Brominated Flame Retardants (BFRs) - gov.scot · A Review of Brominated Flame Retardants (BRFS) in...

Not to be quoted without prior reference to the authors © Crown Copyright 2006

Fisheries Research Services Internal Report No 06/06

A REVIEW OF BROMINATED FLAME RETARDANTS (BFRS) IN THE AQUATIC ENVIRONMENT AND THE DEVELOPMENT OF AN ANALYTICAL TECHNIQUE FOR THEIR ANALYSIS IN ENVIRONMENTAL SAMPLES

Lynda Webster, Marie Russell, Pam Walsham and Colin F Moffat March 2006

Fisheries Research Services Marine Laboratory Victoria Road Aberdeen AB11 9DB

1

A REVIEW OF BROMINATED FLAME RETARDANTS (BFRS) IN THE AQUATIC ENVIRONMENT AND THE DEVELOPMENT

OF AN ANALYTICAL TECHNIQUE FOR THEIR ANALYSIS IN ENVIRONMENTAL SAMPLES

Lynda Webster, Marie Russell, Pam Walsham and Colin F Moffat

FRS Marine Laboratory, 375 Victoria Road,

Aberdeen, AB11 9DB

SUMMARY 1. Due to their persistence, potential for long-range atmospheric transport, high

bioaccumulation and toxicity brominated flame retardants (BFRs) such as polybrominated diphenyl ethers (PBDEs), hexabromocyclododecane (HBCD) and tetrabromobisphenol-A (TBBP-A) are listed on the Oslo and Paris Commission (OSPAR) List of Chemicals for Priority Action.

2. BFRs have been found in the OSPAR Convention area in sediment and fish and in

areas remote from known sources, indicating they are widespread environmental contaminants.

3. Although not produced in Scotland BFRs have been produced within the UK and

have been detected in the UK marine environment. Release from products containing BFRs will be a diffuse source of these compounds to the environment. However, there is currently little information on PBDEs, TBBP-A and HBCD in the UK marine environment, particularly in Scotland, or on their biological effects.

4. As methods for PBDEs are now well established and fully validated, the International

Council for the Exploration of the Sea (ICES) has recommended that the analysis of PBDEs be considered for inclusion in the OSPAR Joint Assessment and Monitoring Programme (JAMP). PBDE data may now be submitted as part of UK National Marine Monitoring Programme (NMMP) on a voluntary basis.

5. An analytical method for the analysis of PBDEs was developed for both sediment

and biota at FRS ML. This method involves the accelerated solvent extraction (ASE) of PBDEs from biota or freeze dried sediment using iso-hexane. Addition of 5% deactivated alumina to the ASE extraction tubes allowed the extraction and clean-up steps to be combined. The extract was then concentrated using a Turbovap prior to analysis by gas chromatography-mass spectrometry (GC-MS) using negative chemical ionisation (NCI). An external standard method was used for quantification.

6. Eighteen PBDE congeners were analysed using GC-NCIMS, namely BDE17, 28, 75,

49, 71, 47, 66, 77, 100, 119, 99, 85/154, 153, 138, 183, 190, 209. Chlorobiphenyl (CB) 198 was assessed as a possible recovery standard to be added to samples prior to extraction.

A Review of Brominated Flame Retardants (BRFS) in the Aquatic Environment

2

7. This method was fully validated for the analysis of PBDEs in biota and sediment. Replicate analysis of the high standard on separate days by GC-NCIMS gave CV% for the tri- to hepta-BDEs of <10% and principally <20% for the low standard. Recoveries for the tri- to hepta-BDEs were generally >75% and CV% <15% for both sediment and biota samples. CB198 gave similar recoveries and precision and therefore was suitable for use as a recovery standard. BDE209 was more problematic with reproducibility being poor (CV% ~ 30%), although similar to other published studies.

8. Limits of detection (LoDs) for the tri- to hepta-BDEs in biota were between 0.05 and

0.07 µg kg-1 wet weight and 0.75 µg kg-1 wet weight for BDE209. For sediment the LoDs were 0.03 µg kg-1 dry weight for all congeners except for BDE99, 85/154 and 209 which have LoDs of 0.09, 0.10 and 1.56 µg kg-1 dry weight, respectively.

9. Analysis of HBCD and TBBP-A gave high detection limits and results were highly

variable. GC-NCIMS could not be used for the quantitative analysis of these compounds. Investigation of liquid chromatography-mass spectrometry (LC-MS) for the analysis of HBCD and TBBP-A is recommended.

INTRODUCTION Hazardous substances included on the Oslo and Paris Commission (OSPAR) List of Chemicals for Priority Action and the Water Framework Directive (WFD) List of priority substances were recently reviewed1. The aim of this review was to identify what data currently exists on hazardous substances in the Scottish marine environment and what data will be required to ensure Scotland meets its international obligations. One of the key recommendations of this report was that: FRS should establish the analytical capability to determine PBDEs (polybrominated diphenyl ethers) in sediment and biota and establish a limited monitoring programme to assess the current situation. In addition, the effects of these compounds on the marine biota should be assessed. This will require the development of suitable techniques1. As part of the 2004-2005 Schedule of Service for the Water Division of the Environment Group, analytical methods for the analysis of brominated flame retardants (BFRs), including polybrominated diphenyl ethers (PBDEs), in biota and sediment were developed with the aim of undertaking a monitoring programme to establish concentrations in the Scottish marine environment. This report updates the hazardous substances review with regards to BFRs and describes the analytical method developed at FRS for the analysis of these compounds in environmental samples. The biological effects, and the application of methods for the assessment of the effects of PBDEs on marine organisms, will be described in a separate report. Production and Use of BFRs Polybrominated diphenyl ethers (PBDE) were one of the most widely used groups of brominated flame retardants (BFRs). Other BFRs include polybrominated biphenyls (PBBs), hexabromocyclododecane (HBCD) and tetrabromobisphenol-A (TBBP-A). BFRs reduce fire hazards by interfering with the combustion of polymeric materials and can be classed as additive or reactive materials2, 3. PBBs were mainly produced in the USA, with only deca-PBB being produced in Europe and, since 1977, only in France2. The production of PBBs in Europe finally ceased in September 2000. PBBs were replaced with PBDEs, HBCD and


3

TBBP-A. PBBs are not included on the OSPAR List of Chemicals for Priority Action and, therefore, are not considered further. Reactive BFRs, such as TBBP-A, are chemically bonded into plastics3. PBDEs and HBCD are used as additives to polymers and resins and are thought to be more easily released to the environment compared to reactive BFRs. Commercial PBDE mixtures are classified according to the degree of bromination4. The penta mixture contains mainly tetra- and penta-BDEs, the octa mixture mainly hexa- to hepta-BDEs and the deca mixture containing mainly deca-BDE. The penta BDE mixture is mainly used in furniture and upholstery, the octa mixture in plastics and the deca mixture in textiles2. HBCD is used in polymers such as polystyrene and in textiles and TBBP-A in electrical equipment2,3. On 30 January 2001 the European Commission issued a proposal to ban the penta technical mixture. This ban was finally put in place on 15 August 2004, restricting the use of the penta and the octa technical mixtures to a limit of 0.1% by mass for all articles placed in the market according to the European Directive 2003/11/EC¹, 24th amendment of 76/769/EEC. Furthermore, the Restriction of Hazardous Substances in Electrical and Electronic Equipment (RoHS) Directive (2002/95/EC) will restrict the marketing of electrical and electronic equipment containing hazardous substances, which includes PBDEs, and should be effective by 1 July 2006. Although penta and octa BDE technical mixtures have been phased out in Europe, production and use continues in other parts of the world such as North America. BFRs can be released to the environment during their production and while manufacturing other products, and during disposal of products containing theses chemicals. In addition, BFRs may continue to leak out of treated material. As the usage of HBCD and TBBP-A began more recently and they continue to be produced it is possible that concentrations in the environment will rise. TBBP-A has not been produced in the UK but is imported. Penta and octa formulations of PBDEs, and HBCD, were manufactured by the Great Lakes Chemical Company at Newton Aycliffe, County Durham in the north east of England. Manufacture of PBDEs at this plant ceased in 1996 and for all BFRs on 23 December 2003. PBDEs and HBCD have the potential for long-range atmospheric transport. In addition release from products containing BFRs will constitute a diffuse source of these compounds to the environment. Due to the high usage of BFRs and their ability to undergo long-range atmospheric transport it is thought likely that these compounds will be present in the Scottish marine environment. At present PBDEs are the most commonly studied BFRs while data on HBCD and TBBP-A is scarce. However, in the Scottish marine environment there is currently little information on any of the BFRs (PBDEs, HBCD or TBBP-A). Water Framework Directive (WFD) PBDEs are on the Water Framework Directive (WFD) Priority List that came into force on 16 December 2001. In addition, penta-BDE is classed as a priority hazardous substance (PHS) and octa-BDE is classed as a priority substance (PS)5. Under the WFD, member states must implement measures to progressively reduce emissions and discharges of PS. The objective for PHS is for the cessation of emissions, discharges and losses within 20 years of adoption of EC proposals for control measures. Persistence, Bioaccumulation and Toxicity of BFRs The inclusion of PBDEs, HBCD and TBBP-A on the OSPAR List of Chemicals for Priority Action2,3 is because of their persistence, potential to bioaccumulate and toxicity as assessed using the OSPAR Dynamic Selection and Prioritisation Mechanism for Hazardous Substances (DYNAMEC)1,4. This procedure has been used since 2000 to identify priority substances for addition to the OSPAR List of Chemicals for Priority Action. Substances are


4

examined against a set of cut-off values for persistence, tendency to bioaccumulate and toxicity, known as the PBT criteria (Table 1). Chemicals are further prioritised for addition to the list on the basis of production volumes and occurrence in the marine environment. In this context PBDEs have recently been included on the OSPAR Co-ordinated Environmental Monitoring Programme (CEMP) on a voluntary basis. This will result in a limited amount of data being gathered for PBDEs across the maritime area and will contribute to assessments of concentrations of these synthetic chemicals in the marine environment. As the ultimate aim of the OSPAR Strategy for Hazardous Substances is to achieve concentrations in the marine environment of near background for naturally occurring [hazardous] substances and close to zero for man-made, synthetic chemicals by 2020, gathering data at an early stage is critical3. In addition, information on effects will be required within the UK. PBDE data may be submitted to the UK National Marine Monitoring Programme (NMMP). There are 209 possible PBDE congeners, depending on the number and positions of the bromine atoms on the two phenyl rings. The PBDE congeners are numbered according to the International Union of Pure and Applied Chemistry (IUAPAC) nomenclature used for chlorobiphenyls (CBs). The full chemical names for the most commonly monitored PBDEs are shown in Table 2. Due to the similarity in structure between PBDEs and CBs, PBDEs are expected to persist in the marine environment and exhibit similar toxic properties2, 6 - 8. The persistence (P) of a substance can be measured as its half-life (T1/2) and, if greater than 50 days in water that substance is considered persistent. A half-life of 600 days has been reported for penta BDE mixture in aerobic sediment (Table 1). PBDEs are highly lipophilic and are more hydrophobic than the corresponding chlorinated compounds. Liability to bioaccumulate (B) can be predicted from the octanol water partition coefficient (Log Kow), or from the bioconcentration factor (BCF). The DYNAMEC criteria state that a compound with a Log Kow of greater than 4 and BCF greater than 2000 l kg-1 has a high potential to bioaccumulate. PBDEs have high (Log Kow > 4) octanol water coefficients ranging from 5.03 for di-BDE to 10.33 for deca-BDE (Table 1 and 2)2, 4. Lower brominated PBDE congeners bioaccumulate to a greater extent than the more highly brominated PBDEs2, 6. This is because of the low uptake rate of the highly brominated PBDEs due to their large size. A bioconcentration factor (BCF) of 27400 l kg-1 has been reported for penta BDE mixture. Available data for deca-BDE suggests this congener does not bioaccumulate to the same extent as the lower brominated PBDEs due to its high molecular mass. Acute toxicity is measured in the laboratory as lethal or sublethal toxicity resulting from intermittent or continuous exposure of a substance for a period of time shorter than the life cycle of the organism. This can be measured as the Lethal Concentration (LC50), the concentration of a chemical in air that kills 50% of the test animals in a given time (usually four hours), using DYNAMEC procedures. Values of ≤ 1 mg l-1 should give rise to concern. Long-term chronic effects are defined as sublethal toxicity resulting from intermittent or continuous exposure of a substance for a period during a substantial proportion of the life cycle. This can be measured as the no-observable effect concentration (NOEC) and the cut off value, using the DYMANEC procedure, is ≤ 0.1 mg l-1. Currently there is a limited amount of chronic and acute toxicity data for brominated flame retardants. The toxicity of PBDEs varies by type due to their differing chemical structures7-10. An NOEC value of 0.0089 mg l-1

was obtained for pentaBDE in fish (Table 1)2. Predicted NOEC and LC50 values for decaBDE are in excess of the aqueous solubility limit of this compound. There is little information on the health effects of PBDEs in humans although animal studies have shown that PBDEs affect thyroid hormone functions and can impair the developing central nervous system and brain7-10. The molecular structure of TBBP-A is similar to that of the thyroid hormone, thyroxines, except the iodine atoms have been replaced with bromines3. TBBP-A is expected to be


