Occurrence of major perfluoroalkyl substances (PFAS) in water samples from a transect of the Indian Ocean along the East African coast

(1)

Örebro University, School of Science and Technology

Master of Chemistry in Environmental Forensics 2019-2020

Occurrence of major perfluoroalkyl substances (PFAS) in water

samples from a transect of the Indian Ocean along the East

African coast

Student: Kjell Hope

Main supervisor: Heidelore Fiedler Assistant supervisor: Leo Yeung Examiner: Mattias Bäckström

(2)

(3)

Occurrence of major perfluoroalkyl substances (PFAS) in water

samples from a transect of the Indian Ocean along the East

African coast

Table of Figures

Figure 1: The currents in the Mozambique Channel as well as the East African coast. The

different colours represent the speed of the currents. ... 5

Figure 2: Sampling locations from the IIOEC 2Figure x: Location of all samples taken during the IIOE2. Each dot represents one sampling location, wherein several samples at different depths were taken. ... 7

Figure 3: Samples analysed from Mozambique ... 11

Figure 4: Samples analysed from Tanzania ... 12

Figure 5: PFOA levels in surface water near the island of Zanzibar ... 17

Figure 7: PFOA concentrations in surface water in Northern Mozambique ... 17

Figure 8: Samples that showed presence of either PFOS or PFHxS besides PFOA ... 18

Figure 9: Variation of temperature with relation to depth below water surface ... 19

Figure 10: PFOS, PFOA and PFHxS levels in Thailand and Uganda ... 21

Figure 11: Graphic depicting the route and levels of PFAS found during the Oden and Endeavor cruises (Figure adapted from Benskin et al., 2012) ... 23

Figure 12: LC/MS chromatogram for native PFASs standard (UNEP-PFAS-SOLN. A). Peak J corresponds to br-PFOS and peak K to L-PFOS ... 29

Figure 13: Chromatogram: separation and quantification of L-PFOS and relevant br-PFOS isomers ... 29

Table of Tables

Table 1: LOQ levels achieved during the current project in pg and pg/L. ... 10

Table 2: Designation of samples as to location and sampling depth (Mozambique) ... 13

Table 3: Designation of samples as to location and sampling depth (Tanzania) ... 14

Table 4: Concentrations of priority PFASs in samples from Mozambique (pg/L) ... 15

Table 5: Concentrations of priority PFASs in samples from Tanzania (pg/L) ... 16

Table 6: Table of some extended parameters measured in the Tanzanian samples ... 19

Table 7: Isomeric composition of native PFOS standard (UNEP-PFAS-SOLN. A) (Wellington Laboratories, Guelph, Canada) ... 29

Table 8: Detected peaks in chromatogram (raw data) ... 30

Table 9: Recoveries of internal standard in samples ... 31

Table 10: Recoveries of internal standard in QC samples ... 31

Table 11: Individual and ∑PFAS concentration (pg/L) in surface ocean water samples obtained for the Malaspina 2010 expedition ... 31

(5)

Abbreviations

BEH Ethylene bridged hybrid

br-PFOS Branched perfluorooctanesulfonic acid

COP Conference of the Parties (here: of the Stockholm Convention on Persistent Organic Pollutants)

EPA Environmental Protection Agency FTCA Fluorotelomer carboxylic acid FTOH Fluorotelomer alcohols

HPLC High performance liquid chromatography IARC International Agency for Research on Cancer IIOEC2 Second International Indian Ocean Expedition

LC Liquid chromatograph(y)

LOD Limit of detection

LOQ Limit of quantification

L-PFOS Linear perfluorooctanesulfonic acid MS/MS Tandem mass spectrometer

MZ Mozambique

OECD Organisation for Economic Co-operation and Development PAH Polycyclic aromatic hydrocarbons

PFAA Perfluoroalkyl acids

PFASs Per- and polyfluoroalkyl substances PFASA Perfluoroalkyl sulfonamides

PFCA Perfluorinated carboxylic acid PFHxS Perfluorohexanesulfonic acid PFOA Perfluorooctanoic acid PFOS Perfluorooctanesulfonic acid PFOSF Perfluorooctanesulfonyl fluoride

(6)

PFSA Perfluoroalkane sulfonate POP(s) Persistent organic pollutant(s)

QC Quality control sample

SPE Solid phase extraction

TZ Tanzania

UNEP United Nations Environmental Program UPLC Ultra-performance liquid chromatography

WAX Weak anion exchange

(7)

1 Introduction

Nowadays, researchers from various disciplines work together to address global challenges in an attempt to link uses of chemicals and their pollution with effects on the environment and ecosystems. Among the organic chemicals, much attention has been given to atmospheric transport of polycyclic aromatic hydrocarbons (PAH) or chlorinated and brominated persistent organic pollutants. These lipophilic organic compounds, due to their hydrophobicity, have minor importance in aquatic media. On the other hand, coastal

environments are severely stressed in many parts of the world as a result of overpopulation, urbanization, excess resource use, and pollution. Climate change is expected to increase the implication of these stressors and could be considered one of the most important challenges facing the world in the 21st century. Monitoring marine mammals and their aquatic

environment gives a window into global trends for water-soluble persistent organic pollutants (POPs) concentrations that are occurring as a result of their use patterns of industrial and agrochemicals, and reduced biodiversity in the oceans. In addition, these chemicals are linked to changes in weather and climate events (Persistent Organic Pollutants in the Marine Food Chain - Our World, 2020). Marine mammals accumulate POPs in their fatty tissues. In some areas of the world, POPs can impact human health as a result of the consumption of these contaminated food sources. Among these POPs, per- and polyfluoroalkyl substances (PFASs) exhibit unique properties which are distinct from the chlorinated or brominated POPs but make them readily available in the marine environment due to their water solubility. The oceans are assumed to be the final sink of PFASs in the environment (Yamashita et al., 2004). Studies state that PFAS can represent excellent tracers of global circulation of oceanic waters, due to their persistence, water solubility, and measurability (Yamashita et al., 2008). Further data is needed to confirm this suggestion. In addition, not much data is available to reflect these concentrations in certain parts of the world such as the areas surrounding the western coast of Africa. It is essential for governments to be aware of possible contamination risks as these pose potential effects on both their growing population base and their

surrounding environments.

The aims of this experiment are 1) to analyse PFASs in water samples collected under the Second International Indian Ocean Expedition II (IIOE2) expedition, which covers off the coasts of Mozambique and Tanzania in the Indian Ocean. This is needed in order to fill the data gap in this area and evaluate if any differences in PFAS concentrations between the waters surrounding Mozambique and Tanzania. In this West Indian Ocean areas, the water current flows west towards the eastern coast of Africa and then splits at the coast into a northern and a southern direction. Researchers are evaluating the variation of various physical and chemicals parameters in these areas using vertical and horizontal profiling thereby capturing different flows of the water currents in that region; 2) to evaluate the variation of the concentrations between the areas using vertical and horizontal profiling by analysing water samples at different sampling locations and different depths which will be applied to find characteristic concentrations of selected priority PFASs. It will be attempted to identify linkages to previous studies and other metadata.

(8)

2 Background

2.1 What are PFASs?

Perfluorooctanesulfonic acid (PFOS) and perfluorooctanoic acid (PFOA) are the best studied perfluoroorganic substances. PFOS and PFOA belong to the family of per- and

polyfluoroalkyl substances, which are abbreviated as “PFASs”. Chemically, PFOS and PFOA are organic acids, with PFOS being a sulfonic acid and PFOA a carboxylic acid and both of them contain a perfluoroalkyl moiety (CnF2n+1−). With respect to carbon chain-length,

for PFOS the n=8 and n=7 for PFOA. Being organic acids in water or other environmental media and depending on the pH, they are present either in their protonated form (and occur as an acid) or anionic form (and with a cation to form salts). Analytically, they are determined as an anion. PFOS, PFOA and other PFASs are not known to occur naturally. They are frequently detected in many matrixes such as water, wildlife, and humans.

PFAS are industrial chemicals that can be grouped into polymeric and non-polymeric substances; for overview, see a review paper by Buck et al., 2011 or OECD documents (OECD, 2013; OECD, 2008). Non-polymer PFASs are partially or fully fluorinated alkylated substances attached to a functional group such as a carboxylic group forming perfluorinated carboxylic acid (PFCA, e.g., PFOA) or fluorotelomer carboxylic acid (FTCA); or with a sulfonic group forming perfluoroalkane sulfonates (PFSAs, e.g., PFOS) or fluorotelomer sulfonic acid (FTSA). The polymeric PFASs are fluorinated polymers consisting of a carbon backbone with fluorines attached to it or a carbon and oxygen backbone with fluorines. The most studied and regulated perfluoroalkyl acids (PFAAs) are PFCAs and PFSAs (see section 2.3).

