lack of trend exhibited by the DOM used in the present study was presumably due to differences in the nature of the DOMpw isolated from the three sites. Kpwdoc was not significantly correlated with 254 (p > 0.05), suggesting that factors other than DOMpw aromaticity may have contributed to the observed intersite differences for anoxic and oxic samples. Effect of Change in Oxic State on Kpwdoc. Aeration of anoxic porewaters induced site-specific changes in the magnitude of Kpwdoc that were significant for samples from two of the three sites. Aeration effected a significant 1.4 log unit drop (p < 0.0005) in Kpwdoc for site A DOMpw. In contrast, site B DOMpw exhibited a significant 0.6 log unit increase (p < 0.005) in Kpwdoc upon aeration. The observed difference between anoxic and oxic Kpwdoc values for site C DOMpw was within measurement error. These results indicate that TeCB would be released from DOMpw at site A but sorbed to a greater extent at site B when anoxic sediments are oxygenated. Our previous investigations of aeration effects on TeCB association with DOMpw in estuarine systems indicated a decrease or no change in Kpwdoc upon aeration (1, 2). Our former results taken together with the present study suggest that the change in TeCB association with DOMpw upon aeration is site-specific and cannot be predicted without more detailed knowledge of dissolved organic matter properties. The application of more sophisticated analytical techniques (e.g., 13C-nuclear magnetic resonance spectroscopy) may yield insight into the aeration-induced alterations in DOM responsible for the change in sorptive behavior. Changes in DOMpw macro-

molecular conformation or supramolecular structure due to aeration may also warrant investigation. Sorption of TeCB to Sediment Organic Matter. Figure 3b presents log Koc values for sediments from the three sites under anoxic and oxic conditions. Variation in Koc values among sites under both anoxic and oxic conditions were less pronounced than that observed for Kpwdoc. Under both anoxic and oxic conditions, site A sediment log Koc values were significantly larger (p < 0.01) than those for the other two sites. Anoxic log Koc values for site B and C sediments were statistically indistinguishable (p > 0.10); under oxic conditions, the log Koc value for site B was higher than that for site C (p < 0.10). Effect of Change in Oxic State on Koc. Exposure of sediment solids to atmospheric oxygen resulted in a significant change in TeCB association with SOM for only one of the three sediment samples. A significant 0.4 log unit decline in Koc was observed for site A sediment (p < 0.01). Koc values for site B and C sediments did not differ significantly before and after exposure to atmospheric oxygen (p > 0.10). In prior work in our laboratory, aeration of anoxic estuarine sediment resulted in a 0.5 log unit decrease in Koc (2). As noted in the discussion of Kpwdoc, additional characterization of aeration-induced changes to sediment organic matter may allow prediction of the effect on HOC sorption. In all cases, values for Koc were higher than those for Kpwdoc. The ratio of Koc to Kpwdoc varied considerably, especially among aerated samples (anoxic Koc:Kpwdoc ) 17-38; oxic Koc:Kpwdoc ) 10-170). The lower capacity of DOMpw to sorb TeCB presumably reflects its more polar nature. Previous studies indicated that DOM is often more polar than particleassociated SOM and therefore displays a lower affinity for HOCs (29-31). The ratio of Koc to Kpwdoc exhibited a significant order of magnitude increase for site A sediment upon aeration (p < 0.005), while the ratio decreased by a factor of approximately 3.5 for site B sediment (p < 0.025). No significant change in this ratio was observed for site C sediment (p > 0.10). Fractional Equilibrium Distribution of TeCB. Figure 4 illustrates the distribution of TeCB between the mobile (i.e., truly dissolved and DOMpw-associated) phases under anoxic and oxic conditions. Both before and after aeration, the overall concentration of mobile species (i.e., Cpwaq + Cpwdoc) was highest at the site with the lowest ionic strength (site A). Aeration of anoxic sediments elicited significant 1.4- and 1.3-fold increases in the total concentration of mobile species at sites A and C (p < 0.0005) and significant changes in Cpwaq and Cpwdoc for sediments from all three sites (p < 0.001). For VOL. 37, NO. 16, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3551

