Environ. Sci. Technol. 2007, 41, 1152-1158
Air-Water Exchange of Polychlorinated Biphenyls in the Delaware River A M Y A . R O W E , † L I S A A . T O T T E N , * ,† MINGE XIE,‡ THOMAS J. FIKSLIN,§ AND STEPHEN J. EISENREICH⊥ Department of Environmental Sciences, Rutgers University, 14 College Farm, Road, New Brunswick, New Jersey 08901, USA, Department of Statistics, Rutgers University, 110 Frelinghuysen Road, Piscataway, New Jersey 08854 USA, Delaware River Basin Commission, 25 State Police Drive, West Trenton, New Jersey 08628, USA, and Toxicology and Chemical Substances Unit, European Commission, Joint Research Centre, TP 582, I-21020 Ispra (VA), Italy
The air-water exchange of polychlorinated biphenyls (PCBs) often results in net volatilization, which is thought to be the most important loss process for PCBs in many systems. Previous investigations of the air-water exchange of PCBs have been hampered by difficulties in treatment of the uncertainty in the calculation of air/water fugacity ratios. This work presents a new framework for the treatment of uncertainty, where uncertainty in physical constants is handled differently from random measurement uncertainty associated with random samples, and it further investigates the sorption of PCBs to colloids (dissolved organic carbon). Simultaneous measurements of PCBs in the air and water of five water quality management zones of the Delaware River were taken in 2002 in support of the total maximum daily load (TMDL) process. Gasphase concentrations of ΣPCBs ranged from 110 to 1350 pg m-3, while dissolved water concentrations were between 420 and 1650 pg L-1. Shallow slopes of log KOC vs log KOW plots indicated a colloidal contribution to the apparent dissolved-phase concentrations, such that a three-phase partitioning model was applied. Fugacity ratios for individual congeners were calculated under the most conservative assumptions, and their values (log-transformed) were examined via a single-sample T-test to determine whether they were significantly less than 1 at the 95% confidence level. This method demonstrated that air-water exchange resulted in net volatilization in all zones over all cruises for all but seven high molecular weight congeners. Calculated net fluxes ranged from +360 to +3000 ng m-2 d-1 for ΣPCBs. The colloidal correction decreased the volatilization flux of ΣPCBs by ∼30%. The decachlorinated congener (PCB 209), exhibited unusually high concentrations in the suspended solids, especially in the southern portions of the river, indicating that there is a distinct source of PCB 209 in the Delaware River.
Introduction The Delaware River acts as a natural boundary between the mid-Atlantic states of Delaware, New Jersey, and Pennsyl* Corresponding author phone: (732 932-9588; fax: (732) 9328644; e-mail:
[email protected]. † Department of Environmental Sciences, Rutgers University. ‡ Department of Statistics, Rutgers University. § Delaware River Basin Commission. ⊥ European Commission, Joint Research Centre. 1152
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 4, 2007
vania, eventually flowing into the Atlantic Ocean (Figure 1). The heavily populated urban areas of Philadelphia, PA, and Camden, NJ, have impacted the Delaware River Estuary via anthropogenic inputs of all types of pollutants. The system has been classified as “impaired” under Section 303(d) of the United States Environmental Protection Agency (USEPA) Clean Water Act for a variety of contaminants such as pesticides, polychlorinated biphenyls (PCBs), mercury, and arsenic (1). As required by the Clean Water Act, a Total Maximum Daily Load (TMDL) for PCBs was developed by the Delaware River Basin Commission (DRBC) (2). The TMDL modeling effort suggested that volatilization is the single most important loss process for PCBs in the river (2). Likewise, volatilization of PCBs from the water body to the atmosphere is thought to be the leading loss process for PCBs in the Great Lakes and the New York/ New Jersey Harbor (3, 4). The main objective of this study was to provide air-water exchange fluxes of PCBs in the Delaware River Estuary as part of the TMDL process. Our research group has published many studies of PCB air-water exchange (5-7). Goss et al. (8) have taken issue with some of these works, and have emphasized the necessity of selecting accurate physical-chemical data for use in performing complex environmental calculations. In particular, the Henry’s Law constant (Kaw) is the central property determining both the direction and magnitude of air-water exchange. Goss et al. argue that the uncertainty inherent in the published values of Henry’s Law constants for PCBs prevents the calculation of the direction of air-water gas exchange (8, 9). They recommend an examination of the work of multiple groups, a checking of thermodynamic consistency among the data, and an analysis of uncertainty in order to assess the error generated throughout the environmental analytical process (8, 10). Our group has argued that the direction of air-water exchange can be determined with some confidence in the context of a rigorous error analysis (11). In light of this controversy, the second important goal of this work was to develop a new framework for dealing with the uncertainty in the calculations of air/ water fugacity ratios and determining the direction of air/ water exchange. Once the direction of air-water exchange is determined using the most conservative values of parameters such as Henry’s Law Constant, the values of Kaw that are thought to be most accurate are employed in an attempt to generate “best” estimates of fluxes.