The strength of the C-F bonds contributes to the extreme stability of perfluorinated alkyl substances (PFAS). In fluoropolymers, their distinctive properties, such as the strong electronegativity and small atomic size of fluorine, the perfluoroalkyl moiety (CnF2n+1−)

imparts enhanced properties to polymeric molecules (e.g., stronger acidity, higher surface activity at very low concentrations, stability, and/or water- and oil-repellence) compared to monomeric hydrocarbons. The stability of the polar covalent carbon-fluorine bond makes them resistant to metabolism and degradation in the environment. This leads to high trophic magnification factors and the potential for accumulation in various species in the food web. Studies on animals have revealed some of their toxicological properties, such as neonatal mortality and carcinogenicity (Ju, Jin, Sasaki and Saito, 2008). The International Agency for Research on Cancer (IARC) has not classified PFOS as to its carcinogenicity to humans (status: May 2020). IARC classified PFOA as possibly carcinogenic to humans (Group 2B) (IARC, 2018). In the EU, under the Regulation on classification, labelling and packaging of substances and mixtures, PFOS and PFOA have a harmonised classification as a suspected carcinogen and presumed human reproductive toxicant (European Commission, 2008). These compounds are also extremely hydrophobic (water repellent) and lipophobic (fat repellent). These two properties make these compounds difficult to break down and are of of great use in several products such as Teflon and Scotchgard. Uses of this include layering on food wrappers like popcorn bags and in aqueous film forming foams (AFFF) which are used for putting out fires.

(9)

2.2 Production

Global production of PFASs occurred in USA, Germany, Italy, Belgium, China, Japan and Russia. Major US manufacturer 3M began phasing out production of PFOS in 2000 and stopped production in 2002. By the early 2000s, Europe and Japan made voluntary efforts to phase out production of PFOA and PFOS. In 2006, the eight major PFOA manufacturers in the USA agreed to phase out production by 2015 under the EPA PFOA Stewardship Program (https://www.epa.gov/assessing-and-managing-chemicals-under-tsca/fact-sheet-20102015-pfoa-stewardship-program). The actual production of PFOS, PFOA, PFHxS and their related substances cannot be quantified and confirmed due to a lack of reporting of production in many countries. The PFOA stewardship programme in the U.S. reported that facility

emissions and product content for PFOA and PFOA-related compounds were eliminated by the year 2015. For actions thereunder, see https://www.epa.gov/assessing-and-managing-chemicals-under-tsca/risk-management-and-polyfluoroalkyl-substances-pfas.

2.3 Regulation

The Stockholm Convention on Persistent Organic Pollutants (POPs), administered by the United Nations Environment Programme (UNEP), is a global treaty to protect human health and environment from adverse effects caused by POPs. The objectives of the Convention shall be achieved through elimination of production, uses and releases of these POPs (Fiedler et al., 2019).

The Stockholm Convention on Persistent Organic Pollutants was concluded in 2001 and entered into force on 17 May 2004 (www.pops.int). The Stockholm Convention is a worldwide legally binding agreement and at the beginning covered 12 initial POPs. The United Nations Environment Programme (UNEP) provides the secretariat of the Stockholm Convention, which is based in Geneva, Switzerland. UNEP is the leading international environment entity that supports the agenda and implementation of

environmental sustainability for the United Nations. The COP, or the Conference of the Parties of the Stockholm Convention, governs the POPs Convention, with its members being the Convention’s Parties. Today, the Stockholm Convention has 184 Parties

(http://www.pops.int/Countries/StatusofRatifications/PartiesandSignatoires/tabid/4500/Defau lt.aspx).

PFASs have not been among the initial 12 POPs. The COP listed PFOS, its salts and perfluorooctanesulfonyl fluoride (PFOSF) into annex B of the Stockholm Convention in 2009; PFOA, its salts and PFOA-related compounds were listed in 2019 in Annex A. PFHxS, its salts and PFHxS-related compounds were recommended for listing in 2021.

The Stockholm Convention has set-up a Global Monitoring Plan to follow concentrations of POPs in the environment and to evaluate the effectiveness of the measures taken under the Convention.

2.4 Transport

PFASs have higher water solubility and lower lipophilicity when compared to the typical chlorinated or brominated POPs, thus the processes driving the long-range transport and cycling of ionic PFASs are different than those described for more hydrophobic pollutants, for instance the polychlorinated biphenyls (IARC, 2016). The neutral and more volatile PFASs like the fluorotelomer alcohols (FTOHs) and perfluoroalkyl sulfonamides (PFASAs)

(10)

undergo atmospheric transport, which after oxidation in the atmosphere (i.e., converted into respective PFSAs/PFCAs depending on their chain length) reach remote oceanic and continental regions through dry/wet deposition (Young and Donaldson, 2007). However, direct transport by ocean currents is thought to be the main way for the global transport of the ionic PFASs from source areas. The ocean is assumed to be the final sink of PFASs in the environment (Yamashita et al. 2004). Seawater measurements of PFASs are very useful for determining the dominant transport pathway, either oceanic currents or atmospheric transport of precursors (Ahrens, Xie and Ebinghaus, 2010). Nevertheless, many uncertainties exist on the main drivers of their global distribution, including the influence of source areas, currents, and biogeochemical cycles, such as the biological pump and degradative processes.

(González-Gaya et al., 2014)

2.5 IIOEC 2

The Second International Indian Ocean Expedition II (IIOE2) is a multi-national programme of the United Nations Intergovernmental Oceanographic Commission (IOC), which

emphasizes the need to research the Indian Ocean and its influence on the climate and its marine ecosystem (https://iioe-2.incois.gov.in/).

The IOC realised that there was a lack of long-term environmental information in the Indian Ocean, especailly for countries surrounding the Indian Ocean. As a result, the IOC decided to declare the beginning of the IIOE2, which came 50 years after the first IIOE. This interest in the area has brought numerous research voyages with state-of-the-art technology. The data collected covers physics, chemistry, plankton, biodiversity, large animals such as whales and seabirds as well as geology.

So far, the information in relation to chemical contamination is restricted to inorganic chemicals such as heavy metals. Organic chemicals including PFASs have not been measured. Two of the PFASs – PFOS and PFOA - are listed in the Stockholm Convention and a third PFAS, PFHxS, is recommended for listing. Countries along the African Indian Ocean coast are parties to the Stockholm Convention and lack information as to the presence of PFAS. This study will generate quantitative information so that parties to the convention can provide information to the Global Monitoring Plan of POPs and make an informed decision as to the proposed listing of PFHxS in 2021 by consulting these data.

2.6 Mozambique Current

This study predominantly focuses on data obtained from samples collected off the coast of Mozambique and Tanzania. As the water from the middle of the Indian Ocean nears the Western African coast, it breaks into two parts. One flows northwards towards Tanzania and the other goes south along the coast of Mozambique. This stretch of water is known as the Mozambique Channel, aptly named as it flows between the mainland Mozambique and the island of Madagascar (Figure 1). This area of ocean contains anticyclonic eddies. Studies show that drifters launched during a cruise described moved in circular paths around the outer edge of the eddies (de Ruijter et al., 2002). These appear to be over 300 km wide and

continue to the bottom of the ocean with varying velocities. They propagate southward at an average speed of about 4.5 km/day. Consequently, a train of three anti-cyclonic eddies is mostly present in the Channel. This produces a southward current at the Mozambique side of the Channel and northward towards the Madagascar side (Sætre and Da Silva, 1984). Below the thermocline (layer of water where temperature changes more rapidly with depth than in the water layers above and below), the direct current observations have revealed a

(11)

Mozambique undercurrent flowing along the continental slope. (de Ruijter et al., 2002). These eddies cause a high connectivity or mixing among marine populations and chemicals.

Figure 1: The currents in the Mozambique Channel as well as the East African coast. The different colours represent the speed of the currents (van Aswegen, n.d.).

2.7 Vertical and Horizontal Profiling

Studies state that PFAS can represent excellent tracers of global circulation of oceanic waters, due to their persistence, water solubility, and measurability. The water solubilities of PFOS and PFOA are 570 mg/L and 3400 mg/L, respectively (OECD, 2002; USEPA, 2002).

Because of these high-water solubilities, the open-ocean water column is suggested to be the final sink for PFOS and PFOA. Concentrations of PFAAs in open-ocean waters have been reliably measured at parts-per-quadrillion levels, by a combination of solid phase extraction and high-performance liquid chromatography (HPLC)-tandem mass spectrometry (Yamashita et al. 2008).

(12)

Horizontal profiling refers to the practice of taking samples with varying distance to possible point sources. These could be from run-off water, discharge wastewater from industrial as well as residential sources. It is important to understand whether or not the concentrations of PFAS decrease with increasing distance from the shore in the coastal and open ocean

environments. Decreasing values show there is no oceanic sources of contamination since the water is diluting the concentrations of the select analytes rapidly.

Studies have shown that in different areas of the world such as the Mediterranean Sea and the Japan Sea, several differences were found in the PFASs profile in the water column

(horizontal profiling) (Yamazaki et al., 2019).

A study on the influence of salinity, pH and sediment characteristics on the sorption and desorption of PFOS in surface waters suggests that PFOS tends to exist as dissolved species in low salinity water such as freshwater, but is sorbed to sediment in high salinity water like in seawater (You, Jia and Pan, 2010). Moreover, it was found that the concentrations in the upper water layer were higher than the lower layer, suggesting that there was an incomplete vertical mixing. This is in part due to seasonal changes in the density of seawater structure. During months such as May and August, it was found that the salinity was diluted by freshwater inputs, which proceeded from rivers and heating of the water surface during the warm season (Weiss et al., 2015). This reason is less of a factor when considering the area of the current study as freshwater deposits contribute very little to the overall volume of the sampling area.

