Phytoremediation of long-term PCB-contaminated soil: A greenhouse feasibility study.

(1)

IN

DEGREE PROJECT THE BUILT ENVIRONMENT,

SECOND CYCLE, 30 CREDITS ,

STOCKHOLM SWEDEN 2016

Phytoremediation of long-term

PCB-contaminated soil: A

greenhouse feasibility study.

JIEYUAN WANG

KTH ROYAL INSTITUTE OF TECHNOLOGY

(2)

(3)

TRITA-LWR Degree Project ISSN 1651-064X

LWR-EX-2016:19

P

HYTOREMEDIATION OF

L

ONG

-

TERM

PCB-

CONTAMINATED

S

OIL

:

A

G

REENHOUSE

F

EASIBILITY

S

TUDY

Jieyuan Wang

(4)

Jieyuan Wang TRITA LWR Degree Project 2016:19

ii © Jieyuan Wang 2016

Degree Project in Environmental Engineering and Sustainable Infrastructure

Done in association with the Plant Mental Research Group, Department of ecology, envi-ronment and plant science of Stockholm University

Division of Land and Water Resources Engineering Royal Institute of Technology (KTH)

SE-100 44 STOCKHOLM, Sweden

(5)

Phytoremediation of long-term PCBs-contaminated soil: A greenhouse feasibility study

iii SUMMARY

Soil contamination, which can be caused by various pollutants, has at-tracted public attention for many years. Polychlorinated biphenyls (PCBs) are priority soil contaminants because of their toxicity and ten-dency to persist in soils, sediments and to escape biological degradation. PCBs were banned in 1970s; but even now, they still exist in the envi-ronment, threating human’s health. Previously, the joint sealants added with PCBs were used in construction. In this study, the long-term PCB-contaminated soils were from a residential area in Västerås Municipality where the PCBs keep leaching from buildings to the soils for several years. The existing PCBs in this area were mainly highly chlorinated PCBs. Due to economic and efficiency factors, phytoremediation was chosen for PCB removal.

The objective of this thesis is to investigate the feasibility of using plants to remove PCBs from the contaminated soil at a greenhouse scale, and to use a site-specific guideline model for the risk assessment of this con-taminated site. After a literature review, four plant species were selected for the greenhouse cultivation, consisting of alfalfa (Medicago sativa L.),

white clover (Trifolium repens L.), horseradish (Armoracia rusticana L.) and

tobacco (Nicotiana tobacum L.). The soil samples from different depths

were divided into three types, and planted with seeds. The plants culti-vated in the greenhouse were maintained for 92 days and then the con-centrations of PCBs in the initial and remediated soils were analyzed by GC-MS.

(6)

(7)

iv SUMMARY IN SWEDISH

Kontaminerad mark, orsakad av olika typer av föroreningar, har fått allt mer uppmärksamhet under senare år. En särskilt uppmärksammad typ av föroreningar är polyklorerade bifenyler (PCB) vilka har hög prioritet för rening på grund av sin toxicitet och höga persistens i jord och sediment beroende på långsam biologisk nedbrytning (Mackova et al., 2006). PCB förbjöds på 1970-talet men de finns fortfarande kvar i miljön och är ett hot mot människors hälsa. Tidigare, innan förbudet, tillsattes PCB till fogmassor som användes till konstruktioner, t.ex. bostadshus. I denna studie användes jord från ett bostadsområde i Västerås stad. Jorden har under många år kontaminerats med PCB från läckande fogmassor i bo-städer. PCB:erna i detta område är huvudsakligen högklorerade PCB:er. För rening av marken har fytoremediering föreslagits, baserat på bl.a ef-fektivitet och ekonomiska faktorer.

Syftet med denna studie är att undersöka möjligheten att använda växter för att avlägsna PCB från den förorenade jorden i växthusförsök, och att applicera platsspecifika riktvärden i en modell för riskbedömning av om-rådet. Efter genomgång av litteraturen valdes fyra växtarter ut för od-lingsförsöken, alfalfa (Medicago sativa L.), vitklöver (Trifolium repens L.),

pepparrot (Armoracia rusticana L.) och tobak (Nicotiana tobacum L.). Jorden

hämtades från det aktuella området i Västerås från olika provtagnings-djup. Jordproverna delades in i tre kategorier och växterna planterades. Växter odlades i växthus under 92 dagar. Koncentrationerna av PCB i de initiala och sanerade jordarna analyserades med GC-MS.

(8)

(9)

v ACKNOWLEDGEMENTS

I owned my deepest gratitude to my supervisor, Prof. Maria Greger, for giving me the opportunity to do this thesis at Stockholm University. Thank you for the insightful guidance and constant encouragement pro-vided to me throughout this whole thesis work. I would also like to thank Prof. Jon Petter Gustafsson at KTH, for all the insightful sugges-tions and generous help right from the beginning of this project. Thank you especially for helping me to figure out the site-specific guideline model.

Special thanks also go to Arifin Sandhi for introducing me to the plant metal research group at Stockholm University, Tommy Landberg for pa-tiently helping me to solve the various problems I met during the exper-iment. Thanks to the Department of Ecology, Environment and Plant Science at Stockholm University for providing such a nice environment to study and work in.

(10)

(11)

Jieyuan Wang TRITA LWR Degree Project 2016:19 vi

T

ABLE OF

C

ONTENT SUMMARY iii SAMMANFATTNING iv ACKNOWLEDGEMENTS v Table of Content vi Abbreviations vii Abstract 1 1. Introduction 1 1.1. Soil contamination 1 1.2. Polychlorinated biphenyls (PCBs) 2 1.3. Phytoremediation 3

1.4. Environmental risk assessment 5

1.5. Literature review of current research 7

1.6. Objective of the study 8

1.7. The selection of study plants 8

2. Materials and Methods 9

2.1. Contaminated site description 9

2.2. Soil collection and characterization 10

2.3. Greenhouse cultivation 10

2.4. Polychlorinated biphenyl (PCB) analysis 11

2.5. Soil properties 13

3. results 13

3.1. Polychlorinated biphenyl (PCB) analysis 13 3.2. Site specific guideline values 15

4. discussion 16

4.1. Polychlorinated biphenyl (PCB) analysis 16

4.1.1. Alfalfa-planted pots 16

4.1.2. Tobacco-planted pots 16

4.1.3. Horseradish-planted pots 16

4.1.4. White clover-planted pots 17

4.1.5. Unplanted pots 17

4.1.6. Other impact factors 17

4.2. Risk assessment for initial and remediated soils 18

5. recommendations 19

5.1. Analysis of microorganisms and plants 19 5.2. Analysis of infiltrated water from pots 19 5.3. TRIAD method for risk assessment 19

6. conclusion 20

References 21

Appendix I – Results from the chemical analysis I

Appendix II – The PCBs II

Appendix III – Map of quaternary deposits in the study area III

Appendix IV – Map of the depth to bedrock in the study area IV

(12)

vii

Abbreviations

PCB: polychlorinated biphenyls PAH: polycyclic aromatic hydrocarbon SGU: The Geological Survey of Sweden EPA: Environmental Protection Agency Cis: Individual guideline value for soil ingestion Cdu: Individual guideline value for skin contact

Cid: Individual guideline value for inhalation of soil particles Civ: Individual guideline value for inhalation of vapors

(13)

1

A

BSTRACT

Polychlorinated biphenyls (PCBs) are persistent organic pollutants, which were banned several decades ago but still exist in the environment, posing a threat to human health. Previously, joint sealants containing PCBs were used in construction. In this study, the long-term PCB-contaminated soils were from a residential area in Västerås Municipality, where the PCBs have been leaching from buildings to the soil for several years. The objective of this thesis is to investigate the feasibility of using plants to remove PCBs from the contaminated soil at a greenhouse scale, and to use a site-specific guideline model for the risk assessment of this contaminated site. After a literature review, four plant species were se-lected for the greenhouse cultivation including alfalfa (Medicago sativa L.),

white clover (Trifolium repens L.), horseradish (Armoracia rusticana L.) and

tobacco (Nicotiana tobacum L.). The plants cultivated in the greenhouse

were maintained for 92 days and then the concentrations of PCBs in the initial and remediated soils were analyzed by GC-MS. The results indi-cate that the selected plant species can enhance the removal of high chlorinated PCBs from soils. In the risk assessment, the contents of PCBs in soils were higher than the calculated site-specific guideline, which means it is urgent to implement measures for protecting residents’ health.

