UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The influence of the high molecular weight fraction of bleached kraft mill effluent on the biological… Bullock, Christopher Mark 1994

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

Item Metadata


831-ubc_1994-0495.pdf [ 2.57MB ]
JSON: 831-1.0075255.json
JSON-LD: 831-1.0075255-ld.json
RDF/XML (Pretty): 831-1.0075255-rdf.xml
RDF/JSON: 831-1.0075255-rdf.json
Turtle: 831-1.0075255-turtle.txt
N-Triples: 831-1.0075255-rdf-ntriples.txt
Original Record: 831-1.0075255-source.json
Full Text

Full Text

THE INFLUENCE OF THE HIGH MOLECULAR WEIGHT FRACTION OF BLEACHEDKRAFT MILL EFFLUENT ON THE BIOLOGICAL ACTIVITY OF ACTIVATED SLUDGEbyCHRISTOPHER MARK BULLOCKB.A.Sc., The University of British Columbia, 1992A THESIS SuBMITTED IN PARTIAL FULFILMENTOF THE REQUIREMENTS FOR THE DEGREE OFMASTER OF APPLIED SCIENCEinTHE FACULTY OF GRADUATE STUDIES(Department of Forestry)to the required standardTHE UNIVY OF BRITISH COLUMBIAtember, 1994We accept this thesis as conforming© Christopher Mark Bullock, 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.(Signature)_____________________Department of 3eEsn2_’tThe University of British ColumbiaVancouver, CanadaDate t99DE-6 (2/88)ABSTRACTTHE INFLUENCE OF THE HIGH MOLECULAR WEIGHT FRACTION OF BLEACHEDKRAFF MILL EFFLUENT ON THE BIOLOGICAL ACTIVITY OF ACTIVATED SLUDGEChris Bullock Supervisor:University of British Columbia Dr. J.N. SaddlerBleached kraft mills have reduced the environmental impact of their effluent discharges by investingin biological treatment processes such as activated sludge treatment. Since the high molecular weight(HMW) fraction of bleached kraft mill effluent (BKME) represents a significant fraction of theadsorbable organic halide (AOX) and chemical oxygen demand (COD) load on biological treatmentsystems, the influence of this material on activated sludge activity was studied. Initially, we usedultrafiltration to fractionate the BKME from a mill producing elemental chlorine free softwood pulp.The HMW material (> 1000 Da) fraction contained 20% of the chemical oxygen demand and 75-85%of the AOX. Most of the nitrogen present in the effluent from the pulping and bleaching process wasdetected in the HMW fraction. Transition metals were associated with the BMW material to a greaterextent than the less charged alkaline earth metals. Batch growth trials with microorganisms from alaboratory-scale activated sludge reactor were performed using the low molecular weight (LMW)fraction as the substrate and supplementing it with varying amounts of the HMW material. While boththe LMW and HMW fractions alone supported little microbial growth, the HMW material stimulatedmicrobial activity in the LMW fraction, by apparently providing the limiting nutrient, nitrogen. Theaddition of the BMW material to the LMW effluent increased the removal of various chlorinatedphenols (6-chiorovanillin (6-CVa), 4,5-dichloroguaiacol (4,5-DCG), and 2,4-dichiorophenol (2,4-DCP)),which had been added (1.0 mgIL) to the effluent fractions. Supplementing the LMW effluent withnutrients also stimulated chlorinated phenol removal. The relative removal rates of the chlorophenolsdecreased in the order 6-CVa > 4,5-DCG > 2,4-DCP. Spiking the unfractionated effluent with up to10 mgfL of 6-CVa, 4,5-DCG, and 2,4-DCP had no effect on the biomass production or substrateremoval. These results seemed to demonstrate that the BMW material in the BKME had a stimulatoryeffect on the microorganisms present in the activated sludge by providing the limiting nutrient, nitrogen.11TABLE OF CONTENTSABSTRACT.iiTABLE OF CONTENTS iiiLIST OF TABLES vLIST OF FIGURES viLIST OF ABBREVIATIONS viiiACKNOWLEDGEMENTS ix1. iNTRODUCTION 11.1. Kraft pulping and bleaching 31.2. Recent modifications to the pulping and bleaching process 41.3. Biological Treatment 61.4. Bleached Kraft Mill Effluent (BKME) Characteristics 101.4.1. Nature of Adsorbable Organic Halide (AOX) 101.4.2. Molecular weight distribution of BKME Low Molecular Weight Fraction High Molecular Weight Organics 141.5. Effect of Biological Treatment on the Effluent 171.5.1. AOX Removal 171.5.2. Chlorinated Phenolic removal 181.5.3. High Molecular Weight Material removal 201.6. Objectives 242. MATERIALS AND EXPERIMENTAL METHODS 262.1. Materials 262.1.1. Bleached Kraft Mill Effluent 262.1.2. Model Chlorinated Phenolic Compounds and Standards 262.2. Analytical Methods 282.2.1. Chlorinated Phenolic Analysis 282.2.2. Chemical Oxygen Demand (COD) 292.2.3. Biochemical Oxygen Demand - 5 day (BOD5) 302.2.4. Volatile Suspended Solids (VSS) 312.2.5. Adsorbable Organic Halide (AOX) 312.2.6. Phenol-Sulphuric Assay 332.2.7. Somogyi-Nelson Reducing Sugar Assay 332.2.8. Protein Determination of Sludge 332.2.9. Total Kjeldahl Nitrogen 352.2.10. Ammonia 352.2.11. Nitrate+Nitrite, (NO3- 02or NOR) 352.2.12. Total Phosphorous 362.2.13. Inductively Coupled Plasma (ICP) Metal Scan 361112.3. Experimental Methods.372.3.1. Effluent Molecular Weight Fractionation 372.3.2. Activated Sludge Reactor Operation 402.3.3. Batch Growth Trials 413. RESULTS AND DISCUSSION 423.1. Activated Sludge Reactor Performance 423.2. Effluent Fractionation and Characterization 433.2.1. Membrane Retention Properties 433.2.2. Ultrafiltration system performance during effluent fractionation anddiafiltration 473.2.3. Molecular Weight Mass Balance 503.3. Biomass Determination 583.4. Chlorophenol Analysis and Calculations 623.5. Batch Growth Experiments 663.5.1. Effect of High Molecular Weight material on microbial growth 663.5.2. Factors influencing growth on the LMW fraction 693.5.2.1. COD storage losses in the LMW fraction 693.5.2.2. Effect of initial COD concentrations on growth 693.5.2.3. Microbial Toxicity of the LMW fraction 723.5.2.4. Nutrient Deficiency in the LMW fraction 733.5.3. Effect and Behaviour of Model Chlorinated Phenolics 763.5.3.1. Effect of Chlorophenolic Concentration 763.5.3.2. Effect of the High Molecular Weight Fraction on ChlorophenolRemoval 784. CONCLUSIONS 915. LITERATURE CITED 93ivLIST OF TABLESTable 1. Important classes of low molecular weight organic compounds identified in effluentfrom mills producing bleached kraft pulp 12Table 2 - Some properties of the model chlorinated phenolic compounds used in thisresearch 25Table 3 - Volumes employed in the diafiltration process 39Table 4 - Chemical characteristics of the effluent fractions 51Table 5 - Nutrient and trace metal content of the different molecular weight fractions 55Table 6 - Chlorophenol parameters used for their analysis 65Table 7 - Substrate utilization, biomass production and yield after 88 hours (from the datashown in Figure 19) 70Table 8 - Effect of nutrient addition to LMW effluent on substrate utilization, biomassproduction and yield 74VLIST OF FIGURESFigure 1 - Flowsheet of a typical activated sludge system used by the pulp and paper industry(from McCubbin, 1992b) 8Figure 2 - Phenolic compounds most frequently identified in spent liquors (adapted from Vosset al, 1981) 14Figure 3 - Important reactions of phenolic lignin structures in the ffrst bleaching stage(R=residual lignin). Adapted from: Mörck et al, 1991 15Figure 4 - Chemical structure of the chlorinated phenolic compounds used in this study 27Figure 5 - Schematic diagram of the laboratory scale activated sludge reactor 40Figure 6 - Ultrafiltration system operating with retentate recycle 44Figure 7 - Retention and percent breakthrough at different volume reductions of (a) vitamin B12,(b) bacitracin before effluent fractionation, (c) dextran, and (d) bacitracin añer effluentfractionation 46Figure 8 - Crossfiow and transmembrane pressure effect on effluent permeate flux 48Figure 9 - Permeate flux during the initial HMW/LMW fractionation process 49Figure 10 - Permeate flux during the diafiltration operation 49Figure 11 - COD mass balance of the entire fractionation and diafiltration process 51Figure 12 - Correlation between the solid COD and VSS measurements made at the start andend of the growth experiments (R2=0.84) 60Figure 13 - GC-ECD chromatography of (a) chlorophenol standards spiked in deionized water(calibration standard) and (b) a typical spiked effluent samples from a growthexperiment 63Figure 14 - Individual molecular weight fractions original effluent concentrations. (a) biomassconcentration; and (b) substrate concentration. Error bars indicate standard samplingerror; n=3 67Figure 15 - Effect of varying amounts of 11MW material added to the LMW fraction (a)biomass concentration; and (b) substrate concentration. Error bars represent standardsampling error; n=3 68Figure 16 - Effect of the addition of HMW material to LMW effluent diluted to the same initialconcentration. (a) biomass production; and (b) substrate utilization. Error bars representstandard sampling error; n=3 71Figure 17 - Glucose removal with varying amounts of LMW effluent 73vFigure 18 - Effect of 100 ig/L, 1.0 mg/L, and 10 mgfL of 6-CV, 2,4-DCP, and 4,5-DCG on:(a) biomass production; and (b) substrate removal. Error bars represent standardsampling error;n=3 77Figure 19 - LMW and LMW+3xHIVIW effluent with and without 1.0 mg/L of 6-CV, 2,4-DCPand 4,5-DCG; (a) biomass production; and (b) substrate removal. Error bars indicatestandard sampling error; n3 80Figure 20 - 6-CV, 4,5-DCG, and 2,4-DCP concentrations in inoculated and abiotic flasks ofspiked LMW and LMW+3xHMW effluent. Error bars indicate standard sampling error;n=3 81Figure 21 - LMW and LMW+3xHMW effluent with and without 1.0 mg/L of 6-CV, 2,4-DCPand 4,5-DCG diluted to same initial COD. Error bars indicate standard sampling error;n=3 84Figure 22 - 6-CV, 4,5-DCG, and 2,4-DCP concentrations in inoculated and ‘abiotic’ LMW andLMW+3xHMW flasks diluted to the same initial COD concentration. Error bars indicatestandard sampling error; n=3 85Figure 23 - LMW, LMW+3x}-IMW, and HMW effluent with and without 1.0 mg/L of 6-CV,2,4-DCP and 4,5-DCG diluted to the same initial COD with nutrients supplemented.Error bars indicate standard sampling error; n=3 87Figure 24 - 6-CV, 4,5-DCG, and 2,4-DCP concentrations in inoculated and ‘abiotic’ flasks ofspiked effluent fractions diluted to the same initial COD and supplemented withnutrients. Error bars indicate standard sampling error; n=3 89viiLIST OF ABBREVIATIONS2,4-DCP 2,4-dichlorophenol4,5-DCG 4,5-dichloroguaiacol4,5,6-TCTMB 4,5,6-trichlorotrimethoxybenzene6-CVa 6-chlorovanillinAA amino acidsADMT air dried metric tonneAOX adsorbable organic halideASB aerated stabilization basinAST activated sludge treatmentBCA bicinchoninic acidBCF bioconcentration factorBKME bleached kraft mill effluentBOD biochemical oxygen demandBSA bovine serum albuminCOD chemical oxygen demandDO dissolved oxygenECF elemental chlorine freeEOX extractable organic halideGC gas chromatographHMW high molecular weightHPLC high pressure liquid chromatographyHRT hydraulic retention timeICP inductively coupled plasma metals scanI( octanol-water partitioning coefficientLC50 lethal concentration (50% mortality)LMW low molecular weightNMWL nominal molecular weight limitPCDD polychlorinated dibenzo-p-dioxinPCDF polychlorinated dibenzofuranPCP pentachlorophenolRRF relative response factorSEC size exclusion chromatographySRT solid retention timeTCF totally chlorine freeTCP trichiorophenolTKN total Kjeldahl nitrogenTMP transmembrane pressureTSS total suspended solidsVSS volatile suspended solidsYNB yeast nitrogen basevrnACKNOWLEDGEMENTSThis thesis is the final product of my research work which would not have been completed without thesupport and assistance of many people. I would like to thank my supervisor, Dr. Jack Saddler, forgiving me the opportunity to work with the excellent people and resources that make up the ForestProduct Biotechnology research group at UBC. The sponsorship by the Science Council of BritishColumbia and Weyethaeuser Canada is gratefully acknowledged.In particular, I am indebted to Paul Bicho for his contributions to the development of this project andhis patient review of my ideas, experiments, and manuscripts. I would also like to thank Vince (the‘King’) Martin for sharing his experience with gas chromatography and effluent chemistry with me.In addition, Dave Scott’s time and dedication assisting with keeping the reactor organisms happy overthe many months of operation were much appreciated. Nancy Thompson’s assistance with negotiatingthe UBC bureaucracy was very helpful. To the many colleagues at FPB, Paprican, and UBC that havenot been named individually, I have valued our association.The support from my family has been incredible. My wife, Marefe, has been remarkable with herunselfish support throughout my education and raising our two wonderful boys, Thomas and Simon.I can only hope to repay her for her patience. Lastly, I would like to thank my parents for theirencouragement throughout my studies.ix1. INTRODUCTIONThe discharge of halogenated organic material into the environment has become a concern in manycountries, as many of these compounds are toxic and persistent. The discovery of chlorinated dioxinsand furans in bleached kraft mill effluent (BKME) in the 1980’s focused attention on organochiorinebyproducts formed during pulp bleaching with chlorine and chlorine dioxide. As a result, manyjurisdictions have considered or developed effluent regulations restricting the discharge of chlorinatedmaterial measured as adsorbable organic halide (AOX) (Reeve and Earl, 1989).Since 1988, the Canadian pulp and paper industry has invested nearly $2 billion on new pulping andbleaching technologies and improved water treatment which has reduced dioxin and furan dischargesby 97% and AOX releases by 67% (CPPA, 1992). While totally chlorine free bleaching technologycontinues to be developed, the substitution of elemental chlorine with chlorine dioxide has been theprimary way of achieving the reductions in AOX and chlorinated dioxin formation. This largeinvestment in increased chlorine dioxide generating capacity indicates that low levels of AOX willprobably continue to be produced in the future.Many pulp mills have invested in biological treatment facilities adapted from municipal sewageapplications to remove readily biodegradable organics, total suspended solids (TSS) and toxicity. Thesesystems have been less successful at removing AOX, chemical oxygen demand (COD), and colour.Biological treatment has demonstrated variable AOX removal ranging between 15-65% (Randle et a?,1991), which indicates the need for a better understanding of the behaviour of AOX material in theprocess to achieve better and more consistent AOX removal.Organochiorines in pulp mill effluent consist of a wide range of compounds that have differentmolecular weights and structures. To simplify the study of the chlorinated organic material,Iultrafiltration fractionation has often been used to separate low molecular weight (<1000 dalton)compounds from high molecular weight (>1000 dalton) material (Kringstad and Lindsträm, 1984). Thelow molecular weight (LMW) fraction is readily analyzed and considered more biologically active.Over 300 low molecular weight (LMW) compounds have been identified in bleached kraft mill effluent(BKME) (McKague et al, 1989; Suntio et at, 1988). Since the virtual elimination of dioxin and fliranat most mills, among the most toxic and persistent of the low molecular weight compounds remainingare the chlorinated phenolic compounds. The fate of these compounds in both the biological treatmentsystem and the receiving waters is an important consideration for managing the impact of effluentdischarges.The high molecular weight material contributes 50-80% of the AOX and has been assumed to bebiologically benign since its large molecular size restricts passage through cell membranes (Kringstadand Lindström, 1984). However, the HMW material has been associated with toxicity to marineorganisms (Higashi et a!, 1992) and contributes a substantial portion of the non-degradable chemicaloxygen demand, AOX and colour which passes through biological treatment plants (Yin et a!, 1990).Although the HMW component of the effluent is believed to be biologically inactive, this material couldhave some indirect effects on the microbial population and the behaviour of more problematiccomponents of the effluent during biological treatment. For example, the HMW material may have astimulatory or inhibitory effect on microbial activity by interacting with trace nutrients or toxic metals(Benes et at, 1976) or physical interference with cell membrane transport (Freeman and Lock, 1992).Similarly, the 11MW material has been shown to interact with chlorinated phenolic compounds(O’Connor and Voss, 1992), altering their bioavailability to microorganisms (Robinson and Novak,1994) and invertebrates (Kukkonen, 1992). The objective of this thesis was to better understand therelationship between the 11MW material, chlorinated phenols and microorganisms found in a biologicaltreatment system.21.1. Kraft pulping and bleachingMany pulping processes are used to produce different pulp and paper products. However, theproduction of chlorinated organics occurs primarily at mills bleaching chemical pulp with chlorine-basedbleaching agents. The most common chemical pulping processes include the acidic sulphite pulpingprocess or the alkali kraft (sulphate) pulping process. Kraft pulping technology has gained popularitybecause of its efficient recovery of pulping chemicals and ability to generate energy from thecombustion of solubilized lignin while producing a high strength pulp.The purpose of chemical pulping is to liberate the cellulose fibres for paper production from thepolymeric lignin matrix that gives rigidity and protection to plants. The lignin biopolymer, whichmakes up 16-24% of hardwoods and 24-33% of softwoods, is a complex of phenyl propane units whichform aromatic ring structures. The aromatic ring structure is relatively stable and recalcitrant in thenatural environment and forms the basis of naturally occurring humic material. Residual lignin in paperproducts is undesirable since it decreases strength properties and absorbs visible light thereby reducingpaper brightness. Both suiphite and kraft pulping processes use chemicals and heat to solubilize anddepolymerize lignin for removal. Approximately 90 to 95% of the lignin can be removed during thepulping process without damaging the cellulose fibres. After the kraft cooking process, the fibres inthe cooked wood chips are separated, screened and washed. Unbleached fibres constitute 40 to 45%of the original wood dry weight. Spent cooking chemicals and solubilized lignin (‘black liquor’) areconcentrated and combusted in the recovery boiler to generate steam for the process and to recover thecooking chemicals.The residual lignin (5-10%) is responsible for the brown colour in the unbleached fibres or ‘brownstock’. A multi-stage bleaching sequence is often used to remove the colour contributed by residualLignin without damaging the fibres. Traditionally bleaching sequences consisted of an initial3delignification step with chlorine (C-stage) followed by an alkali extraction stage (E-stage), a firstchlorine dioxide (Di) or hypochlorite step (H), a second extraction (E2) and a final chlorine dioxide step(D2). The chlorine reacts to further depolymerize and solubilize the residual lignin and forms chlorideions (—9O% of the Cl2) and some organochlorine compounds (Berry, 1993). The high chloride contentof bleach plant effluent poses a corrosion risk if recycled, therefore the effluent with its associatedorganochiorine material is usually discharged into a nearby water body. The recent emphasis onreducing the discharge of chlorinated organics has resulted in many process changes in the pulping andbleaching operations.1.2. Recent modifications to the pulping and bleaching process.Historically, pulp mills bleaching with chlorine produced 5 to 10 kg AOX per air-dried metric tonne(ADMT) of pulp. However, modem pulping and bleaching processes have reduced this to less than 2kg AOXIADMT. The recent investment in new pulping and bleaching technologies has been primarilydriven by the effort to eliminate chlorinated dioxin formation and reduce AOX production. Twostrategies which have been used to reduce the production of AOX are reducing the amount of unwantedorganic material (e.g. lignin) entering the bleach plant; and finding altematives to bleaching withelemental chlorine.To reduce the amount of lignin entering the bleach plant, extended delignification pulping processes andoxygen delignification stages have been incorporated at many mills (Macleod, 1993). A number of newdigester technologies have been developed to effectively reduce the kappa number (a measure of thelignin content) of unbleached kraft pulp by nearly half of the value achieved by traditional cookingsystems. Some examples of these alternative technologies include Ahlstrom/Kamyr’s ExtendedModified Continuous Cooking (EMCC) and Kvaemer Pulping’s IsoThermal Cooking (ITC) forcontinuous digesters, and Sunds Defibrator’s SuperBatch, Beloit’s Rapid Displacement Heating (RDH),4and Voest Alpine Industieanlagenbau’s Enerbatch systems for batch digesters (Patrick et al, 1994). Aneffective oxygen delignification stage followed by advanced pulp washing and pressing can also reducethe final kappa number by a further 50% (Patrick et al, 1994). Puips with lower kappa numbersentering the bleach plant require less bleaching chemicals and generate less organic material fordischarge from the mill.There have been rapid changes in bleaching technologies ever since concerns about dioxin were firstreported. The traditional use of elemental chlorine (Cl2) in the first bleaching stage has been largelysubstituted by chlorine dioxide (C102). The use of 100% chlorine dioxide is referred to as elementalchlorine free (ECF) bleaching. Chlorine dioxide has 2.5 times the oxidizing power of elementalchlorine which reduces the total amount of chlorine (Cl) applied and the amount of AOX produced.In addition, the nature of the AOX produced in the effluent is different (see section 1.4) as chlorinedioxide reacts with the lignin by oxidation rather than substitution (i.e., chlorinationXAxegrd et al,1993). Non-chlorine based bleaching agents and enhancers such as ozone, peroxide, activated oxygen,xylanases, etc. are used to produce totally chlorine free (TCF) pulp. These new bleaching agents haveyet to achieve the same pulp brightness levels that can be achieved with chlorine and