Screening of replacement substances for the brominated flame retardants PBDE, HBCDD and TBBPA

(1)

Screening of replacement substances for the brominated flame retardants PBDE, HBCDD

and TBBPA

Jakob Gustavsson, Henrik Karlsson, Lutz Ahrens, Karin Wiberg

Rapport till Naturvårdsverket Överenskommelse NV-05635-17

Uppsala, 2018-07-06

Department of Aquatic Sciences and Assessment

(2)

NATIONELL

MILJÖÖVERVAKNING PÅ UPPDRAG AV NATURVÅRDSVERKET

ÄRENDENNUMMER AVTALSNUMMER PROGRAMOMRÅDE DELPROGRAM

NV-05635-17 2219-17-014 Miljögifter akvatiskt Utveckling och analys

Screening of replacement substances for the brominated flame retardants PBDE, HBCDD and TBBPA

Rapportförfattare

Jakob Gustavsson, SLU Henrik Karlsson, SLU Lutz Ahrens, SLU Karin Wiberg, SLU

Utgivare

Institutionen för vatten och miljö (IVM) Sveriges lantbruksuniversitet (SLU) Postadress

Box 7050, 750 07 Uppsala Telefon

018-671000 Rapporttitel och undertitel

Screening of replacement substances for the brominated flameretardants PBDE, HBCDD and TBBPA

Beställare Naturvårdsverket 106 48 Stockholm Finansiering

Nationell Miljöövervakning Nyckelord för plats

Screening, punktkällor Nyckelord för ämne Flamskyddsmedel

Tidpunkt för insamling av underlagsdata 2018-02-01 – 2018-04-30

Sammanfattning

Målet med projektet var att identifiera akvatiska punktkällor för flamskyddsmedel (FRs) till den svenska miljön.

Detta gjordes genom att mäta halter i olika typer av vattenflöden och beräkna dagliga flöden av FRs med hjälp av vattenmassaflöden. De verksamheter som undersöktes inkluderade avloppsreningsverk (n = 5), avfallsanläggningar (n = 4), flygplatser (n = 5), industrier (n = 2), dagvatten från stads/industriområden (n = 6) samt lantbruk (n = 2).

Vatten- (n = 42) och partikelprover (n = 42) samlades in under februari-april 2018 och analyserades för totalt 62 olika FRs från tre olika ämnesklasser: halogenerade FRs (HFRs), organofosfater (OPFRs) och polybromerade difenyletrar (PBDEs).

Totalt detekterades 34 olika FRs i minst ett prov. Flest antal olika FRs hittades i provet från en avfallsanläggning (Högbytorp; n = 23), följt av Skavsta flygplats (n = 17), Ärna flygplats (n = 16) och Ryaverket avloppsreningsverk (n = 16). I genomsnitt hittades flest flamskyddsmedel vid avfallsanläggningar (n = 15 ± 5), följt av avloppsreningsverk (n = 13 ± 3), dagvatten (n = 11 ± 2), flygplatser (n = 9 ± 6) och industri (n = 5 ± 1). TDCIPP var det FR som detekterades i flest antal prover (78 % av alla prover), följt av BDE66 (68 %), TEHP (57 %), TCIPP (57 %) och TBOEP (57 %).

Den totala halten FRs i proverna varierade mellan <MQL och 130 000 ng L^-1. Den högsta totala koncentrationen hittades i proverna från Skavsta (130 000 ng L^-1), Vivsta (11 000 ng L^-1), Högbytorp (6 900 ng L^-1) och Henriksdal (4 300-6 600 ng L^-1). Generellt bidrog OPFRs mest till de totala koncentrationerna med i genomsnitt 76 %, jämfört med halogenerade FRs (7 %) och PBDEs (6 %).

De beräknade flödena av FRs varierade mellan <MQL och 1.8 kg dag^-1. Fyra provtagningsplatser hade betydligt högre flöden än övriga platser. Tre av dessa fyra var avloppsreningsverk (Henriksdal: 1.2 - 1.8 kg dag^-1; Ryaverket:

0.67 kg dag^-1; Skebäcksverket: 0.18 kg dag^-1), vilket tyder på att avloppsreningsverk kan vara viktiga spridningsvägar av FRs till den svenska vattenmiljön. Med tanke på studiens screeningkaraktär så bör samtliga rapporterade koncentrationer och flöden tolkas med försiktighet.

(3)

Summary

The aim of this study was to identify aquatic point sources of flame retardants (FRs) to the Swedish environment. This was done by measuring FR levels in different types of water flows and by estimating daily fluxes of FRs using water flow data. The sampled sites included wastewater treatment plants (WWTPs, n = 5), waste treatment facilities (WTFs, n = 4), airports (n = 5), industries (n = 2), storm water from urban/industrial areas (n = 6), and agriculture (n = 2). Water (n = 42) and particulate (n = 42) samples were collected during February-April 2018 and analyzed for in total 62 target FRs from three different classes: halogenated FRs (HFRs), organophosphorus FRs (OPFRs) and polybrominated diphenyl ethers (PBDEs).

In total, 34 different FRs were detected in at least one sample. The highest number of FRs were detected in a sample from a WTF (Högbytorp, n = 23), followed by Skavsta airport (n

= 17), Ärna airport (n = 16), and Ryaverket WWTP (n = 16). The highest average number of detected FRs (± standard deviation) were found for WTFs (n = 15 ± 5), followed by WWTPs (n = 13 ± 3), storm water (n = 11 ± 2), airport (n = 9 ± 6), and industry (n = 5 ± 1).

The most frequently detected FR was TDCIPP (78% of all samples), followed by BDE66 (68%), TEHP (57%), TCIPP (57%), and TBOEP (57%).

Total bulk FR concentrations ranged between <MQL to 130 000 ng L

^-1

. The highest total bulk concentrations were found in the samples from Skavsta airport (130 000 ng L

^-1

), Vivsta …. (11 000 ng L

^-1

), Högbytorp WTF (6 900 ng L

^-1

) and Henriksdal WWTP (4 300-6 600ng L

^-1

). In general, OPFRs contributed the most to the total concentrations with on average 76% of the total bulk concentration followed by HFRs (7%), and PBDEs (6%).

FR fluxes ranged between <MQL and 1.8 kg day

^-1

. Four sites showed considerably higher

total fluxes than the other sites. Out of those, three sites were WWTPs (Henriksdal: 1.2-1.8

kg day

^-1

; Ryaverket: 0.67 kg day

^-1

; Skebäcksverket: 0.18 kg day

^-1

), indicating WWTPs as

important pathways of FRs to the Swedish environment. Considering the screening design

of this study, all reported concentrations and fluxes should be interpreted with care.

(4)

4

1. Introduction

Flame retardants (FRs) are chemicals used in various materials and products to decrease flammability and thereby provide fire protection. The use of FRs dates far back in history [1], but it is with the increased use of synthetic polymers (such as plastics) in recent decades that the use of FRs has grown substantially [2]. Nowadays, FRs are extensively used in combustible materials such as plastics and paper [2] and as a result, FRs are present in many everyday products, such as furniture, carpeting, electronics and building materials [3]. The total worldwide use of FRs amounts to more than 2.25 Million tons per year [4]. Out of this total volume, halogenated FRs (HFRs, containing bromine and/or chlorine) make up approx.

31%, while organophosphorus FRs (OPFRs, containing a phosphate group and sometimes also bromine/chlorine) make up approx. 18% [4]. The remaining FRs being used are mainly inorganic compounds such as aluminum trihydroxide. The main use of FRs is in the production of plastics and resins (~85% of total use), while rubber and textile products account for most of the rest [4, 5]. Unfortunately, FRs used in different products may leach to indoor and outdoor environments during production, use and disposal of the flame-amended products [3]. This results in a continuous exposure for humans to these types of chemicals. As an example, researchers in Norway detected 30 out of 37 targeted compounds when analyzing indoor air and dust from households and classrooms, illustrating the wide occurrence of FRs in the indoor environment [6]. Concentrations are often relatively low, meaning that the exposure is mostly found to be well below effect dose values [7-10]. Nevertheless, long-term, low dose exposure to FRs may still be of health concern [11].