Keywords: PCBs, Phytoremediation, Risk assessment, Site-specific guideline

1. I

NTRODUCTION

1.1. Soil contamination

(14)

2

As Europe has become wealthier, the problem of soil contamination has become more obvious and urgent. The European Commission has adopted the Soil Thematic Strategy (Commission Proposal COM (2006) 231) and proposals for a Soil Framework Directive in 2006 (Commission Proposal COM (2006) 232) specifically to protect soils (Science Communication Unit, 2013). According to the European Environmental Agency (2007), the number of sites in the EU where potentially polluting activities have been carried out is approximately three million. Of these, an estimated 250,000 sites may need urgent remediation (Science Communication Unit, 2013).

Soil remediation refers to various processes designed to remove hazard-ous contaminants from soil. Soil remediation usually requires cleaning at high quality standards, so the soils can consequently benefit commercial cultivation, wild flora and fauna (Technologywater, 2014). According to the remediation location, the techniques of soil remediation consist of ex situ remediation and in situ remediation. The removal methods are di-vided into three groups: biological treatment, thermal treatment and physical-chemical treatment. Biological remediation mainly includes composting, phytoremediation and enhanced bioremediation. The ther-mal treatment involves incineration and vitrification, while the physical-chemical treatment is mainly comprised of soil washing, physical-chemical extrac-tion, soil vapor extracextrac-tion, solidification and chemical reduc-tion/oxidation.

Although there are many efficient soil remediation methods, the selec-tion of proposed method should firstly consider the site specific condi-tions, such as the properties of contamination, composition of the soil, location and structure of the site, groundwater levels, nearby surface wa-ter and the available time and money.

1.2. Polychlorinated biphenyls (PCBs)

(15)

3

The specific PCB congeners differ in their physical properties according to the number and position of chlorines (Mackova et al., 2006). Lower chlorinated PCB congeners (less than 6 chlorines) are more volatile, wa-ter soluble and easily degradable, while higher chlorinated PCBs with greater partition coefficient (log Kow) are sorbed more strongly into the soil organic matter and become less bioavailable to soil microorganisms and plants (McFarland & Clarke, 1989). Furthermore, their lipophilicity allows them to bioaccumulate and biomagnify in higher trophic levels of the food chain (Xu et al., 2010). The marine mammals at the top of the food chain therefore accumulate high levels of PCBs (Mackova et al., 2006).

The wide use of Arochlor and its resistance to degradation has caused PCBs to have a broad geographic distribution. Over thirty years after the production of PCBs ceased, they were still found everywhere even in-cluding the polar regions. In industrial countries, contamination originat-ed from inadequate disposal and leaks from equipment. In remote areas where PCBs were not used, the contamination resulted from atmospher-ic transport (Mackova et al., 2006). Because of this situation, PCBs are still a global problem. Moreover, PCB was classified as a persistent or-ganic pollutant (POPs) in the Stockholm Convention (adopted in 2001). The toxicity of PCBs has been recognized since the 1930s and although effects on adult humans are rather low, chronic exposure to PCBs induc-es serious neurobehavioral, immunological, reproductive and endocrine disorders in children (Aken et al., 2010).

The extreme persistence of PCBs in the environment and their ability to bio-magnify in the food chain causes considerable environmental and human health risks that need remedial action. Besides, PCBs added in caulking, sealants, paints, adhesives and other building materials from 1930s are leached into soils in some regions. Indoor air quality is also se-verely affected by the presence of joint sealants containing PCBs (Kohler et al., 2005). The most common methods for remediation of PCB-contaminated soil are incineration and landfilling, which require excava-tion and transportaexcava-tion of the soil. This has high energy costs and de-stroys the soil matrix (Åslund et al., 2007). These methods are also not well suited for treating large volumes of soil with lower, but still signifi-cant, levels of PCB contamination (Zeeb et al., 2006). As a result, there is a growing interest in developing new remediation technologies such as phytoremediation, which is more environmentally friendly and cost-effective.

1.3. Phytoremediation

(16)

4

Fig. 2. Main mechanisms of phytoremediation (Pilon-Smits, 2005)

While definitions and terminology vary, the different phytoremediation processes can be summarized as in Figure 2. Pollutants in soil are taken up and transported inside plant tissues (phytoextraction) or volatilize in-to the atmosphere (phyin-tovolatilization). Plants change their root envi-ronment, so pollutants are bound to soil particles and/or that pollutants absorbed by roots which will decay to humus in which the pollutant is bound (phytostabilization) (Aken et al., 2010). Furthermore, there are some other processes specific for organic compounds. Plant roots can excrete enzymes, which degrade organics in the rhizosphere (rhizodegra-dation). Plants take up, degrade and transform organics in their biomass (phytodegradation). Roots function as a filter in water removing pollu-tants and cleaning the water (rhizofiltration). The whole plant (sub-merged plants) functions as a filter in water removing pollutants and cleaning the water (phytofiltration). The leaves of plants function as a fil-ter, removing pollutants and cleaning the air (foliar filtration).

Phytoremediation offers several advantages over other remediation strat-egies; low cost because of the absence of energy-consuming equipment and limited maintenance, no or limited negative impact on the environ-ment because of the in situ nature of the process and wide public ac-ceptance as an attractive green technology. Phytoremediation also in-creases the amount of organic carbon in the soil which, in turn, stimulates microbial activity. In addition, the establishment of deep-rooted vegetation helps to stabilize soil. Relative with many traditional remediation engineering techniques, phytoremediation is a fledging tech-nology intended to address a wide variety of surficial contaminants. Ac-cording to a report, approximately 80% of the polluted groundwater are within 20m of the surface (Mackova et al., 2006). This suggests that a significant number of sites are potentially suitable for low cost phytore-mediation applications. Furthermore, phytorephytore-mediation offers potential beneficial side-effects, such as erosion control, site restoration, carbon sequestration and feedstock for biofuel production. As autotrophic or-ganisms, plants use sunlight and carbon dioxide as energy and carbon sources. From an environmental standpoint, plants can be seen as ‘natu-ral, solar-powered, pump-and-treat systems’ for cleaning up contaminat-ed soils (Aken et al., 2010).

