GC-MS screening and PCB analysis
of sediment from central Kattegat
Emma Eriksson
Örebro University
i
Abstract
Five sediment samples were collected in Bua on the Swedish west coast, near two industries, a paper mill, and a nuclear power plant. The two industries use water in their processes and have long been associated with releases of different substances, such as PCBs, and other chlorinated compounds. The environmental impact by the two industries is believed to be significant. The aim of the project was to examine the sediments close to both the water intake and water output to determine if these industrial activities have in any way changed the
composition of the sediments. The sediments were extracted by Soxhlet extraction, followed by a deactivated silica and an acidic silica clean-up and then analysed by using a gas
chromatograph coupled to a mass spectrometer, (GC-MS) with electron ionization, EI+, mode used in full scan mode. Each mass spectra were analysed by comparing them to the NIST database from 1998. The results were inconclusive since the peaks were not properly resolved, causing a poor correlation to the NIST database.
One batch was specifically analysed for polychlorinated biphenyls (PCB) by using an
atmospheric pressure gas chromatograph (APGC) coupled to a mass spectrometer (MS). The PCB analysis provided accurate results, except for the Ringhals intake where the MS became saturated due to the high levels. The river Viskan also showed high levels of PCB. The congener pattern from PCBs found near Ringhals intake resembled an Aroclor pattern from Aroclor 1248. Since the Aroclor pattern is only seen in Ringhals intake, the source is most likely from the small harbour and not from either of the industries.
ii
Table of Content
Abstract ... i
List of Figures ... iii
List of Tables ... iv
1. Introduction ... 1
1.1 Aim and objective ... 1
2. Background and theory ... 2
3. Material and Method ... 4
3.1 Sampling ... 4
3.2 Soxhlet extraction ... 4
3.3 Metallic copper clean-up ... 5
3.4 Silica clean-up ... 6
4. Results and Discussion ... 10
4.1 Ringhals ... 11
4.1.1 Indicative ions for chlorine and bromine ... 13
4.2 Södra Cell Värö ... 15
4.3 Polychlorinated biphenyls ... 17
5. Conclusion ... 25
6. References ... 29
Appendix A- Batch 1, without minisilica ... 31
Appendix B: Batch 1 with minisilica ... 35
Appendix C: Batch 2 ... 37
Appendix D – NIST Database matches and analyses ... 39
iii
List of Figures
Figure 1: Sampling locations, marked as circles, with the industries, marked as squares, and
actual intake and discharge, marked as diamonds (Google Maps 2015). ... 2
Figure 2: Silica columns ... 7
Figure 3: Total ion chromatogram, TIC, of Ringhals input, batch 1 ... 11
Figure 4: TIC of Ringhals output, batch 1 ... 11
Figure 5: TIC of Ringhals intake, batch 2 ... 12
Figure 6: TIC for Ringhals output, batch 2 ... 12
Figure 11: Indicative ions for chlorine, m/z 35 and 37 ... 13
Figure 12: Indicative ions for bromine, m/z 79 and 81 ... 14
Figure 7: TIC of Södra Cell Värö intake, batch 1 ... 15
Figure 8: TIC of Södra Cell Värö output, batch 1 ... 15
Figure 9: TIC of Södra Cell Värö intake, batch 2 ... 16
Figure 10: TIC of Södra Cell Värö output, batch 2 ... 16
Figure 13: PCB levels in sediment samples from the different sites. The blue stars indicates peaks with too high concentration of PCBs which saturated the MS. The red stars indicate peaks with bad peak shape. ... 17
Figure 14: PCB levels in sediment samples from the different sites, the blue stars indicate saturation of the MS, and the red stars indicate peaks with bad peak shape. ... 18
Figure 15: Indicative ions for PCB with one chlorine substitution ... 19
Figure 16: Indicative ions for PCB with two chlorine substitution ... 20
Figure 17: Indicative ions for PCB with three chlorine substitution ... 20
Figure 18: Indicative ions for PCB with four chlorine substitutions ... 21
Figure 19: Indicative ions for PCBs with five chlorine substitutions ... 21
Figure 20: Indicative ions for PCBs with six chlorine substitutions ... 22
Figure 21: Indicative ions for PCBs with seven chlorine substitutions ... 22
Figure 22: Indicative ions for PCBs with eight chlorine substitutions ... 23
Figure 23: Indicative ions for PCBs with nine chlorine substitutions ... 23
Figure 24: Indicative ions for PCBs with ten chlorine substitutions ... 24
Figure 25: TIC from Toluene ... 31
Figure 26: TIC from Soxhlet blank ... 31
Figure 27: TIC from Ringhals input ... 32
Figure 28: TIC from Ringhals output ... 32
Figure 29: TIC from Södra Cell Värös input ... 33
Figure 30: TIC from Södra Cell Värö output ... 33
Figure 31: TIC of sample from the bay ... 34
Figure 32: TIC of toluene blank ... 35
Figure 33: TIC of Soxhlet blank, after use of minisilica ... 35
Figure 34: TIC of sample from the bay, after the use of minisilica ... 36
Figure 35: TIC of toluene ... 37
Figure 36: TIC of Soxhlet blank from batch 2 ... 37
Figure 37: TIC of sample taken in the bay from batch 2 ... 38
Figure 38: NIST database match for Ringhals batch 1, 6.258 min ... 39
Figure 39: NIST database match for Ringhals batch 1, 15.378 min ... 40
Figure 40: NIST database match for Ringhals batch 1, 18.291 min ... 41
iv
Figure 42: NIST database match for Ringhals batch 2, 15.817 min ... 43
Figure 43: NIST database match for Ringhals batch 2, 19.883 min ... 44
Figure 44: NIST database match for Ringhals batch 2, 20.408 min ... 45
Figure 45: NIST database match for Ringhals batch 2, 20.987 min ... 46
Figure 46: NIST database match for Södra Cell Värö batch 1, 16.271 min ... 47
The molecule suggested by the NIST database for the peak at 16.27 min is 2,2’-diethyl-1,1’-biphenyl (Fig. 46). The NIST reference spectra show several similarities, mainly the peaks at m/z 181 and 165. There are two peaks from the reference spectra just above m/z 181, which seems to be missing or have a very low intensity in the sample spectra. The ions at m/z ratio of 181 and 165 seem to be very common for many polyaromatic hydrocarbons, since the first four hits from the NIST database all display the same ions, m/z 181 and 165, at similar intensities. The delta plot show some differences between the sample spectra and the NIST reference spectra. The compound with retention time 16.27 minutes might be a biphenyl or another polyaromatic hydrocarbon, but the functional groups have not been determined Figure 47: NIST database match for Södra Cell Värö batch 1, 18.442 min ... 47
Figure 48: NIST database match for Södra Cell Värö batch 1, 18.520 min ... 49
Figure 49: NIST database match for Södra Cell Värö batch 1, 20.546 min ... 50
Figure 50: NIST database match for Södra Cell Värö batch 2, 5.939 min ... 51
Figure 51: NIST database match for Södra Cell Värö batch 2, 20.004 min ... 53
Figure 52: NIST database match for Södra Cell Värö batch 2, 21.515 min ... 54
Figure 53: NIST database match for Södra Cell Värö batch 2, 32.553 min ... 55
List of Tables
Table 1: Sampling information ... 4Table 2: Examples of Soxhlet extraction methods used in literature ... 5
Table 3: Internal standard added to the samples. ... 5
Table 4: GC-MS full scan conditions ... 8
Table 5: APGC conditions ... 9
Table 6: Swedish EPA classifications of PCB contaminated sediments, µg/g dry weight of sediment (Naturvårdsverket 1999) ... 24
Table 7: Comparison of levels with the Swedish EPA * indicates peaks too large to integrate ** indicate bad peaks ... 24
1
1. Introduction
Paper mills have long been associated with the release of different environmental
contaminants, such as polychlorinated biphenyls (PCB) and polychlorinated dioxin and furans (PCDD/F), especially when the chlorine bleaching method was used (Ratia & Oikari 2014). Urban activities have been a source of contamination for the aquatic environment, and paper mills have been considered to be the sixth largest source of contamination (Ali &
Sreekrishnan 2001). Chlorinated effluent from the pulp industry is an extensive toxic hazard for the aquatic environment (Wulff et al. 1993). The pulp produced only contain around 40-45% of the original raw material. The rest of the raw material is then released with the
effluent, mainly as organic matter. Some suggested by-products released with the effluent are chlorinated products, fibres, fatty acids, lignin, sulphur containing compounds, chlorinated lignin, resin acids, phenols, dioxins, and furans (Ali & Sreekrishnan 2001).