2.8 Analytes of Interest

In this report, three PFASs, namely PFOS, PFOA, and PFAS levels in the Western Indian Ocean surrounding the countries of Madagascar and Tanzania were discussed. These compounds were selected as PFOS and PFOA are listed in the Stockholm Convention (Annex B and A respectively) and PFHxS is under review.

(13)

3 Materials and Methods

3.1 Samples (location and reason for choosing them)

Figure 2: Sampling locations from the IIOEC

Samples were obtained from the Second International Indian Ocean Expedition II (IIOE2). During this expedition, hundreds of water samples were taken along the Mozambique and Tanzanian coast. The samples covered a stretch of coast of approximately 2,000 km. The distances from the coast ranged from 5 km to 100 km. Each sampling location had anywhere from 1 to 4 sampling points, with each sample being taken at different depths. Depths ranged from 3 m (surface sampling) to 1,247 m.

Aliquots of 500 mL, packaged in 500 mL HDPE bottles were shipped by our collaborators at North-West University (Potchefstroom, South Africa) to Örebro University in fall 2019. Upon arrival at Örebro University, samples were stored in a cool room at 6 °C in the dark until analysis. A total of 46 samples were analysed for PFOS, PFOA and PFHxS.

3.2 Chemicals

The chemicals used in the project include methanol (HPLC grade, > 99.99% and LCMS grade, > 99.9%), ammonia hydroxide (25%), acetonitrile (HPLC grade, > 99.99%),

isopropanol ( > 99.0%) which were purchased from Fisher Scientific (Leicestershire, UK). Ammonium acetate (> 99.0%) and glacial acetic acid were purchased from Sigma Aldrich (Darmstadt, Germany). Water used during the analyses was MilliQ quality (18.6 MΩ). Analytical standards consisted of; a PFAS Internal Standard UNEP-PFAS-EXT. A. This is a mixture of ten mass-labelled (13C) perfluoroalkylcarboxylic acids (C4-C12 and C14), three

mass-labelled (13C) perfluoroalkylsulfonates (C4, C6, and C8), perfluoro-1

[13_C

8]octanesulfonamide, and sodium 1H,1H,2H,2H-perfluoro-1-[1,2 13C2]octane sulfonate.

(Wellington Laboratories, Guelph, Canada), a PFAS Recovery Standard UNEP-PFAS-INJ. A which is a mixture of seven mass-labelled (13C) perfluoroalkylcarboxylic acids (C4-C6 and

C8-C11), and two mass-labelled (18O and 13C) perfluoroalkylsulfonates (C6 and C8).

(Wellington Laboratories, Guelph, Canada), and a PFAS Native Standard UNEP-PFAS-SOL. A which is a mixture of eleven native linear perfluoroalkylcarboxylic acids (C4-C14), four

(14)

native perfluoroalkylsulfonates (C4, C6, and C10 linear; C8 linear and branched), native

perfluoro-1-octanesulfonamide, and native sodium 1H,1H,2H,2H-perfluorooctane sulfonate. (Wellington Laboratories, Guelph, Canada). The branched C8 PFSA (PFOS) is a technical

mixture containing approximately 21.1% branched isomers and 78.8% of the linear isomer (Table 7 in Annex).

3.3 Extraction and clean-up

Sample were retrieved from storage in a walk-in cooler set to 4oC. Samples followed the extraction procedure created by United Nations Environmental program (Weiss et al., 2015). Bottles weights were recorded and later used to calculate the amount of water extracted by subtracting the empty bottle weight from the total. Samples, procedure blanks and the quality control (QC) were spiked with 5 µl of 13C labelled internal standard with concentrations ranging from 0.1858 ng/µl to 0.2 ng/µl. In addition, the QC was spiked with 5µl of a 12_C

native standard with concentrations ranging from 0.025 ng/µl to 0.200 ng/µl. Both procedural blanks and QC samples were consisting of 50 mL of MilliQ water; they were treated as the same way as samples

In brief, before the SPE, conditioning, washing and elution solutions were prepared. Samples were loaded onto a preconditioned SPE cartridge at a rate of 1-2 drops per second. After loading the water samples, the cartridges were washed with 25 mM ammonium acetate buffer at pH 4 and dried under vacuum. Two separate fractions were collected for analysis into labelled 10 mL polypropylene tubes. Empty sample bottles were rinsed with 4 mL of

methanol and added to the respective cartridges for the elution of fraction one. Once fraction one had finished eluting, the polypropylene tubes were removed from the manifold and replaced with another set of tubes used for the elution of fraction two using 4 mL of a 0.1% ammonium hydroxide solution.

The extracts were placed in a RapidVap evaporator to concentrate the samples to approximately 0.5 mL for fraction one and 0.2 mL for fraction two. After batch 6, these values were changed to 0.25 mL and 0.1 mL respectively. The RapidVap was set to an internal temperature of 50 °C and a pressure of 300 bar. At the beginning of every evaporation, samples were monitored to ensure that no boiling occurred. Bubbles during evaporation could cause a loss of analytes onto the lid of the evaporator. Samples took between 60 minutes to 4 hours to complete evaporating. The longer evaporation times were caused by the presence of water in the extract due to incomplete drying prior to elution. After concentration, samples were spiked with 5 µl of 13C labelled recovery standard with concentrations ranging from 0.1892 ng/µl to 0.2 ng/µl. Fraction two had either 300 µL or 150 µL of aqueous mobile phase added depending on the batch. They were vortexed to ensure proper mixing and then centrifuged to make sure all aquatic molecules were in the bottom of the tubes.

Both fractions were then transferred to glass LC vials and ready for instrumental analysis using LC-MS/MS.

3.4 Instrumental analysis and quantification

For separation and identification of the targeted compounds, an Acquity Ultra-performance liquid chromatograph (UPLC) (Waters Corporation, Milford, USA) coupled to a Xevo-TQ-S

(15)

tandem mass spectrometer (MS/MS) (Waters Corporation, Milford, USA) was used. The LC was equipped with an ultra-high-performance reverse phase column (Acquity BEH, 1.7 μm C18 particles, 100 mm length, 2.1 mm inner diameter of 2.1 mm) (Waters Corporation

Milford, USA). This column contains ethylene bridged hybrid (BEH) particles, which increase the column stability, which enables a wider usable pH range (1-12). This creates a more versatile and robust column.

Two mobile phases were used during instrumentation. The first mobile phase consisted of a mixture of 2 mM ammonium acetate in MilliQ-water and with 2 mM ammonium acetate in methanol at a ratio of 70:30. The second mobile phase consisted of only 2 mM ammonium acetate in methanol. MS/MS instrumental settings were electrospray ionisation operated in negative mode with a source temperature at 150 °C, desolvation temperature at 400 °C, desolvation gas flow at 800 L/h, a cone gas flow at 150 L/h and a capillary voltage at 0.84 kV. Column temperature was 50 °C.

One quantification standard was prepared for each extraction batch of samples. This standard contained equal concentration of internal standard, recovery standard and natives as the quality control samples. It was not subject to the extraction procedure and was used to quantify the results.

Quantification was performed using standards dissolved in a pure solvent, the standard evaluated on calibration curve where the deviation within ±20%. The calibration curves consisting of a concentration series of 10, 20, 200, 500, 1000, 4000 pg/mL, the deviation of every point from the regression line was less than 20% from the theoretical value, and correlation coefficients R2 >0.99.

Quantification of target PFASs was done using response factor for samples (the area of native PFAS divided by the area of the appropriate mass labelled internal standards) and compared with the response factor for the standard. IS was spiked before extraction to give a recovery-corrected concentration. In current LC separation, not all of the individual PFOS isomers were baseline separated; they were separated into different groups (i.e., 3-, 4-, 5-PFOS, 2-, 6-PFOS, and L-PFOS). Branched isomers of PFOS were quantified using PFOS isomer

standard and reported as the sum of the isomer groups (see section 3.2, or Table 7 and Figure 11 and Figure 12 in the Annex).

The MassLynx software TargetLynx was used to assist with both data acquisition,

quantification and the transformation of data into usable results. Specifically, quantification was done using the appropriate quantification standard.

Data interpretation was made using Microsoft Excel.

3.5 Additional parameters

Metadata to characterize the samples were obtained from Duan van Aswegen (North-West University, South Africa) and included information regarding sampling date, location, oxygen content, electrical conductivity, turbidity, temperature, and pressure which is shown in Table 6.

(16)

3.6 Quality assurance and Quality control (QA/QC)

All extraction and clean up steps were done in a fumehood. All equipment was rinsed with methanol prior to use to avoid contamination. Smaller parts were put into glass beakers and sonicated in methanol to remove possible contamination from previous extractions.

