Declining PCB Concentrations in the U.K. Atmosphere: Evidence and

Declining PCB Concentrations in the U.K. Atmosphere: Evidence and ..... The TOMPs ambient air monitoring network – Continuous data on UK air quality...
0 downloads 0 Views 102KB Size
Environ. Sci. Technol. 2000, 34, 863-869

Declining PCB Concentrations in the U.K. Atmosphere: Evidence and Possible Causes ANDREW J. SWEETMAN* AND KEVIN C. JONES Department of Environmental Science, Institute of Environmental and Natural Sciences, Lancaster University, Lancaster, LA1 4YQ, U.K.

PCB air concentrations have been measured at a meteorological site in northwest England since 1992. Examination of this data set, comprising over 200 data points, suggests that PCB levels are decreasing with average congener specific half-lives ranging from approximately 2 to 6 yr. With the exception of congener 52, which shows the steepest decline, the slopes of other ICES congeners included in this study (i.e., 28, 101, 118, 153, and 138) were not found to be significantly different from each other. A U.K. mass balance model has been used to examine which factors are likely to be controlling present and future air concentrations. This allowed a range of fate scenarios to be examined and the controlling fate processes to be scrutinized. Estimates of fluxes using contemporary soil and air concentrations suggest that the observed longterm decrease of PCB levels in U.K. air is likely to be influenced by several factors, including existing primary emissions and recycling, volatilization from soil, advective losses from the U.K. atmosphere, reaction in the atmosphere, and soil fate processes such as microbial degradation.

Introduction Polychlorinated biphenyls (PCBs) are now ubiquitous within the environment as a result of their widespread production and use and their semivolatile nature. While the atmosphere has provided an effective transport medium, terrestrial soils have provided the largest repository. During the 1950s through the 1970s, net deposition to soils from the atmosphere was occurring as a result of high primary emissions and hence air concentrations (1-3). Following restrictions on production and use, direct emissions to the atmosphere had reduced by the late 1970s/1980s. However, by this time the soil repository had built up to a level that, as atmospheric concentrations declined, it could provide a secondary source of PCBs back to the atmosphere (4). As a result, there is much interest in the long-term fate of PCBs in the environment; notably the balance between primary and secondary (recycling) sources to the atmosphere, their global redistribution (5), and long-term dynamics governed by different physical, chemical, and biologically mediated loss processes. These processes are summarized in Figure 1. Among the processes that “remove” PCBs from the environment and the “pool” available for recycling are (i) reaction in the atmosphere with OH- radicals; (ii) microbially mediated degradation in soil and the formation of irreversibly bound residues in soils and * Corresponding author e-mail: [email protected]; phone: +44 1524 593300; fax: +44 1524 593985. 10.1021/es9906296 CCC: $19.00 Published on Web 01/19/2000

 2000 American Chemical Society

FIGURE 1. Conceptual representation of the fate processes that effect PCBs in the environment. sediments; (iii) burial in soils, sediments, peat, and ice; and (iv) incorporation into the deep oceans. However, at the present time the relative importance of these processes is unclear, and there is a lack of long-term monitoring data on concentrations from which to quantify the “rates of disappearance” of PCBs and to gain clues about the relative importance of the different processes governing these rates. This paper therefore represents a compilation and examination of detailed atmospheric concentration data obtained at one site in the U.K. over several years and assesses the results using a simple mass balance model to evaluate which fate processes may be controlling the current and future fate of PCBs in the environment.

