Click here to load reader

Brominated Flame Retardants Brominated Flame Retardants

  • View
    1

  • Download
    0

Embed Size (px)

Text of Brominated Flame Retardants Brominated Flame Retardants

Flame Retardants
Flame Retardants
The fact that several brominated flame retardants (BFRs) have been found in the environment, sometimes at increasing levels, has caused great concern in many fora.
The time trends studied, indicate increased levels of many BFRs in the environment since the 1970s but the levels of important lower-brominated compounds have begun to decline in the Baltic Sea since voluntary withdrawal of use in a number of countries. However, the human breast milk trend indicates that levels are increasing exponentially, doubling every five years.
The results may indicate that humans are exposed to these substances not just from the diet, but also from exposure to electronic appliances and textiles.
These findings, and many more, are discussed in this report. The results indicate that BFRs may be a new "PCB problem".
Brominated
Brominated
Brominated
E-mail: [email protected]
The author assumes sole responsibility for the contents of this report, which, therefore, cannot be cited as representing,
the viewpoint of the Swedish Environmental Protection Agency. The report has been submitted to external referees for review.
The author’s address: Institute of Applied Environmental Research (ITM)
Stockholm University SE-106 91 Stockholm, Sweden
E-mail: [email protected]
Customer Services SE-106 48 Stockholm, Sweden
Telephone: +46 8 698 12 00 Fax: +46 8 698 15 15
E-mail: [email protected] Internet: www.environ.se
Edition: 800 copies
Preface
The results presented in this report are, in many cases, from Swedish research projects funded by a number of national and international research programmes. These include grants from the Swedish Natural Science Research Council, the Swedish Medical Research Council, the Eu- ropean Commission Environment and Climate Program, the Swedish Environmental Protection Agency’s monitor- ing program (Brominated Flame Retardant Project) and the Persistent Organic Pollutants (POP) research program funded by the Swedish Environmental Protection Agency from 1992 to 1998.
This review has been compiled by Cynthia A. de Wit, Stockholm University, at the request of the steering group of the POP research program.
Swedish Environmental Protection Agency
5
Contents
Abbreviations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Summary in Swedish . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 1.1 What are flame retardants?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 1.2 Which substances are covered in this report? . . . . . . . . . . . . . . . . . . . . . . 12 1.2.1 Polybrominated diphenyl ethers (PBDEs) . . . . . . . . . . . . . . . . . . . . . . . 13 1.2.2 Tetrabromobisphenol A (TBBPA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 1.2.3 Hexabromocyclododecane (HBCD) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 1.2.4 Polybrominated biphenyls (PBBs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2. Brominated flame retardant chemistry . . . . . . . . . . . . . . . 17 2.1 Synthesis of labelled and unlabelled substances . . . . . . . . . . . . . . . . . . . . 17 2.2 Validation of synthetic substances . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 2.3 Physicochemical properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 2.4 Transformation/Decomposition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 2.4.1 Microbial . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 2.4.2 Photolytic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3. Analytical methods for brominated flame retardants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 4. Toxicology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 4.1 Uptake, distribution, metabolism, excretion . . . . . . . . . . . . . . . . . . . . . . . 24 4.1.1 Mammals. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 4.1.1.1 Lower brominated PBDEs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 4.1.1.2 DeBDE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 4.1.1.3 TBBPA. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 4.1.1.4 HBCD. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 4.1.2 Fish and shellfish. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 4.1.2.1 Lower brominated PBDEs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 4.1.2.2 DeBDE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 4.1.2.3 TBBPA. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 4.1.2.4 HBCD. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 4.2 Toxicity/effects. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 4.2.1 PBDEs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 4.2.1.1 In vitro . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 4.2.1.2 In mammals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 4.2.1.3 In fish . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 4.2.2 TBBPA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
6
4.2.2.1 In vitro . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 4.2.2.2 In mammals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 4.2.2.3 In fish . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 4.2.3 HBCD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 4.2.3.1 In vitro . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 4.2.3.2 In mammals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 4.2.3.3 In fish . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
5. Environmental concentrations . . . . . . . . . . . . . . . . . . . . . . . . 41 5.1 Abiotic samples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 5.1.1 Air . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 5.1.2 Water . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 5.1.3 Sewage sludge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 5.1.4 Sediment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 5.2 Biological samples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 5.2.1 Terrestrial ecosystem. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 5.2.1.1 Birds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 5.2.1.2 Mammals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 5.2.1.3 Humans. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 5.2.2 Freshwater ecosystem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 5.2.2.1 Fish . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 5.2.2.2 Birds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 5.2.3 Marine ecosystem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 5.2.3.1 Fish and shellfish . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 5.2.3.2 Birds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 5.2.3.3 Mammals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 5.3 Bioaccumulation, biomagnification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
6. Trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 6.1 Spatial trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 6.2 Temporal trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 6.2.1 Baltic Sea sediment core . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 6.2.2 Pike from Lake Bolmen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 6.2.3 Roach from Lake Krankesjön . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 6.2.4 Guillemot eggs from St. Karlsön. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74 6.2.5 Human breast milk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
7. Summary and conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78 8. References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
7
Abbreviations
MROD Methoxy resorufin-O-deethylase
Summary in Swedish
Flamskyddsmedel används i plast, gummi och textilier för att förhindra bränder. Direkta källor till utsläpp av bromerade flamskyddsmedel som upptäckts i Sverige är knutna till plast- och textilindustrier och förhöjda halter har påvisats i sediment och fisk nedströms sådana anläggningar. Stu- dier av damm och inomhusluft visar att dessa substanser även finns i ar- betsmiljön, möjligen genom läckage från plastdetaljer i datorer och annan elektronisk utrustning. Polybromerade difenyletrar (PBDE) finns i dag överallt i Sverige och till och med i Arktis. Förekomst i utomhusluft indi- kerar att dessa kan spridas via långväga transport. Bromerade flamskydds- medel har även hittats i vatten, sediment och rötslam.
I Sverige är halterna i landlevande däggdjur och fåglar låga, men är hög- re i djur från akvatisk miljö. Relativt höga halter har påvisats i däggdjur och fisk från Östersjön men också från Nordsjön. Betydligt lägre halter har på- visats i norra Sverige och Arktis. PBDEs förekomstmönster liknar alltså det som vi sett hos PCB och DDT.
De högsta halterna av PBDE i Sverige har påvisats i sediment (ng/g torr- vikt) och fisk (µg/g fettvikt) längs Viskan i anslutning till textilindustrier. Även DeBDE and HBCD har påvisats i sediment och HBCD i fisk från Viskan. Höga halter TeBDE, PeBDE, OcBDE och DeBDE har hittats i se- diment från flera brittiska floder där utsläpp från industrier som tillverkar eller använder PBDE pågår (upp till µg/g torrvikt). Fisk i de påverkade flodmynningarna visar också höga halter av TeBDE, PeBDE och OcBDE (upp till µg/g fettvikt).
Lägre bromerade PBDE, OcBDE och HBCD är biotillgängliga och åter- finns i fisk. Upptaget via mag-tarmkanalen i råttor, möss och fisk är högt för lägre bromerade PBDE. Musungars upptag av dessa substanser från bröstmjölk är också högt. Mönstren skiljer sig dock mellan olika arter vil- ket delvis kan bero på skillnader i metabolism. PBDE kan metaboliseras till hydroxylerade PBDE. Upptaget av HBCD från mag-tarmkanalen är högt hos råttor. DeBDE är biotillgänglig även om upptaget är långsamt i fisk men det verkar som den del som tas upp debromeras i fisken till lägre bro- merade PBDE.
2,2',4,4'-TeBDE (BDE-47), 2,2',4,4',5-PeBDE (BDE-99) och 2,2',4,4',6- PeBDE (BDE-100) biomagnificeras i lax, fiskätande fåglar och däggdjur.
9
BDE-47, -99 och -100 har hittats i blod från människa och lax och PBDE finns även i människors fettvävnad och bröstmjölk. Högre halter BDE-47, hexaBDE (BDE-153 och -154), heptaBDE (BDE-183) och DeBDE har hit- tats i personer som arbetar med återvinning och destruktion av elektronisk utrustning jämfört med människor som arbetar framför datorer eller som lokalvårdare. Flera hydroxylerade och metoxylerade PBDE (Te- och PeB- DE) har hittats i lax, strömming, vikare och gråsäl från Östersjön. Metox- ylerade PBDE har också hittats i gädda från en sötvattensjö. Ursprunget till dessa ämnen är ej känt.
Flera lägre bromerade PBDE, inklusive BDE-47, -99 och -100, verkar antingen aktiverande eller hämmande via Ah-receptorn (dioxinreceptorn). Bromkal 70-5DE, en PeBDE teknisk produkt, inducerar Ah-receptorme- dierade leverenzymer, t ex EROD, in vitro och i råttor och regnbåge in vi- vo. BDE-47 och -99 har däremot visats sänka EROD-aktiviteten i regn- bågslever. Hydroxylerade PBDE och TBBPA är potenta agonister för transtyretin (TTR), ett plasmaprotein som transporterar sköldkörtelhormo- ner i blodet. Bromerade analoger till sköldkörtelhormonerna tyroxin (T4) och triiodotyronin (T3) binder också till dessas receptorer. Råttor och möss behandlade med Bromkal 70-5DE eller BDE-47 visar minskade tyroxin- halter och ändringar i immunsystemet. Möss matade med BDE-47 eller BDE-99, 10 dagar efter födsel, visar permanenta förändringar i spontant motoriskt beteende som förvärras med åldern. Liknande exponering till BDE-99 orsakade även effekter på inlärning och minne när djuren blev vuxna.
Vissa bromerade flamskyddsmedel kan alltså inducera eller nedreglera leverenzymproduktion, påverka tyroideahormonsystemet, vara immuno- toxiska samt neurotoxiska när exponering sker vid en känslig tidpunkt för hjärntillväxt. Mycket litet är känt om TBBPA utöver att det in vitro kan binda till de plasmaproteiner som transporterar sköldkörtelhormoner i blo- det. Däremot tycks det inte kunna binda till samma protein in vivo. Nästan inget alls är känt om HBCD's effekter.