Batches of 7 samples were extracted at a time with two procedural blanks as well as one quality control (QC) sample to make a total of 10 simultaneous extractions. Procedural blanks consisting of 50 mL MilliQ water were used to check any contamination during the analysis procedure whereas QC samples consisting of 50 mL MilliQ water were used to assess the extraction efficiency, accuracy, and repeatability of the method. The procedural blanks showed no significant signs of background contamination. The recoveries of the 13C mass labelled standards can be found in the table below. The majority of the recoveries were within the required range of 50% - 150%. Deviation occurred for two analytes, recoveries of PFOA in batch 8 as well as well as of PFHxS in batch 9 (see Table 10 in Annex). If recoveries were

outside this range, the data was assumed to be corrupt and the data was omitted (see Table 9

in Annex).

Quantification standard was injected three times per batch run; in the beginning of each run, halfway through the run and at the end of the run to evaluate the overall performance of the instrument during the analysis. The areas were monitored to ensure a stable intensity (<20%). For this study, the limit of detection was not used as blanks did not show concentrations above the lowest point in the calibration curve. The limit of quantification (LOQ) is defined as the smallest amount or concentration of analyte of interest in the calibration curve that exhibits the signal to noise ratio of the peak is greater than 10 when no detectable

concentration of analyte in the procedural blank. When detectable levels of analytes are found in the procedural blank, average of detectable blank concentrations plus 10 times of the standard deviation of the blank concentrations will be used. Since no detectable

concentrations of analytes of interest were found in the procedural blank, the LOQs of analyte of interest are equal to the lowest point of calibration curve that exhibits a signal to noise ratio greater than 10 (Table 1).

To make this applicable to my samples, the LOQ was divided by the sample volume (approx. 0.5 L) in order to give units of pg/L. The LOQ of an individual analytical procedure is

defined as the lowest amount of analyte in a sample, which can be quantitatively determined with suitable precision and accuracy.

Table 1: LOQ levels achieved during the current project in pg and pg/L.

LOQ pg pg/L PFHxS 9.0 18.0 PFOA 10.0 20.0 3/4/5-PFOS 5.0 10.0 6/2-PFOS 6.0 12.0 L-PFOS 9.0 18.0

(17)

4 Results

4.1 Designation and characterization of samples

In order to comply with the hypothesis, 46 samples were selected for chemical analysis, 21 from the Mozambique coast and 25 from the Tanzania coast. The sampling locations and the associated sampling IDs are shown in Figure 3 and Table 2 for Mozambique (M or MZ) and Figure 4 and Table 3 for Tanzania (T or TZ).

(18)

Figure 4: Samples analysed from Tanzania

These sampling sites varied by both distance from the coast as well as depth. At each chosen location, as little as one and up to four samples were taken at different depths. For the present study, a fraction of these samples was analysed.

The samples from site M1-8 to M1-10 were selected for this study as they were the closest sampling points next to the fourth most populous city in Mozambique, Beira. Just south of this location, two major rivers, the Buzi, and Pungwe river meet and empty into the Indian Ocean.

M4-1 and M5-1 were selected as they were the closest sampling points to where the Zambezi river meets the ocean. Both were needed as the Zambezi river splits approximately 10

kilometres before it reaches the coast. The Zambezi river is the fourth largest river in Africa and is the largest that flows into the Indian Ocean.

Locations M8-1, M9-5, M11-1, and M12-1 were selected to provide background levels in the area as these samples were taken far from any major human settlement or possible industry. M14-1 was selected as this was the northern most sampling point in Mozambique. This location would be the closest to where the Indian Ocean splits along the border of Mozambique and Tanzania.

T1-1 and T1-3 is in the area where the above-mentioned Indian Ocean current meets the west coast of Africa.

(19)

The sample sites T9-1 to T9-7 were the areas closest to the largest city in Tanzania, Dar es Salaam. Due to its increasing growth and population, it was considered and important site. These samples were also situated just south of the island of Zanzibar, which is a major tourist hub in the country.

T10-1 to T10-3 were sampled on the western coast of Zanzibar.

Table 2: Designation of samples as to location and sampling depth (Mozambique) Sample site Depth (m) Sampling Date Sampling time

MZ_M1-08_C 380.3 10/21/2017 13:13 MZ_M1-08_D 889.9 10/21/2017 13:13 MZ_M1-10_A 3 10/21/2017 09:07 MZ_M1-10_B 22.1 10/21/2017 09:07 MZ_M1-10_C 31.7 10/21/2017 09:07 MZ_M1-10_D 1247 10/21/2017 09:07 MZ_M4-01_A 3.4 10/24/2017 14:35 MZ_M9-05_B 40.2 10/28/2017 14:45 MZ_M9-05_C 66.9 10/28/2017 14:45 MZ_M9-05_D 690.3 10/28/2017 14:45 MZ_M11-01_A 4.6 10/29/2017 23:46 MZ_M11-01_C 50.8 10/29/2017 23:46 MZ_M11-01_D 178.4 10/29/2017 23:46 MZ_M12-01_A 3.3 10/30/2017 06:46 MZ_M12-01_B 11.1 10/30/2017 06:46 MZ_M12-01_D 202.2 10/30/2017 06:46 MZ_M14-01_A 3.4 10/31/2017 09:53 MZ_M14-01_B 47.1 10/31/2017 09:53 MZ_M14-01_C 81 10/31/2017 09:53 MZ_M14-01_D 433.5 10/31/2017 09:53

(20)

Table 3: Designation of samples as to location and sampling depth (Tanzania)

4.2 Concentrations of PFOS, PFHxS and PFOS

A total of 46 samples were analysed, 21 from the Mozambique coast and 25 from the Tanzania coast. In general, the concentrations were very low and the measured values from the peaks in the chromatogram are shown in Table 8 in the Annex. Table 4 and Table 5 below

show the concentrations that were detected above the LOQ. All values below the LOQ have the respective LOQ values filled in.

L-PFOS found in the water samples ranged from below the LOQ to 140.9 pg/L; br-PFOS could not be quantified in any of the samples; therefore, the sum PFOS corresponds to L-PFOS. The concentrations of PFHxS and PFOA ranged from below the LOQ to 25.4 pg/L and below the LOD to 63 pg/L respectively.

Out of the samples analysed only two samples showed detectable levels for PFHxS. These two were both from the same sample location off the coast of Tanzania, but at different depths. PFOS had a slightly higher rate of detection at four samples in total, one from

Mozambique and two from Tanzania. PFOA was most commonly present as it was quantified in 29 samples, most of whom were taken from Tanzania.

Sample site Depth (m)Sampling Date Sampling time TZA_T1-01_A 4.3 11/09/2017 00:31 TZA_T9-05_B 37.2 11/05/2017 14:15 TZA_T9-05_C 55.2 11/05/2017 14:15 TZA_T9-05_D 501.6 11/05/2017 14:15 TZA_T9-07_A 3.6 11/05/2017 18:46 TZA_T9-07_B 70.5 11/05/2017 18:46 TZA_T9-07_C 81.2 11/05/2017 18:46 TZA_T9-07_D 669.4 11/05/2017 18:46 TZA_T10-01_A 4.9 11/05/2017 00:46 TZA_T10-01_C 84.9 11/05/2017 00:46 TZA_T10-01_D 225.2 11/05/2017 00:46 TZA_T10-02_A 4.9 11/04/2017 23:11 TZA_T10-02_C 79.5 11/04/2017 23:11 TZA_T10-02_D 500 11/04/2017 23:11 TZA_T10-03_A 4.6 11/04/2017 19:07 TZA_T10-03_B 20.4 11/04/2017 19:07 TZA_T10-03_C 26 11/04/2017 19:07 TZA_T10-03_D 751.2 11/04/2017 19:07 TZA_T1-03_A 3.3 11/08/2017 21:24 TZA_T1-03_B 20.5 11/08/2017 21:24 TZA_T1-03_C 59.2 11/08/2017 21:24 TZA_T1-03_D 830.9 11/08/2017 21:24

(21)

Table 4: Concentrations of priority PFASs in samples from Mozambique (pg/L)

Sample-IDs L-PFOS br-PFOS ΣPFOS PFHxS PFOA

pg/L pg/L pg/L pg/L pg/L MZ_M1-08_C <18 <22 <18 <18 <20 MZ_M1-08_D <18 <22 <18 <18 <20 MZ_M1-10_A <18 <22 <18 <18 <20 MZ_M1-10_B <18 <22 <18 <18 <20 MZ_M1-10_C <18 <22 <18 <18 <20 MZ_M1-10_D <18 <22 <18 <18 <20 MZ_M4-01_A <18 <22 <18 <18 <20 MZ_M4-1_D 37.7 <22 37.7 <18 63.0 MZ_M9-05_B <18 <22 <18 <18 27.6 MZ_M9-05_C <18 <22 <18 <18 30.5 MZ_M9-05_D <18 <22 <18 <18 <20 MZ_M11-01_A <18 <22 <18 <18 28.9 MZ_M11-01_C <18 <22 <18 <18 <20 MZ_M11-01_D <18 <22 <18 <18 38.5 MZ_M12-01_A <18 <22 <18 <18 22.9 MZ_M12-01_B <18 <22 <18 <18 23.2 MZ_M12-01_D <18 <22 <18 <18 24.5 MZ_M14-01_A <18 <22 <18 <18 23.0 MZ_M14-01_B <18 <22 <18 <18 <20 MZ_M14-01_C <18 <22 <18 <18 26.0 MZ_M14-01_D <18 <22 <18 <18 <20