Sources of Atmospheric Concentration Data The following section details the sampling site and the sampling campaigns used to compile the long-term monitoring data discussed in this manuscript. Hazelrigg Site. The sampling site used for these studies is a meteorological station located in a semirural location outside Lancaster on the northwest coast of England, approximately 5 km from the Irish Sea (52°2′ N, 2°45′ W). The data used in this study comprised quarterly, biweekly, weekly, and daily samples from several different studies, which are briefly discussed below. Quarterly Samples. The U.K. Department of the Environment, Transport and the Regions has been supporting a long-term air monitoring network to determine environmental levels of toxic organic micropollutantssthe so-called TOMPs program. The Hazelrigg site has been included in this since autumn 1992 and is still operational. Samples are taken biweekly and the results pooled to provide quarterly averages. Results from the TOMPs study are presented in Coleman et al. (6) and the National Air Quality Information Archive (7). Weekly and Biweekly Air. Samples taken in this section were collected under two projects which have been reported in the literature. The first study, comprising weekly samples, was conducted from July 1995 to September 1995 and is reported in Ockenden et al. (8). The second, longer term study started in January 1996 and continued until December of that year and comprised biweekly samples. The results are presented in Thomas et al. (9). Daily Samples. A study carried out by Lee et al. (10) collected 24-h air samples on 161 occasions during MarchOctober and December 1994 at the Hazelrigg site. This study was undertaken to examine some of the factors that are likely to control PCB air concentrations, including meteorological parameters such as temperature, wind speed, and direction. VOL. 34, NO. 5, 2000 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

863

FIGURE 2. Temporal variation of atmospheric PCB fugacities in U.K. air (1992-1998). Over the study period, ΣPCB concentrations ranged by a factor of 7 between 54 and 375 pg m-3 with individual congener data showing significant but weak relationships between partial pressure and inverse temperature. Air Sampling and Analytical Methodology. All air samples were collected using Hi-Vols (General Metal Works, model GPS1) with glass fiber filters and PUF plugs with typical collection volumes of between 400 and 800 m3. The analytical methods for air samples collected during these studies have been presented elsewhere (9, 10) but generally comprise Soxhlet extraction of combined filter and PUF plug, silica gel chromatography, and quantitation using GC-ECD or GC/ MS. To ensure that the methods were comparable, they were subject to the same rigorous QA/QC procedures that included standard reference materials. Calculation of Fugacities and Multiple Regression Analysis. To allow comparison of air concentration data quantified in samples taken under different conditions, the raw air concentrations were converted to fugacities or partial 864

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 34, NO. 5, 2000

pressures. This was achieved using the following equation:

fair )

Cair (pg m-3) × RT 1E-12 × MW

(1)

where Cair is the vapor-phase concentration; fair is the air fugacity (Pa); T is the temperature in Kelvin; R is the gas constant (8.314 J mol-1 K-1); MW is the molecular weight (g mol-1). Plots of fugacity against time for each of the congeners over the period of the three studies are contained in Figure 2. Despite a high degree of scatter, the regressions showed a strong negative correlation for each of the congeners with significance exceeding 99%. Congener 52 showed the steepest slope (significant at 95%) and hence the shortest half-life, ranging between 1.7 and 2.3 yr. Congeners 28, 101, 118, 153, and 138 showed similar half-lives ranging from 2 to 6 yr with PCB-138 showing the greatest range of values (3-12 yr). It is important to note that although these half-life values

TABLE 1. Regression Statistics for Temporal Trends no. of slope data congener (yr-1) points PCB-28 PCB-52 PCB-101 PCB-118 PCB-153 PCB-138

0.29 0.37 0.25 0.22 0.20 0.14

167 215 200 210 113 139

r2 0.112 0.313 0.142 0.096 0.129 0.073

half-life (yr) (95% CI) significance min max of slope (%) 1.7 1.6 2.1 2.3 2.3 3.0