Materials and Methods Sampling Conditions/Site Characterization. The Delaware River flows out of the Catskill Mountains in New York State for 280 miles before emptying into the Atlantic Ocean off the Mid-Atlantic coast (Figure 1). The Delaware River Estuary begins south of the head of tide at Trenton, NJ, (River Mile 133.4) but the River is not saline until Zone 5, south of Philadelphia. The estuary was divided into water quality zones by DRBC. Zone 1, the non-tidal portion of the river located north of the head of tide at Trenton, is not covered by the TMDL, and therefore, it is not considered here. Zone 2 contains Trenton, NJ. Zone 3 includes Philadelphia, PA, and Camden, NJ. Zone 4 covers the remaining southeastern Pennsylvania to the Delaware border. The industrial city of Wilmington, DE, is found in Zone 5. Zone 6 consists of Delaware Bay. The original TMDL applies to Zones 2-5, but Zone 6 was sampled in the later cruises in order to provide a boundary condition for the TMDL model and because a TMDL for PCBs in Zone 6 is currently being developed. 10.1021/es061797i CCC: $37.00
2007 American Chemical Society Published on Web 01/17/2007
by Nutrient Analytical Services of the Chesapeake Biological Laboratory, University of Maryland. Analytical Methods. Sampling methods are described in ref 12. Prior to extraction, all samples were injected with PCBs 14, 23, 65, and 166 as surrogate standards. Samples were Soxhlet extracted for 24 h in appropriate solvent. XAD and GFF samples were then liquid-liquid extracted with 60 mL of Milli-Q water, to separate the aqueous fraction from the organic fraction. The aqueous fraction was back-extracted three times with 50-mL of hexane and 1 g of NaCl in a separatory funnel. The extracts were then concentrated via rotary evaporation followed by N2 evaporation to a volume of ∼1 mL. The samples were then split with 50% of the extract archived for PAH analysis and the other 50% being used for PCB and pesticide testing. The PCB portion was fractionated on a 2.5% water-deactivated Florisil column with the PCB fraction eluted with 35 mL of petroleum ether (13). The PCB extract was reduced in volume and injected with an internal standard containing PCBs 30 and 204. All samples were analyzed for PCBs by gas chromatography with electron capture detection. Quality Assurance PCB concentrations were corrected for surrogate recoveries. Surrogate recoveries for PCBs 14, 23, and 166, respectively, were as follows: PUF samples, 89 ( 16%, 73 ( 15%, and 79 ( 14%; QFF samples, 91 ( 10%, 80 ( 13%, and 93 ( 15%; XAD-2 water samples, 84 ( 16%, 70 ( 10%, and 80 ( 17%; GFF, 86 ( 10%, 72 ( 9%, and 75 ( 8%. Field blanks and matrix spikes were used for quality control purposes. Only PCB 209 appeared in significant amounts in the field and rinsate blanks and required blank correction.