5

present in an ionised form in the marine environment. TBBP-A has a high octanol water coefficient (Log Kow = 4.5)4 and is therefore likely to bioccumulate. However, the highest measured bioconcentration factor (BCF) is 1,234 l kg-1 and is below the OSPAR cut-off point for bioaccumulation of 2,000 l kg-1. Data for the acute toxicity in fish of TBBP-A has been reported (LC50 = 0.54 mg l-1) and chronic toxicity has been measured with an NOEC value of 0.16 mg l-1. The LC50 value for TBBP-A meets with the DYNAMEC toxicity criterion (≤ 1 mg l 1). Theoretically there are sixteen possible stereoisomers of HBCD; 6 enantiomeric pairs and 4 meso forms11. However, in technical HBCD mixtures only three of the 6 enatiomeric pairs are found, namely α-, β- and γ-HBCD, with the dominant isomer being γ-HBCD. In sediment the γ- isomer also dominates but in biota the major isomer is α-HBCD, β-HBCD is always a minor component. HBCD has a high octanol water coefficient (Log Kow = 5.8) and, the potential to bioaccumulate4. A BCF of 18,100 l kg-1 has been estimated4. It is thought to be persistent; however there is very little information available on this compound. The properties of persistence in the environment, its tendency to bioaccumulate and its toxicity to aquatic and terrestrial organisms have not been fully characterised and the risks to human health have not been fully evaluated2. Environmental Concentrations Law et al. recently reviewed levels and trends of BFRs in the European environment12. Most environmental studies concentrate on the PBDEs and many of these do not include BDE209. There is limited data on concentrations of HBCD and TBBP-A in the marine environment. Table 3 summarises published data on PBDE concentrations in a range of matrices. Polybrominatediphenyl ethers (PBDEs) Penta and octa formulations were manufactured by the Great Lakes Chemical Company at Newton Aycliffe, County Durham in the north east of England up until 1996. A survey of sediment and fish was undertaken from UK rivers and estuaries, including sites close to this plant. The river Tees was found to be heavily contaminated with PBDEs downstream of this plant. Concentrations in sediment ranged from < 0.3 to 368 µg kg-1 dry weight and < 0.6 - 898 µg kg-1 dry weight for BDE47 and BDE99, respectively13. In fish (plaice, flounder, dab) concentrations for BDE47 ranged from <0.3 to 9,500 µg kg-1 lipid weight and for BDE99 <0.6 – 370 µg kg-1 dry weight. Concentrations of BDE47 are normally higher than BDE99 in biota. Recent studies have illustrated that various fish species are capable of metabolising some BDE congeners14. This is thought to involve oxidative and de-bromination steps. De-bromination of BDE99 is thought to result in the formation of BDE47 which would explain the predominance of this congener in biota. DecaBDE (BDE209) was also found in sediment samples in this area, although not produced at this plant, with the highest concentration being 399 µg kg-1 dry weight. BDE209 was not found in biota (fish and shellfish) in this study, suggesting that this congener does not bioaccumulate. PBDE concentrations were also measured in marine top predators from this area15. The highest total concentration (Σ14 congeners) was 6,900 µg kg-1 wet weight (7,667 µg kg-1 lipid weight) found in porpoise blubber. Total concentrations in cormorant livers ranged from 1.8 to 140 µg kg-1 wet weight. Atmospheric transportation is a major pathway for PBDEs into the marine environment. PBDEs have been found to concentrate in the Arctic and bioaccumulate in native animals and humans. They have been found at high concentrations in fish, crabs, Arctic ringed seals and other marine mammals in northern Canada16, and have also been found in fish and mussels from Greenland17. Ikonmou et al. reported concentrations for PBDEs (Σ13 congeners) of between 350 and 2,300 µg kg-1 lipid weight in porpoise blubber from Northern


6

Canada and between 22 µg kg-1 lipid weight and 340 µg kg-1 lipid weight in sole16. Christensen et al. reported PBDEs in mussels and marine fish from Greenland17. The highest BDE concentration (Σ4 congeners) in mussels was 0.11 µg kg-1 wet weight and in marine fish liver 12.0 µg kg-1 wet weight17. PBDEs have been found to contaminate the blubber of sperm whales and long-finned pilot whales in the remote deep waters of the Atlantic18, 19. PBDE concentrations (Σ4 congeners) in a range of marine mammal blubber (sperm whale, minke whale, dolphin, harbour seal) from strandings on the Dutch coast ranged from 78.5 to 7,700 µg kg-1 wet weight. Highest concentrations were in the dolphin and harbour seal (>1,000 µg kg-1 wet weight) while concentrations in sperm whale were around 100 µg kg-1 wet weight18. Nineteen tetra- to hexa-BDEs were measured in the blubber of pilot whale caught off the coast of the Faroe Islands19. Concentrations ranged from 843.2 to 3,260 µg kg-1 lipid weight. In both studies the dominant congener was BDE47. PBDEs (7 congeners) have recently been reported in the penguin food web in Antarctica, confirming their global transport and distribution20. Mean total PBDE concentrations were 5.60, 5.81, 4.57 and 3.07 µg kg-1 lipid weight in krill, rockcod muscle, rockcod homogenate and penguin eggs, respectively. BDE47 accounted for >70% of the total PBDE concentration. PBDEs (BDE28, 47, 66, 85, 99, 100, 138, 153, 154, 184) were measured in the blubber of twelve species of marine mammals stranded in the UK21. The highest total BDE concentration was found in killer whale (16,200 µg kg-1 wet weight) and the lowest in a hooded seal (5 µg kg-1 wet weight). The highest concentration in a bottlenose bottle was 11.6 mg kg-1 wet weight, in a female found off Findochty, Grampian in 1999. The major congeners were BDE47, 99 and 100. BDE183, found in the octa-mix, was not detected in any samples. PBDEs have also been analysed in grey seal pups from the Farne Islands, off north east coast of England22. Total concentrations (BDE17, 28, 35, 37, 47, 49, 71, 85, 99, 100, 120, 153, 154) ranged from 61 to 903 µg kg-1 lipid weight. A wide range of PBDEs (sum of BDE47, 99, 100, 154, 153) has also been found in harbour porpoise sampled from Scottish waters between 2001 and 2003, ranging from 130.6 to 2,213 µg kg-1 lipid weight (mean = 1,029 µg kg-1 lipid weight)23. Concentrations in harbour porpoise from other areas for the same study (Ireland, Holland, France and Spain) were lower with mean concentrations < 1,000 µg kg-1 lipid weight23. PBDEs were found in human breast milk in Sweden24. Recent Swedish research has also found that PBDEs are present in the blood of office workers who use computers, in hospital cleaners and workers at an electronics-dismantling plant. The highest levels were in the latter, demonstrating the role of electrical goods in the contamination25. PBDEs were found in pike from Swedish waters collected between 1960 and 2000, with total PBDE (BDE47, 99, 100, 153, 154) concentrations ranging from 0.06 to 1.60 µg kg-1 wet weight26. Concentrations increased from the 1960s up to the 1980s, followed by a decline. The highest concentration for BDE47 was 0.76 µg kg-1 wet weight in a pike sample collected during 1983. The highest total BDE concentration (1.60 µg kg-1 wet weight) was in a pike sample collected in 1989. Fourteen PBDE congeners were measured in Dungeness crab from various marine sites on the west coast of Canada between 1992 and 200227. Total PBDE concentrations ranged from 8.6 to 660 µg kg-1 lipid weight with highest concentrations being found in crabs from harbours and areas close to population centres. The dominant congener was BDE47. PBDEs were measured in mysid shrimp and sediment from the Scheldt estuary in28. Fifteen PBDEs were measured, including BDE209. Total BDE concentrations ranged from 1,765 to 2,962 µg kg-1 lipid weight for mysid and from 14 to 22 µg kg-1 dry weight for sediment. BDE209 was detected in both the sediment and mysid although concentrations in shrimp were lowest for this congener, indicating the higher brominated congeners are less


7

bioaccumulative. Similar to other studies the penta mixture congeners BDE47, 99 and 100 were the dominant congeners. Thomas et al. investigated the uptake of BDE209 by grey seals by feeding them BDE209 in cod liver oil capsules29. Measurable concentrations of BDE209 were detected at the end of the 3 month study. By the end of the study between 11 and 15% of the cumulative amount of BDE209 ingested by the seal (~320 µg per seal) was stored in the blubber. The authors suggested that this congener can be stored in the tissue and may be more bioaccumulative than was previously thought. PBDEs were measured in the Netherlands in sediment, suspended particulate material (SPM) from a sewage treatment plant and effluents and biota30. SPM was identified as a carrier for the higher brominated PBDEs. High PBDE concentrations were found in sewage treatment plant influents and effluents. Median PBDE concentrations (BDE47, 99, 153 and 209) are shown in Table 3 for marine mussels and sediment collected as part of this study. In both cases BDE47 gave the highest concentration. A median BDE209 concentration of 22 µg kg-1 dry weight was found in sediment. However, BDE209 was not found in any of the fish or mussel samples, highlighting their limited ability to bioaccumulate. Sewage sludge samples from Sweden were analysed for PBDEs and concentrations were found to be variable, with total PBDE (8 congeners) concentrations ranging from Not Detected (ND) to 450 µg kg-1 wet weight 31. Concentrations were highest for penta and deca mix PBDE congeners, with maximum concentrations of 320 and 390 µg kg-1 wet weight for BDE99 and BDE209, respectively. Highest BDE209 concentrations were in areas where there was a possible contribution from the textile industry. PBDEs (40 congeners, mono- up to deca-BDEs) were measured in coastal sediments from Spain32. Total PBDE concentrations ranged from 2.7 – 134 µg kg-1 dry weight, with the most contaminated sediments being from the Barcelona area. BDE209 was the dominant congener (50-99% of the total), followed by BDE47, 99 and 153. PBDEs were measured in sediment from the Niagara River in western New York33. Total PBDE concentrations (sum of nine congeners) ranged from 0.72 to 148 µg kg-1 dry weight with the highest concentrations being found close to wastewater treatment plants. Typically the dominant congener was BDE47 followed by BDE99. PBDEs (BDE28, 47, 99, 100, 153, 154, and 183) were also measured in sediment and fish from the Danube Delta, Romania34. Concentrations in the sediment were below the limit of quantification (LoQ) of 0.1 µg kg-1 dry weight. However, they were detected in zooplankton (1–7.2 µg kg-1 dry weight), and in the muscle of various fish species (<0.1–14.3 µg kg-1 lipid weight), with the main congener being BDE47. Persistent organic pollutants (POPs), including chlorobiphenyls (CBs) and dioxins, are common contaminants in salmon feeds and higher concentrations of CBs and dioxins have been found in farmed salmon when compared to wild salmon35 - 37. Elevated concentrations of dioxins have been reported in Scottish farmed salmon35, 36, and of CBs in the sediment around the cages38. Easton et al. also reported PBDE concentrations in farmed salmon from Canada37. Similar to the CBs and dioxins, higher concentrations of PBDEs (sum of 41 congeners) were found in farmed salmon compared to wild salmon with concentrations ranging from 1.2-4.1 µg kg-1 wet weight and 0.04–0.5 µg kg-1 wet weight, respectively37. PBDEs were also found in salmon feed with a total concentration of 1.9 µg kg-1 wet weight37. Hites et al. undertook an assessment of PBDEs in farmed and wild salmon and also suggested that concentrations were higher in the farmed salmon39. In addition, PBDE concentrations in farmed salmon from Europe were higher than in farmed salmon from North America39. Total PBDE concentrations ranged from < 1 - ~5 µg kg-1 wet weight. One of the highest PBDE (sum of 43 congeners) concentrations was in farmed Scottish salmon (flesh)


8

(~ 4 µg kg-1 wet weight)38. In all cases (farmed and wild) the dominant congener was BDE47 (~50% of total), followed by BDE99 and 100. A small study was recently undertaken looking at PBDE (and HBCD) concentrations in fishery products available on the Irish market, including farmed and wild salmon40. Five farmed salmon fillets from different sites were analysed for PBDEs. These contaminants were found in all samples although for some congeners concentrations were below the limit of detection (LoD). The most abundant congener was BDE47. Total PBDE concentrations (16 congeners) ranged from 2.42 – 5.05 µg kg-1 wet weight. Concentrations in wild salmon were lower (0.71 – 1.01 µg kg-1 wet weight). Concentrations in mackerel and herring from Irish waters were also lower, with ranges of 0.93 – 1.86 µg kg-1 wet weight and 1.46 – 1.77 µg kg-1 wet weight, respectively. Zennegg et al. reported PBDE concentrations in farmed rainbow trout from Swiss lakes41. Total PBDE concentrations (BDE28, 47, 99, 100, 153, 154 and 183) ranged from 12 to 24 µg kg-1 lipid weight (0.74 – 1.3 µg kg-1 wet weight). PBDE concentrations were also determined in wild white fish from the same lakes as part of this study and concentrations ranged from 36 – 165 µg kg-1 lipid weight (1.6 – 7.4 µg kg-1 wet weight). In all cases BDE47 was the dominant congener and with the BDE47/99 ratio being highest in the farmed fish and lower in the wild whitefish. All lakes sampled during this study were close to populated areas and could not be considered as background sites. PBDEs (6 congeners) were measured in the muscle and liver of 4 species of fish (kalashapa, carp, nose-carp and wells) from a lake in Kahramanmaras, Turkey42. Concentrations ranges were not detected (ND) – 6.7 µg kg-1 wet weight and ND to 597 µg kg-1 lipid weight in the muscle and liver, respectively. BDE47 was the dominant congener in both muscle and liver. A Japanese survey of PBDEs in marine products (fish and shellfish) was recently undertaken43. Eleven congeners were measured although not all were detected. Total PBDE concentrations in fish ranged from ND to 0.55 µg kg-1 wet weight and from 0.01 to 0.12 µg kg-1 wet weight in shellfish43. A recent study of fish from European high mountain lakes looked at PBDEs in the muscle and liver tissue of brown trout. Highest concentrations were found in muscle and liver samples from Lochnagar in the north of Scotland with total PBDE concentrations of 1.2 µg kg-1 wet weight (177.0 µg kg-1 lipid weight) and 11.0 µg kg-1 wet weight (366.0 µg kg-1 lipid weight), respectively44. These concentrations were around 10 times higher than the other lakes in the study, including lakes from Greenland, the Alps, the Pyrenees and Norway. Such concentrations are, however, in the lower range of reported concentrations for biota from other environmental studies37, 39, 40. The major congeners present in fish from this high mountain lake study were BDE28, 47, 99, 100, 153 and 154 and as with other studies, BDE47 was the dominant congener in the majority of samples. The pattern of PDBE congeners in the trout from Lochnager was different to that in the fish from other lakes in the study and also to the technical mixture (Bromkal 70-5DE) analysed as a reference, with BDE99 being the dominant congener in the Lochnagar trout. Other persistent organic pollutants (e.g. PCBs) in fish from this loch did not show higher concentrations than the other mountain lakes therefore the phenomenon was specific for PBDEs. The author suggested that this distinct pattern may be due to contamination from a nearby source and could be indicative of high industrial emissions in the UK compared to other European countries44. However, there are no point sources close to Lochnagar. Furthermore, in other UK studies the typical BDE pattern, with BDE47 dominating, is observed. Hexabromocyclododecane (HBCD) and tetrabromobisphenol-A (TBBP-A) HBCD and TBBP-A were measured in the Scheldt estuary in mysid shrimp and sediment28. TBBP-A was only found in trace amounts in mysid and was below the detection limits in sediment. HBCD was detected in both sediment (14 – 71 µg kg-1 dry weight) and mysid


