UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The identification and natural clonal variation of important wood extractives in populus tremuloides Fernandez, Marc Phillip 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_2000-0055.pdf [ 6.81MB ]
Metadata
JSON: 831-1.0089333.json
JSON-LD: 831-1.0089333-ld.json
RDF/XML (Pretty): 831-1.0089333-rdf.xml
RDF/JSON: 831-1.0089333-rdf.json
Turtle: 831-1.0089333-turtle.txt
N-Triples: 831-1.0089333-rdf-ntriples.txt
Original Record: 831-1.0089333-source.json
Full Text
831-1.0089333-fulltext.txt
Citation
831-1.0089333.ris

Full Text

THE IDENTIFICATION AND NATURAL CLONAL VARIATION , OF IMPORTANT WOOD EXTRACTIVES IN POPULUS TREMULOIDES by MARC PHILIP FERNANDEZ B.Sc, The University of Victoria, 1997 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIRMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES (Faculty of Forestry) (Department of Wood Science) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA December 1999 © Marc Philip Fernandez, 1999 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of The University of British Columbia Vancouver, Canada DE-6 (2/88) ABSTRACT A rapid method for the detailed compositional analysis (70 compounds) of quaking aspen (Populus tremuloides Michx.) wood extractives was developed to monitor the differences in these extractives between natural aspen clones. The method involves the removal of increment cores from standing trees, soxhlet extraction of the sampled wood with acetone, and gas chromatography / mass spectrometry (GC-MS) of the resultant extractives. Additionally, preparative chromatographic techniques were developed and employed in order to elucidate the identity of various steryl esters that co-eluted with other components and did not provide characteristic mass spectra by GC-MS. Furthermore, comparisons between high-performance liquid chromatography (HPLC) and GC-MS were made on the basis of suitability for the analysis of high molecular weight and industrially problematic steryl esters. Genetic variation in aspen wood extractives that potentially impact on the utilization of this species for pulp and paper was sought using natural clone tests. Significant (95% confidence) interclonal variation in several wood extractives was found by analysis of variance in the extractives data from nine natural clones. The interclonal variation of biosynthetically related groups was similar, and thus, the total amounts of these groups showed more significant interclonal variation than the individual components. Significant clonal differences were found in the sterols / triterpenes, steryl esters / waxes and triglycerides which are all known pitch culprits in pulp and paper making. Also, extractive compounds known to exhibit toxic effects in aquatic organisms, showed significant differences between natural aspen clones. However, in ii some cases (ie flavonoids) these compounds formed a part of the defense system in the living tree and thus, were affected by the presence of wood decay measured in the 10mm cores. This decay affected the levels of various phenolic extractives in this study and thus, increased the intraclonal variance such that interclonal differences in these phenolics were not significant. The assessment of bound (glycoside) salicylic acid and its precursors, benzoic acid and o-coumaric acid, as a measure of decay resistance, was found to be affected by the presence of decay and thus, was not possible in mature aspen stands where decay is prevalent. iii TABLE OF CONTENTS Page ABSTRACT ii LIST OF TABLES viii LIST OF FIGURES ix TABLE OF ABBREVIATIONS xiii ACKNOWLEDGEMENTS xvi INTRODUCTION 1 1.1 General Introduction 1 1.2 Quaking Aspen 1 1.2.1 Biology of Quaking Aspen 1 1.2.2 Measuring Genetic Variability in Quaking Aspen 3 1.3 Utilization of Aspen in the Pulp and Paper Industry 5 1.3.1 Desirable Feature of Aspen for Pulp and Paper 5 1.3.2 Limits on the Utilization of Aspen for Pulp and Paper 6 1.4 Wood Extractives 8 1.4.1 Wood Extractives in Hardwoods 8 1.4.2 Biosynthesis of Wood Extractives 9 1.5 Wood Extractives and Limits on the Utilization of Aspen 11 1.5.1 Wood Resin and Pitch 11 1.5.2 Wood Extractives and Mill Effluent Toxicity 12 1.5.3 Salicylic Acid and Aspen Decay 13 1.6 Modern Methods for Analysis of Wood Extractives 14 1.6.1 Sampling and Extraction 14 1.6.2 Chromatographic Techniques for the Analysis of Wood Extractives 15 1.6.3 Analysis of Steryl Esters 18 1.6.4 Analysis of Salicylic Acid Glycoside 20 1.7 Objectives 21 DEVELOPMENT OF ANALYTICAL METHODS 21 2.1 Introduction 21 2.2 Materials and Methods 23 2.2.1 Origin of Samples Used for Method Development 23 2.2.2 Sample Preparation and Soxhlet Extraction 24 2.2.3 Determination of Total Weight of Extractives 24 2.2.4 GC-MS Analysis of Total Extractives 25 2.2.5 GC-FID Analysis of Total Extractives 27 2.2.6 Solid-Phase Extraction to Isolate SE/W Fraction 27 2.2.7 Argentation Chromatography Fractionation of SE/W Fraction 28 2.2.8 HPLC Analysis of Steryl Esters 29 2.2.9 GC-MS Analysis of SE/W Fraction 29 IV TABLE OF CONTENTS (CONT'D) Page 2.2.10 Standard Compounds 30 2.2.11 Assessment of Potential Sources of Contamination 30 2.2.12 Analysis of Salicylic Acid 31 2.2.12.1 P-Glucosidase Hydrolysis 31 2.2.12.2 Trifluoroacetic Acid Hydrolysis 32 2.2.12.3 Monosaccharide Analysis 32 2.2.13 Calculations and Statistics 33 2.3 Results and Discussion 34 2.3.1 Sample Preparation and Extraction 34 2.3.1.1 Freeze Drying 34 2.3.1.2 Soxhlet Extraction 35 2.3.1.3 Repeatability of 10 mm Core Measurements 39 2.3.2 Extractives Analysis Techniques 39 2.3.2.1 Analysis of Total Extractives by GC-MS 39 2.3.2.2 Development of a HPLC method for Steryl Esters 42 2.3.2.3 Analysis and Possible Origin of Salicylic Acid in Aspen 48 2.3.3 Identification and Quantification of Compounds by GC-MS 52 2.3.3.1 Identification of Extractive Compounds by GC-MS 52 2.3.3.2 Identification of Steryl Esters by AC-GC-MS 56 2.3.3.3 A HPLC pre-GC Fractionation Technique for Steryl Esters 68 2.3.3.4 Quantification of Extractive Compounds by GC-MS 71 2.3.4 Validation of Underivatized Extractives Analysis by GC-MS 77 2.3.4.1 Reproducibility of GC-MS Method for Underivatized Extractives 77 2.3.4.2 Comparison to Silylated (TBDMS) Extractives 79 2.3.4.3 Comparison to Flame Ionization Detection 82 2.4 Summary 84 2.4.1 Development of Analytical Methods 84 2.4.2 Identification and Quantification of Compounds 85 2.4.3 Validation of the GC-MS Method for Underivatized Extractives 86 CLONAL VARIATION OF WOOD EXTRACTIVES 87 3.1 Introduction 87 3.2 Materials and Methods 88 3.2.1 Clone Selection and Sampling 88 3.2.2 Site Assessments 92 3.2.3 Assessment of Decay 92 3.2.4 Sample Extraction and Total Extractives Analysis 95 3.2.5 Analysis of Salicylic Acid 95 3.2.6 Statistical Calculations and Analyses 95 3.3 Results and Discussion 96 3.3.1 Assumption of Equal Variance and Clonal Variation of Total Acetone Extractives 96 V TABLE OF CONTENTS (CONT'D) Page 3.3.2 Composition of the Acetone Extractives from Aspen 10 mm Cores 99 3.3.3 Interclonal Variation in Components Relating to Pitch Issues 101 3.3.4 Intercloanl Variation in Components Relating to Toxicity Issues 107 3.3.5 Interclonal Variation in Compounds Related to Decay Resistance 109 3.3.6 Correlation of Compounds with Related Biosynthetic Pathways 111 3.3.7 Non-genetic Contribution to Phenotypic Variation 113 3.4 Summary 116 CONCLUSIONS AND RECOMMENDATIONS 117 4.1 Conclusions 117 4.1.1 Analytical Methods 117 4.1.2 Clonal Variation of Wood Extractives 118 4.2 Recommendations 119 4.2.1 Analytical Methods 119 4.2.2 Clonal Variation of Wood Extractives 120 LITERATURE CITED 122 Appendix I: Statistical Formulae 130 Appendix II: List of All Compounds Found in Aspen Acetone Extractives and Quantifiable in a Single Chromatogram by GG-MS Underivatized Method 133 Appendix III: Identification of Various Compounds by Mass Fragmentography 135 Appendix IV: Optimization Curves for ELSD Nebulizer Gas Flow Rate and Drift Tube Temperature 138 Appendix V: GC-MS RIC Profiles of Lipophilic Fraction for CH-F6-02 Before and After 0.5 M KOH Hydrolysis 139 Appendix VI: Mass Spectra of Selected Steryl Esters Found in Aspen 140 Appendix VII: Response Factors (based on heptadecanoic acid) for the Standard Compounds used in Generic Response Factor Determination 143 VI TABLE OF CONTENTS (CONT'D) Page Appendix VIII: Mass Fragmentography of Docosanoic Acid Comparison Between Underivatized and TBDMS-Derivatized 144 Appendix IX: Amounts (mg/kg fd wood) of Each Compound Found in the Acetone Extracts for Each Ramet by GC-MS 145 Appendix X: ANOVA Results for the Quantity of Extractive Components Found in Natural Aspen Clones 149 Appendix XI: Decay Correlation Plots for Benzoic Acid and O-Coumaric Acid 150 Appendix XII: Pearsons Correlation Matrices for Biosynthetically Similar Compounds Found in Aspen 151 Appendix XIII: Chemical Structures 154 vii LIST OF TABLES Table Page 2.1 Solutions of (3-glucosidase hydrolysis in 10.00 mL of deionized water with 2.4 mg of heptadecanoic acid as an internal standard 31 2.2 Percentage weight loss in aspen wood chips (-40 g wet) after freeze drying at ~ 2xl0_1 mBar and -45°C 35 2.3 Repeat extraction of milled cores to determine the completion of extraction using samples CO-F3-08 (relatively sound wood) and CO-K2-03 (relatively decayed wood) 35 2.4 Extraction duplicates for raw wood chip and milled wood samples from CH-F6-02 (-10 g samples, 8 h, 200 mL acetone, 4-5 cycles/h) 36 2.5 Peak areas of replicate samples of 0.25 mg/mL heptadecanoic acid (HDA; RRT = 1.00) and cholesteryl palmitate (CP; RRT = 2.06) in acetone by GC-MS using underivatized acetone extractives method (section 2.2.4) 78 3.1 Number of ramets sampled, age, DBH, and wood properties from nine natural aspen clones in north eastern BC 90 3.2 Decay index developed to grade each core on extent of decay present 94 3.3 Site characteristics for selected aspen clones 113 viii LIST OF FIGURES Figure Page 1.1. Natural occurrence of quaking aspen in North America 2 1.2 Structures of typical compounds found in hardwood extractives 9 1.3 Typical micelles formation of saponifiable extractives allowing solubilization of non-polar components in aqueous environment found in kraft pulping 12 1.4 Schematic of an ion trap mass spectrometer 18 1.5 Schematic of a typical evaporative light scattering detector 20 2.1 Scheme for extractives method development and analyses (sections 2.2.1 to 2.2.9) 23 2.2 Scheme for salicylic acid analyses 33 2.3 Gas chromatography profile of total extractives from chip or milled wood 38 2.4 Typical profile for aspen wood acetone extractives underivatized by GC-MS 40 2.5 Typical profile for aspen wood acetone extractives underivatized by fast GC-MS 42 2.6 Thermal decomposition of cholesteryl palmitate at 150°C sealed in reacti-vials 43 2.7 Typical HPLC-ELSD separation of cholesteryl esters 44 2.8 Calibration curves for cholesterol palmitate by HPLC-ELSD and GC-MS 46 2.9 Resolution comparisons between optimized HPLC and GC-MS methods for steryl ester analysis 47 2.10 Benzoic acid and salicylic acid retention zone (grey area) in the phenolics region of the total extractives profile by GC-MS before and after P-glucosidase treatment 49 2.11 Results of monosaccharide analysis by GC-MS for ethyl acetate soluble acetone extracts before and after TFA hydrolysis 50 2.12 Suggested origin of salicylic acid in aspen from known glucosides salicortin / tremulacin via acid (HC1) dehydration to co-salicyloylsalicin / salicyloyltremuloidin followed by acid hydrolysis to salicylic acid and salicin / tremuloidin 51 2.13 Identification of a low molecular weight phenolic in the total extractives profile by NIST '98 mass spectral match 53 ix LIST OF FIGURES (CONT'D) Figure Page 2.14 Mass spectrum of P-sitosterol authentic standard and classification of an unidentified sterol based on general mass spectral fragmentography 55 2.15 GC-MS RIC profiles of total extractives and the resultant SE/W fraction obtained by solid-phase extraction 57 2.16 GC-MS RIC profiles of total SE/W fraction and argentation chromatography fractions I and II of total SE/W fraction 58 2.17 Relative retention by GC-MS versus fatty acyl carbon number for lanosteryl (3p-hydroxy-8, 24-lanostadiene) esters: lanosteryl myristate (C14:0), lanosteryl palmitate (C16:0), lanosteryl stearate (C18:0), lanosteryl eicosanoate (C20:0), and lanosteryl oleate (C18:1) 59 2.18 GC-MS RIC profiles for argentation fraction I of total SE/W fraction with synthetic a-amyrin myristate, palmitate, stearate, and eicosanoate 61 2.19 Structures of P- and a- amyrin esters with mass spectra of steryl palmitates (C16:0) from argentation fraction I of total SE/W fraction 62 2.20 GC-MS RIC profiles for argentation fraction II of total SE/W fraction and synthetic lupeol myristate, palmitate, stearate, and eicosanoate 63 2.21 GC-MS RIC profiles for total SE/W fraction and argentation fractions II-1 and II-2 of argentation fraction II 64 2.22 GC-MS RIC profile for the purified alkaline hydrolysate of fraction II-l and associated mass (normalized abundance) spectrum of the predominant peak with lanosterol (3p-hydroxy-8, 24-lanostadiene) structure 66 2.23 Structures of 4,4' - dimethylsterols: lanosterol (3p-hydroxy-8, 24-lanostadiene) and butyrospermol (tirucalla-7, 24-dien-3P-ol) 67 2.24 HPLC-ELSD profile of SE/W fraction using a 1:1 (volumetric) pre-detector split where I-IX bands were separately collected for subsequent GC-MS analysis 69 2.25 GC-MS RIC profiles of SE/W fraction and resultant HPLC fractions I-IX 70 2.26 Elimination of interference by co-eluting compounds in RIC chromatogram of a-amyrin by quantitation based on 218 m/z ion 72 2.27 Relative mass spectral response (RIC) for chemically similar components in the fatty acid and steryl esters class verses fatty acid carbon number 75 x LIST OF FIGURES (CONT'D) Figure Page 2.28 Determination of the upper linear response limit of the mass spectrometer with cholesteryl palmitate 76 2.29 Triplicate GC-MS chromatograms for underivatized aspen extractives illustrating the high reproducibility in the developed method 79 2.30 Typical GC-MS profile for TBDMS-derivatized extractives compared with the GC-MS profile for underivatized extractives 81 2.31 Typical GC-FID profile compared with GC-MS profile for underivatized extractives 83 3.1 Biogeoclimatic zone map of BC with Prince George Forest Region overlay showing aspen sampling sites Del Rio in the Dawson Creek Forest District, and Farrell / Kobes in the Fort St. John Forest District 89 3.2 Photographs (100 ISO, Kodachrom™ Slides) of each aspen clone sampled highlighting morphological differences between clones 91 3.3 Images of decayed wood: A = typical decay seen in 10 mm cores; B = typical sound wood seen in 10 mm cores (for reference); C = decay caused by Phellinus tremulae as seen in aspen stem cross-section 93 3.4 The distribution curves for the total weight of extractives for clones Dl,Kl,andF3 97 3.5 One-way ANOVA results for total acetone extractives (% fd wood) 98 3.6 The average amounts of several chemical classes found in the acetone extracts of aspen wood core (10 mm) samples based on GC / MS quantitation 100 3.7 One-way ANOVA results for total sterols / triterpenes (g / kg fd wood) 102 3.8 One-way ANOVA results for total triglycerides (g / kg fd wood) 103 3.9 One-way ANOVA results for total steryl esters / waxes (g / kg fd wood) 104 3.10 One-way ANOVA results for high molecular weight steryl esters / waxes (g/kgfdwood) 105 3.11 One-way ANOVA results for saponifiables / unsaponifiables 107 3.12 Scatter plot of decay index with monoaryl phenolics content (g / kg fd wood) 108 3.13 Scatter plot of decay index with total flavonoid content (g / kg fd wood) 109 xi LIST OF FIGURES (CONT'D) Figure Page 3.14 Relative amounts of bound salicylic acid and free benzoic and o-coumaric acids in CO-D1-03 (decay = 1.5), CO-D1-20 (decay = 2.5) and CO-D2-02 (decay = 5) 111 3.15 Correlation of the amount (mg / kg fd wood) of butyrospermol palmitate with that of other steryl esters 112 3.16 Site factors (moisture class and nutrient regime class) related to the total acetone extractives for aspen clones 114 3.17 Age versus total weight of acetone extractives for aspen clones 116 xii TABLE OF ABBREVIATIONS AC argentation chromatography AC-GC-MS argentation chromatography - gas chromatography / mass spectrometry ADT air dry tonne ANOVA analysis of variance BWBS Boreal White and Black Spruce °C degrees Celsius cc cubic centimetre cGC capillary gas chromatography CI confidence interval cm centimetre CP cholesteryl palmitate (standard) CPPA Canadian Pulp and Paper Association CTMP chemi-thermomechanical pulping DP decomposition products EI electron impact ELSD evaporative light scattering detector FC flash chromatography fd freeze-dried FID flame ionization detector g gram GC gas chromatography GC-FID gas chromatography - flame ionization detection GC-MS gas chromatography / mass spectrometry x i i i TABLE OF ABBREVIATIONS (CONT'D) ha hectare HDA heptadecanoic acid (standard) HPLC high-performance liquid chromatography ITMS ion trap mass spectrometry kg kilogram km kilometre LC50 lethal concentration killing 50% of the test population M molar (mol / L) m metre mg milligram ug microgram mL millilitre uL microlitre mm millimetre um micrometre min minute MS mass spectrometer MW molecular weight (g/mol) m/z mass / charge NIST National Institute of Standards and Technology QI quantitation ions QIR quantitation ion ratio rf radiofrequency TABLE OF ABBREVIATIONS (CONT'D) RIC reconstructed ion current RRT relative retention time SAR systemic acquired resistance SE/W steryl esters and waxes SPE solid-phase extraction TBDMS tert.-butyldimethylsilyl TFA trifluoroacetic acid TLC thin-layer chromatography UV ultraviolet V volts XV ACKNOWLEDGEMENTS I would first and foremost like to thank my supervisors Dr. C. Breuil from the Department of Wood Science and Dr. P. Watson from Paprican for their invaluable guidance and support throughout my thesis project. Also, I would like to extend this appreciation to Dr. P. Bicho and Dr. B. Sithole from Paprican, and Dr. A. Serreqi and Dr. R. Beatson from the Department of Wood Science for their knowledgeable suggestions in the area of wood extractives analysis. Assistance from Robert Leone, Maxewell McRae, Luis Del Rio, Dr. K. Hunt, Shanley Pitts, Charlotte Norris and the valuable advice from Dr. C. Jennison at Varian Canada Inc. are gratefully acknowledged. Finally, I want to thank my family and friends that put up with me and supported me throughout this work. I wish to thank Paprican for their generous in-kind contributions including extended use of their gas chromatograph / mass spectrometer and high-performance liquid chromatography system without which this research would not be possible. NSERC and Paprican provided me with financial support (Industrial Postgraduate Scholarship) throughout this research. XVI 1 INTRODUCTION 1.1 GENERAL INTRODUCTION This thesis covers the development of analytical techniques that may be used to monitor the amounts of various wood extractives, including key pitch forming components, in standing quaking aspen (Populus tremuloides Michx.) trees. Also, presented is a survey of the interclonal variation in wood extractives that may potentially impact on the utilization of this species for the manufacturing of pulp and paper. The introduction of this paper will explain issues and concepts that are important to this study, and provide references to relevant studies found in the literature. The second section of this thesis, entitled "Development of Analytical Methods", is an account of the methods developed to characterize and quantify important wood extractives in standing aspen trees. The third section, entitled "Clonal Variation of Extractives", provides information on the types and amounts of wood extractives in aspen along with results of the interclonal differences observed in these extractives between nine natural aspen clones. Finally, conclusions and recommendations are provided. 1.2 QUAKING ASPEN 1.2.1 Biology of Quaking Aspen Quaking aspen, also commonly known as trembling aspen, is the most widely distributed native tree species in North America, occurring across an impressive 111 degrees of longitude and 47 degrees of latitude as illustrated in Figure 1.1 (Peterson and Peterson 1995). 1 Figure 1.1. Natural occurrence of quaking aspen in North America (taken from Farrar 1997). In Canada, this species occurs in every province making up over 50% of the hardwoods and 11% of the entire Canadian timber resource (reviewed in Chen et al. 1995). In British Columbia where 15% of the Canadian aspen volume exists (MacLeod 1987), aspen is an early-succession species, and occurs in all forested biogeoclimatic zones east of the Coastal Mountains, but it is most prevalent in the Boreal White and Black Spruce (BWBS) zone (Peterson and Peterson 1995). Aspen's unusual mode of asexual reproduction referred to as "suckering" where several genetically identical stems, or ramets, emerge from a common root system, lends this species to a rapid natural regeneration, particularly in disturbed sites (Mitton and Grant 1996; Peterson and Peterson 1995). Furthermore, this mode of reproduction results in the occurrence of natural clones commonly less than 1 ha in area and composed of tens to thousands of ramets (Peterson and Peterson 1995). Aspen is a dioecious species, meaning sexual reproduction occurs between a separate male and female plant (Lindroth and Hwang 1996). Furthermore, aspen is diploid, and therefore each plant has two genes or alleles coding for each protein (Mitton and Grant 1996), such that there is a potential for individuals to be either homozygous (ie both recessive or both dominant, aa or AA) or heterozygous (ie dominant and recessive or recessive and dominant, Aa or aA) for a particular gene. Heterozygotes retain more genetic variability through sexual reproduction, and in aspen, heterozygotes are favored in semi-arid environments where high genetic variation between aspen clones has been observed (Mitton and Grant 1996). This selection of heterozygous individuals is thought to be a contributing factor to the large genetic diversity seen in aspen. Thus, not surprisingly, a large variability in several traits including morphology, time of budbreak and leaf senescence, autumn coloration and resistance to herbivores, pathogens and drought has been observed between aspen clones (Lindroth and Hwang 1996; Peterson and Peterson 1995). Furthermore, an electrophoretic survey of protein (in leaves or buds) variation in aspen concluded that more that 90% of the enzymes analyzed were genetically variable, suggesting that quaking aspen is the most genetically variable species of plant investigated to date (Mitton and Grant 1996). 1.2.2 Measuring Genetic Variability in Quaking Aspen Heritability values for economically important traits such as growth rate and wood properties are useful to tree improvement foresters who aim to improve the forest both for better lumber yields and quality. Heritability or h2, is the proportion of variation of a particular trait, in a population that is attributable to genetic differences (Zobel and Talbert 1984). 3 h2 = r j 2 G / ( G 2 G + 0- 2 E ) (1) Where a2c = phenotypic variance due to genetics; O 2 E = phenotypic variance due to environment. In studies involving natural stands, the genetic and environmental components of variation cannot be separated and thus, no definitive conclusions about degree of inheritance of any characteristic can be made (Zobel and Talbert 1984). Nevertheless, estimates of broad-sense heritabilities for certain traits in aspen based on natural clone tests have been performed by Yanchuk et al. (1984) using a modified version of Equation 1: h 2*CT 2 c/(o 2 c + a2t + o-2e) (2) Where a c = variance among clones; a t = variance among trees withm clones; a e = error (unexplained) variance. Analysis of variance (ANOVA) in natural clone tests provides a comparison of variance among clones verses the variance among trees within clones to determine significant differences between clones. Currently, randomized complete block design is a common experimental design utilized in forest genetic trials (Yong-Bi et al. 1998). With this design, each experimental unit(ie ramef) occurs randomly in each site block (2-4 ha) of varying environmental conditions, and thus, the variability due to environmental differences may be effectively eliminated, and the genetic factor is isolated. As mentioned previously, this segregation of factors is not possible with natural clones, and a completely randomized design (CRD) with pseudoreplication is 4 commonly used with natural clones, where several ramets are randomly sampled (ie replicates) from several different clones (Yanchuck et al. 1984). Pseudoreplication is unavoidable in many experiments and occurs when either treatments are not replicated (though samples may be) or the replicates are not statistically independent (Hurlbert 1984). The danger with pseudoreplication in the sampling of natural clones comes when there is a large environmental variation between the different clones, together with a high environmental dependence on the trait being tested. However, if the testing area contains relatively uniform site properties throughout the sampled clones, environmental factors may be less significant. Also, in Populus species such as aspen, wood quality traits are known to be under strong genetic control with estimates of h2 (broad-sense) determined from natural aspen clones to be 0.35 to 0.43 (Yanchuk et al. 1984). Furthermore, according to Zobel and Talbert (1984), variation in wood quality characteristics between trees is normally due to inheritance rather than environmental differences. 1.3 UTILIZATION OF ASPEN IN THE PULP AND PAPER INDUSTRY 1.3.1 Desirable Features of Aspen for Pulp and Paper Historically in Canada, poplars (largely made up of aspen) were considered to be "weed trees" and used primarily for fuel, excelsior and, to a small degree, for pulpwood (Dunlop-Jones et al. 1991). However, Watson (in litt, 16 Sept. 1999) stated that the value of quaking aspen as a fibre source has been increasingly appreciated, as evidenced by its utilization in the latest five major pulp and papermaking projects commissioned in Canada. Unusual wood chemistry of low lignin (16% extractives-free wood) and high carbohydrates (cellulose = 53% extractives-free wood) 5 makes this species conducive to chemical or mechanical pulping and so attractive to the pulp and paper industry (MacLeod 1987). In Western Canada, 40% or 650,000 ADT/year of aspen pulp is produced via kraft pulping, and the remaining 60% or 915,000 ADT/year is produced by mechanical pulping processes, mainly chemi-thermomechanical pulping or CTMP (Rodden 1999). Aspen provides a high kraft pulp yield, requires a relatively low refiner energy in CTMP and requires little bleaching as a result of its natural brightness, making it an ideal pulpwood species (MacLeod 1987). Furthermore, aspen pulp consists of small-diameter, thin-walled fibres which are ideal for producing high-density paper sheets with a smooth printable surface (Karl 1988). Another attractive feature of aspen is its rapid growth rates and lower age of maturity, resulting in a shorter rotation time (as low as 30 years for pulpwood in Western Canada) compared with commercially important softwoods (Peterson and Peterson 1995). As a result, aspen is very amenable to reforestation and genetic selection (Dunlop-Jones et al. 1991). 