4.2 2.3 4.3 5.5 6.6 11.9

>99 >99 >99 >99 >99 >99

represent the net rate of disappearance from the atmosphere, because the air is in rapid exchange with the surface and is resupplied by it, they reflect a combination of degradative processes in terrestrial/aquatic surfaces and the atmosphere as well as long-range transport, i.e., a “multimedia clearance rate”. Hites and co-workers (11, 12) and others (e.g., refs 7, 13, and 14) have clearly demonstrated that PCB air concentrations can show strong temperature dependency, which provides powerful evidence for their environmental recycling. This highlights the need to remove seasonal or short-term temperature-controlled “noise” from the data set when applicable. To check for evidence of temperature dependence, natural logarithms of the fugacities for each congener were plotted against reciprocal temperature (1/T). A highly significant (>99%) correlation was found for each congener (PCB-28 >95%), although the relationships were statistically weak, generally accounting for 5-10% of the observed variability. Data sets of this type, by their nature, exhibit considerable noise as a result of environmental variables, such as wind speed, air mass origin, short- and long-term temperature fluctuations, and experimental variables such as sampling period and high-volume air sampler reliability. When considered separately and particularly when shortterm events are taken into account, the significance of the temperature dependence increases (e.g., refs 10 and 13). This weak relationship with temperature was expected as this study has combined three data sets that span over several years. Overall, these data suggest that temperature, over the time span covered, exerts a significant but relatively small effect. Assuming that a first-order decay model would be the most appropriate, the calculated fugacities were again log(e) transformed but then regressed with time. Initially, each data point was weighted, based on the number of days sampled before the regression analysis was carried out. However, probability plots of the standardized residuals showed a highly skewed distribution. On further examination of the data, it appeared that each data type (e.g., daily, weekly, quarterly) were exhibiting similar variances despite different exposure times. This observation is difficult to explain as the variance of quarterly averaged samples should be less than daily samples. However, to avoid skewed residuals, the regression analyses were carried out without weighting the data. The resulting regression slopes and their significance are contained in Table 1. To place these half-life values in perspective, we have compared data reported by others in a range of compartments. For example, Baker and Eisenreich (15) and Panshin and Hites (11, 16) found no evidence for a reduction in atmospheric PCB levels over Lake Superior, in Bermuda, and in Bloomington, IN, respectively. Jeremiason et al. (17) and Pearson et al. (18) have provided evidence that concentrations of PCBs in water from the Great Lakes have been reducing with a half-life between 2.5 and 9 yr. Fish samples also taken from the Great Lakes (19, 20) and from the Baltic Sea (21) have shown half-lives ranging from 4 to 10 yr. Bignert et al. (22) reported similar time trends for a wide range of biota

samples taken from Swedish freshwater lakes and the Baltic, which suggested that half-lives ranged from 4 to 14 yr. More recently, Simcik et al. (23) reported congener-specific atmospheric half-lives ranging from 0.5 to 5.9 yr from IADN sites near Lakes Michigan and Erie and near Chicago. The U.K. air clearance rate is generally in line with these data for other environmental matrixes.

The U.K. PCB Mass Balance: Possible Controlling Factors of Long-Term PCB Air Concentrations As mentioned earlier, there are several factors that could potentially affect or control long-term atmospheric PCB concentrations. This section discusses some of these factors and considers which are likely to be the most important. An important finding of this study was that all congeners, with the exception of PCB-52, showed similar long-term atmospheric half-lives. If this observation is correct and there are no substantial congener differences, this suggests that the controlling factor(s) may not be congener specific. However, this observation may also be a result of the noise in the data which has prevented any differences being observed. For the purposes of these calculations, the observations taken at Hazelrigg have been assumed to be typical for the U.K. as a whole. To investigate these processes further, we have calculated a mass balance for the U.K. in 1993 taking into account the processes shown in Figure 1. The mass balance approach was adopted not only to attempt to quantify, and hence rank, a range of potentially important fate processes but to identify any possible shortfalls or missing factors. The processes included were hydroxyl radical (OH) degradation, wet and dry deposition (gaseous and particle associated), volatilization from soil surfaces, long-range transport, and remaining emission sources. The air concentration data for 1993 were calculated from the regression lines taken from Figure 2, while the soil concentration data were taken from a 1993 U.K. wide survey (7). We have assumed for these calculations that the U.K. is covered with bare soil and that vegetation plays no part in the long-term fate of PCBs. However, vegetation may play an important role in the long-term fate of PCBs, for example, photolytic degradation on leaf surfaces may be a significant, but as yet unquantified, additional loss process. The calculation of each of the fate processes included in the mass balance is detailed in the following sections, and the parametrization used is given in Table 2. The results of the calculated mass balance for two selected congeners, PCB52 and PCB-153, are contained in Figure 3. While reading the next section, it is important to bear in mind that, although the overall half-life of individual PCB congeners in U.K. air ranges between 2 and 12 yr, this only represents a total loss of a few kilograms per year from the overall U.K. air burden. Atmospheric Reaction. Anderson and Hites (32) suggested that reaction with OH- radicals is an important congenerspecific loss process for PCBs in the atmosphere with individual congeners exhibiting half-lives on the order of days. Assuming an average atmospheric concentration of 31 pg m-3 for PCB-52 and a half-life of 17 days at 10 °C (32), then the OH- radical reaction would account for a total loss of approximately 125 kg yr-1. However, at the same temperature, PCB-153 has a half-life of 55 days (32), which results in a loss of 6 kg yr-1, based on an ambient concentration of 8 pg m-3. These observations suggest that if the OH- radical pathway were a dominant pathway then the difference in reaction kinetics would most likely result in a significantly greater reduction in the lighter congeners than heavier ones over time. However, the apparent lack of difference between congener half-lives in the measured data would suggest that this fate pathway, although important, is not the dominant influence on the trends observed. VOL. 34, NO. 5, 2000 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