chlorine dioxidewithout sacrificing pulp strength. Therefore, a final chlorine dioxide brightening stage will probablycontinue to be used for some time in the future, extending the production of small amounts of AOX.Increasing environmental pressures and ever-more stringent discharge regulations are making closedcycle or zero effluent technology more attractive. Over the past decade, the overall water consumptionin pulp and paper mills has steadily decreased. While some major breakthroughs have been achievedwith mechanical pulping and recycled operations, completely closing bleached kraft mills has been anelusive goal. Many of the process improvements used to reduce AOX discharges have also greatlyreduced water consumption within the mills. Reducing the kappa number prior to the bleach plant has5reduced bleaching chemical and washing requirements. Many of the new bleaching agents, such asoxygen and ozone, are applied to high consistency puips (>10% pulp by weight), instead of thetraditional low consistency (3%) pulp suspensions. These water reductions have directly impacted onexisting water treatment facilities by reducing pumping requirements, increasing hydraulic residencetimes, and altering the temperature, effluent concentrations and loading. Even if completely closed-cycle mill technology is achieved, it is likely that biological treatment will be required to clean up someeffluent streams prior to recycling.1.3. Biological TreatmentThe pulp and paper industry has invested heavily in biological treatment facilities over the past 30 years,and nearly all North American operations use biological treatment for their effluent discharges. Withover 600 installations in the North American pulp and paper industry, the dominant biological treatmentprocesses are aerobic and include activated sludge treatment (AST) and aerated stabilization basins(ASB) (McCubbin, 1992a). In contrast, anaerobic processes were operating or under construction atonly 40 mills in 1991 (Springer, 1993). Biological treatment facilities have been designed to removebiodegradable organic compounds, suspended solids (TSS), and acute toxicity from pulp mill effluents.However, the desire to remove organochiorines requires a greater understanding of the biologicalprocesses within these systems.While anaerobic treatment has several advantages over aerobic treatment including lower biosolidproduction, decreased electricity requirements, and the production of methane for energy use, the longstart-up and recovery time from upsets and the inability to produce an acceptable effluent for dischargewithout an aerobic polishing step have made the pulp and paper industry reluctant to invest in thistechnology (McCubbin, 1 992b). Anaerobic treatment facilities are usually used for high strength wastessuch as mechanical pulping effluent rather than the relatively weak bleached kraft mill effluent.6However, incorporating anaerobic processes into biological treatment facilities could become moreattractive with the need to remove problematic organochlorines since highly chlorinated compoundsoften require dechlorination before aerobic organisms can mineralize them (Haggblom, 1992). Forexample, this process has been observed in the anoxic sludge bed of aerated lagoons (Bryant et a!,1988).As mentioned previously, aerated stabilization basins are the predominant form of biological treatmentin the North American pulp and paper industry (McCubbin, 1992a). The ASB is simply a large basinwhere microorganisms are allowed to metabolize the soluble organic material in the effluent. Oxygenis provided by mechanical or diffused aeration units and induced surface aeration. Turbulence in thebasin is required to suspend actively growing biomass and to distribute the soluble substrate for theorganisms to metabolize. Biomass produced is allowed to settle and become a substrate for othermicroorganisms which reduces sludge disposal and nutrient requirements. Properly operated ASBsystems use less energy than activated sludge systems and avoid sludge generation. These systemsrequire large amounts of land as the hydraulic retention times range between 5 to 10 days and the depthis usually less than 5 metres.Until recently, nearly all Canadian mills with secondary treatment systems had ASB technology(McCubbin, 1992a). However, many secondary treatment facilities constructed in the last decade haveused activated sludge technology (AST) (Rodden, 1994). This process is a high rate biologicaltreatment process, requiring less space and able to produce final effluent with BOD and TSSconcentrations about half of the lowest attainable by ASB treatment (McCubbin, 1 992a). The activatedsludge process was developed around the turn of the century in England for the treatment of municipalwastewaters. The principle is based on growing a settleable sludge of microorganisms from the solublematerial in the effluent. The sludge is settled in secondary clarifiers and recycled into the aeration basin7to establish a concentrated population of microorganisms. Excess sludge is dewatered and eliminatedby landfiuing or incineration. The very high concentrations ofmicroorganisms, supported with nutrientsand oxygen added to the aeration basin, allow rapid BUD removal and relatively short hydraulicretention times. Activated sludge facilities are usually designed with hydraulic retention times of 6-24hours, resulting in compact installations.A typical primary effluent treatment system and activated sludge facility is shown schematically inFigure 1. The effluent passes through a coarse screen to a primary clarifier where settleable solids areremoved by sedimentation. If necessary, the pH of the clarified effluent is neutralized with sulphuricacid or lime. Nutrients and the recycled sludge are mixed with the effluent as it flows into the aerationtank. The mixed liquor, effluent combined with a high concentration of microbial biomass, isvigorously agitated and aerated with either air or oxygen. A mixed population of bacteria, fungi andprotozoa feed on the organic substances, oxygen and nutrients. The mixed liquor flows to the secondaryclarifiers where most of the biomass settles and the clarified effluent is discharged.Biological treatment facilities have been designed primarily to remove acutely toxic and oxygendemanding substances that would have a direct impact on the aquatic life in the receiving waters.Increasing demands to remove substances of chronic concern (i.e., dioxin and furan, chlorophenols,AOX, etc.) requires a greater knowledge of the removal mechanisms to optimize biological treatmentfor the removal of these pollutants. The results presented in this thesis give some insight on thebehaviour of both the 1-1-MW fraction of the AOX and the chlorinated phenolics when being treated withactivated sludge microorganisms.8micxSLUDGEp(SPOSALFigure 1 - Flowsheet of a typical activated sludge system used by the pulp and paper industry (fromMcCubbin, 1992b).1.4. Bleached Kraft Mill Effluent (BKME) Characteristics1.4.1. Nature ofAdsorbable Organic Halide (AOX)Adsorbable organic halide is an aggregate measure of the amount of chlorine that is bound to organicmaterial. To determine AOX, inorganic chloride is separated from the organochiorine by absorbing theorganic material to activated carbon and rinsing with acidic potassium nitrate solution. Combusting thecarbon and AOX produces HCI which can be measured by microcoulimetric titration. Given thecomplex and heterogeneous nature of pulp mill effluent streams, AOX is a good indicator of the totalamount of chlorinated organic material being discharged. However, AOX does not give any indicationof the nature of the chlorinated compounds being measured.COARSESCREENPREESI—9The behaviour of a given chlorinated compound in the environment depends on the molecular structureand degree of chlorination of the compound and the conditions found in the environment. For example,chlorinated phenolic compounds are considered to be an environmental concern due to their acutetoxicity, lipophilicity, and relatively low biodegradation rates. The acute toxicity (EPA criterion) ofchlorinated phenolic compounds ranges from 4380 gIL for 2-chlorophenol to 20 jig/L forpentachlorophenol (NCASI, 1992). The hydrophobicity of chlorinated compounds increases 4 to 6 timeswith each additional chlorine atom (Solomen et a!, 1993). Hydrophobicity is usually expressed as K0,the octanol-water partitioning coefficient. Since the mechanism of bioconcentration is one ofwater-lipidpartitioning, increasing the K0 has been shown to correlate well with increasing bioconcentrationfactors (BCF) (Mackay, 1982). Phenolic compounds with three or more chlorine atoms have anoctanollwater partitioning coefficient greater than 1000 (logK0>3) and are considered to be lipophilicwith the potential to bioaccumulate (BCF>100) (NCASI, 1994b). While many chlorophenols aresubstantially removed during biological treatment, their half lives in the natural environment have beenmeasured in terms of weeks and months (Solomen et al, 1993).Most of the chlorinated organic (AOX) material in bleached kraft mill effluent is not considered to bevery persistent, lipophilic, or toxic. A more useful measure is Extractable Organic Halogen (EOX)which measures the hydrophobic, low molecular weight material that extracts into a non-polar organicsolvent (Craig et at, 1990). The EOX is usually only 1-3% of the total AOX (Berry, 1992).Furthermore, only a small portion of the FOX is considered to be highly lipophilic and persistent (i.e.,phenolic compounds with three or more chlorine atoms). Many experts have been critical of AOX asa legislative parameter because of the broad nature of this measurement (Fleming et a!, 1990; NCASI,1990).101.4.2. Molecular weight distribution ofBKMEThe organic material in bleached kraft mill effluent exhibits a wide molecular weight distribution,ranging from monomeric compounds to larger complex molecules. The molecular weights of the largestcompounds in the effluent range from 10,000 to 30,000 as assessed by aqueous size exclusionchromatography (SEC) (Mörck et al, 1991; Pellinen and Salkinoja-Salonen, 1985). Organic compoundsthat are retained by a 1000 dalton ultrafiltration membrane have been described as high molecularweight (BMW) (Kringstad and Lindsträm, 1984). Determination of the molecular weight distributionof effluent using aqueous SEC or ultrafiltration has shown that the furnish (i.e., wood species),bleaching sequence, and biological treatment all affect the distribution of compounds observed. Effluentobtained from the production of softwood kraft pulp tends to have a higher proportion of high molecularweight compounds than a corresponding effluent derived from hardwood (Dahlman et a!, 1994; Mörcket a!, 1991). Ultrafiltration studies have shown that the additional oxidation that occurs with oxygendelignification and chlorine dioxide substitution, usually produces a higher proportion of low molecularweight material (O’Connor et a!, 1994; Yin et a!, 1990) compared to effluent from elemental chlorinebleaching. Biological treatment has been shown in several cases to remove primarily the low molecularweight fraction, thus increasing the proportion of high molecular weight material in the treated effluent(Joshi and Hillaby, 1991; Dahlman et a!, 1993a; Yin et a!, 1990; Yu and Welander, 1993). Low molecular weight fractionOver three hundred low molecular weight compounds have been identified in bleached kraft milleffluent by several research laboratories. Most of these findings have been summarized by Suntio eta! (1988). Much of the biodegradable material (Dahlman et a!, 1993a; Joshi and Hillaby, 1991; Pfisterand Sjöström, 1978, 1979a, 1979b; Yin et a!, 1990; Yu and Welander, 1993), and potentiallybioaccumulable and toxic compounds are thought to be associated with the low molecular weightfraction (Sameshima et al, 1979, Leach and Thakore, 1975). Based on their chemical properties, the11compounds can be divided into three main classes: acids, neutral compounds, and phenolic compounds(Table I).Table 1. Important classes of low molecular weight organic compounds identified in effluent frommills producing bleached kraft pulp.Acids Phenolic compoundsAlkanoic acids (e.g. formic acid, Phenolsacetic acid, saturated fatty acids) CatecholsAlkenoic acids (e.g. unsaturated Syringol (hardwoods)fatty acids) VanillinsHydroxy acids SyringaldehydesDibasic acids (hardwood)Aromatic acidsResin acidsNeutral compoundsChlorinated alkanes (mainly chloroform) BenzenesAliphatic alcohols (mainly methanol) VeratrolsAromatic and aliphatic aldehydes PCDDs and PCDFsAromatic and aliphatic ketones SterolsSulphones (mainly chlorinated Terpenes anddimethylsuiphones) terpenoids(adapted from Axegárd et a!, 1993)The acidic compounds can be further subdivided into the following categories: fatty, hydroxy, dibasic,aromatic and resin acids. Formic and acetic acid are the predominant fatty acids in bleached krafteffluent. The resin acids and higher fatty acids originate from the wood extractives and can be foundin BKME at concentrations up to 18 mg/L (Taylor et a!, 1988). Hydroxy acids are likely oxidationproducts derived from the wood carbohydrates. Dibasic acids - oxalic, malonic, succinic, and malicacids - are found in both acid and alkali extraction effluent and originate from residual lignin andcarbohydrates. Aromatic acids, related to the phenolic family of compounds, are relatively low inabundance (Kringstad and Lindsträm, 1984).The dominant neutral compound is methanol. The majority of the methoxyl groups, abundant in lignin,are cleaved during the first chlorination stage to form methanol, contributing about 70% of the12biochemical oxygen demand (BOD) in total bleach plant effluent (Pfister and Sjöström, 1979a, 1979b).Compared to methanol, the total amount of chlorinated compounds in the neutral fraction is very small.Chloroform is the predominant chlorinated neutral compound and is produced at concentrations of upto 0.3 kg/ADMT at mills bleaching with hypochiorite (Axegard et at, 1993). The use of chlorinedioxide has substantially lowered the production of chloroform and chioroacetones (Gergov et at, 1988).OCH3ChiorosyringolsCHOclxCHLIOCH3Chiorosyringaldehydes_-Figure 2 - Phenolic compounds most frequently identified in spent liquors (adapted from Voss et a!,1981).A number of phenolic compounds result from the breakdown of lignin in the pulping and bleachingprocesses. Examples of the various species of phenolic compounds found in effluent from bleachingof softwood and hardwood puips are shown in Figure 2. Mills using chlorine or chlorine dioxide asthe primary bleaching agent discharge phenolic compounds with varying numbers of chlorine atomssubstituted on the aromatic ring. The increasing substitution of chlorine dioxide by many mills in thefirst bleaching sequence has reduced both the amount of phenolic compounds and the degree ofchlorination of the aromatic ring. A recent study by National Council for Air and Stream Improvement.*— Softwoods —.-CI Cl,‘ySCH3H OHChlorophenols ChloroguaiacolsChlorovanillins.4ChiorocatecholsHardwoods13(NCASI, 1994a) showed that after biological treatment mills operating with 100% chlorine dioxidesubstitution discharged less than 6% of the total chlorinated phenolic compounds that were detectedfrom mills operating with less than 50% chlorine dioxide substitution. The installation of extendeddelignification or oxygen delignification facilities results in a further reduction in chlorophenolsproduced since less residual lignin enters the bleach plant (NCASI, 1994a). While the amount ofchiorophenols decreased with the implementation of modern bleaching and pulping technologies, thesecompounds continue to be a concern. High molecular weight organicsThe high molecular weight (HMW) fraction is a major contributor to the chemical oxygen demand(COD), AOX, and colour found in bleached kraft mill effluent (Pfister and Sjöström, 1978, l979b).This 1Th4W material consists of residual lignin and hemicellulose broken down and solubilized in thebleaching process (Axegdrd et at, 1993). Most of the HMW material is solubilized following the firstdelignification stage in the caustic extraction. Size exclusion chromatography has revealed that themolecular weight of this material ranges from 1000 to 30,000 dalton (Dahlman et al, 1993 a). Jokelaand Salkinoja-Salonen (1992) have found that the actual molecular weight of the FIMW material basedon aqueous SEC is less than the apparent weight indicated by ultrafiltration. Dilution of the HIvIWretentate was shown to increase the passage of organic material through the membrane, suggesting thatthe HMW material may not be large covalently bonded structures, but more likely an association ofLMW compounds. However, it is unlikely that the highly oxidized and hydrophillic nature of thematerial will support the formation of micellar aggregates, and the increase in LMW material withdilution could be a ‘washing effect’ resulting in improved ultrafiltration (Dahiman et at; 1993a).The HMW material is composed of degraded, strongly oxidized lignin material with a low degree ofchlorination (sometimes called chiorolignin). This material has lost 80-90% of the aromatic character14found in industrial lignin (Mörck et at, 1991), and the methoxyl and phenolic hydroxyl content (<1.8mmollg) is lower than that found in lignin, while the carboxylic acid content is higher (2.4-6.0 mmollg)(Mörck et at, 1991; Dahlman et a!, 1994). The reduction of methoxyl and phenolic hydroxyl contentin conjunction with the increase in carboxylic acid groups is an expected result based on the ringcleavage reactions shown in Figure 3.CIOCH3 OHOH OHOCH3 I OCH3 dOH 0Figure 3 - Important reactions of phenolic lignin structures in the first bleaching stage (Rresiduallignin). Adapted from: Mörck et a!, 1991.The high carboxylic acid content of the HMW material is largely responsible for the high polarity (andtherefore solubility) of this material. Diercks and Banerjee (1993) have demonstrated the lowlipophilicity of the bulk of the chlorinated organic constituents in bleached kraft mill effluent.The carbohydrate concentration in the 11MW fraction varies with the bleaching sequence and the woodfurnish used. Dahiman et a! (1994) found 3.5-18% (w/w) of the HMW material in effluent fromchlorine dioxide bleaching was present as carbohydrate, while totally chlorine free bleaching effluents15were found to contain 13-48% carbohydrate in the HMW fraction. The amount of carbohydrate in theHMW material resulting from a hardwood furnish is higher than when bleaching softwood species(Mörck et al, 1991; Dahlman et al, 1994). The composition of the resulting monosaccharides after acidhydrolysis of the carbohydrate from the different furnishes indicates they originate from thehemicellulose (Mörck et a!, 1991). The carbohydrates in the HMW fraction are susceptible to biologicaldegradation during biological treatment (Dahlman et a!, 1993a).The HMW material has been shown to be less chlorinated than the LMW fraction (Lindström andOsterberg, 1984; Osterberg and LindstrOm, 1985), and the amount of chlorine decreases an order ofmagnitude when chlorine dioxide is substituted for elemental chlorine (Dahiman et a!, 1993a, 1994;O’Conner et a!, 1994). The degree of chlorination in the HMW fraction ranges from 7-11 Cl atomsper 100 C atoms when chlorine bleaching was used and 0.4-1.1 Cl atoms per 100 C atoms when 100%chlorine dioxide substitution was used. After oxidative degradation, more than 90% of the remainingaromatic material in the HMW material from chlorine dioxide bleaching was shown to be non-chlorinated (Mörck et a!, 1991). The small amount of chlorinated aromatic material found was largelymonochiorinated, with only trace amounts of dichlorinated phenolic material. These results indicate thatthe degradation of the HIvIW material from elemental chlorine free (ECF) bleaching should not lead tothe formation of highly chlorinated monomeric phenolic compounds in the receiving water (Dahlmanet al, 1993a). A comparison of ECF 11MW material to naturally occurring HMW humic substances hasrevealed similar mono- and di-chiorinated structures in nature (Dahiman et a!, 1993 b). The degradationof RMW material from elemental chlorine free bleaching and naturally occurring humic material willlikely result in similar monomeric phenolic degradation products (Axegârd et a!, 1993).The HMW material has generally been shown to be non-toxic (Kringstad and Lindström, 1984; Sâgforsand Stark, 1988). Slight toxicity has been detected using bioluminescent bacteria (i.e., Microtox EC5016values between 100 and 325 mgfL; Brezny et a!, 1993). A polar high molecular mass constituent (MW> 10 kDa) has been found to be toxic to the early life stages of marine animals and a plant species(Higashi et a!, 1992). The difficulty in identif’ing HMW constituents to assess their toxicological andbiological properties has discouraged previous researchers from this area of study. One of theobjectives of this thesis is to assess the influence of the HMW fraction on microbial populations.1.5. Effect of Biological Treatment on the Effluent1.5.1. AOX RemovalWhile activated sludge treatment is operated to maximize the removal of such conventional pollutantsas BOD and TSS, some removal of non-conventional pollutants like AOX and chlorinated phenoliccompounds may occur. A summary by Wilson and Holleran (1992) reported that AOX removalefficiencies vary between 14-65% for activated sludge treatment. The main mechanisms for chlorinatedorganic removal are biodegradation, biosorption, volatilization, and air stripping (Leuenberger et a!,1985; Amy et a?, 1988, Randle et a!, 1991). The removal of these pollutants is highly variable andknowledge regarding the optimum conditions for the removal of these parameters is scarce. Somepractical studies perfonned by Randle et a? (1991) to optimize AOX and chiorophenolic removal forAST and ASB processes, found that increasing both temperature and hydraulic retention time (HRT)improved chlorinated organics removal but solids retention time (SRT) did not. However, Rempel eta? (1992) found AOX and chlorophenol removal improved significantly by extending the SRT ratherthan increasing the HRT. Biodegradation appears to be the main removal mechanism accounting formore than 65% of the AOX removed (Randle et a?, 1991). Less than 10% of the AOX removed wasassociated with the sludge wasted