1.3.2 Limits on the Utilization of Aspen for Pulp and Paper Despite these exceptional qualities, aspen utilization is plagued by a series of complications. Aspen is notorious for giving rise to pitch problems in CTMP and even in kraft pulping. Pitch deposits are hydrophobic in nature, originate from substances native in the wood, and deposit on equipment surfaces in mills, or become imbedded in the final product along with accumulated dirt particles (Allen 1988). When the wood supplied to a fully bleached kraft pulp mill is changed from softwoods to aspen, the incidence of pitch-dirt contaminated pulp may increase by six to ten times (Chen et al. 1995). Furthermore, Allen (1988) reported that it was not uncommon for an 800 ADT/day fully bleached kraft mill pulping aspen to lose two million 6 dollars per year as a result of the cost of additives used to alleviate pitch problems or the devaluation of pulp due to pitch-dirt contamination (Allen 1988). Another serious limitation on the utilization of aspen results from the high susceptibility of this species to decay by fungi (Pausler et al. 1995; Knoll et al. 1993). The most important form of aspen wood decay in Canada is heartwood decay caused by Phellinus tremulae which is present in high frequency in mature (>60 years old) stands (Hiratsuka et al. 1990). This white-rot fungus often causes large decay columns (heart rot), frequently continuing along most of the main stem producing soft / spongy and discolored wood in advance stages of decay (Hiratsuka et al. 1995). Other fungal species important in aspen decay are Peniophora polygonia, which produces pink to brownish pink decay pockets along the stem, and Armillaria species, which produce stringy dark brown butt rots seldom extending more than 1 m above ground (Hiratsuka et al. 1995). Estimates of mature aspen trees with stem decay range from 25% nationwide, to 75% in Western Canada (Knoll et al. 1993). Decayed wood is a problem to the pulp and paper industry for several reasons. According to Watson (in litt., 16 Sept. 1999), wood volume losses through chip production occur with advanced decay material. Also, the presence of decayed wood chips increases the chemical consumption in kraft pulping and decreases the pulp brightness in mechanical pulping due to the presence of discolored wood (Hunt 1995), both of which increase the cost of pulping aspen. A final limitation on the utilization of aspen, particularly for mechanical pulping (ie CTMP) where large volumes of process effluents are generated, is the relatively high toxicity of the natural components in the wood released during pulping. The LC5o (% solution) using Microtox™ bioassay has been calculated to be 5.70% for aqueous wood extracts in aspen, 7 compared to 24.05% and 20.95% in Eucalyptus globulus and Douglas-fir, respectively (Charlet et al. 1997). This implies that aspen aqueous leachates are approximately four times more toxic than leachates from these other commercially important species. Thus, the toxicity of leachates from aspen woodpiles and of process effluents, may be a significant environmental problem to the pulp and paper industry, as well as to the veneer, plywood and oriented-strand board industries which rely heavily on aspen (Taylor et al. 1996). 1.4 WOOD EXTRACTIVES 1.4.1 Wood Extractives in Hardwoods Wood extractives are defined by Fengel (1983), in a general sense, as a large number of different compounds that can be extracted from wood by means of both polar and non-polar solvents. This includes a large variety of compounds grouped into general classes including terpenoids, fats and waxes and their components, and phenolics. Also, extractives may include other components not formally classified into these groups such as carbohydrates, peptides and inorganic compounds. As illustrated in Figure 1.2, the wood extractives of many hardwood species including quaking aspen have previously been reported to contain sterols (A) and triterpenes (1 in B), fatty acids (E) and their esters (B, C and D), monoaryl phenolics (F), and other higher phenolics (G) (Fengel 1983; Serreqi etal. 1998). 8 OH 0 Figure 1.2. Structures of typical compounds found in hardwood extractives (A = [3-sitosterol; B = a-amyrin palmitate; C =tripalmatin; D = linoleic acid, linoleyl ester; E = stearic acid; F = coniferaldehyde; G = kaempferol). 1.4.2 Biosynthesis of Wood Extractives Most extractive compounds are considered to be secondary products in plants since they are biosynthetic derivatives of primary metabolites (Miller 1973a). Biosynthetic pathways of secondary products usually involve several reaction steps catalyzed by enzymes or single multi-enzyme units. As a result, certain extractive compounds may be more closely biosynthetically related to one another, depending on the pathway they originate from. Terpenoids, such as triterpenes commonly found in hardwoods, are characterized by the composition of five carbon isoprene (2-methyl butadiene) units and are produced in plants via a related isoprenoid biosynthetic pathway which in most cases involves a mevalonic acid [1.1] 9 building block (Rowe 19896). Thus, mevalonic acid [1.1] is considered the primary metabolite from which secondary products or terpenes are synthesized. Steroids are biosynthetically related to the triterpenes since they both arise from cyclisation of squalene [1.2] which in turn is produced from mevalonic acid [1.1] via the isoprenoid route (Dey and Harborne 1991). Conversely, fatty acids are biosynthesized via a route involving protein-bound acetoacetate [1.3] and butyrate [1.4] catalyzed by a multi-enzyme unit (Miller 19736). The production of waxes involves further esterification of a fatty acyl intermediate (from fatty acid biosynthesis), and a fatty alcohol [1.5] produced from the corresponding reduced fatty acid (Rowe 1989a). Glycerides or steryl esters are most likely biosynthesized by enzyme catalyzed dehydration coupling of glycerol [1.6] with fatty acids or various sterols and triterpenes with fatty acids, respectively. Phenolic wood and bark extractives form the defense mechanism in most tree species protecting the plant against invading microorganisms or wood-boring insects and even browsing by mammals or birds (Taylor et al. 1996). Monoaryl phenolic compounds (ie F in Figure 1.2) found in wood are derived from phenylpropane [1.7] monomers (Rowe 1989a) which polymerize to form lignin, the amorphous "cement-like" compound found in wood. Other important phenolics in plants, including aspen, are the flavonoids [1.8], some of which are known phytoalexins, or compounds specifically produced in plant tissue in response to microbial attack (Rowe 1989a). Flavonoids are biosynthetically related to the aromatic amino acids, phenylalanine [1.9] and tyrosine [1.10], which are also precursors to all phenylpropanoids [1.7] (Miller 19736). 10 The complexity of the biosynthetic steps mentioned in this section is immense and thus, details have been omitted as this information is presented only to introduce the general biogenetic relationship between extractive compounds. For a more detailed review of the biosynthetic pathways mentioned, the following references are recommended (Miller 1973a; Miller 19736; Rowe 1989a; Rowe 19896; Dey and Harborne 1991; Fengel 1983). 1.5 WOOD EXTRACTIVES AND LIMITS ON THE UTILIZATION OF ASPEN 1.5.1 Wood Resin and Pitch A particular fraction of the wood extractives, referred to as resin, is described as consisting of a mixture of relatively nonpolar, oleophilic, water-insoluble, and low molecular weight nonstructural wood components which, when dispersed in water, tend to coagulate, especially at hydrophobic surfaces (Hillis 1962). In aspen, this includes fatty acids, waxes, fatty alcohols, triterpenes / sterols, steryl esters and glycerides (Sithole et al. 1992). However, kraft mill personnel find it useful to classify resinous extractives into two broad classes based on their ability to form soap under the high alkaline conditions found in kraft pulping. Saponifiables, such as fatty acids, glycerides, and even some steryl esters form soaps under alkaline conditions (Sithole et al. 1992). These soaps generally consist of an ionized negatively charged carboxylate "head" (A) with a non-polar hydrocarbon "tail" (B) and tend to arrange in micelles (see Figure 1.3) in aqueous environments. This arrangement allows for the solubilization of non-polar resin components (C) within the non-polar environment of the micelle centres (Allen 1988). 11 Figure 1.3 Typical micelles formation of saponifiable extractives allowing solubilization of non-polar component in aqueous environment found in kraft pulping. (A = carboxylate anion end; B = non-polar hydrocarbon chain end; C = non-polar resin component). Unsaponifiables or non-soap forming components of resin tend to coagulate and cause pitch problems even in alkaline conditions (Sithole et al. 1992). Saponification aids in the removal of resins during kraft pulping, thereby alleviating pitch problems with most species. However, aspen wood contains a relatively high percentage of unsaponifiables, estimated at 16% (average from data presented in Hillis 1962) of the ether soluble extractives, resulting in a relatively low saponifiables to unsaponifiables ratio of 1.2 (Allen 1988). Moreover, the unsaponifiables in aspen consist predominantly of steryl esters (ie B in Figure 1.2) and some waxes (ie D in Figure 1.2) which are inherently stable and persist through seasoning, alkaline hydrolysis during pulping (Allen et al. 1991) and even bio-treatment with lipid consuming fungi (Chen et al. 19946). 1.5.2 Wood Extractives and Mill Effluent Toxicity Extractive compounds that are thought to be responsible for the acute toxicity of aspen extractives include low molecular weight phenolics such as benzoic acid [1.11] and phenol 12 [1.12], as well as non-phenolics such as nonanoic acid [1.13] (Taylor et al. 1996). Furthermore, phytoestrogens such as sterols and flavonoids can compromise reproductive capacity and elicit toxic effects in livestock, rodents and fish (Zacharewski et al. 1995). A significant amount of research shows that phytosterols including sitosterols (ie A in Figure 1.2), the most abundant and widely distributed sterols in plants, are responsible for the chronic toxicities elicited in aquatic life exposed to mill effluents (Denton et al. 1985; Maclatchy and Van der Kraak 1994; Zacharewski et al. 1995). [3-sitosterol [2.27] is the most abundant compound in the sterol / triterpene alcohol fraction of aspen (Dunlop-Jones et al. 1991; Sithole et al. 1992), and was found to be the main sterol discharged (at an average mass discharge rate of 5.76x10" kg/tonne) from several pulp and paper mills in the US utilizing various pulping processes and furnishes (Cooke/ al. 1997). A myriad of other potentially toxic compounds may result from extractives modification during the pulping process. In particular, chlorinated bleaching agents have been shown to react with certain extractives to produce the environmentally notorious adsorbable organohalogens (AOX) (Servos et al. 1996). 1.5.3 Salicylic Acid and Aspen Decay Salicylic acid [1.14] is commonly found in the leachates during the processing of plant material including aspen wood (Rowe 1989a). In plants such as cucumber and tobacco, salicylic acid [1.14] is thought to be the signaling compound that triggers systemic acquired resistance (SAR) against viral, bacterial or fungal pathogens throughout the plant (Yalpani et al. 1993). Biosynthesis of salicylic acid [1.14] is thought to be initiated with the conversion of phenylalanine [1.9] to t'-cinnamic acid [1.15] catalyzed by phenylalanine ammonia lyase, an 13 enzyme which is a key branch point in the phenylpropanoid pathway that produces lignin as well as other phytoalexins (Horvath and Chua 1994). It has been well documented that phenolic glucosides are present in aspen bark and foliage extractives. These include salicortin [1.16], tremulacin [1.17], salicin [1.18], tremuloidin [1.19], and 1-O-p-coumaroylglucose [1.20] (Pearl and Darling 1971a). In a more recent study by Pausler et al. (1995), based on work by Hubbes (1969), acidic hydrolysis of an uncharacterized glycoside (glycoside W) in the butyl alcohol extract of aspen wood liberated salicylic acid [1.14]. Salicylic acid (3-glucoside [1.21] is known to occur in plants in which salicylic acid [1.14] produces SAR, however, the glucoside is an inactive storage form, which may be quickly hydrolyzed to release SAR inducing salicylic acid [1.14] in the event of microbial invasion (Horvath and Chua 1994). If a similar mechanism exists in aspen, levels of bound salicylic acid may allow one to assess the potential decay resistance of aspen trees, particularly against P. tremulae which is known for its ability to methylate and thus, inactivate benzoic acid [1.11], a salicylic acid [1.14] precursor (Pausler et al. 1995). 1.6 MODERN METHODS FOR ANALYSIS OF WOOD EXTRACTIVES 1.6.1 S ampling and Extraction To increase the number of samples that may be taken in a reasonable time frame, and to reduce the cost and environmental impact of sampling, non-destructive sampling techniques such as increment coring are often used in wood properties studies. Moreover, tree selection for tree breeding programs is normally performed by increment core methods such that vegetative propagation of the sampled tree remains possible (Ona et al. 1996). Although sampling at a single point on the tree (ie increment core taken at breast height) will not necessarily provide one with an absolute measure of the wood properties for the entire tree, single-point sampling is 14 suitable for relative comparisons between trees of similar diameter and sampled in the identical manner. DeBell et al. (1997) found that single increment cores taken from such trees were sufficient for assessing relative differences in tropolone content in western redcedar to predict the decay resistance of individual trees. The Canadian Pulp and Paper Association (CPPA) standard procedure for determining the extractives content of wood and wood pulps involves solvent extraction with a Soxhlet apparatus (CPPA test methods G.13 and G.20). Solvents commonly used for extractives analysis include diethyl ether, ethanol, acetone, dichloromethane, and combinations including ethanol/benzene and ethanol/toluene, where each solvent or combination of solvents removes a slightly different amount of material from the wood (Allen 1988). However, acetone has become the standard solvent in Canada for wood extractives removal since it is more inert and stable, and less harmful from a health and environment perspective than other solvents or combinations (Sjostrom and Alen 1999). Soxhlet extractions involve a refluxing solvent that condenses onto the material to be extracted and then siphons back into the solvent reservoir to repeat the cycle. Several modern extraction techniques such as accelerated solvent extraction and supercritical fluid extraction are currently available and may eventually replace Soxhlet (Sithole 1992). Although these modern techniques are more time efficient and use less solvent, Hughes and Thurbide (1998) have shown that Soxhlet extraction is still more reproducible than these modern extraction techniques. 1.6.2 Chromatographic Techniques for the Analysis of Wood Extractives Methods for the analysis of plant or wood extracts, or any complex mixture of compounds, usually involve some form of chromatography where partitioning of the solutes or analytes 15 occurs between a mobile phase (usually a liquid or a gas) and a stationary phase (liquid coated to the inside of a tube or on the surface of solid particles, or only solid particles) (Harris 1991). Chromatographic techniques typically used for the analysis of plant / wood extracts include high-performance liquid chromatography (HPLC), gel permeation chromatography (GPC), thin-layer chromatography (TLC), gas chromatography (GC), and more recently supercritical fluid chromatography (SFC) (Dey and Harborne 1991). Gas chromatography has been the traditional technique for the analysis and isolation of the lower, more volatile terpenoids (Dey and Harborne 1991). GC involves a gaseous solute (analyte) which is propelled by a gaseous mobile phase (usually helium, nitrogen, or hydrogen gas) within a column in which the stationary phase, a non-volatile liquid, is bonded to the walls or on a fine solid support (Harris 1991). Improved technology in GC fused-silica capillary columns has enabled the use of much shorter (<10 m) columns allowing for the analysis of higher molecular weight and less volatile components such as triglycerides and steryl esters (Sithole etal. 1992). Derivatization of GC analytes is the process of chemically blocking or capping protic sites thereby reducing dipole-dipole interactions and increasing volatility and GC performance (Blau and Halket 1993). For analysis of wood extractives this is usually achieved by methylation (cap = -CH 3), acetylation (cap = -COCH3) or silylation (cap = -Si(R3)) (Peng et al. 1999). However, derivatization procedures increase sample preparation time, reagent costs, and introduce a degree of error into any quantitative analysis. Gas chromatographic analysis of lipophilic wood extractives such as steryl esters, triglycerides and waxes has been performed without any prior derivatization due to the lack of free derivatization sites on these lipids (Gutierrez et al. 1998). 16 A common detector used in GC analysis of wood extractives is the flame ionization detector (FID) which measures the voltage of the current produced by carbenium ions (CHO+) created by the combustion of susceptible analytes in a hydrogen-oxygen flame (Harris 1991). Although the concentration of eluting analyte is strictly proportional to the voltage produced by the FID, no qualitative information on the analyte is provided by this mode of detection. In ion trap mass spectrometry (ITMS) (see Figure 1.4) operating in electron impact (EI) mode, compounds are ionized by a pulse of electrons generated from a heated filament (A) and accelerated through an electron gate (B). Ions are formed and stored in the trap by a radiofrequency (rf) voltage applied to a ring electrode (C). The ions are then released one mass / charge (m/z) ratio at a time by increasing the rf amplitude which destabilizes the ions and causes them to exit the trap (Adams 1989). Finally, the released ions enter the electron multiplier (D) which amplifies the signal and then sends it to the computer which records each ion intensity. Since ITMS utilizes a larger fraction (50%) of the sample for analysis, it is far more sensitive (detection limits as low as 5 pg) than conventional quadrupole mass spectrometer which only uses a fraction (0.1-0.2%)) of the sample (Adams 1989). Thus, ITMS allows characterization and quantification of a large variety of compounds even when present in trace amounts. 17 • Signal Figure 1.4. Schematic of an ion trap mass spectrometer (Varian Canada 1998). A = filament; B = electron gate; C = ring electrode; D = electron multiplier. 1.6.3 Analysis of Steryl Esters The chromatographic separation of phytosteryl esters is plagued by the formation of critical pairs, where a wide variety of sterols esterified with several fatty acids produces various sterol and fatty acid combinations that have the same retention for a given chromatographic system (Rezanka 1992). Thus, not surprisingly, the identification of individual compounds in the steryl esters fraction of aspen by GC is plagued by co-elution of components (Serreqi et al. 1998). Furthermore, steryl esters do not produce a strong molecular ion fragment (M+) by electron impact ionization which is needed to confirm the identity of the intact steryl esters by gas chromatography - mass spectrometry (GC-MS) (Evershed et al. 1987; Rezanka 1992). Evershed 18 et al. (1987) reported that argentation chromatography (AC), which utilizes a AgN03-silica gel column, will fractionate plant steryl esters based on polarity and the degree of unsaturation of both the fatty acyl and steryl moieties. The gas chromatograms of the resultant fractions will contain less co-elution and more pure bands that may be compared to standard compounds for identification. Steryl esters are susceptible to thermal degradation, and in particular, steryl esters containing di-and tri-unsaturated fatty acyl moieties have been known to degrade during GC analysis (Kuksis et al. 1986). HPLC is a variety of chromatography that utilizes theories originally developed for GC to achieve very efficient separations (Lindsay 1989). This technique utilizes a liquid mobile phase, driven through a narrow (4-6 mm i.d.) bore stainless steel column, packed with a fine (<5 um grade) microporous support, at elevated pressures of 20-100 bar (Dey and Harborne 1991). Since elevated temperatures are not needed for separation of analytes, this method is well suited for high boiling point and / or thermally labile analytes. Furthermore, in the event of co-elution, HPLC is amenable to fraction collection which allows for the further purification and identification of co-eluting components. Detection of eluting analytes in HPLC is usually achieved by ultraviolet (UV) absorbance or Refractive Index (RI) detection (Dey and Harborne 1991). However, other more specialized detection methods have been developed to deal with the limitations associated with these detectors. Evershed et al. (1987) reported that the lack of a strong chromophore in steryl esters results in poor UV detection limits for these compounds. More recently, the evaporative light scattering detector (ELSD), as illustrated in Figure 1.5, has been used for the analysis of steryl esters and other non-UV absorbing analytes. (Ferrari et al. 1997; Nordback et al. 1998; Claassen etal. 1998). 19 The operation of an ELSD (see Figure 1.5) involves the atomization of the eluent into a gas stream with a nebulizer (A), followed by the evaporation of the eluent in the heated evaporation chamber (B), and finally measurement of light (C) scattered by the resultant aerosol of non-volatile analyte components as it passes the detector (D) (Oppenheimer and Mourey 1985). Figure 1.5. Schematic of a typical evaporative light scattering detector (Polymer Laboratories 1999). 1.6.4 Analysis of Salicylic Acid Glycoside Pausler et al. (1995) and Hubbes (1969) reported that salicylic acid [1.14] is present as a glucoside in aspen. Analysis of unknown phenolic glycosides can be performed by HPLC separation and quantification of the intact glycoside, followed by hydrolysis and subsequent GC-MS analysis of the separate aglycon and carbohydrate components (Vilegas et al. 1999). A more convenient method for the quantification of a known glycoside involves the hydrolysis of the 20 extract containing the glycoside followed by GC analysis of the hydrolysate for the aglycon (Jonasson et al. 1997). Acid hydrolysis with trifluoroacetic acid (TFA) will provide specific hydrolysis of non-crystalline carbohydrate material (Sjostrom and Alen 1999). However, to increase the specificity of hydrolysis, enzymatic hydrolysis with a glycosidase may be performed if the linkage configuration is known. 1.7 OBJECTIVES There were two primary objectives in this project. The first was to develop analytical methods that enable the rapid identification and quantification of the wood extractives in aspen with particular emphasis on those extractive components that may impact on the utilization of aspen for pulp and paper. This includes the steryl esters, which have not been well characterized in hardwoods such as aspen. It is anticipated that these methodologies would be useful for monitoring pitch culprits and environmentally problematic components to improve process efficiency, product quality and environmental monitoring and protection in pulp mills. The second objective was to conduct a survey of these important extractives in various natural aspen clones to assess their genetic variability in aspen. This information may provide tree breeders with a measure of the genetic diversity in the extractives of naturally occurring aspen that may be exploited for the production of plantation aspen with superior pulpwood qualities. 2 DEVELOPMENT OF ANALYTICAL METHODS 2.1 INTRODUCTION There has been a great deal of discussion in scientific literature on which analytical methods are superior for the quantification and characterization of extractives in wood, pulp and process 21 effluents. However, several studies on the characterization of wood extractives show that capillary gas chromatography (cGC) is by far the most convenient and comprehensive technique available to separate out the individual components in wood extractives for subsequent identification (Del Rio et al. 1998; Orsa and Holmbom 1994; Sithole et al. 1992; Charlet et al 1997). Also, a number of high-performance liquid chromatography techniques have been employed to deal with steryl esters and other high boiling point fatty acid esters found in various fungi, algae and plant extracts (Evershed et al. 1987; Ferrari et al. 1997; Rezanka 1992; Nordback et al. 1997). More recently, studies on the analysis of plant steryl esters and other lipophilic extractives have employed the Evaporative Light-Scattering Detector with excellent results (Ferrari et al. 1997; Nordback et al. 1997; Claassen et al. 1998). Each method has advantages and disadvantages based on the particular application and rarely will one method allow for a complete analysis of a complex sample. In many studies, different chromatographic methods are combined to achieve detailed analysis of a complex sample. This chapter reports on the development and validation of a comprehensive stand-alone GC-MS extractives analysis method. In addition, the utilization of argentation chromatography - gas chromatography / mass spectrometry (AC-GC-MS) to facilitate the identification of the major steryl esters in the total extractives profile is discussed. Also, the development of a stand-alone HPLC-ELSD method for the analysis of a simple mixture of plant steryl esters, and a HPLC-ELSD fractionation method for the analysis of more complex steryl ester mixtures (ie steryl esters in aspen) is presented. A suitable method for the quantification of bound salicylic acid in aspen has also been developed and the origin of salicylic acid [1.14] in aspen is discussed. Initially, this chapter focuses on the development of the basic methodologies for aspen extractives analysis, then on details of the characterization and quantification of various aspen 22 extractives in the GC-MS profile, and finally, on the validation of the underivatized GC-MS procedure. 2.2 MATERIALS AND METHODS 10 MM CORES OR CHIPS Grind MILLED WOOD Extract Filter CRUDE EXTRACT SOLUTION AMBER GLASS VIAL FRACTION Freeze-dry FILTERED EXTRACT SOLUTION Freeze-dry TBDMS-DERJVATIVES CLEAR GLASS VIAL FRACTION Solid Phase Extraction SE/W FRACTION ^ / H P L C , F C / ^ TLC / FRACTIONS Chromatographic Fractionation Figure 2.1. Scheme for extractives method development and analyses (sections 2.2.1 to 2.2.9). 2.2.1 Origin of Samples Used for Method Development The samples used for method development included wood chips (here after referred to as "CH-##(clone)-##(tree number)") from the mature wood (>40 years) from the bottom 6 m (butt) of the stems harvested from nine natural aspen clones near Chetwynd, BC in April and July of 1997. Also used in method development were ten millimeter diameter increment cores taken in June 1998 from the residual trees of the same nine clones (hereafter referred to as "CO-##-##" as in 23 the chip samples). Also, duplicate cores were taken parallel and adjacent to the first core from one to two ramets from some clones. Further details of the increment core sampling is discussed in section 3.2.1. 