865

TABLE 2. Parameter Values for Calculations parameter

value

units

source

Kp PCB-52 Kp PCB-153 TSP F particle density log Koa PCB-52 log Koa PCB-153 H PCB-52 H PCB-135 Wgas PCB-52 Wgas PCB-153 UR Cair(g) PCB-52 Cair(g) PCB-153 atmos height Asoil Wpart PCB-52 Wpart PCB-153 UP UG Csoil PCB-52 Csoil PCB-153 L W a φ θ F soil bulk density foc Koc PCB-52 Koc PCB-153 Dag Dw l

5.29E-04 at 10 °C 3.49E-03 at10 °C 30 2000 9.1 at 10 °C 10.500 at 10 °C 25.1 at 10 °C 22.6 at10 °C 93.7 104.1 2.74E-3 1.06E-13 2.22E-14 1000 2.34E11 180000 150000 10.8 2.2 2.25E-04 5.10E-04 0.01 5 0.2 0.3 0.5 1500 0.09 5.16E+05 3.26E+06 4.320E+03 4.32E-01

m3 µg-1 m3 µg-1 µg m-3 kg m-3 dimensionless dimensionless Pa m3 mol-1 Pa m3 mol-1 dimensionless dimensionless m d-1 mol m-3 mol m-3 m m2 dimensionless dimensionless m h-1 m h-1 µg cm-3 µg cm-3 cm cm dimensionless dimensionless dimensionless kg m-3 dimensionless mL g-1 mL g-1 cm2 d-1 cm2 d-1

calcd from Koa (24) calcd from Koa (24) (25) (25) (26) (26) (27) (27) calcd calcd U.K. meteorological data calcd from regression calcd from regression (25) (27) (28) (28) (25) (29) (3) (3) ( 3) ( 3) (25) (25) (25) (25) ( 3) calcd from Kow (30) calcd from Kow (30) (31) (31)

sociated with particulate material (φ) will be an important factor in controlling which processes dominate and is defined in eq 2: φ)

TSP × Kp where log Kp ) 0.55 log Koa - 8.23 (24) 1 + TSP × Kp

(2)

with φ as the fraction associated with particles, Kp as the gas particle partition coefficient (m3 µg-1), TSP as total suspended particulate material (µg m3), and Koa as the octanol-air partition coefficient. The gas phase dominates for both PCB congeners, although particle association is very temperature sensitive. For example, for PCB-153 the fraction associated with particles can range from 12.7% to 3.4% at 5 and 25 °C, respectively. Wet and dry deposition processes were calculated using eqs 3-6, and the results are contained in Figure 3. The relative importance of these processes is of course dependent on the degree of association with atmospheric particles. Wet Deposition.

