Chiral Persistent Organic Pollutants as Tracers of
Transport and Accumulation Processes
Channa Yath
Student
Master Thesis 30 ECTS Report passed:
Supervisor: Patrik Andersson
Abstract
Fiber bank region in the northern Baltic Sea were contaminated with persistent organic pollutants (POPs) formerly released from the forest products industries. Today POPs in fiber bank sediments are contaminating benthic biota and reach food web. Sediment-biota accumulation experiments were done using contaminated sediments from estuaries in fiber banks region and two invertebrates, Macoma balthica and Marenzelleria spp.. Partitioning among sediment, pore water and biota was measured for polychlorinated biphenyls (PCBs, 41 congeners) and hexachlorobenzene (HCB). The accumulation of PCBs and HCB by biota related to quality of sediment, meaning that they were less bioavailable from sediment with stronger sorption. Marenzelleria spp. was more contaminant than Macoma balthica due to behavior and feeding strategies. Enatiomerspecific analysis of seven chiral PCBs (PCB-91, 95, 132, 136, 149 174 and 176) was conducted to investigate metabolism and partitioning. Some chiral PCBs showed evidence of enantioselective degradation. Most sediment analysis were not conclusive due to chromatographic interferences, but there was generally good agreement between the enantiomer fraction (EF = enantiomer 1/enantiomer 2) in pore water and Macoma balthica for PCBs-132, 136 and 149. In particular, PCB-132, 149 and 174 were clearly enantioselective biodegredation in Macoma balthica. Enantioselective biodegradation was only observed for PCB-149 in Marenzelleria spp. from Kramfors. Other chiral PCBs were below detection in Marenzelleria spp. form Kramfors and the other locations.
III
List of abbreviations
BAF bioaccumulation factor
BSAF biota-sediment accumulation factor
BSAF_Mac_K biota-sediment accumulation factor of Mac_K BSAF_Mac_N biota-sediment accumulation factor of Mac_N BSAF_Mar_K biota-sediment accumulation factor of Mar_K BSAF_Mar_N biota-sediment accumulation factor of Mar_N
Clip concentration of target compound in biota (normalized to lipid
content, ng/g lipid)
CPOM concentration of target compound that it is adsorbed by POM,
ng/g POM
CPW concentration of target compound in pore water, ng/ml
Csed concentration of target compound in sediment ng/g d.w
DDT 1,1,1-trichloro-2,2-bis(4-chlorophenyl)ethane E1 and E2 first and second eluates enantiomers in GC-MS
EF enantiomer fraction
flid lipid fraction in dry weight
fOC organic carbon fraction
GC-HRMS gas chromatography-high resolution mass spectrometry GC-LRMS gas chromatography-low resolution mass spectrometry (MSD) H Henry’s law constants, Pa.m3/mol
HCB hexachlorobenzene
HCHs hexachlorocyclohexanes
KD partitioning coefficient between of sediment and pore water,
ml/g
KOC partitioning coefficient between sediment organic carbon and
pore water, ml/g
Kow octanol-water partition coefficient, dimensionless
KPOM partition coefficient between POM and pore water, ml/g
LOD limit of detection LOQ limit of quantification
m/z mass/charge number
Mac_K Macoma balthica at Kramfors Mac_N Macoma balthica at Norrbyn Mac_O Macoma balthica at Örnsköldsvik Mar_K Marenzelleria spp. at Kramfors Mar_N Marenzelleria spp. at Norrbyn Mar_O Marenzelleria spp. at Örnsköldsvik
MSD Mass selective detector (quadrupole mass spectrometer) NGOs non-governmental organizations
OC organic carbon
PCBs polychlorinated biphenyls
PCDD/PCDF polychlorinated dibenzo-p-dioxins and dibenzofurans
PEL probable effect level
POM polyoxymethylene
POPs persistent organic pollutants PW_K pore water at Kramfors
PW_N pore water at Norrbyn
PW_O pore water at Örnsköldsvik
Se_K Kramfors sediment
Se_N Norrbyn sediment Se_O Örnsköldsvik sediment
IV
V
Contents
Abstract ... I
List of abbreviations ... II
Contents ... V
1.
Introduction ... 1
1.1.
Target compounds ... 3
1.1.1
Polychlorinated biphenyls (PCBs) ... 3
1.1.2
Hexachlorobenzene (HCB) ... 4
1.2.
Species of interested for this thesis ... 5
1.2.1
Marenzelleria spp. ... 5
1.2.2
Macoma balthica ... 5
1.3
Polyoxymethylene ... 6
1.4
Aim of the diploma work ... 6
2.
Popular scientific summary including social and ethical aspects ... 6
2.1.
Popular scientific summary ... 6
2.2.
Social and ethical aspects ... 7
3.
Experiments ... 8
3.1
Chemicals and materials... 8
3.2
Sediment sampling and pre-treatment ... 8
3.3
Test species ... 9
3.4
Bioaccumulation experiment... 9
3.4.1
Sediment analysis... 10
3.4.2
Organic carbon analysis ... 10
3.4.3
Pore water analysis ... 10
3.4.4
Biota analysis ... 11
3.5
Instrumental analysis ... 11
3.6
Quality control... 12
3.7
Data analysis ... 13
4.
Results and Discussions ... 14
4.1
Concentration of PCBs and HCB in sediments ... 14
4.2
Concentration of PCBs and HCB in pore water ... 18
4.3
Concentration of PCBs and HCB in biota ... 21
4.3.1
Bioaccumulation factor ... 22
4.3.2
Biota-sediment accumulation factor ... 25
4.4
Chiral PCBs as fate in benthic zone ... 26
4.4.1
Chiral PCBs in sediments ... 26
4.4.2
Chiral PCBs in pore water ... 28
4.4.3
Chiral PCBs in biota ... 29
5.
Conclusions ... 33
6.
Acknowledgment ... 33
7.
Appendix ... 34
1
1. Introduction
Since the science was developed, new compounds have been produced every year by chemists. These chemicals are produced for different purposes, some used for drugs, pesticides, herbicides, fungicides, clothes, dyes, food additives and other use materials. Most of these chemicals were produced in developed countries. The biggest disadvantages of processing are by-products, wastes and the persistence of these chemicals in environment. The negative effects such as cancer, reduce the reproductive abilities can occur by exposure to this wastes, by-produces and persistent chemicals. Nowadays, these problems are of high concern for our environment, so scientists are trying to solve these problems.
Persistent organic pollutants (POPs) are one group of chemicals that have the most negative effects to environment, including humans and wildlife. According to Stockholm Convention in 2001,” persistent chemicals” means that their half-lives are longer than two months in water and six months in sediments and soils. The convention initially included twelve compounds as POPs: aldrin, chlordane, dieldrin, endrin, heptachlor, hexachlorobenzene (HCB), mirex, toxaphene, polychlorinated biphenyls (PCBs), 1,1,1-trichloro-2,2-bis(4-chlorophenyl)ethane (DDT), polychlorinated dibenzo-p-dioxins and dibenzofurans (PCDD/PCDFs) (UNEP, 2001). Stockholm Convention 2009 added some new chemicals as POPs: hexachlorohexanes (HCH: alpha-HCB, beta-HCH, gamma-HCH, also known as lindane) chlordecone, hexabromobiphenyl, tetra-, penta, hexa- and heptabromodiphenyl ethers, pentachlorobenzene and perfluoroocatane sulfonic acid (PFOS), its salts and perfluorooctane sulfonyl flouride (UNEP, 2009). Many of these compounds were banned long time ago but they are still detected in the environment because they have long half-lives (half-life of some compounds are many years) (Wiberg et al., 2009; Sobek et al., 2015).
After use, POPs are distributed world-wide by water and air flow. Due to their physical and chemical properties, like high hydrophobicity (high octanol-water partition coefficients, log Kow) and low Henry’s law constants (H), POPs, (e.g. PCBs, DDTs,
HCB, HCHs and chlordane) end up in soil and sediment which contain high sorptive materials such as fiber, black carbon, clay and organic carbon (Wiberg, 2002; Yang et al., 2005, Warner and Wong, 2006). These pollutants can diffuse from sediment to pore water and reach biota (Meijer et al., 2006; Zhang et al., 2003; Josefsson et al., 2010; Kupryianchyk et al., 2013, Lydy et al., 2014). However, the mechanism of this transfer is not really understood.
