Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The biotransformation of didecyldimethylammonium chloride by the hyphomycetes gliocladium roseum and… Dubois, Jason 1999

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

Item Metadata

Download

Media
831-ubc_1999-0186.pdf [ 17.39MB ]
Metadata
JSON: 831-1.0089094.json
JSON-LD: 831-1.0089094-ld.json
RDF/XML (Pretty): 831-1.0089094-rdf.xml
RDF/JSON: 831-1.0089094-rdf.json
Turtle: 831-1.0089094-turtle.txt
N-Triples: 831-1.0089094-rdf-ntriples.txt
Original Record: 831-1.0089094-source.json
Full Text
831-1.0089094-fulltext.txt
Citation
831-1.0089094.ris

Full Text

THE B I O T R A N S F O R M A T I O N OF D I D E C Y L D I M E T H Y L A M M O N I U M C H L O R I D E BY THE HYPHOMYCETES  GLIOCLADIUM ROSEUM A N D VER TICILLIUM B VLBILL OSUM by Jason Dubois B.Sc.(1993) University o f British Columbia, Vancouver  A THESIS S U B M I T T E D F O R P A R T I A L F U L F I L L M E N T OF T H E R E Q U I R E M E N T S F O R T H E D E G R E E OF M A S T E R OF SCIENCE in THE F A C U L T Y OF G R A D U A T E STUDIES (Faculty of Forestry, Department of W o o d Science)  We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH C O L U M B I A A p r i l , 1999 ® Jason Dubois, 1999  In presenting this thesis in partial fulfilment o f the requirements for an advanced degree at the University o f British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying o f this thesis for scholarly purposes may be granted by the head o f my department or by his or her representatives. It is understood that copying or publication o f this thesis for financial gain shall not be allowed without my written permission.  Department o f  wloob ^Cie-NlCg  The University o f British Columbia Vancouver, Canada  Date Aftuu 2/5, \9&b  ABSTRACT Two  mould fungi,  Gliocladium roseum and  Verticillium bulbillosum, that  had  demonstrated an inherent ability to degrade didecyldimethylammonium chloride ( D D A C ) , were used in the biodegradation experiments. D D A C was a) added to liquid media or b) used to treat wood pieces that were then subjected to fungal biodegradation. A metabolite from the degradation o f D D A C in the treated wood was discovered from the H P L C chromatogram, where the peak due to D D A C decreased and was replaced by a new peak at a longer retention time. It was isolated by preparative H P L C and further purified.  Proton  nuclear magnetic resonance ( N M R ) and Fourier transformed infrared spectroscopy (FTIR) were used to identify the metabolite.  From the spectra the compound was determined to be a  quaternary ammonium compound ( Q A C ) that differed only slightly from D D A C .  The  metabolite had two methyl groups and two decyl groups attached to the central nitrogen, with a hydroxyl group attached to one o f the decyl chains. The hydroxyl group was not at the terminal carbon o f the chain, suggesting that it was not the initiation o f P-oxidation o f the alkyl chain. The breakdown  structure o f this metabolite suggested that it was not an intermediate in the and  biotransformation.  possible  mineralization o f  DDAC,  but  rather  it  appeared  to  be  a  It was hypothesised that this biotransformation reduced the toxicity o f the  D D A C to fungi. This is supported in the literature, which cites many cases o f moulds employing hydroxylating enzymes to detoxify compounds. The two mould fungi and the basidiomycetes, Postia placenta and Trametes versicolor were used to screen various Q A C s with oxidized alkyl chains to determine their toxicity. It was found that a Q A C that was heavily oxidized along its long alkyl chains ( Q A C - O X I D ) was significantly less toxic to both the mould fungi and the basidiomycetes than D D A C .  ii  TABLE OF CONTENTS ABSTRACT T A B L E OF CONTENTS LIST OF T A B L E S LIST OF FIGURES LIST OF A B R E V I A T I O N S ACKNOWLEDGEMENTS C H A P T E R  ii iii v vi ix xi  1  B A C K G R O U N D  1  1.1  S T R U C T U R E , PROPERTIES, A N D A P P L I C A T I O N S  1.2  ENVIRONMENTAL IMPACTS  4  1.3  QACs A S W O O D  6  1.4  B E N E F I T S OF U N D E R S T A N D I N G Q A C B I O D E G R A D A T I O N  1.5  OBJECTIVES  C H A P T E R 2.1  2.3  2  L I T E R A T U R E R E V I E W  Morphology Metabolism  72 14 15  White-Rot Fungi Hyphomycetes  16 77  B I O D E G R A D A T I O N OF Q U A T E R N A R Y A M M O N I U M C O M P O U N D S  2.3.1 Factors affecting Biodegradation 2.3.1.1 Microbial Considerations 2.3.1.2 Bioavailability 2.3.1.3 Recalcitrance 2.3.2 Techniques for Studying Biodegradation 2.3.2.1 Measuring Q A C loss 2.3.2.2 Measuring Removal, Biotransformation, and Biodegradation 2.3.2.3 Determining Degradative Mechanism 2.3.3 Evidence of Biodegradation 2.3.4 Potential Pathways 2.3.5 Biodegradation Mechanism 2.3.5.1 B-oxidation 2.3.5.2 C - N de-methylation 2.3.5.3 C - N de-alkylation 2.3.5.4 C - N de-arylation 2.4  A N A L Y T I C A L TECHNIQUES U S E D  2.4.1 2.4.2 2.4.3 C H A P T E R  HPLC NMR FT1R  18  19 19 20 22 24 24 25 27 28 31 33 33 34 35 38 39  39 41 44  '.  3  12 12  F U N G A L BIODEGRADATION  2.2.7 2.2.2  8 10  C O M P A R I S O N OF F U N G I A N D B A C T E R I A  2.7.7 2.7.2 2.2  PRESERVATIVES  1  E V A L U A T I O N O F  B I O D E G R A D A T I O N  46  3.1  INTRODUCTION  46  3.2  METHODOLOGY  46  3.2.1  Liquid Culture Experiments  49 iii  3.2.1.1 Media Considerations 3.2.1.2 Specific Liquid Culture Experiments 3.2.2 Wood Substrate Experiments 3.2.2.1 Specific Wood Substrate Experiments 3.3  RESULTS A N D DISCUSSION  3.3.1 3.3.2 3.4  51 53 55 57 59  Liquid Culture Wood Substrate  59 67  CONCLUSIONS  C H A P T E R  4  74  ISOLATION  &  IDENTIFICATION  O F M E T A B O L I T E PI  76  4.1  INTRODUCTION  76  4.2  I S O L A T I O N OF P I  77  4.2.1 4.2.2 4.2.3 4.3  Preparative HPLC Confirmation of Purity Benzene Extraction  IDENTIFICATION OF PI  4.3.1 Methodology 4.3.1.1 H P L C 4.3.1.2 N M R 4.3.1.3 F T I R 4.3.2 Results and Discussion 4.3.2.1 H P L C 4.3.2.2 N M R 4.3.2.3 F T I R : 4.4  77 80 83 85  85 85 86 87 89 89 91 105  -.  CONCLUSIONS  C H A P T E R  5  109 F U N G A L T O L E R A N C E T O OXIDIZED Q A C S  110  5.1  INTRODUCTION  110  5.2  METHODOLOGY  110  5.3  RESULTS A N D DISCUSSION  112  C H A P T E R  6  C O N C L U S I O N S  6.1  CONCLUSIONS  6.2  RECOMMENDATIONS  APPENDIX I A P P E N D I X II  &  R E C O M M E N D A T I O N S  120 120 122  COMPLETE  'H-NMR  SPECTRA  INTEGRATION A N D PROTON COUNTS O F N M R SPECTRA  iv  130 137  LIST OF TABLES Table 3.1. Tests to determine the cause o f murkiness in V o g e l ' s medium after autoclaving  52  Table 3.2. Testing o f potential liquid medium buffers for precipitation reactions with D D A C . Note: + = noticeable murkiness (indicating precipitation) upon the addition o f D D A C . - - clear solution (indicating no reaction). K H P = potassium-hydrogen-phthalate; Citrate = sodium citrate; Acetate = acetic acid 53 Table 3.3. D D A C recovered from flasks (initially containing 12.5 mg) inoculated with V. bulbillosum. D D A C loss measured relative to the day 2 control (the day D D A C was added to the cultures). Day 16 control is an uninoculated control 65 Table 3.4. D D A C recovered from flasks (initially containing 12.5 mg) inoculated with V. bulbillosum. D D A C loss measured relative to the day 3 control (the day D D A C was added to the cultures) 66 Table 3.5. Average concentrations (mg D D A C / g wood) and D D A C disappearance for 8 wood pieces each that were either inoculated with V. bulbillosum, G. roseum, or uninoculated and incubated for a period o f 17 wks. (standard deviation in brackets) 68 Table 3.6. The recovery o f D D A C (mg D D A C / g sawdust) over 11 weeks from the three different sawdust treatments, inoculated with G. roseum. Standard deviations i n brackets  71  Table 3.7. The recovery o f D D A C from S W D I over a period o f 11 weeks and the formation o f metabolite " P I " over the same period. Standard deviations are given brackets 72 Table 4.1 Solubility test for benzyltrimethylammonium chloride ( B T M A C ) , D D A C as Bardac 2280, and dioctyldimethylammonium chloride ( D O D M A C ) in a range o f solvents.. 85 Table 4.2. Retention times o f tertiary amine acid salts and Q A C s injected into the H P L C system. D D A C standards (from same mobile phase) provide reference  90  Table 4.3. Data from ' H - N M R spectra for D D A C , P I , and Q A C - C O O H  101  Table 4.4. Data from ' H - N M R spectra for P I , Q A C - l ° O H and QAC-2°OH  101  Table 5.1. The growth rate o f four fungi over a range (0-2000 ppm) o f concentrations o f Q A C l°OH, QAC-2°OH, Q A C - O X I D , and D D A C 116  LIST OF FIGURES Figure 1.1. General structures o f major types o f alkylammonium compounds (adapted from Nicholas and Preston, 1980)  2  Figure 1.2. Total production o f Q A C s in the U S from 1966 to 1990 (adapted from S R I , 1995). 6 Figure 2.1. Potential biodegradation mechanisms for a Q A C  32  Figure 2.3. Theory o f separations by liquid chromatography (Skoog and Leary, 1992)  40  Figure 2.4. Theoretical ' H - N M R spectrum o f ethanol (adapted from Bruice, 1995)  44  Figure 3.1. The biomass curves for Verticillium bulbillosum and Gliocladium roseum grown i n pure Vogel's liquid medium (no D D A C ) . [data labels show acutal biomass values]. Single replicates used 60 Figure 3.2. Tolerance o f four fungi to D D A C (0-250 ppm) in V o g e l ' s ( K H P ) medium after 14 days incubation, [data labels show acutal biomass values]. Single replicates used. 61 Figure 3.3. Tolerance o f V. bulbillosum and G. roseum to D D A C i n Vogel's medium buffered with K H P and containing silica gel (1.0 g/ 50 ml), [data labels show acutal biomass values]. Single replicates used 62 Figure 3.4. Tolerance o f V. bulbillosum to D D A C (0-300 ppm) after 10 days incubation in Vogel's liquid culture (with acetate buffer) with and without silica gel added to the flask (1.0 g/50 ml), [data labels show acutal biomass values] 63 Figure 3.5. Parallel Recovery o f D D A C and biomass measurement from cultures o f V. bulbillosum to which 250 ppm D D A C was added at day 2. Single replicates used. 64 Figure 3.6. Replicate experiment: Parallel recovery o f D D A C and biomass measurement from cultures o f V. bulbillosum to which 250 ppm D D A C was added at day 2. T w o replicates used 67 Figure 3.7. The disappearance o f D D A C from wood pieces treated with 2% D D A C after 17 weeks. C l and C 2 = uninoculated controls; G l & G 2 = G. roseum; and V I & V 2 = V. bulbillosum. P I = metabolite from D D A C transformation 69 Figure 3.8. The rate o f biodegradation o f D D A C from three different sawdust treatments by G. roseum  70  Figure 3.9. The H P L C chromatogram o f S W D I incubated with G. roseum (G2) for 11 weeks.73 Figure 3.10. The formation o f metabolite P I in time over an 11 month incubation period o f S W D I with G. roseum vi  73  Figure 4.1. Capacity o f the H P L C column for D D A C  78  Figure 4.2. H P L C chromatogram o f the concentrated PI crude extract. The peak at 10.09 min. represents D D A C while that at 12.50 min. represents P I 80 Figure 4.3. Confirmation by H P L C that the crude P I fraction contained no D D A C contamination. A = D D A C - c o n t a i n i n g fraction (10-12 min); B = P I fraction (12-14 min) and C - the post-Pi fraction (14-16 min) 81 Figure 4.4. U V - V i s spectrum o f the concentrated P I extract from S W D week 11 prior to preparative H P L C . (reference cell = H P L C mobile phase). Inset is spectrum o f BTMAC  82  Figure 4.5. U V - V i s spectrum o f the P I fraction collected after preparative H P L C . (reference cell = H P L C mobile phase)  83  Figure 4.6. L i q u i d cell components (adapted from Perkin Elmer)  88  Figure 4.7. ' H - N M R spectra o f A : D D A C , B : Metabolite P I and C : a carboxylated Q A C over the range 0.8-2.5 ppm 97 Figure 4.8. ' H - N M R spectra o f A : D D A C , B : Metabolite P I , and C : Q A C - C O O H over the range 3.0-4.2 ppm , 98 Figure 4.9. ' H - N M R spectra o f A : Q A C - 2 ° O H , B : Metabolite P I , and C : Q A C - l ° O H over the range 0.8-2.5 ppm : 99 Figure 4.10. ' H - N M R spectra o f A : Q A C - 2 ° O H , B : Metabolite P I , and C : Q A C - l ° O H over the range 3.2-4.6 ppm 100 Figure 4.11. F T I R spectrum o f blank solvent C H C 1 over the wavenumber range o f 4000-650 cm"' 106 3  Figure 4.12. F T I R spectra o f PI overlain with D D A C over a wavenumber range o f 4000-650 cm"' 107 Figure 4.13. F T I R spectra o f Q A C - 1 ° O H overlain with Q A C - 2 ° O H over a wavenumber range of 4000-650 cm"' 107 Figure 4.14. F T I R spectra o f P I overlain with Q A C - C O O H over a wavenumber range o f 4000650 cm"' 108 Figure 5.1. The growth (as measured by colony diameter) o f four fungi on malt agar containing 100 ppm D D A C 113  vii  Figure 5.2. The growth rate o f four fungi over a range (100-2000 ppm) o f concentrations o f QAC-OXID 114 Figure 5.3. The growth rate o f four fungi over a range (100-2000 ppm) o f concentrations o f DDAC 114 Figure 5.4. The growth rate o f four fungi over a range (100-2000 ppm) o f concentrations o f QAC-l OH 115 0  Figure 5.5. The growth rate o f four fungi over a range (100-2000 ppm) o f concentrations o f QAC-2°OH 115 Figure 5.6. Agar plates containing D D A C and Q A C - O X I D at concentrations o f 500 ppm (left) and 2000 ppm (right), and incubated for 14 days with G. roseum (top), P. placenta (middle), and T. versicolor (bottom) 117  viii  LIST OF ABREV1ATIONS AAC  alkylammonium compound  ACQ  ammoniacal copper quaternary ammonium compound  BTMAC  benzyltrimethylammonium chloride [BzN(CH ) ] Cr  CCA  chromated copper arsenate  CDCI3  deuterated chloroform  CMC  critical micelle concentration  CTAB  cetyltrimethylammonium bromide [ ( C ) N ( C H ) ] B r "  D, BDMAC  dodecylbenzyldimethylammonium chloride [(C, )BzN(CH ) ] Cr  DDAC  didecyldimethylammonium chloride [(C1 ) N(CH ) ] C1"  DDT  bis(p-chlorophenyl)trichloroethane (an insecticide)  DOC  dissolved organic carbon  DODMAC  dioctyldimethylammonium chloride [ ( C ) N ( C H ) ] C l "  D0 DMAC  dioctadecyldimethylammonium chloride [(C, ) N(CH ) ] C1"  DTDMAC  ditallowdimethylammonium chloride [(C , 67%; C , 31%; C  DTMAB  decyltrimethylammonium bromide [(C )N(CH ) ] Br~  FTIR  Fourier transformed infrared spectroscopy  GC-MS  gas chromatography - mass spectrometry  HDTMAC  hexadecyltrimethylammonium chloride [ C N ( C H ) ] C r  HPLC  high performance liquid chromatography  KHP  potassium-hydrogen phthalate  LAS  linear alkyl benzene sulfonate  L C 50  lethal concentration  2  1 8  +  3  3  +  16  3  3  +  2  3  2  +  0  2  3  2  '>' +  g  2  3  2  +  8  2  3  18  2  1 6  +  10  3  3  +  16  (50%)  ix  3  3  2%) N(CH ) ] C1 +  1 4  2  3  2  LD  5 0  lethal dose (50%)  LDMBAC  lauryldimethylbenzylamrnoniurn chloride [ ( C ) B z N ( C H ) ] C l "  LiP  lignin peroxidase (an enzyme)  MnP  manganese peroxidase (an enzyme)  NMR  nuclear magnetic resonance ( ' H - N M R = proton nuclear magnetic resonance)  OTAC  octadecyltrimethylammonium chloride [ ( C , ) N ( C H ) ] C r  PI  label given to fungal metabolite from D D A C degradation  PAH  polyaromatic hydrocarbons  PCB  polychlorinated biphenyl  PCP  pentachlorophenol  PEEK  polyetheretherkeytone (material used for chromatography tubing).  ppm  parts per million  QAC  quaternary ammonium compound  QAC-COOH  a Q A C with a terminally carboxylated alkyl chain  QAC-l°OH  a Q A C with a primary alcohol on alkyl chain  QAC-2°OH  a Q A C with a secondary alcohol on alkyl chain  QAC-OXID  a Q A C with long, ethoxylated alkyl chains  SCAS  semi-continuous activated sludge  +  12  3  +  8  3  3  S W D (I, II, III) sawdust types treated in different ways TD TMAB  tetradecyltrimethylammonium bromide [ ( C ) N ( C H ) ] B r  TMEAC  trimethylethylammonium chloride [ ( C ) N ( C H ) ] C f  TMS  tetramethylsilane  UV-Vis  ultraviolet-visible light spectroscopy  1 4  2  +  ]4  3  +  2  3  3  3  ACKNOWLEDGEMENTS  I would like to thank, first and foremost, Dr. John N . R . Ruddick for his patience, and his unfailing optimism. The opportunities that I was provided with during this Masters program have gone beyond my expectations, and I know that I have been very fortunate to have had h i m as a supervisor. I am deeply grateful to Dr. Jochen Biirgel for his friendship, and for giving me no illusions about what I was getting into. Je voudrais dire merci a Dr. Colette Breuil pour son aide et ses avis. Elle a ete un membre de comite precieux. I would also like to thank Dr. Alessio Serreqi for his synthesized compounds, and for his approachability and valuable discussions, especially during times when technical support in our lab was lacking. I would like to thank Dr. Reto Riesen for his help in interpreting the N M R spectra and for editing large sections o f the thesis draft, and to Jennifer Jarvis for her editorial labours. I would like to extend a deep thanks to my family and friends for their constant support, understanding, and patience. " Y e s , it is finished now". To my colleagues o f the W o o d Preservation Lab, past and present, thanks for your good humor. It has been a real pleasure working with you. This research was supported by an NSERC/Industrial Postgraduate Fellowship, with Lonza Inc. as the industrial sponsor.  xi  CHAPTER 1  BACKGROUND  1.1  Structure, Properties, and Applications Fatty amines are a class o f compounds based on a nitrogen atom to which organic  substituents, such as methyl, alkyl, or benzyl groups, are attached.  Primary, secondary, and  tertiary amines are formed by the successive replacement o f the hydrogen atoms on the nitrogen by organic substituents.  These amines can react with an acid to form ionic amine acid salts,  which are tetrahedral in shape. When four organic groups are attached to the central nitrogen, a quaternary ammonium compound ( Q A C ) is formed with a stable positive charge.  Four main  types o f Q A C s are manufactured, depending on the bonded organic substituents: alkyltrimethyl-, alkylbenzyldimethyl-,  and  dialkyldimethylammonium , salts.  In  addition,  there  are  alkylpyridinium salts, i n which the nitrogen is incorporated into a benzene ring (Figure 1.1). The amine acid salts and quaternary  ammonium compounds are collectively referred to as  alkylammonium compounds ( A A C s ) . Q A C s are amphiphilic molecules, since they contain both non-polar character due to the hydrophobic alkyl chain, and hydrophilic properties arising from the cation. Because o f this combination Q A C s are surfactants. In aqueous solution, the hydrophobic portion aligns so that it has as little contact with water as possible, while the polar portion o f the molecule associates freely with water. This is thermodynamically most favorable. A s a result, surfactant molecules in aqueous solution tend to accumulate at interfaces where the water contacts solids, air, or nonaqueous liquids, reducing the strong surface tension o f water. For this reason Q A C s can promote foaming, emulsification, and dispersal o f hydrophobic particles within an aqueous medium. 1  ammonia  N  H  1° •  I II III IV V  amine  Alkyldimethylammonium chloride Alkyltrimethylammonium chloride Alkylbenzyldimethylammonium chloride Dialkyldimethylammonium chloride Alkylpyridinium chloride  2  H H  amine  3  R  amine  Cl V  Figure 1.1.  General structures o f major types o f alkylammonium compounds (adapted from Nicholas and Preston, 1980).  Because o f this special property,  surfactants have proven useful in a myriad o f  applications. Q A C s are used i n many industrial sectors as emulsifiers and drilling muds and in the textile industry as anti-static agents (Boethling, 1984). A major use o f Q A C s is as additives in a wide range o f consumer products, including fabric softeners, shampoos, hair conditioners, cosmetics, and room deodorizers (Boethling, 1984; Lewis and Wee, 1983). A s a manufacturer o f Q A C s once explained, "we are the 2% in everything". In addition, many Q A C s exhibit strong  germicidal properties and are employed widely as slimicides in pulp mills, algicides in cooling water systems and swimming pools, and as disinfectants in hospitals (Makrell et al, 1978). A n attractive feature about Q A C s as biocides is that they have very low mammalian toxicity (Isomaa, 1975; Hughes, Millburn, and Williams, 1973) with an acute oral L . D . from 400 to 1000 mg/kg (Butcher et al,  5 0  to rats ranging  1977). A s a result, they are widely used to prevent  microbial spoilage in the food industry, ranging from dairy products to fish processing (Reuvers and Ortiz, 1989; Vallado and Sandine, 1994).  The toxicity o f Q A C s to algae, fungi, and bacteria appears to be closely related to their surfactant  properties.  When the surfactant  properties  o f Q A C s are neutralized through  complexation with anionic surfactants, toxicity is greatly reduced (Gerike et al, 1978; Games et al, 1982). It is widely believed that the antimicrobial activity o f Q A C s occurs as a result o f the disruption o f the cell surface, though the precise mechanism is still under investigation. \ Ahlstrom et al. (1997) found the killing o f the yeast Candida albicans by Q A C s to be closely/ related to the binding o f the compounds to the cells as well as damage to the cell membrane.  Interestingly, anionic surfactants have a much lower toxicity to microbes than cationic surfactants (Swisher, 1987). Bacterial and fungal cell walls and membranes have a net negative charge that would tend to repel anionic surfactants, and attract cationic surfactants such as Q A C s . Many functions essential to the life o f fungal and algal cells are carried out i n membrane-bound organelles such as mitochondria or chloroplasts.  Q A C s that enter the cell could potentially  disrupt these organelle membranes, thus weakening or killing the microbes in this fashion. Steffann et al. (1988) studied the effect o f a benzalkonium bromide on fungal mitochondria, and showed that at very low concentrations this Q A C could impair energy production.  3  In addition, Q A C s may be deleterious to microbes i n ways other than by cell surface disruption.  Anderson and Reynolds (1965) noted that alkylammonium chlorides could act as  competitive inhibitors o f alcohol dehydrogenase enzymes in yeast cells.  Ueno and Y o k o y a  (1996) studied the effect o f Q A C s on the metabolic inhibition o f several microorganisms. They found a correlation between species resistance and differences in cell membrane permeability. Furthermore, lactate dehydrogenase enzymes extracted from both resistant and sensitive species were equally inhibited by cetyltrimethylammonium bromide ( C T A B ) within the concentration range required for growth inhibition. These results, while inconclusive, do not exclude enzyme inhibition as a possible cause o f Q A C toxicity.  Extensive laboratory testing has revealed a correlation between the structure o f the: A A C and its effectiveness against fungi. Butcher and Preston (1973) studied the toxicity o f monoalkyl tertiary amine acid salts with alkyl chain lengths from C - C , to basidiomycete fungi. 8  8  They  found that compounds with alkyl chains o f 12 to 14 carbons were the most toxic to the fungi. When two alkyl chains were present on a Q A C , the maximum toxicity occurred at chain lengths of C - C 8  1 0  (Nicholas and Preston, 1980).  Similar findings were also made for the toxicity o f  Q A C s to the protozoan Plasmodium falciparum, the microbe that causes malaria (Calas et al, 1997).  1.2  Environmental Impacts While Q A C s have been shown to have little toxic effect on mammals, this is not the case  for aquatic organisms. A wide variety o f aquatic organism including fish, mollusks, barnacles, starfish, shrimp, and rotifers, among others, are susceptible to Q A C s in the parts per million (ppm) range (Taft, 1964; Pessoa, 1952; Vallejo-Freire et al, 1954; Knauf, 1973; Huber, 1979; Kappeler, 1982; Waters, 1982; Lewis and Wee, 1983). For example, when juvenile coho salmon 4  were placed in a tank o f static freshwater containing didecyldimethylammonium chloride ( D D A C ) at a concentration o f 1.0 ppm, half o f them died within a 24-hour period (24 h r - L . C . value) (Henderson, 1992). benefits.  50  In some cases this aquatic biocidal activity has shown potential  Q A C s have been examined for the control o f invasive shellfish that threaten  aquaculture (Waller, 1996) and mosquito larvae (Taylor and Schoof, 1967), as well as in applications to boat hulls as anti-fouling agents (His et al, 1996). However, for the most part aquatic organisms are non-target species that may be harmed by Q A C s i n the environment. It has been suggested that the acute toxic effects o f Q A C s on aquatic organisms result from the disruption o f gill structure and function, which may suffocate the organism (Biesinger and Stokes, 1986).  In addition, there has been significant research into the non-lethal toxic  effects o f Q A C s upon important marketable fish species. After exposing juvenile rainbow trout to 0.4 ppm o f D D A C for 24 hours, W o o d et al. (1996) found no gross disruption to gill surfaces, but noted that swimming performance decreased significantly. Johnston et al. (1997) observed that D D A C concentrations as l o w as 0.3 ppm, could disrupt the osmoregulatory balance o f juvenile coho salmon. With any toxin released to the environment, there is a concern that it may bioaccumulate. Knezoich et al. (1989) examined the potential for the Q A C hexadecylpyridinium bromide to accumulate in the tissues o f three aquatic organisms: clams, minnows, and tadpoles. After a 24hour exposure to the Q A C , the gills o f all three species contained Q A C at a concentration 10  5  times that o f the solution while much smaller amounts were detected i n body tissues. However, when returned to fresh water, the concentration o f Q A C decreased steadily, suggesting that the binding process is reversible and that bioaccumulation does not occur. There has been a dramatic increase in the U S production o f Q A C s over recent years (Figure 1.2). From 1970-1990 production has more than tripled from 32 thousand tonnes to 114 5  thousand tonnes (SRI, 1995). Q A C s accounted for 38% o f all cationic surfactants and for 3.4% o f all surfactants produced in the U S in 1991 ( U S I T C , 1993). O f the many Q A C s produced in 1991, bis-(hydrogenated tallow alkyl)dimethylammonium chloride accounted for about one third o f total production and benzyldimethyloctadecylarnmonium chloride for 5%, while the remainder were produced in significantly smaller quantities.  Most o f the applications in which Q A C s are  used ultimately lead to release into the environment.  A s a result o f the increased production o f  Q A C s , they are being detected in lakes, streams, and wastewater where they pose a potential threat to aquatic life.  U . S . P r o d u c t i o n of Q A C s  Figure 1.2. Total production o f Q A C s in the U S from 1966 to 1990 (adapted from S R I , 1995).  1.3  QACs as Wood Preservatives Over the past couple o f decades there has been a trend in w o o d preservation towards  more environmentally friendly preservatives.  Creosote, a mixture o f organic chemicals resulting  from the distillation o f coal tar, contains many polyaromatic hydrocarbons (PAHs). 6  Technical  grade pentachlorophenol (PCP), a chlorinated aromatic molecule, contains traces o f dioxin as an impurity. Although effective as preservatives, these oil-borne compounds have lost considerable market share to chromated copper arsenate ( C C A ) .  While C C A is considered to be safer as a  wood preservative due to its fixation to wood, there is still concern over the long-term environmental acceptability o f using heavy metals such as arsenic and chromium.  This has  focused attention on the development o f alternative preservative systems that do not contain metals. . In this search for effective, more environmentally friendly substitutes, Q A C s have shown considerable promise.  Butcher et al. (1977b) showed that dodecyldimethylbenzylammonium  c h l o r i d e . ( D B D M A C ) was more effective than C C A : i n preventing attack by soft-rot organisms. 12  Butcher and Drysdale (1977) compared Q A C s to C C A against seven standard decay fungi. They found a mixture of dialkydimethylammonium chloride compounds (with chain lengths o f C and 8  C , ) to be the most effective Q A C . 0  This dialkyldimethyl-QAC mix was more effective than  C C A i n protecting wood against F. vaillanti. W o o d containing this Q A C at a retention o f less than 1 kg/m resisted decay, while a C C A retention between 5-10 kg/m was required to achieve 3  3  the same level o f protection. Overall, the dialkyl Q A C was comparable to C C A i n effectiveness against the seven test fungi. Extrapolating laboratory test results for wood preservative performance in the field can be a difficult task as so many factors are involved.  Results may differ greatly depending on the  wood species, the wood type (sapwood or heartwood), the specific chemical, and the specific fungal strain used for testing. Nevertheless, extensive laboratory trials have shown that Q A C s such as D D A C are as effective as C C A in protecting wood against a broad array o f standard decay fungi (Butcher et al, 1977a; Preston and Nicholas, 1982; Butcher, 1979; Preston, 1983; Hedley etal, 1982). 