Results and Discussion FIGURE 1. Map showing five river zones and sites where samples were collected (stars). RM indicates river mile from the mouth of Delaware Bay. Map provided by DRBC. Sampling Methods Sampling occurred during the TMDL model calibration period (September 1, 2001 to March 31, 2003) (2). Simultaneous air and water samples were taken aboard a boat in each of the five zones of the Estuary over four 2-day sampling intensives spanning all four seasons. Each water quality zone was sampled at approximately the same location in each cruise. Duplicate over-water air samples were collected using modified high-volume air samplers (Tisch) that were mounted on the bow of the boat. Pre-combusted, preweighed quartz fiber filters (QFFs, Whatman) captured the aerosol phase while Soxhlet-extracted polyurethane foam plugs (PUFs) collected the gas phase. The air samplers were operated during travel through each zone for about 4 h, which yielded calibrated sample volumes between 84 and 175 m3. Water samples were collected in situ with an Infiltrex 100 (AXYS Systems) sampler on the first cruise. For the remaining cruises, the water sampling system consisted of a 20-L stainless steel holding tank (“Pepsi can”), copper tubing, and a filter head. Pre-combusted glass fiber filters (GFFs, Whatman) with a 0.7 µm pore size collected the particle phase and a Teflon column filled with XAD-2 resin (Amberlite) captured the dissolved phase. Clean XAD columns that were taken to the field and returned to the laboratory without being opened were used as field blanks. Rinsate blanks consisting of 2 L of Milli-Q water were also taken to assess any contamination in the lines of the sampling equipment. Sample volumes processed ranged from 12 to 40 L. Triplicate water samples (grab, 1 L) were collected in glass bottles for total suspended matter (TSM), particulate organic carbon (POC), and dissolved organic carbon (DOC). POC and DOC were analyzed
Meteorological data, TSM, and DOC results are presented in Supporting Information Table S-1. The total suspended matter (TSM) generally increased with distance downstream, ranging from 6.6 to 62 mg L-1 in Zones 2, 3, 4, and 6. Zone 5 exhibited high TSM values, with a maximum of 150 mg L-1 in November of 2002, due to the estuarine turbidity maximum. DOC was relatively constant at about 4.0 mg C L-1 among all zones for the May, August, and November cruises. Higher DOC (averaging 15 mg C L-1) was measured during the March cruise, which is probably indicative of the spring burst of primary production. The salinity also increased with movement downstream, ranging from essentially 0 in Zones 2, 3, and 4, to 0.1 mol L-1 in Zone 5 and 0.3 mol L-1 in Delaware Bay (Zone 6). PCB Concentrations. ΣPCB concentrations in the gas, aerosol, suspended matter, and apparent dissolved phases are given in Supporting Information Tables S2-6. Gas Phase. ΣPCB concentrations were generally lowest during the May cruise, ranging from 113 to 225 pg m-3. These values are typical of the PCB “background” for this region (7). The highest concentrations were observed during August, ranging from 1140 to 1350 pg m-3, due to higher temperatures in summer. Zones 3 and 4, which harbor the urban areas of Philadelphia and Camden, generally had the highest gasphase ΣPCB concentrations over the five cruises, with maximum concentrations for ΣPCBs of about 1300 pg m-3 during the August cruise. These concentrations are lower than the both the 12-month (3300 pg m-3) and summer (5000 pg m-3) average values observed at Camden, NJ (7). These gas-phase samples were dominated by lesser-chlorinated congeners with homologues 3-5 contributing an average of 87% to the total PCB concentrations. The gas-phase concentrations were dominated by homologues 3 and 4, with little or no contribution from groups 7, 8, 9, and 10. This gas-phase homologue pattern is followed closely by the water dissolved phase trend. VOL. 41, NO. 4, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1153
The measured gas-phase PCB concentrations were correlated with temperature via the Clausius-Clapeyron equation:
lnC )
a +b T
(1)
where C is the concentration of the PCB congener in pg m-3, a and b are fitting parameters, and T is the average temperature over the sampling period in Kelvin. Homologues 3-7 were strongly temperature driven (R2 ) 0.82 to 0.87, p < 10-5), while the heavier homologues were less strongly temperature driven (R2 for homologues 8 and 9 were 0.64 and 0.68, respectively). PCB 209 (homologue 10) was below detection limits in all the gas-phase samples. Apparent Dissolved Water Concentrations. These concentrations are those measured from the XAD-2 resin. Other researchers have noted that passage of colloids through a 0.7 µm filter can cause the apparent dissolved concentration of PCBs to be significantly greater than the truly dissolved concentrations (14, 15). This can lead to an apparent decrease in the suspended sediment/water partitioning coefficient (Kd) with increasing TSM (the so-called “solids concentration effect”). The effect of colloids will be discussed later. The highest dissolved phase PCB concentrations were found during the March and August sampling events. The highest apparent dissolved ΣPCB concentration occurred in August in Zone 3 (1700 pg L-1). The November 2002 cruise yielded the