(17)

metabo-Phytoremediation of long-term PCBs-contaminated soil: A greenhouse feasibility study

5

lism, plants usually lack the biochemical pathways necessary to achieve total mineralization of recalcitrant pollutants, such as PAHs and PCBs (Schnoor et al., 1995). Phytoremediation can therefore lead to undesira-ble effects, such as the accumulation of toxic metabolites that may be re-leased to the soil, enter the food chain, or volatilize into the atmosphere (Eapen et al., 2007). In addition, planted trees need several years to reach mature size and in temperate regions, plants have limited activity during the dormant season (Schnoor et al., 1995). Moreover, plants are living organisms with specific oxygen, water, nutrient and pH limits that must be maintained. Researchers still need to establish whether contaminants can accumulate in leaves and be released during litter fall or accumulate in fuel-wood or mulch. It may be difficult to establish the vegetation be-cause of soil toxicity or possible migration of contaminants off site by binding with soluble plant exudates (Schnoor et al., 1995). Possible mi-gration of contaminants off-site by binding with soluble plant exudates is a concern, but to date, none of these problems has been observed. In some situations, regulatory restrictions will not allow contaminants to be left in situ, even when a vegetative cover prevents erosional pathways of exposure (Alkorta & Garbisu, 2001). Phytoremediation is also still in de-velopment. The technology is not yet widely accepted by regulatory agencies and therefore not commonly used.

1.4. Environmental risk assessment

Focus on soil protection in combination with growing pressure of man-agement of contaminated sites increase the need for risk assessment methods for cost-effective characterization of contamination (Ribe et al., 2012). Several countries and regions have developed their own method-ologies for performing risk assessment of contaminated sites (Ribe et al., 2012). The US Environmental Protection Agency’s methodology for ecological risk assessment, the Superfund Program, aims to quantify both potential adverse effects to humans and the ecological risks at con-taminated sites (US Environmental Protection Agency, 1997). In Swe-den, the Environmental Protection Agency (Naturvårdsverket) has re-vised the model for calculating guideline values and published new general guideline values for contaminated soil together with a calculation program for site-specific guideline values (Jones et al., 2006). This is a necessary change from a generic approach using benchmark values to-wards a site-specific risk assessment approach (Ribe et al., 2012).

(18)

Markanvä-Jieyuan Wang TRITA LWR Degree Project 02:11

6

ndning) and for less sensitive land use MKM (Mindre Känslig Mar-kanvändning) respectively (Naturvårdsverket, 2009). The areas consid-ered for sensitive land use are mainly the residential areas, schools and playgrounds, while the less sensitive areas include industrial areas, com-mercial areas and roads. Generally, if the field concentrations exceed the guidelines, a comprehensive risk assessment should be considered (Naturvårdsverket, 2009). Therefore, the site-specific guideline values (SSGVs) will be generated based on the investigated site’s characteristics. This is conducted by the use of the model supplied by the Swedish EPA.

Fig. 3. Exposed pathways considered in the risk assessment (Naturvårdsverket, 2009)

A risk assessment should always consider the recipients that are exposed. The Swedish specific guideline model identifies human health, environ-ment, groundwater and surface water as conservation objectives (Natur-vårdsverket, 2009). As Figure 3 shows, humans are exposed through six pathways, including soil intake, skin contact, inhalation of soil particles, inhalation of vapors, intake drinking water, and intake of vegetables cul-tivated on-site. In addition, the intake of fish from a lake downstream of the contamination source is also a possible pathway. This exposure pathway of the consumption of fish is calculated by the model but does not affect the final guideline value because of the high level of uncertain-ty in the results. The uncertainuncertain-ty stems from the long and complex transport pathways from the point source to a nearby surface water body and to the difficulty in relating adverse health effects with the consump-tion of fish from the water body (Naturvårdsverket, 2009).

(19)

site-Phytoremediation of long-term PCBs-contaminated soil: A greenhouse feasibility study

7

specific guideline, the monitoring plan and mitigation measures would be carried out if necessary.

The human health risk presented by environmental PCBs remains prob-lematic. In any case, risk assessment for PCB-contaminated soil was nec-essary in this study. It was performed using the site-specific guideline model from Swedish EPA. Based on the model, the guideline value was calculated only for the sum of seven PCB congeners, consisting of PCB-28, PCB-52, PCB-101, PCB-118, PCB-138, PCB-153 and PCB-180. The Swedish EPA generally regards the sum of them as 20% of total PCBs (Naturvårdsverket, 2009).

Since the PCBs have a high affinity for soil, the Kd value needs to be considered as well. The Kd value is generally used to estimate the leach-ing of contaminants from the soil, which is assumed to represent the re-lationship between the concentrations of the contaminant in solid phase of the soil to the concentrations in the pore water. For organic contami-nants such as PCBs, the model calculates the Kd value from the ability of the substance to bind to organic carbon (the Koc value) and the amount of organic carbon in the soil. Furthermore, organic pollutants may bind to mobile organic carbon in the soil water and the groundwa-ter. If the concentration of mobile organic carbon is high, it may transport organic pollutants and can thereby increase the leaching of or-ganic contaminants with a high ability to bind to oror-ganic carbon. This may affect the guideline value. So the organic carbon content of soil should also be considered.

According to the Swedish EPA, the KOC value of 7PCB is 22,500, and

the organic carbon content fOC of default value is 0.02

(Naturvårdsver-ket, 2009). The equation below indicates the relation between the Kd value and the Koc value, where Coc is the concentration of the organic compound that is associated with organic matter (Schwarzenbach et al., 2005). It was used to calculate the Kd value for the risk assessment.

w oc oc d oc

C

f

K

=

(Schwarzenbach et al., 2005) (Equation 1)

1.5. Literature review of current research

In recent years, a number of papers have addressed the role of plants in remediating contaminated soils, especially for the PCB-contaminated soil. Several plants have proven to enhance the dissipation of PCBs in soil. Suzuki et al. (1977) reported the uptake and translocation of PCBs in soybeans (Glycine max L.); Iwata and Gunther (1976) in carrots (Daucus carota L.); and Sawhney and Hankin (1984) in beets (Beta vulgaris L.),

tur-nips (Brassica rapa L.), and beans (Phaseolus vulgaris L.). In each case, it was

reported that the lower chlorinated congeners were found more abun-dantly in the shoots than the roots of the various plant species (Iwata & Gunther, 1976; Suzuki et al., 1977; Sawhney & Hankin, 1984). In the re-search of Webber et al. (1990) about PCB uptake and translocation from contaminated sewage sludge, carrots (Daucus carota L.) had the highest

PCB concentration in the plant tissues followed by cabbages (Brassica oleracea var. captata L.) and corn (Zea mays L.). Moreover, alfalfa was found

to selectively support the growth of not only symbiotic nitrogen fixing bacteria but also a variety of PCB degrading bacteria, such as Pseudo-monas fluorescens F113 (Ryslava et al., 2003). Chekol et al. (2004) also conducted the experiments on three legumes and four grass species for PCB remediation, including alfalfa (Medicago sativa L.), flatpea (Lathyrus sylvestris L.), sericea lespedeza (Lespedeza cuneata Dum. -Cours),

(20)

8

switchgrass (Panicum virgatum L.) and tall fescue (Festuca arundinacea

Schreb). They found the presence of plants significantly increased the PCB biodegradation. Zeeb et al. (2006) proved tall fescue is well adapted to a wide range of soils and has the ability to grow in highly PCB-contaminated soils as well.