As a consequence of leakage from various products, FRs are not only ubiquitously spread in the indoor environment but also in the outdoor environment [12]. In addition, many FRs have been found or predicted to exhibit hazardous properties such as being persistent in the environment or to have bioaccumulation potential [13, 14]. Many FRs are also suspected of being toxic to aquatic organisms [15]. One group of FRs that are relatively well-studied are the polybrominated diphenyl ethers (PBDEs). These have been used extensively in the past but are nowadays forbidden for use in new products as a result of their hazardous properties [14]. The ban of those and other legacy FRs (e.g. hexabromobiphenyl, BB-153) has instead led to the introduction of a large number of other FRs (hereafter referred to as alternative FRs) on the global market [16]. Despite the good intention of replacing hazardous FRs with alternatives, many of the alternative FRs have been found to have similar physicochemical properties as the banned FRs [17], and several of them (e.g. BTBPE, HBB and TNBP), have been detected in the environment [12, 18-20]. Therefore, it is important to monitor FRs in the environment, especially since information about what FRs that are being used by product manufacturers (and in what quantities) is not always readily available for researchers and the public.

We recently conducted a literature review on current use FRs [21], where we aimed at

identifying chemicals used as FRs. Within this work, we also developed a multicriteria model

(5)

and used it for prioritization of the identified FRs. The prioritization was based upon FR usage, time trends in the potential of exposure, environmental detection, and previous prioritization lists. The current study was designed as a follow-up study of the previous work with the aim to screen for priority FRs (n = 52) nearby potential point sources in order to determine which FRs are being released to the aquatic environment. A second aim was to calculate FR fluxes from each potential point source in order to identify major FR point sources/pathways to the environment.

2. Materials and methods

2.1 Chemicals

Target FRs were selected based on the prioritization presented in the previous work [21]. The

FRs ranked top-25 were included in the current screening, except ethylene bis-tetrabromo

phtalimide (EBTEBPI, CAS 32588-76-4), pentaerythritol (CAS 115-77-5), tetrabromobisphenol

A-bis(2,3-dibromopropyl ether) (TBBPA-BDBPE, CAS 21850-44-2), and 1,2-bis(2,3,4,5,6-

pentabromophenyl) ethane (DBDPE, CAS 84852-53-9), which were excluded because of

analytical challenges. In addition, 27 FRs that had previously been detected in Swedish rivers

and streams [22, 23] were selected together with 11 legacy compounds, resulting in a total of 62

target compounds. Names, acronyms and CAS-numbers for all targeted FRs are given in

Table 1.

(6)

6

Table 1 FR target compounds (sorted alphabetically by acronym). FRs selected through the previous prioritization work [21] are marked in bold.

Substance Acronym CAS

Halogenated FRs

2,4-Dibromophenol 24-DBP 615-58-7

2,6-Dibromophenol 26-DBP 608-33-3

2,4,6-Tribromophenol 246-TBP 118-79-6

2-Bromoallyl 2,4,6-tribromophenyl ether BATE na

Bis(2-ethyl-1-hexyl)tetrabromo phthalate BEH-TEBP 26040-51-7 1,2-Bis(2,4,6-tribromophenoxy)ethane BTBPE 37853-59-1 1,2-Dibromo-4-(1,2-dibromoethyl)cyclohexane DBE-DBCH 3322-93-8 Hexachlorocyclopentadienyl dibromocyclooctane DBHCTD 51936-55-1

2,2-Dibromovinylbenzene DBS 31780-26-4

Dechlorane Plus, anti isomer aDDC-CO 13560-89-9

Dechlorane Plus, syn isomer sDDC-CO 13560-89-9

2-Ethylhexyl 2,3,4,5-tetrabromobenzoate EH-TBB 183658-27-7

Hexabromobenzene HBB 87-82-1

4,5,6,7-Tetrabromo-1,1,3-trimethyl-3-(2,3,4,5-

tetrabromophenyl)indane OBTMPI 1084889-51-9

Pentabromobenzyl acrylate PBB-Acr 59447-55-1

Pentabromobenzylbromide PBBB 38521-51-6

Pentabromochlorocyclohexane PBCH 87-84-3

Pentabromoethylbenzene PBEB 85-22-3

Pentabromophenyl allyl ether PBPAE 3555-11-1

Pentabromotoluene PBT 87-83-2

Tetrabromobisphenol A TBBPA 79-94-7

1,2,5,6-Tetrabromocyclooctane TBCO 3194-57-8

1,2,3,4-Tetrabromo-5chloro-6-methylbenzene TBCT 39569-21-6

Allyl 2,4,6-tribromophenyl ether TBP-AE 221-913-2

2,3,5,6-Tetrabromo-p-xylene TBX 23488-38-2

Tetrachlorobisphenol-A TCBPA 27360-90-3

Organophosphorus FRs

Bisphenol A bis(diphenyl phosphate) BADP 5945-33-5

2-Ethylhexyl diphenyl phosphate EHDPP 1241-94-7

Cresyl diphenyl phosphate CDP 26444-49-5

ortho-Tritolyl phosphate o-TMPP 1330-78-5

meta-Tritolyl phosphate m-TMPP 1330-78-5

para-Tritolyl phosphate p-TMPP 1330-78-5

Resorcinol bis(diphenyl phosphate) PBDPP 57583-54-7

Tri(2-chloropropyl) phosphate T2CPP 6145-73-9

Tri(3-chloropropyl) phosphate T3CPP 26248-87-3

Tri(2-butoxyethyl) phosphate TBOEP 78-51-3

(7)

Tris(4-tert-butylphenyl) phosphate TBPP 78-33-1

Tris(2-chloroethyl) phosphate TCEP 115-96-8

Tri(1-chloro-2-propyl) phosphate TCIPP 13674-84-5

Tris(1,3-dichloro-isopropyl) phosphate TDCIPP 13674-87-8

Tris(2-ethylhexyl) phosphate TEHP 78-42-2

Triisobutyl phosphate TiBP 126-71-6

Tri(2-isopropylphenyl) phosphate TiPPP 64532-95-2

Tri-n-butyl phosphate TNBP 126-73-8

Tripentyl phosphate TPeP 2528-38-3

Triphenyl phosphate TPHP 115-86-6

Tripropyl phosphate TPP 513-08-6

Tris(tribromoneopentyl) phosphate TTBNPP 19186-97-1

Legacy halogenated FRs

2,4,4’-Tribromophenyl ether BDE28 41318-75-6

2,2',4,4'-Tetrabromodiphenyl ether BDE47 5436-43-1

2,3′,4,4′-Tetrabromodiphenyl ether BDE66 189084-61-5

2,2',3,4,4'-Pentabromodiphenyl ether BDE85 82346-21-0

2,2',4,4',5-Pentabromodiphenyl ether BDE99 32534-81-9

2,2',4,4',6-Pentabromodiphenyl ether BDE100 189084-64-8

2,2',4,4',5,5'-Hexabromodiphenyl ether BDE153 68631-49-2

2,2′,4,4′,5,6′-Hexabromodiphenyl ether BDE154 207122-15-4

2,2',3,4,4',5',6-Heptabromodiphenyl ether BDE183 207122-16-5

Decabromodiphenyl ether BDE209 109945-70-2

3,3',4,4',5,5'-Hexabromobiphenyl BB-153 67774-32-7

2.2 Sampling sites and sampling

Water and suspended particulate matter (SPM) samples were collected directly from or downstream nearby potential point sources in mid-Sweden (Table 2). The potential point source categories included waste water treatment plants (WWTPs, n = 5), airports (n = 5), storm water (n = 6, including one meltwater sample from a snow dump), industries (n = 2), waste treatment facilities (WTFs, including landfills, n = 4), and agricultural land (n = 2). In addition, two rivers (n = 2), one of which showed comparably high FR levels in a previous river screening study [23] were sampled. The sampling was performed in January-April 2018, and the samples were analyzed at the POPs laboratory in the Department of Aquatic Sciences and Assessment (IVM) at the Swedish University of Agricultural Sciences (SLU), Uppsala.