2 forestry companies grew up and the negative effects of the wastes and cutting down the trees from the forestry companies became the big debates between Swedish NGOs (non-governmental organizations) and the companies. After 1993, the Swedish Parliament decided to make the policy to protect the environments (Elliott and Schlaepfer, 2001). Nowadays, the wood industries are separated to sawmill industry, wood board industry, pulp and paper industry, and wood packaging products. The pulp and paper industries in Swedish are the largest companies of the wood industries. Generally, the paper and pulp industries need large quantities of water and use chlorine bleaching and wood preservatives during the processing of the products; that is the main reason for discharging waste water which contains fiber contain some POPs (Apler et al., 2014). The discharging of the waste directly from the companies is the most important factor that has resulted in fiber and POPs accumulation in some coastal areas of the Baltic Sea. These locations that contain high fiber waste are called fiber banks. Due to the POPs are highly hydrophobic compounds and the fiber bank locations are content high organic carbon and POPs, it is importance to study POPs fate at the fiber bank locations.
The Baltic Sea has high fiber content sediment on sea floor and pollution history by POPs so it is one of among highly POPs polluted places (Strandberg et al., 1998). Nowadays, POPs are still high concern because of their negative effect on ecosystems (Wiberg et al., 2009) and pollution histories (Strandberg et al., 1998). For instance, the reproductive abilities of both sexes of gray and ringed seals in the Baltic Sea were decreased during 1980s which was due to the fact that gray and ringed seals were exposed to POPs, especially PCBs and DDTs (Wiberg et al., 2002). These pollutants have increased in benthic organisms, fish and sea eagle eggs in the fiber banks region of the Bothnian Sea (Sundqvist et al., 2009). The concentration of HCB in blue mussel in Nidingen and concentration of 1,1-bis-(4-chlorophenyl)-2,2-dichloroethene (DDE) in Kvädöfjärden perch increased in the last decade (Nyberg et al., 2015). In the Baltic Sea, DDTs and PCDD/PCDFs have been shown to remobilize to the water column where they have contaminated benthic biota and accumulated in the food webs (Sundqvist et al., 2009). Remobilization of PCBs from sediments in the Baltic Sea was hypothesized to be the result of benthic species such as Marenzelleria spp. (Josefsson et al., 2010). This contradicts that the general trend in the Baltic Sea showing that contaminants from past industrial activities are buried in deep sediments with reduced availability. Possible reactivation created an urgent need to understand the processes involved in the recycling of contaminants in benthic zone.
3 have the same atoms and bonds but different 3D structures (non-superimposable mirror images) and no plane of symmetry. Most chiral POPs were produced as racemic (1:1) mixtures of two enantiomers. Abiotic processes can-not change the racemic ratio, while biotic processes involving enzymatic degradation can result in nonracemic enantiomer propertion (Bidleman et al., 2013; Borgå and Bidleman, 2005; Hühnerfuss and Shah, 2009, Lehmler et al., 2010; Wiberg, 2002; Warner and Wong, 2006; Yang et al., 2011). Due to these processes, chiral POPs can be used as tracers of metabolism, bioaccumulation and transport processes in the benthic zone.
1.1. Target compounds
PCBs, chlordanes, HCHs, HCB and DDT were the primary compounds of interested in the present study as these are of high concern and were identified in a pilot study of fiber bank sediments (Darya Kupryianchyk, personal communication). However, due to time limitation, only PCBs and HCB were investigated in this project.
1.1.1 Polychlorinated biphenyls (PCBs)
PCBs are a group of compounds with two phenyls connected by an α-bond and one to ten of chlorines connected to the phenyls (Figure 1 (A)). Since the 1920s, more than 1 million tons of PCBs were produced for many industrial applications, mainly as dielectric fluid (coolants and lubricant) for transformers and capacitors because of they are stable compounds (Wiberg, 2002; Wu et al., 2008). The sources of PCBs into the environment were emissions from industry, leaking transformers and capacitors and other products in which they were used (Wiberg et al., 2009). PCBs are toxic chemicals, with carcinogenic and mutagenic effects on the reproductive, nervous and endocrine system (Shou-Hui et al., 2012). PCBs were restricted by individual countries in the 1980s, and added to the Stockholm Convention in 2001 due to their persistence and negative effects on the environment. The main sinks of PCBs are soil and sediment due to high hydrophobicity (high log Kow) and low Henry’s law constant.
Nowadays, the secondary sources of PCBs are re-volatilization from contaminated soil, sediments, and combustion and distribution by air (Wiberg et al., 2009). The Baltic Sea is one place that has been highly polluted by PCBs. However, contamination by PCBs has decreased in Baltic Sea sediment since the starting of monitoring program in 1970s (Wiberg et al., 2009).
4 on the ortho-positions (position 2, 6, 2’ and 6’) of the biphenyl rings. This hinders rotation about the α-bond between two phenyl rings so they cannot move freely, and can result in asymmetric arrangement of chlorines, depending on substitution position (Shou-Hui et al., 2012; Wiberg, 2002). Figure 1 (B) is an example of atropisomeric PCB-135 (Shou-Hui et al., 2012).
Figure 1: (A) PCB structure and (B) Atropisomers of PCB-135.
1.1.2 Hexachlorobenzene (HCB)
HCB (C6Cl6) is the compound produced by substituting all hydrogen atoms of benzene
are substituted by chlorine atoms (Figure 2). HCB has moderate hydrophobicity (log Kow=5.7) compares with PCBs but it still can end up in soil and sediment. Its relatively
high Henry’s law constant (H=35 Pa.m3.mol-1) results in air and water if compare with
PCBs (Wiberg et al., 2009; Jantunen and Bidleman, 2006). HCB was produced for many purposes such as a fungicide, a lubricant, a wood preservative, a flame retardant processing chemical, a plasticizer for polyvinyl chlorine, and in the production of printing ink (Wiberg et al., 2009; Hirano et al., 2007). Moreover, it is produced from combustion of waste and is a by-product of chemical industrial processes. HCB was banned in the 1980s (Wiberg et al., 2009) and added to the Stockholm convention in 2001.
5
1.2. Species of interested for this thesis
The primary sources of POPs were decreased in the last decade so the secondary sources (soil and sediment) are the largest source of contamination nowadays. There are two ways to contaminate biota by the POPs from sediment, digestion and sorption (Josefsson et al., 2011). Contaminants concentrations may increase or decrease with depth of the sediment in the fiber banks region (Apler et al., 2014) so the biota living at different depths may reflect the depth distribution of contaminants. Due to high sediment populations of Marenzelleria ssp. and Macoma balthica in Baltic Sea (Joelsson, 2013), these organisms were selected for this project.
1.2.1 Marenzelleria spp.
Marenzelleria spp., (Figure 3 (A)) including Marenzelleria vividis, Marenzelleria negleta and Marenzelleria spionide are burrowing polychaetes. These worms were transport by shipping in tanker ballast water from North American to the Baltic Sea in late 1970s and early 1980s. Nowadays, in Baltic Sea, there is large population of Marenzelleria spp. (Bastrop et al., 1997). Normally, the life time of Marenzelleria spp is 3 years if it lives at a depth up to 0.5 m. The principal food is suspension of the surface sediment. Due to the digging of these worms, they can turbate and transport deep sediment and/or pore water to the surface (Joelsson, 2013; Josefsson et al, 2010).
1.2.2 Macoma balthica
6 Figure 3: Marenzelleria spp. (A) and Macoma balthica (B)
Thus, based on the species traits, i.e. the living characteristics of Marenzelleria ssp. and Macoma balthica in the fiber bank and differences in feeding strategies, these two invertebrates are good choices to study the transfer of POPs from sediment to biota.
1.3 Polyoxymethylene
There are many types of passive samplers such as polydimethylsiloxane, polyethylene and polyoxymethylene (POM) (Lydy et al., 2014) to determine the concentration of POPs in water or pore water. POM is a cheap material, shows high adsorption of POPs from pore water (high KPOM), is easy to clean up and has a smooth surface (easy to
clean particles after exposure to sediment), and the sorption of POPs is not related to temperature that experiment (Hawthorne et al., 2009; Lydy et al., 2014). For these reasons POM was chosen to determined concentration of POPs in pore water.
1.4 Aim of the diploma work
The aims of the project are (1) to study sediment-pore water exchange and bioaccumulation of PCBs and HCB in Marenzelleria spp. and Macoma balthica in the fiber banks region of the northern Baltic Sea using quantitative analysis and (2) to apply enantiospecific analysis of chiral PCBs as tracers of PCBs fate and transport processes.