(22)

Table 5: Concentrations of priority PFASs in samples from Tanzania (pg/L)

Samples L-PFOS br-PFOS ΣPFOS PFHxS PFOA

pg/L pg/L pg/L pg/L pg/L TZA_T01-1_A <18 <18 <LOQ <18 <20 TZA_T01-1_D 47.0 <18 47.0 <18 42.2 TZA_T01-1_C 24.0 <18 24.0 <18 47.0 TZA_T9-05_B <18 <18 <18 <18 25.5 TZA_T9-05_C <18 <18 <18 <18 22.7 TZA_T9-05_D <18 <18 <18 <18 30.7 TZA_T9-07_A <18 <18 <18 <18 28.2 TZA_T9-07_B <18 <18 <18 <18 42.9 TZA_T9-07_C <18 <18 <18 <18 22.8 TZA_T9-07_D <18 <18 <18 <18 38.5 TZA_T10-01_A <18 <18 <18 <18 45.6 TZA_T10-01_C <18 <18 <18 <18 39.0 TZA_T10-01_D <18 <18 <18 <18 23.5 TZA_T10-02_A <18 <18 <18 <18 53.6 TZA_T10-02_C <18 <18 <18 <18 28.6 TZA_T10-02_D <18 <18 <18 <18 22.2 TZA_T10-03_A <18 <18 <18 <18 21.0 TZA_T10-03_B <18 <18 <18 <18 36.2 TZA_T10-03_C <18 <18 <18 <18 <20 TZA_T10-03_D <18 <18 <18 <18 <20 TZA_T1-03_A <18 <18 <18 <18 <20 TZA_T1-03_B <18 <18 <18 <18 <20 TZA_T1-03_C <18 <18 <18 23.1 26.1 TZA_T1-03_D <18 <18 <18 25.4 34.8

The concentration of PFOA in surface water are described in Figure 5 and Figure 6. Surface

water was indicated by the depth code “A”, where samples were taken between 3 and 5 m depth. The sampling points correspond to the Sample IDs TZA_T10-01_A, T10-02_A, T10_03_A and TZA_T9-07_A (Figure 5 and Table 5) and MZ_M11-01_A, M12_01_A and

M14_01_A (Figure 6). When focusing on Figure 5, the two sampling points TZA_T10-01_A

and T10-02_A closer to the island are higher than the one furthest out to sea with values of 46 and 54 pg/L compared to the 21 pg/L found in T10_03_A. This value corresponds more with the sample TZA_T9-07_A, which had a PFOA concentration of 28 pg/L. Figure 6 shows

surface water samples along the northern coast of Mozambique. Here the concentration of PFOA was 29, 23 and 23 pg/L.

(23)

Figure 5: PFOA levels in surface water near the island of Zanzibar

Figure 6: PFOA concentrations in surface water in Northern Mozambique

Five samples showed levels above the detection limit as shown in Figure 7. These samples

had depth codes of either C or D, meaning that they were not surface water samples. PFOA was detected at higher concentration than the other PFASs except in TZA_T01-1_D where L-PFOS showed the same concentration. No sample showed detectable levels of all three priority PFASs. Such low detections can be explained by the mass dilution that is cause by the ocean as well as the constant mixing cause by the ocean currents surrounding the two

Zanzibar, Tanzania

46 pg/L

54 pg/L

21 pg/L

28 pg/L

Northern coast of Mozambique

29 pg/L

23 pg/L

(24)

countries. This constant water flow enables pollutants to disperse rapidly until they are either sorbed onto materials and sediment or taken up in biota.

Figure 7: Samples that showed presence of either PFOS or PFHxS besides PFOA

4.3 Other parameters

While measuring the concentrations of the select PFASs, other parameters were examined to determine if they had an effect on the found concentrations (see Table 6). Salinity remained

constant no matter the sampling location or depth though the salinity of the surface and bottom water were a bit different suggesting that the samples were collected where the waters were well mixed rather than being influence from riverine discharge from the coast. Much like salinity, the oxygen levels remained rather constant throughout all of the samples. The only differences between the surface and bottom water were temperature. The concentrations of detectable PFOA in selected samples do not show observable trends with these parameters.

0.00 20.00 40.00 60.00 80.00 100.00 120.00

MZ_M4-01_D TZA_T01-1_C TZA_T01-1_D TZA_T1-03_C TZA_T1-03_D

p

g/L

Sample ID

(25)

Table 6: Table of some extended parameters measured in the Tanzanian samples Samples PFOA pg/L Average Temperature [ITS-90, °C] Average electrical Conductivity [S/m] Pressure [db] Oxygen [mL/L] Salinity [PSU] TZA_T01-1_A <LOQ 27.7 5.7 5 4.2 35.4 TZA_T01-1_C 47.0 26.4 5.5 61 4.1 35.5 TZA_T01-1_D 42.2 10.3 3.8 431 3.5 34.9 TZA_T9-05_B 25.5 27.5 5.6 38 4.2 35.5 TZA_T9-05_C 22.7 26.8 5.6 56 4.2 35.5 TZA_T9-05_D 30.7 10.0 3.8 502 3.6 34.8 TZA_T9-07_A 28.2 27.7 5.7 4 4.1 35.6 TZA_T9-07_B 42.9 27.4 5.6 71 4.1 35.5 TZA_T9-07_C 22.8 26.8 5.6 82 4.1 35.5 TZA_T9-07_D 38.5 8.8 3.7 674 2.2 34.9 TZA_T10-01_A 45.6 26.4 5.5 5 4.1 35.4 TZA_T10-01_C 39.0 26.2 5.5 85 4.0 35.4 TZA_T10-01_D 23.5 12.9 4.1 227 3.3 35.1 TZA_T10-02_A 53.6 26.8 5.5 5 4.1 35.3 TZA_T10-02_C 28.6 26.2 5.5 80 4.0 35.4 TZA_T10-02_D 22.2 10.0 3.8 503 3.6 34.8

Figure 8: Variation of temperature with relation to depth below water surface

As shown in Figure 8, the temperature drops steeply once the depth passes the thermocline.

As this happened with all sampling locations, no correlation could be found linking temperature to the presence of PFAS.

0 5 10 15 20 25 30 0 500 1,000 1,500 Te m p era tu re (oC) Depth (m)

MZ_M1-10

(26)

4.4 Analytical issues as to elution in fractions

As per the method, two fractions were collected for each sample from the OASIC WAX cartridges. The first fraction, which was eluted using methanol to contain the neutral PFASs compounds such as perfluorooctanesulfonamide (FOSA). The second fraction was eluted with 0.1% ammonium hydroxide in methanol to elute the anionic substances such as PFOS, PFOA and PFHxS. While conducting this project, there were discrepancies in which fraction the ionic compounds eluted in. It was shown in two batches where the 13C labelled PFOS, PFOA and PFHxS were found in the first fraction and not the second. This was only the case for the actual samples and not the blanks or QC samples. A reason for this could not be identified as there was no negligible difference in the parameter such as salinity across all the samples. Recoveries remained within the acceptable region, so the data was accepted. Results such as this could suggest that there is no need to separate the two fractions prior to analysis, although this has been done to limit the potential matrix effects of the ocean water.

(27)

5 Discussion:

5.1 Comparison with results from water samples

Within a UNEP monitoring project, water samples from Thailand and Uganda were analysed for PFOS, PFOA and PFHxS. In contrast to the ocean samples from Mozambique and

Tanzania, these are surface water samples. In these samples, all targeted PFASs could be detected and in general, concentrations were much higher (see Figure 9). Although Uganda

and Tanzania are neighbouring countries (sharing Lake Victoria) the concentrations found are extremely different. Uganda showed levels of PFOA ranging from 91 pg/L to 1500 pg/L compared to the <LOQ – 63 pg/L found in this study. Similarly, PFOS and PFHxS were also higher concentrations in the national samples. This is caused due to the sampling locations as the nation sample were taken from river water which does not have the same dilution

potential as the coastal country of Tanzania. Other factors could include population density in the areas measured, use and disposal of PFAS contaminated products, and transport of

pollutants.

Figure 9: PFOS, PFOA and PFHxS levels in Thailand and Uganda

5.2 Comparison with other study with respect to sampling depth

Other studies conducted showed that PFOS and PFOA concentrations in the North Atlantic Ocean ranged from 8.6 pg/L to 36 pg/L and from 52 pg/L to 338 pg/L, respectively. The corresponding concentrations in the Mid Atlantic Ocean were 13 pg/L to 73 pg/L and 67 pg/L to 439 pg/L (Yamashita et al., 2008). The surface waters of the North and Mid- Atlantic Ocean are more contaminated with PFOS and PFOA than are the surface waters of the Indian Ocean with an overall range of <5 pg/L to 11 pg/L for PFOS and PFOA. Concentration of PFOA in the eastern Pacific Ocean ranged from 10 pg/L to 60 pg/L whereas that in the western Pacific Ocean ranged from 140 pg/L to 500 pg/L. Ocean water collected from the southern Indian Ocean contained only trace levels of PFAS (Yamashita et al., 2008). Yamashita et al. (2008) stated that since the ocean is a three-dimensional environment, general monitoring survey of PFAS in the surface water provide very limited information of global kinetics. It was clear that comprehensive survey of not only in surface water but also deep water in oceans and seas are necessary to reconstruct the global kinetics model of PFAS.

