Environ. Sci. Technol. 2007, 41, 4529-4535
A Global Mass Balance Analysis of the Source of Perfluorocarboxylic Acids in the Arctic Ocean FRANK WANIA* Department of Physical and Environmental Sciences, University of Toronto Scarborough, 1265 Military Trail, Toronto, Ontario, Canada M1C 1A4
Whereas the pervasive and abundant presence of perfluorinated carboxylic acids (PFCAs) in the Arctic marine food chain is clearly established, their origin and transport pathway into the Arctic Ocean are not. Either the atmospheric oxidation of volatile precursor compounds, such as the fluorotelomer alcohols (FTOHs), or the longrange oceanic transport of directly emitted PFCAs is seen as contributing the bulk of the PFCA input to the Arctic. Here simulations with the zonally averaged global fate and transport model Globo-POP, in combination with historical emission estimates for FTOHs and perfluorooctanoic acid (PFOA), are used to evaluate the relative efficiency and importance of the two transport pathways. Estimates of the emission-independent Arctic Contamination Potential reveal that the oceanic transport of directly emitted PFCAs is more than 10-fold more efficient than the atmospheric degradation of FTOHs in delivering PFCAs to the Arctic, mostly because of the low yield of the reaction. The cumulative historic emissions of FTOHs are lower than those estimated for PFOA alone by a factor of 2-3, further limiting the contribution that precursor oxidation makes to the total PFCAs load in the Arctic Ocean. Accordingly, when fed only with FTOH emissions, the model predicts FTOH air concentrations in agreement with the reported measurements, but yields Arctic seawater concentrations for the PFOA that are 2 orders of magnitude too low. Whereas ocean transport is thus very likely the dominant pathway of PFOA into the Arctic Ocean, the major transport route of longer chain PFCAs depends on the size of their direct emissions relative to those of 10:2 FTOH. The predicted time course of Arctic seawater concentrations is very similar for directly emitted and atmospherically generated PFCAs, implying that neither past doubling times of PFCA concentrations in Arctic marine mammals nor any future time trends are likely to resolve the question of the dominant source of PFCAs.
PFCAs in Arctic seawater are currently unresolved and two alternative hypotheses have been proposed. The first explains the presence of PFCAs in the Arctic as the result of the global atmospheric dispersion of volatile precursor compounds that are oxidized in the atmosphere to form PFCAs (8). High water solubility of PFCAs leads to rapid wet deposition. Several studies have investigated and confirmed various aspects of this proposed mechanism. Field measurements have provided evidence of the widespread presence of the precursor compounds, namely the fluorotelomer alcohols (FTOHs), in the global atmosphere (9-13). Laboratory studies have established both the kinetics (14) and the mechanism (8) of the atmospheric degradation process that converts FTOHs into PFCAs. Measurements in Arctic snow confirmed that PFCAs are atmospherically deposited throughout the Arctic (15), and calculation with a spatially resolved atmospheric transport and chemistry model reproduced the global scale dispersal of FTOH in the atmosphere and predicted the global perfluorooctanoic acid (PFOA) yield from 8:2-FTOH based on an explicit parameterization of the atmospheric oxidation mechanism (16). The alternative hypothesis purports that the oceanic transport of PFOA directly emitted into the environment as a result of numerous industrial processes can account for the presence of PFOA in the Arctic Ocean and thus in Arctic marine organisms (17, 18). A simple zonally averaged global transport model (18) fed with global scale historical emission estimates (17) predicted global seawater concentrations and doubling times of PFOA in the Arctic which agree with observations (19, 20). Although proponents of either hypothesis do not question the occurrence of both transport pathways, their relative importance is contested. Based on predictions of very slow response in Arctic water concentrations to emission reductions, Armitage et al. (18) concluded that “measurements of PFOA time trends in Arctic biota over the next few decades will help to determine the relative contribution of oceanic and atmospheric transport mechanisms to the presence of PFOA and other PFCAs in the Arctic”. Recently observed decreases in the concentrations of PFOS and PFCAs in Arctic seal are interpreted as confirmation of the importance of an atmospheric pathway, because the response to emission reductions of atmospherically transported compounds is expected to be immediate (6). What has been missing so far is a global modeling study that considers and contrasts both these transport hypotheses with the objective to evaluate their relative efficiency and importance. It also needs to be confirmed that reductions of the emissions of volatile precursor compounds would indeed result in fast declines in Arctic seawater concentrations. Building upon the work by Armitage et al. (18) and employing the same global transport model, this is what this study set out to do.
