Environ. Sci. Technol. 2004, 38, 738-745
PBDEs in European Background Soils: Levels and Factors Controlling Their Distribution ASHRAF HASSANIN,† KNUT BREIVIK,‡ S A N D R A N . M E I J E R , †,§ E I L I V S T E I N N E S , | GARETH O. THOMAS,† AND K E V I N C . J O N E S * ,† Department of Environmental Science, Institute of Environmental and Natural Sciences, Lancaster University, Lancaster LA1 4YQ, U.K., NILU, Norwegian Institute for Air Research, P.O. Box 100, NO-2027 Kjeller, Norway, and Department of Chemistry, Norwegian University of Science and Technology, NO-7491, Norway
Surface soils (0-5 cm) from remote/rural woodland (coniferous and deciduous) and grassland locations on a latitudinal transect through the United Kingdom and Norway were analyzed for polybrominated diphenyl ethers (PBDEs). Concentrations ranged between 65 and 12 000 ∑ALLPBDE ng kg-1 dry weight. PBDE-47, -99, -100, -153, and -154sthe major constituents of the penta-BDE technical productsdominated the average congener pattern of the soils. Indeed, the average congener composition and distribution measured in these European background soils closely matched that reported in the technical pentaBDE product. This is interpreted as evidence that transfer of the congeners present in penta-BDE-treated products from source-air-soil occurs with broadly similar efficiency, perhaps because there has been little weathering/ degradation/alteration of the congener source pattern by processes operating during atmospheric transport or within the soil itself. BDE-183, a marker for the octa-BDE mix, was detected at concentrations ranging from DL median 52 14 29 76 100 67 29 19 14 81 95 90 67 10 86 90 29 86 14 19
28 17 11 14 61 12 15 11 32 20 280 36 15 47 72 22 25 26 24 28 440 610
min
GL UK pg g-1 OM (n ) 21) max
11 44 12 21 9 38 8 48 7 520 6 18 9 73 9 79 11 52 9 180 78 3200 8 470 6 100 26 68 19 600 8 240 9 70 10 900 10 26 7 59 15(6) 5000(9) 65(6) 6000(9)
median 100 55 41 58 260 42 61 37 100 65 990 120 61 180 250 100 110 85 94 86 1500 2000
min
max
28 180 38 66 24 230 35 130 63 2000 28 120 24 290 24 430 34 160 39 710 330 13000 32 1800 30 700 100 270 100 4200 33 940 47 150 60 6600 44 110 49 150 140(6) 20000(9) 610(6) 23000(9)
WL UK pg g-1 DW (n ) 21) %>DL median 76 76 43 48 100 90 10 52 95 95 95 95 90 67 95 100 24 100 0 29
27 21 11 33 490 44 49 23 13 51 900 110 48 35 210 100 60 70 34 1800 2500
min 9 8 8 11 50 9 15 8 13 7 190 11 8 13 38 14 33 10
WL UK pg g-1 OM (n ) 21) max
67 200 120 180 1400 250 82 86 13 1700 3200 360 180 270 1200 420 110 7000
9 110 75(7) 5600(11) 110(7) 12000(8)
DL, detection limit. (site no.) ) location where the highest or lowest sample in each category was measured.