from the reactor (Randle et a!, 1991, Saunamaki et a!, 1991).Several researchers have reported increases in the proportion of the HMW material after biologicaltreatment indicating the preferential removal of low molecular weight AOX (Joshi and Hillaby, 1991;17Dahlman et a!, 1993a; Yin et a!. 1990; Yu and WeLander, 1993). However, this finding is not universalas several papers have shown uniform removal of all molecular weight fractions (Graves et a!, 1993;Struthridge et a!, 1991; LindstrOm and Mohamed, 1988; Jokela et a!, 1993). Jokela et al (1993)suggested that the recalcitrant portion of the AOX is not related to molecular weight or the degree ofchlorination, but more on the basis of extractability into an organic solvent (tetrahydrofuran). The lackof consistent findings about the recalcitrant portion of the AOX exemplifies the need for further researchon the behaviour of the HMW fraction in biological treatment facilities.1.5.2. Chlorinated phenolic removalChlorinated phenolic removal in biological treatment systems is usually greater than the removal ofAOX (Gergov et at, 1988; Struthridge et at, 1991; Randle et at, 1991; Lindström and Mohamed, 1988;Häggblom and Salkinoja-Salonen, 1990). The vapour pressure and water solubility of these compoundssuggest that volatilization and air stripping are not important removal mechanisms (Leuenberger et a!,1985). Although the lipophilicity of the higher chlorinated phenols will result in some adsorption tothe biomass (Gergov et at, 1988), less than 10% of the chlorophenol removal can be accounted for bythe wasting of excess sludge in activated sludge treatment (Gergov et al, 1988, Randle et a!, 1991).This would seem to indicate that biodegradation or biotransformation accounts for the majority ofchlorophenol removal.The biodegradation of chlorinated phenolic compounds has been extensively studied and severalmetabolic pathways have been elucidated (Neilson et a!, 1990; Häggblom, 1992). Mono- anddichlorophenols are hydroxylated to catechols followed by cleavage of the aromatic ring. More highlychlorinated phenols must lose chlorine atoms by reductive dechlorination and hydroxylation prior tocleavage of the aromatic ring. These processes can occur under aerobic conditions (Neilson et a!,1990), but usually occur anaerobically (Hãggblom, 1992). While hydroxylation of chiorophenolics18usually leads to more readily degradable metabolites, 0-methylation reactions may lead to more stable,lipophilic and toxic compounds (Allard et at, 1987; Neilson et a!, 1983, 1987, 1988; Remberger et a!,1986). The 0-methylation reaction is performed under aerobic conditions by some Gram-positivebacteria and forms guaiacols, veratroles and anisoles. These compounds are largely bound to sedimentwhere anaerobic demethylation can occur to produce chlorocatechols (Remberger et a!, 1986). The 0-methylation reaction does not appear to be common since veratroles and anisoles are not reported insignificant quantities in biological treatment facilities.The efficiency of treatment systems at removing different phenolic families (i.e. phenols, guaiacols,catechols, etc.) is variable, making conclusions about relative degradation rates difficult to formulate(Struthridge et at, 1991, Lindström and Mohamed, 1988). Activated sludge treatment appears to bemore effective than aerated lagoons at removing chlorophenols (Lindström and Mohamed, 1988).Higher chlorinated phenolics (tn, tetra and pentachlorinated) are degraded at a slower rate than mono-and dichlorinated compounds (NCASI, 1 994a).1.5.3. High molecular weight material removalThe high molecular weight organic material is a significant component of bleached kraft mill effluent,and can interact with both microorganisms and chemical constituents in biological treatment facilitiesand the receiving environment. Several researchers have studied the mechanisms for the long-termdegradation of this biologically recalcitrant material and have observed the relationships between theHMW fraction and chiorophenolic compounds. In addition, a considerable amount of the researchperformed on naturally occurring humic material and its interaction with micronutrients and hydrophobicpollutants could be relevant to the behaviour of 11MW fraction of effluent (Dahlman et at, 1993b)during biological treatment.19Biosorption of AOX to biomass has been shown to be a significant process within biological treatmentfacilities (Bryant et a!, 1988; Amy et a!, 1988; Tomar and Allen, 1991). Bryant et al (1988) and Amyet a! (1988) have suggested that biosorption to settling biomass may be a necessary transport steprequired for anaerobic dechlorination in the benthal layer of aerated lagoons. Recent work hasdemonstrated the HMW AOX has a high potential for sorbing to biomass (Yan and Allen, 1994) andsediment (Pellinen, 1994).Studies on the long-term fate of the biologically recalcitrant HMW material have shown that abioticprocesses contribute significantly to the mineralization of this material. Over a 16 week period, RoyArcand and Archibald (1993) found substantial and continuing dechlorination of the AOX and similardecreases in colour and apparent molecular size via both daylight-dependent and daylight-independentmechanisms. Microbial processes were not significant in the HMW mineralization observed.Continuing studies (Archibald and Roy-Arcand, 1994) found that daylight activated decolourization wasoxygen dependent while AOX mineralization was not. The FIMW photodegradation process functionsat high and low effluent dilutions in both fresh and salt water. The photolytic processes involvedincluded the replacement of chlorine with hydroxyl groups or hydrogen as well as decarboxylation anddecarbonylation reactions to mineralize organic carbon (Sonnenberg et a!, 1994).Neilson et a! (1983) observed the production of small quantities of two trichioroveratrole isomers andtetrachioroveratroles while incubating BMW material with bacterial monocultures. The veratroles werebiological transformation products of phenolic compounds derived from HMW material. Whether thesecompounds were LMW compounds reversibly bound to the HMW material or biological breakdownproducts of the HMW structure was not determined. Eriksson et a! (1985) observed a similarproduction of catechols, guaiacols and veratroles with a number of bacterial consortia and a white-rotfungus. These authors proposed that the chlorophenolics were formed from the chemical breakdown20of unstable HMW material. However, O’Connor and Voss (1992) using spiked‘3C-chloroguaiacolsuggested the probable mechanism was adsorptionldesorption of low molecular weight chlorophenolsfrom the HMW material. Martin (1993) showed that the HMW material released several chloroguaiacoland chiorovanillin isomers under abiotic conditions, but the mechanism (breakdown versus sorption)was not apparent. Regardless, the amount of chiorophenols released from the HMW fraction in thesestudies was small relative to the amount found in the LMW fraction.The potential of chiorophenol compounds to interact with the 11MW material may alter thebioavailability of the chiorophenols to microbial degradation and affect their toxicity to higherorganisms. Robinson and Novak (1994) found that 2,4,6-trichloro-(‘4C)-phenol (2,4,6-TCP) bound todissolved humic acid (HA) mineralized 5-15% less overall than 2,4,6-TCP free of humic acid by anacclimatized bacterial culture. The rate of degradation of humic acid bound 2,4,6-TCP was muchslower than that of the free 2,4,6-TCP and appeared to be desorption controlled. In addition, the effectof the HMW chlorolignin from bleach kraft mill effluent on the bioavailability of hydrophobic pollutantsto higher organisms has been demonstrated on Daphnia magna (Kukkonen, 1992; Kukkonen and Oikari,1992). The presence of chlorolignin was shown to significantly reduce the bioconcentration factors forseveral hydrophobic organic pollutants in daphnids.The HMW material may impact microbial growth by a number of mechanisms. Freeman and Lock(1992) observed that seasonal variations in the recalcitrant (>1 kDa) material in brown-water riversresulted in different microbial activities of river biofilms. Laboratory and field observations appearedto support the hypothesis that increasing the concentration of natural and synthetic 11MW materialinhibited microbial activity. The proposed mechanism for this effect was the impedance of accessiblesubstrates through the cell membrane. Likewise, the long-term exposure of microbial communities tohumic acids was found to reduce the ability of the microbial population to respond to monosubstituted21phenols (Shimp and Pfaender, 1985). The reason for this behaviour was not identified and a numberof hypotheses were suggested, including the irreversible binding of an enzyme responsible forchlorophenol transport or metabolism to the humic acids. Similar studies by Larsson and Lemkemeier(1989) found that microbial populations from a highly humic lake were more capable of mineralizing‘4C-ring-labelled 3,4-dichiorophenol and 2,4,5-trichlorophenol than cultures from a clear lake. Thisapparently contradictory fmding was believed to be a result of the higher cell concentrations oforganisms which had been acclimatized to utilizing refractory aromatic compounds in the humic lakecompared to the oligotrophic lake. However, pentachlorophenol (PCP) was better metabolized by theclear lake microorganisms, perhaps suggesting the highly hydrophobic PCP in the humic lake water wasbinding to the humic material, and therefore was not bioavailable to the microorganisms.Several researchers have performed elemental analysis on the HMW material and have reportedsignificant nitrogen contents (Mörck et a!, 1991; Dahlman et a!, 1994; O’Conner et at, 1994). Nitrogenis an important nutrient biologically, and its presence will impact both biological treatment operationand the receiving water well-being. The nitrogen content of the HMW material has been reported tovary from 0.2 to 2.1% (w/w) (Mörck et al, 1991; Dahiman et a!, 1994; O’Conner et a!, 1994). Thisnitrogen content is higher than the nitrogen content of wood (0.05 - 0.3%) (Merrill and Cowling, 1966).Over 70% of the nitrogen in wood is found in the form of amino acids (i.e., protein) (Laidlaw andSmith, 1965), and some of these proteins are structurally fixed to the cell wall lignin (Bao et a!, 1992;Whitmore, 1982). Therefore, the fact that the nitrogen is enriched in the 11MW fraction of the effluentcould indicate that it is still associated with the lignin or in the form of large protein molecules.Whether the nitrogen in the HMW material is available to microorganisms in a biological treatmentplant or in the receiving environment has not been reported.22The high carboxyl content of the humic and BKME 1-1MW material not only promotes thehydrophillicity of these large molecules, but also binds many trace minerals or heavy metals that maypromote or inhibit microbial growth (Shimp and Pfaender, 1985; Benes et al, 1976). Dahiman et at(1994) showed that under neutral or alkaline conditions the HMW material behaves as an anionicpolyelectrolyte capable of attracting metal ions, particularly di- and trivalent metals. Whether thechelating ability of the HMW material affects the performance of the biological treatment facility hasnot been established.While considerable research has identified how the 11MW material interacts with other components ofthe effluent and biological activity of organisms in the receiving environment, little has been done toestablish the role of the HMW material within a biological treatment system. The objective of theresearch was to gain some insight on the influence of the HMW material on the microorganisms in anactivated sludge treatment plant.1.6. ObjectivesA better understanding of the behaviour of the various components of the effluent in biologicaltreatment facilities and the receiving environment will help pulp mills develop enhanced methods andprocesses that reduce AOX discharges into the environment. The recent use of AOX as a regulatoryparameter makes the 11MW material an important fraction to study as it contributes a substantialproportion of the AOX. This highly coloured, recalcitrant material is responsible for a significantamount of the chemical oxygen demand and AOX load on a biological treatment system and may affectthe behaviour of the microbial population and other chemical constituents. With the elimination ofdioxin and furan production, chlorinated phenolics are considered the most environmentally significantfraction of the AOX in bleached kraft mill effluents. Understanding the impact of the 11MW fractionon chlorinated phenolic removal should lead to better biological treatment strategies.23The main objective of the thesis was to establish the effect that the 11MW material in a modem bleachkraft effluent has on the growth and metabolism of the microorganisms present in activated sludgeplants. The HMW material may inhibit microbial activity by interfering with cell membrane transport,chelating trace nutrients or concentrating toxic metals. Alternatively, the 11MW material may stimulategrowth by providing an enriched source of nutrients or making toxic compounds less bioavailable.Growth experiments in this thesis set out to determine whether the 11MW material was inhibitory orstimulatory to microorganisms and what the mechanism might be.A second objective was to establish whether the HMW fraction of the effluent altered the behaviour ofchlorophenols with respect to biological removal. Previous work showing the interaction between the11MW material and chiorophenols suggested that the HMW material could affect the biologicalavailability of the chlorophenols to the microorganisms. Consequently, the altered bioavailability couldimpact the microbial toxicity or biodegradability of the chlorophenols.The strategy employed to achieve these objectives was to fractionate effluent, obtained by a modembleached kraft mill using 100% chlorine dioxide substitution, into the LMW and HMW fractions. Thiswas followed by a series of batch growth experiments containing varying amounts of HIVIW materialand spiked chlorophenols. Batch growth experiments, using a microbial population from a laboratory-scale activated sludge reactor, were monitored for biomass production, substrate removal andchlorophenol removal. The model chlorinated phenolics used in this study, 6-chlorovanillin, 2,4-dichlorophenol, and 4,5-dichloroguaiacol, were chosen because they represent the predominantchlorinated phenols found in a modem bleach plant using 100% chlorine dioxide in the first bleachingstage (Solomen et a!, 1993; Spengel et at, 1994). The main properties of these compounds are listedin Table 2.24Table 2 - Some properties of the model chlorinated phenolic compounds used in this research.Compound M.W. LC50 Log K0 Log BCF(glmol) (mg/L)6-chiorovanillin 186.6 NR 1.76 0.592,4-dichiorophenol 163.0 2.6 2.80-3.23 1.41-1.754,5-dichloroguaiacol 193.04 2.3 3.23 1.75(from NCASI, 1992)MW: molecular weightLC50: average toxicity to rainbow troutLog K0: logarithmic octanol/water partitioning coefficientLog BCF: logarithmic bioconcentration factor in fishNR: not reportedWhile these mono- and dichiorophenols are commonly found in modem effluents, their properties arequite different from the higher chlorinated isomers which have received recent research attention. Thesecompounds are not considered to be hydrophobic and may not exhibit the same interactions that havebeen observed with higher chlorinated compounds and FIMW or humic material. In the work presentedin this thesis, I hope to reveal the role the HMW material has on microbial metabolism and thebehaviour of the common chlorophenols found in modem bleach plant effluent.252. MATERIALS AN]) EXPERIMENTAL METHODS2.1. Materials2.1.1. Bleached kraft mill effluentThe source of effluent used in this thesis was a bleached softwood kraft mill operated by WeyerhaeuserCanada in Kamloops, British Columbia. Two 1000 L and several 20 L shipments of untreatedcombined effluent were collected during the course of this thesis. The 20 L effluent samples collectedin the spring and summer of 1992 were used to develop laboratory procedures, while the bench-scaleactivated sludge reactor was maintained on the two 1000 L shipments of effluent which arrived inNovember 1992 and April 1993.All of the results presented in this thesis were obtained using the second 1000 L shipment of untreatedcombined effluent collected over the three day period (April 2-4, 1993). The average production duringthe collection period was 1218 ADMT/day using a furnish of 32% pine, 30% spruce, 18% fir, and 6%cedar. The mill was operating two bleach lines (DEEDED and DEEO(DE)D) with an average effluentdischarge of 142,900m3/day (117m3/ADMT). The effluent was collected after acid and caustic sewerswere combined at the inlet to the primary settling basin. The nutrients, ammonium nitrate (28-0-0)(NPK (%w/v)) and ammonium polyphosphate (10-34-0), were added at a rate of 1.5 L/min and 0.5L/min, respectively.After shipping, the effluent was completely mixed, dispensed into 20 L plastic containers and frozenat -20°C. Prior to use, the effluent was thawed slowly at 4°C over several days.2.1.2. Model Chlorinated Phenolic Compounds and StandardsChlorinated phenolic compounds chosen as model compounds for this research were 6-chlorovanillin,2,4-dichlorophenol, and 4,5-dichloroguaiacol (Figure 4). The 6-chiorovanillin (99+% purity) and 4,5-26CIOCH34,5,6-trichioro-trimethoxybenzene(internal std.)Figure 4 - Chemical structure of the chlorinated phenolic compounds used in this study.dichloroguaiacol (99+% purity) were purchased from Helix Biotech Corp. (Richmond, B.C). The 2,4-dichiorophenol was obtained from Aldrich Chemical Co. (Milwaukee, WI), 2,6-dibromophenol (>97%purity) purchased from Fluka AG Chem. Fabrik was used as a surrogate standard to monitor acetylationand extraction efficiency. The internal standard used was 4,5,6-trichlorotrimethoxybenzene (4,5,6-TCTMB; 99+%) from Helix Biotech Corp. (Richmond, B.C.). Standards were dissolved in methanol(Optima grade, Fisher Scientific, Vancouver, B.C.) except the 6-chiorovanillin which was dissolved inacetone (AnalaR grade, BDH) and the 4,5,6-TCTMB which was dissolved in hexane (Optima grade,Fisher Scientific, Vancouver, B.C.).All solvents used were HPLC grade or equivalent. Dilution water used for all reagents was deionizedwater from a NANOpure ultrapure water system (Barnstead, Dubuque, IA). All reagent chemicals wereanalytical grade from specialist chemical producers.CHOCI OCHOH6-chiorovanillinCI2,4-dichiorophenol 4,5-dichloroguaiacolBrQBr2,6-dibromophenol(surrogate std.)272.2. Analytical Methods2.2.1. Chlorinated phenolic analysisAcelylation and extraction: Chlorinated phenolic samples were acetylated and extracted using a methoddescribed by Eriksson et a! (1985). A 3-5 mL sample was spiked with 2,6-dibromophenol (surrogatestandard) in 50 IlL of methanol and mixed with an equal volume of 0.1 MK2C03 in a 15 ml graduatedcentrifuge tube fitted with a Teflon®-coated screwcap. Acetic anhydride (200 jiL) was added and themixture was shaken gently on a horizontal rocker platform for 30 minutes with periodic venting. Afteracetylation, 1.5 ml of n-hexane were added and the mixture was rocked for a minimum of two hourson a horizontal shaker table. The hexane phase was back-extracted with 5 ml of 0.05 NK2C03 toeliminate residual acetic anhydride which may damage the GC column. One hundred ng of 4,5,6-TCTMB (2 tgJmL in hexane) were added to the hexane extract of the samples as an internal standard.The acetic anhydride was glass distilled twice followed by hexane extraction and theK2C03buffer washexane extracted before use.Gas chromatography: The acetylated and extracted sample was injected (1 tL) on a Hewlett PackardHP5 890 Series II gas chromatograph equipped with an electron capture detector (63Ni) and a DB-5fused silica column (25mm x 30 m, 25i.m film thickness) from J&W Scientific, Inc. Prepurified heliumserved as a carrier gas (1 mL/min), and prepurified nitrogen was used as a make-up gas (30 mL/min).The injector and detector were maintained at 260°C and 350°C, respectively. The temperature programused was: 60°C for two minutes, 3°C/mm. to 180°C, 20°C/mm. to 280°C, and 280°C for ten minutes.Quantification and quality control: Chlorinated phenolic samples were prepared in batches of 15-20samples. Triplicate samples from growth experiments were processed in separate acetylation andextraction batches to remove any bias that may occur with a particular batch. A blank (deionized waterspiked with the surrogate standard) was included in each batch to account for any reagent or glassware28contaminants. In addition, deionized water spiked with known quantities of 6-CVa, 2,4-DCP and 4,5-DCG standards was acetylated and extracted in triplicate with each group of samples. The relativeresponse factor (RRF) of each compound was determined from the spiked deionized water samplesaccording to the equation:RRFa = x isM1s AawhereMa = mass of analyte in the standardM15 = mass of internal standardAa = peak area of analyteA15 = peak area of internal standard.The RRF for the three standard chlorophenoL mixtures were averaged and applied to the samples withinthe same batch using the following equation:[Analyte] = Aa. x RRFa x M15 X %RA18%R = Ass- X RRFss X MisA15 M55whereMa, M55, and M15 = mass of analyte, surrogate and internal standard, respectively.Aa, Ass, and A15 = peak area of analyte, surrogate standard and internal standard respectively%R = percent recovery of surrogate standard in sample.2.2.2. Chemical Oxygen Demand (COD)Chemical oxygen demand (COD) was an aggregate measure of the amount of chemically oxidizablematerial (APHA Standard Method 5220 D - Closed reflux, colorimetric method, 1992). After dilutingthe sample appropriately, 1.2 mL of digestion solution (10.216 g O.D.K2CrO7,167 mlH2SO4, and33.3 g HgSO4made up to 1000 mL with distilled water) and 2.8 mL of sulphuric acid solution (9.58 1g Ag2SO4 to 1000 mL 112S04)were added to 2.0 mL of sample. The mixture was vortexed in test29tubes with Teflon®-coated screwcaps before digesting for two hours at 145°C. Tubes were cooled toroom temperature, and the absorbance (600 nm) was measured on a spectrophotometer (Milton RoySpectronic 1001 plus) with a 10 mm cell path and compared to potassium hydrogen phthalate standardswith a theoretical COD of 0 