0 500 1000 1500 2000 2500

Thailand Thailand Uganda Uganda Uganda Uganda Uganda

p

g/L

Country

National Samples

(28)

They collected water column samples at several depths, from the Labrador Sea, the Mid-Atlantic Ocean, the South Pacific Ocean, and the Japan Sea, to enable an understanding of the hydrodynamics of PFAS, and to understand the global oceanic circulation. The vertical profile of PFAAs in the two water columns from the Japan Sea showed gradual decrease in PFAA concentrations from surface to subsurface layers. Concentrations of PFOS in surface water samples were 15 pg/L, and decreased to 2 pg/L at a depth of 1000 m. At depths below 1500 m, concentrations were at or below the detection limit of 1 pg/L and 6 pg/L, for PFOS and PFOA, respectively. (Yamashita et al., 2008). When compared to the present study, no correlation could be made between depth of the sample taken and the concentrations of PFOS, PFOA and PFHxS. Results found showed that varying concentration could be found at different depths of the same sampling location. In sample TZA_T01-1, the PFOA

concentration at the deepest point was 42.7 pg/L whereas it was not detected in the surface samples. Conversely, TZA_T10-01 had a PFOA concentration of 45.6 pg/L at the surface and 23.5 pg/L at the deepest. These inconsistencies could show that the currents produced from the Indian Ocean provide good mixing within the sampling area as supported by other parameters such as salinity and oxygen levels.

In the Southern Ocean only PFOS was detected, however, the PFOA concentrations are usually higher than PFOS concentrations in open-ocean waters, which can be explained with its different amount of source compounds (Prevedouros, Cousins, Buck and Korzeniowski, 2006) or lifetime of those precursors (D'eon, Hurley, Wallington and Mabury, 2006; Ahrens, Xie and Ebinghaus, 2010).

The study performed by González-Gaya et al., sampled areas near the West African coast. In their study they found that the global median surface seawater concentration for the sum of all the analysed compounds (ΣPFAS) was 1180 ± 1860 pg/L. In the northern hemisphere the median concentration was 708 pg/L± 831 pg/L, while in the southern hemisphere the median concentration was 1620 pg/L± 488 pg/L (González-Gaya et al., 2014). It must be noted that the sampling areas was not the exact same as the current study which could lead to the reasoning why the levels are so vastly different. Overall PFAS levels were much lower in the present study as majority of values fell below the detection limit. The vast difference in the southern hemisphere could be cause by high PFAS levels found elsewhere, such as off the coast of South America.

The most often reported PFASs, such as PFOA and PFOS, show a concentration variability of several orders of magnitude depending on the ocean basin. PFOA and PFOS

concentrations in the Atlantic (PFOA with a median of 44 pg/L and PFOS with a median of 59 pg/L in the North Atlantic, while in the South Atlantic medians were 58 pg/L and 742 pg/L respectively), Pacific (median 25 pg/L of PFOA; 101 pg/L PFOS) and Indian oceans (median 23 pg/L of PFOA; and 89 pg/L of PFOS) (González-Gaya et al., 2014) (see Table 11

in Annex).

5.3 Ratio of PFOS/PFOA in the Atlantic Ocean

In another study when samples from the 2007 Oden and 2009 Endeavor Atlantic Cruises were pooled, a distinct correlation between PFOA/PFOS ratio and latitude appeared. The majority of samples in the northern hemisphere contained PFOA/PFOS ratios of > 1, samples around the equator contain ratios of ~1, and most samples in the southern hemisphere

contained ratios of < 1. Furthermore, the magnitude of the difference tended to increase with increasing distance from the equator. (Benskin et al., 2012)

(29)

These observations are likely explained by a combination of surface water circulation in the Atlantic Ocean and differences in emissions from the continued manufacture and/or use of certain fluorochemicals in the northern versus southern hemispheres. For example, PFOA and perfluorooctanesulfonyl fluoride (PFOSF) based products were manufactured for over 50 years in North America and Europe, but only large-scale manufacturing of PFOA and PFOA precursors (e.g., fluorotelomer alcohols) continues today on these continents. In contrast, manufacture/use of PFOSF based products in South America has continued since the 2002 phase-out. Thus PFOA/PFOS ratios of >1 reflect continued manufacture and use of PFOA and PFOA precursors in the northern hemisphere, while PFOA/PFOS ratios of <1 reflect current use and importance of PFOSF-based products from the southern hemisphere.

PFOA/PFOS ratios of ~1 near the equator and south eastern Atlantic can likely be explained by mixing of northern and southern hemisphere water. (Benskin et al., 2012) as shown in

Figure 10. For detail, see section 9 Annex: Supplementary Information

Figure 10: Graphic depicting the route and levels of PFAS found during the Oden and Endeavor cruises (Figure adapted from Benskin et al., 2012)

According to this knowledge, samples in the current study should exhibit ratios of PFOA to PFOS of < 1, however, this is not the case. In the majority of the samples, PFOA is detected at higher concentrations than PFOS. Even if all PFOS values were changed to reflect the LOD, all of the PFOA found would result in a higher ratio than 1. In only one samples is the amount of PFOS found to be greater than that of PFOA. The reason for the differing results could be due to both the differences in scale of the project, date of sampling (10 years difference) as well as the sampling locations. As shown in the figure above, that expedition did not venture as far north as the sampling points in this project. Their results would be more biased on the data points surrounding South America, which is known to have high usage of PFOS.

(30)

6 Conclusion

For this study, 238 water samples were provided from cruises in the West Indian Ocean. Samples of 500 mL volume were collected with big efforts at 48 locations in the

Mozambique channel and 20 in the Tanzania channel. Typically, four vertical samples were collected up to depth of more than 1000 m, resulting in a total of 166 water samples from Mozambique and 72 from Tanzania. Physical parameters as to salinity, temperature, turbidity, pressure and many others were recorded. From the total of 238 ocean water

samples, 46 samples that were thought to have been impacted by some activities on land that might have resulted in the release of PFASs, were analysed for three priority PFAS. Since the choice of the sampling locations, the sampling equipment and the conditions are not known to us, we have to assume that the integrity of the samples has been maintained throughout the whole process until arrival of the samples at Örebro University.

The majority of the samples had concentrations below the limit of quantification and were several orders of magnitude lower than surface water samples from Thailand and Uganda. This picture is very uniform and coincides with results from coastal and deep ocean reports from other authors. In their studies the most abundant compound detected was PFOA, which is also our finding. Assessment of the physical parameters showed that changes mainly occur at high depths, when pressure increases, and temperature drops sharply. A slight correlation can be implied from these data since the highest PFOA concentrations were found at the deeper sampling points. However, in general, the difference in concentrations at different depths did not seem to factor into the overall distribution of the target PFAS in this project. The concentrations found in this study were several order or magnitude lower than any regulatory values. The effects of these low concentrations under these conditions on aquatic organisms or the environment have not been studied so far and most likely are not to occur. In a follow-up to further elucidate the absence of PFASs in the Mozambique and Tanzania channels, it is suggested to pool remaining water samples according to vertical or horizontal vectors in order to have more volume. By maintaining the low blank levels and with careful concentration of the larger volumes into small injection volumes, it is hoped to obtain more quantifiable values.

The very low levels found in this study may be explained by the huge dilution in oceans of pollutants released from land-based activities or the fact that PFOS, PFOA and PFHxS have not been or are not used widely on the African continent. The answer may be given through the national reports from the countries under the Stockholm Convention on POPs.

(31)

7 Acknowledgements

I would like to thank Henk Bowman and Duan van Aswegen (North-West University, South Africa) for providing the samples analysed in the project and the metadata for the sampling points.

I would also like to thank my supervisors Heidelore Fiedler and Leo Yeung as well as my colleagues Mohammad Sadia and Abeer Baabish for all of the help and guidance given to me over the past months. I really appreciate all the effort to accommodate my studies during this pandemic.

Thank you to everyone at MTM Research Centre who has been with me throughout my academic career.

Finally, thank you to my family and friends, who have always supported me and continued to motivate me.

(32)

8 References

Ahrens, L., Xie, Z. and Ebinghaus, R., 2010. Distribution of perfluoroalkyl compounds in seawater from Northern Europe, Atlantic Ocean, and Southern Ocean. Chemosphere, 78(8), pp.1011-1016.

Benskin, J., Muir, D., Scott, B., Spencer, C., De Silva, A., Kylin, H., Martin, J., Morris, A., Lohmann, R., Tomy, G., Rosenberg, B., Taniyasu, S. and Yamashita, N., 2012.

Perfluoroalkyl Acids in the Atlantic and Canadian Arctic Oceans. Environmental Science & Technology, 46(11), pp.5815-5823.