Methods Introduction Perfluorocarboxylic acids (PFCAs) with longer perfluorochains bioaccumulate in the Arctic marine food chain (1, 2) and accordingly are present in marine predators across the Arctic (1-6). Some PFCAs have been shown to cause developmental delays and cancer in lab animals (7), highlighting the importance of resolving how PFCAs reach the Arctic Ocean. However, the origin and transport pathway of * Phone: +416-287-7225; e-mail:
[email protected]. 10.1021/es070124c CCC: $37.00 Published on Web 05/23/2007
2007 American Chemical Society
Two types of simulations with a zonally averaged global transport model were performed. First, using generic, hypothetical emission estimates, the Arctic Contamination Potential of individual FTOHs and PFCAs was estimated to directly compare the efficiency of various transport pathways into the Arctic. Second, historical emission estimates and future scenarios were employed to reproduce and predict the magnitude, time course, and pathways of PFOA and atmospherically generated PFCAs into the Arctic over the past and future five decades. The Globo-POP Model is a dynamic, compartmental, fugacity-based mass balance model of persistent organic VOL. 41, NO. 13, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4529
chemical fate in the global environment (21). Each hemisphere is represented by five latitudinal bands, consisting of four atmospheric layers, two types of soil, fresh water, and fresh water sediment. The surface ocean is represented by a single, unstratified layer of 200 m depth. Chemical is emitted, degraded, transferred between the compartments of one climate zone, and undergoes meridional transport between climate zones in both atmosphere and oceans. Globo-POP has been used to describe the historical global distribution of R-hexachlorocyclohexane (22) and polychlorinated biphenyls (23), and is the basis of efforts to quantify an organic chemical’s potential to reach and accumulate in the Arctic (24, 25). Recently, the model was used to highlight the potential importance of ocean transport in the global distribution of PFOA (18). Defining the Arctic Contamination Potential of a Degradation Product. It is obvious that absolute amounts and temporal trends of the emissions for PFCAs and precursors are pivotal in any effort to assess the relative importance of the two transport pathways to the Arctic Ocean. It is, however, also clear that this emission information will inevitably be among the most uncertain and contested elements of such an assessment. It is thus enlightening to first assess the relative efficiencies of the two transport mechanisms independent of quantitative emission information. The assessment of long-range transport potential as an extensive chemical characteristic, i.e., one that is independent of the emitted amount, has been the subject of considerable research recently (26), not the least because confirmation of the potential to reach remote regions is one of the criteria for classifying a substance as a Persistent Organic Pollutant under the Stockholm Convention. Many of the indicators of long-range transport that have been proposed and used are unsuitable for the comparative evaluation performed here, because they are restricted to quantifying dispersal in the atmosphere while considering neither the oceanic transport pathway nor the need for atmospheric deposition at the remote location (26). The Arctic Contamination Potential, ACP, as calculated by the Globo-POP model, on the other hand, is ideally suited for the task, because it explicitly strives to assess an organic chemical’s potential to reach high latitudes by atmospheric and oceanic pathways and accumulate in the surface media (soil, water, sediment) of the Arctic (24, 25). Previous ACP calculations were applied to parent compounds and did not consider the fate of degradation products. To apply it to atmospherically generated PFCAs the ACP concept first has to be extended to degradation products. The definition of the absolute Arctic Contamination Potential, eACP, relates the amount of chemical in Arctic surface media MArctic Surface to the cumulative globally emitted amount EGlobal (25).
eACP ) MArctic Surface/EGlobal‚100%
(1)
To compare the eACP of different compounds generic emission scenarios (continuous emissions to either air, fresh water, or soil, zonally distributed according to human population) and fixed simulation lengths (1 or 10 years) have been employed (24, 25). Equation 1 can be used to calculate the eACP of a parent compound. If one then performs additional global fate calculations for daughters being produced as a result of the degradation of the parent, the amount of the degradation products that ends up in the Arctic surface media can be calculated to obtain the eACP of the degradation products:
eACP(daughter) ) MArctic Surface(daughter)/EGlobal(parent)‚100% (2) 4530
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 13, 2007
Because the daughter is not emitted directly, the denominator is still the cumulative emission amount of the parent. It is possible to add the eACP of parent and daughters to calculate a joint eACP:
Joint eACP ) (MArctic Surface(parent) + Σi MArctic Surface(daughter i))/EGlobal(parent) (3) To ensure unit consistency, the amounts in the numerator and denominator of eqs 2 and 3 need to be in units of mol rather than mass units. The Arctic Contamination Potential of Directly Emitted and Atmospherically Generated PFCAs. Throughout the manuscript the acronyms dePFCA and agPFCA are used to distinguish between PFCAs directly emitted into the environment and PFCAs atmospherically generated by oxidation of FTOHs, respectively. Equation 1 was used to calculate the eACP of PFCAs and FTOHs of different chain length directly emitted for 10 years to the environment (eACP10(dePFCA) and eACP10(X:2 FTOH), X ) 6, 8, 10). In the case of dePFCAs, different modes of emission were considered, including exclusive emissions to air, fresh water, soil, and ocean water, as well as a more realistic, mixed mode of emission (31% air, 14% soil, 55% fresh water) (18). The FTOHs were assumed to be exclusively emitted to air. The daily atmospheric degradation rate of an FTOH in mol/h from the eACP calculation was saved to file and read as the emission input file for another fate calculation of a PFCA generated by atmospheric transformation of the FTOH. The degradation rates were either used as issimplying an unrealistic yield of 100%sor reduced using a more realistic yield of the FTOHPFCA reaction between 3 and 10%. The latter varies zonally because of higher abundance in midlatitudes of photooxidants that compete with the PFCA-forming reaction of the FTOHs. The yields, listed in Table S1 in the Supporting Information, were estimated based on the results of atmospheric fate calculations for 8:2-FTOH (Figure 5 in 16). These calculations relied on erroneously low KAW values (27), but it is unclear how this affects the yields. The formation of PFCAs from FTOHs in media other than the atmosphere was ignored. Even though there is evidence of such degradation occurring (28), the information on kinetics and yield is currently insufficient to consider this process in the model. Also, because of the very high volatility of the FTOHs, the atmospheric degradation process is believed to be quantitatively more important than FTOH-PFCA conversions in other media. The results of this second fate calculation were used in eqs 2 and 3 to calculate the eACP of agPFCAs (eACP10(agPFCA)) and the joint eACP of a FTOH and the PFCAs formed by its atmospheric oxidation (e.g., eACP10(8:2-FTOH & agPFOA & agPFNA). Additional calculations were performed to probe the sensitivity of the eACP results to variations in chemical properties. Chemical Input Parameters. The model requires quantitative information on a chemical’s partitioning and degradation properties. Although there has been much discussion about the correct partitioning properties of the FTOHs, the work by Goss et al. (27, 29) has resolved much of these issues. A detailed description of the selection of partitioning coefficients for the model calculations is given in the Supporting Information. A temperature-independent degradation rate constant of the reaction of FTOHs of different chain length with OH radicals of 1.07‚10-12 cm3‚molecules-1‚s-1 (14) was used. FTOHs were assumed to be persistent in all surface media. The partitioning properties of the PFCAs are considerably less well established, and various property estimation techniques not only give widely variable values (Table 6 in 29) but also generally apply to the protonated form of the acids and are thus of little environmental relevance. When
FIGURE 1. Absolute Arctic Contamination Potential after 10 years of continuous emission eACP10 for perfluorinated carboxylic acids of variable KOC directly emitted to air (31%), water (55%), and soil (14%), for fluorotelomer alcohols of variable chain length emitted to air, and for perfluorinated carboxylic acids generated atmospherically from the degradation of FTOHs.
selecting the properties of the PFCAs, we therefore adopted the approach and values given for anionic PFOA in (18). In particular, the lowest log KAW value that can be used in the Globo-POP model (-6.5) is used for all PFCAs, because beyond a certain lower threshold, the actual value of the KAW is of no importance. In fact, lower KAW values would result in non-sensical scavenging efficiency (30). The log KOW value was varied between 1 and 3 to cover the range of KD and KOC values that have been reported for the anionic forms of the PFCAs. The PFCAs were assumed to be perfectly persistent in air and surface media. Emission Information. Global historical emissions of FTOHs up to 2005 were compiled based on information provided in refs 17 and 31. It was assumed that production started in the mid-1970s at levels of 500 metric tons/annum (t/a), stabilized at 1500 t/a throughout the 1980s and early 1990s, to quickly increase to 11 000 t/a during the decade 1995 to 2005. It was assumed that throughout this time period 2% of the production was emitted to the atmosphere of the boreal (5%), temperate (90%), and subtropical (5%) zones of the Northern hemisphere (18). This yields FTOH emission rates that increase from 30 t/a in the early 1990s to more than 200 t/a in 2005, and a cumulative release of approximately 1500 metric tons of FTOHs for the time period 1974 to 2005 (Figure S1 in the Supporting Information). These rates are at the lower end of the range (100-1000 t/a) estimated in ref 14 to reconcile measured atmospheric concentrations and degradation half-life. Even though both production volumes and percent release are clearly uncertain, no maximum or minimum emission scenarios were developed. The linear nature of the model allows scaling the result directly to the assumed emissions, i.e., if, e.g., the percent release is 3% instead of 2%, the model results simply need to be multiplied by 1.5. Minimum and maximum global historical emission estimates for directly emitted PFOA, taken from ref 18, lead to a cumulative emission between 2700 and 5000 metric tons over the time period 1950 to 2005. This implies that even though the global emission rates of PFOA and FTOHs are generally of the same order of magnitude, PFOA has been emitted for a longer period of time and at rates approximately 3 times higher. Only in the past few years, as a result of reductions in PFOA emissions and an increase
in the production of FTOHs, do the estimated emissions for the FTOHs exceed those for PFOA.