9

(562 - 727 µg kg-1 lipid weight). The environmental occurrence of HBCD was also investigated in Sweden45. HBCD was found in a range of matrices (water, soil, sediment, biota) with the highest concentrations found in pike (27 µg kg-1 lipid weight) and in sediment (25 µg kg-1 dry weight). Morris et al. investigated HBCD and TBBP-A in sediment and biota from North Sea estuaries using liquid chromatography- mass spectrometry (LC-MS)46. The highest HBCD and TBBP-A concentrations in sediment were from the River Skerne, County Durham, close to the Newton Aycliffe plant, with concentrations of 1.7 mg kg-1 dry weight and 9.8 mg kg-1 dry weight, respectively. The HBCD pattern was similar to the commercial formulations with the γ-HBCD being the dominant diastereoisomer. Concentrations were also measured in different trophic levels of the North Sea food web46. Most samples contained α-HBCD, typically the dominant congener in biota. Few samples contained TBBP-A; this is in agreement with the lower bioaccumulation of TBBP-A by fish. Concentrations for ΣHBCD ranged from 2.1 to 6.8 mg kg-1 lipid weight in liver and blubber of harbour porpoises and seals with some evidence of biomagnification through the food web. Zegers et al. investigated HBCD concentrations in harbour porpoises and dolphins from western European Seas47. The highest ΣHBCD concentration was in harbour porpoises stranded on the Irish and Scottish coasts of the Irish Seas (2.9 mg kg-1 lipid weight) and on the north west coast of Scotland (5.1 mg kg-1 lipid weight). Concentrations in other areas were < 1.5 mg kg-1 lipid weight. α-HBCD was the only diastereoisomer detected. HBCD was found in harbour porpoise sampled from Scottish waters between 2001 and 200323. Concentrations ranged from 393.3 to 9,592 µg kg-1 lipid weight (mean = 2,354 µg kg-1 lipid weight). HBCD was also measured in farmed salmon as part of the Irish study of BFRs 40. Seven fish fillets from seven different sites were analysed for HBCD in addition to PBDEs. HBCD was found in all samples. The mean upperbound concentration for HBCD was 1.17 µg kg-1 wet weight.

ANALYTICAL METHODOLOGY Precautionary Measures used at FRS ML Special precautions are required when analysing PBDEs, particularly if BDE209 is to be measured, due to their sensitivity to UV light. Incoming light was minimised in the laboratory by placing UV filters on the windows, and lights were kept turned off. All calibration and spiking standards were prepared and stored in amber glassware. The use of plastics was avoided as they can contain BFRs. In addition, BDE209 can adsorb to dust particles and so an ioniser was placed in the laboratory. Cleaning of Glassware, ASE tubes, filters and sodium sulphate BDE209 can stick to glassware (or any other chemically active sites). This can result in contamination of glassware. At FRS ML glassware was washed and dried in a GW 4000 glassware washer (Camlab Ltd., Cambridge, UK). Prior to use, all glassware was rinsed twice with acetone and then twice with iso-hexane, the latter being allowed to evaporate before proceeding. Anhydrous sodium sulphate was washed ultrasonically with dichloromethane (DCM) (2 x 500 ml) followed by iso-hexane (2 x 500 ml) and dried overnight at 150oC. Cellulose filters were washed ultrasonically with DCM (500 ml) followed by iso-hexane (250 ml) and dried overnight at 150oC All accelerated solvent extraction (ASE) cells, caps and collection bottles were solvent washed with acetone followed by iso-hexane, with the latter being allowed to evaporate before proceeding. The lids of the collection bottles were fitted with ultra low bleed septa which were first solvent washed with iso-hexane.


10

Lipid Determination The total lipid content was determined using the Smedes method48. The biota sample (0.5 – 10 g) was weighed into a centrifuge tube and iso-propanol (18 ml) and cyclohexane (20 ml) added. The sample was homogenised then de-ionised water (~13 – 22 ml, depending on the moisture content of the sample) added and the mixture homogenised again. Centrifugation was used to separate the organic extract from the particulate material. A second extraction was carried out with 13% iso-propanol (v/v) in cyclohexane. The two extracts were combined, concentrated by rotary evaporation and the resulting residue reconstituted in the solvent mixture, transferred to an evaporating dish and dried on a hotplate for at least 1 hour at 105oC ± 5oC. The residue was weighed and the lipid content calculated. Sample Preparation Biota An appropriate amount of tissue (equivalent to 300 mg lipid) was mixed with sodium sulphate (~20 g), spiked with recovery standard (CB198), and left overnight before being ground to a fine powder using a mortar and pestle. For spiking experiments, a BFR standard solution containing PBDEs, HBCD, TBBP-A and CB198 was added in place of the CB198 recovery standard. Solvent washed ASE cells (100 ml) were packed as follows: solvent washed filter paper, pre-washed sodium sulphate (10 g), 5% deactivated alumina (30 g), solvent washed filter paper and the biota/sodium sulphate mixture prepared as above. The cell was finally filled to the top with more sodium sulphate then packed down and topped up if required and another filter paper placed on top (Fig. 1). It was essential that cell was tightly packed, so as to reduce the dead volume. Sediment Freeze dried sediment (~20 g) was mixed with sodium sulphate (~30 g), spiked with the recovery standard (CB198) and left overnight before being homogenised using a mortar and pestle. For spiking experiments, a BFR standard solution containing PBDEs, HBCD, TBBP-A and CB198 was added in place of the CB198 recovery standard. Solvent washed ASE cells (100 ml) were packed as follows: solvent washed filter paper, 20 grams of granulated activated copper (30 mesh, Thames Restek, Buckinghamshire), pre-washed sodium sulphate (10 g), 5% deactivated alumina (10 g), solvent washed filter paper, sediment/sodium sulphate prepared as above. The cell was finally filled to the top with more sodium sulphate then packed down and topped up if required and a filter paper placed on top. It was essential that the cell was tightly packed, to reduce the dead volume. A few grains of the granulated activated copper were also put in the base of the ASE sample bottles. If this reacted with sulphur from the sample then a further clean up using additional activated copper was performed. Preparation of standard solutions Standards were obtained as single compound solutions at a concentration of 50 µg ml-1 from LGC Promochem, (Middlesex, UK). These were BDE17, 28, 75, 49, 71, 47, 66, 77, 100, 119, 99, 85, 154, 153, 138, 183, 190, 209, HBCD (α, β, γ isomers) and TBBP-A. Calibration standards were prepared by mixing each of the BDE congener solutions other than BDE209 (200 µl) and making up to 10 ml in iso-hexane to give a 1 µg ml-1 composite stock solution. For BDE209 and TBBP-A, 400 µl of the individual solution was added and for HBCD 600 µl (200 µl of each isomer) to give concentrations of 2 µg ml-1 and 3 µg ml-1,


11

respectively, in the composite stock solution. Dilutions were made to give seven calibration standards with nominal concentrations of 500, 250, 100, 10, 5, 1 and 0.2 ng ml-1. Concentrations in the calibration solutions were double for BDE209 and TBBP-A and three times higher for total HBCD. All calibration standards were analysed in triplicate by GC-MS (in NCI mode). CB198, at a concentration of 1 µg ml-1, was also included in the spiking solution, with the aim of using this as a recovery standard to be added to samples prior to extraction as a check on the method, to give nominal concentrations as for the PBDE calibration standards. Accelerated solvent extraction (ASE) Samples were extracted using an oven temperature of 60oC and a pressure of 1500 psi. Five minutes heating was followed by 2 x 5 min static cycles. The cell flush was 50% total cell volume (i.e. 25 % of the cell volume for each flush = 25ml per flush) with a 60 second purge (using nitrogen) at end of each sample extraction. The extraction solvent used was iso-hexane. Following ASE extraction, the extracts were transferred to Turbovap tubes and the volume reduced to ~0.5 ml at 20ºC, before transferring, with washings, to pre-weighed crimp top amber glass GC vials. The extracts were then concentrated further under a stream of nitrogen to approximately 0.5 ml, re-weighed and the weight recorded before analysis by gas chromatography-negative chemical ionisation mass spectrometry (GC-NCI-MS). For sediment samples, only the extract was passed through a pasteur pipette which was plugged with pre-cleaned cotton wool and three quarters filled with sodium sulphate. This was to catch any particles of copper that might be present in the extract. Analysis by gas chromatography-negative chemical ionisation mass spectroscopy (GC-NCIMS) The concentration and composition of the BFRs were determined by GC-NCIMS using an HP6890 Series gas chromatograph interfaced with an HP5973N MSD, fitted with a cool on-column injector. A Thames Restek STX-500 column (STX-500, 30 m x 0.25 mm i.d., 0.15 µm film thickness, Thames Restek, Buckinghamshire) was utilised, fitted with a Thames Restek Siltek (0.53 mm i.d) 5 m guard column. The injector temperature was initially 120oC. After 2 minutes it was elevated at 100oC per minute to 300oC where it was maintained to the end of the run. The carrier gas was helium, set at a constant pressure of 15 psi. Methane was used as the reagent gas at a pressure of 1.6 bar. The transfer line was held at 280oC and the ion source at 150oC. Injections were made at 120oC and the oven temperature held constant for 2 minutes. Thereafter the temperature was raised at 15oC min-1 up to 205oC. This was followed by a ramp of 6oC min-1 up to a final temperature of 330oC. The MS was set for selective ion monitoring (SIM) with a dwell time of 50 ms. Ions monitored were m/z 78.9 and 80.9 (ions equating to bromine) for all PBDEs, HBCD and TBBP-A and for BDE209 m/z 486.7 and 488.7 (pentabromo phenoxy ions formed from the cleavage of the ether bond) were also monitored. Appendix 1 contains the standard operating procedure (SOP) and method for the analysis of PBDEs in sediment and biota.

METHOD DEVELOPMENT AND VALIDATION Analytical Methods for PBDEs Analytical methods for the analysis of PBDEs in environmental samples are well established

and have been reviewed49, 50. The similarity in structure of the PBDEs to CBs means that techniques used for the analysis of CBs may also be applied to the analysis of PBDEs. An external quality assurance scheme was developed through QUASIMEME (Quality Assurance of Information for Marine Environmental Monitoring in Europe) for the analysis of


12

these compounds in environmental samples51. However, within laboratory CV% tend to be high (10 – 25%) for all PBDEs except the deca-BDE (BDE209) and between 17 and 78% for between laboratory CV%51. The analysis of the deca-BDE is problematic and many laboratories do not analyse BDE209. Between laboratory CV% for this congener ranged from 40 – 256%. BDE209 has a high molecular weight (959.2) and low vapour pressure51. It is prone to photolytic and thermal degradation and is reactive50. Detection limits are always poorer for BDE209 compared to other congeners (~factor of 10 higher). GC-MS Analysis GC-MS is the preferred technique for quantification of PBDEs in environmental samples, using either electron impact (EI) or negative chemical ionisation (NCI)49- 53. Although gas chromatography-high resolution mass spectrometry with electron impact ionisation (GC-HRMS) is the best method to unambiguously identify and quantify PBDEs in environmental standards, the expense and limited availability means that most laboratories have used GC-NCIMS. The majority of recent publications use this technique for the quantification of PBDEs in environmental samples. Ackerman et al. used a 30 m DB-5MS (Agilent Ltd.) column to separate 39 PBDE congeners53. In addition, a comparison was made to the detection of the BDE congeners using low resolution GC-MS in both EI and NCI modes. The lower brominated PBDEs (mono- and di-BDEs) showed better sensitivity in EI mode. However the higher brominated PBDEs (> 3 bromines) gave better sensitivity using NCI mode53. A MS method was developed at FRS ML for the analysis of PBDEs. Individual standard solutions were initially scanned to identify appropriate ions for quantification and confirmation. Chemical ionisation (CI) mode gave the best sensitivity giving approximately a 10 fold reduction in the limits of detection compared to electron impact mode (EI). Although one advantage of using EI is that 13C labelled internal standards may be used, this method was not pursued as sensitivity in EI was particularly poor for the higher brominated congeners. Negative impact chemical ionisation (NCI) showed improved sensitivity compared to positive impact chemical ionisation (PCI). Therefore, NCI was the method of choice. The base ions detected using NCI were the bromine ions (m/z = 79/81) for the tri- to hepta-BDEs and no other major fragment ions were identified from the scanning method that could be used as confirmatory ions. BDE congeners showed the typical 79Br (50.5%) and 81Br (49.5%) isotope distribution pattern. For BDE209 m/z 486.7 and 488.7 were the base ions, with m/z 486.7 being used as the quantification ion and m/z 488.7, 79 and 81 being confirmatory ions. Ions m/z 486.7 and 488.7 are formed by the cleavage of the ether bond to give the pentabromo phenoxy ion. One of the drawbacks of the CI mode is that isotopically labelled standards (13C) cannot be used as internal standards for quantification purposes when only the bromide ions are monitored. In addition larger fragment ions, necessary for confirmation, are only found for BDE209. An external standard method was therefore utilised for the analysis of BDEs. Similar to CBs, 209 BDE congeners are theoretically possible, however, only a small number of these are found in technical mixtures. Tri- through to hepta-BDEs and deca-BDE were selected for analysis; BDE17, 28, 75, 49, 71, 47, 66, 77, 100, 119, 99, 85, 154, 153, 138, 190, 209 (Table 2). This incorporated all the BDEs recommended for analysis at the QUASIMEME BFR54 workshop plus some other commonly monitored PBDEs known to be present in the technical mixtures. Similar to other published methods an STX-500 column (STX-500, 30 m x 0.25 mm i.d., 0.15 µm film thickness) was used for the analysis. Other columns commonly used include HT-8, DB1701 and DB144, 45, 47, 48, 50, 51. Using the temperature programme detailed in the Experimental Section baseline separation was achieved for nearly all PBDEs (Fig. 2). Only BDE85 and 154 did not give baseline separation. As such the areas for the two unresolved peaks were summed to give a