2. Popular scientific summary including social
and ethical aspects
2.1. Popular scientific summary
POPs were produced during the 1920s-1940s for used for various reasons like pesticide, termicides, fungicides, and dielectric fluids and they were banned during late 1970s and early 1980s due to their negative effects like carcinogenicity, developmental effects, reproductive effects and neurodevelopmental effects. POPs are highly hydrophobic and persistent chemicals, therefore, these pollutants are found in high concentrations in fiber bank sediments and biota. In this thesis, PCBs and HCB were studied and these compounds identified in sediments contaminate invertebrates
7 that are living in sediment (Marenzelleria spp. and Macoma balthica) by feeding and diffusion processes. The sediment that had low sorption properties will render higher bioavailability for the compounds by desorption. Moreover, organism characteristics such as ventilation and suspension feeding of surface sediments by Marenzelleria spp. yield higher contamination levels of PCBs and HCB than Macoma balthica which is deposit feeding at sediment surface. The metabolism of PCBs in this study was structure-specific, it means that different structure had different metabolic pathways. In addition, the metabolism of PCBs was also species-specific, it means one chemical had different pathways of biodegradation in different species.
In conclusion, level of POPs in fiber bank sediments depend on the quality of the sediments and characteristics of the studied biota. Moreover, fiber bank sediments need to be continuously monitored because that the composition (qualities) of sediment will change overtime so the contamination pathways of POPs might change overtime.
2.2. Social and ethical aspects
8
3. Experiments
3.1 Chemicals and materials
Heptane (LC-MS Chromasolv) was purchased from Sigma-Aldrich, Germany. Methanol (HPLC grade), iso-octane (analytical reagent grade) and acetone (HPLC grade) were purchased from Fisher Scientific, UK. Dichloromethane (SupraSolve for GC-ECD and FID), toluene (SupraSolve for GC-ECD and FID), cyclohexane (SupraSolve for GC-ECD and FID), sodium azide (extra pure) and sulfuric acid (95-97%, analytical grad) were purchased from Merck, Germany. Granular copper (10+40 mesh) was purchased from Sigma-Aldrich Chemical Company, Inc., USA. Before use, granular copper was cleaned with concentrated hydrochloric acid (37%) until granular cupper was shiny, followed rinsing by three times with Milli-Q water, three times with methanol, three times with cyclohexane and kept in cyclohexane. Granular copper was used on the same day of cleaning. Florisil (100-200 U.S mesh) and hydrochloric acid 37% were purchased from VWR, France. Before use, Florisil was burned at 5500C and
deactivated with Milli-Q water (100 g Florisil:5 g Milli-Q water). Cellulose thimbles from Schleicher & Schuell MicroScience and Cellulose thimbles GE Healthcare Life Science Whatman were purchased from Germany and UK, respectively. POM, 76 µm thickness, was purchased from CS Hyde Company, USA.POM was pre-cleaned with hexane (24 h), then with methanol (24 h) and finally with Milli-Q water. 13C12-labeled
CB-32, 47, 105, 138 and 170 (99% purity) were obtained from LGC Standards GmbH, Germany. Internal standard CB-103 (purity 99.9%) was obtained from Sigma-Aldrich. Individual PCBs (purity 99+ %), HCB (purity 99+ %) and seven chiral PCBs (PCB-91, 95, 132, 136, 149, 174 and 176) (purity 99+ %) were received from AccuStandard (New Haven, CT, USA).
3.2 Sediment sampling and pre-treatment
9 18043’4.80”E), Kramfors (62056’28.03”N, 18043’4.80”E), Norrbyn (63032’86”,
19050’340”E) and each sediment sample was a mixture of multiple grabs (5-6 grabs).
Sediments were sieved with a 500µm mesh sieve, homogenized and stored at 4oC
before the experiment.
Figure 4: Sediment sampling locations (Kramfors, Örnsköldsvik and Norrbyn).
3.3 Test species
Two marine benthic invertebrates (Marrenzelleria ssp. and Macoma bathica) were collected at a clean site, Norrbyn, in July 2015 (one week before the start of the experiment). The animals were acclimatized under test conditions. Randomly selected healthy individuals of about the same size were used for experiment.
3.4 Bioaccumulation experiment
To study sediment-water exchange and bioaccumulation of POPs, a microcosm experiment was set-up using two fiber bank sediments (Kramfors and Örnsköldsvik) and one uncontaminated sediment (Norrbyn) and two benthic species, viz Macoma balthica and Marenzelleria spp. The test system (2L beaker) was filled with 5-7 cm of sediment and then 1 L of sea water was added carefully. Finally, animals, 10 Marenzelleria spp. or 15 Macoma balthica were added to the different systems. The experiment was performed in triplicate, thus, there were 18 systems in total. The experiment was run for four weeks to reach equilibrium (OECD guidelines, 2008). The test was performed in a temperature-controlled room (8oC and 16 h light:8 h dark
cycle).
After 28 days, the experiment was terminated, the water was carefully taken out, and biota were sieved out from the sediment, washed and stored at -20oC before analyses.
10
3.4.1 Sediment analysis
Sediment, 1 g d.w. from Kramfors and Örnsköldsvik and 3 g d.w. from Norrbyn preloaded with recoveries standards (10 µg) was put in the pre-cleaned (Soxhlet extracted with toluene for 4 h) cellulose thimbles and extracted with 200 ml of toluene overnight (at least 16 hours). The extracted was concentrated to 1-2 ml using rotary evaporator (Rotavapor, BÜCHI, Switzerland) and the solvent was exchanged to iso-octane. Afterwards, acid-activated granular copper was used to remove sulfur by adding to extract and stand 1 hour. After that, a column containing 2 g of Florisil was used to clean up the extracts. The column was cleaned with 30 ml of dichloromethane/cyclohexane (20%/80%), then sample was loaded onto the column and eluted with 100 ml of dichloromethane/ cyclohexane (20%/80%). The eluate was concentrated to 1-2 ml using rotary evaporator (Rotavapor, BÜCHI, Switzerland) and the solvent was exchanged to 0.5 ml of iso-octane using a nitrogen evaporator (Organomation, N-EVAPTM111, USA). The internal standard (CB-103) was added and
the sample was analyzed for PCBs and HCB using GC-HRMS for quantitative analysis and GC-LRMS for enantioselective analysis of chiral PCBs.
Note: Sine all sediment samples had chromatographic interferences, additional clean-up step was undertaken with concentrated (95-97%) sulfuric acid. The sediment extracts were diluted with 1.5 ml of cyclohexane, then 1 ml of sulfuric acid was added, the mixture was homogenized and left to stand 1 hour. The organic phase was collected, concentrated and the solvent was exchanged to iso-octane (100 µl) using the nitrogen evaporator.
3.4.2 Organic carbon analysis
Organic carbon (OC) was determined using a robust method of loss-on-ignition (LOI) at 5500C. The sediment sample 1 g (wet weight) was dried at 1050C overnight (at least
16 hours) to remove water and the residue was weighed. The residue was burned at 5500C for 2 hours to determine LOI (Method descriptions (Appendix: 11.1, 2009);
Bojko and Kabala, 2014).
3.4.3 Pore water analysis
11 HCB using GC-HRMS for quantitative analysis and GC-LRMS for enantioselective chiral PCBs.
3.4.4 Biota analysis
Biota extraction was done by Darya Kupryianchyk. In brief, biota samples were extracted with diethyl ether/hexane (1/9) followed by acetone/hexane (5/2), after which the extract was evaporated to dryness. The lipid content was determining gravimetrically and then the lipids were dissolved in hexane. The samples were cleaned on a Florisil column and analyzed for PCBs and HCB GC-HRMS for quantitative analysis and GC-MS for chiral PCBs.
Note: Biota samples were additionally cleaned with sulfuric acid by the process used for sediment samples.
3.5 Instrumental analysis
Samples were analyzed for enantiomers of seven chiral PCBs using GC (Agilent 6890N) equipped with a chiral-phase column (CP-Chirasil-Dex-CB, 25 m x 0.25 mm i.d. (inside diameter), 0.25 μm film (Agilent Technologies)) and low-resolution mass spectrometry (LRMS, Agilent 5975 MSD) with electron impact ionization (EI+). Instrument conditions are presented in table 1:
Table 1: Instrumental conditions for gas chromatography-low resolution mass spectrometry (enantiospecific analysis, GC-MSD)
Conditions Injector port 250 oC, 2 μL injected
Transfer line 250 oC
Carrier gas He at 1.1 mL/min
Oven 90 oC (1 min), 10 oC/min to 160 oC, 1 oC/min to 190 oC (20
min), 10 oC/min to 220 oC (20 min), total running time 81
min.
