Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Application of semipermeable membrane devices (SPMDs) to the monitoring of kraft mill effluents with… Rohr, Annette Christine 1994

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_1994-0124.pdf [ 3.94MB ]
JSON: 831-1.0050402.json
JSON-LD: 831-1.0050402-ld.json
RDF/XML (Pretty): 831-1.0050402-rdf.xml
RDF/JSON: 831-1.0050402-rdf.json
Turtle: 831-1.0050402-turtle.txt
N-Triples: 831-1.0050402-rdf-ntriples.txt
Original Record: 831-1.0050402-source.json
Full Text

Full Text

APPLICATION OF SEMIPERMEABLE MEMBRANE DEVICES (SPMDs) TO THEMONITORING OF KRAFT MILL EFFLUENTS WITH EMPHASIS ONPOTENTIAL FISH-TAINTING COMPOUNDSbyANNETTE CHRISTINE ROHRB.Sc. (Microbiology), The University of British Columbia, 1991A THESIS SUBMITTED IN PARTIAL FULFILMENT OFTHE REQUIREMENTS FOR THE DEGREE OFMASTER OF APPLIED SCIENCEinTHE FACULTY OF GRADUATE STUDIES(Department of Civil Engineering)We accept this thesis as conformingto the required standardTHE UNIVERSITY OF BRITISH COLUMBIAJanuary, 1994©Annette Christine Rohr, 1994In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission._________________________Department of ‘The University of British ColumbiaVancouver, CanadaDate /8, /‘?DE-6 (2/88)ABSTRACTThe use of Semipermeable Membrane Devices (SPMDs) for monitoring pulp milleffluents containing potential fish tainting compounds was investigated. SPMDs are lipid-filled polyethylene bags which appear to mimic bioconcentration phenomena in aquaticorganisms. They have been used as in situ, passive monitors of organic contaminants suchas polynuclear aromatic hydrocarbons (PARs) in aqueous environments, and were used herein a pulp mill setting.Preliminary method development work included investigating the effect of tubingsegment location within the roll of polyethylene on uptake of 2,2’ ,5 ,5 ‘-tetrachlorobiphenyl(TCB), a reference compound. Tubing segment location was determined to have nosignificant effect on degree of uptake. In addition, polyethylene consistency was investigatedby subjecting tubing to various batch solvent extraction conditions. Selected tubing extractswere also analyzed by gas chromatography/mass spectrometry (GCIMS), and thecontaminating compounds were identified. Hexane exposures showed that a 24 hour hexaneextraction with no solvent replacement was sufficient for the adequate removal ofcontaminating compounds associated with the polyethylene tubing.Three model compounds were chosen on the basis of their significant presence inkraft mill effluent as well as their potential propensity to taint fish. Octanol-water partitioncoefficients (K0s) were determined for the compounds, alpha-pinene, guaiacol anddehydroabietic acid, and for 2,2’ ,5 ,5 ‘-TCB. Compound uptake into SPMDs and dialyticrecoveries from SPMDs were determined. Generally, the model compounds were not11sequestered efficiently by SPMDs, and their dialytic recoveries were low. The behaviourof the compounds was explained on the basis of their K,s, polarities, volatilities and watersolubilities.SPMDs were exposed to both untreated and biotreated full-strength unbleached kraftmill wastewater under static conditions. In addition, SPMD exposures were carried outusing several concentrations of treated effluent in continuous flow systems. In both typesof exposures, SPMDs were found to sequester many compounds, as evidenced by gaschromatograms. Of particular interest was the uptake of compounds which may beresponsible for the tainting of eulachons (Thaleichthys pacWcus) in the Kitimat River, B.C.Eulachons are small smelts with a high lipid content which traditionally have been consumedby the Haisla Indians living in the area, but have not been used in this capacity since theearly 1970s due to a noticeable off-flavour. Five compounds identified in tainted eulachonwere identified in SPMD dialysates after GC/MS analysis.Based on the results shown in this research, SPMDs appear to show significantpromise as passive monitors for the detection and possible quantification of taintingcompounds in aqueous environments. Their specificity for nonpolar compounds maydiminish their usefulness in pulp mill settings, where ionic and polar compounds such asresin acids and phenols are common constituents of wastewaters.111TABLE OF CONTENTSpageABSTRACT. .iiTABLE OF CONTENTS. ivLIST OF TABLES. viiLIST OF FIGURESACKNOWLEDGEMENTS1. INTRODUCTION1.1 Motivation of Research1.2 Thesis Organization2. BACKGROUND AND LITERATURE REVIEW..2.1 Introduction2.2 Environmental Monitoring Methods2.2.1 Water Samples2.2.2 Aquatic Organism Tissue Samples2.2.3 Solvent-filled Membranes2.2.3.1 Solvent-filled Dialysis Membranes2.2.3.2 Solvent-filled Polyethylene Membranes2.3 SPMDs as Environmental Monitors2.3.1 Composition of SPMDs2.3.2 Preparation and Deployment of SPMDs2.3.3 Dialytic Cleanup of SPMDs2.3.4 SPMDs as Biological Surrogates2.3.5 Advantages of SPMDs2.3.6 Membrane Transport Theory2.3.6.1 Mechanism of Transport2.3.6.2 Factors Affecting Membrane Transport2.3.7 Equilibrium Partitioning2.3.7.1 Octanol-Water Partition Coefficients2.3.7.2 Lipid-Water Partition Coefficients2.3.8 Comparison of Bioconcentration and Membrane Transport2.3.9 Applications of SPMDs2.4 Overview of Kraft Mill Wastewaters 272.4.1 The Kraft Pulping Process 272.4.2 Chemical Composition of Kraft Effluents 2711255566711121617• • 1718• . 22• 22• 23• . 2526iv4.2 SPMD Methodology.4.2.1 Preparation of SPMDs4.2.2 Cleanup and Dialysis of SPMDs4.3 Analytical Techniques4.3.1 Gas Chromatography4.3.2 Mass Spectrometry4.3.3 High Pressure Liquid Chromatography4.3.4 Resin Acid Analysis4.3.5 Aqueous Sample Extraction4.3.6 Internal Standards4.4 Wastewater Exposures4.4.1 Static Wastewater Exposures4.4.2 Continuous Flow Wastewater Exposures3131323235• . 363639• . 43• • 4444444444• • 4547• . 475. RESULTS AN]) DISCUSSION 575.1 Introduction 575.2 Method Development5.2.1 Verification of Hexane and Triolein Purity.5.2.2 Verification of Consistent Tubing Permeability5.2.3 Effect of Batch Extraction Conditions on Tubing312.5 The Fish Tainting Issue2.5.1 Overview of Kraft Mill Effluent-Induced Tainting2.5.1.1 Introduction2.5.1.2 Mechanism of Tainting2.5.1.3 Identification of Tainting Agents . Sources and Removal of Tainting Agents2.5.2 Eulachon Off-flavour at Kitimat, B.C2.5.2.1 Introduction2.5.2.2 Recent Tainting Evaluations3. RESEARCH OBJECTIVES4. MATERIALS AND EXPERIMENTAL METHODS4.1 Materials4.1.1 SPMD Constituents4.1.2 Model Compounds4.1.2.1 Selection of Model Compounds4.1.3 2,2’ ,5,5’-Tetrachlorobiphenyl4.1.4 Solvents and Standard Solutions .4.1.5 Other Materials4747484949505052525253535557575863v5.2.3.1 Effect of Exposure Time on Tubing Consistency . 635.2.3.2 Effect of Solvent Replacement on Tubing Extractives 665.2.4 Identification of Crystalline Material in Solvent Extracts . 685.2.5 Characterization of Glove Extracts 695.2.6 Discussion of Method Development Work . . . 705.3 Model Compound Work 735.3.1 Determination of K0s by HPLC . . 735.3.2 Determination of Dialytic Recovery 775.3.3 Determination of Total Uptake . 865.3.4 Discussion of Model Compound Work . 1015.4 Wastewater Exposure Work 1065.4.1 Static Wastewater Exposures 1065.4.1.1 Introduction and Goals of Work . 1065.4.1.2 Uptake of Pulp Mill Effluent Contaminants into SPMDs 1075.4.1.3 Effect of Wastewater Treatment on SPMD Uptake 1105.4.1.4 Effect of Wastewater Aeration on SPMD Uptake . . 1155.4.1.5 Effect of Exposure Time on SPMD Uptake 1185.4.1.6 Fish Tainting Significance 1235.4.2 Continuous Flow Wastewater Exposures 1315.4.2.1 Introduction and Goals of Work 1315.4.2.2 Compound Uptake in a Flow-through System 1325.4.2.3 Effect of Wastewater Concentration on SPMD Uptake 1375.4.2.4 Effect of Exposure Time on SPMD Uptake 1395.4.2.5 Fish Tainting Significance 1425.4.3 Discussion of Wastewater Exposure Work 1496. CONCLUSIONS AND RECOMMENDATIONS 1556.1 Conclusions 1556.2 Recommendations 157LITERATURE CITED 159APPENDICES 167Appendix A: Chromatograms from GC and GC/MS Analyses 168Appendix B: Raw Experimental Data 172Appendix C: Compounds Identified in Sample E-NA-9-1 185Appendix D: Temperature Programs Used 187viLIST OF TABLESpageTable 1. Comparison of Experimental Partition Values for 2,2’,5,5’-TCB withLiterature Values 24Table 2. Acidic Compounds in Unbleached Kraft Mill Wastewater 28Table 3. Neutral Compounds in Biologically Treated Kraft Mill Wastewater 29Table 4. Terpenes in Kraft Mill Aerated Lagoons 30Table 5. Tentative GC/MS Identification of 3 GC Peaks of Tainted Eulachon Flesh 42Table 6. Selected Physico-Chemical Properties of Model Compounds and TCB . . 46Table 7. Ultraviolet Maxima for Reference and Unknown Compounds 51Table 8. Experimental Design for Static Wastewater Exposures 54Table 9. Experimental Design for Continuous Flow Wastewater Exposures 55Table 10. Compounds Identified in Triolein 58Table 11. Extractives from Hexane Extraction of Sample E24 68Table 12. Compounds Identified in Crystalline Material 69Table 13. Compounds Identified in Polyethylene Glove Extracts 70Table 14. Pertinent Data for K0 Determination: Reference Compounds 74Table 15. Pertinent Data for K0 Determination: TCB and Model Compounds . . . . 74Table 16. Compounds Identified in Dialysis Laboratory Blanks 84Table 17. Partial List of Compounds Identified in Sample E-NA-9-1 109Table 18. Potential Tainting Agents Identified in SPMD Static Kraft Effluent ExposureExtracts 123Table 19. Compounds Identified in Dialysates of SPMD Exposed to 100% TreatedEffluent for 9 Days (Sample #1) 134viiTable 20. Compounds Identified in Sample #2, not Sample #1 135Table 21. Potential Tainting Agents Identified in SPMD Continuous Flow WastewaterExposure Extracts 142Table 22. Chemical Analysis of Eurocan Effluent Samples Collected Between March 22and April 1, 1993 151Table B-i. Polyethylene Permeability Data 172Table B-2. Dialytic Recovery Data 173Table B-3. Rinsate Analysis Data 173Table B-4. Total Uptake Data, Including Non-Conventional SPMDs 174Table B-5. Distribution of Test Compounds in SPMDs and Aqueous Phases 175Table B-6. Static Wastewater Exposures: Effect of Treatment (1 day, NA) 176Table B-7. Static Wastewater Exposures: Effect of Treatment (9 day, A) 177Table B-8. Static Wastewater Exposures: Effect of Aeration 178Table B-9. Static Wastewater Exposures: Effect of Exposure Time (E/NA) 179Table B-b Static Wastewater Exposures: Effect of Exposure Time (I/A) 180Table B-li Static Wastewater Exposures: Uptake of Potential Tainting Agents . . . . 181Table B-12 Continuous Flow Wastewater Exposures: Effect of Effluent Concentration 182Table B-13 Continuous Flow Wastewater Exposures: Effect of Exposure Time . . . . 183Table B-14 Continuous Flow Wastewater Exposures: Uptake of Potential TaintingAgents 184viiiLIST OF FIGURESFigure 1.Figure 2.Figure 3.Figure 4.Figure 6.Figure 7.Figure 8.Figure 9.Figure 10.Figure 11.Figure 12.Figure 13.Figure 14.Figure 15.page101315333746545662• . . . 65• . . . 67• . . . 75• . . . 79• . . . 83878889949899100108111Photographs of Prepared SPMDsComparison of Dioxin and Furan Uptake into Clams and SPMDsComparison of PAR Uptake into Mussels and SPMDsGas Chromatograms of Strongly Tainted and Mildly Tainted Salmon FleshFigure 5. Map of Area Surrounding Eurocan Pulp and Paper Co.Structures of the Model Compounds and TCBPhotograph of Static Wastewater Exposure SetupDilution Setup for Continuous Flow Exposures2,2’,S,S’-TCB Uptake into SPMDs vs. Tubing Segment LocationEffect of Exposure Time on Tubing ConsistencyEffect of Solvent Replacement on Tubing ConsistencyK0 Determination by HPLC (Reference Compounds)Dialytic Recovery of Model Compounds and TCBDistribution of Test Compounds in Dialysate and Rinsate(a) Uptake of 2,2’,S,S’-TCB into SPMDs(b) Uptake of Alpha-Pinene into SPMDs(c) Uptake of Guaiacol into SPMDsTest Compound Uptake into Non-Conventional SPMDs(a) Distribution of 2,2’,5,5’-TCB(b) Distribution of Aipha-Pinene(c) Distribution of GuaiacolGas Chromatograms of Blank and Exposed SPMD DialysatesGas Chromatograms of Samples E-A-9 and I-A-9FigureFigureFigureFigure16.17.18.19.ixFigure 20.Figure 21.Figure 22.Figure 23.Figure 24.Figure 25.Figure 26.Figure 27.Figure 28.Figure 29.Figure 30.Figure 31.Figure 32.Figure 33.Figure 34.Figure 35.Figure 36.Figure 37.Figure A- 1.Figure A-2.Figure A-3.Figure A-4.112114116117119120122124125126127128138140141144145146147148168169170171Effect of Wastewater Treatment on SPMD Uptake (Samples ElI-NA-i).Effect of Wastewater Treatment on SPMD Uptake (Samples E/I-A-9)(Part 1) Effect of Wastewater Aeration on SPMD Uptake(Part2)Effect of Exposure Time on SPMD Uptake (Samples E-NA-1/3/9) .Effect of Exposure Time on SPMD Uptake (Samples I-A-1/3/9)Effect of Exposure Time on SPMD Uptake (Samples I-A-1/3/9)Uptake of 2-methyl-5-(1-methylethyl)bicyclo[3.1 .0]hex-2-eneUptake of 4-methyl- 1-(1-methylethyl)bicyclo[3. 1. 0]hexaneUptake of alpha-pineneUptake of 1-methyl-4-(1-methylethyl)cyclohexaneUptake of (z)-3 ,7-dimethyl- 1,3, 6-octatrieneEffect of Effluent Concentration on SPMD Uptake(Part 1) Effect of Exposure Time on SPMD Uptake (Cont.Flow @100%)(Part2)Uptake of 2-methyl-5-(l-methylethyl)bicyclo[3. 1 .0]hex-2-eneUptake of 4-methyl- 1-(1.-methylethyl)bicyclo[3. 1 .0]hexaneUptake of aipha-pineneUptake of l-methyl-4-(1-methylethyl)cyclohexaneUptake of (z)-3 ,7-dimethyl- 1,3, 6-octatrieneGCIMS Total Ion Count for HexaneGCIMS Total Ion Count for Neat Triolein Dissolved in Hexane . . .Gas Chromatograms of Polyethylene Glove ExtractsGCIMS Total Ion Count for SPMD Blank DialysatexACKNOWLEDGEMENTSThanks are extended to my supervisor, Dr. Eric Hall, and to my co-supervisor, Dr.Ken Hall, for their guidance in this work. I would also like to thank Paula Parkinson, SusanHarper and Timothy Ma for their technical assistance.I wish to thank Brian Stevens and Mike Martins of Eurocan Pulp and Paper Co.,Kitimat, B.C. for their assistance with SPMD deployment.Finally, I am grateful to my family and friends, especially my parents, for their moralsupport during the course of this work.This research was funded by an NSERC Postgraduate Scholarship. Additionalfunding was provided by the Natural Sciences and Engineering Research Council of Canada,the Council of Forest Industries of B.C., the B.C. Ministry of the Environment and theUniversity of British Columbia, through the NSERC/COFI Industrial Chair in ForestProducts Waste Management.xi1. INTRODUCTION1.1 Motivation of ResearchThe purpose of this research was two-fold. First, the application of semipermeablemembrane devices (SPMDs) to pulp mill effluent was to be investigated. Althoughconsiderable research had been carried out using SPMDs to sequester polynuclear aromatichydrocarbons (PARs) and other environmental contaminants, as well as work comparingSPMD uptake with that of aquatic organisms such as mussels, none had yet been undertakento investigate SPMD application in a pulp mill setting. From regulatory, economic, labourintensity and reliability perspectives, SPMDs may prove valuable as in situ monitors ofcontaminants in kraft mill effluent; therefore, proving their scientific and technical feasibilityis important. Not only can they be placed in environments too hostile for living organisms,but also their cleanup and analysis is relatively straightforward. With further methoddevelopment, modelling could be carried out to determine the time-averaged concentrationsof kraft mill pollutants in receiving waters.The second motivation behind this work was the tainting of eulachons, also knownas “oolichans” or “candlefish”, in the Kitimat River near Kitimat, B.C. The eulachonsspawning downstream of the effluent outfall of Eurocan Pulp and Paper Co. have a taintedflavour that prevents their consumption by the native Haisla Indians, who traditionally usetheir flesh and grease as food sources.Although a separate study by Beak Consultants has tentatively identified somepotential tainting agents in tainted eulachon flesh, their identities are presently unconfirmed.SPMDs could prove useful in this situation by sequestering the tainting agents and thus1facilitating their identification. Ultimately, SPMDs could be placed in a variety of in-milllocations in an attempt to determine the source(s) of the tainting compound(s). In addition,by monitoring its effluent with SPMDs, Eurocan could obtain an integrated measure of thetainting propensity of its effluent.1.2 Thesis OrganizationChapter 2 gives the background pertaining to environmental monitoring, SPMD use,the composition of kraft mill effluent and fish tainting. Chapter 3 outlines the objectives ofthe research. Chapter 4 describes the experimental techniques and materials used in thisresearch. In Chapter 5, the results obtained from the experiments performed are presentedand discussed. Finally, Chapter 6 gives conclusions and recommendations for further study.The experimental data are presented in the Appendices. Appendix A containschromatograms from gas chromatographic (GC) and gas chromatographic/mass spectrometric(GC/MS) analyses. Appendix B gives raw experimental data from GC chromatograms.Appendix C contains a compound list from GC/MS analysis of a wastewater exposuresample. Finally, Appendix D shows the temperature programs used in the variousexperiments.22. BACKGROUND AND LITERATURE REVIEW2.1 IntroductionMonitoring of environmental contamination is currently an area of great interestand research activity. Due to increasingly stringent regulations regarding the dischargeof waste materials, the need for more efficient monitoring methods is evident. Pulp milleffluents, in particular, are the objects of much scrutiny.Intense consideration of pulp mill effluents is warranted. Aquatic toxicity ofeffluents, both acute and sublethal, has been reported in numerous instances in theliterature (reviews by McLeay, 1987; Owens, 1991). In addition, flavour impairment offish has resulted from exposure to pulp mill effluents (Shumway and Chadwick, 1971;Cook et al., 1973; Whittle and Flood, 1977; Heil et al., 1989). Much of the interest inthis area concerns kraft mill effluents, because of the prevalence of the kraft pulpingprocess.A monitoring system applicable to pulp mill contaminants would be beneficial formonitoring both toxic compounds and potential off-flavour agents. Such a monitor wouldbe required to accurately mimic the uptake of compounds into aquatic organisms and besimple to use. In particular, a monitoring technique is required by Eurocan Pulp andPaper Co. in Kitimat, B.C., where eulachons exposed to kraft effluent from the mill havenoticeable flavour impairment (Beak Consultants, 1991; 1992; 1993b).The Semipermeable Polymeric Membrane Device (SPMD) could potentially serveas a monitor of this type. Recently developed by research chemists at the NationalFisheries Contaminant Research Center in Columbia, Missouri (Huckins et al., 1990a),3SPMDs show considerable promise as passive, in situ samplers for aqueous organiccontaminants. They consist of a model lipid enclosed in a polymeric membrane, andappear to mimic bioconcentration phenomena occurring in aquatic organisms (Prest et a!.,1992; Lebo et a!., 1992; Huckins et a!., in press; Shigenaka and Henry, in press).Already they have been utilized in a variety of environmental settings, and work hasbegun to develop mathematical models for the determination of time-weighted aqueousconcentrations from levels detected in SPMDs (Huckins et a!., in press). Due to theirusefulness as integrative monitors, a U.S. patent is pending, and a commercial laboratorywill likely begin marketing SPMDs in the near future (J.Lebo, personal communication,1993).The following section will discuss alternate environmental monitoring methods, aswell as the composition, preparation, deployment, cleanup, analysis and advantages ofSPMDs. The theory of membrane transport and equilibrium partitioning will then becovered. Recent and future applications of SPMDs will be mentioned. A brief overviewof the composition of kraft mill wastewaters will be presented. Finally, the issue of fishtainting will be discussed, with specific mention of the eulachon tainting problem atKitimat, B.C.42.2 Envfronmental Monitoring Methods2.2.1 Water SamplesWater samples can be directly analyzed, but this tends to be impractical,especially if the contaminant is present only at trace concentrations. The difficultiesinherent in the collection and extraction of large volumes of water, the loss of analytes byvolatilization and sorption, and the possibility of changing water quality or characteristicsdue to handling make this logistically impractical. In addition, the resultingmeasurements only represent the analytes present at the time of sample collection, not theanalytes present over a longer period of time. Therefore, contamination events whichonly occur episodically or sporadically may go undetected.2.2.2 Aquatic Organism Tissue SamplesFish or other aquatic organisms, either caged or from natural populations, mayalso be used as monitors of aqueous contamination. Bivalves such as mussels, oystersand clams, have been used in this capacity in Northern Europe and the United Statessince the late 1960s, and are being adopted by an ever-increasing number of nations(National Academy of Sciences, 1980). Most recently, the California State MusselWatch Program has been implemented (Prest et al., 1992).However, like water sampling, the tissue analysis approach is problematic.Background contaminant levels in the organism may erroneously be attributed to theexposure site. Also, uptake by biota is by no means constant. Organism-related factorssuch as age, size, sex, dietary uptake, lipid composition and metabolic activity, as well as5environmental factors such as pH, temperature, suspended solids and dissolved oxygen,can affect bioconcentration processes. Many compounds are biologically degraded ormodified, so there may be no detectable levels of the original targeted compound in thetissue. Fish and other organisms cannot survive in hostile or toxic environments, makingtheir use as biomonitors somewhat limited. Organisms such as clams and mussels areunable to differentiate between dissolved and particulate material, since they filter andingest particles. Therefore, biomagnification processes may obscure truebioconcentration phenomena. Complex biological tissues are also difficult to analyze andrequire complicated cleanup steps.2.2.3 Solvent-filled Membranes2.2.3.1 Solvent-filled Dialysis MembranesSome researchers have utilized regenerated cellulose membranes enclosing anonpolar solvent (hexane) as environmental dosimeters (SOdergren, 1987; SOdergren,1990; Herve et al., 1991; Johnson, 1991). However, they do not appear to accuratelysimulate biological uptake. Nonpolar, lipophilic contaminants most likely to accumulatein organisms cannot easily traverse the membrane because of its hydrophilicity. In fact,when Herve et al. (1991) compared solvent-filled dialysis membranes with mussels, theyfound that the mussels accumulated chlorinated organic compounds much more effectivelythan the membrane surrogates. Similarly, although Sodergren (1987) reported aconcentration factor in solvent-filled dialysis membranes of approximately 250 forpolychiorinated biphenyls (PCB5), Bruggeman et al. (1981) found a bioconcentration6factor in goldfish ranging from 0.4 x 106 to 1.5 x 106 for the same compounds.Another study by Gray and Spacie (1991) compared the partitioning of twopesticides into SPMDs and hexane-fihled cellulose membranes with bioconcentrationfactors (BCFs) in fish. The study showed that the relative concentration factors forSPMDs were closer to the BCF for fish than the hexane-fihled cellulose membranes. Solvent-filled Polyethylene MembranesThe solvent can also be enclosed in polyethylene, rather than regenerated cellulose(Hassett et al., 1989). Such a device, given the acronym PISCES (passive, in situconcentration/extraction sampler), has a few significant problems. First, since thesolvents used are of low molecular weight, they tend to diffuse out of the bags into thesurrounding aqueous phase, thus preventing the contaminants from attaining trueequilibrium. Second, there is some question as to the realism of solvent as a simulator ofbiological lipid. There was a discrepancy between concentration factors of certaincompounds in the solvent and their octanol-water partition coefficients (Fish et al., 1989;Hassett et al., 1989; McEachren et al., 1991). Since the partition coefficient (K) of acompound describes its partitioning from an aqueous phase to an organic phase, thiswould imply that the PISCES may not be a realistic simulator.72.3 SPMDs as Environmental Monitors2.3.1 Composition of SPMI)sSemipermeable membrane devices consist of a thin film of a large molecularweight, neutral lipid enclosed by a low density, non-porous polymer such aspolyethylene, plasma-treated silicon or polypropylene. Usually the polymer is in theform of layflat tubing. The term “non-porous” does not mean completely impermeable;rather, there are transient channels of diameter 5-10 A (Huckins et a!., 1990a). Thesechannels are created by random thermal motions of polymer chains, and allow transportof molecules across the membrane. Lieb and Stein (1969) noted that the transport ofnon-electrolytes across synthetic membranes is similar to transport across much morecomplex biomembranes.The model lipid most often used in SPMDs is tnolein, or glycerol trioleate. Itrepresents the largest mass fraction of neutral lipids in freshwater fish (Henderson andTocher, 1987). It also has a low melting point (4.9°C) so it is a liquid at roomtemperature. In addition, it is commercially available in purified form. These attributes,combined with Lieb and Stein’s work, would suggest that, by virtue of their composition,SPMDs might realistically mimic biological uptake.2.3.2 Preparation and Deployment of SPMDsThe details of the procedure are presented in the Materials and Methods section ofthis thesis, but a brief synopsis is given here. The method developed by Huckins et a!.