7  However, when QAC-treated wood was tested in the field, the results did not reflect initial laboratory observations. Butcher et al. (1979) noticed signs of decay on the surface of QAC-treated wood stakes shortly after being installed in the field. These stakes did not perform as well as CCA-treated wood over a 30-month period. After only one year, wood stakes treated with a mixture of various QACs failed in field trials (Ruddick, 1981). In a subsequent field trial, wood stakes treated with D D A C at a retention of 3.2 kg/m failed after two years (Ruddick, 3  1983).  Morris and Ingram (1988) observed significant decay of QAC-treated wood with  retention values as high as 11 kg/m after a 6-year period. In further field studies, Preston et al. 3  (1987) found that the addition of copper as a co-biocide improved the performance of QACs only slightly. Investigations ruled out the leaching of D D A C from the wood or its uneven distribution in the wood as the cause of premature failure in the treated wood stakes (Ruddick and Sam, 1982). A number of mould fungi (hyphomycetes) were isolated from antisapstain-treated lumber (Doyle, 1995), which demonstrated the ability to tolerate (Doyle and Ruddick, 1993; Zheng, 1995) and degrade D D A C (Zheng and Ruddick, 1995). It was hypothesized that the alteration of the Q A C by certain mould and staining fungi may have detoxified them to a level where woodrotting basidiomycetes could grow and break down the wood. In further field studies, Ruddick (1987) found that modification of QACs through the addition of co-biocides prevented colonization by staining fungi, and greatly enhanced the wood-protecting performance.  1.4  Benefits of understanding QAC biodegradation To be taken into consideration as a new wood preservative, a compound must satisfy  several criteria. The ideal wood preservative should have the following properties. It is effective at protecting wood at low concentrations and specifically kills the target organisms {e.g. basidiomycetes), while having little effect on non-target organisms (e.g. mammals and fish). It is 8  strongly fixed to the wood so that minimal leaching occurs while being inexpensive at the concentrations needed to protect the wood. It has no effect on the physical or aesthetic properties of the treated wood, such as strength loss, checking, or discoloration, and does not persist in the environment once the service life o f the treated wood has ended.  C C A is an effective and  inexpensive preservative, but it discolors the wood, is potentially harmful to non-target species, and persists in the environment. While colourless and more environmentally benign, the major problem with D D A C is its biodegradability. i Despite the poor results i n field stake tests, D D A C performed well in above ground L joint tests, and was approved and standardized as a preservative for above ground use. In British Columbia, D D A C is the principle component o f most antisapstain formulations currently registered and in use by the industry (Henderson, 1992).  In addition to its effectiveness i n  protecting wood against fungal attack, D D A C has the advantages o f being colourless and having .a low mammalian toxicity.  Another QAC-containing wood preservative, ammoniacal copper  quaternary ammonium compound ( A C Q ) , has been developed for ground contact applications. However, it has the disadvantage o f strongly discoloring the wood upon application and has not been widely used to date in Canada, but is very widely used in Japan and Scandinavia (Ruddick, personal communication). Because laboratory tests have shown that Q A C s are toxic to a wide range o f aquatic organisms (Kappeler, 1982; Waters, 1982; Lewis and Wee, 1983), the B . C . Ministry o f Environment, Land and Parks has regulated limits for the storm water discharge o f D D A C at 400 pg/L (0.4 ppm) ( B C M E L P , 1994).  However, the toxicity tests upon which this regulation is  based were conducted under carefully controlled laboratory conditions.  In natural aquatic  environments, there are so many other physical, chemical, and biological variables present, that it is difficult to predict what w i l l occur, based on the laboratory results. Sediments and suspended  9  solids provide a large surface area for sorption of DDAC and have been shown to have a significant impact on reducing QAC toxicity (Krzeminski et al, 1973; Larson and Vashon, 1983). Compounds that complex or chelate QACs have also been shown to reduce toxicity (Gerike et al, 1978; Masuda et al, 1976). In addition, most QACs (including DDAC) are susceptible to biodegradation. Specific laboratory experiments are often poor models of what actually occurs in the environment. This is well illustrated by the vastly different results obtained in laboratory testing and field testing of DDAC as a wood preservative. It is unlikely that QACs pose as great a threat to the aquatic environment as the laboratory tests suggest.  Ventullo and Larson (1985)  performed in situ research on the effects of several long-chain (C to C ) QACs on aquatic 12  ]8  microbial communities in a lake. They tested mono- and dialkyl- QACs over a concentration range of 0.001-10 ppm.  In general, they concluded that the exposure to QACs at the  concentrations tested posed little threat, but resulted in adapted microbial communities within the lake which were less sensitive to toxic effects and more active in terms of biodegradation of the QACs.  1.5  Objectives  The global objective of this thesis was to determine how the two moulds first isolated by Andress Doyle and subsequently studied by Yu Zheng, were degrading DDAC. Confirmation of the metabolic pathway employed by these fungi is an important piece of information that would aid QAC-manufacturers by giving insights into what makes these compounds more effective or more biodegradable. Knowledge of the biotransformation of DDAC is an important first step to finding ways to improve the efficacy of QACs as biocides. Furthermore, this information will contribute to the growing body of knowledge concerning the fate of DDAC in the environment. 10  To achieve the global objective several specific objectives were defined:  •  Examine rates of D D A C degradation in liquid culture and wood substrates  •  Generate significant quantities of metabolite from D D A C degradation  •  Isolate the primary, cationic metabolite by preparative H P L C  •  Identify this metabolite through N M R and FTIR analysis  •  Determine whether or not the primary metabolite is less toxic to fungi than  11  CHAPTER 2 LITERATURE  REVIEW  No research has been conducted in which fungi are used as the agents for quaternary ammonium compound (QAC) degradation, prior to that reported by Professor Ruddick's research group at UBC (Burgel et al, 1996a, Burgel et al, 1996b, Dubois et al, 1997, Dubois et al, 1998, Zheng and Ruddick, 1995). Most of the literature examines sewage sludge or bacteria, either in pure culture or in a consortium, as the inocula for QAC biodegradation experiments. However, fungi have demonstrated the ability to successfully degrade a wide range of compounds, both natural and xenobiotic. This review compares bacteria and fungi in terms of their different morphology and metabolic capabilities, evaluates the available literature on the use of fungi as biodegrading agents, takes a comprehensive look at QAC biodegradation studies, and covers the basic principles behind the analytical techniques used in this research project.  2.1  Comparison of Fungi and Bacteria  2.1.1 Morphology When comparing fungal and bacterial cells several differences are immediately apparent. Perhaps the most noticeable difference is in the size of the cells. Bacterial cells vary in size and shape but are usually orders of magnitude smaller than fungal cells. A typical E. coli "rod" has dimensions of 0.5 um x 2 urn whereas fungal cells range from 2-200 p.m in diameter (Brock and Madigan, 1991). Bacterial cells may aggregate into chains, sheets, or clusters, but may also be found unattached as single cells. Fungi on the other hand, may assume a "yeast-like" growth state where they grow as single cells, but in most cases newly produced cells remain attached to  12  the old to form long filamentous hyphae.  Hyphal branching results in many long intertwining  chains that develop into a thick mycelium, often visible without the use o f a microscope.  Bacteria have an advantage over the fungi i n degrading suspended  solids.  Their  diminutive size allows them to fit into smaller spaces and crevices where a biodegradable compound may reside, inaccessible to the fungal mycelium. A l s o the small bacterial cell size gives it a greater surface area, per unit cell volume, than the fungal cell for taking in chemicals per unit biomass.  However, in solid matrices such as agar, soil, and wood, fungi have an  advantage. Penetrating apical hyphal growth permits active exploration; while a broad arsenal o f enzymes secreted from the growing tip enables the fungus to exploit many substrates, especially non-dissolved substrates (Kendrick, 1985).  Hyphae are especially well adapted to a substrate  like wood which contains many "tunnels" in the form o f resin canals, pitted tracheids, and vessel elements (Tsoumis, 1968). While several bacteria possess flagella that give them some degree o f motility, they are generally dependent on the physical transport o f their medium (e.g. water flow) to achieve their ubiquity.  This may give fungi a preferential role i n the degradation o f  preservatives in treated wood prepared for disposal, or certain advantages in remediating contaminated soils where there is little mixing.  Cells — fungal, bacterial, or otherwise ~ are largely a collection o f dissolved substances encased in a flexible, water permeable membrane. When living in a dilute aqueous environment, osmotic pressures as high as 53 kPa (Brock and Madigan, 1991) may occur within the cell easily enough to burst the membrane.  In a dry environment, water loss would cause the cell to  deflate. These forces are mitigated by the cell wall, which gives shape, rigidity, and strength to the cell. Fungal and bacterial cells differ considerably in the structure and composition o f their cell walls.  13  The distinguishing feature o f the bacterial cell wall is the presence o f peptidoglycan, a sugar polymer cross-linked with amino acids.  T w o main cell wall types prevail i n bacteria,  distinguished by their Gram stain. Gram-positive cells have a cell wall composed almost entirely o f peptidoglycan, while gram-negative cell walls have a smaller proportion o f peptidoglycan and an additional outer membrane. The fungal cell wall, on the other hand, is a thick, rigid structure that is composed almost entirely o f chitin, a cross-linked polysaccharide (Kendrick, 1985). This strong chitinous cell wall is waterproof and allows the fungus to grow i n very dry environments, giving the fungus another advantage over bacteria in colonizing solid substrates.  2.1.2 Metabolism Metabolism is the sum o f all biochemical reactions catalysed by an organism. reactions can be divided into two main groups:  These  degradative (catabolic) and biosynthetic  (anabolic). The catabolic reactions are o f primary interest for the purpose o f biodegradation. Bacteria have a much greater metabolic diversity than fungi.  In addition to obtaining  energy from the oxidation o f organic compounds (heterotrophy), some bacteria are capable o f harnessing the energy from light (photosynthesis), while still others can oxidize inorganic compounds such as iron as an energy source (lithotrophy).  When organic compounds are  oxidized as a source o f energy, bacteria again display an impressive diversity. Whereas fungi rely on oxygen to grow on organic compounds through conventional aerobic respiration, certain bacteria employing anaerobic respiration can grow in the absence o f oxygen, using N 0 , S 0 " , _  3  2  4  C 0 , or a range o f other inorganic and organic electron acceptors (Brock and Madigan, 1991). 2  Pseudomonas is a genus o f gram-negative bacterium that can utilize a wide variety o f organic compounds as a carbon and energy source. different  compounds (Silver et al,  1990).  Some strains have been observed to use over 100 Perhaps it is not surprising then that most 14  microorganisms isolated from sludge that were capable o f using Q A C s as nutrients were species o f this genus.  Though lacking the metabolic diversity o f the bacteria, fungi play a very important role as biodegraders in nature.  Fungi, using cellulase and lignin peroxidase enzymes, play the major  part in decomposing the voluminous woody and plant material o f the world's forests. Cellulose and lignin are large polymeric molecules that are tightly associated to make up the structural bulk o f most woody plant tissue. Due to the large size o f these complexes, microbes cannot transport them across their cell walls to metabolize them.  Fungi overcome this problem by secreting  enzymes that break down the cellulose and lignin into smaller molecules, which are then taken up by the cell. Extracellular enzyme secretion is a major difference between fungi and bacteria, as the bacteria generally retain their degradative enzymes within the cell.  ;  In bacteria, the  periplasmic space between the cell wall and cell membrane contains hydrolytic enzymes that function i n the initial degradation o f food molecules. However, the impermeability o f the outer membrane o f gram-negative bacteria to enzymes acts to inhibit them from being secreted (Brock and Madigan, 1991).  2.2  Fungal Biodegradation To date, the bacteria have been the dominant kingdom o f organisms studied in terms o f  biodegradation and bioremediation. However, the use o f fungi in this capacity has generated considerable interest recently as evidenced by the numerous journal articles published on this topic since 1994.  Fungi have evolved enzymes capable o f digesting some very recalcitrant  substances in nature. Keratin, which makes up hair and feathers; chitin, the main component o f insect and crustacean exoskeletons; cellulose, the major component o f most plant debris; and lignin, found in wood, are very resistant molecules that succumb to fungal degradation (all but 15  lignin are also degraded by bacteria). It has been demonstrated that fungi are able to break down many xenobiotic compounds ranging from polyaromatic hydrocarbons ( P A H ) to chlorinated contaminants such as pentachlorophenol (PCP).  A t the forefront o f fungal biodegradation  research are the white rot fungi. However, mould fungi have also been shown to have potentially important talents for biotransformation.  2.2.1 White-Rot Fungi White-rot fungi are basidiomycetes capable o f degrading both the cellulose and lignin compounds that make up the bulk o f woody plant tissue. M a n y bacteria and fungi possess the enzymes required to break down cellulose, but few microbes have the enzymes to degrade lignin. L i g n i n is a very complex amorphous polymer, derived o f phenylpropane units cross-linked by ether bridges (Tsoumis, 1968).  This complex, variable structure poses certain challenges for  degradative enzymes specific for the precise shapes o f their substrate.  Phanerochaete  chrysosporium is a white-rot fungus that is highly efficient at degrading lignin in the presence o f another carbon and energy source. P. chrysosporium has developed a number o f enzymes for lignin biodegradation that are highly non-specific. The primary enzymes used for the process are lignin peroxidase (LiP) and manganese peroxidase (MnP).  Within lignin there are many structures similar to toxic aromatic compounds (Kennes et al,  1994).  It has been suggested that the non-specificity o f these enzymes makes P.  chrysosporium a prolific degrader o f xenobiotic compounds as well. A m o n g the compounds that P. chrysosporium is known to be able to partially degrade are P A H (including benzo(a)pyrene, benzo(a)anthracene, and pyrene), several P C B ' s , 2,3,7,8-tetrachlorodibenzo-p-dioxin, D D T , and lindane (Alexander, 1994).  Most o f these compounds are important contaminants, many o f  which are not readily degraded by bacteria. 16  2.2.2  Hyphomycetes The hyphomycetes, also known as moulds, are a large group o f relatively simple fungi  that lack a sexual reproductive cycle. Moulds also lack the lignin degrading enzymes that play an important role in the biodegradation o f xenobiotics by white-rot fungi.  However, the  biotransformation o f several recalcitrant compounds by hyphomycetes has been observed. A s with the 'white-rot' fungi, an important group o f compounds degraded by hyphomycetes are the P A H compounds. Chrysene is a 4-ringed P A H that is strongly recalcitrant, with only a few documented reports o f its microbial degradation. Penicillium  sp. and a Syncephalastrum  Kiehlmann et al. (1996) isolated four hyphomycetes (3 sp.) that were capable o f oxidizing chrysene.  Using  radiolabeled chrysene, they isolated 3 metabolites by H P L C and identified one as trans-1,2dihydroxy-l,2-dihydrochrysene (a chrysene molecule with 2 hydroxyl groups attached). Pyrene, a P A H closely related to chrysene, was found to be metabolized by a strain o f  Penicillium  glabrum (Wunder et al, 1997). Several metabolites were isolated from the transformation and identified as various hydroxylated and methoxylated pyrene structures.  Wunder and his group  postulated that the metabolism o f P A H by non-white-rot fungi involves a cytochrome P-450 monooxygenase enzyme system.  M o u l d fungi have also been observed to degrade xenobiotic compounds other than P A H .  For  example, organophosphonates are xenobiotic compounds with a C - P bond; i n nature carbon and phosphorus are joined through a C - O - P bond.  Several o f these compounds, including the  herbicide glyphosate, were used to screen for the ability o f 26 soil borne fungi to degrade them (Krzysko-Lupicka et al,  1997).  From the screening, strains o f Trichoderma  harzianum,  Aspergillus niger, and Scopularopsis sp. could degrade all organophosphonates tested at 10 m M when they were used as a sole source o f phosphorus. 17  The bacterial degradation o f organophosphonates  has already been well-established  (Krzysko-Lupicka et al, 1997); however, the study o f fungi in this role is relatively new. While fungi w i l l probably not rival bacteria for the spotlight on bioremediation as a whole, they do hold considerable promise for certain situations, especially for the biodegradation o f persistent P A H compounds.  2.3  Biodegradation of Quaternary Ammonium Compounds Microorganisms degrade organic compounds for three main reasons:  1. the compound may be used by the microbe as a nutrient for growth (food source)  2. the compound may be modified to reduce its toxicity (detoxification)  3. the compound may be transformed with no benefit to the microbe (co-metabolism) A large number o f synthetic compounds can be utilized by microbes to support their growth. Q A C s may act as sources o f carbon, nitrogen, and energy for appropriate organisms.  When a  Q A C is the sole source o f an essential nutrient, there is a selective advantage to organisms that can utilize it.  When a toxic compound is present there is selective advantage to microbes that can remove the offending  compound, or alter it to reduce  or remove its toxic properties.  Numerous  transformations can be done to reduce the toxicity, often only involving a slight modification o f the molecule (Alexander, 1994), for example, the methylation o f P C P . When detoxification occurs, the degradation is often incomplete; once the concentration is below the non-toxic threshold, the microbe derives no advantage in further degradation.  18  Co-metabolism may best be described as fortuitous biodegradation i n which enzymes produced to break down a particular substrate also by chance degrade a pollutant compound with some structural similarity.  This is an important phenomenon i n the biodegradation o f xenobiotics,  especially for transformations caused by white-rot fungi (Kennes et al, 1994).  The ideal goal o f bioremediation is to achieve ultimate biodegradation (mineralization) i n which the original organic pollutant is completely broken down into C 0 , H 0 and inorganic 2  2  salts. However, biodegradation events do not always lead directly to mineralization. Often some of the resulting intermediates are resistant to further biodegradation and thus tend to accumulate. Sometimes the metabolites are less desirable (increased toxicity) than the mother compound (e.g. methylation o f P C P to anisoles).  In many cases a microorganism w i l l not 'break-down' a  pollutant molecule per say, but rather alter it, or biotransform it, by adding functional groups or by other means.  2.3.1 Factors affecting Biodegradation 2.3.1.1 Microbial Considerations In the environment numerous factors determine whether or not a compound w i l l undergo biodegradation. Firstly, appropriate microorganisms — those capable o f degrading Q A C s — must be present in the system.  Microbes representing a wide range o f metabolic capabilities are  ubiquitous in natural soil and aquatic systems, though often in small numbers. There is usually an initial delay in significant biodegradation until the population o f capable microbes grows to an appreciable size.  In order for this to occur, the conditions must be appropriate for the  microorganisms to grow and degrade the compound. Essential nutrients and trace elements must be present for growth to occur. However, temperature, p H , appropriate aeration, moisture, and even the amount of solar radiation may act to limit microbial growth (Boethling, 1984). 19  Microbes may also be inhibited by toxic substances, and are susceptible to competition with other bacteria and fungi for essential growth factors. M a n y Q A C s exhibit antimicrobial activity and w i l l prevent the growth o f microorganisms at high concentrations (above the toxic threshold).  Many metabolic enzymes within bacteria and fungi are inducible, and as such they are produced only when the substrate they act upon is available.  Acclimation is the process o f  exposing microbes to a compound at a gradually increased concentration. Ginkel et al. (1992) grew  cells  of  the  same  bacterial  isolate  hexadecyltrimethylammonium chloride ( H D T M A C ) .  separately  on  acetate  and  on  Later, the growth rate was measured for  both acclimated cell types on various substrates. Cells raised on H D T M A C , when transferred to acetate as a sole source o f carbon and energy had a growth rate only 2 5 % that o f the original cells grown on H D T M A C ; while the acetate-acclimated cells could degrade acetate. at a 5 times greater rate than they could degrade H D T M A C . Acclimation can have a profound influence on the biodegradation o f Q A C s , as supported by the literature. In acclimated biological treatment systems, the removal o f Q A C at non-toxic concentrations (Boethling,  1984).  Gerike  hexadecyltrimethylammonium  et  bromide  concentration was gradually raised. sewage  sludge  to  16  al.  (1978) (at  found  15 ppm) was  should generally exceed 90%  that  >  90%  removed, but  of  the  initial  only when  the  Janicke and Hilge (1979) found that after acclimation o f  ppm  of  ditallowdimethylammonium  chloride  and  alkylbenzyldimethylammonium chloride, a net elimination o f 96% could be achieved.  2.3.1.2 Bioavailability In order for contaminant compounds to undergo biodegradation they must be accessible to the microbes with the requisite catabolic enzymes. It is widely believed that compounds must 20  be in solution to be metabolized (Swisher, 1987), as the enzymes responsible for biochemical reactions rely on the very specific shape imparted on them through interactions with the water they are dissolved in.  Certain processes such as chelation, complexation, precipitation and  adsorption can remove Q A C s from solution and potentially make them more difficult for microbes to access (Alexander, 1994).  Q A C s are surfactants, and as such they have an affinity for sorption onto solid substances. Q A C s have been found to bind strongly to glass, sewage, proteins, and soil (Barbara and Hunter, 1965; Lawrence, 1970; MacQuillen, 1950; Swisher, 1987). Q A C s also readily bind onto the cell walls o f bacteria and yeast (Fujita and Koga, 1966; Krzeminski et al, Mackrell and Walker, 1978).  1973;  Clay minerals and soil organic matter have a large number o f  negatively charged sites which allow Q A C s to interact: electrostatically with their cationic nitrogen as well as through V a n der Walls forces and hydrophobic interaction (Theng, 1974). In an aqueous environment the presence o f sediments and other suspended solids can bind Q A C and effectively remove it from solution. Games et al. (1982) and Sullivan (1983) determined that the rate at which Q A C s are adsorbed is significantly faster than the rate at which they are degraded. Greater than 99% O T A C was adsorbed by wastewater solids within 30 minutes, while the halflife for O T A C primary biodegradation was about 2.5 hours (Games et al, 1982).  In a similar manner, Q A C s have a strong affinity to complex with anionic surfactants. This phenomenon is especially important in wastewater systems where the average concentration of anionic surfactants greatly exceeds (by an order o f magnitude) that o f cationic surfactants (Swisher, 1987; Boethling, 1984). The formation o f these complexes eliminates the surfactant properties o f the original compounds.  Their reduced solubility in water, causes them to  21  precipitate. Thus sorption and complex formation act to limit the amount o f compound that is bioavailable, and limit or control the rate o f biodegradation.  A large number o f experiments from the literature shows that biodegradation o f Q A C s is in fact enhanced by sorption and complexation. There have been several studies which clearly demonstrate that the addition o f sediment can encourage the biodegradation o f Q A C s at much higher levels than can be biodegraded without the sediment (Krzeminski et al, 1973; and Larson and Vashon,  1983).  In a similar manner, the addition o f anionic complexing agents can be used  to enhance the biodegradability o f Q A C s (Gerike et al, 1978; Masuda et al, 1976; and Games et  al, 1982). It has been noted that the aquatic toxicity o f many anionic and cationic surfactants is significantly reduced by complexation (Knauf,  1973)i  This is l i k e l y due to the removal o f the  Q A C s surfactant properties.  2.3.1.3 Recalcitrance Recalcitrance is the inherent resistance o f a compound to degradation.  A s enzyme  activity is often highly specific for its substrate, recalcitrance may result from chemical structures that do not easily fit into the active sites o f enzymes. The two major structural aspects o f Q A C s that influence their resistance to biodegradation are the numbers o f alkyl or aryl (aromatic) substituents attached to the central nitrogen, and the length o f these alkyl chains. Nishiyama et al.  (1995)  studied the degradation o f a series o f straight-chain alkyltrimethylammonium  bromides, in which the alkyl chain length ranged from C to C , by measuring 0 8  l g  2  consumption.  After 28 days incubation with a sludge inoculum, 47% o f the C compound was degraded, while 8  the amount o f 0 consumed steadily decreased with increasing chain length, there being only 7% 2  degradation o f C , . 8  22  Ginkel and Kolvenbach (1991) examined the recalcitrance o f a number o f Q A C s using oxidation rates as a measure o f biodegradation. They found that the resistance to biodegradation increased to a small extent as the alkyl chain length increased from C dialkylammonium compounds.  1 0  to C  l g  in the mono- and  They also found that the recalcitrance increased to a much  greater degree with a greater number o f long alkyl substituents attached to the nitrogen atom. In fact, a clear hierarchy o f recalcitrance can be seen among Q A C s depending on their substituents. For Q A C s containing linear hydrophobic chains the recalcitrance increases from left to right in Equation 1. (Boethling, 1984):  RMe N 3  +  < RBzMe N 2  +  < R Me N 2  2  +  < RPy  +  < R MeN 3  (1)  +  where: R = linear alkyl chain, M e = methyl, B z =; benzyl group, P y = pyridinium +  This relationship has been consistently demonstrated in the literature.  Masuda (1976)  measured oxygen uptake after 10 days incubation with various Q A C s at 20 ppm. When grown on decyltrimethyl- and decylbenzyldimethylammonium chloride, the degradation approached 100%, while only 50% degradation occurred when grown on didecyldimethylammonium chloride and 4% for decylpyridinium chloride. Cruz and Garcia (1979) examined the time required for 5 ppm of selected Q A C s to decrease by 50%. One to two days was required for alkyltrimethyl and alkylbenzylammonium chlorides, while dialkyldimethylammonium chlorides required 2 to 3 days and trialkylmethylammonium compounds took approximately 7 days. Dean-Raymond and Alexander (1977) used oxygen uptake to study the degradation o f several Q A C s . that their  microbial  culture  could  only  degrade the  alkyl  trimethyl  They found  compounds,  the  benzyltrimethyl or dialkyldimethyl compounds remaining undegraded. Enriched cultures grown on tetramethylammonium chloride were not able to utilize tetraethylammonium chloride, triethylmethylammonium chloride, or diethyldimethylammonium chloride (Ghisalba and Kiienzi,  23  1983).  However, Mackrell and Walker  (1978) found  that their Pseudomonas  adapted to grow on ethyltrimethylammonium chloride.  