median
min
max
66 49 34 86 890 98 58 60 34 87 1700 210 90 85 410 180 110 159
14 17 18 26 300 30 32 20 34 25 800 66 39 18 130 79 50 20
110 210 120 200 3100 330 85 230 34 240 4300 570 320 270 2600 880 130 15000
110 2900 4000
40 230 450(7) 9600(6) 510(7) 26000(8)
WL Norway pg g-1 DW (n ) 24) %>DL median 75 63 42 71 100 75 25 42 8 75 92 92 83 46 54 92 46 54 0 0
35 29 28 30 250 23 33 26 23 31 360 58 36 44 51 46 33 25 710 970
min 7 8 7 10 12 10 14 8 19 11 63 18 9 20 11 13 17 9
max 100 49 54 99 860 71 67 43 27 90 1400 230 150 140 270 310 110 130
90(23) 2600(36) 130(28) 3000(18)
WL Norway pg g-1 OM (n ) 24) median 46 41 36 50 420 41 40 42 33 54 570 120 58 66 87 70 66 51 1100 1500
min 27 20 21 26 65 20 21 28 28 28 120 38 27 28 28 14 29 20
max 140 190 70 120 1100 86 70 63 39 100 1500 240 170 140 290 340 120 140
300(38) 2600(36) 530(37) 3700(33)
FIGURE 2. Total PBDE concentration vs latitude: (A) expressed as pg/g dry weight; (B) pg/g organic matter. (GL, grassland; WL, woodland). the two countries. Air PBDE concentrations measured at Mace Head, a remote location on the west coast of Ireland, gave no indication of LRAT events bringing “pulses” of PBDEs over the Atlantic Ocean (22)sdespite the fact that North America dominates the current global usage of penta-BDE (3, 23). It is believed that the levels and distribution of PBDEs measured in these European soils is largely a response to European scale emissions, supplemented by advected air from the west and environmental processes. Jaward et al. (17) reported the results of a European passive air sampling campaign conducted at over 70 locations. UK urban centers had the highest values detected in Europe (consistent with the UK’s relatively high PBDE usage), while values in northern Norway were among the lowest. Urban: remote rural differences exceeded a factor of 700 across Europe in the passive sampling survey. On the basis of these urban/rural and geographical differences observed for PBDEs in European air and the properties/usage patterns of PBDEs, UK urban soil values are likely to be much higher than measured in these samples. Correlations were performed between the major PBDE congeners and the PCB homologues reported in these soils previously (15) (see Supporting Information Table 1). There was a very strong positive relationship between ∑ALLPBDEs and ∑PCBs (p < 0.001) (and most PBDE congeners and the different PCB homologues), supporting the view that the broad source areas (i.e., urban locations), atmospheric dispersal, deposition, and retention in soils of these two compound classes have much in common. The ∑ALLPBDE concentrations and their range are also very similar to that observed for PCBs in the same soils (180-22 000 ng of ∑PCB/ kg DW) (15). This is an interesting comparison. About 1.3 million t of PCBs was manufactured and used globally (24, 25) with the peak usage of ∼75 000 t occurring in 1970. It has been estimated that Europe accounts for ∼12% of the total global demand for PBDEs (2); the European Union risk assessment for penta-BDE estimated that ∼10 000 t was used
in 1989 (1). Therefore global penta-PBDE production may have reached a similar peak production/use value to that of PCBs, although cumulative global PBDE production has not yet been of the same order as for PCBs. Despite this, soil concentrations of the two compound classes measured here are very similar. The passive sampling survey mentioned earlier (17) showed that European PCB air concentrations (range of ∼201700 pg m-3) still exceed those of the PBDEs (0.5-250 pg m-3). Despite this, as just noted, PBDE and PCB soil concentrations are very similar. Soil concentrations are a complex function of emissions, advective transport, net deposition, and net degradation/burial processes. Cumulative European PCB emissions are probably greater than for PBDEs. However, PBDEs may show greater partitioning to the soil because they have higher octanol-air partition coefficients (KOA) values than the equivalent PCBs (26). For a given air concentration, the soil PBDE concentration would be expected to exceed that of the equivalent PCB congener, assuming that the air and the surface soil were approaching equilibrium. Very little information exists at present about PBDE behavior in soils. Estimates using the software EPIWIN indicated that the expected half-lives of nine selected PBDE congeners were all less than half a year (9). Litz (27) conducted studies where high amounts (∼50 mg kg-1) of PBDE were added to soils and reported aerobic and anaerobic degradation occurred over periods of weeks and inferred bound residue formation. Half-lives of PBDEs in soils may therefore be much shorter than for the PCBs (9). However, it is noteworthy that congeners of the technical penta-PBDE mixture have accumulated to similar levels as the PCBs in these surface soils and that the pattern of the penta-BDE mixture is preserved in the soils (see below). Further work is therefore required to clarify the balance between the rates of atmospheric emissions and deposition and soil degradation of PBDEs. VOL. 38, NO. 3, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