(1990a) consists of taking segments of polyethylene tubing from a roll and extracting8them for 24 hours in hexane, thereby removing any additives or contaminants which maybe present. A small amount (usually 1 mL) of triolein is then pipetted into the tubing andspread into a thin film down the length of tubing. Both ends are heat-sealed and then resealed together to produce a closed-loop configuration. Photographs of prepared SPMDsare shown in Figure 1.Certain compounds may be added to the triolein prior to its placement in thetubing. Because triolein is an unsaturated triglyceride, it may undergo oxidation over aperiod of time which could affect the behaviour of compounds entering the SPMD(Huckins et al., 1990a). Therefore, an anti-oxidant chemical may be added in smallamounts to the triolein to prevent such modification. In addition, biological colonizationof SPMDs deployed in the field may occur (“Aufwuchs colonies”) and hinder analyteuptake (Huckins et al., 1990a). It has been reported that biofouling of the membrane canreduce compound uptake by 20-30%, but that colonization only occurs to a significantdegree after about 7 days of exposure (B. Poulton, unpublished material, 1992). IfSPMDs are to be exposed for a period longer than this, the biofouling problem can becircumvented by the addition of low levels of pesticides to the triolein prior toincorporation into SPMDs, or to the polymer before extrusion into tubing. In addition,regular dipping of deployed SPMDs into a pesticide solution may reduce biologicalgrowth (Huckins et al., 1990a). Conceivably, the deposition of an ultra-thin (5-50 nm)film of plasma polymer on the polyethylene surface might also reduce biofouling, sincecolonization is difficult on an irregular surface (H. Yasuda, unpublished information,1992).9Figure1.PhotographsofPreparedSPMDs.10Storage and transport of SPMDs should ideally be in sealed, amber glass jarsunder argon gas, due to the propensity of SPMDs to sequester air-borne contaminants(Zajicek et al., unpublished information, 1992).2.3.3 Dialytic Cleanup of SPMDsTraditionally, organic analytes of relatively low molecular weight are separatedfrom lipid by size exclusion or gel permeation chromatography (Stalling et a!., 1972), atime-consuming and complicated procedure. With the recent development of a polymericfilm dialytic extraction method for lipid cleanup (Huckins et a!., 1990b), SPMD analysiscan be carried out with a minimum of time and effort. The technique involves placingSPMDs in a nonpolar hydrocarbon solvent such as hexane and allowing dialysis to occur.Contaminant molecules diffuse out of the lipid into the dialytic solvent, whereas thelarger, more polar lipid molecules remain within the SPMD. Residue analysis can thenbe carried out directly, although extremely complex samples may require additionalcleanup. This method has several advantages over gel permeation chromatography,including reduced solvent use, ability for scaling up or down to accommodate varyingsample masses, simplicity and low cost. Also, the technique could potentially beautomated. Preliminary testing of the method indicated good dialytic recoveries oforganochlorine compounds, ranging from 88-101% (Huckins et a!., 1990b).The dialytic cleanup procedure can also be used to separate contaminants from thetriolein prior to its incorporation into SPMDs if an extremely high degree of purity isdesired (Huckins et a!., 1990b). This is accomplished by adding triolein to the tubing,11usually a quantity sufficient to make the tubing fairly turgid, heat-sealing the ends andallowing dialysis to occur for a period of time (usually 24 hours). The triolein is thenremoved from the tubing and incorporated into SPMDs. In an extrapolation of the aboveidea, organic solvent dialysis can also be used to separate contaminants from other lipidsamples, particularly lipids originating from organisms, thereby simplifying the analysisof complex biological samples (Meadows et al., 1993).In their work on optimization of dialysis, Meadows et al. (1993) found that an80:20 mixture of hexane:dichloromethane produced the highest recoveries, closelyfollowed by hexane alone. They also concluded that dialysis was optimized at a surfacearea of 15 cm2 of polyethylene per mL of lipid, and at a dialytic solvent:lipid ratio of30:1. Daily replacement of the dialytic solvent produced higher recoveries, due to theincreased concentration gradient. Agitation of SPMDs during dialysis, as well asreducing dialytic temperature, appeared to have little effect on analyte recoveries.2.3.4 SPMDs as Biological SurrogatesIn comparison with other simulators of biological uptake such as solvent-filleddialysis membranes, uptake into SPMDs has been shown to mimic fairly closely uptakeby aquatic organisms. In a study by Prest et al. (1992), SPMDs and clams weredeployed on the Sacramento and San Joaquin Rivers and compared on the basis ofsequestered pesticides, PCBs, dioxins and furans. In general, the uptake profiles werefairly similar, but levels of chlorinated organics were about 1.6 times higher in the clamsthan the SPMDs. In addition, the clams seemed to preferentially accumulate12hexachiorinated PCBs, while SPMDs contained higher levels of tetra- and pentachlorinated PCBs. This is in agreement with the tendency of other bivalves topreferentially degrade lower chlorinated congeners (Tanabe and Tatsukawa, 1987). Interms of dioxins and furans, their uptake profiles into both clams and SPMDs werevirtually identical (see Figure 2).I,,.4S7.I,10S.4.___16z2•I,—IFigure 2. Comparison of Dioxin and Furan Uptake into Clams and SPMDs.(from Prest et al., 1992. Used with permission.)LJ I I I— ———2 •‘Dioxin $g$-F U ran a013SPMDs and transplanted mussels were deployed on Smith Island, Prince WiffiamSound, Alaska, in 1992 to assess the bioavailability of PARs three years after the ExxonValdez oil spill (Shigenaka and Henry, in press). Uptake into both monitors wascompared, and is shown in Figure 3. It was found that PAR accumulation levels weresimilar, but the distribution profiles varied. Mussels accumulated alkylated phenathrenes,especially C-3 phenanthrene, whereas SPMDs contained higher levels of the lighterPARs, naphthalenes and fluorenes. This indicates that PAR absorption into SPMDs anduptake by mussels do not occur by the same process. This is expected, since compounduptake into SPMDs is purely a passive process, while uptake into living organisms alsoinvolves active processes. For example, one route of entry into mussels is by ingestionof food, a mechanism of uptake obviously not available to SPMDs. However, althoughPAR distribution differed in the two monitors, total target PAR results in mussels andSPMDs were significantly correlated, thereby demonstrating the usefulness of SPMDs assimulators of biological uptake.Another recent study comparing PAR accumulation by mussels with uptake intoSPMDs indicated similar results. Although uptake profiles were fairly similar, themussels accumulated predominantly alkylated dibenzothiophenes and alkylatedphenanthrenes, while the SPMDs also sequestered a high proportion of alkylatednaphthalenes (Lebo et al., 1992). The SPMDs were also preferentially enriched in thelower molecular weight compounds compared to the mussels, again likely due to thefacilitated elimination of these compounds from bivalves.14PERCENTofPERCENTofTOTALTARGETPAHTOTALTARGETPAIl.-t)t.)(1)--k)()o(110(110(1100(I’OUIo(noONaphthaleneIINaphthalene1Ic4c..ijC-2___C2j___C-3C-304IC4FluoreneIFluoreneIC-IC-iC-24iC-2C-3C-3DibenzothiopheneIDibenzothiophenejIC.IC-laC-21C-2C-3PhenanthreneiPhenanthreneC-2C-2I(I03C-3AnthraceneAnthraceneFluorantheneFluorantheneIPyrenePyrenedc-iC-i$Benzo(a)A.n(hraceneBenzo(a)Ai)thraceneIIChryseneChryseneC-iC-i0.2C-2Benzo(l))I?lLloranlheneBenzo(b)FluorantheneIBenzo(a)PyreneBenzo(a)Pyrenelndeno(a,h)Anthracene-Indeno(a,h)AnthraceneICDibenzo(a,h)An(hraceneDibenzo(aJi)AJ-)gracefle1Benzo(g,h,i)PeryleneBenzo(g,h,i)peiylene1ImIn work carried out by Huckins et al. (in press), SPMDs were exposed to2,2’ ,5,5’-TCB under the same conditions as the goldfish in the Bruggeman et al. work(1981). The concentration factor of 1.0 X 106 determined in the SPMDs correspondedvery closely to the value of 0.4-1.5 x 106 determined by Bruggeman et al.At present, there is little information on the use of SPMDs as pulp mill effluentmonitors. Paasivirta et al. (1992, unpublished information) filled polyethylene tubingsegments with bleached pulp mill effluent and Soxhlet extracted them in hexane or ethylacetate. GC/MS analysis identified a number of small molecular weight compounds, bothchlorinated and non-chlorinated. An attempt at placing SPMDs near pulp mills in CentralFinland was not entirely successful, with significant Aufwuchs growth evident and anescape of triolein from some SPMDs.2.3.5 Advantages of SPMDsSPMDs have many advantages over other integrative monitoring approaches,including simplicity of use and low cost. They can be made “clean” prior to use, thusavoiding background contamination which may cause errors in the interpretation ofresults. They can concentrate contaminants which might be metabolized by biota,monitor episodic contamination problems and determine the presence and bioavailabilityof contaminants, since they mimic biological uptake. Polymeric membranes such aspolyethylene are also more resistant to organic solvents and biological degradation thancellulose, which is rather fragile. Cleanup of SPMDs by organic solvent dialysis issimple and straightforward. Because SPMDs can only take up dissolved material, there16is less ambiguity as to whether a compound was associated with particulate matter or in adissolved state.When monitoring trace levels of contamination, the ideal monitoring tool must becapable of concentrating the contaminant efficiently so that it can be detected. SPMDsappear to be advantageous in this aspect. In comparing uptake into clams (wet weightbasis) with uptake into SPMDs (total weight basis), Prest et al. (1992) found clams to beless efficient on a volumetric basis than SPMDs at concentrating contaminants. Also,since the SPMD blanks were much cleaner than the clam blanks, the interferences inSPMD analysis were fewer.2.3.6 Membrane Transport Theory2.3.6.1 Mechanism of TransportDue to the non-porous nature of the polymeric film which results in very smalland transient cavities within the polymer, it is generally believed that molecules crossingthe membrane are “solubilized” by the polymer, rather than diffusing per se (Hwang andKammermeyer, 1975; Comyn, 1985). The dependence of uptake on this process isillustrated by the following equation (Underhill and Feigley, 1991):P=DS (1)where P is the permeability coefficientD is the molecular diffusion constantS is the analyte solubility in the polymerThe maximum breadth of molecule that can be transported across the polymericfilm is 10 A, the diameter of the randomly-forming cavities. This allows size exclusion17of any undissolved or particulate-associated analyte by the polyethylene membrane.SPMDs are thus monitors of only dissolved organic contamination.Analyte transport across the membrane appears to be the rate-limiting step inSPMD uptake (Huckins et al., 1990a). This assertion is based on the fact that diffusionof neutral compounds through non-porous polymers is a slow process (Hwang andKammermeyer, 1975). Also, there is a rapid increase of analyte levels in the polymermatrix following exposure, thus showing that the transfer of analyte from the bulkwater/polyethylene interface to the matrix is not difficult (Huckins et al., 1990a). Factors Affecting Membrane TransportIn terms of membrane-related factors, the thickness of the polymer directly affectsuptake, since an increase in thickness results in an increase in the permeabilitycoefficient, according to:P = (mass of analyte) (membrane thickness) (2)(surface area of membrane) (time) (concentration drop across membrane)where P is the permeability coefficient of the polymer (Yasuda, 1967).It is therefore important to verify constant thickness in a batch (roll) of tubing to preventdiscrepancies in uptake due to variations in thickness. It is also important to use tubingof thickness >25 m, since a thinner polymer may be too fragile and not feasible forSPMD use (J. Huckins, personal communication, 1992).The density of the polymer, as well as the crystallinity and chain orientation,18when increased, result in decreased permeability. Low density polyethylene usually hascrystallinity <60%, and a density of <0.928 g/mL. High density polyethylene, withcrystallinity up to 95%, has low permeability (Roff et al., 1971). Increased chainorientation, which occurs when the polymer chains are parallel to each other, tends todecrease diffusional transport due to the reduction in available transport channels (Hwangand Kammermeyer, 1975).The additives incorporated into polyethylene can affect its permeability as well,especially if the compounds are nonpolar, in which case they tend to make the membraneeven more nonpolar and increase the transfer of nonpolar compounds. A proposed ideafor increasing the permeability of the polyethylene film to nonpolar analytes is to coextrude it with triolein (H. Yasuda, unpublished information, 1992). Besides increasingthe hydrophobicity of the membrane, the presence of additives also decreases chainorientation and therefore increases transport rates.Environmental factors also play a role in the transport of compounds across thepolymeric membrane. An increase in temperature increases polymer permeability bydecreasing both crystallinity and density. In addition, analyte molecules have higherkinetic energies and therefore diffuse at a faster rate.The presence of Aufwuchs colonies may affect uptake into SPMDs, as mentionedearlier, by reducing flux through the polymeric membrane.Analyte characteristics also affect their uptake into SPMDs. Generally, it isbelieved that molecular weight is inversely correlated with diffusion rates in bothbiological and synthetic membranes (Collander, 1954; Lieb and Stein, 1969). In fact,19equations relating diffusion coefficients to molecular weight in the two membranes arevery similar:in biological membranes (Collander, 1954),Dm = D0MSrn (3)where Dm is the diffusion coefficientD0 is the calculated diffusion coefficient for a compound ofunit molecular weightsm is a parameter for membrane mass selectivityin synthetic membranes (Lieb and Stein, 1969),Dm = Dm°M (4)where Dm is the diffusion constantDm° and sm are constants characteristic of the polymer usedM is the diffusant molecular weight relative to methanolThe similarity of these two equations lends more credibility to the hypothesis thatSPMDs can realistically simulate biological uptake. Furthermore, work by Huckins et al.(1990a) verified that uptake varies significantly based on the molecular weight parameter.SPMDs were exposed to two tetrachlorinated PCB congeners (292 da), fenvalerate (420da) and mirex (546 da). Mirex failed to attain equilibrium concentrations (defined as<10% change between sample periods of 7 days) in the SPMDs after 21 days, whereasthe PCB congeners reached maximum concentrations in only about 7 days. Anotherstudy (Huckins et al., 1990b) investigating the dialytic cleanup of SPMDs showed similarresults. Naphthalene (128 da, 2 rings) reached 90% dialytic recovery in 4 hours, whiledibenz[a,h]anthracene (278 da, 5 rings) required 24 hours to reach the same level. Notunexpectedly, mirex residues required greater than 48 hours to attain equilibrium uptake.20Because of the 10 A maximum diameter transport channel, compounds withmolecular breadth exceeding that value cannot access the lipid within the polyethylene.Huckins et al. (1990a) suggested that mirex, with a length of 9.82 A and a breadth of9.2 A, may be the maximum size able to be absorbed by SPMDs.The polarity of substituents on analyte molecules significantly affects their abilityto accumulate in SPMDs. According to Roff et a!. (1971), the hierarchy of ease ofdiffusion of low molecular weight compounds (<200 da) in polyethylene, from least tomost, is as follows: alcohols < acids < nitro-derivatives < aldehydes < ketones <esters < ethers < hydrocarbons. Highly polar compounds are the least likely to diffusethrough the hydrophobic polymer.An analyte’s octanol-water partition coefficient (K0), and especially its trioleinwater partition coefficient (K), dictate the extent to which the analyte will beconcentrated in an SPMD, as well as how easily it will enter the polyethylene matrix.These parameters will be discussed further in Sections and, the vapour pressure, water solubility and Henry’s Law constant (equal tovapour pressure divided by water solubility) of a compound affect its transport acrosspolymeric membranes. If a compound has a high vapour pressure coupled with lowwater solubility, it will therefore have a large Henry’s Law constant and will tend to betaken up efficiently into SPMDs.212.3.7 Equffibrium Partitioning2.3.7.1 Octanol-Water Partition Coefficients (I()Equilibrium partitioning is the process by which an organism comes intoconcentration equilibrium with compounds present in its environment. There is muchinterest in this phenomenon, since it is responsible for the bioconcentration of chemicalsin aquatic environments. In particular, much work has been done regarding theequilibrium octanol-water partition coefficient, or K0, since this parameter determinesthe degree to which a compound will partition out of the aqueous environment and intoorganic material. The K0 is therefore a useful physico-chemical parameter for assessingthe lipophilicity of organic compounds (Leo et al., 1971), and is used to characterizepartitioning of environmental contaminants into biota as well as into soils and sediments(Hawker and Connell, 1989). It is a common index for chemical hazard assessmentbecause of its good correlation with bioconcentration factors (BCFs) (Neely et al., 1974;Veith et a!., 1979) and aquatic toxicity (Liu et a!., 1982).The octanol-water partition coefficient specifically refers to the ratio of theconcentration of an analyte in octanol to that in water, after equilibrium has beenattained. It can be measured by using the shake-flask method, in which spiked octanol isshaken with water and the concentrations in each phase measured. Alternatively, highpressure liquid chromatography (HPLC) can be used, in which there is a correlationbetween the K0 of a compound and its capacity factor (derived from retention times)when using a C18 reverse-phase column (Veith et a!., 1979; Xie et a!., 1984). Once aK0 has been determined, it is usually expressed logarithmically as log K0, or log P.22Many authors have established inverse correlations between K0 and watersolubility, thereby facilitating estimation of one of the parameters if the other is known(Briggs, 1981; Miller et al., 1985). In addition, there appears to be a correlationbetween log K, and the bioconcentration factor (BCF) of a compound. For example,the log K0 of 2,2’,5,5’-TCB is 6.1 (Shiu and Mackay, 1986), which corresponds to aconcentration factor in octanol of 1.26 x 106. Recall that Bruggeman’s work (1981)found a bioconcentration factor of 0.4-1.5 X 106 in goldfish, thereby illustrating theability of K0 to predict BCF. Lipid-Water Partition Coefficients (KLW)The partitioning behaviour of an analyte into a lipid, instead of octanol, isgoverned by its KLW. One might speculate that a correlation would exist between theK0 of a compound and the KLW. In fact, Platford (1979) compared experimentallyderived triolein-water partition coefficients (Ks) and K0s for three organic compounds,hexane, benzene and carbon tetrachloride. It was found that the Ks were 2-3 timeslarger than the K0s, indicating that there was indeed a correlation. Similarly, in otherresearch (Chiou, 1985) the Ks for 38 slightly water-soluble organic compounds weredetermined and compared to known K0s. A significant correlation was identifiedbetween the two parameters. It was also found that when the determinedbioconcentration factors in fish were expressed on the basis of lipid content instead oftotal body weight, they were almost equal to the Ks, thus indicating the predictiveability of K for BCF.23Huckins et al. (1990a) reported similar fmdings using SPMDs. They generatedKs and KLWS (grass carp lipid) for 2,2’,5,5’-TCB after a 21 day exposure, andcompared them to Chiou’s K0 and K values. Good agreement was found between thevalues, shown in Table 1.Table 1. Comparison of Experimental Partition Values for 2,2’,5,5’-TCB withLiterature Values (from Huckins et al., 1990a).Type of Partition Method of ValueCoefficient DeterminationK0 shake-flask 6.46 x i0 (Chiou, 1985)K direct partitioning 4.17 X i0 (Chiou, 1985)K SPMD-derived 1.02 X 106 ± 9.80 x 10’KLW SPMD-derived 6.63 x i0 ± 1.50 x i05The difference in K is most likely a result of the differences in methods used for K0determination. Chiou used a direct partitioning method whereby triolem was simplyexposed to TCB in solution, while Huckins et al. enclosed the triolein in polyethylene.Interestingly, when grass carp lipid in SPMDs was spiked with 2,2’ ,5 ,5 ‘-TCB andequilibrium reached by that route, water concentrations of TCB only reached 55% ofthose resulting from a spiked water SPMD exposure (Huckins et al., 1990a). Obviously,the two pathways to steady-state are not completely reversible.242.3.8 Comparison of Bioconcentration and Membrane TransportIt can easily be shown that there is a phenomenological link between thebioconcentration of contaminants in aquatic organisms and polymeric membranepermeability, a conclusion also reached by Lieb and Stein (1969). Beginning with theexpression for uptake and depuration of contaminants by fish (Mackay and Hughes,1984),CT = C (K1/2)(1-exp(-K2t)) (5)where CT is the analyte concentration in the target organism (fish)C is the analyte concentration in waterK1 is the uptake rate constantK2 is the clearance rate constantThe ratio ofK1K2 is equivalent to the bioconcentration factor, and is also approximatelyequal to KLW, assuming C is constant and that t approaches infinity. If we substitute CTwith CL (to apply to SPMD5), we then obtain:CL = CWKLW (l-exp(-K2t)) (6)Further substituting Kmw/Kmi for KLW, and K,iPA/Vj for K2, where Kmw is themembrane-water partition coefficient, K is the membrane-lipid partition coefficient, P isthe permeability coefficient, A is the membrane surface area and V1 is the volume oflipid within the SPMD, we obtain:CL = (1-exp (KmiPAs/Vi)(t)) (7)Kmi25which is identical to the expression developed by Yasuda (1967) for the transport ofnonpolar compounds across a polymeric film. The strong similarity evident hereindicates that a similar process is occurring in bioconcentration of contaminants andtransport across synthetic membranes.2.3.9 Applications of SPMDsThe potential applications of SPMDs are numerous. Their major use at presentappears to be the monitoring of lipophilic contaminants in aqueous environments. In thiscapacity, they can be used to confirm bioaccumulation mechanisms in other organismsand verify bioavailability, as well as provide integrative monitoring for such lipophiliccompounds. A distinct advantage of SPMDs is their ability to be deployed in harshenvironments which would be too hostile for biomonitors such as fish. In this way, theycan serve as “artificial” aquatic organisms, or biological surrogates, when direct tissuesampling is impractical or even impossible.The use of SPMDs for the cleanup of bulk lipids has previously been mentioned(Huckins et al., 1990b; Meadows et a!., 1993). Such a dialytic enrichment applicationcould prove especially useful for separating organic contaminants from biological lipids.Another interesting potential application of SPMDs is passive air sampling.From residues quantified both within the SPMD and in laboratory air over a 28-dayperiod of SPMI) exposure, Zajicek et a!. (unpublished material, 1992) determined an airsampling rate of 10.7 m3/day/g triolein. Furthermore, during the exposure period,laboratory air samples were taken and demonstrated a 40% decrease in PCB levels,indicating that perhaps SPMDs are capable of “cleaning” contaminated air.262.4 Overview of Kraft Mill Wastewaters2.4.1 The Kraft Pulping ProcessKraft pulping, also known as “alkaline” or “sulfate” pulping, is currently thedominant pulping process in use worldwide because it can produce strong pulp from awide variety of wood species, while allowing for chemical recovery. The basic processconsists of cooking wood chips with sodium hydroxide and sodium sulfide at atemperature of 160-180°C and pH of 13-14. This acts to solubilize wood lignin(Biermann, 1993).Much of the adverse environmental impact associated with kraft mills, includingthe usual parameters of BOD5 (5-day biological oxygen demand), toxicity and colouroriginates from either the spent cooking liquor, called black liquor, or from variouscondensates of process gases.2.4.2 Chemical Composition of Kraft WastewatersKraft wastewaters are complex mixtures containing hundreds of organiccompounds. These compounds can be broadly classified into two groups: (1) steamvolatile compounds present in condensates, and (2) non-volatile compounds present inblack liquor, pulp wash water and bleach plant effluent in bleached kraft mills. Themajority of these compounds originate from the wood, either as lignin degradationproducts or wood extractives.There are several main classes of compounds which can be found in krafteffluents. Sulfur-bearing compounds, such as hydrogen sulfide, methyl mercaptan,27dimethylsulfide, dimethyldisulfide and dimethyltrisulfide, are formed from sodium sulfidein the cooking liquor (Blackwell et al., 1979). Thiophene is also present, but its origin isunknown.Acids, including resin acids, fatty acids and saccharinic acids, are present in krafteffluents at significant levels, and originate from the wood itself. Resin acids have beenidentified as the main contributors to fish toxicity in whole-mill effluent (Rogers et al.,1975). Various acidic compounds which have been identified in unbleached kraft effluentare shown in Table 2.Table 2. Acidic Compounds in Unbleached Kraft Mill Wastewater. (adaptedfrom llrutfiord et a!., 1975)p-tolyl-valeric acidpalmitic acidstearic acidoleic acidlinoleic acidpimaric acidsandaracopimaric acidisopimaric acidabietic aciddehydroabietic acidneoabietic acidformic acidacetic acidglycolic acid2-hydroxybutyric acidlactic acid3-deoxythreopentonic acid3-deoxyerythropentonic acidalpha-glucoisosaccharinic acidbeta-glucoisosaccharinic acidalpha-glucometasaccharinic acidbeta-glucometasaccharinic acidalpha-galactometasaccharinic acid28The neutral organic compounds, a large and heterogeneous group which includesterpenes and low molecular weight hydrocarbons, aldehydes, ketones and alcohols, areimportant constituents of kraft effluent. Since these compounds tend to be volatile, theyare present in foul condensates resulting from the condensation of process gases fromdigesters and multiple-effect evaporators. Foul condensates have been shown tocontribute significantly to many of the negative environmental effects of kraft millopeiations (Blackweli et al., 1979). A list of representative neutral organic compoundsidntifled in biotreated bleached kraft effluent, along with their approximateconcentrations, is shown in Table 3.Table 3. Neutral Organic Compounds Identified in Biologically TreatedBleached Kraft Mill Wastewater. (from Vo&s, 1984. Used ivithpermission.)spptoi1ma.concn. ,g/Lp..kno. eompd d.ntUT.d mu ring.1 2.5.dUnethy1.2-cop.M1noiI 6 (9) 2—102 dici.thyl uisuUld.• ?(Qd (6)3• 10(1)4 t.4.dfcldoeob.fll*fl. NQ5 2.3-d!.thy12-cyc1cp.aLrnoo. 10(6) 5-20S•fli 10 (1)1 20(8) 9-40S Wcpb.n. 4(1) 2-69 2.arIthcpbe I m 2-1010 hyI.2-dop.t..iion. 9(8) 1-2011 !anchon. 10(5) 12012 23.4.S.1.tza.th114.qdO. 4(9) 2-10p.nhIDøM’13 tm.thy(-2ydøpinticni ?IQ (7)14 (.Dchyt aJcoo1’ 7(4) 2-1015 cydoh.xylfdine .c.too. 30(4) 8-8016 wudentUTid eatp.. 10(3) 2-2011 campbor 40(7) 3—10018 2.aqthyI.3410pt09$Cytle 30(2) 20-30potaon.19 2.ptopioytbiOph.O1 7 (7) 2—1020 udiiWtid te,p.. a)cobol 30(7) 2-6021 to-4-L 50(4) 10-10022 ideitied .e:. siceat 50 (5) 2—10023 200 (2) 5-30024 ,erbeoo.’ 