mg02/L, 450 mg02/L and 900 mg02/L. All samples and standards wereanalyzed in triplicate.Biomass (solid COD): The COD method was adapted as a sensitive measure of biomass requiring arelatively small sample volume to substitute for volatile suspended solids. A sample with less than 2mg of solid COD was filtered through a Whatman AH-934 (1.5 urn retention) glass fibre filter that hadbeen combusted at 600°C for 15 minutes prior to use. Filter and sample were rinsed with 10 ml ofdistilled water to remove residual soluble COD. The filter and the retained solids were transferred toscrewcap test tubes with 2.0 mL of distilled water. Digestion of the sample was performed as describedabove. Prior to reading the absorbance at 600 nm, the samples were centrifuged (14,000 rpm; 3minutes) in Eppendorf tubes to remove suspended glass fibres.2.2.3. Biochemical Oxygen Demand - 5 day (BODBiochemical oxygen demand (BOD) measures the amount of biodegradable organic material (APHAMethod 5210, 1992). An appropriate sample volume (Vs) was placed into 300 mL glass bottles withground glass stoppers. Each bottle was completely filled with dilution water, consisting of tap waterthat had been aerated at room temperature for several hours and supplemented with standard phosphate,magnesium, calcium, and iron nutrient solutions. Sample bottles and 3 dilution water blanks wereseeded with 200 jiL of mixed liquor from the activated sludge laboratory reactor (see section 2.3.2).All sample dilutions were prepared in triplicate. The initial dissolved oxygen (DO) content of thesamples (D1) and seeded blanks (B1) were recorded using a YSI DO meter (model 57) equipped witha BOD probe (model 5730) with a built-in stirrer calibrated with air-saturated dilution water. The30bottles were securely sealed and incubated at 20°C for five days. The final DO content of the samples(D2) and seeded blanks (B2) was measured. The BOD was calculated using the formula:BOD(mg021L) = [(D2 - D1) - (82 - B1)] x 300 mLVsNote: Dilutions with a final DO > 6 mg O2IL or < 2 mg 02/L were not considered.2.2.4. Volatile Suspended Solids (VSS)Volatile suspended solids (VSS) measures the amount of filterable organic material in a sample (APHAmethod 2540 G, 1992). Whatman 934-AH (1.5 jim retention) glass fibre filters were placed in goochcrucibles and ashed at 600°C for 15 minutes, cooled in a desicator and weighed (I). A sample volume(V) containing a minimum of 20 mg of VSS was filtered through the filter with a vacuum flask. Thegooch crucible was dried overnight at 105°C, cooled in a desicator and weighed (OD) before ashingthe sample at 600°C for 15 minutes. The final weight (A) was determined and the followingcalculations performed:Volatile suspended solids, VSS (mg/L) = (OD - A)/VTotal suspended solids, TSS (mgIL) (OD - 1)/VNon-volatile suspended solids, NVSS (mg/L) = (A - 1)/V2.2.5. Adsorbable Organic Halide (AOX)The analysis of adsorbable organic halide of the effluent fractions was performed on a Mitsubishi TotalOrganic Halogen Analyzer Model TOX- 10 at both MacMillan Bloedel Research and CanFor Researchand Development. Chlorinated organic material absorbed to activated carbon and the halogen contentwas determined by combusting the carbon at 1000°C and titrating the combustion gases for HC1microcoulometrically. The instrument was set-up according to the manufacturers specifications and31calibrated according to the Canadian Standard Method (1989). The calibration procedure involvedseveral steps:(i) titration cell calibration: 10 pL of 1.856 g/L 2,4,6-trichiorophenol solution (10.tg Cl/lOpL) werecombusted and titrated,(ii) activated carbon blanks: a pair of packed columns, each containing 40 mg of activated carbon,were washed with nitrate solution (8.2 g of potassium nitrate acidified to pH 2 withconcentrated nitric acid, diluted to 1000 mL) and were combusted and titrated,(iii) adsorption efficiency: 10 iL of the trichiorophenol solution (10 jiL ClJ1 0 L) was diluted in50 mL of deionized water acidified with nitric acid and passed through the pair of activatedcarbon columns.Each calibration step was performed in duplicate.Effluent samples were acidified with nitric acid and diluted with deionized water to appropriate AOXconcentrations (between 20-200 jig/L). Diluted samples (50 mL) were passed (3.3 mL/min.) throughtwo 40 mg of granular activated carbon beds that were held in glass columns plugged with ceramicwool at both ends. After adsorption of the sample, the columns were washed with nitrate solution (8.2g potassium nitrate acidified to pH 2 with concentrated nitric acid and diluted to 1000 mL withdeionized water) to remove the inorganic halide. Columns were emptied into sample boats forpyrolysis. During combustion of the activated carbon, the chloride content of the combustion gases wasdetermined by coulometric titration. The sum of the halide content of the two columns minus theactivated carbon blank halide content was used to determine the AOX concentration based on theoriginal sample volume. The halide content of the second column should not have exceeded 10% ofthe total halide measurement to ensure column overload did not occur.322.2.6. Phenol-Sulphuric AssayThe phenol-sulphuric assay measures total carbohydrate (Wood and Bhat, 1988). A 50 1.tL sample wasdiluted with 450 jiL of 2.5% phenol and mixed with 1.25 mL of concentratedH2S04. After the mixturecooled, the absorbance at 490 nm was measured and compared to the dextran standards (0-150 pjg).2.2.7. Somogyi-Nelson Reducing Sugar AssayGlucose determinations were performed using the method described by Wood and Bhat (1988).Somogyi Reagent 1(180 gNa2SO4,15 g Rochelle salt (sodium potassium tartrate), 30 gNa2CO3,and20 g NaHCO3 dissolved in 1000 mL of boiled deionized water) and Somogyi Reagent 11(45 g sodiumsulphate, 8 g copper sulphate5H2Oin 250 mL of boiled deionized water) were mixed in a ratio of 4:1.One millilitre of the combined Somogyi reagent was added to 1.0 mL of sample or standard containingglucose (2.5-100 jig) and water. After the addition of 1 mL of Nelson reagent (50 g(1’14)6Mo7024H, 42 mL of concentrated H2S04, 6 g NaAsO2 in 1000 mL deionized water), thesolution was heated in a boiling water bath for 15 minutes, cooled and mixed on a Vortex mixer. TwomL of distilled water were added and mixed prior to reading the absorbance at 520 nm. The absorbancewas translated to glucose equivalents using measurements of glucose standards at 0, 10, 25, 50, 75 and100 jig/mL concentrations.2.2.8. Protein Determination of SludgeThe use of protein assays to quantify the amount of microorganism (sludge) growth was explored.Three different protein assays were evaluated for their ability of determine the biomass content ofsludges. Samples were hydrolysed with equal volumes of 1 M NaOH at 100°C for 10 minutes priorto protein determination.33Lowry (modified) Protein Assay: The Lowry protein assay was based on the complexing of copper withpeptide bonds and has been modified by Pederson (1977). Reagent A included 0.5 g CuSO45H2Oand1.0 g of trisodium citrate in 100 mL of water. Reagent B was a mixture of 20 g Na2CO3and 4 g ofNaOH dissolved in 1000 mL of water. A working solution (A+B) containing 1.0 mL of Reagent A and50 mL of Reagent B was prepared prior to each assay. A 1.0 mL volume of A+B reagent was addedto a 200 jL volume of sample containing 1-30 .tg of protein. The mixture was vortexed and incubatedat room temperature for 10 minutes. One hundred jiL of Folin & Ciocalteu Phenol reagent (iN) wasadded to the test tube and vortexed immediately. After incubating for one hour at room temperature,the absorbance of the solution at 750 nm was measured and translated to bovine serum albumin (BSA)equivalents (1-30 tg).Bicinchoninic Acid (BCA) Protein Assay: The BCA protein assay by Smith et al (1985) was based ona similar principle as the Lowry assay (Stoschek, 1990). This assay was modified for use with a 96-well microplate reader. The assay was calibrated using concentrations of bovine serum albumin (BSA)ranging from 0 to 500 mg/L. A 20 jiL sample (diluted with distilled water, if necessary) and standardswere transfeffed to microplate wells in triplicate. A working solution containing 500 iL of coppersulphate solution (0.4 g ofCuSO45H2Oin 10 mL water) and 25 mL of BCA solution (1 g of sodiumbicinchoninate, 2 gNa2CO3,0.16 gNa2C4HO,0.4 g NaOH, and 0.95 g NaHCO3 in 100 mLof water adjusted to pH 11.25 with 1OM NaOH), was mixed together. A 200 jiL volume of the workingsolution was combined with the samples and standards. The plate was covered with a clear plasticadhesive film and incubated at 60°C for 30 minutes before reading the absorbance at 562 nm with amicroplate reader (Molecular Devices, THERMOmax model). The absorbances were translated to BSAequivalents.34Bio-Rad (Bradford) Protein Assay: The Bio-Rad protein assay was a commercialized method developedby Bradford (1976) adapted for use with a microplate reader. The assay was calibrated with a bovineserum albumin (BSA) standards ranging from 0 to 25 pgImL. Hydrolysed samples were diluted tobelow 25 jig/mL with 0.1 M phosphate buffer. Dye Reagent Concentrate (Bio-Rad, Hercules, USA)was dispensed (50 iL) to the plate wells followed by 200 jtL of sample or standards. After incubatingat room temperature for five minutes to one hour, the plate was read at 595 nm using a microplatereader (Molecular Devices, THERMOmax model).2.2.9. Total Kjeldahl NitrogenTotal Kjeldahl nitrogen (TKN) samples were analyzed by the U.B.C. Environmental Engineeringlaboratory and Analytical Service Laboratories Ltd in Vancouver, B.C. according to Standard Method4SOONorg (APHA, 1992). TKN measures nitrogen in the trinegative state. Samples were digested withH2S04,K2S04,and HgSO4 catalyst to convert amino nitrogen, free ammonia and ammonium-nitrogento ammonium sulphate ((NH4)2S0. The two laboratories used different techniques to measure theammonium content of the digested samples. The Environmental Engineering laboratory method raisesthe pH with concentrated buffer to convert the ammonium cation to ammonia, which was heated withsalicylate and hypochlorite to produce a blue colour proportional to the ammonia concentration. TheAnalytical Services Laboratory simply used an ammonium ion selective electrode to measure thenitrogen content. Nitrogen in the form of azide, azine, azo, hydrazone, nitrate, nitrite, nitro, nitorso,oxime and semi-carbazone was not included in the TKN results.2.2.10. AmmoniaAmmonia content of the samples was determined by the U.B.C. Environmental Engineering laboratoryby using the colorimetric technique used in the TKN method on filtered samples without the digestionprocedure.352.2.11. Nitrate+Nitrite, (N03-2or NO)The NO content of the effluent samples was determined by the U.B.C. Environmental Engineeringlaboratory using QuikChem Method No. 10-107-04-1 -Z (Lachat instruments, 1990). Passing the effluentsample through a copperized cadmium column quantitatively reduced any nitrate to nitrite. The reducednitrite plus the original nitrite was determined by diazotizing with sulfanilamide and coupling with N-(1-naphthyl)ethylenediamine dihydrochioride. The resulting complex absorbed at 520 nm. The sampleswere calibrated with KNO3 and NaNO2 standards ranging from 0.10 mg N/L to 25 mg N/L.2.2.12. Total PhosphorusTotal phosphorus samples were analyzed by Analytical Service Laboratories Ltd in Vancouver, B.C.according to Standard Method 4500-P/Ascorbic Acid (APHA, 1992). Ammonium molybdate andpotassium antimony! tartrate reacted in an acid medium with orthophosphate to form phosphomolybdicacid that was reduced to an intensely coloured molybdenum blue by ascorbic acid. The absorbancesat 880 nm of the treated samples were determined and compared with blanks and standard phosphatesolution (219.5 mg anhydrous KH2PO4diluted to 1000 mL with deionized water).2.2.13. Inductively Coupled Plasma (ICP) Metal ScanICP metal determinations were performed by Analytical Service Laboratories Ltd in Vancouver, B.C.according to Standard Method 3120 B (APHA, 1992). Total metal content was determined followingsample digestion. The principle of atomic emission spectroscopy was based on the fact that energyaddition to atoms will cause outer electrons to jump to higher energy orbitals. When the electronsreturn to their lower state orbital, the discrete energy release resulted in spectra unique to each element.In practice, a sample aerosol was generated and carried into a plasma field consisting of a flowingstream of argon gas ionized by an applied radio frequency field typically oscillating at 27.1 MHz. Theconstituent atoms were subjected to temperatures of about 6000 to 8000 K exciting atomic emission36efficiently and producing ionic emission spectra. The light produced was quantitatively measured witha polychromator at all configured wavelengths. The ICP instrument was calibrated to identily andquantifr 32 elements listed in Table 5 (p. 55).2.3. Experimental Methods2.3.1. Effluent molecular weight fractionationThe fractionation and diafiltration of the effluent was performed using a Pellicon Filter system(Millipore, Bedford, MA). The system included a 16 L/min Procon (Vane) pump and acrylic filterholder outfitted with a pressure gauge on the inlet and outlet. The membrane filter used was a 0.46 m2polyethersulphone (PES) Filtron Omega 1 kDa NMWL Centrasette membrane.The membrane’s performance was tested against the manufacturer’s specifications. Membrane retentionproperties were tested with vitamin B12, bacitracin, and dextran solutions. Permeate flux wasdetermined by timing 500 mL volumes of permeate. Samples of the filtrate and the retentate werecollected to determine the amount of breakthrough at different solute concentrations.Vitamin B12. Vitamin B12 (MW 1350 Da) was monitored by absorbance at 550 nm. Eight litres ofvitamin 12 solution (0.25 g/L) were fractionated with the ultrafiltration membrane. The ultrafiltrationsystem was operated with retentate recycle at an inlet pressure of 210 kPa and an outlet pressure of 70kPa.Dextran. The dextran used for this experiment was from Leuconostoc spp. (MW 6000 Da; FlukaBiochemika 31388). Dextran concentration was determined using the phenol-sulphuric assay. A 4 Lvolume of 0.5 g/L dextran solution comprised the test solution. The ultrafiltration system was operatedunder the same conditions as the vitamin 12 retention test.37Bacitracin. The antibiotic, bacitracin (73,000 units/g), was recommended by the manufacturer for testingthe molecular weight cut-off of the 1 kDa membrane. Bacitracin (MW 1400 Da) concentration wasfollowed by monitoring the absorbance at 250 nm. The test solution was 5 L of 0.5 g bacitracinlL indistilled water. The ultrafiltration system was operated and sampled under the same conditions as thevitamin B12.Effluent Pretreatment: Sixty litres (60 L) of frozen, untreated effluent were thawed at 4-6°C for 60hours. To separate solids from the effluent prior to ultrafiltration, the effluent was centrifuged (10800g; 10 mill; 4°C) in 3 L batches using a Sorvall RC24 centrifuge equipped with a GS-3 rotor. Thesupernatant was vacuum filtered using 15 cm diameter Whatman 934-AH (1.5 p.m retention) glass fibrefilters and the pellet discarded. The centrifugation and filtration process was performed over two days,and the effluent was kept at 4°C throughout the process.HMWILMW Separation: The initial fractionation of the prefiltered effluent was completed in 8 hours.The ultrafiltration system was operated with a transmembrane pressure of 300 kPa and a crossflow of1.7 L/min (AP 55 kPa). The filtration rate was monitored for each litre of permeate. An initialvolume of 57 L of effluent was concentrated to a final volume of 4.0 L. The 53 L of filtrate was frozenat -20°C in 3.2 L volumes in amber glass solvent bottles.BMW Diafiltration: In order to remove any residual LMW material from the HMW retentate, theBMW concentrate was diafiltered three times by diluting with deionized water and reconcentrating theHMW retentate with the membrane. The volumes of the retentate and diafiltrate and the resultingLMW dilution factor for each step is outlined in Table 3.38Table 3 - Volumes employed in the diafiltration process.Diafiltration Original D.I. water Diafiltrate Dilutionstep volume (L) added (L) removed (L) factor1 4.00 17.35 16.90 4.82 4.45 18.00 17.00 4.13 5.45 18.00 21.70 8.5Final 2.75 --- --- 168Note: Dilution Factor, D = Original volume + D.I. water addedFinal volumeFinal dilution factor = D1 x D2 x D3Assuming no membrane interactions or binding with the HMW material, any residual LMW compoundsin the retentate should have been diluted to a level 168 times lower than their initial concentration.The final diafiltered HIv[W concentrate was made up to 3.8 L with membrane and container rinse waterto produce a 11MW fraction 15 times more concentrated than the original effluent. The diafilteredHMW concentrate was split into ten 380 mL samples and frozen at -20°C in polypropylene bottles.Membrane Cleaning: The membrane was cleaned by flushing a solution of 0.2% (wlv) Tergazyme(Alconox mc, NY, USA) for 3 0-60 minutes at 40-50°C and rinsed with a minimum of 15 L of distilledwater with a single retentate pass, followed by 12 L of distilled water permeate flushing with retentaterecycle. The membrane was cleaned following each retention test and the initial IIMW/LMWseparation and the following HIVIW diafiltration operation.392.3.2. Activated sludge reactor operationA laboratory-scale activated sludge reactor operated for a period of two years was the source ofacclimatized microorganisms for batch experiments in this study. The reactor consisted of a 5 Laeration basin constructed from a 25 cm long piece of a 20 cm diameter acrylic cylinder glued onto 1cm thick acrylic base (Figure 5). The aeration basin was surrounded with a 25 cm diameter acryliccylinder to provide a water jacket to maintain a constant temperature (about 22°C). The aeration basinwas attached to a 2 L inverted polyethylene bottle which served as a clarifier. A central overflow pipein the clarifier controlled the liquid level in both the clarifier and aeration basin. Sludge from theclarifier bottom was returned to the aeration basin with a recycle pump operating on a timer (15 secondson/lO minutes off). A 25 mL volume/lO minute cycle was adequate to maintain a constant sludgeblanket. A solids retention time (SRT) of 10 days was controlled by wasting 500 mL (10% of aerationFigure 5 - Schematic diagram of the laboratory scale activated sludge reactor.Sludge recycleEffluent40basin volume) of mixed liquor directly from the aeration basin each day with a pump operating on a10 minute cycle. Feed was pumped from a 4°C walk-in refrigerator at a rate of 2.5 L/day to maintaina 48 hour hydraulic retention time (HRT).2.3.3. Batch Growth TrialsA series of batch growth trials was performed to evaluate the effect of the HMW fraction and spikedchiorophenolic compounds on the biological activity of activated sludge microorganisms. Most growthtrials were performed with an initial volume of 1.0 L of substrate in a 2 L Erlenmeyer flask. Someexperiments, requiring less sampling, were scaled to 150 mL volumes in 300 mL Erlenmeyer flasks.Treatments were inoculated with 0.5% (v/v) mixed liquor from the laboratory-scale activated sludgereactor providing an initial VSS concentration of 1 - 8 mgfL. To observe any abiotic effects,uninoculated controls were spiked with 0.5% (wlv) sodium azide and 0.2% (w/v) bacitracin todiscourage microbial activity. The flasks were agitated on a reciprocal shaker table (40 rpm) in the darkat 21°C.Flasks were monitored for substrate consumption (soluble COD), biomass production (solid COD) andoxygen uptake rate over a period of seven days. At each sample time, three samples werevolumetrically pipetted to separate glass fibre filters to determine the solid COD concentration. Thecomposite filtrate was subsampled in triplicate for soluble COD content. Chiorophenol-spiked flaskshad 3-5 mL samples collected using glass Pasteur pipettes in triplicate and frozen at -10°C for futurechiorophenol analysis.413. RESULTS AND DISCUSSIONThe work described in this thesis was pursued to reveal the possible influence the HMW componentof BKME might have on the microbial population present in an activated sludge treatment plant. Inorder to achieve this objective, a laboratory-scale activated sludge reactor was developed to provide asource of inoculum. In addition, 60 L of effluent were fractionated into its LMW and HMWcomponents and methods for determining biomass and chiorophenol concentrations during the growthexperiments were developed. After this preliminary work, a series of batch growth trials wereperformed to evaluate the effect of the HMW material on biomass production, substrate removal andchiorophenol removal.3.1. Activated Sludge Reactor PerformanceThe microbial population from a laboratory-scale activated sludge reactor was used as the source ofinocula for batch growth experiments. To monitor the health of the microorganisms during the courseof the experiments, the aeration basin biomass concentration and the removal of COD and BOD weredetermined weekly. The COD removal efficiency ranged from 47-80% with an average value of 68%obtained during the period of the batch experiments. The BOD removal efficiency never dropped below93%, indicating the sludge effectively removed the readily degradable components of the effluent.During the experimental period, the feed rate to the reactor was decreased to extend the use of the samebatch of effluent for additional experiments. As a result, the biomass concentration decreased from2500 mg VSS/L to about 1000 mg/L. The change in biomass concentration did not have any effect onthe amount of biomass growth observed in flasks of unfractionated effluent which were run as controlsfor each experiment. A preliminary growth experiment demonstrated that the amount of inoculum (inthe range used here) had no effect on the rate of substrate removal or net growth. Apparently, theamount of inoculum used was large enough to eliminate any effect the inoculum concentration might42have on the lag period. These results indicate the laboratory activated sludge reactor was a satisfactorysource of acclimatized microorganisms which were probably representative of what would be found ina full scale treatment plant. The next objective was to fractionate effluent into its LMW and HMWcomponents to provide the