Water (2-200 L, depending on the amount of particles) was pumped through glass fiber

filters (Whatman™ Glass Microfiber Filters GF/F™, 297 mm diameter, 0.7 μm pore size) in

the field. Two consecutive filters were used at each sampling site and used for the analysis of

particulate-bound FRs. An aliquot (approx. 12 L) of the filtered water was collected in

stainless steel cans and used for the analysis of apparently dissolved FRs (truly dissolved FRs

(8)

8

+ FRs bound to particles and dissolved organic matter passing through the filter, <0.7 μm).

Before sampling, the filtration equipment was rinsed with sample water by pumping water through the system for approx. 5 min before mounting the filter. Sample containers were rinsed three times with sample water before being completely filled. At two sampling sites (viz. Ärna and Volvo), bulk water samples were provided by others to us, and filtered in the laboratory. Flow-proportional water samples were collected at two of the WWTPs (Kungsängsverket and Henriksdal) and filtered in the laboratory together with the sample from the Högbytorp recipient. In total, 42 filter and 42 water samples were analyzed from in total 29 sites, including 12 duplicates, 10 field blanks, and 2 laboratory blanks.

In addition to samples for FR analysis, separate samples (1 L polypropylene bottles)

were collected for analysis of SPM and total organic carbon (TOC). Water temperature and

pH were measured in the field. Water samples were stored at +8°C, and filter samples were

stored at -18°C until analysis. Field water blanks consisted of Millipore water that was

brought to the sampling site, where the sample containers was opened shortly, and after this

the sample was treated the same way as a real sample. Field filter blanks were prepared by

exposing a clean filter shortly to the air at the sampling site before being wrapped and treated

the same way as a real sample. In addition, two laboratory blanks (one filter and one water

sample) were prepared by filtering approx. 10 L of Millipore water in the laboratory using the

same filtration equipment as in the field. The laboratory blanks were then further treated in

the same way as ordinary samples.

(9)

Table 2 Sampling sites.

Site Category City Coordinates Description Shooting range Agriculture Nyköping N6520123,

E611699

Stream draining agricultural land where biosludge have been applied to fields. Biosludge storage approx. 1 km from stream but no obvious connectivity between sludge storage and the sampled stream.

Sörtuna gård Agriculture Nyköping N6532721,

E642174 Ditch draining agricultural land where no biosludge have been applied to fields.

Halmsjöbäcken Airport Stockholm N6613878,

E662496 Stream downstream Arlanda airport.

Issjön Airport Göteborg N6390716,

E337335 Lake downstream Landvetter airport.

Issjöbäcken Airport Göteborg N6389055,

E337861 Stream downstream Landvetter airport.

Skavsta Airport Nyköping N6518284,

E611819 Outflow from Skavsta airport before water is reaching wetland.

ST airport, east Airport Sundsvall N6934027,

E626068 River Indalsälven. Sampled close to the outlet from Sundsvall-Timrå airport. Also downstream of e.g. the outlet from Vivsta stormwater pond.

ST airport, west Airport Sundsvall N6934270,

E625101 River Indalsälven. Sampled close to outlet from Sundsvall-Timrå airport. Also downstream of e.g. the outlet from Vivsta stormwater pond.

Ärna Airport

(military) Uppsala N6642812,

E644819 Outflow from Ärna military airport.

Volvo Car industry Göteborg N6401189,

E316760 The effluent from Volvo WWTP. Further treated in the municipal WWTP (Ryaverket).

Skutskär IN Pulp industry Uppsala N6725355,

E631112 River water (River Dalälven) intake for use during pulp production.

Skutskär OUT Pulp industry Uppsala N6726582,

E631369 Outflow from the pulp industry WWTP.

Nyköpingsån Re-sampled

river Nyköping N6522686,

E611114 Outlet from Lake Långhalsen. Located in an area where biosludge are applied to fields. Upstream of outlet from Skavsta airport.

Snow dump Snow dump Uppsala N6642221,

E648788 Meltwater from city snow dump site.

Järnbrottsdammen Stormwater Göteborg N6393236,

E316930 Outflow from sedimentation pond. Mixed catchment with ~60 000 vehicles day^-1, larger roads, residential and industrial areas.

Skebäcksdammen Stormwater Örebro N6570417,

E514666 Pond draining mixed urban catchment with some industries.

Essingeleden Stormwater Stockholm N6578577,

E671281 Well mainly draining a heavily trafficked road (Essingeleden) with a yearly average of 154 000 vehicles per day.

Uppsala

stormwater pond Stormwater Uppsala N6636651,

E653034 Sedimentation pond. Catchment with larger roads, retail and car workshops.

Vivsta Stormwater Sundsvall N6932966,

E618594 Catchment draining industrial area and forest. Sample taken before sedimentation pond.

Hovgården WTF Uppsala N6647132,

E655068 Outflow from WTF's WWTP. Household waste, biosludge storage, landfill for non-hazardous waste, sorting of e.g. industrial waste and construction material.

Fläskebo WTF Göteborg N6397378,

E330429 Landfill for mainly surface materials (Swedish: schaktmassor). Outlet from sedimentation pond.

Fortum WTF Örebro N6552930,

E515570 Downstream Fortum waste solutions WTF with e.g. high-temperature incineration, landfill and treatment of hazardous waste.

Högbytorp WTF Stockholm N6604154,

E647824 Landfill leachate before reaching sedimentation pond. Permission for landfill of both hazardous and non-hazardous waste.

(10)

Högbytorp

recipient WTF Stockholm N6602735,

E649007 Downstream Högbytorp WTF.

Kungsängsverket WWTP Uppsala N6636923,

E648871 Outflow from Uppsala's main WWTP, treating around 2200 m³/h.

Storvreta WWTP Uppsala N6650589,

E650635 Outflow from Storvreta WWTP with capacity of treating 220 m³/h.

Ryaverket WWTP Göteborg N6399261,

E314830 Outflow from Göteborg's main WWTP, treating around 14 000 m³/h.

Skebäcksverket WWTP Örebro N6570596,

E514822 Outflow from Örebro's main WWTP, treating around 2000 m³/h.

Henrikdal WWTP Stockholm N6578832,

E676949 Outflow from Stockholm's main WWTP, treating around 11 000 m³/h.

WTF: waste treatment facility: WWTP: waste water treatment plant

(11)

2.3 Analysis

All samples were analyzed at the POPs-lab at IVM (SLU, Uppsala), except TOC samples that were analyzed by the Geochemistry laboratory at IVM (SLU, Uppsala). The samples were analyzed according to methods described previously [22] with some minor modifications. Briefly, the extraction of water samples (apparently dissolved phase) was performed using solid phase extraction (SPE) with Oasis

^®

HLB cartridges (6 g, Waters, Massachusetts, USA). Approx. 15 h before sample extraction, samples were fortified with 100 µL isotopically labeled internal standard (IS) mixture (c = 200-1000 pg μL

^-1

) followed by vigorous shaking. Before sample loading, cartridges were preconditioned with 20 mL methanol (MeOH) followed by 20 mL MilliPore water.

