Environ. Sci. Technol. 2002, 36, 1959-1964
PBCDD and PBCDF from Incineration of Waste-Containing Brominated Flame Retardants GUNILLA SO ¨ DERSTRO ¨ M* AND STELLAN MARKLUND Environmental Chemistry, Umeå University, S-901 87 Umeå, Sweden
Brominated organic flame retardants comprise a large, heterogeneous group of compounds that are useful but also potentially damaging to the environment. In this investigation, three common brominated flame retardants were cocombusted with an artificial municipal solid waste in a pilotscale fluidized bed incinerator. Combustion conditions ensured degradation of the flame retardants, but stable brominated organic compounds such as polybrominated dibenzo-p-dioxins and polybrominated dibenzofurans were formed in the cooling flue gases. The incineration of fuels that contain both chlorine and bromine leads to the formation of a complex mixture of polybrominated/ chlorinated dibenzo-p-dioxins and dibenzofurans. More bromination than chlorination was observed in these experiments, and the chlorination levels increased when bromine was added to the fuel. The most favored reactions led to the formation of dibromodichlorodibenzo-p-dioxin and dibromodichlorodibenzofuran. Theoretical calculations show that there is more Br2 than Cl2 in the flue gas when the two halogens are present at equimolar levels, because chlorine is mainly found in the form of HCl. This may explain the higher bromination level. BrCl is also present, which drives the chlorination observed when bromine is added.
Introduction Brominated Flame Retardants. The use of brominated flame retardants is rising and strongly linked to their presence in appliances that are being produced in increasing levels for use in homes and offices such as TVs, computers, and other pieces of electronic equipment. Flame retardants are also used in diverse materials such as fabrics, paints, and furnishings for car interiors. They are relatively inexpensive to produce and are miscible with most plastics. Over the years, more and more brominated flame retardants have been reaching waste or metal recycling treatment facilities, where they often undergo thermal treatment. The physicochemical nature of brominated flame retardants makes them very effective for preventing fires, but it can complicate deliberate efforts to burn them. The molecular structure of the most commonly used brominated flame retardants has evolved over the years; polybrominated biphenyls (PBBs), analogues of the corresponding polychlorinated biphenyls (PCBs), were among the first to be commonly used in products. However, a poisoning accident in Michigan (1) reduced the use of PBBs and initiated moves in favor of other * Corresponding author phone: +46-90-786 66 64; fax: +46-90786 66 64; e-mail:
[email protected]. 10.1021/es010135k CCC: $22.00 Published on Web 03/27/2002
2002 American Chemical Society
compounds. Today, some of the most frequently used brominated flame retardants are the polybrominated diphenyl ethers (PBDEs). Because of the extended use of brominated flame retardants in products, the level of bromine in waste will increase, with a delay of several years as the products become waste. In Japan, 20 000 tonnes of brominated flame retardants were used in 1986, and this figure had increased to 51 450 tonnes by 1994 (2). The bromine content in Swedish waste has generally been very low in the past; it accounted for only approximately 0.004% by weight, compared to 0.75% for chlorine in 1992 (3). However, in the electronic waste treatment and metal recycling processes, the plain plastic and metal-containing parts of most products are usually separated. Both fractions, which can contain brominated flame retardants, are then often combusted batchwise. During this batchwise combustion, the levels of organic bromine will increase dramatically and may well rise to match the chlorine