MS Ion source 230 oC, quadrupole 150 oC
Selected ion mass (m/z)
5-Cl PCBs: 326, 328 (338 and 340 for 13C12-labeled
surrogate standard)
6-Cl PCBs: 360, 362 ( 372 and 374 for 13C12-labeled
surrogate standard)
7-Cl PCBs: 394, 396, (406 and 408 for 13C12-labeled
surrogate standard)
12 capillary column, 0.25 mm i.d., 25 µm film. Instrumental conditions for HCB and PCBs analysis are presented in Table 2.
Table 2: Instrumental conditions of gas chromatography-high resolution mass spectrometer (GC-HRMS)
Conditions
Injector port 2µl, splitless (split after 1.5 min), 2600C
Transfer line 3000C
Carrier gas Helium at 1.2 ml/min
Oven 80 oC, then 20 oC min-1 to 200 oC, 6 oC min-1 to 230 oC,
then 3 oC min-1 to 322 oC, total run time 30.6 min.
Selected ion mass (m/z)
HCB: 283.8102 and 285.8072
3-Cl PCBs: 255.9614, 257.9584 (168.0017 and 269.996 for 13C12-labeled surrogate standard)
4-Cl PCBs: 289.9224, 291.9194 (301.9626 and 303.9597 for 13C12-labeled surrogate standard)
5-Cl PCBs: 325.8804, 327.8775 (337.9207 and 339.9178 for 13C12-labeled surrogate standard)
6-Cl PCBs: 359.8414, 361.8385 (371.8817 and 373.8788 for 13C12-labeled surrogate standard)
7-Cl PCBs: 393.8025, 395.7995 (405.8428 and 407.8398 for 13C12-labeled surrogate standard)
8-Cl PCBs: 427.7635 and 429.7606 9-Cl PCBs:463.7216 and 465.7187 10-Cl PCB: 497.6826 and 499.6797
3.6 Quality control
13 within the range of ±10% of the (m/z)/((m+2)/z) standard ratio, that compound did not have interference.
3.7 Data analysis
Total organic carbon or organic carbon (OC, g) was calculated from LOI (g) as in equation 1 (Bojko and Kabala, 2014):
OC= LOI/2 (1)
The concentration in pore water was calculated from the partitioning coefficient between pore water and POM for each compound that performs as in equation 2:
=
(2)
Where KPOM and are the concentration of target compound that it is adsorbed by POM (ng/g), the partition coefficient between the concentration on the POM (L/kg) and concentration in pore water (Appendix 1), and the concentration of PCBs or HCB in pore water (ng/ml), respectively (Hawthorne et al., 2009; Lydy et al., 2014).
The concentration of individual PCBs and HCB in sediment (Csed, ng/g d.w) and pore
water were used to calculate the partition coefficient between sediment and pore water (KD, ml/g), which describes the sorption ability of sediment (equation 3).
KD =
(3)
KOC (organic carbon-water partition coefficient, ml/g) was calculated by equation 4,
which normalizes KD to the fraction of organic carbon in the sediment.
KOC =
(4)
Where fOC is organic carbon fraction (Zhang et al., 2013).
The concentration of individual PCBs and HCB in pore water and biota were used to calculate the bioaccumulation factor (BAF) as equation 5:
BAF =
(5)
Clip is lipid-normalized concentration of PCBs and HCB in the biota (ng/g lipid)
(Zhang et al., 2013; Wu et al., 2008).
The concentrations of each PCB congener and HCB in biota and sediment were used to calculate the biota-sediment accumulation factor (BSAF), and this factor was normalized to organic fraction in sediment (fOC) and lipid fraction in biota (flip) that
shown as in equation 6 (Josefsson et al., 2011). BSAF =
(6)
14 Peak areas of each enantiomer pair of chiral PCBs were used to calculate the enantiomer fraction (EF) as in equation 7.
EF=
(7)
E1 and E2 are the first and second eluating enantiomers that are separated by the chiral column. If EF= 0.5 (E1=E2), the enantiomer ratio is called racemic (Borgå and Bidleman, 2005; Hühnerfuss and Shah, 2009).
4. Results and Discussions
4.1 Concentration of PCBs and HCB in sediments
The total concentration of 41 congeners of PCBs were 4.09 ± 0.04 ng/g d.w. (range 0.018-0.522) in Norrbyn sediment and 47.7 ± 3.17 ng/g (range 0.07-5.5 ng/g) in Kramfors sediment. The concentration of seven most abundant PCBs (ƩPCB7: PCB-28, 52, 101,118, 138, 153 and 180) was 1.81 ± 0.03 ng/g and 20.4 ± 1.45 ng/g in Norrbyn and Kramfors sediments, respectively. Even after additional clean up steps, it was not possible to analyzed PCBs in Örnsköldsvik sediment due to chromatographic interferences. According to the Swedish classification of contaminated sediments, Norrbyn sediment belongs to Class III (moderately polluted) and Kramfors sediment to Class V (highly polluted) (Naturvårdsverket, 1999).
The results of chemical analysis ƩPCB7 in Kramfors sediment were similar to those
reported previously for this location. For example, in Apler et al., (2014) report, ƩPCB7
was range from 6.1 to 72 ng/g (average was 30 ng/g). PCB concentration in sediment in different locations around the world is presented in Table 3. The concentration of Norrbyn and Kramfors sediments were in the middle range of contaminated sediments around the world. Concentration of PCBs in Norrbyn and Kramfors sediments were below the probable effect level (PEL) (277 ng/g, Sediment Quality Guideline) (Sapozhnikova and Schlenk, 2004). PEL is the concentration level above which adverse effects frequently (Sapozhnikova and Schlenk, 2004).
15 around 20% (nona and decachlorinated congeners were not detected), it showed that concentration of PCBs was increased from tri-chlorinated to hexa-chlorinated and decreased from hexa-chlorinated to octa-chlorinated (Figure 5 and Appendix 3). This pattern is similar to that in Aroclor 1260 (Commercial name of PCBs) indicating that the source of PCBs might be from Aroclor 1260 (Appendix 3). This pattern is similar to those observed in previous studies at the Baltic Sea (Zhang et al., 2013; Magnusson et al., 2006; Sobek et al., 2015) and NW Mediterranean (Sole et al., 2013) but different from that reported on the Singapore’s coast sediment where the concentration of PCBs was increased with increasing degree of chlorination (Wurl and Obbard, 2005). The high abundance of penta-, hexa-, and hepta-PCBs and the deviation of PCBs patterns from Arochlor 1260 (Appendix 3) can be explained by the degradation processes (reductive dechlorination) occurring in the sediment (Sobek et al., 2015).
Table 3: PCB concentration (ng/g d.w.) in sediments in this study and other locations around the world.