gaseous Fwet,g ) WgasURCair(g)Asoil

FIGURE 3. Calculated PCB mass balances for the contemporary U.K. situation (1993). Volatilization flux is calculated as a 5-year average and is net of soil microbial degradation with a half-life of 20 years. Deposition Processes. Deposition processes have been divided into wet and dry deposition, each having a gaseous and particulate component. The fraction of chemical as866

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 34, NO. 5, 2000

(3)

where Fwet,g is the gaseous wet deposition flux, mol d-1; Wgas is the scavenging ratio for the gaseous phase; UR is the precipitation rate, m d-1; Cair(g) is the gaseous phase concentration, mol m-3; and Asoil is the U.K. surface area with

Wgas )

RT H

where R is the gas constant, 8.314 J mol-1 K-1; T is the

temperature, K; and H is the Henry’s law constant, Pa m3 mol-1

particle bound Fwet,p ) WpartURCair(p)Asoil

(4)

where Fwet,p is the particle-associated wet deposition flux, mol d-1; Wpart is the particle washout ratio; and Cair(p) is the particle-phase concentration, mol m-3. Dry Deposition.

particle bound Fdry,p ) UpCair(p)Asoil

(5)

where Fdry,p is the particle-associated dry deposition flux, mol d-1; and Up is the dry particle deposition flux, m d-1.

gaseous Fdry,g ) UGCair(g)Asoil

(6)

where Fdry,g is the gaseous dry deposition flux, mol d-1; and UG is the dry gaseous deposition velocity, m d-1. Dry gaseous deposition is the dominant deposition process for both congeners, accounting for 135 and 28 kg yr-1 for PCB-52 and PCB-153, respectively. However, wet particle deposition is also important for PCB-153, especially at lower temperatures. Using mean ambient concentration data, these calculations give values for deposition to the surface (assuming entirely bare soil) of 160 and 60 kg yr-1 for PCB-52 and PCB-153, respectively (see Figure 3). Surface/Air Exchange. To calculate the magnitude of the volatilization flux of PCB congeners from U.K. soils, a flux model based on Jury et al. (33) has been employed, see eqs 7 and 8:

volatilization flux Jbs ) -Csoil exp(- µt)xDE/πt{exp(-L2/4DEt) -

exp(-(L + W)2/4DEt)} (7)

where

DE )

2 [(a10/3DagH′ + θ10/3Dw l )/φ ]

(FbfocKoc + θ + aH′)

(8)

where Jbs is the volatilization flux, mol d-1; Dag is the air-gas diffusion coefficient, cm2 d-1; Csoil is the soil concentration, 2 mol m-3; Dw l , is the water-liquid diffusion coefficient, cm d-1; L is the incorporation depth, cm; H′ is the dimensionless Henry’s law constant; W is the width of contaminant band, set at L cm for this application; Koc is the organic carbon partition coefficient, cm3 g-1; t is time, days; foc is the fraction of organic carbon; a is the volumetric air content; F is the bulk density, g cm-3; θ is the volumetric water content; and φ is the soil porosity. For the theory behind these calculations, the reader is directed toward Jury et al. (31-33). The model was run for PCB-52 and PCB-153 for a 5-yr period, and an average flux was calculated, which yielded a flux of 125 and 110 kg yr-1 for each congener, respectively. These calculations suggest that for the contemporary situation, under the prescribed conditions, volatilization from soil surfaces approximately equals deposition fluxes, accounting for ∼80% of the PCB52 flux and ∼180% of the PCB-152 flux. This suggests that within the errors associated with calculations of this type