levels. Brominated Dioxins and Furans from Incineration. It has already been demonstrated that brominated flame retardants can form polybrominated dibenzo-p-dioxins (PBDDs) and polybrominated dibenzofurans (PBDFs) during combustion and pyrolysis processes (4-6). The formation of PBDDs and PBDFs from brominated flame retardants in combustion and pyrolysis has been examined in several investigations. Most studies have focused on thermal treatment experiments in quartz reactors with fuel mixtures consisting of only a few components. In all of these experiments, the formation of PBDDs and PBDFs was demonstrated. Funcke et al. (7) performed full-scale experiments at the TAMARA experimental reactor in Germany with low levels of bromine in the fuel. They found that the formation of polybromochlorodibenzo-p-dioxins (PBCDDs) and polybromochlorodibenzofurans (PBCDFs) was correlated with the bromine input. However, the formation rates of PBCDD/Fs were much higher than the corresponding rates in comparable pure chlorine experiments. Lemieux and Ryan (8) noted a similar discrepancy. Other experiments at the TAMARA combustor showed no formation of PBCDD/F when bromine was present in the fuel (9). A series of systematic experiments in a quartz reactor were performed by Kanters and Louw (10) in which more PBDD/Fs were found to be formed from brominated phenols than PCDD/Fs from corresponding chlorophenols. It is likely that PBDDs and PBDFs are formed in a similar way as PCDDs and PCDFs but with different rates of formation. Different reaction pathways have been suggested for PCDDs and PCDFs, suggesting that the chlorinated dibenzofurans are formed in de novo reactions and the chlorinated dioxins from precursors. The formation pathway, whereby chlorinated dibenzofurans are generated in combustion systems, has been established by Wikstro¨m (11). She showed that the unchlorinated dibenzofurans were formed in the combustion zone like other PAHs. The second step in the formation process took place in the cooling zone (600-250 °C), where HCl was converted to Cl* by a catalyst such as copper, and the Cl then substituted a hydrogen atom in the dibenzofuran molecule. As only minor amounts of dibenzo-p-dioxin are formed in the PAH process in contrast to dibenzofurans, most chlorinated dioxins are formed via chlorophenol precursors, but they can undergo further chlorination steps. Environmental Aspects of Halogenated Dioxins and Furans. Polychlorinated dibenzo-p-dioxins (PCDDs) and polychlorinated dibenzofurans (PCDFs) have been thoroughly investigated from toxicological and environmental perspectives. PCDDs and PCDFs in relatively low doses are VOL. 36, NO. 9, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1959
TABLE 1. Fuel Halogen Composition, Combustion Parameters, and Halogen Inputs in the Experimentsa halogen balance fuel composition flame retardant
fuel H A1 A2 B D C E F G I a
DeBDE, high HBCD, high DeBDE, low HBCD, low TBBP-A, low TBBP-A, low reproducible run
bromine
chlorine
Br/Cl % in % Cl % Br molar feed flue feed % in w/w w/w ratio g/h gas g/h fluegas 0.1 0.75 0.75 0.75 0.75 0.75 0.75 0.75 0.1 0.75
0 0 0 0.87 1.7 1.7 0.87 0.87 0.87 1.7
0:0 0:1 0:1 0.5:1 1:1 1:1 0.5:1 0.5:1 1:0 1:1
0 0 0 21 17 10 9 9 10
25 43 31 17 42 11
0.1 8 15 9 8 9 8 8 0.1
35 52 5.5 44 65 65 45 65
combustion parameters
T T T °C °C fuel particle CO2 CO SO2 NO2 °C % ppm ppm ppm bed combustion sampling kg/h g/m3
O2 % 8.4 9.0 10 11 9.6 8.3 9.9 9.2 8.2 9.3
13 12 10 15 12 13 10 12 13 12
3 11 283 820 1 107 269 804 1 82 278 785 573 33 260 779 314 2 326 739 244 7 351 773 26 0 302 769 24 0 322 771 36 0 338 790 599 3 243 715
803 804 807 817 814 813 814 812 808 823
300 301 297 299 298 296 299 299 297 298
1.1 1 2.2 1.2 1 1.1 1 1 1.1
2.0 1.4 1.0 2.4 1.6 1.7 1.2 1.5 2.0 1.9
DeBDE, decabromodiphenyl ether; HBCD, hexabromodiphenyl ether; TBBP-A, tetrabromobisphenol A.