Studier över tid visar ökade halter av PBDE och HBCD i miljön sedan 1970-talet. Tidstrenden i sillgrissla visar att nivåerna av BDE-47, -99 och -100 i Östersjön har börjat sjunka efter mitten av 1980-talet medan tids- trenden i gädda från Bolmen indikerar att halterna inte har förändrats sedan början av 1980-talet. I bröstmjölk däremot har halterna ökat exponentiellt med en dubblering var femte år. Detta kan indikera att exponeringssitua-
10
tionen är olika i Östersjön, sötvattensekosystem och för människor. Män- niskor exponeras inte bara via maten utan även från den miljö man lever i, i hemmet och på arbetsplatsen, vilket kan påverka den samlade expo- neringen.
PBDE, TBBPA och HBCD finns i miljön. De tas upp av levande orga- nismer och åtminstone de lägre bromerade PBDE biomagnificeras. TBB- PA och PBDE och/eller deras metaboliter är biologiskt aktiva. Halterna av PBDE verkar öka, och trenden i människor indikerar att ökningen kan vara snabb. Om inte denna trend bryts kommer den att leda till att belastningen i vilda djur och i människor når nivåer som orsakar effekter. Vår kunskap om dessa substanser, deras källor, toxicitet och beteende i miljön är än så länge mycket begränsad, vilket gör riskbedömningen osäker. Särskilt PBDE är ett miljöproblem på grund av deras persistens, lipofilicitet och bioackumulerbarhet. Kunskapen om TBBPA och HBCD är så begränsad att det idag är omöjligt att ens uppskatta vilken risk de utgör. Tillsammans indikerar dessa resultat att vi kan stå inför ett nytt ”PCB problem”.
11
1. Introduction
1.1 What are flame retardants? Flame retardants are substances used in plastics, textiles, electronic circu- itry and other materials to prevent fires. Some of the technical flame retard- ant products contain brominated organic compounds including polybromi- nated biphenyls (PBBs), polybrominated diphenyl ethers (PBDEs), tetra- bromobisphenol A (TBBPA) and hexabromocyclododecane (HBCD). The structures for these are shown in Figure 1. Many of these substances are persistent and lipophilic and have been shown to bioaccumulate.
Some brominated flame retardants are additives mixed into polymers and are not chemically bound to the plastic or textiles and are therefore able to leak out of products and into the environment. Additives include PBBs,
Figure 1. The chemical structures of a) polybrominated diphenyl ethers, b) hexabromocyclododecane (HBCD), c) tetrabromobisphenol A (TBBPA), and d) polybrominated biphenyls.
12
PBDEs and HBCD. Others, such as TBBPA, are reactive and are bound to the material chemically. However, some of the reactive flame retardant may not have polymerized and can thus leak out of the product.
Brominated flame retardants as such are not produced in Sweden. For the most part, these substances are imported into Sweden in finished prod- ucts and goods or in materials used to make products and goods. The cur- rent environmental levels may partly be the result of leaching from first generation flame-retarded products dumped in landfills over a decade ago. Knowledge about these substances is very limited and hinders environ- mental authorities from carrying out adequate risk assessments.
The major companies producing brominated flame retardants are Ethyl Corporation (USA), Great Lakes Chemical Corporation (USA and UK), Dead Sea Bromine (Israel) and Eurobrom (Netherlands). Other companies include Riedel de Haen (Hoechst Group), Ceca (ATOCHEM, France), Po- tasse et Produit Chimiques (Rhone Poulenc Group), Warwick Chemicals (UK), Albemarle S.A. (Belgium) as well as Nippo, Tosoh and Matsunaga all from Japan (KEMI, 1994a; WHO/IPCS, 1994b). The total world pro- duction of all brominated flame retardants is approximately 150 000 metric tons/year. Forty percent is distributed to North America, 30% to the Far East and 25% to Europe (KEMI, 1994a).
1.2 Which substances are covered in this report? PBBs have not been prioritized in this report for several reasons. Due to a major PBB-poisoning accident in the USA in 1973, substantial knowledge of the toxicology of some PBBs is now available (KEMI, 1990; WHO/ IPCS, 1994a). For a recent review of PBBs see de Boer et al. (2000). Also, the environmental levels of PBBs are low in Sweden. Thus, the knowledge base for risk assessment of PBBs in Sweden is more satisfactory. The ma- jor emphasis of this report is therefore on PBDEs, and to some extent on TBBPA and HBCD, as knowledge about these substances is incomplete or in some cases, almost non-existent. Other brominated organic substances are also of concern, such as polybrominated dibenzo-p-dioxins and furans. PBDD/Fs are formed, for example, when plastics containing brominated flame retardants are heated (welding of mats, melting of polymers). How- ever, coverage of PBDD/Fs is outside the scope of this report. Interested
13
readers are referred to recent reports on PBDEs and PBDD/Fs by WHO/ ICPS for more information (WHO/ICPS, 1994b; 1997; 1998).
1.2.1 POLYBROMINATED DIPHENYL ETHERS (PBDES) There are theoretically 209 PBDE congeners. Technical PBDE products are produced by brominating diphenyl ether in the presence of a catalyst. The major technical products contain mainly pentaBDEs, octaBDEs or de- caBDE, but contain other PBDEs as well. The general compositions of the technical products are given in Table 1. The individual PBDE congeners are numbered according to the IUPAC system used for numbering PCBs based on the position of the halogen atoms on the rings.
Table 1. The general compositions of PBDE-based flame retardants given in percent of BDE congeners present (WHO/IPCS, 1994b).
PBDEs are mainly imported to Sweden incorporated in plastics used in production or in finished products such as TVs, computers, electrical equipment, household appliances, cables, furniture, textiles, etc. Annual worldwide production of penta-, octa- and decaBDE technical products in 1990 was estimated to be 4 000, 6 000 and 30 000 metric tons respectively (Arias, 1992). Consumption of PBDEs for 1999 in the European Union was estimated to be 150 metric tons penta-, 400 metric tons octa- and 7 000 metric tons decaBDE technical products (De Poortere, 2000). Accumulat- ed amounts in electric products still in use in 1994 in the Nordic countries were estimated to be 250 metric tons of PeBDE and 5500 metric tons of OcBDE and DeBDE combined (Hedelmalm et al., 1995). The estimated amounts supplied to Sweden alone in 1991 were 17 metric tons of PeBDE and 626 metric tons of OcBDE and DeBDE combined (Hedelmalm et al., 1995). The combined total annual amounts of penta-, octa- and decaBDE technical products recently imported to Sweden are given in Table 2. For 1993, approximately 20 metric tons of PBDE were imported to Sweden
Congener percent
Technical Product
Tetra- BDEs
Penta- BDEs
Hexa- BDEs
Hepta- BDEs
Octa- BDEs
Nona- BDEs
Deca- BDE
OcBDE 10–12 44 31–35 10–11 <1
DeBDE <3 97–98
14
and the same amount was imported in plastics (KEMI, 1995). The amount of PBDE imported in ready-made products the same year was estimated by the Swedish National Chemicals Inspectorate to be approximately 400 metric tons (KEMI, 1995).
Based on this information, technical products containing DeBDE are es- timated to be used in larger quantities than those containing TeBDEs and PeBDEs.
Table 2. The import of brominated flame retardants to Sweden in metric tons (KEMI, 1999). Due to confidentiality regulations, only the sum of penta-, octa- and decabrominated prod- ucts as well as the sum of TBBPA and its carbonate oligomer are available.
*Only decabrominated products were imported.
1.2.1.1 Lower brominated PBDEs (TeBDEs, PeBDEs, HxBDEs) PBDEs are lipophilic and have some structural similarities to PCB and PCDD/F. Bromkal 70-5DE is one of several PeBDE technical products that contains lower brominated PBDEs such as 2,2',4,4'-TeBDE (BDE-47), 2,2',4,4',5-PeBDE (BDE-99) (Sundström and Hutzinger, 1976), 2,2',4,4',6- PeBDE (BDE-100), as well as two newly identified triBDEs, one addition- al tetraBDE, one additional pentaBDE, three hexaBDEs and one hep- taBDE (Sjödin et al., 1998). PBDEs were first discovered in Sweden in fish samples taken downstream from several textile industries on the River Viskan (Andersson and Blomkvist, 1981). In some countries, industries are voluntarily replacing the lower brominated PBDEs with other flame retard- ants.
1.2.1.2 Higher brominated PBDEs (Hepta-NonaBDEs, DecaBDE) OcBDE technical products are used in acrylonitrile butadiene styrene, polycarbonate and thermosets. DeBDE products are used in most types of synthetic materials including textiles and polyester used for printed circuit boards (OECD, 1994). DeBDE is the most widely used technical product of the PBDEs. DeBDE has been shown to photolytically degrade in labo- ratory experiments (Watanabe and Tatsukawa, 1987; U. Örn, Stockholm
Year 1993 1994 1995 1996 1997 1998
PBDE 22 90 20 79 123 41*
HBCD 50 80 90 83 125 89
TBBPA 488 450 258 291 303 269
15
University, personal communication; Sellström et al., 1998a) and form lower brominated PBDE. If this process is significant, it could lead to the formation of TeBDEs, PeBDEs and HxBDEs, which are known to accu- mulate in living organisms.
1.2.2 TETRABROMOBISPHENOL A (TBBPA) TBBPA is covalently bound to plastic and is used in electronic circuit boards. Annual worldwide production has been estimated to be 50 000 metric tons/year (KEMI, 1994a). Estimated accumulated amounts of TBB- PA in products in the Nordic countries in 1994 were 4000 metric tons (Hedelmalm et al., 1995). The estimated amount supplied to Sweden alone in products in 1991 was 334 metric tons. More recent information on the import of TBBPA to Sweden is given in Table 2. One moiety of TBBPA’s molecular structure is similar to that of the thyroid hormone thyroxine, ex- cept that the iodine atoms have been replaced by bromines (see Figure 1c, Figure 2). The dimethylated derivative of TBBPA (MeTA) may have some use as a flame retardant, but may also be the result of methylation of TBB- PA in sediment (Watanabe et al., 1983).