741
General Comments on the PBDE Congeners and Patterns in the Soils. The soil PBDE mixture was dominated by congeners 47, 99, 100, 153, and 154, those defined as the ∑5PBDEs. The ∑5PBDE typically constituted ca. 65% of the ∑ALLPBDE in the soils (see Table 1). A wide range of other congeners were also detected, making up the remaining ∼35% of the ∑ALLPBDE. Among these, congeners 66, 77, 119, 138, and 166 were generally the most important. BDE-183 will be considered separately below. However, several others (17, 28, 32, 35, 49, 71, 75, and 85) were detected too (Table 1). The mixture and relative abundance of PBDEs in the soils broadly matches that reported in the technical penta-PBDE product. Sjo¨din et al. (19) reported the following for Bromkal 70-5DE (% w/w): PBDE-47, 37%; -99, 35%; -100, 6.8%; -153, 3.9%; -154, 2.5%; -85, 1.6%; -138, 0.41%; -66, 0.22%; -28, 0.11%; -17, 0.022%. The soils data gave the following average composition: PBDE-47, 21%; -99, 40%; -100, 6%; -153, 8.7%; -154, 4.4%; -85, 2.4%; -138, 1.5%; -66, 3%; -28, 1.1%; -17, 1.4%. Assuming that Bromkal 70-5DE is representative of penta-BDE technical formulations (19, 28), two important observations are therefore put forward. First, the transfer processes from source-air-soil operate with broadly similar efficiencies across all the penta-PBDE congeners; indeed, Lee et al. (22) reported a similar congener distribution in rural European air samples, analyzed in our laboratory with the same methods as reported here. Second, overall there has been little weathering/degradation/alteration of the congener source pattern by processes operating during atmospheric transport or within the soil itself; PBDEs must therefore be quite stable in the terrestrial environment and in air. The first of these observations is important. The primary usage of penta-BDE has been as an additive flame retardant to polyurethane foam (PUF) (3, 16) used in furnishings, cars, textiles, etc. These materials therefore represent major reservoirs of penta-PBDEs. It is envisaged that congeners can transfer from these reservoirs to the ambient environment via two potential pathways. First, Hale and co-workers have reported that PUF crumbles as it ages to form a dust, which could be dispersed into the environment (23, 29). Second, it has been shown that PBDEs can volatilize from PUF to air (30). Indeed, this has been demonstrated across the range of congeners present in the technical penta-PBDE mix (30), even though their vapor pressures and KOA differ substantially (31). The second of the observations made above is also interesting. Once released to the ambient environment, PBDEs exist in the atmosphere in the vapor and particle phases. Lower brominated congeners have a greater proportion in the gas phase at a given temperature than their higher brominated counterparts (32). Differences in gas-particle partitioning can lead to differences in atmospheric reactivity and LRAT potential (5, 9). In addition, deposition processes (dry gaseous, dry particulate, wet) operate with different efficiencies for different compounds in the environment and are affected by environmental conditions. However, comparison of the composition of technical penta-PBDE, the composition of ambient air, and the average composition of European background soils suggests that the net “efficiency” of these processes is again broadly similar across the range of congeners. Some differences are apparent, however: the most obvious is that the PBDE-47:99 ratio reported in Bromkal 70-5DE is about 1:1 (19), while the composition in the soils analyzed here averaged ca. 1:1.8. Because it is one of the major components, the relative decline of PBDE-47 from the mixture in soils elevates the contribution of all the other congeners. It is unclear why PBDE-47 shows this decline. Possible explanations may be that it undergoes proportionately greater degradation and/or LRAT to outside the European region or that other technical formulations, with different ratios, impact the pattern too. 742
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 3, 2004
FIGURE 3. Comparison of land cover for the 5 main congeners: (A) results expressed as pg/g dry weight; (B) pg/g organic matter.