Results and Discussion Model Results Using Hypothetical Emission Scenarios. The absolute Arctic Contamination Potentials after 10 years of continuous emission eACP10 for dePFCAs (blue bars) and FTOHs (red bars), and for agPFCAs generated atmospherically from the degradation of FTOHs (green bars) are compared in Figure 1. Tables S2-S4 detail the sensitivity of these results to chemical input parameters. Arctic Contamination Potential of FTOHs Emitted to the Atmosphere. Despite their relatively high persistence in air (estimated atmospheric half-life of 20 days, 14) the 8:2 and 10:2-FTOH have a very low eACP10air (0.0038% and 0.0018%, respectively, Figure 2, red bars, the superscript denotes the mode of emission). They are “flyers”: too volatile to deposit appreciably to the Earth’s surface, even at the low temperatures prevalent in the Arctic (25). The eACP10air of 6:2-FTOH is 1 order of magnitude higher (0.039%), because its lower KAW allows for some partitioning to the oceans. In fact, 6:2FTOH, and even more so 4:2-FTOH, classify as “multihoppers” that can reach the Arctic by repeated cycles of deposition and evaporation between atmosphere and oceans (25). The eACP10air(FTOH) is sensitive to the value and the temperature dependence of KAW and the degradation rate in water, but insensitive to KOA, KOC, and the degradation rate in other environmental compartments (Table S2). A higher atmospheric degradation rate reduces the eACP10air(FTOH), because it limits atmospheric transport to higher latitudes (Table S2). Arctic Contamination Potential of PFCAs Directly Emitted into the Environment. We estimate identical eACP10 values for all PFCAs of intermediate chain length (nC ) 6-11) that are emitted into the environment in the same way, because they behave as conservative tracers in water (Figure 2, blue bars). The eACP10a/s/w(dePFCA) values of 1.08% are quite high, reflecting their efficient transport with ocean currents (18). PFCAs are assumed to be perfectly persistent in water, and low KAW and KOC ensure that they can neither volatilize nor settle to the deeper ocean with solid particles. The eACPs of VOL. 41, NO. 13, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4531
FIGURE 2. Estimated historical and predicted future fluxes of PFCAs into (top of the red areas) and out of (top of the blue areas) the Arctic Ocean derived with Globo-POP. Panels A and B show the oceanic flow of PFOA directly emitted into the global environment. Panels C and D display the oceanic flow of PFCAs atmospherically generated by oxidation of FTOHs, and panels E and F show the deposition of atmospherically generated PFCAs. The historical net flux of directly emitted PFOA (red area in A) is much higher than the net flux of atmospherically generated PFCAs (red areas in C and E). If emissions of FTOHs and PFOA hypothetically ceased completely in 2006, atmospheric deposition of PFCAs to the Arctic Ocean would stop immediately (F), whereas the ocean would be a net source of both directly emitted (red area in B) and atmospherically generated PFCAs (red area in D) to the Arctic Oceans for decades to come. dePFCAs are not very sensitive to the mode of emission (Figure S2, Table S3). Emitting a PFCA exclusively to air or ocean water leads to higher values (eACP10air(dePFCA) ) 1.33%, eACP10οcean(dePFCA) ) 1.82%), whereas exclusive emission to land and fresh water reduces them (eACP10s°il(dePFCA) ) 0.91%, eACP10water(dePFCA) ) 0.99%). These eACP values are of the same order of magnitude, i.e., no matter what the mode of emission, dePFCAs will eventually find their way to the sea and therefore to the Arctic Ocean. Emission to media other than the ocean retards that transport, but does not prevent it. The eACP(dePFCA) is neither very sensitive to the partitioning properties (Figure S3, Table S3), which is the reason for the limited influence of chain length. Only a 10fold increase in the default KOC value (to logKOC ) 2.54) results in a small decline in the calculated eACP10a/w/s(dePFCA) from 1.08 to 0.92%, suggesting that with increasing chain length dePFCAs may have a slightly lower potential to reach the Arctic. Arctic Contamination Potential of PFCAs Atmospherically Generated from FTOHs. The eACP of agPFCAs is extremely 4532
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 13, 2007
dependent on the assumed yield of the degradation reaction. If all of the 8:2-FTOH were quantitatively converted into PFOA and PFNA, the eACP10air(agPFOA&agPFNA) is 1.66% and thus of the same order of magnitude as the eACP10air(dePFOA) of 1.33%. However, using the more realistic yields of the transformation (Table S1, 16), the eACP10air(agPFOA or agPFNA) is reduced to 0.062%. Even though the ACP of the parent-FTOH was very dependent on chain length (Figure 1, red bars), the eACP of the degradation products is much less so. The PFCAs generated from various FTOHs have very similar eACP10air values ranging from 0.05 to 0.08% (Figure 1, green bars). This is because the amount of the FTOHs in air is much larger than the amount in the surface media and therefore less sensitive to the value of the FTOH’s KAW and other chemical properties (Table S4). A shift in the distribution from air to water with decreasing KAW, i.e. decreasing chain length of the FTOH, reduces the amount in air only slightly, but increases the amount in water substantially. The amount of FTOH in air determines the rate of PFCA formation, whereas the amount in surface media is dominating the