1.2.3 HEXABROMOCYCLODODECANE (HBCD) HBCD is produced by bromination of cyclododecane in a batch process. HBCD has been used for about 20 years. Approximately 11 000 metric tons are incorporated in articles made of polystyrene and in textile back- coating for the EU market. It is used in foams and expanded polystyrene. End products include upholstered furniture, interior textiles, automobile
Figure 2. The chemical structures of the thyroid hormones 3,3',5,5'-tetraiodo-L- thyronine (thyroxine or T4) and its 5'-deiodinated congener 3,3',5-triiodo- thyronine (T3).
16
interior textiles, car cushions, insulation blocks in trucks and caravans as well as in building materials such as house walls, cellars, roofs and parking decks, against frost heaving in roads and railway embankments, packaging material, video cassette recorder housing and electric equipment (Marga- reta Palmquist, National Swedish Chemicals Inspectorate, personal com- munication). The amounts of HBCD imported to Sweden for the last few years are given in Table 2.
1.2.4 POLYBROMINATED BIPHENYLS (PBBS) In 1973, a commercial flame retardant containing PBBs was accidently mixed into feed for dairy cattle, livestock and poultry in the state of Mich- igan, USA (KEMI, 1990; WHO/IPCS, 1994a). The feed was used widely, leading to widespread PBB-contamination of milk, meat and eggs and poi- soning in animals. Over 9 million people were exposed to PBBs from food. Because of this widespread exposure, research was funded to better under- stand the toxicology of PBBs, and poisoned animals and exposed humans have been studied as well. The effects of PBBs were found to be essentially the same as those seen for PCBs.
Technical hexabrominated biphenyl (HxBB) is banned in North Ameri- ca and in Europe. Technical decabrominated biphenyl (DeBB) is still pro- duced in Europe, but production will be discontinued after 2000 (De Poor- tere, 2000).
17
2. Brominated flame retardant chemistry
2.1 Synthesis of labelled and unlabelled substances Considerable work has been carried out to synthesize pure PBDE conge- ners for use as standards, for toxicology studies and in the identification of unknown substances in environmental samples. Within several research programs, a number of both unlabelled and radiolabelled PBDEs have been synthesized for these purposes (Örn et al., 1996; Hu, 1996; Jakobsson et al. 1996; Örn, 1997; Örn et al., 1998; Marsh et al., 1999). This work is sum- marized in Bergman, (1999) and in Table 3.
2.2 Validation of synthetic substances Using synthesized single congeners of some PBDEs, it has been possible to better quantify the components of the technical product Bromkal 70- 5DE and to identify the previously unknown Pe1BDE (2,2',4,4',6-PeBDE) in this product. According to previously published results, Bromkal 70- 5DE consisted of 41% 2,2',4,4'-TeBDE, 45% 2,2',4,4',5-PeBDE and 7% 2,2',4,4',6-PeBDE (Sundström and Hutzinger, 1976). Using the new stand- ards, it has been found that the composition is actually 37%, 35% and 6.8% for these three PBDEs, respectively (Sjödin et al., 1998). Bromkal 70-5DE also contains 1.6% of 2,2',3,4,4'-PeBDE (BDE-85) and 3.9, 2.5 and 0.41% of HxBDEs -153, -154 (2,2',4,4',5,6'-HxBDE) and -138, respectively. A number of other standards that have been produced have also been ana- lyzed in order to use them in identifying and quantifying more PBDEs in environmental samples on a congener-specific basis.
2.3 Physicochemical properties The PBDEs have low vapor pressures and are very lipophilic with Log Kows (octanol-water partitioning coefficients) ranging from 5.9-6.2 for TeBDEs, 6.5–7.0 for PeBDEs, 8.4-8.9 for OcBDEs and 10 for DeBDE (Watanabe and Tatsukawa, 1990). TBBPA has a Log Kow of 4.5 (WHO/
18
IPCS, 1995). The dimetylated derivative of TBBPA (MeTA) has a Log Kow
of 6.4, making it more lipophilic than the parent compound (Watanabe and Tatsukawa, 1990). The Log Kow for HBCD is 5.8 (IUCLID, 1996).
Table 3. Names and congener numbers of PBDEs synthesized.
Name Congener number
Reference
2-bromodiphenyl ether BDE-1 Jakobsson et al., 1996; Marsh et al., 1999
3-bromodiphenyl ether BDE-2 Jakobsson et al., 1996; Marsh et al., 1999
4-bromodiphenyl ether BDE-3 Jakobsson et al., 1996; Marsh et al., 1999
2,4-DiBDE BDE-7 Jakobsson et al., 1996; Marsh et al., 1999
2,4'-DiBDE BDE-8 Jakobsson et al., 1996; Marsh et al., 1999
2,6-DiBDE BDE-10 Jakobsson et al., 1996; Marsh et al., 1999
3,4-DiBDE BDE-12 Jakobsson et al., 1996
3,4'-DiBDE BDE-13 Jakobsson et al., 1996; Marsh et al., 1999
4,4'-DiBDE BDE-15 Jakobsson et al., 1996; Marsh et al., 1999
2,2',4-TrBDE BDE-17 Jakobsson et al., 1996; Marsh et al., 1999
2,3',4-TrBDE BDE-25 Jakobsson et al., 1996; Marsh et al., 1999
2,4,4'-TrBDE BDE-28 Jakobsson et al., 1996; Marsh et al., 1999
2,4,6-TrBDE BDE-30 Jakobsson et al., 1996; Marsh et al., 1999
2,4',6-TrBDE BDE-32 Jakobsson et al., 1996; Marsh et al., 1999
2',3,4-TrBDE BDE-33 Marsh et al., 1999
3,3',4-TrBDE BDE-35 Jakobsson et al., 1996
3,4,4'-TrBDE BDE-37 Jakobsson et al., 1996
2,2',4,4'-TeBDE BDE-47* Örn et al., 1996; Jakobsson et al., 1996; Örn, 1997; Marsh et al., 1999
2,2',4,5'-TeBDE BDE-49 Marsh et al., 1999
2,2',4,6'-TeBDE BDE-51 Jakobsson et al., 1996; Marsh et al., 1999
2,3',4,4'-TeBDE BDE-66 Jakobsson et al., 1996
2,3',4',6-TeBDE BDE-71 Jakobsson et al., 1996
2,4,4',6-TeBDE BDE-75 Jakobsson et al., 1996; Marsh et al., 1999
3,3',4,4'-TeBDE BDE-77 Marsh et al., 1999
2,2',3,4,4'-PeBDE BDE-85* Örn et al., 1996; Örn,, 1997
2,2',4,4',5-PeBDE BDE-99* Örn et al., 1996; Örn, 1997
2,2',4,4',6-PeBDE BDE-100 Jakobsson et al., 1996; Marsh et al., 1999
19
*Radiolabelled congeners using 14C were also synthesized.
PBDEs are persistent, have low water solubility, high binding affinity to particles and a tendency to accumulate in sediments. HBCD also has low water solubility (IUCLID, 1996), and probably also has an affinity for par- ticles and sediments.
2.4 Transformation/Decomposition 2.4.1 MICROBIAL
A study of anaerobic microorganisms’ ability to break down DeBDE to lower brominated PBDE in sediment was carried out during 1994. DeBDE was applied to anaerobic sediment which was then inoculated with micro- organisms enriched from a PBDE-contaminated sediment. The sediment was then divided into smaller samples and allowed to gently shake. Sam- ples were analyzed at different time points but no breakdown of DeBDE was seen during the experimental time of four months. The experiment was extended by letting one aliquot of sediment continue incubation. Subsam- ples were analyzed at several time points but no breakdown could be seen after an incubation of 2 years (unpublished results, Ulla Sellström; de Wit, 1995; 1997).
Studies of TBBPA in sediments have shown that a dimethylated deriva- tive is also found (Sellström and Jansson, 1995). The origin of this meth- ylated TBBPA is not completely understood, however it probably has
2,3,4,5,6-PeBDE BDE-116 Jakobsson et al., 1996; Marsh et al., 1999
2,3',4,4',6-PeBDE BDE-119 Jakobsson et al., 1996
2,2',3,3',4,4'-HxBDE BDE-128 Örn et al., 1996; Örn, 1997
2,2',3,4,4',5'-HxBDE BDE-138 Örn et al., 1996; Örn, 1997
2,2',3,4,4',6'-HxBDE BDE-140 Marsh et al., 1999
2,2',4,4',5,5'-HxBDE BDE-153 Örn, et al., 1996; Örn, 1997
2,2',4,4',5,6'-HxBDE BDE-154 Marsh et al., 1999
2,3,4,4',5,6-HxBDE BDE-166 Jakobsson et al., 1996; Marsh et al., 1999
2,2',3,4,4',5,6-HpBDE BDE-181 Jakobsson et al., 1996; Marsh et al., 1999
2,3,3',4,4',5,6-HpBDE BDE-190 Jakobsson et al., 1996
Name Congener number
Reference
20
some use as a flame retardant (WHO/IPCS, 1997) but may also be due to methylation of TBBPA by microorganisms in the sediment (Watanabe et al., 1983).
2.4.2 PHOTOLYTIC
Previous studies indicate that DeBDE is debrominated by UV light and sunlight to lower brominated PBDE (to triBDE with UV light and to tetraBDE with sunlight) but it is not known if this happens in the environ- ment (Norris et al., 1973, 1975a; Watanabe and Tatsukawa, 1987; Ulrika Örn, Department of Environmental Chemistry, Stockholm University, pers. comm.).
Recent laboratory studies of the photolytic breakdown of DeBDE showed that DeBDE in toluene and applied to silica gel is successively de- brominated by UV light to lower brominated PBDE (down to TeBDEs) and that this occurs very rapidly (Sellström et al., 1998a). The half-life in toluene was less than 15 minutes. DeBDE-treated sand, soil and sediment were exposed to UV light in the laboratory or to sunlight outdoors for dif- ferent time periods and then analyzed. Results from the sand series exposed to both UV and sunlight show that DeBDE is photolytically debrominated in the same manner as in solution and on silica gel, but that the debromina- tion time course proceeds more slowly for both UV and sunlight exposure (half-life of 12 and 37 hours, respectively). Similar results were obtained for sediment, with a half-life for DeBDE of 53 hours for UV-exposure and 81 hours for sunlight exposure and with TeBDEs appearing at the longest exposure time (244 hours). Anomalous results were obtained for the soil samples exposed to sunlight, making interpretation difficult. However, the results for soil samples treated with DeBDE and exposed to UV-light in the laboratory indicated a half-life of 185 hours and the debromination process was the same as that seen for all the other matrices tested.