However, most of these previous studies have been conducted with pot-ted plants in the greenhouse with intentionally Arochlor-contaminapot-ted soil. Comparing with the aged polluted soil, PCBs do not strongly bind to soil particles in man-made contaminated soil. Hence the removal effi-ciency might be higher than that in long-term polluted soil. In addition, since the greenhouse conditions can be operated manually, there are still some differences between it and field trial. The variable weather and temperature also could affect the plant growth. An investigation of using alfalfa inoculated with Rhizobium in field had a significant decrease after 3-month cultivation (Xu et al., 2010). Furthermore, Mackova et al. (2009) summarized 10 years of studies using plants and their cooperation with microorganisms in the root zone to remove PCBs from the long-term contaminated soil for both pot and field experiments, which indicated that tobacco (Nicotiana tabacum), nightshade (Solanum nigrum), alfalfa

(Medicago sativa) and horseradish (Armoracia rusticana) have a good

poten-tial for PCB removal. Additionally, a pilot-scale field trial of phytoreme-diation of PCBs that was implemented by Åslund et al. (2007) showed that it is possible to accumulate significant amounts of PCBs in the shoots of pumpkin (Cucurbita pepo ssp pepo cv. Howden), sedge (Carex normalis) and tall fescue (Festuca arundinacea).

1.6. Objective of the study

The PCB-contaminated site in this study was located in a residential area, where the PCBs from joint sealants of buildings were continiosly leached into the soils. The most likely cause for leaching is natural weathering of joint sealants (Herrick et al., 2007). Therefore, the remediation method should be long-term, efficient and economical. Among the available re-mediation methods used for PCB-contaminated soil, phytorere-mediation is a good choice for this scenario.

Thus, the objective of this master thesis is to investigate the feasibility of using plants to remove PCBs from soils. It is preliminary research per-formed at a greenhouse scale and later it will be extended into the field scale if positive results can be acquired after the greenhouse cultivation. The specific aims of this study were to answer the following research questions:

 Can the selected plant species remove PCBs from soils?

 Which species can remove the most PCBs among the selected plant

species?

 Which specific PCB congeners can be removed the most?  Does the initially contaminated soil have a health risk for people?  Does the remediated soil still have a health risk for people?

1.7. The selection of study plants

(21)

9

research that has been done about phytoremediation of long-term PCB- contaminated soil in the greenhouse or in the field. Based on the litera-ture review, it was concluded that alfalfa, white clover, tobacco and horseradish could grow with about 20% removal efficiency of PCBs in greenhouse (Mackova et al., 2009). Due to the common properties of long-term PCB-contaminated soils and experimental feasibility, alfalfa (Medicago sativa L.), white clover (Trifolium repens L.), tobacco (Nicotiana to-bacum L.) and horseradish (Armoracia rusticana L.) were chose for this

in-vestigation.

2. M

ATERIALS AND

M

ETHODS

2.1. Contaminated site description

Fig. 4. The location of the PCB-contaminated site (Google Maps, 2016)

Because of the wide use of PCBs in caulking and elastic sealant materials between the 1950 and the 1970s, PCBs have been leached from the house facades down to the soils for more than half a century (Herrick et al., 2007). In Sweden, sealant-containing PCBs also have been used from 1956 to 1973, mainly in building façade constructions in joints between cement blocks, around windows, balconies, etc. (Corner et al., 2002). The study area (59°35'56.0"N, 16°30'32.0"E) is a PCB-contaminated site located in a residential area of Västerås Municipality. Västerås is a city in central Sweden, located on the shore of Lake Mälaren in the province Västmanland. This contaminated site is between two standard six-floor detached apartment houses in a parking environment (Figure 4). Before the construction of these residential buildings, this area was plain agricul-tural soils. The area has never been used for any industries, which im-plies that the background concentration of PCBs should be zero. There-fore, all the PCBs found in the soils near the buildings are from the leaching of joint sealants.

(22)

10

2.2. Soil collection and characterization

The soil samples were dug out from three spots (No. 1, No. 2 and No. 3) in the study area, which are at a distance of three meters from each other and one meter from the house walls. This work was done by Structor Miljöteknik, an engineering consultancy company.

The soil samples were divided into three kinds of soils for the following experiment. The initial soil A was a mixture of soils at the depth of 30-40 cm from three spots (No. 1, No. 2 and No. 3). The initial soil B was a mixture of soils at the depth of 40-50 cm from spot No.3, while the ini-tial soil C was a mixture of soils at the depth of 20-40 cm from spot No.3. The initial soil A and C was brown with some small gravels, but the initial soil B looked different. It was black and seems to contain some humic substances. The soil samples were also totally dried when we re-ceived them.

Table 1: The soil properties of initial contaminated soil smaples Sample Description (depth) PCB-138 (mg/kg) PCB-153 (mg/kg) PCB-180 (mg/kg) 7 PCB sum (mg/kg) Total PCBs (mg/kg) Organic matter (%) pH Initial soil A (30-40 cm) 0.0039 0.0053 0.0036 0.013 0.065 3.56 6.05 Initial soil B (40-50cm) 0.0044 0.0047 0.0034 0.013 0.065 9.73 5.28 Initial soil C (20-40 cm) 0.0049 0.0061 0.0043 0.015 0.075 3.25 5.96 As Table 1 shows, there is no obvious variation of total PCBs concentra-tion with the soil depth. The initial soil A and C have similar organic matter content and pH values, while the initial soil C has higher organic matter content and lower pH. As for the specific PCB congeners, only seven PCB congeners were measured. The reason is explained in section 2.4. The initial concentrations of low chlorinated PCB congeners (PCB-28, PCB-52, PCB-101 and PCB-118) were all below the test limit 0.002 mg/kg (dry weight), so the high chlorinated PCB congeners (PCB-138, PCB-153 and PCB-180) were regarded as the main pollutants in soil samples. PCB-153 had the highest concentration levels of the PCB con-geners in all the soil samples.

2.3. Greenhouse cultivation

The seeds of alfalfa (Medicago sativa L.), white clover (Trifolium repens L.),

horseradish (Armoracia rusticana L.) and tobacco (Nicotiana tobacum L.)

were produced by Nelson Garden AB. Apart from the unplanted pots,

(23)

11

Table 2: The original setting of experimental pots

Sample No. Soil Amount (g) Planted species

Initial soil A 3 50.0015 Unplanted 8 50.0321 Unplanted 1 50.0714 Alfalfa 2 50.0526 Alfalfa 4 50.0282 Tobacco 10 50.0882 Tobacco 5 50.0747 White clover 9 50.0376 White clover 6 50.0652 Horseradish 7 50.0924 Horseradish Initial soil B 15 50.0501 Unplanted 16 50.0535 Unplanted 11 50.0615 Alfalfa 12 50.048 Alfalfa 13 50.0987 Tobacco 14 50.0657 Tobacco Initial soil C 25 50.0300 Unplanted 26 50.0683 Unplanted 17 50.0167 Alfalfa 18 50.0805 Alfalfa 19 50.0115 Tobacco 20 50.0647 Tobacco 21 50.0752 White clover 22 50.0261 White clover 23 50.0434 Horseradish 24 50.0476 Horseradish

Greenhouse cultivation of selected plants was carried out between 9 February 2016 and 10 May 2016 (a period of three months), at the De-partment of Ecology, Environment and Plant Science at Stockholm Uni-versity in Sweden.

The temperature of the greenhouse was kept at 23°C and the moisture content was set as 60%. The plants growth mainly depended on the sun-light and manual watering every day. Also the lamps in the greenhouse were open during the day without sunlight. No fertilizer was added dur-ing the whole greenhouse cultivation.