lowest dissolved phase PCB concentrations, with an average ((standard deviation) of 780 ( 83 pg L-1. No obvious correlation between river flow (measured at Philadelphia) and PCB concentrations was observed in this data set, although DRBC did observe that PCB concentrations were diluted at high flow rates in the larger data set of ambient PCB concentrations collected for the model calibration (16). Average flows were 2300, 4300, 350, and 4000 cfs during the March, May, August, and November cruises, respectively. Zones 3 and 4, which are directly downstream of the Philadelphia-Camden urban area, had the highest concentrations in all four cruises. The congener profiles of these dissolved phase samples were dominated by the tri-, tetra-, and penta-chlorinated congeners. Water Particle Phase PCB Concentrations. These were highest during the March cruise, with ΣPCB concentrations ranging from 1970 to 4900 pg L-1. Water particle ΣPCB concentrations were lowest during the August cruise, averaging ((SD) 880 ( 245 pg L-1. As for the River regions, there was no particular zone that had high concentrations over all five cruises, although Zone 5 had the highest average value of 2700 ( 1600 pg L-1. No obvious correlation was observed between particle phase PCB concentrations and river flow or dissolved PCB concentration. Homologues 4, 5, and 6 tend to dominate the water particle samples, with the very high molecular weight groups, 9 and 10, playing larger roles in Zones 5 and 6. Organic-carbon normalized ΣPCB concentrations in the suspended matter ranged from 0.61 to 3.51 µg gOC-1, with most values falling between 1 and 2 µg gOC-1. Normalizing to OC reduces the relative standard deviation of the particle-phase ΣPCB concentrations by ∼20% in Zones 2, 3, 4, and 6, and by nearly 30% in Zone 5. PCB 209. Another interesting aspect of the Delaware River is the presence of the decachlorinated congener, PCB 209, which comprised 1, there is net absorption from the air to the water and if f < 1, there is net volatilization from the water to the air. If f ) 1, then the system is at equilibrium. Bruhn et al. (9) used this equation and estimated the uncertainty (coefficient of variance) in Ca, Cd, and Kaw. The propagated uncertainty in f was greater than 100%, primarily because of the high coefficient of variation for Kaw (∼80%). They, therefore, concluded that it was not possible to determine the direction of the net airwater exchange for most PCB congeners. We developed a different approach to determining the fugacity ratios, which is more rigorous in that it treats the measurement uncertainty in parameters such as Cd and Cg differently from the systematic uncertainty in Kaw. First, the correction for sorption to DOC was applied, that lowers Cd:
Cd )
Cd,a 1 + KDOC ‚ DOC
(5)
So that the fugacity ratio becomes this:
f)
Cg ‚ (1 + KDOC ‚ DOC) Cd,a ‚ Kaw
(6)
In this equation, DOC, KDOC, Cd,a, and Cg were obtained from simultaneous measurements. Thus to calculate the fugacity ratios, the sample-specific, DOC, KDOC, Cd,a, and Cg were used, rather than using average values of KDOC. The key parameter in the calculation of the fugacity ratios is the Henry’s Law constant, Kaw. Because Kaw is a physical constant that is knowable (if not readily measurable), its uncertainty should not be propagated along with the random measurement uncertainty associated with Cg and Cd. Instead, the error analysis approach taken here is to calculate the fugacity ratios under the most conservative assumptions (i.e., those that produce the highest value of f when f < 1, and the lowest value of f when f > 1), and then to determine whether the calculated fugacity ratios were different from 1 at the 95% confidence level. In order to apply the most conservative conditions in calculating f, it is necessary to use the lowest reasonable value for Kaw. The primary criticism of our previous air-water exchange work presented by Goss et al. (8) regarded the selection of appropriate Henry’s Law values for use in air-water gas exchange calculations in order to avoid misinterpretation of results (8). To obtain the most conservative (highest) fugacity ratio, the minimum Henry’s Law value for each congener
was chosen from among literature values from four studies that determined Kaw experimentally (23-26) and one study in which available experimental data was evaluated (10). The Kaw values from these five studies are presented in Supporting Information Table S-7. Kaw varies with both temperature and salinity. Kaw should be corrected for temperature via eq 7:
ln KawT2 ) ln Kaw -
[ ](
∆Haw 1 1 R T1 T2
)
(7)
where ∆Haw is the enthalpy of phase change, R is the universal gas constant, T1 is the temperature at which the Henry’s Law constant was measured and T2 is the target temperature. The temperature correction was applied to all but the August data, because at lower temperatures the correction lowers Kaw, while at higher temperatures, the correction increases Kaw. The enthalpy measurements of Bamford et al. (26) were used because this was the only study to report ∆H for all 209 PCB congeners. Normally, a salinity correction would also be applied to Kaw:
Kaw,salt ) Kaw ‚10(Ks ‚[salt])
(8)