isolates could be  They suggested the close-packed  tetrahedral structure o f the higher substituted homologues was likely protecting the central nitrogen atom against enzymatic degradation.  2.3.2 Techniques  for Studying  Biodegradation  In principle the study o f biodegradation is straight forward — add the compound o f interest to microorganisms and observe its fate.  In practice difficulties arise due to the many  environmental factors that influence biodegradation. Studies o f Q A C biodegradation often fail to take all o f these factors into account making the results difficult to interpret.  Especially  important is the tendency o f Q A C s to adsorb onto solids. M a n y techniques have been developed to characterize different aspects o f biodegradation, but no one single technique is available that can give a comprehensive picture o f what is happening.  Nevertheless, a combination o f  approaches can provide evidence from different perspectives to cumulatively build a strong case.  2.3.2.1 Measuring QAC loss A common approach for studying biodegradation is to measure the loss o f Q A C over time. A s Q A C s are not directly visible, analytical techniques must be used to "see" them and determine their concentration. These methods exploit a specific property o f the compound, such as its cationic charge, but may not detect modification o f the original compound which do not impart on this basic property.  Therefore they are only useful i n determining the primary  degradation o f the Q A C and not the extent o f degradation that it undergoes. A major advantage of this procedure though is that primary degradation can be followed, whether the microorganism is using the Q A C as a sole source o f carbon and energy, detoxifying the Q A C , or degrading it cometabolically. 24  Colourimetric techniques have been a prime method for the detection o f cationic surfactants. This method involves adding an anionic dye, such as disulfine blue, that complexes with cationic surfactants in solution. The dye and Q A C form an ion pair that when extracted with chloroform, partitions to the organic phase (Swisher, 1987). The uncomplexed dye remains in the aqueous phase as a halide salt. The concentration o f Q A C is then determined by the colour intensity in the organic phase.  Problems are encountered with this technique because o f  interference with other organic cations (e.g. amines) that may react with the dye. Several methods for the detection and quantification o f Q A C s have also been developed based on H P L C , but their effectiveness depends on a complete extraction o f Q A C from solution.  Despite its usefulness in monitoring primary degradation, measuring the change in concentration o f the target compound has limitations. Firstly, where adsorption is also, present it is often impossible to distinguish between biodegradation and reduction due to adsorbance. Secondly, there is no indication o f the degree to which degradation takes place. The original compound may accumulate as a slightly modified metabolite that may be less desirable than the original compound.  2.3.2.2 Measuring Removal, Biotransformation, and Biodegradation The extent o f biodegradation can be determined in an experimental system by nonspecific methods such as loss o f organic matter and respirometry.  Measuring the loss o f  dissolved organic matter is a fairly simple procedure. The culture is filtered or centrifuged and the supernatant analyzed for total organic matter. However, as with measuring the loss o f the specific Q A C , this technique is susceptible to major errors due to adsorption and precipitation. Also some soluble organic compounds are secreted by the microbes as a natural part o f their metabolism, which can lead to erroneous results. 25  Respirometry is a useful technique for measuring the extent o f biodegradation o f a substance.  It is based on the principle that biodegradation is essentially an oxidative process  (Smith, 1967).  In a closed system the microbial respiration is measured as, either the 0  2  consumed or the C 0 emitted, when the culture is grown on a substrate. After correcting for 2  endogenous respiration, the amount o f degradation after a certain incubation period is calculated as the amount o f 0  2  consumed, relative to the theoretical amount o f 0  2  that would be necessary  for complete oxidation. For example, the complete oxidation o f 180 mg o f glucose (1 mole) would consume 192 mg o f 0  2  (6 moles) and evolve 264 mg o f C 0 (6 moles) (Equation 2). In 2  practice, the values measured actually correspond quite well to the theoretical values (Swisher, 1987).  C Hi 0 6  2  6  + 60  2  —6  C0  2  + 6H 0 2  (2)  When examining Q A C biodegradation, oxygen uptake respirometry has the main advantage o f being independent o f the state o f the test Q A C whether dissolved, precipitated, or adsorbed.  Most other analytical systems require the Q A C to be dissolved. Although it can  provide a good idea o f the extent o f biodegradation, respirometry does have its limitations. Firstly, the Q A C that is to be degraded must be used as a sole source o f carbon and energy, requiring that the microbial culture be capable o f metabolizing it this way.  This rules out  respirometry as a means o f determining degradation rates o f compounds that are degraded cometabolically.  Secondly, respirometry vastly underestimates  metabolism. In certain cases bacteria w i l l use 0  2  the complexity o f microbial  to oxidize ammonia or amino compounds to  nitrate (Swisher, 1987). When measuring C 0 evolution it is important to note that not all o f the 2  degraded Q A C is converted into C 0 , some is incorporated into the biomass. 2  26  Thirdly, while  respirometry can measure the completeness o f biodegradation, it can not determine how the compound is degraded, or identify any intermediate metabolites that may be produced.  2.3.2.3 Determining Degradative Mechanism Neither measuring the specific loss o f a compound, nor the extent o f biodegradation through respirometry, gives insights into how the compound is degraded.  B y collecting and  identifying metabolites from Q A C degradation, insights into the degradative pathway can be gained. G C - M S and ' H - N M R are powerful analytical tools often used i n identifying metabolites. However, many intermediates i n the breakdown o f a compound have a very short lifetime. The success in identifying metabolites usually requires that the metabolites accumulate i n the growth culture either as the rate-limiting step in a slow reaction, or due to the inability o f the degradingmicrobes to completely mineralize the compound.  It is possible to confuse compounds that arise from regular bell metabolism with those that may be degradation products. The use o f radioisotope-labelled Q A C s , with either is invaluable in removing this uncertainty (Games  al,  1982).  l 5  N or C , l 4  Radiolabel-tracer studies are  also valuable in determining the fate o f the Q A C during degradation.  B y labelling different  carbon atoms within a Q A C , differences can be observed in their conversion to C 0 , their 2  incorporation into biomass, or their release in the solution as soluble metabolites or as undegraded Q A C . Sullivan (1983) used  1 4  C to label a Q A C at three different positions within the  molecule to determine which structures on the molecule were more susceptible to degradation.  Once a hypothesis of the degradative mechanism is suggested, enzyme assays may provide useful information.  Demonstrating the presence o f the requisite enzymes in the  degrading organism for a particular biodegradation pathway, strengthens the likelihood o f the proposed pathway.  Ginkel et al. (1992) used enzyme assays to show that their 27  Pseudomonas  strain had the enzymes required to cleave the alkyl chain from H D T M A C and degrade it through B-oxidation.  2.3.3 Evidence of Biodegradation Primary biodegradation (Swisher, 1987) is a term used to describe the biological transformation o f a molecule to the extent that its characteristic properties are changed or it no longer responds to the analytical technique used to monitor it.  According to this broad  definition, primary biodegradation may involve only a slight modification o f the Q A C molecule or it may involve complete degradation, depending on the analysis. I f the analytical technique used to detect the initial compound is very specific, then small changes to the molecule may escape being detected.  With such an analysis, one could only determine i f a compound was  being transformed, but not the degree to which degradation was taking place.  It essentially  means that biodegradation has occurred to some extent. Conversely, mineralization (or ultimate biodegradation) is used to describe the complete decomposition o f an organic compound into C 0 , water, biomass, and inorganic salts (Swisher, 1987). These terms are useful as boundaries 2  that define the range o f biodegradation.  From the results o f simple experiments, it can be difficult to fully understand what is happening. When only 60% o f the original compound is degraded, why is the remaining 40% not removed?  When 100% primary biodegradation occurs, are all o f the Q A C molecules  degraded to the same extent or is there a range o f metabolites, which are oxidized to differing degrees? Are any abiotic processes (e.g. adsorption) involved in the loss o f Q A C ? When 50% o f the theoretical 0 uptake is observed, is the microbe using the 0 2  2  to preferentially degrade a part  o f the molecule (e.g. the alkyl chain) or is the oxidation indiscriminant, attacking all Q A C  28  functional groups equally? When reviewing the literature on Q A C biodegradation, care must be taken in interpreting the results, especially the earlier work done prior to 1980.  Gerike et al. (1978) investigated the biodegradation o f D B D M A C using an activated I 2  sludge system.  The elimination o f the Q A C was monitored colourimetrically using disulfine  blue and by dissolved organic carbon ( D O C ) .  Under these conditions, the D B D M A C was I 2  reduced by 54%. It was concluded that this Q A C was degradable at least at a primary level, but both the disulfine blue and D O C measurements were subject to interference by adsorption and may have reported higher biodegradation rates than actually occurred.  . Based upon oxygen uptake respirometry, Masuda et al. (1976) observed that 80-100% o f alkylbenzyldimethylammonium salts (alkyl chains: C  g  to C ) and 50% o f didecyldimethyl M  ammonium chloride ( D D A C ) were biodegraded in 10 days, while higher homologues were not degraded at all. The addition o f sodium dodecyl sulfate, an anionic surfactant, increased the biodegradation o f D D A C to 80% in 20 days and that o f hexadecylbenzyldimethylammonium chloride to about 4 5 % in 15 days.  Larson  and  Vashon,  (1983)  confirmed  the  biodegradation  of  dioctadecyldimethylammonium chloride ( D 0 D M A C ) in river water at 5 and 50 ppb, based l g  upon the amount o f  l 4  C0  labelled uniformly with  , 4  2  produced during biodegradation o f the alkyl chains which were  C . Without any sediment the  , 4  C 0 production was much less (10 and 2  20% o f the theoretical value) than that achieved when 5 g/L o f sediment was present (65%).  Sullivan  (1983)  studied  the  degradation  o f ditallowdimethylammonium chloride  ( D T D M A C ) in a "semi-batch" activated sludge system, using disulfine blue to measure loss o f Q A C and  1 4  C radiolabel to follow its fate. In three replicate experiments, samples o f D T D M A C  were radiolabeled: a) at the methyl position, b) at the C , position o f the C 29  1 8  chain (adjacent to the  nitrogen), and c) uniformly throughout the alkyl chain. The formation o f C 0 from each set o f 1 4  2  labelled Q A C was monitored. The D T D M A C rapidly and extensively disappeared (95%) from the solution, but the radioactivity remained on the sludge indicating adsorption had occurred. Subsequently, over the remaining 35 days, it was biodegraded as indicated by the production o f 1 4  C0 . 2  Between 20 and 50% o f the D T D M A C was mineralized by day 39, based upon the  production o f labelled carbon dioxide.  This observation confirmed that each o f the three major carbon positions in the molecule could be oxidized, and that ultimate biodegradation was possible for this Q A C .  Based upon  radio thin layer chromatography, Sullivan (1983) found the concentrations of.any. intermediates to be very l o w during the biodegradation, with no accumulation in the sludge. H e suggested that the intermediates have a short life and that the primary degradation process, during which the Q A C is removed from the solution, must therefore limit the rate o f degradation.  A  comprehensive  study  focusing  on  the  fate  of  a  single  QAC,  octadecyltrimethylammonium chloride ( O T A C ) in wastewater treatment was conducted by Games et al. (1982). They studied both the adsorption and biodegradation o f O T A C in a semicontinuous activated sludge ( S C A S ) system and in a respirometry test measuring C 0 evolution. 2  When 20 ppm O T A C was added to an inoculum o f raw wastewater, it was not degraded. In fact, the O T A C toxicity inhibited the endogenous C 0 production. However, when 20 ppm O T A C 2  was complexed with equimolar amounts o f an anionic surfactant ( L A S ) , 80% o f the theoretical C0  2  evolution was produced. This agrees with earlier work suggesting that anionic surfactants  can have a mitigating effect on the toxic (bacteriostatic) properties o f cationic surfactants when the two are complexed.  30  The S C A S system contained a large proportion o f biosolids and the O T A C was removed from solution by both adsorption and biodegradation. Adsorption proved to be the more rapid process as more than 98% o f the O T A C was removed from solution after 90 minutes. However, primary biodegradation was also quite rapid with a half-life o f about 2.5 hours. Results from tracer studies using C-labelled O T A C , showed that the distribution o f Q A C i n solution relative l4  to that bound to solids remained constant during biodegradation (i.e. in dynamic equilibrium). N o metabolites with appreciable half-lives were detected, so this study could not propose a specific degradation mechanism. However, experiments with O T A C radiolabeled at the methyl position and the alkyl carbon adjacent to the nitrogen revealed some interesting results.  A  significantly greater amount o f radioactivity was incorporated into the biomass from the alkyllabelled than the methyl-labelled O T A C .  Despite the interfering factors that may weaken experimental results, data from the literature as a whole aptly shows that a wide range o f Q A C s undergo biodegradation. The role o f sediments and anionic surfactants in enhancing the biodegradation is reinforced, as is the recalcitrance o f the various Q A C structures.  However, there is a shortage o f research  demonstrating the mechanistic pathways by which Q A C s are broken down.  2.3.4  Potential Pathways Quaternary ammonium compounds are limited in their structural complexity. Four units  are bonded to the central nitrogen. These may be any combination o f methyl, linear or branched alkyl, or aryl functional groups. pyridinium structure.  In addition, Q A C s may be formed by incorporating a  Due to its simple structure there are relatively few  degradative  mechanisms that can be proposed to occur during the biological breakdown o f Q A C s . One o f the mechanisms commonly advanced, suggests that the alkyl chains are subject to degradation by (331  oxidation.  To initiate this process, the organism will oxidize the end o f the alkyl chain (co-  oxidation) after which B-oxidation o f the attached alkyl group occurs. Alternatively, the alkyl chain could first be cleaved from the nitrogen to form a fatty aldehyde and a tertiary amine. The  aldehyde would be further oxidized to a fatty acid from which P-oxidation would proceed. The other functional groups (methyl, aryl) are also susceptible to this C - N fission. This reaction forms various aldehydes, and stepwise reduces the amine from quaternary ammonium, to tertiary, secondary, and primary amines and finally to an inorganic ammonium ion.  I f a benzene  substituent is present on the Q A C , cleavage and ring opening would have to occur for the complete degradation. A combination o f these various degradative mechanisms would be required for complete degradation o f the Q A C (Figure 2.1). A single species o f microbe may not have the metabolic capability to completely destroy a Q A C , but may instead degrade it to an intermediate that accumulates. The initial step o f the biodegradation pathway is therefore important in deterrnining the mechanistic route utilized by a particular organism.  Ring cleavage  CH  Oxidation of alkyl chain:  N / d  V  CH  a. terminal-oxidation b. P-oxidation  C-N fission 3  c  Figure 2.1. Potential biodegradation mechanisms for a Q A C  32  c. dealkylation d. demethylation e. dearylation  2.3.5 Biodegradation Mechanism 2.3.5.1 (3-oxidation In all cells, be they plant, animal, bacterial, or fungal, fatty acids are catabolised through (3-oxidation.  The process involves a number o f enzymes, which attach coenzyme A to the  carboxyl group o f the fatty acid. The alkyl chain is then sequentially broken down by two carbon units to yield acetyl groups which are used by the cells for energy or biosynthesis (Stryer, 1988). P-oxidation is almost certainly the method used to break down the alkyl chain o f Q A C s . What is in question is whether the alkyl chain is first cleaved from the central nitrogen with P-oxidation starting at the a-carbon; or whether the alkyl chain is first terminally oxidized (co-oxidation) into a carboxyl group with P-oxidation proceeding from the end o f the alkyl chain while still attached to the nitrogen. Dean-Raymond and Alexander (1977) were the only group to find strong evidence that Poxidation o f the alkyl chain was occurring while attached to the Q A C . They employed respirometry (oxygen depletion) to study the ability o f soil and sewage microbes to utilize Q A C s . O f the ten compounds examined, only the alkyltrimethyl ammonium salts (decyltrimethyl ammonium bromide ( D T M A B ) and hexadecyltrimethylammonium bromide ( H D T M A B ) ) were capable o f supporting growth. The other Q A C s tested were either dialkyl or benzylated Q A C s that are known to be more difficult to degrade. D T M A B at 100 u.g/ml was extensively, but not completely, degraded by the inoculum within 10 days. The concentration o f D T M A B (measured by a specific colourimetric method) dropped to undetectable limits over the incubation period. However, the oxygen uptake data indicated the degradation o f the compound was not complete. G C / M S was used to identify two prominent peaks in the gas chromatogram o f the culture extract. The metabolites were identified as 9-carboxynonyl- and 7-carboxyheptyltrimethylammonium compounds.  33  O f the organisms isolated, none could grow on decyltrimethylammonium by themselves. However, a mixture o f two pure strains, o f a Pseudomonas and a Xanthomonas species, were able to grow on the amended medium using the Q A C as a sole carbon source.  In nutrient  utilization tests, Xanthomonas cells (supplied with certain growth factors) could rapidly and completely oxidize D T M A B and trimethylamine.  Methyldecylamine and decanoic acid were  also readily degraded with no lag phase, but only about half o f the theoretical quantity o f oxygen was consumed. These results strongly suggest that the D T M A B was being degraded by initial oxidation o f the terminal alkyl chain, followed by B-oxidation, resulting in trimethylamine.  2.3.5.2 C - N de-methylation The main alternative to P-oxidation o f an alkyl chain as the first step in a Q A C degradative pathway is for the C - N bond to be cleaved.. This may involve demethylation, dealkylation, or dearylation depending on what types o f organic substituents are bound to the nitrogen.  Hampton and Zatman (1973) studied the metabolism o f tetramethylammonium chloride by a Gram-negative, non-motile bacterium (strain 5H2) isolated from soil. It was found to utilize tetramethylammonium chloride and trimethylamine as the sole carbon and energy Complete  biodegradation  of  tetramethylammonium  chloride  via  the  sources.  intermediates  of  trimethylamine, and trimethylamine-N-oxide to methylamine was proposed.  Ghisalba and Kiienzi (1983) isolated a number o f bacteria from the sludge o f an industrial sewage treatment plant which were capable o f utilizing trimethylethylammonium chloride ( T M E A C ) . The bacteria were characterized as Pseudomonas sp. and could also grow on a wide range o f Q A C and amine substrates as a sole carbon and nitrogen source, including methylamine, dimethylamine,  trimethylamine,  dimethylethylamine 34  and  tetramethylammonium  chloride.  Bacterial growth was measured by the optical density o f the liquid culture or by dry weight biomass. The growth rate on the amines was more rapid than on either o f the Q A C s examined, suggesting that the first stage o f Q A C degradation was rate limiting.  A  significant drop  in the  p H was  observed  in unbuffered  shake  cultures  of  trimethylethylammonium chloride (5 g/L) as the sole carbon source. This was thought to be due to the formation o f hydrochloric acid. It was found that the trimethylethylammonium chloride was biodegraded and utilized without the accumulation o f intermediate degradation products, but with the formation o f biomass and C 0 and the release o f N H C 1 and HC1. Ghisalba and Kiienzi 2  4  (1983) established the following stoichiometry for the biodegradation o f T M E A C (Equation 3.):  C  H  3 \  /CH CH 2  N CH / CH / 3  + \ r u C H  3  Pseudomonas biomaSS  C l " "  •  r C  u Mn 5 9 2 H  N  +  10 C 0  2  +  2 NH CI 4  + HCI  0  (3)  X  3  2.3.5.3 C - N de-alkylation Nishiyama et al. (1995) examined the biodegradation o f several alkyltrimethylammonium halides by an activated sludge.  They used two separate methods to measure biodegradation:  oxygen consumption respirometry, and dissolved organic carbon ( D O C ) .  Interestingly, the  degradation results from these two methods differed considerably. The D O C results suggested 100%  biodegradation o f all alkyltrimethylammonium salts tested after 28 days.  The 0  2  consumption results over the same period varied for each Q A C , but corresponded to significantly less  biodegradation,  ranging  from  7%  for  octadecyltrimethylammonium to  47%  for  octyltrimethylammonium. This disparity i n the results for the different methods is not likely due  35  to Q A C adsorption onto the suspended solids in the growth chamber, which would reduce the amount o f dissolved organic matter measured, and give erroneously high biodegradation values.  Nishiyama et al. (1995) used proton nuclear magnetic resonance ( ' H - N M R ) and ion chromatography  to  analyze  the  culture  for  identification  and  quantification  of  alkyltrimethylammonium salts and their degradation products. The formation o f trimethylamine was  observed  after  40  hours  incubation  with  tetradecyltrimethylammonium  bromide  ( T D T M A B ) . Subsequently the formation o f dimethylamine after 48 hours and o f methylamine after 76 hours o f incubation, was recorded. B y day five neither the original T D T M A B nor any o f the amine degradation products could be detected. The same pattern o f amine formation (all i n the form o f acid salts) was also observed for decyl and dodecyltrimethylammonium bromide. These observations clearly support the cleavage o f the long alkyl chain as the initiating step for alkyltrimethylammonium halides.  Using enriched cultures, Ginkel et al. (1992) screened microorganisms from domestic sewage sludge.  A bacterium (strain B l ) , capable o f utilizing hexadecyltrimethylammonium  chloride ( H D T M A C ) as a sole source o f carbon and energy, was isolated. Tentatively identified as a Pseudomonas species, strain B l grew on a l l o f the tested alkyltrimethylammonium compounds  (C  1 2  to  C ). 2 2  However,  none  of  the  alkylbenzyldimethyl-  and  dialkyldimethylammonium salts tested were able to support the growth o f the strain B 1 .  In nutrient utilization studies, the bacterium could grow on a wide range o f possible intermediates o f H D T M A C degradation, such as hexadecanal, hexadecanoate, (Ginkel et al,  1992).  However, none  o f the  and acetate  amines tested (hexadecyldimethylamine,  hexadecylamine, trimethylamine and methylamine) could serve as growth substrates for this bacterium, regardless o f whether strain B l was previously acclimated on H D T M A C or not. Gas 36  chromatography was used to measure trimethylamine concentration over time, when a liquid culture o f Pseudomonas  was grown with H D T M A C .  The result was the growth-associated  stoichiometric formation o f trimethylamine from the H D T M A C .  3 C H  3  ( C H  2  )  1  4  C H  _  2  N  + —  CrH NADH + H ^ +  NAD  +  3  3  .  ^  C H  Oo j ^ H  2  (1) 0  O C H  ( C H  3  2  )  1  4  N(CH )  C  3  H  NAD  another bacterium  H 0  +  2  NADH + H ^  (2)  i  +  NH  3  ( C H  2  )  1  4  +  C0  2  +  H 0 2  biomass  C ^  OH *  3  +  O C H  3  Enzymes:  P-oxidation  (1) M o n o o x y g e n a s e (2) A l k a n a l d e h y d r o g e n a s e  C0  2  Figure  + 2.2.  H 0 2  +  biomass  Pathway proposed by Ginkel et al. (1992) for the biodegradation o f hexadecyltrimethylammonium chloride by a strain ( B l ) o f a Pseudomonas species  37  The inability o f strain B l to utilize trimethylamine suggested that the bacterium was cleaving the alkyl chain from the central nitrogen o f the H D T M A C .  Ginkel et al. (1992)  proposed the pathway o f initial bacterial cleavage o f the alkyl chain to form trimethylamine and hexadecanal. The decanal would be further oxidized to hexadecanoate, followed by complete degradation of the alkyl chain through P-oxidation (Figure 2.2). Assays with cell extracts o f strain B l have revealed the presence o f necessary enzymes for the bacterium to metabolize the alkyl chain by way o f P-oxidation. The degradation o f trimethylamine would require another microorganism.  Ginkel and his research team examined the biodegradation o f a number o f fatty tertiary amines following the same experimental approach used to study H D T M A C degradation (Ginkel et al, 1993; Ginkel et al, 1995; Ginkel and Kroon, 1993; and K r o o n et al, 1994). For each o f these amines they proposed the same mechanism (shown in Figure 2.2). A l l o f the pure bacterial cultures they isolated from enrichment o f sewage sludge first cleaved the alkyl chain from the nitrogen, and then degraded it leaving the resulting secondary amine to accumulate.  2.3.5.4 C-N de-arylation Krzeminski  et  al.  (1973)  examined  the  biodegradation  of  alkyl(Ci -  C| )benzyldimethylammonium chloride (Hyamine 3500) by monitoring the generation o f 6  from the  1 4  C label in the methylene group o f the benzyl substituent.  than 90% o f the  1 4  2  1 4  C0  2  They reported that more  C appeared as carbon dioxide with less than 10% being found in the effluent.  Using a continuous activated sludge system, Fenger et al. (1973) were able to reduce the amount o f tetradecylbenzyldimethylammonium chloride detected colourimetrically by up to 85%.  Through  a  combination  of  infrared  38  spectroscopy  and  gas  chromatography,  tetradecyldimethylamine,  benzoic acid,  and  acetic acid  were  identified as  metabolites.  Intermediate degradation products from further degradation (e.g. dimethylamine) were not detected.  This suggests that cleavage o f the aromatic group from the central nitrogen occurs  preferentially to de-methylation or de-alkylation as the first step in the degradation o f benzylated QACs.  Tolerant bacteria, isolated from the sludge conditioned with tetradecylbenzyldimethyl-  ammonium chloride, were almost exclusively Pseudomonas and Comamonas species.  2.4  Analytical Techniques Used  This section briefly reviews the basic priciples behind the three analytical chemistry techniques that were used to isolate and identify the metabolite from D D A C biotransformation.  2.4.1  HPLC H i g h performance liquid chromatography ( H P L C ) was used extensively in this research  to quantify D D A C loss, to discover the metabolite P I , and to isolate P I from the D D A C in the extract. H P L C also contributed to a degree in the identification o f P I .  Chromatography is a means o f separating a mixture o f substances into its substituent compounds.  L i q u i d chromatographic separations are achieved in a column o f closely packed  particles (referred to as the solid phase). These tiny particles are usually made o f silica, and are often coated with some sort o f surface chemistry, such as non-polar C groups.  l 8  chains or polar N H  2  Sample compounds are added and washed (eluted) though the column with a steady,  continuous flow o f eluent (referred to as the mobile phase).  The compounds i n the sample  require various volumes o f the mobile phase to pass through the column before they elute. Elution depends on the affinities o f a sample compound to dissolve in the mobile phase and to bind to the solid phase. The affinity o f a compound for a solvent or a solid-phase chemistry is a 39  specific property o f the compound that depends on its molecular structure, size, functional groups, and ionic moieties. When the flow o f eluent is constant, the compounds w i l l emerge from the end o f the column at distinct retention times (Figure 2.3).  (a)  Sample  Mobile phase  T  * f  IIP  •ma B A.  •  Packed column  ML  •  I'.  § i ?  mm*  M  if  m  5  -Detector  to  g M  a  u  Figure 2.3. Theory o f separations by liquid chromatography (Skoog and Leary, 1992). The retention times are the key to identification. Under uniform running conditions, with the same column, mobile phase, pressure, and flow rate, a specific compound w i l l elute at a characteristic, and reproducible retention time. If two unknown compounds are believed to be the same compound and have both the same retention times under two different H P L C running conditions, they are likely identical compounds.  The molecular size and polarity o f the sample compounds are the properties upon which liquid chromatographic separations are primarily based.  A s D D A C and other similar aliphatic  Q A C s are highly polar, ionic compounds o f relatively l o w molecular weight (<10 ), ion 3  exchange chromatography was used for separation. 40  Therefore, a strong cation exchange resin  was used as the solid phase in conjunction with a moderately polar liquid phase.  Under this  system, the affinity o f a compound for the solid phase is influenced by the charge and size o f the sample molecule.  The greater the charge, the greater the attraction for binding the cation  exchange sites. The smaller the molecular size, the greater the attraction for binding the cation exchange sites due to less steric interference and the easier for the charged molecule to interact with the exchange site.  In an ion exchange column, the retention time for the Q A C s depends on the strength and concentration o f the counter ion, the strength o f the sample ion being injected into the column, and the polarity o f the mobile phase.  The separation o f a compound is based upon structural  features such as its overall size and shape, its polarity, and polar functional groups or ionic moieties. Aliphatic Q A C s elute in order o f decreasing size as smaller Q A C s have less steric hindrance and can interact more easily with the solid phase exchange sites.  2.4.2 NMR Nuclear magnetic resonance ( N M R ) is based on the principle that nuclei are spinning, charged particles capable o f establishing a small magnetic field.  The magnetic properties o f  certain nuclei (e.g. ' H , C , and P ) are such that when placed i n an external magnetic field, their 1 3  3 !  own magnetic fields w i l l align with or against the applied field. A t any given magnetic field strength, the energy difference between oppositely aligned nuclei, corresponds to a discrete frequency o f electromagnetic radiation that w i l l be absorbed by these nuclei to produce a signal.  A one-dimensional N M R spectrum is a plot o f the frequencies at which absorbance occurs versus the intensity o f the absorbance. N M R absorptions occur at different frequencies for different species of nuclei ( ' H , C ) , and for nuclei o f the same species in different molecular 13  41  environments.  The  energy  difference  between  oppositely  aligned  protons  increases  proportionally as a function o f the applied magnetic field strength. For a 200 M H z instrument, proton peaks occur over a range o f about 2000 H z , while on a 400 M H z instrument, this range becomes 4000 H z . Therefore the proton peaks are plotted in units o f parts per million (ppm) as the difference between their frequency and the frequency o f a reference compound, commonly tetramethylsilane ( T M S ) , divided by the operating frequency o f the N M R instrument.  For  example, a proton with an absorption difference o f 200 H z on a 200 M H z instrument w i l l appear at a chemical shift o f 1 ppm.  Proton N M R ( ' H - N M R ) is a powerful analytical tool in the identification o f organic compounds. The chemical shift o f a proton is heavily influenced by its local environment, which provides information about the presence o f structural elements.  In addition, the signals o f  adjacent protons interact and can indirectly provide some indication o f the carbon skeleton o f a compound. Several pieces o f important information are revealed by a ' H - N M R spectrum which help elucidate the chemical structure o f an unknown compound:  1. The number and type o f equivalent protons. Due to differences in the electron density that surrounds them, protons within a molecule exist in different chemical environments.  Electrons, like nuclei, are also spinning, charged  particles that produce a small magnetic field o f their own that counteracts an applied field. A s a result, protons surrounded by high electron density are said to be "shielded". The more shielded a proton is, the smaller the magnetic field that it "senses", and the lower the frequency o f radiation required to produce a signal. Within a molecule, functional groups that are strongly electron withdrawing such as carboxyl, hydroxyl, and ammonium groups play a role in removing electron density from around  42  a proton "deshielding" it.  Deshielding results in a higher energy difference in the proton  requiring an increased frequency to produce a signal. The strength o f the functional group and its proximity to the affected protons determines the degree to which the protons w i l l be deshielded and consequently shifted.  A s a result, each signal in an ' H - N M R spectrum corresponds to a group o f equivalent protons.  In addition, the degree to which a signal is shifted along the frequency axis gives  information about the types o f protons responsible for the peak, whether methyl ( - C H ) , 3  methylene (-CH -), or methine (-CH-), and the functional groups near to these protons. 2  2. The relative number o f protons between each peak. The vertical axis o f an N M R spectrum has no units.  The areas under each peak are  proportional to the number o f protons, which give rise to the peak. If the proton number o f one peak can be assigned, then the absolute number o f all the integration can be calculated. For example, a ratio o f methyl ( C H - ) to methylene (-CH -) protons o f 3:16 indicates that i f the actual 3  2  numbers of methyl protons are 6 then the methylene protons w i l l be 32. This is also helpful in situations where there is an overlap o f many similar groups, and not each signal is separated. 3. Adjacent protons. Proton signals are subject to the influence o f protons in different chemical environments that are located on adjacent carbon atoms. The result is a splitting o f the peaks. The multiplicity of the resulting split peaks provides further information about the structure o f a compound. Peaks w i l l be split into a number o f sub-peaks one greater than the number o f non-equivalent protons on the adjacent carbon atom (multiplicity = n + 1; where n = the number o f nonequivalent protons on the adjacent carbon). For example, the three protons in a methyl group are equivalent to each other as they are in exactly the same chemical environment. However, these  43  methyl protons w i l l not be equivalent to a pair o f methylene protons. If the methylene protons are attached to the adjacent carbon atom to the methyl group the methyl protons w i l l form a signal that w i l l be split by the methylene protons into (2)+l = 3 subpeaks or a triplet.  The information gleaned from an ' H - N M R spectrum is best illustrated with an example. In a compound such as ethanol ( C H C H O H ) there are three sets o f non-equivalent protons as 3  2  indicated by the presence o f three peaks (Figure 2.4). The peaks are shifted downfield according to how much they are deshielded. The methyl protons are the most shielded and the furthest upfield, followed by the methylene protons and the hydroxyl proton. The O H functional group plays a strong role in the shift o f C H , which would otherwise be closer to C H . Tables o f N M R 2  3  chemical shifts for protons in different environments have been published. Integration o f the peaks reveals a 3:2:1 ratio o f protons. The methyl peak is split into a triplet by the presence o f the two adjacent methylene protons and the methylene peak is correspondingly split into a quartet by the three methyl protons. 'H=3 'H=2  /  ^  'H=l  ^  Downfield  Figure 2.4. Theoretical ' H - N M R spectrum o f ethanol (adapted from Bruice, 1995).  2.4.3 FTIR Covalently bonded atoms can be thought o f as marbles attached by a spring. The spring allows the marbles to stretch and bend in several planes and the frequency of these vibrations is governed by both the properties o f the spring and the mass o f the marbles. Because the atoms i n 44  a molecule have a large variety o f masses and the covalent bonds connecting them have different strengths, there is a wide range o f vibration frequencies present for any compound. Each o f these vibrations corresponds to a specific frequency o f infrared radiation that can be absorbed by the bond to produce a signal on the I R spectrum. A s a result the I R spectrum o f a compound yields a wealth o f information and is often very complex.: . - „'  The intensity o f IR absorption bands varies from weak to moderate to strong.  The  intensity o f absorption is proportional to the strength o f the dipole moment. For example, the C H bond has a weak dipole and has weak absorption, while a carbon-oxygen double bond has a strong dipole moment and has strong absorption.  Furthermore, the bond absorption's are  additive, so that a large number o f weak C - H bonds in a molecule w i l l produce a large absorption peak.  Infrared spectra are usually plotted as percent transmittance (ordinate) over a frequency range (abscissa). For IR spectra, frequencies are usually reported in wavenumbers (cm" ) rather 1  than in Hertz (s ). The frequency range o f interest lies between 1.2 x 10 H z and 1.2 x 10' H z , 1  14  3  that is more conveniently expressed as 4000 cm" to 400 cm" . I R spectra can roughly be split 1  into two regions.  1  The functional group region (4000 cm" to 1000 cm" ) generally contains 1  1  absorptions due to bond stretches as they require more energy and thus a higher frequency. The fingerprint region (1000 cm" to 400 cm" ) contains many complex signals generally resulting 1  1  from bond bending that are unique for each compound.  For the scope o f this thesis, we are  interested primarily in the functional groups from the I R spectra, as a complement to the N M R spectra.  45  CHAPTER 3 EVALUATION  3.1  OF  BIODEGRADATION  Introduction The primary goal o f this overall research project was to determine the mechanism by  which moulds degrade D D A C through the identification o f metabolic by-products.  Before  metabolites could be identified, it was first necessary to generate and isolate fungal metabolites from D D A C degradation. Therefore, experiments using liquid culture and wood substrate were good starting points. It was important to get a sense o f the fungal interaction with Q A C s in different substrates, and to find the experimental conditions most appropriate for the generation and recovery o f metabolites. Useful information could also be generated by determining i f there were differences in the way fungi degrade Q A C s in aqueous solution, compared to when they were bound or fixed onto a solid substrate such as wood. Some fungi may be better adapted to tolerate and degrade D D A C in solid, versus liquid, substrates.  3.2  Methodology  Chemicals: studies.  Didecyldimethylammonium chloride ( D D A C ) was the sole Q A C used in these  It was obtained as Bardac 2280, a commercially available formulation (80% D D A C ,  10% ethanol, 10% water) supplied by Lonza Inc. Acetonitrile (Fisher, H P L C grade) was used for the extraction o f D D A C and as the main component o f the H P L C mobile phase. Formic acid (Fisher) was  used  to  lower the  p H o f the  acetonitrile used  for  DDAC  extractions.  Benzyltrimethylammonium chloride ( B T M A C ) , was obtained from A l d r i c h , and used in the H P L C mobile phase.  46  Fungi:  Four fungi were employed for liquid culture experiments involving D D A C -  degradation.  Verticillium bulbillosum (TC-NP-13) and Gliocladium roseum Banier complex,  two hyphomycete species, were isolated by Doyle (1995) from DDAC-treated lumber. Fusarium avenaceum (S-l-1), another hyphomycete, and Zygorrhyncus moelleri (S-l-6), a zygomycete, were isolated by Zheng (1995) from DDAC-treated wood stakes at the U B C field site. A l l o f these fungi were identified by the National Identification Service, Agriculture Canada, Ottawa. For experiments involving the degradation o f D D A C from treated wood, only V. bulbillosum and G. roseum were used.  E x t r a c t i o n of D D A C :  Attempts to extract D D A C from the liquid portion o f the culture  recovered only a very small fraction.  Based on this result and the abundant evidence o f the  strong affinity o f D D A C to adsorb to solid surfaces (Lawrence, 1970; Swisher, 1987; Mackrell and Walker, 1978), the assumption was made that most o f the D D A C was bound to the solids (silica gel and fungal biomass). Zheng (1995) demonstrated that greater than 90% o f the D D A C added to Vogel's medium with silica gel, was bound to the silica gel. In addition, tolerance studies revealed that V. bulbillosum was unable to grow at 50 ppm D D A C , when no adsorbing substrate (e.g. silica gel) was present. Therefore, in growth flasks containing silica gel and 300 ppm  D D A C , presumably enough D D A C  was bound to the silica to leave the  effective  concentration in the liquid below 50 ppm.  The culture solids (silica gel and biomass) were first oven dried and pulverized into a powder with mortar and pestle. The powder was added to a centrifuge tube together with 10 m l of acetonitrile (pH adjusted to 3.0), and the tube was shaken vigorously on a vortex machine for 1 minute. It was then placed in an ultrasonic bath (40 H z ) for one hour. After centrifuging at 3500 rpm for 15 min, the supernatant in the tube was removed and 10 ml o f fresh extracting  47  solvent was added.  This extraction procedure was repeated three times and the supernatants  collected. Using a syringe, the extract was filtered through a 0.45 urn P T F E filter into a vial, and analyzed for D D A C by H P L C .  The harvested sawdust was weighed out into 0.3 g amounts and placed in 10 m l centrifuge tubes, to which acetonitrile (adjusted to p H 3.0 with formic acid) was added.  The  tubes were placed i n an ultrasonic bath (40 H z ) for two hours, then centrifuged at 3500 rpm for 15 min. The supernatant was carefully removed by syringe, filtered through a 0.45 u m P T F E filter and analyzed by H P L C .  H P L C system:  The H P L C system consisted o f a Spectra Physics 1200 series pump, which  carried mobile phase through the column to. a variable wavelength U V detector Physics).  (Spectra  Stainless steel and polyetherether keytone ( P E E K ) tubing was used to connect the  components. A strong cation exchange Partisil S C X (Waters/Millipore) column, with a length o f 10 cm and a diameter o f 8 mm, was used to separate the Q A C s .  The solid phase packing  consisted o f irregularly shaped, 10 um particles o f porous silica, to which a benzene sulfonic group was bound.  The mobile phase consisted o f acetonitrile and ultrapure water (70:30) acidified with 1% acetic acid.  In addition, 0.8 g/L o f B T M A C (Aldrich) was added as a counter ion and  chromophore to provide a UV-absorbing background.  The mobile phase was prepared in one  litre batches and was both filtered through a 0.45 u m nylon filter (Millipore) and stirred under vacuum to remove dissolved gases prior to running through the H P L C .  The degassed mobile  phase was carried through the column at a flow rate o f 1.5 ml/min, and required about one hour to equilibrate and form a stable baseline.  Samples were first dissolved in acetonitrile, then  injected using an H P L C syringe (Hamilton). To prevent problems due to air bubbles, it was 48  necessary to overfill the syringe, tap out the air bubbles and overload the 20 u l injection loop with 40-50 u l o f sample.  A n indirect photometric method was used to detect and quantify D D A C based on the method o f Larson and Pfeiffer (1983).  The U V detector was set at 262 nm, as this was the  wavelength o f greatest absorption for B T M A C . The absorption at this wavelength was recorded over time on the Spectra physics integrator.  A continuous UV-absorbing background was  established by the B T M A C in the mobile phase. D D A C and other n o n - U V absorbing aliphatic Q A C s appeared as negative peaks against this UV-absorbing background.  Identification was revealed though the retention time o f the compound, and the concentration was proportional to the height o f the peak. Peak height was chosen over peak area as a means o f determining the concentration as for the most part the peaks were sharp and uniform and the peak heights provided more manageable numbers than the areas. D D A C was quantified by comparing the peak heights on the chromatogram with the standard curve from known D D A C concentrations.  3.2.1 Liquid Culture Experiments Medium Composition:  The liquid media experiments were carried out in V o g e l ' s liquid  medium. This medium was chosen based on previous research into the fungal interaction with Q A C s in liquid media (Zheng, 1995; Burgel et al. 1996a). V o g e l ' s medium was prepared by dissolving (% w/w): K H P 0 2  (0.05%), C a C l  2  4  (0.1%), N a H P 0 2  4  (0.05%), M g S 0  . 2 H 0 (0.04%), potassium-hydrogen-phthalate 2  4  . 7 H 0 (0.05%), N H N 0 2  4  3  (0.3%), soluble starch (2%),  micronutrients (Vogel, 1956), and vitamins (Montenecourt, 1977), in distilled water. The initial p H o f the medium was adjusted to 6.0 using N a O H . The medium was allocated into 125-ml or  49  250-ml erlenmeyer flasks, and sterilized by autoclaving at 121°C for 40 minutes. thermal sensitivity o f the vitamins, they were not autoclaved. Instead they were  Due to the  filter-sterilized  and added to the autoclaved medium once it was cooled to room temperature.  G r o w t h C o n d i t i o n s : The method o f inoculation varied between experiments in an attempt to find a balance between reproducibility and ease.  Flasks were inoculated by one o f three  methods:  1. Spore suspensions were prepared by adding sterile, distilled water (-10 ml) to an agar plate containing a mature fungal colony. Spores and some hyphae were dislodged from the plate by carefully massaging the colony with a blunt, sterile glass rod.  A n aliquot (1 ml) o f the  suspension was pipetted into the flasks v i a sterile pipette, to provide a quick and simple inoculum that was variable between batches.  2. Agar plugs, were cut from the growing edge o f fungal colonies with a 3-mm borer and placed in flasks containing the medium. This was another simple method; however, there was variation between experiments due to differences in the age o f the colonies.  3. Mycelial suspensions o f known biomass were the most labour intensive o f inocula to prepare, but also the most accurate in terms o f consistent characterization. Fungal cultures were grown up in Vogel's liquid medium for three days, then poured into sterile tubes (50 ml) and centrifuged at 3500 rpm for 15 min. The supernatant was then discarded and the pellet was washed with sterile, distilled water and re-centrifuged. A small amount (10 ml) o f sterile distilled water was used to resuspend the clean, fungal pellet and a 1 m l aliquot o f the suspension was placed on a glass micro fibre filter for oven drying at 103°C for 2 hours. Once the concentration o f the suspension was determined, it was diluted to the appropriate concentration with sterile, distilled water and used to inoculate experimental flasks with a known amount o f biomass. 50  In early experiments, potassium-hydrogen phthalate (0.3%) was used to buffer V o g e l ' s ( K H P ) medium, while later acetic acid (0.3%) was used to buffer the V o g e l ' s ( A ) medium. The liquid culture (50 ml) was added to Erlenmeyer flasks (125 m l or 250 ml), which acted as discrete growth chambers.  The flasks were incubated in an environmental incubator/shaker  (New Brunswick G24) at 25°C and shaken at 200 rpm.  Biomass:  The fungal biomass was determined directly by filtering the incubated culture  through pre-weighed glass microfibre filters (0.5 u m pore size). The culture was centrifuged at 3500 rpm for 15 minutes prior to filtration, to aid the process. The culture solids were then ovendried at 103°C to a constant weight. The biomass was determined by the difference o f the total dry weight from the weight o f the filter and silica gel.  3.2.1.1 Media Considerations Autoclave-induced Precipitation When all o f the liquid medium components were added together and autoclaved, the medium became cloudy and a coarse, white precipitate settled to the flask bottom. It had been hypothesized by Zheng (1995) that the precipitation was due to the insolubility o f the starch, which prior to autoclaving formed a dilute slurry.  However, when a 2% starch slurry was  prepared with no other media components, a clear solution appeared after autoclaving. Tests in which selected components were omitted from the media enabled the cause o f the precipitation during autoclaving to be determined (Table 3.1). N o precipitation occurred in any test where calcium chloride (CaCl ) was absent, even with all other media components present. A slight 2  precipitation developed in flasks with calcium chloride and starch (test #3) and calcium chloride and glucose (test #4). The greatest degree o f suspended solids occurred where the two phosphate  51  salts and calcium chloride were combined and autoclaved (test #5).  It was reasoned that the  elevated temperature and pressure o f the autoclave allowed the calcium and phosphate to react to form an insoluble calcium phosphate. To overcome this problem, a concentrated stock solution of C a C l » 2 H 0 was prepared, and autoclaved separately from the media. 2  2  The appropriate  amount o f this stock solution was added to the flask containing the medium once it had cooled, and no precipitate formed.  Table 3.1. Tests to determine the cause o f murkiness i n V o g e l ' s medium after autoclaving  Test # 1 all salts starch no CaCI  2 all salts glucose no CaCI  clear  clear  2  2  3  4  ~  —  •  •  5 Na,HP0 KH P0 CaCI  6 4  starch glucose = CaCI CaCI Results: cloudy cloudy particulate 2  2  2  4  2  —  starch —  clear  DDAC precipitation After completion o f a number o f experiments, it was found that V o g e l ' s medium could also react with D D A C to form a suspended white precipitate. This reaction was not evident in liquid cultures containing some mycelial biomass prior to the addition o f D D A C , or in liquid medium where silica gel was added.  It was hypothesized that the majority o f D D A C became  bound to the silica gel or the culture solids leaving only a very small amount o f D D A C in solution, to react with the dissolved media components.  A s with the autoclave-induced  precipitation, simple tests were conducted to determine which component o f the medium was reacting with the D D A C . phthalate ( K H P ) buffer.  It was concluded that D D A C reacted with the potassium-hydrogen Several buffers were tested to replace the K H P i n the liquid medium  (Table 3.2).  52  A s the component responsible for the precipitation was the buffering agent, it was hypothesized that precipitation could be a solubility problem related to p H . Therefore the K H P and potential replacement buffers were acidified with HC1 to p H ' s ranging from 3.0 to 7.0. Phosphate buffer was discarded as it aggravated the calcium phosphate precipitation. K H P precipitated with D D A C over the range o f p H 3.0 to 7.0, and citrate buffer (0.3% sodium citrate)  Table 3.2. Testing o f potential liquid medium buffers for precipitation reactions with D D A C . Note: + = noticeable murkiness (indicating precipitation) upon the addition o f D D A C . - = clear solution (indicating no reaction). K H P = potassium-hydrogenphthalate; Citrate = sodium citrate; Acetate = acetic acid.  Buffer 30  35  4.0  4.5  PH 50  5.5  60  65  7.0  KHP Citrate  +  -  + +  + +  + +  +' .+  + +  + +  + +  + +  Acetate  -  -  -  -  -  -  -  -  -  reacted with D D A C over the same range o f p H except at p H 3.0.  The acetate buffer (0.3%  sodium acetate) did not react with D D A C to form a visible precipitate at any o f the p H ' s tested. Therefore it was used in later experiments to replace the potassium-hydrogen phthalate buffer in the V o g e l ' s liquid medium.  3.2.1.2 Specific L i q u i d C u l t u r e Experiments Biomass.  To observe the effect that D D A C had on the normal growth and functioning o f the  fungal colonies without D D A C present, reference growth curves for V. bulbillosum and G. roseum were obtained in V o g e l ' s ( K H P ) medium, in 125 ml Erlenmyer flasks. The flasks were harvested at one to two day intervals and their biomass was measured.  Tolerance 1. The tolerances o f four fungal species to D D A C i n liquid culture were examined. V. bulbillosum, G. roseum, F. avenaceum, and Z. moelleri were inoculated by spore suspension  53  and grown in V o g e l ' s ( K H P ) medium in 125 m l flasks. After one day o f growth for the fungi to adjust to the medium, D D A C was added to the flasks to give a final concentration o f 50, 100, 200, and 250 ppm. The cultures were grown for 14 days, after which they were harvested and the biomass in each flask was measured.  Tolerance 2. This experiment examined the DDAC-tolerance o f V. bulbillosum and G. roseum when the liquid culture was supplemented with silica gel (1.00 g/ flask). The setup was identical to the first tolerance experiment, except for the presence o f silica gel in the flasks and the greater range o f D D A C concentrations: 0, 100, 250, 500, 750, and 1000 ppm.  Tolerance 3. When it was discovered.that D D A C was reacting with K H P o f the liquid medium, the tolerance o f V. bulbillosum to D D A C was examined in the acetate buffered medium. V. bulbillosum was inoculated with agar plugs into V o g e l ' s ( A ) medium in 125 m l flasks. The culture was acclimated for three days prior to addition o f D D A C to bring the flask concentrations to 0, 50, 100, 200, and 300 ppm. The flasks were harvested and the biomass measured after 10 days.  For comparison, the tolerance o f V. bulbillosum was examined i n the same medium to which silica gel (1.00 g/flask) was added. The culture was grown for three days prior to addition of D D A C to make up flask concentrations o f 0, 50, 100, 200, and 300 ppm. These cultures were harvested after 16 days growth.  Biodegradation 1.  Due to its high tolerance to D D A C under the liquid culture conditions  used, V. bulbillosum was chosen for the experiments to determine the biodegradation o f D D A C . Mycelial suspensions (1.0-1.5 mg biomass) were used to inoculate 50 m l o f V o g e l ' s ( A ) medium in 250 m l Erlenmyer flasks. Each flask was supplemented with 1.00 g o f oven-dried silica gel to reduce the impact o f the D D A C on the fungus. The cultures were acclimated to the medium for 54  2 days, following which, 12.5 mg o f D D A C was added to each flask to give a final D D A C concentration o f 250 ppm in each flask. The flasks were harvested at one and two day intervals over a period o f 16 days. From each flask, both the biomass and amount o f D D A C remaining were measured.  Three control flasks were prepared. For the first control, a flask at day 2 was harvested within an hour o f D D A C addition to allow the efficiency o f the extraction to be determined. A second control flask was uninoculated, but had D D A C added. It was left until day 16 to check on the extraction efficiency, and to confirm that the D D A C was not being degraded abiotically. A third control flask, set up in the same manner as the others, but receiving no D D A C , was used to compare the effect o f the D D A C on fungal growth.  Biodegradation 2.  This experiment was a replica o f the previous experiment (Biodegradation  1) and differed only slightly. The fungi were acclimated for three days instead o f two, and two flasks were used to generate each data point, instead o f only one as previously reported.  3.2.2 Wood Substrate Experiments Wood:  Southern yellow pine sapwood radial sections (2 m m x 10 m m x 40 mm) were  prepared with alternating bands o f early wood and latewood parallel to the 10 m m face.  Some  sections were ground into 20 mesh sawdust i n a Wiley m i l l . Treatment: The wood and sawdust samples to be vacuum impregnated with D D A C were placed in a beaker and covered with a wire mesh screen that was weighted down with glass stoppers. The beaker was then placed in a glass dessicator that was evacuated for 30 minutes. A t the end o f this period, the dessicator was disconnected from the pump and attached to a hose which drew the D D A C treating solution into the beaker as a result o f the vacuum. The wood (or sawdust) was allowed to soak in the treating solution for 30 minutes. Following this soaking 55  period, the wood pieces were removed from the beaker and patted dry with a paper towel. The treated sawdust was placed in a Buchner funnel and the excess solution was removed by vacuum filtration.  