13

combined concentration for BDE85 and 154. BDE100 and 119 were separated at the baseline by only 6 secs but could be quantified separately by careful integration. The linear range of the method was 0.20 ng ml-1 to 500 ng ml-1 for the tri- to hepta-BDEs when injecting 1 µl. BDE209 gives a smaller and broader peak compared to other PBDEs and could not be detected at the lowest levels. The linear range for BDE209 was from 2 ng ml-1 to 1000 ng ml-1. Thinner films (0.1 µm) on the GC column have been found to improve the detection of BDE209, as have shorter columns. However, the use of a shorter column (10 m) showed little improvement to the detection of BDE209. CB198 was also included in the calibration standards to assess if this chlorobiphenyl (CB) congener could be used as a recovery standard. CB198 is used as one of the recovery standards for CB analysis at FRS ML and is known not to be present in environmental samples. The ion used for quantification was the molecular ion, m/z 430, with confirmatory ions at m/z 35 and 37. The linear range of this compound was similar to the PBDEs, 0.2 ng ml-1 to 500 ng ml-1. Degradation of PBDEs, particularly BDE209, can occur on the GC. The presence of a hump or rising baseline before BDE209 is an indication of degradation during injection, whereas the presence of lower brominated BDE (nona-, octa- and eventually other lower brominated BDEs) indicates possible degradation during extraction and clean-up50. This rising baseline was observed on a number of occasions. To minimise this, the GC liners and injection syringe were changed regularly (once per week), with both syringe and liner being silanised before use. The choice of retention gap can also have an effect on the degradation of BDE209 during analysis. Initially, retention gaps were prepared by cutting a fused silica column into approximately 25 cm lengths and treating them with dimethyl dichlorosilane solution. Varying results were obtained using these retention gaps, therefore, commercially available pre-columns were also investigated. Firstly, varying lengths of methyl deactivated fused silica tubing were investigated, and these gave better results than the shorter fused silica retention gap. Siltek deactivated fused silica retention gaps (0.53 mm i.d. x 5 m length) (Thames Restek, Buckinghamshire) were also investigated, and these proved to give the best and most consistent results for BDE209. The injector end of the retention gap was trimmed periodically when the area of BDE209 decreased relative to the area of CB198. Analysis of HBCD and TBBP-A Analyses of HBCD and TBBP-A is less straightforward than the analyses of PBDEs and a different approach is normally required. HBCD can be determined by GC-MS, but the analysis can be problematic. The uncertainty is greater than for BDEs and standard deviations are high. In addition, the three main HBCD diastereoisomers found in technical mixtures cannot be separated by GC and a total concentration only can be determined. A liquid chromatography (LC) method is required to separate the three diastereoisomers, with separation of enantiomers being possible with a chiral HPLC column11. HBCD degrades at 240oC, therefore, there may be significant losses of HBCD during GC analysis11. A derivatisation step is normally required for analysis of TBBP-A by GC, to the diacetylated derivative2. Few publications analyse TBBP-A and HBCD along with the PBDEs by GC-MS. Verslycke et al. analysed for TBBP-A and HBCD along with the PBDEs using GC-NCIMS, but gave no details of the method or validation data28. Oberg et al. analysed PBDEs and HBCD using high resolution GC-MS but again no information on the method performance was given31. However, LC-MS is the preferred technique for both HBCD and TBBP-A and eliminates the need for a derivatisation step for TBBP-A11. Morris et al. determined TBBP-A and HBCD in sediments and biota using LC-MS47. LC-MS has been used to determine TBBP-A is sediment and sewage sludge55. Analysis of HBCD and TBBP-A using the GC-NCIMS method for PBDEs was possible, however, reproducibility and sensitivity were poor. HBCD gave a broad unresolved peak due to the three diastereoisomers. Both compounds were separated from the PBDEs,


14

however, HBCD and TBBP-A were not always recovered and were generally only detected at higher concentrations (standard solutions of concentrations > 20 ng ml-1 for TBBP-A and > 5 ng ml-1 for total HBCD). Even above these concentrations it was not always possible to detect HBCD and TBBP-A as both compounds can breakdown during the GC separation. It may be possible to use this method to screen samples for high concentrations of HBCD and TBBP-A, however, investigation of an LC-MS method for the analysis of these compounds is recommended. Extraction and Clean-Up A range of extraction methods have been used for the extraction of BFRs from environmental samples. These include the more traditional methods such as Soxhlet and the newer automated methods such as accelerated solvent extraction (ASE)50. However, most laboratories are still using the traditional Soxhlet extraction. Supercritical fluid extraction (SFE) has also been applied however the reproducibility was poor compared to Soxhlet50. Alumina and silica columns as well as gel permeation chromatography have all been used for clean-up of the extract. Destructive methods for lipid removal such as saponification and sulphuric acid treatment have also been investigated; however these methods can result in the degradation of the higher brominated PBDEs. ASE has the advantage of using less solvent, being fully automated and taking less time than traditional methods of extraction such as Soxhlet. Ashizuka et al. used an ASE method for the extraction of PBDEs in marine products43. Hexane was used for the extraction, followed by a separate clean-up involving a sulphuric acid wash and silica and florisil columns. Recoveries of between 57.7 and 78.5% were achieved. Samara et al. also used ASE for the extraction of PBDEs from sediment, followed by an alumina column clean up using hexane:DCM (1:2, v/v)33. A further advantage of using ASE extraction is that it is possible to combine the clean up with the extraction, especially where mass spectrometry is being used as the detection method. Methods have been developed by Lund University for online clean-up and fractionation of dioxins, furans and PCBs with ASE for food, feed and environmental samples56. The first method utilises a fat retainer for the on-line clean-up of fat. Silica impregnated with sulphuric acid, alumina and florisil have all been used as fat retainers. The second method uses an online carbon column for the fractionation of dioxins, furans and planar PCBs. However, problems have been highlighted with the analysis of BDE209 which can be lost during ASE extraction through adsorption on to the extraction system tubing and therefore recoveries tend to be poor for this congener. The ASE technique using fat retainers has been further developed at the Netherlands Institute for Fisheries Research (RIVO) for the extraction and clean-up of samples for PBDE analysis. An ASE method using fat retainers was investigated at FRS ML for the extraction of BFRs from biota and sediment. The ASE conditions are given in the Experimental section and the packing of the ASE cells is shown in Figure 1. Alumina (5% deactivated) was used as the fat retainer and it was found that this could be used for the removal of lipid from biota samples and to clean up freeze dried sediment samples. If a higher alumina deactivation level was used then lipid was found to breakthrough, if a lower deactivation level was used then PBDEs were also retained in the ASE cell and the extraction was less efficient. Using 30 g of 5% deactivated alumina 300 mg of lipid could be retained in the ASE cell. Similarly, using a greater amount of alumina resulted in lower recoveries of PBDEs and a smaller amount resulted in lipid breakthrough. For freeze dried sediment less alumina could be used (10 g). No further clean-up steps were required. A non-polar solvent must be used to avoid lipid breakthrough, therefore, iso-hexane was used for the extraction. Using this method it was essential to first determine the total lipid content, using the Smedes48 method


15

for lipid determination, to enable the amount of tissue to be extracted to be calculated. Following extraction by ASE, the iso-hexane extract was concentrated, using a Turbovap system, to ~ 0.5 ml before GC-NCIMS analysis. Method Validation for PBDEs For the validation of the method the following was investigated: • linear response range • precision of standards and samples • limit of detection • recovery • blank values Linear response range The linear response range was assessed by the triplicate analysis of seven standards ranging from 0.2 ng ml-1 to 500 ng ml-1 for the tri- to hepta-BDEs. The injection volume was 1 µl. BDE209 could not be detected at the lowest levels therefore the concentration range was 2 to 1000 ng ml-1. Linear calibration curves gave correlation coefficients of at least 0.996. Precision of standards and samples For the determination of analytical precision, replicate low (3.5 ng ml-1) and high (200 ng ml-1) standards of the working range were analysed on separate days. Table 4 shows the precision for the BFRs at both concentrations. CV% for the PBDEs were generally <10% for the high standard and <20% for the low standard. Precision of the whole method was calculated through the replicate analysis of a sample (NIST CRM1946 Lake Superior fish tissue) on separate days. This material is not certified for PBDEs but it does contain PBDEs. CV% were generally <15% for the tri- to hepta-BDEs but higher for BDE209 (~30%). Limits of detection The limits of detections (LoD) of individual PBDE congeners were determined through the repeat analysis (n =7) of a low spiked sample consecutively on the same day. Biota (mussels, 8 g) and freeze dried sediment (20 g) were spiked with 1 ng of each individual congener (2 ng for BDE209) and left overnight before extraction. Following ASE extraction and concentration to ~ 0.5 ml, samples were analysed by GC-NCIMS in one batch. The mean concentration and standard deviation for each congener were calculated and the LoDs calculated from 4.65 x S.D. (Table 4). The LoDs for biota samples were between 0.05 and 0.10 µg kg-1 wet weight for the tri- to hepta-BDEs and 0.75 µg kg-1 wet weight for BDE209 (Table 4). For sediment the LoDs were 0.03 µg kg-1 dry weight for all congeners except for BDE99, 85/154 and 209 which had LoDs of 0.09, 0.10 and 1.56 µg kg-1 dry weight, respectively (Table 4). Few publications give method LoDs. Samara et al. reported LoDs in sediment ranging from 0.90 µg kg-1 dry weight for BDE28 to 1.4 µg kg-1 dry weight for BDE154 (BDE209 was not measured) using an ASE extraction followed by alumina clean-up and analysis by GC-MS in EI mode33. Corsolini et al. reported a method LoD of 0.01 µg kg-1 wet weight for individual BDE congeners in biota using a Soxhlet extraction, acidified silica clean-up and analysis by GC-NCIMS, again BDE209 was not measured20. Erdogrul et al. report method LoDs ranging from 0.02 to 0.1 µg kg-1 wet weight for individual BDE congeners (tri- to hepta-BDE congeners) in fish using a Soxhlet extraction, acidified silica clean-up and analysis by GC-NCIMS42. Hites et al. reported LoDs of between 0.001 to 0.01 µg kg-1 wet weight for individual BDE congeners in fish using a soxhlet extraction, gel


16

permeation chromatography clean-up and analysis by high resolution GC-MS39. Vives et al. calculated LoDs from the SD of the mean of real samples x 344. In fish muscle the LoDs for individual BDE congeners ranged from 0.009 – 0.013 µg kg-1 wet weight. Recovery Recoveries were calculated through replicate analysis of PBDEs in spiked salmon liver (~0.5 g), salmon flesh (~1.5 g), Mytilius edulis (mussels, 8 g) and freeze dried sediment (20 g) samples on separate days. The spiking solution was added to the sample and sodium sulphate mixture prior to drying overnight. Both sediment and biota were spiked with 40 ng of each PBDE congener, except for BDE209 where the spike was 50 ng. The samples were then extracted by ASE, concentrated by Turbovap and analysed by GC-NCIMS. Recoveries are shown in Table 5 and typical chromatograms are shown in Figures 3 and 4. Mean recoveries for individual PBDE congeners were 82.73 – 93.28% (CV% = 13.79 – 32.04, n = 6), 73.88 – 89.86% (CV% = 8.93 – 31.36, n = 4) and 66.94 – 88.28% (CV% = 1.96 – 29.05, n = 4) for fish liver, fish flesh and mussels, respectively. For sediment, recoveries ranged from 75.4 – 100.1% (CV% = 12.77 – 30.98, n = 11). In all case CV% were highest for BDE209, this analysis is often problematic (see previous sections) with reproducibility being poor despite the precautions taken. Covaci et al. reviewed analytical methods from the literature and concluded that precision of between 10 and 20% is typical for the tri-to hepta-BDEs and around 25% for BDE20948. CB198 was also present in the spiking solution with the aim of assessing if this compound could be used as a recovery standard which could be added to samples prior to extraction as a check on the method. Recoveries for CB198 were similar to the PBDEs, in addition it did not interfere with any of the BDE congeners on the GC chromatogram. This CB is not found in environmental samples and is one of the recovery standards used in routine CB analysis at FRS ML. Therefore, this CB congener is a suitable recovery standard. Procedural Blanks With each batch of samples a procedural blank was analysed. The matrix/sodium sulphate mixture (Fig. 1) was replaced with sodium sulphate and extracted by ASE, concentrated by Turbovap and analysed by GC-NCIMS. Concentrations of all but BDE209 were below the LoDs. BDE209 was detected in all procedural blanks with final concentrations in the GC vial ranging from 3-30 ng ml-1 initially but normally < 10 ng ml-1. A typical chromatogram is shown in Figure 5. BDE209 is known to stick to any active sites, such as on glassware and possibly also in the ASE system, resulting in problems with procedural blanks. The use of Siltek™ treated extraction cells (Thames Restek, Buckinghamshire) will be investigated in the near future.