HCB concentration in Norrbyn and Kramfors sediments were 0.00471 ± 0.0003 ng/g and 0.0009 ± 0.0002 ng/g, respectively. The Baltic Sea was the place that not directly polluted by HCB product (Vogel, 2015), so in this case the contaminant source of HCB was the long rang air transportations. Meijer et al., (2006) found that the high altitude lake (lower temperature than low altitude lake and no contaminated history) had more contamination of low chlorinated PCBs than low altitude lake, so it contradicted that the deposited from the air of these congeners was the main contaminant source. Similarity, in this study, HCB has similar hydrophobicity and higher Henry law’s
Location Year ƩPCBs (ng/g) Reference
Baltic Sea, Sweden
Norrbyn
Kramfors
4.09 ± 0.04
47.7 ± 3.17 this study
Baltic Sea, Sweden 2010 the coastal areas 0.71- 28
the offshore area 1.2 -16 Sobek et al., 2015 The NW Mediterranean,
England 2011 5.5 Sole et al., 2013
Minjiang River Estuary,
Southeast China 1999 34.5 Zhang et al., 2003
Canada and UK 2002 0.03–23 Wong et al., 2009
Lake Ontario, Canada 1998 100 Marvin et al., 2003
Salton Sea, California,
USA 2000-2001 217 Sapozhnikova and Schlenk, 2004 Mekong River delta,
South Vietnam 2003-2004 0.039-9.2 Minh et al., 2007 Lower St. Lawrence
Estuary, Canada 1993-1994 40 Lebeuf and Nunes, 2005 Western Scheldt river,
16 constant if compared to low chlorinated congeners of PCBs (Appendix 1), so the deposition of HCB from the air was the dominant contaminant source to Norrbyn and Kramfors sediments. Moreover, concentration of HCB in Norrbyn sediment was higher than concentration in Kramfors sediment because Norrbyn location has lower temperature (lower temperature might have more HCB deposit from air) than Kramfors location (historical weather for Sweden). When comparing the result of HCB to the classifications with Swedish coast guideline (Appendix 2), both Kramfors and Norrbyn sediments belong to class II, i.e. the low level of pollution of HCB at these locations. Moreover, levels of HCB in this study were also lower than the PEL (20 ng/g) (Marvin et al., 2004) indicated low level of pollution by HCB in the study areas. Moreover, HCB in Kramfors sediment was lower the one previously reported (Apler et al., 2014; Vogel, 2015) and the concentration of HCB seems decrease with time. When comparing HCB in the Baltic Sea and in other locations in the world, HCB levels at present study locations were lower than many HCB contaminated places around the world as shown in table 4.
Table 4: HCB concentration (ng/g d.w.) in sediments in this study and other locations around the world.
Location Year HCB (ng/g) Reference
Baltic Sea, Sweden
Norrbyn
Kramfors
0.0047 ± 0.0003
0.0009 ± 0.0002 this study Baltic Sea, Sweden 2010 0.36-0.84 Sobek et al., 2015 The NW Mediterranean,
Endland 2011 4.1 Sole et al., 2013
Kamogawa River, Japan 2004 0.30-0.92 Hirano et al., 2007
The Han river, Korea 2005 1.48 Kim et al., 2009
Ontario lake and Erie
lake, Canada 1997-1998 0.4-0.7 Marvin et al., 2004
Mekong River delta,
South Vietnam 2003-2004 0.016 Minh et al., 2007
Lower St. Lawrence
Estuary, Canada 1993-1994 1.3 Michel and Nunes, 2005 Western Scheldt river,
17 Figure 5: Concentration (ng/g d.w.) of each congener of PCBs and HCB in Norrbyn (Nor) and Kramfors sediment (Kra).
0 1 2 3 4 5 6 PCB -1 7 PCB -1 8 PCB -2 8 PCB -3 1 PCB -3 3 PCB -4 4 PCB -4 9 PCB -5 2 PCB -7 0 PCB -7 4 PCB -8 2 PCB -8 7 PCB -9 5 PCB -9 9 PCB -1 01 PCB -1 05 PCB -1 10 PCB -1 18 PCB -1 28 PCB -1 32 PCB -1 38 PCB -1 49 PCB -1 51 PCB -1 53 PCB -1 56 PCB -1 58 PCB -1 69 PCB -1 70 PCB -1 71 PCB -1 73 PCB -1 80 PCB -1 83 PCB -1 87 PCB -1 91 PCB -1 94 PCB -1 95 PCB -1 98 PCB -2 05 PCB -2 06 PCB -2 08 PCB -2 09 H CB Csed (n g/ g d. w )
Concentration of PCBs and HCB in Norrbyn and Kramfors sediments
Nor Kra
Tri Tetra Penta Hexa Hepta Octa Nona
18
4.2 Concentration of PCBs and HCB in pore water
Non-depletive condition was the most important when using passive sampler techniques. Hawthorn et al., (2009) suggested that non-depletive condition, the ratio of sediment concentration/POM should be below 200:1. In this thesis, the ratio sediment concentration/POM of all PCBs congeners and HCB were range from 0.019 to 11.9, so using POM as passive sampling technique was acceptable.
Pore water is the compartment that closest to sediment, therefore, there is a dynamic contaminant exchange among them. The total concentration of PCB in pore water was the highest in Örnsköldsvik sediment (44.3 ± 4.73 pg/L and the concentration of individual congeners ranging from 0.092 to 5.97 pg/L), followed by sediment Kramfors sediment (28.4 ± 4.54 pg/L, range from 0.0326 to 6.89 pg/L), and then sediment Norrbyn sediment (27.1 ± 6.56 pg/L, range from 0.057 to 4.42 pg/L) (Figure 6). The concentration of PCBs in pore water decreased with increasing of chlorinated for Norrbyn pore water and Kramfors pore water while octa to decachlorinated congeners were not detected (Figure 6 and Appendix 4). In addition, the lower chlorinated congeners seem to have higher water solubility from sediment to pore water than higher chlorinated. However, these results contrasted with Örnsköldsvik pore water, the concentration of PCBs increased from trichlorinated to pentachlorinated and decreased from pentachlorinated to heptachlorinated while octa to decachlorinated was not detected. In this case, degree of chlorination of PCBs was not enough to describe the desorption capacity of PCBs from sediment to pore water. Therefore, desorption and sorption behavior of PCBs was studied using sediment-water partition coefficients (KD and KOC). The concentration of PCBs in pore water of
19 Figure 6: PCBs and HCB concentration (pg/L) in pore water (PW) in Norrbyn (N), Kramfors (K) and Örnsköldsvik (O) sediments.
0 1 2 3 4 5 6 7 8 PCB -1 7 PCB -1 8 PCB -2 8 PCB -3 1 PCB -3 3 PCB -4 4 PCB -4 9 PCB -5 2 PCB -7 0 PCB -7 4 PCB -8 2 PCB -8 7 PCB -9 5 PCB -9 9 PCB -1 01 PCB -1 05 PCB -1 10 PCB -1 18 PCB -1 28 PCB -1 32 PCB -1 38 PCB -1 49 PCB -1 51 PCB -1 53 PCB -1 56 PCB -1 58 PCB -1 69 PCB -1 70 PCB -1 71 PCB -1 73 PCB -1 80 PCB -1 83 PCB -1 87 PCB -1 91 PCB -1 94 PCB -1 95 PCB -1 98 PCB -2 05 PCB -2 06 PCB -2 08 PCB -2 09 H CB CPW (p g/ L)
Pore water concentration of PCBs and HCB in Norrbyn, Kramfors and Örnsköldsvik sediments
20 HCB was detected in all pore water samples in this study. The concentration of HCB in pore water increased from Kramfors (0.026 ± 0.0009 pg/L), Norrbyn (0.041 ± 0.001 pg/L) to Örnsköldsvik (2.25 ± 0.039 pg/L). HCB in pore water in this study was lower than Palmer Long Term Ecological Research (0.63-6.7 pg/l) (Zhang et al., 2013) for Norrbyn pore water and Kramfors pore water while Örnsköldsvik pore water was the same as HCB in pore water at Palmer Long Term Ecological Research. In addition, HCB concentration of overlying water at Godthåbsfjord, Greenland (50 pg/L) (Carlsson et al., 2012) was higher than in this thesis. Greenland and Baltic Sea were the places that did not use organochlorine pesticides (Carlsson et al., 2012; Vogel, 2015) but these places are contaminated with HCB, so the air long range transportation might be the source of HCB to overlying water and pore water in these places.
To study sorption behavior of the fiber banks sediment, PCBs and HCB concentration in pore water and sediment were used to calculate sediment-water partition coefficient KD (Figure 7). Log KD correlate well with log KOW (R2 = 87% and 95% for
Norrbyn and Kramfors, respectively). These results indicate the less hydrophobic compounds tended to dissolve in water to a greater extent than more hydrophobic compounds that makes them more bioavailable for uptake by different aquatic species (Lydy et al., 2014).
21 Figure 7: Sediment-water distribution coefficients (KD) of HCB and PCBs were plotted
against hydrophobicity (log KOW) of each compound in Norrbyn (Nor) and Kramfors
(Kra) sediments.
Figure 8: Organic carbon normalized sediment-water distribution coefficients (KOC) of
HCB and PCBs were plotted against hydrophobicity (log KOW) of each compound in
Norrbyn (Nor) and Kramfors (Kra) sediments.
4.3 Concentration of PCBs and HCB in biota
PCBs and HCB are highly hydrophobic compounds, the concentrations of these compounds in biota were normalized with lipid content in biota (Appendix 6).