However, besides the expected plants, weeds grew in some pots as well and these were removed at once whenever identified. After 92 days of plant growth, the soils were sampled. The soils close to the roots were collected and divided into two sets. One was stored at -20℃ in the freezer for soil property analysis (organic matter content and pH value), and the other was stored for PCB analysis. The PCB removal efficiency (R) was calculated as:

% 100 0 0− × = C C C R t (Equation 2)

where C0 and Ct were soil initial and residual concentration of PCBs

re-spectively after 92 days of the greenhouse experiment

2.4. Polychlorinated biphenyl (PCB) analysis

(24)

12

the total PCB content of industrial PCB mixtures in the past (Saba, 2012). Moreover, the Swedish EPA also set the guideline value for the sum of these seven congeners in risk assessment and assumes its value to be 20% of the total PCB content (Naturvårdsverket, 2009).

Table 3: Polychlorinated biphenyl (PCB) congeners chosen for analysis (IUPAC number is the number given to the compound by the International Union of Pure and Applied Chemistry)

IUPAC Compound Name Chemical structures 28 2,4,4'-Trichlorobiphenyl (C12H7Cl3) 52 2,2',5,5'-Tetrachlorobiphenyl (C12H6Cl4) 101 2,2',4,5,5'-Pentachlorobiphenyl (C12H5Cl5) 118 2,3',4,4',5-Pentachlorobiphenyl (C12H5Cl5) 138 2,2',3,4,4',5'-Hexachlorobiphenyl (C12H4Cl6) 153 2,2',4,4',5,5'-Hexachlorobiphenyl (C12H4Cl6) 180 2,2',3,4,4',5,5'-Heptachlorobiphenyl (C12H3Cl7)

The initial and remediated soil samples were all sent to ALS Scandinavia

AB for PCBs analysis on 12th-May 2016. The test method performed in

ALS Scandinavia AB is based on the Swedish standard SS-EN 16167:2012, which is generally used for the quantitative determination of seven selected polychlorinated biphenyls (PCB-28, PCB-52, PCB-101, PCB-118, PCB-138, PCB-153 and PCB-180) in sludge, sediment and soil (Swedish Standards Institute, 2012). The analysis limit of PCB is 0.002 mg/kg (dry weight).

(25)

13

with a mass selective detection (MS) and a DB-XLB capillary column (30 m×0.25 mm i.d., and 0.25 µm film thickness). As for the GC condi-tions, the column was temperature-programmed at 60°C initially (1 min), followed with two ramps at 15°C /min to 90°C and then at 30°C /min to 290°C. The injector and detector temperatures were maintained at 290°C and 300°C respectively. The column was operated with an ul-trapure helium carrier gas at a flow rate of 1.0 ml/min. The whole run-ning time for each sample was set at 22.67 min. Results were calculated from the residual amounts of each congener peak of the sample, com-paring with the respective peaks of the controls. The analytical error was <25-30%.

2.5. Soil properties

The analysis of organic matter content and pH value for the initial and remediated soil samples was performed in the laboratory of the plant-metal group, Department of Ecology, Environment and Plant Science, Stockholm University. To 1.0 g of fresh soil, 5 ml water was added, and the suspensions were shaken for two hours, followed by one hour of sedimentation. Then the pH value was measured using a Metrohm 744 pH meter. As for the content of organic matter of soil samples, 5.0 g fresh soil was first dried at 105°C in the oven for 24 hours, and then heated at 550°C for two hours. The moisture was removed after drying at 105°C, whereas the loss-on-ignition was determined after 550°C heat-ing and assumed to represent the organic matter content. The detailed results are shown in Appendix I.

3.

RESULTS

3.1. Polychlorinated biphenyl (PCB) analysis

Because of the limited test budget, it was not possible to send all the soil samples from each living plants pot for PCB analysis. Therefore the re-mediated soils from each two parallel pots were bulked to one sample for PCB analysis. The concentrations in Table 4 represent the average values of two parallel pots, except alfalfa planted pot 1 and 2. They were used to check if there was any obvious difference between two parallel pots. All the data in Table 4 was from the final report provided by ALS Scandinavia AB.

Table 4: Amount of 7PCB taken up and percent removed Sample No.

(Initial con.)

Sample Description 7 PCB sum (mg/kg) 7 PCB removal (%) Soil A (0.013 mg/kg) 1 Alfalfa 0.016 0 2 Alfalfa 0.011 15.4%

4,10 Tobacco Plant did not germinate 5 White clover 0.0093 28.5% 9 White clover Plant did not germinate 6 Horseradish 0.0048 63.1% 7 Horseradish Plant died

3,8 Unplanted 0.015 0 Soil B (0.013 mg/kg) 11,12 Alfalfa 0.055 0 13,14 Tobacco _0.012 7.7% 15,16 Unplanted 0.011 15.4% Soil C (0.015 mg/kg) 17,18 Alfalfa _0.016 0

19,20 Tobacco _{Plant did not germinate} 21,22 White clover _0.013 13.3% 23,24 Horseradish _0.014 6.7%

(26)

14

Although the plants were given sufficient sunlight and water every day in the greenhouse, no tobacco plants germinated in soil A and C (pot 4, 10, 19 and 20). In planted soil A pots, pot 9 with white clover did not ger-minate, and the horseradish died in pot 7. In the remaining experimental pots, all plants were alive during the whole cultivation period.

As for the soil A, the removal of 7PCBs was greatest in horseradish-planted soil (63.1%), followed by white clover-horseradish-planted soil (28.5%) and alfalfa-planted soil (15.4%). In contrast, the sum of the seven PCB con-geners had increased by 15% in the unplanted control.

In soil B, only tobacco germinated and grew well. It removed 7.7% of 7PCBs after three months. Furthermore, the highest decrease of 7PCBs was in unplanted pots (15.4%), which was contradictory to soil A. Com-pared with the initial concentration, the content of 7PCBs in alfalfa-planted pots increased dramatically. These abnormal data should be dis-carded in the discussion.

Surprisingly, in soil C the horseradish and the white clover did not per-form as well as in soil A. The white clover removed 13.3% of PCBs, fol-lowed by the horseradish with only 6.7% removal efficiency. On the contrary, the 7PCBs in unplanted soil C increased by 13.3%.

Table 5: The removal efficiency of highly chlorinated PCB conge-ners Sample Description PCB-138 (mg/kg) Removal (%) PCB-153 (mg/kg) Removal (%) PCB-180 (mg/kg) Removal (%) Soil A Alfalfa 0.0052 0 0.0065 0 0.0046 0 Alfalfa 0.0035 10.3 0.0045 15.1 0.0027 25.0 White clover 0.0032 18.0 0.0036 32.1 0.0025 30.6 horseradish 0.0023 41.0 0.0025 52.8 <0.002 60.0 Unplanted 0.0052 0 0.0062 0 0.0037 0 Soil B Alfalfa 0.013 0 0.017 0 0.023 0 Tobacco 0.004 9.1 0.0051 0 0.0032 5.9 Unplanted 0.0036 18.2 0.0044 27.9 0.0028 17.7 Soil C Alfalfa 0.0051 0 0.0061 0 0.0045 0 White clover 0.0047 4.1 0.0053 13.1 0.0034 20.9 Horseradish 0.0043 12.2 0.0059 3.23 0.0038 11.6 Unplanted 0.0056 0 0.0065 0 0.0044 0 According to the detailed results of specific PCB congeners in Appendix II, the low chlorinated PCB congeners (PCB-28, PCB-52, PCB-101 and PCB-118) in all samples were below the test limit 0.002 mg/kg. So the removal efficiency in this study was only considered for highly chlorinat-ed PCB congeners (PCB-138, PCB-153 and PCB-180).