2.2.2 Sample Preparation and Soxhlet Extraction Initially, the sampled wood (cores were segregated into lengths not exceeding 7cm to fit into the freeze drier vessels) was placed in 150 mL freeze drier flasks, re-frozen for ~1 hour at <0°C, and then freeze-dried on an Edwards Pirani 1001 freeze drier (Edwards High Vacuum, Crawley, Sussex, England) for 24 h at -45°C and ~2xl 0"1 mBar (unless stated otherwise). Next, the dry material was promptly ground into wood meal with a 2 mm screen on a Thomas wood grinder (Arthur H. Thomas Co., Phila., PA, USA), and weighed (±1 mg) to obtain a freeze-dried (fd) weight. Approximately 5 grams (unless stated otherwise) of each sample was weighed into each soxhlet thimble for extraction. The samples were soxhlet extracted for 8 h at a rate of 4-5 cycles/h (unless stated otherwise) using 200 mL of certified ACS acetone. To minimize cross-contamination, soxhlet apparatus was rigorously cleaned with soap, water and acetone after every sample batch and the thimbles acid washed (diluted reagent grade HNO3) after every two batches. 2.2.3 Determination of Total Weight of Extractives The resultant acetone extracts (-200 mL) were then concentrated to -10 mL on a rotary evaporator (Buchi, Switzerland) using a water line aspirator for vacuum and a 30°C water bath. This residue was then passed through a pasture pipette packed (-2 cm) with glass wool to filter out particulate matter from the sample, and transferred to a pre-tared 16 mL screw cap vial with a Teflon™ liner. Each sample was split into two equal aliquots. One aliquot was loaded in an amber glass vial (the amber color retards photolytic changes in the sample) for total extractives 24 analysis, and the other in a clear vial for analysis of the lipophilic fraction (ie steryl esters). The extract residue (~10 mL) in the vials was then further concentrated down to <1 mL by passing a stream of nitrogen into the vial while heating at no greater than 40°C. Then, each of the extracts (with the vial cap loosened off) was placed in a 150 mL freeze drier flask and frozen for >1 h at <0°C. Next, the samples were freeze-dried for 24 h (~2xl0~' mBar, -45°C), removed from the freeze drier, and quickly capped tightly and weighed (±0.01 mg). The percent extractives (E) was then calculated based on the freeze-dried wood weight using Equation 3: E = I(A-B)«100]/W (3) Where: A = freeze-dried weight of the wood extractives; B = freeze-dried weight of the residue in a blank determination (soxhlet extraction without material loaded in thimble); W = freeze-dried weight of the test specimen (ie milled wood). 2.2.4 GC-MS Analysis of Total Extractives Samples in the amber screw-top vials were suspended at a concentration of 5.000 mg/mL in an internal standard solution of 0.250 mg/mL heptadecanoic acid (99%; Sigma, St. Louis, MO, USA) in certified ACS grade acetone. The standard solution was added using a self-filling buret at 23°C, ensuring that there was no significant evaporation/concentration of the solution during the addition. Next, the sample solutions in capped amber vials were sonicated for 60 min on a Branson 3210 Ultra-Sonicator (Danbury, CT, USA). For GC-MS analysis of underivatized extractives, approximately 1.5 mL of each sample solution was transferred to a 2 mL amber auto-sampler vial and capped (screw top with Teflon™ septum). 25 For the analysis of the total extractives as tert. -butyldimethylsilyl ( T B D M S ) derivatives, 200 u L of stock sample solution (containing 1 mg o f extractives) was added to a 2.0 m L amber auto-sampler vial , and this solutions was concentrated with a stream of nitrogen while heating (40°C). The resultant extract was then re-suspended in 100 u L of H P L C grade acetonitrile and 100 u L of N-methyl-N-(^rr.-butyldimethylsilyl)-triflouro-acetamide (silylation reagent; Sigma, St. Louis, M O , U S A ) was added. The reaction mixture was capped for 20 minutes at room temperature, then diluted with 1 m L of H P L C / G C - M S grade dichloromethane, and loaded on the auto-sampler for analysis. The G C - M S system was a Varian 3800 gas chromatograph coupled to a Varian Saturn 2000 ion trap mass spectrometer (Walnut Creek, C A , U S A ) . The gas chromatograph contained a D B -X L B i t d capillary column (10m x 0.25 mm, 0.25 um fi lm; J & W Scientific, Folsom, C A , U S A ) , a deactivated 1078 fritted splitter inlet sleeve (Restek, Brockville, O N , C A N ) and a Varian 1079 injector (Walnut Creek, C A , U S A ) . The M S electron multiplier voltage was set at 2150 V , an ionization time of 25,000 us, running in electron impact mode, with transfer line, trap and manifold temperatures of 350°C, 250°C and 50°C, respectively. Data acquisition sampling frequency was 60 H z for a scan range of 50 to 650 m/z. Samples were injected via a Varian 8200 auto-sampler (Walnut Creek, C A , U S A ) fitted with a 10 u L syringe. For the underivatized samples, 1 uL was injected with a split ratio of 1:60, and injector temperature of 320°C. A continuous flow rate of 1.6 mL/min of helium (chromatographic grade) was used with a temperature profile starting at 50°C for 3 min followed by a 10°C/min ramp to 340°C, which was held for 36 min, then finally a 10°C/min ramp to 360°C (thermal maximum of D B - X L B ) . For the T B D M S derivatives, 2 u L was injected with a split ratio of 1:20, and an injector temperature o f 300°C. A continuous flow rate of 1.6 mL/min 26 was used with a temperature profile starting at 50°C for 3 min followed by a 6°C/min ramp to 340°C, which was held for 20 min, then finally a 10°C/min ramp to 360°C (thermal maximum of D B - X L B i t d ) . 2.2.5 G C - F I D Analysis of Total Extractives The gas chromatograph - flame ionization detector used was a HP-5890A (Palo Al to , C A , U S A ) equipped with the identical column used on the Varian G C - M S (section 2.2.4), a deactivated splitter sleeve (packed with glass wool) designed to be used with the 5890A injector. The FID was set at 350°C, to be analogous to the transfer-line at 350°C in the G C - M S method. Samples were injected via a H P 7673A auto-sampler (Palo Al to , C A , U S A ) fitted with a 10 u L syringe. Two microlitres of the underivatized samples were injected with a split ratio of 1:20 and injector temperature of 320°C. A constant head pressure of 483 mBar provided a flow rate of 1.2-1.9 mL/min of hydrogen (UHP grade) carrier gas throughout the analysis. The air, make up gas (UHP grade N 2 ) , and hydrogen fuel ( U H P grade) were set at 350 mL/min , 10 mL/min, and 30 mL/min, all within settings recommended by the manufacturer for the 5890A. The FID flame was run rich (high fuel to air ratio) to enhance the sensitivity of the higher molecular weight, less combustible components (ie steryl esters), as recommended by the manufacturer. The temperature profile used for the total extractives analysis was identical to that used in the G C -M S method underivatized extractives (section 2.2.4). 2.2.6 Solid-Phase Extraction to isolate S E / W The steryl esters and waxes fraction was isolated from the total acetone extracts using a slightly modified version of a solid-phase extraction (SPE) method developed by Chen et al. (1994a). Initially, the freeze-dried total acetone extracts (~70 mg) in the clear glass vial were extracted with H P L C grade chloroform (3x2 mL) , run through a 3 cc cartridge column loaded with magnesium sulphate (reagent grade), and the solvent was evaporated under a stream of nitrogen 27 gas while heating (40°C). The residue was then freeze-dried overnight and weighed (only for HPLC fractionation). Cholesteryl palmitate (99%; Sigma, St. Louis, MO, USA) was added as a recovery standard at 5% by weight (only for HPLC fractionation). The freeze-dried residue was then extracted with chloroform (3xl00uL) and loaded onto a 3 cc Varian Bond-Elut™ aminopropyl-silica gel column (Harbor City, CA, USA) and eluted with 7 mL of 1:5 CHCl3:Hexane (all solvents were HPLC grade) to produce fraction 'A1. This fraction, containing triglycerides, and steryl esters / waxes (SE/W), confirmed by TLC (250-um layer, AL SIL G/UV plates; Whatman, Maidstone, Kent, England) developed in hexane-diethyl ether-acetic acid (70:30:1 v/v/v; all solvents reagent grade) and visualized by ammonium molibdate ((NH4)6Mo7024«4H20 dissolved in 4:1 EtOH:H2S04; all chemicals reagent grade - Fisher Scientific, Fair Lawn, NJ, USA) scorching, was then evaporated with nitrogen and heat (40°C) as before. The residue from fraction 'A' was then reconstituted in 0.5ml of HPLC grade hexane, loaded onto a new Bond-Elut™ column, and extracted with 12 mL of HPLC grade hexane to elute the SE/W fraction. The purity of this fraction was confirmed by TLC as before. Finally, the fraction was evaporated with nitrogen and heat (40°C), and then freeze-dried overnight. 2.2.7 Argentation Chromatography Fractionation of SE/W Fraction Flash chromatography (FC) and TLC fractionation of the SE/W fraction was completed by Alex Serreqi as reported in Serreqi et al. (1999). Argentation-silica gel FC was used to separate the total SE/W fraction by polarity and degree of saturation into 3 fractions referred to as fractions I, II and III. Fraction II was further separated into II-1 and II-2 using argentation-silica TLC. The fractions were collected by scraping the bands from the plate. Furthermore, a purified hydrolysate of Fraction II-1 was prepared by alkaline hydrolysis of this fraction followed by the isolation of the major triterpene in this hydrolysate as described in Serreqi et al. 1999. 28 2.2.8 H P L C Analysis of Steryl Esters The total S E / W fraction or cholesterol ester standards (99%; Sigma, St. Louis, M O , U S A ) were dissolved in hexane or chloroform (solvents were H P L C grade) at various concentrations. Next 5- lOuL of this solution were injected on to a Mil l ipore Waters 600 H P L C (Milford, M A , U S A ) equipped with a Waters Spherisorb™ S3 ODS2 C18 reverse phase column (4.6 x 150 mm, 3 pm pore size; Milford, M A , U S A ) , a Varex Mark IIA E L S D (Rockville, Maryland, U S A ) and a three way stream-splitting valve (set at 1:1 volumetric split for fraction collection). The nebulizer gas flow rate and drift tube temperature were previously optimized for split and non-split modes. The analyte was isocratically eluted with acetonitrile-tetrahydrofuran (67:33; H P L C grade and degassed by 0.22 pm filtration) at a flow rate of 1.00 ml/min. Chromatograms were acquired digitally on a computer running a L a b V I E W acquisition program and the resultant peaks were viewed with Peak-Fit 4.02 software (SPSS, Chicago, IL, U S A ) . Time verses response data was filtered at 1% with a Loess algorithm to remove spiking due to electronic noise. 2.2.9 G C - M S Analysis of S E / W Fraction The total S E / W fraction, F C , T L C or H P L C fraction of the total S E / W , or the synthetic steryl ester standards were dissolved in hexane or chloroform (both solvents were H P L C grade) at various concentrations. Subsequently, 1-5 u L of these solutions were injected onto the G C - M S system described previously (section 2.2.4) with a split o f 1:60, and injector temperature of 300°C. A continuous flow rate of 1.6 mL/min of helium (chromatographic grade) was used with a temperature profile modified from Sithole (in l i t t , 6 July 1998) staring at 200°C for 1 min followed by a 15°C/min ramp to 300°C, then a 5°C/min ramp to 320°C, and finally, a 2°C/min ramp to 325°C which was held for 48 min. 29 2.2.10 Standard Compounds Alpha-amyrin was obtained from ICN Pharmaceuticals Inc. (Costa Mesa, CA, USA), and all other authentic standards were obtained from Sigma (St. Louis, MO, USA) unless otherwise stated. For quantification purposes, standards of 99% purity or greater were used. The only exception was stigmasterol (95%; Aldrich, Milwaukee, WI, USA). All standards were weighed (±0.01 mg) using a Mettler AE240 analytical balance (Mettler, Switzerland) promptly after being stored over silicon bead desiccant at the recommended storage temperature for each compound. Solutions were made up to 50.00 mL - 250.00 mL in a volumetric flask at room temperature in HPLC grade hexane or chloroform, or certified ACS acetone. Since steryl ester compounds are not readily available from chemical suppliers, synthetic standards of various steryl esters were synthesized by Alex Serreqi as outlined in Serreqi et al. (1999). 2.2.11 Assessment of Potential Sources of Contamination Various potential sources of sample contamination were assessed for their contribution to artifacts in the total extractive chromatograms. These included, the Ziplock™ bags used to store the 10 mm cores prior to extraction, the soxhlet extractors, storage vials and caps, Teflon™ lined Tygon™ tubing used for the addition of internal standard, and GC autosampler vials and septa. All potential sources were extracted with minimal amount of acetone, or dipped in acetone solution for a few minutes. The resultant extracts were then analyzed by GC-MS using the underivatized extractives method (section 2.2.4), and any resultant peaks were compared to peaks of similar retention time in the sample chromatograms. 30 2.2.12 Analysis of Salicylic Acid ; Assessment of salicylic acid glycoside content in the extracts was performed by determining the quantity of free salicylic acid [1.14] (aglycon) by GC-MS following hydrolysis of the glycoside. 2.2.12.1 (3-Glucosidase Hydrolysis Using freeze-dried acetone extracts from CH-F6-02 (as it was abundant unlike the core material), p-nitrophenyl P-D-cellobioside (99%; Sigma, St. Louis, MO, USA), crude P-glucosidase as Novozym 188 (Novo Nordisk, Franklinton, NC, USA), and heptadecanoic acid (99%; Sigma, St. Louis, MO, USA), solutions as seen in Table 2.1, were dissolved in 16 mL glass vials using 10.00 mL of deionized water. Table 2.1. Solutions for P-glucosidase hydrolysis in 10.00 mL of deionized water with 2.4 mg of heptadecanoic acid as an internal standard Solution Weight of Amount of P-Glucosidase Substrate (mg) (uL) Extractive blank 4.7 0 p-nitrophenyl P-D-cellobioside blank 2.4 0 Enzyme blank 0 25.0 Extractive reaction 4.7 25.0 p-nitrophenyl P-D-cellobioside reaction 2A 25.0 The solutions were mixed and stored capped at room temperature for 100 minutes to allow sufficient time for the hydrolysis of P-glucosidic bonds, and release of the aglycon. Subsequently, the water was removed by freeze drying for two days. Then, 3 mL of certified ACS acetone was added, and the solution passed through a magnesium sulphate cartridge (10 cc cartridge 1/3 packed with reagent grade MgS04) under vacuum to remove traces of water. The solutions were then analyzed by GC-MS by the underivatized total extractives method (section 2.2.4). A salicylic acid [1.14] standard (99.9%; Fisher Scientific, Fair Lawn, NJyUSA) was also run for unequivocal identification of salicylic acid [1.14] in the samples. 31 2.2.12.2 Trifluoroacetic Acid Hydrolysis The procedure used for TFA hydrolysis of glycosides was a modification of the method reported by Jonasson et al. (1997). Approximately 10 mg (±0.1 mg) of freeze-dried extract was added to 2.5 mL of optima grade ethyl acetate, sonicated for 10 minutes, cooled in the refrigerator (~5°C), and filtered through a pasture pipette packed with glass wool to remove all suspended solids. One millilitre of the filtrate was used for monosaccharide analysis (see Section 2.2.12.3), another 1 mL fraction was added to a 1.5 mL reacti-vial for TFA hydrolysis and the remainder was used for GC-MS analysis (underivatized extractives method section 2.2.4). In the reacti-vial, the ethyl acetate was evaporated with nitrogen and heat (50°C). One millilitre of 2 M trifluoroacetic acid (prepared with deionized water from 99.9% TFA; Fisher Scientific, Fair Lawn, NJ, USA) was added, the vial was capped with a Teflon™ septum cap, then heated at approximately 75°C for 4 hours. Excess TFA was removed by evaporation with nitrogen while cooling the mixture from 75°C. One millilitre of ACS certified acetone was added to the hydrolysate and a sub-sample of the solution was taken for direct GC-MS analysis (underivatized extractives method section 2.2.4), and another for monosaccharide analysis. 2.2.12.3 Monosaccharide Analysis Monosaccharide analyses were performed on the ethyl acetate extracted acetone extracts before and after TFA hydrolysis (see Figure 2.2). Analyses were performed by Paprican staff using a standard protocol involving alditol acetate derivatization (modified from Blakeney et al. 1983) followed by GC-MS analysis. The GC-MS conditions were as stated in 2.2.4 except utilizing a DB-225 column (30 m x 0.25 mm; J&W Scientific, Folsom, CA, USA) and samples were injected (1 uL) with a 1:40 split and 260°C injector, starting with column temperature of 175°C held for 1 min, then a 6.0°C/min ramp to 240°C which was held for 1 min. The resultant 32 monosaccharides were identified by comparisons to authentic standards (95-99%; Aldrich, Milwaukee, Wl, USA) analyzed under the identical conditions. TFA H Y D R O L Y S A T E Figure 2.2. Scheme for salicylic acid [1.14] analyses. E 2 versus Ei should provide information on (3-glucosidic aglycons that are GC analyzable; T 2 versus Ti should provide information on moieties in ethyl acetate soluble, acid labile compounds; S2 versus Si should provide information on sugars bound to aglycons released by TFA hydrolysis. 2.2.13 Calculations and Statistics Relative Percent Difference (RPD) was calculated as the difference in the two values divided by their mean and multiplied by 100%). Resolution of chromatographic peaks was calculated by separation between peaks divided by average peak width (Atr/wavg.) as in Harris (1991). All error ranges in numerical data equate to the 95% (ie a = 0.05) confidence intervals (CI) calculated using Microsoft Excel™ 97. Relative error is equal to half of the 95% CI divided by the mean 33 value and multiplied by 100%. Regressions were performed by Excel 97™ for a quadratic (k=2) model. See Appendix I for relevant statistical formulae. 2.3 RESULTS AND DISCUSSION 2.3.1 Sample Preparation and Extraction 2.3.1.1 Freeze Drying It was found that grinding the chips or cores in a dry state simplified grinder cleaning and minimized cross contamination between samples and thus, freeze drying was performed on the solid wood prior to grinding. Experiments run to determine the sufficient amount of time for freeze drying were performed using wood chips since the 10 mm wood cores in this experiment were not abundant (each core only provides 5-10 g of freeze-dried material). Table 2.2 illustrates that freeze drying solid aspen wood for 24 h was sufficient to dry the wood to a constant weight. All relative percent differences between the weight loss at 24 h and that at 48 h presented in Table 2.2 were less than 0.2%. Furthermore, for one sample (CH-F7-03 trial#l), the weight actually was shown to increase with added freeze drying suggesting that the weight loss difference between the 24 h and 48 h drying was within the error introduced in weighing the samples. 34 Table 2.2. Percentage weight loss in aspen wood chips (-40 g wet) after freeze drying at -2x10 mBar and -45°C Sample (trial #) 24 hour weight 48 hour weight Relative percent loss loss difference between (% green wood) (% green wood) 24 h and 48 h (%) CH-K1-05 (trial #1) 44.995 45.034 0.09 CH-K1-05 (trial #2) 45.220 45.235 0.03 CH-F6-02 (trial #1) 44.007 44.068 0.14 CH-F6-02 (trial #2) 44.324 44.342 0.04 CH-F7-03 (trial #1) 45.834 45.824 -0.02* CH-F7-03 (trial #2) 44.776 44.776 0 * Appears to have gained weight but difference is so small that it is within error of method Also apparent in Table 2.2 is a measurable difference in moisture loss between material from the same tree. This is not surprising as the sub-samples from the same parent material may have had slightly different initial moisture contents as a result of slight compositional differences. However, this will not affect the accuracy of the final dry weight determination as long as the freeze drying was performed to completion. 2.3.1.2 Soxhlet Extraction Experiments were run to determine how complete and reproducible the soxhlet extractions were, as well as, to determine the difference in quantity and quality of extracting chipped wood verses milled wood, and the results are presented in Table 2.3 and 2.4. Table 2.3. Repeat extraction of milled cores to determine the completion of extraction using samples CO-F3-08 (relatively sound wood) and CO-K2-03 (relatively decayed wood) Sample Percent extractives Percent extractives Relative percent after 8h after 13h difference between 8 h (% fd wood) (% fd wood) and 13 h (%) CO-F3-08 1.980 2.022 2.10 CO-R2-03 3.652 3.737 0.62 The extractions were all visually monitored to be at least 4-5 cycles/h. However, some of the variation introduced by extraction may be due to differences in extraction rates between the 35 samples, which may be equivalent to differences in reflux times at a constant rate. Table 2.3 shows that after the 8 hour extraction, an additional 5 hours (total 13 hours) extraction removes very little or negligible amounts of extractives. GC-MS analysis (underivatized total extractives method section 2.2.4) of the additional extractives obtained from 8-13 hours revealed that the components extracted are mainly sinapyl [2.13] and coniferyl [2.8] alcohols as well as flavonoids [1.8], particularly in the heavily decayed sample. These results suggest that an extraction for 8 h at 4-5 cycles/h removes virtually all the extractives of interest at high efficiency. Furthermore, many extractives that are of interest, particularly the esterified fatty acid compounds (ie steryl esters, waxes, and glycerides), are thermally labile and thus, a shorter extraction time would reduce the decomposition and loss of these components from the samples. A reasonable compromise between accuracy / completion and preservation of thermally labile components was therefore reached at 8 hours and 4-5 cycles/h. Table 2.4 illustrates that there is some variation (~5%) in percent extractives between replicate extractions of wood samples. As in the freeze drying experiments, the replicate extractions seen in Table 2.4 were performed using chips since the cores in this experiment were not abundant. Table 2.4. Extraction duplicates for raw wood chip and milled wood samples from CH-F6-02 (~10 g samples, 8 h, 200ml acetone, 4-5 cycles/h) Sample (trial #) State Weight of Percent Relative Percent extractives extractives difference between (mg) (% fd wood) trials (%) CH-F6-02 (trial #1) Milled 345.5 3.266 CH-F6-02 (trail #2) Milled 360.3 3.433 4,99 CH-F6-02 (trial#l) Chips 129.3 1.292 CH-F6-02 (trial #2) Chips 122.6 1.167 5X)8 Blank , N/A Ll 0.01 l a N/A aPercent of blank residue based on 10 g of fd wood meal used in non-blank runs. 36 The observed error between extraction duplicates includes errors due to freeze drying and weighing. However, since both these errors would have contributed « 1 % to the results, the resultant error seen in Table 2.4 was caused mainly by the extraction process. Nevertheless, the overall reproducibility of soxhlet extraction observed in this study was well within the repeatability of 6.5% published by CPPA standard G.13 and G.20 for soxhlet extraction of wood. Hughes and Thurbide (1998) reported a relative standard deviation of only 1.1% in the soxhlet process which implies a repeatability of far less that 6.5%. However, this value was obtained with TMP fibre (spruce) which is more uniform in composition than milled wood. Hence, a large part of the error introduced by the extraction process was most likely due to the heterogeneous nature of wood. The blank run reported in Table 2.4 illustrates that the weighing process and possible extraction of contaminants from the soxhlet apparatus (although minimized by rigorous cleaning between samples) may account for <1% of the error observed between duplicates. The results in Table 2.4 for CH-F6-02 show that the extraction of milled wood removes 2.7 times more material than using intact chips. This was the primary reason for grinding the samples into a higher surface-area / volume milled wood. It is hypothesized that the grinding disrupts intact wood cells so that the lumen will be more efficiently extracted. To assess the differences in extractives quality as a result of the different material forms (wood chip or milled wood), a GC-MS profile of the total extractives from each material form was run and is shown in Figure 2.3. 37 0) C/3 Retention Time (min) Figure 2.3. Gas chromatography profile of total extractives from chip (A - offset by +0.5 min) milled (B) wood. luL at 5.000 mg/mL of CH-F6-02. GC-MS: 50°C held for 3 minutes, 10°C/min ramp to 340°C, held for 30 minutes, and a 10°C/min ramp to 360°C, with an injector temperature of 320°C. A 1:60 split was used, and a 690 mBar head pressure of helium carrier gas. MS manifold, trap, and transfer line temperatures were 50°C, 200°C and 300°C, respectively. It is apparent from Figure 2.3 that the resultant extractives from both material forms was very similar in quality. Both profiles contain all the same major components or chromatographic features. However, there was some variation in the relative amounts of certain components between the two chromatograms as would be expected considering that milling will expose the interfibre surfaces (ie middle lamella and interior of fibre wall) to more thorough extraction. 38 2.3.1.3 Repeatability of 10 mm Core Measurements In order to measure over all experimental error, duplicate cores were removed from certain sample trees over the nine clones. This repeatability included all experimental error, as well as error due to within tree variation, the later of which should be quite small for cores taken adjacent to one another according to the results of DeBell et al. (1997). Thus, this value should give a good estimate of the overall reproducibility in the 10 mm core extractives evaluation method. The repeatability of the total weight of acetone extractives analysis based on the duplicated increment cores was determined to be 4±1%. As has already been shown in section 2.3.1.2, this amount of variability may be attributable mainly to experimental error during soxhlet extraction. 2.3.2 Extractives Analysis Techniques 2.3.2.1 Analysis of Total Extractives by GC-MS The GC-MS method developed in this study provided excellent separation and identification of most extractives in aspen including those relevant in the pulp and paper processing of this species. Forty-four compounds have been identified and an additional 26 classified in the single chromatogram. Figure 2.4 illustrates the total extractives chromatogram obtained in this work, and highlights some of the major peaks present in this profile (see Appendix II for complete list of components). Several compounds implicated in pitch formation during the pulping of aspen (ie fatty acids, sterols / triterpenes, steryl esters / waxes and triglycerides) are particularly well resolved and represented in quantifiable amounts. 39 G O OH .tu u 17 Fatty Acids& Monoglycerides 5 Low M W Phenolics 2 3 Sterols 12 Waxes& Diglycerides 10 9 11 3 15 16 Steryl Esters Triglycerides 10 20 30 40 50 Retention Time (min.) 60 Figure 2.4. Typical profile for aspen wood acetone extractives underivatized by G C - M S . l u L at 5.000 mg/mL of CO-D4-02 by underivatized extractives method section 2.2.4. 1 = ;7-hydroxybenzoic acid[2.6]; 2 = coniferyl alcohol[2.8]; 3 = palmitic acid[2.12]; 4 = sinapyl alcohol[2.13]; 5 = internal standards (heptadecanoic acid); 6 = linoleic acid[2.15]; 7 = eicosanoic acid[2.17]; 8 = docosanoic acid[2.18]; 9 = sakuranetin[2.20]; 10 = naringenin[2.22]; 11= p-sitosterol acetate[2.23]; 12 = p-sitosterol[2.27]; 13 = 24-methylenecycloartenol[2.32]; 14 = unidentified sterols; 15 = various waxes; 16 = unidentified steryl esters; 17 = palmitate (C16:0) steryl esters[2.34], [2.35], [2.36], and [2.37]; 18 = stearate (C18:0) steryl esters [2.38], [2.39], [2.40] and [2.41]; 19 = unidentified triglyceride; 20 = butyrospermol eicosanoate [2.42]; 21 = trilinolein [2.43]. The separation of compounds in aspen acetone extractives relevant to this study was maximized by careful optimization of temperature programs and carrier gas flow conditions. The D B - X L B column used is a low polarity column (between the polarity of a D B - 5 and a DB-17) with a bonded and cross-linked stationary phase, featuring excellent inertness for active compounds and a high temperature limit of 360°C ( J & W Scientific 1998) making it an ideal column for the analysis of underivatized semi-volatile wood extractives. 