Since PCBs and PCDD/Fs are included in the Stockholm Convention on Persistent Organic Pollutants (POP), they have been well monitored globally for the past 20 years (United Nations 1995). A POP is a substance that remains in the environment for a long period of time. It accumulates in adipose tissue, and are toxic to humans and all living organisms. A majority of the known POPs have a low solubility in water, but they dissolve easily in adipose tissue and therefore accumulate in any substance that display lipophilic properties (Secretariat of the Stockholm Convention n.d.). Some POPs display different characteristics, such as poly fluorinated compounds, that have a lower degree of hydrophobicity (Lau et al. 2007).
Since the beginning of the 1990s, the pulp industry have changed from chlorine bleaching to chlorine-free bleaching methods, or elemental chlorine free bleaching. Decreasing levels of PCBs and PCDD/Fs have been demonstrated, although they are still present as contaminants in sediments surrounding paper mills (Ratia & Oikari 2014).
In another study fish from the unheated and heated area near a nuclear power plant was compared, and found that perch showed significantly higher levels of PCBs. However, there were no significant difference for another fish species (Edgren et al. 1981)
The cooling water from the nuclear power plant Ringhals, does normally only affect the surface water. Surface water is considered water down to 7 meters depth. Under extreme conditions during the winter, the sediments can be exposed to the heated water. The heat increase will affect the uptake of PCBs and DDT. The first studies on the subject showed that the uptake were twice as big in fish when the temperature increased from 5°C to 15°C. The initial study was performed in a laboratory, and the effect was much less prominent in the environment (Elforsk 2009).
1.1 Aim and objective
The aim of this project was to compare the sediments from the inlet and outlet of the selected industries. This approach was chosen to determine if the changed characteristics of the water (such as heat increase and organic components) have affected the composition of the
sediments. This will be examined through a GC-MS screening where the chromatographic pattern will be examined and through a specific PCB analysis of the sediments. The sediment near the two industries was also compared to a sample from the nearby bay, to determine if any potential contamination can be linked to either of the two industries. The PCB analysis is
2 also to be compared to known congener profiles of Aroclor to possibly link the PCBs to a specific Aroclor.
2. Background and theory
Bua is a village situated 20 km north of Varberg, on Värö peninsula. In the surrounding area there are two larger industries, Södra Cell Värö, a paper mill, and Ringhals, a nuclear power plant. Both industries use the sea and river water in their processes. Ringhals uses the water from the sea as cooling water for the reactors. The water is only used for heat exchange and is not a part of the nuclear fission itself. Södra Cell Värö uses water from Viskan, a larger river that ends in Kattegatt, in its process. The water is after usage released into the sea from a 2 km long pipeline. The pipeline is situated at near 20 meters depth and the release takes place through 180 holes along the pipeline (Eriksson 2015).
Figure 1: Sampling locations, marked as circles, with the industries, marked as squares, and actual intake and discharge, marked as diamonds (Google Maps 2015).
Södra Cell Värö was built in 1978. From the beginning until 1990, the bleaching used chlorine, when they started the transition to chlorine free bleaching. In 1993, the bleaching process at Södra Cell Värö was totally free of chlorine, being one of the first paper mills in the wold to apply a chlorine free bleaching method. The chlorine bleaching was replaced with oxygen bleaching. The oxygen bleaching is then followed by further bleaching with hydrogen peroxide, oxygen and peracetic acid, if further beaching is requested. In 2002, Södra Cell Värö installed a facility for biological cleaning of the waste water (Södra Cell Värö 2014). In 2010, Södra Cell Värö was the first paper mill in the world to be independent of fossil fuels in their normal production (Södra Cell Värö 2015). Södra Cell Värö produces around 425 000
3 metric tonnes of paper pulp each year of witch about 90% is exported, mainly to Europe. Södra Cell Värö mainly produce paper pulp, but in recent years they have also started
producing green electricity, biofuel, and long distance heating to the surrounding area (Södra Cell Värö 2014).
Previous investigations have been conducted by Södra Cell Värö in 1989-1996, where they analysed 2,3,7,8-tetrachlorodibenzodioxin (TCDD), 3,4,3’,4’-tetrachlordibenzofuran (TCDF), 1,2,3,7,8-pentachlorodibenzofuran (PeCDF), and extractable organic chlorine (EOCl). The samples were taken from an affected point near the industry and included crabs and common whelk, Buccinum undatum (Stibe 2015; Nationalencyklopedin n.d.). The samples were compared to a reference site in Fladen. The study ended soon after Södra Cell Värö changed bleaching process from the chlorine bleaching to oxygen bleaching since the levels from the affected station approached the levels in the reference samples (Stibe 2015; Södra Cell Värö 2014).
Ringhals is situated on the northern side of Värö peninsula, and produces 20% of the electricity consumed in Sweden today. The construction of the power plant started in 1969, and in May 1975 the first reactor was put into operation. There are currently four reactors in operation, Ringhals 1 through 4. In 2002, Ringhals became certified according to
Environmental Product Declaration (EPD), which means that they for each kWh produced can calculate the environmental effects. The sea water is used as cooling water for the steam produced in the reactors. When operating at full power, around 170 cubic meters of seawater is used per second for the cooling of the reactors. (Vattenfall 2012a).
The cooling water leaves the reactor around 10°C warmer than the initial temperature.
Ringhals have constantly tried to minimize the different emissions to the environment, such as regular and dangerous waste, chemicals, cooling water discharge water treatment and
radioactive emission. Ringhals has since 1990 been registered with Eco Management and Audit Scheme (Emas) and since 1998 it is also certified according to ISO 14001 (Vattenfall 2012b). Their ammonia emission to water is less than one tenth of the natural amount found in the sea. The boric acid release is near one hundredth of the natural content found in the sea. Furthermore, chlorine is released at around 0.1 ppm per year (Vattenfall 2012b).
PCB are considered to be persistent and toxic in the environment. The PCBs have different degrees of chlorination, from one to ten, where each chlorine is attached to a carbon in a biphenyl backbone. PCB is a mixture of 209 different congeners. When producing PCBs, they were sold under the trade name Aroclor. Commensally, there were nine different formulations sold, all containing different weight percentage of chlorine. Aroclor were used in a vast number of purposes, such as dielectric fluid, inks, and pesticide extenders (Murphy & Morrison 2007).
4
3. Material and Method
3.1 Sampling
Duplicate samples for each site were acquired by using an Ekman sediment grab sampler. The samples were collected from five different locations along the coastline. Each location was chosen so that it could be compared to the other locations.
Table 1: Sampling information
Sample Sediment type Location description Coordinates Depth (m) Ringhals intake
Clay Small harbour N57°16.069’ E012°06.511’
2 Ringhals
output
Sand Open water, near coastline N57°15.078’ E012°05.351.’ 5 Södra Cell intake
Clay Open water, middle of river N57°13.678’ E012°12.768’ N57°13.633’ E012°12.804’ 3 Södra Cell output
Clay Open water, near coastline
N57°12.843’ E012°09.502’
4
Bay Sand Open water, near
coastline
N57°14.666’ E012°06.323’
3.5
The sample vials were filled to the best capacity in amber glass jars, 250 ml, and kept frozen until the extraction took place. The sampling point from Södra Cell Värös intake, were acquired from two different locations, from both sides of a delta at the end of the river. 3.2 Soxhlet extraction
Before extraction began, the samples were thawed and dried overnight. The sediments were spread out on aluminium foil and left in room temperature in the fume hood until dryness. There are several problems associated with the drying technique, mainly the loss of analyte. Drying times of only 4 hours have been associated with 14 – 23 % loss of PCBs from sediments. This, when the temperature was 20°C and the relative humidity was 25%. The main part of the total PCB losses, up to 90%, occurred after the first 4 hours of drying. The heat, air flow, water content and grain size are believed to affect the amount of PCBs in sediments. The congener pattern found in the sediments may also be altered in the drying process. The change in congener pattern are most distinct in sediments with PCB source from lighter Aroclors (Chiarenzelli et al. 1996).