The highest total concentration of PCBs in Marenzelleria spp. was observed in Kramfors sediment, 746 ±205 ng/g lipid (ƩPCB7= 383 ±127 ng/g), followed by
Örnsköldsvik, 579 ±198 ng/g (ƩPCB7= 243±98.3 ng/g), and finally Norrbyn, 350±223
ng/g (ƩPCB7= 164 ±112 ng/g), which does not resemble well PCB patterns in sediment
and pore water, i.e. the total concentration of PCBs in Marenzelleria spp. in Örnsköldsvik sediment was lower than in Kramfors even though the concentrations of PCBs in Örnsköldsvik sediment (using result from Apler et al., (2014) and pore water were higher than in Kramfors sediment and pore water.
As for Macoma balthica, the total concentration of PCBs in this species decreased in order Örnsköldsvik, 2340±515 ng/g (ƩPCB7= 1100 ±23 ng/g), Norrbyn, 473 ±164 ng/g
y = 1.4856x - 2.9295 R² = 0.8764 y = 1.6307x - 2.7836 R² = 0.9501 0 2 4 6 8 10 5 5.5 6 6.5 7 7.5 8 Lo g KOC (m l/g) Log KOW
Log KOC versus log KOW
Nor Kra y = 1.4856x - 4.1166 R² = 0.8764 y = 1.6307x - 3.9843 R² = 0.9501 0 1 2 3 4 5 6 7 8 5 5.5 6 6.5 7 7.5 8 Lo g K D (m l/g) Log KOW
Log KD versus log KOW
22 (ƩPCB7= 241±85.1 ng/g) to Kramfors352 ±100 ng/g (ƩPCB7= 175 ±51.2 ng/g).
Interestingly, the total concentration of PCBs in Macoma balthica in Norrbyn sediment was higher than in Kramfors, whereas PCBs concentration in sediment and pore water was higher in Kramfors than in Norrbyn. A possible explanation of this can be the sorption ability of sediment (KD and KOC). As it was discuss above, the sorption
capacity of PCBs in Kramfors sediment was stronger than in Norrbyn sediment so the directly desorption of PCBs in Norrbyn sediment to biota was higher compared to Kramfors sediment.
When comparing the total concentration of PCBs between two species, the total concentration of PCBs in Marenzelleria spp. in Norrbyn and Örnsköldsvik were lower than total concentration in Macoma balthica in Norrbyn and in Örnsköldsvik, respectively, however, total concentration in Marenzelleria spp. in Kramfors was higher than Macoma balthica in Kramfors. These complex results indicated that the different feeding between Marenzelleria spp. (surface suspension feeding) and Macoma balthica (deposit feeding) can-not explain well the total concentration of PCBs in biota while the other studies like Zhang et al., (2013) and Kaag et al., (1997) found that the different feeding strategies had affect to accumulation of PCBs when compared the total concentration of PCBs in biota.
The concentration of HCB in all Marenzelleria spp. was below detection limits, while HCB was detected in all Macoma balthica samples. This could be due to differences in sample volume, i.e. grams for Macoma balthica and milligrams for Marenzelleria spp., and lipid content,~20% for Macoma balthica vs. ~1% for Marenzelleria spp.. The highest HCB concentration in Macoma balthica was observed in Örnsköldsvik (9.18±1.13 ng/g lipid) followed by Norrbyn (2.56 ±0.1 ng/g) and then in Kramfors (2.46±0.24 ng/g) sediments resembling the same trend as in pore water and sediments (Örnsköldsvik > Norrbyn > Kramfors).
4.3.1 Bioaccumulation factor
The concentration of PCB congeners in pore water and in biota were used to calculate bioaccumulation (BAF) (some researchers call it KLipid) to measure the extent of
chemical sharing between an organism and the surrounding environment (Figure 9). As discussion above, low hydrophobic compounds had high concentration in pore water and they should have high concentration in biota but in Figure 9 showed highly hydrophobic compounds had high concentration in biota. BAF is the tool to describe the partitioning behavior of PCBs between pore water and biota, as we can see from Figure 9, the potential for a chemical to bioaccumulate in organisms is correlated with the octanol-water partition coefficient (KOW) (R2 was ranging from 60% to 95%).
24 between log BAF and log KOW because the differences of biota that maybe they can
have different type of lipid contain so they don not have the same BAF.
Figure 9: Log BAF of PCB for Marenzelleria spp. (Mar) and Macoma balthica (Mac) were plotted against log KOW in (A) Norrbyn (N), (B) Kramfors (K), and (C)
Örnsköldsvik (K) sediments.
The most probable explanation for the observed differences in accumulation can be species-specific characteristic of Marenzelleria spp. and Macoma balthica. The main exposure pathway of PCBs to invertebrates is sediment ingestion (Josefsson et al., 2011). Macoma balthica uses siphon to take in the food from the surface sediment. At the same time Marenzelleria spp. always ventilate and take in the food at the surface sediment (Josefsson et al., 2011). It goes up and down in the sediment column, digs sediment, pumping water in and out, thus, in this way facilitating desorption of PCBs from sediment to water. Moreover, low hydrophobic compounds tend to have higher solubility in water than higher hydrophobic compounds. That might be the reason that
25 low hydrophobic compounds in Marenzelleria spp. were higher than in Macoma balthica (Figure 9).
In Figure 9 showed that BAF was higher than KOW, indicated that the partitioning to
lipids is higher than octanol. This result differed from the previous study that they found that the sorption capacity of lipid and sediment was nearly the same (log BAF= 0.95 ×log KOW +0.06) (Zhang et al., 2013). Differences results might the differences
species, so the lipid composition also might be differences.
4.3.2 Biota-sediment accumulation factor
The BSAFs of Marenzelleria spp. in Norrbyn and Kramfors sediments range from 2.05 to 11.8 and 0.193 to 8.03, respectively. The BSAFs in this study were higher than previous study (range from 1.4 to 6.8) and also higher the equilibrium levels (range from 1.1 to 1.5) (Josefsson et al., 2011). However, most BSAFs of Macoma balthica in Kramfors sediment was in the range of equilibrium range (except PCB-49, 28 and 191) (Figure 10). The reason that the BSAFs of Marenzelleria spp. in Norrbyn sediment of the present study was higher than previous study might be the sorption capacity of sediment of previous study was higher than Norrbyn sediment. In Figure 10 showed that the BSAFs of Marenzelleria spp. in Kramfors were lower than Norrbyn sediment, these results can explain well with sorption strengths that the KOC in Kramfors
sediment was higher thanin Norrbyn. Both BSAFs of Marenzelleria spp. in Kramfors and Norrbyn sediments had high BSAFs for less hydrophobic compounds and more hydrophobic compounds and low BSAFs for intermediate hydrophobic compounds. Low hydrophobic compounds had higher BSAFs than intermediate hydrophobic might be the behaviors of Marenelleria spp. (ventilate and feeding at surface sediment), so low hydrophobic compounds can have more desorption capacity from sediment to pore water and reach biota or directly from sediment to biota (more detail in section 4.3.1). High hydrophobic compounds had higher BSAFs than intermediate compounds because the sorption capacity of lipid for more hydrophobic compounds was stronger than less hydrophobic compounds.
26 Maccoma balthica in Kramfors and Norrbyn show that low hydrophobiccompounds and high hydrophobic compounds had high BSAFs indicate that these compounds had more metabolisms if comparing with intermediate hydrophobic compounds (Pruell et al., 1993; Tracey and Hansen, 1996).
The BSAFs of Marenzelleria spp. in Kramfors and Norrbyn sediments were higher the BSAFs in Macoma balthica in Kramfors and Norrbyn, respectively. This result should be explained well by the behaviors and lipid in biota. Again, Marenzelleria spp. ventilates and takes in suspended sediment at surface while Macoma balthic just uses siphon to take in deposit materials from surface sediment. On the other hand, lipid in Marenzelleria spp. might be had more sorption capacity than Macoma balthica.
Figure 10: BASF of PCBs for Marenzelleria spp. (Mar) and Macoma balthica (Mac) were plotted against log KOW in Norrbyn (N) and Kramfors sediments (K).
4.4 Chiral PCBs as fate in benthic zone
Seven chiral PCBs (PCBs-95, 91, 136, 149, 132, 176 and 174) were selected to study their fate in benthic zone. PCB-91 and 176 were not detected in all compartments so these were excluded from discussion. Moreover, low concentration and chromatographic interferences in the GC-LRMS method made difficult to describe and relate the EFs of chiral compounds between each compartment and the result should be that interpreted with caution. T-test was used in this thesis to compare means two EFs of chiral PCBs.