40 Gutierrez et al. (1998) determined the effect of column length on the resolution and detection of eucalyptus extractives by cGC and concluded that one must seek an optimum of maximized resolution (brought about by an increase in column length) versus minimized exposure of the analyte to high temperatures (brought about by a decrease in column length). The cGC method developed in this study utilizes a moderate temperature ramp of 10°C/min with a relatively short column (10 m), together with a long temperature hold at 340°C to allow for the elution of high MW and boiling point compounds such as steryl / triterpene esters and triglycerides. This method was sufficient to provide high resolution without severe decomposition of thermally labile components near the end of the run. Also, to ensure that no material eluted after 70 minutes, an analysis using a longer final temperature hold at 340°C (>70 minutes) was also performed on a few samples. With the prevalence of auto-samplers and more robust cGC columns, a run time of 70 minutes is reasonable for the detailed analysis of extractives in a large number of samples. However, it was determined that this analysis time could be reduced by half without a significant loss in resolution by doubling the temperature ramp rates (see Figure 2.5). Although the resolution of low molecular weight components (ie simple phenolics, fatty acids and sterols) was reduced, the resolution of the main pitch forming compounds in aspen, the steryl esters and waxes, was relatively unaffected. Thus, this fast cGC method may be useful for rapid and / or routine monitoring of various deresination treatments on aspen wood chips or logs including biological treatments (ie Cartapip™ 97 developed by Blanchette et al. 1991) and seasoning. 41 Fatty Acids& Monoglycerides Vi a o OH Vi <D U 3 L o w M W Phenolics Sterols Waxes& Diglycerides Steryl Esters 10 15 20 Retention T i m e (min.) Triglycerides Figure 2.5. Typical profile for aspen wood acetone extractives underivatized by fast G C - M S . l u L at 5.000 mg/mL o f CO-D4-02. G C - M S : identical to section 2.2.4 except temperature ramps, start at 50°C, held for 3 min, ramped at 31.6°C/min held for 17 min, then at 25°C/min to 360°C (total time = 30 min). The separation of the high boiling point lipophilic constituents such as steryl esters and triglycerides shown in Figure 2.5, is superior to that seen in several studies utilizing c G C for similar compounds and analysis time (Wallis and Wearne 1999; Orsa and Holmbom 1994; Del Rio etal. 1998). 2.3.2.2 Development of a H P L C Method for Steryl Esters The rationale for the development of a stand-alone H P L C method for the analysis of steryl esters is that certain steryl esters are known to be thermally labile and may decompose by G C analysis due to the high temperatures (>100°C) employed to achieve separation with this technique. In 42 Figure 2.6, the thermal decomposition of cholesteryl palmitate is illustrated by the increase in the ratio of decomposition products (DP; including palmitic acid [2.12] and cholesterol derivatives) to cholesterol palmitate (CP) with exposure to heat (150°C). 2 0 4 0 6 0 8 0 1 0 0 1 2 0 Heating time at 150°C (min) Figure 2.6. Thermal decomposition of cholesteryl palmitate at 150°C sealed in reacti-vials. 0.2 uL of 1.00 mg/mL in hexane by underivatized extractives method section 2.2.4. DP = peak area of decomposition products, CP = peak area of intact cholesteryl palmitate. An improved HPLC method for the analysis of steryl esters was developed based on that previously reported by Evershed et al. (1987). After careful optimization in terms of both detector response (Appendix IV), and mobile phase composition (isocratic elution, acetonitrile-tetrahydrofuran at 67:33), the best resolution and sensitivity obtained for cholesteryl esters (used as readily available model compounds for steryl esters) is illustrated in Figure 2.7. 43 o Q (—1 —I 1 h— 7 8 9 Retention T ime (min) 1 0 n Figure 2.7. Typical HPLC-ELSD separation of cholesteryl esters. 5ul injected at 1.00 mg/mL each for (1) cholesteryl myristate (C14:0), (2) cholesteryl palmitate (C16:0) and (3) cholesteryl stearate (C18:0). HPLC: drift tube temperature = 155°C, nebulizer gas flow rate = 35 mL/min - otherwise as in section 2.2.8. The relative elution of components as shown in Figure 2.7 is consistent with cholesteryl ester results from Evershed et al. (1987). A short run time (ie 11 minutes with C14:0, C16:0 and C18:0 cholesteryl esters) is a highly desirable feature for a routine analysis method such as those used in monitoring process effluents. For quantitative purposes, the advantages of the ELSD include rapid analysis, high reproducibility (relative standard deviation around 1%), and low detection limits in the middle or upper nanogram-range (Onken 1998). In this work, the detection limit, or amount of compound required to produce a peak with signal to noise ratio of 2 (Harris 1991), of the optimized ELSD was found to be 44.8 ng based on cholesteryl palmitate determined by extrapolation of the 44 calibration curve (Figure 2.8 A). The detection limit, based on cholesteryl palmitate, by GC-MS was much higher at 170 ng (determined by extrapolation of calibration curve seen in Figure 2.8 B), which may in part be due to thermal degradation of this thermally labile compound during GC. Also, the calibration curves for cholesteryl palmitate shown for both ELSD and MS detection methods in Figure 2.8 illustrate the improved precision that HPLC-ELSD offers over GC-MS for the quantitation of cholesteryl palmitate and potentially steryl esters. The R2 for the ELSD second order polynomial curve (R2 = 0.9999) is slightly higher than that of the MS curve (R2 = 0.9960), suggesting that there is more variability in the quantitation of CP by GC-MS than by HPLC-ELSD. The HPLC-ELSD system had a higher analyte capacity and thus, the ELSD calibration curve extends to analyte amounts passed that of the MS detector. Hence, the gain in accuracy with the ELSD may be in part due to the lower uncertainty associated with measuring larger amounts of analyte. However, in Figure 2.8, there is a larger spread of data points about the calibration curve for the MS compared with that of the ELSD, even when comparing similar analyte weights. Ultimately, the ELSD is inherently more precise for the quantitation of cholesteryl palmitate and potentially steryl esters. 45 1.5E+00 1.2E+07 Figure 2.8. Calibration curves for cholesterol palmitate (CP) by HPLC-ELSD (A) and GC-MS (B). (A) was 5 uL of cholesterol palmitate in hexane; HPLC: drift tube temperature = 155°C, flow rate = 35 mL/min - otherwise as stated in section 2.2.8. (B) was 3 uL of cholesteryl palmitate in hexane; GC-MS: 200°C for 1 min, 15°C/min to 300, then, 10°C/min to 330°C, then 7°C/min to 335°C which was held for 60 min; injector was ramped from 230°C to 300°C at 200°C/min at time of injection; 690 mBar of He carrier gas; multiplier voltage was 1850 V; manifold, trap and transfer line temperatures were 50°C, 200°C and 300°C respectively - otherwise as stated in experimental section for SE/W in section 2.2.9). R = coefficient of multiple determination. 46 The resolution for these three cholesteryl esters obtained by HPLC is compared with that obtained by GC-MS in Figure 2.9. Figure 2.9. Resolution comparisons between optimized methods by HPLC (A - from Figure 2.7; offset +12.5 min) and GC-MS (B) methods for steryl ester analysis (2 uL of 1.00 mg/mL for each cholesteryl myristate (1), cholesteryl palmitate (2) and cholesteryl stearate (3) in hexane. GC-MS: 200°C for 1 min, 15°C/min to 300, then, 10°C/min to 330°C, then 7°C/min to 335°C which was held for 60 min; injector was ramped from 230°C to 300°C at 200°C/min at time of injection; 690 mBar of He carrier gas; multiplier voltage was 1850 V; manifold, trap and transfer line temperatures were 50°C, 200°C and 300°C respectively - otherwise as stated in section 2.2.9). Despite the benefits of a shorter run time in HPLC analysis (11 min) as compared with cGC (40 min), the resolution of cholesteryl esters for the optimized HPLC method is approximately half that obtained by the cGC method. As it has been already seen, even in cGC analysis, there is co-47 elution of components in the SE/W fraction of aspen (see Figure 2.4). Hence, the optimized HPLC method, which showed lower resolution for cholesteryl esters than the corresponding GC-MS method, would produce even less pure band separation of the total SE/W fraction in aspen (Figure 2.24 shows a chromatogram of the total SE/W fraction by HPLC). However, the method developed here would be suitable for accurate and rapid analysis of steryl ester mixtures that are relatively simple in composition (probably 2 or less sterols esterified to a variety of normally even numbered fatty acids). 2.3.2.3 Analysis and Possible Origin of Salicylic Acid in Aspen Salicylic acid [1.14] is known to be present as a (5-glucoside in tobacco plants (Horvath and Chua 1994). P-glucosidase is an enzyme that is specific for the cleavage of P-glucosidic linkages such as in p-nitrophenyl P-D-cellobioside used in this study to determine the extent of hydrolysis. The formation of large amounts of p-nitrophenol after enzyme treatment was confirmed by GC-MS. The same p-glucosidase treatment on aspen wood extractives liberated benzoic acid [1.11], but no salicylic acid [1.14] was found by GC-MS analysis (see Figure 2.10). 48 Retention Time (min) Figure 2.10. Benzoic acid [1.11] (#1) and salicylic acid [1.14] retention zone (gray area) in the phenolics region of the total extractives profile by GC-MS before (Ei) and after (E ) (3-glucosidase treatment. 1 uL at 1.50 mg/mL by underivatized total extractives method section 2.2.4. Note: no salicylic acid was found in retention zone (Grey area). These results suggest that salicylic acid [1.14] may not be linked to glucose in a P-linkage. Ethyl acetate extraction followed by TFA hydrolysis released several low MW phenolic moieties including salicylic acid [1.14] and o-coumaric acid [2.2] in all three samples analyzed (CO-D1-03, CO-D1-20 and CO-D2-02). Surprisingly, no increase in free benzoic acid [1.11] was observed after the TFA hydrolysis in all three samples. It is known that there are differences in stability for different glycosidic linkages (Sjostrom and Alen 1999), and the TFA hydrolysis 49 conditions used in this work may not have been sufficient for the hydrolysis of bound benzoic acid. Monosaccharide analysis of the ethyl acetate fractions prior to and following TFA hydrolysis confirmed the release of saccharides potentially originating from glycosides in aspen. In Figure 2.11, the S i chromatogram shows all free monosaccharides present in the acetone extracts. Thus, the increased peaks in the hydrolysate chromatogram (S ) should reflect the increase in sugars resulting from the hydrolysis of non-polymeric (polymeric carbohydrate material would not be soluble in cold ethyl acetate) carbohydrates by TFA hydrolysis. 7 8 9 10 Retention Time (min) Figure 2.11. Results of monosaccharide analysis by GC-MS for ethyl acetate soluble acetone extracts before (Si) and after (S ) TFA hydrolysis. 1 uL for the same amount of starting material (4.00 mg/mL of total extracts) analyzed by monosaccharide method section 2.2.12.3. 1 = xylose, 2 = rhamnose, 3 = mannose, 4 = galactose, 5 = glucose, 6 = Unknown, 7 = Inositol (internal standard). All identified peaks confirmed by authentic standards except rhamnose (see Appendix III for mass fragmentography). Sample: CO-D1-03. 50 As seen in Figure 2.11 glucose, rhamnose and xylose were seen to increase following the TFA hydrolysis. Jonasson et al. (1997) also reported an increase in non-polymeric glucose and rhamnose following TFA hydrolysis of MeOH / water extracts of aspen known to contain phenolic glycosides. From the results presented previously, it is possible that salicylic acid [1.14] in aspen is not linked to glucose in a [3-linkage, or that it is merely a derivative of acid hydrolysis from known aspen glycosides, salicortin [1.16] and tremulacin [1.17], as suggested by Pearl and Darling (1971b) (see Figure 2.12 for reaction scheme). F(R=H) E G(R=2-0-benzoyl) Figure 2.12. Suggested origin of salicylic acid in aspen from known glucosides salicortin (A) / tremulacin (B) via acid (HC1) dehydration to co-salicycloylsalicin (C) / salicyloyltremuloidin (D) followed by acid hydrolysis to salicylic acid (E) and salicin (F) / tremuloidin (G) (Pearl and Darling 1971b). 51 However, no direct evidence of the hydrolytic conversion of C/D to F/G and E (Figure 2.12) was presented in the Pearl and Darling (1971b) study. If salicylic acid [1.14] in aspen is not associated with a (3-glucoside such as in other plants (Horvath and Chua 1994), it is possible that it is linked to rhamnose as a rhamnoside since this sugar was observed after hydrolysis of the ethyl acetate soluble extractives. If this is the case in aspen, salicylic acid [1.14] may still be involved in systemic acquired resistance against various pathogens including P. tremulae. In either case, the TFA hydrolysis method presented is a suitable method for the analysis of salicylic acid [1.14] in aspen. 2.3.3 Identification and Quantification of Compounds by GC-MS 2.3.3.1 Identification of Extractives by GC-MS Eluting compounds in the total extractives profile were identified based on unequivocal match with mass spectra of the 1998 National Institute of Standards and Technology (NIST '98) mass spectral library. The "fit" search method was used as recommended by Varian when co-eluting peaks are present with a high fit result being 800 and above to a maximum of 1000. Agreement between elution order and molecular properties (ie molecular weight, polarity, etc.), and the possibility of artifacts introduced from various sources of contamination (which were assessed as described in section 2.2.11) were considered. Any suspicious compound assignments made by the NIST matching program were rejected. A typical example of an NIST assignment is represented in Figure 2.13. 52 *i3 a> a> o N cj ' r " 1 Co "e3 121 138 ° -5 T3 a> ' 1 - 1 Co 13 -3 is < 65 93 39 Ll_ [M-OH]+ — 121 [M]+ 138 [M-COOH] + 65 93 o PI Q 100 B 200 300 Mass / Charge (m/z) Figure 2.13. Identification of a low molecular weight phenolic in the total extractives profile by NIST '98 mass spectral match. A = mass spectrum of compound at 11.9 min (RRT = 0.6758) from underivatized total extractives chromatogram (sample: CO-K2-06); B = NIST '98 mass spectrum for />hydroxybenzoic acid [2.6]; C = difference spectrum, fit = 978. Note: arrows illustrate fragment(s) of molecule that is/are lost to result in specified m/z value. Para-hydroxybenzoic acid [2.6] has been previously reported in aspen (Pausler et al. 1995). Other compounds that eluted near p-hydroxybenzoic acid [2.6] included 4-hydroxy-2-methylacetophenone [2.4] (fit = 879), and j^ -hydroxy-hydrocinnamic acid [2.7] (fit = 931). These compounds are very similar in chemical structure, and thus their similar retention time is not surprising. When the fit was less than 800, the identification was confirmed by running authentic standards (when available) under the identical GC-MS conditions used with the samples. This provided 53 both retention time and mass spectra confirmation of the compound. All ten compounds confirmed by authentic standards are indicated under "ID method / fit" in the list of all identified components (Appendix II). In some cases, complete compound identification (and assignment of a chemical structure) was not possible based on the total extractives profile due to the limit of the NIST '98 mass spectral library, severe co-elution of compounds, and / or because only trace amounts of the compound were present in the sample resulting in a spectrum affected by high background noise. However, if the compound was found to exist in every aspen sample, and did not appear to be a contaminant (by comparison to extracts from several potential contamination sources - section 2.2.11), compound identification was attempted by mass spectral fragmentography and retention time groupings. Mass fragmentography was used to successfully identify several compounds (see Appendix III for fragmentography illustrations). In several other cases, complete compound identification was not possible based on the mass spectrum from the total extractives chromatogram. However, these compounds were classified into various broad chemical classes (referred to as "MS classification" in Appendix II) based on general mass spectral features characteristic of certain chemical class including the fatty acids, flavonoids, sterols / triterpenes, diglycerides, and triglycerides. An example of this method of classification is shown in Figure 2.14 for an unknown sterol / triterpene. The fragmentation of sterols (ie P-sitosterol [2.27]), is conveniently divided into ions resulting from the cleavage of the ring system and those resulting from side chain losses (Dey and Harborne 1991). This is illustrated in Figure 2.14 where the three predominant peaks occurring between 350 m/z and 450 m/z result from fragments produced by side chain losses, and 54 predominant peaks occurring at m/z values below 350 m/z result primarily from fragments produced by cleavage of the sterol ring system. |M - OH - CH31+ 2tr 162 31 105 57 i l l ill 256 I Ill' |M-C7HnO]+| 273 304 I Ill U I Illl 355 .mk , ill! JuL 100 2 0 0 3 0 0 4 0 0 Mass / Charge (m/z) Figure 2.14. Mass spectrum of P-sitosterol [2.27] authentic standard (60% pure) (A) and classification of an unidentified sterol (B) based on general mass spectral fragmentography. 1 = peaks arising from loss of side chain; 2 = peaks arising from ring system cleavage. Note: arrows illustrate fragment(s) of molecule that is/are lost to result in specified m/z value. These general fragmentography patterns served to complement retention time data which, in some cases, are used as the sole basis for chemical classification of unidentified compounds (Nugent et al. 1977; Sithole et al. 1992; Orsa and Holmbom 1994). Using only retention time grouping in the classification of unknowns into various broad chemical classes will undoubtedly 55 lead to false classifications since several chemical classes were shown to overlap in this work (see Appendix II). Additional evidence in support of the more difficult classifications for certain glycerides (saponifiables), which coeluted with steryl esters (unsaponifiables), was obtained by examining the disappearance of the glyceride peaks (saponifiables) after a 0.5 M potassium hydroxide (KOH) hydrolysis in ethanol at 60-70°C for 4 hours (see Appendix V for chromatograms). Finally, the steryl esters that eluted in the total extractives GC-MS profile were identified using additional chromatographic techniques prior to GC-MS analysis as will be presented in the following section 2.3.3.2 Identification of Steryl Esters by AC-GC-MS A typical total extractives profile and corresponding SE/W fraction profile for aspen is shown in Figure 2.15. 56 10 20 30 40 50 60 70 Retention Time (min) Figure 2.15. GC-MS RIC profiles of total extractives (A) and the resultant SE/W fraction (B) obtained by solid-phase extraction (SPE). (A) was luL at 5.000 mg/mL of CO-Kl-04 by underivatized extractives method section 2.2.4. (B) was luL at 1.00 mg/mL of CO-K1-04 SE/W fraction by SE/W method section 2.2.9. The first two fractions (I and II) resulting from argentation chromatography of the SE/W fraction yielded several peaks which corresponded to those in the total SE/W fraction profile as seen in Figure 2.16. 57 20 30 40 50 Retention Time (min) Figure 2.16. GC-MS RIC profiles of total SE/W fraction (A) and argentation chromatography fractions 1(B) and 11(C) of total SE/W fraction. (A) was luL at 1.00 mg/mL of SE/W fraction from CO-K1-04 and (B) and (C) were luL at 1.00 mg/mL for SE/W fractions I and II of CH-K1-05 all run by SE/W method 2.2.9. These nine peaks (numbered in Figure 2.16) in the steryl ester region of the total extractives profile matched well with those from the argentation fractions by mass spectra and retention time. From the mass spectra of these nine compounds it becomes apparent that the mass spectra are very similar for three groups of triplets (2, 6 and 10; 3, 7 and 11; and 4, 8 and 12 - see Appendix VI for mass spectra). 58 The EI mass spectrum of a steryl ester is characteristic of the sterol moiety derived from the initial loss of the fatty acid moiety ([M-RC02H]+) in the EI ionization process (Evershed et al. 1987). Furthermore, the relative gas chromatographic retention time of various cholesteryl esters will give information on the fatty acyl group (Evershed et al. 1987). A curvi-linear relationship was shown to exist between retention time by cGC and fatty acyl carbon number for fatty acyl esters of lanosterol (3P-hydroxy-8, 24-lanostadiene) on the low-polarity DB-XLB column used in this work (see Figure 2.17). The retention times of the lanosteryl esters increased with an increase in chain length of the fatty acyl moiety, whereas unsaturation of the fatty acyl moiety had the effect of decreasing the retention time slightly, as seen by comparison of the retention time of the stearate (C18:0) with that of the oleate (C18:1) esters. 0.9 -I 1 1 1 1 13 15 17 19 21 Carbon # on Fatty Acid Moiety Figure 2.17. Relative retention by GC-MS versus fatty acyl carbon number for lanosteryl (3P-hydroxy-8, 24-lanostadiene) esters: lanosteryl myristate (C14:0), lanosteryl palmitate (C16:0), lanosteryl stearate (C18:0), lanosteryl eicosanoate (C20:0), and lanosteryl oleate (C18:l). All runs were luL at 0.50 mg/mL of steryl ester in hexane run by SE/W method section 2.2.9. *Relative to lanosteryl myristate (Retention time = 20.152 min). 59 The carbon number versus relative retention time plot shown in Figure 2.17 provides a relative basis for steryl ester fatty acid moiety identification. However, steryl esters with different sterol / triterpene moieties w i l l have slightly different absolute retention times. Peng et al. (1999) found that the main esterified fatty acids in the S E / W fraction of aspen were palmitic (CT6:0) [2.12], linoleic (C18:2) [2.15], stearic (C18:0) [2.16] and eicosanoic (C20:0) [2.17] in order of decreasing amounts. Considering the previous mass spectral matches in the three distinct peak groups (2, 6 and 10; 3, 7 and 11; 4, 8 and 12), the first group should consist of a particular sterol / triterpene esterified with palmitic [2.12], stearic [2.16] and eicosanoic [2.17] acids. The second group should consist of another sterol / triterpene in these three esterified forms, and the third group yet another sterol / triterpene in these three esterified forms. Alpha and p-amyrin [2.31, 2.28], and lupeol [2.44] are the major triterpene alcohols found in the hydrolysate of the S E / W fraction in aspen wood extractives (Peng et al. 1999). Synthesis of a-amyrin esters and lupeol esters from myristate (C14:0) to eicosanoate (C20:0) and comparison to the argentation fractions confirmed the identity of many steryl esters in these fractions (Figure 2.18 and 2.20). 60 l l * V i 1 . / I / TV 20 30 40 50 Retention Time (min) Figure 2.18. GC-MS RIC profiles for argentation fraction I (A) of total SE/W fraction with synthetic a-amyrin myristate (B), palmitate [2.37] (C), stearate [2.41] (D), and eicosanoate [2.47] (D). (A) was luL at 1.00 mg/mL fraction I from CH-K1-05 and (B) to (E) were 2uL at 0.50 mg/mL of each steryl ester in hexane all run by SE/W method in section 2.2.9. *not identified in the sample. The a-amyrin esters were unequivocally identified as shown in Figure 2.18 with arrows indicating retention time and mass spectral match. However, for each a-amyrin ester, especially the palmitate, stearate, and eicosanoate, there is a peak of lesser intensity that occurs at a slightly earlier retention time in fraction I. As seen in Figure 2.19, the mass spectrum of this peak was 61 remarkably similar to that of the a-amyrin esters, together with the fact that they both occur in the same argentation fraction suggests it may be a steryl ester of another amyrin isomer. m/z Figure 2.19. Structures of P-(A) and a-(B) amyrin esters with mass spectra of steryl ester palmitates (C16:0) from argentation fraction I of total SE/W fraction. Note: arrows illustrate fragment(s) of molecule that is/are lost to result in specified m/z value. As mentioned previously, P-amyrin [2.28], another common plant triterpene (Kochhar 1983) was found in the hydrolysate of aspen steryl esters in a prior study. Thus, these three peaks (peak numbers 2, 6 and 10 from Figure 2.16) were assigned as P-amyrin palmitate (C16:0) [2.35], stearate (C18:0) [2.39] and eicosanoate (C20:0) [2.45] based on mass spectral identification of the triterpene moiety and carbon number versus relative retention time (ie Figure 2.17) for the fatty acyl moiety assignments as performed in previous steryl ester identifications. 62 Lupeol palmitate [2.36], stearate [2.40] and eicosanoate [2.46] were unequivocally identified in argentation fraction II as shown in Figure 2.20 with arrows symbolizing retention time and mass spectral agreement between peaks. These compounds correspond to peaks 3, 7 and 11 in Figure 2.16. 20 30 40 50 Retention Time (min) Figure 2.20. GC-MS RIC profiles for argentation fraction II (A) of total SE/W fraction and synthetic lupeol myristate (B), palmitate [2.36] (C), stearate [2.40] (D), and eicosanoate [2.46] (E). (A) was luL at 1.00 mg/mL of CH-K1-05 and (B) to (E) were 2-4uL at 0.100-0.50 mg/mL in hexane all run by SE/W method section 2.2.9. *not identified in the sample. 63 The final major steryl ester peaks found in the total extractives profile were identified by comparing G C / M S profiles of further argentation fractions (fraction II-1, and II-2) of fraction II with that of the total S E / W fraction (see Figure 2.21). 1 20 30 40 50 Retention Time (min) Figure 2.21. G C - M S R I C profiles for total S E / W fraction (A) and argentation fractions II-1 (B) and II-2 (C) of argentation fraction II. (A) was l u L at 1.00 mg/mL of total S E / W from CO-K1-04 and (B) and (C) were l u L at 1.00 mg/mL o f further argentation fractions of fraction II from CH-F7-03 all run by S E / W method section 2.2.9. 64 Fraction II-2 contained the lupeol esters already identified, however, fraction II-1 contained three major peaks that corresponded to peaks (1, 5 and 9 in Figure 2.21) in the steryl ester region of the total extractives profile. In accordance to the steryl ester retention index (Figure 2.17), these steryl esters should consist of a palmitate (C16:0), stearate (C18:0) and eicosanoate (C20:0). Hydrolysis of Fraction II-1 and subsequent isolation of the major sterol / triterpene yielded the chromatogram shown in Figure 2.22. The mass spectrum of this sterol found in fraction II-1 matched well (fit = 831) with that of lanosterol (3 p-hydroxy-8, 24-lanostadiene) found in the NIST '98 library. However, when a lanosterol standard was run by the identical method and similar concentration, the lanosterol standard eluted 0.183 min earlier than the unknown sterol. Furthermore, lanosterol is reported to be an important intermediate in sterol biosynthesis only for non-photosynthetic plants (Rowe 19896). Butyrospermol (tirucalla-7, 24-dien-3p-ol) [2.29], which is an isomer of lanosterol, is known to occur in woody plants (Rowe 19896; Kochhar 1983). 