After the whole water sample (approx. 10 L) had been loaded, the SPE cartridge was dried by centrifugation (1000 rpm for 10 min). The sample container was rinsed three times with 50 mL dichloromethane (DCM), which was subsequently used for elution of the cartridge. Prior to the first rinsing, another 10 mL DCM was used to rinse the inner wall of the SPE cartridge. Thus, the total elution volume was 160 mL. Extracts were further dried using a Horizon DryDisk

^®

membrane followed by evaporation to ~1 mL using a TurboVap

^®

II evaporator. Clean-up was performed using alumina as previously described [24] with the addition of ~1 g sodium sulphate on top of the alumina to achieve close to water-free samples. Finally, extracts were solvent exchanged into toluene using an N-Evap

^®

solvent evaporator and concentrated to ~0.5 mL. Prior to analysis, extracts were fortified with 10 μL of a recovery standard (RS, c = 1000 pg μL

^-1

). For a few samples, the clean-up procedure was repeated (water samples from Skavsta, Kungsängsverket, shooting range, Högbytorp, Vivsta, Volvo, and Issjön; filter samples from: Skavsta and Vivsta) because of poor peak shape in the GC-analysis (likely due to matrix interferences).

Filter samples were extracted using Soxhlet Dean-Stark extraction with toluene for >20 hours.

Before extraction, 100 μL of IS mixture (c = 200-1000 pg μL

^-1

) were fortified directly onto the filter placed in the Soxhlet. The two filters from each site were extracted together. After extraction, the extracts were treated in the same way as described for the water sample extracts with the exception that no drying with DryDisk was done on filter extracts as the water content was limited.

Instrumental analysis was done according to a previously published method [24] using a gas chromatograph (GC) coupled to tandem mass spectrometry (MS/MS; Agilent Technologies, GC 7890A Triple Quad 7010). The column was a 15 m DB-5MS column (J&W Scientific, Agilent Technologies, i.d. 250 μm, film thickness 0.10 μm). The constitutional isomers TCIPP, T2CPP and T3CPP were quantified as sum of compounds (i.e. ΣTCIPP/T2CPP/T3CPP), as they could not be separated due to co-elution and similar fragments.

2.4 Quality assurance/Quality control (QA/QC)

A signal-to-noise (S/N) ratio ≥3 with a matching ion ratio (ion ratio within ±20) between

sample and calibration was required for positive detection for both target compound and internal

standard. Retention time in sample and calibration solution was required to match (within ± 0.1

min) for both target compound and internal standard. The isotope dilution quantification method

was used to correct for losses occurring during sample preparation and analysis. In general,

recoveries of internal standards were highly variable (Tables A1 and A2 in Appendix), likely as a

(12)

12

consequence of the complexity of the various matrices sampled. Therefore, a threshold of 5%

recovery of the IS was used as a criteria for reporting. Following the high variability in recoveries, all presented concentrations and fluxes should be interpreted with care.

Field blanks (water, n = 5; filter, n = 5) were used to calculate method detection limits (MDLs) using the following formula:

𝑀𝐷𝐿 = 𝑎𝑣𝑒𝑟𝑎𝑔𝑒 𝑐𝑜𝑛𝑐𝑒𝑛𝑡𝑟𝑎𝑡𝑖𝑜𝑛 𝑜𝑓 𝑏𝑙𝑎𝑛𝑘𝑠 + 3 ∗ 𝑠𝑡𝑎𝑛𝑑𝑎𝑟𝑑 𝑑𝑒𝑣𝑖𝑎𝑡𝑖𝑜𝑛 𝑜𝑓 𝑏𝑙𝑎𝑛𝑘𝑠

If no peaks were found in the blanks, the lowest point in the calibration curve was used as the MDL. Method quantification limits (MQLs) were calculated from MDL with the following formula:

𝑀𝑄𝐿 = 𝑀𝐷𝐿 3 ∗ 10

The laboratory blanks were assessed as ordinary samples with the obtained concentrations

presented together with the ordinary samples. The average field blank levels ranged between not

detected to 300 ng absolute for both water and filter samples. The method detection limits (MDLs)

ranged from 0.010 to 140 ng L

^-1

and from 0.00058 to 39 ng L

^-1

for water and filter samples,

respectively (Table 3).

(13)

Table 3 Blank concentrations and method detection limits (MDLs).

^a

Water Filter

Compound Average blank (ng absolute, n = 5)

MDL (ng L^-1)

Average blank (ng absolute, n = 5)

MDL (ng L^-1)

HFR

24-DBP ND 3.8 ND 0.88

26-DBP ND 1.9 ND 0.44

aDDC-CO ND 0.010 ND 0.0022

ATE ND 0.029 ND 0.0066

BATE ND 0.029 ND 0.0066

BB-153 ND 0.029 4.7 0.26

BEH-TEBP ND 1.9 ND 0.44

BTBPE ND 0.24 ND 0.055

DBE-DBCH ND 0.010 ND 0.0022

DBHCTD ND 0.24 ND 0.055

DBS ND 0.24 ND 0.055

EHTBB ND 0.24 ND 0.055

HBB ND 0.010 ND 0.0022

OBTMPI ND 7.6 ND 1.8

PBB-Acr ND 0.24 0.13 0.020

PBBBr ND 1.9 ND 0.44

PBEB ND 0.029 ND 0.0066

PBPAE ND 0.24 ND 0.055

PBT ND 0.010 0.048 0.0048

sDDC-CO ND 0.048 ND 0.011

TBBPA ND 19 ND 4.4

TBCO ND 0.24 ND 0.055

TBCT ND 0.029 ND 0.0066

TBP 300 140 6.2 0.95

TBX ND 0.010 ND 0.0022

TCBPA ND 19 0.72 0.050

OPFR

BADP 17 7.3 0.23 0.022

CDP ND 1.9 ND 0.44

EHDPP 2.6 1.6 0.013 0.0020

mTMPP ND 0.010 ND 0.0022

oTMPP ND 0.010 0.0059 0.00058

pTMPP ND 0.24 300 39

RDP ND 7.6 ND 1.8

TBOEP ND 0.24 ND 0.055

TBPP ND 0.010 ND 0.0022

TCEP 1.3 0.86 4.1 0.36

TCIPP 120 53 46 3.2

TDCIPP 1.2 0.40 1.7 0.16

TEHP ND 0.010 ND 0.0022

TIBP ND 0.010 ND 0.0022

TiPPP 9.7 2.9 0.89 0.086

TNBP 75 28 9.2 0.85

(14)

14

TPeP ND 19 ND 4.4

TPHP 18 6.9 11 0.55

TPP ND 0.029 0.37 0.057

TTBNPP ND 0.029 ND 0.0066

PBDE

BDE28 ND 0.029 4.4 0.24

BDE47 ND 0.048 0.041 0.0063

BDE66 ND 0.048 0.11 0.012

BDE85 ND 0.24 0.075 0.012

BDE99 ND 0.24 0.11 0.017

BDE100 ND 0.048 0.041 0.0063

BDE153 ND 0.24 ND 0.055

BDE154 ND 0.24 0.068 0.011

BDE183 ND 1.9 ND 0.44

BDE209 37 12 ND 0.88

a ND = not detected

2.5 Calculation of river fluxes

Fluxes of FRs were calculated based on the measured concentrations (>MQL) of individual

target compounds multiplied with the measured or modeled water mass flow for each sampling site

at the day of sampling (if available). The water mass flows were in most cases obtained from the

sampled facility’s internal measurements (for details, see Table A16 in Appendix). If the flow for

the specific day of sampling was unavailable, monthly/yearly flow data were used to estimate the

fluxes. For some sites (Issjön, Sörtuna, Vivsta, Uppsala stormwater pond, snow dump, and shooting

range), no flow data was available and thus no fluxes are reported for those sites. Please note that

the calculation of fluxes is subject to large uncertainties due to uncertainties in both measured

concentrations and in water flow over time. Thus, the estimated fluxes should be considered

indicative rather than conclusive.