For the first batch, dryness was achieved after 12 hours and for batch 2 after 48 hours. Different amounts of sediment were dried between the different batches and that can account for the difference in drying time. The Soxhlet extraction chambers were cleaned by starting an extraction with clean thimbels and toluene while the samples were drying. The Soxhlet blank, which is a method blank were treated in the same way as the samples. The wash continued for 24 hours before the extractions were stopped at reflux and the toluene was discarded. The dry sediments were then weighed, and 10 g of sediments were placed in each thimble. The
decision to use 10 g of sediments was based on previous studies which used similar amount and method for extraction of sediments (Table 2).
5 Table 2: Examples of Soxhlet extraction methods used in literature
Amount of sediment used
Solvent Time (h) Reference
5 g n-hexane:DCM 1:1 v/v 16 (Parera et al.
2004) 10 g Hexane, acetonitrile, toluene/methanol, acetone:hexne 1:1, 2-propanol, cyclohexane, or acetone:hexane 16 – 24 (US EPA 1996) 1 g Methanol 18 h (Jurado-Sánchez et al. 2013)
2 g Chloromethyl: Acetone 18 h (Hawthorne et al.
2000)
Since the sediments contained varying amounts of water, the water present in the sediments had to be estimated with the intent of having 10 g of sediment present for the analysis. For batch 1, the sample representing Södra Cell Värös intake, only 8.17 g were used since the amount of water in the sediment was underestimated. After the drying process were completed, only 8.17 g of sediment were present and used for the analysis.
A PCB internal standard was added to the sediments, 25 µl was added (Table 3), and the extraction started. The extraction was then discontinued after 24 hours and the toluene used for the Soxhlet extraction was evaporated using a rotary evaporator until ~1 ml of extract remained in the round bottom flask. The rotary evaporator was washed with hexane before usage. The temperature of the water bath was 45°C, and the pressure was 1000 mbar for hexane and 80 mbar for toluene.
Table 3: Internal standard added to the samples. Compound Amount (pg µl-1) 13C PCB #70 30 13C PCB #101 30 13C PCB #118 30 13C PCB #105 30 13C PCB #153 30 13C PCB #138 30 13C PCB #156 30 13C PCB #180 30 13C PCB #170 30 13C PCB #194 30 13C PCB #206 30 13C PCB #209 30 13C PCB #47 3 13C OCDD 5
3.3 Metallic copper clean-up
If any of the samples at this stage had a yellow discolouration, metallic copper was added to the eluate to react with sulphur that can disturb the analysis at higher levels. The copper was added to the round bottom flask and allowed to react. The extract was then removed from the
6 round bottom flask and placed in a new flask. The copper was rinsed with hexane three times to ensure that most of the analyte had been transferred to the new round bottom flask. The new round bottom flask was evaporated until ~1 ml of the extract remained.
Batch 1 was at this point split, around half of the solvent was removed and stored in 8 ml amber glass vials. The partition occurred in case the clean-up method were to be altered and the extract would be needed for a second clean up.
3.4 Silica clean-up
The extract was then added to a 10% deactivated silica for purification. The deactivated silica was prepared by weighing 70 g of silica, and adding 7 ml of MilliQ to the silica. The silica was shaken for three hours on a shaker, and additionally shaken by hand every 15 minutes under the three hours to ensure that the silica was homogenized. This method for the silica preparation was used since there was no access to the rotary shaker at the time of the experiment. The setup can be seen in fig. 2
To prepare the deactivated silica column, a piece of glass wool was placed at the bottom of a column and rinsed with ethanol, hexane, and dichloromethane, DCM. To the column, 10 g of silica was added, followed by a layer of anhydrous sodium sulphate to remove traces of water from the organic solvents used.
The columns were then rinsed with 40 ml of hexane before the addition of the sample. The eluates were collected after the samples were added to the column. After the samples were added the round bottom flasks were rinsed three times with hexane and added to the silica. The analytes were then eluted by addition of 100 ml hexane. The extract was then evaporated using the rotary evaporator until ~ 1 ml of the extract remained.
For the first batch, the extracts were transferred to 8 ml amber glass vials, and the previous vials washed with hexane three times. The extracts were then evaporated until near dryness with nitrogen gas. The extracts were then transferred to GC vials and 25 µl of toluene was added. The 8 ml amber glass vials were rinsed 3 times with hexane. The GC vials were then placed under nitrogen gas.
7 Figure 2: Silica columns
The samples were then analysed using GC-MS. When evaluating the results, it was noticed that the samples were too impure to provide useful mass spectra and chromatograms, see Appendix A. The samples were after the GC-MS analysis further purified by using a minisilica column.
The minisilica column were prepared in a Pasteur pipette, with glass wool acting as a plug at the end. The pipette and the glass wool was initially rinsed with ethanol, hexane, and DCM. The last drops of DCM were pushed out, the column was then packed. Around 2 cm of KOH silica and 2 cm 40% H2SO4 silica were added to the pipette (Fig. 2). The silica was then covered with sodium sulphate. The minisilica column was then rinsed with two pipettes of hexane before the sample was added. The collection vials were placed under the pipettes as soon as the samples were added. The analytes were eluted with 2 pipette volumes of hexane. The eluate was collected in 8 ml amber glass vials and then evaporated until near dryness before being transferred to the GC vial, in the same manner as stated previously.
The chromatograms were matched to the NIST library. The NIST database match comes in four parts, one part gives the name of the suggested structure, denoted hit list, and a visual representation of the first compound on the hit list. The third part is the mass spectra of the sample and the first three hits on the hit list. The fourth part is a delta plot where the sample spectra and the NIST reference spectra is compared to each other. For an ideal match, the delta plot should show no peaks. On the y-axis, the positive side indicates a peak present in the sample and negative side represents a peak present in the reference spectra.
8 Batch 2 was analysed in the same manner as the previous batch, the only difference was that the GC-MS analysis only took place after the minisilica clean-up, not before and after. Table 4: GC-MS full scan conditions
Inlet
Injection type Splitless
Volume 1 µl
Heater 250°C
Pressure 12.29 psi
Total flow 24.3 ml/min
Purge flow
Split vent 50.0 ml/min @ 2 min Gas saver 20.0 ml/min @ 4.0 min
Column
Constant flow 14.40 psi
Flow 1.4 ml/min
Average velocity 44 cm/sek
Manufacturer specifications
Model number J & W 122-5032
Column DB-5
Capillary
30 m * 250 µm * 0.25 µm nominal
Oven settings
Oven ramp Temperature
increase °C/min
Next °C Hold
Initial 90 2 min
Ramp 1 7 320 5 min
Ramp 2 0
Run time 39.86 min
Solvent delay 4 min
GC-MS
Brand Hewlett & Packard
Model GC 6890 plus series
Model MS 5973
Ionization mode Positive Electron ionization (EI+)
Scanning range Batch 1: m/z 50-500
9 Batch 2 was also analysed for PCBs using atmospheric pressure gas chromatography, APGC. Table 5: APGC conditions
Inlet
Injection type Splitless
Pulse time 1 min
Heater 280°C
Pulse pressure 450 kPA
Column
Constant pressure 14.40 psi
Flow 1.4 ml/min
Average velocity 44 cm/sek
Manufacturer specifications
Model number Cat #: 12623
Serial #: 941009
Column Rtx®-5MS
Capillary
30 m * 0.25 mmID * 0.25 µm no
Oven settings
Oven ramp Temperature
increase °C/min
Next °C Hold time
Initial 180
Ramp 1 3.50 260 0 min
Ramp 2 6.50 300 0 min
Run time 31.01 min
Solvent delay 3 min
Ion source
Transferline Temperature 280°C
Make up gas Nitrogen
Make up gas flow 370 ml/min
APGC
GC Agilent 7890
10
4. Results and Discussion
The results from the full scan experiment are presented as chromatograms that can be seen in Appendix A, B and C. The NIST spectral database from 1998 was used to determine which compounds were present in the sediments from the full scan experiment. Each match was of varying quality since many of the mass spectra obtained did not match well with the
suggested spectra.
The chromatograms from the intake and output from the same industry and batch are compared to each other to determine which peaks are present in the output but not in the input. This, in order to determine which compounds might have been released from the two industries, the chromatograms were examined, and attempts were made to identify peaks that appears in the output but not in the input. The chromatographic pattern were examined and each pair of chromatograms provided a varying number of peaks that can only be found in the output. . This approach resulted in several peaks that were examined, and the peaks with the highest probability of obtaining a good match were chosen from each sample. The
chromatograms are examined for peaks only present in the output and NIST matches are generated for each peak. The NIST comparisons with the highest probability of providing and accurate match have been presented in the thesis. The selection process is based on the delta plot and if the main peaks, such as molecular ions or indicative ions, are present in the spectra.