65 [M-CH 3 -OH] + 394 109 241 69 159 411 M + il 1 1 I 299 426 i l 1,1 i hJlli 1,1 II ll J 11. JUL J J ii, il 1, ,i ii ill. Jut Iii ill. In ii ii ,. J 100 200 25 [M-CH 3] + 300 " 400 500 m/z + + 26 27 28 Retention Time (min) 29 600 30 ure 2.22. GC-MS RIC profile for the purified alkaline hydrolysate of fraction II-1 and associated mass spectrum (normalized abundance) of the predominant peak with lanosterol (3 P-hydroxy-8, 24-lanostadiene) structure. luL at 0.50 mg/mL of the hydrolysate in hexane. GC-MS: temperature profile was started at 50°C for 3 minutes followed by at 10°C/min ramp to 340°C which was held for 5 minutes -otherwise as stated in SE/W method section 2.2.9. Note: arrows illustrate fragment(s) of molecule that is/are lost to result in specified m/z value. 66 Figure 2.23. Structures of 4,4' - dimethylsterols: lanosterol (3(3-hydroxy-8, 24-lanostadiene) (A) and butyrospermol (tirucalla-7, 24-dien-3p-ol) [2.29] (B). *H - N M R analysis of the purified hydrolysate of fraction II-1 by Serreqi et al. (1999) was compared with ' H - N M R data published in Dey and Harborne (1991) and confirmed the assignment as butyrospermol (tirucalla-7, 24-dien-3P-ol) [2.29]. Thus, peaks 1, 5 and 9 shown in Figure 2.21 from the steryl ester region of the total extractives profile correspond to butyrospermyl palmitate [2.34], butyrospermyl stearate [2.38], and butyrospermyl eicosanoate [2.42], respectively. Again, the fatty acyl assignments were based on the retention time of each steryl ester as in the assignments made previously. Finally, according to Serreqi (m Utt, 25 M a y 1999) analysis of the third argentation fraction by G C - F I D resulted in 12 poorly resolved peaks. Serreqi et al. (1999) reported that for aspen, argentation fraction III made up 37% of the total steryl esters / waxes fraction based on gravimetric analyses. Fraction III needs further purification before elucidation of the compounds present in this fraction is possible. The other major fatty acid found by Peng et al. (1999) in the hydrolysate of the S E / W fraction was the di-unsaturated linoleic acid [2.15]. The linoleate esters should occur in fraction III according to the principles of AgN03-s i l ica gel chromatographic separation. For compounds of 67 equal polarity separated by argentation-silica gel chromatography, the more unsaturated compounds will be retained longer due to the complexation of double bonds to the silver cation. According to the carbon number index in Figure 2.17 for the GC-MS method developed for SE/W, linoleate esters should elute slightly before the stearate esters and they would most likely appear as partially resolved side bands in the total SE/W profile. Also, it is possible that linoleic steryl esters would not be seen by GC as steryl esters containing di- and tri-unsaturated fatty acyl moieties have been know to degrade during GC analysis (Kuksis et al. 1986), despite attempts to avoid analyte decomposition in this study by the use of a short cGC column. 2.3.3.3 A HPLC pre-GC Fractionation Technique For Steryl Esters In the total extractives profile by GC-MS, it is possible to monitor quantitative differences in nine identified major steryl ester peaks (see Appendix II - [2.34] to [2.42]). However, for some studies, a more detailed compositional analysis is needed. Although the argentation-silica gel chromatography allows the separation and identification of the complex mixture of steryl esters, quantifications performed via flash chromatography or thin-layer chromatography were based on gravimetric analysis. This required relatively large amounts of SE/W material (around 30 mg) in order to obtain quantitative results according to Serreqi (in litt., 25 May 1999). Furthermore, these methods are susceptible to error since fractions may not be completely eluted in flash chromatography, and in TLC entire fractions may not be quantitatively removed from the plate surface resulting in recoveries as low as 65% based on cholesteryl oleate (Evershed et al. 1987). Such limitations contributed to an average error of ±35.5% (based on four replicates starting with approximately 35 mg of SE/W) in the FC fraction weights according to Serreqi (in litt., 25 May 1999). However, since HPLC is a non-destructive analysis method suitable for the separation of some steryl esters, a pre-detector fractionation valve was installed on the HPLC - ELSD system 68 allowing for the collection of fractions for subsequent analysis by G C - M S . Combining the two techniques provided excellent results for detailed compositional analysis of S E / W fraction of aspen. The E L S D was re-optimized under the new pre-detector split mode (using approximately 1:1 volumetric split) using the same procedure as outlined in Appendix IV . The optimum nebulizer gas flow rate and drift tube temperature were 32 mL/min and 130°C, respectively. A typical chromatogram of the S E / W fraction obtained under these conditions with the previously optimized mobile phase composition of 67:33 acetonitrile to tetrahydrofuran is shown in Figure 2.24. <u Vi S3 O Vi tu p-1 Q GO w VI ^ VII VIII IX 10 15 Retention Time (min) 20 Figure 2.24. H P L C - E L S D profile of S E / W fraction using a 1:1 (volumetric) pre-detector split where I-IX bands were separately collected for subsequent G C - M S analysis. 10 uL atlO mg/mL S E / W fraction of CH-F6-02. H P L C - E L S D : volumetric split of 1:1 was used, nebulizer gas flow rate and drift tube temperature were 32 mL/min and 130°C - otherwise as in S E / W H P L C method section 2.2.8. 69 Nine bands were suitably resolved to enable separate collection with minimal cross contamination. Each band was collected for a total of three runs, concentrated down, and re-suspended in a minimum amount of hexane for G C - M S analysis (see Figure 2.25). 10 20 3 0 4 0 5 0 Retention Time (min) Figure 2.25. G C - M S R I C profiles of the S E / W fraction (A) and resultant H P L C fractions I-IX (B-J). (A) was 1 p i at 1.00 mg/mL S E / W fraction of CH-F6-02; (B) to (J) were l u L injection at unknown concentrations of H P L C fractions of CH-F6-02 S E / W dried under nitrogen and resuspended in hexane all run using S E / W method section 2.2.9. #1-12 correspond to steryl esters identified in section 2.3.3.2. * additional components that have not been identified but have been purified in the H P L C - G C - M S procedure making identification and quantification possible. RS = recovery standard (cholesteryl palmitate). 70 The GC-MS profiles for certain HPLC fractions (I, III, and IV) indicated that the solutions were too dilute to obtain sufficiently intense peaks. Consequently, the mass spectra of the resultant peaks were too faint and thus, affected by high noise which would not allow for reliable compound identification. The most important feature to notice in Figure 2.25 is the pure bands, as indicated by symmetrical peak shape approximating Gaussian distributions, obtained after HPLC fractionation. This method allowed for a detailed compositional account of at least 12 identified steryl esters and at least three more unidentified components (indicated by * in Figure 2.25) in the SE/W fraction of aspen acetone extractives. These pure bands obtained in the HPLC - GC-MS approach would allow for precise quantification of individual components in the SE/W fraction. Assuming a universal response factor for steryl esters with the ELSD, the peak area of each fraction in the HPLC chromatogram would be a function of (quadratic equation of calibration obtained with standard steryl or cholesteryl ester such as seen in Figure 2.8 A) of the amount of total steryl esters in that band. Then, the peak area ratios in the GC-MS chromatograms will be equivalent to the ratio of the amounts of individual components in each HPLC fraction. 2.3.3.4 Quantification of Extractive Compounds by GC-MS Once identified or classified, each compound was then quantified without interference of co-eluting components using mass spectral detection. As the ion trap mass spectral detector captured information on the intensity of fragment ions from 50-650 m/z for each eluting compound, one or more particular fragment ions or quantitation ions (QI) can be selected in order to carry out quantitative calculations (see Appendix II for specific QI used for each 71 compound). Quantitation ions were carefully selected such that they occurred only in the compound of interest and not in the co-eluting compounds, and hence, they were used to quantify many compounds free of interference from co-eluting components. An example of how successful this technique was in eliminating interference due to the co-elution of adjacent compounds is shown in Figure 2.26. C • a-Amyrin v / \ / \ A 0 , / \ / \ RIC .2 a-Amyrin^^ / 1 B N OO CN i i i i i i i 27.4 27.6 27.8 28.0 Retention Time (minutes) Figure 2.26. Elimination of interference by co-eluting compounds in R I C chromatogram of a-amyrin [2.31] ( A ) by quantitation based on 218 m/z ion (B). From G C - M S total extractive profile of C O - D 2 - 0 4 . The quantitation ion(s) based peak area (AQI(A)) was then directly proportional to the concentration ( C A ) for that compound in the sample as seen in Equation 6 which results from the 72 substitution of Equation 4 stated by Jennison (in litt. 8 Sept. 1999) into Equation 5 (adapted from Harris 1991). ARICC?) = ( A Q i ( ? ) ) x (QIR(?)) (4) (AR,C (A)) / (AR,C (.S)) = F a x [(CA) / (C, S ) ] (5) Where, F A = the response factor for a particular analyte ( A ) measured on a particular detector relative to the internal standard (IS). This Value is determined empirically by running a known solution of authentic standard with the internal standard; QIR (quantitation ion ratio) = the total ion intensity of the 50-650 m/z mass spectrum divided by the sum of intensities of the QI. Substituting Equation 4 into 5: In order to calculate the amount of a particular compound relative to the amount of wood, or M A (ie mg/kg of freeze-dried wood) Equation 7 was used. C A = [(AQ,(A)) / (AQ, (,S))] x [(QIR(A)) X ( C I S ) ] / [ ( F A ) x ( Q I R ( I S ) ) ] (6) M A = [ C A / 5 mg/mL] x [E / 1 0 0 ] x [1,000,000 mg/kg] (7) Where E is percentage of extractives removed from the wood based on the freeze-dried weight of wood (see section 2.2.3). 73 The QIR of a particular compound was calculated from the mass spectrum for that compound found in the NIST library or from the mass spectrum of the standard compound. However, if the compound was not successfully identified by mass spectrometry, but only classified, an estimate of the QIR was obtained directly from the sample mass spectrum. In every case, AQI was converted to ARIC (see Equation 4) since the RIC area response factors may be very similar for chemically similar groups. This way, determination of F A for every compound identified can be avoided, by assigning a generic response factor for each chemical class. Several studies (Sithole et al. 1992; Orsa and Holmbom 1994; Nugent et al. 1977) reported that assigning a generic response factor to each group of structurally similar compounds was sufficient for the quantification of various wood extractives by GC-FID. In this work, it was found that fatty acids, palmitic [2.12] (C16:0), heptadecanoic (C17:0) and stearic [2.16] (C18:0) acid, produced similar RIC responses with small variations which were linear with carbon number (see Figure 2.27). Similar results were also found with the cholesteryl esters (cholesteryl myristate / C14:0, palmitate / C16:0 and stearate / C18:0), however, the average absolute response was much higher than in the fatty acids (see Figure 2.27). 74 § 6 cu <2 5 5 o = o. a) w o 4 0 c * 8 a> ^  3 1 2 0) (0 ^ * o 1 0 • cholesteryl esters • fatty acids • • • • • • 12 13 14 15 16 17 Fatty Ac id Carbon Number 18 19 Figure 2.27. Relative mass spectral response (RIC) for chemically similar components in the fatty acid (carbon # 16, 17 and 18 refer to palmitic [2.12], heptadecanoic, and stearic acids [2.16], respectively) and steryl ester (carbon #14, 16 and 18 refer to cholesteryl myristate, palmitate, and stearate, respectively) versus fatty acid carbon number. Fatty acids were 1 uL at ~ 0.4 mg/mL in acetone and cholesteryl esters were 1 uL at -0.7 mg/mL in chloroform both set of standards run by total extractives method section 2.2.4. Relative to palmitic acid. This linear trend suggests that one could determine response factors for various compounds within a chemical class by running only a few standard compounds from that class to obtain a linear calibration versus carbon number. However, in this work, other compound classes (ie flavonoids, sterols / triterpenes, etc.) did not allow for carbon number plots and thus, a representative compound was used to establish an average response factor for each chemical class. The standard compounds and corresponding response factor (based on heptadecanoic acid) are listed in Appendix VII. Also, the response factors used in the quantitation of each compound identified or classified in the total extractives profile is listed in Appendix II. 75 The relationship shown in Equation 5 between concentration of an analyte compound (CA) and the respective peak area (ARIC(A)) by G C - M S is only linear for a finite range known as the lower through to the upper linear detection limits. This range is specific for each detector, and Harris (1991) suggests that for mass spectral detection this range may span as many as five orders of magnitude (105). The maximum peak height (detector response or R I C ) for linear quantification was determined to be 6.5 x 106 R I C units by running a series of cholesteryl palmitate solutions (see Figure 2.28). 1.E+07 2000 Concentration (pg / mL) Figure 2.28. Determination of the upper linear response limit of the mass spectrometer with cholesteryl palmitate. 3 uL injection in hexane. GC-MS: 200°C for 1 min, 15°C/min to 300, held for 6 minutes, 10°C/min to 330°C, then 7°C/min to 335°C which was held for 60 min. Injector was ramped from 230°C to 300°C at 200°C/min at time of injection; 690 mBar of He carrier gas; multiplier voltage was 1850 V. 76 All RIC peak heights from the eluting components in the samples (5.000 mg/mL) were less than the maximum RIC peak height of the upper linear response limit of 6.5x106 units. Jennison (in litt., 8 Sept. 1999) stated that certain compounds, especially chlorinated and aromatic compounds, may exhibit a greater space charging effect in ITMS due to intermolecular repulsion when overloaded, lowering the upper detection limit for these compounds. However, compounds that were in danger of occurring in quantities above their upper detection limit, namely steryl esters, fatty acids, sterols, and triglycerides, were all similar in chemical composition to cholesteryl palmitate. Furthermore, these compounds were neither chlorinated nor aromatic, and thus, should exhibit similar charge spacing effects compared to cholesteryl palmitate. The lower response limit (ie limit of detection) or the amount of compound (ie mg) that produces a signal to noise ratio of two (Harris 1991), was controlled by the computer integration parameters. These parameters were set such that quantitation ion peaks measuring approximately less than twice the background noise, were not detected. 2.3.4 Validation of Underivatized Extractives Analysis by GC-MS 2.3.4.1 Reproducibility of GC-MS Method for Underivatized Extractives To assess the reproducibility of the GC-MS method for the total underivatized acetone extractives profiles in aspen wood, a standard solution containing 0.25 mg/mL of heptadecanoic acid (HDA) and 0.25 mg/mL cholesteryl palmitate (CP) was injected several times prior to sample analysis. The results of nine consecutive trials are presented in Table 2.5, showing that the quantitation error was within 10 %. 77 Table 2.5. Peak areas of replicate samples of 0.25 mg/mL heptadecanoic acid (HDA; RRT = 1.00) and cholesteryl palmitate (CP; RRT = 2.06) in acetone by GC-MS using underivatized acetone extractives method (section 2.2.4) HDA CP CP/HDA (Peak area xlO6) (Peak area xlO7) (Ratio) 4.39 1.40 3.18 4.38 1.32 3.00 4.76 1.52 3.21 4.71 1.17 2.50 5.14 1.24 2.41 4.30 1.09 2.53 4.77 1.42 2.98 5.06 1.17 2.31 4.69 1.26 2.68 ±4.84a ±8.36a ±9.55a 'relative error (for a 95% confidence interval) Major sources of error would include integration error as a result of computer calculated peak areas (Papas and Delaney 1987), decomposition of cholesteryl palmitate in acetone while stored on the auto-sampler prior to analysis, injection volume error and split ratio error. However, injection volume error should not be present in the cholesteryl palmitate / heptadecanoic acid value since in a two component mixture, one component acts as an internal standard. To visually illustrate this high reproducibility in terms of the underivatized aspen acetone extractives profile, Figure 2.29 shows triplicate chromatograms obtained by three independent injections of the same sample solution. 78 C c p, 1/3 u 3 10 20 30 40 50 Retention Time (min.) Figure 2.29. Triplicate GC-MS chromatograrns (green, red - offset by +1.5 min, and black -offset by +3 min) for underivatized aspen extractives illustrating the high reproducibility in the developed method. luL at 5.000 mg/mL for CH-F3-02 by total extractive method section 2.2.4. The reproducibility across all compounds is very high, and the ratio of steryl / triterpene esters and triglycerides to the rest of the components remains constant throughout the three chromatograrns. 2.3.4.2 Comparison to Silvlated (TBDMS) Extractives Components that may not be detected via the underivatized extractives method include compounds with a normal boiling point of greater than 320°C, the isothermal temperature of the gas chromatograph injector. Compounds such as sterols, triterpenes and some phenolics, which are of interest in this work, may be bound up as high molecular weight glycosides in aspen and other hardwoods (Jonasson et al. 1997), and hence, may not be seen in the underivatized 79 chromatograrn. Figure 2.30 illustrates the tert. -butyldimethylsilyl derivatized extractives chromatogram compared to the underivatized chromatogram for the same sample (seen in red in Figure 2.30). TBDM-silylation was used since the resultant TBDMS derivatives retain most of the desirable features of trimethylsilyl ethers, yet they are much more stable under hydrolytic conditions (Quilliam et al. 1980). This is important considering that wood extractives are generally hydrophilic, and trace amounts of water are inevitably present during the silylation. In the silylated extractives profile, enhanced sensitivity and improved peak shape is seen for certain components such as benzoic acid (#1), and the fatty acids (#7, #8, #9, #11, #12, and #13). This is likely a result of reduced interference caused by the partial disappearance of non-silylated components (due to a lack of available active sites in these compounds) in the mixture, together with the high yield and enhanced volatility of the resultant TBDMS esters of these mono-carboxylated compounds (Blau and Halket 1993). Furthermore, producing a TBDMS -derivatized extractives profile by GC-MS will give additional mass spectral data with which to unequivocally assign structures to components in an unknown mixture. This is illustrated in Appendix VIII where the interpretation of trace amounts of docosanoic acid in the underivatized spectrum was affected with high background noise, but together with the respective TBDMS derivative spectrum, the identification was unequivocal. The abundant ion fragments at [M-15]+ and [M-57]+ typically produced by TBDMS esters of fatty acids (Blau and Halket 1993), provides high sensitivity, as well as, characteristic information from which assignments can be easily made. 80 9 10 20 30 40 50 60 Retention Time (min.) Figure 2.30. Typical GC-MS profile for TBDMS - derivatized extractives compared with the GC-MS profile for underivatized extractives (shown in red; taken from Figure 2.4). 2 uL at -0.83 mg/mL (depends on derivatization yield) of CO-D4-02 using derivatized total extractives method section 2.2.4. *silylation contaminants (from blank); 1 = TBDMS - benzoic acid (authentic standard); 2 = monosaccharide (MS classification); 3 = unknown; 4 = TBDMS - 4-hydroxybenzoic acid (NIST '98 - fit = 850); 5 = monosaccharide (MS classification); 6 = TBDMS - nonanedionic acid (NIST '98 - fit = 869); 7 = TBDMS - palmitic acid (authentic standard); 8 = internal standard (TBDMS - heptadecanoic acid); 9 = TBDMS - linoleic acid (NIST '98 - fit = 941); 10 = unknown; 11 = TBDMS - eicosanoic acid (authentic standard); 12 = TBDMS - docosanoic acid (authentic standard); 13 = TBDMS -C23:0 - C26:0 fatty acids (NIST '98 - fits >800); 14 = various underivatized and some TBDMS - derivatized sterols / triterpene alcohols (MS classification); 15 = TBDMS - naringenin (authentic standard); 16 = TBDMS - sakuranetin (MS fragmentography - see Appendix III); 17 = underivatized palmitate steryl esters (Appendix II); 18 = underivatized stearate steryl esters (Appendix II). However, in the TBDMS - derivatives chromatogram, significantly less information is seen than in the underivatized spectrum. A comprehensive evaluation of all the detected components in 81 the derivatized chromatogram revealed that there are no newly identified components as a result of the silylation aside from the monosaccharides (#2 and #5 in Figure 2.30), which are of lesser interest to this study. Thus, the underivatized extractives profile contained more information than a silylated extractives profiles and it is concluded that derivatization is an unnecessary extra step in the preparation of samples for the analysis of extractives in aspen. The underivatized spectrum still did not represent all of the extractives that were removed from the aspen cores, and it was calculated that only 56% of the extractives removed by acetone were represented in quantifiable amounts in the total extractives chromatogram. Components present in the acetone extractives that are not analyzable by conventional GC include large macromolecules found in wood such as low volatility / large MW polyaromatic compounds, as well as, polypeptides and polysaccharides. Also, mineral and ash extractives (ie Ca, K, Mg, etc) which may account for 0.1-1.0 % of the dry wood in hardwoods of temperate zones (Fengel 1983) are not GC analyzable. However, these inorganic compounds, found predominantly in the sapwood, are only slightly soluble'in acetone (Sjostrom and Alen 1999) and hence should not be present in significant amounts in the acetone extractives of the aspen cores. Also, certain compounds present in the acetone extractives may decompose before they volatilize, and thus, were not seen by the GC-MS method in this work. 2.3.4.3 Comparison to Flame Ionization Detection Flame ionization detectors are commonly used in extractives analysis by GC. Furthermore, compounds which do not produce molecular fragments <650 m/z by EI ionization at 70eV will not be detected by the ITMS detection system used in this study. Thus, a comparison of the response between ITMS and an FID for the compounds analyzed in this work is appropriate. 82 Figure 2.31 shows a typical underivatized extractives profile for the same aspen sample run by the GC-MS method and by an analogous GC-FID method. Retention times may not correspond exactly since the GC-FID system carrier gas flow rate increased from 1.2 to 1.9 mL/min during the run whereas in the GC-MS system constant flow programming allowed a constant flow rate of 1.6 mL/min. Vi fl O ft Vi o o w tu Q Fatty Acids& Monoglycerides Low MW Phenolics Sterols Waxes& Diglycerides 1/ Triglycerides 10 20 30 40 Retention Time (min.) 50 60 Figure 2.31. Typical GC-FID profile (A) compared with the GC-MS (B) profile for underivatized extractives. (A) was 2uL at 5.000 mg/mL of CO-K7-02 by total extractives method for GC-FID section 2.2.5. (B) was luL at 5.000 mg/mL of CO-K7-02 by total extractives method section 2.2.4. 1 - heptadecanoic acid (internal standard); 2 = butyrospermol palmitate [2.34]. No new chromatographic features were seen in the FID chromatogram, however, the relative responses for various compounds differed between the two detection methods. Also, the signal / noise (S/N) ratio for heptadecanoic acid internal standard (peak #1; RRT = 1.00) was 130 for the 83 FID, and 288 for the M S based on 270, 227, 185 and 129 m/z quantitation ions. The S/N ratio for butyrospermol palmitate [2.34] (peak #2; R R T = 2.25) was 78 for the FID, and 225 for the M S based on the 394 m/z quantitation ion. To obtain an acceptable chromatogram with the FID, the injection volume was doubled (2 uL instead of 1 uL used with the M S ) and the split ratio reduced by one third (1:20 vs. 1:60). Both of these changes w i l l more than double the level of analyte reaching the detector in the FID run. Taking these differences into account, the sensitivity for heptadecanoic acid and butyrospermol [2.29] was much greater by mass spectrometry than by flame ionization detection. Furthermore, the FID does not provide any qualitative information on the eluting analyte other than retention time (using Kovats index), therefore, the M S is far superior for the quantification and characterization of many unknown compounds found in aspen wood extractives. 2.4 S U M M A R Y 2.4.1 Development of Analytical Methods A rapid and accurate method for assessing the extractives content in trees from increment cores was developed. The method involves freeze drying the wood samples for 24 hours, grinding the core through a 2 mm mesh screen, extracting with acetone for 8 hours at 4-5 cycles/h, and finally concentrating the extract solution by rotary evaporation, removing the residual solvent with nitrogen and freeze drying the residue to obtain the dry weight of extractives. The method requires small amounts of material (<5 g dry wood) allowing one to examine the extractives profile of a standing tree and assess the levels of various extractives that may potentially impact on the utilization of aspen. The reproducibility of the method was determined to be 4 ± 1 % based on the total extractives weights from replicate cores taken from sample trees. 84 The wide range of compound classes in aspen does not necessarily lend itself well to single whole extractives analysis. However, a GC-MS method that allowed for the accurate quantification and characterization of up to 70 compounds in aspen wood acetone extracts with minimal sample preparation was developed. Forty-four compounds were identified, and 26 additional compounds were classified into fatty acid, flavonoid, sterol / triterpene, diglyceride, and triglyceride classes. The total extractives profile by GC-MS accounted for on the average 56% of the weight of total acetone extractives in aspen and all the components that were targeted in this study were quantifiable in a single chromatogram aside from salicylic acid [1.14]. Specific enzymatic hydrolysis of the salicylic acid glucoside was attempted using P-glucosidase, however, no additional salicylic acid [1.14] was found in the hydrolysate. A suitable method for the analysis of bound salicylic acid was developed involving TFA hydrolysis of the ethyl acetate soluble fraction of the total extractives and subsequent GC-MS analysis of the hydrolysate for free salicylic acid [1.14]. However, the origin of salicylic acid [1.14] in aspen is still unclear. It may be present as a glycoside, or it may be an artifact derived from acid hydrolysis of salicortin [ 1.16] and tremulacin [ 1.17]. 