substrate for the batch growth experiments.3.2. Effluent Fractionation and CharacterizationPrevious research on BKME has frequently described the molecular weight distribution of the effluenton the basis of retention by ultrafiltration membranes (Pfister and Sjöström, 1978, 1979a, 1979b;Dahiman et a!, 1993a, 1993b, 1994; Yin et a!, 1990; Mörck et a!, 1991; O’Connor et a!, 1994). Thegrowth experiments described in this thesis required 60 L of BKME to be fractionated into its low andhigh molecular weight components. While the filtrate could be assumed to contain only LMWcompounds, the concentrated retentate (raw HMW material) required diafiltration with distilled waterto remove most of the residual LMW compounds.This portion of the research involved characterizing the ultrafiltration membrane retention properties,monitoring the membrane performance during the fractionation and diafiltration process, and finallycharacterizing the resulting effluent fractions.3.2.1. Membrane Retention PropertiesMembranes selectively retain compounds based on a number of properties including molecular weight,charge, shape, etc. (Mulder, 1991). A membrane’s nominal molecular weight limit (NMWL) is basedon the size distribution of the membrane pores and is determined on the basis of the passage orretention of different molecular weight test compounds. Prior to using the 1000 dalton membrane forfractionating the BKME, the membrane was tested for its retention of vitamin B12, bacitracin, anddextran. Vitamin B12 and bacitracin were compounds used by the membrane manufacturer (Filtron43Technology Corp., Clinton, MA, USA) to establish the 1 kDa NMWL of the membrane (Omega 1 kDaCentrasette membrane) used in this work. Vitamin B12 has a molecular weight of 1350 Da and hasoften been used to test the retention properties of I kDa cellulose acetate membranes (Millipore,Bedford, MA; Amicon, Beverly, MA). According to the Filtron Technology Corp., the polyethersulfone(PES) membrane used in this work has variable (20-80%) retention of vitamin B12 due to changes invitamin 12 conformation with pH and pressure. Bacitracin was an antibiotic protein with a molecularweight of 1400 Da which was recommended by Filtron to test the retention of their 1 kDa PESmembrane used in this thesis. Dextran was used to monitor the retention of larger molecules (MW6000 Da) which may have linear configurations. The purpose of these experiments was to establishwhether the membrane was performing according to the manufacturer’s specifications during theeffluent fractionation process.The ultrafiltration system was operated with a retentate recycle configuration (Figure 6). The datapresented in Figure 7 shows the concentration of the test compounds in the filtrate and retentate atdifferent volume reductions (volume reduction = filtrate volume/initial volume). The right hand axisshows the percent breakthrough (% breakthrough = filtrate concentration/retentate concentration x 100)at the various volume reductions. Higher breakthrough occurred when the compound was poorlyretained indicating that the NMWL may be larger than the molecular weight of the test compound.Vitamin B12 (MW 1350 Da) was found to be very poorly retained with only about 8-12% retention(Figure 7a). According to manufacturer specifications, these results were not unusual for this membraneas vitamin 12 retention was reported to range between 20-80% depending on the ultrafiltrationoperating conditions, such as temperature, pH, and pressure, that were used.44FiltrateFeedBacitracin (MW 1400 Da) behaved quite differently from the vitamin B 12 Under the same operatingconditions, 93% of the bacitracin was retained by the membrane (Figure 7b). The large difference inretention of compounds with a molecular weight difference of only 50 Da suggested that the retentionof different compounds was not due to molecular weight alone. Factors such as molecular charge,polymerization or shape were likely affecting retention rather than the 50 Da size difference.Dextran (MW 6000 Da) exhibited about 85% retention (Figure 7c). The lower retention of thedextran compared to the much smaller bacitracin molecule could have been a result of the linear natureof the dextran molecule, resulting in the ability to pass through much smaller pores than smallerglobular compounds. Otherwise, since the molecular weight of dextran was estimated from adistribution of different sized dextran compounds, the material passing through the membrane may havebeen LMW dextran.RetentateUltrafiltrationmembraneFigure 6 - Ultrafiltration system operating with retentate recycle.45a) b)100 1400 1001200803200 800) )100060 60800-°— RetentateZ3 —-F kthrough. 100 40 8 ca 600 408E C(U (U400 CDz.50 [Retentatel % Breakthroug 20 20[Filtratel 2000 0 0 a0.0 02 0.4 0.6 0.8 1.0 0.0 02 0.4 0.6 0.8 1.0Volume Reduction Volume Reductionc) d)1500 100 100140080400801200-200% Braarough1000___60 .. . 60___ __0)0)800% BreakthroUgi 40 40 aCC600CDCDz.20 20200______________________________0 0 00.0 02 0.4 0.6 0.8 1.0 0.0 02 0.4 0.6 0.8 - 1.0Volume Reduction Volume ReductionFigure 7 - Retention and percent breakthrough at different volume reductions of (a) vitamin B12, (b)bacitracin before effluent fractionation, (c) dextran, and (d) bacitracin after effluent fractionation.46The results of these tests verified the membrane performed according to the manufacturer’sspecifications for both vitamin B12 and bacitracin. However, the variation in retention of the differentcompounds emphasized that ultrafiltration membranes do not select on molecular weight alone andmolecular weight distribution results based on membrane separations should not be viewed as absolute.For the purpose of this thesis, the material in the effluent which was retained by the membrane wasconsidered to be heavier than 1000 Da.When the fractionation and diafiltration of the effluent had been completed, the membrane was cleanedand stored in sodium azide. Several days later, the membrane was reinstalled and tested again forbacitracin retention to confirm the membrane’s integrity (Figure 7d). While most of the bacitracin wasretained, the retention was not as high as the initial retention test. The decrease in membraneperformance could have been a result of different operating conditions (i.e., temperature, pH, pressures,etc.), damage caused during the washing and handling of the membrane after the diafiltration process,or shrinking and/or swelling of the membrane during storage. However, the membrane was believedto have been fully functional for the entire effluent fractionation and diafiltration process, since the fluxthroughout these operations did not increase as shown in the next section.3.2.2. Ultrafiltration system performance during effluent fractionation and diafiltration.During the fractionation process and diafiltration operation, the ultrafiltration system was closelymonitored. In particular, the permeate flux was determined over 1 L intervals. Initially, flux wasmeasured at different crossflow velocities and transmembrane pressures (TMP) to find the operatingconditions that maximize the flux.The flux was strongly related to the transmembrane pressure as expected, and inversely related tocrossflow velocity at all transmembrane pressure tested (Figure 8). Therefore, transmembrane pressures472.52 -G- -o10.50 I I1.5 2 2.5 3 3.5 4Crossflow (L/min.)I-’-70kPa — 140 kPa-O 210 kPa-A-280 kPaIFigure 8 - Crossflow and transmembrane pressure effect on effluent permeate flux.were set near the highest pressure recommended by the manufacture (i.e., 400 kPa), while the crossflowvelocity was minimized to achieve maximum filtrate flux. The fractionation and diafiltration operationswere performed at a transmembrane pressure of 300 kPa and a crossflow velocity of 1.7 L/min.During the initial fractionation of the 60 L of BKME, the flux remained constant demonstrating minimalmembrane fouling during the process (Figure 9). This result indicated good membrane performanceeven at volume reductions over 90% where the retentate COD concentration had increased from 2100mg/L to greater than 8000 mgfL. The filtrate COD values increased slightly at the end of thefractionation process from an overall average concentration of 1650 mg/L to 1950 mg/L for the last 1L of filtrate.483.2.41rih’-i.1i.1-.51.8 -:-0.6 -01 2 3 4 5 6 7Time (hours)Figure 9 - Permeate flux during the initial HMW/LMW fractionation process.3.C4>s++18 “..‘+x12U-0.6001st diafiltration 2nddiafittratiOn 3rd diafiltration2 4 6 8Time (hours)Figure 10 - Permeate flux during the diafiltration operation.49Unlike the fractionation process, diafiltration did not exhibit a steady flux (Figure 10). An increase inthe flux was observed at the start of each diafiltration step. A decrease in the flux occurred during thefirst diafiltration step, but was absent in the second and third diafiltration steps. The initial flux increaseat the start of each step probably indicated a decrease in foulants on the membrane surface as dilutionwater was added. The flux decrease in the first diafiltration step likely indicates the formation of alayer of foulants.Overall, the 1 kDa PES ultrafiltration membrane behaved consistently, which should have resulted ina good separation of the LMW and HMW components. The consistent flux during the fractionationoperation indicated that good crossflow turbulence was maintained to prevent a ‘dynamic’ membraneor organic gel layer forming on the membrane which would have reduced the effective molecularweight cut-off. Once the reproducible operation of the membrane was assured, we wanted to determinehow the different components of the effluent were fractionated during the ultrafiltration operation.3.2.3. Molecular Weight Mass BalanceEffluent fractions were analyzed for COD, BOD, and AOX content (Table 4). Using the COD data,a complete mass balance could be performed (Figure 11). The resulting balance accounted for 98% ofthe total COD, indicating little loss or degradation of the effluent during the fractionation process. Theinitial centrifugation and coarse filtration of the effluent removed 10.5% of the original COD, whilediafiltration of the HIvIW removed about 6% of the original COD.Based on the soluble (filtered) COD of the whole effluent, the LMW material which passed throughthe membrane accounted for 70.7% of the COD. The diafiltered HMW fraction retained by the 1000dalton membrane constituted 20.6% of the soluble COD. The remaining soluble COD was lost duringthe three diafiltrate rinses.50Table 4 - Chemical characteristics of the effluent fractions.Sample Volume COD (mg02/L) BOD (mgO2/L) AOX (mg C1/L)Original volume(L) Actual Original Actual Original 1 2Unfiltered 57.0 2372 2372 494 494 5.6 4.2Prefiltered 57.0 2124 2124 462 462 5.3 3.3HMW 4.0 8230 577 - - - -Diafiltered 3.8 6230 415 225 16 4.5 2.5HMWLMW 53.0 1615 1500 205 191 0.8 0.4lstdiafiltrate 16.9 340 101 - - 0.8 -2nd diafiltrate 17.0 81 24 - - - -3rd diafiltrate 20.7 39 14 - - - -Actual - concentration at volume indicated.Original - concentration when diluted to the originalAOX (1) MacMillan Bloedel Research(2) CanFor Research and Development57.0 L volume.UnfilteredEffluentI 135.200 mg10.5% 189.5%FilteredEffluentI•4_ 121100mg70.7%(uwfiltrate85600mg27.2%(RawI IIMW.i 32900mg2.1%17.5% 4.2% 2.5%Diafiltered r 1st ‘ ( 2nd ‘ ( 3rdHMW diafiltrate diafiltrate diafiltrate24900 mg 5700 mg J I 1400mgJ 800mgJFigure 11 - COD mass balance of the entire fractionation and diafiltration process.51The BOD results did not yield as good a mass balance as COD (Table 4). Accurate BODmeasurements were more difficult to obtain. As BOD was more likely to degrade in storage, this mayhave accounted for some of the imbalance observed. The BOD appeared to account for about 25% ofthe COD detected in the whole effluent, with a much greater amount found in the LMW fraction.The AOX content of the effluent fractions was determined at two different laboratories and both resultsare reported in Table 4 at the original effluent concentrations. Both laboratories used similar AOXanalyzers (Mitsubishi Total Organic Halogen Analyzer Model TOX- 10) and, while the values wereslightly different, the proportions of AOX in the different fractions were similar from both analyses.The HMW fraction appears to contain between 75 and 85% of the soluble AOX. The AOXdeterminations of the unfiltered effluent tended to be more variable because of the difficulty in obtaininga representative proportion of the suspended material, but the results presented indicate 5 to 21% of thetotal AOX was removed by filtering the effluent. Based on mill production data, the AOX content of5 mgfL translates to a production rate of 0.6 kg AOXIADMT. This amount of AOX generation wassubstantially lower than the 5-10 kg AOX/ADMT typically produced at a bleach kraft mill a decadeago. Other modern softwood bleached kraft mills that have eliminated the use of elemental chlorinein the bleach plant (ECF bleaching) have reported AOX loads of 0.1 to 1.0 kg AOXIADMT (Axegârdet a!, 1993; Dahlman et a!, 1993 a), so the effluent used in this thesis appeared to be representative ofother ECF mill effluents.The proportion of HMW COD and AOX obtained from the effluent was about 20% and 80% of thefiltered effluent, respectively, which differed from results presented in other papers. Typically, thedegree of chlorination in the HMW component was less than the LMW fraction when expressed as C:Clratios (Lindsträm and Osterberg, 1984; Osterberg and Lindsträm, 1985). Therefore, since COD wasrelated to the amount of carbon, one would have expected the proportion of COD to be higher than the52proportion of AOX in the 1-1MW fraction of the effluent. The COD:AOX ratios (based on the mass ofCl and 02) of the effluent fraction were 500:1, 120:1, and 2500:1 for the filtered whole effluent and the1-1MW and LMW fractions, respectively. Assuming that all the COD was derived from themineralization of organic carbon (i.e., 1 mole 02 = 1 mole C), the elemental ratios of C:Cl should havebeen 451:1, 108:1 and 2251:1 for the filtered whole effluent and the HMW and LMW fractions,respectively. The degree of chlorination in the 1-1MW fraction was similar to that found in othereffluents from mills using 100% chlorine dioxide substitution (Dahlman et al, 1993 a, 1994; O’Connoret a!, 1994). However, the LMW fraction had an extremely low degree of chlorination. While noobvious reason for this low degree of chlorination in the LMW fraction is readily apparent, a coupleof explanations are presented for discussion. First, a paper by Sullivan and Douek (1993) demonstratedthat a significant portion of the first chlorination stage (C-stage) AOX was sensitive to pH adjustmentwhile the caustic extraction stage (E-stage) AOX was more resistant. The first chlorine or chlorinedioxide stage (C or D-stage) is the source of most of the LMW AOX (Pfister and Sjöström, 1978) andthe caustic extraction (E-stage) contains most of the BMW AOX (Pfister and Sjöström, 1978; Lindströmand Osterberg, 1984), so the pH sensitivity of the C-stage AOX suggests the LMW AOX is more pHsensitive than the HMW AOX. Similarly, Slimak et a! (1993) found that the medium molecular weightAOX (1 kDa> MW> 10 kDa) was more readily hydrolysable than the HMW AOX (MW> 10 kDa).If the LMW AOX was chemically unstable, a large portion of the LMW AOX may have beendechlorinated or hydrolysed when the acid and caustic sewers were combined upstream of where theeffluent was sampled. A second explanation for the high COD:AOX ratio in the LMW fraction couldhave been an elevated COD concentration resulting from a large non-bleach plant source of LMW CODwhich would not have contained any AOX.The high COD concentration in the LMW fraction appeared to be significantly higher than thatpreviously reported in the literature. Typically, the proportion of LMW COD in whole mill effluent53ranged between 40-60% (DahLman et al, 1993a, 1994) unlike the 70% found in this effluent. It waspossible the use of different membranes under different operating conditions would have altered theamount of material considered to be LMW. Many of the results presented in the literature wereperformed with cellulose acetate membranes which may be less porous than the polyethersulfonemembrane used in this study. Solomen et a! (1993) noted that the proportion of HMW AOX materialfound in bleach plant effluent decreased from 70-80% for chlorine bleaching (Lindsträm and Osterberg,1984; Osterberg and Lindström, 1985) to about 50% with high chlorine dioxide substitution (Osterbergand Lindström, ‘1985; Pfister, K. and E. Sjöström, 1978). These results indicated increased oxidationwith the increased use of chlorine dioxide (dO2) and this was clearly demonstrated by O’Connor eta! (1994). Similarly, on a COD basis, the small proportion of the HMW material in this effluent couldhave been a result of increased oxidation that occurred during the modern elemental chlorine free (ECF)bleach process (DEEO(DE)D and DEEDED).Further analysis of the molecular weight fractions for nitrogen, phosphorus and trace metals revealedsome interesting differences in the partitioning of various compounds (Table 5). The total Kjeldahlnitrogen (TKN) concentration was three times greater in the HMW fraction than that obtained in theLMW fraction. Ninety-two percent (92%) of the TKN in the LMW fraction was in the form ofammonia, which was probably added to the effluent by the pulp mill operators in the form ofammonium nitrate and ammonium polyphosphate. Based on data provided by the mill technical staff,nutrient addition contributed 2.6 mg/L of ammonia, 2.1 mg/L of nitrate, and 2.7 mg/L of phosphorus.Both the nitrogen estimates were higher than the amount measured in the whole mill effluent. Sincethe effluent was collected over a 3 day period and transporting and freezing the effluent took another2 days, some nitrogen losses may have occurred. Nitrogen could be converted to various forms or belost by volatilization. For example, Standard Methods (APHA, 1992) recommended that ammoniasamples be preserved below pH 2 (i.e., convert to NH4j to prevent volatilization. Similarly,54Table 5 - Nutrient and trace metal content of the different molecular weight fractions.Nutrient/Metal Unfractionated LMW HMW1(mg/L) (mg/L) (mg/L)Total Kjeldahl Nitrogen 10.5 2.5 6.93Ammonia, NH3 2.49 2.31 0.201Nitrates, NO 0.21 0.08 0.13Total Phosphorus 2.8 1.6 0.39Aluminum <0.20 <0.20 0.101Antimony <0.20 <0.20 <0.013Arsenic <0.20 <0.20 <0.013Barium 0.125 0.075 0.0126Beryllium <0.005 <0.005 <0.0003Bismuth <0.10 <0.10 <0.0067Boron <0.10 <0.10 <0.0067Cadmium <0.010 <0.010 <0.00067Calcium 67.1 38.2 4.91Chromium <0.015 <0.015 0.0037Cobalt <0.015 <0.015 <0.0010Copper 0.018 0.012 0.0091Iron 0.373 <0.030 0.163Lead <0.050 <0.050 <0.0033Lithium <0.015 <0.015 <0.0010Magnesium 4.58 2.31 0.174Manganese 0.5 14 0.306 0.0489Molybdenum <0.030 <0.030 <0.0020Nickel <0.020 <0.020 0.00 14Phosphorous 1.24 1.33 0.135Potassium 8.0 4.1 <0.13Selenium <0.20 <0.20 <0.013Silicon 1.55 0.816 0.154Silver <0.015 <0.015 <0.0010Sodium 275 152 0.740Strontium 0.150 0.085 0.0084Thallium <0.10 <0.10 <0.0067Tin <0.30 <0.30 <0.30Titanium 0.011 <0.010 0.0051Tungsten <0.10 <0.10 <0.0067Vanadium <0.030 <0.030 <0.0020Zinc 0.008 0.009 0.0113All concentrations are based on the original effluent volume. The HMW fraction was analyzed at 1 5xthe original concentration, resulting in lower detection limits.55denitrification could occur under anaerobic conditions. Since the original effluent was pH 6.5 and the5 day collection and transport period was sufficient to produce anoxic conditions, these mechanismscould account for some ammonia or NO losses. Another explanation could be the uptake of nitrogenby microorganism to form proteins.Since most of the nitrogen added as ammonia (a component of the TKN) was found in the LMWfraction and little ammonia was found in the HMW fraction (Table 5), the organic nitrogen (TKN) inthe HMW fraction must have been derived from the wood furnish or pulping process. A large portionof the structural protein in wood has been found bound to the lignin (Bao et a!, 1992; Whitmore, 1982).Since the BMW fraction was derived from lignin solubilized in the bleaching process, this structuralprotein could be the source of most of the original nitrogen in the effluent and would be expected tobe enriched in the HMW fraction. On a COD basis, the proportion of organic nitrogen (TKN minusNH3) in the HMW fraction was about 1.6% (w/w). Assuming the COD mass is similar to dry weight,this result was comparable to elemental analysis data reported by other researchers in the range of 0.2-2.1 (wlw) of the BMW material (Mörck et a!, 1991; Dahlman et at, 1994; O’Connor et a!, 1994).These nitrogen contents were higher than the proportion of nitrogen reported to be in whole wood (0.05-0.3%) (Merrill and Cowling, 1966) which would seem to support the lignin protein enrichmentsuggestion. However, the harsh conditions of the pulping and bleaching operations would be expectedto alter the form of the protein. Whether the HMW organic nitrogen was BMW protein or another formof nitrogen associated with the 11MW organic structure was not known.NO (nitrate-nitrite), which was not included in the TKN measurement, was found to a greater extentin the BMW fraction (65%). Nitrate or nitrite, which was usually ionized and highly soluble, wouldhave been expected to pass through the ultrafiltration membrane into the LMW fraction unless the NOinteracted with the BMW material. In contrast, the phosphorus content of the LMW fraction was four56times greater than the HMW portion. The total phosphorus measured in the whole mill effluent wasvery close to the amount added as a nutrient. This suggested that the original phosphorus content ofthe effluent was very low.Similarly, the partitioning of the trace metals analyzed varied considerably. For example, the transitionmetals, aluminum, chromium, iron, nickel and zinc had a greater affinity for the HMW fraction, whilethe alkaline earth metals, calcium, barium, magnesium, and strontium were found to a greater extentin the LMW effluent. In addition, manganese, silicon, sodium, and copper were found largely in theLMW fraction. These results indicated the ions with the greatest positive charge (the transition metals)had the greatest tendency to associate with the HMW material. In related work, Dahlman et al (1994)looked at the concentration of the common metal ions (K4, Na, Ca2,Mg2, and Fe3) in the HMWfraction and