that air and soil are close to steady-state conditions. Further support for this observation comes from fugacity calculations carried out by Cousins and Jones (34) using contemporary U.K. air and soil concentration data, which also suggested that near-equilibrium conditions exist for many congeners. They calculated soil-air fugacity ratios ranging from 3.5 for PCB-28 to 4.4 for PCB-153 (assuming a soil organic carbon fraction of 9%, soil bulk density of 1500 kg m-3, and temperature of 10 °C). Being close to equilibrium suggests that little or no net diffusive transfer between soil and air is occurring. However, small changes in temperature (soil surface or air) or air concentration resulting from air mass movement can change this position in either direction. Remaining Primary Sources. There are few data on remaining primary PCB sources in the U.K., although a recent estimate suggested an annual emission to air of 5300-6100 kg ΣPCB (35). Assuming a typical U.K. Aroclor production ratio of 10:5:2 for 1242, 1254, and 1260, respectively (36), this would result in an annual release of 210-240 kg of PCB-52 and between 160 and 180 kg for PCB-153. However, these calculations are extremely tentative as they assume that the releases are not influenced by the physicochemical properties of each congener, which would not be the case if the releases were a result of volatilization from transformer oils or landfills. If these releases are significant, then the apparent lack of steep slopes from the ln(P) vs 1/T plots suggests that they are not acting on a scale local to Hazelrigg, rather they indicate a diffuse source or sources supplying the U.K. atmosphere. Long-Range Transport. It is recognized that the U.K. has been (and probably still is) a net exporter of PCBs via longrange transport, since it was a major producer and user of Aroclors. A method combining frequency distributions of wind directions for six sectors and typical air concentrations for each sector has been used to calculate net advective loss of PCBs from the U.K. atmosphere (37). These calculations suggest that as a result of dilution with air masses from the Arctic and Atlantic regions, the U.K. exports approximately 30-40% ‘dirtier’ air to its neighboring regions than it imports. If these calculations are correct, then this accounts for a loss of approximately 280 kg yr-1 of PCB-52 and 80 kg yr-1 for PCB-153. If these figures, along with emission estimates, are included in the mass balance calculations then there is a deficit, i.e., output via reaction, deposition, and advection exceeds input via emission and volatilization by approximately 220 kg for PCB-52. For PCB-153, there is a surplus, i.e., input exceeds output by 125 kg. Soil Processes. The losses from soil via microbially mediated degradation, organic matter occlusion, and burial are difficult to quantify. For example, there are very few measurements of soil biodegradation half-life in the literature, and those there are do not adequately differentiate biodegradation from biodegradation and volatilization. However, our model suggests that a short soil degradation half-life would quickly reduce soil concentrations and consequently reduce the amount available for volatilization. A half-life of 15-20 yr is required to sustain a significant input to the atmosphere. Estimated Errors Associated with Mass Balance Calculations. It is important with mass balance calculations of this type that, where possible, range estimates for each process should made. As a result, each of the mass balance components has been examined and a range of values assigned. The results are contained in Table 3. It is clear that long-range transport and advective exchange are important processes for the U.K. The range of values assigned for atmospheric emission are relatively small; however, this greatly understates the uncertainty involved with their estimation. The overall mass balance calculations suggest VOL. 34, NO. 5, 2000 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

867

TABLE 3. Range Estimates for Mass Balance Calculationsa min (kg yr-1)

max (kg yr-1)

PCB-52 Input emission to atmosphere 210 volatilization from soil 60 advective inflow 390 total IN 660

240 190 430 860

PCB-52 Output atmospheric degradation 120 total deposition 150 advective outflow 720 total OUT 990

130 170 790 1090

PCB-153 Input emission to atmosphere 160 volatilization from soil 40 advective inflow 70 total IN 270

180 180 125 485

PCB-153 Output atmospheric degradation 4 total deposition 40 advective outflow 125 total OUT 169

8 80 225 313

a Balance for PCB-52: output exceeds input by 130-430 kg. Balance for PCB-153: input exceeds output by 45-320 kg.

that for PCB-52 calculated atmospheric loss processes exceed calculated sources by 130-430 kg. For PCB-153, the calculations suggest that estimated sources exceed loss processes by 45-320 kg. Given the uncertainties involved, this represents reasonable accounting, and as our understanding of individual processes improves so should the mass balance estimates.