20 (3) 10—2025 pipcrit3C 10 (3) 10—2026 id.iflid 10 (6) 2—2027 dicydob.xYI*iaI 60 (2) 30-100IS dIcb)ate.pci3-e1 6 (8) 2—2020 d.thyI pbLS*laCI 50(9) 20-10030 pchi-t.2.d1o1 10(5) 3-3031 un(d.ngilTd S.beai(g eopcund 10(8) 3-2032 dii,utyI pbt*iac. NQ (8)33 dibuy( p6thaZ44 70 (9) 10-10034 jiiei (0 (7) 2—3029Terpenes are initially present in wood, and may react during pulping to form otherterpenes. The effluent concentration of terpenes is reduced significantly when turpentinerecovery is practiced (Blackwell et a!., 1979). A list of common terpenoid compoundswhich have been identified in kraft mill secondary treatment systems is shown in Table 4.Table 4. Terpenes Identified in Kraft Mill Aerated Lagoons. (adapted fromWilson and Hrutfiord, 1975)chloroformalpha-pinenesantenecamphenesabinenebeta-pinenemyrceneaipha-phellandrene1 ,4-cineolelimonene1, 8-cineolebeta-phellandrene3-carenep-cymeneterpinolenefenchonecamphorlinaloolfenchyl alcoholterpinene-4-olbomeolaipha-terpineolanetholeThe main phenolic compounds include phenol, o-, m- and p-cresol, vaniffin,acetovanillone, syringol and guaiacol. These compounds are probably formed fromphenolic structural elements in lignin during digestion (Wilson and Hrutfiord, 1971).30Although the compounds listed in the above tables have been identified in certain kraftmill wastewaters, it must be noted that wastewaters can be extremely variable, with theircomposition depending on many factors, such as the type of wood pulped and the effluenttreatment system, including modifications such as turpentine collection.2.5 The Fish Tainting Issue2.5.1 Overview of Kraft Mill Effluent-Induced Tainting2.5.1.1 IntroductionThe first documented case of fish absorbing off-flavour compounds from theiraqueous environments was in 1936 (Thaysen and Pentelow, 1936). Since then, therehave been numerous cases of fish tainting, including many caused by exposure to pulpmill effluents (Shumway and Chadwick, 1971; Cook et al., 1973; Whittle and Flood,1977; Gordon et al., 1980; Paasivirta et a!., 1983). Most of the research carried out onthis subject has concerned kraft mill discharges. Tainting has occurred with both treatedand untreated wastewaters, from both bleached and unbleached mills, in both marine andfreshwater environments, and is obviously of great concern because it can render manyrecreationally and commercially important fisheries useless.Unfortunately, despite much work on the issue, there has not been a great deal ofprogress. There are many discrepancies between tainting studies, so that the degree andextent of tainting originating from different pulp mills vary widely. Furthermore, theidentification of specific tainting agents is complicated by the complex chemical matrixpresent in pulp mill wastewaters.312.5.1.2 Mechanism of TaintingOff-flavour compounds can be taken up by fish via the following routes (Persson,1984):(1) absorption through the gills;(2) absorption through the alimentary canal; and(3) absorption on or through the skinOf these routes, gill contact seems to be the most likely route of uptake, although whilefish are feeding, absorption through the gut seems to be important (Persson and York,1978).It is unlikely that tainting agents bioconcentrate to a large degree, since theygenerally have relatively low K0s. For example, two compounds known to cause off-flavours in fish, thiophenol and 4-thiocresol, have log K0s of 2.52 and 3.18,respectively (Heil et al., 1989). Generally, compounds noted for their ability to persistand accumulate in the aquatic environment have log K0s > 5.5 (Chiou et al., 1977).However, bioconcentration processes must be occurring to some degree for an off-flavourto be detected. Identification of Tainting AgentsGenerally, tainting agents are expected to be relatively volatile. When stronglytainted and mildly tainted salmon were analyzed using gas chromatography, it was foundthat the total area of all volatile peaks from the strongly tainted sample was about 5 timesgreater than that from the sample with the less noticeable off-flavour (Berg, 1983).Figure 4 shows GC chromatograms of the two samples.32Figure 4. Gas Chromatograms of Volatile Compounds from (a) Strongly TaintedSalmon Flesh, and (b) Mildly Tainted Salmon Flesh (from Berg, 1983.Used with permission.)1•1 I.(a)(b)$ II I. IIII33The compounds identified by GC/MS as being present in the volatile fractionincluded terpenes and their derivatives as well as alkyl- and alkenylbenzenes. Terpeneshave previously been suggested as possible tainting compounds (Rogers, 1978).Total reduced sulfur (TRS) compounds are suspected to cause off-flavours due totheir low odour thresholds. TRS compounds, which include hydrogen sulfide,dimethylsulfide, dimethyldisulfide and methyl mercaptan, may collectively be present inuntreated kraft mill effluent at 8-60 mg/L levels (Cook et al., 1973). When oneconsiders the fact that catfish exposed to 15 mg/L dimethylsulfide for only 10 minutesdeveloped a noticeable off-flavour (Maligalig et a!., 1975), it is clear how tainting couldoccur. However, at a mill located in Cornwall, Ontario, the implementation of pollutionabatement measures designed specifically to remove TRS compounds actually appeared toworsen the tainting problem (Liem et a!., 1977).Polynuclear aromatic sulfur heterocyclic (PASH) compounds have been linked tooff-flavours in fish exposed to oilsands wastewaters (Koning and Hrudey, 1988).Apparently, there was a positive correlation between the concentrations of organosulfurcompounds such as benzothiophene, dibenzothiophene and their ailcylated derivatives andthe detectability of a taint.Resin acids have been suggested as potential off-flavour agents, but no informationexists on their tainting propensity (McLeay, 1987). It is known that resin acids canbioconcentrate to some degree in fish, especially in tissues such as liver, kidney and brain(McLeay, 1987). However, it has been suggested by Oikari and Holmbom (1986) thatresin acids are rapidly made water-soluble and excreted by fish through the bile.34Phenol derivatives and associated compounds, both chlorinated and non-chlorinated, have received much attention as possible tainting agents (Persson, 1984).Among those mentioned as having off-flavour properties in fish are isopropyiphenols,diisopropylphenols, methylisopropyiphenols, thiophenols, guaiacol and syringol (Cooket a!., 1973; Hell and Lindsay, 1988). Chlorinated phenolic compounds have also beenstrongly implicated as tainting agents in some studies (Paasivirta et a!., 1983; Kleinet a!., 1987), but not in others (Kovacs et a!., 1984). Sources and Removal of Tainting AgentsThe literature seems to implicate foul condensates as major sources of taintingpropensity (Brouzes and Naish, 1979; Gordon et a!., 1980). Exposure to 0.2% foulcondensate has been reported to cause off-flavours in both rainbow trout and perch (Liemet a!., 1977). More specifically, Cook et a!. (1973) indicated that the evaporatorcondensate, digester foul condensate and recovery furnace flue gas condensate were theprocess streams most likely to contain off-flavour compounds.A number of actions can be taken to reduce the levels of tainting agents in kraftmill effluents. In terms of physical treatment, steam stripping removes about 95% ofTRS compounds (Autopro Canada, 1991), and has been shown to decrease the taintingpropensity of foul condensate (Liem et a!., 1977). Similarly, intensive aeration using aturbulent contact absorber removed TRS compounds and improved the odour of theeffluent, although it did not significantly reduce the degree of tainting (Cook et a!.,1973).35Biological treatment has been shown to be effective in reducing the taintingpropensity of effluent (Shumway and Chadwick, 1971; Cook et al., 1973). Cook et al.(1973) reported a dramatic improvement in perch flavour after activated sludge treatmentof 1 % foul condensate for 24 hours. Similar results were shown for aerated lagoontreatment of 4% condensate for 5 days. However, work by Gordon et al. (1980) showedno significant reduction in tainting propensity of bleached kraft mill effluent afterbiotreatment, although BOD and toxicity were reduced effectively. Therefore, some ofthe tainting agents may not be readily biodegradable.2.5.2 Eulachon Off-Flavour at Kitimat, B.C. IntroductionEurocan Pulp and Paper Co. is an unbleached kraft mill located in Kitimat, B.C.The mill utilizes softwoods, including lodgepole pine, spruce, balsam fir and hemlock.Wastewater treatment is via two batch settling ponds and an aerated stabilization basinwith 5 day hydraulic retention time. Effluent discharge is to the Kitimat River, whichdrains into the Douglas Channel 3.2 km from the mill (Autopro Canada, 1991). Figure 5shows a map of the area. According to Warrington (1987), the effluent has a residencetime of 1 to 2.5 hours in the river.36FIgure 5. Map of the Area Surrounding Eurocan Pulp and Paper Co., Kitimat, B.C.Haisla ReserveKitimatRiver I37Since 1972, native Haisla Indians have been unable to use eulachon fish capturedin the Kitimat River downstream of Eurocan’s effluent discharge due to noticeable off-flavours. The Pacific eulachon (Thaleichthys pacficus) is a small, oily smelt having alipid content of — 15% wet weight (Rogers et al., 1990). This high lipid content,combined with the coincidence of the spawning period with the Kitimat River annual lowflow, predisposes the eulachons to flavour impairment by exposure to Eurocan effluent.Since the Haisla have historically used the flesh and rendered grease of eulachons asimportant food sources, they have been forced to collect eulachons from the Kemano,Kildala and Kitlope Rivers (Beak Consultants, 1991).Because eulachons do not feed in fresh water, the only uptake route forcompounds in the Kitimat River is by direct environmental contact, rather than byingestion of contaminated food. A recent Fraser River study reported the use of thePacific eulachon as an environmental monitor, due to its high lipid content (Rogers et al.,1990). The study found a correlation between eulachon body levels of certainorganochiorine compounds and the distance travelled upstream, indicating that eulachonscan bioconcentrate environmental contaminants.The off-flavour detected in the eulachons at Kitimat has been investigated severaltimes (Beak Consultants, 1991). In 1972, the degree of tainting of sockeye salmonexposed to Eurocan effluent was found to be directly correlated to the effluentconcentration. The following year, it was found that the effluent could taint eulachonseven at the river mouth, and that the degree of tainting increased with increasing effluentconcentration. Similar findings were reported in a 1975 study.382.5.2.2 Recent Tainting EvaluationsIn 1991, Beak Consultants carried out a study to determine if eulachons exposedto Eurocan final effluent at concentrations present in the Kitimat River would developoff-flavours (Beak Consultants, 1991). A preliminary assessment of water quality in thearea indicated that the concentration of total phenols near the river bottom upstream anddownstream of the Haisla Reserve exceeded receiving water quality objectives. It wasthought that this may indicate a potential role for phenols as tainting agents.In the main component of the study, eulachons were collected from the KemanoRiver and exposed to treated effluent concentrations of 0.63%, 1.25%, 2.5%, 5% and10% for 1, 3, 9 and 27 hours. Upon taste panel evaluation, it was found that exposureof fish to 3% effluent for >27 hours, or exposure to >3% effluent would result inincreasing percentages of test panellists noticing flavour impairment. Effluentconcentrations <3% were considered unlikely to cause tainting. Meanwhile, effluentdispersion studies performed by Beak had shown that during low river flows, effluentconcentrations along the shore of the Haisla Reserve were 3-4%, corresponding to thetainting threshold determined in the taste evaluations.A similar study was carried out in 1992 (Beak Consultants, 1992). In this study,eulachons were exposed to 0.63%, 1.25%, 2.5%, and 5% effluent for 3, 9, 27 and 96hours to determine the effect of a longer exposure time on tainting propensity. It wasfound that exposure to 3% effluent for >4 hours resulted in noticeable off-flavour, andfrom extrapolation, that the tainting threshold would be 0.3%. Coincidentally, while thetainting propensity of Eurocan effluent increased in 1992, the resin and fatty acid and39phenol levels in the effluent also increased (phenols increased from 0.01 mg/L in 1991 to0.2-0.3 mg/L in 1992).Another tainting study was carried out the following year (Beak Consultants,1993b) to assess the effect of the implementation of a turpentine recovery system andimproved brownstock washing on eulachon off-flavour. In addition, the study was aimedat determining a more accurate tainting threshold by exposing fish to lower concentrationsof effluent for longer periods of time to reflect the time eulachons actually spend in theKitimat River during spawning.Eulachon were exposed to 0.2%, 0.6%, 1.25% and 2.5% effluent for 3, 9, 27, 96and 240 hours. It was found that eulachons exposed to 1.25% effluent for 3 hours, and0.63% effluent for 9 and 27 hours, had a noticeable off-flavour. Tainting was alsodetected in fish exposed to 0.2% effluent for 96 hours.The results of the 1993 study showed that the tainting propensity of Eurocaneffluent had increased since the 1992 study, despite the in-plant modifications to increaseeffluent quality. This may have been due to an increase in the concentration of taintingchemicals in the effluent in 1993. However, it may have been that the taste panel, whichin 1993 was comprised of Haisla familiar with eulachon taste, was more sensitive to offflavours compared to the 1992 panellists, who had not been exposed to eulachon flavourto such a large degree.Chemical analyses of the effluent in both years showed a general decrease in resinand fatty acids, total phenols and BOD5 from 1992 to 1993. In addition, analysis ofgrease samples from eulachons exposed to 2.5% effluent for 27 hours showed that40terpene levels remained fairly constant between the two years, thus indicating thateffluent quality did not change significantly between 1992 and 1993.In another component of the study, an attempt was made to fractionate the taintingchemicals to facilitate identification. This was done by nitrogen-stripping control andtainted eulachon grease at 4 different temperatures (50°C, 100°C, 130°C and 160°C)into either mineral oil, which was then tasted by panellists, or hexane, which was thenanalyzed by GC/MS. Panellists detected an off-flavour at all temperatures except 160° C.GC/MS analysis showed 3 peaks, denoted A, B and C, which were identified in theextract of tainted grease which were not found in the control grease. The possiblecompounds and their respective probabilities are shown in Table 5.The 1993 study indicated that a reduction in tainting chemicals of 97% would berequired to ensure flavour acceptability, a significant increase from the 75% and 90%reductions determined in the 1991 and 1992 studies, respectively. Positive identificationof the tainting chemicals is required in order to target specific mill process streams andimplement controls so that ultimately the agents can be eliminated from the effluent.41Compound MatchQualityPeak A: 3 ,6,6-trimethyI-bcyc1o[3.1. lJhept-2-cnc2 832-mcthyl-S-(1-methylcthyl)-bicyclo[3.L .Ojhcx-2-cne 784-methyl-1-(1-methylethyl)-bicyclo[3. I .O]hexane 72Peak B: (3-methylbutylidcne)-cyclopentanc 781-methyl.4-(1-mcthyLethyl)-cyclohcxanc’ 904-methyl-l-(l-methylethyl)-cyclohcxanc’ 831 ,7,7-trimethyl-bicyclo[2.2. ljheptanc’ 83Meth-i(8)-ene’ 833-methyl-6-(1-methylethyl)-cyclohexane’ 831-mcthyl4-(1-methylcthylidcnc)-cyctohcxane’ 743 ,7.7-trirnçthyl-bicyclo[4. I .OJheptan& 74Peak C: 1, 1-dimethyl-2-(3-methyl-l ,3-butadienyl)-cyclopropanc1 86-methylcne-I-(1-methylcthyl)-cyclohexane’ 72(1S,3R.6R)-(-).-4-.carcnc’ 802-methyl-5-(1-mcthy!ethyl)-bicyclo(3.1.Ojhex-2-cne’ 80bcta-phcllandrcn& 78a1pha-pnene 78(E)-3 ,7-dimethyl-1 ,3 ,6-octatricn& 784-rnethyl-1-(1-methylethyl)-bicyclo[3 .1 .Ojhcxane’ 724-rnethylcne-1-(1-methylcthyl)-bicyclo[3. 1.0]hcxanc’ 723,7-dimcthyl-1,3,6-octadicn-3-oI acetate’ 72(Z)-3,7-dimethyl-1 ,3,6-.octathenc’ 72Table 5. Tentative GC/MS Identification of 3 Gas Chromatogram Peaks fromTainted Eulachon Flesh (from Beak Consultants, 1993a).423. RESEARCH OBJECTIVESAfter studying the available literature and determining the research areas requiringfurther study, as well as considering the situtation at Eurocan Pulp and Paper Co., thefollowing research objectives were established:(1) To investigate the background preparation steps in SPMD use. In particular, the batchextraction step was considered important, as it removes contaminating substances from thepolyethylene tubing.(2) To determine the octanol-water partition coefficients, uptake into SPMDs and dialyticrecovery from SPMDs of three model compounds which are known kraft mill effluentconstituents and potential tainting agents. This examination would also include 2,2’ ,5 ,5 ‘-TCB, a standard whose behaviour in SPMDs is well-characterized and could serve both toverify the technique and provide a basis for comparison.(3) To expose SPMDs to kraft mill wastewater in both static and continuous flow systems,in order to assess their ability to sequester contaminants in a real environmental situation.Of particular interest were compounds potentially responsible for the tainting of eulachonsin the Kitimat River.434. MATERIALS AND EXPERIMENTAL METHODS4.1 Materials4.1.1 SPMD ConstituentsLow-density layflat polyethylene tubing (2.54 cm wide, 63.5 m thick) wasobtained from Cope Plastics Inc., St. Louis, MO. and arrived in the form of 366m (1200ft) rolls. Triolein (1,2, 3-tri[cis-9-octadecenoyl] glycerol, 99%) was purchased fromSigma Chemical Co., St. Louis, MO. Due to its sensitivity to heat and light, it wasstored at -15°C in a foil-covered flask.4.1.2 Model CompoundsDehydroabietic acid (8,11, 13-abietatrien- 1 8-oic acid, >99%) was procured fromHelix Biotech Corp., Richmond, B.C. and refrigerated at 4°C. Guaiacol (2-methoxyphenol, 98%) was obtained from Aldrich Chemical Co., Milwaukee, WL (+)-alpha-pinene ([1S,5S-2,6, 6-trimethylbicyclo [3.1.1] hept-2-ene, 99%) was obtained fromSigma Chemical Co., St. Louis, MO. Selection of Model CompoundsThe three model compounds were chosen for two reasons. First, they are usuallypresent in significant quantities in softwoods (Swan, 1973; Fengel and Wegener, 1983)and are therefore common constituents of kraft mill effluent (Wilson and Hrutfiord, 1971;Hrutfiord et al., 1975; Rogers et al., 1975; Voss, 1984). Dehydroabietic acid is presentin untreated kraft effluent at 990-5780 g/L, and treated effluent at <20-1930 ugIL44(McLeay, 1987). Guaiacol is often present in kraft effluent at concentrations of 10 mg/L(Wilson and Hrutfiord, 1971). Terpenoid compounds, of which aipha-pinene is animportant constituent (Fengel and Wegener, 1983), are often present in untreated effluentat 8 mgIL, and treated effluent at 1 mgIL in mills with turpentine collection systems(Hrutfiord et al., 1975).Second, the three selected model compounds are representative of three groups ofpotential fish-tainting compounds: resin acids, phenols and terpenes (Cook et at., 1973;Brouzes and Naish, 1979). Since they are from three different classes of compounds,they have widely differing physico-chemical properties as shown in Table 6. Theirmolecular structures are shown in Figure 2,2’ ,5,5’-Tetrachlorobiphenyl (TCB)2,2’,5,5’-TCB was obtained from Ultra Scientific Co., Kingstown, RE and usedlargely as a standard to which the behaviour of the model compounds could be compared.There were several reasons to use this particular compound. First, initial work onSPMDs involved the use of TCB (Huckins et al., 1990a and 1990b); therefore, thebehaviour of TCB in SPMDs was well-established. In addition, TCB is considered to benon-dissociable (U.S. EPA, 1991). This is important because often a non-ionized speciesis more easily absorbed into aquatic organisms than its ionized form, resulting in uptakebeing dependent upon environmental pH (Saarikiski and Viluksela, 1982; Saarikiski etat., 1986). Similarly, uptake into SPMDs depends on the polarity and charge of thecompound. Finally, TCB is lipophilic, with literature log K,s of 5.81 (Chiou et at.,451982) and 6. 10 (Shiu and Mackay, 1986). Therefore, it would be expected to partitionwell into triolein. In fact, work by Chiou et al. (1982) has shown log (trioleinwater) for 2,2’,5,5’-TCB to be 5.62.Table 6. Selected Physico-Chemical Properties2,2’ ,5,5’-Tetrachlorobiphenyl (TCB).Index (1983) unless otherwise noted.of Model Compounds andAll information from MerckCompound Formula Molecular Form Boiling Solubility VapourWeight at Point in Water Pressure(da) Room ( ° (Pa)TempDehydroabietic C20H90 301.42 Solid N/A 4.9 mg/L’ nilacidAipha-pinene C10H6 136.23 Liquid 155 Practically 400.72insolubleGuaiacol C7H802 124.13 Liquid 204 16 gIL 14.692,2’,5,5’-TCB C12H614 292 Solid N/A 30 I.Lg/L4 0.00491 Nyren and Back, 19582 at 20 C, from manufacturer Material Safety Data Sheet3 at 25 C, from manufacturer Material Safety Data Sheet4 Shiu and Mackay, 19862aDehydroabietic acid Guaiacol Aipha-pinene 2,2’, 5,5’ -TCBFigure 6. Structures of the Model Compounds and TCB.464.1.4 Solvents and Standard SolutionsAll solvents were HPLC grade or equivalent and were obtained from either BDHInc., Toronto, Ont. or Fisher Scientific Co., Vancouver, B.C. Standard solutions wereprepared in acetone.4.1.5 Other MaterialsAmber glass jars (Qorpak, 500 mL) were purchased from VWR Scientific Ltd.,London, Ont. Spectra/Por dialysis tubing closures (40 mm) were obtained fromSpectrum Medical Industries, Houston, TX. An impulse heat sealer was used to seal theSPMDs. Glass weights fashioned from 5mm diameter glass rods were used to keep theSPMDs submerged.4.2 SPMD Methodology4.2.1 Preparation of SPMDsLayflat tubing was cut into 123 cm (46”) long segments. To remove any additivesor contaminants that may have been present, the tubing was batch extracted before beingfabricated into SPMDs. This was done by adding 25 mL of hexane to each tube, holdingboth ends closed and gently moving the hexane throughout the tube to contact the innersurfaces. Eight tubes at a time were then submerged in 1 L of hexane in a clear glass jarand allowed to extract for 24 hours. The segments were drained and allowed to air dryon hexane-rinsed foil. Handling of the tubing was kept to a minimum; when handlingwas necessary, polyethylene gloves were used to reduce the likelihood of contaminant47leaching from latex or nitrile gloves. Until required for SPMD preparation, tubing wasstored in sealed amber glass jars with hexane-rinsed foil lined caps.To prepare the SPMDs, 1 mL (0.91 g) of triolein was pipetted into the tubingsegment about 10 cm from one end. The tubing was then double heat-sealed directlyabove the tnolein and laid on an acetone-rinsed glass sheet. Using a polypropyleneroller, the triolein was spread into a thin film 86 cm (34”) long. The tubing was againdouble heat-sealed at the bottom after voiding as much air as possible, and, depending onthe intended use of the SPMD, it was threaded through either 1 or 3 glass clips tofacilitate submersion. Finally, after rotating one end of the segment 180 degrees, the 2ends were heat-sealed together. This Möbius strip configuration minimized the likelihoodof the tubing sticking together in the aqueous environment and consequently reducing theexposed surface area. Due to the propensity of SPMDs to sequester air-bornecontaminants, they were stored at 4°C in sealed amber glass jars to minimize theircontact with laboratory air.4.2.2 Cleanup and Dialysis of SPMDsAfter removal of the SPMDs from aqueous exposure, they were stored in sealedamber jars and refrigerated until further processing. Cleanup consisted of first addingapprox. 75-80 mL of hexane to each jar and shaking gently for 15 seconds. The SPMDwas then rinsed under cold tap water to remove any external material; this was onlyimportant in effluent exposures, although for consistency, rinsing was always carried out.Each SPMD was then successively rinsed with isopropanol and acetone to remove all48traces of water.When the SPMD had dried thoroughly, a dialysis tubing closure was placed on theheat seal and the SPMD was placed in a 250 mL clear glass jar for dialysis. 210 mL ofhexane was added to each jar and a piece of hexane-rinsed foil with a 2-inch slit was usedto cover the jar. The SPMD was fed through the slit and the tubing closure was used toanchor the SPMD and prevent the heat seal from contacting the hexane. Such contactmay lead to weakening of the seal (Huckins et al., 1990b). Another piece of foil wasthen used to cover the jar again before placing on the lid. The jars were then left for 48hours at room temperature. According to work reported by Meadows et al. (1993),dialysis temperature has little effect on analyte recovery, but at higher temperatures, alarger volume of solvent enters the SPMD, and lipid carryover is also higher.After 48 hours, the SPMDs were removed and disposed of. The dialysates werethen rotoevaporated to a small volume and, depending on the experiment, concentratedfurther under a gentle stream of nitrogen.4.3 Analytical Techniques4.3.1 Gas ChromatographyGas chromatography (GC) was carried out on two Hewlett Packard instruments, a5880A and a 5890 Series II, both equipped with flame ionization detectors (Fifi). A 30m, DB-5, 0.32 ID fused silica column with a film thickness of 1m was used (J&WScientific, Folsom, CA). Helium served as the carrier gas at a linear velocity of 20 cm/sat 290 C. The makeup gas for the FID consisted of helium @ 20 mL/min, hydrogen @4930 mL/min and air @ 400 mL/min. The temperature program varied depending upon theexperiment.4.3.2 Mass SpectrometryGas chromatographic/mass spectrometric (GC/MS) analyses were performed on aHewlett Packard 5985 GC/MS System with a DB-5 column as specified above. The ionsource temperature was 200 C, the interface temperature was 250°C and the scan rangewas 45-450 amu. As with GC analyses, the temperature program depended on theexperiment.In addition to GC/MS analysis, solid probe analysis was performed on acrystalline solid. The sample was deposited on a solid probe, which was inserted into themass spectrometer and slid into the vicinity of the ion source. To assist volatilization ofthe sample, the probe was heated from ambient temperature to 200°C at 25°/minute.4.3.3 High Pressure Liquid ChromatographyOctanol-water partition coefficients were determined by high pressure liquidchromatography (HPLC). An HP 1050 HPLC System was used with an ODS Hypersil 5m, 125mm by 4mm C18 reverse-phase column at 25°C. The HPLC system utilized aUV detector operated at specific wavelengths for each compound. The mobile phase wascomprised of 75% methanol: 25% aqueous buffer (0.06 M ammonium phosphate) at aflow rate of 1.0 mL/min. The buffer was acidified to pH 2.5 with phosphoric acid toanalyze the ionizable compounds, while the neutral compounds were analyzed with the50buffer at pH 7.5. Both buffers were filtered through a 0.45 m filter. The HPLCinjection volume was 5 L.Five reference compounds of known K0 at concentrations of 20 mg/L in HPLCgrade methanol were used for calibration. The compounds were as follows: p,p’-DDT,naphthalene, fluoranthene, toluene and pentachiorophenol. The four unknown compoundswere 2,2’ ,5 ,5 ‘-TCB, aipha-pinene, guaiacol and dehydroabietic acid, and were atconcentrations of 20 mg/L, except dehydroabietic acid, which was at 25 mg/L. Methanolwas used to determine the unretained volume.The UV absorption maximum for each compound was determined on a Beckman DBG UV spectrophotometer to increase sensitivity for HPLC UV detection. The UV maximadetermined are shown in Table 7.Table 7. Ultraviolet Maxima for Reference and Unknown Compounds.Compound Wavelength(nm)p,p’-DDT 235naphthalene 272toluene 270fluoranthene 284pentachlorophenol 220methanol 200dehydroabietic acid 2072,2’,5,5’-TCB 220guaiacol 220alpha-pinene 213514.3.4 Resin Acid AnalysisDehydroabietic acid analysis involved derivitization using diazomethane according to aPaprican in-house method based on the original method of De Boer and Backer (1954).DHA is converted to its methyl ester using diazomethane as the derivitizing agent as follows:R-COOH + CH2N R-COOCH3 + N2 (8)(resin acid) (diazomethane) (methyl ester) (nitrogen)Following derivitization, the extracts were concentrated to 1 mL and analyzed by GCand/or GC/MS.4.3.5 Aqueous Sample ExtractionGuaiacol was acetylated prior to extraction from aqueous samples using SM K2C03and acetic anhydride according to the method of Voss et al. (1981). Aipha-pinene and TCBdid not require any specialized treatment before extraction.4.3.6 Internal StandardsFor most of the research, 1 ,4-dichlorobenzene, chloronaphthalene andchioroanthracene were used as internal standards. These compounds have GC retention timeswhich correspond well to those of the test compounds; therefore, they were used to generateresponse factors for test compound and TCB quantification. 