also found that monovalent metal ions did not associate with the HMW material since they“are only capable of forming comparatively weak electrostatic linkages to anionic positions in organicmaterial”. Divalent metal ions tended to be more closely associated with the HIvIW material and ironwas largely found in the HMW fraction. The high carboxylic acid content of the 11MW materialprovided many opportunities for the trivalent metal ions to form strong chelate complexes (Dahlmanet a!, 1994). Similarly, researchers have shown the same interactions between metal ions and naturallyoccurring humic material, which also contained a high proportion of carboxyl groups (Benes et a!,1976). The ability of HMW material to attract these trace elements may affect microbial growth in abiological treatment facility or in the receiving water since these compounds could be either limitingnutrients or toxicants.We next planned to carry out growth experiments to determine the possible influence the 11MW materialhad on microbial growth and chiorophenol removal. However, before this work could be performed,methods to determine biomass content and chiorophenol concentrations required some development.573.3. Biomass DeterminationDetermining the biomass concentration over the course of a batch growth trial was a fundamentalparameter which we used to study the growth of the microbial population. The need to take multiplesamples from individual flasks over the course of the experiment limited the volume available for eachsample collected. The traditional biomass measure, VSS, required a larger sample than was possiblefor these experiments, so a considerable amount of work was performed to develop a more sensitivemeasure of biomass.The flocculating nature of activated sludge made the use of optical density impractical. Similarly,standard microbial plate counting methods were considered ineffective because of the long timerequirements and the assay was limited to measuring only microorganisms capable of growing on solidsubstrate. As a more sensitive chemical measure of the activated sludge biomass was sought, threeprotein assays and filterable chemical oxygen demand were evaluated for correlation with VSS.The growth of microorganisms requires the production of proteins to perform the biochemical reactionswhich maintain the organisms. Because pulp and paper effluent should not contain notable amountsof protein, the amount of protein was expected to increase with the growth of microorganisms as theyconvert nitrogen to protein. The Lowry, Bio-Rad, and bicinchoninic acid (BCA) protein assays weretested on different concentrations of hydrolysed biomass and compared to the corresponding VSS value.The Bio-Rad protein assay was found to be the simplest method for determining protein and exhibitedless variability than the Lowry method and the BCA protein assay. However, filtered effluent with noVSS, was found to exhibit an high background level of about 180 mg/L of BSA equivalent protein.This high measurement indicated interference by some component of the effluent. Attempts to eliminatethe background interference by centrifuging the solid biomass and resuspending twice with distilled58water before hydrolysing with I M NaOH at 100°C for 10 minutes were unsuccessful. The ratio ofcentrifuged B io-Rad protein to VSS increased from 0.2 to 0.5 mg proteinlmg VS S when activatedsludge was diluted with filtered effluent (i.e., no VSS) from 100% to 20%. The change in responseindicates some interference from the filtered effluent even with the centrifuge clean-up step. The useof protein assays as a surrogate for VSS measurements was not pursued any further.The use of COD to measure the biomass was found to give the best results. Initial work using thedifference between the total COD and the COD of glass filtered samples found that the resulting ‘solid’COD value had high variability as a result of combining the variability of two COD readings. Thisphenomenon was greatest at low VSS concentrations where the difference between the total and filtrateCOD was small. Direct measurement of the COD of biomass samples retained on glass fibre filters bydigesting both the sample and filter together was successfully attempted provided that: (a) the glass fibrefilters were combusted at 600°C before use; (b) the filtered sample was rinsed with about 10 mL ofdistilled water to displace any retained soluble COD; and, (c) the digested filter and sample wascentrifuged to remove any suspended glass fibres that may have interfered with the absorbance readings.Furthermore, low concentrations of ‘solid’ COD could be measured by simply filtering larger volumesof effluent. This technique reduced the variability and increased the precision of measurements takenat low concentrations in contrast to calculating the difference between the total and soluble COD.The ‘solid’ COD method was compared to VSS values using both activated sludge diluted to differentconcentrations with filtered effluent and biomass taken at different stages of a growth experiment (i.e.,lag phase, exponential growth, and stationary phase). The good correlation between the VSS and glassfiltered COD indicated that the COD of the filtered solids was related to the dry weight of thecombustible solids and was not dependent on the stage of growth of the organisms. The ratio of CODand VSS was found to range between 1.1 and 1.4 mg COD/mg VSS. VSS measurements and591,000800-Jc,)g 60000C-)4000(I)20000 100 200 300 400 500 600VSS (mg/L)Figure 12 - Correlation between the solid COD and VSS measurements made at the start and end ofthe growth experiments (R2=O.84).corresponding solid COD values were taken at the start and end of the batch growth experiments werecompared. The VSS data from treatments using prefiltered effluent at the start of the growthexperiments were calculated from the inoculum VSS concentration divided by the amount of dilution(0.5% v/v). Most VSS measurements were based on single determinations, while all of the solid CODdata were the mean of triplicate samples. The correlation (least squared analysis) revealed an R2 valueof 0.84 and a slope of 1.1 mg solid COD/mg VSS (Figure 12). The ratio of the COD to the dry weightis similar to the theoretical COD of many organic compounds. The high R2 value verified that the solidCOD method was a good indicator of VSS concentration.The advantages of this method to determine biomass included the shorter time requirement (only 3-4hours instead of 24 hours required for VSS) and the smaller sample size - only 0.5-2 mg of biomass60instead of the minimum 200 mg recommended for VSS in the Standard Methods (APHA, 1992). Thesmaller amount of biomass required substantially decreased the filtration requirements, althoughsampling variability was found to increase with large diameter flocs and high solid concentration thatrequired small sample volumes. Sample volumes less that 2.0 mL were not recommended because ofthe difficulty in obtaining representative samples. This gave the method an upper detection limit of 900mg COD/L. The lower detection limit was only limited by the amount of sample available to get 0.5mg of solid COD. This method was found to be useful in monitoring the production of filterablebiomass in batch growth experiments which never exceeded 500 mg COD/L and could be applied tomonitoring the solids content of other wastewaters.The remaining method to be developed prior to the batch growth experiments was the analysis of themodel chiorophenols; 6-chiorovanillin (6-CVa), 2,4-dichiorophenol (2,4-DCP), and 4,5-dichloroguaiacol(4,5-DCG).3.4. Chlorophenol Analysis and CalculationsTo detennine the effect that chiorophenols (6-CVa, 2,4-DCP, and 4,5-DCG) had on the activated sludgemicroorganisms and the effect of the HMW material on chiorophenol removal, a method for monitoringchlorophenols was required. The method described by Eriksson et al (1985) was able to achieve goodchromatographic separation of the spiked chiorophenols in the effluent (Figure 13). The eliminationof contaminants by glass distilling all reagents and hexane rinsing all glassware prior to handling thesamples was largely successful. However, several unidentified peaks observed in Figure 13 weredetermined to be contaminants present in the chlorophenol standards spiked into the sample. Unspikedeffluent samples did not contain these peaks. Unfortunately, the quantification of the spikedchiorophenols in the effluent samples was complicated by poor recoveries of the ehiorophenol standards61spiked in deionized water (calibration standards). This section describes some of the experimentsperformed to resolve this problem.Deionized water spiked with chiorophenol standards, which was going to serve as the calibrationstandard, was found to have a lower recovery of the surrogate standard (2,6-DBP) than the equivalentspiked effluent samples. For example, if the amount of the surrogate standard found in the calibrationstandard was assumed to be 100%, the recoveries of the surrogate standard in the effluent samplesvaried between 150% to 500%. Analyzing unspiked LMW, HMW and whole mill effluent samplesshowed that no background peaks interfered with the compounds of interest. Therefore, the greaterrecovery of surrogate standard found in the effluent samples compared to the deionized water wasbelieved to due to incomplete acetylation and/or extraction in the spiked deionized water standards. Toverify this hypothesis, the total chiorophenol concentration in the aqueous phase of standards spiked indeionized water was monitored by absorbance at 280 nm. After preparing the standards using the samemethod as the effluent samples from the growth experiments, 90% of the chlorophenols remained inthe aqueous phase after the acetylation and extraction procedure, indicating that incomplete acetylationand/or extraction had occurred.The effluent matrix appeared to have some properties that enhanced the acetylation reaction or theextraction efficiency of the chlorophenols when compared to a deionized water matrix. To determinewhether ionic strength was an important factor, the recoveries of spiked chlorophenols in deionizedwater and water saturated with sodium chloride were compared. No difference was observed betweenthe two treatments indicating that the ionic strength of the effluent was not the cause of the improvedrecoveries. Since the acetylation reaction was pH dependent, the pH of the effluent fractions anddeionized water buffered with 0.1 MK2C03 buffer was measured. All the samples had a similar pHof 10, which did not reveal why the deionized water had depressed recoveries. Therefore, since an62• 0 0a.1.9 x 104,5-DCG IS1.7 X IO 2.4-DCp 6-CVaInIS: internal standard1.5 10:SS: surrogate standard0 1.3x101.1 xlL0.9 xlO0 10 20 30 40 50Retention Time1.9x104 b.ss Is41.7 x 100g0 1.5x10:1.3x101.1x104’4 V0.9x10 V V0Retention TimeVFigure 13 - GC-ECD chromatography of (a) chiorophenol standards spiked in deionized water(calibration standard) and (b) a typical spiked effluent sample from a growth experiment.63explanation for the poor recoveries in the calibration standards was not evident, a procedure wasdeveloped to try to correct the chlorophenol data produced during the growth experiments.As mentioned in section 2.2.1, samples collected in triplicate at each sample time were acetylated andextracted individually in different batches with their own set of standards, usually in triplicate. Thisprocedure minimized bias that could result from differences between standards in each batch, butincreased sample variability within a triplicate sample. All data were first corrected for the solventvolume by normalizing the data with the peak area of the internal standard. To account for anydifferences between batches of acetylations and extractions, the amount of each chlorophenol compoundwas determined using the relative response factors calculated from the calibration standards in thatbatch. Even though surrogate recovery values were often greater than 150%, applying the surrogaterecovery correction factor was found to reduce the variability between triplicates of samples andbetween sample points within a treatment. The resulting values were not accurate, so the samples werestandardized against time “0”. Since the concentration of the chlorophenols was known at the start ofeach experiment, the uncorrected data from ‘time 0’ for all of the spiked treatments were averaged (n= 3 replicates x no. of treatments or greater than 10% of all sample data) and divided by the actualamount of each chlorophenol present to make a correction factor. This factor was applied to the entireexperimental data set. The relative retention times (RRT), relative response factors (RRF), detectionlimits and correction factors applied to the chlorophenol compounds are listed in Table 6.While this method of quantifying the data was not ideal, the corrected data allowed some conclusionsto be made about the relative removal rates of the spiked chlorophenols in batch growth experiments.With the effluent fractionation completed and the biomass and chlorophenol methods developed,experiments to evaluate the interactions between the 11MW material, chiorophenols and activated sludgemicroorganisms could be performed.64Table 6 - Chiorophenol parameters used for their analysis.DetectionCompound RRT RRF Correction Factor LimitExp. 1 Exp. 2 Exp. 3 Q.tgfL)2,4-DCP 0.69 0.012-0.039 1.89 1.61 2.13 852,6-DBP (SS) 0.79 0.03-0.17 - - - -4,5-DCG 0.90 0.012-0.043 2.33 1.63 3.55 1506-CV 0.95 0.002-0.006 1.11 0.92 1.20 2503,4,5-TCTMB (IS) 1.00 1.00 - - - -RRT: relative response timeRRF: relative response factor3.5. Batch Growth ExperimentsA number of growth experiments were conducted to determine the influence of the high molecularweight fraction and chiorophenols on the growth and metabolism of the microorganisms present inactivated sludge. The first set of experiments observed the effect of the 11MW fraction on microbialgrowth and substrate consumption. Once the role of the HMW material was established, subsequentexperiments were carried out to determine the influence of the model mono- and dichlorinated phenoliccompounds on the microbial population and how the 11MW material affected chlorophenol removal.3.5.1. Effect of high molecular weight material on microbial growthInitial experiments using whole mill effluent and the LMW and 11MW fractions alone at the originaleffluent concentrations, resulted in good growth in the whole effluent, but poor growth on both theindividual LMW and 1-1MW fractions (Figure 14). Biomass production and substrate removal wasrelatively low for both individual molecular weight fractions compared to the whole effluent. The poorgrowth on the 11MW fraction was expected since other researchers had already demonstrated thismaterial was biologically recalcitrant (Eriksson et al, 1985; Eriksson and Kolar, 1985; Bourbonnais and655002400-10:0 25 50 75 180 185—rn—Whole8003 —S2 4009 rnv--- --V—._.JJ V0• • I • I • •/I •0 25 50 75 180 185Time (hours)Figure 14 - Individual molecular weight fractions original effluent concentrations. (a) biomassconcentration; and (b) substrate concentration. Error bars indicate standard sampling error; n=3.66Paice, 1987; Roy-Arcand and Archibald, 1993) and the low initial COD concentration may have beeninsufficient to support an effective microbial population. The poor growth obtained with the LMWfraction was not expected, but was reconfirmed and explained in subsequent experiments.Combining the HMW material with the LMW fraction revealed a synergistic effect between the twofractions (Figure 15). While little growth occurred when using the LMW fraction alone, the additionof the 11MW material at both the original (LMW+HMW) and three times (LMW+3xFflvIW) the originalconcentration more than doubled the amount of growth and substrate utilization. One possibleexplanation for the poor growth on the LMW fraction was that some COD losses occurring in the LMWeffluent during storage may have reduced the amount of biodegradable organic material available to themicroorganisms. Another explanation was that the lower initial COD concentration obtained when theLMW fraction was used alone compared to the flasks to which HMW material had been added mayhave resulted in a less complete microbial population unable to fully utilize the substrate. Alternatively,a component of the LMW fraction which was toxic to the microorganisms may have been able to bindwith the 11MW material to make it less bioavailable or there could have been a nutrient deficiency inthe LMW that was being compensated for by the HMW material. All of these hypotheses are examinedmore fully in the next section.3.5.2. Factors influencing growth on the LMWfraction3.5.2.1. COD storage losses in the LMWfractionWhile frozen at -20°C, the COD value of the LMW effluent dropped from 1600 mg/L afterfractionation, to 500-900 mg/L at the start of some initial growth experiments (including Figure 14 andFigure 15). The LMW effluent fraction had been frozen in amber glass solvent bottles to minimizephotolytic degradation and sorption or leaching that may have occurred with plastic containers.Unfortunately, many of the bottles cracked and experienced some losses from evaporation and67500400300o 25 50 75 1601652400 Whole2000 •—A--- LM W+HM W: :::::0 25 50 75 160165Time (hours)Figure 15 - Effect of varying amounts of HMW material added to the LMW fraction (a) biomassconcentration; and (b) substrate concentration. Error bars represent standard sampling error, n3.68sublimation. However, two uncracked bottles were recovered and the original 1600 mg/L of COD wasstill detected six months after the fractionation process. This effluent was used for the remainingexperiments. The following experiments show that the LMW fraction still supported relatively littlegrowth by itself (Figures 19, 22, and 24), indicating that the ‘missing’ COD was not the cause of thepoor growth shown in Figure 14 and Figure Effect of initial COD concentrations on growthThese earlier experiments suggested that the higher initial COD concentration obtained with the additionof the HMW material may have resulted in a more diverse population of microorganisms which couldbetter utilize the available substrate and produce more biomass. As a result, in subsequent experiments,different effluent treatments were diluted to the same initial COD concentration to try to minimize thiseffect. However, the addition of the FIIvIW material resulted in better growth and greater substrateutilization (Figure 16), even when the initial COD concentrations were equalized. This would appearto discredit the hypothesis that the higher initial COD in the treatments with HMW material supporteda more diverse microbial population better able to use the available substrate.Closer examination of the data in Figure 16 shows the combination of the LMW and the ilIvIW fractionresulted in better substrate removal (62% for LMW+3xHMW compared to 33% and 26% for LMW and3xHMW alone, respectively) and biomass production increased by over 300% (Table 7). The substrateremovals of the single effluent fractions were not additive since the combined LMW+3xHMW wasdiluted to the same initial concentration. These results seem to indicate a synergistic effect in whicheither the HMW detoxified a LMW toxicant or the HMW material provided a limiting nutrient. Thenext two sections describe experiments designed to determine which of these hypotheses was correct.69500——Whole—•--- LMW-..A---- LMW+HMW2 400 —V— LMW+3HMW—O—3xHMNC4300A3 2000U) 100 0--. 00 —. , . - Io 25 50 75 1601652400:::o 25 50 75 160165Time (hours)Figure 16 - Effect of the addition of 11MW material to LMW effluent diluted to the same initialconcentration. (a) biomass production; and (b) substrate utilization. Error bars represent standardsampling error; n=3.70Table 7 - Substrate utilization, biomass production and yield after 88 hours (from the data shownin Figure 19).Substrate BiomassTreatment Removal Production Yield(mg/L) (%) (mg/L) (%)Whole 983 50 410 41.7LMW 414 33 98 23.6LMW+HMW 644 50 297 46.1LMW+3xHMW 798 62 300 37.63xHMW 286 26 90 31.4Subsirate Removal, S: Soluble COD1 - Soluble CODf Biomass Production, B: Solid CODf- Solid COD% Removal: - CODfx 100% Yield: B x 100%COD1 S3.5.2.3. Microbial Toxicity of the LMWfractionThe presence of a microbial toxicant in the LMW fraction that became less toxic or bioavailable in thepresence of the HMW material was proposed as a possible explanation for the poor growth on theLMW fraction alone. Several papers have reported that HMW chiorolignin and humic material can alterthe bioavailability of toxicants to aquatic organisms (Kukkonen, 1992; Kukkonen and Oikari, 1992).To test for the presence of a LMW toxicant, varying amounts of LMW effluent were added to activatedsludge microorganisms growing on a synthetic media. The media contained glucose (0.1 % w/v), areadily degradable carbon source not normally present in the effluent, and the nutrients, 0.48% (wlv)NH4CI and 0.17% (wlv) Yeast Nitrogen Base without Amino Acids. The nutrients provided nitrogen,phosphorus and most essential trace elements required for microbial growth (Difco Laboratories, 1984).Therefore, if the LMW effluent contained an inhibitory substance, the rate of glucose removal wouldhave been expected to decrease with increasing concentrations of the LMW effluent. From the glucoseconcentration results (Figure 17), it was apparent that the glucose removal rate increased with theaddition of the LMW fraction, proving that the LMW fraction was not inhibitory, but in fact stimulatedglucose removal. Although, the organisms may have been better acclimatized to using the LMW71effluent as the carbon source, it is probable that the increased glucose removal may have been a resultof a trace nutrient present in the LMW effluent that was not provided by the YNB or NH4C1 nutrientbase.