CONCLUSIONS 1. Brominated flame retardants (BFRs) have a high usage and can undergo long-range

atmospheric transport. There is limited data on BFRs in the Scottish marine environment but it is thought likely that they will be found, both in biota and sediments from Scottish waters.

2. Following a review of hazardous substances in the Scottish marine environment one

of the key recommendations was that FRS ML should establish the analytical capability to determine PBDEs in sediment and biota and establish a limited monitoring programme to assess the current situation.


17

3. A review of analytical methods used for the analysis of BFRs (PBDEs, HBCD, TBBP-A) was undertaken. GC-NCIMS is the most commonly used technique for this analysis. A range of extraction methods have been used including traditional Soxhlet and newer automated methods such as accelerated solvent extraction (ASE). Most publications employed a separate clean up using gel permeation, alumina or silica column or sulphuric acid treatment.

4. One advantage of using ASE is that the extraction and clean-up may be combined by

adding fat retainers to the extraction cell. This technique was investigated at FRS for the extraction of BFRs from sediment and biota.

5. Using 5% deactivated alumina in the ASE cell it was possible to extract and clean-up

sediment and biota (= 300 mg lipid) samples. No further clean-up steps were required with the extract simply concentrated prior to analysis by GC-NCIMS.

6. This method was validated for both biota (fish flesh and liver and mussels) and

sediment. 7. Replicate analysis of high and low standards on separate days by GC-NCIMS gave

CV% for the tri- to hepta-BDEs of <10% and <20%, respectively. 8. Recoveries for the tri- to hepta-BDEs from sediment and biota were mainly between

75 and 95%. CV% from analysis on separate days were <15%. 9. BDE209 gave poorer detection limits and reproducibility. CV% for spiked samples

analysed on separate days were ~30%. However this is similar to other published work and problems with this analysis have been previously reported.

10. The LoDs for biota samples were between 0.05 and 0.07 µg kg-1 wet weight for the tri

to hepta-BDEs and 0.75 µg kg-1 wet weight for BDE209. For sediment the LoDs were 0.03 µg kg-1 dry weight for all congeners except for BDE99, 85/154 and 209 which had LoDs of 0.09, 0.10 and 1.56 µg kg-1 dry weight, respectively.

11. Analysis of HBCD and TBBP-A gave high detection limits and results were highly

variable. GC-NCI-MS could not be used for the quantitative analysis of these compounds. Investigation of an LC-MS method for the analysis of HBCD and TBBP-A is recommended.

REFERENCES 1. Webster, L., Robinson, C.D., Gubbins, M.J., Davies, I.M. and Moffat, C.F. 2004. A

review of hazardous substances in the Scottish marine environment. FRS Internal Report No 01/04.

2. OSPAR Commission. 2001. Certain Brominated Flame Retardants-Polybrominated

diphenyl ethers, polybrominated biphenyls, hexabromo cyclododecane, OSPAR Priority Substances. OSPAR Commission, London.

3. OSPAR Commission. 2005. Tetrabromobisphenol-A-Update, OSPAR Priority

Substances. OSPAR Commission, London. 4. OSPAR Commission. 2002. Provisional Instruction Manual for the dynamic

selection and prioritisation mechanism for hazardous substances.


18

5. Water Framework Directive. Directive 2000/60. EC of the European Parliament and

of the Council, 2000. 6. de Wit, C.A. 2002. An overview of brominated flame retardants in the environment.

Chemosphere, 46, 583–624. 7. McDonald, T.A. 2002. A perspective of the potential health risks of PBDEs.

Chemosphere, 46, 745–755. 8. Pijnenburg, A.M.C.M., Everts, J.W., de Boer, J. and Boon, J.P. 1995.

Polybrominated biphenyl and diphenyl ether flame retardants: Analysis, Toxicity and environmental occurrence. Review of Contamination and Toxicology, 141, 1-26.

9. Palm, A. Cousins, I. T. Mackay, D. Tysklind, M. Metcalfe C. and Alaee, M. 2002.

Assessing the environmental fate of chemicals of emerging concern: a case study of the polybrominated diphenyl ethers. Environ. Pollut., 117, 195–213.

10. Eriksson, P. Jakobsson E. and Fredriksson, A. 1998. Developmental neurotoxicity

of brominated flame retardants, polybrominated diphenyl ethers and tetrabromo-bis-phenol A. Organohalogen Compounds, 35, 375-377.

11. Law, R.J. Kohler, M. Heeb, N.V. Gerecke, A.C. Schmid, P. Voorspoels, S. Covaci, A.

Becher, G. Janak, K. and Thomsen, C. 2005. Hexabromocyclododecane challenges scientists and regulators. Environ. Sc. Technol., 281A – 287A.

12. Law, R.J., Allchin, C.R., de Boer, J., Covaci, A., Herzke, D., Lepom, P., Morris, S.

and de Wit, C.A. 2004. Levels and trends of brominated flame retardants in the European environment. Proceedings of the Third International Workshop on Brominated Flame Retardants BFR2004, June 6–9, 2004 Toronto, Canada. 79-104.

13. Allchin, C.R., Law, R.J. and Morris, S. 1999. Polybrominated diphenylethers in

sediments and biota downstream of potential sources in the UK. Environ. Pollut., 105, 197-207.

14. Veltman, K., Hendriks, J., Huijbregts, M., Leonards, P., van-den Heuvel-Greve. M.

and Vethaak, D. 2005. Accumulation of organochlorines and brominated flame retardants in estuarine and marine food chains: field measurements and model calculations. Mar. Poll. Bull., 50, 1085–1102.

15. Law, R. J. Allchin, C. R. Bennett, M. E. Morris S.and Rogan, E. 2002.

Polybrominated diphenyl ethers in two species of marine top predators from England and Wales. Chemosphere, 46, 673–681

16. Ikonmou, M.G., Rayne, S., Fischer, M., Fernandez, M.P. and Cretney, W. 2002.

Occurrence and congener profiles of polybrominated diphenyl ethers (PBDEs) in environmental samples from coastal British Columbia, Canada. Chemosphere, 46, 649-663.

17. Christensen, J.H., Glasius, M., Pecseli, M., Platz, J. and Pritzl, G. 2002.

Polybrominated diphenyl ethers (PBDEs) in marine fish and blue mussels from Southern Greenland. Chemosphere, 47, 631–638.


19

18. de Boer, J., Wester, J., Pastor, P.G., Rodriguez, I.D., Lewis, W.E. and Boon, J.P. 1998. Polybrominated biphenyls and diphenylethers in sperm whales and other marine mammals - a new threat to ocean life? Organohalogen Compounds, 35, 383-386.

19. Lindstrom, G. Wingfors, H. Dam, M. and Bavel, B.V. 1999. Identification of 19

polybrominated diphenyl ethers (PBDEs) in long-finned pilot whale (Glonbicephala melas) from the Atlantic. Arch. Environ. Contam. Toxicol., 36, 355–363.

20. Corsolini, S. Covaci, A. Ademollo, N. Focardi, S. and Schepens, P. 2006.

Occurrence of organochlorine pesticides (OCPs) and their enantiomeric signatures, and concentrations of polybrominated diphenyl ethers (PBDEs) in the Adelie penguin food web, Antarctica. Environ. Pollut., 140, 371–382.

21. Law, R.J., Allchin, C.R. and Mead, L.K. 2005. Brominated diphenyl ethers in twelve

species of marine mammals stranded in the UK. Mar. Poll. Bull., 50, 356-359. 22. Hall, A.J. Kalantzi, O.I. and Thomas, G.O. 2003. Polybrominated biphenyl ethers

(PBDEs) in grey seals during the first year of life-are they thyroid hormone endocrine disruptors. Environ. Pollut., 126, 29–37.

23. Learmonth, J.A. 2005. Life history and fatty acid analysis of harbour porpoises

(phocoena phocoena) from Scottish waters, PhD thesis, Aberdeen University, Aberdeen.

24. Noren, K. and Meironyte, D. 2000. Certain organochlorine and organobromine

contaminants in Swedish human milk in perspective of past 20–30 years. Chemosphere, 40, 1111–1123.

25. Sjodin, A., Hagman, L., Klasson-Wehler, E., Jakobsson, E. and Bergman, A. 1999.

Flame retardant exposure: Polybromianted diphenyl ethers in blood from Swedish workers. Environ. Health Persp., 107, 643–647.

26. Kierkegaard, A., Bignert, A., Sellstrom, U., Olsson, M., Asplund, L., Jansson, B. and

Wit, C.A. 2004. Polybrominated diphenyl ethers and their methoxylated derivatives in pike from Swedish waters with emphasis on temporal trends (1967–2000). Environ. Pollut., 130, 187-198.

27. Ikonomou, M. Fernandez, M.P. and Hickman, Z.L. 2006. Spatio-temporal and

species-specific variation in PBDE levels/patterns in British Columbia’s coastal waters. Environ. Pollut., 140, 355-363.

28. Verslycke, T.A., Vethaak, A.D., Arijis, K. and Janssen, C.R. 2005. Flame retardants,

surfactants and organotins in sediment and shrimp of the Scheldt estuary (The Netherlands). Environ. Pollut., 136, 19–31.

29. Thomas, G.O., Moss, S.E.W., Asplund, L. and Hall, A.J. 2005. Absorption of

decabromodiphenyl ether and other organohalogen chemicals by grey seals (Halichoerus grypus). Environ. Pollut., 133, 581–586.

30. de Boer, J., Wester, P.G., van der Horst, A. and Leonards, P.EG. 2003.

Polybrominated diphenyl ethers in influents, suspended particulate matter, sediments, sewage treatment plant and effluents and biota from the Netherlands. Environ. Polutl., 122, 63–74.


20

31. Oberg, K., Warman, K. and Oberg, T. 2002. Distribution and levels of brominated

flame retardants in sewage sludge. Chemosphere, 88, 805–809. 32. Eljarrat, E., De La Cal, A., Larrazabal, D., Fabrellas, B., Fernandez-Alba, A.R.,

Borrull, F., Marce, R.M. and Barcelo, D. 2005. Occurrence of polybrominated diphenylethers, polychlorinated dibenzo0p-dioxins, dibenzofurans and biphenyls in coastal sediments from Spain. Environ. Pollut., 136, 493–501.

33. Samara, F., Tsai, C.W. and Aga, D.S. 2006. Determination of potential sources of

PCBs and PBDEs in sediments from the Niagara River. Environ. Pollut., 139, 489-497.

34. Covaci, A. Gheorghe, A. Hulea, O. and Schepens, P. 2006. Levels and distribution

of organochlorine pesticides, polychlorinated biphenyls and polybrominated diphenyl ethers in sediment and biota from Danube Delta, Romania. Environ. Pollut., 140, 136–149.

35. Jacobs, M., Ferrario, J. and Byrne, C. 2002. Investigation of polychlorinated

dibenzo-p-dioxins and dibenzo-p-furans and selected coplanar biphenyls in Scottish farmed Atlantic salmon (Salmo salar). Chemosphere, 47, 183–191.

36. Hites, R., Foran, J.A., Carpenter, D.O., Hamilton, M.C., Knuth, B.A. and

Schwager, S.J. 2004. Global assessment of organic contaminants in farmed and wild salmon. Science, 303, 226–229.

37. Easton, M.D.L., Lusziniak, D. and Von der Geest, E. 2002. Preliminary examination

of contaminant loadings in farmed salmon, wild salmon and commercial salmon feed. Chemosphere, 46, 1053–1074.

38. Pirie, D. 2001. Report on the occurrence and significance of polychlorinated

biphenyls (PCBs) in marine sediments from Scottish fish farm locations. SEPA internal report (unpublished).

39. Hites, R., Foran, J.A., Schwager, S.J., Knuth, B.A., Hamilton, M.C. and

Carpenter, D.O. 2004. Global assessment of polybrominated diphenyl ethers in farmed salmon. Env. Sci. Technol, 38, 4945-4949.

40. Tlustos, C., Pratt, I., McHugh, B., Tyrell, L., Cooper, H., Duffy, C. and McGovern, E.

2005. Investigation into levels of polybrominated diphenyl ethers (PBDEs) and hexabromocyclododecane (HBCD) in fishery products available on the Irish market, Dioxin 2005 Conference, Toronto, EMG-Bromonitated Flame Retardants IV, 636–639.

41. Zennegg, M., Kohler, M., Gerecke, A.C. and Schmid, P. 2003. Polybrominated

diphenyl ethers in whitefish from Swiss lakes and farmed rainbow trout. Chemosphere, 51, 545–553.

42. Erdrogrul, O., Covaci, A. and Schepens, P. 2005. Levels of organochlorine

pesticides, polychlorinated biphenyl and polybrominated diphenyl ethers in fish species from Kahramanmaras, Turkey. Environ. Int., 32, 703-711.


21

43. Ashizuka, Y., Nakagawa, R., Tobiishi, K., Hori, T. and Iida, T. 2005. Determination of polybrominated diphenyl ethers and polybrominated dibenzo-p-dioxins/dibenzofurans in marine products. J. Agric. Food Chem., 53, 3807–3813.

44. Vives, I., Grimalt, J.O., Lacorte, S., Guillamón, M. and Barcelo, D. 2004.

Polybromodiphenyl ether flame retardants in fish from lakes in European high mountains and Greenland. Environ. Sci. Technol., 38, 2338-2344.

45. Remberger, M., Sternbesk, J., Palm, A., Kaj, L., Stromberg, K. and Brorstrom-

Lunden, E. 2004. The environmental occurrence of hexabromocyclododecane in Sweden. Chemosphere, 54, 9–21.