2.4.2 Identification and Quantification of Compounds The AC-GC-MS approach to elucidate the structures of individual intact steryl esters in the total SE/W fraction was successful in unequivocally identifying twelve steryl esters. Once identified, nine of these steryl esters were quantifiable in the total extractives profile without interference from co-eluting components using quantitation ion based peak areas. Although linoleate steryl esters have been reported to occur in aspen wood, these esters were not found in this work. 85 These compounds may have decomposed throughout the GC-MS analysis and thus, alternate methods must be explored in order to quantify any thermally unstable esters. The HPLC-ELSD method yielded good results for the separation and quantification cholesteryl esters. However, for the detailed analysis of more complex samples such as aspen SE/W fraction, it was determined that the HPLC-ELSD method may serve as a pre-GC fractionation procedure. This combined method (HPLC - GC-MS) method would allow the characterization and quantification of 12 to 15 compounds found in SE/W fraction of the aspen acetone extractives. 2.4.3 Validation of the GC-MS Method for Underivatized Extractives The reproducibility of the developed GC-MS method for total extractives was found to be excellent at less than 10% error in replicate peak heights. The biggest source of error was likely integration error as a result of inaccuracies in the computer-determined peak areas. The underivatized extractives method was validated by comparing the results to those of TBDMS - derivatized extractives. It was found that the derivatization did not result in the detection of any important extractives not already present in the underivatized chromatogram. Thus, derivatization was considered an unnecessary time consumption and added expense in the analysis of wood extractives by GC-MS for this study. Comparisons between GC-MS and GC-FID illustrated the robustness of the developed method as it was transferable to less sophisticated chromatographic systems. In addition, the MS was much more sensitive than the FID for the compounds of interest in this study. The added cost of the 86 MS was justified by improved sensitivity, and more importantly, the MS provided detailed qualitative information which was necessary to make the 70 compound assignments in the total extractives profile despite the inevitable co-elution of some compounds. 3 CLONAL VARIATION OF EXTRACTIVES 3.1 INTRODUCTION There are at least seven publications that deal with the clonal variation and heritability of wood properties such as wood density and fibre length in quaking aspen (reviewed in Yanchuk et al. 1984). However, a comprehensive search of the available literature has revealed that little has been published on the natural variation of wood extractives in aspen. Only a few studies exist that address the natural clonal variation of the total weight of wood extractives in fast-growing hardwoods such as aspen and eucalyptus (Yanchuk et al. 1988; Sandstrom et al. 1996). However, there are almost no studies on the natural clonal variability of the composition and quantity of different chemical classes of aspen wood extractives. A study by Ayer (1995) reported on the clonal differences in the levels of bound salicylic acid in aspen wood, however, only 3 clones (3 to 5 ramets / clone) were examined providing very limited conclusions. This chapter provides a survey of the interclonal variability in various wood extractives found in aspen that affect the utilization of this species and that may potentially be used as clonal selection criteria for future plantations. 87 3.2 MATERIALS AND METHODS 3.2.1 Clone Selection and Sampling Ten and five millimeter increment cores were taken at breast height (BH) from 118 ramets (trees) over 21 natural aspen clones by Paprican in May of 1996. As the clones had not yet flushed (leafing) that year, clonal distinctions were based on morphological differences such as bark colour and pattern, and even branch angle when needed. The clones were all located in North Eastern British Columbia in the Boreal White and Black Spruce biogeoclimatic zone as shown in Figure 3.1. Five clones were located within an 8.5 km radius in the Del Rio site [121°W and 56°20'N] in the Dawson Creek TSA (see Figure 3.1). Eight clones were located within a 5 km radius at the Farrell [122°W and 56°40'N] site and eight within a 2.5 km radius at the Kobes Creek site [122°W and 56°40'N], both of which occurred within the Fort Saint John TSA (see Figure 3.1). For these 21 clones, estimates of clonal average age, DBH, fibre length and wood density were determined previously by Hunt et al. (1999) using 10 mm and 5 mm cores extracted at BH. For this study, nine of these clones were selected to include clones of maximum, moderate, and minimum average fibre lengths and wood densities (see Table 3.1). In June of 1998, 2-13 ramets per clone (see Table 3.1) were non-destructively sampled by the removal of a 10 mm increment core at breast height (BH; ~1.3 m) in a west to east orientation across the entire diameter of each stem. The ramets within each clone were selected based on absence of excessive damage by grazing or previous coring as indicated by presence of exuded pitch. Also, duplicate samples were taken from a few ramets from most clones to assess the overall reproducibility of the extractives determination. 88 Figure 3.1. Biogeoclimatic zone map of BC with a Prince George Forest Region overlay showing aspen sampling sites Del Rio (A) in the Dawson Creek Forest District, and Farrell / Kobes (B) in the Fort St. John Forest District (adapted from British Columbia Ministry of Forests 1994). 89 Table 3.1. Number of Ramets sampled, age, D B H , and wood properties from nine natural aspen clones in north eastern B C Clone ID / Site Number of Average Average Wood properties3 ramets ramet age3 ramet - density (kg/m3), sampled (years) D B H a fibre length (mm) (#duplicates) (cm) D l / Del Rio 13(2) 76 28.4 402, 1.05 D2 / Del Rio 10(2) 73 28.8 382, 0.99 D4 / Del Rio 6(2) 67 24.9 338, 0.87 F3 / Farrel 6(3) 81 28.0 374, 0.92 F5 / Farrel 3(0) 106 29.1 358, 0.84 F7 / Farrel 2(1) 77 29.9 409, 0.89 K I / Kobes Creek 5(0) 100 31.9 395, 0.96 K 2 / Kobes Creek 3(1) 85 40.6 318, 0.95 K 7 / Kobes Creek 2(2) 119 31.6 345, 1.04 determined previously by Hunt et al. (1999) The identification of ramets was based on intraclonal similarities in gross morphological features. These included bark color, pattern, texture and surface scars, as well as, leaf size, and stem form, all known to be good distinguishing features for aspen clones (Peterson and Peterson 1995). Interclonal differences in several of these morphological features is apparent in the photographs taken of each clone seen in Figure 3.2. The 10 mm cores were placed into separate zip-lock™ plastic bags with minimal handling, and kept in a cooler packed with ice until they arrived at the laboratory. A black and white electronic scan was obtained for each core using an Alphalmager 2000 (Alpha Innotech, Leandro, C A , U S A ) after which the cores were frozen (-10°C) until extraction. 90 Figure 3.2. Photographs (100 ISO, Kodachrom™ Slides) of each aspen clone sampled highlighting morphological differences between clones. Note: tree with red "X" superimposed in F5 is from adjacent clone. 91 3.2.2 Site Assessments To account for environmental variability between clones, site assessments were performed by a certified consultant for each of the nine clones sampled in this experiment. The assessments were done in accordance to the Biogeoclimatic Ecosystem Classification (BEC) system for site identification and interpretation. Soil moisture and nutrients regimes (edatopes) were determined by assessing the vegetation, drainage / topography, and soil properties present at each site. Site index calculations were based on the heights of several trees in each site. 3.2.3 Assessment of Decay Microbial analysis of the 10 mm core material for common decay fungi was attempted, however, no decay fungi where observed. It is thought that the freezing and thawing of the cores, may have destroyed any viable fungi present in the material. Thus, the extent to which the material in each 10 mm core was decayed was assessed in the laboratory by macroscopic features in the wood as described by Hiratsuka et al. (1995). Characteristics that were indicative of the most common and important cause of aspen trunk rot in Alberta, type Ai(too soft for pulping) and / or A2(firm enough for pulping) due to Phellinus tremulae fungus (Hiratsuka et al. 1990), were seen in most decayed samples (Figure 3.3). The presence of the second most common cause of decay in aspen, Peniophora polygonia, in the 10 mm decadent material is very likely, but could not be confirmed in this study, due to the lack of distinct decay characteristics for this fungus. P. polygonia does not cause large columns of advanced decay as does P. tremulae, however, it is found more often in decayed and discolored wood (Hiratsuka et al. 1990). The amount of decay in each core was assessed visually by the extent to which characteristics of "Type A " decay and general discoloration were present. A 92 "decay index" number was then assigned to each core based on this visual inspection (see Table 3.2). Figure 3.3. Images of decay caused by Phellinus tremulae: A = typical decay seen in 10 mm cores; B = typical sound wood seen in 10 mm cores (for reference); C = decay cause by Phellinus tremulae as seen in aspen stem cross-section (taken from Hiratsuka et al. 1995). Note: prominent black lines around and within decayed areas are diagnostic for rot caused by this fungus. 93 Table 3.2. Decay index developed to grade each core on extent of decay present Decay Index Number Description Example 1 Sound wood 2 Minimal discoloration / wood intact 3 Moderate discoloration / wood intact 4 Moderate discoloration / wood frail 5 Severe discoloration and wood frail This visual decay index method is by no means a replacement of a comprehensive decay evaluation including microbial analysis and identification and quantification of major decay organisms. However, the criteria used to visually evaluate each core are known to correlate well with the amount of decay and stain organisms present in a sample. 9 4 3.2.4 Sample Extraction and Total Extractives Analysis Samples were prepared and extracted as outlined in section 2.2.2. and 2.2.3 for 10 mm increment cores. GC-MS analysis was performed on underivatized extractives and carried out exactly as outlined in section 2.2.4. for underivatized extractives. Analysis of resultant chromatograrns was completed as per section 2.3.3.1 and 2.3.3.4 for characterization and quantification of various components (see Appendix II) in the GC-MS profile. Saponifiables / Unsaponifiables was calculated based on information in Allen (1988) and Hillis (1962) using Equation 9. Sap. / Unsap. = [FA + DG + TG] / [SE + W + S/T] (9) Where FA = fatty acids; DG = diglycerides; TG = triglycerides; SE = steryl esters; W = waxes; S/T = sterols / triterpenes (all in mg/kg of freeze-dried wood). 3.2.5 Analysis of Salicylic Acid Analysis of bound salicylic acid [1.14] was performed via the TFA hydrolysis method for three samples (see section 2.2.12 part 2.2.12.2), two of moderate (CO-D1-03) and mild (CO-D1-20) decay from clone Dl and one of severe decay (CO-D2-02) from clone D2. 3.2.6 Statistical Calculations and Analyses All error ranges in numerical data equate to the 95% confidence intervals (unless otherwise stated) calculated using Microsoft Excel™ 97. All a errors used in F-tests or multiple comparison test are 0.05. Analysis of Variance was calculated using the ANOVA Model 95 Estimation procedure in SYSTAT™ 7.0 (SPSS, Chicago, IL, U S A ) assuming a completely randomized design model shown in Equation (10). 7, =M + TJ+ei/ (10) Where Yy is the ith observation (ramet) on t h e t r e a t m e n t (clone); p. is the overall true mean; ty is the effect of the jth treatment; and Sy random error present in the ith observation on the jth treatment. The multiple range tests were performed with SYSTAT® 7.0 using the Bonferroni test. A l l correlations were performed using the SYSTAT™ 7.0 correlation procedure. See Appendix I for relevant statistical formulae. 3.3 R E S U L T S A N D D I S C U S S I O N 3.3.1 Assumption of Equal Variance and Clonal Variation of Total Acetone Extractives Due to the limit o f increment cores that may be sampled, processed and the resultant data analyzed within a reasonable budget and time frame, a large number of ramets were sampled only from a select few clones, and only a few ramets were sampled from the rest of the clones. Thus, in the A N O V A process, intraclonal variation, largely weighted on these well sampled clones, is effectively extrapolated via the mean squared error ( M S E ) to other clones from which few samples were. However, this procedure is only valid i f the variance within each treatment (ie clone) is known to be similar (Hicks 1993). To illustrate that this requirement of uniform variance was met for the samples in this experiment, the distribution curve for the total weight of extractives (one of several random variables measured in this study) is shown in Figure 3.4 for one clone from each of the three sites (Del Rio , Kobes and Farrel). Clones D l , K I and F3 are shown since these are the clones from which the most ramets were taken from each area. 96 1.00 2.00 3.00 4.00 1.00 2.00 3.00 4.00 % Extractives % Extractives 1.00 2.00 3.00 4.00 % Extractives Figure 3.4. The distribution curves for the total weight of extractives for clones D l (A), KI (B), and F3 (C). In each case, the width of the distribution curve serving as a measure of the relative intraclonal variance, is remarkably similar. It is known that the non-experimental error component of intraclonal variance is due to environmental factors in clonal wood species (Zsuffa 1975). Intraclonal patterns of wood property variation should remain relatively constant within a species for a given geographical location (reviewed in Yanchuk et al. 1984). Furthermore, variance due to experimental error will contribute to the intraclonal variation of each clone equally, since this error is random and the samples were all treated identically. Thus, it is not surprising that D l , KI and F3 show 97 equivalent variances, and the assumption of equal variances holds true for the clones sampled in this experiment. The one-way ANOVA results for total weight of extractives for all nine aspen clones sampled is presented in Figure 3.5. 0.00 I 1 1 1 1 1 1 1 1 1— d1 d2 64 f3 f5 f7 k1 k2 k7 Clone Figure 3.5. One-way ANOVA results for total acetone extractives (% fd wood). dl*f3, k l ; d2*f5; d4*f3,f5; f3*f5,k2;f5*f7,kl,k7; kl*k2. F-Ratio = 8.40 (p=0.000); maximum / minimum = 1.75. Note: error bars represent the standard error of the mean for each clone. A large interclonal variability is seen in the average weight of extractives for each clone from as high as 4% to just over 2% (fd wood). This is also indicated by a high F-ratio of 8.4 (Ftabuiated = 2.18) indicating that there is a significant treatment effect. Furthermore, the Bonferroni multiple 98 comparison tests indicated various significant interclonal differences in total extractives levels. Yanchuk et al. (1988) also found significant interclonal differences in the total acetone extractives in aspen. Unfortunately, the total amount of extractives is not a very meaningful parameter to be used in the selection of superior clones for pulp and paper production. However, trees or clones with higher extractives content would be avoided if possible by mill personnel due to the fear that higher extractive levels will lead to higher chemical (alkali) consumption in kraft pulping, and aggravation of pitch problems in mechanical pulping, and a reduced pulp yield in both cases. 3.3.2 Composition of the Acetone Extractives from Aspen 10 mm Cores The amounts of 44 identified compounds and 26 additional classified components extracted with acetone were monitored in the nine clones to provide information on the interclonal variation in extractives quality (see Appendix IX for amounts of each compound found in each ramet). The composition of the 10 mm core wood extractives as determined by GC-MS is illustrated in Figure 3.6. The most abundant material extracted from the cores were the steryl esters and waxes at an average of over 5 g per kg of fd core material. Also high in abundance were the sterols / triterpenes, glycerides, and fatty acids. 99 8.00 T3 g 6 .00-Figure 3.6. The average amounts of several chemical classes found in the acetone extracts of aspen wood core (10mm) samples based on GC-MS quantitation, (note: error bars represent standard deviation) However, these amounts serve as only relative measures for these components between trees sampled at BH with 10 mm cores and of similar DBH (see Table 3.1). The increment core sampling method has a unique sampling bias and hence, the results can only be compared with those obtained using the same sampling procedure. As a result, the amounts seen in Figure 3.6 may differ from those reported by studies that utilize wood sampled via methods other than increment coring at BH. Some unique compounds, not classified into any of the major groups shown in Figure 3.6 were found in this work. Nine-oxononanoic acid [2.5], not previously reported in the literature to occur in aspen, was found in significant amounts (189 mg/kg fd wood) in this study. Nugent et al. (1977) reported that oxidized fatty acid material may be present in an unknown fraction 100 (-0.2% of dry wood) of aspen wood acetone extractives. Thus, 9-oxononanoic acid [2.5] may be the oxidized derivative of linoleic acid, which was found in large amounts in this study and has a reactive double bond on the 9-carbon. Also , vitamin E or a-tocopherol [2.25], previously reported to occur in aspen bark (Nekrasova 1982), was found in relatively minor amounts (37 mg/kg) in every sample. Vitamin E is thought to prevent the autooxidation of unsaturated lipids (ie linoleic acid - C l8 :2 ) in plants (Miller 1973a). 3.3.3 Interclonal Variation in Components Relating to Pitch Issues , A s discussed previously, steryl esters, sterols / triterpenes, triglycerides and fatty alcohols, in particular, are known to participate in pitch formation during the pulp and paper processing of aspen. The interclonal variation in total sterols / triterpenes, triglycerides and steryl esters / waxes is shown in figures 3.7, 3.8 and 3.9 respectively, as these were the pitch culprits found in the largest amounts in aspen wood. Results in Figures 3.7, 3.8 and 3.9 show that there are significant differences between the clones for components that are important in pitch propensity prediction. In the steryl esters / waxes case, this difference was almost two fold between the minimum and maximum amount for the nine clones sampled in this project. High levels of steryl esters have been reported in aspen pitch deposits (Allen et al. 1991). Steryl esters are known to be non-polar, hydrophobic, and non-structural components of wood, together with the fact that they make up a large fraction of the unsaponifiables in aspen (Hil l is 1962) makes them a good candidate for predicting pitch propensity in aspen for both kraft and mechanical pulping methods. 101 09 Cm © o % * H * d1 d2 d4 f3 f5 f7 k1 k2 k7 Clone Figure 3.7. One-way ANOVA results for total sterols / triterpenes (g/kg fd wood). dl*d2; d2rf5. F-ratio = 3.42 (p = 0.004); maximum / minimum = 1.63. Note: error bars represent the standard error of the mean for each clone. 102 d1 d2 d4 f3 f5 f7 k1 k2 k7 Clone Figure 3.8. One-way ANOVA results for total triglycerides (g/kg fd wood), d l^k l . F-ratio 3.99 (p = 0.001); maximum / minimum = 7.8. Note: error bars represent the standard error of the mean for each clone. 103 09 Si S3 a > S3 ^ « s i u 1/3 9.24 7.67 6.10 4.53 2.96 i — i — i — i — i — r j i L _i i i L d1 d2 d4 f3 f5 f7 k1 k2 k7 Clone Figure 3.9. One-way ANOVA results for total steryl esters / waxes (g/kg fd wood). d l*k l ; d2*kl. F-ratio = 2.82 (p = 0.013); maximum / minimum = 1.90. Note: error bars represent the standard error of the mean for each clone. In this study, four main sterols were found linked to palmitic [2.12], stearic [2.16] and eicosanoic [2.17] acids in aspen. Longer, higher molecular weight steryl esters are known to have a higher melting point (Ginsburg et al. 1983) and thus, could potentially cause greater pitch problems by coagulating at higher temperatures. Furthermore, the presence of double bonds in sterols / triterpene alcohols and fatty acids or their esters, may exert a pro-oxidative action on these components (Peng et al. 1999; Kochhar 1983), thereby making them more prone to modification and decomposition during bleaching stages in pulp processing. Hence, the quality of this fraction may also have bearing on the degree of pitch forming potential of a specific clone. 104 Figure 3.10 shows the interclonal variation of large molecular weight steryl esters with saturated fatty acyl moieties (C18:0 and C20:0) that may have a higher tendency to coagulate and lead to pitch problems. Figure 3.10. One-way ANOVA results for high molecular weight steryl esters / waxes (g/kg fd wood). dl*kl ,d2*kl. F-ratio = 3.46 (p = 0.004); maximum / minimum = 2.65. Note: error bars represent the standard error of the mean for each clone. Figure 3.10 shows that the interclonal variation pattern for the large molecular weight steryl esters is very similar to that of the total SE/W fraction (Figure 3.9) despite the fact that they make up less than 23% of this fraction. This supports the notion that the composition of this fraction does not vary significantly between clones, and there is a close biosynthetic relationship between the steryl ester compounds. 2.38 0.46 d1 d2 d4 f3 f5 f7 k1 k2 k7 Clone 105 In kraft pulping, mi l l personnel use the ratio of saponifiables to unsaponifiable resin to predict the pitch propensity that a particular pulpwood species may have. This parameter has little significance in terms of the biosynthetic origin of compounds, however, it has an industrial importance. The average value for this ratio in the aspen cores was found to be 0.65+0.03, which was significantly lower than that published by Al l en (1988) of 1.2 for fresh aspen wood (estimated using 2 cm thick disks taken from base, 2m from base and top o f stem). This discrepancy is not surprising considering the increment core method provides a linear average of the extractives across the stem at B H , giving this method a larger heartwood bias compared with sampling round wood sections (ie disks). Furthermore, aspen heartwood has been reported to contain higher amounts o f steryl esters (the major unsaponifiables found in aspen) and lower amounts of triglycerides (a major saponifiable found in aspen) than sapwood which would result in a lower saponifiables / unsaponifiables value (Nugent et al. 1977). Nevertheless, in this work, comparisons were made between samples which were taken in an identical fashion (described in section 3.2.1). The interclonal variation in saponifiables / unsaponifiables is seen in Figure 3.11. Figure 3.11 shows that a significant interclonal variation was present in saponifiables / unsaponifiables, and clone K 2 would be selected over clone D l due to the higher ratio in the former. Pulpwood with a higher saponifiables / unsaponifiables ratio would provide better deresination during kraft pulping, and thus lead to less pitch problems in the mi l l . 106 1.00 0.00 I 1 1 1 1 1 1 1 1 !— d1 62 64 f3 f5 f7 k1 k2 k7 Clone Figure 3.11. One-way ANOVA results for saponifiables / unsaponifiables. dl*k2. F-ratio = 3.59 (p = 0.003); maximum / minimum = 1.34. Note: error bars represent the standard error of the mean for each clone. 3.3.4 Interclonal Variation in Components relating to Toxicity Issues Extractive compounds that are thought to contribute to the acute and / or chronic toxicity of aspen extractives include low molecular weight phenolics such as benzoic acid [1.11], phenol [1.12], and flavonoids [1.8], as well as, non-phenolics such as nonanoic acid [1.13] and phytosterols, especially P-sitosterol [2.27]. The interclonal variation of total phytosterols (including triterpenes) is shown in Figure 3.7. The results showed that significant interclonal differences were apparent, and thus, selection of lower sterol containing clones should result in the production of lower sterol containing propagules. However, no significant interclonal variability occurred in P-sitosterol [2.27] (p = 0.70; see Appendix X for ANOVA) as a single 107 compound, most likely as a result of the higher uncertainty or error associated with measuring a single compound. Although several monoaryl phenolics were detected in this study, no significant interclonal variation (p = 0.053; see Appendix X for ANOVA) was found for the total fraction. Furthermore, as would be expected with these compounds, the level of some monoaryl phenolics and all detectable flavonoids increased with core decay as seen in Figures 3.12 and 3.13. The monoaryl phenolics that showed no increase with core decay included the lignin precursors coniferyl alcohol [2.8] and aldehyde [2.9], as well as, sinapyl alcohol [2.13] and aldehyde [2.14]. 0.40 2 3 4 Decay Index Figure 3.12. Scatter plot of decay index with monoaryl phenolics content (g/kg fd wood). Note: the selected monoaryls include o-coumaric acid [2.2], salicyl alcohol [2.3], 4-hydroxy-2-methylacetophenone [2.4],^-hydroxy-hydrocinnamic acid [2.7] andp-hydroxycinnamic acid [2.10]. 108 4.00 3.00 h 8 I to WD © WD 2.00 h 1.00 r-0.00 0 1 2 3 4 Decay Index Figure 3.13. Scatter plot of decay index with total flavonoid content (g/kg fd wood). Particularly in older forests, the environmental factor of disease, such as the presence of heart rots or root rots, will affect wood properties (Zobel and van Buijtenen 1989). In this work, the presence of decay added to the intraclonal variability of flavonoids and certain monoaryl phenolics and therefore, reduced the detectable interclonal differences in these components. Ultimately, heavily decayed aspen wood, will produce more toxic mill effluent when pulped than sound wood due to elevated levels of phenolics including flavonoids (up to 6-8 times more flavonoids in extremely decayed cores than in sound wood) associated with decayed wood. 3.3.5 Interclonal Variation in Compounds Related to Decay Resistance As previously mentioned and shown in Figures 3.12 and 3.13, the level of the flavonoids and some monoaryl phenolics groups increased with core decay implying that these compounds are produced by the tree as part of its induced resistance to the pathogen or they are metabolites 109 produced by the invading organisms. Salicylic acid [1.14], is known to induce systemic acquired resistance in many plants, and thus, play a vital role in the plant's decay resistance. Synthesis of salicylic acid [1.14] may occur via two different pathways, as has been demonstrated in young tomato seedlings. In non-infected plants, salicylic acid [1.14] is formed from ^-cinnamic acid [1.15] via benzoic acid [1.11], however, upon infection, biosynthesis from ^-cinnamic acid [1.15] via o-coumaric acid [2.2] is favored (reviewed in Yalpani et al. 1993). Conversely, in the tobacco plant, salicylic acid [1.14] is formed via benzoic acid [1.11] in infected and non-infected plants (Yalpani et al. 1993). Both benzoic [1.11] and o-coumaric [2.2] acids occurred in free form in the acetone extractives of the core wood (see Appendix II). However, salicylic acid [1.14], which is thought to be present as a glucoside in aspen (Hubbes 1969; Pausler et al. 1995), was not detected in free form in this work. Levels of bound salicylic acid were assessed in three samples by acid hydrolysis, and the results are shown in Figure 3.14 together with the relative levels of potential salicylic acid [1.14] precursors benzoic [1.11] and o-coumaric [2.2] acids. Figure 3.14 shows a general increase in these three monoaryl phenolics with decay. However, the benzoic acid [1.11] content in these three samples did not follow the same positive correlation (ie CO-D1-20) with decay as did salicylic acid [1.14] and o-coumaric acid [2.2] and in fact, the decay correlation coefficient (R) for benzoic acid [1.11] considering all 50 samples was only 0.37, but 0.47 for o-coumaric acid [2.2] (see Appendix X I for correlation plots). These results are consistent with the findings in tomato seedlings that in infected plants, salicylic acid [1.14] is synthesized from the o-coumaric acid [2.2] precursor and not benzoic acid [1.11]. However, these results do not take into account any benzoic acid [1.11] or o-coumaric acid [2.2] that may be present as a glycoside and thus, may be misleading. 