where Kaw is the Henry’s Law constant at 0 salinity, Kaw,salt is the salinity-corrected Henry’s Law constant, Ks is the Setschenow constant, and [salt] is the molar salinity of the water. For purposes of calculating f, this correction was not applied since it increases Kaw. The salinity correction was, however, employed for the calculation of fluxes below. The calculated fugacity ratios displayed a log-normal distribution (Figure 2). Thus they were log transformed and examined to determine whether log f < 0 at the 95% confidence level using a single-sample t-test. Figure 2 displays a histogram of the log f values for all data points (all congeners in all cruises and Zones), demonstrating that the median value of log f is much less than 0. This test was also performed on a congener-specific basis using the pooled data for all measurements on all cruises in all Zones. The t-test indicated that, at a 95% confidence interval, the net movement of PCBs for all but seven high molecular weight congeners (174, 177, 202 + 171 + 176, 180, 201, 194, 209) is from the river to the air, or net volatilization. Because of the way in which the data was pooled, this conclusion applies to the overall behavior of the congeners on a net basis in all Zones and all seasons. It is possible that under some specific conditions and over short time periods the net flux could reverse and result in gaseous deposition, but integrated over the whole river (Zones 2-6) and over the course of the year, the net behavior of these congeners results in volatilization. It was not expected that the river would act as a source of PCBs to the air in all zones and overall seasons. Both Philadelphia and Camden are old, heavily populated cities with many landfills, abandoned manufacturing sites, and operating industrial facilities. Based on the high year-round air concentrations measured at the Camden NJADN site (27), it was thought that nearby Zone 3 could be a sink for PCBs from the atmosphere. The PCB TMDL water quality model does show that under summer conditions, when high air temperatures lead to very high gas-phase PCB concentrations in Zone 3, the net air-water exchange flux does become absorptive for short periods of time (days) in that Zone. Fluxes. As described above, the determination of the direction of the air-water exchange flux is not trivial. Having established that the net flux results in volatilization, however, it is logical to proceed to calculate the magnitude of the fluxes. This calculation is even more complicated than the calculation of the fugacity ratio because it involves the mass transfer coefficient, vaw, a parameter that is associated with VOL. 41, NO. 4, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1155
FIGURE 2. Histograms of fugacity ratios (top panel) and log fugacity ratios (bottom panel) for all congeners in all samples. a high degree of uncertainty (28). Despite this complexity, the calculation of fluxes is absolutely essential to the construction of the ecosystem mass balances for PCBs in most surface waters, and certainly in the Delaware. The fluxes presented here are the best estimates possible given the state of the science and are undertaken to further our understanding of the fate of PCBs in the Delaware River. For these calculations, many of the assumptions and approaches used above to calculate the most conservative fugacity ratios are abandoned. In particular, the salinity and temperature corrections to Kaw are applied to all samples, and the Kaw values chosen are different. For the flux calculations, we believe it is logical to use Kaw values from one study to ensure uniformity, and we persist in our belief that the studies of Bamford et al. (26, 29), which measured not only Kaw but also ∆Haw, represent the best available data set for calculating temperature-corrected values of Henry’s law constants for all 209 PCBs. Kaw was corrected for salinity via eq 8. As in other studies (5, 30), the Setschenow constant, KS, was assumed to be 0.3 for all congeners. Salinity was negligible in Zones 2, 3, and 4 but it increased Henry’s law by 7% in zone 5 and 25% in zone 6. PCBs in the truly dissolved phase and the gas-phase participate in air-water gas exchange. In order to undergo air-water exchange, the compound must travel across both the air and water boundary layers of the air-water interface. Each compound has its own mass transfer coefficient, vaw, which is a measure of its resistance to travel through these 1156
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 4, 2007
gaseous and liquid films (31). vaw accounts for the compound’s resistance to transfer in both air (va) and water (vw) (31):
1 1 1 ) + vaw vw vaKaw
(9)
The framework for calculating va and vw has been published (5) and will not be repeated here. The mass transfer coefficients are primarily a function of the compound’s physical-chemical properties (primarily molecular weight and Henry’s Law constant) and wind speed. Values of va and vw calculated here were similar to those previously published (see Table 3 of ref 5) The gaseous absorption flux (Fa in ng m-2 day-1) quantifies movement of compounds from the atmosphere into the water column, and is calculated as follows:
Cg Fa ) vaw Kaw
(10)
Following the trends in Cg, the absorption fluxes were lowest during the May cruise and highest during the November cruise (Table 1), and Zone 4 generally had the highest absorption fluxes among all seasons, while Zone 6 commonly had the lowest.