46. Morris, S., Allcin, C.R., Zegers, B.N., Haftka, J.H., Boon, J.P., Belpaire, C.,

Leonards, P.E.G., Van Leeuwen, S.P.J. and De Boer, J. 2004. Distribution and fate of HBCD and TBBP-A brominated flame retardants in North Sea estuaries and aquatic food webs. Environ. Sci. Technol., 38, 5497–5504.

47. Zegers, B.N., Mets, A., Van Bommel, R., Minkenberg, C., Hammers, T., Kamstra, J.,

Pierce, G.J. and Boon, J. P. 2005. Levels of hexabromocyclododecane in harbour porpoises and common dolphins from western European Seas, with evidence for stereoisomer-specific biotransformations by Cytochrome P450. Environ. Sci. Technol., 39, 2095–2100.

48. Smedes, F. 1999. Determination of total lipid using non-chlorinated solvents.

Analyst, 124, 1711-1718. 49. de Boer, J., Allchin, C.R., Zegers, B. and Boon, J.P. 2001. Method for the analysis

of polybrominated diphenyl ethers. Trends in Anal. Chem., 20, 591–599. 50. Covaci, A., Voorspoels, S. and de Boer, J. 2003. Determination of brominated flame

retardants, with emphasis on polybrominated diphenyl ethers (PBDEs) in environmental and human samples- a review. Environ. Int., 29, 735–756.

51. de Boer, J. and Cofino, W.P. 2002. First world-wide interlaboratory study on

polybrominated diphenyl ethers (PBDEs). Chemosphere, 46, 625–633. 52. Korytar, P., Covaci, A., de Boer, J., Gelbin, A. and Brinkman, U.A.Th. 2005.

Retention of 126 polybrominated diphenyl ether congeners and two Bromkal technical mixtures on seven gas chromatographic columns. Journal of Chromatography A, 1065, 239–249.

53. Ackerman, L.K., Wilson, G.R. and Simonich, S.L. 2005. Qualitative analysis of 39

polybrominated diphenyl ethers by isotope dilution GC/Low-resolution MS. Anal. Chem., 77, 1979–1987.

54. The third International Workshop on Brominated Flame Retardants, Toronto, 2004. 55. Saint-Louis, R. and Pelletier, E. 2004. LC-ESI-MS-MS method for the analysis of

tetrabromobisphenol A in sediment and sewage sludge. Analyst, 129, 724–730. 56. Sporring, S., Wiberg, K., Bjorklund, E. and Haglund, P. 2003. Combined

extraction/Clean-up strategies for fast determination of PCDD/Fs and WHO PCBs in food and feed samples using accelerated solvent extraction. Organohalogen Compounds, 60- 65, Dioxin 2003, Boston.


22

TABLE 1

Persistence, bioaccumulation and toxicity data for BFRs along with the DYNAMEC selection Criteria

Brominated flame retardants DYNAMEC

criteria Tetra- and

penta-BDEs; penta mix

Hexa- and hepta-

BDEs; Octa mix

Deca-BDE; deca mix

HBCD TBBP-A

Half-life 50 days 600 days (aerobic sediment)

Log Kow ≥ 4 6.60 – 7.32 7.24 – 8.27 10.33 5.8 4.5 BCF ≥2000

(l kg-1) 27400 18,100 1,234 (fish)

L(E)C50 ≤ 1 (mg l-1)

>500 (fish) 0.54 (fish)

NOEC (fish)

≤ 0.1 (mg l-1)

0.0089 (fish) 0.16 (fish)


23

TABLE 2 Brominated flame retardants (BFR) analysed as part of this study along with their degree of bromination, chemical name and the octanol water coefficient (Log KOW), where available.

BFR Number of Br Name Log

KOWBDE17 3 2,2’,4-tribromodiphenyl ether 5.74 BDE28 3 2, 4,4’-tribromodiphenyl ether 5.94 BDE75 4 2, 4,4’,6-tetrabromodiphenyl ether BDE49 4 2, 2’,4,5’-tetrabromodiphenyl ether 6.60 BDE71 4 2,3’,4’,6-tetrabromodiphenyl ether 6.81 BDE47 4 2,2’,4,4’-tetrabromodiphenyl ether BDE66 4 2,3’,4,4’-tetrabromodiphenyl ether BDE77 4 3,3’,4,4’-tetrabromodiphenyl ether BDE100 5 2,2’,4,4’,6-pentabromodiphenyl ether 7.24 BDE119 5 2,3’,4,4’,6-pentabromodiphenyl ether BDE99 5 2,2’,4,4’,5-pentabromodiphenyl ether 7.32 BDE85 5 2,2’,3,4,4’-pentabromodiphenyl ether BDE154 6 2,2’,4, 4’,5, 6’-hexabromodiphenyl ether 7.82 BDE153 6 2,2’,4,4’,5,5’-hexabromodiphenyl ether 7.90 BDE138 6 2,2’,3, 4,4’,5-hexabromodiphenyl ether BDE190 7 2,3,3’,4,4’,5,6-heptabromodiphenyl ether BDE183 7 2,2',3,4,4',5',6-heptabromodipheny l ether 8.27 BDE209 10 2,2’,3,3’,4,4’,5,5’,6,6’-decabromodiphenyl ether 10.33 TBBP-A 4 Tetrabromobisphenol-A 4.5 Dimethyl-TBBP-A 4 Dimethyl Tetrabromobisphenol-A 6.4 HBCD 6 Hexabromocyclododecane 5.8


24

TABLE 3 Summary of PBDE concentrations reported in environmental samples. Concentrations are given in µg kg-1 and reported either on a wet weight (ww), lipid weight (lw) or dry weight (dw) basis. The number of PBDE congeners analysed is shown in brackets.

Matrix Area Year sampled

ΣPBDE concentration (µg kg-1)

Reference Number

Sediment UK 1999 < 0.3 – 368 (BDE47, dw); <0.6 – 898 (BDE99, dw)

13

Fish UK 1999 ND – 9,500 (BDE47, lw); ND – 370 (BDE99, lw)

13

Cormorants England and Wales

1996 - 2000 1.8 – 140 (14, ww) 15

Harbour porpoise

England and Wales

1996 - 2000 ND – 6,900 (14, ww) 15

Porpoise blubber

Canada 1991 - 1993 350 – 2,300 (13, lw) 16

Sole Canada 1992, 2000 22 – 340 (13, lw) 16 Marine fish Greenland 1.2 – 12.0 (4, ww) 17 Mussels Greenland ND - 0.11 (4, ww) 17 Sperm whale/ dolphin/harbour seal blubber

Holland 1995 78.5 - 7,700 (4, lw) 18

Pilot whale blubber

Faroe Islands

1994, 1996 843.2 – 3,260 (19, lw) 19

Krill Antarctica 1995 - 2000 4.8 – 6.2 (7, lw) 20 Rockcod (whole)

Antarctica 1995 - 2000 2.4 – 9.5 (7, lw) 20

Rockcod muscle

Antarctica 1995 - 2000 1.96 – 7.11 (7, lw) 20

Penguin eggs

Antarctica 1995 - 2000 0.91 – 6.00 (7, lw) 20

Marine mammals

UK 1992 - 2002 5 -16,200 (10, lw) 21

Grey seal pups

NE England 1998, 1999 61 – 903 (13, lw) 22

Harbour porpoise

Scotland 2001 - 2003 130.6 – 2,213 (5, lw) 23

pike Sweden 1967 - 2000 0.06 -1.6 (5, ww) 26 Dunganess crab

Canada 1992 - 2002 8.6 – 660 (14, lw) 27

Mysid shrimp

Scheldt 2001 1765 – 2,962 (15, lw) 28

sediment Scheldt 2001 14 – 22 (15, dw) 28 Marine Mussels

Holland 1999 1.2 (BDE47);0.5 (BDE99); <0.1 (BDE153; <4 (BDE209) (dw)

30

Sediment Holland 1999 1.1 (BDE47);0.6 (BDE99); <0.7 (BDE153; 22 (BDE209) (dw)

30


25

Matrix Area Year sampled

ΣPBDE concentration (µg kg-1)

Reference Number

Sewage sludge

Sweden 1999, 2000 ND – 450 (8, ww) 31

Sediment Spain 2002 2.7 – 134 (40, dw) 32 Sediment US 2003 0.72 – 148 (9, dw) 33 Sediment Romania 2001 <LoD (7, dw) 34 Zooplankton Romania 2001 1.0 – 7.2 (7, dw) 34 Fish (various)

Romania 2001 <LoD – 14.3 (7, lw) 34

Farmed salmon

Canada 1999, 2000 1.2 - 4.1 (41, ww) 37

wild salmon Canada 1999, 2000 0.04 – 0.5 (41, ww) 37 Farmed and wild salmon

Europe and North America

2002 <1- 5 (43, ww) 39

Farmed salmon

Ireland 2.42 - 5.05 (16, ww) 40

Wild salmon Ireland 0.7 -1.01 (16, ww) 40 mackerel Ireland 0.93 – 1.86 (16, ww) 40 herring Ireland 1.46 – 1.77 (16, ww) 40 Rainbow trout

Sweden 0.74 – 1.3 (7, ww) 41

Fish (kalashapa, carp, nose-carp, wells) muscle

Turkey 2003 ND – 6.7 (6, ww) 42

Fish (kalashapa, carp, nose-carp, wells) muscle

Turkey 2003 ND – 597 (6, lw) 42

Fish Japan 2004 ND – 0.55 (14, ww) 43 shellfish Japan 2004 0.01 – 0.12 (14, ww) 43 Brown trout (liver)

European high mountain lakes and Greenland

2000, 2001 0.11 – 11 (6, ww) 44

Brown trout (flesh)

European high mountain lakes and Greenland

2000, 2001 0.07 -1.2 (6, ww) 44


26

TABLE 4 Precision of low (3.5 ng ml-1; n = 7) and high (200 ng ml-1;n = 9) standards and method LoD determined from the replicate (biota n = 7; sediment n = 11) analysis of a spiked sample at a low concentration (4.65 x SD).

LoD biota (µg kg-1 wet

weight)

LoD sediment (µg kg-1 dry

weight)

High standard %CV

Low standard %CV

BDE17 0.05 0.03 8.06 17.88 BDE28 0.05 0.03 7.63 18.85 BDE75 0.06 0.03 7.66 13.80 BDE49 0.05 0.03 7.49 14.28 BDE71 0.06 0.03 8.92 13.41 BDE47 0.07 0.03 6.77 15.73 BDE66 0.05 0.03 7.40 16.13 BDE77 0.05 0.03 7.38 15.14 BDE100 0.05 0.03 8.40 23.24 BDE119 0.05 0.03 7.93 24.66 BDE99 0.06 0.09 8.62 17.25 BDE85/154 0.10 0.10 8.16 16.26 BDE153 0.06 0.03 8.68 15.52 BDE138 0.05 0.03 9.46 18.42 BDE183 0.06 0.03 9.36 16.85 BDE190 0.05 0.03 10.42 16.70 BDE209 0.75 1.56 9.55 31.33

TABLE 5

Summary of results from spiking experiments. Fish liver and flesh, mussels and sediment were spiked with PBDEs (fish liver, 80 µg kg-1, n = 6; flesh, 27 µg kg-1, n = 4; mussel, 5 µg kg-1, n = 4; sediment, 2 µg kg-1, n = 11) and the recovery and CV% calculated for each PBDE congener. CB198 was also included in the spiking solution to assess if this could be used as a recovery standard

Fish liver Fish flesh Mussels Sediment % recovery CV% %recovery CV% %recovery CV% %recovery CV%

BDE17 85.38 14.32 80.48 8.93 71.37 4.23 77.07 12.77BDE28 87.13 14.67 82.02 9.82 74.46 4.33 77.44 15.50BDE75 88.92 16.01 80.92 10.36 75.34 3.16 75.43 15.25BDE49 90.69 14.84 81.80 9.90 79.70 4.12 76.79 16.12BDE71 93.28 13.79 89.86 20.02 88.28 22.41 100.11 13.06CB198 82.73 12.24 73.88 6.90 66.94 5.05 80.00 10.68BDE47 82.73 14.66 73.88 9.11 66.94 3.51 83.65 12.96BDE66 90.96 14.95 82.50 9.28 75.45 2.40 82.61 13.12BDE77 92.41 15.48 83.00 9.45 76.64 1.96 79.74 16.32BDE100 92.56 14.98 84.12 8.99 77.82 4.18 75.65 16.62BDE119 92.49 14.54 84.19 8.73 77.23 3.16 91.01 15.72BDE99 90.92 14.36 84.40 9.77 77.23 3.03 81.00 16.52BDE85/154 92.49 14.51 84.19 10.02 77.23 2.52 81.38 14.95BDE153 92.72 14.12 83.70 10.14 76.41 3.88 84.09 14.10BDE138 90.33 14.35 83.11 11.97 76.30 7.81 84.60 14.88BDE183 89.24 14.45 82.86 10.82 76.60 7.41 84.26 13.46BDE190 87.25 16.94 79.50 15.25 73.24 12.45 80.20 16.57BDE209 84.56 32.04 78.95 31.36 72.83 29.05 84.86 30.98

Figure 1 Packing of accelerated solvent extraction (ASE) tubes for the extraction of PBDEs from biota. For sediment samples a layer of activated copper is placed at the bottom of the tube to remove sulphur.

Na 2SO 4

Na 2SO 4

PBDEs

Matrix / Na 2SO 4

Filter paper

Filter paper

Fat retainer (alumina)Filter paper

Filter paper

Na 2SO 4

Na 2SO 4

Figure 2 Chromatogram of a PBDE standard analysed by GC-NCIMS using an STX-500 column. Baseline separation was not achieved for BDE85 and 154. Sensitivity was lowest for BDE209. CB198 was added as a recovery standard.