Summary of Comments on Mass Balance The overall U.K. mass balance, including remaining emission and losses via advection, suggests that for PCB-52 losses exceed sources/inputs by ∼130-430 kg yr-1, whereas for PCB153 input exceeds loss by ∼45-320 kg yr-1. These differences are considerably greater than the observed overall decreases of 2.7 and 0.4 kg yr-1 for PCB-52 and PCB-153, respectively. This is probably a reflection of the uncertainty associated with calculations of this type with the greatest uncertainties being the estimation of remaining sources and the net advective losses from the U.K. atmosphere. These processes are potentially very significant contributors to the U.K. mass balance and hence need further effort to accurately quantify them. Again, it is important to bear in mind that although the overall half-life of individual PCB congeners ranges between 2 and 12 yr, this represents a total loss of a few kilograms from the overall U.K. air burden; hence, these losses are far less than the uncertainty involved in the mass balance calculations. It is stressed that it is difficult to accurately quantify each of the fate processes described in this paper to allow comparison as a result of the highly variable nature of the environment they are describing. However, calculations of this type can reveal which processes may exhibit the greatest influence in a contaminants’ overall fate. As a result of the calculations presented in this paper, we suggest that deposition of PCBs from the U.K. atmosphere is currently approximately balanced by volatilization from soil. The importance of degradative processes such as atmospheric OH- radical reaction is still unclear. Remaining primary emissions (or those resulting from recycling) appear to be still supporting the U.K. atmospheric burden, but in the long term, with the continual reduction of these sources, it will be fate processes within soil that will most likely control the long-term fate of PCBs, their supply to the atmosphere, and “drive” their global cycling. 868

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 34, NO. 5, 2000

Acknowledgments We are grateful to the U.K. Department of the Environment, Transport and the Regions for funding our research on atmospheric measurements and environmental fate modeling of persistent organic contaminants.