[Glucose] (mg/L)1,00080060040020000 20 40 60 80Time (hours)100Figure 17 - Glucose removal with varying amounts of LMW effluent.723.5.2.4. Nutrient Deficiency in the LMWfractionThe increased growth in the LMW effluent with the addition of the HMW material strongly suggestedthat a limiting nutrient in the LMW fraction was present in the 11MW material. The chemical analysisof the two molecular weight fractions revealed that the majority of the total Kjeldahl nitrogen (65%)was found in the HMW component. In contrast, phosphorus was largely found in the LMW fraction.This suggested that nitrogen may be the limiting nutrient in the LMW fraction. Growth experimentswere performed to establish whether the provision of nitrogen improved growth (Table 8). Biomassproduction and substrate removal improved by 350% and 650% respectively when nitrogen (0.48% w/vNH4CI) was supplemented to the LMW effluent. The addition of 0.17% (w/v) Yeast Nitrogen Base,a phosphate and trace element source, alone did not improve growth, but did increase growth slightlyin the presence of excess nitrogen. These results established that the LMW fraction is nitrogendeficient, but was readily degradable (nearly 70% of soluble COD) once nutrients are provided. Theaddition of nutrients changed the composition of microbial growth from a flocculating culture to a finedispersed suspension. The decrease in biomass yield observed with nutrient addition (Table 8) may bean result of microorganisms passing through the glass fibre filters which were used to detennine thesolid COD (see also Figure 21, p. 83).Since the low molecular weight fraction was nitrogen limited, the higher concentration of nitrogen inthe HIvIW material would explain the increased growth that resulted from the addition of the HMWmaterial. The large size of the 1-1MW material generally reduced the accessability of this material tomicroorganisms, but the nitrogen present in the HMW fraction appeared to be available to themicroorganisms. Further research is required to establish the chemical nature of the HIVIW nitrogen andenzymatic mechanism used by the microbial population to use the nitrogen. The high proportion ofnative nitrogen found in the 11MW fraction (section 3.2.3) should be of practical interest to mills whichare considering the use of ultrafiltration for the removal of HMW AOX and a water reduction measure73Table 8 - Effect of nutrient addition to LMW effluent on substrate utilization, biomass production andyield.Substrate BiomassNutrients Removal Production Yield(mgi) (%) (mg!L) (%)None 65 9.1 20.6 31.5YNB (w/o amino acids) 74 9.9 17.4 23.5NH4C1 423 55.8 72.4 17.1NH4C1 + YNB (w/o AA) 483 67.3 130 27.1Substrate Removal, S: Soluble CODE - Soluble CODE Biomass Production, B: Solid CODf Solid CODR% Removal: - CODx 100% Yield: B x 100%COD Sin the mill. The removal of a significant amount of the native nitrogen will impact on the health of thebiological treatment system and increase the demand for additional nitrogen supplements. Processengineers should also consider the effect that combusting the ultrafiltration concentrates (i.e., HMWmaterial) in the recovery boiler may have on NO air emissions.The fact that the majority of the nitrogen in the effluent is associated to the HMW material could affectthe distribution of nitrogen within ASB operations. Yan and Allen (1994) found up to 70% of theHMW AOX absorbed to living or dead biomass. If the BMW material absorbs to sludge whichconsequently settles to the bottom of the lagoon, the sludge blanket may serve as a nitrogen sinkresulting in nitrogen deficiencies in the suspended layer. Similarly, the apparent enrichment of nitrogen(compared to the LMW fraction) in the recalcitrant HMW material could be significant in the receivingenvironment. The HMW material may serve as a carrier of nitrogen through the treatment system intothe receiving environment where it will influence the natural microbial community and biota. Pellinen(1994) found that BMW AOX (>20 kDa) had great potential to sorb to sediments so the benthal layerwould probably serve as a nitrogen sink. Ecological studies examining the impact of pulp and papereffluents in the environment should consider the behaviour of the HMW nitrogen in the nutrient cycle.74Mills discharging into nitrogen limited environments such as marine receiving waters may benefit fromfurther research in this area.3.5.3. Effect and behaviour of model chlorinated phenolicsOnce the role of the FIMW material on microbial growth and metabolism was established, the effectsof HMW material and activated sludge microorganisms on the behaviour and removal of chlorinatedphenols were then studied. Three model chlorinated phenolic compounds, 6-CVa, 2,4-DCP and 4,5-DCG, were chosen to represent the predominant chiorophenols found in modem ECF bleaching effluent.These compounds were spiked at elevated concentrations into effluent to determine whether thechlorophenols themselves had any affect on the microorganisms or if the 11MW fraction altered thebiological removal of the chiorophenols. Effect of Chiorophenolic ConcentrationTo establish the impact that these three model chlorinated phenolic compounds might have on microbialactivity, growth experiments with the three compounds spiked at three concentrations into unfractionatedeffluent were performed. The 6-CVa, 4,5-DCG and 2,4-DCP were each spiked at 100 j.g/L, 1.0 mg/Land 10 mg/L concentrations into whole mill effluent. The three treatments were compared to anunspiked whole mill flask serving as a control. The chlorinated phenolic compounds had no significanteffect on the production of biomass or consumption of substrate (Figure 18). Surprisingly, the highestgrowth was observed in the treatment with the highest concentration of chlorophenols (3x 10 mg/L).However, a repeat of this experiment showed slightly lower growth in this treatment relative to the othertreatments indicating that these differences in biomass were not significant. These experimentsestablished that the mono- and dichiorinated phenols in this experiment did not affect microbial growthor COD removal at concentrations up to 10 mg/L each. This was notably in excess of chlorophenolconcentrations expected in BKME. Typical concentrations of total chiorophenols in untreated effluent75500A400300I: 1002500——ocP1:E 10002 *I500 • • • I/i0 25 50 75 155160Time (hours)Figure 18 - Effect of 100 jig/L, 1.0 mg/L, and 10 mg/L of 6-CV, 2,4-DCP, and 4,5-DCG on: (a)biomass production; and (b) substrate removal. Error bars represent standard sampling error;n3.76produced by mills using 100% chlorine dioxide substitution ranged between 0.2 to 2 g/ADMT (Axegârdet a!, 1993). Assuming that the average water consumption was about 90m3/ADMT (CPPA, 1992),the total chiorophenol concentration would have ranged from 2.2-22 j.ig/L or about 1400 to 14,000 timesless than levels shown in this experiment. This implies that the mono- and di-chlorinated phenols foundin modem ECF bleached kraft mills are unlikely to have any detrimental effects on biological treatmentsystems. Having established the lack of effect by the chiorophenols on the microorganisms, experimentswith and without HMW material were performed to determine the effect of this material onchlorophenol removal. Since the HMW material was found to have a significant impact on overallsubstrate removal and biomass growth, the HMW material was expected to have an similar effect onchlorophenol removal. Effect of the High Molecular Weight Fraction on Chiorophenol RemovalAfter the effects of the HMW fraction and the chiorophenols on the microorganisms had beendetermined, the interactions of the two treatments were then studied. Experiments were performed withthe three chiorophenols spiked into the LMW fraction and the LMW+3xHMW mixture at an initialconcentration of 1.0 mg/L. Unspiked treatments were monitored to confirm that the chlorophenol spikesdid not have any effects on the microbial population. In addition, uninoculated treatments spiked with1.0 mg/L of each chiorophenol were inhibited with 0.2% sodium azide and 0.5% bacitracin to observeany abiotic removal of the chiorophenols. The experiments were monitored for biomass content,substrate (soluble COD) and chiorophenol concentration.Experiments to study the effect of the BMW material on chiorophenol removal in the nitrogen limitedLMW fraction were performed twice; with and without dilution to the same initial COD. Since themajor influence of the HMW material was the provision of nitrogen, an experiment with supplementednutrients was performed to determine if the BMW material could have any additional effects on the77degradation of the chiorophenols and the growth of the microorganisms. This experiment wasperformed with the treatments diluted to the same initial COD.Effect of HMW material on chlorophenol removal in unsupplemented LMW effluent (no dilution)The treatments in the first experiment were not diluted to the same initial COD concentration andnutrients were not added. The growth data presented in Figure 19 showed similar results to thoseobtained in previous experiments. The provision of 11MW material resulted in much greater biomassconcentrations and substrate removal than the nitrogen-limited LMW fraction alone, while the additionof the chlorophenols did not effect microbial growth or substrate removal in either treatment.Monitoring the individual chlorinated phenol concentrations in the chlorophenol-spiked inoculated and‘abiotic’ treatments indicated some differences in the behaviour of each compound and gave someinsight about the removal mechanisms. The uninoculated treatments inhibited with sodium azide andbacitracin showed no decrease in the chlorophenol concentration (Figure 20). Similarly, no significantamount of chlorophenol was removed in the inoculated LMW fraction alone. However, thechlorophenols decreased in the inoculated LMW+3xHMW flask with the quickest rate of removal forthe 6-CVaa and slowest for the 2,4-DCP.The main mechanisms proposed for the removal of these types of compounds in an effluent treatmentsystem include abiotic (light, volatilization, etc.), sorption, mineralization and biotransformationfunctions (Leuenberger et al, 1985; Randle et a!, 1991). Abiotic removal of the chiorophenols did notoccur as shown by the uninoculated control flasks. Sorption to either the dissolved HMW material orthe biomass did not appear to be a factor. The chiorophenol did not bind to the 11MW material in anunextractable manner since there were no drops observed in chlorophenol concentration in theuninoculated LMW+3xBMW flask. Similarly, the steady chlorophenol concentration in the inoculated785001 ——Whole-I •--LMWLMW+CPI LMW+3HMW. LMW+3HMW+CF’ ,Y3000200_./•r10--0 25 50 75 160 165240120— 80J—.:: ::: : :: :: :U,0 • •o 25 50 75 160 165Time (hours)Figure 19 - LMW and LMW+3XHMW effluent with and without 1.0 mg/L of 6-CV, 2,4-DCP and 4,5-DCG; (a) biomass production; and (b) substrate removal. Error bars indicate standard sampling error;n=3.792400200-J1&0.4.0• • I • I • I0 25 50 75 1&)16524002000-J1OO10ZLZEEjUwE:::-400• —• • • //a.,o 25 50 75 11652400• LMW• LMW+3HMW200 C LM W (abio tic)0 LMW+3HMW(abiotic):::Time (hours)Figure 20 - 6-CV, 4,5-DCG, and 2,4-DCP concentrations in inoculated and abiotic flasks of spikedLMW and LMW+3xHMW effluent. Error bars indicate standard sampling error; n3.80LMW flask, which contained biomass, indicated that the chlorophenols were not sorbing to the biornass.These observations differ from research using more highly chlorinated phenolics (Kukkonen, 1992;Larsson and Lemkemeier, 1989; Robinson and Novak, 1994; Kukkonen and Oikari, 1992). The lowerhydrophobicity of these mono- and dichlorinated compounds could explain the difference from reportswith higher chlorinated compounds. The log octanol/water partition coefficients for the compounds hasbeen estimated to be 1.76 for 6-CVa, 3.23 for 4,5-DCG and 2.80-3.23 for 2,4-DCP (NCASI, 1992).These values indicate these compounds range from being ‘essentially non-hydrophobic’ to ‘slightlyhydrophobic’ (Solomen et a!, 1993). Further evidence that biosorption was not involved in the removalof the chlorophenols was the fact that the relative removal rates in the inoculated LMW+3xHIVIW flaskdid not correspond to the degree of hydrophobicity. For example, the 6-CVa is the least hydrophobic,but was removed the quickest. Consequently, the disappearance of the chlorophenols in the inoculatedLMW+3xHMW flask appeared to be due to either biological mineralization or biotransformation.The influence of the HIvIW material to improve the biodegradation of these chlorinated phenoliccompounds was most likely due to the increased growth that was supported by the provision of nitrogen.However, the presence of the 11MW material possibly favoured the growth of chlorophenol-degradingbacteria. Larsson and Lemkineier (1989) found that di- and trichlorophenols were more rapidlymineralized by bacterial populations obtained from humic lakes than those found in clear water lakes.This result was more probably a result of differences in the microbial population, with the bacteria inthe humic environment being better acclimatized to utilizing aromatic substances. Since the inoculumsource was the same for both the LMW and LMW+3xHMW treatments, differences in acclimatizationshould not have been a factor. However, the different conditions in the treatments would have favouredparticular organisms and would have produced distinct microbial populations.81Effect of HMW material on chiorophenol removal in unsupplemented LMW effluent (Diluted to sameinitial COD)The previous experiment without supplemented nutrients was repeated except the treatments werediluted to the same initial COD concentrations. Again, the biomass growth and substrate removal weresignificantly improved with the addition of the HMW material (Figure 21). The LMW fraction showedmore growth and substrate removal than the previous experiment and the dilution of the LMW+3xHMWtreatment to the same initial COD concentration resulted in a less notable difference in the amount ofgrowth compared to the LMW fraction. Both treatments with spiked chlorophenols resulted in slightlyless growth, but virtually identical substrate removal indicating no effect by the chlorophenols. Theseresults confirm previous conclusions regarding the provision of a limiting nutrient by the HMWmaterial, and the tolerance of the microbial population of 1.0 mgfL of the three chlorinated phenols.The chiorophenol results of this experiment (Figure 22) did not follow all of the trends observed in theprevious experiment. The variability between triplicate samples and between sampling times for a givenflask was higher as a result of some difficulties experienced with the chiorophenol analysis. None ofthe chlorophenols in the uninoculated treatments appeared to be removed over the course of theexperiment with the exception of the 6-CVa in the uninoculated LMW+3x1{MW flask on the lastsample point. This treatment experienced a 25% drop in the total COD between the last two samplepoints indicating that the flask did not remain ‘abiotic’ throughout the experiment. The lack ofchiorophenol removal in the uninoculated treatments confirmed that abiotic processes such asvolatilization, photodegradation, hydrolysis and sorption did not occur.In this experiment, the 6-CVa was removed in both the inoculated LMW and LMW+3xHMWtreatments. While the LMW+3xHMW flask removed the 6-CVa more quickly, the improved growthin the LMW fraction compared to the previous experiment appeared to have stimulated the removal of82400-—A— LMW- LMW+CP300- - LMW+3xHMW . •-. •-- LMW-’-3xHMW+CP200 V10:0 25 50 75 160 16515001:1:::Cl)300-V. • I • I • I .1/1.10 25 50 75 160 165Time (hours)FIgure 21 - LMW and LMW+3xHMW effluent with and without 1.0 mgIL of 6-CV, 2,4-DCP and 4,5-DCG diluted to same initial COD. Error bars indicate standard sampling error; n’3.8330002500-Ji4zi{Jj0 25 50 75 160 16516001200• - -800• I-J402’516001200C):i0.. eoo004C’ 4000- •0 25U LMN• LMN+3xHMND LMJV (unirctdated)0 LMN+3xHMN(uninocutated)• -,, I • I50 160 16575 160 165Time (hours)Figure 22 - 6-CV, 4,5-DCG, and 2,4-DCP concentrations in inoculated and ‘abiotic’- LMW andLMW+3xHMW flasks diluted to the same initial COD concentration. Error bars indicate standardsampling error; n=3.84the 6-CVa. This result probably indicated that the greater chiorophenol removing ability of theLMW+3xHMW combination used in the previous experiment was more a result of increased microbialactivity, rather than some selective pressure that the HMW material was exerting of the microbialpopulation. The 2,4-DCP, which had experienced little degradation in the previous experiment, did notexhibit any removal by any of the treatments used in this experiment. Similarly, the 4,5-DCG did notundergo any significant losses in any of the treatments. Overall, chiorophenol removal was less in thisexperiment. It was possible that differences in the composition of the microbial population in theinoculum may have affected the chiorophenol removing ability.Effect ofHMW material on chlorophenol removal on nutrient supplemented LMW effluentEarlier experiments established that the LMW fraction was nitrogen limited, and the increased growthobserved with the provision of the HMW material was a result of the higher nitrogen concentration inthe HMW fraction. To determine if the HMW material impacted on chiorophenol removal whennutrients were not limiting, an experiment was performed using supplemental nutrients. The additionof 0.17% (wlv) Yeast Nitrogen Base (without amino acids) and 0.48% (w/v) NH4C1 provided most ofthe essential nutrients, in excess of microbial requirements (Difco Laboratories, 1984). In addition tothe treatments performed in the previous two experiments, a flask with HMW material alonesupplemented with nutrients and spiked with chiorophenols was included. With this treatment, wehoped to establish whether the poor growth obtained previously on the HMW material alone was indeedan indication of biological recalcitrance, or a result of a nutrient limitation. Since chlorophenols hadalready been shown not to affect the biomass growth or substrate removal, the model chlorophenolswere spiked in this treatment to observe their behaviour without any LMW material present.The addition of nutrients reduced the difference in biomass growth between the LMW andLMW+3xHMW treatments (Figure 23). The resulting growth was more dispersed, without the floc85200M LMWLMW+CPLMW+3xHMW‘‘ LMW+3xHMW+CF150 • HMW+CPo 25 50 75 165 170I OOCo 25 50 75 165 170Time (hours)Figure 23 - LMW, LMW+3xHIvIW, and 11MW effluent with and without 1.0 mg/L of 6-CV, 2,4-DCPand 4,5-DCG diluted to the same initial COD with nutrients supplemented. Error bars indicate standardsampling error; n=3.86formation found under nutrient limiting conditions. The low overall biomass concentration and dropin measured biomass after 75 hours may have been an artifact of the more dispersed growth (i.e.,incomplete retention on the glass fibre filters). The addition of nutrients substantially increased theamount of soluble COD consumed in the LMW flask. The HMW material alone showed little growthor substrate loss even when nutrients were provided in excess, confirming the previous assertion thatthe RMW material was not readily degraded by microorganisms.The chiorophenol concentrations in the spiked LMW, LMW+3xHMW and HMW treatments are shownin Figure 24. The 6-CVa was found to be the most readily degraded chlorophenol, as demonstrated inprevious experiments. The provision of nutrients appears to have increased the rate of removal in allof the inoculated treatments. Both the LMW and LMW+3xHMW treatments removed the 6-CVa belowdetectable levels within 50 hours. The HMW material alone, while not supporting much growth, alsoshowed that 6-CVa was degraded to below detectable levels within 75 hours. The 6-CVa has a highlyreactive aldehyde group which could make it more susceptible to transformation. None of theuninoculated treatments showed any decrease in 6-CVa concentrations, indicating the removal in theinoculated treatments was a biological process.The 4,5-DCG did not show any removal in the first 90 hours, and only decreased significantly in theinoculated LMW treatment after 160 hours. The lack of 4,5-DCG removal in the inoculatedLMW+3xHMW flask contrasts the removal observed in the first experiment (Figure 20). The fact thatthe 6-CVa was removed more quickly in this experiment might suggest that the 4,5-DCG removalshould also have been quicker if the overall chiorophenol removal was related to the amount ofmetabolic activity. Since the 4,5-DCG was removed more slowly in this experiment, it suggested thatthe removal mechanisms of the two compounds were different. For example, different organisms orbiochemical pathways were probably responsible for the degradation of each compound, and the872000I &‘OI :r. • • I • i I •0 25 50 75 1165S LMN C LMN20 • LMN+3xHMN 0 LMN+3xHMNA HMJV HMN* Uninoo4ad ard ithibid1zoTime (hours)Figure 24 - 6-CV, 4,5-DCG, and 2,4-DCP concentrations in inoculated and ‘abiotic’ flasks of spikedeffluent fractions diluted to the same initial COD and supplemented with nutrients. Error bars indicatestandard sampling error; n=3.88provision of nutrients improved 6-CVa removal to a greater extent than 4,5-DCG removal. Thepresence of nutrients appeared to have selected for non-DCG degrading bacteria.The 2,4-DCP showed no decrease in any of the treatments. In fact, all of the inoculated treatmentsdisplayed slight increases in the 2,4-DCP concentrations. This rise in 2,4-DCP concentration wasprobably a result of analytical error, as 2,4-DCP was not a likely biotransformation product of 6-CVaor 4,5-DCG.In summary, chiorophenol removal did not occur in any of the abiotic controls indicating that biologicalprocesses (mineralization or biotransformation) were responsible for any disappearance. Biosorptionwas not responsible for any decrease in chiorophenol concentration since each experiment hadtreatments with biomass present that did not experience any drop in chiorophenol concentration. In allthree experiments, the 6-CVa was found to be the most readily degraded chlorophenol, while 2,4-DCPwas the least degraded. Chlorophenol removal was largely dependent on the overall amount ofmetabolic activity (i.e., biomass production and substrate removal). Under nutrient limiting conditions,the 11MW material stimulated microbial activity and chiorophenol removal. The lack of 2,4-DCPremoval in the time period and biomass concentrations studied in this thesis makes it impossible tomake any conclusions about the influence of the 11MW material on its biological removal. The 11MWmaterial did not appear to affect the bioavailability of the chlorophenols to the microorganism. Themono- and dichlorophenols used in this study had lower K0 values than the compounds reported tohave had altered bioavailability in the presence of humic material (Kukkonen, 1992; Kukkonen andOikari, 1992; Robinson and Novak, 