eACP10(FTOH). In fact, eACP10air(agPFCA) depends strongly neither on the properties of the parent FTOH nor on those of the PFCAs themselves (Table S4). Comparison of the Efficiency of Different Pathways of PFCAs into the Arctic. When contrasting the results of the eACP calculations for the FTOHs, dePFCAs, and agPFCAs (Figure 1), it transpires that after 10 years of continuous release, the eACP of directly emitted PFCAs is about 16 times higher than that of PFCAs generated by atmospheric degradation of FTOHs, mostly because of the low yield of the reaction. This suggests that for the atmospheric pathway to be of comparable importance for explaining the presence of PFCAs in the Arctic Ocean as the oceanic pathway, the emission rate of FTOHs and other PFCA precursor compounds would need to be at least an order of magnitude higher than the direct emission rate of PFCAs. One of the major simplifications of the above discussion is the focus on the eACP after 10 years of emissions, which ignores the temporal trend in the calculated eACP. The eACP of dePFCAs, in particular, continues to increase after 10 years (Figure S4), reflecting the very slow process of meridional ocean mixing which occurs on the scale of decades (18). An interesting outcome of the eACP calculations is that also the eACP of agPFCAs continues to increase after 10 years, although the eACP of their precursors does not (Figure S4). The explanation is that a substantial fraction of the agPFCAs also reaches the Arctic by oceanic transport. The calculations by Wallington et al. (Figure 3 in 16) have revealed that the highest concentrations of PFOA are predicted to occur “over the Atlantic and Pacific Oceans, North Africa, and the Arctic during summer” and it is safe to assume that much of the agPFOA will be deposited to the Northern Atlantic and Pacific Ocean, from where it will undergo slow but steady transport to higher latitudes. When repeating the eACP calculation in a global model environment without ocean mixing (setting the meridional eddy macrodiffusivity in the oceans to zero), the eACP of 8:2-FTOH remains unchanged, whereas the eACP of agPFOA is indeed significantly reduced (eACP10air(agPFOA) from 0.067% to 0.037%, Figure S5). Model Results Using Historical Emission Scenarios. Evaluating the Simulated FTOH Air Concentrations. In the context of this study, the most important result of the GloboPOP simulation of the global fate of the FTOHs with historical emissions is the atmospheric degradation rate, which in turn is controlled by the calculated atmospheric concentrations. These had lower values in summer than in winter and increased in accordance with the emissions (Figure S6). Air concentrations of the FTOHs measured in the Northern Hemisphere (9-13) can serve in the evaluation of these results. The calculated air concentrations were not dependent on whether the properties of 6:2, 8:2, or 10:2-FTOH were used. Because the emission estimates are for the sum of the FTOHs, the calculated air concentrations can therefore be compared with the sum of measured FTOH concentrations, which generally includes 6:2, 8:2, and 10:2-FTOH. The model calculates zonally averaged air concentrations, whereas the reported measurements, with some exceptions (11, 13), tend to be relatively close to urban areas, which are expected to have concentrations higher than the zonal average. This is evident in spatially resolved atmospheric model simulations, which shows much higher 8:2-FTOH concentrations in Europe and Eastern North America (Figure 2 in 16). Zonally averaging is much less of a concern when comparing polar air concentrations. For the years 2001-2005 the model predicts zonally averaged air concentrations in the lowest air layer in the range of 20-90, 6-50, and 2-30 pg/m3 for the Northern temperate, boreal, and polar zones, respectively. Arctic air concentrations (11) for July 2005 average 25 pg/m3 (range 7-55 pg/m3), which is in good agreement with calculated
concentrations for the N-Polar zone of 15 pg/m3 (range 5-32 pg/m3) for 2005. Concentrations in more remote temperate locations tend to be between 10 and 100 pg/m3 (9, 10), whereas cities in North America and Europe tend to have levels above 100 pg/m3 (10, 12). This again compares favorably with the predictions for the Northern temperate zone, which are in the same range as those reported for Long Point and Winnipeg. This suggests that annual FTOH emission rates on the order of 100-200 t for the time period 2000 to 2005 are reasonable and in the correct order of magnitude. Wallington et al. (16) assumed a global emission rate of 1000 t/a for 8:2-FTOH alone, which is more than a factor of 5 higher than the rates used in this study for the sum of the FTOHs. They report a calculated 8:2 FTOH air concentration value of “(0.5-5) × 105 molecules/cm3 in remote ocean and Arctic locations in the Northern Hemisphere”, which converts to 38-385 pg/m3. This is clearly much higher than the measured air concentrations reported for this area (11) and is further indication that an emission rate of 1000 t/a is too high. Magnitude of Atmospheric and Oceanic Transport Fluxes of PFCAs into the Arctic Ocean. Globo-POP was used to estimate the flux of agPFCAs into the Arctic Ocean, both by deposition from the atmosphere and by inflowing ocean water. Because emission estimates are for the sum of FTOHs, the model results are for the sum of the PFCAs atmospherically generated from those FTOHs, without regard to the chain length. These fluxes were further compared with the oceanic inflow of dePFOA (16), because only emission estimates for this particular PFCA are available. Results are displayed in Figure 2. All three types of fluxes increase in the latter half of the 20th century in response to increasing emissions of both FTOHs and PFOA. Whereas the ocean fluxes are continuous, the atmospheric deposition flux is strongly seasonally variable with maxima in summer, when Arctic OH radical concentrations are high and FTOH oxidation is occurring. The predicted seasonality in deposition flux agrees with higher concentrations in Arctic snow deposited during spring and summer (15). The atmospheric deposition flux and the net ocean flux of agPFCAs are of similar magnitude, increasing from about 20 to 30 kg/a in the 1980s and early 1990s to more than 100 kg/a in 2005. Wallington et al. (16) estimate a deposition flux of agPFOA of 400 kg/year, which is a factor of 5-10 higher than our estimate for all of the PFCAs atmospherically generated from FTOHs, consistent with their 8:2-FTOH emission rate (1000 t/a) that is also higher by approximately the same factor as our total FTOH emission during 20002005 (100-200 t/a, Figure S1). In other words, when normalized by the emission rate, the two model simulations give similar estimates of atmospheric deposition of agPFCAs to the Arctic. When compared with the estimated net influx of dePFOA, which during 2000-2005 was on the order of 9-20 t/asin agreement with the estimate of 8-23 t/a given in (18) using the same model and emission estimatesthe influx of agPFCA is negligible. For 2000 and 2005, we estimate a total input of agPFCAs of 110 and 260 kg/a, respectively, which is almost 2 orders of magnitude smaller than the oceanic influx of dePFOA alone. Figure S7 shows the time course of the calculated Arctic seawater concentrations for dePFOA and agPFCAs. The concentrations of dePFOA increase from 50 to 70 pg/L between 2000 and 2005, which is in excellent agreement with the measured concentrations in the Greenland Sea (19), as had been previously noted (18). The concentrations of agPFCAs in the Arctic Ocean are more than 2 orders of magnitude lower (0.25-0.4 pg/L between 2000 and 2005, Figure S7). This implies that the amount of PFCAs atmospherically generated from FTOHs is insufficient to explain measured seawater concentrations. VOL. 41, NO. 13, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4533