4.4.1 Chiral PCBs in sediments
The biotic dechlorination or metabolism of PCBs is the main processes of changing the EF of chiral PCBs in sediments. EF of chiral PCBs for each sediment and standards
0 2 4 6 8 10 12 14 5 6 7 8 BSAF Log KOW
BSAF versus log KOW
27 were plotted in Figure 11. T-test was used to compare mean EFs between target compound and standard to know the enantioselective biodegradation in sediment. EFs of PCB-95 in Norrbyn (t=1.35, p =0.267) and Kramfors (t= -1.79, p=0.324) sediments were racemic (i.e. did not differ from the standard), there was not enantioselective degradation of PCB-95 in Norrbyn and in Kramfors sediments. Örnsköldsvik sediment had EF for PCB-95 bigger than standard (t=16.8, p=1.2 x 10-8)
indicated that the second enantiomer had transformation higher rate than first enantiomer. Similarity, PCB-136 in Norrbyn (t= 2.67, p= 0.228) and Kramfors (t=2.27, p= 0.108) sediments were racemic while Örnsköldsvik (t= -23.5, p= 3.9 x 10-7)
sediment seem to be first enantioselective transformation but this congener had chromatographic interference (more detail in section 4.4.2). For PCB-149, racemic was found in Norrbyn sediment (t=1.18, p=0.32) and the first enantioselective biodegradation was found in Kramfors sediment (t= -23.5, p=0.0002). PCB-174 also did not differ from the other chiral congeners, the racemic was found in Norrbyn (t= 2.82, p= 0.07) and Örnsköldsvik (t= -0.29, p= 0.78) sediments and the nonracemic (second enantiomer dominant) in Kramfors sediment (t= -8.41, p=0.001). These results suggested that enantioselective biodegradation can occur in sediments.
Wong et al., (2001) and (2007) found that the enantiomer enrichment of each chiral PCBs was different for each sediment locations and also different depths within same sediment core. They suggested that the different depths of sediment had different kinds and the population of microbes. Dang et al., (2010) also found that the enantiomer enrichment of PCBs-136, 91 and 95 differed with composition of sediments (which contained different microbes and fungi). A review reported both aerobic and anaerobic microbes can enantioselectively degrade chiral PCBs (Lehmler et al., 2010. The previous studies of enantioselective degradations also found the similar results. In this case, the deviation of some chiral PCBs in the present study from racemic might be from microbe activities in sediments. On the other hand, if we compared one chiral chemical with different sediments, we found that from one sediment sample to the other sediment samples had different EF values (e.g. PCB-95 in Örnsköldsvik sediment had second enantioselective degradation but it did not have enantioselective degradation in Norrbyn and Kramfors sediments), even though this thesis did not determine the microbial community and activities of microbes, it also contradicted that sediment samples might be had different kinds of microbes. Future experiments should include the population and kind of microbes. Moreover, the activities of microbes should be including (Wong et al., 2007).
28 0 0.1 0.2 0.3 0.4 0.5 0.6 PCB-95 PCB-136 PCB-149 PCB-174 EF
Chiral PCBs at sediments and standard
Se_N Se_K Se_O std
*
*
*
*
PCBs-136 (2,2’,3,3’,6,6’-). PCB-174 (2,2’,3,3’,4,5,6’-) can be meta-dechlorinated to chiral PCB-149 (2,2’,3,4’,5’,6-) and then para-dechlorinated to PCB-95 (2,2‘,3,5‘,6-). PCB-132 (2,2’,3,3’,4,6’-) can be meta-dechlorinated to chiral PCB-91(2,2‘,3,4‘,6-). On the other hand, PCB-132 also can meta-dechlorinated to chiral PCB-89 (2,2’,3,4,6’-) (Wong et al., 2001 and 2007). In this thesis, the chiral PCB-174, 149 and 95 were detected in all sediment samples so the nonortho-dechlorinated of these congeners shown the same as previous study. PCB-176 was not detected in any sediment samples but PCB-136 was detected in all sediment samples even though interference prevented determination of the EFs in some case. It might be that all PCB-176 was para-dechlorinated to PCB-136. PCB-91 was not detected in any sediment samples while the PCB-132 detected in all sediment samples. This result might be because all PCB-132 was not meta-dechlorinated to PBC-91 but preferred meta-dechlorination to chiral PCB-89. Moreover, if we compared the EFs through the daughters-parents pair (Figure 11), the EF values for each congener at the same location sample were almost not the same EF values (e,g. PCB-174 and 149 in Kramfors sediment were nonracemic while PCB-95 in Kramfors sediment was racemic) showed that the each enantiomer had different pathways of enantioselective degradation although daughters-parents pairs.Figure 11: Enantiomer fraction of all chiral PCBs in sediments (Norrbyn sediment: Se_N, Kramfors sediment: Se_K, Örnsköldsvik sediment: Se_O) and standard (std). The error bars were standard deviation. The star signs (*) mean nonracemic.
4.4.2 Chiral PCBs in pore water
29 sediment. In short, only PCB-95 and 149 in Norrbyn and PCB-136 and 174 at Örnsköldsvik were kept to describe the fate of chiral PCBs between pore water and sediment (Figure 12 and more detail in Appendix 10).
Figure 12: Enantiomer fraction (EF) of chiral PCBs in sediments (Se) and pore waters (PW) in Norrbyn(N) and Örnsköldsvik (O).
Again, sediment is a sink and also a secondary source of PCBs since PCBs was banned (Wiberg et al., 2009). PCBs in sediment can be desorption into pore water. Moreover, abiotic processing of chiral PCBs can-not change their EFs (Bidleman et al., 2013; Hühnerfuss and Shah, 2009). Due to this concept, the EFs of chiral PCBs in pore water should be the same as in sediment. The EFs of chiral PCB-174 in Örnsköldsvik pore water and sediment was the same (t= -1.02, p= 0.335) and indicated that PCB-174 was racemic (Figure 12). PCB-95 and 149 in Norrbyn pore water were the same as in Norrbyn sediment (PCB-95, t= -0.514, p= 0.642 and PCB-149, t= -0.04, p= 0.941) (Figure 12) and also indicated racemic compounds. These results indicated that the source of these PCBs was from sediment and desorption process did not change EFs. In contrast, the EF of PCB-136 in Örnsköldsvik pore water was significantly different (t= 14.4, p= 6.7 x 10-6) from sediment sample (Figure 12). The EF of PCB-136 in Örnsköldsvik pore water
was 0.569 ± 0.025 and the EF of the same chemical in Örnsköldsvik sediment was 0.411 ± 0.009, indicated that sediment was not the main source for this contaminant into pore water. On the other hand, the result for PCB-136 might have the same problem as quantitative analysis that Örnsköldsvik sediment had chromatographic interferences during instrumental analysis so the conclusion for chiral 136 in this thesis was PCB-136 in Örnsköldsvik sediment had interference.
4.4.3 Chiral PCBs in Biota
All biota samples were collected from Norrbyn and the quantitative analysis of Norrbyn sediment showed PCBs contamination, so it was not strange that PCBs were detected in biota samples before the bioaccumulation experiment. Chiral PCB-149 and 132 were
P C B -95 P C B -95 P C B -1 36 PCB -1 36 P C B -1 49 P C B -1 49 P C B -174 PCB -174 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Se_N PW_N Se_O PW_O
EF
30 detected in Macoma balthica while no chiral PCBs were detected in Marenzelleria spp., probably because of small samples available for analysis (Figure 13 and Appendix 10). PCB-149 in Norrbyn Maccom balthica was racemic, the same as in Norrbyn sediment and pore water, indicated that there was no enantioselective degradation. PCB-132 was detected in Macoma balthica but the EFs of PCB-132 could not be determined in Norrbyn sediment because of interferences and it was not detected in Norrbyn pore water. However, EF>0.5 (t= 10.9, p=0.0001) suggests that the source is nonracemic PCB-132 in Norrbyn sediment.