The treated wood (or sawdust) was left to air dry overnight. It was then sealed in small, plastic bags and sterilized by y-radiation (2.5 Mrad) at the U B C Biomedical Research Centre. Radiation was chosen over autoclaving for sterilization, since there was concern that autoclaving may influence the degradation o f D D A C by removing or altering the wood extractives.  G r o w t h Conditions: A solution o f 2 % malt agar (Difco) was poured into disposable petri dishes (100 m m x 15 mm standard, Fisher). It was necessary to prevent the wood pieces and sawdust from contacting the agar surface as moisture would be drawn into the wood from the agar causing water-logging.  Fungi do not grow well in water-saturated wood due to the limited  oxygen content. A sterile piece o f rubber mesh was placed on top o f the agar in each plate to support the wood pieces above the agar. To prevent contact between the agar and the sawdust, circular sheets (8 cm diameter) o f cellophane (Bio-Rad) were cut and placed over the agar, and the sawdust was placed on top o f the cellophane. The cellophane acted to prevent the sawdust from absorbing excessive moisture from the agar, while still allowing the fungus to grow.  Agar cores (3 mm) taken from the edge o f a growing colony o f G, roseum were placed on top o f the cellophane to inoculate the plates. After the fungal colony had covered the plate, 1.00 g o f sterile oven-dried sawdust was carefully spread over the plate surface. The solid wood was placed on the mesh and inoculated by spore suspension (1 ml) that was added dropwise directly onto the wood and agar.  A l l plates were sealed with parafilm to prevent moisture loss, and  incubated at 25°C.  56  3.2.2.1 Specific W o o d Substrate Experiments Wood:  This experiment was designed to examine the biodegradation o f D D A C within  treated wood by two moulds, G. roseum and V. bulbillosum.  The wood pieces were treated with  an aqueous 2 % D D A C treating solution, and placed on the mesh-covered agar in petri plates. Four wood pieces were placed within each petri dish. The plates were grouped so that 10 plates were inoculated with each fungus, and five plates were left as uninoculated controls. The wood pieces were harvested after 17 weeks incubation, dried, ground into sawdust, extracted with 3.0 ml acetonitrile, and analysed for D D A C as mentioned previously (Sect. 3.2). The uninoculated controls were present to confirm that any observed degradation o f D D A C was the result o f biological activity. It was also important to determine the efficiency o f the D D A C extraction technique.  This was done by comparing the amount recovered after a  period o f time with the initial retention determined from the procedure. The oven-dried weight of each test sample prior to treatment and the corresponding wet weight shortly after treatment were recorded.  This weight uptake, together with the concentration o f a known amount o f  treating solution measured before and after treatment, enabled the D D A C retention within the wood pieces to be calculated.  Sawdust:  Three sawdust types were identified based on their method o f treatment and the  resulting D D A C retentions: Sawdust I ( S W D I) was treated with a 0.5% aqueous D D A C solution.  Sawdust II ( S W D II) was treated with a 1.0% aqueous D D A C solution. Sawdust III ( S W D III) was formed by treating wood pieces with a 1.0% aqueous solution. The wood pieces were then dried and ground into sawdust.  57  DDAC  Small samples (0.6 ml) o f the D D A C treating solution were taken before and after the sawdust treatment. The amount o f solution taken up by the different sawdust treatments was determined by the difference in the wet and dry weights o f each sawdust.  The D D A C  concentration o f the treating solution prior to, and after treatment, along with the amount o f solution taken up by the sawdust gave an approximation o f the D D A C retained i n each sawdust batch. For S W D III, wood pieces were treated in the same manner as the other two sawdust treatments.  The weight for each wood piece was recorded before and after treatment and the  retention was based upon the assumption that the wood pieces take up D D A C and water in roughly the same proportion.  This assumption was verified by measuring the  DDAC  concentration o f the treatment solution used to treat the wood pieces before and after treatment and finding that the concentration remained the same (1.3%) after treatment.  The sawdust was harvested weekly for the first five weeks, after which it was harvested every two weeks.  Triplicate plates o f each sawdust type were recovered at each harvest to  estimate the variation o f D D A C biodegradation within the different sawdust treatments. During harvesting, the sawdust from an individual plate was gently scraped off the cellophane and oven dried at 103°C. D D A C was extracted from the sawdust using 6.0 m l o f acetonitrile and determined by the H P L C method mentioned previously (Sect. 3.2).  Controls:  Control A was identical to the sample plates, except that the sawdust contained no  D D A C . This control allowed the potential for interference o f regular fungal metabolites during D D A C analysis to be identified. plates.  Control B contained DDAC-treated sawdust in uninoculated  This provided a reference for the biodegradation o f D D A C from each o f the sawdust  types. In addition, the efficiency o f the extraction procedure could be determined.  58  3.3  Results and Discussion  3.3.1 Liquid Culture The primary purpose o f these experiments was to develop a system in liquid culture that would allow V. bulbillosum and G. roseum to produce metabolites from D D A C biotransformation under controlled conditions. Due to limitations in equipment, single replicates were often used for the liquid culture experiments.  However, the single replicate experiments were often run several  times, and the overall trends were reproducible. Biomass.  The biomass curves for both V. bulbillosum and G. roseum within V o g e l ' s ( K H P )  liquid medium to which no D D A C was added are shown in Figure 3.1. It was apparent from visual observation that V. bulbillosum was capable o f producing more biomass i n a shorter time than G. roseum under these growth conditions.  V. bulbillosum  grew as thick white mycelial,  spheres that filled up the flask, making the culture opaque, while the growth o f G. roseum culture was relatively sparse. In addition, the G. roseum flasks acquired a bright yellowish colour due to pigment production. V. bulbillosum attained a maximum biomass o f 7.8 mg/ml using starch as a carbon source by day four, which was double that produced by G. roseum after 12 days. The growth curve for V. bulbillosum  showed the classic shape for microbial growth. A  lag phase occurred from inoculation until day two, during which time the fungus adjusted to the medium, synthesizing the appropriate enzymes to make use o f media components.  After the  second day, biomass was produced exponentially (log phase), until the fourth day when the fungal biomass reached a maximum. The stationary phase followed with a slight decline o f the biomass as some condition within the growth flask had become limiting.  The biomass  production o f G. roseum also followed such a multi-stage growth curve, but it was less distinct.  59  Days - -•- - V. bulbillosum  —a— G. roseum  Figure 3.1. The biomass curves for Verticillium bulbillosum and Gliocladium roseum grown in pure Vogel's liquid medium (no D D A C ) . [data labels show acutal biomass values]. Single replicates used. Tolerance 1. Even though V. bulbillosum proved to be a more prolific fungus than G. roseum for growth in Vogel's medium, o f greater importance was how it compared to G. roseum, F. avenaceum and Z. moelleri in terms o f tolerance to D D A C .  The aim o f this experiment was to  find the best fungus for degrading D D A C under these conditions and a high tolerance to D D A C was an important step in choosing an organism. V. bulbillosum proved to be the most tolerant o f the fungi, showing no decrease in biomass over the tested range o f 0-250 ppm (Figure 3.2). G. roseum was sensitive to D D A C at concentrations between 100-200 ppm, dropping roughly 80% from the n o - D D A C control biomass. The biomass o f F. avenaceum fluctuated, but was significantly lower at 100 ppm than the n o - D D A C control o f 6.6 mg/ml.  Z. moelleri, the only zygomycete tested, showed rapid  growth in the liquid medium, and nearly reached its maximum biomass by the time that the D D A C  60  was added. For the no D D A C control, this fungus achieved a biomass o f 3 mg/ml. This biomass slowly decreased as the D D A C concentration increased, reaching 2.0 mg/ml at 250 ppm D D A C .  0.0 -I 0  1  1  1  1  1  50  100  150  200  250  D D A C C o n c e n t r a t i o n (ppm)  - - • - - V. bulbillosum —•—  G. roseum -  - F. avenaceum  •  Z. moelleri  Figure 3.2. Tolerance o f four fungi to D D A C (0-250 ppm) in Vogel's ( K H P ) medium after 14 days incubation, [data labels show acutal biomass values]. Single replicates used. This experiment was done using Vogel's medium with K H P as a buffer.  Therefore the  D D A C would be expected to have precipitated through reaction with the K H P , so that the concentrations reported would overestimate the amount o f D D A C available to the fungus.  It  would be anticipated that the biomass would decrease for the parallel experiment conducted with acetate buffer. Tolerance 2. To achieve significant levels o f D D A C degradation it was necessary to grow the fungi in as high a concentration o f D D A C as possible. However, the concentration o f D D A C that would be added to the solution was limited by the tolerance o f the fungi. Methods were sought to increase the amount o f D D A C that would be used in each experiment. The addition o f silica gel  61  to the liquid medium had been successfully applied in previous studies (Zheng, 1995).  This  method was based on the premise that most o f the D D A C would rapidly bind to the silica gel and be slowly released into aqueous solution as the D D A C in the solution was degraded, to maintain the equilibrium o f free to bound D D A C .  7  <l 6.4  6 - \  \  »  5 4-  __44  V4.9  DDAC (ppm) - -•- - V. bulbillosum  — D — G. roseum  Figure 3.3. Tolerance o f V. bulbillosum and G. roseum to D D A C in Vogel's medium buffered with K H P and containing silica gel (1.0 g/ 50 ml), [data labels show acutal biomass values]. Single replicates used. The results from this experiment confirmed that in liquid culture G. roseum is much more sensitive to D D A C than V. bulbillosum  (Figure 3.3). G. roseum failed to grow at or above 500  ppm D D A C , and at 250 ppm D D A C only 50% o f the biomass o f the n o - D D A C control was produced. This may be compared to V. bulbillosum, o f 5 mg/ml) at  1000 ppm D D A C .  which showed significant growth (biomass  Consequently,  biodegradation experiments o f D D A C in liquid culture.  62  V. bulbillosum  was chosen for  the  s  6.0  1  5.0  to (A  £ o m  4.0 3.0 2.0 1.0  V-H  <_0_6__  0.5  0.4  0.0 100  50  150  200  250  300  DDAC (ppm) - No silica gel Figure 3.4.  silica gel  Tolerance o f V. bulbillosum to D D A C (0-300 ppm) after 10 days incubation in Vogel's liquid culture (with acetate buffer) with and without silica gel added to the flask (1.0 g/50 ml), [data labels show acutal biomass values].  This experiment was carried out using silica gel and Vogel's medium buffered with K H P . Therefore, the actual aqueous concentrations o f D D A C to which the fungus was exposed would be significantly lower than reported.  In addition, the K H P buffer would cause D D A C to  precipitate out o f solution for an even smaller concentration o f D D A C available to the fungus. Tolerance 3.  When Vogel's medium was buffered with acetate in place o f the K H P , the D D A C  did not precipitate out and the reported concentration should represent the amount o f D D A C present in the aqueous phase.  Figure 3.4 shows the biomass produced in D D A C  tolerance  experiments where V. bulbillosum was grown in Vogel's medium buffered with acetate. In one experiment the medium was supplemented with silica gel (1.00 g/flask) while in the comparison experiment no silica gel was present.  V. bulbillosum showed a good tolerance to D D A C in the  presence o f silica gel. A t 300 ppm D D A C , V. bulbillosum produced an oven-dried biomass o f 5.5 63  mg/ml. However, with no silica gel present and no K H P to precipitate the D D A C , the threshold of the fungus was below 50 ppm, the lowest concentration tested.  D D A C R e m o v a l 1.  Metabolic intermediates are often short-lived, unstable compounds.  The  goal o f this experiment was to show the rate o f D D A C loss relative to the production o f fungal biomass. B y doing this one could determine at what stage the greater part o f the D D A C was being degraded and when to harvest for recovering the metabolites.  Figure 3.5.  Parallel Recovery o f D D A C and biomass measurement from cultures o f V. bulbillosum to which 250 ppm D D A C was added at day 2. Single replicates used.  Based on the day 2 control ( D D A C  added to acclimated culture and harvested  immediately), the extraction technique recovered 6 3 % o f the original D D A C .  Unrecovered  D D A C may have been lost on glassware surfaces, bound to the glass microfibre filters, as well as irreversibly bound to the culture solids. Tests examining the extraction o f D D A C from silica gel  64  confirmed that the D D A C loss due to volatilization during oven-drying (103°C for 8-16 hr) was negligible.  Table 3.3.  D D A C recovered from flasks (initially containing 12.5 mg) inoculated with V. bulbillosum. D D A C loss measured relative to the day 2 control (the day D D A C was added to the cultures). Day 16 control is an uninoculated control.  Day  Biomass  Recovered  Removal  (mg/ml)  D D A C (mg)  D D A C loss %  2 (control)  1.67  7.91  0  4  2.08  7.44  5.9  7  5.57  7.19  9.1  9  6.20  5.16  35  11  7.14  4.74  40  13  6.43  4.22  47  16  5.44  4.53  43  16 (control)  N/A  6.44  19  From the uninoculated control, harvested at day sixteen, 52% o f the original D D A C (6.4 mg) was recovered (Table 3.3). This yield, lower than the day two control, may have been the result o f non-biological breakdown (e.g. photodegradation) over time.  O n the other hand this  lower D D A C recovery may be a result o f stronger binding o f the Q A C s to the silica gel over time.  It is also possible that fungal products (e.g. organic acids) in the liquid culture (day 2  control) provided an environment more conducive to extracting the D D A C from the solids than the uninoculated control (day 16).  The effect o f D D A C addition (250 ppm) to the medium was to slow the rate o f growth o f V. bulbillosum. In the presence o f D D A C the maximum biomass was not reached until day 11 (Figure 3.5), whereas in liquid culture without D D A C , maximum biomass was achieved in four  65  days (Figure 3.1). The shape o f the growth curve o f V. bulbillosum grown with 250 ppm D D A C was similar to that o f the control culture. Both showed a lag and an exponential growth phase, followed by a stationary phase. In addition, the maximum biomass produced (7.0 mg/ml) by V. bulbillosum did not seem to be affected by the presence o f D D A C .  Over the two week incubation period, the amount o f D D A C decreased by nearly 50%, with the greatest removal taking place near the end o f the fungal exponential growth phase. During this phase, between days 7 to 9, there was a 25 % decrease in D D A C (Figure 3.5). Relatively small levels o f D D A C degradation were recorded from day 2 to day 7 (10%), and again from day 9 though day 16 (8%).  Biodegradation 2.  The D D A C extraction i n the second biodegradation experiment was not as  efficient (42%) as in the first (63%). Otherwise, the results were well replicated. The growth curve was similar to that o f Biodegradation 1, with a maximum biomass o f 6.8 mg/ml obtained at  Table 3.4.  D D A C recovered from flasks (initially containing 12.5 mg) inoculated with V. bulbillosum. D D A C loss measured relative to the day 3 control (the day D D A C was added to the cultures).  Day  Biomass  Recovered  Degradation  (mg/ml)  DDAC (mg)  DDAC loss %  3 (control)  2.2  5.3  0  5  3.3  5.3  0  7  4.7  4.2  21  9  6.4  2.7  49  11  6.6  2.8  47  13  5.8  2.2  58  15  6.0  3.2  40  66  E  Day ^•DDAC  Biomass  Figure 3.6. Replicate experiment: Parallel recovery o f D D A C and biomass measurement from cultures o f V. bulbillosum to which 250 ppm D D A C was added at day 2. T w o replicates used. day 11. In addition, the greatest decrease in D D A C recovery occurred between day 7 and day 9, with smaller losses before and after (Table 3.4 and Figure 3.6).  3.3.2 Wood Substrate Solid W o o d :  Table 3.5 gives the average disappearance o f D D A C . while Figure 3.7 shows the  initial D D A C retention calculated and the final D D A C retention extracted from wood pieces on two  plates each o f uninoculated control, and those inoculated with G. roseum, and V.  bulbillosum. The difference in initial D D A C retention o f the wood pieces reflects the inherent variability within the wood. After 17 weeks incubation, both G. roseum and V. bulbillosum were able to remove some o f the D D A C from the wood pieces. This was evidenced by a greater loss o f D D A C from wood pieces incubated with fungus, than in the uninoculated controls.  67  The  recovery o f D D A C from the control samples ranged from 89-91%, revealing that the extraction technique from wood was far more efficient than that from the liquid culture solids. Table 3.5. Average concentrations (mg D D A C / g wood) and D D A C disappearance for 8 wood pieces each that were either inoculated with V. bulbillosum, G. roseum, or uninoculated and incubated for a period o f 17 wks. (standard deviation in brackets)  DDAC (pre--incubation) mg DDAC/ g SDST  DDAC (post-incubation) DDAC disappearance mg DDAC/ g SDST %*  Control  24.3  (3-6)  21.9  (3.2)  9.9%  (0.2)  V. bulbillosum  25.4  (1.1)  17.5  (2.7)  31.3%  (7-7)  G. roseum  27.8  (0.9)  20.9  (0.6)  24.7%  (4.5)  * Note: average of % disappearance from individual wood pieces  In addition to a loss o f D D A C , a metabolite (designated P I ) was discovered in the harvested wood pieces incubated with both fungi, but not i n any uninoculated controls.  The  metabolite was identified by the presence o f a new peak in the H P L C chromatogram at 11.2 minutes relative to the D D A C peak at 9.0 minutes. A similar metabolite peak was discovered earlier by Btirgel et al. (1996b) i n similarly treated wood pieces subjected to the same fungi. A greater degree o f removal occurred with G. roseum (27-36%) than with V. bulbillosum (22-25%). The Gliocladium wood extracts also had a greater amount o f metabolite present than in the wood pieces incubated with Verticillium.  During this incubation period, G. roseum showed better  qualitative growth on the surface o f the wood pieces than V. bulbillosum.  From microscopic  observations, Biirgel et al. (1996a) found that G. roseum could penetrate into wood pieces treated with D D A C , whereas V. bulbillosum growth was restricted to the wood surface. This may give G. roseum an advantage in degrading D D A C from solid wood substrates.  The D D A C removal observed in these results was for treated wood pieces incubated for a period 17 weeks (118 days).  Considering the lengthy incubation time, one would expect a  68  a greater amount o f removal. The cause for the low removal was likely due to the high initial retention o f the D D A C (-2.5% by weight) in the wood that approached the toxic threshold for D D A C treated wood.  C1  C2  G1  G2  V1  V2  Figure 3.7. The disappearance o f D D A C from wood pieces treated with 2% D D A C after 17 weeks. C l and C 2 = uninoculated controls; G l & G 2 = G. roseum; and V I & V 2 = V. bulbillosum. P I = metabolite from D D A C transformation. Sawdust:  Previous experiments examining the removal o f D D A C in wood pieces by moulds  revealed that G. roseum could remove more o f the Q A C than V. bulbillosum on the solid substrate.  In addition, it appeared to grow more quickly (by visual observation). However,  neither fungus was able to remove more than 35% o f the D D A C over a seventeen week period. A s mould growth is limited to the wood surface, it was believed that one reason for the limited degradation was the inability o f the fungus to access the more deeply penetrated D D A C . Sawdust containing D D A C was incubated with G. roseum as it was hypothesized that the greater surface area would enhance the access to, and degradation o f the D D A C . intervals and the amount o f D D A C loss was recorded.  69  Sawdust was harvested at  A. Kinetics of Degradation: For the first two weeks of this experiment there was little or no D D A C removal detected in any of the sawdust treatments (Figure 3.8).  During these first two weeks, typical fungal growth  occurred on the agar, presumably as the rich agar-nutrients were being consumed. For this period the fungi had little need of the sawdust for nutritional purposes. From week 2 to week 9 the degradation followed a linear path for the three samples. By the second week, G. roseum had probably used up most of the readily available nutrients in the medium and began to utilize alternative sources of nutrients. The sapwood sawdust would contain starch and other nutrients that could be used by the fungus, however, the fungus would also have to overcome the effects of the D D A C in order to access this nutritional source. During the last two weeks of the experiment (week 9-11) the degradation seemed to level off for SWD I and SWD III. This could  0  2  4  6  8  10  12  Week  SWD I - ^ - S W D II - ^ - S W D III Figure 3.8. The rate of biodegradation of D D A C from three different sawdust treatments by G. roseum. 70  be attributed to colony atrophy as the agar was becoming dry and aged, and the fungus had exhausted its available food sources. SWD III was more variable than SWD I in its linear degradation. From week 2 through week 9 the linear regression R value for SWD III was 0.825 2  while that of SWD I was 0.983.  Table 3.6.  The recovery of DDAC (mg DDAC/g sawdust) over 11 weeks from the three different sawdust treatments, inoculated with G. roseum. Standard deviations in brackets. SWD  SWD  I  II  SWD  III  Week  0  19 (1.1)  28 (1.0)  19 (1.7)  1  20(1.1)  29 (1.0)  19 (1.7)  2  17(1.1)  29 (1.0)  20(1.7)  3  18 (0.8)  27(1.0)  15 (0.7)  4  16 (1.6)  26 (2.1)  14 (0.6)  5  14 (0.3)  -  14(1.9)  7  11 (1.9)  -  11 (1.0)  9  8 (0.2)  -  10 (2.5)  11  9 (0.9)  21 (2.0)  10 (2.0)  This is likely due to the more leterogenous distribution olf DDAC in SWD III as compared to SWD I. During treatment of the wood pieces, most of the DDAC will bind to the surface and the concentration will decrease rapidly with penetration into the wood, leaving the centre with reduced amounts of DDAC. Therefore SWD III is likely composed of some particles that are heavily treated, some that are only partially treated and some that have no DDAC at all. This helps to explain the shape of degradation for this sawdust treatment. From week 2 to week 3 there was a rapid drop in DDAC content as the sawdust fragments with partial DDAC treatment are handily depleted of QAC by the fungus.  The remaining DDAC containing  fragments have a higher DDAC content and are therefore degraded at a much slower rate (Table 71  3.6). S W D I, treated as a sawdust, was more homogeneous and therefore had a smoother, more uniform degradation than S W D III.  B . Rate o f Degradation:  The rate o f degradation for the three sawdust treatments was greatest i n sawdust I (-1400 ppm/week), followed by sawdust III (-1200 ppm/week) and sawdust II (-750 ppm/week) over the period from week 2 through week 9. The slow degradation rate o f S W D II is probably due to its high retention o f D D A C that approaches the toxic threshold for the fungus.  Table 3.7. The recovery o f D D A C from S W D I over a period o f 11 weeks and the formation o f metabolite " P I " over the same period. Standard deviations are given brackets.  Dl) \ (  Metabolite " I M "  Week  (mg D D A C / g sawdust)  (mg IM.'g sawdust)  0  19(1.1)  0(0)  1  20(1.1)  0 (0)  2  17 (1.1)  0(0)  3  18 (0.8)  0(0)  4  16 (1.6)  1.9 (1.0)  5  14 (0.3)  2.0 (0.4)  7  11 (1.9)  3.3 (0.5)  9  8 (0.2)  3.5 (0.6)  11  9 (0.9)  5.0 (1.0)  C. B y product Formation:  The H P L C method used to detect D D A C also detected a metabolite (PI). It appeared on the chromatogram as a peak with a retention time near 12.5 minutes, shortly after the D D A C  72  peak at 9.9 minutes (Figure 3.9). The metabolite did not appear until week 4 in both S W D I and S W D III, and reached a similar concentration o f 4.8 mg/g and 4.2 mg/g at week 11 for S W D I  Figure 3.9. The H P L C chromatogram o f S W D I incubated with G. roseum (G2) for 11 weeks.  Figure 3.10. The formation o f metabolite P I in time over an 11 month incubation period o f S W D I with G. roseum. and S W D III respectively (Table 3.7; Figure 3.10). The metabolite was also formed in S W D II, but required a longer incubation time than the other two sawdust treatments.  73  This metabolite  (PI) was not detected in any o f the uninoculated controls, confirming that it is the result o f a biological transformation o f D D A C .  3.4 Conclusions Differences between the growth o f the two test fungi G. roseum and V. bulbillosum in the liquid and wood substrates were observed.  V. bulbillosum grew more quickly, and to a greater  degree, than G. roseum in the liquid medium. It showed much greater tolerance to D D A C within the liquid culture than G. roseum. O n the other hand, G. roseum was better adapted to growth on solid substrates such as wood and agar.  O n these solid substrates it appeared to grow more  rapidly, produce thicker, more vigorous colonies and degrade more D D A C than V. bulbillosum. D D A C was degraded by V. bulbillosum i n liquid culture with silica gel present, but no metabolite was found.  However, this does not mean that metabolites were not formed.  The  methods used to search for metabolites were limited to what was retained by the H P L C ionexchange column, and noticable with the detection system.  Furthermore, i n oven-drying the  culture solids to determine the biomass, there was a possibility that volatile by-products were lost. This was unlikely though, as D D A C controls bound to silica gel did not evaporate after oven drying, and one would suspect that other Q A C and tertiary amine acid salt by-products that would be subject to retention on the H P L C column, would also have a l o w volatility.  The  method for extracting D D A C from the silica gel was poor (only 65% o f the control D D A C recovered). Therefore the metabolites may not have been extracted, either due to very strong binding to the culture solids, or more likely, by increased water solubility that would allow more metabolite to partition into the bulk liquid than onto the culture solids.  On a wood substrate, G. roseum grew better than V. bulbillosum. Both fungi produced a metabolite that was detected on the H P L C chromatogram at a retention time o f around 12.5 74  minutes. As the metabolite for both fungi occurred at the same retention time, it is likely that the initial degradation mechanism for both fungi is the same.  75  CHAPTER 4 ISOLATION 4.1  & IDENTIFICATION  of METABOLITE  P1  Introduction A metabolite P I was formed i n DDAC-treated sawdust from the biodegradation o f  D D A C by the fungus G. roseum (Sect. 3.3.2). Determining its identity would provide strong evidence as to the biodegradative pathway being used. However, before the chemical structure could be deciphered by instrumental analyses, the compound first had to be isolated from the remaining D D A C and other metabolites or wood extractives that were present in the colonized sawdust. Preparative H P L C was chosen as the technique for separating P I from the remaining D D A C , as the H P L C system used to detect the metabolite P I already proved effective in separating the two compounds. After the preparative H P L C separation, the metabolite fractions were checked for purity. O f special concern was contamination by D D A C as its presence in the PI fraction could easily confuse the interpretation o f N M R spectra.  In addition, an extraction  using benzene was used as reduce the amount o f B T M A C present.  After isolating the metabolite P I from the sawdust extract, two features were apparent that limited the analytical techniques that could be used for its identification: the amount o f P I , and its purity. There was only about 10 mg o f P I available (based on peak heights compared to the D D A C standard curve) in the inoculated sawdust, and with losses at each step o f purification, only 0.8 mg o f 'pure' P I was obtained. In addition, a significant amount o f B T M A C remained in the P I sample from the preparative H P L C procedure.  Therefore, analytical methods for  identifying P I were sought that were non-destructive so that the small amount o f P I sample  76  could be used again. H P L C was used to detect and isolate the metabolite, and also provided some information about its composition.  