TABLE 3. Input Parameters for Diffusive Flux Modela site A parameter

unit

log Kd k1 Dm Dpwdom Db [DOC]pw [DOC]w Zw Cw R L Coc,L Cw

kg-1

L s-1 cm2 s-1 cm2 s-1 cm2 s-1 mgC L-1 mgC L-1 cm ng cmw-3 cm cm ng gOC-1 ng cmw-3

site B

site C

anoxic

oxic

anoxic

oxic

anoxic

oxic

4.7 0.00465 4.93 × 10-6 3.00 × 10-6 6.30 × 10-6 20.7 1.0 0.01 0 0.01 16 27 500 0

4.3 0.00465 4.93 × 10-6 3.00 × 10-6 6.30 × 10-6 18.0 1.0 0.01 0 0.01 16 27 500 0

4.0 0.00534 4.93 × 10-6 3.00 × 10-6 6.30 × 10-6 9.71 1.0 0.01 0 0.01 16 27 500 0

4.1 0.00534 4.93 × 10-6 3.00 × 10-6 6.30 × 10-6 8.85 1.0 0.01 0 0.01 16 27 500 0

4.0 0.0108 4.93 × 10-6 3.00 × 10-6 6.30 × 10-6 5.48 1.0 0.01 0 0.01 16 27 500 0

3.8 0.0108 4.93 × 10-6 3.00 × 10-6 6.30 × 10-6 5.43 1.0 0.01 0 0.01 16 27 500 0

a Abbreviations: C , , concentration of TeCB sorbed in/on sediment solids at the depth of L; C , TeCB concentration in the bulk water phase; oc L w Db, bioturbation coefficient; Dm, TeCB aqueous molecular diffusivity; Dpwdoc, diffusivity of TeCB associated with porewater DOC; [DOC]pw, porewater DOC concentration; [DOC]w, bulk water DOC concentration; k1, desorption rate constant; L, depth of active sediment bed; R, radius of particle aggregates; Zw, thickness of diffusive boundary layer.

site A, aeration induced a 12.7-fold increase in Cpwaq and a 2.8-fold decrease in Cpwdoc (p < 0.0005) due to simultaneous release of TeCB from particle-associated and dissolved organic matter (i.e., Kpwdoc and Koc decreased) and a decrease in [DOM]pw. For sites B and C, aeration resulted in significant increases in the amount of TeCB associated with DOMpw (p < 0.0005) and significant decreases in truly dissolved concentrations (p < 0.0005). In all cases the fraction of TeCB sorbed to particle-associated SOM exceeded 0.9994 (data not shown). Diffusive Flux Modeling. Diffusive flux from bed sediments to the overlying water was simulated using the data from each of the three sites under anoxic and oxic conditions. Input parameters for the sediment flux model are presented in Tables 1-3. TeCB molecular diffusivity (Dm) was estimated from the molal volume of TeCB at its boiling point and the dynamic viscosity of water at 20 °C (32). DOM diffusivity in sediment porewater depends on macromolecular size and conformation. For the purpose of simulating diffusive flux, an average DOMpw molecular mass of 2708 u was used. This value is slightly less than the midpoint of the range reported by Chin et al. (33) and reflects the predominance of lower molecular weight DOM in Mugu Lagoon porewater (1). To account for the effect of TeCB associated with DOMpw on the diffusivity of porewater DOM, a final molecular mass of 3000 u was used. Because our goal was to examine the effect on diffusive flux of aeration-induced changes in TeCB sorption, we used constant values for the bioturbation coefficient (Db), diffusive boundary layer thickness (Zw), aggregate particle radius (R), active sediment depth (L), and TeCB concentration in the sediment organic matter at depth L (Coc,L) and in the overlying water (Cw). Although bioturbation is typically less under anoxic conditions (34), the value of Db was held constant to allow examination of changes in diffusive flux induced by alteration of sediment oxic state. For the purpose of the simulation, we set the dissolved organic carbon concentration in the bulk water phase ([DOC]w) at 1 mgC L-1. The diffusive boundary layer thickness (Zw) was calculated from Dm, the kinematic viscosity (ν), and the friction velocity (u*): Zw ) 12ν2/3Dm1/3/u* (16) and held constant for all model simulations. The bulk water column TeCB concentration (Cw) was assumed to be zero. Changes in oxic state were assumed to extend to the depth of the biologically active layer (L). The concentration of TeCB sorbed to sediment organic matter at the depth of biologically active layer (Coc,L) was the average value obtained in the batch equilibrations. To examine the importance of facilitated transport by porewater DOM, we simulated TeCB diffusive flux with and without this phase. 3552