(15)

3. Results and discussion

3.1 Detection frequency

At least one and up to 23 FRs were detected (>MDL) at each of the sampled sites (Figure 1).

The highest number of FRs were detected in the sample from the Högbytorp waste treatment facility

(WTF) (n = 23, sample taken upstream of sedimentation pond), followed by Skavsta airport (n =

17), Ärna airport (n = 16), Fortum WTF (n = 16) and Ryaverket WWTP (n = 16). When comparing

the different main categories of potential point sources, the highest average number of detected FRs

(± standard deviation) were found for WTFs (n = 15 ± 5), followed by WWTPs, (n = 13 ± 3), storm

water (n = 11 ± 2), airport (n = 9 ± 6), industry (n = 5 ± 1), river (n = 4 ± 1) and agriculture

(n = 1 ± 0). Two FRs (TDCIPP and TEHP) were detected (>MDL) in the lab blank sample but at

comparably low concentrations (< 1.1 ng L

^-1

).

(16)

16

Figure 1 Number of detected FRs (>MDL) at each site (filter + water). Orange bar = agricultural sites, light blue = airport sites, green = industrial sites, yellow = river sites, grey = stormwater sites, dark blue = waste treatment facility (WTF) sites, black = waste water treatment plant (WWTP) sites, and purple = the laboratory blank sample

.

0 5 10 15 20 25

Shooting range Sörtuna Halmsjöbäcken Issjöbäcken Issjön Skavsta Skavsta ST airport, east ST airport, west Ärna Skutskär OUT Skutskär OUT Volvo Volvo Nyköpingsån Skutskär IN Essingeleden Järnbrottsdammen Järnbrottsdammen Skebäcksdammen Snow dump Uppsala stormwater pond Vivsta Fläskebo Fortum Hovgården Hovgården Högbytorp Högbytorp recipient Henriksdal (flow-int) Henriksdal (grab) Kungsängsverket Kungsängsverket Ryaverket Skebäcksverket Storvreta Lab blank

Number of detected FRs

(17)

The five most commonly detected FRs (>MDL) were TDCIPP (OPFR, 78% of all samples), BDE66 (PBDE, 68%), TEHP (OPFR, 57%), TCIPP (OPFR, 57%), and TBOEP (OPFR, 57%) (Figure 2). The most commonly detected HFR was TCBPA (54%).

Figure 2 Detection frequency (%) of all detected FRs (>MDL). Orange bar = organophosphorus FRs (OPFRs), blue = halogenated FRs (HFRs), and green = polybrominated diphenyl ethers (PBDEs).

3.2 FR concentration

Total bulk FR concentrations (>MQL) ranged from <MQL to 130 000 ng L

^-1

. However, given that this study is a screening study, investigating samples from a wide range of locations with extremely varying matrices that highly affects the recoveries of analytes and internal standards (section 2.4), all presented concentrations should be interpreted with care. Moreover, a few samples (Tables A11-A14 in Appendix) showed concentrations exceeding the highest point of the calibration curve, resulting in additional uncertainty to the highest observed concentrations. The highest total bulk concentrations were found in the samples from Skavsta airport (130 000 ng L

^-1

), Vivsta stormwater pond (11 000 ng L

^-1

), Högbytorp WTF (6 900 ng L

^-1

), and Henriksdal WWTP (6600 and 4 300 ng L

^-1

) (Figure 3). Note that the two water samples from Skavsta appeared to contain high amounts of an organic solvent (possibly glycol from the deicing of airplanes) resulting in a total organic carbon (TOC) concentration of >800 mg L

^-1

. Extra care should thus be taken when interpreting the concentrations from Skavsta, as the high content of organics may have impacted the extraction and instrumental analysis. The laboratory blank samples had a total FR concentration of 1.7 ng L

^-1

, which was considerably lower than most samples. In general, OPFRs contributed the most to the total concentrations with on average 76% of the total bulk concentration followed by HFRs (on average, 7%) and PBDEs (on average, 6%).

0 10 20 30 40 50 60 70 80 90

TDCIPP BDE66 TEHP TCIPP TBOEP TCEP EHDPP TCBPA PBT BDE99 BDE100 TIBP TNBP BADP TPHP TTBNPP mTMPP TiPPP 246TBP oTMPP BDE154 sDDC-CO TBX RDP BDE47 BDE153 TBPP PBB-Acr pTMPP BDE85 BDE183 ATE BB-153 TBBPA

De tec ti on fr eque nc y (% )

(18)

18

Figure 3 Total bulk FR concentrations (ng L^-1) for each of the sampled sites. Error bars represent the standard deviation (n = 2). A) Including Skavsta airport. B) Excluding Skavsta airport (for visibility).

0 20000 40000 60000 80000 100000 120000 140000 160000

Shooting range Sörtuna Halmsjöbäcken Issjöbäcken Issjön Skavsta ST airport, east ST flygplats, west Ärna Skutskär OUT Volvo Nyköpingsån Skutskär IN Essingeleden Järnbrottsdammen Skebäcksdammen Snow dump Uppsala stormwater pond Vivsta Fläskebo Fortum Hovgården Högbytorp Högbytorp recipient Henriksdal (flow-int) Henriksdal (grab) Kungsängsverket Ryaverket Skebäcksverket Storvreta Lab blank

T otal bulk conc entra ti on (ng L

-1

)

0 2000 4000 6000 8000 10000 12000

Shooting range Sörtuna Halmsjöbäcken Issjöbäcken Issjön ST airport, east ST flygplats, west Ärna Skutskär OUT Volvo Nyköpingsån Skutskär IN Essingeleden Järnbrottsdammen Skebäcksdammen Snow dump Uppsala stormwater pond Vivsta Fläskebo Fortum Hovgården Högbytorp Högbytorp recipient Henriksdal (flow-int) Henriksdal (grab) Kungsängsverket Ryaverket Skebäcksverket Storvreta Lab blank

T otal bulk conc entra ti on (ng L

-1

)

A

B

(19)

3.3 Composition profile

In total, 34 different FRs were detected (>MDL) in at least one sample. Out of the detected

FRs, 8 were alternative HFRs, 17 were OPFRs and 9 were legacy HFRs. Out of the 21 FRs that

were selected based on the previous prioritization list [21], 2 HFRs and 13 OPFRs were detected in

at least one sample. Legacy HFRs (PBDEs and BB-153) constituted in most cases a minor fraction

of the total amount FRs detected, showing the importance of other non-restricted, alternative FRs to

environmental levels and especially of OPFRs, which were the dominating class of FRs at most

sampled sites (Figure 4).

(20)

Figure 4 Compositional profile (including compounds detected >MDL) of ∑alternative HFRs, ∑OPFRs and ∑legacy HFRs (i.e. PBDEs and BB-153) for each site.