As Table 5 shows, in soil A where horseradish was planted, the highest decrease observed was for PCB-180 (60%), followed by PCB-153 (53.0%) and PCB-138 (41.0%). In addition, in the soil where alfalfa was planted (pot 2), the most removed pollutant, again, was PCB-180 (25%), fol-lowed by PCB-153 (15.1%) and PCB-138 (10.3%). But in the soil where white clover was planted, PCB-153 decreased the most (32.1%), fol-lowed by PCB-180 (30.6%) and PCB-138 (18.0%).

(27)

de-Phytoremediation of long-term PCBs-contaminated soil: A greenhouse feasibility study

15

creased by 27.9%, which is larger than the removal of PCB-138 (18.2%) and PCB-180 (17.7%).

However, PCB-138 reduced most (12.2%) in soil C where horseradish was planted, followed by PCB-180 (11.6%). Yet in the soil where white clover was planted, PCB-180 (20.9%) and PCB-153 (13.1%) were re-moved the most. There was no decrease of PCBs in the soil where alfalfa was planted, as well as in the unplanted soil.

3.2. Site specific guideline values

In order to calculate the site specific guideline values for risk assessment, the model NV Soil Guideline from the Swedish EPA was used in this work. The instruction regarding the use of the model was provided by the Swedish EPA (Naturvårdsverket, 2009). Since the contaminated site is in a residential area, the KM scenario was chosen for pre-defined sce-nario. According to the Swedish EPA, the generic value of 7PCB in the scenario KM is 0.008 mg/kg. Then several assumptions were made re-garding the future landuse and exposure in the site. Considering PCBs as organic compounds, the site specific parameters also include the Kd-value. The assumed values used in the model have been compiled in Ta-ble 6 below.

Table 6: The assumed site-specific values used in the NV Soil Guideline model

Parameters for the model Site-specific values KOC-value for 7PCB (l/kg) 22500

Kd-value for 7PCB (l/kg) 4400 Amount of organic carbon (kg/kg) 0.02 Water Content (%) 25 Hydraulic conductivity (m/s) 10-7 Length of contaminated site (m) 20 Width of contaminated site (m) 5

Depth to contaminant (m) 0.2 The aquifers thickness (m) 0.5 Exposure time for adults and children (day/year) 365 Plant intake from the site (%) 10

Table 7: The results of 7PCB risk assessment in the NV Soil Guideline model

Possible pathway Single pathway concentrations (mg/kg)

Intake of soil 0.5 Dermal contact 1.3 Inhalation dust 560 Inhalation vapors 6.3 Intake of drinking water 0.33

Intake of plants 0.12 Integrated health value 0.007

(28)

16

Generic guideline 0.008 Final guideline 0.007

Deciding factor Health value adjusted as regards exposure from other sources

The results of risk assessment are described in Table 7 above. The final site-specific guideline of 7PCB is 0.007 mg/kg, which is lower than the generic guideline for sensitive land use. Protection of groundwater was not considered in this area. The governing factor for this guideline value is the health value had been adjusted with regard to exposure from other sources

4.

DISCUSSION

4.1. Polychlorinated biphenyl (PCB) analysis

Generally, the removal efficiency of PCBs from soil varies for different plants and soils, usually depending on both bacteria and plant actions. Comparing with the unplanted pots, most planted pots show higher re-moval efficiency of PCBs. It means that the selected plant species can enhance the removal of PCBs from soils in this study. As for the specific PCB congeners, the decrease mainly occurred for PCB-153 and PCB-180. These highly chlorinated PCB congeners may be reduced to low chlorin-ated PCB congeners, which are more volatile and degradable.

4.1.1. Alfalfa-planted pots

As for the alfalfa-planted pots, the contents of 7PCBs went up in pot 1, but declined in pot 2. The analytical error (<25-30%) might result in this consequence. Moreover, during the greenhouse cultivation, the alfalfa germinated earlier than other plants and grew well for almost two months, though some shoots died in the last month. Therefore, the less lively alfalfa might slow down the PCBs’ degradation and result in a low removal efficiency (15.4%). In spite of the low efficiency in this study, it was still in agreement with the previous paper that the alfalfa can en-hance the removal of PCBs from the rhizosphere soil (Xu et al., 2010). Additionally, in the alfalfa-planted soil B, strange data of 7PCBs was ob-tained. This might be related to the high analytical uncertainty as well. It also indicates that the phytoremediation process is quite complex, which could not only be affected by the plant species. The soil properties and nutrient availability also play an important role in phytoremediation, which can influence the growth of plants and bacteria (Li et al., 2013). 4.1.2. Tobacco-planted pots

Considering the tobacco-planted pots, only the plants in soil B germinat-ed. This might be due to the high organic matter content (9.7%) of soil B, which is almost three times that in other soils. The tobacco seems to re-quire more humic substances for growth. However, Ryslava et al. (2003) proved that tobacco (Nicotiana tobacum L.) has a high potential to

accu-mulate PCBs in its tissue in field experiments. The highest decrease measured in tobacco-planted soil was 34.0% (Ryslava et al., 2003). Therefore, tobacco is more sensitive to the soil properties.

4.1.3. Horseradish-planted pots

(29)

con-Phytoremediation of long-term PCBs-contaminated soil: A greenhouse feasibility study

17

tact area between plants and soils. In addition, previous research indi-cates that horseradish in comparison with tobacco supports the growth of rhizosphere microflora better (Mackova et al., 2009).

4.1.4. White clover-planted pots

Though the white clover performed with 28.5% removal in soil A, one of the two parallel pots did not germinate. Moreover, the white clover that survived had fewer roots than that of horseradish. However, the better removal efficiency (13.3%) in comparison with horseradish was observed in soil C. Though the horseradish still had more roots than the white clover in soil C, soil properties had a more significant impact on the results. In addition, the activities of bacteria could affect the conse-quences as well. According to previous literature, white clover with its well-developed root system could create optimal conditions for biodeg-radation and its dense canopy could minimize PCB volatilization from soils (Vasilyeva et al., 2010). Therefore, white clover can help to stabilize PCBs in soils.

4.1.5. Unplanted pots

The opposite results were observed in the unplanted soil A. The increase of 7PCBs can be explained by the fact that other untested PCB conge-ners might be transformed to tested PCB congeconge-ners through reductive dehalogenation by anaerobic bacteria in soil. What’s more, the analytical error could have caused this as well. However, in soil B, the highest de-crease (15.4%) was found in unplanted pots. This could be because soil B was obtained from deep soil layers, which had more microorganism degradation of PCBs. Some specific bacteria can degrade highly chlorin-ated PCB congeners to low chlorinchlorin-ated PCB congeners, which are more volatile into the atmolysphere. Another possible reason for the reduction could be the UV light degradation, caused by the light from sun and lamps.

4.1.6. Other impact factors

(30)

18

4.2. Risk assessment for initial and remediated soils

Fig. 5. Comparison of total PCB concentration among the initial soils, remediated soils, the unplanted soils and the site-specific

guideline

Based on the assumption that 7PCB is 20% of the total concentration of PCBs in the guideline model, the concentrations of 7PCB were convert-ed to the total concentration of PCBs. The generic guideline value of to-tal PCBs for sensitive land use is 0.04 mg/kg (Naturvårdsverket, 2009), while the site-specific guideline value was obtained through the model NV Soil Guideline as 0.035 mg/kg. This is slightly lower than the generic value, because the PCBs accumulate easily in this contaminated site. On the one hand, the PCBs in the sealant materials of the buildings keep leaching into the soils because of the natural weathering (Priha et al., 2005). On the other hand, PCBs have a high affinity for soils and are lip-ophilic (Smith et al., 2007). For this reason it is almost impossible for PCBs to be transported by groundwater or surface runoff.