Proton N M R spectroscopy proved to be a very  powerful, non-destructive tool for the elucidation o f the chemical structure o f P I . However, it was not sufficient in and o f itself for positive identification. Infrared spectroscopy was used i n conjunction with N M R primarily to identify oxidized functional groups within P L  4.2 Isolation of P1 4.2.1 Preparative HPLC Preparative H P L C takes advantage o f the inherent separation o f chemicals on the H P L C column to isolate compounds from a mixture. Instead o f identifying a compound by its retention time under defined running conditions, or quantifying a compound by comparing the integration of its peak with that o f standard concentrations, preparative H P L C collects the liquid fractions as they come off o f the column. True preparative H P L C employs large columns for high volume separations that are operated under conditions unsuitable for analytical purposes. The isolation of metabolite P I by H P L C was a laborious task performed on the same analytical system used to detect it, which would more accurately be described as "semi-preparative".  The week 11 harvest o f S W D I (Sect. 3.3.2) was chosen as the source to isolate P I from, as it had the greatest ratio o f P L D D A C (0.52). There was about 1.8 g o f S W D I that had been incubated for 11 weeks, which contained approximately 8.5 mg o f extractable P L The P I was extracted from the sawdust as in Section 3.2.2. Six test tubes were set up with 0.3 g o f sawdust each. To each tube 6 ml o f acetonitrile (pH 3.0, adjusted with formic acid) was added.  The  acetonitrile containing the extract was collected; however, only 27 m l o f the total 36 m l added to the tubes was recovered. The rest remained absorbed by the sawdust. The solvent in the extract was removed by evaporation in a rotary evaporator, and then redissolved in 1 m l o f acetonitrile. 77  The concentration o f the D D A C and PI in this concentrated, crude extract was estimated by H P L C to be 8200 ppm and 4600 ppm, respectively. This would equate to only 3.6 mg o f PI remaining from the original 8.5 mg.  To determine the limits o f the column for concentrated samples, a series o f D D A C standards were prepared at different concentrations:  0.5%, 1.0%, 2.0%, 2.5%, 5%, and 10%.  The peak height was measured from each o f the standards, and plotted against the D D A C concentration. From 0-2.5% D D A C , the relationship between peak height and concentration was  12000  D D A C ( % )  Figure 4.1. Capacity o f the H P L C column for D D A C .  roughly linear (Figure 4.1). However, after 2.5%, the peak height for D D A C did not increase any further even with large increases in concentration.  Therefore, the H P L C column had a  maximum capacity for D D A C o f around 2.5%. A t this concentration, samples injected into the column would saturate the ion exchange sites and any more D D A C contained in higher concentrations would not bind to the column, but would be expelled in the void volume. It was  78  therefore necessary to keep the concentrated crude extract o f the metabolite below a D D A C concentration o f 2.0%.  The amount o f P I that could be isolated by this semi-preparative H P L C system was also limited by the size o f the injection loop. A n injection loop o f 20 p i meant that at the highest concentration o f crude extract, 0.16 mg o f P I could be injected at a time. This would require about 20 injections o f 15 minutes each. Due to the originally small sample o f P I i n the extract, it was necessary to minimize losses as much as possible.  A s the H P L C injections required  overfilling the injection loop to prevent the introduction o f air bubbles into the column, a vial was placed at the output port o f the injection loop overflow and the excess extract was recovered to be re-injected onto the column.  The H P L C system and running conditions for the semi-preparative isolation were the same as for analysis (Sect. 3.2). The only difference in equipment was the addition o f a fraction collector. A small section o f P E E K tubing was connected to the liquid output from the U V detector.  The liquid output from the H P L C passed through the U V detector and into a waste  beaker until a sample was injected into the H P L C . A t this time the liquid output was redirected into the test-tubes held by the fraction collector. L i q u i d fractions were collected in the test tubes at 2-minute intervals. The fraction o f primary interest, containing P I , was collected between 12 and 14 minutes (Figure 4.2).  There was a considerable amount o f B T M A C in the mobile phase (800 mg/L). A s B T M A C was present as the sole non-volatile component in the mobile phase, it would not be removed in the rotary evaporator. It was uncertain at first whether or not the B T M A C , present in the crude P I fraction, would interfere with the identification o f PI by proton N M R .  From the  literature values, the N M R shifts for the protons in B T M A C should be easily identifiable and not  79  interfere with the interpretation o f P I proton N M R spectra. A comparison o f the proton N M R spectra o f D D A C and B T M A C standards confirmed this.  Figure 4.2.  H P L C chromatogram o f the concentrated P I crude extract. The peak at 10.09 min. represents D D A C while that at 12.50 min. represents P L  4.2.2 Confirmation of Purity Because o f the similar H P L C retention times between D D A C and P I , it was expected that the metabolite P I was very similar in structure to that o f D D A C . Consequently, it was important to completely separate all o f the D D A C from the crude P I extract. To confirm that the semipreparative H P L C method had achieved this, the P I fraction (collected from 12-14 min) was checked by analytical H P L C for the presence o f any D D A C .  A l l o f the PI fractions from the  numerous injections were combined in a round bottom flask in which the solvent o f the mobile phase was removed by a rotary evaporator. The small amount o f dried residue remaining was then redissolved in 1.0 m l o f pure acetonitrile. In addition, the fractions collected immediately  80  before P I (10-12 min) and immediately following P I (14-16 min) were also examined for purity (Figure 4.3).  Figure 4.3.  Confirmation by H P L C that the crude P I fraction contained no D D A C contamination. A = DDAC-containing fraction (10-12 min); B = P I fraction (12-14 min) and C = the post-Pi fraction (14-16 min).  It was uncertain whether or not extractives i n the wood would pose a problem for purification. W o o d contains many complex extractives, some o f which were present in the crude PI extract. This was evident from the light yellow colour o f the fresh sawdust extract that turned dark reddish brown upon concentration. However, most o f the wood extractives are found in the heartwood and the wood from which PI was extracted was southern yellow pine sapwood. This sawdust should therefore contain relatively small concentrations o f extractives. In addition, there is very little nitrogen in wood.  The only cationic biomolecules that would be likely to be  retained on our H P L C column would have to be based on nitrogen. Based on this argument the prospect o f separating the wood extractives from P I appeared to be promising.  During  preparative H P L C , the yellow colour from each o f the injections appeared i n the first fraction that was collected, suggesting that most o f the extractives came out in the void volume, and were not retained by the column.  81  i-1  mmmm-i  1  Be Figure 4.4.  Wavelength (nm)  :  (  4ee.  E$@  U V - V i s spectrum o f the concentrated P I extract from S W D week 11 prior to preparative H P L C . (reference cell = H P L C mobile phase). Inset is spectrum o f BTMAC.  Nevertheless, U V - V i s analysis was used to check for contamination of.the crude P I fraction by wood extractives.  W o o d extractives often contain UV-absorbing chromophores  while D D A C and presumably P I do not absorb U V or visible light. Figure 4.4 shows the large amount o f UV-absorbing material that is present in the concentrated P I extract. A s can be seen in Figure 4.5, most o f this UV-absorbing material has been removed after preparative H P L C . The peak that remained from - 2 3 5 to 275 nm occurred over the same absorbing range as B T M A C , and probably contained a considerable amount o f it. B T M A C was the only non-volatile compound in the mobile phase. Therefore once the PI fractions were collected from the H P L C , combined, and the solvent removed, a considerable amount o f B T M A C was still present. Initially it was not thought that its presence in the sample would pose a problem during identification o f P I , as the N M R shifts o f the protons on B T M A C were significantly downfield from D D A C , and the locations where P I protons were anticipated.  82  However, due to the much greater concentration o f B T M A C than P I , initial N M R experiments were unsuccessful.  4.0686  I A/1  i  1  *  V  i  e  - * 1.0000_  I  ,  / i  J  8,ee8e-^—-r--—.— 200.BB  Figure 4.5.  —  T  '  {  300.80  Wavelength (nm)  -^P——. .: '  ~'""~~r~--^==^ 400.00  U V - V i s spectrum o f the P I fraction collected after preparative H P L C . cell = H P L C mobile phase)  (reference  4.2.3 Benzene Extraction In the crude P I fraction collected from 12-14 m i n there was about 25 times as much B T M A C as there was P L This was due i n part to the small amount o f crude extract per injection into the column, and in part to the constant amount o f B T M A C (800 ppm) i n the mobile phase. Even though the N M R proton shifts were different for these two compounds, the proton signals for B T M A C dwarfed those o f the P I metabolite. Therefore the amount o f B T M A C present i n the P I fraction had to be reduced so that N M R could be employed i n the identification o f P I .  It was not necessary to remove all o f the B T M A C from the fraction. It was only required that the ratio o f B T M A C to P I which was about 25 to 1 after preparative H P L C be reduced to about 1:1. Therefore, a simple liquid extraction was attempted.  83  This first required finding a  solvent that would preferentially dissolve one o f the components, but not the other. A range o f solvents was selected that spanned a wide range o f polarities. The solvents tested in order o f decreasing polarity were distilled water, methanol, acetonitrile, tetrahydrofuran,  methylene  chloride, benzene, and hexane. For the solubility tests, aliquots o f B T M A C (200 mg), o f D D A C (0.2 ml) as Bardac 2280, and o f solid dioctyldimethylammonium chloride ( D O D M A C ) (20 mg) were placed in separate test tubes. To each o f the tubes one o f the solvents was added (1 ml), and the dissolution o f the solids or the miscibility o f Bardac 2280 was observed (Table 4.1). I f the solids did not dissolve within 5 minutes, more solvent was added i n 1 -ml increments until the solid dissolved or the volume reached 10 m l . D D A C and D O D M A C were used to estimate the likely solubility o f P I in different solvents, as the structure o f P I was believed to be similar.  O f the eight solvents tested, only benzene and tetrahydrofuran were suitable for extracting PI from B T M A C in the crude P I fraction.  Benzene completely dissolved D O D M A C and  Bardac 2280 in 1 m l , while B T M A C showed no signs o f dissolving in a volume o f 10 m l benzene.  The liquid extraction consisted o f adding a 25 m l aliquot o f benzene into the round  bottom flask containing the dried P I fraction. The flask was capped with a stopper and shaken for 5 minute, before the extracting solvent was transferred to another round-bottom flask. This was repeated twice, the benzene from each extraction was combined, and then evaporated in a rotary evaporator to produce a dry residue.  The refined P I fraction was then dissolved in  acetonitrile and a sample analyzed by H P L C confirmed that the benzene extraction did not alter the compound or introduce contaminants, as only a single peak was present at the retention time of PI (12.5 min).  84  Table 4.1  Solubility test for benzyltrimethylammonium chloride ( B T M A C ) , D D A C as Bardac 2280, and dioctyldimethylammonium chloride ( D O D M A C ) in a range o f solvents.  S O U BIU | \ SOLVENT  Water  BTMAC  BAKI).\( 22S0  DODMAC  (200 nm)  (0.2 ml)  (20 nm)  1 Complete 1 Complete  Methanol Acetonitrile Tetrahydrofuran Methylene Chloride Ben/ene Hexane  ml dissolution ml dissolution  1 ml Slow dissolution 5 ml N o dissolution 10 m l Partial dissolution 10 ml No dissuliiiiiin 10 m l N o dissolution  1 ml Miscible 1 ml Miscible 1 ml Miscible 1 ml Miscible 1 ml Miscible 1 nil Miscible 2 ml Not miscible  1 Complete 1 Complete  ml dissolution ml dissolution  1 ml Complete dissolution 1 ml Complete dissolution 1 ml Complete dissolution 1 ml CVmplele dis^oluiimi 10 m l N o dissolution  It was estimated that there was roughly 8.5 mg o f extractable P I i n the S W D I that had been incubated for 11 weeks. Through each step o f extraction and purification, some losses o f P I were observed.  After the benzene extraction to remove excess B T M A C from the crude P I  fraction, only about 0.8 mg o f relatively pure P I remained.  4.3 4.3.1  Identification of P1 Methodology  4.3.1.1 H P L C This section was not designed as an experiment to provide evidence as to the identity o f P I , but is rather a collection o f useful observations and results obtained over time along the way. For details into the H P L C system used, see sect 3.2.  4.3.1.2 NMR Chemicals Several Q A C s were used as standards for comparison o f N M R spectra with the metabolite P I : D D A C , B T M A C , Dodecyldimethylhydroxypropylammonium bromide ( Q A C l ° O H ) and dodecyldimethylhydroxy-2-propylammonium bromide ( Q A C - 2 ° O H ) were obtained as white powders synthesized by Albemarle Corporation. In addition a carboxylated Q A C , 11-carboxyundecanyldecyldimethylammoniun chloride ( Q A C - C O O H ) , was synthesized by Dr. Alessio Serreqi (Forest Products Biotechnology, U B C ) . It was prepared as follows: N,N-dimethyldecylamine (0.20 g, 1.1 mmol) and tributylamine (0.34 ml, 1.4 mmol) were added to a solution o f 11-bromoundecanoic acid (0.31 g, 1.2 mmol) in dry dioxin (3 ml), and refluxed for 18 hours. Ethyl acetate was added to the reaction mixture and a precipitate formed which was filtered and washed with N,N-dimethylformamide ( D M F ) . Drying in a dessicator in the presence o f P 0 2  5  gave the quaternary ammonium product as a white  amorphous solid (0.30 g, 0.67 mmol, 6 1 % yield). The PI metabolite sample contained a significant amount o f B T M A C .  To determine  whether or not this would pose a problem to the identification o f P I , a mixture o f D D A C and B T M A C (1:1) was prepared for N M R analysis.  Sample Preparation Hydrogen-containing solvents generate interfering peaks on N M R spectra. therefore very important to remove them from the sample prior to analysis.  It was  Water poses a  particular problem as it is ubiquitous in the atmosphere and is attracted by the hygroscopic Q A C samples as well as the glass N M R tubes. Water contamination appears as a broad, variable peak in the N M R spectrum that can make interpretation difficult. thoroughly dried.  86  For this reason the samples were  A n aliquot (~2 g) o f Bardac 2280 was placed i n a round bottomed flask and the water and ethanol were removed i n a rotary evaporator. After removing as much solvent as possible, the resulting paste was placed in a freeze drier (Edwards E 2 M 8 ) overnight. The other Q A C samples to be used were already in powder or paste form and were dried directly in the freeze drier overnight. After freeze drying, each sample was immediately transferred to an N M R tube (5 m m outer diameter; Norell X R - 5 5 ) that had been oven dried at 103°C overnight to remove any residual water adhering to the glass surface.  Approximately 2 m l o f deuterated chloroform  (CDC1 ) (D, 99.8%, Cambridge Isotope Laboratories) was added to dissolve the dried Q A C 3  within the tube and the sample was taken for N M R analysis within 2 hours.  N M R analysis: Samples were run on instruments at the N M R spectroscopic laboratory (Dept. o f Chemistry, U B C ) . The spectra o f pure D D A C , B T M A C , as well as the 1:1 mixture o f these two Q A C s , were recorded on a 200 M H z instrument (Briiker A C - 2 0 0 ) ; while the metabolite P I , Q A C - C O O H , and both hydroxylated Q A C s were examined using a 400 M H z instrument (Briiker WH-400). The higher the operating frequency, the greater the resolution o f the N M R spectra. This increased resolving power was very important for obtaining good spectra o f P I and Q A C C O O H where there was only a small amount o f the substance for analysis. A s the deuterated solvent contained a small fraction o f chloroform (0.2%), no T M S was added as a standard. The N M R instruments were set to lock onto the C H C 1 solvent signal, which produced a narrow peak 3  in all spectra at 7.25 ppm.  4.3.1.3 FTIR Chemicals: Samples o f the following quaternary ammonium compounds were analyzed by F T I R : P I metabolite, D D A C , B T M A C , Q A C - l ° O H , Q A C - 2 ° O H , Q A C - C O O H (Sect. 4.3.1.2). 87  For  comparison o f D D A C and PI spectra the D D A C was mixed with a B T M A C in the ratio estimated to be present in the P I sample.  Chloroform (Fisher) was used as the solvent for  dissolving the Q A C samples.  Sample preparation:  S T O P P E R  I>l2-!2ftj  Figure 4.6. L i q u i d cell components (adapted from Perkin Elmer)  F T I R analysis was performed using a liquid cell, which had the advantages over a pellet or mull i n that it could be used for a very small amount o f sample, which could be recovered afterwards. The liquid cell (Perkin Elmer) consisted o f a sandwich o f polished sodium chloride plates held i n an assemblage o f steel plates, gaskets, and spacers (Figure 4.6). A s there are no covalent bonds in sodium chloride, the I R absorption is minimal over the range o f 4000 - 650 cm"  First the samples were placed in the rotary evaporator to remove as much water as possible prior to being freeze-dried overnight. They were then transferred to a desicator until they were ready to be analyzed. The vigorous drying o f the Q A C samples was important for two main reasons. 88  Water contamination would appear on I R spectra as a broad peak that would hide any O - H stretch resulting from a hydroxyl group. Moreover, the sodium chloride plates are very sensitive to water and w i l l dissolve i f too much water is present in the sample.  F T I R analysis: The dried samples were dissolved in chloroform to a final concentration o f 25 mg/ml. A syringe was used to place a thin film o f sample solution between plates o f polished sodium chloride in the liquid cell (Perkin Elmer). The liquid cell was rinsed between samples with three m l o f chloroform solvent and then a spectrum was run o f the solvent to check for purity. The analysis was carried out on a Perkin Elmer (1600 Series) F T I R over the range 4000 c m to 650 -1  cm'.  4.3.2 Results and Discussion 4.3.2.1 HPLC The retention time o f a compound in an H P L C system under a certain set o f running conditions is a relatively arbitrary value. It is reproducible i f the conditions remain exactly the same; however, slight differences between mobile phases from batch to batch, aging effects o f the column, and daily temperature fluctuations make some variation in retention time inevitable. Therefore, there is a need to run standards ( D D A C ) for comparison. The concentration o f the sample plays a minor role in increasing the retention times.  The larger the concentration, the longer the  retention time, due to the larger peak. Although it is difficult to standardize the retention time o f an unknown compound, an idea o f its retention time relative to D D A C provides some information about its identity. Early thoughts into the identity o f P I , excluded tertiary amines as it was believed they would not be detected on the cation exchange column. In fact, tertiary amines become acid salts  89  Table 4.2. Retention times o f tertiary amine acid salts and Q A C s injected into the H P L C system. D D A C standards (from same mobile phase) provide reference. Sample Chemical Structure  Retention time (min)  D D A C Standard Peak Height  Retention time (min)  Peak Height  9 • 12.5  440  9.53  1781  10.32  1197  17.95  843  10.17  1243  16.15  1040  10.17  1243  11.33  229  10.00  2730  10.0  955  Metabolite PI  Didecylmethylammonium chloride  \ \ / \ / \ / \ / \ /  + N  \  Decyldimethylammonium chloride  \ / \ / \ / \ / \ / \ /  N  \  Dodecyldimethylammonium chloride  Dioctyldimethylammonium chloride at the p H o f the mobile phase, and are detected by this H P L C method.  This opened the  possibility that P I was a product o f alkyl chain cleavage. However, the much greater retention time for the tertiary amine (decyldimethylammonium chloride) than metabolite P I suggests that  90  it is unlikely P I is the result o f alkyl-chain cleaving. Table 4.2 shows the retention times o f several tertiary amine acid salts along with D D A C standards.  The  tertiary amine acid salts followed the same general pattern as the  saturated,  unoxidized, aliphatic Q A C s . In other words, the longer the alkyl chains, the shorter the retention time due to the greater steric inhibition with the binding sites. The carboxylated Q A C s did not appear to fit the same size/retention time pattern as the unoxidized aliphatic Q A C s .  This likely  has something to do with interference by the carboxyl groups. It is interesting to observe that didecylmethylammonium chloride has a shorter retention time than the D D A C , even though it is a slightly smaller molecule. This is likely the result o f D D A C having a stronger charge due to the methyl on the nitrogen rather than the hydrogen coordinate bond in the tertiary amine acid salt.  A s can be seen from the retention times in table 5.2.1 this H P L C system is very sensitive  to changes in structure. Therefore it is not expected that P I w i l l differ greatly from D D A C in structure.  4.3.2.2 NMR A l l o f the complete ' H - N M R spectra are contained i n Appendix I, with abstracted subspectra shown in Figures 4.7 through 4.10.  Appendix II contains the integration o f each peak and  describes how the proton counts were calculated. In addition, chemical structures are provided in the text, with numbered carbon atoms for clarity, where appropriate. 5  D D A C (Spectrum 1)  _  L  J  The ' H - N M R spectra for D D A C , dioctyldimethylammonium chloride ( D O D M A C ) , and a mixture o f dialkyl(50%, C , C ; 25% C , C , ; 25% C C )dimethylammonium chlorides (Bardac 0  8  0  0  8  91  g  2250, Lonza Inc.) all showed common features, which are characteristic for these dialkyl Q A C s . The spectra consisted o f three distinct regions where resonance by protons occurred.  Region #1, located furthest upfield at 0.8-0.9 ppm, was the result o f resonance by the methyl protons at the end o f the alkyl chain. Integration o f the peak, which was split into a very distinct triplet by the pair o f adjacent methylene protons, accounted for six protons.  Region #2 resulted from the absorbance o f the methylene protons within the alkyl chain, and occurred slightly downfield from the methyl peak over the range o f 1.2-1.8 ppm. It was composed o f a large peak and shoulder totaling 28 protons over 1.2-1.5 ppm (i.e. the in-chain protons o f C3 to C8), and a smaller broad multiplet peak o f four protons over 1.6-1.8 ppm. This smaller peak represented the methylene protons attached to C 2 (two carbons away from the ammonium nitrogen) o f both alkyl chains.  Region #3 resulted from the absorption o f the protons, both methyl and methylene, on carbon atoms adjacent to the nitrogen. Peaks accounting for 10 protons were found over the range o f 3.3-3.6 ppm. The six methyl protons formed a sharp singlet at 3.35 ppm, while the four methylene protons (bound to C,) were shifted only slightly downfield at 3.4-3.6 ppm.  In  addition, the D D A C spectrum showed a sharp singlet at around 2.0 ppm. This was believed to be caused by a small amount o f remaining water.  B T M A C (Spectrum 2) The  1  ""  3  ^  J  ' H - N M R spectrum for B T M A C was relatively simple as there were only five sets o f  chemically equivalent protons.  A sharp singlet at 3.4 ppm was the result o f the nine methyl  protons adjacent to the nitrogen. Another smaller singlet at 5.0 ppm resulted from the pair o f methylene protons between the nitrogen and the phenyl ring. The five aromatic protons gave a 92  characteristic absorption pattern o f two multiplet peaks shifted well downfield: a multiplet for the ortho and para protons at 7.45 ppm, and a doublet o f doublets at 7.6 ppm for the meta protons. A small peak at 2.7 ppm was suspected to be due to water contamination.  Combination o f D D A C and B T M A C (Spectrum 3) A s expected, the solution containing D D A C and B T M A C produced a spectrum that was the combination o f the peaks o f the individual spectra.  A s the methyl protons adjacent to the  nitrogen were common in both compounds, there was an overlap in the N M R peak associated with this group, which occurred at 3.4 ppm. However, the contribution to the proton integration arising from B T M A C could be determined from the integration o f the unique B T M A C peaks. This allowed the contribution o f the B T M A C to the overlapped peak to be subtracted, leaving the remaining integration due to D D A C to be identified. Consequently, it was not necessary to remove all o f the B T M A C in the P I sample after preparative H P L C , prior to N M R analysis. Furthermore, in this spectrum o f D D A C + B T M A C a proton resonance due to water was located at 2.3 ppm, which was between that found i n the pure D D A C and B T M A C spectra. Metabolite P I (Spectrum 4) The spectrum o f P I was similar to that o f D D A C and the other dialkyldimethyl Q A C s , having peaks in the same three regions with similar chemical shifts. However, upon closer observation, the N M R resonances for each o f these three regions differed from those o f the D D A C spectrum. The methyl region was split into a multiplet in P I as opposed to a clean triplet in D D A C .  The  resonance due to the methylene protons was a combination o f five partially resolved peaks, while only three were observed in the D D A C spectrum. In addition, the peak integration revealed that there were two protons less in the total methylene integration o f P I (30 protons) compared with that o f D D A C (32 protons).  The protons on the carbon ( C l ) adjacent to the nitrogen in P I  produced a broader resonance than that observed for D D A C extending over the range 3.2-3.8 93  ppm.  In addition, the resonance envelope contained four peaks with a total integration o f 12  protons instead o f the 10 protons determined for D D A C .  Almost all o f the water had been  removed from the sample and consequently only an extremely small peak appeared at 2.0 ppm. One unexpected large, sharp single resonance peak at 0.1 ppm was tentatively attributed to stopcock grease (Silverstein, p 215) that must have entered the N M R sample tube during workup of the sample in the rotary-evaporator.  +  Q A C - C O O H (Spectrum 5)  _ l  w  O  J  In addition to the N M R resonances due to the terminal methyl, the methylene, and the protons on the carbons adjacent to the nitrogen, all o f which were characteristic o f ' H - N M R spectra for dialkyldimethyl Q A C s , there was an additional peak in the Q A C - C O O H spectrum that was diagnostic for carboxylated alkyl chains.  A t 2.3-2.4 ppm a triplet, which integrated to two  protons, corresponded to the methylene protons on the carbon adjacent to the terminal C O O H . A peak far downfield (near 11 ppm) resulting from the strongly deshielded carboxyl proton, did not appear on the spectrum.  However, this peak is often absent form the 1 H - N M R spectra o f  carboxylated compounds as the carboxyl proton dissociates easily. The alkyl methyl component of the spectrum was the same shape as that in D D A C (a triplet at 0.85 ppm), but the integration corresponded to only three protons, instead o f the six present in D D A C . The methylene region (1.2-1.8 ppm) o f the Q A C - C O O H showed four distinct peaks compared to three for D D A C and five for P I . The peaks resulting from the methyl and methylene protons adjacent to the nitrogen in the carboxylated spectrum had the same general shape as that in the P I spectrum. However, in the Q A C - C O O H spectrum they were completely separated, whereas in the P I spectrum these  94  peaks are somewhat merged. The multiplet resonances located at 2.3 and 4.0 ppm are considered to be due to contamination arising from residual products o f the Q A C synthesis.  ^2^CH -OH  Br  Q A C - l ° O H (Spectrum 6) 1 3' The N M R spectrum for Q A C - l ° O H differed from that o f D D A C mainly through the 2  influence o f functional group signals that spread the peaks out. The terminal alkyl methyl peak (arising from protons bound to C12) was still present as a triplet at 0.85 ppm and contained three protons. Four peaks accounting for a total o f 22 protons were found in region #2 (1.2-2.1 ppm), including a singlet peak at 2.0 ppm, not seen in any o f the previous spectra.  