TBBPA is also photolytically decomposed when exposed to UV light, both in the absence and presence of hydroxyl radicals (Eriksson and Jakobs- son, 1998). The main breakdown product is 2,4,6-tribromophenol (Figure 3). A number of other decomposition products are also found and some of these have been tentatively identified as di- and tri-bromobisphenol A, di-
21
time(h)
Expon. (TBBPA)
Figure 3. Decomposition of TBBPA and formation of tribromophenol (TBP) (Eriksson and Jakobsson, 1998).
22
3. Analytical methods for brominated flame retardants In most analytical methods for brominated organic compounds, the sample is first extracted with an organic solvent. Lipids can be removed using sul- furic acid treatment or gel permeation chromatography methods. In some cases the extract needs to go through a further clean-up step using some form of column chromatography to remove interfering substances. The pu- rified extract is then analyzed with gas chromatography using electron cap- ture detection (ECD), mass spectrometry using the negative ions formed at chemical ionization (MS-ECNI) (Jansson et al., 1987; Jansson et al. 1991; Sellström et al., 1993a; 1998a; Nylund et al. 1992) or high resolution gas chromatography-masspectrometry (HRGC-MS)(Meironyté et al., 1999). The higher brominated PBDEs have longer retention times and are often analyzed using a shorter GC column (Sellström, 1996; Sellström et al., 1998a). For a more detailed review of analytical methods, see de Boer et al. (2000).
TBBPA is separated from the neutral components by treatment with a basic water solution. The pH of this water solution is then adjusted to be acidic and TBBPA is extracted with an organic solvent. Before analysis, TBBPA is derivatized to its diacetylated derivative. Dimethylated TBBPA (MeTA) is analyzed with the PBDEs (Sellström and Jansson, 1995).
Recovery experiments have been performed for TeBDE (BDE-47), PeBDE (BDEs -99, -100) and DeBDE (BDE-209) in fish muscle and for BDEs -47, -99, -100, -209 and HBCD in sediment (Sellström et al. 1998b). The results showed recoveries of 111–114% for fish muscle and 106-140% for sediment (Sellström et al., 1998b).
It is difficult to quantify HxBBs because of interference by HxBDEs in the analysis. An HPLC method was therefore developed to separate PBBs from PBDEs using an HPLC amino column. ”Clean” fish samples were spiked with known amounts of several brominated flame retardants (tetra- hexaBBs, methylated TBBPA, TeBDE, PeBDE, HxBDE, DeBDE, HB- CD), extracted and the extracts fractionated using the new separation meth- od. To further validate the method, a few fish and sediment samples from River Viskan were extracted and the extracts separated using the new
23
method. The method was successful in separating the PBBs from PBDEs and the levels of PBBs were found to be very low in these samples (Sell- ström et al., 1994).
Hydroxylated PBDE are separated from the neutral components by treatment with a basic water solution and then derivatized with diazometh- ane. Methoxy-PBDE are found in the neutral fraction. The analysis is car- ried out using gas chromatography-mass spectrometry or GC-ECD (Asp- lund et al., 1997; 1999a; Haglund et al., 1997).
24
4. Toxicology Thorough reviews of previous studies of the toxicology of PBBs, PBDEs and TBBPA can be found in the relevant WHO/IPCS reports (1994a; 1994b; 1995) and in Darnerud et al. (1998) for PBDEs.
4.1. Uptake, distribution, metabolism, excretion 4.1.1 MAMMALS
4.1.1.1 LOWER BROMINATED PBDES
The distribution of 14C-labelled 2,2',4,4'-tetrabromodiphenyl ether (BDE- 47), 2,2',3,4,4'-pentabromodiphenyl ether (BDE-85), and 2,2',4,4',5-penta- bromodiphenyl ether (BDE-99) was studied in C57BL mice using whole- body autoradiography, and the latter two congeners were also used in milk transfer studies during the neonatal period, by use of liquid scintillation technique (Darnerud and Risberg, 1998).
The autoradiograph findings in this study showed that the uptake from the gastrointestinal tract was effective. Organs and tissues with high radio- activity concentrations were the fat depots, liver, adrenal and ovary, lung and (shortly after administration) the brain. At longer post-injection time, the apparent concentration in most tissues was considerably lower, and ra- dioactivity was still present in white and brown fat depots. Studies in preg- nant mice showed a low fetal uptake of the compounds. No significant dif- ference in distribution between the three studied congeners was observed. Breast milk transport of the pentaBDE congeners was substantial and 30– 40% of the administered single dose of radioactivity was found in the suck- ling offspring litter after four days. At the same time point, the plasma lev- els in the neonates were more than two times that of the mothers, although the absolute levels were low. The concentration in tissues and milk was in about the same range as that earlier observed for PCB congeners of a sim- ilar degree of halogenation, when administered in equimolar doses.
Uptake, distribution, metabolism and excretion of BDE-47 has been studied in rats and mice dosed orally with 14C-labelled BDE-47 (Klasson- Wehler et al., 1996; Örn, 1997; Örn and Klasson-Wehler, 1998). In rats,
25
14% was excreted in feces and less than 0.5% in urine during 5 days. Of the amount excreted, 79% was the parent compound and 21% correspond- ed to metabolites. Treated rats retained 86% of the administered dose after 5 days and the highest concentrations were found in adipose tissue, where the 14C corresponded to the parent BDE-47. All tissues analyzed contained parent BDE-47. Liver also contained low concentrations of five hydroxy- lated metabolites. Two of these were found in plasma (Örn och Klasson- Wehler, 1998).
In mice, 20% of the dose was excreted in feces and 33% in urine. Of the amount excreted, 15% was parent compound and 85% corresponded to metabolites. Treated mice retained 47% of the administered dose after 5 days and similar concentrations were found in both adipose tissue and liv- er. Covalently bound metabolites were found in liver (12% bound to mac- romolecules, 16% to lipids), in the lungs (28%) and kidneys (4%). Low concentrations of five hydroxylated metabolites were also observed in the liver. Three of these were seen in plasma as well (Örn och Klasson-Wehler, 1998). The conclusions drawn from this study were that BDE-47 was ab- sorbed well by both the rat and mouse, but the rate of metabolism and ex- cretion varied considerably.
Tissue disposition, metabolism and excretion of BDE-99 has been stud- ied in bile-cannulated and uncannulated male rats dosed orally with 14C- labelled BDE-99 (Hakk et al., 1999; Larsen et al., 1999). Feces was the major route of elimination in both groups (43% of the administered dose in uncannulated and 86% in cannulated rats after 72 h). Of the amount excret- ed in feces, more than 90% was parent BDE-99. Cumulative excretion of metabolites in both groups of rats was less than 1% of the administered dose into urine and 3.7% into bile. In the uncannulated rats, 6.3% of the 14C in urine was protein bound and was associated with alpha-2-microglobulin. In cannulated rats, 28–47% of the 14C in bile was bound to an unidentified protein of 79 kDa. In the uncannulated rats, 39% of the administered dose was retained and the highest concentrations were found in adipose tissue, skin and adrenals. Covalently bound 14C was observed in feces indicating the formation of reactive metabolic intermediates. Two monomethoxy pentabromodiphenyl ether metabolites and two de-brominated monometh- oxy tetrabromodiphenyl ether metabolites were indicated in feces. Two monohydroxy pentabromodiphenyl ether metabolites, two dihydroxy
26
pentabromodiphenyl ether metabolites and possibly two thio-substituted pentabromodiphenyl ether metabolites were characterized in bile.
In a recent study, 17 PBDE congeners (BDEs -15, -28, -30, -32, -47, -51, -71, -75, -77, -85, -99, -100, -119, -138, -153, -166 and -190) were incubat- ed individually with rat hepatic microsomes from rats treated with beta- naphthaflavone, phenobarbital or clofibrate (Meerts et al., 1998b). The original congeners and the metabolites formed were then tested for their ability to compete with the thyroid hormone thyroxine (T4) for binding to human transthyretin in vitro. A number of hydroxylated PCBs are structur- al analogues of T4 and are known to compete with T4 in this system. The structural requirements for binding to transthyretin are hydroxy-substitu- tion in para- or meta-positions of one or both phenyl rings with adjacent halogen substitution (Lans, 1995). X-ray diffraction studies have shown that several phenolic organohalogen compounds bind with the hydroxy group in the central channel of the transthyretin molecule.
Results showed no competition with the parent compounds, but consid- erable potency for several of the metabolites, indicating the metabolism of PBDE to hydroxylated PBDE (see 4.2.1.1 for results). No binding compe- tition was seen for several of the higher brominated PBDE such as BDEs - 138, -153, -166 and -190 after incubation with the microsomes. This may indicate that these congeners are not readily metabolized.
4.1.1.2 DeBDE The uptake of 14C-DeBDE has previously been shown to be low following oral administration to rats, and between 90% to 99% of the dose was elim- inated in the feces and gut (Norris et al., 1975b; El Dareer et al., 1987). A 2-year feeding study of DeBDE in rats confirmed a low accumulation, however, the small portion that reached the adipose tissue, measured as the total bromine content, remained unaffected for 90 days of recovery (Norris et al., 1975b; Kociba et al., 1975). When 14C-DeBDE was injected intra- venously, 74% of the dose was found in feces and gut contents after 72 hours. Of the excreted material, 63% was metabolites and 37% was the par- ent compound (El Dareer et al., 1987).
4.1.1.3 TBBPA TBBPA fed to rats was eliminated primarily in feces (95%) and 1% was eliminated in urine within 3 days (WHO/IPCS, 1995).