10 L of the internal standardmix, containing 2000 mg/L of each compound, was spiked directly into each GC vial prior toanalysis.524.4 Wastewater Exposures4.4.1 Static ExposuresUnbleached kraft mill wastewater samples (20 L) were obtained from Eurocan Pulpand Paper Co., Kitimat, B.C. and stored at 4CC. Two types of wastewater were used: (1)influent samples taken prior to secondary treatment, and (2) effluent samples takenimmediately before discharge to the Kitimat River (after secondary treatment by aeratedstabilization basin). SPMD exposure to both wastewaters was carried out for 1, 3 and 9days.Twenty-four aliquots (1 L each) of each wastewater were placed in 1 L clear glassMason jars. Also, 6-1 L aliquots of distilled water were placed in jars to serve as blanks.Half of the exposure containers were maintained under sealed, anaerobic conditions, whilethe other half were aerated to avoid dissolved oxygen depletion. It was hypothesized thatunder anaerobic conditions, certain effluent constituents might be chemically altered and thustheir SPMD uptake characteristics changed. However, due to the danger of volatilization oforganics in the aerated samples, it was decided to perform the exposures under both sets ofconditions.The aeration apparatus consisted of a molecular sieve column connected to the labbench air supply, a 4-way aquarium-style brass connector valve (airflow splitter) and a glassfrit impinger in each jar. Mixing was thus provided by gentle bubbling through the jar.Mixing in the anaerobic samples was achieved by using a magnetic stir plate and stirring thesamples for about 10 minutes each day. A schematic of the experimental design is shown inTable 8 and a photograph is shown in Figure 7.53FollowingSPMDexposureandtheusualdialysis,extractswereconcentratedto2mL.A1mLaliquotwasfirstderivitizedforresinacidanalysisand1mLwasanalyzeddirectlybyGCandGC/MS.Table8.ExperimentalDesignforStaticWastewaterExposures.Shownarenumberofsamplesforeachexperimentalcondition.INFLUENTEFFLUENTBLANKExposureAeratedNon-AeratedNon-AeratedNon:imeaeratedaeratedaeratedlday2222113days222211idays222211Figure7.PhotographofStaticWastewaterExposureSet-up.544.4.2 Continuous Flow ExposuresSPMDs were exposed to various final effluent concentrations in a continuous flowsystem at Eurocan Pulp and Paper Co., Kitimat, B.C. A serial diluter system maintainedeffluent concentrations of 0, 0.2, 0.6, 1.25 and 2.5% in 450 L exposure tanks, with aminimum flow rate of 2L/min. The maximum retention time of diluted effluent in thecontainers was therefore 225 minutes. SPMDs were also exposed to 100% effluent anduntreated water (0% effluent) from the Kitimat River. Exposures were carried out for 3, 9,27, 96 and 240 hours, although samples were not exposed for every concentration/exposuretime combination possible. The experimental sampling plan is shown in Table 9.Table 9. Experimental Design for Continuous Flow Wastewater Exposures.Exposure time is on the top, concentration on left. Shown are number ofSPMDs deployed at each combination.3 9 27 96 240hours hours hours hours hours0% 1 0 1 1 10.2% 2 0 0 1 00.6% 0 0 1 0 21.25% 0 1 0 0 02.5% 1 0 1 2 0100% 2 1 2 0 2TOTAL SPMDs = 22Eulachons were being simultaneously exposed as part of a tainting study carried out atthe time, and the SPMDs were placed in the tanks containing the fish, rather than in thediluter setup pails, to avoid outlet plugging. Control valves were set so that flows to each55downstream tank were constant, and dilutions were verified by sodium concentration ratios.A schematic of the dilution arrangement is shown in Figure 8.Figure 8. Dilution Set-up for Continuous Flow Wastewater Exposures.SUPPLYFROM RIVER(UNTREATED)TO FISH EXPOSURE TANKS565. RESULTS AND DISCUSSION5.1 IntroductionThe study consisted of three phases. First, method development work wasundertaken to verify the purity of the triolein and hexane, to determine the permeabilitycharacteristics of the polyethylene and to investigate the initial batch extraction step inSPMD preparation more thoroughly. This initial work was necessary for confirmation ofthe validity of the technique.The second phase was comprised of experiments with the model compounds and2,2’ ,5 ,5 ‘-TCB. Octanol-water partition coefficients were determined, and the totaluptake and dialytic recovery of the compounds were investigated. These experiments,carried out under controlled laboratory conditions, enabled characterization of the basicperformance of SPMDs in the context of pulp mill effluent compounds.The final phase consisted of SPMI) exposures, under both static and continuousflow conditions, to Eurocan effluent. The ultimate goals of this work were to applySPMDs in practice in a pulp mill setting and to assess their performance. In addition,suspected fish tainting agents were specifically targeted in order to determine their abilityto be absorbed by SPMDs.5.2 Method Development5.2.1 Verification of Hexane and Triolein PurityTo ensure that levels of impurities in the hexane and triolein were minimal, bothwere subjected to GC/MS analysis before use. Hexane was injected directly into the gas57chromatograph, while neat triolein was diluted in a small amount of hexane prior toinjection.The total ion count for hexane showed a large peak corresponding to cyclohexaneplus a few minor impurities, none of which were identified by the GC/MS database.Similarly, contaminants in the triolein were present only in extremely trace quantities (seeAppendix A for both spectra). The identified compounds found in triolein are shown inTable 10.Table 10. Compounds Identified hi Triolein.toluenetetrachioroethene3-butenylbenzeneethylbenzeneisocyanatobenzene1, 10-decanediolnonanoic acid1-methyl-3-(1-methylethyl) cyclopentaneundecanoic acid2-hydroxy-cyclopentadecanonebis-(2-ethylhexyl) phthalateThe low level of triolein contamination is consistent with its 99% purity level. Inaddition, some of the contamination may be due to column leaching or minor glasswarecontamination. Since the levels of impurities present were low, the hexane and trioleincould be used without any further purification or treatment.5.2.2 Verification of Consistent Tubing PermeabifityIt was important to verify that the tubing used to assemble the SPMDs in theresearch was consistent and would yield reproducible results. In other words, the effect58of the location of tubing segment within the roll of polyethylene on compound uptake wasto be investigated.One important variable inherent in the use of tubing was polymer permeability,since compound uptake is related to the permeability coefficient P (recall Equation (7) inSection, such that an increase in P causes an increase in lipid analyteconcentration. If segments of tubing of differing permeability were used to assembleSPMDs, there could be significant differences in uptake. At the dilute concentrationsused in the laboratory, the error resulting from such permeability variations could be verylarge indeed. Furthermore, the likelihood of variations in permeability within the 366 m(1200 ft) roll of tubing would appear to be high, since the manufacture of this material isnot carried out under rigidly controlled conditions.Other researchers have utilized other techniques to determine consistent tubingcharacteristics. Prior to using tubing for SPMDs, Huckins et al. (l990a) measured TCBuptake in a standardized exposure similar to the one carried out in this work, whilePaasivirta et al. (1992, unpublished material) determined thickness by infraredspectroscopy, since permeability is related to thickness (recall Equation (2) in section2.3.6.2).In the present work, SPMDs were assembled using tubing segments taken fromevenly spaced 38.1 m (125 ft) points throughout the polyethylene roll, except for 2additional “replicate” samples at each of the 0 m, 152.4 m (500 ft) and 342.9 m (1125 ft)marks. These “replicates” were intended to determine the variability in uptake betweendirectly adjoining segments. All SPMDs were exposed to 10 g/L 2,2’,5,5’-TCB in 1 L59glass jars for 7 days. This time period was chosen because it had been shown byHuckins et al. (1990a) to be the time required for TCB to reach its maximumconcentration in the lipid. The SPMDs were then dialyzed in hexane, and dialysatesrotoevaporated and made up to 10.0 mL. Assuming 100% uptake, the concentrateddialysates would contain 500 g/L TCB. Uptake was quantified by gas chromatography,and the results graphically displayed in Figure 9. The replicate samples are expressed asmeans of the three values.Because only one exposure time was investigated, there remained the possibility ofvariations in permeability being manifested in fluctuations in short-term uptake. Themaximum TCB concentration in the lipid is constant; therefore, only the rate of uptakechanges significantly with permeability variation. However, if the exposure time used isless than or equal to the time required to reach equilibrium, then differences in the totalamount of TCB absorbed will reflect differences in permeability characteristics. Since 7days is known to be the time required for TCB to reach steady-state, an exposure time of7 days should provide some information on the constancy of tubing permeability.However, if one wished to more thoroughly investigate the aspect of permeability as itrelates to the kinetics of uptake, one could carry out similar exposures to this work, butinclude more time points at shorter exposure times to assess the short-term uptake ofTCB. Hence, the rate of uptake could be compared between tubing segments.It appeared that TCB uptake was fairly constant throughout the roll, implying thatthere was little difference in the characteristics of the tubing from location to location.Regression analysis showed a slope of -0.008 with an R squared value of 0.02, indicating60a virtually horizontal line and therefore no observable trend through the roll.Furthermore, the coefficient of variation of the pooled samples was low, at 5.1%.In examining the replicate samples more closely, reproducibility was lower thanfor the separated samples. Coefficients of variation were 11.1 %, 12.2% and 12.1% forthe replicates at the 0 m, 152.4 m and 342.9 m marks, respectively.In addition to verifying consistency of tubing characteristics, this experimentprovided a baseline uptake value for TCB to which the uptake of the model compounds infuture work could be compared. The average TCB level in the concentrated dialysateswas 435.0 LgIL, corresponding to a recovery of 87.0%. Recoveries measured inindividual SPMD exposures ranged from 73.7% to 100.2%.This work enabled us to discount variations in tubing consistency as contributingto variations in SPMD uptake. If uptake into SPMDs incorporating tubing segments fromwidely differing locations had been found to vary, it would have been necessary tocharacterize the tubing in some other way prior to use in SPMDs to establish consistency.This would have been tedious and time-consuming; therefore, the verification ofconsistency performed here considerably simplified future work.61600cL400300200•l.0100150200250300350400Distancefrombeginningof roll(m)Figure9.2,2’,5,5’-TCBUptakeintoSPMDsvs.TubingSegment Location.5.2.3 Effect of Batch Extraction Conditions on Tubing5.2.3.1 Effect of Exposure Time on Tubing ConsistencySince polyethylene is an organic polymer, it was speculated that its dissolutioncould occur under prolonged solvent exposure during the batch extraction and dialysissteps, although polyethylene is generally not soluble in any solvent at temperatures lowerthan 50-60°C (Aggarwal and Sweeting, 1957). Such breakdown of the tubing could posea significant problem for three reasons. First, after the initial batch extraction step, thepermeability characteristics of the tubing might be altered and thus uptake of compoundswould be affected. Second, the presence of breakdown products in the dialysate wouldcomplicate analysis and result in extremely cluttered chromatograms. One would expectthere to be many peaks, since the ethylene polymers would be of varying lengths andwould therefore have varying GC retention times. Finally, the breakdown products maybe the same compounds that are under investigation as analytes. This would of courseresult in an overestimation of the quantities of these compounds in the sample.In actual fact, this investigation into the possibility of dissolution was initiated dueto the extremely messy chromatograms from an earlier experiment in which petroleumether and hexane had been used as dialytic solvents in model compound work. Thesechromatograms were virtually useless due to the sheer number of peaks present, leadingus to hypothesize that perhaps the tubing was dissolving in the solvents.Tubing segments were extracted for 8, 24, 48, 72 and 144 hours in hexane. Ifdissolution was occurring, one would expect the gas chromatograms of the extracts tobecome progressively more cluttered with peaks, with each peak constituting a possible63breakdown or dissolution product. However, the chromatograms of the 8 hour exposure(Figure lOb) and the 144 hour exposure (Figure lOc) appeared to be very similar in bothnumber of peaks as well as distribution and size of peaks, indicating that no dissolutionwas occurring. There obviously were many compounds leaching from the tubing asevidenced by a comparison of the sample chromatograms with the solvent blank (FigurelOa), but these compounds were probably additives used in the manufacture of the tubing.The identity and significance of these compounds will be discussed later.64Ur(a)(b)6 * ON OVI START FINAl.. TLTIL I3V: START PRGN RATL 221.73 22.2?2.231 3.?I17.31Figure 10. Effect of Exposure Time on Tubing Consistency. (a) refers to thesolvent blank, (b) refers to an 8 hour exposure, and (c) refers to a 144hour exposure.2.2317.44(c)2.23: fl:3.ON 3’,: START FINAL rL.L I1 .655.2.3.2 Effect of Solvent Replacement on Tubing ExtractivesTo ensure that contaminating additives were sufficiently removed from tubingsegments during the 24 hour batch extraction, the hexane in the extraction containers wassuccessively replaced after 24, 48, 72 and 96 hours. The hexane acquired for eachtubing segment for each exposure time was then analyzed and compared to solvent blankswhich had been exposed for the same length of time under identical conditions, butwithout tubing.All the solvent blanks were very clean, as shown by their gas chromatograms.Figure 11(a) shows an example solvent blank chromatogram from a 96-hour exposure.Figures 11, (b) through (e), are chromatograms of the 24 hr, 48 hr, 72 hr and 96hr samples, respectively. The extraction samples showed many compounds for the first(24-hour) exposure, but following the first hexane replacement, very little additionalmaterial was extracted. One can thus conclude that 24 hours is sufficient for clean-up ofthe tubing by solvent extraction. Furthermore, the results of this experiment verify theresults of the previous solvent exposure work (Section, in which the issue oftubing dissolution was investigated. Clearly, the polyethylene did not dissolve in thisexperiment either, since the gas chromatograms did not display progressively increasingnumbers of peaks, but rather decreasing numbers of peaks.66(c)(d)(e)ifjLjLiFigure 11. Effect of Solvent Replacement on Tubing Consistency. (a) refers to thesolvent blank, (b) refers to the first (24 hour) hexane exposure, (c) tothe second (48 hour) exposure, (d) to the third (72 hour) exposure and(e) to the fourth (96 hour) exposure.j(a)(b)167A GC/MS analysis was performed on the hexane from the first 24 hour exposureto identify the contaminants extracted. Many hydrocarbons were found, as evidenced bythe prevalentC4H97ion in the mass spectra (57 amu). Although not all the gaschromatographic peaks could be identified by the GC/MS database, almost all of thespectra contained this ion, which is characteristic of hydrocarbons. Possibly thehydrocarbons were compounds comprising a lubricant or other plastic additive. Thecompounds identified are listed in Table 11.Table 11. Extractives from Hexane Batch Extraction, 24 hour Exposure (SampleE24).toluene4, 6-dimethylundecane2,2 ,6-trimethyldecane2-methyltetradecane2, 6,6-tnmethyloctane2,2, 3-trimethylnonane2 ,2-dimethyldecane2,5-dimethylundecane2,6-dimethylundecane2,2,7-trimethyldecane3 ,4-dimethyldecane2,2,4,10,12, 12-hexamethyl-7-(3 ,5 ,5-trimethylhexyl)-tridecanepropanoic acid, octyl ester2, 3-dimethylnonane2,2,4,15,17, 17-hexamethyl-7, 12-bis-(3 ,5 ,5-trimethylhexyl)-octadecane2,7,7-trimethyldecane2, 6-bis (1, 1-dimethylethyl)-4-ethylphenol5.2.4 Identification of Crystalline Material in Solvent ExtractsIn all solvent/batch extraction work with unextracted tubing, a white crystalline solidwas observed in concentrated extracts. This material was speculated to be a mixture of long-68chain, high-boiling hydrocarbons which, due to their length, would form crystals whenconcentrated in a small volume of solvent (super-saturation).After solid-probe analysis by GC/MS, the compounds comprising the solid materialwere identified as follows in Table 12.Table 12. Compounds Identified as Comprising the Solid Crystalline Material inSolvent Extracts.Compound No. of Carbons(e)-3-octadecene 8cyclododecane* 121dotriacontano1* 321 -hentetracontanol 416-cyclohexyl-6-cyclohexyl-dodecane 12(e)-5-eicosene 20*indicates most prominent constituents of crystalline materialAs shown in the table above, the compounds were long-chain compounds as expected.The compound 1-dotriacontanol is a C32 alcohol, 1-hentetracontanol is a C41 alcohol and (e)5-eicosene is a C20 alkene. This may explain their crystallization in hexane extracts.5.2.5 Characterization of Glove ExtractsBecause polyethylene gloves were used for handling the SPMDs, it was decided todetermine the types of compounds which could be extracted from the gloves, especially sincethe gloves were often used in the presence of solvents. Two exposures were performed: one69in which a glove was dipped into hexane for 5 minutes, and one in which a glove wassubmerged for 48 hours. From the gas chromatograms, there appeared to be manycompounds being extracted from the gloves in both exposures (see Appendix A forchromatograms). GC/MS analysis was carried out on the 5 minute exposure sample becausethat was deemed to be a typical exposure time during SPMD preparation and manipulation.Table 13 shows the compounds identified.Table 13. Compounds Identified in Polyethylene Glove Extracts.i-butyl-2-(2-methylpropyl) cyclopropane2 ,4-dimethyleicosanel-(l-methylethyl)-2-(2-methyl- 1-methylenepropyl) cyclopropane3-ethyl-i ,4-hexadiene3 .alpha. -methyl-cis-hexahydrophthalide1 ,2-dithiane4-methyl-benzenepropanol2,3 ,5-trimethyldecane1,1 ‘-(6-methoxy-2,5-benzofurandiyl) bis-ethanonedecylcyclohexane2,3, 8-trimethyldecaneheneicosylcyclopentanei-chloro-3-(phenylethenyl) benzeneBecause of the obvious source of possible contamination the gloves presented,handling of SPMDs with gloves was minimized in all successive experiments.5.2.6 Discussion of Method Development WorkThe solvent exposure component of the method development work, in particular,warrants some discussion. The major problem encountered during this portion of the study(batch extractions, crystalline material characterization and glove exposures) was the70multitude of extractable compounds which were identified. There did not appear to be anyconsistency among the compounds in the various solvent samples. Specifically, one wouldhave expected the compounds present in the crystalline material to also have been present inthe simple tubing extraction in hexane, since both were derived from the same source. Ifprecipitation had been occurring to form the solid, there would still have been a proportionof the compounds in solution. One consideration is that in all extraction work, the tubingwas taken directly from the roll, which was stored in the laboratory under less thancontaminant-free conditions and was not certified additive-free by the manufacturer.Therefore, one would expect a certain degree of contamination. Furthermore, there aremany additives such as lubricants, extruders and antioxidants which are present in plastics.Of particular note is 2, 6-bis-(l, 1-dimethylethyl)-4-methylphenol, known as BHT, one of itstrade names. BHT was identified in the hexane tubing extract, corresponding to the largepeak at 17.44 minutes on the gas chromatogram (Figure 10). This compound is anantioxidant which is commonly added to plastics in order to minimize potential polymerdegradation by oxidative processes (Vargo and Olson, 1985).One important observation in the GC/MS work on tubing extracts was that thecompounds identified were not pulp-mill relevant. In other words, they were not deemed tobe “typical” pulp mill effluent constituents. This is useful information, as it enabled thesecompounds to be ignored if they appeared in future experiments. It must be noted, however,that the identification of compounds by mass spectrometry is not always absolute andtherefore cannot be taken as positive evidence of the presence of a particular chemical.Since it was not the goal of this research to thoroughly probe the behaviour of71polyethylene in solvents, no further work was carried out to identify more positively thecompounds found in the extracts, and no additional experiments were performed to furthercharacterize the potential dissolution or degradation processes. It. a goal, however, toestablish the effects of various parameters on the batch extraction step in order to carry outthe procedure most effectively. In addition, a determination of the effect of variations intubing segment location and thus potential permeability differences on uptake was desired.Both these goals were accomplished, with the following conclusions reached:(1) The polyethylene characteristics throughout the roll of tubing appeared tobe constant; therefore, no further characterization was required.(2) Dissolution or degradation of the polyethylene tubing in hexane was notevident.(3) An extraction period of 24 hours was sufficient for adequate additiveremoval from polyethylene tubing.(4) Polyethylene glove use should be minimized due to the number of solublecompounds which can be extracted.(5) Hexane extracts of polyethylene tubing did not contain pulp mill-releventcompounds.725.3 Model Compound Work5.3.1 Determination of K0,s by HPLCOctanol-water partition coefficients (K0 or P) for the three model compounds and2,2’ ,5 ,5 ‘-TCB were determined by high pressure liquid chromatography (HPLC). Thereis a correlation between the log of the retention times of organic chemicals on apermanently bonded reverse-phase (C18) HPLC column and their log K0s (Veith et al.,1979). Chiou (1985) has reported that the equilibrium partition coefficients of variousorganic compounds were similar in both octanol-water and triolein-water systems. Thisimplies that knowledge of the K0 of a compound would enable the prediction of uptakebehaviour into SPMDs.The retention time of the unretained compound (methanol) was 1.11 minutes.Capacity factors were calculated using the following equation:k= (trto)/to (9)where tr is the retention time of the test compound and to is that of methanol.The UV maxima were presented in Materials and Methods. The retention times,capacity factors (k), logs of capacity factors (log k) and logs of partition coefficients (logP) are given in Table 14 for the reference compounds.A plot of log k vs. log P is given in Figure 12. Using regression analysis, thefollowing linear relationship was established:log k = O.263(log P) - 0.572 (10)Table 15 shows the pertinent data as in Table 14, as well as the calculated K0values for the unknown compounds.73Table 14. Pertinent Data for K0 Determination: Reference Compounds.Compound Ret.time k log k log P(minutes)p,p’-DDT 17.73 14.97 1.18 6.361naphthalene 3.36 2.03 0.31 3372fluoranthene 9.04 7.14 0.85 5.52pentachiorophenol 7.27 5.55 0.74 5.27toluene 2.76 1.49 0.17 2.691 Chiou et at., 19822 Hansch and Fujita, 19643 Miller et at., 19854Xieetat., 19845 Fujita et at., 1964Table 15. Pertinent Data for K0 Determination: TCB and Model Compounds.Compound Ret.time k log k log P(minutes)Dehydroabietic 18.33 15.51 1.19 6.70acid2,2’,5,5’-TCB 14.59 12.14 1.08 6.30gualacol 1.40 0.26 -0.59 -0.068aipha-pinene 12.48 10.24 1.01 6.0274C)— Table 15, we can construct a hierarchy of log K0s. Dehydroabietic acidappears to have the highest partition coefficient (6.70), followed by TCB (6.30), alphapinene (6.02) and finally guaiacol (-0.068). This is reasonable considering that TCB andaipha-pinene ase neutral, non-ionizable compounds and would therefore be expected topartition well from water into an organic material. DHA, being an acidic compound,might at first glance be somewhat of an anomaly with its very high log K0. However,at the pH used (2.5), virtually all of the DHA would be in its undissociated form, sinceits PKa is 7.25 (McLeay et al., 1979). It is also a very hydrophobic molecule with atendency to adsorb to surfaces in aqueous environments (Drobosyuk et al., 1982).Therefore, at the low pH at which it was run by HPLC, it is not surprising that itspartition coefficient was determined to be large.Guaiacol’s extremely low K0 is also expected, since it is a polar compound withhigh water solubility, and it is known that water solubility and log K0 are inverselycorrelated. The K0 determined here was lower than values given in the literature forpartitioning into materials other than octanol. These include 1.36 for ether (Lindberg,1958), 0.96 for neatsfoot oil (Doerr and Fiddler, 1970) and 0.30 for mineral oil (Doerrand Fiddler, 1970). Clearly, the difference between the experimentally-derived value andthe literature value for mineral oil partitioning is minimal, although the other literaturevalues show greater discrepancies. The reason for this deviation is unknown; however,since the reference compound pentachiorophenol appeared to behave consistently to theother reference compounds (see Figure 12), there did not appear to be a problem withphenolic compound partitioning.76The log K0 value for 2,2’ ,5,5’-TCB of 6.30 compares well with the literaturevalues of 6.1 (Shiu and Mackay, 1986) and 5.81 (Chiou et a!., 1982), thereby verifyingto some degree the validity of the results determined here for the model compounds.Based on the partition coefficients determined here, one would predict the degreeof uptake of the model compounds to follow this order:dehydroabietic acid > 2,2’ ,5 ,5 ‘-TCB > aipha-pinene > guaiacolSubsequent experiments investigating model compound behaviour in SPMDs wereundertaken in an attempt to confirm this hypothesis.5.3.2 Determination of Dialytic RecoveryThe polymeric film dialysis step for SPMD analysis is important because finalquantification depends on the tendency of compounds to diffuse out of triolein intohexane, the dialytic solvent. Therefore, this step was investigated more thoroughly byincorporating triolein, spiked with the model compounds and TCB, into SPMDs prior todialysis.Triolein was spiked at 80 g TCB/0.91 g triolein (ie. per SPMD) and100 ,g/0.9lg triolein of each of alpha-pinene, guaiacol and dehydroabietic acid. Dialysiswas carried out for 48 hours in hexane. Three SPMDs with spiked triolein were dialyzedfor 48 hours under static conditions (no replacement of hexane, single-stage dialysis),while the dialytic solvents of three more were replaced after 24 hours (two-stagedialysis). SPMDs with unspiked triolein served as controls for both types of exposures.Following rotoevaporation of the hexane dialysates to 10 mL, 1.0 mL aliquots of each77sample were added to GC vials, along with 10 L of the internal standard mix.Recoveries of the test compounds for each type of dialysis are shown in Figure 13.Coefficients of variance for the recoveries of each compound in the replicates were low,ranging from 1.1 % to 7.8%, with the average being 5.1 %.The highest recovery from the spiked triolein was observed for 2,2’,5,5’-TCB, at72.9% for the single-stage dialysis. The other test compounds followed in this order:DHA (57.3%), guaiacol (51.8%) and lastly, alpha pinene (44.3%). Several possibleexplanations for the observed results were examined. As previously mentioned (recallequation (2) in section, molecular size/weight affects the transport rate throughpolymeric membranes (Lieb and Stein, 1969). Higher molecular weight compounds takea longer time to reach the same dialytic recovery as do lower molecular weightcompounds. Furthermore, Lieb and Stein (1969) reported that increasing molecularweight of non-electrolytes is inversely correlated with diffusion rate in non-porouspolymers. However, since only two of the test compounds are true non-electrolytes(TCB and alpha pinene), one would not necessarily expect this molecular weightdependency to hold true in this