TABLE 1. ΣPCB Fluxes for all Cruises in all Zonesa ΣPCB fluxes (ng m-2 d-1) corrected corrected zone absorption volatilization volatilizaton net flux net flux March
2 3 4 5 6
-76 -160 -180 -140 NA
750 1700 2600 1500 NA
2 3 4 5 6
-73
970 2000 1700 2700 950
2 3 4 5 6
-110
2 3 4 5 6
-180
-49 -45 -43
-160 -130 -61
-190 -95 -68
590 1700 2500 1200 NA
670 1600 2400 1300 NA
510 1500 2100 1100 NA
630 1300 1200 1500 550
900 2000 1600 2700 910
560 1400 1000 1700 610
1900 2800 3700 1200 420
1100 2100 1900 580 160
1800 2700 3600 1100 360
1000 2000 1700 450 110
2200 2600 3200 2200 910
November 1500 2100 2100 870 680
2000 2500 3000 2100 850
1300 1900 1900 780 610
May
August
a The corrected volatilization and corrected net flux values were corrected for sorption of PCBs to DOC (see text).
The gaseous volatilization flux (Fv) quantifies movement of compounds from the water column into the atmosphere, and is calculated as follows:
Fv ) vaw ‚ Cd
(11)
The volatilization fluxes were lowest during the March sampling cruise and they were highest during the November expedition (Table 1). These trends are due to the trends in wind speeds, which were generally lowest in March and highest in November. The net diffusive gas exchange flux (Fnet) is calculated by subtracting the absorption flux from the volatilization flux, which is equivalent to the following equation:
(
Fnet ) vaw
Ca Cd Kaw
)
(12)
A positive flux indicates net volatilization from the water column to the air, while a negative flux value signifies net absorption from the atmosphere into the water column. Net fluxes of ΣPCBs and homologues were calculated as the sum of fluxes of individual congeners. In keeping with the fugacity ratios, net fluxes for ΣPCBs and most homologues were positive for all Zones during all seasons, and highly chlorinated homologues were calculated to have small fluxes, indicating they were close to equilibrium. As Table 1 demonstrates, the absorptive fluxes were typically less than 10% of the volatilization fluxes. The ΣPCB fluxes were highest during the August and November 2002 sampling expeditions, with both cruises having an average flux value around 1250 ng m-2 day-1. The net fluxes were lowest during the March trip, with an average value of 870 ng m-2 day-1 over all zones. Because absorption fluxes are small, these trends are driven by the volatilization flux, which is in turn driven by wind speed. Generally, the tri- and tetrachlorinated homologues dominated the flux profiles, accounting for more than 70% of the total flux. The measured
dissolved phase PCBs were corrected for DOC sorption, as discussed above, so all calculations performed here include only the “truly” dissolved PCBs and not those sorbed to colloids. This correction reduces the magnitude of the overall flux by an average of 35% and decreased volatilization flux by 30%. Total PCB fluxes were higher than those calculated for Raritan Bay (400 ng m-2 d-1), but were lower than those calculated for New York Harbor (2100 ng m-2 d-1) (5). The average ΣPCB fluxes seen here are higher than those seen in Baltimore Harbor (350 ng m-2 d-1) and Chesapeake Bay (330 ng m-2 d-1) and are closer to the maximum fluxes (1240 ng m-2 d-1 and 1150 ng m-2 d-1, respectively) calculated at those sites using the same Kaw values and deriving vaw using the same method presented here (28). Lake Michigan exhibited much lower ΣPCB fluxes (24-220 ng m-2 d-1) than those seen in the Delaware, but that study used different Kaw values, which will also affect vaw (32). Comparison with TMDL Model. A crude approximation of the yearly loss of PCBs due to volatilization can be obtained by multiplying the average fluxes (over all cruises) for each Zone by the surface area of each Zone and by 365 days. This results in an annual loss of ∼380 kg ΣPCBs y-1 and ∼60 kg y-1 of penta-PCB. The TMDL model estimates an annual volatilization loss of ∼20 kg y-1 of penta-PCB. Because the TMDL assumed ΣPCBs ) 4 × penta-PCB based upon ambient water concentrations in Zones 2-5, this translates to ∼90 kg/yr for ΣPCBs, although this conversion factor probably does not apply to the volatilization flux. In this study, the ΣPCB flux was about 6.6 times the penta-PCB flux. Nevertheless, the volatilization fluxes calculated here are within a factor of 4 of those calculated by the PCB TMDL water quality model. This level of agreement is not unreasonable considering that the TMDL model used different algorithms for calculating va and vw and a much wider range of wind speed, temperature, and PCB concentration conditions. According to the DRBC TMDL model, volatilization is the most important loss process for penta-PCBs. Approximately 60 000 mg of penta-PCBs are lost from the system due to volatilization each day, compared to 44 000 mg penta-PCB d-1 lost due to sediment burial. Water advected out of the river accounts for the removal of about 725 mg d-1 of penta-PCBs. Both studies are in agreement, however, that the net direction of air/water exchange for all zones is from the water to the air.