carcinogenic, terratogenic, and cause chloro-acne. High levels of exposure cause an irreversible starvation syndrome. PBDD/ Fs and PBCDD/Fs have also been compared, to some extent, in toxicological studies, and various clear tendencies have been noted. Generally, changing the halogen does not alter their toxicity much (12). However, the differences in toxicity are heavily dependent on the effect being considered. The differences in molecular size and electronegativity between the corresponding species can either decrease or increase toxicity, depending on the mechanism involved. No European studies have reported any findings of PBCDD/F in the environment (13), but one investigation found PBDF in sewage sludge (14). A Japanese analysis from 1999 reported high levels of both PBCDF and PBDD/F in atmospheric deposition (15). Aims of the Study. The purposes of this study were to compare the formation of PCDD/Fs, PBCDD/Fs, and PBDD/ Fs in the combustion of municipal solid waste (MSW) with and without three different brominated flame retardants and to study the formation pathways for PBDDs, PBDFs, PBCDDs, and PBCDFs. As yet, very little is known about how the brominated flame retardants in waste affect the emissions of halogenated pollutants in MSW incineration. The data gained from pilot-scale combustions, such as those studied here, cannot be directly extrapolated to a full-scale plant but can provide some indications of how to deal with the combustion of electronic scrap. The chosen levels of bromine were high as compared to normal levels of bromine in MSW or industrial waste and were regarded as “worst-case scenarios” for the batchwise incineration of waste products with flame retardants.
Experimental Section Experimental Setup. Three common brominated flame retardants were selected for these tests: decabromodiphenyl ether (DeBDE), tetrabromobisphenol A (TBBP-A), and hexabromocyclododecane (HBCD). Altogether, 10 experiments were performed: nine using the experimental setup described here, with fuels A-H, and one repetition of an earlier unpublished experiment, with fuel I, to check the method’s reproducibility. To obtain an appropriate fuel matrix, we used a well-documented artificial fuel as a base (11). This fuel mimics “normal” Swedish/European MSW with respect to both its content and its tendency to form PCDD/Fs when combusted. The chlorine content in the waste was the same as that generally found in ordinary waste, 0.75%. This fuel was burned twice, without any addition of brominated compounds (A1 and A2). In the following experiments, bromine was added to two different levels to the 1960
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 9, 2002
reference fuel. The “high” bromine level was 1.7% w/w (fuels B and D) and was equal to the chlorine content found in “normal” MSW on a molar basis, and the “low” level, 0.87% w/w (fuels C, E, and F) corresponded to half of the chlorine content. DeBDE and HBCD were combusted in the high bromine level experiments, and all three flame retardants were combusted in the “low” bromine level experiments. In one experiment, no chlorine was present, but there was a “low level” of TBBP-A (fuel G). Finally, one combustion with neither chlorine nor bromine in the MSW-pellets was performed (fuel H). The composition of all the fuels can be seen in Table 1. Analytical Limitations. A total of 210 congeners of both PBDD/Fs and PCDD/Fs can be theoretically formed. The relative retention times of all of the PCDD and PCDF congeners have been determined on the GC column used in our experiments, allowing for all the isomers to be identified. For the PCDD/Fs, PBCDD/Fs, and PBDD/Fs, there are 5020 possible congeners. The large numbers of different congeners give rise to considerable analytical difficulties. One major problem is that not all PBDD/F and PBCDD/F isomers have been synthesized. Therefore, the relative retention times for brominated species have not been as comprehensively described in the literature as the corresponding times for their chlorinated analogues have been. Thus, it is likely that isomer-specific analyses will include some overestimations because coelution may occur. Because of the lack of standards, we have focused in this study on tetrahalogenated compounds, namely, PCDDs, PCDFs, monobromotrichlorodibenzo-p-dioxins (MoBrTriClDDs), monobromotrichlorodibenzofurans (MoBrTriClDFs), dibromodichlorodibenzop-dioxins (DiBrDiClDDs), dibromodichlorodibenzofurans (DiBrDiClDFs), tribromomonochlorodibenzofurans (TriBrMoClDFs), tetrabromodibenzo-p-dioxins (TeBrDDs), and tetrabromodibenzofurans (TeBrDFs). The total amounts generated of all isomers within the different groups of congeners with four halogens have been calculated in this study. Fuel Composition. The artificial MSW pellets used in these experiments contained paper, plastics, organic matter, metals (Cu 60 mg/kg), and 0.75% chlorine (organic: inorganic molar ratio, 1:1). All ingredients were thoroughly mixed before pelletizing to give all pellets the same basic composition. Brominated Fuels. Pre-prepared MSW pellets were mixed with three brominated flame retardants purchased from chemical vendors. DeBDE and HBCD supplied by Fluka and TBBP-A from Riedel de Hae¨n were all of technical grade. Because the fuel base was already mixed and pelleted, the flame retarding fuel was prepared by making a slurry of the