situation. In fact, the molecular weight hierarchy of thetest compounds is DHA > TCB > alpha-pinene > guaiacol, an order which differssignificantly from the order of dialytic recovery determined in this work.780- a) > 0 0 G)IISingledialysisTwo-stagedialysisFigure13.DialyticRecoveryof ModelCompoundsand2,2’,5,5’-TCB.100• 9O80 7O60’50’40’30-’20-’10-’ 0-i-aipha-pinene_4I__I_—__I2,2’,5,5’-TCBguaiacolDHAAnother explanation for the recoveries observed is based on the polarity orionizability of the test compounds. Since polyethylene is a nonpolar polymer, ionized orpolar compounds tend to have difficulty traversing the membrane, especially since thephase into which they partition is also nonpolar (hexane). Thus, one would expect TCBand aipha-pinene to have the highest dialytic recoveries, since they are the least polar.Dehydroabietic acid is also very hydrophobic; therefore, fairly high recoveries would alsobe expected. However, aipha-pinene unexpectedly exhibited the lowest recovery of allthe compounds. The low recovery of guaiacol was expected due to its polarity.The low recoveries of aipha-pinene, and to a lesser extent guaiacol, are likelyrelated to their volatilities. Both are fairly low-boiling hydrocarbons (boiling points156 C for aipha-pinene, 206°C for guaiacol) with relatively high vapour pressures (400.7Pa for aipha-pinene, 14.69 Pa for guaiacol). Guaiacol is steam-volatile in kraft mills,causing it to be emitted into the atmosphere and producing a characteristic odour in theimmediate vicinity (Wilson and Hrutfiord, 1971). Volatilization may have occurredduring various stages of SPMD preparation, such as mixing the model compoundstandard solution into the triolein prior to pipetting into SPMDs and rotoevaporating theextracts after SPMD removal. Analysis of the rotoevaporated hexane would verify thispotential source of analyte loss.There seemed to be higher recovery of analytes from the two-stage dialyses thanfrom the single-stage, with an average increase of 13.0% and ranging from 7.1% to22.9%. This is not surprising, since as equilibrium is approached, the concentrationgradient across the polyethylene film decreases, and thus the driving force for diffusion80similarly decreases. If the dialytic solvent is replaced, then in order for equilibrium to beregained, more analyte will have to cross the polymeric film, resulting in higherrecoveries. However, the modest increase in recoveries observed with two-stage dialysisdid not warrant the use of that technique for the remainder of the work, due to increasedsolvent usage and sample manipulation. Therefore, for the remaining research, single-stage dialyses were carried out.This work compares favourably to research by Meadows et al. (1993), in which anincrease in TCB recovery from 70.5% with one 48 h dialysis to 86.7% with two 24 hdialyses was reported. However, when the volumetric ratio of dialytic solvent:lipid wasincreased to 40:1 in the single-stage dialysis to equal the two 20:1 samples comprising thetwo-stage dialysis, recovery was increased to 81.4%. These data show the importance ofcumulative volume in determining the effect of solvent replacement.Through the use of internal standards, a measure of the analytical instrumentalvariation was determined. This was calculated to be 4.5% for l,4-dichlorobenzene,3.5% for chloronaphthalene and 3.8% for chioroanthracene. From this, we can concludethat instrumental variation did not contribute significantly to overall experimental error.Since dialytic recoveries of the test compounds were not close to 100%, it wasnecessary to determine whether any of the analytes had remained in the tnolein duringdialysis. To do this, two SPMDs were opened and rinsed out with hexane. The sampleschosen were single-stage dialysis samples, since they exhibited lower dialytic recoveriesand would therefore be expected to contain a higher proportion of the test compounds inthe triolein than the two-stage dialysis samples. It was hoped that a mass balance could81be closed around the two samples, ie. total analyte spiked into lipid = analyte indialysate + analyte remaining in lipid. Due to the procedural complexity of DHAanalysis, it was decided not to quantify DHA in the rinsate and rather target the otherthree compounds.Only one of the analyzed samples is shown (Figure 14), since the profiles werevirtually identical. The results indicate that, in fact, a significant proportion of eachanalyte (6.4% for aipha-pinene, 4.5% for guaiacol and 8.8% for TCB, given as averagesof 2 replicates) remained within the triolein and did not partition out into hexane.Unfortunately, however, a mass balance could not be closed around the system as washoped, with the maximum cumulative recovery being that of TCB, at 86.8%. Thisapparent low analyte recovery will be discussed in more detail in Section G) > 0 0 a)Figure14.Distributionof TestCompoundsinDialysateandRinsate.aipha-pinenegualacol2,2’,5,5’-TCBIIDialysateRinsateOne of the laboratory blanks (unspiked triolein, two-stage dialysis) was furtheranalyzed by GC/MS. This analysis was carried out for two reasons. First, it was ofinterest to investigate the types of contaminants present in the blank. The tubingincorporated into the SPMD used here had been pre-extracted in hexane, and thus wouldnot be expected to show the same degree of contamination found in the methoddevelopment solvent extraction work. However, there was a higher degree of manualmanipulation inherent in this SPMD exposure, implying the possibility of increasedcontamination. Second, it was necessary to assess the potential for lipid carryover, aphenomenon recently investigated by Meadows et al. (1993). In that study, it appearedthat a small amount of triolein made its way through the polyethylene film and out intothe dialytic solvent. Table 16 shows the compounds identified in the present study (seeAppendix A for the GC/MS Total Ion Count).Table 16. Compounds Identified in Dialysis Laboratory Blanks.2-methyl-2-propenoic acid, octyl esterisocyanatocyclohexane1 ,4-dichlorobenzene (internal standard)nonanal1 -chloronaphthalene (internal standard)1-chloroanthracene (internal standard)(z)- 1 3-docosenoic acid, methyl ester(z)-15-tetracosenoic acid, methyl estern-methyl-n-( 1 -oxododecyl) glycine4-methyl-pentanamide(z)—3-tetradecen-5-ynehexanedioic acid1 -dotriacontanol3-nitro- 1, 2-benzenedicarboxylic acid(ethoxymethoxy) cyclohexane2 ,4-diethyl-5 , 6-dimethyl- 1,3 ,2-dioxaborinane84There appeared to be no overlap between the contaminant list identified for thetriolein (Table 10) or the tubing (Table 11) and that identified for the laboratory blank(Table 16). The most likely source of the contaminants is the high degree ofexperimental manipulation involved in SPMD use. This includes contamination fromglassware, addition of spiked solutions to triolein, rotoevaporation etc. However, the twofatty acids (z)-13-docosenoic acid and (z)-15-tetracosenoic acid may have been lipidcontaminants or degradation by-products. During the course of this work, a sticky filmwas observed on the outside of SPMDs upon their recovery from exposure to an aqueoussample. This phenomenon was also reported by Petty et al. (1992), who suggested thatthe film may be comprised of (a) methyl ester(s). They noted that the material was nottriolein, but that certain lipid by-products could traverse the polyethylene membrane in asimilar fashion to the lipid carryover described earlier. Furthermore, Lebo et al. (1992)reported lipid by-products such as methyl oleate (the methyl ester of oleic acid) in SPMDdialysates. According to Huckins et al.(1990a), fatty acids represent a portion of themembrane-permeable lipid present in analyte dialysates, leading to speculation that thetwo fatty acids found in the laboratory blank were attributable to lipid degradation orcontamination. The source of these by-products might have been photooxidation oftriolein, which is known to be light-sensitive (from Material Safety Data Sheet).If lipid carryover was occurring, one would expect to find significant quantities ofoleic acid and glycerol, the constituents of triolein, in the dialysate. Derivitization offatty acids is required for their detection on the gas chromatograph, but since thisprocedure was not carried out, the presence of oleic acid was not confirmed. Glycerol85was not evident in the dialysate, indicating that lipid carryover did not occur.The dialytic recovery work illustrated some problems associated with the use ofpolymeric film dialysis for the analysis of the model compounds used in this research.However, the technique should not be completely abandoned, since there were indicationsthat the low recoveries observed may have stemmed from a lack of detection, rather thana lack of recovery. More discussion of this will follow in Section Determination of Total UptakeThe next stage of the model compound study attempted to quantify the uptake ofTCB and the model compounds into SPMDs from the surrounding aqueous environment.Since quantification was via dialysis into hexane, this work involved both partitioning jtriolein, as well as dialysis out of triolein. The inclusion of an additional step presentedan additional opportunity for analyte loss; therefore, one might expect analyte recoveriesto be lower than determined in the dialytic recovery experiment described earlier.SPMDs were exposed to 10 g1L and 100 gIL solutions of each of the testcompounds in distilled water (pH 7) for 1, 2, 4 and 7 days. Duplicates of eachcombination of concentration and exposure time were included to provide some measureof the variability in the experiment. Laboratory blanks for each exposure time, consistingof SPMDs in distilled water only, were also provided. As with the dialytic recoveryexperiment, internal standards were added to the final 1.0 mL volume of concentrateddialysate in GC vials. The percent uptake of each compound into the SPMDs is shown inFigures 15(a), (b) and (c). Note that guaiacol was not detected in any of the dialysates.86100- 90-80-70-6OExposureTime(days)•loppbexposure‘E100ppbexposureFigure15(a).Uptakeof 2,2’,5,5’-TCBintoSPMDs.100- 90-80-70-6O-50-(U40-D30- 0ExposureTime(days)—10ppbexposure100ppbexposureFigure15(b).UptakeofAipha-PineneintoSPMDs.100- 90 80 706O50 40DExposureTime(days)—10ppbexposure100ppbexposureFigure15(c).UptakeofDehydroabieticAcidintoSPMDs.One can conclude that TCB was sequestered most efficiently (44.5% after 7 days@100,.g/L), followed by DRA (19.6%), alpha-pinene (12.2%) and fmally guaiacol,which was not taken up at all. This order differs from that determined in the dialyticrecovery experiment (TCB > DHA > guaiacol > aipha-pinene), although only in the orderof the last two compounds. Recall from Section that polarity of substituentgroups strongly affects compound uptake. The order of increasing polyethylenediffusivity (Roff et al., 1971) is alcohols, acids, nitro-derivatives, aldehydes, ketones,esters, ethers and finally hydrocarbons. This order correlated well to the order of testcompound uptake, with the exception of alpha-pinene.Guaiacol, having significant polarity, was not expected to be taken up well, andthis was indeed observed. According to Huckins et al. (1990a), even moderatelylipophilic compounds (K0 between 0.1 and 2 x 10) should be sequestered to somedegree. Guaiacol’ s K0 was determined to be 0.86; therefore, on the basis of octanolwater partition coefficients, it should be almost completely excluded from the SPMD.Furthermore, guaiacol’s pKa is 9.98 (Dictionary of Organic Compounds, 1982), whichmeans that at the neutral pH at which the experiment was run, only about 0.1 % would beunionized, with the remainder being charged. Consequently, the poor uptake of guaiacolwas not surprising, since ionized compounds are not absorbed efficiently into SPMDs.Aipha-pinene is also somewhat of an anomaly, since according to both log Ppredictions and the rankings given by Roff above, it should have been taken up easily.Again, based on its vapour pressure, it is likely that much of it was volatilized duringstirring of the exposure containers or rotoevaporation.90Dehydroabietic acid was taken up fairly well, but its absorption did not reflect itshigh log K0 determined in Section 4.3.1. Based on that value, it should have exhibitedthe highest uptake. Again, this was probably a result of the pH at which the experimentwas performed. The K0 determination was performed at a very low pH, at whichvirtually all the DHA was nondissociated. In the total uptake experiment, performed ataround pH 7, only about 60% of the DHA was unionized, with 40% in the acidic form(McLeay et al., 1979). In effect, this means that only 60% of the DHA was availablefor uptake.The TCB behaved consistently in all three of the model compound experiments. Itwas predicted to have been easily sequestered by its high log P, and in fact, it was themost easily taken up of the four test compounds. Furthermore, it had the highest dialyticrecovery, correlating well to the other two results. However, the total uptake of TCBinto SPMDs at 10 1hg/L for 7 days was only 52.5%, whereas in the permeabilityexperiment using TCB (Section 4.2.2), under identical experimental conditions, anaverage recovery of 87% was observed. This discrepancy in recoveries is difficult tointerpret. Possibly the presence of the other analytes in the total uptake experiment insome way interfered with the uptake of TCB. Alternatively, it is possible, althoughunlikely, that there was a variation in the permeability of the tubing used between the twoexperiments. Different rolls of tubing were used for the two experiments, since the rollused for the permeability work had been somewhat fragmented by taking segmentsthroughout the roll. Since the permeability study had shown that tubing permeability wasconstant, the study was not repeated when using the next roll of tubing for subsequent91work. Therefore, it is possible that the permeability differed significantly from the rollused to generate the 87% value.The general trend displayed in Figure 15(a), (b) and (c) is one of increasinganalyte uptake with time. However, in all 3 graphs, there is a noticeable decrease in the4-day uptake value. Although care was taken to be consistent, this could perhaps beattributed to a handling discrepancy on day 4, especially since the decrease occurs for allthree analytes at both exposure concentrations. Note that according to the uptake curvefor TCB, 7 days did not appear to be sufficient time to reach steady-state concentration,since uptake was still increasing at that time. This lends more credibility to theconclusion of constant tubing permeability reached in Section 5.2.2, since if time toequilibrium is actually greater than 7 days, differences in permeability would have beenevident after a 7 day exposure period.It appeared that for TCB, and to a lesser extent alpha-pinene, the lower exposureconcentration (10 Lg/L) resulted in higher uptake efficiency than the higher concentration(100 gIL). Since aqueous concentrations should not affect the proportional uptake ofcompounds, the observed result is difficult to explain. If any one of the compounds hadbeen at concentrations above its maximum solubility, then there would have been aproportion of the analyte not in its dissolved form. In fact, 2,2’,5,5’-TCB has a watersolubiity of 30 g/L (Huckins et al., 1990a); therefore, at a 100 g/L level, a proportionof the TCB would not have been in dissolved form and may not have been available foruptake into SPMDs (only dissolved material can traverse the membrane). Similarly,alpha-pinene is virtually insoluble in water. Still, one would expect that eventually92equilibrium would have been established, and the relative uptake at each concentrationwould have been similar. However, due to the low solubility of both TCB and alphapinene, a portion of each may have been adsorbed to the container walls, thus making thecompounds unavailable for SPMD uptake.In addition to these conventional SPMDs, the experiment was conducted at 100Lg/L for 7 days with SPMDs containing no triolein (tubing only), and SPMDs with thetriolein concentrated in one place, rather than spread into a thin film. Polyethylene is anonpolar organic carbon matrix; therefore, incorporation of hydrophobic contaminantsinto the polymer might be expected. TCB has a membrane-water partition coefficient(log Kmw) of 4.6, which, although lower than its K0 and K, would still predict theaccumulation of compounds within the matrix. Polypropylene, a similar polymer topolyethylene, has been used as an integrative sampler for polynuclear aromatichydrocarbons in the aquatic environment (Black et al., 1982). It is thus conceivable thatcertain compounds would simply adsorb onto the surface of the polyethylene or enter thepolymer matrix. The purpose of the triolein-less SPMDs, then, was to determine if thiswas occurring. The purpose of the SPMDs with the triolein concentrated in one placewas to determine the effect of lipid-occupied SPMD surface area on uptake. The resultsof both types of non-conventional SPMDs are displayed in Figure 16, with uptake valuesshown as averages of two duplicates.93IITubingonlyTrioleinnotspread100 90 80 70 600-50C’,4J a D30a20 10•Alpha-pinene2,2’,5,5’-TCBDHAFigure16.TestCompoundUptakeintoNon-Conventional SPMDS.From Figure 16, it is obvious that the test compounds, especially TCB, weresequestered fairly efficiently into or onto the tubing itself. TCB was recovered at 46.7%,DHA at 12.6% and aipha-pinene at 12.9%. Work by Huckins et a!. (1990a) utilizingradiolabeled compounds showed that 25-45% of the analytes (depending on thecompound) were present in the polymer matrix after 21 day exposures. One wouldexpect a high proportion of the analytes to be present in the membrane itself, since thediffusion of non-electrolyte compounds through non-porous polymers has been reported tooccur slowly (fiwang and Kammermeyer, 1975). According to Huckins et al. (1990a),only 2% of the membrane-associated residues of mirex were present on the exteriorsurface of the SPMD after a 4-day exposure, thus implying that the analyte detected inthe tubing-only dialysates was largely present within the matrix as opposed to adsorbed tothe surface.Once again, TCB was the most largely recovered analyte from the tubing. Thiswas expected, due to its hydrophobicity and consequent affinity for the polyethylene.Dehydroabietic acid and alpha-pinene recoveries were lower, DHA probably because ofits polarity, and alpha-pinene probably due to loss through volatilization. Guaiacol wasnot recovered at all, again not surprisingly, since it diffuses very poorly in polyethylene(Roff et a!., 1971). Interestingly, the recoveries of all 3 detected compounds were nearthe corresponding values for conventional SPMDs exposed under identical conditions(100 g/L for 7 days). This means that the tubing alone sequestered the test compoundsto the same degree as the triolein-fihled SPMDs.Similar results were observed for the SPMDs in which the triolein was not spread95into a thin film but instead concentrated in one place (Figure 16). In fact, the recoveriesfor these samples were the highest among all the samples exposed, including bothconventional SPMDs and tubing-only SPMDs. TCB recovery was 52.9%, comparedwith 44.5% for the lOOg/L, 7 day exposure of a conventional SPMD. Likewise, alphapinene was recovered at 16.6%, higher than the conventional SPMD recovery of 12.2%.Finally, DHA had a recovery of 19.4%, virtually identical to the 19.6% observed withthe usual method. These results are rather surprising, since one would expectconventional SPMDs to be more efficient in sequestering contaminants due to their largersurface area of lipid-occupied membrane. However, in light of the observed affinity ofpolyethylene for the test compounds, particularly TCB, it appears that the tubing itselfsequesters a large proportion of the analyte, and then the small area of triolein-occupiedtubing sequesters some more, which results in higher uptake than regular SPMDs.As in the dialytic recovery experiment, it was decided to attempt to close a massbalance around the uptake step by measuring the analyte remaining in the aqueousexposure containers. It was hoped that since the recoveries of the compounds from theSPMDs were low, the remaining analyte would be found in the exposure water. Only 5samples were chosen, and they were samples which had exhibited low SPMD uptakeefficiencies, leading to the speculation that there would be higher levels of the testcompounds remaining behind in the exposure water.Each 1 L sample was first extracted with hexane to sequester alpha-pinene andTCB. Then, guaiacol was acetylated with acetic anhydride (method given in Materialsand Methods section) and the water sample extracted again with hexane. For simplicity,96DHA analysis was not performed. The results of the analysis of both the SPMDdialysate and the aqueous phase are given in Figures 17(a), (b) and (c). Thenomenclature of the samples is as follows:100/1 day/i = 100 ppb for 1 day, sample #1 of 2100/1 day/2 = 100 ppb for 1 day, sample #2 of 2100/2 day/i = 100 ppb for 2 days, sample #1 of 2100/4 day/2 = 100 ppb for 4 days, sample #2 of 2tubing 1 = tubing only (no triolein), sample #1 of 2Guaiacol was present in significant quantities in the aqueous samples, up to50.2% of the spiked amount (recall that no guaiacol was measured in any of the SPMDdialysates). TCB was recovered at a higher percentage from the SPMDs themselves,having no measurable residue present in 3 of the aqueous samples. Similarly, moreaipha-pinene was located in the SPMDs than in the water. The results of these aqueoussample extraction were expected, since TCB has a very low aqueous solubility, whileguaiacol is quite soluble in water. However, two of the samples contained no measurablelevel of aipha-pinene whatsoever (in either the SPMD or the water). The lack of alphapinene in the samples can probably once again be attributed mainly to volatilization,especially in the water samples, where venting of the separatory funnel could easily haveallowed the escape of a large proportion of the aipha-pinene volatilized by vigorousshaking during extraction. Volatilization during rotoevaporation has already beenmentioned as a possible additional mechanism of analyte loss.97as 4-I 0 4-’ 0 o C 0 I100-- 90. 80 70 60-’ 50- 40- 30-v 20- 10-p 0.00—I__l100/1day/i100/2days/itubing1100/1day/2100/4days/2SampleIISPMDAqueousphaseFigure17(a).Distributionof 2,2’,5,5’-TCBinSPMDsandAqueousSamples.SampleIISPMDAqueousphaseas 0 4-’ ‘I 0 0 0 I100-’ 90-p 80 70 60 50 40 30 20 10• 0100/1day/iSC SCI100/1day/2100/2days/i100/4days/2-711tubing1Figure17(b).Distributionof Aipha-PineneinSPMDsandAqueousSamples.SampleIJSPMDAqueousphase4-J 0 I100- 90- 80- 7O- 6O- 5O- 40- 3O 20 1o.C CjI I100/iday/iI1OÔ/2days/1tubing1100/1day/2100/4days/2Figure17(c).DistributionofGuaiacol inSPMDsandAqueousSamples.Unfortunately, once again a mass balance could not be closed on the system. Ofthe 5 samples which underwent both dialysate and aqueous extraction, the maximumcumulative recovery was that of TCB, at 46.9%. Huckins et al. (1990a) reportedsomewhat similar findings in a radiometric study also using 2,2’,5,5’-TCB.Approximately 55% was found in the triolein, 27% in the membrane matrix, 0.1% on theexternal surface of the SPMD, 0.1 % in the water and 0.3% on the container surface,resulting in a net total of only 82.52%. Although this is higher than the total recoverydetermined in the present work, it nevertheless represents a loss of TCB during theprocedure.Mention should be made as to the variability of the data in the total uptakeexperiment. In general, the variability in the peak areas for each set of two replicateswas low, with the average coefficient of variance for aipha-pinene, TCB anddehydroabietic acid measurements being 15.6%, 19.4% and 24.4%, respectively.5.3.4 Discussion of Model Compound WorkThe behaviour of the test compounds in this work was undoubtedly influenced bya number of factors. First, the polarity of substituent groups on compounds affects theiruptake into and recovery from SPMDs. The uptake of the test compounds correspondedwell to the ranking of diffusivity in polyethylene, based on polarities, given by Roff(1971). TCB and aipha-pinene, being neutral compounds, were expected to be taken upmost efficiently, with guaiacol and dehydroabietic acid sequestered ineffectively, if at all.There was a concern with poor recovery of both aipha-pinene and guaiacol, which will be101discussed later in this section. Second, the octanol-water partition coefficient of acompound influences uptake and dialysis processes. Guaiacol, with an extremely lowK0, was not present at significant levels in dialysates from either spiked SPMD exposureor regular aqueous exposures. Third, the solubilities of the test compounds appear to berelated to their propensity to be taken up into SPMDs, although this would be expected,given the extensive reports of correlations between K0 and water solubility. Guaiacolhas a fairly high solubility, and was not sequestered well. TCB, on the other hand, isextremely insoluble in water, and exhibited the highest recoveries. Dehydroabietic acidhas low solubility as well, and its uptake reflected this to some degree. Again, alphapinene, with its very low solubility, unexpectedly was not taken up well. However, asmentioned earlier, this may reflect volatilization losses in the measurement procedure,rather than its uptake behaviour. Finally, analyte molecular weight, over the rangestudied here, did not appear to have a significant effect on analyte uptake.Dehydroabietic acid is the heaviest of the test compounds, but did not exhibit the highestrecoveries. This is likely due to its ionizability, which causes a breakdown in theexpected size/weight correlation with membrane transport.A significant problem evident in the model compound work concerned pooranalyte recoveries. Obviously, analyte was either being lost somewhere in the system, orthe analytical recovery was not 100%. One possible explanation, volatilization, hasalready been mentioned with respect to aipha-pinene and guaiacol. The rotoevaporationprocedure, as well as sample concentration under nitrogen, both provide opportunities forlow-boiling compounds such as guaiacol and alpha-pinene to volatilize and be lost. Both102of these compounds have relatively high vapour pressures, and are thus quite volatile.Since they were also the compounds with the lowest total recoveries after rinsate andaqueous sample analysis, volatilization seems to be a logical explanation. Huckins et al.