The resonance at 2.05 ppm i n the Q A C - l ° O H peak probably represented the two methylene protons (attached to C 2 ' ) i n the middle o f the short propanol chain. If so, it would be split by the methylene protons on either side which would yield a triplet o f triplets or a nine-peak multiplet.  These would be unresolved, resulting in the broad single resonance observed.  (Appendix I, Spectrum 6).  Four peaks accounting for a total o f 13 protons were located in the region between 3.24.3 ppm. The proton resonance from the methyl and methylene groups adjacent to the nitrogen were completely separate, as i n the Q A C - C O O H spectrum.  However, integration o f the  methylene peak at 3.4 ppm accounted for only the two protons on the dodecyl chain (attached to C l ) . The methylene group ( C l ' ) on the propanol chain would be expected to be deshielded by the proximity o f the O H group, causing it to be shifted downfield to 3.7-3.8 ppm. A s well, the methylene protons (C3') adjacent to the terminal hydroxyl group were shifted to a similar position due to the deshielding effect o f the nitrogen. The result is a peak cluster at 3.7-3.8 ppm containing four protons. This peak may have arisen from the overlapping o f two triplet peaks, as 95  each methylene signal should a triplet. The single peak at 4.1 ppm resulted from absorption by the sole hydroxyl proton.  N  Q A C - 2 ° O H (Spectrum 7)  1  2' O  H  12  Br'  j _  QAC-2°OH was structurally identical to Q A C - l ° O H , except that the hydroxyl group was located in the middle o f the propyl chain rather than at the end o f the chain. There are three main differences between the ' H - N M R spectrum o f QAC-2°OH compared to that o f Q A C - l ° O H . The resonance due to the methylene protons (1.2-1.8 ppm) is less broad than in Q A C - l ° O H as it lacks the peak found in the Q A C - l ° O H spectrum at 2.0-2.1 ppm and has a broad doublet at 1.6-1.8 ppm  containing two protons.  This doublet arises from the methylene protons (C2) on the  dodecyl chain, two carbons from the nitrogen. The integration to two protons arises because the corresponding methylene on the propyl chain has been replaced by a methine (-CH(OH)-) group to which the secondary hydroxyl is attached. The net integration o f the region 1.2-2.1 ppm in the ' H - N M R spectrum o f QAC-2°OH was 23, while the Q A C - l ° O H spectrum had 22 protons in this region. In addition to the 20 methylene protons within the dodecyl chain (C2-C11), this proton count in QAC-2°OH included the three methyl protons (C3') on the terminus o f the propyl chain, which was shifted downfield by the adjacent O H group and appears as a sharp doublet at 1.25 ppm.  96  98  !  Jt h  IT  •S  •  A  1  1  -i  1  •  i  j  /  fV  fa  r-•  \  e*  1  1 i  1  1  2.0  1.5  1 j1  I IF  '  ii  •  I  1  1-0  •  1-  Figure 4.9. ' H - N M R spectra o f A : Q A C - 2 ° O H , B : Metabolite uP I , and C : Q A C - l ° O H over the range 0.8-2.5 ppm. 99  Table 4.3. Data from ' H - N M R spectra for D D A C , P I , and Q A C - C O O H DDAC Prolan T \ pe  N-( 11 R-c  11  B (  11  -K  i (mil  IN-CII,  Metabolite PI  O \C -( O O l l  Shift (ppm)  Multi  Int  Shill (ppm)  Multi  Int  Shiit (ppm)  Multi  Int  0.85  t  6  0.85  P  6  0.85  t  3  1.2-1.8  m(3)  32  1.2-1.8  m(5)  30  1.2-1.8  m(4)  44  -  -  -  -  -  -  2.35  t  2  3.3-3.55  m(3)  10  3.3-3.5  m(2)  11  3.35  s  6  3.5  m(4)  4  -  -  -  4N-CH.-R  included above  K-C'I1(()M)-K  -  -  included above  -  3.5-3.8  m(2)  1  Note: M u l t i is multiplicity o f peaks where s = singlet, d = doublet, t = triplet, q = quadruplet, p = pentuplet, m = multiple peaks (# o f peaks). Int. is integration o f peaks i n terms o f relative number o f protons Table 4.4. Data from ' H - N M R spectra for P I , Q A C - l ° O H and QAC-2°OH  o\< -mn  Met;il)olite PI Proton 1 \ pi-  K-i II  Shift (ppm)  Mu lli Int  0.85  P  IM II -R  1.2-1.8  m(5)  30 .  \ ( II  3.3-3.5  m(2)  10  Shift (ppm)  Multi  0.85  t  6  1.2-2.1  QAC-yOH  Shift (ppm) 3  m(4) 22  Multi Int  0.85  t  3  1.2-1.8  m(4)  23  3.3  s  6  3.35  s  6  3.4  m(3)  2  3.5  m(2)  2  - \ - t H;-I<  included above  rN-Cll2-CH,-CI!i-0[l  -  -  -  3.7-3.8  q  4  -  -  -  -  -  -  -  -  -  3.5-3.6  m(5)  3  K-< ll(Olh-K  3.5-3.8  m(2)  1  -  -  -  -  -  -  K-> )ll  3.3-3.5  -  1  4.15  s  1  4.5  s  1  \-(  ll;-(ll-()||  Note: M u l t i is multiplicity o f peaks where s = singlet, d = doublet, t = triplet, q = quadruplet, p = pentuplet, m = multiple peaks (# o f peaks). Int. is integration o f peaks i n terms o f relative number o f protons  The integration o f the protons on the carbons adjacent to the nitrogen resulted i n 11 for QAC-2°OH, compared to 12 for the corresponding region o f Q A C - T O H .  The methyl protons  adjacent to the nitrogen accounted for six protons, while the peak for the methylene protons adjacent to the nitrogen only accounted for two protons.  A multiplet at 3.5-3.6 ppm, which  integrated to three protons, was believed to be a combination o f the methylene protons between  101  the nitrogen and the hydroxylated methine and the methine proton.  The hydroxyl proton  resonance was observed at 4.5 ppm.  Discussion The metabolite P I clearly resulted from the biological transformation o f D D A C as it could not be detected i n the uninoculated controls. Therefore it was expected to be related to D D A C through an oxidation o f one o f the alkyl chains. Biirgel et al. (1996) hypothesized that the metabolite P I contained an oxidized group in the form o f either an aldehyde or carboxyl, at the end o f the alkyl chain, due to (3-oxidation. The results described by Dean-Raymond and Alexander  (1977)  supported  this  prediction,  with  the  discovery  of  9-  carboxynonyltrimethylammonium chloride and 7-carboxyheptyltrimethylammonium chloride as metabolites from Q A C biodegradation. However, integration o f the alkyl methyl peak i n the ' H N M R spectrum o f P I revealed that both methyl groups remained intact (i.e. six protons). This meant that i f P I resulted from an oxidation reaction o f D D A C , the oxidized carbon was not located at the end o f the alkyl chain. The ' H - N M R spectrum o f a carboxylated quaternary ammonium compound ( Q A C C O O H ; Appendix I, spectrum #5) was run to compare with that o f P I . A close examination o f the methylene regions o f D D A C , P I , and Q A C - C O O H revealed that P I was indeed different from D D A C , but did not contain a carboxylated alkyl chain similar to that found in Q A C - C O O H (Figure 4.1).  Carboxyl groups may be readily detected by the resonance due to the carboxyl  proton that occurs far downfield at 10-13 ppm. However, this proton may exchange with traces of water in the sample, and is not be detected.  However, carboxyl groups attached to an alkyl  chain would cause the adjacent methylene protons to be shifted further downfield to 2.3-2.4 ppm. In addition the proton resonance would be split into a triplet, due to interaction with the adjacent methylene protons. This peak was clearly seen i n the Q A C - C O O H spectrum, but was not found 102  in the P I spectrum, confirming the lack o f a carboxyl group in the metabolite. Oxidation to form carboxyl groups represents the limit o f oxidation prior to cleavage o f the carbon-carbon bond. More limited oxidation would result in either alcohol or aldehyde groups being produced.  A secondary hydroxyl group generated on an alkyl chain o f D D A C during biodegradation would still generate a peak at -0.85 ppm with two alkyl methyl groups (6 ' H ) .  Secondary  hydroxylation could also account for the loss o f the 2 ' H from the methylene region o f P I compared to that o f D D A C (Table 4.1).  Both the hydroxyl proton ( O H ) and the methine (-  C H ( O H ) - ) proton would be shifted downfield out o f the methylene region and into the proton resonance associated with the C - N . The lH-resonance due to the hydroxylated methine would be observed ca. 3.6-4.0 ppm, while the chemical shift for the hydroxyl proton fluctuates widely from 0.5-5 ppm (Silverstein, 1991; Bruice, 1995). In the ' H - N M R spectrum P I there is a small peak at 3.6 ppm, which is absent i n the spectra o f D D A C and Q A C - C O O H (Figure 4.2).  In  addition, the integration o f the protons on carbons bound to the nitrogen confirms 12 protons are present in the P I spectrum.  Ten protons result from methyl (6) and methylene (4) protons  adjacent to the nitrogen. The remaining two protons under the broad, merged peak could be due to the methine and hydroxyl proton. To investigate this further, two samples o f hydroxylated Q A C s were prepared for N M R , one with terminal alkyl hydroxylation ( Q A C - l ° O H ) and the other with a secondary O H group (QAC-2°OH).  The spectra o f Q A C - l ° O H and QAC-2°OH were compiled as examples o f hydroxylated Q A C s so that their spectral features could be compared with those o f P I . It was also useful to determine whether any differences between the spectra o f the two hydroxylated Q A C s could be attributed to the position o f the O H group on the alkyl chain. In general the chemical shifts o f these hydroxylated Q A C spectra supported the hypothesis that P I contained a secondary alcohol.  103  However, these Q A C s were o f limited usefulness as models for comparison with P I . Although they were dialkyl Q A C s , the second alkyl chain was only three carbons long. It is probable that there are important differences between Q A C s with hydroxylation on short alkyl chains and intermediate alkyl chains such as that proposed for P I .  The hydroxylated Q A C nmr spectra did not explain why the P I spectrum has such a varied methylene region (Figure 4-3) as the equivalent region o f both Q A C - l ° O H and Q A C 2°OH were clear with distinct peaks. effects.  This could be attributed to short chain vs. long chain  O n a propanol chain, the hydroxyl group would exert a significant effect on the six  protons attached to the three carbon atoms, but with longer chains a greater number o f protons may be influenced.  Assuming that the essential modification in P I is the insertion o f a hydroxyl group in the alkyl chain, the question then arises: where is it located on the alkyl chain? It is clear that it is not on the carbon adjacent to or second from the nitrogen. Equally the spectra do not support the hydroxylation o f the terminal methyl group, based on the presence o f the six methyl protons. Close inspection o f the alkyl methyl peak o f the P I spectrum reveals that it is split into a pentuplet rather than the usual triplet. This could be the result o f the O H being on the second carbon from the end creating a single methine proton.  The splitting o f the methyl peak that  would result would be a doublet that when overlapped with the triplet o f the other methyl protons could yield a pentuplet.  Conversely, the pentuplet could also result from the overlapping o f two slightly misaligned triplet peaks, in which case the hydroxyl need not be on the second carbon from the end. From the QAC-2°OH spectrum it was seen that even though there were two alkyl methyl groups present, only one had protons generating a signal at 0.85 ppm. The other was shifted 104  under the methylene peak.  The deshielding o f the methyl in this case results from both the  nearby hydroxyl and the relatively close nitrogen due to short chain length. If the nitrogen was primarily responsible for the methyl shift, then the O H group in P I could be on the second carbon from the end o f the decyl chain. However, i f the hydroxyl plays a significant role i n shifting the methyl, then it is unlikely that the O H in P I is at this second from the end position. Silverstein (1991) suggests that the p shift (the shift due to deshielding o f a functional group on the adjacent carbon) o f a methyl group influenced by a hydroxyl is 0.33 ppm. This would be added to the value for a regular methyl group (0.85 ppm) to obtain a theoretical chemical shift o f around 1.18 ppm. This suggests that the hydroxylation o f P I is located somewhere in the centre of the alkyl chain.  4.3.2.3 FTIR The chloroform solvent produced a characteristic I R spectrum (Figure 4.11) that was similar to the chloroform spectrum published by Hummel and Scholl (1969). The chloroform peaks could then be subtracted from the other spectra to determine which peaks were due to the sample.  From the N M R spectra obtained previously, it was seen that P I and D D A C are fairly similar compounds. A comparison o f the IR spectrum o f PI to D D A C revealed differences in five peaks. The major difference was the presence o f two moderate peaks in P I at 1097 cm' and 1  1016 cm" , which were absent in the D D A C spectrum (Figure 4.12). The peak at 1097 cm" was 1  1  likely due to absorption from a carbon-oxygen stretch that takes place around 1100 cm" (Bruice, 1  1995).  In fact, this peak corresponds well to the C - O stretch o f a secondary alcohol, which  absorbs at 1110 ± 30 cm"' (Roeges, 1994). Primary alcohols have this same bond stretch at 1045 ± 45 cm" . However, the strong absorption at 1016 cm" is a bit puzzling. The literature mentions 1  1  105  few bonds that produce strong absorptions at this frequency.  One possibility is for CH2-O-P  covalently bound to phosphate.  4000  3600  3200  2800  2400  2000  1800 CM-1  1600  1400  1200  1000  800  650  Figure 4.11. F T I R spectrum o f blank solvent CHCI3 over the wavenumber range o f 4000-650 c m ' . 1  There was also a broad weak peak at 3319 cm" in the P I spectrum that was absent in the 1  D D A C spectrum. This is a characteristic I R frequency at which O - H stretches absorb. A s the OH bond has a large dipole moment, the I R absorption o f this bond is often strong.  Weak O - H  stretch peaks can occur in samples that contain a small amount o f water contamination. However, the intensity o f absorption from the hydroxyl groups in alcohols is also variable depending on hydrogen bonding effects. This peak coupled with that o f the C - O stretch at 1097 cm" strongly suggests that P I possesses a hydroxyl group. 1  106  — y % 75. 70. 65. 60. 55. 50. 45. 40. 35. 30. 25. 20. 15. ] . 5.  - T  4000  3600  3200  2800  2400  2000  1800 CM-1  1  1600  1400  1200  1  1000  f800  650  Figure 4.12. F T I R spectra o f P I overlain with D D A C over a wavenumber range o f 4000-650 cm"'  4000  3600  3200  Figure 4.13. F T I R spectra o f QAC-1°OH overlain with QAC-2°OH over a wavenumber range o f 4000-650 c m . 1  The I R spectra o f the hydroxylated Q A C standards (Figure 4.13) have the characteristic O - H stretch peak around 3300 cm" . For both hydroxylated Q A C s this peak is slightly larger than 1  107  that o f P I , but is not particularly strong. Besides hydrogen bonding effects, the greater intensity of these peaks could result from the greater ratio o f O - H : C - H bonds in the  QAC-OH  compounds relative to P I . Even though all three compounds have a single hydroxyl, there are more C - H bonds in P I (44) than in either Q A C - O H (33) which in a sense can "dilute" the observed absorption for the O - H .  The proximity o f the O H to the nitrogen in both Q A C - O H ' s  may also play an interfering role.  Figure 4.14. F T I R spectra o f P I overlain with Q A C - C O O H over a wavenumber range o f 4000650 cm" . 1  Carbon-hydrogen bonds form weak dipoles and individually produce weak absorptions around 2900 cm" , but their effect is cumulative. The slightly larger peak in the D D A C spectrum 1  is likely due to a greater number o f C - H bond (48) than P I (44). However, another possibility is that there is a slightly higher concentration o f D D A C present in the sample solution than P I . The final difference between D D A C and PI spectra is a small stubby peak present in PI at 1719 cm" , which is absent in D D A C . 1  This frequency is indicative o f a carbon-oxygen double  108  bond, but the intensity is very much weaker.  This can be seen in the Q A C - C O O H spectrum  (Figure 4.14). Carbon-carbon double bonds also produce weak signals near this frequency and this peak could be due to a slightly larger concentration o f B T M A C i n the P I sample.  The  spectrum o f Q A C - C O O H showed no peak from 3200-3600 cm" where a strong absorbtion due to 1  an O - H stretch would be expected.  In addition, the sharp absorption at around 2850 cm"  strongly suggests that Q A C - C O O H  contained an aldehyde rather than a carboxyl group.  1  However, even as an aldehyde, Q A C - C O O H still demonstrated that P I does not contain a C = 0 double bond.  4.4  Conclusions From the small difference in H P L C retention times it was suspected that the metabolite  PI  was not greatly different from D D A C .  P I was successfully isolated from D D A C by  preparative H P L C , and purified further with a benzene extraction. The relatively pure metabolite sample was then analyzed by ' H - N M R and F T I R .  B y comparing the N M R spectrum o f P I with various known Q A C samples, it was discovered that P I was not the result o f terminal oxidation as suspected, nor was it the result o f a carbon-nitrogen fission event.  From the N M R results, hydroxylation along the length o f the  alkyl chain remained a possibility though. F T I R was used to detect the presence o f functional groups within the PI structure. P I contained two regions o f infra-red absorbance at around 3300 and 1100 cm" that were indicative o f a secondary O H group. Taking all data into account, it was 1  concluded that the structure o f P I was the same as D D A C with a hydroxyl group attached somewhere in the centre o f the alkyl chain.  109  CHAPTER 5 FUNGAL  5.1  TOLERANCE  to OXIDIZED  QACs  Introduction From the N M R and F T I R analyses, the metabolite P I appeared to be a Q A C with a  secondary hydroxyl group located near the end o f one o f the decyl chains. It was hypothesized that hydroxylation o f the alkyl chain lowered the toxicity o f the Q A C to fungi, reducing the effectiveness o f D D A C and other Q A C s as preservatives.  Ideally, we would have liked to  synthesize the putative metabolite, but the necessary resources were unavailable. Instead, several available hydroxylated Q A C s were used for agar plate toxicity screening.  The aim o f these  screening experiments was to determine whether or not hydroxylated Q A C s were less toxic, to mould and decay fungi compared to D D A C .  5.2  Methodology  Chemicals:  Three readily available Q A C s containing hydroxyl groups were used to test fungal  tolerance: Q A C - O X I D [cocoalkyl ethoxylated(15) ammonium chloride], a commercial product received bromide],  from and  Akzo  Nobel;  QAC-2°OH  Q A C - l ° O H [dodecyldimethyl-3-hydroxypropylammonium  [dodecyldimethyl-2-hydroxypropylammonium bromide]  were  commercial products obtained from Albemarle.  Q A C - O X I D was chosen as it was originally believed to have hydroxyl groups at the end of long alkyl chains. It was later discovered that Q A C - O X I D was an ethoxylated Q A C , highly oxidized along the alkyl chains. QAC-2°OH was used as it possessed a secondary alcohol group, but on a short alkyl chain. Q A C - l ° O H had the same structure as Q A C - 2 ° O H except that the  110  hydroxyl group was at the end o f the short chain and allowed a comparison between the terminal and secondary alcohol. In addition to the three Q A C s mentioned above, D D A C (Bardac 2280, Lonza) was screened as a reference along with agar control plates to which no Q A C was added.  Fungi:  For the screening test, two moulds, Gliocladium roseum, and Verticillium  bulbillosum; and two basidiomycetes Trametes versicolor, and Postia placenta were used. Agar plates were inoculated with a 3 m m agar core taken from the actively growing mycelium at the edge o f a fungal colony.  Setup:  In addition to a set o f control plates containing no Q A C , agar plates were prepared  with the following concentrations o f the four Q A C s : 100, 500, 1000, 1500 and 2000 ppm. Pyrex bottles (500 ml) were filled with 245 g o f 2 % malt agar (Difco) which was then autoclaved at 121°C for 45 minutes. After sterilization, the agar was cooled i n the laminar flow hood. To avoid the possibility o f thermal degradation, the Q A C s were not autoclaved. Rather, once the agar was sufficiently cool (40-50°C), 5 m l o f an aqueous stock solution o f Q A C was added to the bottles through a sterile syringe filter (0.45 um) to make up the final concentration in the agar. The agar was then mixed in the bottles and poured into disposable Petri plates (Fisher). The plates were left to cool and solidify in the flow bench prior to inoculation. A mycelial core from the growing edge o f a fungal colony from stock culture plates was placed centrally on the Q A C plates.  The inoculated plates were labelled, wrapped in parafilm, and incubated at 25°C. Over a period o f two weeks the fungal growth was measured every two days as the colony diameter. A s not all fungi produced perfectly round colonies, a pair o f straight lines transecting the inoculum plug were drawn on the bottom o f each plate ensuring that the points o f measurement for each  ill  plate were consistent.  Triplicate plates were prepared for each fungus  at each Q A C  concentration and the results were averaged.  This agar plate screening method was used as a quick and simple technique to get a 'sense' for the tolerance o f fungi to the Q A C s o f interest. This test method produced results for fungal tolerance to Q A C s at several concentrations after only two weeks as compared to several months required to get meaningful results from decay fungi i n soil block tests. However, this test only measured the rate at which the fungal colony diameter increases. It did not measure the amount o f fungal biomass produced, nor did it take into account the interactions o f the Q A C and fungus with wood. In addition, not all fungi grow i n uniform colonies and this poses a difficulty in measurement.  5.3  Results and Discussion The data from the measurements o f fungal colony growth over time were compiled into  graphs similar to Figure 5.1.  For each Q A C , six graphs were prepared, one for each  concentration and from these graphs the fungal growth was seen to be generally linear. Linear regressions were carried out for each fungus at each Q A C concentration and from this the growth rate was taken as the slope (mm/day). The growth rates were condensed into Table 5.1 and Figures 5.2 through 5.5. O n the control plates containing no Q A C s , the basidiomycetes grew more rapidly (11.3 and 15.7 mm/day for P. placenta and T. versicolor, respectively) than the moulds (3.0 and 5.3 mm/day for V. bulbillosum and G. roseum, respectively). A s expected, the growth rate o f all fungi decreased as the Q A C concentration in the agar was increased (Table 5.1). N o significant difference in toxicity was observed between Q A C - 2 ° O H and Q A C - l ° O H for each o f the four fungi tested. These two Q A C s , which were hydroxylated on the short alkyl 112  chain, were more toxic than D D A C to the two basidiomycetes tested. A t concentrations as low as 500 ppm these Q A C s prevented any growth o f P. placenta or T. versicolor, while to achieve the same effect using D D A C , a concentration between 1500-2000 ppm was required. However, these two fungi were slightly less inhibited by Q A C - 2 ° O H and Q A C - l ° O H than D D A C .  0 I 0  .  •  1  .  3  6  9  12  — \  15  days  <—G.roseum --"--V.bulbillosum  Figure 5.1.  A  P.placenta —»—T.versicolor  The growth (as measured by colony diameter) o f four fungi on malt agar containing 100 ppm D D A C .  Q A C - O X I D , a Q A C with long, heavily oxidized alkyl chains, was much less toxic than D D A C , QAC-2°OH or Q A C - l ° O H . A l l four fungi showed significant growth at 2000 ppm Q A C O X I D (Figure 5.2) - the highest concentration tested.  V. bulbillosum and G. roseum showed  only a 31 % and 20% respective drop in growth rate relative to the control plates.  113  (ppm) • P. placenta  Figure 5.2.  • V. bulbillosum  • G. roseum  The growth rate o f four fungi over a range (100-2000 ppm) o f concentrations o f QAC-OXID.  • P. placenta  Figure 5.3.  • T. versicolor  • 7. versicolor  • V. bulbillosum  • G. roseum  The growth rate o f four fungi over a range (100-2000 ppm) o f concentrations o f DDAC. 114  (ppm) • P. placenta  Figure 5.4.  • T. versicolor  • V. bulbillosum  • G. roseum  The growth rate o f four fungi over a range (100-2000 ppm) o f concentrations o f QAC-l°OH.  (ppm) • P. placenta  Figure 5.5.  • T. versicolor  • V. bulbillosum  • G. roseum  The growth rate o f four fungi over a range (100-2000 ppm) o f concentrations o f QAC-2°OH. 115  Table 5.1.  The growth rate o f four fungi over a range (0-2000 ppm) o f concentrations o f Q A C l°OH, QAC-2°OH, Q A C - O X I D , and D D A C .  Growth Rate (mm/day) V. bulbillosum  G.  roseum  (ppm) 0 100 500 1000 1500 2000  3.0 2.5 2.4 2.1 1.9 1.6  5.3 4.0 3.3 2.8 2.0 1.4  3.0 2.2 2.1 1.9 1.7 1.5  5.3 4.2 3.6 3.2 2.2 1.7  15.7 3.6 0.0 0.0 0.0 0.0  11.3 0.7 0.0 0.0 0.0 0.0  15.7 5.8 0.6 0.0 0.0 0.0  QAC-OXID 3.0 2.8 2.5 2.4 2.4 2.0  5.3 4.6 4.4 4.3 4.4 4.2  (ppm) 0 100 500 1000 1500 2000  11.3 0.1 0.0 0.0 0.0 0.0  QAC-2°OH  (ppm) 0 100 500 1000 1500 2000  T. versicolor  QAC-1°OH  (ppm) 0 100 500 1000 1500 2000  P. placenta  11.3 .6.2 4.5 2.8 1.5 1.2  15.7 5.9 2.4 1.3 1.1 0.8  11.3 2.8 1.1 0.5 0.2 0.0  15.7 3.1 2.8 1.9 0.4 0.5  DDAC 3.0 2.5 2.1 1.7 1.3 1.0  5.3 3.4 2.3 1.8 1.5 1.1  The other two Q A C s with hydroxylated short chains, were comparable to or more toxic than D D A C . This suggests that the greater the degree o f oxidation along the length o f an alkyl chain, the lower the toxicity o f the Q A C to fungi. This agrees with the concept that the toxicity of Q A C s to fungi is generally due to their strong surfactant nature, which disrupts the fungal cell  wall. Surface active agents exist when strong polar and non-polar moieties are attached within a single molecule. This dual nature causes them to accumulate at surface interfaces, where both the hydrophilic and hydrophobic forces will be satisfied. A n y change to the structure o f a surfactant molecule that would lessen this difference in polarity between these groups will reduce its strength as a surfactant and presumably its toxicity to microbes and aquatic life.  B y attaching a polar  hydroxyl group near the end o f a long hydrocarbon, the fungus is increasing the water solubility o f the Q A C and reducing its activity.  Figure 5.6.  Agar plates containing D D A C and Q A C - O X I D at concentrations o f 500 ppm (left) and 2000 ppm (right), and incubated for 14 days with G. roseum (top), P. placenta (middle), and T. versicolor (bottom).  117  Essentially no difference was observed between the toxicity o f QAC-2°OH and Q A C l ° O H to the four fungi.  These two Q A C s were structurally identical with one exception - the  attachment o f the hydroxyl group i n the former is secondary, whereas in the latter it is terminal. Comparing the toxicity o f the two hydroxylated Q A C s with that o f Q A C - O X I D , it would appear that the location and the degree o f oxidation within the alkyl chain has a greater effect on determining toxicity than does a terminal O H compared with a secondary O H displaced one carbon from the end.  However, it would still be interesting to see i f there was a significant  difference in toxicity between a terminal and a secondary hydroxyl group when attached to a long ( C - C ) alkyl chain. g  16  A s an alternative to directly reducing toxicity, this biotransformation o f D D A C to an oxidized metabolite may indirectly reduce the effectiveness o f D D A C as a wood preservative in the field due to increased leaching. B y hydroxylating the long hydrophobic alkyl chain, it could be reasonably hypothesized that its aqueous solubility would increase. Perhaps this would also weaken the fixation o f the Q A C in wood and lead to increased leaching.  Ruddick (1986) examined the influence o f staining fungi on the decay resistance o f alkyldimethylbenzylammonium chloride. In this test, QAC-treated wood was subjected to a "soup" suspension o f various staining fungi isolated from D D A C treated stakes.  Following  exposure to this soup, some stakes were then leached, while others were left unleached and both sets o f stakes were then exposed to three decay fungi.  Results from this study showed no  significant difference in toxic limits o f QAC-treated wood between the leached and unleached wood that was pre-exposed to staining fungi.  This suggests that the biotransformation by G.  roseum and V. bulbillosum is not increasing the leachability o f D D A C .  118  If the biotransformation o f D D A C caused it to leach more easily from the wood, then the toxicity o f metabolite P I would not be as important a factor in the failure o f D D A C in the field. A more leachable Q A C would be washed from the wood exposing the substrate to wood rotting basidiomycetes. However, these tests have shown that Q A C - O X I D is less capable o f slowing the growth rate o f two mould and two wood rotting fungi. Previous soil block tests showed that the basidiomycetes T. versicolor, P. placenta, and G. trabeum were much more tolerant to Q A C O X I D than D D A C and other non-oxidized Q A C s (Tang, 1994). In addition, other agar screening tests o f Q A C - 2 ° O H and Q A C - l ° O H against G. roseum and V. bulbillosum were in agreement with these tests - slightly less toxic than D D A C to G. roseum and about the same toxicity as D D A C to V. bulbillosum (Ruddick, personal communication).  