27
14C-labelled TBBPA was fed to rats with and without bile duct cannula- tion (Larsen et al., 1998). Bile, urine and feces were collected every 24 hours for three days, after which the rats were killed and organs and tissues sampled. Three conjugated metabolites were found in the bile – a diglu- curonide, a monoglucuronide and a glucuronide-sulfate ester. These repre- sented 34, 45 and 21% of the radioactivity excreted in bile during the first 24 hours, respectively. In the cannulated rats, 71% of the dose was excreted in the bile within 72 hours. In uncannulated rats, 95% of the 14C-labelled TBBPA was excreted in the feces as TBBPA. However, there was a delay in fecal excretion. This is the result of enterohepatic circulation of TBBPA where the biliary metabolites are deconjugated and reabsorbed from the lower intestine, reconjugated and re-excreted in the bile. About 2% of the dose remained in the uncannulated rat after 72 hours and highest levels were found in the large intestine (1%), small intestine (0.6%), lung (0.2%) and carcass (0.2%).
Uptake and distribution of 14C-labelled TBBPA in pregnant rats after oral exposure on gestational days 10 to 16 has recently been studied (Meerts et al., 1999). The major portion of radioactivity was excreted in feces (79.8%). Only 0.83% of the total administered dose was found in the tissues of the dams and 0.34% in the fetuses. Highest maternal levels were found in the carcass (0.37%) and liver (0.26%).
4.1.1.4 HBCD Rats fed a single oral dose of 14C-labelled-HBCD rapidly absorbed the sub- stance (Yu and Atallah, 1980). HBCD was readily distributed in the entire body with highest concentrations found in adipose tissue, followed by liv- er, kidney, lung and gonads. The half-life was two hours. HBCD was rap- idly metabolized and 72% of the dose was eliminated via feces and 16% by urine within 72 hours. Four metabolites were found but no information on their structures was given. HBCD absorption followed a two-compartment open model system, with the central compartment consisting of blood, muscle, liver, kidney and other non-adipose tissues, and the peripheral compartment consisting of fat tissue. Elimination from fat was slower than for the central compartment.
Rats fed HBCD daily for 5 days showed no urinary excretion of HBCD (Ryuich et al., 1983). Average daily fecal excretion was 29–37% of the ad- ministered amount. HBCD was found to accumulate in adipose tissue in
28
this study. Experiments using a loop of the upper jejunum suggest that HBCD can be absorbed from the intestine.
4.1.2 FISH AND SHELLFISH
4.1.2.1 LOWER BROMINATED PBDE A study of uptake, accumulation and excretion of BDE-47, -99 and -153 from water was carried out in blue mussels (Edulis mytilus) (Gustafsson et al., 1999). Several PCB congeners were included for comparison (tri-hexa- CBs). The uptake clearance rates were found to be approximately ten times higher for BDE-47 and -99 than for BDE-153 and the PCB congeners. Depuration rates were similar for all three PBDEs indicating no depend- ence on hydrophobicity, but were correlated to hydrophobicity for the PCBs.
A dietary uptake study of TeBDE, PeBDE and HxBDE has been carried out in pike (Burreau et al., 1997). Pike were fed with rainbow trout that had been injected with a mixture containing BDE-47, -99 and -153 as well as selected PCB and PCN congeners. The uptake efficiencies of all three PB- DEs were high, with BDE-47 showing the highest uptake (more than 90% of the given dose) from the gastrointestinal tract. Uptake efficiencies for BDE-99 and – 153 were 62% and 40% repectively. The uptake efficiency of BDE-47 was higher than for the tri- to hexaCBs and the hexa- to octaC- Ns tested. The uptake efficiencies were higher than expected considering the size and lipophilicity of the compounds. It was concluded that uptake of these substances from the gastrointestinal tract may be facilitated by cotransport with lipids and/or proteins through a mediated or even active transport mechanism.
In a distribution study, several pike were fed rainbow trout previously injected with 5 microcuries of 14C-2,2',4,4'-TeBDE (BDE-47). Pike were killed after 9, 18, 36 and 65 days and whole-body autoradiography per- formed. The results showed accumulation of the labelled TeBDE in liver, gall blader, kidneys, brain, chorion of the eye, in the perivisceral adipose tissue and along the spinal column (Figure 4) (Burreau and Broman, 1998; Burreau et al., 2000).
Asplund et al. (1997; 1999a) have found hydroxylated and methoxylat- ed PBDE in Baltic salmon blood plasma, and Haglund et al. (1997) found methoxylated PBDE in ringed and grey seal, salmon and herring from the
29
Baltic Sea. Kierkegaard et al. (1999) found methoxy-BDE-47 in pike from Lake Bolmen, a freshwater lake. It is not clear what the source of these are but one possibility may be due to metabolism either in the organisms or by microorganisms. Natural production by invertebrates or algae can not be ruled out.
4.1.2.2 DeBDE A major argument made by flame retardant producers for the use of DeBDE is that the molecule is so large that it isn’t bioavailable and there- fore will not be accumulated by living organisms. To determine if this was the case, a dietary uptake study of DeBDE in rainbow trout was carried out (Kierkegaard et al., 1994; Kierkegaard et al., 1999a). Some biological ef- fects were also measured. Trout were fed with food containing DeBDE (10 mg DeBDE/kg/day) for 0, 16, 49 and 120 days. One group was treated with clean food for 71 days after 49 days of exposure to study elimination.The DeBDE concentrations in muscle ranged from <0.6 ng/g fresh weight after 0 days, to 38 (±14) ng/g fresh weight after 120 days. Corresponding liver concentrations were <5 and 870 (±219) ng/g fresh weight. A number of or- ganic brominated substances, characterized as hexa- to nonaBDEs, in- creased in concentration with exposure length.
Gall Bladder Muscle Eye
Skin
Figure 4. Tape section radiogram showing the distribution of radiolabelled 2,2',4,4'-TeBDE in pike 9 days after dietary exposure (Burreau and Broman, 1998).
30
After the 71 day depuration period, DeBDE levels declined, but the lev- els of some of the lower brominated congeners were unaffected during the same period. This may indicate that DeBDE is metabolized via a reductive debromination process to the lower brominated diphenyl ethers and/or that lower brominated PBDE present in the technical product are selectively ab- sorbed. However, no HxBDEs were detected in the technical product and there was a pronounced shift in increasing peak heights for the first-eluting congeners in the fish compared to the technical product they were fed. The estimated uptake of DeBDE was 0.02-0.13% based on DeBDE concentra- tions in muscle and an estimation of the concentrations of the metabolites formed compared to the total dose administered.
A number of physiological and biochemical variables were investigated. Liver body index (indicating increased liver weight) and plasma lactate levels increased in fish exposed for 120 days and in the depuration group. The number of lymphocytes was significantly lower after 120 days expo- sure compared to controls. DeBDE did not affect ethoxyresorufin-O- deethylase (EROD), ethoxycoumarin-O-deethylase (ECOD) or transketo- lase activity and no DNA adducts were seen.
4.1.2.3 TBBPA TBBPA is rapidly taken up from water via the gills in fish. When fish are placed in clean water, the compound is rapidly eliminated (WHO/IPCS, 1995).
4.1.2.4 HBCD Thirty fathead minnows (Pimephales promelas) were exposed to HBCD in water for 32 days (Veith et al., 1979). Five fish were removed for analysis after 2, 4, 8, 16, 24 and 32 days of exposure. The bioconcentration factor was estimated to be 18 100 (log BCF 4.26). Based on these results, HBCD seems to have high bioaccumulating potential.
In a study along the River Viskan, HBCD was found in both sediments and in pike (Sellström et al., 1998b), indicating that HBCD is bioavailable. It is probable that the levels found in pike are due to a combination of uptake via the gills and via food (gastrointestinal tract).
31
4.2 Toxicity/effects As has been stated previously, more detailed reviews of the toxicology, in- cluding the toxicity and effects of PBBs, PBDEs and TBBPA, can be found in the relevant WHO/IPCS reports (1994a; 1994b; 1995).
4.2.1 PBDES
4.2.1.1 IN VITRO A PeBDE technical product, Bromkal 70-5DE, has weak dioxin-like tox- icity as measured in rat H-4-II E hepatoma cells (Hanberg et al., 1991). Using a recombinant H-4-II E rat hepatoma cell line having Ah-receptor mediated expression of a luciferase reporter gene (the CALUX as- say)(Aarts et al., 1995; Murk et al., 1998), a number of individual PBDE congeners have been tested for their potency to activate/deactivate the Ah receptor (Meerts et al., 1998a). In order to study antagonism, the same PBDE congeners were also tested in the presence of 2,3,7,8-tetrachloro- dibenzo-p-dioxin (TCDD). Seven of the 17 PBDE congeners (BDEs -32, -85, -99, -119, -153, -166, -190) tested showed ability to activate the Ah- receptor. Potencies could only be determined for BDE-166 and BDE-190 and these are in the same range as the mono-ortho PCB congeners 105 and 118 (Sanderson et al., 1996). Some congeners such as BDE-85, -99 and -119 showed both agonist and antagonist activities depending on the concentration tested. Nine congeners, including BDEs -15, -28, -47, -77 and -138, showed antagonist activities against TCDD. The observed an- tagonism may be due to competition between PBDEs and TCDD at the Ah-receptor level.
Studies of the potency of 17 PBDE congeners and their hydroxylated metabolites for competitive binding to human transthyretin in vitro have recently been carried out (Meerts et al., 1998b). Transthyretin is the thyroid hormone transport protein present in plasma and has a binding site for the thyroid hormone thyroxine (T4). The parent compounds were incubated with rat hepatic microsomes from rats treated with beta-naphthaflavone (NF), phenobarbital (PB) or clofibrate (CLOF). The parent compounds and the metabolites formed from the different microsomal incubations were then tested for their ability to compete with T4 for binding to transthyretin. The results are given in Table 4 and show that a number of metabolites were potent competitors for transthyretin. The parent compounds showed
32
no competition. For example, after PB-microsomal incubation of 2,2',4,6'- tetraBDE (BDE-51), hydroxylated PBDE metabolites displaced T4 from transthyretin with fairly high potency (Meerts et al., 1998b). The results in- dicate that hydroxylated metabolites of PBDE may be potent competitors of T4 and could disrupt normal thyroid hormone function in wildlife and humans if present.