The Source of PFCAs in the Arctic Ocean and the Implications for Time Trends. The global mass balance analysis presented here thus confirms the finding by Armitage et al. (18) that long-range ocean transport of directly emitted PFOA is an important, and very likely the dominant, source of this substance to the Arctic Ocean. Two factors limit the relative importance of the atmospheric oxidation, transport, and deposition pathway. First, the low yield of the atmospheric oxidation of FTOHs is responsible for the low overall efficiency of agPFOA to reach and accumulate in the Arctic, as was quantified in the eACP calculations (Figure 1). Second, historically FTOHs emissions are not sufficiently high relative to the emission from direct sources of the manufacture and use of PFOA (Figure S1). Even though emission estimates are notoriously uncertain, it is hard to imagine that the emission estimates for FTOHs are so much too low and/or those for PFOA so much too high that atmospheric oxidation of FTOHs could possibly be the dominant source of PFOA to the Arctic Ocean. In particular, the agreement between model-predicted and measured PFOA ocean water concentrations (18) and the agreement between model-predicted and measured FTOH air concentrations noted here, essentially excludes the possibility that either emission estimate is erroneous by an order of magnitude. As it is mostly the longer PFCAs that accumulate in the Arctic food chain, it is important to establish whether the dominance of the oceanic pathway only applies to PFOA or also to PFCAs with longer perfluorochains. This is currently difficult to answer conclusively because no estimates of the amounts of longer chain PFCAs directly emitted to the environment exist. However, the above ACP calculations indicate that the emissions of FTOHs with 12 carbons and more would have to be more than 10-fold higher than the direct emissions of longer chain PFCAs to be the dominant source of those PFCAs in remote regions. It is unknown how much FTOHs with 12 and more carbons contribute to the total FTOH emissions. However, the measured FTOH composition in the atmosphere indicates that the relative contribution of 10:2 FTOH to the total emissions is likely on the order of 10-25% (11-13). This suggests that up to a quarter of the agPFCAs that was estimated to be deposited to the Arctic (i.e., 10-25% of 154 kg/a in 2006 or 15-40 kg/a) may be PFDA and PFUnA. These fluxes are consistent with, although at the lower end of, the range obtained by extrapolating measurements of PFDA and PFUnA concentrations in snow to the whole Arctic (15). Most of the PFCAs generated in the atmosphere by degradation of FTOHs are expected to be deposited to midlatitude Northern hemisphere oceans. Therefore, both dePFCAs and agPFCAs reach the Arctic Ocean via ocean transport. The response of Arctic Ocean water concentrations to reductions in the emissions of both PFCAs and volatile PFCA precursor compounds is thus not expected to be immediate, but considerably delayed (Figure S7). In fact, predictions using hypothetical emission scenarios with no or greatly reduced emissions after 2005 show continued oceanic influx of both dePFCAs and agPFCAs into the Arctic Ocean (right side of Figure 2). Consequently, the time trends of Arctic seawater concentrations of dePFOA and agPFCAs are quite similar and are predicted to peak around 2030 (Figure S7). This also means that measurements of time trends in Arctic biota over the next few decades may not allow us to distinguish the relative contribution of dePFCAs and agPFCAs to the Arctic. Nevertheless, time trends of PFCAs in Arctic marine mammals have been cited as providing evidence of the dominance of atmospheric oxidation and transport to the Arctic Ocean (6, 20). In particular, doubling times of PFCAs in polar bears sampled between 1972 and 2002 were judged too short to implicate transport via ocean currents (20). 4534
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 13, 2007
However, the model-predicted rates of Arctic Ocean water concentration increase during that time period are very similar for dePFOA (approximately 9-10 years) and agPFCAs (6-9 years) (Figure S5) and are in the same range as the measured doubling times (3.6-31.1 years, 20). Recent declines in the concentrations of PFOS in ringed seal from the Arctic were judged to be “most probably the result of declining concentrations in sea water” (6). If the concentrations of PFCAs in Arctic marine mammals appear to show a rapid response to reductions in the emissions of FTOHs, it may be because the time trends in marine mammals are not reflecting the time trends in the average Arctic Ocean water concentrations (6). Strong stratification may limit the exposure of the arctic marine food chain to seawater flowing into the Arctic and subsiding under a thin surface layer of fresher water, which is strongly influenced by atmospheric inputs to the Arctic Ocean and the Eurasian and North American watersheds of its tributaries (32). Also, factors other than seawater concentrations influence contaminant concentration time trends in biological samples (33) and in particular a rapidly changing Arctic climate may seriously confound such trends in seals and bears (34). Progress in the identification of the sources of PFCAs and other perfluorinated alkyl compounds in Arctic biota will depend on (1) compound-specific, spatially and temporally resolved estimates of the direct emissions of PFCAs of different chain length and of the emissions of FTOHs of different chain length and other volatile precursor compounds, (2) concentrations of PFCAs of different chain length in Arctic Ocean water as a function of depth and season, and (3) global transport models with descriptions of water transport and mixing in the Arctic Ocean that are more sophisticated than those in Globo-POP.