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 16, 2003

FIGURE 5. Effect of change in oxic state on diffusive flux of TeCB from bed sediments, FD (mg m-2 yr-1), (a) neglecting influence of DOM and (b) considering influence of DOM. Values given represent mean ( standard deviation (n ) 3); significance of increase or decrease: *, p < 0.10; **, p < 0.025. Model Results: Effect of Oxic State on TeCB Diffusive Flux. Figure 5a displays predicted diffusive fluxes when DOMpw-facilitated transport was ignored. Values for the diffusive flux, desorption enhancement factor, and mass transfer resistances are presented in Table 4. Predicted diffusive fluxes were significantly higher for site C than for site A (p < 0.05) under both anoxic and oxic conditions. FD was also significantly higher for site B than site A under anoxic conditions (p < 0.005). These differences were due primarily to the higher Koc values at site A. Similar differences in FD among sites were obtained in simulations explicitly considering DOMpw-facilitated transport (Figure 5b). When DOMpw-facilitated transport was neglected, aeration of anoxic estuarine sediments resulted in a significant 110 mg m-2 yr-1 increase in diffusive flux at site A (p < 0.025)

TABLE 4. Effect of Change in Oxic State on Diffusive Flux (FD), Desorption Enhancement (Ψ), and Mass Transfer Resistances (rs, rw, rtotal) site A parameter

unit

anoxic

FD Ψ rs rw rtotal

mg m-2 yr-1 s cm-1 s cm-1 s cm-1

75 ( 35 14 800 ( 7 100 161 ( 77 2 030 2 190 ( 80

FD Ψ rs rw rtotal Γa

mg m-2 yr-1 s cm-1 s cm-1 s cm-1 -

98 ( 62 1 700 ( 940 134 ( 86 1 530 ( 710 1 670 ( 710 1.3 ( 1.0

a

site B oxic

anoxic

site C oxic

anoxic

oxic

Case 1: DOMpw-Facilitated Transport Ignored 185 ( 43 295 ( 89 249 ( 74 5 990 ( 1 400 3 920 ( 1 180 4 800 ( 1 450 399 ( 93 647 ( 197 528 ( 161 2 030 2 030 2 030 2 430 ( 90 2 670 ( 200 2 560 ( 160

390 ( 184 11 800 ( 5 630 317 ( 151 2 030 2 340 ( 150

507 ( 240 8 900 ( 4 220 421 ( 200 2 030 2 450 ( 200

Case 2: DOMpw-Facilitated Transport Considered 188 ( 57 300 ( 105 266 ( 126 4 700 ( 1 110 3 150 ( 960 2 520 ( 860 391 ( 94 638 ( 196 507 ( 197 2 000 ( 460 1 990 ( 460 1 880 ( 870 2 390 ( 470 2 630 ( 500 2 390 ( 890 1.0 ( 0.4 1.0 ( 0.5 1.1 ( 0.6

406 ( 208 8 880 ( 4 230 310 ( 149 1 940 ( 450 2 250 ( 470 1.0 ( 0.7

537 ( 326 5 980 ( 2 910 410 ( 207 1 900 ( 880 2 310 ( 900 1.1 ( 0.8

Γ, flux enhancement by porewater dissolved organic matter; Γ ) FD,w DOMpw/FD,w/o DOMpw.

representing a nearly 2.5-fold rise in FD (Figure 5a). Aerationinduced changes in FD at sites B and C were not significant (p > 0.10). Changes in FD in simulations neglecting DOMpwfacilitated transport reflected aeration-induced alterations in Koc alone. The strongest influence on diffusive flux was the inverse dependence of FD on Koc (cf. eq 7). Koc exerts a less significant, more indirect and opposing effect through the desorption enhancement factor contained in the sedimentside mass transfer resistance (cf. eq 5). The lack of significant change in FD at sites B and C upon aeration was not surprising given the negligible effect exposure to atmospheric oxygen had on the magnitude of Koc at these two sites. For site A, diffusive flux was predicted to increase upon aeration because TeCB would be released from particle-associated organic matter (i.e., Koc decreased) into the mobile porewater phase as the truly dissolved compound. In the system modeled, water-side mass transfer resistance dominated overall mass transfer resistance (Table 4). This mass transfer resistance remained unaltered because Zw was fixed in the model simulations and Dm is constant at a given temperature and viscosity. Figure 5b displays simulated diffusive fluxes when DOMpwfacilitated transport was considered. When DOMpw-facilitated transport was included in model simulations, a significant 89 mg m-2 yr-1 aeration-induced increase in diffusive flux at site A was obtained (p < 0.10), representing nearly a doubling of FD. For sites B and C, changes in FD upon aeration were not significant (p > 0.10). The predicted fluxes were of the same order of magnitude as those estimated by Achman et al. (35) for the lower Hudson River Estuary. As with the simulations ignoring the influence of DOMpw on diffusive flux, water-side mass transfer resistance dominated overall mass transfer resistance. We note that the mean values for diffusive flux, desorption enhancement and mass transfer resistances presented in Table 4 differ from those reported by ref 2. In propagating errors through the flux calculations, the model was reconstructed. In doing so, it was discovered that k1 was not calculated correctly by ref 2 because bulk density was used instead of the solid-water ratio. The values in Table 4 represent corrected values. Model Results: Importance of DOMpw-Facilitated Transport. The relatively minor contribution of DOMpw-facilitated transport to the overall diffusive flux of TeCB became apparent when model results including the influence of porewater DOM are compared to those ignoring it (Table 4). Flux enhancement by DOMpw (Γ) was defined as the ratio of flux in the presence of DOMpw to that in its absence. Although in some cases Γ exceeded unity, the increases in diffusive flux due to DOMpw-facilitated transport were not statistically significant (p < 0.10).