Table 4. Inhibition of T4-transthyretin binding in vitro by PBDE metabolites obtained after incubation with rat hepatic microsomes induced by phenobarbital (PB), beta-naphthafla- vone (NF) or clofibrate (CLOF). Inhibition potencies are given from the undiluted extract. ++=60% inhibition, +=20% inhibition, – =0–20% inhibition. (From Meerts et al., 1998b).
Transthyretin carries T4 in the plasma to the target tissues, where T4 is then deiodinated to 3,3',5-triiodothyronine (T3) (Figure 2). T3 then inter- acts with two sub-types of thyroid hormone receptors (THRs) designated alpha and beta. The T3-THR complex can then bind to response elements on the DNA that regulate the transcription of thyroid hormone activated
BDE congener tested
15 ++ ++ –
28 ++ + +
30 ++ ++ ++
32 – + +
47 ++ – –
51 ++ – +
71 + + +
75 ++ + +
77 ++ + +
85 + – –
99 + – –
100 ++ – –
119 ++ – –
138 – – –
153 – – –
166 + – –
190 – – –
33
genes. T4 can also interact with THR but has only about 10% of the poten- cy of T3. To determine the potency of hydroxylated PBDE to bind to THR- alpha and THR-beta, several brominated structural analogues of T4 and T3 were synthesized: 4-hydroxydiphenyl ether, 4-hydroxy-2',4',6'-tribro- modiphenyl ether (III), 3-bromo-4-hydroxy-2',4',6'-tribromodiphenyl ether (IV) and 3,5-dibromo-4-hydroxy-2',4',6'-tribromodiphenyl ether (V) (Fig- ure 5) (Marsh et al., 1998).
The results showed that the highest affinity was found for the T3 ana- logue (IV), with the T4 analogue (V) showing about one third of the affin- ity of the T3 analogue. The affinity of analogue III was low and was lowest for 4-hydroxydiphenyl ether (Marsh et al., 1998). The brominated ana- logues had lower affinities than T3 and T4, probably because of the lack of the 4-carboxyl group. The results indicate that hydroxylated metabolites of PBDE may not only disrupt the normal transport of T4 to target tissues, but may also be able to bind to the thyroid hormone receptors, thus influencing the regulation of thyroid hormone dependent genes within the cell nucleus.
Mitogen-induced DNA synthesis and immunoglobulin synthesis by hu- man lymphocytes in vitro was examined after exposure to purified BDE- 47 and -85. No effects on mitogen-induced proliferation or immunoglobu- lin synthesis were observed (Fernlöf et al., 1997). The results indicate that proliferation and immunoglobulin synthesis are insensitive to the direct ac- tion of polybrominated diphenyl ethers. Exposure to different PCB conge- ners (CB-77, CB-118, CB-153) also gave no effects.
BDE-47 was shown to induce a statistically significant increase in intra- genic recombination when studied in one of two tested in vitro assays using mammalian cells (Helleday et al. 1999). This may indicate that BDE-47
Figure 5. Synthetic pathway and chemical structures of several hydroxylated PBDE (Marsh et al., 1998).
34
can induce cancer via a non-mutagenic mechanism, similarly to other en- vironmental contaminants such as DDT and PCB.
4.2.1.2 In mammals Technical PBDE products are able to induce both phase I and phase II de- toxification enzymes in the liver. Regarding the cytochrome P450 (CYP) mediated phase I metabolism, CYP1A1 and 1A2 are induced; this could be shown by the increased activity of liver microsomal 7-ethoxyresorufin O- deethylase (EROD) after Bromkal 70 (a commercial pentaBDE) exposure in Wistar rats (von Meyerinck et al. 1990) and in H-4-II E cells (Hanberg et al., 1991). Other enzymes that are used as indicators of microsomal phase I activity were also induced by PBDEs (technical pentaBDE prepa- rations in rats) including benzphetamine N-demethylation, p-nitroanisole demethylase, aryl hydrocarbon hydroxylase (AHH) and benzo(a)pyrene hydroxylase (von Meyerinck et al., 1990; Carlson 1980a; b). Some of the enzymes were induced in a long-term oral administration study in rats at a concentration as low as ca 1 µmol/kg (technical pentaBDE), and the en- zymes remained induced 30–60 d after termination of exposure (Carlson, 1980b). DeBDE seems to have low enzyme-inducing potency. However, as CYP1A1 and 1A2 are typically induced by halogenated dioxin-like compounds, possible contaminants with Ah-receptor binding affinity present in technical PBDE mixtures could be responsible for the enzyme induction seen in these studies.
In interaction studies of PBDE, PCB and chlorinated paraffins (CP), mi- crosomal enzyme activities were studied in rats (Hallgren and Darnerud, 1998). The results showed that the pure congener BDE-47 (2,2',4,4'-tetra- BDE) increased EROD and MROD activities only somewhat (to twice the control levels), indicating that this substance has a low CYP1A1/2 induc- ing activity. In the same study, PCB (Aroclor 1254) markedly induced EROD and MROD. Earlier results on a commercial PBDE mixture (Bro- mkal 70) showed a rather strong induction of EROD and MROD in rats, which suggests that the Bromkal mixture contained CYP 1A1/2-inducing substances, probably present as contaminants (unpublished studies in Darnerud and Sinjari, 1996).
Regarding other microsomal enzymes, PROD levels (which mirror CYP2B activities) were measured in rats exposed to BDE-47 (Hallgren and Darnerud, 1998). In this case the PROD levels were dose-dependently el-
35
evated up to 10 times in exposed animals, and the levels were about the same as those found after PCB exposure. As the microsomal enzyme in- duction results are indicative of operational metabolic systems, PBDE may therefore, at least to some extent, be transformed by CYP2B in the phase I metabolism step.
In studies on phase II induction, three different PBDE fractions, i.e. a low (24% tetra, 50% penta) and a high (45% hepta, 30% octa) brominated mixture, and the DeBDE congener only, were tested. Daily oral adminis- trations (14 days, 0.1 mmol/kg body wt.) of both mixtures, but not DeBDE, resulted in a long-lasting induction of uridine diphosphate glucuronyl transferase (UDPGT) activity in rats (Carlson, 1980a).
Short-term feeding studies using PeBDE in rats led to changes in the liv- er (increased weight, hepatocytomegaly) and thyroid (hyperplasia) (Norris et al., 1975b; El Dareer et al., 1987; Great Lakes Chem. Corp, undated a). No carcinogenicity studies have been performed for TeBDE or PeBDE.
The Bromkal 70-5DE product, containing mainly Te- and PeBDE, caus- es decreased thymus weight, increased liver/body weight ratios in mice and decreases in the thyroid hormone thyroxine in rats and mice (Fowles et al., 1994; Darnerud and Sinjari, 1996). Decreases in thyroxine were also seen when rats and mice were treated with the single congener BDE-47, but no effects on thyroid stimulating hormone were seen for BDE-47 or Bromkal 70 (Darnerud and Sinjari, 1996).
In subsequent studies the interactive effects of different organohalogen compounds (PCB, PBDE and CP) on thyroxine hormone levels and micro- somal enzyme activities were tested (Hallgren and Darnerud, 1998). Fe- male rats were orally exposed to single compounds or combinations daily during 14 days. The results show that PCBs (Aroclor 1254) and PBDEs (BDE-47) significantly reduce the T4 levels in rats, in the actual exposure interval (6–18 mg/kg body weight/day), and that Aroclor 1254 results in the strongest effect, when administering the substances orally in isomolar concentrations. EROD and MROD, but to a lesser extent PROD and UD- PGT, activites correlated to T4 effects, which could indicate that glucuro- nidation of T4 is not a major factor in explaining the observed decrease in T4 plasma levels. Regarding the mixed BDE-47 + CP group, a synergistic decrease in free T4 levels, and increase in EROD activity, was observed. As organisms are exposed to these environmental chemicals as mixtures, the observed interactive effects are of interest.
36
It is known that hydroxylated metabolites of PCB can compete with thy- roxine for the binding site on the thyroxine-carrying protein transthyretin in plasma (Brouwer et al., 1990). The effects on the thyroid gland (hyper- plasia, decreased thyroxine levels) seen with PBDE could also be due to ef- fects of hydroxylated metabolites, which have been found as metabolites in both mice, rats and fish. Some hydroxylated PBDEs have structural sim- ilarities to thyroxine. In preliminary experiments, metabolites formed in rat liver from BDE-47, -99 and -153 have been shown to be better competitors for the transthyretin binding site than the parent compounds, which indi- cates they could compete with thyroxine as well (Brouwer and Murk, per- sonal communication, cited in Örn, 1997).
Short-term feeding studies with high doses of DeBDE in rats led to liver lesions and thyroid hyperplasia. Long-term exposure to DeBDE was also found to induce thyroid hyperplasia, hepatocellular and thyroid adenomas and carcinomas in mice (Great Lakes Chem. Corp, undated b; 1987). Workers producing DeBDE and decabromobiphenyl had a statistically sig- nificant increase in hypothyroidism (Bahn et al., 1980).
Immunotoxicity was studied after oral treatment with Bromkal 70-5 DE or BDE-47 in rats and mice (Darnerud and Thuvander, 1998). In mice, BDE-47 caused reduced splenocyte number as reflected in decreased num- bers of CD45R+, CD4+ and CD8+ cells in spleens. In mice treated with Bromkal 70-5 DE, absolute numbers of double negative thymocytes were significantly lower than in controls and mice also showed reduced produc- tion of IgG. No effects were seen in rats. Thus, BDE-47 and Bromkal 70- 5 DE, which contains BDE-47, both seem to be immunotoxic in mice.
Studies have shown that there is a critical phase in neonatal mouse brain development when the brain is particularly susceptible to effects of low- dose exposure to toxic substances such as PCB, DDT, pyrethroids, organo- phosphates, paraquat and nicotine (Eriksson, 1997). This critical phase is known as the ”brain growth spurt” (BGS) and disruption leads to persistent disruption in adult brain function. The BGS occurs at different time points in different mammalian species (Davison and Dobbing, 1968). In rats and mice it occurs in the first 3-4 weeks of life (neonatal period) whereas in hu- mans it occurs during the third trimester of pregnancy and throughout the first 2 years.