110 saliclyic acid benzoic acid o-coumaric acid (bound) Figure 3.14. Relative amounts of bound salicylic acid [1.14] and free benzoic [1.11] and o-coumaric acids [2.2] in CO-D1-03 (decay = 1.5), CO-D1-20 (decay = 2.5) and CO-D2-02 (decay = 5). Nevertheless, the levels of free benzoic [1.11] and o-coumaric [2.2] acid are still appreciably affected by decay and it is known that aspen trees infected with P. tremulae have a higher benzoic acid [1.11] content than those not infected (Pausler et al. 1995). It seems that due to the strong relationship with decay, levels of bound salicylic acid and precursors benzoic [1.11] and / or o-coumaric [2.2] acid (in free form), would not allow the assessment of decay resistance in the mature aspen clones sampled. 3.3.6 Correlation of Compounds with Related Biosynthetic Pathways Various extractive compounds originate from a related biosynthetic path and thus, should show a relationship in their amount found in wood. Butyrospermol palmitate [2.34] was the predominant steryl ester detected in aspen acetone extractives, making up 35% of the total steryl esters detected. Figure 3.15 illustrates the excellent correlation (R > 0.83 and Rcriticai = 0.279) 111 between levels of butyrospermol palmitate [2.34] and those of P-amyrin [2.35] and a-amyrin [2.37] palmitates along with levels of butyrospermol stearate [2.38] and eicosanoate [2.42]. 800 700 600 1 cs Cu 500 a 400 S < < 300 Qi — 200 1 1 1 u o o 0 0 o 0 0 % o O 0 0 0D% 8 ° o OO o — ^ os>o e o ° ° o 0 -0 " o R 1 1 = 0.84 " 1 2000 1000 2000 3000 4000 Butyrospermol Palmitate ti OS | 1500 ss c S I 1000|-< cs { 5001 o o ° n n 0 9 8 _ L o o R= 0.85 1000 2000 3000 4000 Butyrospermol Palmitate 2000 99 -ti 1500h u a a on O -1000h ea 500 h 0 1000 2000 3000 4000 Butyrospermol Palmitate 400 cs e | 300 e u 5 "3 Z a. 200 o u I* 100 CQ <*p 8 o 0 a > ° 0 R = 0.91 i 1000 2000 3000 4000 Butyrospermol Palmitate Figure 3.15. Correlation of the amount (mg/kg fd wood) of butyrospermol palmitate [2.34] with that of other steryl esters. R = correlation coefficient. This positive correlation was also seen with other biosynthetically related compounds such as the fatty acids, flavonoids, and free sterols / triterpenes to some extent (see Appendix X I I ) . This 112 suggests that one may simply monitor the levels of one prevalent component in a chemical class to obtain an estimate of the relative levels for that chemical class between samples. However, to improve the correlation between the level of the measured component and the level of the entire class, and to minimize the error associated with measuring the levels of single compounds, it is recommended that more than one compound from each class be monitored. Nevertheless, this correlation procedure may speed up the analysis time by GC allowing for more clones to be assessed, ultimately, providing more confident results. 3.3.7 Non-genetic Contribution to Phenotypic Variation Environmental factors which may affect wood properties such as extractives, include climate, soil, and overall site, as well as the influence of trees on one another within forest stands (Zobel and van Buijtenen 1989). Site assessments may be done qualitatively by examining site soil and vegetation characteristics or quantitatively by determining the potential for wood production from standing trees in the sites (ie site index) (Watts 1983). Quantitative and qualitative site assessment data is provided in Table 3.3 for each of the nine clones examined in this study. Table 3.3. Site characteristics for selected aspen clones Clone Climate Moisture Nutrient Regime Site Index Dl Moist / warm 4 C/D 20.0 D2 Moist / warm 4 B/C 22.8 D4 Moist / warm 4 C 21.2 F3 Moist / warm 4/5 C/D 18.8 F5 Moist / warm 3/4 B 17.3 F7 Moist / warm 5 D 21.0 KI Moist / warm 3/4 B 17.5 K2 Moist / warm 4 D 22.2 K7 Moist / warm 3/4 B 20.0 Note: all parameters listed are comparable within the BWBS biogeoclimatic zone only. Table 3.3 shows that climate was consistent throughout all clones and thus, should not contribute to environmental variability between clones. Qualitative data, listed under "Moisture" and 113 "Nutrient Regime" in Table 3.3, should provide the most useful information on the effects of site on wood extractives. Site index, although listed in Table 3.3, will not be considered since growth parameters such as tree height, which was used to calculate these values, may be partially influenced by genetics. Figure 3.16 shows that there were no apparent trends between site quality factors and the measured variable (total extractives). Figure 3.16. Site factors (moisture class and nutrient regime class) related to the total acetone extractives for aspen clones. *Nutrient regime class cross reference: 2 = B, 2.5 = B/C, 3 = C, 3.5 = C/D, and 4 = D. Note: error bars represent the standard error of the mean for each site rating. Also, Zobel and van Buijtenen (1989) stated that site has little effect on other wood characteristics including density in diffuse-porous hardwoods such as aspen. In this study, site was not a significant source of variation, and the large genetic variability in aspen is a far more logical explanation for the interclonal differences observed. 114 Disease, and in particular, heart rot caused by P. tremulae and other decay organisms, contributed to the intraclonal variability in the quantity of various extractives in this work. This effect has been illustrated in the extreme case for flavonoids and certain monoaryl phenolics. Furthermore, invading microorganisms in aspen are known to produce hydrolytic enzymes, which break down neutral resins such as glycerides as well as steryl esters to some degree (Dunlop-Jones et al. 1991). Hence, decay most likely contributed to the intraclonal variability seen in the quantity of lipophilic compounds, but to a lesser extent than occurred in the phenolics. Finally, as seen in Table 3.1, there was some variability in clonal age, which may impact on the levels of the various extractives monitored. To show that extractives did not depend on age in the mature aspen clones used in this work, age versus total weight of extractives is plotted in Figure 3.17 for each of the 50 ramets sampled in this work. It is clear from Figure 3.17 that no correlation exists between age and total extractives for the ramets sampled in this work. Hence, the interclonal variability observed would have been caused by factors other than clone age. 115 w c c • a e H T 3 o o 5.00 4.00 r-3.00 h 2.00 1.00 50 100 Ramet Age (years) 150 Figure 3.17. Age verses total weight of acetone extractives for aspen clones. R = 0.07 < Rcriticai = 0.273. 3.4 SUMMARY Interclonal variability was apparent not only in the total weight of extractives, but also in different chemical classes such as sterols / triterpenes, triglycerides, and especially steryl esters / waxes which may affect the pitch propensity of particular clones. Thus, not surprisingly, there was a significant interclonal effect in saponifiables / unsaponifiables, which may allow the selection of clones with a higher more favorable ratio for kraft pulping. Also, a positive correlation between the levels of certain biosynthetically related compounds was observed, especially within the steryl esters, fatty acids, flavonoids and sterol / triterpene classes. The presence of significant interclonal variability in the sterol / triterpene compounds implies that the level of these compounds may be used as criteria for the selection and propagation of 116 lower sterol / triterpene pulpwood which may reduce the detrimental estrogenic effects of mi l l effluents on aquatic organisms. Other compounds known to be toxic to aquatic life including monoaryl phenolics and flavonoids, were found to be extremely dependent on the environmental variable of decay and thus, interclonal variation in these components were not detectable. Also , the assessment of clonal decay resistance by levels of bound salicylic acid was not possible due to the prevalence of decay in the ramets as indicated by decay observed in the 10 mm cores. Finally, it was determined that ramet age, and site factors did not significantly impact on the total extractives level and thus, the observed interclonal variability in extractives may be attributed primarily to genetic differences between the clones. 4 C O N C L U S I O N S A N D R E C O M M E N D A T I O N S 4.1 C O N C L U S I O N S 4.1.1 Analytical methods A rapid and accurate method has been developed that enables the evaluation of various extractives in standing aspen trees. The G C - M S method, which does not rely on prior sample derivatization, provides information on 70 semi-volatile compounds found in aspen acetone extractives, including nine steryl esters which have been identified. This procedure would be useful for the monitoring of pitch culprits and environmentally problematic components to improve process efficiency, product quality and environmental monitoring and protection in pulp mills. Additionally, a H P L C - E L S D fractionation method combined with G C - M S was developed and allowed for the separation and quantification of 15 steryl esters in aspen, 12 of which have been identified. This method would be well suited to laboratory monitoring of deresination treatments including biotreatment with various fungi and seasoning experiments. Finally, a 117 method for the analysis of salicylic acid [1.14] in aspen was developed. However, the origin of salicylic acid [1.14] could not be determined and its involvement in systemic acquired resistance in aspen still remains unconfirmed. The rapid and comprehensive G C - M S extractives method developed, may be used in the selection of superior natural clones for tree improvement programs. M u c h attention in tree improvement efforts has been diverted to hybridization with a view to produce hybrids that have benefits from both parents. However, hybridization of tree species for the enhancement of one or a few commercially important traits may lead to losses in other important traits (ie pest resistance, etc) in the resultant progeny. The advantage of propagating natural genotypes is that natural selection has already created superior individuals, one must simply identify the clones that feature desirable traits and propagate the ramets. 4.1.2 Clonal Variation of Wood Extractives Natural genetic diversity in the levels of various extractives has been observed in the natural clones of aspen sampled in this work. These variations as measured by the F-ratio (interclonal / intraclonal variance), and the maximum interclonal range, was highest for triglycerides, and moderate for sterols / triterpenes and steryl esters / waxes. These results suggest that it would be possibly to identify and select aspen clones that produce lower amounts of these compounds. This may lead to the breeding of pulpwood trees that give rise to less pitch problems during pulping. Also , the cultivating of propagules from lower sterols / triterpenes containing clones may result in less environmental problems associated with effluent discharge. Bound salicylic acid as determined by T F A hydrolysis, and its potential precursors, benzoic acid and o-coumaric acid, were found to be affected by the presence of decay. A s a result, no 118 significant interclonal differences were seen in the salicylic acid [1.14] precursors. Ayer (1995) reported that there was no apparent clonal difference in levels in bound salicylic acid between 3 clones sampled at 3 to 5 ramets each. Furthermore, levels of phenolic glycosides are known to depend on plant gender, and since quaking aspen is a dioecious species, individual clones may be either male or female (Lindroth and Hwang 1996). Ultimately, it seems that there are to many uncertainties with salicylic acid [1.14] in aspen, and as a result, it may not be a good predictor of inherent decay resistance in natural aspen clones. The levels of certain monoaryl phenolics and all flavonoids were found to be affected by the presence of decay in the nine mature aspen clones. It was determined that phenylpropanoids [1.7] that appeared more closely related to lignin biosynthesis did not increase with decay. Those that did increase with decay included benzoic acid [1.11], o-coumaric acid [2.2], salicyl alcohol [2.3], 4-hydroxy-2-methylacetophenone [2.4],p-hydroxy-hydrocinnamic acid [2.7], p-hydroxycinnamic [2.10] and flavonoids [1.8]. These compounds are most likely associated with the defense mechanism in aspen. Furthermore, many of these compounds have been previously isolated from black galls in aspen, a feature that is though to play a role in pathogen resistance in this species (Pausler et al. 1995). 4.2 R E C O M M E N D A T I O N S 4.2.1 Analytical Methods Despite the high reproducibility obtained with the G C - M S method, the reproducibility could be enhanced with the use of more than one internal standard to reduce the variability caused by random errors in the splitting of different chemical classes during the split injection. Another 119 possibility would be to attempt direct-on column injection to eliminate split ratio variability in the results and possibly minimize thermal decomposition of thermally labile analyte. The H P L C - E L S D pre-GC fractionation must be performed on more samples, and the resultant unknown steryl esters should be identified by the mass spectrum and retention time analysis. For a successful G C - M S analysis of these fractions, it is recommended that more than three replicate fractions be taken for each H P L C band. The EI (70eV) mass spectrum of a steryl ester is characteristic of the sterol / triterpene moiety. The fatty acid moiety can then be confirmed by retention time comparisons to synthetic standards to confirm the identity of the intact steryl ester. Alternately, i f a quadrupole mass spectrometer capable of analyzing high mass ions (>650 m/z) is available, chemical ionization (ie NH3) with a mass spectral scan of up to 850 m/z may allow identification of the intact steryl ester without the need to synthesized standards. This identification would be based on the mass-to-charge value of the molecular ion ([M] + ) provided that the sterol / triterpene moiety is known. 4.2.2 Clonal Variation of Wood Extractives The study should be expanded to include many more clones, and a larger number of ramets per clone so that a more comprehensive evaluation of the natural variation in aspen wood extractives may be obtained. Also , clones from semi-arid sites unlike the sites in this experiment are know to yield greater interclonal variability due to the selection of heterozygous individuals in these sites. Therefore, the extractives variability between extremely different environments may be compared to determine i f similar results are obtained. Also , the sampling of younger (<60 years old) stands may provide decay free material from which interclonal variation in phenolics and in particular, salicylic acid [1.14] may be assessed. However, it is recommended that a parameter other than salicylic acid [1.14] be used for the evaluation of inherent decay resistance, due to the 120 complexity in accurately measuring the genetic predisposition of a single compound, especially one as difficult to analyze and determine the origin of as salicylic acid [1.14]. Finally, separation of heartwood and sapwood in each core may account for some intraclonal variability (seen as unexplained error in this experiment) and improve the sensitivity of the interclonal tests. However, due to the limited amount of material provided by 10 mm increment cores, a second core should be taken from each tree and combined with the first to reduce the large errors associated with weighting minute amounts of extractives. 121 L I T E R A T U R E C I T E D Adams, R.P. 1989. Identification of essential oils by ion trap mass spectroscopy. Academic Press, San Diego, California, pp. 1-13. Al len , L . H . , B . B . Sithole, J . M . MacLeod, C L . Lapointe, and F.J. McPhee. 1991. The importance of seasoning and barking in the kraft pulping of aspen. J. Pulp Paper Sci . 17(3): 85-91. Al len , L . H . 1988. Pitch control during the production of aspen kraft pulp. Pulp Paper Can. 89(10): 87-91. Ayer, W . A . 1995. Benzoic acid, salicylic acid and black galls on aspen. In Proceedings of the Aspen Stain and Decay Workshop, 21 Mar. , Edmonton, Al ta . pp. 27-31. British Columbia Ministry o f Forests. 1994. Forest, Range and Recreation Resource Analysis, 1994. Victoria, B . C . p. 28, 39. Blakeney, A . B . , P.J. Harris, R .J . Henry, and B . A . Stone. 1983. A simple and rapid preparation of alditol acetates for monosaccharide analysis. Carbohydr. Res. 113:291-299. Blanchette, R . A . , R . L . Farrell, Y . Hader, J .E. Merritt, R . A . Snyder and P . A . Wendler. 1991. European Patent 0387, 187, A Z . ^ Blau, K . , and J . M . Halket. 1993. Handbook of Derivatives for Chromatography. 2 n d ed. John Wiley & Sons, Chichester, Great Britain, pp. 52, 75-85. Charlet, P., G . Lenone, B . Joseleau, and P. Chareyre. 1997. Analyses of extractives from different wood species. In Proceedings of the 9 t h International Symposium on Wood and Pulp Chemistry, June 9-12, Montreal, Q C . pp. 15-1 - 15-4. Chen, T., C. Breuil , S. Carriere, and J .V . Hatton. 1994a. Solid-phase extraction can rapidly separate l ipid classes from acetone extracts of wood and pulp. Tappi J. 77(3): 235-240. Chen, T., Z . Wang, Y . Gao, C. Breuil , and J .V . Hatton. 19946. Wood extractives and pitch problems: analysis and partial removal by biological treatment. Appita J. 47(6): 463-466. Chen, T., Z . Wang, Y . Zhou, C. Breuil , O . K . Aschim, E . Yee, and L . Nadeau. 1995. Using solid-phase extraction to assess why aspen causes more pitch problems than softwoods in kraft pulping. Tappi J. 78(10): 143-149. Claassen, F .W. , C. van de Haar and T A . van Beek. 1998. H P L C analysis of low molecular weight constituents in wood. In Proceedings of the 7 t h International Conference on Biotechnology in the Pulp and Paper Industry. June 16-19, Vancouver, B C . pp. B45-B47. Cook, D . L . , L . LaFleur, A . Parrish, J. Jones and D . Hoy. 1997. Characterization of plant sterols from 22 U S pulp and paper mills. Water Sci . Technol. 35(2/3): 297-303. 122 DeBell, J.D., J.J. Morrell, and B.L. Gartner. 1997. Tropolone content of increment cores as an indicator of decay resistance in western redcedar. Wood Fibre Sci. 29(4): 364-369. Del Rio, J., A. Gutierrez, F.J. Gonzalez-Vila, F. Martin, and J. Romero. 1998. Characterization of organic deposits produced in the kraft pulping of Eucalyptus globulus wood. J. Chromatogr. A, 823: 457-465. Denton, T.E., W.M. Howell, J.J. Allison, J. McCollum, and B. Marks. 1985. Masculinization of female mosquitofish by exposure to plant sterols and Mycobacterium smegmatis. Bull. Environ. Contam. Toxicol. 35: 627-632. Devore, J.L. 1995. Probability and Statistics for Engineering and the Sciences. 4th ed. Wadsworth, Inc., Belmont, CA. pp. 29, 293, 399, 411, 541, 510, 509. Dey, P.M., and J.B. Harborne (Series editors). 1991. Methods in Plant Biochemistry. Vol.7. Academic Press, San Diego, pp. 1-43, 331-359, 369-471. Dunlop-Jones, N., H. Jialing, and L.H. Allen. 1991. An analysis of the acetone extractives of the wood and bark from fresh trembling aspen: implications for deresination and pitch control. J. Pulp Paper Can. 17(2): J60-J66. Evershed, R.P., V.L. Male, and L.J. Goad. 1987. Strategy for the analysis of steryl esters from plant and animal tissues. J. Chromatogr. 400:187-205. Farrar, J.L. 1997. Trees in Canada. 3rd ed. Fitxhenry & Whiteside Ltd., Ottawa., Ont. p. 346. Fengel, D., and G. Wegener. 1983. Wood Chemistry, Ultrastructure, Reactions. Walter de Gruyter & Co., Berlin, pp. 132-226. Ferrari, R.Ap., W. Esteves, K.D. Mukherjee, and E. Schulte. 1997. Alteration of sterols and steryl esters in vegetable oils during industrial refining. J. Agric. Food Chem. 45: 4753-4757. Ginsburg, G.S., D. Atkinson, and D.M. Small. 1983. Physical properties of cholesteryl esters. Progress in Lipid Res. 24:135-167. Gutierrez, A., J.C. del Rio, F.J. Gonzalez-Vila, F. Martin. 1998. Analysis of lipophilic extractives from wood and pitch deposits by solid-phase extraction and gas chromatography. J. Chromatogr. A, 823: 449-455. Harris, D.C. 1991. Quantitative Chemical Analysis. 3rd ed. W.H. Freeman and Company, New York. pp. 601,610-710. Hicks, R. H. 1993. Fundamental Concepts in the Design of Experiments. 4th ed. Oxford University Press, New York. pp. 49-120. Hillis, W.E. 1962. Wood Extractives and their Significance to the Pulp and Paper Industries. Academic Press, New York. pp. 331-363, 423. 123 Hiratsuka, Y., D.A. Gibbard, O. Bakowsky, and G.B. Maier. 1990. Classification and measurement of aspen decay and stain in Alberta. Forestry Canada Information Report NOR-X-314. Hiratsuka, Y., T. Stokes, P. Chakravarty, and D.J. Morgan. 1995. A field guide to classify and measure aspen decay and stain. UBC Press, Vancouver, BC. pp. 1-7. Horvath, D.M. and N. Chua. 1994. The role of salicylic acid in systemic acquired resistance. Curr. Opin. Biotechnol. 5:131-136. Hubbes, M. 1969. Benzoic and salicylic acids isolated from a glycoside of aspen bark and their effect on Hypoxylon pruinatum. Can. J. Bot. 47:1295-1301. Hughes, D. and K. Thurbide. 1998. Rapid determination of pulp extractives using pressurized fluids. Pulp and Paper Research Institute of Canada report, PGRLR# 714. Hunt, K. 1995. Impact of decay and stain on mechanical pulping. In Proceedings of the Aspen Stain and Decay Workshop, 21 Mar., Edmonton, Alta. pp. 7-17. Hunt, K., W.Y. Gee, A. Hussein, S.M. Reath, and PA. Watson. 1999. Kraft pulping opportunities from Canadian aspen clones. Pulp and Paper Research Institute of Canada report. In press. Hurlbert, S.H. 1984. Pseudoreplication and the design of ecological field experiments. Ecol. Monogr. 54(2): 187-211. J&W Scientific. 1998. J&W 1998 Catalog, Technical Reference and Cookbook. J&W Scientific Inc., Folsom, CA. pp. 59, 331-380. Jonasson, R.G., B. Fuhr, B.B. Sithole, and L.H. Allen. 1997. Glucosides in aspen wood kraft pulping. In Proceedings of the 9th International Symposium on Wood and Pulp Chemistry, June 9-12, Montreal, QC. pp. 43-1 - 43-5. Karl, W. 1988. Aspen: quickly becoming today's preferred wood species for pulp. Pulp Paper J. 41(7): 23. Knoll, C.S., B.M. Wong, and D.N. Roy. 1993. The chemistry of decayed aspen wood and perspectives on its utilization. Wood Sci. Technol. 27(6): 439-448. Kochhar, S.P. 1983. Influence of processing on sterols of edible vegetable oils. Prog. Lipid Res. 22(3): 161-188. Kuksis, A., J.J. Myher, L. Marai, J.A. Little, R.G. McArthur, and D.A.K. Roncari. 1986. Fatty acid composition of individual plasma steryl esters in phytosterolemia and xanthomatosis. Lipids 21(6): 371-377. Lindroth, R.L., and S.-Y. Hwang. 1996. Clonal variation in foliar chemistry of quaking aspen {Populus tremuloides Michx.). Biochem. Syst. Ecol. 24(5): 357-364. 124 Lindsay, S. 1989. High Performance Liquid Chromatography. John Wiley & Sons, Chichester, Great Britain, pp. 1-8. MacLeod, J.M. 1987. Chemical pulping of aspen: possibilities and realities. Pulp and Paper Research Institute of Canada report, MR# 115. MacLatchy, D.L. and G. van der Kraak. 1994. The plant sterol (3-sitosterol decreases reproductive fitness in goldfish. In poster abstracts of 2nd International Conference on Environmental Fate and Effects of Bleached Pulp Mill Effluents, 6-10 Nov., Vancouver, BC. p. 79. Miller, L.P. (Editor). 1973a. Phytochemistry (Vol. Ill): inorganic elements and special groups of chemicals. Van Nostrand Reinhold Company, New York. pp. 195-220, 327-350. Miller, L.P. (Editor). 19736. Phytochemistry (Vol. II): organic metabolites. Van Nostrand Reinhold Company, New York. pp. 171-309, 344-399. Mitton, J.B., M.C. Grant. 1996. Genetic variation and the natural history of quaking aspen. Bioscience, 46(1): 25-31. Nekrasova, V.B. 1982. Vitamin concentrate from aspen bark - a bioactive feed additive. Lesokhim. Podsochka, 6(10). (abstract only - translated to English) Nordback, J., E. Lundberg, and W.W. Christie. 1998. Separation of lipid classes from marine particulate material by HPLC on a polyvinyl alcohol-bonded stationary phase using dual-channel evaporative light-scattering detection. Mar. Chem. 60:165-175. Nugent, H.M., L.H. Allen and H.I. Bolker. 1977. Effect of seasoning on the acetone extractives composition of the wood from black spruce, jack pine and trembling aspen. Trans. Tech. Sect. CPPA, 3(4): 103-109. Ona, T., T. Sonoda, K. Ito, M. Shibata, Y. Tamai, and Y. Kojimo. 1996. Use of the radially divided increment core method to assess pulpwood quality for eucalypt breeding in E. camaldulensis and E. globulus. AppitaJ. 49(5): 325-331. Onken, J., and R.G. Berger. 1998. Evaporative light scattering detector (ELSD) for analysis in food chemistry. Deutsche Lebensmittel-Rundschau, 94(9): 287-292. Oppenheimer, L.E. and T.H Mourey. 1985. Examination of the concentration response of evaporative light-scattering mass detectors. J. Chromatogr. 323: 297-304. Orsa, F., and B. Holmbom. 1994. A convenient method for the determination of wood extractives in papermaking process waters and effluents. J. Pulp Paper Sci. 20(12): 361-366. Papas, A. N., and M.F. Delaney. 1987. Evaluation of chromatographic integrators and data systems. Anal. Chem. 59: 54A-59A. 125 Pausler, M . , W . A . Ayer, and Y . Hiratsuka. 1995. Benzoic acid, salicylic acid, and the role of black galls on aspen in protection against decay. Can. J. For. Res. 25:1479-1483. Peng, J., R. Leone, A . N . Serreqi, and C. Breuil . 1999. Are aspen sterols and steryl esters changed structurally by kraft pulping and bleaching?. Tappi J. 82(1): 204-211. Pearl, I .A., and S.F. Darling. 1971a. Hot water phenolic extractives of the bark and leaves of diploid Populus tremuloides. Phytochem. 10:483-484. Pearl, I .A., and S.F. Darling. 19716. The structures o f salicortin and tremulacin. Phytochem. 10:3161-3166. Peterson, E . B . , and N . M . Peterson. 1995. Aspen Managers' Handbook for British Columbia. Canadian Forest Service and the British Columbia Ministry o f Forests report, F R D A# 230. Polymer Laboratories. 1999. Webpage: http://www.polymerlabs.com/elsschem.html. Polymer Laboratories Inc., Amherst, M A , U S A . Quill iam, M . A . , K . K . Ogilvie, K . L . Sadana and J .B. Westmore. 1980. Study of rearrangement reactions occurring during gas chromatography of tert. -butyldimethylsilyl ether derivatives of uridine. J. Chromatogr. 194:379-386. Rezanka, T. 1992. Analysis of sterol esters from alga and yeast by high-performance liquid chromatography and capillary gas chromatography-mass spectrometry with chemical ionization. J. Chromatogr. 598: 219-226. Rodden, G . (Editor). 1999. Pulp and Paper Canada Annual Directory. Pulp and Paper Canada, Pointe-Claire, Q C . Rowe, J.W. (Editor). 1989a. Natural Products of Woody Plants I: Chemicals Extraneous to the Lignocellulosic Cel l Wal l . Springer-Verlag Berl in Heidelberg, Germany, pp. 274-393, 533-569. Rowe, J .W. (Editor). 19896: Natural Products of Woody Plants II: Chemicals Extraneous to the Lignocellulosic Cel l Wal l . Springer-Verlag Berl in Heidelberg, Germany, pp. 691-839. Sandstrom, M . , R. Badenlid and J. Sjostrom. 1996. Variation in wood extracts of Eucalyptus globulus clones. In Proceedings of the 4 t h European Workshop on Lignocellulosics and Pulp, 8-11 Sept., Stresa, Italy, pp. 362-365. Serreqi, A . N . , R. Leone, and C. Breuil . 