Acknowledgments This work was funded by a grant from the Delaware River Basin Commission. Support from the New Jersey Agricultural Experiment Station is also gratefully acknowledged. We thank the members of A.A.R.’s dissertation committee (Kevin Farley, John Reinfelder, and Chris Uchrin) for their useful comments.
Supporting Information Available Tables showing meteorological data and water quality parameters, atmospheric gas-phase ΣPCB concentrations, aerosol phase ΣPCB concentrations, apparent dissolved water ΣPCB concentrations, particulate water ΣPCB concentrations, and published Henry’s Law constants. This material is available free of charge via the Internet at http://pubs.acs.org.
Literature Cited (1) NJDEP. PCB/Dioxin Fish Consumption Advisory; New Jersey Department of Environmental Protection: Trenton, NJ, 2002. (2) Fikslin, T. F.; Suk, N. Total Maximum Daily Loads for Polychlorinated Biphenyls (PCBs) for Zones 2-5 of the Tidal Delaware River; Delaware River Basin Commission: West Trenton, NJ, 2003. (3) Simcik, M. F.; Zhang, H.; Eisenreich, S. J.; Franz, T. P. Urban contamination of the Chicago/Coastal Lake Michigan atmoVOL. 41, NO. 4, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1157
(4)
(5)
(6)
(7)
(8)
(9) (10)
(11)
(12) (13) (14)
(15)
(16)
(17)
sphere by PCBs and PAHs during AEOLOS. Environ. Sci. Technol. 1997, 31, 2141-2147. Farley, K. J.; Thomann, R. V.; Cooney, T. F. I.; Damiani, D. R.; Wands, J. R. An Integrated Model of Organic Chemical Fate and Bioaccumulation in the Hudson River Estuary; The Hudson River Foundation: New York, NY, 1999. Totten, L. A.; Brunciak, P. A.; Gigliotti, C. L.; Dachs, J.; Glenn, IV, T. R.; Nelson, E. D.; Eisenreich, S. J. Dynamic air-water exchange of polychlorinated biphenyls in the New York-New Jersey Harbor Estuary. Environ. Sci. Technol. 2001, 35, 38343840. Totten, L. A.; Gigliotti, C. L.; Offenberg, J. H.; Baker, J. E.; Eisenreich, S. J. Reevaluation of air-water exchange fluxes of PCBs in Green Bay and southern Lake Michigan. Environ. Sci. Technol. 2003, 37, 1739-1743. Totten, L. A.; Gigliotti, C. L.; VanRy, D. A.; Offenberg, J. H.; Nelson, E. D.; Dachs, J.; Reinfelder, J. R.; Eisenreich, S. J. Atmospheric concentrations and deposition of polychlorinated biphenyls to the Hudson River Estuary. Environ. Sci. Technol. 2004, 38, 25682573. Goss, K.; Wania, F.; McLachlan, M. S.; Mackay, D.; Schwarzenbach, R. P. Comment on “Reevaluation of air-water exchange fluxes of PCBs in Green Bay and southern Lake Michigan”. Environ. Sci. Technol. 2004, 38, 1626-1628. Bruhn, R.; Lakaschus, S.; McLachlan, M. S. Air/sea gas exchange of PCBs in the southern Baltic Sea. Atmos. Env. 2003, 37, 34453454. Li, N.; Wania, F.; Lei, Y. D.; Daly, G. L. A comprehensive and critical compilation, evaluation, and selection of physicalchemical property data for selected polychlorinated biphenyls. J. Phys. Chem. Ref. Data 2003, 32, 1535-1590. Baker, J. E.; Totten, L. A.; Gigliotti, C. L.; Offenberg, J. H.; Eisenreich, S. J.; Bamford, H. A.; Huie, R. E.; Poster, D. L. Response to comment on “Reevalutaion of air-water exchange fluxes of PCBs in Green Bay and southern Lake Michigan”. Environ. Sci. Technol. 2004, 38, 1629-1632. Rowe, A. A. Interactions Of Polychlorinated Biphenyls With The Air, Water, And Sediments Of The Delaware River Estuary. Ph.D. Dissertation, Rutgers University, NJ, 2006. Kucklick, J. R.; Harvey, H. R.; Ostrom, P. H.; Ostrom, N. E.; Baker, J. E. Organochlorine dynamics in the pelagic food web of Lake Baikal. Environ. Toxicol. Chem. 1996, 15, 1388-1400. Poster, D. L.; Baker, J. E. Influence of submicron particles on hydrophobic organic contaminants in precipitation. 