(1990a) suggested volatilization as a possible reason for a poor mass balance around theirTCB system. In a later section of this thesis, Section, additional evidence will beprovided that aipha-pinene is easily volatilized. In light of these findings, it seemsreasonable that a different method for extract concentration should be tried in the future.Clearly, analytical recovery with rotoevaporation is rather low for the more volatilecompounds, even at the low rotoevaporator water temperature of around 55°C which wasused.The less volatile compounds, TCB and dehydroabietic acid, may have adsorbed toglassware container walls due to their hydrophobicity. In fact, this has been previouslydocumented by Droboysuk et al. (1982) for dehydroabietic acid.Another possible sink for the analyte is the polyethylene matrix itself.Compounds may have remained within the matrix during dialysis and therefore were notquantifiable in the dialysate. However, it has been reported by Huckins et al. (1990b)that such losses are negligible. Apparently, <2% of various radiolabeled nonpolaranalytes were present in the polymer matrix following dialysis. Since one of the analytesused in that work was 2,2’ ,5,5’-TCB, it seems unlikely that the membrane was a sink forthat particular compound in the present work. In fact, when one considers the fact that apolar compound such as guaiacol would be even less likely to adsorb to the polyethyleneor remain in the membrane matrix after dialysis, one can conclude that the membrane103matrix probably is not a significant source of analyte loss.Loss of analyte due to sample manipulation is another very likely explanation forthe incomplete mass balances observed, and one also given by Huckins et al. (1990a) forhis observed “missing” analyte. SPMD preparation and analysis procedures are multistep in nature and fairly labour-intensive, thereby presenting many opportunities for bothcontamination as well as analyte loss. Much care was taken to ensure quantitativetransfer of solutions and the preparation of accurate standard solutions; however, therewill always be some degree of experimental error.To verify any of the above proposed explanations for the lack of analyteaccountability, one could break down the process into components and try to close a massbalance around each step. For example, in the dialysis investigation in Section 4.3.2,one could rinse out an SPMD containing spiked triolein with hexane and analyze thehexane to determine if analyte loss was occurring even before deployment of the SPMDin the dialytic solvent, perhaps by volatilization from the spiked lipid. Similarly, asmentioned earlier, one could analyze the rotoevaporated and re-condensed hexane toverify the presence of any analytes there.The conclusions of the model compound work were as follows:(1) Volatilization may have been a contributing factor to the low analyterecoveries observed with both the dialytic recovery experiment and the totaluptake experiment. This may be particularly relevant to the most volatile test104compounds, guaiacol and aipha-pinene.(2) The orders of analyte uptake into SPMDs and dialysis out of SPMDs could beexplained on the basis of analyte polarities, octanol-water partition coefficients andwater solubilities. Molecular weight did not seem to be a factor involved in testcompound uptake or recovery.(3) Two-stage dialyses resulted in higher dialytic recoveries than single-stagedialyses.(4) A portion of each test compound (6.4% for alpha-pinene, 4.5% for guaiacoland 8.8% for TCB) remained within the triolein during dialysis of spiked SPMDs.(5) There may have been some degree of lipid degradation or contaminationoccurring in the triolein, as evidenced by fatty acids in the dialysis blanks.(6) Tubing alone sequestered the test compounds to the same degree asconventional, lipid-filled SPMDs.(7) SPMDs with triolein concentrated in one area showed higher uptake efficiencythan conventional SPMDs.(8) Significant amounts of guaiacol were detected in the residual aqueous exposuresamples, but not enough to account for 100% of the spiked amount. Guaiacol wasnot detected in any of the SPMD dialysates.1055.4 Wastewater Exposure Work5.4.1 Static Wastewater Exposures5.4.1.1 Introduction and Goals of WorkThe results of the model compound work were encouraging enough to warrantfurther investigation of SPMDs with respect to pulp mill effluent. Thus, SPMD effluentexposures were carried out in an attempt to address the following issues:(a) what types of compounds, if any, SPMDs absorb from pulp milleffluent;(b) the effect of wastewater treatment on SPMD uptake;(c) the effect of wastewater aeration on SPMD uptake;(d) the effect of exposure time on SPMD uptake; and,(e) what significance, if any, the work has with respect to the eulachontainting problem at Eurocan Pulp and Paper Co.Each of these issues will successively be addressed in the following sections.The procedure, given in detailed form in Materials and Methods, consisted ofexposing SPMDs to 100% biologically treated or untreated unbleached kraft mill effluentfor 1, 3 and 9 days. Half the samples were aerated. The other half were unaerated, butwere stirred daily to provide some mixing. After the designed exposure time, theSPMDs were dialyzed and dialysates concentrated and subjected to GC analysis. Selectedsamples were analyzed by GC/MS.GC/MS identification of the compounds present was rather difficult due to the106complexity of the gas chromatograms. Furthermore, one must keep in mind that massspectrometry only tentatively identifies compounds. Throughout Section 5.4, moreconsideration should be given to specific peaks and their variability/distribution undercertain conditions, rather than the actual identity of the peaks, although that will bediscussed as well.Nomenclature of the samples which will be discussed in this section is as follows:untreated wastewater = I; treated wastewater = Eunaerated wastewater = NA; aerated wastewater = Aexposure time = 1, 3, or 9 days5.4.1.2 Uptake of Pulp Mill Constituents into SPMDsFrom the gas chromatograms generated, it was obvious that the SPMDs absorbednumerous compounds from the wastewater samples. Figure 18 shows 2 gaschromatograms, one from a blank (SPMD in distilled water only) and one from an SPMDexposed to untreated, non-aerated effluent for 9 days. The multitude of peaks in the 9day exposure sample indicates the high sequestering ability of the SPMDs.The distribution of peaks was such that there were higher peak areas at longer GCretention times and lower peak areas at shorter retention times, likely because latereluting compounds are less volatile. This would tend to make them more lipophilic andtherefore more easily sequestered by SPMDs.A list of compounds identified by GC/MS in one of the treated, unaeratedwastewater samples (sample E-NA-9-1) is presented in Appendix C. That particularwastewater sample was deemed to be the most representative of Eurocan final effluent.A shorter version of the list, consisting of only the larger peaks, is shown in Table 17.107(a)(b)Figure 18. Gas Chromatograms of Dialysates from (a) Blank (SPMD in DistffledWater Only), and (b) SPfvID Exposed to Untreated, UnaeratedWastewater for 9 days.131108Table 17. Partial List of Compounds Identified in Sample E-NA-9-1.isocyanatocyclohexane2-methyl-5-(1-methylethyl)-bicyclo [3.1.0] hex-2-ene1 ,2-dimethylcyclopentane1 -methyl-4-(1-methylethyl)-cyclohexanel ,7,7-trimethyl-(. + I-. )-bicyclo [2.2.1]heptan-2-onealpha., alpha. 4-trimethyl-3-cyclohexene- 1-methanol.alpha., 4-dimethyl-benzenemethanol2,3,4,7,8, 8a-hexahydro-3 ,6, 8 ,8-tetramethyl- lh-3a,7-methanoazulene1 -methylene-3-(1-methylethylidene)-cyclopentane6-ethenyl-6-methyl- 1-(l-methylethyl)-3-(1-methylethylidene)-(s)-cyclohexene7,11 -dimethyl-3-methylene-(e)- 1,6, 10-dodecatriene4-(l , 5-dimethyl- 1 ,4-hexadienyl)- 1 -methyl-cyclohexene1,2 ,4a,5 , 8, 8a-hexahydro-4 , 7-dimethyl- 1-(1-methylethyl)-naphthalene2,6-dimethyl-6-(4-methyl-3-pentenyl)-bicyclo [3.1.1] hept-2-ene1,2,3,4 ,4a,7-hexahydro- 1, 6-dimethyl-4-(1-methylethyl)-naphthalene1 -isopropyl-4 , 5-dimethyl-cyclopentene2-ethylidene- 1, 1-dimethylcyclopentane1, 3-dimethyl-tricyclo [,7] decane2-(fench-2-yl)-fenchanebis(2, 2-dimethyl-3(z)-(1-methylprop-2-enyl)-cyclopropyl) zinc(e)-3-tetradecen-5-yne22-ethyl- 15,1 6-dimethoxy-4 ,25-secoobscurinervan-4-ol-25-acetateFrom the table above, it appears that numerous compounds were taken up by theSPMDs, and that the majority of these compounds were hydrocarbons of some type.Ionizable or poiar compounds do not appear on the list, with the exception of twoalcohols. This is not surprising, based on the model compound work reported in Section5.3.Other notable compounds found in the dialysate and listed in Appendix C includeisocyanatocyclohexane, a blank constituent, and (z)-15-tetracosenoic acid, methyl ester, afatty acid present in the dialysis blank (Section 5.3.3). There were also two metalliccompounds identified, a zinc compound and a silver salt.109Interestingly, when comparing the contents of Table 17 with the brief overview ofkraft effluent composition given in Section 2.4.2 (Tables 2, 3 and 4), there does notappear to be significant overlap. However, one must keep in mind that the compoundslisted in Section 2.4.2 are only representative compounds taken from the hundreds ofcompounds present in pulp mill wastewaters. Effect of Wastewater Treatment on SPM1) UptakeEurocan Pulp and Paper Co. utilizes secondary treatment via two settling pondsand an aerated stabilization basin. One would expect to observe a general decrease in thenumbers and sizes of peaks in the treated effluent SPMD dialysates as compared tountreated samples. This was indeed the case. The chromatograms of the treated effluentSPMD dialysates were cleaner than the untreated effluent samples, as shown in Figure19(a) and (b), chromatograms of 9 day, aerated exposures to both untreated and treatedwastewaters, respectively. Figure 19(c) is the 9 day, aerated blank.Because of the number of compounds present, it was not feasible to assess eachpeak for area reductions. Therefore, in order to quantitatively determine the effect ofbiological treatment on SPMD uptake, several compounds from the gas chromatogramswere selected randomly for analysis across the retention time profile.There appeared to be a reduction in individual peaks areas in the treated effluentexposures. This is illustrated in Figure 20, which displays GC peak areas for SPMDexposures to untreated and treated, non-aerated wastewater for 1 day. The selectedcompounds are identified only by their GC retention times, due to the length of some ofthe compound names. The identity of some of these compounds will be discussed later.110(a)(b)(c)Figure 19. Effect of Wastewater Treatment on SPMI) Uptake. Shown are gaschromatograms of SPMD exposures to (a) untreated aerated and(b) treated aerated wastewaters for 9 days. Chromatogram (c) is anSPME) blank.I111700600D500400300ifiLI1Ii[,LF,hi,18.6735.1343.0747.1256.3361.1524.4839.4543.5449.2160.2762.89RetentionTimeofCompound(minutes)UntreatedEffluent IITreatedEffluentLaboratoryBlankFigure20.Effectof WastewaterTreatment onSPMDUptake(SamplesE/I-NA-i).Due to their presence in the laboratory blank, we can speculate that thecompounds eluting at 18.67 minutes and 56.33 minutes were contaminants resulting fromthe tubing or the method. In fact, GC/MS analysis identified the compound with aretention time of 18.67 minutes to be isocyanatocyclohexane and that with a 56.33 minuteretention time to be 7-tetradecyne. Recall that isocyanatocyclohexane was also detectedin the dialysis laboratory blank (Section 5.3.2).The reduction in peak area after biotreatment was even more evident in the 9 dayexposure, shown in Figure 21. This phenomenon is most likely due to the higher uptalcewhich resulted from the extended exposure time used. Also, higher levels of organicswere present in the effluent prior to treatment, and the likelihood of baseline levels ofcompounds remaining after treatment was high; therefore, reduction in uptake would bemore dramatic with longer exposure times. Due to the number of compounds in the 9day exposures, Figure 21 is only a partial display of data, with a more complete list inAppendix B. The compounds eluting at 60.27 mm. and 61.15 mm. were reduced by78% and 100%, respectively. These compounds were identified as pentacyclotetradecaneand l,2,3,4,4a,9, lO,lOa-octahydro-l,4a-dimethyl-7-(l-methyl)-l-phenanthrenecarboxaldehyde, respectively. Note that isocyanatocyclohexane (18.67minutes) was present at similar levels in the blank extracts and in the wastewaterexposures; therefore, it was obviously not an effluent constituent.In terms of compound reduction distribution, it appeared to be fairly consistent.That is, differences in compound uptake with or without treatment seemed to occurevenly over the range of GC retention times used.113(I) 4-’ ci) as G) as 0) 02013421000 900800700600500400300200100 0.EHEI.18.6735.1343.0747.1256.336124.4839.4543.5449.2160.27RetentionTimeof Compound(minutes).1 62.89UntreatedEffluent 11TreatedEffluentLaboratoryBlankFigure21.EffectofWastewaterTreatmentonSPMDUptake(SamplesE/I-A-9).Eurocan’s aerated stabilization basin has a hydraulic retention time of 5 days;therefore, treated and untreated wastewater samples taken simultaneously do not representthe same initial wastewater. However, since the composition of pulp mill effluent tendsto be fairly constant, barring a black liquor spill or other episodic event, a qualitativecomparison can still be made. Treatment of wastewater appears to decrease uptake intoSPMDs, both by decreasing the numbers of available compounds absorbed and bydecreasing the levels of individual compounds sequestered. Effect of Wastewater Aeration on SPMD UptakeAeration was carried out during SPMD exposure to avoid the establishment ofanaerobic conditions which might cause compound modification. The growth ofanaerobic bacteria tends to cause the production of acids and TRS compounds such ashydrogen sulfide. In this way, the exposure of SPMDs to unaerated wastewater mightsignificantly alter the distribution and identities of absorbed compounds. Nontheless, onemight also expect that the more volatile compounds, those with lower GC retention times,might be removed by aeration, while compounds eluting later might not be removed tosuch a large degree, if at all.Figure 22 is presented in two parts to facilitate interpretation, and shows the effectof aeration on uptake into SPMDs. In both parts, a reduction in peak area with aerationwas observed, with the total area for the peaks displayed decreasing from 22181 units(unaerated) to 19813 units (aerated). This area reduction was more noticeable forcompounds with lower retention times, as shown in Figure 22, Part 1. The compoundsin Part 2 showed less significant reductions, probably because of their lower volatilities.115Ci)4-’ C D (‘3 U) (‘3 (‘3 U)-0\(ti U) a-500450400350300250200150100• 50/////_//__////////////LZZi18.6819.34jNon-aeratedAerated/__/IIIIi24.538.6943.0835.1439.44RetentionTimeof Compound(minutes)43.8643.54Figure22.Part1.Effectof WastewaterAerationonSPMDUptake.G) ci) U) U-//11/___/L-:1/-9/1/_-JJ,II47.1349.24III55.4658.7747.8350.825636/ / / / / /6000-/5000-/4000-/3000-/2000-i1000- 0•IZ? Non-aeratedI /“Aerated61.2160.3162.94RetentionTimeof Compound(minutes)Figure22.Part2.EffectofWastewaterAerationonSPMDUptake.In terms of compound degradation and/or modification under anaerobic conditions,there were 29 compounds present in the non-aerated samples which were not present inthe aerated samples. The majority of these compounds were found at shorter retentiontimes, again suggesting that aeration served to volatilize the more volatile compounds inthe wastewater. Since only 4 of the 29 compounds have significant peak areas greaterthan 50, the likelihood of complete volatilization under aeration was high. Of course,based on the data shown, there was no way to conclusively determine if the compoundsin question were degradation products or were simply volatilized. To further investigatethis, one would have to analyze the wastewater prior to placing it under either aerated orunaerated conditions, and then compare the compounds identified in the wastewater withthose identified in the SPMDs. It should be noted, however, that when the SPMDsdeployed in the anaerobic exposure jars were removed after a 9 day exposure, a strong,noxious, sulfurous odour was evident, indicating the likelihood of anaerobic degradation. Effect of Exposure Time on SPMD UptakeAlthough only 3 exposure times were used, there was a significant temporal trendin uptake (Figures 23 and 24). Figure 23 is for non-aerated treated effluent, while Figure24 is for aerated, untreated effluent. The majority of the targeted compounds exhibitedthe expected increased SPMD uptake with exposure time (recall from equation (7) inSection 2.3.8 that exposure time is directly proportional to analyte concentration inSPMDs). The temporal trend was more obvious for the later-eluting compounds. Therewere higher peak areas displayed in Figure 24 than in Figure 23 (note different y-axisscales), again indicating that treatment has a significant effect on compound uptake into SPMDs.11818.6735.1440.547.8154.6656.3558.7261.2024.4939.4347.1149.21554657.8660.2962.91RetentionTime(minutes)1day3daysII9days(I) 4-. C a) 0) a-1600 :::dl1000 800600[A400200 nScFigure23.EffectofExposureTimeonSPMDUptake(SamplesE-NA-1/3/9).Cl).1-a C as ci) as as ci) as a) 0uuu’J500004000300020001000 0fiLJ______F56.3558.7661.2[1B1ifl1,18.6735.1440.5047.8154.6624.4939.4347.1149.2155.46RetentionTime(minutes)57.8660.2962.91r1day3daysII9daysFigure24.EffectofExposureTimeonSPMDUptake(Samples I-A-1/3/9).To more clearly display the temporal relationship of SPMD uptake, 3 compoundswere selected from the gas chromatograms on the basis of their significant levels inSPMD dialysates. The peak areas of the compounds were plotted to investigate theirabsorption patterns (Figure 25). The identities of the compounds are as follows:55.46 minutes = alpha-ethenyldecahydro-5-(hydroxymethyl)-alpha,5, 8a-t- 1-naphthalenepropanol58.76 minutes unidentified61.20 minutes = 1,2,3 ,4,4a,9, 10, loa-octahydro-l ,4a-dimethyl-7-(l-methyl)-l-phenanthrenecarboxaldehydeThe three compounds showed similar SPMD uptake patterns, with peak areasincreasing with time. In Figures 23, 24, and 25, which illustrate the effect of exposuretime on uptake, the general pattern observed with the majority of the compounds wasa typical uptake curve. Even with only three data points for each compound, it appearedthat compound uptake did not level off in the usual way, ie. peak area increases from day3 to day 9 were not as significant as from day 1 to day 3. If there were no equilibriumphenomena, one would expect the increase in peak area moving from day 3 to day 9 tobe significantly higher, since the difference is 6 days, while the first exposure “window”,day 1 to day 3, is only a 2 day exposure. In fact, this was observed, almost as if noequilibration was occurring. This may be related to the fact that the time required toreach steady-state absorption depends on the molecular weight of the absorbed compounds(Huckins et al., 1990a). The compounds which exhibited the most dramatic increases inuptake from day 3 to day 9 were for the most part compounds with longer GC retentiontimes, indicating lower volatility and greater lipophilicity. In addition, they would tendto be higher molecular weight compounds, implying that they may require longerexposure times to reach steady-state conditions.121L’JU) 4—I C (‘S a) (‘S (‘S G) (‘S G) 0—55.46minutes58.76minutes—A—61.20minutesRetentionTime(minutes)Figure25.Effectof ExposureTimeonSPMDUptake(Samples I-A-1/3/9).AlternateDataPresentation. Fish Tamtmg SignificanceThere were 5 compounds identified in the laboratory effluent exposure sampleswhich were also presented as possible matches for peaks on gas chromatograms of taintedeulachon extracts (see Table 5, Section (Beak Consultants, 1993a). Theirprobable identities and GC retention times are shown in Table 18. In addition, theuptake of each compound into SPMDs was plotted as a function of time (Figures 26, 27,28, 29 and 30).Table 18. Potential Tainting Agents Identified in SPM]) Static Kraft EffluentExposure Extracts.GC Retention Time (minutes) Tentative Compound Identity15.97 2-methyl-5-(1-methylethyl)-bicyclo [3.1.0] hex-2-ene17.95 4-methyl-i-( 1-methylethyl)-bicyclo [3.1.0] hexane19.34 aipha-pinene24.49 1-methyl-4-(1-methylethyl)-cyclohexane36.18 (z)-3 , 7-dimethyl-1,3, 6-octatriene12320-18-U) 14-D Cu-a) a)ExposureTime(days)—I,AI,NA—Ar-E,A)KE,NAFigure26.Uptakeof 2-methyl-5-(1-methylethyl)-bicyclo[3.1.O]hex-2-ene intoSPMDs.16-14-(I)-C D culOG)s—BCu G) as G)4. :1I8I-’.:ExposureTime(days)—I,AI,NA-k-E,A)KE,NA1Figure27.Uptakeof 4-methyl-1-(1-methylethyl)-bicyclo[3.1 .O]hexaneintoSPIVIDs.80-70-(I) 5O(ES ci) 30ExposureTime(days)j •I, AGI,NA—Ar-E,A)l(E,NAFigure28.UptakeofAlpha-PineneintoSPms.500-450-.400-Cl) •350-DI-U)-25OExposureTime(days)—I,A‘iiI,NA—A-E,A>KE,NAFigure29.UptakeofI-methy1-4-(1-methy1ethyIcyc1ohexane intoSP’ffls.00(I) D U) as U) I as G) 0—I,AI,NA-A--E,A‘<E,NAExposureTime(days)Figure30.Uptakeof(z)-3,7-dimethyl-1,3,6-octatrieneintoSPrS4Ds.In general, there was increased absorption of each compound with SPMDexposure time. The highest uptake occurred with the untreated, unaerated samples,followed by the untreated, aerated samples. Both of the treated wastewater samplesshowed little uptake; in fact, of the five compounds, only (z)-3 ,7-dimethyl-1,3,6-octatriene and l-methyl-4-(l-methylethyl)-cyclohexane were detected at all in SPMDsexposed to treated, aerated wastewater. Coincidentally, these were the compounds withthe longest retention times. From the reduced uptake of the suspected tainting agentsfrom treated effluent, we can conclude that they were removed to a substantial degreeduring secondary treatment.Recall that in the earlier model compound work (Section 4.3), there was a concernwith the low recoveries of aipha-pinene. In the laboratory wastewater exposure work, itis notable that alpha-pinene exhibited a significant decrease in uptake following aeration(Figure 28). Obviously, aipha-pinene is easily volatilized, which verifies to some degreethe explanation for its poor recovery noted in the model compound studies.Some of the uptake curves in Figures 26-30 did not demonstrate increasing uptakewith increasing exposure time. All of the compounds except 2-methyl-5-(l-methylethyl)bicyclo[3. 1. 0]hex-2-ene exhibited either decreasing or constant uptake withtime for the untreated, aerated wastewater exposure SPMDs. Under the other exposureconditions, the compounds showed the expected positive temporal correlation.Furthermore, the treated, aerated wastewater samples showed negligible uptake, and,when uptake did occur, it declined with exposure time. Therefore, one can conclude thatthe factor responsible for the observed lack of positive correlation was the presence of129aeration. Apparently, aeration causes a gradual decrease in the levels of the five potentialtainting agents. This substantiates the information presented in Section, whichillustrated how wastewater aeration tended to reduce gas chromatogram peak areas.In terms of other known or suspected tainting agents such as phenols and totalreduced sulphur (TRS) compounds, none were found in the SPMD dialysates, althoughthe average phenol concentration in Eurocan’s final effluent during 1992 was 0.2-0.3mg/L (Beak Consultants, 1993a). However, from the model compound experiments, onewould not expect phenols to be significantly absorbed by SPMDs. On the other hand,TRS compounds, particularly the non-polar species, would be expected to accumulate inSPMDs.It is interesting that there appeared to be a match between possible compoundsaccumulated by the eulachon in the Beak study and some of the compounds sequesteredby the SPMDs. As mentioned in the literature review/background section of this thesis,similarities among the types of accumulated compounds have been observed in otherstudies (Prest et al., 1992; Lebo et al., 1992; Shigenaka and Henry, in press). Thislends some credibility to the feasibility of using SPMDs as “artificial aquatic organisms”.1305.4.2 Continuous Flow Wastewater Exposures5.4.2.1 Introduction and Goals of WorkFrom the static wastewater exposures, the usefulness of SPMDs in sequesteringlipophilic compounds from pulp mill effluent was demonstrated. However, to augmentthe findings from these static exposures and to gain a more thorough understanding ofSPMD absorption from wastewater, it was decided to perform flow-through exposures.Such dynamic exposures would provide a more realistic comparison between theabsorption of suspected tainting agents into SPMDs and uptake by eulachons, sinceeulachons in the Kitimat River are continuously exposed to diluted wastewater.Ultimately, it was hoped that the same compounds identified in the static effluentexposures would be verified as being absorbed into SPMDs in the flow-through system.From March 22 to April 1, 1993, Beak Consultants carried out another taintingstudy which included exposing eulachons to various concentrations of mill effluent usinga flow-through diluter setup (Beak Consultants, 1993b). It was decided to simultaneouslyexpose SPMDs to the same wastewater dilutions as the eulachons to minimizeexperimental manipulation.In the Beak 1991 study (Beak Consultants, 1991), a threshold taintingconcentration of 3-5% effluent was identified for an exposure time of at least 27 hours.This was lowered to 0.3% for 96 hours in the 1992 study. These low values, coupledwith the fact that eulachons are typically exposed to 1-3% effluent at the traditionalfishing grounds on the Kitimat River, required that the 1993 study be carried out at fairlylow effluent concentrations. SPMD and eulachon exposures, then, were completed at1310.2%, 0.6%, 1.25% and 2.5% effluent, as well as at 100%. Blanks were also exposedat 0% (untreated river water taken upstream of the effluent outfall). Unfortunately, dueto the necessity of utilizing the available dilution system, concentrations between 2.5%and 100% could not be investigated.Exposure times were also varied (3, 9, 27, 96 and 216 hours) to assess thetemporal effect of effluent exposure on uptake into SPMDs. The experimental design forthis work was presented in Materials and Methods.Two items must be noted concerning the analyses of the continuous flow samplesand the presentation of data. First, a different gas chromatograph was used in this workthan in the static exposure analyses, making comparison between the two types ofexposures difficult due to the variability in instruments. Peak areas were also expresseddifferently (continuous flow GC traces had areas as high as millions, whereas staticexposure outputs were only on the order of thousands). Second, different temperatureprograms were used in the two types of analyses, resulting in problems with matchingpeaks. These two factors combined meant that it was virtually impossible to compare theuptake of the same compounds in both types of exposures. Compound Uptake in a Flow-through SystemThere were a few considerations that applied specifically to the dynamicexposures. First, because some of the SPMDs were deployed in duplicate, and alsobecause the SPMDs in the dilution apparatus used were exposed to wastewater which hadbeen pre-exposed to SPMDs further “upstream”, one had to assume that the SPMDs had132no effect on each other. More specifically, it was assumed that uptake into one SPMDwas not influenced by uptake into any other SPMD, either simultaneously deployed in thesame dilution chamber or deployed upstream in the dilution system. Since the wastewaterretention times in the exposure chambers were not overly long, with the maximum being3.75 hours, it is probably safe to make this assumption. Second, because eulachons wereconcurrently being exposed to wastewater, it was assumed that their presence had noeffect on SPMD uptake. This included any waste products released by the fish, as wellas compound accumulation and/or modification in their tissues.Mass spectrometric analyses were performed on both 9 day (216 hour), 100%effluent exposure replicates (this sample was deployed in duplicate). A complete list ofthe compounds identified in Sample #1 is given in Table 19. It was consideredunnecessary to present the compounds found in Sample #2, since there was significantoverlap; however, compounds identified only in Sample #2 are listed in Table 20. Manyof the same compounds identified in the static exposures were also observed in thecontinuous flow exposures, as evidenced from a comparison of Table 17 and Appendix Cwith Tables 19 and 20.133Table 19. Compounds Identified in J)ialysates of an SPMD Exposed to 100%Treated Effluent for 9 days (Sample #1).toluene2-methyl-5-(1-methylethyl) bicyclo[3. 1.0] hex-2-ene1 ,4-dimethyl-5-(1 -methylethyl)cyclopenteneisocyanatocyclohexane4-methyl- 1-(1 -methylethyl)bicyclo[3. 1. 0]hexanep-dichlorobenzene (Internal Standard)3-methyl-6-(1-methylethylidene)cyclohexane1-methyl-3-(1-methylethyl)benzene1,3, 3-trimethyl-tricyclo[2 .2.1.02, 6]heptane(e)-2-hepten- 1-011 -methyl-4-(1-methylethyl)cyclohexanealpha,alpha-dimethyl-benzene propanoic acid, ethenyl esterchloronaphtalene (Internal Standard)(z)-3 ,7-dimethyl- 1,3 ,7-octatriene1,2,3,4,5,6,7, 8-octahydro- 1,4,9, 9-tetramethyl-4 ,7-methanozaulene3,3,7 ,7-tetramethyl-5-(2-methyl- 1-propenyl)tricyclo[4. 