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Shooting range Sörtuna Halmsjöbäcken ST airport, west ST airport, east Issjöbäcken Issjön Skavsta Skavsta Ärna Volvo Volvo Skutskär OUT Skutskär OUT Skutskär IN Nyköpingsån Essingeleden Snow dump Järnbrottsdammen Järnbrottsdammen Skebäcksdammen Uppsala stormwater pond Vivsta Fläskebo Fortum Hovgården Hovgården Högbytorp Högbytorp recipient Henriksdal (flow-int) Henriksdal (grab) Kungsängsverket Kungsängsverket Ryaverket Skebäcksverket Storvreta Lab blank

∑Alternative HFRs ∑OPFRs ∑Legacy HFRs

(21)

When only comparing alternative HFRs (246-TBP, ATE, PBB-Acr, PBT, sDDC-CO, TBBPA, TBX,

TCBPA) with legacy HFRs (thus, excluding all OPFRs; BDE47, BDE66, BDE85, BDE99, BDE100,

BDE153, BDE154, BDE183, BB-153), the contribution of the alternative HFRs to the composition

profiles were highly variable with an average (± standard deviation) of 68 ± 36% of total FRs,

excluding OPFRs) (Figure 5). For two categories (agriculture and industry), no legacy HFRs were

detected. For the other categories, alternative HFRs contributed with, on average, 78 ± 35, 45 ± 36,

60 ± 40, and 56 ± 23% of total FRs (excluding OPFRs) for airports, stormwater, WTFs and WWTPs,

respectively.

(22)

Figure 5 Compositional profile (including compounds detected >MDL) of ∑alternative HFRs and ∑legacy HFRs (i.e. PBDEs and BB-153) from each site. Missing bar: no HFR detected (>MDL).

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

∑Alternative HFRs ∑Legacy HFRs

(23)

Many similarities in the composition profiles were found between the various samples (Figure 6: OPFRs, Figure 7: HFRs, Figure 8: PBDEs). For example, three out of the five sampled WWTPs (Ryaverket, Skebäcksverket and Storvreta) showed similar OPFR composition profiles with mainly TCIPP, TEHP, TDCIPP, TBOEP, and TCEP detected (Figure 6). The OPFR composition profiles from Henriksdal WWTP resembled the three previously mentioned WWTPs (but with a comparably large fraction of TNBP and no TDCIPP detected), while Kungsängsverket WWTP deviated from the rest with mainly two FRs (TDCIPP and TCIPP) dominating the profile. For the HFRs (Figure 7), Henriksdal, Ryaverket, and Skebäcksverket WWTPs showed identical composition profiles with only one HFR (TCBPA) detected. The two WWTPs located in Uppsala (Kungsängsverket and Storvreta) showed similarities in their HFR profiles, with PBT as the predominant compound. For PBDEs (Figure 8), all WWTPs showed similar composition profiles with mainly BDE66, BDE99, and BDE100 detected.

Regarding the WTFs, both Hovgården and Högbytorp had TCIPP as one of the major compounds in their OPFR profiles (Figure 6). However, one of the Hovgården replicates also showed a comparably large fraction of TCEP. TCEP was detected in both replicates (with a similar peak size) but was excluded from one of the samples due to low IS recovery. Thus, Hovgården and Högbytorp were different in terms of TCEP. The OPFR profile of Högbytorp recipient showed similarities to Högbytorp, indicating that the FRs detected in this stream may derive from the WTF;

however, total bulk concentrations in the recipient were approx. 230 times lower than at the WTF.

For the other two WTFs (Fortum and Fläskebo), the OPFR composition profiles were different, possibly reflecting treatment of other types of products. No obvious similarities between the WTFs could be observed in the HFR composition profiles (Figure 7). For PBDEs, Fortum and Högbytorp showed similar profiles, and BDE66 and BDE99 dominated the composition profile of all WTFs except Fläskebo (Figure 8). No PBDEs were detected in the Högbytorp recipient.

The stormwater samples from Essingeleden and the snow dump site showed similar OPFR composition profiles with TCIPP as the predominant FR together with TBOEP (Figure 6). The sample from Essingeleden derived mainly from a heavily trafficked road, and thus the similar profiles may indicate that the snow at the snow dump site was largely affected by traffic as well.

Also, Järnbrottsdammen storm water pond had a large fraction of TCIPP (visible in one of the replicates while the concentration in the other replicate were <MDL, despite a relatively large peak). Uppsala stormwater pond showed similarities to both Skebäcksdammen stormwater pond (dominated by TDCIPP, TEHP, and TCIPP) and Vivsta stormwater pond (dominated by TCIPP, BADP and TPHP), despite Skebäcksdammen and Vivsta showing relatively different profiles compared to each other. No obvious similarities between the stormwater sites could be observed in the HFR composition profiles (Figure 7). For the PBDE profiles, Essingeleden, Uppsala stormwater pond and Vivsta showed identical profiles with only BDE66 detected, while Järnbrottsdammen and Skebäcksdammen had similar profiles with BDE66, BDE99, and BDE100 detected (Figure 8).

For the airport samples, similar OPFR composition profiles were observed between Halmsjöbäcken and ST airport (west) with TCEP, TDCIPP and TEHP detected (Figure 6).

Interestingly, the outgoing water from Skutskär (Skutskär OUT) showed a more complex

composition profile with a higher number of FRs detected (up to 6 FRs) compared to the inlet water

(Skutskär IN, 3 FRs detected), indicating that a few FRs (mainly TCBPA (Figure 7) and TBOEP

(24)

24

(Figure 6)) are released from the pulp industry, however, the total concentrations were comparably

low (total bulk FR concentration < 29 ng L

^-1

).

(25)

Figure 6 Compositional profile (including concentrations >MDL) of OPFRs for each site. Missing bar: no OPFR detected (>MDL).

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

TDCIPP TEHP TCIPP TBOEP TCEP EHDPP TIBP TNBP BADP TPHP TTBNPP mTMPP TiPPP oTMPP RDP TBPP pTMPP

(26)

Figure 7 Compositional profile (including concentrations >MDL) of HFRs for each site. Missing bar: no HFR detected (>MDL).

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

246TBP ATE BB-153 PBB-Acr PBT sDDC-CO TBBPA TBX TCBPA

(27)

Figure 8 Compositional profile (including concentrations >MDL) of PBDEs for each site. Missing bar: no PBDE detected (>MDL).

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

BDE47 BDE66 BDE85 BDE99 BDE100 BDE153 BDE154 BDE183

(28)

28

3.4 Fluxes

FR fluxes were calculated based on measured concentrations (>MQL) and ranged for total bulk

FRs between <MQL and 1.8 kg day

^-1

. Again, the calculations of fluxes are associated with large

uncertainties, and thus fluxes need to be interpreted with care. Four sites showed considerably

higher total fluxes than the other sites (Figure 9). Not surprisingly, three out of the four sites with

the highest fluxes were WWTPs (Henriksdal: 1.8 kg day

^-1

(grab sample), 1.2 kg day

^-1

(flow-

integrated sample); Ryaverket: 0.67 kg day

^-1

; Skebäcksverket: 0.18 kg day

^-1

). Other sites with

comparably high fluxes were Skavsta airport (0.21 ± 0.27 kg day

^-1

), Kungsängsverket (0.011 ±

0.0030 kg day

^-1

), Högbytorp (0.0038 kg day

^-1

), and Storvreta (0.0037 kg day

^-1

).

(29)

Figure 9 Total bulk FR flux (g day^-1) for each site (for sites with available water mass flow data). A) Including all sites.

B) Excluding the five highest fluxes (for better visibility).