10.00 15.00 20.00 25.00 30.00 35.00 40.00 45.0

10000

20000

30000

40000

50000

60000

70000

80000

90000

100000

110000

120000

130000

140000

150000

160000

Time-->

Abundance

TIC: BFR1277.D\data.ms8

4

70

9

BDE20

BDE19
BDE1
CB19

BDE85/15

0 50.00 55.00 60

Figure 3 Chromatogram of spiked biota (salmon liver) sample extracted by ASE and analysed by GC-NCIMS for PBDEs and with CB198 being used as a recovery standard.

10.00 1

10000

20000

30000

40000

50000

60000

70000

80000

90000

100000

110000

120000

130000

140000

150000

160000

Time-->

Abundance

8

Figure 4 Chromatogr

analysed byrecovery sta

.

10.00 1

1000

2000

3000

4000

5000

6000

7000

8000

9000

10000

11000

12000

13000

14000

15000

16000

17000

18000

19000

20000

21000

22000

23000

24000

Time-->

Abundance

CB19

5.00 20.00 25.00 30.00 35.00 40.00 45.00 50.00 55.00 60.00

TIC: BFR1277.D\data.ms

BDE190

BDE209

am of spiked sediment sample extracted by ASE and GC-NCIMS for PBDEs and with CB198 being used as a ndard.

5.00 20.00 25.00 30.00 35.00 40.00 45.00 50.00 55.00 60.00


CB198

BDE190

BDE209

Figure 5 Chromatogram of a procedural blank extracted by ASE and analysed

by GC-NCIMS for PBDEs and with CB198 being used as a recovery standard. Note the expanded scale on the Y-axis resulting in base-line noise evident in the chromatogram.

.

10.00 15.00 20.00 25.00 30.00 35.00 40.00 45.0

300

320

340

360

380

400

420

440

460

480

500

520

540

560

580

600

620

640

660

680

Time-->

Abundance


8

9

BDE20

CB19

0 50.00 55.00 60.00

APPENDIX 1

Methods and Standard Operating Procedure (SoP) for the analysis of PBDEs in Sediment and Biota

i

Determination of PBDEs in Biota and Sediment - M2900 1. Introduction and Scope This method describes the determination of polybrominated diphenyl ethers (PBDEs) in biota and sediment. The analysis incorporates tri to deca-brominated congeners. This does not cover all of the many PBDE compounds that exist. The concentration range of the method is from the limit of detection to 40 ng g-1 (wet weight) for mussels and fish muscle, 750 ng g-1 (wet weight) for fish liver and 20 ng g-1 (dry weight) for freeze dried sediment. 2. Principle of the Method PBDEs are extracted from biota and sediment using an accelerated solvent extraction (ASE). The extract is concentrated using a TurboVap. The addition of a fat retainer (5% deactivated alumina) means no further clean-up is required for the biota. For the sediment activated copper is added to both the ASE tube and the collection bottle and the resultant extract is passed through sodium sulphate to ensure the removal of any particles of copper. Quantitative analysis is carried out by gas chromatography with mass selective detection (GC-MSD) in chemical ionisation mode (CI) using a Thames Restek STX-500 column fitted with 5 m Siltek™ deactivated pre-column (Thames Restek). An external standard method is used for quantification of samples. The GC-MSD is calibrated using seven different concentrations of a solution containing 18 PBDEs. 3. Reference Materials LRM110, contaminated cod liver oil, containing most determinands. LRM140, a freeze dried sediment. 4. Reagents See SOPs for reagents used. 5. Equipment Gas chromatograph with on column injector and mass selective detector (EN 757 & EN 751 Room C125) Dionex Accelerated solvent extractor (ASE 300; EN 1241) TurboVap (EN1198 or 1199) 6. Environmental Control See individual SOPs. 7. Interferences Special precautions are required for PBDE analysis, particularly if BDE209 is to be measured, due to its sensitivity to UV light. Incoming light is minimised in the laboratory by placing UV filters on windows, and keeping lights turned off. In addition, BDE209 can adsorb to dust particles, therefore, an ioniser is placed in the laboratory. BDE209 can also stick to glassware (or any other active site) which can

ii

result in contamination of glassware. All glassware is therefore solvent washed with acetone followed by iso-hexane. PBDE analysis requires a stable environment, avoiding contamination of samples and reagents eg contact with fingers, dirty equipment, plastics, newly painted surfaces. All new batches of iso-hexane and are checked for contamination as outlined in SOP 1620 and analysed by gas chromatography with flame ionisation detection (GC-FID), as described in SOP1610. 8. Sampling and Sample Preparation Samples are logged into the laboratory according to SOP 60. Samples are sub-sampled and mixed according to SOP 0367 (section 8). 9. Analytical Procedure 9.1 A procedural blank and LRM are analysed with each batch of samples. 9.2 The extraction of PBDEs is carried out using accelerated solvent extraction

(ASE) as detailed in SOP0367. 9.3 The extracts are concentrated using Turbovap as outlined in SOP0560. and

transferred into a pre-weighed GC vial. The weight of the GC vial and cap are recorded before and after the transfer of sample on record sheet B314.

9.5 Analysis is performed by GC-MSD as outlined in SOP2911. Sequences are

set up as in SOP 1265 and results are quantified using SOP 1260. 9.6 Calibration standards, required for quantitative analysis, are prepared as

described in SOP2910. 10. Calculation of Results The GC-MSD is calibrated and results are calculated using the HP data analysis software as described in SOPs 1260 and 2841, using an external standard method. BDE85 and BDE154 do not have baseline separation at low levels therefore the peak areas are summed and a summed concentration is calculated. The correlation coefficient should be greater than 0.996 for the calibration curves. The retention times of the compounds in the calibration standards are also used to confirm retention times and identities of peaks in the LRM and the samples. LRM data arre monitored by plotting results on Shewhart charts with limits at ±2x and ±3x S.D. Results are calculated using the PBDE template (B315 – biota; B316 - sediment). The weight of the sample and the weight of the sample in the GC vial are inserted in the table. The software programme carries out the blank subtraction and calculates the concentration of each PBDE to give a figure in µg kg-1 wet weight of biota and dry weight of sediment. 11. Precision, accuracy and practical detection limits Recoveries were calculated by spiking sediment and biota (fish liver and muscle) samples before extraction. Limits of detection are calculated by multiplying the standard deviation of the mean of a low sample by 4.65. 12. Reports

iii

http://nts5/scripts/qwbcgi.exe/docfetch?id=154&db=ukas




http://nts5/scripts/qwbcgi.exe/docfetch?id=1109db=ukas












A hardcopy of all data should be obtained and submitted to the Technical Manager along with other relevant documentation (SOP 1350). Batches of results are electronically archived (SOP 030) to duplicate CDs via an internal CD writer. CDs are labelled with archive dates, files checked, group name and contents, one is given to Quality Management for archive and the other is stored in Room C125. Paper copies of chromatograms are kept for one year. 13. Safety Safety for all relevant procedures are provided in the appropriate SOPs detailed above, with reference to risk assessments. 14. Literature references See SOPs 0367, 0371, 2911 and 2910.

15. Uncertainty of Measurement Sampling: Sampling not part of method. Samples are analysed and results reported on the samples as received – outwith uncertainty calculations. Subsampling: Processing – Error due to inhomogeneity of sample is minimised by mixing thoroughly in sample container - negligible contribution to uncertainty. Injection on Rheodyne – Assume sample in vial is homogenous - negligible contribution to uncertainty. Injection on GC-MS – Assume sample in vial is homogenous – negligible contribution to uncertainty. Storage: Samples are stored deep frozen to minimise degradation. Reagent purity: All solvents are from Rathburn Chemicals and of at least HPLC Grade, considered sufficient – uncertainty accounted for in validation data. Other chemicals are at least Analar quality, considered sufficient – uncertainty accounted for in validation data. Chemical standards used in the preparation of calibration solutions are of the highest purity available at time of purchase. Final concentrations of the calibration solutions have been corrected for purity- uncertainty accounted for in the validation data. Instrument effects: All syringes are solvent washed between samples. Weight – Tolerance of balance – balances check weight tolerances 0.05% and 0.002%, 2,3 and 4 decimal places used, sufficient for accuracy required. Uncertainty accounted for in validation data. Volume – Pipettes used for calibration standards calibrated to <1%. Uncertainty accounted for in validation data. Temperature – Thermometer to measure rotary evaporator water bath temperature calibrated to <1oC. Uncertainty accounted for in validation data. Timer – Timer for HPLC flow calibrated to < 2 sec. Uncertainty accounted for in validation data.

iv







Environmental conditions: Contamination is minimised by the use of dedicated accommodation, equipment and glassware for organic analysis. Glassware is also separated during cleaning – uncertainty accounted for in validation data. Computational Effects: Integration of peaks by means of instrument software. Concentrations calculated by means of external standard calibration using instrument integrations. Manual checks of peak integrations are made for each sample, negligible contribution. Blank Correction: A procedural blank is analysed with each batch of samples. No contribution to uncertainty. Operator Effects: Only trained personnel may perform method unsupervised. Variations between operators are accounted for by control chart data. Uncertainty accounted for in validation data. Random Effects: These will be accounted for by validation data. Summary of validation data Combined uncertainty: Systematic component:% recovery of spike= Y. Y-100 = Z/2% = CsSpike added to all biota is 40 ng (50 ng for PBDE209) and for sediment is 20 ng (25 ng for PBDE209). Random component (CV% Shewhart chart) = Cr(random component of GC-MSD2 is from CV% of precision of biota and sediment samples as no Shewhart chart data available) Assume linear summation and a value of K=2: Combined standard uncertainty = (Cs

2 + Cr2)0.5 ng

Expanded uncertainty = 2*(Cs

2 + Cr2)0.5 ng

The reported expanded uncertainty is based on an uncertainty multiplied by a coverage factor of K= 2, providing a level of confidence of approximately 95%.

v

GC-MSD2 (EN751) FISH LIVER Compound Systematic

component Random Component

Expanded Uncertainty

% % % BDE17 -7.31 14.32 32.16 BDE28 -6.43 14.67 32.04 BDE75 -5.54 16.01 33.88 BDE49 -4.65 14.84 31.11 BDE71 -3.36 13.79 28.39 BDE47 -4.52 14.66 30.68 BDE66 -3.79 14.95 30.85 BDE77 -3.72 15.48 31.84 BDE100 -4.54 14.98 31.31 BDE99 -3.76 14.54 30.04 BDE85 -3.64 14.36 29.63 BDE154 -4.83 14.51 30.59 BDE153 -5.38 14.12 30.23 BDE138 -6.37 14.35 31.41 BDE190 -7.72 14.45 32.77 BDE183 -14.87 16.94 45.08 BDE209 -16.45 32.04 72.04 GC-MSD2 (EN751) FISH MUSCLE Compound Systematic



% % % BDE17 -9.76 8.93 26.46 BDE28 -8.99 9.82 26.62 BDE75 -9.54 10.36 28.17 BDE49 -9.10 9.90 26.89 BDE71 -5.07 20.02 41.31 BDE47 -8.75 9.11 25.27 BDE66 -8.50 9.28 25.17 BDE77 -7.94 9.45 24.69 BDE100 -7.80 8.99 23.80 BDE119 -7.91 8.73 23.55 BDE99 -8.15 9.77 25.45 BDE85/154 -8.45 10.02 26.20 BDE153 -8.57 10.14 26.55 BDE138 -10.25 11.97 31.51 BDE190 -10.53 10.82 30.19 BDE183 -12.11 15.25 38.94 BDE209 -9.70 31.36 65.65

vi

GC-MSD2 (EN751) MUSSEL Compound Systematic



% % % BDE17 -14.32 4.23 29.86 BDE28 -12.77 4.33 26.97 BDE75 -12.33 3.16 25.45 BDE49 -10.15 4.12 21.90 BDE71 -5.86 22.41 46.33 BDE47 -12.28 3.51 25.54 BDE66 -11.68 2.40 23.84 BDE77 -11.09 1.96 22.52 BDE100 -11.38 4.18 24.25 BDE119 -11.38 3.16 23.63 BDE99 -11.79 3.03 24.36 BDE85/154 -11.85 2.52 24.24 BDE153 -11.70 3.88 24.66 BDE138 -13.38 7.81 30.99 BDE190 -13.58 7.41 30.94 BDE183 -15.36 12.45 39.54 BDE209 -15.51 29.05 65.86 GC-MSD2 (EN751) SEDIMENT Compound Systematic

component Random Component Expanded

Uncertainty % % % BDE17 -12.29 13.07 35.87 BDE28 -11.28 14.77 37.17 BDE75 -12.28 14.87 38.57 BDE49 -11.60 14.53 37.20 BDE71 0.06 14.95 29.90 BDE47 -8.17 13.58 31.69 BDE66 -8.69 13.61 32.30 BDE77 -10.13 17.39 40.26 BDE100 -12.22 20.43 47.61 BDE119 -4.70 18.01 37.23 BDE99 -9.50 16.50 38.08 BDE85/154 -9.31 15.83 36.73 BDE153 -7.93 15.39 34.62 BDE138 -7.89 15.41 34.63 BDE190 -7.49 15.69 34.78 BDE183 -9.90 18.07 41.21 BDE209 -7.57 47.26 95.72

vii

Summary of validation data – raw data archived as Batch 2806 BIOTA

recovery in spiked fish liver %

recovery in spiked fish muscle %

recovery in spiked mussel %

Precision of biota (CRM) sample (CV%)