FIGURE 6. Representative results from analyses of the sensitivity of diffusive flux (FD) to the magnitude of (a) Kpwdoc and (b) Koc. Sensitivity Analyses. Sensitivity analyses were conducted to examine the influence of Kpwdoc and Koc on diffusive flux. Results for site A are presented in Figure 6 for illustrative purposes. Sensitivity analysis indicated that, for the systems modeled, DOMpw-facilitated transport would begin to increase TeCB diffusive flux substantially as Kpwdoc values exceeded approximately 106 L kg-1 (Figure 6a). The Kpwdoc value for anoxic site A DOMpw was close to this value, but the enhancement of flux due to DOMpw-facilitated transport was not significant (p < 0.10). Diffusive flux was much more sensitive to the magnitude of Koc. Over the Koc range of 103-105 L kg-1, diffusive flux changed relatively little in the system modeled (Figure 6b). However, at Koc values greater than 105 L kg-1, FD decreased rapidly. Diffusive flux was relatively insensitive to [DOC]pw at realistic porewater dissolved organic carbon concentrations (data not shown). Chen (16) conducted a series of sensitivity VOL. 37, NO. 16, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3553

analyses and showed that the modeled diffusive flux was highly sensitive to Db, foc, Koc, and Zw. Environmental Implications. Dredging operations and resuspension events during storm events alter the oxic state of estuarine sediments and may result in altered sorption of nonpolar organic chemicals to particle-associated and porewater DOM. The impact of aeration on HOC diffusive flux depends on the sign and magnitude of alterations in Kpwdoc, Koc, and [DOC]pw. Model simulations suggested that aeration of anoxic sediments results in site-specific changes in diffusive flux. In two of the cases examined, aeration-induced changes in FD were negligible. At the low ionic strength site, however, aeration was predicted to effect a doubling of diffusive flux. Accurate prediction of HOC diffusive flux from estuarine bed sediments requires selection of appropriate experimentally determined sorption coefficients and in some cases may require consideration of facilitated transport by porewater DOM (i.e., for highly sorptive DOMpw) and/or extremely hydrophobic organic contaminants. The changes in HOC sorption upon sediment aeration observed in this study varied by site and for Kpwdoc were not clearly related to changes in aromaticity. These findings highlight the need to determine site-specific values for sorption coefficients under appropriate conditions in cases where remediation activities alter sediment oxic state. Our results suggest that, in some cases, remediation efforts may mobilize HOCs to a greater extent than predicted by models not considering changes in oxic state. The effect of oxic state on HOC diffusive flux from freshwater sediments also warrants investigation.

Acknowledgments David Kimbrough (Castaic Lake Water Agency) performed the ICP/AES analyses. Shelley Anghera, Medhi Wangpaichitr, and Monique Myers are thanked for their assistance in sample collection. We thank Ed Ruth for the PCB analysis on the sediment solids. We thank three anonymous reviewers for helpful comments that improved the quality of this manuscript. This research was funded by a grant from the Office of Naval Research (Grant N00014-96-1-0079).

Supporting Information Available A table displaying sediment porewater concentrations of metals and phosphorus before and after aeration. This material is available free of charge via the Internet at http:// pubs.acs.org.

Literature Cited (1) Pedersen, J. A.; Gabelich, C., Lin, C. H.; Suffet, I. H. Environ. Sci. Technol. 1999, 33, 1388-1397. (2) Pedersen, J. A.; Schweitzer, L. E.; Lin, C.-H.; Suffet, I. H. Isr. J. Chem. 2002, 42, 109-118. (3) Karickhoff, S. W. J. Hydraul. Eng. 1984, 110, 707-735. (4) Schwarzenbach, R. P.; Gschwend, P. M.; Imboden, D. M. Environmental Organic Chemistry, 2nd ed.; John Wiley & Sons: New York, 2003. (5) Valsaraj, K. T.; Thoma, G. J.; Porter, C. L.; Reible, D. D.; Thibodeaux, L. J. Water Sci. Technol. 1993, 28, 139-147.