The chromatograms seen in Appendix A, batch 1 before the minisilica clean-up, have not been matched to the NIST library, due to the poor peak resolution. Appendix B show the total ion chromatogram, TIC, from batch 1 for toluene, the Soxhlet blank and the sample from the bay. Appendix C show the TIC for toluene, the Soxhlet blank and the sample from the bay for batch 2.
11
4.1 Ringhals
When comparing the samples from Ringhals batch 1 (Fig. 3,4) some peaks appeared in the output, but not in the input. Some of these peaks were suggested to be 2,3-dihydro-1,1,3-trimethyl-1H-indene, chlorocycloheptane, 2,3,4,6-tetramethyl-bibenzyl and, erucylamide.
Figure 3: Total ion chromatogram, TIC, of Ringhals input, batch 1
Figure 4: TIC of Ringhals output, batch 1
The samples from Ringhals batch 2 (Fig. 5,6) were n-methyl-n-phenyl benzamine, 3-(1,1´-biphenyl-4-YL)butanenitrile, 1H-indene, 2,3-dihydro-1,1,3-trimethyl-3-phenyl and, 1(2-bromoethyl)-4-methoxy-benzene. R1 Time 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 37.50 % 0 100 LRH_150528_003 Scan EI+ TIC 3.26e7 27.23 13.41 13.28 13.24 5.22 13.55 21.49 20.68 20.03 18.56 17.68 16.95 14.45 16.17 18.67 27.08 21.60 24.24 23.38 24.26 25.45 28.07 28.15 28.21 28.47 31.00 30.82 31.23 32.25 R2 Time 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 37.50 % 0 100 LRH_150528_004 Scan EI+ TIC 3.95e7 13.77 13.73 13.47 13.35 5.24 10.84 12.74 8.44 11.71 13.81 24.76 22.69 19.91 18.47 15.29 16.05 18.29 21.39 24.04 25.67 27.03 28.02 28.72 30.13 32.94
12 Figure 5: TIC of Ringhals intake, batch 2
Figure 6: TIC for Ringhals output, batch 2
The NIST database output can be seen in Appendix D together with an analysis of the spectra (Fig. 38-45).
These substances are highly possible matches for some of the peaks found in the output of Ringhals. Each match is of varying quality, and the peaks with the highest probability of providing an accurate match have been included in the thesis.
Erucylamide is used as a slip and antiblocking agent for polyolefins (Fine Organics 2012). Polyolefins is a polymer made from any simple alkene and is for example used in detergent to strengthen the package (Fine Organics 2012; Vasile 2000). Erucylamide is also found in leaf essential oil from Citrus medica. Since C. medica has its origin in Bangladesh (Bhuiyan et al. 2009), it is unlikely to be found in sediment in Kattegat from the C. medica plant.
R1 Time 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 37.50 % 0 100 LRH_150613_003 Scan EI+ TIC 5.29e7 16.50 16.47 7.54 6.38 5.92 4.99 15.95 15.27 11.12 9.52 8.81 13.15 11.46 17.07 29.90 18.89 17.22 29.67 28.49 26.42 25.99 24.27 20.52 23.71 21.91 25.64 27.29 34.38 31.89 30.40 32.61 36.22 36.91 R2 Time 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 37.50 % 0 100 LRH_150613_004 Scan EI+ TIC 5.67e7 18.58 18.52 17.91 17.65 15.82 4.10 7.76 18.86 27.66 18.97 21.02 26.97 19.58 21.09 26.08 24.44 21.18 23.85 24.67 28.12 36.38 28.55 28.91 30.39 34.89 32.34 36.77
13 For the other compounds, no information on their uses or origin have been found. Since no information have been obtained about the other substances, it is difficult to determine their origin.
4.1.1 Indicative ions for chlorine and bromine
Indicative ions for chlorine and bromine have been examined from the sample from Ringhals intake (Fig. 11). There are some peaks present in both chromatograms for chlorine, m/z 35, and 37. This supports the presence of chlorine at several retention times, among them 5.92, and 16.22 minutes. The noise levels are much higher in the chromatogram displaying m/z 35. The signal for m/z 37 is nearly ten times higher than for m/z 35 for the main peak, this suggests that the peak with retention time 16.07 and 16.45 is from an impurity. The other peaks present, 5.92, 7.54, 24.25, 25.88, 26.68, and 27.92, seems to conform to the theoretical distribution of chlorine. In conclusion, chlorine seems to be present in the sample, since several peaks appear with similar retention times.
Figure 7: Indicative ions for chlorine, m/z 35 and 37
The indicative ions for bromine (Fig. 10), show some peaks with retention time between 15.27 and 19.55 minutes, since peaks are present in both chromatogram, it supports the theory that bromine is present in the sample from Ringhals. The isotope ratios for bromine are acceptable for some peaks, such as 15.27, where the isotope ratio are nearly ideal. For the main peak, 16.83 min, the isotope ratio does not resemble bromine, the abundance for the lighter isotope is 32.9%. This is far from the ideal value, suggesting that the peak at 16.82 min does not contain bromine. In conclusion, there are evidence of bromine being present in the sample from Ringhals.
R1 Time 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 % 0 100 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 % 0 100 LRH_150613_003 Scan EI+ 35 6.56e3 27.92 24.25 5.92 24.24 16.22 16.21 5.95 16.03 7.54 22.52 16.24 22.51 20.52 18.33 17.11 25.88 26.68 27.96 28.56 29.69 30.36 32.53 35.53 33.38 36.93 LRH_150613_003 Scan EI+ 37 8.09e4 16.45 16.36 16.32 16.19 16.07 16.51 16.94 24.27 18.73 22.53 25.88 27.91 29.90 31.92 34.29
14 Figure 8: Indicative ions for bromine, m/z 79 and 81
R1 Time 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 % 0 100 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 % 0 100 LRH_150613_003 Scan EI+ 79 5.98e5 16.83 7.54 4.99 5.59 16.50 16.46 15.27 9.97 9.52 11.11 13.15 16.98 30.42 27.56 27.34 17.43 24.30 26.32 19.55 28.47 31.10 32.62 33.75 36.91 LRH_150613_003 Scan EI+ 81 1.22e6 16.83 16.48 15.28 5.61 32.54 31.45 30.43 29.70 28.10 16.98 26.45 33.15 34.82 36.91
15
4.2 Södra Cell Värö
When comparing the samples from Södra Cell Värö batch 1 (Fig. 7,8) some peaks appeared in the output, but not in the input. Some of these peaks were suggested to be 2,2’-diethyl-1,1’-biphenyl, (2-methylphenyl)-methyl-carbamic acid, 2,6,10,14-tetramethyl-nonadecane, and 2,6,10,14,18-pentamethyl-eicosane.
Figure 9: TIC of Södra Cell Värö intake, batch 1
Figure 10: TIC of Södra Cell Värö output, batch 1
The samples from Södra Cell Värö batch 2 (Fig. 9,10) were suggested to be 1-chloro-3-methyl-benzene, (E)-stilbene, 3,4,5,6-tetramethyl-phenanthrene and, erucylamide.
VB1 Time 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 37.50 % 0 100 LRH_150528_005 Scan EI+ TIC 4.37e7 13.76 13.66 13.43 13.23 5.27 12.70 11.24 8.46 9.86 13.89 21.71 21.63 21.31 20.67 20.57 27.03 27.00 24.61 23.29 26.46 29.23 27.17 28.38 31.24 31.18 31.11 33.07 31.27 37.22 34.78 VB2 Time 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 37.50 % 0 100 LRH_150528_006 Scan EI+ TIC 3.60e7 15.76 15.63 14.94 14.17 13.19 15.79 25.41 24.23 15.94 18.43 16.83 23.24 23.22 18.52 21.38 18.64 32.43 25.53 27.09 27.58 28.75 29.70 32.98 33.56
16 Figure 11: TIC of Södra Cell Värö intake, batch 2
Figure 12: TIC of Södra Cell Värö output, batch 2
The NIST database output can be seen in Appendix D together with an analysis of the spectra (Fig. 46-53).
These substances are highly possible matches for some of the peaks found in the output of Södra Cell Värö. Each match is of varying quality, and the peaks with the highest probability of providing an accurate match have been included in the thesis. The substances may have its origin from the paper mill, since the long carbon chains may be a component from the raw material used in the process.