Acknowledgments This work was funded by E.I. Du Pont de Nemours & Co., Inc.
Supporting Information Available Additional figures and tables showing model results and sensitivity of model results to input parameter changes. This material is available free of charge via the Internet at http:// pubs.acs.org.
Literature Cited (1) Martin, J. W.; Smithwick, M. M.; Braune, B. M.; Hoekstra, P. F.; Muir, D. C. G.; Mabury, S. A. Identification of long-chain perfluorinated acids in biota from the Canadian Arctic. Environ. Sci. Technol. 2004, 38, 373-380. (2) Tomy, G. T.; Budakowski, W.; Halldorsen, T.; Helm, P. A.; Stern, G. A.; Friesen, K.; Pepper, K.; Tittlemier, S. A.; Fisk, A. T. Fluorinated organic compounds in an eastern Arctic marine food web. Environ. Sci. Technol. 2004, 38, 6475-6481. (3) Smithwick, M.; Mabury, S. A.; Solomon, K. R.; Sonne, C.; Martin, J. W.; Born, E. W.; Dietz, R.; Derocher, A. E. ; Letcher, R. J.; Evans, T. J.; Gabrielsen, G. W.; Nagy, J.; Stirling, I.; Taylor, M. K.; Muir, D. C. G. Circumpolar study of perfluoroalkyl contaminants in polar bears (Ursus maritimus). Environ. Sci. Technol. 2005, 39, 5517-5523. (4) Bossi, R.; Riget, F.; Dietz, R. Temporal and spatial trends of perfluorinated compounds in ringed seal (Phoca hispida) from Greenland. Environ. Sci. Technol. 2005, 39, 7416-7422. (5) Kannan, K.; Yun, S. H.; Evans, T. J. Chlorinated, brominated, and perfluorinated contaminants in livers of polar bears from Alaska. Environ. Sci. Technol. 2005, 39, 9057-9063. (6) Butt, C. M.; Muir, D. C. G.; Stirling, I.; Kwan, M.; Mabury, S. A. Rapid response of Arctic ringed seals to changes in perfluoroalkyl production. Environ. Sci. Technol. 2007, 41, 42-49. (7) Lau, C.; Butenhoff, J. L.; Rogers, J. M. The developmental toxicity of perfluoroalkyl acids and their derivatives. Toxicol. Appl. Pharmacol. 2004, 15, 231-241. (8) Ellis, D. A.; Martin, J. W.; De Silva, A. O.; Mabury, S. A.; Hurley, M. D.; Sulbaek, Andersen, M. P.; Wallington, T. J. Degradation
(9)
(10)
(11) (12)
(13)
(14)
(15)
(16)
(17) (18)
(19)
(20)
(21) (22)
of fluorotelomer alcohols: A likely atmospheric source of perfluorinated carboxylic acids. Environ. Sci. Technol. 2004, 38, 3316-3321. Martin, J. W.; Muir, D. C. G.; Moody, C. A.; Ellis, D. A.; Kwan, W. C.; Solomon, K. R.; Mabury, S. A. Collection of airborne fluorinated organics and analysis by gas chromatography/ chemical ionization nass spectrometry. Anal. Chem. 2002, 74, 584-590. Stock, N. L.; Lau, F. K.; Ellis, D. A.; Martin, J. W.; Muir, D. C. G.; Mabury, S. A. Polyfluorinated telomer alcohols and sulfonamides in the North American troposphere. Environ. Sci. Technol. 2004, 38, 991-996. Shoeib, M.; Harner, T.; Vlahos, P. Perflourinated chemicals in the Arctic atmosphere. Environ. Sci. Technol. 2006, 40, 75777583. Jahnke, A.; Ahrens, L.; Ebinghaus, R.; Temme, C. Urban versus remote air concentrations of fluorotelomer alcohols and other polyfluorinated alkyl substances in Germany. Environ. Sci. Technol. 2007, 41, 745-752. Jahnke, A.; Berger, U.; Ebinghaus, R.; Temme, C. Latitudinal gradient of airborne polyfluorinated alkyl substances in the marine atmosphere between Germany and South Africa (53 °N-33 °S). Environ. Sci. Technol. 2007, 41, 3055-3061. Ellis, D. A.; Martin, J. W.; Mabury, S. A.; Hurley, M. D.; Sulbaek, Andersen, M. P.; Wallington, T. J. Atmospheric lifetime of fluorotelomer alcohols. Environ. Sci. Technol. 2003, 37, 38163820. Young, C. J.; Furdui, V. I.; Franklin, J.; Koerner, R. M.; Muir, D. C. G.; Mabury, S. A. Perfluorinated acids in Arctic snow: New evidence for atmospheric formation. Environ. Sci. Technol. 2007, 41, DOI: 10.1021/es0626234. Wallington, T. J.; Hurley, M. D.; Xia, J.; Wuebbles, D. J.; Sillman, S.; Ito, A.; Penner, J. E.; Ellis, D. A.; Martin, J.; Mabury, S. A.; Nielsen, O. J.; Sulbaek Anderson, M. P. Formation of C7F15COOH (PFOA) and other perfluorocarboxylic acids during the atmospheric oxidation of 8:2 fluorotelomer alcohol. Environ. Sci. Technol. 2006, 40, 924-930. Prevedouros, K.; Cousins, I. T.; Buck, R. C.; Korzeniowski, S. H. Sources, fate, and transport of perfluorocarboxylates. Environ. Sci. Technol. 