3554

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 16, 2003

(6) Valsaraj, K. T.; Verma, S.; Sojitra, I.; Reible, D. D.; Thibodeaux, L. J. J. Environ. Eng. 1996, 122, 722-729. (7) Hunchak-Kariouk, K.; Schweitzer, L.; Suffet, I. H. Environ. Sci. Technol. 1997, 31, 639-645. (8) Orem, W. H.; Gaudette, H. E. Org. Geochem. 1984, 5, 175-181. (9) Stumm, W. Chemistry of the Solid-Water Interface; John Wiley & Sons: New York, 1992. (10) Rose, A. L.; Waite, T. D. Environ. Sci. Technol. 2002, 36, 433444. (11) Kristensen, E.; Ahmed, S. I.; Devol, A. H. Limnol. Oceanogr. 1995, 40, 1430-1437. (12) Holmer, M. Estuarine Coastal Shelf Sci. 1999, 48, 383-390. (13) Burdige, D. J. Org. Geochem. 2001, 32, 487-505. (14) Skoog, A.; Hall, P. O. J.; Hulth, S.; Paxeus, N.; van der Loeff, M. R.; Westerlund, S. Geochim. Cosmochim. Acta 1996, 60, 36193629. (15) Coates, J. D.; Chakraborty, R.; O’Connor, S. M.; Schmidt, C.; Thieme, J. Acta Hydrochim. Hydrobiol. 2000, 28, 420-427. (16) Chen, H. W. M.S. Thesis, Massachusetts Institute of Technology, Cambridge, MA, 1993. (17) Standard Methods for the Examination of Water and Wastewater, 18th ed.; Greenberg, A. E., Clesceri, L. S., Eaton, A. D., Eds.; American Public Health Association: Washington, DC, 1992. (18) Gauthier, T. D.; Seitz, W. R.; Grant, C. L. Environ. Sci. Technol. 1987, 21, 243-248. (19) Traina, S. J.; Novak, J.; Smeck, N. E. J. Environ. Qual. 1990, 19, 151-153. (20) Chin, Y.-P.; Aiken, G.; O’Loughlin, E. Environ. Sci. Technol. 1994, 28, 1853-1858. (21) Chin, Y.-P.; Aiken, G. R.; Danielsen, K. M. Environ. Sci. Technol. 1997, 31, 1630-1635. (22) Perminova, I. V.; Grechishcheva, N. Y.; Petosyan, V. S. Environ. Sci. Technol. 1999, 33, 3781-3787. (23) Annual Book of ASTM Standards, Vol. 04.08; American Society of Testing and Materials: Washington, DC, 1999; pp 208-212. (24) Ball. W. P. Ph.D. Dissertation, Stanford University, Stanford, CA, 1989. (25) Xertex-Dohrmann. DC-80 TOC Analyzer Operations Manual, 11th ed.; 1996. (26) Annual Book of ASTM Standards, Vol. 11.05; American Society of Testing and Materials: Washington, DC, 1999; pp 478-484. (27) Thibodeaux, L. J. Environmental Chemodynamics, 2nd ed.; John Wiley & Sons: New York, 1996. (28) Lin, C.-H. M. Ph.D. Dissertation. University of California, Los Angeles, CA, 2001. (29) Brannon, J. M.; Myers, T. E.; Gunnison, D.; Price, C. Environ. Sci. Technol. 1991, 25, 1082-1087. (30) Koelmans, A. A.; Lijklema, L. Chemosphere 1992, 25, 313-325. (31) Koelmans, A. A. Environ. Toxicol. Chem. 1995, 14, 2015-2016. (32) Tucker, A. W.; Nelken, L. H. In Handbook of Chemical Property Estimation Methods; Lyman, W. J., Reehl, W. F., Rosenblatt, D. H., Eds.; American Chemical Society: Washington, DC, 1990; pp 17-1-17-25. (33) Chin, Y.-P.; McNichol, A. P.; Gschwend, P. M. In Organic Substances in Sediments and Water; Baker, R. A., Ed.; CRC Press: Boca Raton, FL, 1991; pp 107-126. (34) Bosworth, W. S.; Thibodeaux, L. J. Environ. Prog. 1990, 9, 211217. (35) Achman, D. R.; Brownawell, B. J.; Zhang, L. C. Estuaries 1996, 19, 950-965.

Received for review August 12, 2002. Revised manuscript received April 4, 2003. Accepted April 12, 2003. ES026048P