It appears that the fungal hydroxylation o f D D A C by G. roseum and V. bulbillosum acts to reduce its toxicity. There is a considerable amount o f evidence in the literature that points to hydroxylation as a detoxifying mechanism used by microorganisms, and moulds in particular (Mahato and Majumdar, 1993). Davidse (1976) found that in liquid culture Aspergillus nidulans was hydroxylating methyl benzimidazol-2-yl carbamate ( M B C ) , a toxic breakdown product o f the fungicide benomyl. He found the hydroxylated metabolite to be less toxic than the M B C . Breskvar et al (1995) demonstrated that Rhizopus nigricans was capable o f detoxifying several steroids via a cytochrome-based enzyme complex.  When the enzyme was inhibited by  metyrapore, the fungus became significantly more sensitive to steroids, particularly progesterone and deoxycorticosterone. Hydroxylation has also been noted in plants as a means o f detoxifying herbicides (Owen, 1989).  119  CHAPTER 6 CONCLUSIONS  6.1  &  RECOMMENDATIONS  Conclusions  With large increases in use over the last three decades, there is a much greater potential for Q A C s to escape into the environment. A major use o f Q A C s in British Columbia is a component in wood preservative formulations.  However, due to certain organisms interacting with the  Q A C s , their effectiveness in this field has been limited. Therefore, it is important to understand the interaction o f quaternary ammonium compounds with microorganisms. This thesis focussed on the biotransformation o f didecyldimethylammonium chloride ( D D A C ) by moulds.  In  particular the overall goal was to determine the mechanistic pathway the fungi were using to transform D D A C by isolating and identifying a metabolite from the process.  The ability o f the moulds to degrade D D A C varied considerably depending on the growth conditions. In liquid substrates, V. bulbillosum  grew more rapidly than G. roseum. In addition,  it produced a greater amount o f biomass, could tolerate greater concentrations o f D D A C , and removed more D D A C than G. roseum.  However, in solid substrates such as wood, G. roseum  removed more D D A C and appeared to grow faster than V. bulbillosum.  It is possible that each  fungus has a different mechanism for removing the D D A C in liquid or on solid substrates, but it appears from the metabolite peaks generated on the H P L C chromatograms, that both V. bulbillosum  and G. roseum have the same initial biotransformation mechanism for D D A C i n  wood.  120  The retention time o f the metabolite peak on the chromatogram was close to D D A C suggesting that it was structurally very similar.  The metabolite was successfully isolated by  preparative H P L C and further purified, before being analyzed by N M R and F T I R . The protonN M R spectrum o f the metabolite compound confirmed that it was very similar to D D A C , likely with a hydroxyl group attached to one o f the alkyl chains. The presence o f the hydroxyl group was confirmed by F T I R , and the metabolite was tentatively identified as a Q A C containing two methyl and two decyl chains, with a hydroxyl group attached to one o f the decyl chains somewhere in the middle o f the chain.  Instead o f an intermediate on the way to complete mineralization o f D D A C , it was appeared that the fungus carried out this biotransformation as a detoxification mechanism. This was supported by the literature, which revealed hydroxylation as a common mechanism used by mould fungi to reduce the toxicity o f compounds. To test this concept an experiment was set up to examine the tolerance o f mould and decay fungi to compounds similar in structure to P I . There was only a very small amount o f relatively pure P I that remained after analysis, and the identified compound could not be obtained or synthesized within the scope o f this research project.  Therefore the screening test was carried out on available Q A C s , which served as  models, two that were hydroxylated on short alkyl chains, and one that was oxidized along the length o f long alkyl chains. The toxicity o f the short chained-hydroxylated Q A C s was similar to D D A C ; however, the Q A C with oxidized long alkyl chains was significantly less toxic to all four of the fungi tested than D D A C .  121  6.2  Recommendations One  generated.  o f the major limitations to this research was the small amount o f metabolite A s each purification step involved inherent losses o f the metabolite, a highly pure  sample was not attained. In addition, only non-destructive analysis such as N M R and F T I R were conducted so that the same, relatively pure sample could be re-used.  A n y future work carried out on the biodegradation o f Q A C s by mould fungi should begin with obtaining a relatively large amount o f purified metabolite.  This would require the  development o f a large scale-up, perhaps i n a fermentor with a continuous fungal culture (as opposed to the batch cultures used here) in order to be feasible. Once a large amount o f pure metabolite is available, a number o f powerful analytical techniques such as  l 3  C - N M R and Mass  Spectrometry can be used to positively identify the compound:  Another approach could be to synthesize compounds similar to the suspected metabolite and use them for comparative analysis. A series o f closely related compounds to the D D A C breakdown product could be used for further tolerance testing to get an idea o f how the oxidized Q A C structure relates to fungal toxicity.  The metabolite P1 was obtained solely from the fungal alteration o f D D A C that had been used to treat wood. Metabolites were not successfully isolated from the liquid cultures.  The  development o f a method for isolating and purifying metabolites from the liquid cultures could provide insight into differences in how the fungi deal with D D A C in a liquid matrix as opposed to wood.  122  Literature Cited Ahlstrom, B . , M . Chelminska-Bertilsson, R . A . Thompson and L . Edebo. 1997. Submicellar complexes may initiate the fungicidal effects o f cationic amphiphilic compounds on Candida albicans. Antimicrob. Agents Chemother. 41:544-550. Alexander, M . 1994. Biodegradation and Bioremediation. Academic Press, Inc. CA  San Diego,  Anderson, B . M . and M . L . Reynolds. 1965. Inhibition o f yeast alcohol dehydrogenase by alkylammonium chlorides. B i o c h i m . Biophys. Acta. 96:45. Barbaro, R. D . and Hunter, J. V . 1965. biodegradability. Purdue Conf. 20: 189-196.  The effect  o f clay minerals  BCMELP. 1994. Antisapstain Chemical Waste Control Regulation. Environment, Lands and Parks. Victoria, B . C .  on  surfactant  B . C . Ministry o f  Bezalel, L . , Y . Hadar, P.P. F u , J.P. Freeman and C . E . Cerniglia. 1996. Initial oxidation products in the metabolism o f pyrene, anthracene, fluorene, and dibenzothiophene by the white rot fungus Pleurotus ostreatus. A p p l . Environ. Microbiol. 62:2554-2559. Biesinger, K . E . and G . N . Stokes. 1986. Effects o f synthetic polyelectrolytes on selected aquatic organisms. J. Water Pollut. Control Fed. 58:207-213. Boethling, R. S. 1984. Environmental fate and toxicity i n wastewater treatment o f quaternary ammonium surfactants. Water Res. 18: 1061-1076. Bogan, B . W . and R . T . Lamar. 1995. One-electron oxidation in the degradation o f creosote polycyclic aromatic hydrocarbons by Phanerochaete chrysosporium. 61:2631-2635. Breskvar, K . , Z . Ferencak, and T. Hudnik-Plevnik. 1995. The role o f cytochrome P45011ain detoxification o f steroids in the filamentous fungus Rhizopus nigricans. J. Steroid Biochem. Molec. B i o l . 52:271-275. Brock, T . D . and M . T. Madigan. 1991. Biology o f Microorganisms. Prentice H a l l : N e w Jersey. (55-64,562-610,753-756). Bruice, P . Y . 1995. Organic Chemistry. Prentice-Hall. Englewood Cliffs, N . J. 675.  537-554,633-  B u l l , J.P. 1997. Immunoassays for two quaternary ammonium compounds, benzyldimethyldodecylammonium chloride ( B D D 1 2 A C ) and didecyldimethylammonium chloride ( D D A C ) . MSc.-thesis, University o f British Columbia, Faculty o f Forestry, Dept. W o o d Science. Burgel, J.L., Dubois, J.W. and Ruddick, J.N.R. 1996a. Bioremediation o f surfactant contaminated waste. Internat. Res. Group on W o o d Preserv. Doc. N o . IRG/WP/96-50070. Burgel, J.L., Dubois, J.W. and Ruddick, J.N.R. 1996b. Proposed degradation pathway for quaternary ammonium compounds by mould fungi. Internat. Res. Group on W o o d Preserv. Doc. N o . IRG/WP/96-10166. 123  Butcher, J . A . 1979. Testing new preservatives for protection o f wood exposed in above ground situations. Mat. Org. 14:43-53. Butcher, J.A., A . F . Preston and J. Drysdale. 1979. Potential o f unmodified and copper-modified alkylammonium compounds as groundline preservatives. N e w Zealand J. o f For. S c i . 9:348358. Butcher, J . A . and A . F . Preston. 1978. Toxicity o f tertiary amine acetates against basidiomycetes and soft-rot fungi. N e w Zealand J. o f For. Sci. 8:397-402. Butcher, J . A . and J. Drysdale. 1977. Relative tolerance o f seven wood-destroying basidiomycetes to quaternary ammonium compounds and copper-chrome-arsenate preservative. Mat. Org. 12:271-277. Butcher, J . A . , A . F . Preston and J. Drysdale. 1977a. Initial screening trial o f some quaternary ammonium compounds and amine salts as wood preservatives. For. Prod. J. 27:19-22. Butcher, J . A . , M . E . Hedley and J. Drysdale. 1977b. Comparison o f a quaternary ammonium compound and copper-chrome-arsenate as wood preservative. For. Prod. J. 27:22-25. Cain, R . B . 1976. Surfactant biodegradation in waste waters. C h . 18. Treatment o f industrial effluents. ( A . G . Callely, C . F . Forster, and D . A . Stafford, Eds) Wiley, N e w York. Calas, M . , G . Cordina, J. Bompart, M . Benbari, T. Jei, M . L . Ancelin and H . V i a l . 1997. Antimalarial activity o f molecules interfering with Plasmodium falciparum phospholipid metabolism-structure-activity relationship. J. M e d . Chem. 40:3557-3566. Chaplin, C . E . 1951. 29:373-382.  Observations on quaternary ammonium disinfectants.  Can. J. Botany.  Cooper, J.C. 1988. Review o f the environmental toxicity o f quaternary ammonium halides. Ecotoxicol. Environ. Safety. 16:65-71. Cuta, J. and Hanusova, J. 1964. Hygienic problems o f detergents. I V . Biochemical oxidation o f anionic and nonionic surfactants. Cesk. H y g . 9:507-516. Davidse, L . C . 1976. Metabolic conversion o f methyl benzimidazol-2-yl carbamate ( M B C ) in Aspergillus nidulans. Pest. Biochem. Physiol. 6:538-546. Dean-Raymond, D . and Alexander, M . 1977. Bacterial metabolism o f quaternary ammonium compounds. A p p l . Env. Microbiol. 33(5): 1037-1041. Doyle, A . K . 1995. Factors which influence the performance o f alkylammonium compounds as wood preservatives. Ph.D.-thesis, University o f British Columbia, Faculty o f Forestry, Dept. Wood Science. Doyle, A . K . and Ruddick J.N.R. 1993. Isolation and identification o f non-decay fungi affecting the performance o f alkylammonium compounds. Internat. Res. Group on Wood Preserv. Doc. N o . IRG/WP/93-10020. Dubois, J.W. and J.N.R. Ruddick. 1998. The fungal degradation o f quaternary ammonium compounds in wood. Internat. Res. Group on Wood Preserv. D o c . N o . IRG/WP/10263.  124  Dubois, J.W. and J.N.R. Ruddick. 1997. Aspects o f the fungal degradation o f quaternary ammonium compounds in liquid culture. Internat. Res. Group on Wood Preserv. Doc. N o . IRG/WP/30160. Fenger, B . H . , Mandrup, M . , Rohde, G . and Sorenson, J. C . K . 1973. Degradation o f a cationic surfactant in activated sludge pilot plants. Water Res. 7: 1195-1208. Fujiya, T. and Koga, S. 1966. The binding o f a cationic detergent by yeast cells, in relation to its germicidal action. J. Gen. A p p l . M i c r o b i o l . 12: 229-237. Games, L . M . , K i n g , J. E . and Larson, R. J. 1982. Fate and distribution o f a quaternary ammonium surfactant, octadecyltrimethylammonium chloride, in wastewater treatment. Env. S c i . Technol. 16:483-488. Gawel, L . J. and Huddleston, R. L . 1972. The biodegradability o f low concentrations o f quaternary ammonium antimicrobials by bacteria. Presented at the A O C S Nat. Meeting, L o s Angeles. 22-23 A p r i l . Gerike, P., Fischer, W . K . and Jasiak, W . 1978. Surfactant quaternary ammonium salts i n aerobic sewage digestion. Water Res. 12: 1117-1122. Ghisalba, O. and Ktienzi, M . 1983. Biodegradation and utilization o f quaternary alkylammonium compounds by specialized methylotrophs. Experientia. 39: 1264-1271. Ginkel, C . G . van and Kolvenbach, M . 1991. Relations between the structure o f quaternary alkyl ammonium salts and their biodegradability. Chemosphere 23: 281-289. Ginkel, C . G . van., V a n Dijk, J. B . and Kroon, A . G . M . 1992. Metabolism o f hexadecyltrimethylammonium chloride i n Pseudomonas Strain B l . A p p l . Envir. Microbiol. 9: 3083-3087. Ginkel, C . G . van, C A . Stroo, and A . G . M . Kroon. 1993. Biodegradability o f ethoxylated fatty amines and amides and the non-toxicity o f their biodegradation products. Tenside Surfact. Deterg. 3:390-398. Ginkel, C . G . and A . G . M . Kroon. 1993. Metabolic pathway for the biodegradation octadecylbis(2-hydroxyethyl)amine. Biodegradation. 3:435-443.  of  Ginkel, C . G . van, M . A . Pomper, C A . Stroo, and A . G . M . Kroon. 1995. Biodegradability o f fatty amines; utilization o f the alkyl chain by isolated microorganisms. Tenside Surfact. Deterg. 3:355-359. Ginkel, C . G . van. 1996. Complete degradation o f xenobiotic surfactants by consortia o f aerobic microorganisms. Biodegradation. 7:151-164. Hampton, D . and Zatman, L . J. 1973. The metabolism o f tetramethylammonium chloride by bacterium 5H2. Biochem. Soc. Trans. 1: 667-668. Hedley, M . L . , K . Tsunoda and K . Nishimoto. 1982. Evaluation o f alkylammonium compounds for use as low toxicity wood preservatives in Japan. Wood S c i . 68:37-46. Henderson, N . D . 1992. A review o f the environmental impact and toxic effects o f D D A C . Report Prepared for Environmental Division, B . C . Ministry o f Environment, Lands and Parks, Victoria, B . C . 125  His, E . , R. Beiras, F. Quiniou, A . C . S . Parr, M . J . Smith, M . J . Cowling and T. Hodgkiess. 1996. The non-toxic effects o f a novel antifouling material on oyster culture. Wat. Res. 30:2822-2825. Huber, L . 1979. The behaviour o f cation-active surfactants from laundry softeners in water and wastewater. Munch. Beitr. Z . Abwass. Fisch. F l u . B i o l . 31:203-215. Hughes, R . D . , P. Millburn, and R . T . Williams. 1973. Molecular weight as a factor in excretion of monoquaternary ammonium cations in the bile o f rat, rabbit and guinea pig. Biochem. J. 136:967. Isomaa, B . 1975. Absorption, distribution and excretion o f [ 1 4 C ] C T A B , a ammonium surfactant i n the rat. Cosmet. Toxicol. 13:231-237.  quaternary  Janicke, W . and Hilge, G . 1979. Biodegradation behavior o f anionic-cationic surfactant complexes under aerobic and anerobic conditions in sewage or sludge treatment. Tenside 16: 117-122. Johnston, B . D . , J . M . Seubert and C . J . Kennedy. 1998. Biochemical effects o f didecyldimethylammonium chloride ( D D A C ) exposure and osmoregulatory stress on juvenile coho salmon, Onchorynchus kisutch. A r c h . Environ. Contam.Toxicol. 34:275-279. Kappeler, T. U . 1982. The aquatic toxicity o f distearyldimethylammonium chloride ( D S D M A C ) and its ecological significance. Tenside Deterg. 19: 169-176. Kendrick, B 1985. The Fifth Kingdom. Mycologue Publications. Waterloo, O N . 21-32 Kennes, C . and J . M . Lema. 1994. Simultaneous biodegradation o f p-cresol and phenol by the basidiomycete Phanerochaete chrysosporium. J. Indust. Microbiol. 13:311-314. Kiehlmann, E . , L . Pinto, and M . Moore. 1996. The biotransformation o f chrysene to trans-1,2dihydroxy-l,2-dihydrochrysene by filamentous fungi. Can. J. M i c r o b i o l . 42:604-608. Knauf, W . 1973. Determination o f detergent toxicity to water organisms. Tenside Deterg. 10: 252-255. Knezovich, J.P., M . P . Lawton and L . S . Inouye. 1989. Bioaccumulation and tissue distribution of a quaternary ammonium surfactant in three aquatic species. B u l l . Environ. Contam. Toxicol. 42:87-93. Kroon, A.G.M., Pomper, M . A . and van Ginkel, C . G . 1994. Metabolism dodecyldimethylamine by Pseudomonas M A 3 . A p p l . Microbiol. Botechnol. 42:134-139.  of  Krzeminski, S. F., Martin, J. J. and Brackett, C . K . 1973. The environmental impact o f a quaternary ammonium bactericide. Household Pers. Prod. Ind. 10(3): 22-24. Krzysko-Lupicka, T., W . Strof, K . Kubs, M . Skorupa, P. Wieczorek, B . Lejczak, and P. Kafarski. 1997. The ability o f soil-borne fungi to degrade organophosphonate carbon-to-phosphorus bonds. A p p l . Microbiol. Biotechnol. 48:549-552. Larson, R . J . 1983. Comparison o f biodegradation rates in laboratory screening studies with rates in natural waters. Residue Rev. 85:159-171. Larson, R. J. and Vashon, R. D . 1983. Adsorption and biodegradation o f cationic surfactants in laboratory and environmental systems. Devs. Ind. Microbiol. 24: 425-434. 126  Lawrence, C . A . 1970. Germicidal properties o f cationic surfactants. Cationic surfactants. (E. Jungermann, Ed) pp 491-426. Marcel Dekker, N e w York. Lewis, M . A . and Wee, V . T. 1983. Aquatic safety assessment for cationic surfactants. Envir. Toxic. Chem. 2: 105-118. MacGregor, D . R . and P . R . Elliker. 1958. A comparison o f some properties o f strains o f Pseudomonas aeruginosa sensitive and resistant to quaternary ammonium compounds. Can. J. Microbiol. 4:499-503. Mackrell, J. A . and Walker, J. R. L . 1978. The biodegradation o f quaternary ammonium compounds. Int. Biodeterior. B u l l . 14(3): 77-83. MacQuillen, K . 1950. The bacterial surface. 1. Effect o f cetyltrimethyl-ammonium bromide on the electrophoretic mobility o f certain Gram-positive bacteria. Biochem. Biophys. Acta. 5: 463. Mahato, S.B., and I. Majumdar. 1993. Current Trends in Microbial Steroid Biotransformation. Phytochem. 34:883-989. Masuda R., Machida, S. and Kanno, M . 1976. Studies on the biodegradability o f some cationic surfactants. Surfactant Congr. 7:129-138. Montenecourt, B . S . and D . E . Eveleigh. 1977. Semiquantitative plate assay for determination o f cellulase production by Trichoderma viride. A p p l . Env. Microbiol. 33:178-183. Morris, P.I. and J. Ingram. 1988. Stake test o f wood preservatives at Westham Island, 1987 data. Forintek Canada Corp. Report: 87-02-17-10-E051, Vancouver, B . C . Nicholas, D . D and A . F . Preston. 1980. Evaluation o f alkylammonium compounds as potential wood preservatives. Proc. A m . Wood Pres. Assn. 76:13-21. Nishiyama, N . , Toshima, Y . and Ikeda, Y . 1995. Biodegradation o f alkyltrimethyl- ammonium salts in activated sludge. Chemosphere. 30(3):593-603. Owen, W . J . 1989. Metabolism o f herbicides-detoxification as a basis o f selectivity. In Herbicides and Plant Metabolism. A . D . Dodge (ed.). Society for Experimental Biology Seminar Series 38. Cambridge University Press. 171-196. Pessoa, S.B. 1952. Notes on some molluscicidal substances. Folia clin. biol. 18:137-141. Preston, A . F . , P.J. Walcheski, P A . M c K a i g and D . D . Nicholas. 1987. Recent research on alkylammonium compounds in the U . S . Proc. A m . W o o d Pres. Assn. 83:331-348. Preston, A . F . 1983. Dialkyldimethylammonium halides as wood preservatives. Chem. Soc. 60:567-570.  J. o f A m . O i l  Preston, A . F . and D . D . Nicholas. 1982. Efficacy o f a series o f alkylammonium compounds against wood decay fungi and termites. Wood and Fiber 14:37-42. Reuvers, B . A . , G . Ortiz, M . Ramos, and M . M . De Pozuelo. 1989. Rapid high-performance liquid chromatographic methods for the determination o f bencetonium chloride residues in fish products; confirmation by thin-layer chromatography. Roeges, N . P . G . 1994. A Guide to the Complete Interpretation o f Infrared Spectra o f Organic Structures. John Wiley and Sons Ltd. West Sussex, England. P. 66, 128-129, 261. 127  Ruddick, J.N.R. 1987. A field evaluation o f modified and unmodified alkylammonium compounds. Internat. Res. Group on Wood Preserv. Doc. N o . IRG/WP/3436. Ruddick, J.N.R. 1986. The Influence o f Staining Fungi on the Decay Resistance o f W o o d Treated with Alkyldimethylbenzylammonium Chloride, Mat. u. Org. 21:139-149 Ruddick, J.N.R. 1983. Field testing o f alkylammonium wood preservatives. Group on Wood Preserv. Doc. N o . IRG/WP/3248. Ruddick, J.N.R. 1981. Testing o f alkylammonium compounds. Preserv. Doc. N o . IRG/WP/2152. 12p.  Internat. Res.  Internat. Res. Group on W o o d  Ruddick, J.N.R. and A . R . H . Sam. 1982. Didecyldimethylammonium chloride — a quaternary ammonium wood preservative: its leachability and distribution in four softwoods. Mat. Org. 17:299-313. Ruiz Cruz, J. and Dobarganes Garcia, M . C . 1979. Pollution o f natural waters by synthetic preservatives. X V . Relation between structure and biodegradation o f cationic surfactants in river water. Grasas Aceites 30:67-74. Silver, S., Chakrabarty, A . M . , Iglewski, B . , and S. Kaplan. (Eds.) 1990. Pseudomonas: Biotransformations, Pathogenesis, and Evolving Biotechnology. American Society for Microbiology, Washington, D . C . Silverstein, R . M . , G . C . Bassler and T . C . M o r r i l l . 1991. Spectrometric Identification o f Organic Compounds. Wiley, N e w York. Skoog, D . A . and J.J. Leary. 1992. Principles o f Instrumental Analysis (4th ed.). College Publishing. Forth Worth, T X .  Saunders  Smith, R . S . 1967. Carbon dioxide evolution as a measure o f attack o f wood by fungi and its application to testing wood preservatives and sapstain preventives. A n n . o f A p p l . B i o l . 59:473479. SRI. 1995. Chemical Economics Handbook. Park, C A .  Stanford Research Institute International, M e n l o  Steffann, S., J.P. Joly, B . Loubinoux, and P. Dizengremel. 1988. Effects o f an alkylammonium bromide on oxidative and phosphorylative properties o f mitochondria isolated from Agaricus bisporus. Pest. Biochem. Physiol. 32:38-45. Stryer, L . 1988. Biochemistry (3rd ed). W . H . Freeman and Company. N e w York. Sullivan, D . E . 1983. Biodegradation o f cationic surfactants in activated sludge. Water Res. 17: 1145-1151. Swisher, R. D . 1987. Surfactant Biodegradation. Volume 18. Second Edition. Marcel Dekker, N e w York. Ueno, M . and F. Yokoya. 1996. Inhibition o f microbial metabolism by C T A B (cetyltrimethylammonium bromide) and its effect on cytoplasm enzymes. Rev. M i c r o b i o l . 27:171-176.  128  Taft, C . H . 1946. Some observations on the effects o f a mixture o f high molecular alkyldimethylbenzylammonium chlorides on various invertebrates. Tex. Rep. B i o l . M e d . 4: 27-34. Tang, H . and J.N.R. Ruddick. 1994. Evaluating the potential o f amine chemicals for use as wood protecting agents. Part 1. Investigation o f cation component o f quaternary ammonium compounds. Internat. Res. Group on W o o d Preserv. Doc. N o . IRG/WP/94-30049. 7p. Taylor, R . T . and H . F . Schoof. 1967. The larvicidal activity o f several liquid detergents and quaternary ammonium compounds. Mosquito News. 27:486-487. Theng, B . K . G . 1974. The Chemistry o f Clay-Organic Reactions. Wiley, N e w York. Tsoumis, G . 1968. Wood as a Raw Material: source, structure, chemical composition, growth, degradation, and identification. Permagon, Oxford. 276 p. U S I T C . 1993. Synthetic Organic Chemicals. Publication 2607.  United States International Trade Comission  Valladao, M . and W . E . Sandine. 1994. Quaternary ammonium compounds in milk: detection and reverse-phase high performance liquid chromatography and their effect on starter growth. Dairy Sci. 77:1509-1514. Vallejo-Freire, A . , O.F. Ribeiro, and I.F. Ribeiro. 1954. Quaternary ammonium compounds as molluscacides. Science 119:470-472. Ventullo, R . M . and R . J . Larson. 1986. Adaptation o f aquatic microbial communities to quaternary ammonium compounds. A p p l . Environ. M i c r o b i o l . 51:356-361. Vogel, H . J . 1956. A convenient growth medium for Neurospora (medium N ) . M i c r o b . Gen. B u l l . 13:42-43. Waller, D . L . , S.W. Fisher and H . Dabrowska. 1996. Prevention o f zebra mussel infestation and dispersal during aquaculture operations. Prog. Fish Cult. 58:77-84. Waters, J. 1982. Addendum to the paper on "the aquatic toxicology o f D S D M A C and its ecological significance". Tenside Deterg. 19:177 Wood, A . W . , B . D . Johnston, A . P . Farrell, and C . J . Kennedy. 1996. Effects o f didecyldimethylammonium chloride ( D D A C ) on the swimming performance, gill morphology, disease resistance, and biochemistry o f rainbow trout (Onchorynchus mykiss). Can. J. Fish. Aquat. Sci. 53:2424-2432. Wunder, T., J. Marr, S. Kremer, O. Sterner, and H . Anke. 1997. 1-Methoxypyrene and 1,6dimethoxypyrene: two novel metabolites in fungal metabolism o f polycyclic aromatic hydrocarbons. A r c h . Microbiol. 167:310-316. Zheng, Y . 1995. The biodegradation o f quaternary ammonium compounds by mould fungi. Masters-thesis, University o f British Columbia, Faculty o f Forestry, Dept. W o o d Science. Zheng, Y . and Ruddick, J.N.R. 1995. The effect o f didecyldimethylammonium chloride on the growth o f different strains of the mould fungus Gliocladium roseum. Internat. Res. Group on Wood Preserv. Doc. N o . IRG/WP/95-10105  129  APPENDIX I  Complete 'H-NMR Spectra  130  131  133  134  135  Spectrum #6: OAC-l°OH  Spectrum #7: OAC-2°OH  137  A P P E N D I X II  Integration and Proton Counts of N M R Spectra  138  INTEGRATION O F T H E NMR S P E C T R A Spectrum # 1:  DDAC  1 — i  lpeak*htt(mm)J 0.8-0.9  26  3  1.2- 1.5  120  14  1.5-1.8  17  2  3 . 3 - 3.6  43  5  6 28 4 10  Spectrum # 2 : B T M A C g s h i f t r i p p m ) ! ipeakinti(mm)] i o w e s t p a t i o l 3.3-3.5 149 9  9  5.0-5.1  33  2  7.35-7.5  50  3  3  7.6-7.7  33  2  2  2  S p e c t r u m # 3: M i x t u r e ( D D A C 5 0 % a n d B T M A C 5 0 % ) W M t H M M i I » l i i t t l d dHjCpunt lpeaKtntf(mm)I 0.8-0.9 3 IP"" 17 1.2- 1.5 14 28 1.5- 1.8 3 . 3 - 3.6  81  12  3.3-3.5  29 29  5.0-5.1  58  7.35-7.5 7 . 6 - 7.7  7 10 7  2  4  5  10  9  9  2  2  3  3  2  2  Interpretation of Peak Integration: T h e height o f e a c h p e a k i s p r o p o r t i o n a l t o t h e n u m b e r of p r o t o n s r e s o n a t i n g at that shift (frequency).  T h e r e f o r e , if t h e n u m b e r of p r o t o n s in a m o l e c u l e i s k n o w n f o r a p a r t i c u l a r shift  (e.g. C H - m e t h y l p r o t o n s at 0 . 8 - 0 . 9 p p m ) , t h e n t h e n u m b e r of p r o t o n s p r e s e n t i n t h e o t h e r p e a k s 3  c a n be determined a s they are proportional. EXAMPLE: In s p e c t r u m # 1 ( D D A C ) , t h e integration of t h e m e t h y l proton p e a k at 0 . 8 - 0 . 9 p p m i s 2 6 m m . k n o w n that t h e r e a r e t w o a l k y l m e t h y l g r o u p s in D D A C that c o n t a i n a total of 6 protons. Therefore the height/proton of the integration for this s p e c t r u m is: 2 6 m m / 6 protons = 4 . 3 mm/proton.  F r o m this the other protons counts c a n be determined.  [ T h e p e a k at 1 . 2 - 1 . 5 p p m i s 1 2 0 m m / 4 . 3 m m / p r o t o n = 2 8 protons] Spectrum # 5 :  QAC-COOH  p e a k l h t l m m ) ! IfowestiKatios i l f l p o u n t f  d!8-o.sT~  8  3  3  1.2-1.5  96  36  36  1.5-1.8  21  8  8  2.3-2.4  5  2  2  3.3-3.4  16 11  6 4  6  3.4-3.6  139  4  It i s  Spectrum #4: P1 (+ BTMAC as contamination)  liowestiRatio 0.8-0.9 1.2-2.0 3.3-3.7  15 74 30  2 10 4  6 30 12  3.3-3.7  35  9  9  2 3 2  2 3 2  65 5.0-5.1 7.35-7.5 7.6-7.7  8 12 8  Determination of Proton Count in P1 In cases where two different molecules are present in a sample, the NMR spectrum may contain overlapping peaks in which the protons from each compound contribute to the integration. If one knows the proton count of a peak caused by only one of the compounds, the integration per proton can be calculated for this compound. In the case of spectrum # 4, where the structure of one of the compounds is known (BTMAC), the height of the overlapping peak at 3.3-3.7 ppm due to the BTMAC protons (35 mm) can be subtracted from the overall height (65 mm) to determine the integration due to the protons of the unknown compound P1 (65-35 = 30 mm). As the different compounds are present in the sample at different concentrations, the height per proton for each compound will be different. For the remaining integration of the P1 sample, it is known that the peak at 0.8-0.9 ppm is due to alkyl methyl protons (R-CH3). This leaves the options of 3 or 6 protons. If only 3 protons were present at this peak then the integration per proton would be [15 mm/ 3 protons = 5 mm/proton]. This would cause the other peaks to contain only 15, and 6 protons respectively. However, it is difficult to fit these protons into the shift at which they occur (e.g. 6 protons on C adjacent to N instead of 10). If two alkyl methyl groups were present with 6 protons the integration per proton would be 2.5 mm/proton which provides a much more reasonable fit.  Spectrum #6: ^ shift (ppm) 0.8-0.9 1.2-1.4 1.7-1.8 2.0-2.1 3.2-3.3 3.4-3.5 3.7-3.8 4.1-4.2  QAC-1°OH  peak ht (mm) Lowest Ratio 12 3 72 18 2 8 2 8 24 6 12 3 4 16 4 1  1  H Count 3 18 2 2 6 3 4 1  Spectrum # 7: QAC-2°OH ishiftfppm) 0.8-0.9 1.2-1.4 1.6-1.8 3.3-3.4 3.4-3.5 3.5-3.6 4.5  1 L, /  V  Itfeak|htf(mm)} 14 95 9 27 9 9 5 140  3 21 2 6 2 2 1  3 21 2 6 2 2 1  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

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"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0089094/manifest

Comment

Related Items