From previous research about soil contamination from PCB-containing buildings, the highest PCB concentrations were in areas closest to the buildings, and they declined as the distance increased (Herrick et al., 2007). Priha et al. (2005) sampled soil around 11 buildings used PCB-containing caulking and found that the total PCB concentrations in soil ranged from 0.11 to 26.9 mg/kg. In this study, the soil samples were col-lected one meter from the buildings. The concentration of total PCBs in the study area was about 0.07 mg/kg (Table 1), which is lower than that in previous research.

However, compared to the calculated site-specific guideline, the initial concentrations of PCBs in all soil samples were almost twice the guide-line values. As for the remediated soils, most of them were still higher than the site-specific guideline as well, except that the horseradish plant-ed soil A was below the site-specific guideline values (Figure 5). But in soil C, the horseradish did not perform as well as that in soil A. Further studies on the use of horseradish to remove PCBs are still needed. This implies that the remediated soils still pose a threat to the health of resi-dents.

(31)

19

people’s health. Moreover, due to the continuous leaching from the con-struction, the concentration of PCBs will keep increasing in the future if no efficient remediation methods are implemented. The biological deg-radation and volatilization of PCBs are limited and quite slow processes, especially for highly chlorinated congeners. Therefore, the remediation of PCB-contaminated soil in this site is necessary and urgent. Although the study results from the greenhouse cultivation seem not as good as expected, phytoremediation did have a positive effect to remove PCBs from contaminated soils. If the cultivation time could be longer, the con-sequences might be better.

5.

RECOMMENDATIONS

5.1. Analysis of microorganisms and plants

Due to time and cost limitations, this study only analyzed the variation of PCBs in soil samples and it is still difficult to identify what caused the changes in the PCB concentration. Generally, these variation were decid-ed by microorganisms and plants together. As provdecid-ed by previous stud-ies, there are large numbers and diversity of PCB-degrading bacteria found in contaminated soils (Luo et al., 2007). Both anaerobic and aero-bic bacteria observed can degrade PCBs in soils (Xu et al., 2010). Anaer-obic bacteria may partially dehalogenate the highly chlorinated congeners, while some aerobic bacteria are capable of oxidatively attacking the ring systems of moderately chlorinated congeners (Rein et al., 2007). There-fore, analysis of bacterial species and counts could help us understand the process of phytoremediation better.

In addition, the PCBs in the roots and shoots of plants were not meas-ured in this study. Such data would be helpful to analyze the specific processes that happened in the plants. If the PCBs accumulated in the plants, the disposal of these toxic plants will be another serious problem. Furthermore, the number and amount of soil samples could have been higher. Each pot only had 50 g soil with 30 seeds. After three-months of growth, most plants still did not have plenty of roots and shoots, and even some of them didn’t germinate. If more soil samples and longer greenhouse cultivation had been provided for, the removal efficiency might perhaps have been better.

5.2. Analysis of infiltrated water from pots

During the greenhouse experiment, the infiltrated water from pots were collected by small bottles once a month. All the bottles were marked and stored in a freezer. However, they were not sent for PCB analysis be-cause of the limited budget. As for further studies, it could be necessary to analyze the PCB content in the infiltration water from pots. Although PCBs are hydrophobic, the slight difference in the PCB concentrations between the planted pots and planted pots could help us to identify if the plants can stabilize the PCBs in soils. If so, it would prevent the spread of contamination and hence provide for groundwater protection.

5.3. TRIAD method for risk assessment

(32)

evi-Jieyuan Wang TRITA LWR Degree Project 02:11

20

dence (LoE) including chemistry, toxicology and ecology (Jensen & Mesman, 2006).

6.

CONCLUSION

The effects of different plant species on the PCB removal of a long-term contaminated soil were evaluated in a greenhouse feasibility study. Com-pared to the unplanted control pots, the selected plants (alfalfa, tobacco, white clover and horseradish) did enhance the removal of PCBs in the contaminated soils. The horseradish and the white clover had the highest removal efficiency among the plant species tested in a greenhouse-scale cultivation. In detail, among all the seven tested PCB congeners, the re-moval of the highly chlorinated PCB congeners (PCB-138, PCB-153 and PCB-180) was higher than that of low chlorinated PCB congeners (PCB-28, PCB-52, PCB-101 and PCB-118). In addition, the site-specific guide-line value (0.035 mg/kg) of 7PCB in the study area was calculated by the Swedish site-specific guideline model. This showed that the initial and remediated concentration of 7PCB were all higher than the calculated guideline value, which means there is still a health risk for the residents living in that area.

However, the concentrations of the tested PCB congeners are quite low and they are below or just close to the detection limit. This caused a high analytical error. Therefore, there was no significant decrease of PCBs in this experiment. The study, however, suggested that horseradish and white clover can be focused on for further studies. Additionally, the op-posite results obtained in the experiment indicate again that phytoreme-dation is a complex process and that not only the plant species, but also the soil properties can have a crucial impact on the removal of PCBs. Therefore it is reasonable to assume that the same plant species per-formed differently in different soil samples.

(33)

21

R

EFERENCES

About education, 2016. Polychlorinated Biphenyl (PCB) Chemical Structure.

[Online] Available at:

http://chemistry.about.com/od/factsstructures/ig/Chemical-Structures---P/Polychlorinated-Biphenyl---PCB.htm

[Accessed 26 6 2016].

Aken, B. V., Correa, P. A. & Schnoor, J. L., 2010. Phytoremediation of

polychlorinated biphenyls: New trends and promises. Environmental

Science and Technology, pp. 2767-2776.

Alkorta, I. & Garbisu, C., 2001. Phytoremediation of organic contaminants in soils. Bioresource Technology, 3 1, pp. 273-276.

Åslund, M. L. W., Zeeb, B. A., Rutter, A. & Reimer, K. J., 2007. In situ phytoremediation of polychlorinated biphenyl (PCB) contaminated soil.

Science of the Total Environment, 25 1, pp. 1-12.

Balfanz, E., Fuchs, J. & Kieper, H., 1993. Sampling and analysis of polychlorinated biphenyls (PCB) in indoor air due to permanently elastic sealants. Chemosphere, 26(5), pp. 871-880.

Bunyan, P. J. & Page, J. M. J., 1978. Polychlorinated biphenyls: The effects of Structure on the Induction of Quail Hepatic Microsomal Enzymes. Toxicity and Applied Pharmacology, pp. 507-518.

Burken, J. G. & Schnoor, J. L., 1998. Predictive relationships for uptake of organic contaminants by hybrid poplar trees. Environmental Science and Technology, pp. 3379-3385.

Corner, R. Sundahl M., Ek-Olausson B., Tysklind M., 2002. PCB indoor air and dust in building in Stockholm. Indoor Air , pp. 141-146.

Eapen, S., Singh, S. & D'Souza, S. F., 2007. Advances in development of transgenic plants for remediation of xenobiotic pollutants. Biotechnology Advances, pp. 442-451.

Herrick, R. F., Lefkowitz, D. J. & Weymouth, G. A., 2007. Soil

contamination from PCB-containing buildings. Environmental Health

Perspectives, Volume 115, pp. 173-174.

Herrick, R. F. et al., 2004. An unrecognized source of PCB

contamination in schools and other buildings. Environmental Health

Perspectives, 112(10), pp. 1051-1053.