0 200 400 600 800 1 000 1 200 1 400 1 600 1 800 2 000

Ha lmsj öbä cke n Issjöb äc ken S ka vsta S T a irpor t, e ast S T a irpor t, we st Är na S kut skä r I N S kut skä r O UT Volvo Nyköpingså n Jä rnbrott sda mm en Essi ng elede n ᵟ Uppsa la st or m w ater pond S ke bä cksda mm en ᵟ F lä ske bo Hovgå rde n Högbytor p Hög by torp re cipi ent ᵟ F ortum He nriksda l (f low- int) He nrik sd al (gra b) Kungsä ngs ve rke t R ya ve rke t S ke bä cksve rke t S tor vre ta

T otal F R flux (g da y

-1

)

0 2 4 6 8 10 12 14 16

H al msjöbä cke n Issjöbä cke n ST a irp or t, e ast ST a irp or t, west Är na S kut skä r I N S kut skä r O UT Volvo Nyköpingså n Jä rnbrott sda mm en Ess in gele de n Uppsa la st or m w ater pond S kebä cks damm en F lä ske bo Hovgå rde n Högbytor p Högbytor p r ec ipi ent F ortum Kungsä ngs ve rke t S tor vre ta

T otal F R flux (g da y

-1

)

A

B

(30)

30

4. Conclusions

In total, 34 out of 62 targeted FRs were detected within this study, illustrating the vast

number of FRs being used in various products and leaching to the aquatic environment. Not

surprisingly, many FRs were detected in effluent water from WTFs, reflecting the end of life

treatment of many flame amended products. Many FRs were also detected in effluent water from

WWTPs, likely as result of leakage from flame amended products used in daily life. In general,

the highest FRs concentrations were associated with OPFRs, which often dominated the

compositional profiles. Similarly, 10 out of the 15 most frequently detected FRs were OPFRs,

again showing the importance of this group of compounds. Altogether, this indicates a wide use

and spread of OPFRs and warrants future studies on this group of chemicals. Interestingly, three

PBDEs were among the 15 most frequently detected FRs, showing that despite the ban, PBDEs

are still being emitted to the environment, although concentrations were generally low compared

to OPFRs. Finally, four out of the five sites with the highest fluxes were WWTPs, indicating the

importance of WWTPs as pathways of FRs to the environment.

(31)

31

References

1. Cho, Y.-J., Synthesis and evaluation of five-membered dioxaheterocycles containing flame-retarding elements as initiators for vinyl polymerization, in Science of Advanced Materials. 2011, Central Michigan University: Mount Pleasant, Michigan.

2. Guerra, P., et al., Introduction to brominated flame retardants: Commercially products, applications, and physicochemical properties. The handbook of environmental chemistry:

Brominated flame retardants, 2011. 16: p. 1-17.

3. Barber, J.L., et al., A review of the analysis of novel brominated flame retardants. Journal of Chromatography A, 2012. 1219(0): p. 15-28.

4. IHS consulting 2017, information retrieved from https://www.flameretardants-online.com, accessed 2018-05-15.

5. IHS consulting 2014, information retrieved from https://www.flameretardants-online.com, accessed 2017-10-02.

6. Cequier, E., et al., Occurrence of a Broad Range of Legacy and Emerging Flame Retardants in Indoor Environments in Norway. Environmental Science & Technology, 2014. 48(12): p. 6827-6835.

7. Ali, N., et al., Occurrence of alternative flame retardants in indoor dust from New Zealand: indoor sources and human exposure assessment. Chemosphere, 2012. 88(11): p. 1276-82.

8. Van den Eede, N., et al., Multi-residue method for the determination of brominated and organophosphate flame retardants in indoor dust. Talanta, 2012. 89: p. 292-300.

9. Fromme, H., et al., Organophosphate flame retardants and plasticizers in the air and dust in German daycare centers and human biomonitoring in visiting children (LUPE 3). Environ Int, 2014.

71: p. 158-63.

10. Abdallah, M.A.-E. and A. Covaci, Organophosphate Flame Retardants in Indoor Dust from Egypt:

Implications for Human Exposure. Environmental Science & Technology, 2014. 48(9): p. 4782-4789.

11. Bergman, A., et al., State of the Science of Endocrine Disrupting Chemicals - 2012. World Health Organization (WHO), 2012.

12. de Wit, C.A., D. Herzke, and K. Vorkamp, Brominated flame retardants in the Arctic environment - trends and new candidates. Science of the Total Environment, 2010. 408(15): p. 2885-918.

13. EFSA, Scientific Opinion on Emerging and Novel Brominated Flame Retardants (BFRs) in Food, in EFSA Journal 2012: European Food Safety Authority (EFSA).

14. Stapleton, H.M., et al., Novel and high volume use flame retardants in US couches reflective of the 2005 PentaBDE phase out. Environmental Science & Technology, 2012. 46(24): p. 13432-13439.

15. van der Veen, I. and J. de Boer, Phosphorus flame retardants: properties, production, environmental occurrence, toxicity and analysis. Chemosphere, 2012. 88(10): p. 1119-53.

16. Bergman, A., et al., A novel abbreviation standard for organobromine, organochlorine and organophosphorus flame retardants and some characteristics of the chemicals. Environment International, 2012. 49: p. 57-82.

17. Liagkouridis, I., A.P. Cousins, and I.T. Cousins, Physical–chemical properties and evaluative fate modelling of ‘emerging’ and ‘novel’ brominated and organophosphorus flame retardants in the indoor and outdoor environment. Science of The Total Environment, 2015. 524–525: p. 416-426.

18. Muir, D.C.G. and C.A. de Wit, Trends of legacy and new persistent organic pollutants in the circumpolar arctic: Overview, conclusions, and recommendations. Science of The Total Environment, 2010. 408(15): p. 3044-3051.

19. Martínez-Carballo, E., et al., Determination of selected organophosphate esters in the aquatic environment of Austria. Science of The Total Environment, 2007. 388(1–3): p. 290-299.

20. Schlabach, M., et al., Brominated flame retardants (BFR) in the Nordic environment. 2011, TemaNord: Norwegian Institute for Air Research (NILU), IVL Swedish Environmental Research Institute.

21. Gustavsson, J., et al., Replacement substances for the brominated flame retardants PBDE, HBCDD, and TBBPA, Report to Swedish EPA. 2017.

22. Gustavsson, J., et al., Seasonal trends in legacy and alternative flame retardants in river water in a

(32)

32

boreal catchment, manuscript published in doctoral thesis Acta Universitatis agriculturae Sueciae no. 2017:90, Levels and trends of flame retardants in the Swedish environment, Gustavsson J., Swedish University of Agricultural Sciences. In prep.

23. Gustavsson, J., et al., Screening of organic flame retardants in Swedish river water. Science of The Total Environment, 2018. 625: p. 1046-1055.

24. Gustavsson, J., et al., Development and comparison of gas chromatography–mass spectrometry

techniques for analysis of flame retardants. Journal of Chromatography A, 2017. 1481: p. 116-126.

(33)

Appendix

Table A1 Recovery (%) of internal standards (IS) in water samples.

Sample M-BDE47 M-BDE99 M-BDE100 M-BDE153 M-BDE154 M-BDE183 M-BDE209 M-TNBP M-TPHP M-HBB M-EHTBB M-aDDC-CO

Essingeleden 95 63 57 27 24 15 2 207 109 112 50 24

Field blank 197 93 52 1 0 1 6 153 167 124 0 31

Field blank 91 52 35 0 0 1 7 79 71 67 0 14

Field blank 169 89 60 2 6 0 4 248 231 94 4 22

Field blank 162 22 51 1 0 0 6 198 177 91 0 23

Field blank 93 59 74 23 35 17 0 59 60 102 79 22

Fläskebo 106 78 66 21 22 0 3 219 69 157 69 17

Fortum 58 60 106 40 44 34 1 49 52 68 71 41

Halmsjöbäcken 69 73 85 59 63 48 1 40 18 58 90 48

Henriksdal

(flow-int) 78 63 26 10 12 1 4 156 143 116 14 11

Henriksdal (grab) 134 26 87 7 10 0 12 396 231 78 0 11

Hovgården 61 47 35 17 18 14 2 98 56 69 48 12

Hovgården 71 56 59 36 37 33 5 47 59 70 44 20

Högbytorp 54 23 22 4 8 5 0 9 15 75 18 4

Högbytorp recipient 157 6 0 4 8 0 20 299 293 34 0 13

Issjöbäcken 99 36 27 2 10 0 2 126 143 68 3 10

Issjön 0 1 26 0 0 0 0 9 0 0 0 0

Järnbrottsdammen 79 61 57 35 41 28 2 60 46 74 40 33

Kungsängsverket 30 9 9 4 5 3 0 20 4 20 7 4

Lab blank 153 92 345 4 0 1 7 342 275 137 5 9

Nyköpingsån 82 80 91 50 58 37 2 35 48 80 83 36

Ryaverket 67 58 59 13 39 25 4 56 70 81 44 32

Shooting range 0 0 2 0 0 0 0 5 0 0 0 0

Skavsta 83 25 33 6 10 8 0 42 40 92 22 11

Skavsta 84 67 62 37 50 32 0 17 8 90 53 39

(34)