1.0.02 ,4]heptanealpha-methyl-alpha-2-propynyl-benzenemethanoloctahydro-7-methyl-3-methylene-4-cyclopenta[1 , 3]cyclopropa[1 ,2]benzene6, 6-dimethyl-3-methylene-bicyclo[3. 1. llheptane(e)-7, 1 1-dimethyl-3-methylene-1 ,6, lO-dodecatriene1,2,3,5,6, 8a-hexahydro-4 , 7-dimethyl- 1-(1-methylethyl)naphthaleneaipha-cubebene1,2,3 ,4a,7-hexahydro- 1, 6-dimethyl-4-(1-methylethyl)naphthalene1,2,3,4 ,4a, 5,6, 8a-octahydro-7-methyl-4-methylene- 1-( 1-methylethyl)naphthalene1,2,3 ,4-tetrahydro- 1, 6-dimethyl-4-( 1-methylethyl)naphthalene(z)-3 ,7-dimethyl- 1,3, 6-octatriene1,2, 4a,5 , 6, 8a-hexahydro-4 , 7-dimethyl- 1-( 1-methylethyl)naphthalene2 ,4-dimethylquinoline1,3, 3-trimethyl-tricyclo[, 6]heptane2 ,4-dimethyl-2 ,4-heptadienol1 -methyl-3-(1 ‘-methylcyclopropyl)cyclopentenealpha-oxo-3-cyclohexene- 1-acetic acid1,2,3,4,5,6,7, 8-octahydro- 1-methylphenanthrene2, 6-dimethyl-6-(4-methyl-3-pentenyl)bicyclo[3. 1. 1]hept-2-enethujopsene1,2,3,4 ,4a, 7,8, 8a-octahydro- 1, 6-dimethyl-4-(1-methylethyl) 1 -naphthalenol6-methyl-2-methylene-6-(4-methyl-3-pentenyl)bicyclo[3. 1. ljheptane1, 6-dimethyl-4-(1 -methylethyl)naphthalene1 -(2-thienyl)- 1 -hexanone3-phenyl-2-butanone3, 5-dimethyl- lh-pyrazole1341 -isopropyl-4 ,5-dimethyl-cyclopentene2 ,4-dimethyl-2 ,4-heptadienaltricyclo[4.3. 1.13, 8jundecane- 1-carboxylic acid2-methyl- 1-(octahydro-3a-methyl- lh-inden- l-yl)-i -propanone2-butyidecahydro-naphthalene(z,z)-5, 10-pentadecadien- 1-01chioroanthracene (Internal Standard)[s-(z)]-3 ,7, 1 1-trimethyl- 1,6, 1O-dodecatrien-3-ol8, 13-epoxy-5-beta, 8-beta. h,9-beta. h, 1O-alpha-labd-14-ene3-bromo-cyclodecene2-[(z)-3-hexenyl]-1-methyl-3-methylen- 1 -cyclohexenegermacrene B2,1 ,3-benzothiadiazole1, 4-dimethoxyanthracene7-hexadecyneedulan I(all-z)-5 ,8,1i, 14-eicosatetraenoic acid, ethyl ester4-(1 , 5-dimethyl- 1 ,4-hexadienyl)- 1-methyl-cyclohexene2, 6-di-tert-butyl-4-ethylphenol (BHT)Table 20. Compounds Identified in Sample #2 but not in Sample #1.4-methyl-2-propyl- 1-pentanolalpha-pinene(e, z)-2-hexenoic acid, 3-hexenyl ester2-methyl-5-( 1-methylethyl)bicyclo[3. 1. Ojhex-2-enebenzoic acid, methyl ester4-methyl-cis-cyciohexanol1, 7,7-trimethyl-bicyclo[2 .2. 1]heptan-2-onealpha,alpha,4-trimethyl-3-cyciohexene-1-methanol4, 6-dimethylundecane4-(1, 1-dimethylethyl)benzenemethanol1 -nitrosoadamantane2-(1-methylethyl)thiophene4-methyl-4-phenyl-2-pentanone1 -cyclopropyl-2-(4-pyridinyi)ethanoneoctahydro-7-methyl-3-methylene-4-(1-methylethyl)-lh -cyclopenta[1 , 3jcyclopropa[1 ,2]benzene(z)-4-tridecen-6-yne3-methyl-i -phenyl- 1 -butanone(lr)-2 ,2-dimethyl-3-methylene-bicyclo[2.2. 1]heptanecyclohexyloxybenzene135copaene6, 6-dimethyl-bicyclo[3. 1. 1]heptan-2-one3-phenyl-2-butanone1-isopropyl-4,5-dimehtyl-cyclopentene1 ,4-cyclohexanedimethanol7-pentyl-bicyclo[4. 1 .O]heptane1-(2-methylene-3-butenyl)-1-( 1-methylenepropyl)cyclopropane(z)-3-heptadecen-5-yneoctahydro- 1,4,9 ,9-tetramethyl-lh-3a,7-methanoazulene2,3,8 ,9-tetrahydro-2 ,9-dimethyl-naphtho[2, 1-sigma; 3 ,4-sigma’]difuran6, 6-dimethyl-bicyclo[3. 1. 1]heptane-2-methanol(1-butoxy- 1-methylethyl)benzene4-( 1 ,5-dimethyl- 1 ,4-hexadienyl)- 1-methyl-cyclohexene2-( 1, l-dimethylethyl)anthracene1-(2-methylene-3-butenyl)-1-( 1-methylenepropyl)cyclopropaneThe majority of the compounds identified were hydrocarbons and aromaticcompounds. Two sulfur-containing compounds were identified: 2-(1-methylethyl)-thiophene and l-(2-thienyl)-1-hexanone. Various alcohols were found, including somemethanol, pentanol, cyclohexanol and heptanol derivatives. No resin acids wereobserved, although a few other acids such as an acetic acid derivative were noted. Thiswas surprising, considering the model compound work in which dehydroabietic acid wassequestered fairly efficiently. This will be discussed further in the following section.In general, the blanks (0% effluent) for the various exposure times were fairlyclean, although there were a few contaminants such as isocyanatocyclohexane, as well asthe expected internal standards (p-dichlorobenzene, chloronaphthalene andchioroanthracene). One would expect some degree of contamination due to the presenceof eulachon in the exposure tank.1365.4.2.3 Effect of Effluent Concentration on SPMD UptakeAgain, as in the static exposures, selected GC peaks were targeted for additionalanalysis. Figure 31 shows uptake versus effluent concentration for the 27 hourexposures. Four of the targeted compounds, those with retention times of 7.80, 28.00,51.17 and 66.41 minutes, were found in the blanks at levels comparable to the othersamples. This indicates that they were experimental artifacts and therefore not kraft mill-related. The compound eluting at 66.41 minutes, especially, exhibited extremelyanomalous uptake, with the peak area in the 0.6% effluent sample approximately threetimes that of the 100% exposure.In general, the observed trend in the uptake of the compounds which were nQiidentified in the blanks was one of increasing uptake with wastewater concentration. Thisis intuitively expected, and also predicted based on equation (8) in Section 2.3.8, wherethe dependence of analyte concentration in SPMD lipid was described. Several of theselected compounds were not found at all in the 0.6% exposures, and were identified atlow levels only in the 2.5% samples. In this case, SPMDs were not overly effective insequestering very low levels of the targeted compounds, at least not with an exposuretime of only 27 hours.Due to the necessity to using the available diluter setup, concentrations between2.5% and 100% were not able to be investigated. Had this been possible, the correlationbetween uptake into SPMDs and effluent concentration could have been characterizedmore thoroughly.137U) a) as as G) (‘5 G) 00.6%effluent2.5%effluentII100%effluentU)-D (‘5 U) D 0-F900800700600500400300200-100 0007.80’28.5437.3244.8652.6528.0033.2542.2951.1766.41RetentionTime(minutes)Figure31.ContinuousFlowWastewaterExposures:Effectof WastewaterConcentrationonSPMDUptake@27hrs. Effect of Exposure Time on SPMD UptakeAs with the static laboratory exposures, it was expected that an increase in uptakewould be observed with an increase in exposure time, and this was indeed the case in thecontinuous flow exposures. Gas chromatogram peaks were selected from the 100%effluent exposure, targeted and plotted with respect to exposure time, shown in Figure32. Recall that the peaks chosen for Figure 32 do not correspond to the same peakstargeted in the static wastewater exposure work (Section 4.4.1) due to the use of adifferent temperature program and gas chromatograph. Due to the experimental designand the necessity of utilizing the system already in place, it was not possible toinvestigate exposure time effects with lower effluent concentrations.For clarity, Figure 32 is presented in two parts. In Part 1, showing uptake ofcompounds with shorter retention times, the effect of time does not appear to be assignificant as in Part 2, showing longer retention times. As mentioned earlier, this wasmost likely due to the higher lipophilicity of the later-eluting compounds displayed in Part2. The compounds with short retention times simply do not efficiently accumulate inSPMDs. Also note the differing scales in the two figures.139RetentionTime(minutes)3hours9hoursIl27hours[.j240hoursFigure32.Part1.ContinuousFlowWastewaterExposures:EffectofExposureTimeonSPMDUptake@100%Wastewater.5216652728302U) 4-I C a) I c:s a) U)C’)-D (1 (I) 0-c b5004504003503OO250200150100 50 0[11C7.80’15.2428.0014.6419.1428.5433.2530.0334.371.4-U) C1-----i411ihjrjLh1LHinui,___36.9542.2944.8651.1966.3737.3243.1246.7952.65RetentionTime(minutes)3hours9hoursII27hours[:I240hoursFigure32.Part2.ContinuousFlowWastewaterExposures:EffectofExposureTimeonSPMDUptake@100%Wastewater. Fish Tainting SignificanceThe five suspected tainting agents identified in the laboratory static exposureswere also detected in the continuous exposures. Again, due to the different temperatureprofile used in this work, the retention times of the compounds differed from those in thestatic experiments; therefore, a list of the compounds and their new retention times ispresented below in Table 21. The uptake curves for the compounds into SPMDs exposedto 100% wastewater are shown in Figures 33, 34, 35, 36 and 37.Table 21. Potential Tainting Agents Identified in SPMI) Continuous FlowWastewater Exposure Extracts.GC Retention Time Tentative Compound Identity(minutes)12.61 2-methyl-5-(l-methylethyl)-bicyclo[3. 1. 0]hex-2-ene14.95 4-methyl- 1-(1-methylethyl)-bicyclo[3. 1. Ojhexane15.59 aipha-pinene19.18 1 -methyl-4-( 1-methylethyl)-cyclohexane28.61 (z)-3 , 7-dimethyl-1,3, 6-octatrieneNote that the compounds were not detected at effluent concentrations lower than100%, with the exception of the substituted octatriene, which was observed at very lowlevels in the 2.5% effluent exposures for both 27 and 96 hours. Obviously, the suspectedoff-flavour compounds were not strongly sequestered by SPMDs at the exposure timesused. As shown in the figures, the compounds appeared to follow a somewhat typicaluptake trend, with uptake reaching an apparent plateau.142There appeared to be a correlation between uptake of the tainting agents and theirGC retention times. The compound with the longest retention time, (z)-octatriene,exhibited the highest uptake, and in general, the uptake of the other compounds followedthe same pattern. This correlation of uptake with retention time might be somewhatexpected, since the it is possible to generalize that the later-eluting compounds are morelipophilic and would tend to accumulate in SPMDs at a higher level, although this is byno means a hard-and-fast rule.It is worthwhile to compare the uptake of the suspected tainting agents intoSPMDs exposed under both static and continuous flow conditions. Because the y-axisscales are different (different gas chromatographs), levels of uptake cannot be compared;however, uptake patterns can be. Whereas in the static exposures, it was noted thatuptake did not appear to level off with time, this effect was less noticeable in thecontinuous flow exposures. That is, compound uptake into SPMI)s exposed continuouslyto wastewater seemed to reach more of a plateau, which indicates that they may havebeen nearing steady-state.A measure of the variability of the data was assessed by calculating coefficients ofvariance (COVs) for individual compounds in samples which were deployed in duplicate.It was found that COVs ranged from 4.2% to 12.9%, with the average being 6.7%.Although these values were based on only two measurements, they do indicate thatreproducibility was reasonable under the conditions used.143(I) 4-’ C D cu U) Cu U) L. Cu ci)Figure33.Uptakeof 2-methyl-5-(1-methylethyl)-bicyclo[3.1.0]hex-2-eneintoSPMDsExposedto100%WastewaterinaContinuousFlowSystem.16 14 12 10•U)-o Cu (1) :3 0 b0-——6- 4-—2 0I. 02469Exposuretime(days)30-25Cl)20Ci)I-c0303cUD0)0I CU-a) 05-—0146789Exposuretime(days)Figure34.Uptakeof4-methyl-1-(1-methylethyl)-bicyclo[3.1.0]hexaneintoSPMDsExposedto100%WastewateriiiaContinuousFlowSystem.:3CoCts(s%—0 :30)0LExposuretime(days)Figure35.UptakeofAipha-PineneintoSPMDsExposedto100%WastewaterinaContinuousFlowSystem.70-65--_60-—--—Ci) C D.—.cc3___—.asDQ)0<I—45-—.——.(‘3 G)40-—35 30- 01289Exposuretime(days)Figure36.Uptakeof l-methyl-4-(1-methylethyt)-cyclohexaneintoSPMDs Exposedto100%WastewaterinaContinuousFlowSystem.550-500-450-U)400—-350—ciiCuu ao300-<F250-a)00200-—150——100-014679Exposuretime(days)Figure37.Uptakeof (z)-3,7-dimethyl-1,3,6-octatrjene intoSPMDsExposedto100%WastewaterinaContinuousFlowSystem.5.4.3 Discussion of Wastewater Exposure WorkThe wastewater exposures demonstrated the feasibility of monitoring kraft milleffluents using SPMDs. Many compounds were sequestered by SPMDs; however, asexpected, these compounds tended to be hydrocarbons with low polarities. The fivesuspected tainting agents fall into this group, making SPMDs good monitors for thosecompounds, although at low effluent concentrations, much longer exposure times shouldbe used in order to sequester substantial levels of the compounds.Many of the major constituents of kraft effluent identified in the literature werenot sequestered by SPMDs. For example, although guaiacol is often present in kraft milleffluent at concentrations of 10 mg/L (Wilson and Hrutfiord, 1971), it was not detectedin any of the SPMD dialysates. This of course was predicted by the model compoundwork. Similarly, although methanol often comprises 0.5% of digester relief condensates,it was detected only in substituted form and at low concentrations in SPMDs. Very fewacids were identified in the SPMDs, although kraft effluent typically contains significantlevels of these compounds, especially resin acids, fatty acids and p-tolyl-valeric acid(Hrutfiord et al., 1975).Table 22 presents the chemical analysis of Eurocan final effluent during thecontinuous flow exposure period of March 22 to April 1, 1993. Wastewater sampleswere collected for each eulachon exposure period in order to ascertain the levels to whichthe fish were exposed during the study. From this table, it is clear that the SPMDsexposed to effluent in the continuous flow system did not reflect its chemicalcomposition. Specifically, the lack of resin acids found in SPMD dialysates is puzzling.149The total uptake experiment (Section 5.3.3) demonstrated that SPMD exposure to 10 and100 pgIL dehydroabietic acid resulted in some accumulation. However, continuous flowexposures to final effluent, which contained DHA concentrations of 230-380 gIL ofDHA during the exposure period (Table 22) did not result in uptake. In fact, total resinacids ranged from 1172 to 2025 jig/L, and were not evidenced at all in SPMDs. Thismay be due to the low solubility of resin acids, which causes them to adsorb to apolarsurfaces (Drobosyuk et al., 1982). The resin acids in the wastewater may have adsorbedto suspended solids such as biomass, thus making them unavailable to SPMDs. A similarexplanation may be forwarded for the lack of fatty acids detected in the SPMDs, althoughtotal fatty acids in Eurocan effluent during the exposure period ranged from 767 to 1787LgIL.According to Table 22, phenol levels were fairly low during the exposure period,averaging around 0.005 mg/L; therefore, it is not surprising that individual phenoliccompounds were not taken up, notwithstanding their low affinity for SPMDs. Also, itshould be noted that phenol levels appeared to decrease from 1992 (0.2-0.3 mg/L in1992), while the tainting propensity of the effluent increased.150Table 22. Chemical Analysis of Eurocan Effluent Samples Collected BetweenMarch 22 and April 1, 1993. (Beak Consultants, 1993b)(h..r.P.nm..n Urnt. 3 15 11 72 14 94 lMd.I.R.d. Add.Ab4.d.CaedchydtO.blcllcD4ddotodcy*36b4.Ue‘U—:ixopmidcN.o.b1611Sandm.dcT.Id R.d. Add..t 420 530 360ugh.. 240 320 300‘VT.‘Vt 270 360 320‘Vtugh. 116 123 121ugh?. 16 106 OSiigit 310 430 760Ugh?. 53 14 63‘Vj 1495 2023 1551530 430 410 330330 360 310 230320 330 254 176104 64 64 .1396 30 76 49460 430 360 21072 76 65 521962 1860 1639 Add9.:G-Diddoro.sdd AddLncL.nckicLindeni.MyT1ticOlnic?ahrnilcStenne AddT44.I F.tyAn.daO.M.tbyopodccapne Addugh?.“Vt“IlLUghUgh?.us/i‘Vtugh.vt‘Vt% R.nq• &lew Method Detection LinkNM. P.enncter net menetrnd• to be nnfled by tepoac nnalynis151Alcmirn.n mgil. NM LU 1.78 NM NM 146 Li14d,e. mgi?. NM 0.07 0.06 NM NM 0.07 0.06&yfl1um mg/I. NM . NM NMOunu mg/?. NM 0.14 004 NM MM 0.2 0.12C.dtuiun mg/?. NM • NM NMCdcwm mg/i MM 32 29 9130 MM 30 26mg/i NM . . NM NMCob.k mg/I. NM . . NM NMCoppct mg/i NM . 0.045 NM NM 3Q35Iron mg/I. 9130 0.7 0.6 NM NM 0.66 0.5Lead mg/I. NM • NM NM -Magnetiw. mg/I. NM L4 1.05 9430 NM 1.4 1.2$Mnepnne mgi?. NM 0.71 0.69 NM NM 0.72 0.64Molybdeaum mg/I. NM . • NM NMNickel mg/I. NM . - NM NM?olaa.lum mg/I. NM II 17.7 MM NM 10.2 14.3Sleet mg/I. KM . • NM NMStrodw. mg/i 3430 0.12 0.1 NM NM 0.1 0.1Thaw... w.JL NM . . NM NM . -Tin mg/i NM . . NM NMTlmthaa mg/I. NM 0.01 0.01 NM MM 0.01 001Vanaum mgit 3430 . NM NMZinc mg/I. MM 0.14 0.05 NM NM 0.06 0.07C..ndinal flna.4.nTn.. Color TCU 650 720 440 660 710 610 NMPhonols mg/i 0.064 0.064 0.064 0.064 0. 0063 0.066TouIDIuclecdSoUà mg/i. 1100 1100 1130 1140 1110 2180 NMTotal XpIdaM Nitrogen mg/I. 7 74 7.6 0.3 7.4 74 7.7DIumleedlojum mg/?. 270 290 290 230 280 290 3105, 43 57 33 57 59 4130 90 116 104 115 90 9030 32 63 52 90 4011 106 96 730 210 403 130: 10 19 13 26 13150 160 150 290 230 360 170810 230 300 370 750 390 580540 110 130 330 213 320 1431704 767 159 1565 1237 1717 1019104 97 95 83 92 96 103Terpenes, also major constituents of kraft effluent and often present at 8 mg/Llevels (Hrutfiord et al., 1975), were identified to some degree in SPMD dialysates. Themost common terpene in kraft effluent, aipha-terpineol, was not detected, although otherterpenes were identified, such as alpha- and beta-pinene, aipha-cubebene, thujopsene,labdene, copaene and azulene. In the model compound work, it was not clear whetheralpha-pinene was simply not taken up into SPMDs or was lost due to volatilization,although evidence obtained in the static wastewater exposure work indicated that the latterwas a more likely explanation. There was a similar problem in the wastewater exposurework. Since no analysis of the wastewater for terpenoid compounds was carried out,other than by SPMD exposure, it is unclear whether there were significant levels ofterpenes in the wastewater, or whether the SPMDs were simply not absorbing theterpenoids present. Eurocan had started up a turpentine recovery system in 1992, and itwas assumed that this system would decrease the terpene levels in the condensate andthus minimize their tainting potential. Therefore, the small numbers of terpenes in theSPMD samples may be a result of improved removal from the wastewater stream. Toverify this, one would have to analyze the wastewater for terpenes and compare theresults with the SPMD dialysates.It appeared that the suspected tainting agents were present at relatively low levelsin the wastewater, compared to other contaminants. In the static exposures, this wasevident from the significant reduction in the quantities of these compounds absorbed bySPMDs following secondary treatment (Figures 26-30). In the continuous flowexposures, there appeared to be minimal uptake of the tainting agents at low effluent152concentrations, even for long exposure times.Interestingly, although aipha-pinene only exhibited around 15-20% uptake in themodel compound work, it was detected in the wastewater exposures, especially exposuresto full-strength untreated effluent for the longest exposure period of 9 days. However,simply because absorbed levels were higher does not mean that uptake was higher, sincethere likely were higher levels of alpha-pinene in the effluent than in the total uptakeexperiment, which was carried out at concentrations of 10 and 100 ppb. Also, thewastewater matrix is much more complicated than the simple distilled water set-up in thelaboratory, which may affect uptake.Another difference to consider in the wastewater exposures as compared to thelaboratory work was the possibility of Aufwuchs growth. Upon removal of SPMDs fromdeployment in wastewater (both static and continuous flow systems), there was evidenceof some biological growth on the polyethylene surface. This “slime” layer was morenoticeable at higher wastewater concentrations for longer exposure times, but wasconsidered to be minimal and thus not significantly affect uptake.The conclusions from the wastewater exposure work were as follows:(1) Wastewater treatment reduced the levels of compounds sequestered bySPMDs, as well as the numbers of individual compounds.(2) Wastewater aeration also reduced compound levels, and seemed topreferentially remove the more volatile compounds eluting at shorter GC retention153times.(3) An increase in exposure time resulted in an increase in compound uptake inboth static and continuous flow wastewater exposures.(4) Five potential off-flavour compounds in eulachon were identified in SPMDsstatically exposed to wastewater, but were largely removed during secondarytreatment. The same five compounds were also identified in the flow-throughsystem, but only in the SPMDs exposed to 100% effluent.(5) The effect of varying wastewater concentration on SPMD uptake in acontinuous flow system was to increase uptake, although this was difficult toassess in the absence of test concentrations between 2.5% and 100%.(6) SPMDs sequestered large numbers of compounds from kraft mill effluent.Most of these were hydrocarbons and other neutral compounds. Very fewionizable or polar compounds were detected in dialysates.1546. CONCLUSIONS AND RECOMMENDATIONS6.1 ConclusionsThis work examined the feasibility of using SPMDs as monitors of kraft milleffluent contaminants, with particular interest in the uptake of potential fish-taintingcompounds. As such, they were assessed in the context of both model compounds andactual wastewaters. In addition, the early steps in SPMD preparation were investigated.From this examination, the following conclusions can be drawn:(1) Layflat polyethylene tubing could be used directly from the roll, since permeabilityappeared to be constant intra-roll. The possibility of variations in permeability inter-rollwas not investigated.(2) One 24 hour solvent batch extraction of tubing prior to incorporation into SPMDswas sufficient to adequately remove contaminating substances from the tubing.(3) In general, dialytic recovery of some of the model compounds from SPMDs waspoor. This may have been due to volatilization of aipha-pinene, and to a lesser extent,guaiacol.(4) Two-stage dialyses resulted in slightly higher dialytic recoveries than single-stagedialyses.(5) Tubing alone (no triolein), as well as SPMDs with triolein concentrated in one area,rather than spread into a thin film, sequestered the test compounds to the same or higherdegree as conventional SPMDs.155(6) Uptake into SPMDs exposed to kraft mill wastewater in static and continuous flowsystems was affected by wastewater treatment, wastewater aeration, exposure time andwastewater concentration.(7) The compounds sequestered from kraft effluent tended to be nonpolar, withcompounds such as resin acids and phenols not detected in the dialysates of exposedSPMDs.(8) Five potential tainting agents in eulachons were identified in SPMDs exposed to krafteffluent. Although they were only detected at low levels, the SPMDs were exposed towastewater for a maximum of 9 days in both static and continuous flow experiments;therefore, with sufficient exposure time, higher uptake may occur.(9) Overall, SPMDs performed well in the capacity of monitoring nonpolar compounds,but their application to polar molecules was limited. However, their usefulness may liein their ability to sequester lipophilic compounds, thus providing an integrated measure ofthe levels of these compounds in aqueous environments. In this capacity, SPMDs wouldbe extremely useful for the monitoring of potential tainting agents at Eurocan Pulp andPaper Co., Kitimat, B.C, since such agents are likely to be relatively nonpolar, asevidenced by their accumulation in eulachons.1566.2 RecommendationsBased on the findings of this study and a survey of the current literature, anumber of recommendations for further work in this area can be made. These aresubdivided into two major areas of interest in this work: SPMD application to kraft milleffluent constituents and SPMD use in tainting agent monitoring.I. Application to Kraft Effluent Constituents(1) Further model compound work should be carried out under continuous flowconditions to determine how, if at all, this would affect their behaviour and recoveriesinto SPMDs.(2) A broader range of model compounds of kraft mill origin should be assessed.Specifically, it would be interesting to examine the uptake of chlorinated phenoliccompounds into SPMDs, since the higher levels of chlorine decrease the polarities of thecompounds.(3) A more detailed examination of uptake kinetics of various compounds is required.With sufficient study, expressions for time-averaged aqueous concentrations ofcontaminants could be elucidated.(4) Water quality parameters such as pH, salinity, temperature and the presence of highmolecular weight compounds which may complex with analytes, should be studied todetermine their effects on SPMD uptake. Specifically, the effect of pH on the uptake ofionizable compounds should be investigated.157(5) Additional wastewater exposures should be carried out at a wider range of exposuretimes and wastewater concentrations, with multiple samples at each condition to allowstatistical evaluations such as linear regression and analysis of variance.II. Tainting Agent Monitoring(1) SPMDs should be deployed in the Kitimat River, preferably for a 3-4 week timeperiod corresponding to the time eulachons are exposed to diluted wastewater duringspawning.(2) Eulachons and SPMDs should be simultaneously deployed in the Kitimat River todetermine uptake correlations. If both eulachons and SPMDs showed similar uptake,both in terms of compounds sequestered and levels of compounds, then the use ofSPMDs as “artificial eulachons” would be valid.(3) SPMDs should be deployed in various process streams within the mill itself in anattempt to pinpoint the source of the tainting agents.(4) The use of purified eulachon grease in SPMDs instead of triolein or other model lipidshould be assessed in the context of the tainting problem at Eurocan.158LITERATURE CITEDAggarwal, S.L. and O.J. Sweeting. (1957) Polyethylene: Preparation, Structure andProperties. Chem. Rev. 57:665-742Autopro Canada Ltd. (1991) Study of Remedial Alternatives to Abate Sources of Colour,Odour, Foam and Fish Flavour Impairment in the Mill Final Effluent, Volume 1. Reportprepared for Eurocan Pulp and Paper Co., Kitimat, B.C.Beak Consultants Ltd. (1991) Fish Tainting Potential of Eurocan Effluent in the KitimatRiver, British Columbia. Report prepared for Eurocan Pulp and Paper Co., Kitimat, B.C.Beak Consultants Ltd. (1992) Evaluation of the Fish-Tainting Propensity of Eurocan FinalEffluent. Report prepared for Eurocan Pulp and Paper Co., Kitimat, B.C.Beak Consultants Ltd. (1993a) Tainting Evaluation of Eurocan Final Effluent- StudyOverview 1991-1993. Report prepared for Eurocan Pulp and Paper Co., Kitimat, B.C.Beak Consultants Ltd. (1993b) The Effect of Eurocan Effluent on Oolichan Flavour - 1993.Report prepared for Eurocan Pulp and Paper Co., Kitimat, B.C.Berg, N. (1983) Chemical and sensory analysis of off-flavours in fish from polluted riversin Norway. Wat. Sci. Tech. 15:59-65Biermann, C.J. (1993) Essentials of Pulping and Papermaking. Academic Press Inc., SanDiego, CA.Black, J.J., T.F. Hart Jr. and P.1. Black. (1982) A novel integrative technique for locatingand monitoring polynuclear aromatic hydrocarbon discharges to the aquatic environment.Environ. Sci. Technol. 16:247-250Blackwell, B.R., W.B. MacKay, F.E. Murray and W.K. Oldham. (1979) Review of kraftfoul condensates: sources, quantities, chemical composition and environmental effects. Tappi62(10) :33-37Briggs, G.G. (1981) Theoretical and experimental relationships between soil adsorption,octanol-water partition coefficients, water solubilities, bioconcentration factors and theparachor. J. Agric. Food Chem. 29:1050-1059Brouzes, R.J.P. and V.A. Naish. (1979) Fish tainting - methodology and identification oftainting components in a kraft mill. Presented at the Environment Improvement Conferenceof the Technical Section, Canadian Pulp and Paper Association, Victoria, B.C.159Bruggeman, W.A., L.B.J.M. Matron, D. Kooiman and 0. Hutzinger. (1981) Accumulationand elimination kinetics of di-, tn- and tetra-chiorobiphenyls by goldfish after dietary andaqueous exposure. Chemosphere 10:811-832Chiou, C.T., V.H. Freed, D.W. Schmedding and R.L. Kohnert. (1977) Partition coefficientand bioaccumulation of selected organic chemicals. Environ. Sd. Technol. 11:676-682Chiou, C.T., D.W. Schmedding and M. Manes. (1982) Partitioning of organic compoundsin octanol-water systems. Environ. Sci. Technol. 