Since erucylamide is found near both Ringhals and Södra Cell Värö, either its source is of a natural origin, or the chemical is released from either of the industries and is transported with the currents in the sea.
For the other compounds, no information on their uses or origin have been found. Since no information have been obtained about the other substances, it is difficult to determine their origin. VB1 Time 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 37.50 % 0 100 LRH_150613_005 Scan EI+ TIC 5.49e7 17.07 16.85 15.96 7.54 7.36 4.11 6.42 5.97 5.05 12.53 7.75 8.84 9.99 13.64 31.96 29.77 28.27 28.24 20.97 17.37 17.94 19.21 25.10 24.50 21.93 26.16 28.31 31.17 34.44 33.01 35.29 36.25 VB2 Time 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 37.50 % 0 100 LRH_150613_006 Scan EI+ TIC 5.52e7 19.46 18.08 15.80 4.09 10.05 20.05 26.9227.78 21.31 26.22 26.12 24.46 24.19 22.32 28.41 28.69 36.60 29.00 35.53 35.05 33.34 30.43 32.56 34.03
17
4.3 Polychlorinated biphenyls
The results from the APGC can be seen in Table 8, appendix E. The results from the PCB analysis (Fig. 13), are shown below. The peaks representing Ringhals intake saturated the detector, and only the base of the peaks were present. Therefore the peaks were not integrated. There were no attempts made to integrate the peaks since the levels were believed to be completely misguided if prefrormed. The peaks from all samples are also shown in graphical form below in Fig. 14, which is a cropped axis of Fig. 13.
In table 5 in appendix E, it can be seen that the blank contain high levels of the analysed PCBs. This might be due to contaminations that occurred in the sample preparation. The samples may contain additional analytes than the PCB that may interfere with the PCB
analysis, this is seen in the large shift in retention time which could be up to 0.5 minutes. This problem should also affect the samples, and should be kept in mind when examining the results. Due to the high levels found in the blank, the levels are only indicative.
Figure 13: PCB levels in sediment samples from the different sites. The blue stars indicates peaks with too high concentration of PCBs which saturated the MS. The red stars indicate peaks with bad peak shape.
0 2000 4000 6000 8000 10000 12000 PCB 28 PCB 52 PCB 101 PCB 118 PCB 153 PCB138 PCB 180 p g/g d .w . s ed im en t
18 Figure 14: PCB levels in sediment samples from the different sites, the blue stars indicate saturation of the MS, and the red stars indicate peaks with bad peak shape.
The sample from Ringhals intake, PCB #101, #118, #153 and #138 saturated the detector, since only the base of the peaks were seen in the chromatograms. PCB #180 from Ringhals intake is several times larger than the other sites, supporting the saturation theory. The intake of both industries seems to contain higher levels of PCBs. Although, this may be a biased statement, both intake points come from places believed to contain higher levels of
contaminants. Ringhals intake comes from a small harbour near the intake where the extra activity associated with the harbour might increase the PCB levels. This is due to a sampling error, since the sediments were too well packed to obtain samples closer to the actual input. The only sediment that were obtained from Ringhals intake were from the harbour, initially the effect of the harbour was unknown.
The intake for Södra Cell Värö was taken from the river Viskan, which passes through Borås, a town associated with the textile industry. Kattegat can therefore receive contaminants from Viskan, as it absorbs pollutants from the nearby land and industries as it passes through Sweden, and deposits them in the sea (Forsman & Edlund n.d.). Since the intake is situated in the river Viskan, no other sampling point was available for the intake from Södra Cell Värö. PCB #118 showed signs of co-elution in Ringhals output and Södra Cell Värös output, and they were therefore not integrated in those samples. In general, the peaks had a poor peak shape and the co-elution is most likely from several PCBs, so obtaining any useful
information is hard.
In another report from Swedish EPA, the levels of several congeners of PCBs (PCB #28, #69, #73, #52, #84, #89, #90, #101, #106, #118, #153, #138, #164, #163, and #180) show a
maximum level of 25000 for the PCBs previously mentioned in pg g-1 d.w in the Baltic Proper region in 1990 (Wiberg et al. 2013). The PCB 180 levels from Ringhals intake for alone account for around 40% of the concentration in the highest level found in the study.
0 100 200 300 400 500 600 700 800 900 1000 PCB 28 PCB 52 PCB 101 PCB 118 PCB 153 PCB138 PCB 180 p g/g d .w . s ed im en t
19 A study showed that the mean concentration along the Baltic Sea were for PCB #101 210 pg g-1 d.w., PCB #118 was 130pg g-1 d.w., PCB #153 was 400 pg g-1 d.w., PCB #138 was 110 pg g-1 d.w and for PCB #180 it was 220 pg g-1 d.w. in 1985 (Sobek et al. 2015).
For PCB #101 the levels are all lower than the mean.
For PCB #118 all sampling sites were lower than the levels from the Baltic Sea coast. For PCB #153, all but Södra Cell Värös intake were lower than the mean. The intake from Södra Cell Värö had around 700 pg g-1 d.w, the highest level at a sampling site in the Baltic Sea was 13 000 pg g-1 d.w.
For PCB #180 the levels were lower than the mean from the Baltic Sea, except for Södra Cell Värös intake and Ringhals intake. The levels near Södra Cell Värö are close to the mean value from the study. The levels near Ringhals intake are much higher than the mean level for PCB #180, but the highest level measured in the study was 17000 pg g-1 d.w. from one of the sampling sites. Ringhals intake are much lower than the highest level measured. The PCB levels from Ringhals intake are believed to be the sampling point with the highest
concentration of PCBs. The levels from Ringhals intake saturates the detector causing the top of the peak to be removed in some cases the base were nearly invisible. No attempt at
integrating the levels of those PCBs have been made.
Since Ringhals intake showed high levels, 11071 pg/g sediment of PCB #180, the remaining PCBs were believed to be seen in the chromatograms from the GC-MS screening.
The indicator ions for PCBs with one chlorine substitution can be seen fig. 15. The two ions displayed are m/z 188 and 190, these are the two ions with the highest and second highest intensity. Since the peaks displayed in either chromatogram are not present in the other chromatogram, the single chlorine substituted PCBs are most likely found in levels below the GC-MS limit of detection, (LOD).
Figure 15: Indicative ions for PCB with one chlorine substitution
In fig. 16 the indicative ions for PCBs with two chlorine substitutions. Some peaks are visible in both chromatograms, such as 22.53. There is a large peak disturbing the chromatogram displaying m/z 224, suggesting that the chromatogram suffers from impurities. Since this peak is visible in the GC-MS full scan experiment, the levels are believed to be quiet high.
R1 Time 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 % 0 100 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 % 0 100 LRH_150613_003 Scan EI+ 188 5.80e5 16.98 15.28 14.78 16.91 16.83 17.02 17.12 36.11 33.30 34.15 LRH_150613_003 Scan EI+ 190 8.61e4 22.30 22.08 12.06 20.21 19.03 12.91 31.45 31.43 28.45 27.63 26.27 23.97 23.78 33.28 33.75 36.91 34.17
20 Figure 16: Indicative ions for PCB with two chlorine substitution
In fig. 17 the indicative ions for PCBs with three chlorine substitution can be seen. Several peaks can be seen in both chromatograms, especially peaks around retention time 24.26 and 27.32 minutes. Since the peaks can be seen with the GC-MS full scan, it indicates high levels, supporting the saturation theory from the PCB analysis with the APGC. Since the standards for the APGC analysis are not visible in the samples when using the GC-MS approach, the peaks cannot be matched to a specific PCB.
Figure 17: Indicative ions for PCB with three chlorine substitution
Fig. 18 show the two largest ions for PCBs with four chlorine substitutions. They show a similar pattern in both chromatogram, suggesting that PCBs with three chlorine atoms are present in high levels. Peaks with retention time of 22.53, 24.27, 27.25, 27.92, 29.18, 29.34 minutes are possible PCBs.