FIGURE 1. Umeå pilot-scale fluidized bed reactor. The combustion takes place in the freeboard, and the flue gases are cooled in the convector. crystalline flame retardant in toluene. The slurry was poured over the pellets in such a way that each pellet was coated with flame retardant. The toluene was evaporated from the pellets in a hood overnight. This procedure was repeated for each of the brominated fuel mixtures. Combustion. A pilot-scale fluidized bed incinerator was used for these experiments (Figure 1). The incinerator is 2.5 m high, and the convector section is 15 m long and has an effect of 5 kW. It has been described by Wikstro¨m et al. (16). In these experiments, it was supplied with approximately 1 kg fuel/h, the bed and combustion zone temperatures were 740-820 °C and 800-820 °C, respectively, and the residence time in the convector zone was 2-3 s. The reactor was preheated with propane and electrical heaters. Samples were taken from the midsection of the convector at 300 °C. Between experiments, the reactor and convector were cleaned in order to minimize memory effects. Otherwise, halogen-containing ash might have affected the subsequent experiment. Sampling and Analytical Procedure. During the incineration of the different fuels, samples were taken to analyze a range of combustion products. Inorganic combustion gases were analyzed on-line with an electrochemical apparatus for determining SOx, NOx, O2, and CO. CO2 and CO were analyzed simultaneously by an IR instrument. The particulate matter content of the flue gas was measured by capturing dust on a filter (an hour after the start of the solid fuel combustion) and determining its mass gravimetrically. The filters were heated to 130 °C and cooled in a desiccator before weighing. Cl2, HCl, Br2, and HBr were taken by a flue-gas bypass through a buffer. During the first hour of combustion, the reactor was allowed to equilibrate. Semivolatile organic matter was sampled using a sampling train that was initially designed for PCDD/F analysis but was also suitable for sampling other organic compounds such as PCBs, PAHs, chlorophenols, chlorobenzenes, and PBCDD/Fs. This is a European Committee for Standardization (CEN)-certified sampling train that has been described in detail by Marklund et al. (17, 18). Briefly, it consists of a cooled probe, a condensate bottle, an impinger with ethylene glycol, and a polyurethane foam adsorbent with an aerosol filter. Samples were taken in two 1-h periods: first after 2 h (Fuel1 samples) and the second after 3 h (Fuel2 samples). O2 levels were measured in outgoing gas after the pump to ensure that no air leaks diluted the samples. Before sampling, the train was “spiked” with 13C-labeled PCDDs, PCDFs, PentaBCDF, TeBDD, and TeBDF to act as internal standards. The measurements were performed according to the EU-standard protocol. After sampling, the samples were stored dark and cold. Before analysis, they were then cleaned-up according to standard European protocols (17) except for the liquid
extractions. For these, the condensate and impinger solutions were diluted with an excess of water and extracted with an SPME-disk (C-18 covered glass fiber disk, Supelco). PCDDs, PCDFs, TeBCDDs, and TeBCDFs were separated on a SP2330 column, and the samples were then analyzed on a VG 70250-S HRGC/HRMS, with a resolution of 8.000 Da. The masses used were monitored for the possible overlap of brominated diphenyl ethers, and the ratios of chloro/bromo clusters were checked. Calculations. The halogenated dioxins and furans were analyzed using an isotope dilution method. PCDD/F contents were calculated against the corresponding 13C-labeled internal standards, TeBCDD/F against 13C MoBrTriClDD/F, and TeBDD/F against 13C TeBDD/F. In analysis of organic pollutants, compounds are commonly analyzed in terms of picograms per gram, but here we also calculated the molar balances between PCDD/F, PBCDD/F, and PBDD/F in order to compare the reactivity of bromine and chlorine in this context.