FIGURE 4. Surface and subsurface PBDE concentrations at selected locations (pg/g dry weight). The UK land surface area is 2.3 × 1011 m2, of which only ∼3% is described as “urban”. If it is assumed that the median rural soil ∑ALLPBDE concentration is 1000 ng kg-1, the UK urban surface soils contain concentrations which areson averages100 times higher, and the bulk density of surface soils averages 1 g cm-3, then the estimated ∑ALLPBDE burden in UK soils is of the order of ∼10 t. This represents only
FIGURE 5. Percent contribution of selected congeners vs latitude. ∼0.3% of the estimated cumulative penta-BDE usage in the UK (16). Factors that alter the “general” background soil PBDE mixture are discussed later, but this section has shown that overall these result in relatively subtle alterations of the pattern found in the technical penta-PBDE formulation and air. Comments on BDE-183. As noted earlier, BDE-183 is a marker for the octa-PBDE mix. In most soils, it made a minor contribution to the ∑ALLPBDE concentration, but it was a major component in some samples from northern England. These locations are the closest to major population centers of all the transect sites; in those locations, it contributed up to 50% of the ∑ALLPBDE. Typically, however, it constituted -99/-100 > -153 (9). Ter Schure et al. (34) also provided evidence for the latitudinal fractionation of PBDEs in the Scandinavian Peninsula in their study of biological specimens. Fractionation of compounds can be caused by differences in the LRAT potential of compounds following a primary emission. As just noted, for example, heavier PBDEs may be more prone to scavenging/deposition processes from a given air mass than lower molecular weight congeners. However, a more pronounced fractionation of the mixture of compounds in soil could develop if there were repeated temperature-driven air-surface exchange (the “grasshopper effect”; 35), with lighter molecules undergoing more “hops” than their heavier counterparts (36). Clues about whether this process has occurred for PBDEs can be found by examining the soil data in more detail, as discussed in the next section. Observations on Relationships with SOM. SOM dominates POP air-soil partitioning behavior. %SOM has previously been shown to influence the concentrations of POPs in global background soils (14). Indeed, this UK-Norway transect provided evidence that the more volatile PCBs are moving toward equilibrium with the OM burden of the soil compartment on a European regional scale (15). Plots of POP concentration against %SOM provide useful information 744
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 3, 2004
on POP air-soil exchange and their tendency to hop. If regressions give a slope of zero, soil POP concentrations are obviously affected by factors other than SOM, such as proximity to sources of deposition processes. This situation would be observed for “sticky” (e.g., particle-bound) compounds that are not re-emitted from soil. In contrast, if a steep slope is obtained, this indicates that the POP can reenter the atmosphere (preferentially from soils of lower %SOM), re-deposit, and tend to be retained in soils of higher %SOM content. Repeated air-soil exchange would result in steeper slopes over time (14, 36). Similarly, at any one time, different compounds would be expected to give different slopes, with steeper slopes for the POPs that are more prone to hopping. Figure 6 presents regression data of soil concentration versus %SOM for PBDE-47, -99, -100, and -153. There is clearly a lot of scatter on the data, indicating that factors other than %SOM have a major influence on soil concentrations. Presumably proximity to sources and atmospheric transport are such factors. However, it suggests that PBDE-47 and -99 have undergone some hopping to higher %SOM soils, while the stickier PBDE-153 has not. It should be noted that there is a potentially confounding factor in this relationshipsthe location of soils of different SOM content. Some of the soils with the highest %SOM are in the far north of Norwaysfurthest from sources.
Acknowledgments We are grateful to the Egyptian Ministry of Higher Education for doctorate funding for A.H.