R1 Time 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 % 0 100 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 % 0 100 LRH_150613_003 Scan EI+ 222 2.19e5 22.53 20.21 19.03 16.47 20.40 36.22 24.27 24.34 27.93 27.24 29.6730.23 32.79 33.73 LRH_150613_003 Scan EI+ 224 9.88e4 34.37 22.53 19.96 19.21 18.61 17.05 15.05 22.44 22.00 23.90 31.89 24.03 24.32 29.37 28.57 24.38 29.68 30.77 32.81 34.44 34.53 35.87 36.96 R1 Time 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 % 0 100 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 % 0 100 LRH_150613_003 Scan EI+ 256 7.36e5 24.26 23.70 24.96 25.88 26.68 27.32 31.94 31.10 27.92 LRH_150613_003 Scan EI+ 258 2.43e5 24.26 24.16 23.32 21.56 20.24 31.10 24.96 25.88 30.42 26.68 30.22 27.32 28.69 32.13 33.09 33.75 35.43
21 Figure 18: Indicative ions for PCB with four chlorine substitutions
Fig. 19 show the two largest indicative ions for PCBs with five chlorine substitutions. Many peaks are present in both chromatograms, such as the peaks with retention time 24.27, 24.96, 25.88, 26.68, and 27.32 minutes. Those peaks are suggested to be PCBs with four chlorine substitutions.
Figure 19: Indicative ions for PCBs with five chlorine substitutions
Fig 20 show the two largest indicative ions for PCBs with six chlorine substitutions. Several peaks are present in both chromatograms, among them the peaks with retention time 25.73, 26.50, 27.25, 27.92, and 28.58 minutes. R1 Time 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 % 0 100 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 % 0 100 LRH_150613_003 Scan EI+ 292 4.08e5 24.28 22.53 27.92 27.25 25.73 29.18 29.34 LRH_150613_003 Scan EI+ 294 2.03e5 24.28 22.53 22.32 27.92 27.25 24.77 34.38 31.89 29.68 32.56 35.84 R1 Time 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 % 0 100 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 % 0 100 LRH_150613_003 Scan EI+ 326 1.14e6 25.88 24.96 24.27 26.68 27.32 28.36 33.18 29.69 LRH_150613_003 Scan EI+ 328 7.27e5 25.88 24.96 24.26 26.68 27.32 29.67 28.29 30.40 33.19
22 Figure 20: Indicative ions for PCBs with six chlorine substitutions
Fig. 21 show the two main indicative ions for PCBs with seven chlorine substitutions. Peaks with retention time 27.47, 28.29, 28.87, 29.69, and 30.37 minutes, among other are suggested to be PCB with six chlorine substitutions.
Figure 21: Indicative ions for PCBs with seven chlorine substitutions
Fig. 22, 23 and 24 show indicative ions for PCBs with eight, nine and ten chlorine substitutions. All six chromatogram show noise at the end of the chromatogram, and the levels of PCBs are believed to be below LOD for the GC-MS screening.
R1 Time 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 % 0 100 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 % 0 100 LRH_150613_003 Scan EI+ 360 7.38e5 27.92 27.25 26.50 25.73 28.57 30.37 31.66 LRH_150613_003 Scan EI+ 362 5.83e5 27.92 27.25 26.50 25.73 28.58 31.58 30.22 32.06 34.07 R1 Time 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 % 0 100 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 % 0 100 LRH_150613_003 Scan EI+ 394 8.87e4 29.69 28.87 28.29 27.47 32.99 30.37 32.80 32.04 33.00 35.42 34.26 35.49 35.83 LRH_150613_003 Scan EI+ 396 8.57e4 29.69 28.88 28.29 27.48 25.49 30.37 32.95 32.22 31.95 34.45 35.66 36.16 36.84
23 Figure 22: Indicative ions for PCBs with eight chlorine substitutions
Figure 23: Indicative ions for PCBs with nine chlorine substitutions
R1 Time 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 % 0 100 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 % 0 100 LRH_150613_003 Scan EI+ 430 2.50e4 35.51 34.58 34.19 33.84 32.67 31.98 30.72 36.72 36.99 LRH_150613_003 Scan EI+ 428 3.98e4 35.51 35.37 33.85 33.79 33.27 32.41 36.78 36.95 36.97 R1 Time 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 % 0 100 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 % 0 100 LRH_150613_003 Scan EI+ 464 5.60e3 36.23 35.63 34.72 33.92 33.76 32.71 36.99 LRH_150613_003 Scan EI+ 461 1.02e4 36.66 36.61 35.10 34.42 33.80 33.54 36.86 36.95
24 Figure 24: Indicative ions for PCBs with ten chlorine substitutions
Swedish EPA have proposed five levels of PCB concentrations, which is no levels, low levels, average levels, high levels and very high levels. Ringhals intake is 5 times higher than the level characterised as very high levels for PCB #180. The saturated peaks are most likely of higher concentrations than PCB #180 and are therefore most likely placed in the very high levels category.
Table 6: Swedish EPA classifications of PCB contaminated sediments, µg/g dry weight of sediment (Naturvårdsverket 1999)
PCB congener No levels Low levels Average levels
High levels Very High levels #101 0 0 – 0.16 0.16 – 0.60 0.60 – 2.0 >2.0 #118 0 0 – 0.15 0.15 – 0.60 0.60 – 2.0 >2.0 #153 0 0 – 0.030 0.030 – 0.30 0.30 – 3.5 >3.5 #138 0 0 – 0.30 0.30 – 1.2 1.2 – 4.1 >4.1 #180 0 0 – 0.10 0.10 – 0.40 0.40 – 1.9 >1.9
Table 7: Comparison of levels with the Swedish EPA * indicates peaks too large to integrate ** indicate bad peaks
PCB congener Ringhals intake Ringhals output Södra Cell Värö intake Södra Cell Värö output Bay
#101 * Low Low Low Low
#118 * ** Low ** Low
#153 * Average High Average Average
#138 * Low Average Low Low
#180 Very high Low Low Average Low
Comparing the results obtained to the data from Swedish EPA, the values are shown as levels instead of numerical values (Table 7). The result indicates that there are PCBs present in the sediments, in high levels at some of the sampling sites, Södra Cell Värö intake, and at very high levels in some samples, Ringhals intake. The industries seem to have no effect on the
R1 Time 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 % 0 100 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 % 0 100 LRH_150613_003 Scan EI+ 498 2.13e3 36.90 35.87 35.41 35.07 33.46 33.02 31.33 29.47 28.29 22.08 16.72 4.72 11.44 9.59 6.38 16.23 19.01 21.56 24.56 26.79 36.98 36.99 LRH_150613_003 Scan EI+ 500 1.15e3 36.73 36.49 35.13 34.43 31.97 31.60 24.13 16.42 15.84 10.40 6.25 5.79 7.99 9.09 10.86 12.13 14.69 16.87 18.59 17.83 19.87 22.58 30.28 29.07 27.35 25.90 33.74 33.72 36.96 36.99
25 PCB concentration in sediments surrounding the two industries since the levels are lower in the output than the input. The levels from Ringhals intake indicates very high levels of PCBs, since peaks are visible in the GC-MS experiment. PCB #180 is believed to have the lowest concentration of all the PCBs measured. PCB #180 is according to the Swedish EPA very high suggesting that the levels for the other PCBs indicative very high levels.
The levels indicated with an asterisk, probably falls in the category of very high levels, since they have a higher concentration than PCB #180 in the GC-MS screening, and PCB #180 falls under the category of very high levels.
The levels proposed by the Swedish EPA have in recent years been out dated, but it is still applied routinely.
The pattern of the congeners may by lined to a specific Aroclor. The conclusion that PCB #180 which was the only peak small enough to integrate, is supported since PCB #180 is a heptachlor, shown in fig. 21, and only noise is visible, suggesting that the levels of the other PCBs with two, three, four, five, and six chlorine substitutions are higher than the PCB #180 levels from the APGC.
For the patter determination, the chromatograms from the GC-MS indicative ions were used. The chromatogram with the highest intensity belonged to the heptachlorobiphenyls. The second highest intensity was hexachlorobiphenyls, followed by trichlorobiphenyls and tetra chlorobiphenyls. The highest intensity is plotted, and the other peaks are plotted by their relative intensity, providing a plot used to match against the congener patterns from EPA. When examining the PCB congeners, the highest peaks belong to heptachlorobiphenyls, PCB #40 – PCB #81. The Aroclor patterns from EPA show higher levels of PCBs with other degrees of chlorination, such as Aroclor 1016 and 1242, where the maximum concentration is between PCB #15 and PCB #30. Aroclor 1254a and 1254g have a maximum between PCB #110 and #120. Aroclor 1260 have its maximum around PCB #180. Neither of those match the pattern from the sample collected from Ringhals intake. The Aroclors that match these high levels of heptaklorobifenyls, is Aroclor 1248a and 1248g from EPA. Aroclor 1248a and 1248g display a very similar pattern, and since the specific PCB congener cannot be
determined from the GC-MS, the Aroclor cannot be differentiated between Aroclor 1248a and 1248g (US EPA n.d.)