Buck, R.C., Franklin, J., Berger, U., Conder, J.M., Cousins, I.T., De Voogt, P., Jensen, A.A., Kannan, K., Mabury, S.A., and van Leeuwen, S.P., 2011. Perfluoroalkyl and polyfluoroalkyl substances in the environment: terminology, classification, and origins. Integrated

environmental assessment and management 7, 513-541.

Costantini, D., Blévin, P., Herzke, D., Moe, B., Gabrielsen, G., Bustnes, J. and Chastel, O., 2019. Higher plasma oxidative damage and lower plasma antioxidant defences in an Arctic seabird exposed to longer perfluoroalkyl acids. Environmental Research, 168, pp.278-285. de Ruijter, W., Ridderinkhof, H., Lutjeharms, J., Schouten, M. and Veth, C., 2002.

Observations of the flow in the Mozambique Channel. Geophysical Research Letters, 29(10), pp.140-1-140-3.

D'eon, J., Hurley, M., Wallington, T. and Mabury, S., 2006. Atmospheric Chemistry ofN-methyl Perfluorobutane Sulfonamidoethanol, C4F9SO2N(CH3)CH2CH2OH: Kinetics and Mechanism of Reaction with OH. Environmental Science & Technology, 40(6), pp.1862-1868.

European Commission, 2008. For references section: Regulation (EC) No. 1272/2008 on classification, labelling and packaging of substances and mixtures. OJ L 353, 31.12.2008, p. 1. Consolidated text to 01.01.2020:

https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:02008R1272-20200101

Fiedler, H., Kallenborn, R., Boer Jacob, d., and Sydnes Leiv, K. (2019). The Stockholm Convention: A Tool for the Global Regulation of Persistent Organic Pollutants. Int Chemistry International, pp. 4

González-Gaya, B., Dachs, J., Roscales, J., Caballero, G. and Jiménez, B., 2014. Perfluoroalkylated Substances in the Global Tropical and Subtropical Surface Oceans. Environmental Science & Technology, 48(22), pp.13076-13084.

IARC (2018). Perfluorooctanoice acid, Vol 110-01 (Lyon, France: IARC). www.iarc.org IARC (2016). Polychlorinated Biphenyls and Polybrominated Biphenyls. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans Volume 107.

https://publications.iarc.fr/131

Ju, X., Jin, Y., Sasaki, K. and Saito, N., 2008. Perfluorinated Surfactants in Surface, Subsurface Water and Microlayer from Dalian Coastal Waters in China. Environmental Science & Technology, 42(10), pp.3538-3542.

(33)

Maina, J., Gamoyo, M., Adams, V., D'agata, S., Bosire, J., Francis, J. and Waruinge, D., 2019. Aligning marine spatial conservation priorities with functional connectivity across maritime jurisdictions. Conservation Science and Practice, 2(2).

McCarthy, C., Kappleman, W. and DiGuiseppi, W., 2017. Ecological Considerations of Per- and Polyfluoroalkyl Substances (PFAS). Current Pollution Reports, 3(4), pp.289-301. Nobels, I., Dardenne, F., Coen, W. and Blust, R., 2010. Application of a multiple endpoint bacterial reporter assay to evaluate toxicological relevant endpoints of perfluorinated compounds with different functional groups and varying chain length. Toxicology in Vitro, 24(6), pp.1768-1774.

OECD, 2013. OECD/UNEP Global PFC Group, Synthesis paper on per- and polyfluorinated chemicals (PFCs). Environment, Health and Safety, Environment Directorate, OECD. Available at: https://www.oecd.org/env/ehs/risk-management/PFC_FINAL-Web.pdf

OECD, 2018. Toward a New Comprehensive Global Database of Per- and Polyfluoroalkyl Substances (PFASs): Summary Report on Updating the OECD 2007 List of Per- and

Polyfluoroalkyl Substances (PFAs). In OECD Environment, Health and Safety Publications Series on Risk Management No 39, P.a.B. Environment Directorate Joint Meeting of the Chemicals Committee and the Working Party on Chemicals, ed. (Paris, France: Environment Directorate, ORGANISATION FOR ECONOMIC COOPERATION AND

DEVELOPMENT, Paris 2018), pp. 24. Available at

https://www.oecd.org/officialdocuments/publicdisplaydocumentpdf/?cote=ENV-JM-MONO(2018)7&doclanguage=en

Ourworld.unu.edu. 2020. Persistent Organic Pollutants In The Marine Food Chain - Our World. [online] Available at: <https://ourworld.unu.edu/en/persistent-organic-pollutants-in-the-marine-food-chain> [Accessed 15 March 2020].

Prevedouros, K., Cousins, I., Buck, R. and Korzeniowski, S., 2006. Sources, Fate and Transport of Perfluorocarboxylates. ChemInform, 37(11).

Sadia, M., Yeung, L. and Fiedler, H., 2020. Trace level analyses of selected perfluoroalkyl acids in food: Method development and data generation. Environmental Pollution, 263, p.113721.

Sætre, R. and Da Silva, A., 1984. The circulation of the Mozambique channel. Deep Sea Research Part A. Oceanographic Research Papers, 31(5), pp.485-508.

Theobald, N., Caliebe, C., Gerwinski, W., Hühnerfuss, H. and Lepom, P., 2011. Occurrence of perfluorinated organic acids in the North and Baltic seas. Part 1: distribution in sea water. Environmental Science and Pollution Research, 18(7), pp.1057-1069.

UNEP (May 10, 2020):

http://www.pops.int/Countries/StatusofRatifications/PartiesandSignatoires/tabid/4500/Default .aspx

(34)

van der Schyff, V., Kwet Yive, N., Polder, A., Cole, N. and Bouwman, H., 2020. Perfluoroalkyl substances (PFAS) in tern eggs from St. Brandon's Atoll, Indian Ocean. Marine Pollution Bulletin, 154, p.111061.

Weiss, J., de Boer, J., Berger, U., Muir, D., Ruan, T., Torre, A., Smedes, F., Vrana, B., Clavien, F. and Fiedler, H., 2015. PFAS Analysis In Water For The Global Monitoring Plan Of The Stockholm Convention. Geneva: Chemicals Branch United Nations Environment Programme (UNEP) Division of Technology, Industry and Economics.

Yamashita, N., Kannan, K., Taniyasu, S., Horii, Y., Okazawa, T., Petrick, G. and Gamo, T., 2004. Analysis of Perfluorinated Acids at Parts-Per-Quadrillion Levels in Seawater Using Liquid Chromatography-Tandem Mass Spectrometry. Environmental Science & Technology, 38(21), pp.5522-5528.

Yamashita, N., Taniyasu, S., Petrick, G., Wei, S., Gamo, T., Lam, P. and Kannan, K., 2008. Perfluorinated acids as novel chemical tracers of global circulation of ocean waters.

Chemosphere, 70(7), pp.1247-1255.

Yamazaki, E., Taniyasu, S., Ruan, Y., Wang, Q., Petrick, G., Tanhua, T., Gamo, T., Wang, X., Lam, P. and Yamashita, N., 2019. Vertical distribution of perfluoroalkyl substances in water columns around the Japan sea and the Mediterranean Sea. Chemosphere, 231, pp.487-494.

Yeung, L., Dassuncao, C., Mabury, S., Sunderland, E., Zhang, X. and Lohmann, R., 2017. Vertical Profiles, Sources, and Transport of PFASs in the Arctic Ocean. Environmental Science & Technology, 51(12), pp.6735-6744.

You, C., Jia, C. and Pan, G., 2010. Effect of salinity and sediment characteristics on the sorption and desorption of perfluorooctane sulfonate at sediment-water interface. Environmental Pollution, 158(5), pp.1343-1347.

Young, C. and Donaldson, D., 2007. Overtone-Induced Degradation of Perfluorinated

(35)

9 Annex: Supplementary Information

Figure 11: LC/MS chromatogram for native PFASs standard (UNEP-PFAS-SOLN. A). Peak J corresponds to br-PFOS and peak K to L-PFOS

Table 7: Isomeric composition of native PFOS standard (UNEP-PFAS-SOLN. A) (Wellington Laboratories, Guelph, Canada)

Abbreviation Name Percent composition

L-PFOS Potassium perfluoro-1-octansulfonate 78.8 78.8 1-PFOS Potassium 1-trifluoromethylperfluoroheptanesulfonate 1.2

21.1 2-PFOS Potassium 2-trifluoromethylperfluoroheptanesulfonate 0.6

3 PFOS Potassium 3-trifluoromethylperfluoroheptanesulfonate 1.9 4 PFOS Potassium 4-trifluoromethylperfluoroheptanesulfonate 2.2 5 PFOS Potassium 5-trifluoromethylperfluoroheptanesulfonate 4.5 6 PFOS Potassium 6-trifluoromethylperfluoroheptanesulfonate 10 5,5 PFOS Potassium 5,5-di(trifluoromethyl)perfluorohexanesulfonate 0.2 4,4 PFOS Potassium 4,4-di(trifluoromethyl)perfluorohexanesulfonate 0.03 4,5 PFOS Potassium 4,5-di(trifluoromethyl)perfluorohexanesulfonate 0.4 3,5-PFOS Potassium, 3,5-di(trifluoromethyl)perfluorohexanesulfonate 0.07