16:4-10Chiou, C.T. (1985) Partition coefficients of organic compounds in lipid-water systems andcorrelations with fish bioconcentration factors. Environ. Sci. Technol. 19:67-62Collander, R. (1954) The permeability of Nitella cells to non-electrolytes. Physiol. Plant.7:420-455Comyn, J., ed. (1985) Polymer Permeability. Elsevier Applied Science Publishers Ltd.,New York, NY. p.383Cook, W.H., F.A. Farmer, O.E. Knistiansen, K. Reid, J. Reid and R. Rowbottom. (1973)The effect of pulp and paper effluents on the taste and odour of the receiving water and thefish therein. Presented at the 58th Annual Meeting of the Technical Section of the CanadianPulp and Paper Association, Montreal, Quebec.De Boer, T.J. and H.J. Backer. (1954) A new method for the preparation of diazomethane.Rec. Tr. Chim. 73:229-234Dictionary of Organic Compounds. (1982) Fifth edition, volume 4. Chapman and Hall,Toronto, Ont.Doerr, R.C. and W. Fiddler. (1970) Partition ratios of some woodsmoke phenols in twooil:water systems. J. Agr. Food Chem. 18(5):937-939Drobosyuk, V.M., M.R. Borodulina, Y.A. Krylatov, S.L. Talmud. (1982) The dispersivepower of resin acids and fatty acids. Journal ofApplied Chemistry of the U.S.S.R. 55(9):1935-1939Fengel, D. and G. Wegener. (1983) Wood: Chemistry, Ultrastructure, Reactions. Waterde Gruyter & Co., Berlin.Fish, C.L., I.S. McEachran, J.P. Hassett. (1989) 198th Meeting of the American ChemicalSociety, Division of Environmental Chemistry. Abstract No. 181, pp. 498-501Fujita, T., J. Iwasa and C. Hansch. (1964) A new substituent constant, pi, derived frompartition coefficients. J. Am. Chem. Soc. 86:5775-5 180160Gordon, M.R., J.C. Mueller and C.C. Walden. (1980) Effect of biotreatment on fishtainting propensity of bleached whole mill effluent. Canadian Pulp and Paper AssociationTransactions, March 1980, pp.TR2-TR8Gray, M.A. and A. Spacie. (1991) Estimating bioconcentration potential using syntheticmembranes: a comparison of hexane-filled cellulose membranes, triolien-filled polyethylenemembranes and fish. Presented at the 12th Annual Meeting of the Society of EnvironmentalToxicology and Chemistry, November, 1991, Seattle, WA.Hansch, C. and T. Fujita. (1964) p-sigma-pi analysis: a method for the correlation ofbiological activity and chemical structure. J. Am. Chem. Soc. 86:1616-1626Hassett, J.P., M. Force and H. Song. (1989) A passive in-situ sampler for organiccompounds in water. Proceedings of the National ACS Meeting, Miami Beach, FL. 5:496-497 (extended abstract)Hawker, D.W. and D.W. Connell. (1989) A simple water/octanol partition system forbioconcentration investigations. Environ. Sci. Technol. 23:961-965Heil, T.P., N.A. Lane and R.C. Lindsay. (1989) Sensory properties of thio- andalkyiphenols causing flavour tainting in fish from the Upper Wisconsin River. J. Environ.Sci. Health B24(4):361-388Heil, T.P. and R.C. Lindsay. (1988) Volatile compounds in flavour tainted fish from theUpper Wisconsin River. .1. Environ. Sci. Health B23:489-512Henderson, J.R. and D.R. Tocher. (1987) The lipid composition and biochemistry offreshwater fish. Prog. Lipid Res. 26:281-347Herve, S., R. Paukku, J. Paasivirta, P. Heinonen and A. Sodergren. (1991) Uptake oforganochlorines from lake water by hexane-filled dialysis membranes and by mussels.Chemosphere 11:997-1001Hrutfiord, B.F., T.S. Friberg, D.F. Wilson and J.R. Wilson. (1975) Organic compoundsin aerated stabilization basin discharge. Tappi 58(10):98-100Huckins, J.N., M.W. Tubergen and G.K. Manuweera. (1990a) Semipermeable membranedevices containing model lipid: a new approach to monitoring the bioavailability of lipophiliccontaminants and estimating their bioconcentration potential. Chemosphere 20(5):533-552Huckins, J.N., M.W. Tubergen, J.A. Lebo, R.W. Gale and T.R. Schwartz. (1990b)Polymeric film dialysis in organic solvent media for cleanup of organic contaminants. J.Assoc. Off Anal. Chem. 73(2):290-293161Huckins, J.N. (research chemist at the National Fisheries Contaminant Research Center,Columbia, MO), personal communication, 1992.Huckins, J.N., G.K. Manuweera, J.D. Petty, D. Mackay and J.A. Lebo. Lipid-containingsemipermeable membrane devices for estimation of average aqueous concentrations oforganic contaminants: background, theory and model development. Chemosphere, in press.Hwang, S.T. and K. Kammermeyer, eds. (1975) Membranes in Separations. Robert E.Krieger Publishing Co., Inc. Malabar, FL. p.559Johnson, G.D. (1991) Hexane-fluled dialysis bags for monitoring organic contaminants inwater. Environ. Sci. Technol. 25(11): 1897-1902Klein, P., J. Pans, J. Knuutinen. (1987) Taste and odour impairment in fish of the LakePaeijaenne. Presented at Lake Paeijaenne Symposium, Jyvaeskylae, Finland. In: Biol. Res.Rep. Univ. Jyvaeskylae, vol.10Koning, C.W. and S.E. Hrudey. (1988) Sensory and Chemical Analysis of Fish Taintingby Exposure to Oilsands Wastewaters. Report prepared for the Alberta ResourceManagement Division. Report L- 112.Kovacs, T.G., R.H. Voss and A. Wong. (1984) Chlorinated phenolics of bleached kraftmill origin: an olfactory evaluation. Water Res. 18(7):911-916Lebo, J.A. (research chemist at the National Fisheries Contaminant Research Center,Columbia, MO), personal communication, 1993Lebo, J.A., J.L. Zajicek, J.N. Huckins, J.D. Petty, P.H. Peterman and R.W. Gale. (1992)Use of semipermeable membrane devices for in situ monitoring of polycyclic aromatichydrocarbons in aquatic environments. Chemosphere 25(5): 697-718Lieb, W.R. and W.D. Stein. (1969) Biological membranes behave as non-porous polymericsheets with respect to the diffusion of non-electrolytes. Nature 224:240-243Liem, A.J., V.A. Naish and R.S. Rowbottom. (1977) An evaluation of the effect of inpianttreatment systems on the abatement of air and water pollution from a hardwood kraft pulpmill. CPAR Project Report No. 484-1, Environmental Protection Service, Ottawa, Ont.Lindberg, J.J. (1958) The relation between the distribution in ether and water and theinfrared spectrum of several guaiacyl and syringyl compounds. Soc. Sci. Fennica,Commentationes Phys.-Math 21(7): 1-13Liu, P., K. Thomson, K.L.E. Kaiser. (1982) Quantitative structure-toxicity relationship ofhalogenated phenols on bacteria. Bull. Environm. Contam. Toxicol. 29:130-135162Mackay, D. and Hughes, A.I. (1984) Three-parameter equation describing the uptake oforganic compounds by fish. Environ. Sci. Technol. 18:439-444McEachren, I.S., C.L. Fish, J.P. Hassett. (1991) 201st Meeting of the American ChemicalSociety, Division of Environmental Chemistry. Abstract No. 86, pp. 367-370.McLeay, D.J. and Associates. (1987) Aquatic toxicity of pulp and paper mill effluent: areview. BPS Report 4/PF/ 1, Environment Canada, Ottawa, Ont.McLeay, D.J., C.C. Walden, J.R. Munro. (1979) Influence of dilution water on the toxicityof kraft pulp and paper mill effluent, including mechanisms of effect. Water Res.13:151-158Maligalig, L.L., J.F. Caul, R. Bassette, O.W. Tiemeir. (1975) Flavouring live channelcatfish (Ictalurus punctatus) experimentally: effects of concentration and exposure time. J.Food Sci. 40:1242-1250Meadows, J., D. Tillitt, J. Huckins and D. Schroeder. (1993) Large-scale dialysis of samplelipids using a Semipermeable Membrane Device. Presented at the Seventh Annual OzarkPrarie SETAC Meeting, Carbondale, IL, May 1992. In: Chemosphere 5:1993-2006Merck Index, J. Windholz, ed. (1983) Merck Index, 10th edition. Merck & Co., Inc.,Rahway, New Jersey.Miller, M.M., S.P. Wasik, G-L. Huang, W-Y. Shiu and D. Mackay. (1985) Relationshipsbetween octanol-water partition coefficients and aqueous solubility. Environ. Sci. Technol.19:522-529National Academy of Sciences. (1980) The International Mussel Watch. Report of aworkshop sponsored by the Environmental Studies Board Commission on Natural Resources.Washington, D.C.Neely, W.B., D.R. Branson and G.E. Blau. (1974) Partition coefficients to measurebioconcentration potential or organic chemicals in fish. Environ. Sci. Technol. 8:1113-1115Nyren, V. and E. Back. (1958) The ionization constant, solubility product and solubility ofabietic and dehydroabietic acid. Acta Chemica Scandinavia 12:1516-1520Oikari, A. and B. Holmbom. (1986) Assessment of water contamination by chlorophenolicsand resin acids with the aid of fish bile metabolites. In: Aquatic Toxicology andEnvironmental Fate, ASTM STP 921. T.M. Poston, R. Prudy, ecis. Ellis Horwood Ltd.Owens, J.W. (1991) The hazard assessment of pulp and paper effluents in the aquaticenvironment: a review. Environ. Toxicol. Chem. 10:1511-1540163Paasivirta, J., J. Knuutinen, J. Tarhanen, T. Kuokkanen, K. Surma-Aho, R. Paukku, H.Kaariainnen, M. Lahtipera and A. Veijanen. (1983) Potential off-flavour compounds fromchloro-bleaching of pulp and chiorodisinfection of water. Wat. Sci. Tech. 15:97-104Paasivirta, J., H. Palm, S. Herve, P. Heinonen, H. Prest, J. Koistinen, T. Hyotylainen.(1992) Polyethylene membranes: experiences in studies on pulp mill wastes in thelaboratory and in the environment. Unpublished material.Persson, P.-E. (1984) Uptake and release of environmentally occurring odourouscompounds by fish. Water Res. 18(10): 1263-1271Persson, P.-E. and R.K. York. (1978) Notes on muddy odour. II. Uptake of 2-methylisoborneol by rainbow trout (Salmo gairdneri) in continuous flow aquaria. AquaFennica 8:89-90Petty, J.D., J.N. Huckins, J.A. Lebo, J.L. Zajicek, J.C. Meadows. (1992) Current statusof lipid-containing semipermeable membrane devices as environmental dosimeters.Unpublished material.Platford, R.F. (1979) Glyceryl trioleate-water partition coefficients for three simple organiccompounds. Bull. Environ. Contam. Toxicol. 21:68-73Poulton, B. (1992) Design considerations for field deployment of SPMDs in aquaticsystems. Unpublished material.Prest, H.F., W.M. Jarman, S.A. Burns, T. Weismuller, M. Martin and J.N. Huckins.(1992) Passive water sampling via semipermeable membrane devices (SPMD5) in concertwith bivalves in the Sacramento/San Joaquin River delta. Chemosphere 25(12): 1811-1823Roff, W.J., J.R. Scott and J. Pacitti. (1971) Handbook of Common Polymers. CRC Press,Cleveland, OH.Rogers, I.H., J.C. Davis, G.M. Kruzynski, H.W. Mahood, J.A. Servizi and R.W. Gordon.(1975) Fish toxicants in kraft effluents. Tappi 58(7): 136-140Rogers, I.H. (1978) Environmental effects of terpenoid chemicals: a review. J. Amer. OilChem. Soc. 55:113A-118ARogers, I.H., I.K. Birtwell and G.M. Kruzynski. (1990) The pacific eulachon (ThaleichthyspacWcus) as a pollution indicator organism in the Fraser River Estuary, Vancouver, BritishColumbia. Sd. Tot. Env. 97/98:713-727Saarikiski, J. and M. Viluksela. (1982) Relation between physicochemical properties ofphenols and their toxicity and accumulation in fish. Ecotoxicol. Environ. Safety 6:501-5 12164Saarikiski, J., R. Lindstom, M. Tyynela and M. Viluksela. (1986) Factors affecting theabsorption of phenolics and carboxylic acids in the guppy (Poecilia reticulata). Ecotoxicol.Environ. Safety 11:158-173Shigenaka, G. and C.B. Henry, Jr. Use of mussels and semipermeable membrane devicesto assess bioavailability of residual polynuclear aromatic hydrocarbons three years after theExxon Valdez oil spill. Third Symposium on Environmental Toxicology and RiskAssessment, ASTM STP 1219, in press. J.S. Hughes, G.R. Biddinger and E. Mones, eds.American Society for Testing and Materials, Philadelphia, PA.Shiu, W.Y. and D. MacKay. (1986) A critical review of aqueous solutions, vaporpressures, Henry’s Law constants and octanol-water partition coefficients of polychiorinatedbiphenyls. J. Phys. Chem. Ref Data 15:911-929Shumway, D.L. and G.G. Chadwick. (1971) Influence of kraft mill effluent on the flavourof salmon flesh. Water Res. 5:887-1003Sodergren, A. (1987) Solvent-filled dialysis membranes simulate uptake of pollutants byaquatic organisms. Environ. Sd. Technol. 21(9): 855-859Sodergren, A. (1990) Monitoring of persistent, lipophilic pollutants in water and sedimentby solvent-filled dialysis membranes. Ecotoxicol. Env. Safely 19:143-149Stalling, D.L., R.C. Tindle and J.L. Johnson. (1972) Apparatus for automated gelpermeation cleanup for pesticide residue analysis: applications to fish lipids. Anal. Chem.44:1768-1773Swan, E.P. (1973) Resin Acids and Fatty Acids of Canadian Pulpwoods - A Review of theLiterature. Environment Canada Forestry Service Publication VP-X- 115.Tanabe, S. and R. Tatsukawa. (1987) Mussels as bioindicators of PCB pollution: a casestudy on uptake and release of PCB isomers and congeners in green-lipped mussels (Pernaviridis) in Hong Kong waters. Environ. Pollut. 47:41-62Thaysen, A.C. and F.T.K. Pentelow. (1936) The origin of an earthy or muddy taint in fish.2. The effect on fish of the taint produced by an odouriferous species of Actinomycetes.Ann. Appi. Biol. 23:105-109Underhill, D.W. and C.E. Feigley. (1991) Boundary layer effect in diffusive monitoring.Anal. Chem. 63:1011-1013U.S. Environmental Protection Agency. (1991) Quantitative structure activity relationshipdatabase. Environmental Research Laboratory, Duluth, Minnesota.165Vargo, J.D. and K.L. Olson. (1985) Identification of antioxidant and ultraviolet lightstabilizing additives in plastics by liquid chromatography/mass spectrometry. Anal. Chem.57:672-675Veith, G.D., N.M. Austin and R.T. Morris. (1979) A rapid method for estimating log Pfor organic chemicals. Wat. Res. 13:43-47Voss, R.H., J.T. Wearing and A. Wong. (1981) A novel gas chromatographic method forthe analysis of chlorinated phenolics in pulp mill effluents. In: Advances in the Identificationand Analysis of Organic Pollutants in Water, Volume 2 (L.H. Keith, ed.). Ann ArborScience Publishers, Ann Arbor, Michigan.Voss, R.H. (1984) Neutral organic compounds in biologically treated bleached kraft milleffluents. Environ. Sd. Technol. 18:938-946Warrington, P. (1987) Skeena-Nass area lower Kitimat River and Kitimat Arm water qualityassessment and objectives technical appendix. B.C. Ministry of Environment.Whittle, D.M. and K.W. Flood. (1977) Assessment of the acute toxicity, growthimpairment and flesh tainting potential of a bleached kraft mill effluent on rainbow trout(Salmo gairdneri). J. Fish. Res. Board Can. 34:869-878Wilson, D.F. and B.F. Hrutfiord. (1971) Sekor IV. Formation of volatile organiccompounds in the kraft pulping process. Tappi 54(7): 1094-1098Wilson, D.F. and B.F. Hruffiord. (1975) The fate of turpentine in aerated lagoons. Pulpand Paper Canada 76(6):91-93Xie, T.M., B. Hulthe and S. Folestad. (1984) Determination of partition coefficients ofchlorinated phenols, guaiacols and catechols by shake-flask GC and HPLC. Chemosphere39: 445-459Yasuda, H. (1967) Basic consideration of permeability of polymer membranes to dissolvedoxygen. J. Polymer Sci., Part A-i, 5:2952-2956Yasuda, H. (1992) Thoughts on polymeric systems for the next generation SPMDs.Unpublished material.Zajicek, J.L., J.D. Petty and J.N. Huckins. (1992) Passive sampling of PCBs in laboratoryair: results of initial investigations using SPMDs. Unpublished material.166APPENDICESAPPENDIX A: Chromatograms of selected gas chromatographic and mass spectrometricanalysesAPPENDIX B: Experimental dataAPPENDIX C: Compounds identified in SPMD wastewater exposure extractsAPPENDIX D: Temperature programs used for GC and GC/MS analyses167APPENDIX A: CHROMATOGRAMS FROM GC AN]) GC/MS ANALYSESIntensity03. oe+ooi-4---2. 5e+0072.Oe+007-1.5e+0071. Oe+007-5e+006 -Hexane TIC5 10 15 20 25 30 35 40 45 50 55 60—3.0e+0071-2. 5e+007—2.Oe+0071.5e+007[1. Oe+007L 5e+006Figure A-i. GCIMS Total Ion Count for Hexane.5 10 15 20 25 30Time35 40 45 50 55 60168Intens i.ty0370000035000003000000-j2500000200000015000001000000—500000-Oil TIC5 10 15 20 25 30 35 40 45 50 55 603700000350000030000002500000200000015000001000000500000Figure A-2. GCIMS Total Ion Count for Neat Triolein Dissolved in Hexane.0 5 10 15 20 25 30 35 40 45 50 55 60Time169(a)zz—:5.3:3:3 :n— :3 2fl.35— 13.3633 :9533 .591 31(b)3 .5::Figure A-3. Gas Chromatograms of Glove Extracts. (a) hexane into which aglove was dipped for 5 minutes; (b) hexane in which a glove wassubmerged for 24 hours.17053.,:.Intensity tJR TIC4.Oe+0075 10 15 20 25 30 35 40 45 501-4.Oe÷0073. Se+007 3. 5e+0073.Oe+007-i 3.Oe+0072.5e+007 2.5e+0072.Oe+007 -2.Oe+0071.5e+007 1.5e+0071.Oe+007 1.Oe+0075e+006 5e+006Oe+000-________I 45L-Oe+000TimeFigure A-4. GCIMS Total Ion Count of SPMD Blank Dialysate (unspikedSPM]), 2-stage dialysis).171APPENDIX B: RAW EXPERIMENTAL DATATable B-i. Polyethylene Permeability Data (Figure 9).distance from distance frombeginning beginning [2,2’,5,5’-TCBJSample # of roll (ft) of roll (m) in dialysate (ppb)1-1 0 0 460.81-2 4 1.22 532.61-3 8 2.44 405.72 125 38.1 427.53 250 76.2 419.44 375 114.3 397.95-1 500 152.4 400.85-2 504 153.6 490.55-3 508 154.8 368.46 625 190.5 465.67 750 228.6 465.68 875 266.7 427.49 1000 304.8 435.410-1 1125 342.9 500.810-2 1129 344.1 398.810-3 1133 345.3 384.511 1250 381 413.5172Table B-2. Dialytic Recovery Data (Figure 13).Sample a-pinene guaiacol TCB DHArecovery (%) recovery (%) recovery (%) recovery (%)S-S-i 42.3 52.4 76.8 54.3S-S-2 48.5 53 74.3 59.9S-S-3 42.1 50 67.5 58S-R-i 48.2 55.7 71.3 67S-R-2 56.9 66.2 79.5 85.1S-R-3 48.7 57.8 84.7 71.1where S-S are spiked, static (one-stage) dialysesand S-R are spiked, replaced (two-stage) dialysesTable B-3. Rinsate Analysis Data (Figure 14).dialytic rinsate totalrecovery (%) recovery (%) recovery (%)alpha-pinene 42.3 7.5 49.8guaiacol 52.4 4.9 57.3TCB 76.8 10 86.8173Table B-4. Total Uptake Data (including non-conventional SPMDs) (Figures 15 and 16).Sample a-pinene TCB DHAuptake (%) uptake (%) uptake (%)10/lday/1 14 34.7 15.910/lday/2 0* 0* 10.410/2days/1 7.3 39 24.2lOI2daysI2 13.1 52 24.410/4days/1 9 37.3 18.5lOI4days/2 9.5 36.2 15.310/7days/1 14.9 54.4 2510/7daysl2 11.1 50.5 22.2100/lday/1 1.2 11.3 6.1100/lday/2 0 19.8 7.3100/2days/1 6.5 24.6 14.5100/2days/2 8.7 34.8 42.1100/4daysll 0 16.5 16.3lOOI4daysI2 0 22.2 8.5100/7daysll 12.1 49.8 20.7lOOI7daysl2 12.2 39.2 18.5tubingl 13.3 46.9 6.6tubing2 12.5 46.5 18.6non-spreadi 17.5 60.9 19.4non-spread2 15.6 44.8 19.4where * indicates sample was rotoevaporated to dryness174Table B-5. Distribution of Test Compounds in SPMDs and Aqueous Phases (Figure 17).Sample paiacol (%) aipha-pinene (%) TCB100/lday/142.5 26 0lOOIldayl243.8 0 19.3100/2days/110 14 0100/4days/2 if32.1 0 7.7tubmgl50.2 0 0Shaded areas contain SPMD values; unshaded areas contain aqueous phase values.175Table B-6. Static Wastewater Exposures: Effect of Treatment on SPMD Uptake in1 Day, Unaerated Exposures (Figure 20).ret time untreated treated lab.blank18.67 79.4 80 105.824.48 245.1 111.3 5.3135.13 134.9 95.7 039.45 77.19 60.18 2.9343.07 88.6 23.1 043.54 42.62 4.73 047.12 97.8 44.3 3.2149.21 613.2 150.4 056.33 117.7 104.7 69.160.27 267 207.3 061.15 531.6 448.7 33.162.89 162.3 171.9 10.6176Table B-7. Static Wastewater Exposures: Effect of Treatment on SPMD Uptake in 9Day, Aerated Exposures (Figure 21).ret.time untreated treated lab.blank18.67 89.3 36.8 72.524.48 247.7 13.2 7.934.46 69.8 32.5 035.13 117.1 29 8.935.49 113.2 48.9 038.7 93.5 32 039.45 387 103.6 3.439.77 84 24.1 040.51* 74.1 35.4 48.443.07 112.9 5.6 3.343.54 123.2 2.2 1.444.81 76.3 5.7 045.22 80.4 8.7 047.12 265.8 26.4 5.547.82 98.5 12.3 049.21 612.3 26.7 052.76 81.6 19.7 054.06 183.9 67.3 0.754.68 130 132.3 2.756.33 464.4 100.3 057.79 375 17.2 060.27 2083.4 430.4 060.59 150.2 16.2 061.15 4402.3 0 062.89 1226.1 155.3 065.85 79.2 0 0* indicates BHT (tubing additive)177Table B-8. Static Wastewater Exposures: Effect of Aeration on SPMD Uptake in 9Day Exposures to Untreated Wastewater (Figure 22).Part 1Part 2Ret.time non-aerated aerated18.68 89.7 89.319.34 70.1 21.924.5 468.2 247.735.14 354.2 117.238.69 144.6 93.539.44 492.5 38743.08 331.2 112.943.54 174.6 123.243.86 235.5 77.347.13 367.3 265.847.83 249.2 98.549.24 1797.8 612.350.82 253.4 12755.46 5582.7 5599.656.36 800.3 464.458.77 3174.5 366460.31 2031.7 2083.461.21 4148.8 4402.362.94 1414.3 1226.1178Table B-9. Static Wastewater Exposures: Effect of Exposure Time on SPMD Uptakein Exposure to Treated, Unaerated Wastewater (Figure 23).ret.time I day 3 days 9 days18.67 87.4 82 48.424.49 67.2 98.7 99.435.14 101.1 104.7 154.139.43 69.8 128.2 198.840.5 66.5 49.4 50.447.11 52.4 62.8 60.247.81 90.3 135.7 164.949.21 175.2 184.6 148.954.66 71 145.9 207.355.46 772 1286.2 1414.456.35 129.4 211.4 210.457.86 88 95.1 158.958.72 460.1 744.3 919.760.29 302.8 486.4 580.561.20 603.6 1114.5 1587.962.91 246.6 498.8 654.8179Table B-1O. Static Wastewater Exposures: Effect of Exposure Time on SP]vII) Uptakein Exposure to Untreated, Aerated Wastewater (Figures 24 and 25).ret.time 1 day 3 days 9days18.67 119.3 100.9 89.324.49 329 300 247.735.14 108.5 102.5 117.239.43 263.9 189 38740.50 51.84 73.44 74.0747.11 210.4 190.3 265.847.81 35.53 44.04 98.5349.21 756.6 643.3 612.354.66 65.39 75.99 130.0455.46* 3130.2 3349 5599.656.35 362.6 373.9 464.457.86 247.03 266.4 374.9558.76* 1528.8 1949 366460.29 1331.2 1092.6 2083.461.2* 2268 2368.5 4402.362.91 610.4 702.3 1226.1where * indicates values used for Figure 25180Table B-il. Static Wastewater Exposures: Uptake of Potential Tainting Agents intoSPMDs (figures 26-30).sample 2M5 4M1 AP 1M4 Z-OCTI-A-i 0 0 16.65 329.03 28.02I-A3 2.87 2.13 12.16 300.2 26.74I-A-9 4.85 0 20.06 184.38 24.88I-NA-i 4.53 5.86 22.57 261.18 19.43I.NA-3 8.5 7.53 34.12 312.38 32.4I-NA-9 18.66 15.28 70.12 468.2 68.65E-A-i 0 0 0 45.61 6.28E-A-3 0 0 0 47.12 14.44E-A-9 0 0 0 13.15 3.41E-NA-l 0 0 4.69 67.21 17.4E-NA-3 0 0 4.01 58.46 21.72E-NA-9 2.22 2.07 8.88 99.36 27.85where 2M5 = 2-methy1-5-(l-methylethyl)-bicyclo[3 .1 .Ojhex-2-ene4M1 = 4-methyl-i-(1-methylethyl)-bicyclo[3.1 .0]hexaneAP = aipha-pinene1M4 = i-methyl-4-(1-methylethyl)-cyclohexanez-oct = (z)-3,7-dimethyl-1,3,6-octatriene181Table B-12. Continuous Flow Wastewater Exposures: Effect of WastewaterConcentration on SPMD Uptake @ 27 Hours (Figure 31).ret.time 0.6% 2.5% 100%, #1 100%, #2 100%, avg7.8 20999 32624 31603 18061 26811.528 82579 104687 102156 105586 10387128.54 0 6112 393362 368315 38083933.25 9655 40757 413628 499155 45639237.32 0 0 195594 203185 19939042.29 0 9603 286131 325128 30563044.86 0 10769 570690 546454 55857251.17 75742 90401 89163 87994 88578.552.65 0 14224 236707 308710 27270966.41 805576 31680 337662 96388 2170251820C.)EC.)000-..C.)C.)C)-•C)reitime3 his,#13hrs,#23hrs,avg9his27hrs,#127his,#227hrs,avg240his,#1240his, #2240his,avg7.8255752222423899.5295423160310589210960293101465514.64310343336332198.52958432177296913093431535305533104415.24619904962555807.55855068719665096761457114718026445819.14335973026731932632256890967363681363886336989379262890550729728176189525105586102156103871115779128834122306.528.541103921175161139542705%368315393362380838.5521665482366502015.530.03320813489833489.590942134793137874136333.535147031375633261333.25484318088064655.547974499155413628456391.532674182728302299786034.373590738575372413358725347477993657332360404263639336.95438441008107232723971143078135929339503713220481087816120493237.32444346738155907.5130744203185195594199389.535949331515333732342.29373876940653396.5171582325128286131305629.5882512742245812378.543.12106885149180128032.5304872450742475747463244.595585084025289805144.86132780169209150994.536717454645457069055857212124871084435114846146.79516737745664564.5167630280218270813275515.573126464402868764651.1976658640507035475304891638799488578.59202310269097356.552.65504034977750090122860308710236707272708.511883261020886110460666.373fl315014843689.5114586337662963882170251248393244778643C1)00-lTable B-14. Continuous Flow Wastewater Exposures: Uptake of Potential TaintingCompounds into SPMDs (Figures 33-37).sample 2M5 4M1 AP 1M4 Z-OCT2.5/27 0 0 0 0 61122.5/96/1 0 0 0 0 6153100/3/1 0 0 0 33597 110392100/3/2 0 0 0 30267 117516100/3/AVG 0 0 0 31932 113954100/9 6831 8942 0 63225 270596100/27/1 7265 9411 5160 68909 368315100/27/2 7846 7515 4743 67363 393362100/27/AVG 7556 8463 4952 68136 380839100/240/1 15100 27269 21140 38863 521665100/240/2 14233 23676 18475 36989 482366100/240/AVG 14667 25473 19808 37926 502016where 2M5 = 2-methyl-5-(1-methylethyl)bicyclo[3.1 .Olhex-2-ene4M1 = 4.-methyl-1-(1-methylethyl)bicyclo[3.1.OjhexaneAP = alpha-pinene1M4 = 1-methyl-4-(1-methylethyl)-cyclohexaneZ-OCT = (z)- 3,7-dimethyl-1,3,6-octatriene184APPENDIX C: COMPOUNDS IDENTIFIED IN SAMPLE E-NA-9-12-ethyl-3-methyl-oxetanecyclohexanonetolueneisocyanatocyclohexane2-methyl-5-(1 -methylethyl) bicyclo [3.1.0] hex-2-ene3 ,7-dimethylnonane1 ,2-dimethylcyciopentane2, 3-dimethyiphenol1-methyl-4-(1-methylethyl) cyciohexane1,7,7-trimethyi-(. +/-.)-bicycio [2.2.11 heptan-2-onepentadecanealpha., alpha,4-trimethyi-3-cyciohexene- 1-methanolalpha., 4-dimethyl-benzenemethanol3-methyi-6-(1-methylethyi)-cis-2-cyclohexen-1-ol(e)-3-methyl-2-nonene5-methyl-2-(1-methylethyl)phenol1 -(2-hydroxy-5-methylphenyl) ethanone2-(fench-2-yl) fenchane2,3,4,7,8, 8a-hexahydro-3 ,6,8, 8-tetramethyl- ih-3a,7-methanoazulene1-methylene-3-( 1 -methylethylidene)-cyclopentane6-ethenyl-6-methyl-i-( 1-methylethyi)-3-( 1 -methylethylidene)-(s)-cyclohexene1,2,3,5,6,7,8, 8a-octahydro- 1, 8a-dimethyi-7-(i-methylethenyl) naphthalene2 ,7-dimethyl-(e)-3-octen-5-yne3 ,7-dimethyl-(z)- 1,3, 6-octatriene5-( 1 s*)] 1, 3-cyclohexadiene7,1 1-dimethyl-3-methylene-(e)-1,6, 10-dodecatrieneaipha-cubebeneia,2, 3,5,6, 7,7a,7b-octahydro- 1,1,7 ,7a-tetramethyl- lh-cyclopropa [a]-naphthalene1,2 ,4a,5 , 6, 8a-hexahydro-4 ,7-dimethyl- 1-( 1-methylethyl)-naphthalene4-( 1, 5-dimethyl- 1 ,4-hexadienyl)-1-methyl-cyclohexene1,2,3,4 ,4a, 5,6, 8a-octahydro-7-methyl-4-methylene- 1-(1-methylethyl) naphthalene1,2, 4a,5, 8, 8a-hexahydro-4,7-dimethyl- 1-( 1-methylethyl) naphthalene2, 6-dimethyl-6-(4-methyl-3-pentenyl)-bicyclo [3.1.1] hept-2-ene4, 8-dimethyiquinoline1-methyl-3-(1 ‘-methylcyclopropyl) cyclopentene3, 8-dimethyl-5-(i-methylethyl)-1 ,2-naphthalenedione1,1,4, 8-tetramethyl-cis,cis,cis-4 ,7, lO-cycloundectriene1,2,3,4 ,4a,7-hexahydro- 1, 6-dimethyl-4-(i-methylethyl) naphthalene1,2,3,4 ,4a,7, 8, 8a-octahydro- 1, 6-dimethyl-4-( 1 -methylethyl)- i-naphthalenol2-ethylidene- 1,7, 7-trimethyl-(z)-bicyclo [2.2.1] heptane1 ,4-dimethyl-7-(1-methylethyl)-azulene185l-isopropyl-4,5-dimethyl-cyclopentene3 ,5-dimethylphenol, methylcarbamate1 -ethyl-2 ,3,4,5, 6-pentafluorobenzene2-ethylidene- 1, 1-dimethylcyclopentane2-isopropyl- 1 ,3-dimethylcyclopentanepentadecane1-(2-methylene-3-butenyl)-1-(1-methylenepropyl) cyclopropane8-methylene-dispiro [] undecane1,3-dimethyl-tricyclo [,7] decane1 -methyl-4-(methylsulfonyl) bicylo [2.2.2] octane8, 13-epoxy-5.beta., 8.beta.h., 9.beta.h., 1O.alpha. -labd-14-ene.alpha., 4-dimethyl-. alpha. -(4-methyl-3-pentenyl)-3-cyclohexene- 1-methanol8-methylene-dispiro [] undecane3,7,1 1-trimethyl-[s-(z)]-1,6, 1O-dodecatrien-3-ol(z)- 15-tetracosenoic acid, methyl ester7,11 -dimethyl-3-methylene-(e)- 1,6, 10-dodecatrienebis(2 ,2-dimethyl-3(z)-(1 -methylprop-2-enyl)-cyclopropyl)-zinctricyclo [ ,7]decane- 1 -carboxylic acid, silver (1+) salt1-(2-methylene-3-butenyl)-1-(1-methylenepropyl)-cyclopropane(e)-3-tetradecen-5-yne25-acetate-22-ethyl- 15,1 6-dimethoxy-4 ,25-secoobscurinervan-4-olmeso- 1 ,2-di- 1 -adamantyl- 1 ,2-dicyclohexylethanone186APPENDIX D: TEMPERATURE PROGRAMS USED1. 45°C 1 minute, 4°/minute to 270°C @ 10 minutes-used for hexane/triolein purity determination, polyethylene permeabilitydetermination, investigation of the effect of solvent replacement on tubingconsistency, and glove extract determination.2. 40°C @ 1 minute, 15°/minute to 250°C @ 5 minutes-used for determination of the effect of solvent exposure time on tubing consistency.3. 45°C @ 1 minute, 8°/minute to 290°C @ 15 minutes-used for dialytic recovery and total uptake determinations, including aqueous sampleanalysis.4. 45°C @ 1 minute, 8°/minute to 290°C @ 4 minutes-used for static wastewater exposure analysis.5. 45°C @ 1 minute, 8°/minute until 3 minutes, then 5°/minute until 20 minutes, then3°/minute to 290°C-used for continuous flow wastewater exposure analysis.6. 45°C @ 1 minute, 15°/minute until 10.3 minutes, then 2°/minute until 22.3 minutes,then 3°/minute-used for resin acid analysis.187


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items