31 13), so the enantiomer selective degradation was not observe. The EF of PCB-136 was the same between in pore water and in Macoma balthica in Örnsköldsvik (t= 2.49, p= 0.057), so this congener did not have enantioselective biodegradation. The last chiral congener detected in Macoma balthica was PCB-174. PCB-174 had racemic in pore water (t= 1.53, p= 0.14) and in Örnsköldsvik sediment (t= -0.29, p= 0.78) but was nonracemic (EF>0.5, t= 18.7, p= 3.7 x 10-8) in Macoma balthica in Örnsköldsvik (Figure 13) indicated that
PCB-174 had second enantiomer selective biodegradation in Macoma balthica. Depend on the results, there are two suggestions should be the critical thinking that we can see some chemicals had enantiomer selective degradation but some chiral PCBs did not have. The first one was the exposure time of invertebrates to chiral PCBs. Exactly, 28 days was enough time to equilibrate between each compartment in bioaccumulation experiment for Macoma balthica but this period might not enough time to enantioselective degradation for some chiral PCBs. On the other hand, the low concentration and less activity of cytochrome P-450 enzyme (detoxify enzyme) in invertebrates can be the reason of enantioselective degradation of some chiral PCBs in Macoma balthica were not observed (Lehmler et al., 2010).
32 Figure 13: Enantiomer fraction (EF) of chiral PCBs in sediment (Se), pore water and in biota, Macoma balthica (Mac) and Marenzelleria spp.(Mar) in Norrbyn (N), Kramfors (K) and Örnsköldsvik sediments. The error bars were standard deviation and the star signs (*) mean enantioselective biodegradation. P C B -95 P C B -95 P C B -1 36 P C B -1 36 P C B -1 49 P C B -1 49 P C B -1 49 P C B -1 49 P C B -1 49 P C B -1 49 PCB -1 49 P C B -1 49 P C B -1 49 P C B -1 32 P C B -1 32 P C B -1 32 P C B -1 32 P C B -174 P C B -174 PC B -174 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Se_N PW_N Mac_N Se_K PW_K Mac_K Mar_K Se_O PW_O Mac_O Mac_T0
EF
EF chiral PCBs at all compartments
*
*
*
33
5. Conclusions
This thesis showed that the PCBs source of contamination of the Norrbyn sediment and the Kramfors sediment might be Aroclor 1260 while the source HCB was mainly long-range air transportation. KD and KOC showed that the sorption capacity for the
Norrbyn sediment was weaker than the Kramfors sediment, which might be due to the different compositions of sediments. KD and KOC indicated that the less hydrophobic
compounds had higher desorption capacity to pore water than more hydrophobic compounds.Moreover, the sorption strength of hydrophobic compounds to sediment also had an effect on bioaccumulation of hydrophobic compounds to invertebrates. If sediment had low sorption capacity for hydrophobic compounds, the bioavailability at that location was high. Behavioral characteristic of Marenzelleria spp. affected the remobilization of hydrophobic compounds especially the chemicals that were less hydrophobic. Marenzelleria spp. had higher BSAF values most probably due to their behavior and suspension feeding if compared to deposit feeding strategy of Macoma balthica.
Chiral analysis showed that enantioselective biodegradation of chiral PCBs occurred in the studied sediments samples. Complex EFs of chiral PCBs in sediments indicated that different sediment might have different microbial communities and activities. Pore water likely contaminated chiral PCBs from the sediment as matrixes showed identical enantiomer fractions indicating desorption process from sediment to pore water. In this thesis it was shown that enantioselective degradation was congener-specific for PCBs in Macoma balthica while we were not able to observe species-specific because Marenzelleria spp. as only PCB-95 was detected among the chiral PCBs.
6. Acknowledgment
34
7. Appendix
Appendix 1: Physical chemical properties, viz. Henry law’s constant (H), and octanol-water partition coefficient (log KOW), and POM/pore water partition coefficient (log
KPOM) for PCBs and HCB.
PCB congeners Congeners IUPAC name a log K
ow a log KPOM -log He
35 180 2,2’,3,4,4’,5,5’- 7.02 6.67b 3.969 183 2,2’,3,4,4’,5’,6- 7.20 6.59b 3.696 187 2,2’,3,4’,5,5’,6- 7.17 6.44b 3.693 191 2,3,3’,4,4’,5’,6- 7.55 6.99b 3.876 194 2,2’,3,3’,4,4’,5,5’- 7.80 7.18b 4.174 195 2,2’,3,3’,4,4’,5,6- 7.56 6.83c 3.926 198 2,2’,3,3’,4,5,5’,6- 7.62 6.86c 3.812 205 2,3,3‘,4,4’,5,5’,6- 8.00 7.05c 4.059 206 2,2’,3,3’,4,4’,5,5’,6- 8.09 7.30c 4.059 208 2,2’,3,3’,4,4’,5,6,6’- 7.71 7.47c 3.777 209 decachlorobiphenyl 8.18 8.18d 3.948 HCB
log Kow log KPOM H
HCB 5.5d 4.96d 35f
a Hawker and Connell, 1988 b Hawthorne et al., 2009
c Log KPOM were calculated from log KPOM = 0.791 log Kow + 1.018 (Hawthorne et al.,
2009)
d Vogel, 2015
e Dunnlvant et al., 1992
f Jantunen and Bidleman, 2006
Appendix 2: Study sediments according to Swedish classification of contaminated sediments (ng/g) (Naturvårdsverket, 1999).
Class I
No-level Class II low-level Class III Mid-level Class IV High-level Class V Very high-level
HCB 0.00 0-0.04 0.04-0.2 0.2-1 >1
ƩPCB7 0.00 0-1.3 1.3-4 4-15 >15
Appendix 3: PCB homologue group in Kramfors (Kra) and Norrbyn (Nor) sediment compared to a commercial mixture Aroclor 1260*.
*Aroclor 1260: CDC (chemical and physical information of PCBs)
05 1015 2025 3035 4045 50 Pe rce n ta ge %
Percentage of clorinated pattern at Norrbyn sediment, Kramfors sediment and Arochlor 1260
Kra Nor
36 Appendix 4: Relative abundance of PCB homologue groups in pore water in Kramfors, Norrbyn and Örnsköldsvik sediments.
Appendix 5: Organic carbon fraction Norrbyn, Kramfors and Örnsköldsvik sediments. Organic carbon fraction
Se_N Se_K Se_O
OC (%) 6.56 ±0.24 6.32 ± 0.35 10.63 ± 0.41
fOC 0.066 0.063 0.106
Appendix 6: Lipid fraction (w.w) of Marenzelleria spp. and Macoma balthica. Lipid fraction
Mar_N Mar_K Mar_O Mac_N Mac_K Mac_O
lipid
contain (%) 1.05 ± 0.09 1.23 ± 0.24 1.04 ±0.13 16.60 ± 1.01 22.85 ±4.01 17.68 ±0.82
flid 0.011 0.012 0.011 0.166 0.228 0.177
Appendix 7: log concentration of each PCB congener in sediments (Se_N: Norrbyn sediment, Se_K: Kramfors sediment) plotted against with log KOW.
0 20 40 60
Tri Tetra Penta Hexa Hepta Octa-deca
Pe rce n ta ge (%)
Percntage of chlorinated patterns of PCBs in pore waters PW_N PW_K PW_O R² = 0.3674 R² = 0.3594 -2.5 -2 -1.5 -1 -0.5 0 0.5 1 5 5.5 6 6.5 7 7.5 8 Lo g C once n tra ion log Kow
Log concentration versus log Kow in sediments
37 Appendix 8: Lipid normalized concentration of PCB congeners in biota, Marenzelleria spp. and Macoma balthica in Norrbyn, Kramfors and Örnsköldsvik sediments. -1 -0.5 0 0.5 1 1.5 2 2.5 3 P C B -1 7 P C B -1 8 P C B -2 8 P C B -33 P C B -44 P C B -49 P C B -52 P C B -7 0 P C B -7 4 P C B -8 2 P C B -8 7 P C B -95 P C B -99 P C B -1 01 P C B -1 05 P C B -1 10 P C B -1 18 P C B -1 28 P C B -1 32 P C B -1 38 P C B -1 49 P C B -1 51 P C B -1 53 P C B -1 56 P C B -1 58 P C B -1 69 P C B -1 70 P C B -1 71 P C B -1 73 P C B -1 8 0 P C B -1 8 3 P C B -1 8 7 P C B -1 91 P C B -1 94 P C B -1 95 P C B -1 98 P C B -2 05 P C B -2 06 P C B -2 08 P C B -2 09 Lo g Cli p (n g/ g l ip id )
Log concentation of each PCB in biotas
Mar_K Mar_N Mar_O Mac_O Mac_N Mac_K