Iwata, Y. & Gunther, F., 1976. Translocation of the polychlorinated biphenyl Arochlor 1254 from soil into carrots under field conditions.

Archives of Environmental Contamination and Toxicology, pp. 44-59.

Jensen, J. & Mesman, M., 2006. Ecological risk assessment of contaminated land- Decision support for site specific investigations, The Netherlands: National

Institute for Public Health and the Environmnet .

Jerald L. Schnoor, L. A. L. S. C. M. N. L. W. L. H. C., 1995.

Phytoremediation of organic and nutrient contaminants. Environmental

Science and Technology, pp. 318-323.

Jones, C. et al., 2006. Improved Ecological Risk Assessments, Sweden: Swedish

Environmental Protection Agency.

Kohler, M. et al., 2005. Joint sealants: An overlooked diffuse source of polychlorinated biphenyls in buildings. Environmental Science and Technology,

pp. 1967-1973.

(34)

22

Luo, W., D'Angelo, E. M. & Coyne, M. S., 2007. Plant secondary metabolites, biphenyl, and hydroxypropyl-β-cyclodextrin effects on aerobic polychlorinated biphenyl removal and microbial community structure in soils. Soil Biology & Biochemistry, pp. 735-743.

M., F. G. et al., 1996. Comprehensive, quantitative, congener-specific analyses of eight Arochlors and complete PCB congener assignments on DB-1 capillary GC columns. Chemosphere, pp. 603-623.

Mackova, M. et al., 2006. Phytoremediation of Polychlornated Biphenyls.

Phytoremediation Rhizoremediation, pp. 143-167.

Mackova, M. et al., 2009. Phyto/rhizoremediation studies using long-term PCB-contaminated soil. Environmental Science Pollution Reserch, 13 10,

pp. 817-829.

McFarland, V. A. & Clarke, J. U., 1989. Environmental occurrence, abundance, and potential toxicity of polychlorinated biphenyl congeners: Considerations for a congener-specific analysis. Environmental Health Perspectives, Volume 81, pp. 225-239.

Naturvårdsverket, 2009. Riskbedömning av förorenade områden (Risk assessment for contaminated sites), Sweden: Swedish Environmental Protection Agency.

Panagos, P., Liedekerke, M. V., Yigini, Y. & Montanarella, L., 2013. Contaminated sites in Europe: Review of the current situation based on data collected through a European network. Journal of environmental and public health, 23 5.

Pilon-Smits, E., 2005. Phytoremediation: Annual review of plant biology,

Colorado State University: Biology Department.

Plevrakis, V., 2014. Comparison of risk assessment methods for polluted soils in Sweden, Norway and Danmark, Stockholm University: Department of

Physical Geography and Quaternary Geology.

Priha, E., Hellman, S. & Sorvari, J., 2005. PCB contamination from polysulphide sealants in residential areas—exposure and risk assessment.

Chemosphere, pp. 537-543.

Rein, A. et al., 2007. Degradation of PCB congeners by bacterial strains.

Appl Microbiol Biotechnol, pp. 469-481.

Ribé, V. et al., 2012. Applying the triad method in a risk assessment of a former surface treatment. Journal of Hazardous Materials, pp. 207-208.

Ryslava, E. et al., 2003. Study of PCB degradation in real contaminated soil. Fresenius Environ Bull, 12, pp. 296-301.

Ryslava, E. et al., 2003. Study of PCB degradation in real contaminated soil. Fresenius Environmental Bulletin, pp. 296-301.

Saba, E., 2012. Detection and Quantification of PCB in soil using GC/MS- method development and education for users, Stockholm: KTH Royal Institute

of technology.

Safe, S., Safe, L. & Mullin, M., 1985. Polychlorinated biphenyls: congener-specific analysis of a commercial mixture and a human milk extract. Journal of agricultural and food chemistry, 33(1), pp. 24-29.

Sawhney, B. L. & Hankin, L., 1984. Plant contamination by PCBs from amended soils. Journal of Food Protection, pp. 232-236.

Schnoor, J. L. et al., 1995. Phytoremediation of organic and nutrient contaminants. Environmental Science and Technology, pp. 318-323.

Schwarzenbach, R. P., Gschwend, P. M. & Imboden, D. M., 2005.

Environmental Organic Chemistry. 2nd edition ed. New Jersey:

(35)

23

Science Communication Unit , 2013. Soil contamination: Impacts on human health, s.l.: European Comission's Directorate General Environment.

Shayler, H., McBride, M. & Harrison, E., 2009. Sources and impacts of contaminants in soils, College of agriculture and lisfe science of Cornel

University: Department of crops and soil sciences.

Smith, K. E., Schwab, A. P. & Banks, M. K., 2007. Phytoremediation of Polychlorinated Biphenyl (PCB)-contaminated Sediment: A Greenhouse Feasibility Study. Journal of Environmental Quality, 9 1, pp. 239-244.

Suzuki, M., Aizawa, N., Okano, G. & Takahashi, T., 1977. Translocation of polychlorobiphenyls in soil into plants: A study by a method of culture of soybean sprouts. Archives of Environmental Contamination and Toxicology, pp. 343-352.

Swedish Standards Institute, 2012. Sludge, treated biowaste and soil- Determination of polychlorinated biphenyls (PCBs) by gas chromatography with mass selective detection (GC-MS) and gas chromatography with electron-capture detection (GC-ECD), Sweden: SSI Standards Limited .

Technologywater, 2014. Types of Soil Remediation Techniques and Methods.

[Online]

Available at:

http://www.technologywater.com/post/104761516702/types-of-soil-remediation-techniques-and-methods [Accessed 7 6 2016].

US Environmental Protection Agency, 1997. Ecological risk assessment

guidance for superfund: Process for designing and conducting ecological risk assessments- Interim Final, United States Environmental Protection Agency:

Solid Waste and Emergency Response.

Vasilyeva, G. et al., 2010. Dynamics of PCB removal and detoxification in historically contaminated soils amended with activated carbon.

Environmental Pollution, pp. 770-777.

Xu, L. et al., 2010. Enhanced removal of polychlorinated biphenyls from alfalfa rhizosphere soil in a field study: The impact of a rhizobial inoculum. Science of the Total Environment, pp. 1007-1013.

(36)

Phytoremediation of long-term polychlorinated biphenyls (PCBs)-contaminated soil: A green-house feasibility study

A

PPENDIX

I

–

RESULTS FROM THE CHEMICAL ANALYSIS

Sample No. Soil Description Water Content

(%) Organic Matter (%) pH - Initial soil A / 3.56 6.05 - Initial soil B / 9.73 5.28 - Initial soil C / 3.25 5.96 1 Alfalfa 25.38 3.23 6.53 2 Alfalfa 23.57 4.86 6.25 3 Unplanted 26.09 3.20 6.25

4 Tobacco Not germinate

5 White clover 22.68 3.47 6.15

6 Horseradish 19.74 2.91 6.45

7 Horseradish Plant died

8 Unplanted 21.92 3.31 6.39

9 White clover Not germinate

11 Alfalfa 39.73 8.64 6.27 12 Alfalfa 34.35 7.65 6.11 13 Tobacco 36.22 10.02 5.67 14 Tobacco 38.74 9.13 5.95 15 Unplanted 38.63 9.37 6.04 16 Unplanted 39.18 9.30 6.00 17 Alfalfa 19.12 3.09 6.23 18 Alfalfa 25.83 3.24 6.08

(37)

(38)

III

(39)

IV

(40)

V

(41)

TRITA LWR Degree Project ISSN 1651-064x