1994) which may explain the lack of an effect. In conclusion, themain effect the HMW material had on chlorophenol removal appears to be the provision of nitrogenwhich stimulated overall microbial activity.894. CONCLUSIONSThe results demonstrated that the high molecular weight (HMW) material (>1000 Da) present inbleached kraft mill effluent had a stimulatory effect on the microbes present in the activated sludgesystem. The addition of HMW material to the low molecular weight (LMW) fraction improved biomassproduction, substrate removal and chlorophenol removal. The increased microbial activity observedseemed to be a result of the high organic nitrogen content found in the HMW material. Although theLMW effluent was readily degradable (70% COD removal), it was severely nitrogen limited.Alternatively, the HMW material was biologically recalcitrant, but contained nitrogen in a form that wasreadily available to microorganisms. The mono- and dichiorinated phenols, 6-CVa, 2,4-DCP and 4,5-DCG, were shown to be non-toxic to activated sludge at concentrations up to 10 mg/L. This is greatlyin excess of any concentrations observed in modern bleach plant effluent. The 6-CVa was the mostreadily removed compound while 2,4-DCP removal was insignificant over a 7 day period.Fractionating the effluent, derived from a modem kraft mill bleaching softwood pulp using an elementalchlorine free bleaching sequence, revealed the HMW fraction contained only 20% of the COD, but 75-85% of the AOX. The LMW material appeared to be chlorinated to a smaller degree than previouslyreported in the literature. The majority (65%) of the organic nitrogen (TKN) was found in the HMWfraction. The TKN in the LMW fraction accounted for the nitrogen added to the effluent as fertilizer,indicating the HMW organic nitrogen originated from the wood furnish. Whether the nitrogen was aHMW protein or another form of nitrogen bound to HMW lignin breakdown products has not yet beenestablished. We showed that the HIVIW material retained multivalent transition metals, in accordancewith previous reports that the HMW material could behave like an anionic polyelectrolyte (Dahimaneta!, 1994).90These findings suggest that the high nitrogen content of the HMW material has a major effect onnutrient transport in biological treatment facilities. The significant potential for biosorption of the 11MWmaterial to sludge suggests that the sludge blanket in aerated lagoons could serve as a nitrogen sink.Since the HIvIW material appeared to be quite recalcitrant, this material may serve as a carrier ofnitrogen into the receiving environment where it would influence the natural biota. Further researchinto the nature of the association between the HMW material and the nitrogen would be useful inpredicting what role the HMW material has on nitrogen transport and cycling. In addition, pulp millsconsidering the use of ultrafiltration of the E-stage liquor as a water conservation strategy may wish toconsider the effect of concentrating and combusting the HMW concentrate on both the biologicaltreatment plant operation (i.e., nutrient addition) and NO air emissions.The LMW effluent was non-toxic to the activated sludge microbial population. In particular, thepredominant chlorophenols in modem BKME did not inhibit microbial activity at concentrations severalorders ofmagnitude greater than the highest values that could be expected in typical untreated effluents.These findings indicate that chlorinated phenols liberated during the bleaching process should notdisrupt the ability of the biological treatment system. Therefore, it is unlikely that high chiorophenolconcentrations will result in system failures in a biological treatment facility.915. LITERATURE CITEDAllard, A.-S., M. Remberger and A.H. Neilson (1987). Bacterial 0-methylation of halogen-sustitutedphenols. Applied and Environmental Microbiology, 53(4):839-845.American Public Health Association (APHA) (1992). Standard Methods for the Examination of Waterand Wastewater. (18th ed.) Washington, DC.Amy, G.L. C.W. Bryant, B.C. Alleman, W.A. Barkley (1988). Biosorption of organic halide in a kraftmill generated lagoon. Journal Water Pollution Control Federation, 60(8): 1445-1453.Archibald, F.S. and L. Roy-Areand (1994). Photolysis of kraft bleachery effluent high molecular weightorganochiorine and color. 1994 International Environmental Conference, Portland OR, April 17-20,1994, TAPPI proceedings, pp 259-270.Axegárd, P., 0. Dahlmann, I. Haglind, B. Jacobson, R. MOrck, and L. Stromberg (1993). Pulpbleaching and the environment - situation 1993. Nordic Pulp and Paper Research Journal, 4(8):3 65-378.Bao, W., D.M. O’Malley and R.R. Sederoff (1992). Wood contains a cell-wall structural protein.Proceeding of the National Academy of Science US.A. 89:6604-6608.Benes, P., E.T. Gjesssing and E. Steinnes (1976). Interactions between humus and trace elements infresh water. Water Research, 10:710-716.Berry, R. (1992). Adsorbable organically bound halogen - an overview. In World Pulp and PaperTechnology, ed. F. Roberts. Sterling Publications International Ltd. London. pp.51-55.Bourbonnais, R. and M.G. Paice (1987). The fate of14C-labelled high molecular weight chlorinatedlignin and chromophoric material during microbial treatment of bleached kraft mill effluent. Journalof Wood Chemistry and Technology, 7(1): 51-64.Bradford, M.M. (1976). A Rapid and Sensitive Method for the Quantification ofMicrogram Quantitiesof Protein Utilizing the Principle of Protein-Dye Binding. Analytical Biochemistry, 72:248-254.Brezny, R., T.W. Joyce, B. Gonzalez and M. Slimak (1993). Biotransformations and toxicity changesof chlorolignins in soil. Environmental Science and Technology, 27(9): 1880-1884.Bryant, C.W., G.L. Amy, R. Neill and S. Ahmad (1988). Partitioning of Organic Chlorine between bulkwater and benthal interstitial water through a kraft mill aerated lagoon. Water Science and Technology,20:73-79.Canadian Standard Method for the Determination of Adsorbable Organic Halides in Waters andWastewaters: jointly prepared by Environment Canada and Paprican. November, 1989.CPPA (1992). Canadian Pulp and Paper Association - Reference Tables 1992, CPPA Press, Montreal,Quebec. -92Craig, G.R., P.L. Orr, J.L. Robertson and W.M. Vrooman (1990). Toxiciy and bioaccumulation ofAOX and EOX. Pulp and Paper Canada, 91(9):39-45.Dahlman, 0., I. Haglind, R. MOrck, F. de Sousa and L. Stromberg (1993a). Chemical composition ofeffluents from chlorine dioxide-bleaching of kraft pulps before and after secondary effluent treatment.EUCEPA Symposium: “Pulp and Paper Technologies for a Cleaner World”. Paris, France, April 27-29,Vol. 1, pp. 193-215.Dahlman, 0., R. Mörck, P. Ljungquist, A. Reimann, C. Johansson, H. Borén, and A. Grimvall (1993b).Chlorinated structural elements in high molecular weight organic matter from unpolluted waters andbleached-kraft mill effluents. Environmental Science and Technology, 27:1616-1620.Dahlman, 0., A. Reimann, L. Stromberg, and R. MOrck (1994). On the nature of high molecular weighteffluent materials from modem ECF- and TCF-bleaching. 1994 Intemational Bleaching Conference,Vancouver, B .C. June 13-16, proceedings (Canadian Pulp and Paper Association), 123-132.Diercks, R. and S. Banerjee (1993). Lipophilicity of a synthetic bleached kraft effluent. EnvironmentalScience and Technology, 27(12):2427-2429.Difco Laboratories (1984). Difco manual - Dehydrated culture media and reagents for microbiology(10th ed.), Difco Laboratories Inc., Detroit, MI. ISBN 9-9613169-9-3.Eriksson, K.-E. and M.-C. Kolar (1985). Microbial degradation of chlorolignins Environmental Scienceand Technology, 19(11):1086-1089.Eriksson, K.-E., M.-C. Kolar, P.O. Lyungquist, and K.P. Kringstad (1985). Studies on microbial andchemical conversions of chiorolignins. Environmental Science and Technology, 19:1219-1224.Fleming, B.!., T. Kovacs, C.E. Luthe, R.H. Voss, R.M. Berry, and P.E. Wrist (1990). A discussion ofthe use of the AOX parameter as a tool for environmental protection. Paper presented at CPPAWorkshop. January 1990.Freeman, C. and M.A. Lock (1992). Recalcitrant high-molecular-weight material, an inhibitor ofmicrobial metabolism in river biofilms. Applied and Environmental Microbiology, 58(6):2030-2033.Gergov, M., M. Priha, E. Talko, 0. Valttila, A. Kangas, and K. Kukkonen (1988). Chlorinated organiccompounds in effluent treatment at kraft mills. Tappi Journal, 71(12): 175-184.Graves, J.W., T.W. Joyce, and H. Jameel (1993). Effect of chlorine dioxide substitution, oxygendelignification, and biological treatment on bleach-plant effluent. Tappi Journal, 76(7): 153-158.Haggblom, M. and M. Salkinoja-Salonen (1990). Biodegradability of chlorinated organic compoundsin pulp bleaching effluents. Water Science and Technology, 24(3/4): 161-170.Häggblom, M. (1992). Microbial breakdown of halogenated aromatic pesticides and related compounds.FEMS Microbial Reviews 103:29-72.Higashi, R.M., G.N. Cherr, J.M. Shenker, J.M. Macdonald and D.G. Crosby (1992). A polar highmolecular mass constituent of bleached kraft mill effluent is toxic to marine organisms. EnvironmentalScience and Technology, 26(1 2):24 13-2420.93Jokela, J.K. and M. Salkinoja-Salonen (1992). Molecular weight distributions of organic halogens inbleached kraft pulp mill effluents. Environmental Science and Technology, 26(6): 1190-1197.Jokela, J.K., M. Lame, M. Ek and M. Salkinoja-Salonen (1993). Effect of biological treatment onhalogenated organics in bleached kraft pulp mill effluents studied by molecular weight distributionanalysis. Environmental Science and Technology, 27(3):547-557.Joshi, B.K and B.J Hillaby (1991). Effects of process improvements on pulp mill effluentcharacteristics. Environmental fate and effects of bleached pulp mill effluents, Stockholm, Sweden.November 19-21, 1991, proceedings, SEPA Report 4031, 101-109.Kringstad, K.P and K. Lindström (1984). Spent liquors for pulp bleaching. Environmental Science andTechnology, 1 8(8):236a-248a.Kukkonen, J. (1992). Effects of lignin and chiorolignin in pulp mill effluents on the binding andbioavailability of hydrophobic organic pollutants. Water Research, 26(11): 1523-1532. -Kukkonen, J. and A. Oikari (1992). Effects of kraft lignin and chlorolignin on the binding andbioavailability of benzo(a)pyrene to Daphnia magna straus. Bulletin ofEnvironmental Contaminationand Toxicology. 48:781-788.Lachat Instruments Inc. (1990). Methods manual for the Quikchem® Automated Ion Analyzer.Milwaulkee, WI.Laidlaw, R.A. and G.A. Smith (1965). The protein of the Scots Pine (Pinus sylvestris), Holzforschung,19(5):129-143.Larsson, P. and K. Lemkemeier (1989). Microbial mineralization of chlorinated phenols and biphenylsin sediment-water systems from humic and clear-water lakes. Water Research, 23(9): 1081-1085.Leach, J.M. and A.N. Thakore (1975). Isolation of toxic constituents of kraft pulp mill effluents.Environmental Protection Service, Environment Canada, CPAR Project No. 11-5. Ottawa, Canada.Leuenberger, C., W. Giger, R. Coney, J.W. Graydon and E. Molnar-Kubica (1985). Persistentchemicals in pulp mill effluents: Occurrence and behavior in an activated sludge treatment plant. WaterResearch, 1 9(7):885-894.Lindström, K. and M. Mohamed (1988). Selective removal of chlorinated organics from kraft mill totaleffluents in aerated lagoons. Nordic Pulp and Paper Research, 3(1 ):26-33.Lindström, K. and F. Osterberg (1984). Characterization of the high molecular mass chlorinated matterin spent bleach liquors (SBL) - Part 1. Alkaline SBL. Holzforschung, 38:201-212. VMacleod, J.M. (1993). Extended delignification: a status report. Appita, 46(6):445-451.Mackay, D. (1982). Correlation of bioconcentration factors. Environmental Science and Technology,1 6(5):274-278.Martin, V. (1993). The formation of chlorophenolics from high molecular weight chlorinated organics(>400 dalton) isolated at a bleached kraft mill. M.Sc. Thesis, University of Guelph, Canada.94MeCubbin Consultants Inc. (1 992a). Review of Literature on Characteristics of Effluent from Pulp andPaper Mills in Northern River in Northern River Basins of Alberta, B.C. and Northwest Territories.Prepared for Northern River Basin Study, Alberta Environment, Edmonton, Alberta.McCubbin Consultants Inc. (1 992b). Best available technology for the Ontario pulp and paper industry.Prepared for the Water Resources Branch, Ontario Ministry of Environment, Queen’s Printer forOntario.McKague, A.B., M. Jan and K.P. Kringstad (1989). An up-to-date list of compounds identified inbleachery effluent as of January 1989. Paper presented at 5th International Symposium of WoodPulping Chemistry. Myrtle Beach, SC.Merrill, W. and E.B. Cowling (1966). Role of nitrogen in wood deterioration: Amounts and distributionof nitrogen in tree stems. Canadian Journal ofBotany, 44:1555-1580.Mörck, R., A. Reimann and O.Dahlman (1991). Characterization of high molecular weight organicmaterials in modern softwood and hardwood bleached kraft mill effluents. Environmental fate andeffects of bleached pulp mill effluents, Stockholm, Sweden. November 19-21, 1991, proceedings, SEPAReport 4031, 155-163.Mulder, M. (1991). Basic Principles of Membrane Technology. Kiuwer Academic Publishers,Dordrecht, Netherlands.NCASI (1990). A review of the characteristics and signficance of adsorbable organic halide (AOX)in wastewaters from U.S. pulp and paper mills. Special Report No. 90-10, September 1990.NCASI (1992). A compilation of data on chemical and biological properties of 28 chlorinated phenoliccompounds. Special Report No. 92-12, December 1992.NCASI (1 994a). An assessment of the significance of discharge of chlorinated phenolic compoundsfrom bleached kraft pulp mills. Technical Bulletin No. 658, March 1994.NCASI (1994b). Aquatic life water quality screening values for chlorinated phenolic compounds.Technical Bulletin No. 659, March 1994.Neilson, A.H., A.-S. Allard, P.-A. Hynning, M. Remberger and L. Landner (1983). BacterialMethylation of chlorinated phenols and guaiacols: Formation of veratroles from guaiacols and high-molecular-weight chlorinated lignin. Applied and Environmental Microbiology, 45(3):774-783.Neilson, A.H., A.-S. Allard, C. Lindgren and M. Remberger (1987). Transformations ofchloroguaiacols, chloroveratroles, and chlorocatechols by stable consortia of anaerobic bacteria. Appliedand Environmental Microbiology, 53(1 0):25 11-2519.Neilson, A.H., C. Lindgren, P.-A. Hynning, M. Remberger (1988). Methylation of halogenated phenolsand thiophenols by cell extracts of gram-positive and gram-negative bacteria. Applied andEnvironmental Microbiology, 54(2): 524-530.Neilson, A.H., A.-S. Allard, P.-A. Hynning, M. Remberger and T. Viktor (1990). The environmentalfate of chlorophenolic constituents of bleachery effluents. TAPPI Journal, 73(3):239-247.95O’Connor, B.I., T.G. Kovacs, R.H. Voss, P.H. Martel, and B. Van Lierop (1994). A laboratoryassessment of the environmental quality of alternative pulp bleaching effluents. Pulp and Paper Canada,95(3 ):47-56.O’Connor, B.I. and R.H. Voss (1992). A new perspective (sorptionldesorption) on the question ofchlorolignin degradation to chlorinated phenolics. Environmental Science and Technology, 26(3):556-560.Osterberg, F. and K. Lindström (1985). Characterization of the high molecular mass chlorinated matterin spent bleach liquors (SBL) - Part II. Acidic SBL. Holzforschung, 38:149-158.Patrick, K.L., J. Young, K.H. Ferguson, and A. Harrison (1994). Closing the Loop - The effluent-freepulp and paper mill. Pulp & Paper, (Special editorial supplement) March 1994Pedersen, G.L. (1977). A Simplification of the Protein Assay Method of Lowry et a! which is generallyapplicable. Analytical Biochemistry 83:346-356.Pellinen, J. (1994). Sorption of high molecular weight chlorolignin on sediment. Chemosphere28(10): 1773-1789.Pellinen, J. and M. Salkinoja-Salonen (1985). Aqueous size exclusion chromatography of industriallignins. Journal of Chromatography 322:129-138.Pfister, K. and E. Sjöström (1978). Characterization of spent bleaching liquors - Part 1. Ultrafiltrationof effluents from conventional and oxygen bleaching sequences. Svensk Papperstidning, 81(6): 195-205.Pfister, K. and E. Sjöström (1979a). Characterization of spent bleaching liquors - Part 2. Compositionof material dissolved during chlorination (CEH sequence). Paperi ja Puu, 61(4a):220-230.Pfister, K. and E. Sjöström (1 979b). Characterization of spent bleaching liquors - Part 3. Compositionof material dissolved during alkali extraction (CEH sequence). Paperi ja Puu, 61 (4):3 67-370.Randle, W.G., E.R. Hall, and A.Y. McKinlay (1991). Optimization of biological treatment systems forenhanced removal of AOX and chiorophenolic compounds. 1991 International EnvironmentalConference, proceedings (TAPPI), 433-441.Reeve, D.W. and P.F. Earl (1989). Chlorinated organic matter in bleached chemical pulp production:Part 1 - Environmental impact and regulation of effluents. Pulp and Paper Canada, 90(4):65-69.Remberger, M., A.-S. Allard and A.H. Neilson (1986). Biotransfonnations of chloroguaiacols,chlorocatechols, and chloroveratroles in sediments. Applied and Environmental Microbiology, 51(3):552-558.Rempel, W., 0. Turk and J.E.G. Sikes (1992). Side-by-side activated sludge pilot plant investigationsfocusing on organochlorines. Journal ofPulp and Paper Science, 1 8(3):J77-J85.Robinson, K.G. and J.T. Novak (1994). Fate of 2,4,6-trichloro-(‘4C)-phenol bound to dissolved humicacid. Water Research, 28(2):445-452. - -96Rodden, G. (1994). Meeting new effluent standards in the focus for next year. Pulp & Paper Canada,95(6):13-27.Roy-Arcand, L. and F.S. Archibald (1993). Effect of time, daylight, and settling pond microorganismson the high molecular weight fraction of kraft bleachery effluents. Water Research, 27(5):873-88 1.Sâgfors, P.E. and B. Stark (1988). High molar mass lignin in bleached kraft pulp mill efflUents. WaterScience and Technology, 20:49-58.Sameshima, K., B. Simson, and C.W. Dence (1979). The fractionation and characterization of toxicmaterials in kraft spent bleaching liquors. Svensk papperstidning, 82(6): 162-170.Saunamäki, R., K. Jokinen, R. Järvinen and M. Savolainen (1991). Factors affecting the removal anddischarge of organic chlorine compounds at activated sludge treatment plants. Water Science andTechnology 24(3/4):295-307.Shimp, R. and F.K. Pfaender (1985). Influence of naturally occurring humic acids on biodegradationof monosubstituted phenols by aquatic bacteria. Applied and Environmental Microbiology, 49(2):402-407.Slimak, M., R. Brezn9, and T.W. Joyce (1993). Organic chlorine structures in chlorolignin.Holzforschung, 47(5)369-374.Smith, P.K., R.I. Krohn, G.T. Hermanson, A.K. Mallia, F.H. Gartner, M.D. Provenzano, E.K. Fujimoto,N.M. Goeke, B.J. Olson, and D.C. Klenk (1985). Measurement of Protein Using Bicinchoninic Acid.Analytical Chemistry, 150:76-85.Solomen, K., H. Bergman, R. Hugget, D. Mackay, and B. McKague (1993). A review and assessmentof the ecological risks associated with the use of chlorine dioxide for the bleaching of pulp, preparedfor the Alliance for Environmental Technology by a Science Review Panel.Sonnenberg, L.B., S.J. Watts, R.C. Eanes, and M.A. Contos (1994). Photolysis of chlorinated organiccompounds in bleach plant effluents exposed to sunlight. 1994 International Environmental Conference,Portland OR, April 17-20, 1994, TAPPI proceedings, pp. 359-369.Spengel, D.B., B. Bicknell, D.F. Anderson, M. Smith, D.G. Bodien (1994). A comparison ofchlorinated phenolic compound concentrations and loadings in bleach plant and treatment systemsamples at eight mills. 1994 International Pulp Bleaching Conference, Vancouver, B.C., June 13-16,1994, CPPA Press, pp.133-144.Springer, A.M. (1993). Industrial Environmental Control : Pulp and Paper Industry. 2nd ed., TAPPIPress, Atlanta, GA.Struthridge, T.R., D.N. Campin, A.G. Langdon, K.L. Mackie, P.N. McFarlane and A.L. Wilkins (1991).Treatability of bleached krafl pulp and paper mill wastewaters in a New Zealand aerated lagoontreatment system. Water Science and Technology, 24(3/4):3 09-317.Stoscheck, C.M. (1990). Quantification of Protein. In Methods in Enzymology, vol. 182, Guide toProtein Purification, ed. M.P. Deutscher, San Diego , Academic Press.97Sullivan, J. and M. Douek (1993). Effect of chemical treatments of bleachery effluents on AOXreduction. Pulp & Paper Canada, 94(12):165-171.Suntio, L.R., W.Y. Shiu, and D. Mackay (1988). A review of the nature and properties of chemicalspresent in pulp mill effluents. Chemosphere, 17(7):1249-1290.Taylor, B.R., K.L. Yeager, S.G. Abernethy and G.F. Westlake (1988). Scientific criteria document fordevelopment of provincial water iuality objectives and guidelines: resin acids. 50 pp., Ontario Ministiyof Environment, Water Resources Branch, Toronto, Ontario. ISBN 0-7729-4347-8, Queen’s Printer forOntario.Tomar, P. and D.G. Allen (1991). Removal of organochiorines from kraft pulp mill effluent by anaerated lagoon. Water Pollution Research of Canada, 6:101-117.Voss, R.H., J.T. Wearing, and A. Wong (1981). A novel gas chromatographic method for the analysisof chlorinated phenolics in pulp mill effluents. Chapt. in Advances in the Identification and Analysisof Organic Pollutants in Water. ed. L.H. Keith. Ann Arbor Science Publishers, Ann Arbor, MI, Vol.II, p. 1059.Whitmore, F.W. (1982). Lignin-protein complex in cell walls of Pinus elliottii: Amino acidconstituents. Phytochemistry, 21 (2):3 15-318.Wilson, D.G. and M.F. Holloran (1992). Decrease of AOX with various external effluent treatments.Pulp and Paper Canada, 93(12):136-142.Wood, T.M. and K.M. Bhat (1988). Methods for Measuring Cellulase Activities. In Methods inEnzymology, vol. 160, Biomass - Part A: Cellulose and Hemicellulose, ed. W.A. Wood and S.T.Kellogg, San Diego, Academic Press.Yan, G. and D.G. Allan (1994). Biosorption of high molecular weight organochlorines in pulp milleffluent. Water Research 28(9): 1933-1941.Yin, C.-F., T.W. Joyce, and H.-M. Chang (1990). Reduction of pollution load from bleach plant byapplying oxygen bleaching, chlorine dioxide substitution, and by ultrafiltration and sequential biologicaltreatment. 24th EUCEPA Conference, Stockholm, Sweden, May 8-11, pp. 152-173.Yu, P. and T. Welander (1993). A comparison of biological and thermo-alkaline treatment for theremoval of chlorinated organic matter from kraft bleach plant effluent. Nordic Pulp and PaperResearch, 8(4):360-364.98


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