34

Skebäcksverket 80 20 58 0 1 2 10 182 139 52 1 10

Skutskär IN 121 32 48 2 1 0 3 56 126 86 3 15

Skutskär UT 92 55 52 12 17 13 14 594 81 131 49 9

Skutskär UT 64 44 26 11 9 5 7 800 84 50 10 6

Snow dump 202 58 48 0 1 0 0 439 310 106 0 23

ST airport, east 98 33 41 1 1 1 1 42 88 93 4 16

ST airport, west 117 53 63 2 0 1 19 131 143 116 0 15

Storvreta 116 21 34 5 1 1 3 105 118 98 2 13

Sörtuna 84 72 76 2 7 0 8 180 104 89 14 13

Uppsala

stormwater pond 115 21 56 0 0 0 3 112 140 97 1 29

Vivsta 105 22 37 2 7 0 0 15 13 56 10 6

Volvo 170 97 17 29 31 14 0 29 36 41 66 55

Volvo 170 58 61 13 19 16 0 49 32 80 45 19

Ärna 128 45 48 2 1 1 11 234 130 100 1 24

Skebäcksdammen 76 29 37 3 1 0 3 258 130 35 0 10

(35)

Table A2 Recovery (%) of internal standards (IS) in filter samples.

M-BDE47 M-BDE99 M-BDE100 M-BDE153 M-BDE154 M-BDE183 M-BDE209 M-TNBP M-TPHP M-HBB M-EHTBB M-aDDC-CO

Essingeleden 166 102 124 48 46 57 16 30 48 243 83 20

Field blank 78 87 89 98 98 104 62 0 1 90 91 110

Field blank 60 75 71 72 78 76 43 2 7 71 67 95

Field blank 74 83 80 85 91 99 79 1 12 83 82 100

Field blank 71 78 77 84 87 94 111 0 3 84 79 110

Field blank 63 75 76 84 88 87 62 8 19 60 78 100

Fläskebo 64 61 67 62 75 80 65 2 10 71 68 79

Fortum 81 76 70 50 57 90 509 58 51 122 72 84

Halmsjöbäcken 116 81 93 29 86 54 65 10 26 105 64 64

Henriksdal (flow-int) 288 237 265 249 284 192 348 100 189 384 242 205

Henriksdal (grab) 81 72 82 77 69 37 62 26 33 128 76 29

Hovgården 94 63 61 35 58 82 316 73 44 134 46 138

Hovgården 105 66 68 40 61 85 454 17 19 131 69 183

Högbytorp 60 59 58 54 55 80 131 4 12 62 68 63

Högbytorp recipient 43 60 38 68 49 80 2 2 17 57 74 50

Issjöbäcken 80 75 73 74 77 76 93 8 27 104 73 86

Issjön 107 36 62 95 83 69 135 4 26 146 47 108

Lab blank 89 87 99 102 112 106 123 0 1 105 93 134

Nyköpingsån 83 75 18 66 76 32 113 2 24 106 81 35

Ryaverket 94 92 111 92 85 38 2 59 43 111 98 15

Shooting range 74 76 80 79 79 63 57 0 3 90 77 86

Skavsta 83 107 66 107 103 180 1051 105 18 126 89 316

Skavsta 149 107 78 87 98 224 0 0 2 97 88 298

Skebäcksdammen 86 62 65 41 45 57 117 0 3 132 64 59

Skebäcksverket 80 62 76 48 73 48 104 2 15 112 66 46

Skutskär IN 70 77 67 82 86 82 61 10 28 86 79 92

(36)

36

Skutskär OUT 79 72 80 76 83 52 62 11 22 107 84 60

Snow dump 62 57 49 52 44 28 1.8 8.6 20 76 67 31

ST airport, east 95 92 90 92 98 73 112 0 15 124 97 93

ST airport, west 86 64 71 52 46 29 96 19 42 120 69 42

Storvreta 73 52 6 65 69 45 75 3 21 91 57 63

Sörtuna 80 78 82 52 87 79 79 0 7 91 83 95

Uppsala stormwater pond 70 48 42 37 38 72 221 11 12 121 49 78

Vivsta 0 123 34 53 51 148 0 0 8 0 5 184

Volvo 84 70 77 90 91 112 132 4 23 112 73 102

Volvo 86 58 74 89 89 107 101 2 20 110 62 80

Ärna 65 64 70 74 78 78 73 0 6 77 67 104

(37)

Table A3 Concentrations (>MDL, ng L

^-1

) of HFRs and PBDEs in water phase samples.

246TBP BB-153 PBB-Acr PBT TBBPA TBX TCBPA BDE47 BDE66 BDE99 BDE100 BDE209

Halmsjöbäcken <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL

Fläskebo <MDL <MDL <MDL 0.070 <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL

Järnbrottsdammen <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL 0.13 <MDL <MDL <MDL

Järnbrottsdammen 330 ^a <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL

Skebäcksverket <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL

Skebäcksdammen <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL

Fortum <MDL <MDL <MDL 0.077 <MDL <MDL <MDL <MDL 0.48 <MDL <MDL 23^a

Sörtuna <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL

Nyköpingsån <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL

Skavsta <MDL <MDL <MDL 9.1 <MDL <MDL <MDL <MDL 0.20 <MDL <MDL <MDL

Skavsta <MDL <MDL <MDL 0.51 <MDL <MDL <MDL <MDL 0.11 <MDL <MDL <MDL

Kungsängsverket <MDL 3.6 7.7 27 <MDL 0.47 <MDL 0.42 1.6 2.3 0.84 <MDL

Shooting range <MDL <MDL <MDL 7.3^a <MDL 1.4 <MDL <MDL <MDL <MDL <MDL <MDL

Högbytorp 230^a <MDL <MDL 0.15 <MDL 0.045 <MDL <MDL 0.65 0.52 <MDL <MDL

Högbytorp recipient <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL

Skutskär IN <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL

Skutskär OUT <MDL <MDL <MDL 0.24 24^a <MDL 24 <MDL <MDL <MDL <MDL <MDL

Skutskär OUT <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL

Essingeleden <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL

Henriksdal (flow-int) <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL Henriksdal (grab) <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL

Storvreta <MDL <MDL <MDL 0.73 <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL

Uppsala stormwater pond <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL

Snow dump <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL

ST airport, west <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL ST airport, east <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL

Vivsta <MDL <MDL <MDL 0.75 <MDL 0.051 <MDL <MDL 0.19 <MDL <MDL <MDL

Ärna 370^b <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL

Volvo <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL

Volvo 280^a <MDL <MDL 2.0 <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL

(38)

38

Hovgården <MDL <MDL <MDL 0.014 <MDL <MDL <MDL <MDL 0.40 <MDL 0.049 <MDL

Issjön <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL

Ryaverket <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL 0.17 <MDL <MDL <MDL

Issjöbäcken <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL

Kungsängsverket <MDL <MDL <MDL 0.094 <MDL 0.023 <MDL <MDL <MDL <MDL <MDL <MDL

Lab blank <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL <MDL

aExcluded from all calculations due to IS recovery <5%; ^bMore uncertain concentration due to exceedance of the calibration curve.