To study possible neurotoxicity of brominated flame retardants, BDE- 47, BDE-99 or TBBPA were administered orally to neonatal mice on day
37
10 (Eriksson et al., 1998). Doses administered are given in Table 5. Several tests of behavior, locomotion, activity and memory were performed with the treated individuals several months later. Results showed that BDE-47 and -99 both induced permanent aberrations in spontaneous motor behav- ior which worsened with age (Table 5). Similar effects have been seen in mice exposed neonatally to similar molar doses of some ortho-substituted PCB and co-planar PCB (Eriksson et al., 1991; Eriksson and Fredriksson, 1996a; 1996b; 1998). Neonatal exposure to BDE-99 also affected learning and memory functions in the adult animal. No effects were seen for TBBPA.
Table 5. Neurotoxicological effects seen in mice several months after administration of a single dose of specific PBDE congeners and TBBPA on day 10 after birth. Results from Eriksson et al., 1998).
+ indicates permanent aberrations; – indicates no effects
In a follow-up study, Eriksson et al. (1999) investigated whether there is a critical time in neonatal mouse brain development for induction of the neurotoxic effects of BDE-99. One single oral dose of 8 mg/kg body weight (14 µmol/kg body weight) was administered to 3-day, 10-day and 19-day-old mice. Spontaneous behavioral tests were performed after 4 months. The mice exposed to BDE-99 on day 10 showed significant behav- ior aberration, as was previously seen, and mice exposed on day 3 showed similar aberration but to a lesser degree. The mice exposed on day 19 showed no significant change from the controls.
Uptake and retention of BDE-99 in the brain was also studied by admin- istering 14C-labelled BDE-99 to 3-day, 10-day and 19-day-old mice (Eriks- son et al., 1999). The amounts of radioactivity found in the brain were
Substance Dose in mg/kg body weight
Dose in µmol/kg body weight
Spontaneous motor behavior disfunction
Learning and memory disfunction
BDE-47 low 0.7 1.4 – –
BDE-47 high 10.5 21.1 + –
BDE-99 low 0.8 1.4 + –
BDE-99 high 12.0 21.1 + +
TBBPA low 0.75 1.4 – –
TBBPA high 11.5 21.1 – –
38
measured at 24 h and 7 days after administration. The retention of BDE-99 was similar to what has been observed after neonatal exposure to CB 52, CB 153 and DDT (Eriksson, 1997). The retention of BDE-99 in mice ex- posed on day 3 indicates that the effects seen may be due to the amount still present in the brain on day 10. The neurotoxic effects seem to involve changes in the cholinergic system as mice given BDE-99 on day 10 and then challenged as adults with a low dose of nicotine behave completely the opposite of controls. From these studies, it was concluded that the win- dow for permanent effects of BDE-99 and BDE-47 is day 10 in neonatal mice (Eriksson et al., 1998; 1999).
Female rats given BDE-47 orally for a period of two weeks were then killed and the choroid plexus of the brain removed, homogenized and in- cubated with 125I-T4 (Sinjari et al., 1998). Compared to controls, there was a dose-dependent reduction in the binding of 125I-T4 to the choroid plexus. In contrast, in vitro incubation of rat choroid plexus with BDE-47 revealed no competitive inhibition of labelled T4 binding. This indicates that BDE- 47 metabolites can cross the blood-brain barrier and bind to the choroid plexus T4-binding sites. This in turn could cause the interference of T4 transport to the brain, with risks for effects on neural development.
4.2.1.3 In fish Microinjection of Bromkal 70-5DE in rainbow trout larvae led to weakly induced EROD activity (Norrgren et al., 1993). Three-spined stickleback fed Bromkal 70-5DE showed weakly induced liver EROD activity, fatty livers and a reduction in spawning success (Holm et al., 1993). Microinjec- tion of 2,2',4,4'-TeBDE, 2,2',3,4,4'-PeBDE or 2,2',4,4',5-PeBDE into new- ly fertilized rainbow trout eggs in an early life stage mortality bioassay showed no effects compared to 2,3,7,8-tetrachlorinated dibenzo-p-dioxin (Hornung et al., 1996). However, 3,3',4,4'-tetrabrominated biphenyl and 3,3',4,4',5,5'-hexabrominated biphenyl were ten-fold more potent than the identically chlorinated biphenyls.
In a recent study, rainbow trout were fed food containing either 2,2',4,4'- TeBDE (BDE-47) or 2,2',4,4',5-PeBDE (BDE-99) for 6 and 22 days to study biological effects. Both congeners were found to significantly inhibit EROD activity in the liver with BDE-47 being most powerful in this effect (Tjärnlund et al., 1998). A 35% reduction of GSH-reductase was also ob- served after 22 days. Hematocrit and blood glucose also showed small but
39
statististically significant changes after 6 days. No differences were seen in condition factor, liver somatic index, spleen somatic index, numbers of leucocytes, thrombocytes, granulocytes, lymphocytes or hemoglobin lev- els. When injected into fertilized fish eggs, no effects were seen on thia- mine use.
4.2.2 TBBPA
4.2.2.1 IN VITRO
Studies of the potency of TBBPA for its competitive binding to human transthyretin in vitro have shown that TBBPA has the highest potency of all brominated and chlorinated substances tested so far. It is up to 25 times more potent than T4 (Brouwer, 1998). However, this effect was not seen in in vivo studies in pregnant rats, see below (Meerts et al., 1999).
TBBPA was tested for its activity in inducing intragenic recombination in two in vitro assays using mammalian cells but caused no effect (Helle- day et al., 1999).
4.2.2.2 In mammals TBBPA fed orally to mice and rats showed low or no effects on behavior, weight gain, mortality, organ abnormalities or hematology (WHO/IPCS, 1995).
The thyroid transport protein, transthyretin, may play a major role in de- livering T4 from the mother to the fetus across the placental barrier as well as across the blood-brain barrier, where it is converted to T3, an essential hormone for normal brain development (Calvo et al., 1990; Southwell et al., 1993). TBBPA has been shown to competitively bind to human tran- sthyretin in vitro with high affinity. To study the effects of TBBPA on thy- roid hormone transport, including across the blood-brain barrier, pregnant rats were given 14C-labelled TBBPA on days 10 to 16 of gestation (Meerts et al., 1999). Blood plasma thyroid hormone and thyroid stimulating hor- mone levels were measured as well as 125I-T4-competition binding on ma- ternal and fetal plasma transthyretin. No effects were seen on total T4, free T4 or total T3 levels in dams or fetuses. Thyroid stimulating hormone in- creased significantly in fetal plasma by 196% but no effect was seen in dams. No 14C label was detected on transthyretin and there was no shift in binding of 125I-T4, which would be expected if TBBPA had bound to trans-
40
thyretin. Therefore, TBBPA was concluded to not bind to transthyretin in vivo. No selective accumulation of TBBPA-related activity was found in the fetal brain.
4.2.2.3 In fish Bluegill sunfish exposed to TBBPA in water became irritated and exhibit- ed abnormal swimming behavior. Rainbow trout exhibited irritation, twitching, erratic swimming, dark discoloration and labored respiration. Fathead minnow showed reduced survival of young at hatch and reduced survival and growth after 30 days (WHO/IPCS, 1995). There is no infor- mation regarding uptake, metabolism or effects in vitro, in mammals or in fish of the methylated derivative of TBBPA (MeTA).
4.2.3 HBCD
4.2.3.1 IN VITRO
The effects of HBCD on intragenic recombination were studied in two in vitro assays using mammalian cells (Helleday et al. 1999). HBCD caused statistically significant increases in recombination frequency in both test systems, indicating that it may induce cancer via a non-mutagenic mecha- nism, similarly to other environmental contaminants such as DDT and PCB.
4.2.3.2 In mammals HBCD is not acutely toxic to rats or mice when given orally or in rats when inhaled. No acute dermal toxicity is seen in treated rabbits. Chronic expo- sure in rats leads to increased liver weights, and in one study, thyroid hy- perplasia. Chronic exposure in female rats led to inhibited oogenesis (Zel- ler and Kirsch, 1969). In chronic oral studies in pregnant rats, HBCD was found to suppress maternal food consumption and increase maternal liver weight in the highest dose group (1% of diet), but no effects were seen on the offspring (IUCLID, 1996). The quality of most of these studies is, how- ever, under question.
4.2.3.3 In fish No information is available on the toxicity of HBCD to fish.
41
5. Environmental concentrations A summary of environmental concentrations is given in Table 17 at the end of this section.
5.1 Abiotic samples Products In laboratory experiments to test if all TBBPA is actually bound, filings from a circuit board were extracted with solvent and analyzed and the re- sults showed that a small percentage of the TBBPA was not bound (Sell- ström and Jansson, 1995). Thus TBBPA can also leak from products into the environment.
Qualitative analyses have shown the presence of HBCD in styrofoam chips used as packing material (Amelie Kierkegaard and Ulla Sellström, unpublished results).
5.1.1 AIR
In 1979, DeBDE was identified in air particulate in the vicinity of plants manufacturing brominated flame retardants (Zweidinger et al., 1979). Wa- tanabe and his coworkers found predominantly DeBDE in airborne dust from the Osaka region, in Japan (Watanabe et al., 1995). In samples from Taiwan and Japan in the vicinity of metal recycling plants, various tri-, tet- ra-, penta- and hexaBDEs were detected in air (Watanabe et al., 1992). Concentrations ranged from 23–53 pg/m3 in Taiwan and 7.1–21 pg/m3 in Japan.
One explanation for the wide distribution of lower brominated PBDEs found in the Swedish environment could be long range transport in air. To test this, two high-volume (48-hour) air samples previously collected with- in the Swedish Dioxin Survey, one from Hoburgen in southern Sweden (Gotland) and one from Ammarnäs in northern Sweden, were analyzed. The sampling conditions for these two samples are given in Table 6.
The results showed quantifiable amounts of BDE-47, -99 and -100 and HBCD in both samples (Bergander et al., 1995). The total PBDE levels
42
were approximately 1 and 8 pg/m3. Highest levels of BDE-47 were found on polyurethane foam plugs (gas phase) while higher levels of BDE-99 and -100 and HBCD were found on filters (particulate phase) (Figure 6). This corresponds well with the physicochemical properties of these substances, as BDE-47 is more volatile than the other compounds, and would be ex- pected to be in the gas phase to a larger extent. No DeBDE was found, but the detection level for DeBDE is much higher than for the lower brominat- ed PBDEs. Because the samples were taken at different times of year and the trajectories of the air packets were not stable over the sampling period, it is not possible to compare the concentrations between the two sites. However, the fact that these substances were found in air samples from both sites supports the hypothesis that these substances are spread via long- range transport.