2006, 40, 32-44. Armitage, J.; Cousins, I. T.; Buck, R. C.; Prevedouros, K.; Russell, M. H.; MacLeod, M.; Korzeniowski, S. H. Modeling global-scale fate and transport of perfluorooctanoate emitted from direct sources. Environ. Sci. Technol. 2006, 40, 6969-6975. Caliebe, C.; Gerwinski, W., Theobald, N.; Hu ¨ hnerfuss, H. Occurence of perfluorinated organic acids in the water of the North Sea and Arctic North Atlantic. Poster presented at Fluoros Toronto, Canada, 2005. Smithwick, M.; Norstrom, R. J.; Mabury, S. A.; Solomon, K.; Evans, T. J.; Stirling, I.; Taylor, M. K.; Muir, D. C. G. Temporal trends of perfluoroalkyl contaminants in polar bears (Ursus maritimus) from two locations in the North American Arctic, 1972-2002. Environ. Sci. Technol. 2006, 40, 1139-1143. Wania, F.; Mackay, D. A global distribution model for persistent organic chemicals. Sci. Total Environ. 1995, 160/161, 211232. Wania, F.; Mackay, D.; Li, Y.-F.; Bidleman, T. F.; Strand, A. Global chemical fate of R-hexachlorocyclohexane. 1. Evaluation of a
(23)
(24)
(25)
(26)
(27)
(28)
(29)
(30) (31)
(32)
(33)
(34)
global distribution model. Environ. Toxicol. Chem. 1999, 18, 1390-1399. Wania, F.; Daly, G. L. Estimating the contribution of degradation in air and deposition to the deep sea to the global loss of PCBs. Atmos. Environ. 2002, 36, 5581-5593. Wania, F. Assessing the potential of persistent organic chemicals for long-range transport and accumulation in polar regions. Environ. Sci. Technol. 2003, 37, 1344-1351. Wania, F. Potential of degradable organic chemicals for absolute and relative enrichment in the Arctic. Environ. Sci. Technol. 2006, 40, 569-577. Fenner, K.; Scheringer, M.; MacLeod, M.; Matthies, M.; McKone, T. E.; Stroebe, M.; Beyer, A.; Bonnell, M.; Le Gall, A. C.; Klasmeier, J.; Mackay, D.; van de Meent, D.; Pennington, D.; Scharenberg, B.; Suzuki, N.; Wania, F. Comparing estimates of persistence and long-range transport potential among multimedia models. Environ. Sci. Technol. 2005, 39, 1932-1942. Goss, K.-U.; Bronner, G.; Harner, T.; Hertel, M.; Schmidt, T. C. The partition behavior of fluorotelomer alcohols and olefins. Environ. Sci. Technol. 2006, 40, 3572-3577. Dinglasan, M. A. J.; Ye, Y.; Edwards, E. A.; Mabury, S. A. Fluorotelomer alcohol biodegradation yields poly- and perfluorinated acids. Environ. Sci. Technol. 2004, 38, 28572864. Arp, H. P. H.; Niederer, C.; Goss, K.-U. Predicting the partitioning behavior of various highly fluorinated compounds. Environ. Sci. Technol. 2006, 40, 7298-7304. Lei, Y. D.; Wania, F. Is rain or snow a more efficient scavenger of organic chemicals? Atmos. Environ. 2004, 38, 3557-3571. DuPont Global PFOA Strategy - Comprehensive source reduction; Presented at the U.S. EPA OPPT, Jan. 31, 2005; U.S. EPA public docket AR226-1914; U.S. EPA: Washington, DC. MacDonald, R. W.; Barrie, L. A.; Bidleman, T. F.; Diamond, M. L.; Gregor, D. J.; Semkin, R. G.; Strachan, W. M. J.; Li, Y.-F.; Wania, F.; Alaee, M.; Alexeeva, L. B.; Backus, S. M.; Bailey, R.; Bewers, J. M.; Gobeil, C.; Halsall, C. J.; Harner, T.; Hoff, J. T.; Jantunen, L. M. M. Lockhart, W. L.; Mackay, D.; Muir, D. C. G.; Pudykiewicz, J.; Reimer, K. J.; Smith, J. N.; Stern, G. A.; Schroeder, W. H.; Wagemann, R.; Yunker, M. B. Contaminants in the Canadian Arctic: 5 years of progress in understanding sources, occurrence and pathways. Sci. Total Environ. 2000, 254, 93-234. Bignert, A.; Goethberg, A.; Jensen, S.; Litzen, K.; Odsjoe, T.; Olsson, M.; Reutergaardh, L. The need for adequate biological sampling in ecotoxicological investigations: A retrospective study of twenty years pollution monitoring. Sci. Total Environ. 1993, 128, 121-139. Macdonald, R. W.; Harner, T.; Fyfe, J. Recent climate change in the Arctic and its impact on contaminant pathways and interpretation of temporal trend data. Sci. Total Environ. 2005, 342, 5-86.
Received for review January 17, 2007. Revised manuscript received April 10, 2007. Accepted April 13, 2007. ES070124C
VOL. 41, NO. 13, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4535