Supporting Information Available One table showing correlations between PBDEs and PCBs. This material is available free of charge via the Internet at http://pubs.acs.org.
Literature Cited (1) Commission of the European Communities. Risk Assessment of Diphenyl Ether, Pentabromoderivative (Pentabromodiphenyl
(2) (3) (4) (5) (6) (7) (8) (9) (10)
(11)
(12)
(13) (14) (15) (16) (17)
Ether); CAS Number: 32534-81-9; EINECS Number: 251-0842; Rapporteur: United Kingdom, 2000. BSEF. An Introduction to Brominated Flame Retardants; Bromine Science and Environmental Forum 28; 2000. Alaee, M.; Arias, P.; Sjo¨din, A.; Bergman, A. Environ. Int. 2003, 29, 683-689. Watanabe, I.; Sakai, S. Environ. Int. 2003, 29, 665-682. Klecka, G.; et al. Evaluation of persistence and long-range transport of organic chemicals in the environment; SETAC Special Publication Series; SETAC: Pensacola, FL, 2000. Rahman, F.; Langford, K. H.; Scrimshaw, M. D.; Lester, J. N. Sci. Total Environ. 2001, 275, 1-17. de Wit, C. Chemosphere 2002, 46, 583-624. Law, R. J.; Alaee, M.; Allchin, C. R.; Boon, J. P.; LeBeuf, M.; Lepom, P.; Stern, G. A. Environ. Int. 2003, 29, 757-770. Wania, F. Dugani, C. B. Environ. Toxicol. Chem. 2003, 22, 12521261. DETR. Risk Reduction Strategy and Analysis of Advantages and Drawbacks for Pentabromodiphenyl Ether; Report for the U.K. Department of the Environment, Transport and the Regions; 2000. UNECE. Protocol on Persistent Organic Pollutants under the 1979 Convention on Long-Range Transport Air Pollution; United Nations Economic Commission for Europe (ECE/EB. Air/60): 1998. UNEP. Preparation of an Internationally Legally Binding Instrument for Implementing International Action on Certain Persistent Organic Pollutants; United Nations Environment Programme, UNEP/POPs/Inc.1/6: United Nations: New York, 1998. Duarte-Davidson, R.; Sewart, A. P.; Alcock, R. E.; Cousins, I.; Jones, K. C. Environ. Sci. Technol. 1997, 31, 1-11. Meijer, S. N.; Ockenden, W. A.; Sweetman, A. J.; Breivik, K.; Grimalt, J. O.; Jones, K. C. Environ. Sci. Technol. 2003, 37, 667672. Meijer, S. N.; Steinnes, E.; Ockenden, W. A.; Jones, K. C. Environ. Sci. Technol. 2002, 36, 2146-2153. Alcock, R. E.; Sweetman, A. J.; Prevedouros, K.; Jones, K. C. Environ. Int. 2003, 29, 691-698. Jaward, F. M.; Farrar, N.; Harner, T.; Sweetman, A.; Jones, K. C. Environ. Sci. Technol. 2004, 38, 0000-0000.
(18) Gouin, T.; Thomas, G. O.; Cousins, I.; Barber, J.; Mackay, D.; Jones, K. C. Environ. Sci. Technol, 2002, 36, 1426-1434. (19) Sjo¨din, A.; Jakobsson, E.; Kierkegaard, A.; Marsh, G.; Sellstro¨m, U. J. Chromatogr. A 1998, 822, 83-89. (20) Ockenden, W. A.; Sweetman, A.; Prest, H. F.; Steinnes, E.; Jones, K. C. Environ. Sci. Technol. 1998, 32, 2795-2803. (21) Meijer, S. N.; Ockenden, W. A.; Steinnes, E.; Corrigan, B. P.; Jones, K. C. Environ. Sci. Technol. 2003, 37, 454-461. (22) Lee, R. G. M.; Thomas, G. O.; Jones, K. C. Organohalogen Compd. 2002, 58, 193-196. (23) Hale, R. C.; Alaee, M.; Manchester-Neesvig, J. B.; Stapleton, H. M.; Ikonomou, M. G. Environ. Int. 2003, 29, 771-779. (24) Breivik, K.; Sweetman, A.; Pacyna, J. M.; Jones, K. C. Sci. Total Environ. 2002, 290, 181-198. (25) Breivik, K.; Sweetman, A.; Pacyna, J. M.; Jones, K. C. Sci. Total Environ. 2002, 29, 199-224. (26) Gouin, T.; Harner, T. Environ. Int. 2003, 29, 717-724. (27) Litz, N. J. Plant Nutr. Soil Sci. 2002, 165, 692-696. (28) Huber, S.; Ballschmiter, K. Fresenius J. Anal. Chem. 2001, 371, 882-890. (29) Hale, R. C.; La Guardia, M.; J.; Harvey, E.; Mainor, M. T. 2002, Chemosphere 2002, 46, 729-735. (30) Wilford, B. H.; Thomas, G. O.; Alcock, R. E.; Jones, K. C.; Anderson, D. R. Organohalogen Compd. 2003, 61, 219-222. (31) Harner, T.; Shoeib, M. J. Chem. Eng. Data 2002, 47, 228-232. (32) ter Schure, A. F. H.; Larsson, P. Atmos. Environ. 2002, 36, 40154022. (33) Wania, F.; McLachlan, M. Environ. Sci. Technol. 2001, 35, 582590. (34) Ter Schure, A. F. H.; Larsson, P.; Merila, J.; Jonsson, K. I. Environ. Sci. Technol. 2002, 36, 5057-5061. (35) Wania, F.; Mackay, D. Environ. Sci. Technol. 1996, 30, 390-396. (36) Gouin, T.; Mackay, D.; Jones, K. C.; Harner, T.; Meijer, S. N. Environ. Pollut. 2004, 128, 139-148.
Received for review September 12, 2003. Revised manuscript received November 15, 2003. Accepted November 20, 2003. ES035008Y
VOL. 38, NO. 3, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
745