1. Concentrations and distributions of polycyclic aromatic hydrocarbons and polychlorinated biphenyls in rainwater. Environ. Sci. Technol. 1996, 30, 341-348. Jiang, X.; Martens, D.; Schramm, K. W.; A., K.; Xu, S. F.; Wang, L. S. Polychlorinated organic compounds (PCOCs) in waters, suspended solids and sediments of the Yangtse River. Chemosphere 2000, 41, 901-905. Yagecic, J.; Cavallo, G.; MacGillivray, R.; Liao, D.; Suk, N.; Fikslin, T. J. Calibration of the PCB Water Quality Model for the Delaware Estuary for Penta-PCBs and Carbon; Delaware River Basin Commission: West Trenton, NJ, 2003. Frame, G. M.; Cochran, J. W.; Boewadt, S. S. Complete PCB congener distributions for 17 Aroclor mixtures determined by 3 HRGC systems optimized for comprehensive, quantitative,
1158
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 4, 2007
(18)
(19) (20) (21) (22) (23) (24)
(25)
(26)
(27)
(28)
(29) (30) (31) (32)
congener-specific analysis. J. High Resolut. Chromatogr. 1996, 19, 657-668. Baker, J. E.; Toaspern, M.; Liebert, D.; Ko, F. Decachlorobiphenyl: using a non-Aroclor PCB as a tracer of estuarine processes. Poster presented at the Society of Environmental Toxicology Chemistry Annual Meeting, Portland, OR,2004. Gamboa, J. A.; Bohe, A. E.; Pasquevich, D. M. Carbochlorination of TiO2. Thermochim. Acta 1999, 334, 131-139. Hornbuckle, K. C.; Achman, D. R.; Eisenreich, S. J. Over-water and over-land polychlorinated biphenyls in Green Bay, Lake Michigan. Environ. Sci. Technol. 1993, 27, 87-98. Karickhoff, S. W.; Brown, D. S.; Scott, T. A. Sorption of hydrophobic pollutants on natural sediments. Water Res. 1979, 13, 241-248. Burkhard, L. P. Estimating Dissolved Organic Carbon Partition Coefficients for Nonionic Organic Chemicals. Environ. Sci. Technol. 2000, 34, 4663-4668. Murphy, T. J.; Mullin, M. D.; Meyer, J. A. Equilibration of polychlorinated biphenyls and toxaphene with air and water. Environ. Sci. Technol. 1987, 21, 155-162. Dunnivant, F. M.; Elzerman, A. W. Quantitative structureproperty relationships for aqueous solubilities and Henry’s Law constants of polychlorinated biphenyls. Environ. Sci. Technol. 1992, 26, 1567-1573. Brunner, S.; Hornung, E.; Santl, H.; Wolff, E.; Piringer, O. G.; Altschuh, J.; Bruggemann, R. Henry’s Law constants for polychlorinated biphenyls: experimental determination and structure-property relationships. Environ. Sci. Technol. 1990, 24, 1751-1754. Bamford, H. A.; Poster, D. L.; Huie, R. E.; Baker, J. E. Using extrathermodynamic relationships to model the temperature dependence of Henry’s Law constants of 209 PCB congeners. Environ. Sci. Technol. 2002, 36, 4395-4402. Totten, L. A.; Panangadan, M. V.; Eisenreich, S. J.; Cavallo, G. J.; Fikslin, T. J. Direct and indirect atmospheric deposition of PCBs to the Delaware River Watershed. Environ. Sci. Technol. 2006, 40, 2171-2176. Bamford, H. A.; Ko, F. C.; Baker, J. E. Seasonal and annual airwater exchange of polychlorinated biphenyls across Baltimore Harbor and the northern Chesapeake Bay. Environ. Sci. Technol. 2002, 36, 4245-4252. Bamford, H. A.; Poster, D. L.; Baker, J. E. Henry’s Law constants of polychlorinated biphenyl congeners and their variation with temperature. J. Chem. Eng. Data 2000, 45, 1069-1074. Brownawell, B. J. Ph.D. Thesis, Massachusetts Institute of Technology, 1986. Schwarzenbach, R. P.; Gschwend, P. M.; Imboden, D. M. Environmental Organic Chemistry; John Wiley and Sons, Inc.: New York, NY, 1993. Hornbuckle, K. C.; Sweet, C. W.; Pearson, R. F.; Swackhamer, D. L.; Eisenreich, S. J. Assessing annual water-air fluxes of polychlorinated biphenyls in Lake Michigan. Environ. Sci. Technol. 1995, 29, 869-877.
Received for review July 28, 2006. Revised manuscript received November 3, 2006. Accepted November 14, 2006. ES061797I