5. Conclusion
The aim of this study was to determine if there were any differences between the intake and output of the two industries and if potential contaminants linked to either industry can be found in the bay. This was done by applying a full scan approach. With this approach, numerous errors were encountered. Many of the peaks were not sufficiently resolved to provide reliable results, and most of the results cannot give any conclusive information regarding the presence of contaminants in the samples. Often, the type of compound group present can be matched with the reference library, but the specific compound cannot be determined. The samples from Ringhals generally contained longer carbon chains, whereas Ringhals generally contained more PAHs.
The high levels found in the Soxhlet blank may be due to the fact that the sample was
26 in the Soxhlet blank. The levels in the Soxhlet blank are higher than the levels in some
samples, so any levels below that of the Soxhlet blank may be used as indicative levels of the PCB concentration.
Since the study is performed by using a screening method, identification of analytes may not be feasible due to the co-elution of different analytes with similar retention time. Some compounds will co-elute and overlapping chromatogram and mass spectra will most likely be obtained. Overlapping mass spectra will then be much harder to match to databases since the mixed mass spectra will generate a different spectra and the possible matches will not fit the mass spectra suggested by the database. Not only will the purity of mass spectra be of issue, many compounds have several congeners, often with similar mass spectra and they cannot be differentiated by only using the mass spectra. For this kind of study, it is impossible to acquire standards for all compounds analysed, and therefore each match to the database is a tentative identification. There are other indications that can be used for the confirmation of the
compound than the mass spectra and retention time. If possible, a MS/MS spectra, and 13C standards for important ions can be used to further strengthen the conclusion. With the MS/MS approach, the limits of detection, LOD, are lowered, but all other analytes are lost in the process. The MS/MS approach does lower the LOD, but the risk of overcharging the detector increases with higher levels. The spectra from the full scan, all follow the criteria from the mass spectra, and the retention time in many cases seems plausible.
The screening method provided some good matches for each chromatogram. The chromatogram do indicate differences in the sediments from the five locations. Each chromatogram show different peaks, suggesting that the industries have affected the
composition of the sediments. Since the peaks have not been identified, to maintain within the scope of the thesis, only differences in the pattern have been examined.
The main part of the thesis was to determine what substances have been released by the industries, and that was proven to be difficult. The industries could not be linked to any compounds found in the sediments. But the general difference in chromatographic pattern suggests that there is a difference in the sediments. The differences in the sediments might be worth additional studies with another approach. There appear to be less extractable material in the outtake of the two industries, suggesting that they have not affected the superficial
environment surrounding Värö peninsula. Since the samples acquired for the intakes are from sites believed to show additional contamination, such as a harbour for Ringhals intake, and the fabric industry for Södra Cell Värö intake. Again, the intake for Ringhals should have been acquired closer to the actual intake and not in the harbour, but due to sampling errors, this was unavailable. The sediments were packed too hard for the Ekman sediment grabber to acquire a sample. The results show a general decrease in the intensity in the chromatogram from the Ringhals comparison, but the decrease might be from the additional elements added by the harbour. The same problem may arise with the water from Södra Cell Värö, the contamination from the previous releases in Viskan.
Several of compounds suggested by the NIST library, any information on the origin of the compound were unattainable. In many instances medical safety data sheets (MSDS) have been found, and they do not provide information on the sources of the compounds. To maintain within the scope of the thesis, the sampling was limited to five locations.
27 small river that flows into the bay in the middle of the peninsula, which should have been investigated as well.
When running the APGC, there are some limitations with the standards, as the standard added only contains a few different congeners of PCBs, and the analysis is thereby limited to those PCBs alone. The large time shift that appeared in retention time due to the high content in the samples made the identification via retention time nearly impossible. The samples were analysed in the end of a long sequence, causing several of the peaks to tail. The samples were also highly contaminated, since no further clean-up took place between the GC-MS screening and the APGC analysis, causing the retention time to shift. This shift in retention time makes the identification of PCB congeners hard. Since the retention time from the PCB standard does not match the retention times in the samples.
The industries seems to not affect the levels of PCB in the sediments, since they should be elevated at the output sites, and that is not the case. The PCB levels are much lower in the output locations suggesting that the industries do not affect the levels of PCBs in the sediments. Ringhals output, Södra Cell Värös output and the bay seem to contain similar levels of PCBs, suggesting that the sites are exposed to similar sources.
Since an Aroclor pattern can be determined from the GC-MS analysis, the levels of PCB #101, PCB #118, PCB #153, and PCB #138, are most likely higher than the levels found for PCB # 180. The Aroclor found near Ringhals intake probably have its origin in the activity associated with the harbour, not the neighbouring industries. This since the elevated levels are only found near Ringhals intake, and all other points display lower levels.
The initial intention was to compare the samples taken in the bay to the findings in the other samples; this was proven to be difficult, it has been excluded as a part of the experiment. Sop for further investigations, the bay should be more extensively compared to the two industries to determine if any releases takes place and affects the industries. Another possibility is to widen the search and examine more than the indicator PCBs. This may be done by using principal component analysis to determine the source of the PCB congeners. It would also be helpful in the Aroclor pattern recognition, since the levels in this thesis is only based on the peak height from the GC-MS screening.
To continue the studies, several additional sample should be examined for PCBs, which seems to be present at quite high levels in some of the sample locations. Preferably, a clean-up adapted for the analysis of PCBs should be used. Another possibility is to examine the exposure to seals, fish, crabs, or clams to determine the impact on wildlife. Preferably crabs and clams, since they live in the sediments. Water samples is another possibility to continue the study.
The results are inconclusive, so further studies are needed to achieve reliable results. The sampling should have taken place at another location than the harbour for the Ringhals intake. Since this location most likely do not represent what can be found closer to the intake, further from the harbour. The harbour near Ringhals intake is in a small bay, probably with limited water exchange, causing the levels released by the boats to increase. If sediment samples are to be acquired again, the Ringhals intake should not be taken from the small harbour but closer to the actual intake, on open water. The congener profile found in the harbour is probably due to the activity in the harbour. This since a similar profile is present in all sites
28 except for Ringhals intake, suggesting that the release of Aroclor is due to the activity in the harbour.
29
6. References
Ali, M. & Sreekrishnan, T.R., 2001. Aquatic toxicity from pulp and paper mill effluents: A review. Advances in Environmental Research, 5(2), pp.175–196.
Bhuiyan, M.N.I. et al., 2009. Constituents of Peel and Leaf Essential Oils of Citrus Medica L. Journal of Scientific Research, 1(2), pp.6–11.
Chiarenzelli, J. et al., 1996. Volatilization of polychlorinated biphenyls from sediement drying at ambient conditions. Chemosphere, 33(5), pp.899–911.
Edgren, M., Olsson, M. & Reutergårdh, L., 1981. A one year study of the seasonal variations of sDDT and PCB levels in fish from heated and unheated areas near a nuclear power plant. Chemosphere, 10(5), pp.447 – 452.
Elforsk, 2009. Miljöeffecter av stora kylvattenutsläpp - Erfarenheter från de svenska kärnkraftverken,
Eriksson, S.G., 2015. Personal communication.
Fine Organics, 2012. Material safety data sheet - Erucamide. , pp.1–8. Forsman, A. & Edlund, L.-E., Viskan. Available at:
http://www.ne.se/uppslagsverk/encyklopedi/l%C3%A5ng/viskan [Accessed July 1, 2015].
Google Maps, 2015. Google maps.
Hawthorne, S.B. et al., 2000. Comparisons of Soxhlet extraction, pressurized liquid extraction, supercritical fluid extraction and subcritical water extraction for
environmental solids: Recovery, selectivity and effects on sample matrix. Journal of Chromatography A, 892(1-2), pp.421–433.
Jurado-Sánchez, B., Ballesteros, E. & Gallego, M., 2013. Comparison of microwave assisted, ultrasonic assisted and Soxhlet extractions of N-nitrosamines and aromatic amines in sewage sludge, soils and sediments. Science of the Total Environment, 463-464, pp.293– 301. Available at: http://dx.doi.org/10.1016/j.scitotenv.2013.06.002.
Lau, C. et al., 2007. Perfluoroalkyl acids: A review of monitoring and toxicological findings. Toxicological Sciences, 99(2), pp.366–394.