Literature Cited (1) Jones, K. C.; de Voogt, P. Environ. Pollut. 1999, 100, 209-221. (2) Rapaport, R. A.; Eisenreich, S. J. Environ. Sci. Technol. 1988, 22, 931-941. (3) Lead, W. A.; Steinnes, E.; Bacon, J. R.; Jones, K. C. Sci. Total Environ. 1997, 193, 229-236. (4) Harner, T.; Mackay, D.; Jones, K. C. Environ. Sci. Technol. 1995, 29, 1200-1209. (5) Wania, F.; Mackay, D. Sci. Total Environ. 1995, 161, 211-232. (6) Coleman, P. J.; Lee, R. G. M.; Alcock, R. E.; Jones, K. C. Environ. Sci. Technol. 1997, 31, 2120-2124. (7) Department of the Environment, Transport and the Regions. National Air Quality Information Archive 1996/7; National Environmental Technology Centre: 1997. (8) Ockenden, W. A.; Prest, H. F.; Thomas, G. O.; Sweetman, A. J.; Jones, K. C. Environ. Sci. Technol. 1998, 32, 1538-1543. (9) Thomas, G. O.; Sweetman, A. J.; Ockenden, W. A.; Mackay, D.; Jones, K. C. Environ. Sci. Technol. 1998, 32, 936-942. (10) Lee, R. G. M.; Jones, K. C. Environ. Sci. Technol. 1999, 33 (5), 705-712. (11) Panshin, S. Y.; Hites, R. A. Environ. Sci. Technol. 1994, 28, 20082013. (12) Hillery, B. R.; Basu, I.; Sweet, C. W.; Hites, R. A. Environ. Sci. Technol. 1994, 31, 1811-1816. (13) Halsall, C. J.; Gevao, B.; Howsam, M.; Lee, R. G. M.; Ockenden, W. A.; Jones. K. C. Atmos. Environ. 1999, 33, 541-552. (14) Honrath, R. E.; Sweet, C. I.; Plouff, C. J. Environ. Sci. Technol. 1997, 31, 842-852. (15) Baker, J. E.; Eisenreich, S. J. Environ. Sci. Technol. 1990, 24, 342-352. (16) Panshin, S. Y.; Hites, R. A. Environ. Sci. Technol. 1994, 28, 20012007. (17) Jeremiason, J. D.; Hornbuckle, K. C.; Eisenreich, S. Environ. Sci. Technol. 1994, 28, 903-914. (18) Pearson, R. F.; Hornbuckle, K. C.; Eisenreich, S. J.; Swackhamer, D. L. Environ. Sci. Technol. 1996, 30, 1429-1436. (19) Hesselberg, R. J.; Hickey, J. P.; Nortrup, D. A.; Willford, W. A. J. Great Lakes Res. 1990, 16, 21-129. (20) Borgmann, U.; Whittle, D. M. J. Great Lakes Res. 1991, 17, 368381. (21) Moilanen, R.; Pyysalo, H.; Wickstrom, K.; Linko, R. Bull. Environ. Contam. Toxicol. 1982, 29, 334-340. (22) Bignert, A.; Olsson, M.; Persson, W.; Jensen, S.; Zakrisson, S.; Litze´n, K.; Eriksson, Ha¨ggberg, L.; Alsberg, T. Environ. Pollut. 1998, 99, 177-198. (23) Simcik, M.; Basu, I.; Sweet, C.; Hites, R. A. Environ. Sci. Technol. 1999, 33, 1991-1995. (24) Mackay, D. Multimedia Environmental Models; Lewis Publishers: Chelsea, MI, 1991. (25) Finizio, A.; Mackay, D., Bidleman, T.; Harner, T. Atmos. Environ. 1997, 31, 2289-2296. (26) Mackay, D.; Shui, W.-Y.; Ma, K.-C. Illustrated handbook of physical-chemical properties and environmental fate for organic chemicals. Volume 1, Monoaromatic hydrocarbons, chlorobenzenes and PCBs; Lewis Publishers: Chelsea, MI, 1992. (27) Countryside Information System. Produced for the U.K. Department of the Environment, Transport and the regions by the Institute of Terrestrial Ecology, 1994. (28) Swackhamer, D.; McVeety, B.; Hites, R. Environ. Sci. Technol. 1988, 22 (6), 664-672. (29) Schro¨der, J.; Welsch-Pausch, K.; McLachlan, M. Atmos. Environ. 1997, 31, 2983-2989. (30) Anderson, P. N.; Hites, R. A. Environ. Sci. Technol. 1996, 30, 1756-1763. (31) Halsall, C. J.; Barrie, L. A.; Fellin, P.; Muir, D. C. G.; Billeck, B. N.; Lockhart, L.; Rovinsky, F. A.; Kononov, E. A.; Pastuhov, B. Environ. Sci. Technol. 1997, 31, 3593-3599.

(32) Jury, W.; Spencer, W.; Farmer, W. J. Environ. Qual. 1983, 12, 558-564. (33) Jury, W.; Russo, D.; Streile, G.; El Abd, H. Water Resour. Res. 1990, 26, 13-20. (34) Cousins, I. T.; Jones, K. C. Environ. Pollut. 1998, 102, 105-118. (35) Dyke, P. H.; Stratford, J. Organohalogen Compds. 1998, 36, 365368. (36) Harrad, S.; Sewart, A. P.; Alcock, R. E.; Boumphrey, R.; Burnett, V.; Duarte-Davidson, R.; Halsall, C.; Sanders, G.; Waterhouse, K.; Wild, S. R.; Jones, K. C. Environ. Pollut. 1994, 85, 131-146.

(37) Sweetman, A. J.; Jones, K. C. Modelling historical emissions and environmental fate of PCBs in the U.K. In Persistent Bioaccumulative Toxic Chemicals: Fate and Exposure; American Chemical Society: Washington, DC, in press.

Received for review June 2, 1999. Revised manuscript received October 26, 1999. Accepted November 30, 1999. ES9906296

VOL. 34, NO. 5, 2000 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

869