BDE-17 85.38 80.48 71.37 21.8 BDE-28 87.13 82.02 74.46 13.7 BDE-75 88.92 80.92 75.34 22.0 BDE-49 90.69 81.80 79.70 12.2 BDE-71 93.28 89.86 88.28 21.4 BDE-47 90.96 82.50 75.45 3.5 BDE-66 92.41 83.00 76.64 9.0 BDE-77 92.56 84.12 77.82 13.5 BDE-100 90.92 84.40 77.23 11.3 BDE-119 92.49 84.19 77.23 BDE-99 92.72 83.70 76.41 6.2 BDE-85/154 90.33 83.11 76.30 11.5 BDE-153 89.24 82.86 76.60 7.7 BDE-138 87.25 79.50 73.24 23.3 BDE-183 84.56 78.95 72.83 20.0 BDE-190 70.27 75.78 69.28 23.1 BDE-209 67.09 80.60 68.98

LOD (spiked biota) (ng/g wet weight)

90% STD %CV

10% STD %CV

precision of spiked fish liver %CV

precision of spiked fish muscle %CV

precision of spiked mussel %CV

PBDE-17 0.05 8.06 17.88 14.32 8.93 4.23 PBDE-28 0.05 7.63 18.85 14.67 9.82 4.33 PBDE-75 0.06 7.66 13.80 16.01 10.36 3.16 PBDE-49 0.05 7.49 14.28 14.84 9.90 4.12 PBDE-71 0.06 8.92 13.41 13.79 20.02 22.41 CB-198 0.07 7.41 12.82 12.24 6.90 5.05 PBDE-47 0.07 6.77 15.73 14.66 9.11 3.51 PBDE-66 0.05 7.40 16.13 14.95 9.28 2.40 PBDE-77 0.05 7.38 15.14 15.48 9.45 1.96 PBDE-100 0.05 8.40 23.24 14.98 8.99 4.18 PBDE-119 0.05 7.93 24.66 14.54 8.73 3.16 PBDE-99 0.06 8.62 17.25 14.36 9.77 3.03 PBDE-85/154 0.10 8.16 16.26 14.51 10.02 2.52 PBDE-153 0.06 8.68 15.52 14.12 10.14 3.88 PBDE-138 0.05 9.46 18.42 14.35 11.97 7.81 PBDE-183 0.06 9.36 16.85 14.45 10.82 7.41 PBDE-190 0.05 10.42 16.70 16.94 15.25 12.45 PBDE-209 0.75 9.55 31.33 32.04 31.36 29.05

viii

SEDIMENT

LOD (spiked low matrix sediment (ng/g DW)

recovery of spiked sediment %

precision of spiked sediment %CV

PBDE-17 0.03 75.43 13.07 PBDE-28 0.03 77.44 14.77 PBDE-75 0.03 75.43 14.87 PBDE-49 0.03 76.79 14.53 PBDE-71 0.03 100.11 14.95 PBDE-47 0.03 83.65 13.58 PBDE-66 0.03 82.61 13.61 PBDE-77 0.03 79.74 17.39 PBDE-100 0.03 75.55 20.43 PBDE-119 0.03 91.01 18.01 PBDE-99 0.09 81.00 16.50 PBDE-85/154 0.10 81.38 15.83 PBDE-153 0.03 84.09 15.39 PBDE-138 0.03 84.60 15.41 PBDE-183 0.03 84.26 15.69 PBDE-190 0.03 80.20 18.07 PBDE-209 1.56 84.86 47.26 Accelerated Solvent Extraction (ASE) of Sediments and Biota for the Determination of Trace Organics-SOP 0367 1. Introduction and Scope Weighed samples of sediment and biota are placed into extraction cells and extracted with iso-hexane for CBs, OCPs and PBDEs. Lipid removal from samples is achieved by the addition of 5% deactivated alumina to the extraction cells. 30 g alumina can be used to remove 300 mg of lipid of samples for PBDE analysis or CB analysis. If OCPs are also to be extracted less alumina should be used and further clean-up steps are needed. 2. Principles of the Method Samples are extracted on the Accelerated Solvent Extractor (ASE 300) under elevated temperatures and pressures. Solvent is added to an extraction cell and heated before being flushed into a collection bottle. 5% Deactivated alumina can be added to the extraction cell to remove lipid from biota samples. 3. Reference Materials In each batch of samples (maximum 12) a laboratory reference material and a procedural blank must be analysed. LRM 110 Danish Cod Liver Oil is analysed with biota samples (CBs, OCPs and PBDEs). LRM 140 is analysed with sediment samples (CBs and OCPs).

ix

The use of LRMS are recorded in organic reference material worksheet (B 102, CBs and OCPs only) 4. Reagents Anhydrous Sodium Sulphate (Analar, granular) Copper powder – Aldrich approximately 40 mesh iso-Hexane Rathburn Chemicals (aliphatic and aromatic hydrocarbon free) Acetone Rathburn Chemicals (aliphatic and aromatic hydrocarbon free) Alumina (Aluminium oxide 9 standardised) – Merck 1.01097.1000 Distilled water – ultra pure 50 µg kg-1 standard containing CB 35, CB 53, CB 112, CB 151, CB 198 and CB 209. If an alternative supplier is used, an equivalent grade of reagent is used 5. Equipment Dionex Sample Cells (various sizes) Dionex cellulose filters Dionex glass fibre filters Dionex cell filter insertion tool Sample collection bottles Balance (3 decimal place) Glass Beakers (various sizes) Glass jars and lids (250 ml) Spatulas Forceps Aluminium foil Solvent reservoir bottle Dionex ASE 300 (EN 1241) Calibrated syringe Ultra sonic bath Drying oven (EN 547) Measuring cylinders (various sizes) Glass stirring rod Test tubes Pastuer pipettes Pipettors Refrigerator Pestle Air Purifier

6. Environmental Control The extraction process is carried out on the bench in the spark proof room (Room 505) with the fume cupboards in operation. Measuring and transfer of solvent is undertaken in the fume cupboard. 7. Interferences PBDE and CB/OCP analysis requires a stable environment, avoiding contamination of samples and reagents eg. contact with fingers, dirty equipment. Cellulose filters are only handled with forceps. All glassware/ extraction cells are solvent washed with either acetone or iso-hexane prior to use. Nitrile gloves are to be worn while working in the laboratory.

x

http://nts5/scripts/qwbcgi.B 102

The addition and transfer of extracts and standards is by means of either glass pipette tips or calibrated glass syringe to prevent contamination. PBDE extraction is carried out with as little light in the room as possible - with the window blinds down and the lights off. Ensure the air purifier is switched on to reduce possible contamination from particulates in the air. 8. Sampling and Sample Preparation All samples should be defrosted at room temperature and exposure to direct sunlight or heat should be avoided. Once defrosted samples are cut up finely with a knife and/or homogenised using a Kenwood MiniChopper. Samples are then transferred glass jars using a solvent washed metal spatula. The remainder of the sample should be returned to the deep freeze as soon as weighing of all samples in the batch is completed. 9. Analytical Procedure 9.1 Cleaning of equipment/materials 9.1.1 Filters, Empty two packs of cellulose or glass fibre filters into a glass beaker

(500 ± 100 ml), add dichloromethane (250 ± 50 ml) and cover with aluminium foil. Sonicate for at least 10 minutes before discarding the solvent into a chlorinated waste solvent bottle and replace with iso-hexane (250 ± 50 ml). Sonicate for at least 15 minutes before emptying the solvent into a unchlorinated waste solvent bottle. Allow excess solvent to evaporate by standing the beaker in the fume cupboard for at least 30 minutes. Transfer the beaker containing the filters to a drying oven and dry at not more than 90oC for at least 2 hours. Transfer the cleaned filter to a clearly labelled solvent washed aluminium can.

9.1.2 All extraction cells, caps and collection bottles are solvent washed with

acetone or iso-hexane followed by iso-hexane, with the latter being allowed to evaporate. The collection bottles lids are fitted with ultra low bleed septa which have been solvent washed with iso-hexane. Cells and caps are numbered 1 – 12 (s, small, m, medium and L, large). Ensure each cell is fitted with the appropriately numbered lid.

9.2 Sample preparation 9.2.1 Sediment sample Preparation 9.2.1.1 Sediment samples are freeze dried and sieved as per SOPs

(SOP 0110, SOP 0120) 9.2.1.2 Into a solvent washed glass jar weigh the appropriate amount of

sediment: suggested weights for different sediment types are shown below.

Clean offshore marine sediment: 20-40 g Coastal or estuarine sediment: 15-25 g Sludges or spoils: 5-15 g Highly contaminated sediment sites: 5-15 g LRM 140: 10 ± 1 g

xi



9.2.1.3 Solvent washed sodium sulphate (SOP 1643) should be added to the

sediment in the ratio of ~ 1:1. 9.2.1.4 Record weights in the organic worksheet B 244 (CBs and OCPs) or B 314 (PBDEs)

9.2.1.5 Mix the sample and sodium sulphate using a solvent washed spatula. Add recovery standard to the jar as per section 9.3.

9.2.1.6 Place lid on glass jar and store overnight in a refrigerator.

9.2.1.7 Steps 9.2.1.2 - 9.2.1.6 are repeated for all samples . The procedural blank consists only of sodium sulphate and recovery standard.

9.2.1.8 Remove from the fridge and grind with a pestle wrapped in aluminium foil or a spatula. 9.2. 2 Biota sample preparation. 9.2.2.1 The % lipid of biota samples should be determined first using the Smedes method (M 0890).

9.2.2.2 For PBDE or CB/OCP analysis an appropriate amount of sample equivalent to 300 mg of lipid is used in the extraction. The weight of tissue required is determined in worksheet B 245 for CBs and B 314 (PBDEs)

Approximate weights for various tissue samples are shown below: Fish muscle, lean: 10-15 g Fish muscle, fatty: 0.5 - 2 g Fish liver: 0.5-3 g Mussel: 8-12 g Cod liver oil: 0.1- 0.3 g LRM 110: 0.5 g ± 0.2 g The sample is weighed in a glass jar and the weight recorded in the

organic Worksheet B 244 for CBs/OCPs and B 314 for PBDEs. Individual samples are cut into small pieces using solvent washed, forceps, scalpel or scissors. Pooled samples may have been previously homogenised (SOP 130).

9.2.2.3 Solvent washed sodium sulphate (SOP 1643) is added to the sample to allow drying. Use between 20 and 40 g of sodium sulphate. Mix the sample and sodium sulphate using a solvent washed spatula. Add recovery standard to the jar as per section 9.3. The sample, recovery standard and sodium sulphate are mixed again with a spatula.

9.2.2.4 Place lid on glass jar and store overnight in a refrigerator.

9.2.2.5 Steps 9.2.2.2-9.2.2.4 are repeated for all samples. The procedural blank consists of only 30 ± 5 g sodium sulphate and recovery standard. The forceps, scalpel and scissors are rinsed with acetone and dried with blue roll between samples. The waste acetone is emptied into a non-chlorinated solvent waste bottle.

xii











9.2.2.6 Remove from the fridge and grind with a pestle wrapped in aluminium foil for at least 2 minutes. 9.3 Addition of recovery standard 9.3.1 The recovery standard is prepared using (SOP 290), (SOP 310) and

(SOP 320). 9.3.2 Break open ampoule containing recovery standard. Pour into a 10ml test tube and stopper. 9.3.3 Add 1ml recovery solution by means of a calibrated syringe into the jar containing the sample and sodium sulphate. Repeat for all samples. 9.4 Filling of extraction cells

9.4.1 The size of cell used will depend upon the amount of sample to be extracted,

for samples < 25 g the 66 ml cell should be used and > 25 g the 100 ml cell should be used. Note the procedural blank and LRM should be put in the same size extraction cell as the samples. It is permissible to use an alternative size of cell if necessary as long as the LRM and procedural blank are extracted in the same size cells.

9.4.2 Cells and caps are numbered, ensure cap and cell body numbers correspond.

Cells are filled as per schematic Figure 1.

Filter paper x 1

Figure 1. Schematic of ASE cell for the extraction of PBDEs, CBs and OCPs

from biota and sediment. 9.4.3 Unscrew the top cap from the cell body. Place 2 filters in the cell at a slight

angle (Dionex ASE 300 manual 3-6). Place the insertion tool over the filters and slowly push the insertion tool into the cell. Ensure the filter is in full contact with the cell.

Sodium sulphate (10 g)

Fat retainer 5% Alumina)

Filter paper x1

Sodium sulphate mixed with sample (sufficient to dry)

Fill any void with sodium sulphate

Filter paper x 2

xiii




Note: Do not place the filter in the bottom cap before installing the cap, this creates an improper seal and allows leaks. 9.4.4 To the cells add solvent washed sodium sulphate (SOP 1643), 10 ± 1g via a

funnel. Add 30 ± 1g, via a funnel, of 5% deactivated alumina (SOP 0430) if extracting PBDEs or CBs only from biota samples. If OCPs are to be extracted in addition to CBs from biota then use 15 ± 1g, via a funnel, of 5% deactivated alumina. If sediment is being extracted for PBDEs, CBs or OCPs then use 15 ± 1g, via a funnel, of 5% deactivated alumina.

9.4.5 Place another filter in the cell and push down on to the alumina using the

insertion tool. Add the sample to the cell (see section 9.2.1 and 9.2.2) via a funnel. Any void volume is filled using more sodium sulphate and the whole cell is tamped down using the insertion tool – top up with sodium sulphate if necessary before adding a further filter on top before hand tighten the top and bottom lids of the cell. DO NOT USE A WRENCH OR OTHER TOOL TO TIGHTEN THE CAP.

9.4.6 Proceed to extraction by ASE 300 SOP 0371. 10. Calculation of Results

Not relevant 11. Precision, Accuracy and Practical Detection Limits Not relevant. 12. Results Not relevant 13. Safety AI143

xiv




http://nts5/scripts/qwbcgi.exe/docfetch?id=210&db=safety

Brominated Flame Retardants (BFRs) - gov.scot · A Review of Brominated Flame Retardants (BRFS) in...

Documents

Transcript of Brominated Flame Retardants (BFRs) - gov.scot · A Review of Brominated Flame Retardants (BRFS) in...