Murphy, B.L. & Morrison, R.D., 2007. Introduction to environmental forensics 2nd ed., Elsevir Inc.
Nationalencyklopedin, Valthornssnäcka. Nationalencyklopedin. Available at:
http://www.ne.se/uppslagsverk/encyklopedi/l%C3%A5ng/valthornssn%C3%A4cka [Accessed June 21, 2015].
30 Parera, J., Santos, F.J. & Galceran, M.T., 2004. Microwave-assisted extraction versus Soxhlet
extraction for the analysis of short-chain chlorinated alkanes in sediments. Journal of Chromatography A, 1046(1-2), pp.19–26.
Ratia, H. & Oikari, a., 2014. Vertical distribution of AhR-activating compounds in sediments contaminated by modernized pulp and paper industry. Water Research, 50, pp.27–34. Available at: http://linkinghub.elsevier.com/retrieve/pii/S004313541300986X.
Secretariat of the Stockholm Convention, What are POPs?
Sobek, A. et al., 2015. Baltic Sea sediment records: Unlikely near-future declines in PCBs and HCB. Science of The Total Environment, 518-519, pp.8–15. Available at:
http://linkinghub.elsevier.com/retrieve/pii/S0048969715002466. Stibe, L., 2015. Email, personal communication.
Södra Cell Värö, 2015. Massafabrik i ständig utveckling. Available at:
http://www.sodra.com/sv/Massa/Vara-massabruk/Sodra-Cell-Varo/ [Accessed June 20, 2015].
Södra Cell Värö, 2014. Välkommen till Södra Cell Värö - En av världens största och mest moderna massaindustrier,
United Nations, 1995. Stockholm convention. US EPA, Aroclor plot EPA. Available at:
http://www.epa.gov/osw/hazard/tsd/pcbs/pubs/aroclor.htm [Accessed July 30, 2015]. US EPA, 1996. EPA method 3540c. , pp.1–8.
Vasile, C., 2000. Handbook of Polyolefins (2nd Edition) 2nd ed., New York: CRC Press. Vattenfall, 2012a. Ringhals – The largest Nordic power plant,
Vattenfall, 2012b. Technical information on Ringhals,
Wiberg, K. et al., 2013. Managing the Dioxin Problem in the Baltic Region With Focus on Sources To Air and Edible Fish,
Wulff, F. et al., 1993. A Mass-Balance Model of Chlorinated Organic Matter for the Baltic Sea: A Challenge for Ecotoxicology. AMBIO, 22(1), pp.27–31. Available at:
31
Appendix A- Batch 1, without minisilica
Figure 25: TIC from Toluene
Figure 26: TIC from Soxhlet blank
Toluen Time 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 37.50 % 0 100 LRH_150527_008 Scan EI+ TIC 6.53e6 4.41 39.10 36.93 35.70 34.75 34.34 33.67 37.18 39.45 39.65 Blank Time 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 37.50 % 0 100 LRH_150527_009 Scan EI+ TIC 5.11e7 4.29 4.37 14.99 14.18 13.99 4.46 4.54 13.57 15.24 26.85 26.13 25.05 22.70 21.79 20.86 30.11 28.13 30.28 31.54 33.25
32 Figure 27: TIC from Ringhals input
Figure 28: TIC from Ringhals output
R1 Time 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 37.50 % 0 100 LRH_150527_010 Scan EI+ TIC 2.71e7 21.73 13.37 13.20 13.16 18.39 13.55 18.27 18.17 18.16 18.07 21.66 21.51 20.64 20.26 20.19 21.95 22.06 22.29 25.77 24.35 27.08 29.1329.80 31.41 34.52 32.48 35.15 R2 Time 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 37.50 % 0 100 LRH_150527_011 Scan EI+ TIC 2.80e7 13.69 13.64 13.30 13.23 13.20 13.19 4.30 13.14 4.31 13.13 14.17 28.18 27.24 27.20 21.80 21.66 21.63 18.32 18.28 21.60 18.35 27.13 27.11 24.16 22.89 24.49 25.47 28.30 30.43 28.37 34.68 34.07 30.83 33.94 35.31 37.34 35.55 37.38
33 Figure 29: TIC from Södra Cell Värös input
Figure 30: TIC from Södra Cell Värö output
VB1 Time 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 37.50 % 0 100 LRH_150527_012 Scan EI+ TIC 3.35e7 13.88 13.52 13.28 13.22 13.19 13.16 13.14 5.45 5.37 5.61 13.99 18.16 16.54 16.52 16.50 16.48 18.13 18.12 17.92 18.33 21.84 21.16 18.36 20.72 20.66 18.57 23.92 30.65 28.52 24.25 27.09 24.72 24.74 26.20 30.62 29.32 31.3532.58 34.43 VB2 Time 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 37.50 % 0 100 LRH_150527_013 Scan EI+ TIC 3.19e7 4.31 4.62 17.38 16.07 15.63 4.72 14.57 14.31 17.95 17.96 24.48 24.47 18.57 21.79 21.72 21.65 21.62 21.60 24.12 24.11 24.08 25.31 31.13 30.72 30.70 30.60 27.11 31.16
34 Figure 31: TIC of sample from the bay
F Time 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 37.50 % 0 100 LRH_150527_014 Scan EI+ TIC 3.54e7 4.50 4.95 5.02 16.80 5.03 16.23 14.97 13.42 17.41 18.00 24.46 18.04 18.19 21.66 21.62 24.09 22.81 25.35 30.70 30.67 30.64 30.95
35
Appendix B: Batch 1 with minisilica
Figure 32: TIC of toluene blank
Figure 33: TIC of Soxhlet blank, after use of minisilica
Toluen Time 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 37.50 % 0 100 LRH_150528_001 Scan EI+ TIC 2.33e6 36.98 36.84 36.46 36.31 36.20 35.23 35.18 4.45 34.82 34.74 37.57 37.81 38.39 38.74 39.83 Blank Time 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 37.50 % 0 100 LRH_150528_002 Scan EI+ TIC 3.65e7 14.82 13.98 13.58 13.44 13.29 27.35 14.83 27.23 24.62 23.92 22.85 22.41 27.44 28.01 28.19 28.38 28.44 28.52
36 Figure 34: TIC of sample from the bay, after the use of minisilica
F Time 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 37.50 % 0 100 LRH_150528_007 Scan EI+ TIC 3.63e7 15.99 14.26 14.08 13.44 13.26 13.12 12.66 11.44 11.33 16.06 25.40 24.21 16.14 24.05 23.24 22.74 18.42 16.92 18.49 21.35 32.43 25.51 27.02 29.68 32.96 33.54
37
Appendix C: Batch 2
Figure 35: TIC of toluene
Figure 36: TIC of Soxhlet blank from batch 2
Toluene Time 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 37.50 % 0 100 LRH_150613_001 Scan EI+ TIC 2.31e7 4.12 4.15 6.55 4.34 5.24 Blank Time 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 37.50 % 0 100 LRH_150613_002 Scan EI+ TIC 4.93e7 18.24 18.20 18.10 17.58 16.79 15.85 4.20 7.64 6.01 11.58 27.68 26.96 18.52 18.83 21.02 19.79 26.01 25.52 21.13 24.74 28.55 36.56 35.00 28.89 34.00 32.53 31.81 36.89
38 Figure 37: TIC of sample taken in the bay from batch 2
F Time 5.00 7.50 10.00 12.50 15.00 17.50 20.00 22.50 25.00 27.50 30.00 32.50 35.00 37.50 % 0 100 LRH_150613_007 Scan EI+ TIC 6.07e7 4.10 17.77 17.65 17.56 17.44 15.93 4.32 7.60 7.40 6.07 8.88 13.66 27.71 17.84 26.94 21.01 19.37 17.93 21.11 26.33 26.14 24.47 22.29 29.95 35.01 30.48 33.45 31.75 35.55
39
Appendix D – NIST Database matches and analyses
The NIST database suggested 2,3-dihydro-1,1,3-trimethyl-1H-indene for the mass spectra generated at 6.26 min (Fig. 38). The spectra from the NIST library resembled the spectra obtained, but there were small variations between the reference and the sample spectra. The other hits from the NIST library were also resembling the mass spectra from the sample, therefore an identification was not possible. The conclusion that can be drawn was that the structure most likely have an indene, or an aromatic structure. Many of the suggested matches from the NIST library all contain indene, benzene, or naphthalene as the suggested backbone for the structure. The proposed analyte most likely have an aromatic structure, with a methyl substitution.