Support for this theory has also come from newer results. Air samples were collected at one rural site in southern England (Stoke Ferry) and one
Table 6. Sampling conditions for two high-volume air samples.
Site Volume sampled Temperature °C Sampling date
Ammarnäs 1475 m3 –4 January, 1991
Hoburgen 1736 m3 +15 July, 1990
0
1
2
3
4
5
6
BDE-47 BDE-99 BDE-100 HBCD
Figure 6. Approximate concentrations of PBDE and HBCD in air samples from two sites in Sweden (Bergander et al., 1995).
43
semi-rural site in northwestern England (Hazelrigg) during 1997 (spring, summer, autumn and winter) and analyzed for PBDEs (Peters et al., 1999a; b). Detectable concentrations of tri- to heptaBDEs were found and the sum concentrations of BDE-47, -99 and -100 were 7–28 pg/m3 at Hazelrigg and 11–67 pg/m3 at Stoke Ferry (A. Kierkegaard, Stockholm University, per- sonal communication). PBDE has also been measured in several archived air samples from the Arctic (Alert, Northwest Territories, Canada and Dunai Island, eastern Siberia) taken between January 1994 and January 1995 (Alaee et al., 1999 and M. Alaee, Dept. of Environment, NWRI, Can- ada, personal communication). The sum concentrations of several di- to hexaBDEs were 1–4 pg/m3 at Alert most of the year, but 28 pg/m3 in July, 1994. Sum concentrations in air samples from Dunai, were somewhat low- er than at Alert, with the highest level also found in summer (7–8 pg/m3). BDE-47 and -99 were the major congeners found.
Bergman et al. (1997) developed a sampling technique for sampling air particulates in the working environment. They determined a number of brominated flame retardants on air particulates in rooms containing com- puters and other electrical equipment. All particulate samples contained TBBPA, BDE-47 and BDE-99. The presence of these substances in air par- ticulates shows that these substances are leaking into the indoor environ- ment from electronic devices and therefore exposing humans.
In another study, air was sampled in an electronics dismantling plant, in an office with computers and outdoors (Bergman et al., 1999; Sjödin et al., 1999b) and analyzed for several PBDE as well as TBBPA. Highest concen- trations of all substances were found in air from the electronics dismantling plant. Concentrations in the office were not detectable or 400-4000 times lower and not detectable concentrations were found for the outdoor air. Mean concentrations found in air at the dismantling plant were 2.5 pmol/ m3 (1250 pg/m3) for BDE-47, 4.6 pmol/m3 (2600 pg/m3) for BDE-99, 6.1 pmol/m3 for BDE-153, 26 pmol/m3 for BDE-183, 38 pmol/m3 (36500 pg/ m3) for BDE-209 and 55 pmol/m3 (29900 pg/m3) for TBBPA (Bergman et al., 1999). Samples were also taken near a plastics shredder at the disman- tling plant to identify possible point sources (Sjödin et al., 1999b). Concen- trations of the PBDEs were found to be 4–10 times higher in proximity of the shredder when compared to the air samples at other sites in the dismant- ling plant.
44
5.1.2 WATER
Plastic parts and textiles are flame-retarded in cars. Scrapped cars are taken apart at car dismantling centers, so that parts that can be recycled are re- moved. However, some materials, possibly containing PBDEs, cannot be reused and are placed in landfills. A car parts removal center in the city of Halmstad has a landfill where leach water is collected in a lagoon for sed- imentation. The water eventually is led to the sewage treatment plant in Halmstad (see sewage sludge results below).
In order to see if PBDEs, especially DeBDE, are leaching from the land- fill, samples of incoming leach water to the lagoon as well as lagoon sludge were collected and analyzed for PBDEs. This was the first attempt to ana- lyze a water sample for these substances, which required some methods de- velopment at the extraction stage. The results are preliminary and lowest possible concentrations (Amelie Kierkegaard, Stockholm University, per- sonal communication; de Wit, 1995; 1997).
The results showed low PBDE concentrations in both the water and sludge samples. The leach water contained 0.3 ng/liter (sum of BDEs -47, -99 and -100) and the sludge, 2 ng/g dry weight (sum of BDEs -47, -99 and -100). Neither sample contained octa-, nona- or DeBDE. Therefore, this particular industry does not seem to be the source of the high DeBDE con- centrations in the sewage sludge from Halmstad.
5.1.3 SEWAGE SLUDGE
Sewage sludge samples collected in 1988 from Ryaverket sewage treat- ment plant in Gothenburg, Sweden, were analyzed for PBDEs. Concentra- tions of BDE-47, -99 and -100 together were around 20–30 ng/g dry weight (Nylund et al., 1992) with BDE-47 and -99 present at similar levels. The proportions seen in sewage sludge are similar to those seen in the technical PeBDE product Bromkal 70-5DE. The levels were similar to those found in Germany by Hagenmeier et al. (1992), who found tri- to heptaBDEs in sewage sludge, with concentrations of Te- and PeBDEs (not designated which congeners) of 0.4–15 ng/g.
Several digested sewage sludge samples were collected in Sweden in 1991 within a project administrated by the Swedish EPA. The samples were taken from the sewage treatment plant at Halmstad. The purpose of the project was to analyze sewage sludge for a number of organic environ-
45
mental toxins including PBDEs and to study what happens to these sub- stances when the sludge is plowed into soil as fertilizer.
One of these sludge samples was analyzed for PBDEs and results showed that the sum of concentrations of BDEs -47, -99 and -100 was ap- proximately at the ppb level (ng/g), similar to the levels found in a sample analyzed previously from Ryaverket (Nylund et al., 1992). However, the concentration of DeBDE was 1000 times higher than for other PBDEs, at the ppm level (µg/g). This sample also contained high concentrations of nona- and octaBDEs (A. Kierkegaard, personal communication; de Wit, 1995).
In 1988, a sewage sludge sample was collected from a treatment plant receiving leach water from a landfill where wastes from a plastics industry are placed (Klippan). This plastics industry uses TBBPA. For comparison, a second sewage sludge sample was collected from a treatment plant (Rim- bo) having no known sources of TBBPA connected to it (Sellström and Jansson, 1995). The samples were analyzed for TBBPA and MeTA and re- sults showed that MeTA was not detectable, but TBBPA levels were 56 and 31 ng/g dry weight. The samples were also analyzed for BDE-47, -99 and -100. The sum of these three PBDE congeners in the Klippan sample was 45 ng/g dry weight and for the Rimbo sample, 119 ng/g dry weight (Sellström, 1999).
Sewage sludge samples from sewage treatment plants in Stockholm, Sweden, have recently been analyzed for TeBDE, PeBDE, DeBDE, HBCD and TBBPA (Sellström et al., 1999; Sellström, 1999). The results are given in Table 7 below. The concentrations of BDE-47, -99, -100 and -209 do not differ as much between plants as do TBBPA and HBCD.
Table 7. Mean concentrations of several PBDE (n=4), HBCD (n=4) and TBBPA (n=2) in sewage sludge from three treatment plants in Stockholm in ng/g dry weight.
5.1.4 SEDIMENT Previous studies in Japan have found TeBDE, PeBDE, HxBDE and DeBDE in river sediments (Watanabe et al., 1986; 1987; 1995). Concen-
Treatment plant BDE-47 BDE-99 BDE-100 BDE-209 HBCD TBBPA
Sickla 78 98 24 220 21 3.6
Bromma 80 100 25 270 54 8.6
Loudden 36 56 13 170 19 45
46
trations of TeBDE and PeBDE together were 21–59 ng/g dry weight. DeBDE was found in concentrations ranging from <25 ng/g to 11600 ng/g dry weight (Environmental Agency Japan, 1983; 1989; 1991).
The upper layer of a sediment core collected in the southern Baltic Sea (Bornholm Deep) was analyzed for BDEs -47, -99 and -100 and contained 0.52 ng/g dry weight (2.9 ng/g ignition loss) (sum of three congeners)(Ny- lund et al., 1992). In a more recent study, twenty surficial sediment sam- ples taken from numerous sites in the Baltic Sea during 1993 were ana- lyzed for BDEs -47, -99 and -100 within a Helsinki Commission (HEL- COM) sediment baseline study (Jonsson and Kankaanpää, 1999). Results showed low levels in all samples, from not detected to 5.4 ng/g ignition loss based on the sum of three congeners (Table 8).
Table 8. Concentrations of 2,2'4,4'-TeBDE (BDE-47), 2,2',4,4',5-PeBDE (BDE-99) and 2,2',4,4',6-PeBDE (BDE-100) in sediment samples from several sites in the Baltic Sea. Concentrations are given in ng/g ignition loss (ng/g IG) and all samples were analyzed in duplicate. < means not detected at this level.
Sample site Ignition loss % BDE-47 BDE-99 BDE-100
Bothnian Sea 8 <1.6 1.2 *
Gulf of Finland 10 1.4–1.9 1.1–1.9 0.36–0.48
Gulf of Finland 13 <1.4 <1.2 <0.33
Gulf of Finland 6 <1.9 <1.6 <0.43
Gulf of Finland 10 <1.7 <1.4 <0.38
Gulf of Finland 16 1.4–1.9 1.1–1.2 0.32
Bothnian Bay 12 <1.6 <1.1 *
Åland Sea 9 1.6–1.7 1.2–2.4 1.3
Northern Baltic Proper 21 <1.1 <1.0 <0.28
Bothnian Bay 9 <1.8 <1.2 *
Kieler Bight 13 3.2–3.4 0.63–0.72 0.48–0.70
Lithuanian Coast 13 1.8–2.3 1.7–2.2 0.37–0.53
Gotland Deep 8 <1.4 <1.3 <0.33
Gdansk Bay 15 1.6 1.3 0.31
Härnösand 8 <2.0 <1.4 *
Kattegatt 5 <1.4 0.55-0.76 *
Riga Bight 3