Results and Discussion System and Combustion Conditions. With some exceptions, the pilot reactor provided almost identical combustion conditions for all experiments performed. Small differences in temperature were seen in the bed between fuels with and without flame retardants, because the flame retardants caused a cooling effect that cooled the bed somewhat. However, the temperature further up in the reactor was the same both with and without the flame retardants. The combustion conditions are listed in Table 1. Another difference noticed between the chlorinated and the brominated fuels was that CO levels were higher and SO2 levels were lower when bromine was present in the fuel. This decrease in SO2 level can be explained by the Br2 reducing reaction
Br2 + SO2 + H2O f SO3 + 2HBr The first sample (designated with a subscript 1) and second sample (designated with a subscript 2) taken during each experiment have different emission levels because the deposition of fly ash within the reactor increases with time. However, the first samples from all of the experiments are comparable with each other, and so are all of the second samples. The second experiment with normal MSW and without the bromine additive (A21 and A22) showed a strong deviation in the PCDD/PCDF ratio (>1) generated as compared to all of the other samples in the series (∼0.2) and also to the previously described MSW combustions (11). The reason for the deviation is unknown. Because of these discrepancies, experiments A21 and A22 are considered to be VOL. 36, NO. 9, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1961
TABLE 2. Results of PCDD, PCDF, TeBCDD, TeBCDF, TeBDD, and TeBDF Analyses (ng/m3 and pmol/m3)a ng/m3 ∑ TCDF ∑ PeCDF ∑ HxCDF ∑ HpCDF OCDF ∑ TCDD ∑ PeCDD ∑ HxCDD ∑ HpCDD OCDD pmol/m3 ∑ moBr triCl DF ∑ diBr diCl DF ∑ triBr moCl DF ∑ teBr DF ∑ moBr triCl DD ∑ diBr diCl DD ∑ teBr DD a
H1
H2
A11 A12 A21 A22
B1
B2
D1
D2
C1
C2
E1
E2
F1
F2
G1
G2
I2
0.2 0.6 nd nd nd nd nd nd nd nd
0.2 0.6 0.3 nd nd nd nd nd nd nd
6.2 56 28 8.7 nd 1.12 6.9 4.4 nd 0.8
8.6 73 47 15 1.5 1.4 10 8.1 4.9 1.9
3.6 17 5.9 1.2 nd 1.8 16 10 4.3 1.5
3.2 nd nd nd nd nd nd
2.1 nd nd nd nd nd nd
nd nd nd nd nd nd nd
nd nd nd nd nd nd nd
21 24 nd nd 14 nd nd
I (old)
4.7 21 7.6 1.9 nd 4.7 39 28 14 3.3
2.2 11 6.8 1.7 1.4 0.2 1.0 1.7 0.7 1.6
3.3 27 13 2.6 nd 0.4 2.6 1.1 0.7 nd
4.0 23 13 5.4 nd 0.5 3.5 2.3 1.2 0.5
3.0 34 17 5.9 0.8 0.4 3.6 2.5 1.4 1.9
2.5 20 11 3.4 nd 0.3 2.4 1.7 0.7 nd
5.7 35 15 4.4 2.0 1.0 3.3 2.9 1.3 3.0
3.2 23 13 3.6 nd 0.7 3.7 2.2 nd 0.3
4.0 25 14 4.0 nd 0.7 3.7 2.7 1.6 0.5
3.2 25 15 5.8 2.5 0.6 3.3 2.9 1.9 2.3
4.4 24 15 5.5 2.7 0.7 3.3 2.4 1.8 1.2
nd nd nd nd nd nd nd nd nd nd
nd nd nd nd nd nd nd nd nd nd
na na na na na na na na na na
na na na na na na na na na na
12 nd nd nd 2.5 nd nd
471 3600 809 2300 21 96 nd
273 1310 328 190 12 38 nd
226 1290 253 nd 9.2 43 nd
448 2460 619 375 25 91 22
550 1770 209 nd 35 54 nd
262 1240 292 239 13 44 nd
282 1040 100 nd 17 36 nd
228 616 106 121 12 26 nd
321 1.6 nd nd 20 155 nd
287 1230 214 153 19 49 nd
nd nd 54 501 nd nd nd
nd 45 94 715 nd nd 55
764 4520 1350 1350 28 109 76
776 6500 na 2500 84 433 54
nd, not detected; na, not analyzed; X1, first sample; X2, second sample.
TABLE 3. Theoretical Molar Fractions of Some Bromine and Chlorine Compounds in Flue Gas at an Input of Equal Molar Amounts of Bromine and Chlorine HCl Cl2 HBr Br2 BrCl
1050 K
800 K
550 K
300 K
0.19