Figure 12: Chromatogram: separation and quantification of L-PFOS and relevant br-PFOS isomers S1 Time 6.60 6.80 7.00 7.20 7.40 7.60 7.80 8.00 8.20 8.40 8.60 8.80 % 0 100

TQS_161026_LY_003 3: MRM of 6 Channels ES-

TIC 7.73e6 7.89 7.66 7.58 7.32 br-PFOS L-PFOS

(36)

Table 8: Detected peaks in chromatogram (raw data)

Sample ID L-PFOS br-PFOS Sum PFOS PFOA PFHxS 3., 4-, 5-PFOS 2-, 6-PFOS MZ_M1-08_C 5.3 2.2 7.5 17.8 9.3 2.2 0.0 MZ_M1-08_D 7.6 2.7 10.2 16.4 8.4 2.7 0.0 MZ_M1-10_A 4.4 5.0 9.4 19.5 7.8 4.4 0.6 MZ_M1-10_B 4.6 1.0 5.7 20.6 7.1 1.0 0.0 MZ_M1-10_C 1.8 2.0 3.8 20.2 6.4 1.0 1.0 MZ_M1-10_D 3.0 2.6 5.6 14.9 9.0 2.2 0.4 MZ_M4-01_A 5.0 1.4 6.4 19.7 9.4 1.2 0.2 MOZ_M04-1_D 37.7 8.6 46.4 63.0 0.0 8.62 MZ_M9-05_B 6.9 4.3 11.2 27.6 10.8 3.0 1.2 MZ_M9-05_C 5.5 2.1 7.6 30.5 14.2 2.1 0.0 MZ_M9-05_D 0.0 3.3 3.3 17.6 14.9 2.2 1.0 MZ_M11-01_A 4.6 3.3 7.8 28.9 14.4 3.3 0.0 MZ_M11-01_C 0.0 2.1 2.1 22.3 2.1 0.0 MZ_M11-01_D 0.0 6.4 6.4 38.5 9.5 6.4 0.0 MZ_M12-01_A 0.0 0.2 0.2 22.9 0.2 0.0 MZ_M12-01_B 0.0 4.6 4.6 23.2 8.4 2.8 1.8 MZ_M12-01_D 4.4 1.8 6.2 24.5 7.6 1.8 0.0 MZ_M14-01_A 5.1 4.9 9.9 23.0 11.1 3.6 1.2 MZ_M14-01_B 4.8 3.6 8.4 20.6 7.8 3.0 0.6 MZ_M14-01_C 5.6 3.3 8.9 26.0 7.4 3.3 0.0 MZ_M14-01_D 2.3 2.7 5.0 14.5 7.4 2.7 0.0 TZA_T1-01_A 4.1 2.9 6.9 16.7 6.7 2.9 0.0 TZA_T01-1_C 24.0 8.6 32.6 47.0 0.0 8.62 TZA_T01-1_D 47.0 16.2 63.2 42.2 0.0 16.16 TZA_T1-03_A 9.3 1.9 11.1 20.2 1.9 0.0 TZA_T1-03_B 5.4 3.0 8.5 20.0 3.0 0.0 TZA_T1-03_C 4.8 1.8 6.6 26.1 23.1 1.0 0.8 TZA_T1-03_D 1.8 3.0 4.8 34.8 25.4 3.0 0.0 TZA_T02-3_D 14.2 8.6 22.8 42.2 0.0 8.62 TZA_T9-05_B 9.0 0.2 9.2 25.5 0.0 0.2 TZA_T9-05_C 7.2 3.2 10.4 22.7 8.6 1.0 2.2 TZA_T9-05_D 4.3 0.9 5.1 30.7 8.1 0.2 0.6 TZA_T9-07_A 9.8 0.8 10.6 28.2 13.0 0.0 0.8 TZA_T9-07_B 7.7 0.9 8.6 42.9 10.3 0.0 0.9 TZA_T9-07_C 6.7 0.2 6.9 22.8 7.3 0.0 0.2 TZA_T9-07_D 3.1 1.2 4.3 38.5 2.6 0.6 0.6 TZA_T10-01_A 0.0 0.0 0.0 45.6 14.8 0.0 0.0 TZA_T10-01_C 0.0 9.5 9.5 39.0 4.3 9.5 0.0 TZA_T10-01_D 0.9 5.0 5.8 23.5 9.1 2.4 2.6 TZA_T10-02_A 42.4 5.9 48.3 53.6 7.1 5.9 0.0 TZA_T10-02_C 0.0 5.3 5.3 28.6 6.9 5.3 0.0 TZA_T10-02_D 4.7 3.3 8.0 22.2 6.7 2.7 0.6 TZA_T10-03_A 11.8 10.4 22.2 21.3 9.7 7.6 2.7 TZA_T10-03_B 5.0 0.0 5.0 36.2 0.0 0.0 TZA_T10-03_C 4.1 1.6 5.8 20.6 1.6 0.0 TZA_T10-03_D 11.8 3.5 15.4 20.3 1.0 2.5

(37)

Table 9: Recoveries of internal standard in samples

Sample ID IS_PFHxS IS_PFOS IS_PFOA Sample ID IS_PFHxS IS_PFOS IS_PFOA

% Recovery % Recovery MZ_M1-08_C 96.7 91.3 95.2 TZA_T1-01_A 85.2 80.9 75.8 MZ_M1-08_D 108.7 101.7 92.1 TZA_T9-05_B 139.4 127.5 146.5 MZ_M1-10_A 117.5 108.9 92.3 TZA_T9-05_C 109.8 123.3 114 MZ_M1-10_B 90.7 102.9 109.2 TZA_T9-05_D 136.4 140.8 138.8 MZ_M1-10_C 127.5 124.1 107.3 TZA_T9-07_A 143.3 140.9 136 MZ_M1-10_D 125.4 119.1 108.3 TZA_T9-07_B 124.5 117.1 128.5 MZ_M4-01_A 101 94.6 78.4 TZA_T9-07_C 136.6 117.7 139.1 MZ_M9-05_B 110.9 107.1 81.5 TZA_T9-07_D 145 145 125.1 MZ_M9-05_C 54.1 62.4 77.5 TZA_T10-01_A 66.1 40.9 103.8 MZ_M9-05_D 76.8 86.2 91 TZA_T10-01_C 125.4 219.3 128.8 MZ_M11-01_A 79.1 77.7 72.1 TZA_T10-01_D 143.6 320.5 137.6 MZ_M11-01_C 13.9 47.3 90.5 TZA_T10-02_A 139.4 309.3 148.6 MZ_M11-01_D 78.8 81.1 72.9 TZA_T10-02_C 135.4 210.8 136.1 MZ_M12-01_A 3.6 21.1 76.9 TZA_T10-02_D 132.8 126.8 142.8 MZ_M12-01_B 82.9 82.2 61.8 TZA_T10-03_A 131.6 166.3 136.9 MZ_M12-01_D 65.3 70 68.2 TZA_T10-03_B 1.3 12.8 86.9 MZ_M14-01_A 73 74.8 60.3 TZA_T10-03_C 4.8 34.4 156.8 MZ_M14-01_B 60 76.1 66 TZA_T10-03_D 0.9 9.5 81 MZ_M14-01_C 71.7 71.5 66.5 TZA_T1-03_A 3.3 18.4 76.5 MZ_M14-01_D 99.8 99 100.7 TZA_T1-03_B 1.1 9.6 68.2 TZA_T1-03_C 24.7 61.8 107.7 TZA_T1-03_D 27.1 65.5 107.5

Table 10: Recoveries of internal standard in QC samples

Internal standard QC 2 QC 4 QC 5 QC 7 QC 8 QC 9

% Recovery

PFHxS 105 61 72 144 129 47

PFOS 102 64 75 136 136 71

PFOA 100 86 71 140 155 95

Table 11: Individual and ∑PFAS concentration (pg/L) in surface ocean water samples obtained for the Malaspina 2010 expedition

PFHxS PFOA PFOS ΣPFAS

Ocean (median) pg/L pg/L pg/L pg/L Atlantic 21.1 45 191 645 Indian 32.4 25.4 101 527 Pacific 6.65 23.2 88.5 329 Hemisphere (median) North 11.2 33.9 69.3 708 South 15.5 33 274 1620

Occurrence of major perfluoroalkyl substances (PFAS) in water samples from a transect of the Indian Ocean along the East African coast

Örebro University, School of Science and Technology

Master of Chemistry in Environmental Forensics 2019-2020

Occurrence of major perfluoroalkyl substances (PFAS) in water

samples from a transect of the Indian Ocean along the East

African coast

Occurrence of major perfluoroalkyl substances (PFAS) in water

samples from a transect of the Indian Ocean along the East

African coast

Contents

Table of Figures

Table of Tables

Abbreviations

1 Introduction

2 Background

3 Materials and Methods

4 Results

MZ_M1-10

5 Discussion:

National Samples

6 Conclusion

7 Acknowledgements

8 References

9 Annex: Supplementary Information