1998. Identification of steryl esters in aspen wood by capillary G C . In Proceedings of the 7 t h International Conference on Biotechnology in the Pulp and Paper Industry. June 16-19, Vancouver, B C . pp. B7-B10. 126 Serreqi, A . N . , R. Leone, L . F . Del Rio , S. M e i , M . Fernandez, and C. Breuil . 1999. Identification and quantification of important steryl esters in aspen wood. J. A m . O i l Chem. Soc. Submitted for Publication. Servos, M . R . , K . R . Munkittrick, J .H. Carey, and G . L . van der Kraak. (Editors). 1996. Environmental Fate and Effects of Pulp and Paper M i l l Effluents. St. Lucie Press, Delray Beach, Florida, pp. 21-31, 159-166, 297-314. Sithole, B . B . , J .L. Sullivan, and L . H . Al len . 1992. Identification and quantification of acetone extractives of wood and bark by ion exchange and capillary G C with a spreadsheet program. Holzforschung, 46(5): 409-416. Sithole, B . B . 1992. Modern methods for the analysis of extractives from wood and pulp: a review. Appita J. 45(4): 260-264. Sjostrom, E . , and R. Alen . 1999. Analytical Methods in Wood Chemistry, Pulping, and Papermaking. Springer-Verlag Berl in Heidelberg, Germany, pp. 48-50, 125-148. Taylor, B.R. , J.S. Goudey, and N . B . Carmichael. 1996. Toxicity o f aspen wood leachates to aquatic life: laboratory studies. Environ. Toxicol . Chem. 15(2): 150-159. Varian Canada. 1998. Techniques of G C - M S Workshop Manual. Varian Canada Inc., Mississauga, O N , Canada. Vilegas, W. , M . Sanommiya, L . Rastrelli, and C. Pizza. 1999. Isolation and structure elucidation of two new flavonoid glycosides from the infusion of Maytenus aquifolium leaves. Evaluation of the antiulcer activity of the infusion. J. Agric . Food Chem. 47: 403-406. Wallis , A . F . A . , and R . H . Wearne. 1999. Analysis of resin in eucalypt woods and pulps. Appita J. 52(4): 295-299. Watts, S.B. 1983. Forestry Handbook for British Columbia. 4 t h ed. D . W . Friesen & Sons Ltd. , Cloverdale, B C . p. 424. Yalpani, N . , J. Leon, M . A . Lawton, and I. Raskin. 1993. Pathway of salicylic acid biosynthesis in healthy and virus-inoculated tobacco. Plant Physiol. 103: 315-321. Yanchuk, A . D . , B .P . Dancik and M . M . Micko . 1984. Variation and heritability of wood density and fibre length of trembling aspen in Alberta, Canada. Silvae Genetica, 33(1): 11-16. Yanchuk, A . D . , I. Spilda, and M . M . M i c k o . 1988. Genetic variation of extractives in the wood of trembling aspen. Wood Sci . Technol. 22:67-71. Yong-Bi , F., G . Peter, Y . Clarke, G . Namkoong, and A . D . Yanchuk. 1998. Incomplete block designs for genetic testing: statistical efficiencies of estimating family means. Can. J. For. Res. 28: 977-986. 127 Zacharewski, T.R., K. Berhane, B.E. Gillesby and B.K. Burnison. 1995. Detection of estrogen-and dioxin-like activity in pulp and per mill black liquor and effluent using in vitro recombinant receptor / reporter gene assays. Environ. Sci. Technol. 29:2140-2146. Zobel, B.J., and J.P. van Buijtenen. 1989. Wood Variation Its Causes and Control. Springer-Verlag Berlin Heidelberg, Germany, pp. 189-217. Zobel, B., and J. Talbert. 1984. Applied Forest Tree Improvement. John Wiley & Sons, Inc., New York. pp. 50-51,118-141. Zsuffa, L., 1975. Broad sense heritability values and possible genetic gains in clonal selection of Pinus grifflthii McClelland x P. strobus L. Silvae Genetica, 24(4): 85-88. 128 APPENDICES 129 APPENDIX I STATISTICAL FORMULAE (TAKEN FROM DEVORE 1995) The standard error of the mean or Smean is then: s = — mean I 130 APPENDIX I (CONT'D) A*h-degree polynomial regression formulae: Regression model: Pi =b0+bixi+b2x? +... + bkxtk Where X, is the independent variable, and the coefficients (ie b^) are solved by: (insert 13.10 from p. 541 in Devore) The coefficient of multiple determination (R ) is: S S E= l Z ( y > - ^ ) 2 SST Where is the actual value of the dependent variable. (insert correlation formulae here!) 131 APPENDIX I (CONT'D) Bonferroni's Test formulae: CD = t a •. ,—1<VMSE) MSE* [JP+JJ Were CD is the critical difference for one comparison; Jp and Jq denote the total number of observations in each treatment being compared; m is the number of comparisons made; and df^SE is the degrees of freedom of the mean squared error = I(J-1). if CD< X P -X difference is significant if CD> X.-Xc difference is not significant 132 APPENDIX. II LIST OF COMPOUNDS FOUND IN ASPEN ACETONE EXTRACTIVES AND QUANTIFI-ABLE IN A SINGLE CHROMATOGRAM BY GC-MS UNDERIVATIZED METHOD compound ID (common name or IUPAC relative ID method/fit quantitation ion(s) / occurrence in: name) / reference number (Appendix XIII) retention time* QIR / response factor P. tremuloides*, woody plants0 benzoic acid /1.11 0.3915 authentic standard 105/4.80/2.25 4" p-ethylphenol / 2.1 0.3933 NIST / 909 107/2.66/2.25 9C o-coumaric acid / 2.2 0.4087 NIST / 926 120,91,65/ 1.65/2.25 9C salicyl alcohol / 2.3 0.4852 NIST / 885 124,106,78/ 1.83/2.25 9d 4-hydroxy-2-methylacetophenone / 2.4 0.5270 NIST / 879 150,135,107/ 1.71 /2.25 -9-oxononanoic acid / 2.5 0.6654 NIST / 796 83/9.18/1.00 -p-hydroxybenzoic acid / 2.6 0.6758 NIST / 978 121,138/ 1.69/2.25 4d p-hydroxy-hydrocinnamic acid / 2.7 0.7583 NIST/931 107,166/2.26/2.25 -coniferyl alcohol / 2.8 0.8312 NIST / 862 180,137/3.98/2.25 1E coniferyl aldehyde / 2.9 0.8388 MS fragmentography 178, 135/6.17/2.25 9= p-hydroxycinnamic acid / 2.10 0.8578 NIST / 944 164,147/3.42/2.25 4" ferulic acid / 2.11 0.9071 NIST/ 901 194,133/3.08/2.25 9e palmitic acid / 2.12 0.9441 authentic standard 256,129/ 17.3/ 1.00 3d sinapyl alcohol / 2.13 0.9811 MS fragmentography 210,167/4.56/2.25 l e sinapyl aldehyde / 2.14 0.9848 NIST / 942 208,165/5.60/2.25 l e linoleic acid / 2.15 1.0379 authentic standard 280,67/8.75/ 1.00 3d stearic acid / 2.16 1.0502 authentic standard 284,129/17.1 / 1.00 3d fatty acid 1 1.0852 MS classification 67,150/ 10.2/ 1.00 -eicosanoic acid / 2.17 1.1488 authentic standard 312,269,129/ 15.3/ 1.00 3d docosanoic acid / 2.18 1,2406 authentic standard 340,297,87/ 14.0/ 1.00 3d fatty acid 2 1.2739 MS classification 262,81 / 12.8/ 1.00 -unknown 1.2967 N/A 272/4.95/N/A -flavonoid 1 1.3137 MS classification 107,195/3.94/ 1.65 -liqnoceric acid / 2.19 1.3270 MS fragmentography 368,185,129/ 18.0/ 1.00 6d sakuranetin /2.20 1.3554 NIST / 929 285,167/4.44/1.65 9e 1-docosanol/2.21 1.3781 NIST / 840 97,83/5.80/ 1.00 3d flavonoid 2 1.3867 MS classification 167/7.80/ 1.65 -naringenin / 2.22 1.3923 authentic standard 271,153/5.19/ 1.65 4d sterol / triterpene 1 1.4179 MS classification 157/35.1 /0.98 -sterol / triterpene 2 1.4482 MS classification 394/34.7/0.98 -3-sitosterol acetate / 2.23 1.4559 NIST / 935 382/39.1 /0.98 -kaempferide / 2.24 1.4608 NIST / 846 300/6.15/ 1.65 -vitamin E / 2.25 1.4682 NIST / 804 431 / 4.72/0.98 8d (bark) kaempferol / 2.26 1.5025 NIST / 968 286/6.21 /1.65 l e 3-sitosterol / 2.27 1.5412 authentic standard 213,397/39.1 /0.98 3d sterol / triterpene 3 1.5449 MS classification 425 / 26.9 / 0.98 -3-amyrin / 2.28 1.5535 NIST / 740 203,218 / 5.86 / 0.98 9e butyrospermol / 2.29 1.5554 other3 394/21.5/0.98 9= cycloartenol /2.30 1.5687 NIST / 824 394 / 62.2 / 0.98 3d a-amyrin / 2.31 1.5725 authentic standard 218/19.1/0.98 sterol / triterpene 4 1.5762 MS classification 174,411 /4.98/0.98 -24-methylenecycloartanol / 2.32 1.5867 NIST / 889 408,407/35.5/0.98 3d citrostandienol /2.33 1.5904 NIST / 737 285/8.32/0.98 6d sterol / triterpene 5 1.6046 MS classification 218,411/18.1/0.98 -sterol / triterpene 6 1.6482 MS classification 152,429/9.05/0.98 -sterol / triterpene 7 1.6558 MS classification 396/38.1 /0.98 -diqlyceride 1 1.6663 MS classification 530,81 /20.0/0.57 -sterol / triterpene 8 1.6786 MS classification 181 /15.0/0.98 -133 APPENDIX II (CONT'D) compound ID (common name or IUPAC name) / reference number (Appendix XIII) relative retention time* ID method/fit quantitation ion(s) / QIR / response factor occurrence in: P. tremuloides", woody plants6 diglyceride 2 1.8255 MS classification1" 531,408/25.0/0.57 -waxes 1.8512 MS classification RIC/1/3.62 1" diglyceride 3 1.8957 MS classification11 559/7.73/0.57 -diglyceride 4 1.9440 MS classification1" 579,394/ 13.7/0.57 -steryl / triterpene ester 1 1.9970 MS classification 408/41.0/4.02 -steryl / triterpene ester 2 2.1307 MS classification" 392/6.55/4.02 -diglyceride 5 2.1545 MS classification1" 394/ 10.3/0.57 -triglyceride 1 2.1886 MS classification 256/ 16.6/4.22 -butyrospermol palmitate / 2.34 2.2493 section 2.3.3.2 (#1) . 394/9.07/4.02 -3-amyrin palmitate / 2.35 2.2795 section 2.3.3.4 (#2) 203 / 7.34 / 4.02 -triglyceride 2 2.2966 MS classification 394/25.3/4.22 -lupeol palmitate / 2.36 2.3212 section 2.3.3.2 (#3) 203 / 36.1/4.02 7d a-amyrin palmitate / 2.37 2.3544 section 2.3.3.2 (#4) 218/14.3/4.02 triglyceride 3 2.4302 MS classification 410/19.4/4.22 -butyrospermol stearate / 2.38 2.4852 section 2.3.3.2 (#5) 109,394/9.60/4.02 -3-amyrin stearate / 2.39 2.5278 section 2.3.3.2 (#6) 203 / 8.87 / 4.02 -lupeol stearate / 2.40 2.5942 section 2.3.3.2 (#7) 392/25.3/4.02 7d a-amyrin stearate / 2.41 2.6273 section 2.3.3.2 (#8) 203,218/ 10.1/4.02 7" steryl / triterpene ester 3 2.6454 MS classification 201,408/18.2/4.02 -triglyceride 4 2.6862 MS classification 548/81.4/4.22 -steryl / triterpene ester 4 2.6985 MS classification 424/61.6/4.02 -butyrospermol eicosanoate / 2.42 2.8151 section 2.3.3.2 (#9) 394/9.77/4.02 -trilinolein / 2.43 3.7004 authentic standard 600/31.3/4.22 -Relative to heptadecanoic acid at 17.584 minutes. amass spectral matched with sterol derived from the base hydrolysate of butyrospermol esters. bsupporting evidence given by disappearance of peaks after alkaline hydrolysis (Appendix V) Supporting evidence given by persistence of peaks through alkaline hydrolysis (Appendix V) References: 1 = Fengel (1983); 2 = Hillis (1962); 3 = Sithole et al. (1992); 4 = Pausler et al. (1995); 5 = Hubbes (1969); 6 = Dunlop-Jones et al. (1991); 7 = Serreqi et al. (1998); 8 = Nekrasova (1982); 9 = Rowe (1989a) or Rowe (19896). 134 APPENDIX III IDENTIFICATION OF VARIOUS COMPOUNDS BY MASS FRAGMENTOGRAPHY (Note: arrows indicate portion(s) of molecule lost to result in specified fragment ion) 135 A P P E N D I X III (CONT 'D) sakuranet in-TBSMS derivative 63 coniferyl aldehyde | M - C 2H 2 Or 135 I M - M e O ] 147 M + 178 [M - O H ] + 189 203 218 136 A P P E N D I X III (CONT 'D) sinapyl alcohol M e O y O M e [M - C 2 H 4 0] + 167 M+ 210 182 | M - OH]+ I 193 137 APPENDIX IV OPTIMIZATION CURVES FOR ELSD NEBULIZER GAS FLOW RATE (A) AND DRIFT TUBE TEMPERATURE (B) TO 35 ML/MIN AND 155°C (5 ul OF 1 MG/ML SOLUTION OF CHOLESTERYL MYRISTATE (CM), CHOLESTERYL PALMITATE (CP) AND CHOLESTERYL STEARATE (CS) IN HEXANE) 5.0E-02 0.0E + 00 20 30 40 50 60 Nitrogen Flow Rate (ml / min) 6.0E-02 0.0E+00 145 150 155 160 165 170 Drift Tube Temperature (Celsius) 138 APPENDIX V GC-MS RIC PROFILES OF LIPOPHILIC FRACTION FOR CH-F6-02 BEFORE (A) AND AFTER (B) 0.5 M KOH HYDROLYSIS AT 60-70°C FOR 4 HOURS (RUN USING SE/W METHOD) APPENDIX VI MASS SPECTRA OF SELECTED STERYL ESTERS FOUND IN ASPEN Scan 1656 from d:\misc expt\98013 se-w1.sms Sped 1 BP2D3(9006=100%) 98013sew1.sms Scan 2138 from d:\misc expt\98013 se-w1.sms Sped 2 BP 203 (576=100%) 98013 se^»1 sms 35.617 min. Scan: 2138 Chan: 1 Ion: 12177 us RC: 7676 BC 133 176 JL 4 U -408 1.1 111. I Scan 2866 from d:\misc expt\98013 se-w1.sms Sped 1 BP 108(61=100%) 98013se^1.sms 47.751 min. Scan: 2866 Chan: 1 Ion: 21122 us RC: 1825 BC 10 140 APPENDIX VI (CONT'D) Scan 1740 from d:\misc expr\98013 se-w1 .sms Sped 1 BP95 (2855=100%) 98013 se^Lsms 161 203 28.983 min. Scan: 1740 Chan: 1 Ion: 3859 us RIC: 89521 BC 341 365 ll il 11, I Scan 2266 from d:\misc expt\98013 se-wlsms Sped 2 BP 95 (263=100%) 98013 se-v»1.sms 37.750 min. Scan: 2266 Chan: 1 Ion: 13697 us RIC: 6960 BC JI T II I II I i 11 i lil nl L ll ill I I ni'i'll I T U Scan 3030 from d:\misc expt\98013 se-w1 .sms Sped 1 BP 150(145=100%) 98013se^1.sms 50.483 min. Scan: 3030 Chan: 1 Ion: 21301 us RIC: 2133 BC 366 409 492 _ L _ 11 'Fl NHrn T i I  i I  r 141 A P P E N D I X V I ( C O N T ' D ) Scan 1792 from d:\miscexpt\98013 se-w1.sms Sped 1 BP218(5724=100%) 98013 s&wtsms 29.861 min. Scan: 1792 Chan: 1 Ion: 3132us RIC 118288 BC Scan 2341 from d:\misc expf\98013 se-wlsms Sped 2 BP203 (331=100%) 98013 se-w1.sms 3.000 min. Scan: 2341 Chan: 1 Ion: 11327usRIC: 8799 BC ytn II nil :iy ninyi ir^iiii ii^l IH*l)|i j » y ^ 8 mi 409 J L 1 462 400 Scan 3142 from d:\misc expt\98013 se-w1.sms Specll . BP203(40=100%) 98013 se-v»1.sns 52.351 min. Scan: 3142 Chan: 1 Ion: 16201 us RIC: 1232 BC 12 VT m i 11 i i 453 494 142 APPENDIX VII RESPONSE FACTORS (BASED ON HEPTADECANOIC ACID) FOR THE STANDARD COMPOUNDS USED IN GENERIC RESPONSE FACTOR DETERMINATION. (Note: error bars represent 95% CI based on seven replicates) / # / 143 APPENDIX VIII MASS FRAGMENTOGRAPHY OF DOCOSANOIC ACID COMPARISON BETWEEN UNDERIVATIZED (A) AND TBDMS-DERIVATIZED (B) (Note: arrows indicate portion(s) of molecule lost to result in specified fragment ion) 144 APPENDIX IX AMOUNTS (MG/KG FD WOOD) OF EACH COMPOUND FOUND IN THE ACETONE EXTRACTS FOR EACH RAMET BY GC-MS ComDOund K7-02 K7-03 K2-07 K2-06 K2-03 K1-08 K1-09 K1-06 K1-07 K1-04 F7-08 F7-02 benzoic acid /1.11 3.52 7.18 11.63 6.40 8.46 17.26 5.04 6.38 7.22 94.00 9.71 10.17 p-ethvlpheno! / 2.1 0.18 7.61 25.09 14.99 26.36 4.25 1.57 0.54 1.67 0.37 0.93 2.52 o-coumaric acid / 2.2 0.30 1.85 1.56 7.14 3.78 2.93 0.78 0.81 1.00 3.38 2.48 2.03 salicvl alcohol / 2.3 0.98 4.51 3.58 3.42 4.82 7.79 2.08 1.19 1.67 12.84 2.17 9.05 4-hydroxy-2-methylacetophenone / 2.4 1.14 2.17 3.75 3.42 3.78 2.67 1.52 202 1.17 5.70 4.16 1.83 9-oxononanoic acid / 2.5 233.48 105.80 81.95 116.11 115.07 79.25 62.92 87.51 60.60 81.51 37.07 109.82 p-hydroxybenzoic acid / 2.6 34.15 40.26 49.86 32.95 53.52 38.73 37.46 59.28 24.53 85.14 22.55 65.04 p-hvdroxy-hydrocinnamic acid / 2.7 1.77 4.31 23.05 3.99 28.75 1.91 1.48 0.41 0.67 1.01 1.18 0.74 coniferyl alcohol / 2.8 156.27 170.77 116.05 81.04 162.27 102.52 122.24 136.76 84.82 135.58 135.55 148.34 coniferyl aldehyde / 2.9 11.56 13.44 16.98 12.66 13.81 13.84 13.41 22.33 10.35 39.24 13.12 16.24 p-hvdroxycinnamic acid / 2.10 2.74 10.54 1.85 28.53 9.89 7.32 1.60 0.97 3.12 16.23 1.42 1.85 ferulic acid / 2.11 4.49 3.97 4.30 3.26 4.18 3.93 2.56 7.38 3.77 7.36 4.36 3.55 palmitic acid / 2.12 375.53 372.11 1 449.66 356.10 400.87 379.05 396.75 683.16 332.35 692.64 282.06 354.38 sinapyl alcohol / 2.13 120.03 103.72 119.05 32.76 74.32 44.10 96.14 145.29 70.27 137.54 115.94 74.73 sinapyl aldehyde / 2.14 13.10 12.04 21.71 11.08 12.49 15.50 16.27 27.71 12.16 50.10 19.03 14.17 linoleic acid / 2.15 1.865.06 2,038.75 2,727.17 1,635.80 2,199.30 1,527.42 2,224.98 2,116.73 1,330.16 1,315.95 888.54 1,624.19 stearic acid / 2.16 68.00 66.83 60.50 39.51 49.86 63.23 67.36 128.64 57.07 141.24 51.01 56.02 fatty acid 1 60.47 60.69 86.73 37.70 41.10 43.74 65.12 90.80 45.47 75.25 54.26 51.87 eicosanoic acid / 2.17 48.28 52.12 41.52 30.23 36.15 36.32 35.79 97.39 38.88 104.46 30.51 53.75 docosanoic acid / 2.18 63.22 75.61 57.03 53.53 75.89 57.28 62.19 122.85 55.86 127.52 37.65 56.37 fatty acid 2 0.00 189.53 533.94 677.23 1,001.28 6.00 199.03 10.18 2.93 0.00 8.23 4.68 unknown 11.92 22.30 18.54 15.66 27.31 22.63 19.50 26.32 9.32 33.89 16.85 15.35 flavonoid 1 1.37 19.88 0.93 8.03 23.70 3.33 0.82 0.40 1.02 2.66 1.20 2.47 lignoceric acid / 2.19 72.26 102.49 46.91 79.44 154.45 43.79 70.17 135.82 53.65 161.91 43.17 45.14 sakuranetin / 2.20 51.21 278.70 411.09 446.25 406.47 195.37 34.56 15.30 21.55 92.91 75.02 44.14 1-docosanol / 2.21 76.27 99.24 82.12 86.66 90.78 61.48 104.46 127.60 78.48 168.23 115.30 25.93 flavonoid 2 22.66 46.04 24.23 27.91 29.99 18.59 2.69 2.14 3.05 5.33 7.31 25.05 naringenin / 2.22 47.63 278.03 397.76 582.34 1,291.29 188.67 37.62 7.87 14.19 31.92 33.27 73.74 sterol / triterpene 1 54.95 41.64 61.99 54.80 52.46 32.61 40.52 63.01 43.47 91.34 42.14 48.85 sterol / triterpene 2 396.07 184.31 379.89 290.66 305.06 130.59 214.98 456.67 161.88 579.05 230.65 225.81 beta-sitosterol acetate / 2.23 299.02 132.32 198.06 317.11 164.99 253.18 205.41 240.87 111.86 292.71 327.05 124.51 kaempferide / 2.24 21.44 92.18 58.21 50.06 61.19 15.93 4.55 3.38 4.16 16.61 15.71 14.63 vitamin El 2.25 33.91 49.02 38.60 25.84 33.24 55.54 61.89 96.89 48.30 70.65 68.87 35.52 kaempferol / 2.26 7.80 38.06 2.33 56.39 173.82 13.70 0.31 0.00 0.88 6.51 0.00 3.14 beta-sitosterol / 2.27 1,316.80 1,393.55 1,899.39 1,463.14 1,313.26 1,156.31 1,299.02 1,777.34 1,316.67 1,332.41 779.04 1,601.26 sterol / triterpene 3 114.72 156.56 171.13 191.99 137.32 48.71 63.81 98.55 85.99 143.59 81.50 61.57 beta-amvrin/2.28 100.98 125.00 86.23 68.57 85.78 91.22 115.11 123.72 81.59 62.14 51.26 79.95 butyrospermol / 2.29 52.31 132.94 280.47 156.17 342.23 132.86 242.15 162.59 80.10 97.36 41.20 60.10 cycloartenol / 2.30 208.75 208.51 194.70 141.76 154.50 171.28 143.45 223.71 113.23 83.44 46.61 106.16 alpha-amvrin / 2.31 162.40 215.68 189.09 144.83 180.09 155.43 200.22 263.20 145.18 170.31 112.24 120.36 sterol / triterpene 4 325.13 255.54 476.60 340.29 291.72 166.47 195.00 424.37 208.81 702.43 307.56 169.88 24-methvlenecvcloartanol / 2.32 205.29 209.68 208.67 139.12 210.46 115.54 144.31 200.21 131.06 134.18 145.14 148.72 citrostandienol / 2.33 59.71 101.93 104.03 65.78 108.16 87.32 .110.09 116.97 85.02 103.04 , 17.01 69.44 sterol / triterpene 5 119.78 111.75 75.48 66.91 84.16 84.62 100.49 139.53 97.83 179.19 121.19 45.56 sterol / triterpene 6 52.22 47.58 21.79 12.01 15.89 42.02 65.92 97.18 61.08 58.17 20.85 29.71 sterol / triterpene 7 251.47 267.40 307.40 383.85 432.75 128.52 166.25 298.98 207.85 571.28 157.85 148.01 diglyceride 1 8.10 313.51 412.29 593.46 967.60 29.01 263.35 0.00 8.15 0.00 5.56 0.00 sterol / triterpene 8 533.75 511.73 447.75 245.16 300.85 482.55 377.87 630.08 308.72 618.39 642.29 438.68 diglyceride 2 148.21 256.37 568.07 335.29 473.96 193.63 150.59 405.75 164.73 487.63 487.57 146.72 waxes 206.68 241.77 244.58 87.48 90.91 134.18 193.31 465.24 264.78 387.45 220.42 258.51 diglyceride 3 109.15 102.26 264.15 157.33 137.02 143.29 194.77 387.45 168.05 682.37 308.49 38.78 diglyceride 4 166.98 147.38 250.52 181.93 174.28 105.76 167.43 270.12 154.47 228.83 104.35 141.00 steryl / triterpene ester 1 15.73 18.54 12.82 21.20 14.65 16.80 10.94 15.67 11.73 16.16 10.97 13.82 steryl / triterpene ester 2 118.21 85.70 133.02 87.69 66.60 90.06 83.45 177.77 118.63 191.70 104.94 109.82 diglyceride 5 703.47 984.74 2,258.89 780.10 1,241.83 1,256.13 1,796.71 2,001.11 1,319.61 2,440.61 1,032.56 849.80 triglyceride 1 54.48 46.11 32.10 26.02 20.86 36.42 23.68 65.43 34.44 97.53 65.39 38.15 butyrospermol palmitate / 2.34 1,197.30 1,527.22 3,437.94 1,392.72 1,904.62 1,910.10 2,530.17 3,520.04 2,088.35 3,704.22 1,686.86 1,493.68 beta-amyrin palmitate / 2.35 410.70 512.74 565.20 220.05 262.60 354.21 469.76 778.37 364.97 735.55 480.49 321.23 triglyceride 2 46.75 78.69 223.07 80.74 93.14 106.38 138.23 214.66 116.00 123.95 107.80 36.25 lupeol palmitate / 2.36 444.42 630.61 664.65 353.26 368.30 416.91 499.26 787.81 391.54 792.81 516.60 378.63 alpha-amvrin palmitate / 2.37 728.61 911.25 1,073.10 527.61 623.32 726.50 973.28 1,720.07 801.68 1,486.61 804.20 579.91 triglyceride 3 29.92 34.46 40.81 19.99 58.20 33.42 51.87 66.48 17.34 72.74 30.04 41.26 butyrospermol stearate / 2.38 480.42 694.23 1,790.51 612.44 996.03 887.53 1,285.90 1,838.81 758.70 1,283.34 835.52 746.97 beta-amvrin stearate / 2.39 107.40 107.48 319.83 41.54 100.14 123.19 326.46 395.86 63.46 155.89 107.55 • 101.43 lupeol stearate / 2.40 22.52 25.69 21.91 19.53 25.22 23.17 18.08 58.26 24.94 41.21 26.95 30.72 alpha-amvrin stearate / 2.41 177.62 150.45 474.21 43.77 88.76 216.85 406.36 505.68 158.87 260.27 256.53 86.97 stervl / triterpene ester 3 139.12 214.49 406.01 113.92 190.22 133.05 237.11 225.92 84.89 96.06 197.53 162.55 triglyceride 4 85.81 90.07 185.85 48.76 53.36 171.62 198.57 156.24 111.91 76.43 48.93 38.55 steryl / triterpene ester 4 239.76 150.60 133.34 112.77 81.61 133.79 128.14 272.62 146.02 508.19 252.11 141.75 butyrospermol eicosanoate / 2.42 96.86 120.78 234.85 36.81 146.83 152:53 224.16 324.27 177.08 244.43 148.77 135.68 trilinolein / 2.43 794.72 1,160.10 1,703.68 233.23 149.64 869.79 1,388.28 2,298.42 1,909.98 2,985.99 2,469.28 1,190.35 145 APPENDIX IX (CONT'D) Compound F5-02 F5-05 F5-04 F3-07 F3-04 F3-08 F3-05 F3-06 F3-01 D4-06 D4-09 D4-08 D4-03 benzoic acid /1.11 7.91 18.56 42.36 9.55 7.56 2.86 3.76 4.34 14.06 126.32 20.14 18.92 41.57 p-ethylphenol / 2.1 23.01 1.90 58.27 2.53 6.62 0.20 2.11 0.66 23.00 15.81 9.01 34.54 16.36 o-coumaric acid / 2.2 2.12 11.82 31.37 1.57 1.93 0.19 0.64 0.68 7.48 61.00 15.40 18.71 38.31 salicyl alcohol / 2.3 222 4.70 14.64 3.07 3.78 1.68 3.66 2.85 3.92 27.76 9.63 9.49 11.35 4-hydroxy-2-methylacetophenone / 2.4 2.55 3.06 15.98 2.54 2.21 1.04 1.75 1.52 4.86 16.09 5.52 6.09 7.78 9-oxononanoic acid / 2.5 250.72 328.54 214.58 78.74 114.41 92.54 74.99 65.86 140.77 31.81 64.96 51.24 162.16 p-hydroxvbenzoic acid / 2.6 75.87 47.19 131.60 45.10 25.07 30.63 30.13 28.95 42.21 41.86 44.36 41.71 42.41 p-hvdroxv-hydrocinnamfc acid / 2.7 3.85 1.53 75.21 1.28 6.80 0.15 1.30 0.31 2.06 106.54 2.07 49.37 65.50 coniferyl alcohol / 2.8 134.35 70.63 115.32 162.65 151.34 192.49 163.84 145.70 82.32 50.08 77.75 59.35 65.67 coniferyl aldehyde / 2.9 18.65 11.08 31.80 17.57 15.19 15.36 16.72 19.06 17.19 11.66 14.97 11.00 12.74 p-hydroxycinnamic acid / 2.10 1.60 89.76 55.97 1.39 2.42 0.30 1.60 0.82 34.32 177.27 29.43 66.69 167.33 ferulic acid/ 2.11 4.17 4.02 9.28 3.46 2.78 3.05 3.32 2.77 5.48 9.16 3.24 3.48 3.78 palmitic acid / 2.12 579.55 534.65 792.50 298.37 302.73 296.77 251.18 312.33 414.52 293.93 356.58 305.49 318.89 sinapyl alcohol / 2.13 90.73 17.51 53.15 99.85 99.97 122.55 102.03 121.09 29.90 3.84 30.59 16.25 4.20 sinapyl aldehyde / 2.14 21.56 11.24 28.46 19.60 17.07 17.06 17.98 21.88 15.86 7.09 13.45 8.85 8.28 linoleic acid / 2.15 2.574.23 922.70 2,000.89 2,250.56 873.39 1,639.18 1.620.38 2,122.32 2,029.93 1,082.78 2,192.89 830.98 1,007.39 stearic acid / 2.16 113.87 136.66 192.94 54.32 62.37 49.83 41.47 47.88 93.37 37.33 30.61 36.71 43.71 fatty acid 1 97.02 45.93 118.45 57.71 32.06 49.42 29.34 37.00 55.24 36.95 53.08 24.64 26.99 eicosanoic acid / 2.17 64.96 77.89 • 142.01 52.14 35.56 33.16 31.53 27.22 58.05 35.36 34.03 37.83 37.19 docosanoic acid / 2.18 74.21 100.20 149.38 47.95 37.38 42.84 52.72 41.15 75.04 75.32 62.82 59.71 78.61 fatty acid 2 17.32 5.75 14.20 119.04 33.65 4.74 17.91 7.53 13.06 9.03 16.46 4.89 1.45 unknown 29.59 15.43 52.23 26.66 18.93 11.88 16.54 12.25 21.25 30.46 28.27 2326 20.83 flavonoid 1 6.61 15.85 66.61 3.25 1.98 0.35 1.64 0.33 34.30 35.56 19.38 36.09 50.18 lignoceric acid / 2.19 95.07 135.24 289.24 67.59 59.52 39.70 73.13 57.36 162.95 87.20 71.70 81.56 115.06 sakuranetin / 2.20 150.50 459.03 924.20 93.05 140.11 56.65 80.81 68.49 728.45 901.67 345.40 338.40 870.20 1-docosanol / 2.21 71.71 77.48 118.88 50.77 65.28 64.82 55.22 76.07 81.69 56.24 36.65 65.63 42.66 flavonoid 2 62.40 78.16 262.31 12.92 16.45 23.67 10.00 16.54 89.10 403.44 211.69 124.46 529.59 naringenin / 2.22 138.47 809.26 1,708.70 169.92 131.28 20.00 95.87 54.09 729.47 1,152.01 420.50 469.16 1,525.37 sterol / triterpene 1 81.94 74.53 164.00 43.68 44.72 48.23 38.40 35.52 83.74 57.72 66.21 52.25 74.95 sterol / triterpene 2 522.66 158.23 613.03 175.54 105.63 166.14 129.73 198.55 425.54 189.72 64.83 194.27 164.83 beta-sitosterol acetate / 2.23 593.26 188.36 485.04 166.77 251.39 141.47 144.44 120.14 285.29 147.15 292.09 181.41 130.65 kaempferide / 2.24 43.54 47.60 240.38 21.52 26.21 6.48 10.46 5.87 130.50 161.94 77.58 70.61 135.70 vitamin E/2.25 48.84 28.76 70.18 72.42 34.92 66.07 40.45 52.23 35.00 27.36 61.75 27.46 29.74 kaempferol / 2.26 26.89 123.50 425.17 16.62 5.82 0.40 1.71 0.54 128.24 182.33 49.30 81.56 255.95 beta-sitosterol / 2.27 1.451.91 859.40 1,673.94 1,488.56 1,360.26 1,565.22 1,225.29 1,466.08 1,599.65 1,140.19 1,916.99 1,332.87 1,407.14 sterol / triterpene 3 148.27 201.76 135.63 65.23 142.11 90.49 104.65 62.37 114.14 51.95 57.17 83.02 59.22 beta-amyrin / 2.28 158.31 102.69 145.90 96.08 72.71 86.99 91.31 81.57 92.43 53.71 99.52 68.41 90.35 butyrospermol / 2.29 210.65 135.45 230.39 126.73 101.66 59.94 104.45 51.29 138.88 83.74 189.38 110.65 131.84 cycloartenol / 2.30 338.39 203.04 337.73 91.65 129.76 94.58 128.93 83.45 98.11 101.35 199.66 149.04 104.66 alpha-amyrin / 2.31 309.95 189.92 277.52 226.55 157.83 201.71 174.11 181.09 194.48 88.22 180.48 119.56 141.77 sterol / triterpene 4 324.40 331.21 485.22 211.67 287.16 212.20 179.62 180.96 304.18 191.33 182.79 214.12 186.93 24-methylenecycloartanol / 2.32 347.69 233.79 281.07 179.20 148.27 203.50 240.38 126.50 259.94 160.12 206.32 171.04 180.63 citrostandienol / 2.33 139.34 48.52 141.17 74.34 72.16 96.60 36.05 85.96 91.66 40.06 107.76 36.74 85.89 sterol / triterpene 5 184.63 115.74 165.40 96.83 56.41 80.24 65.00 78.72 75.64 40.56 61.47 46.80 51.15 sterol / triterpene 6 102.52 52.87 107.67 60.88 17.17 66.14 48.72 42.74 63.83 16.55 64.69 30.22 47.49 sterol / triterpene 7 179.71 175.25 357.85 190.80 225 32 219.14 205.19 191.00 224.24 298.29 157.20 258.37 269.55 difllvceride 1 12.92 14.10 7.40 185.15 102.90 3.00 69.62 6.47 48.03 0.00 36.40 11.51 0.00 sterol / triterpene 8 544.36 259.41 1,094.34 680.88 604.96 593.65 677.21 617.56 600.51 292.67 378.55 293.45 186.13 difllvceride 2 283.26 259.87 670.55 233.40 225.42 199.53 201.71 236.93 268.02 254.92 114.17 242.13 189.04 waxes 176.52 66.33 230.71 302.48 99.83 268.13 176.34 273.96 194.27 110.61 166.69 91.95 94.22 difllvceride 3 216.68 142.61 550.83 78.09 98.51 75.67 65.32 94.90 97.55 86.40 62.23 127.43 88.37 dialvceride 4 359.12 296.86 380.10 144.14 140.61 194.88 144.97 122.11 245.11 63.72 171.90 115.78 117.31 steryl / triterpene ester 1 23.96 30.10 58.01 20.32 24.67 17.05 14.85 13.95 32.73 4.29 20.67 3.24 9.29 steryl / triterpene ester 2 191.40 105.13 243.30 71.37 120.63 106.62 79.06 90.00 129.55 56.19 113.60 86.46 97.05 difllvceride 5 1,314.65 512.78 1,405.58 1,129.02 622.82 1,054.59 754.99 964.15 982.13 1,118.45 1,858.20 856.78 939.41 triglyceride 1 65.02 60.07 86.07 29.24 44.94 26.78 32.22 32.23 46.96 25.75 26.79 35.87 32.55 butyrospermol palmitate / 2.34 2,217.24 1,085.36 3.138.15 1,679.40 1,061.13 1,586.17 1,285.63 1,581.44 1,884.08 1,689.96 2,643.66 1,275.98 1,405.10 beta-amyrin palmitate / 2.35 605.14 211.25 792.53 276.61 204.10 297.66 268.00 319.52 199.03 205.42 347.31 168.51 226.87 triglyceride 2 154.79 64.33 215.56 57.94 56.47 62.24 69.06 102.97 24.34 56.85 142.46 10.70 42.01 lupeol palmitate / 2.36 782.55 481.41 1,110.76 430.37 373.73 417.73 413.44 411.43 514.32 336.36 464.33 296.89 341.96 alpha-amyrin palmitate / 2.37 1,270.24 555.26 1,525.32 756.14 526.48 684.78 593.26 723.91 883.06 391.01 779.90 406.84 500.03 triglyceride 3 103.10 72.30 131.40 22.87 33.04 26.61 12.68 24.72 16.76 1.70 45.13 22.11 27.86 butyrospermol stearate / 2.38 1,110.96 562.67 1,646.34 1,073.98 490.06 940.17 686.86 968.49 1,081.30 495.99 1,069.12 392.69 521.59 beta-amvrin stearate / 2.39 79.63 7.30 238.09 159.78 36.62 154.82 114.81 195.02 36.37 10.86 46.77 15.18 20.47 lupeol stearate / 2.40 39.1-1 35.27 64.08 23.18 36.10 31.68 18.08 24.38 42.70 10.35 23.50 21.94 23.10 alpha-amvrin stearate / 2.41 496.44 79.37 244.72 307.43 86.53 207.79 200.94 210.33 389.29 126.63 134.26 78.58 63.54 steryl / triterpene ester 3 274.18 109.03 453.13 299.05 125.41 229.82 158.71 229.64 280.64 100.23 199.05 76.38 76.95 triglyceride 4 87.39 37.21 130.07 40.84 35.85 74.07 26.69 51.74 84.11 86.51 47.74 75.66 66.56 steryl / triterpene ester 4 200.33 253.34 362.76 121.51 169.01 136.78 106.94 107.55 172.57 130.35 89.93 143.00 126.90 butyrospermol eicosanoate / 2.42 167.75 89.82 223.09 174.65 110.43 151.41 130.47 124.53 190.63 126.57 183.79 94.91 101.57 trilinolein / 2.43 426.68 103.36 646.08 1,477.18 442.81 2,041.29 747.54 1,141.09 624.22 516.56 1,034.87 307.47 335.68 146 A P P E N D I X I X ( C O N T ' D ) Compound D4-02 D4-07 D2-16 02-02 D2-04 D2-01 D2-13 D2-11 D2-12 D2-17 D2-05 D2-08 P1-12 benzoic acid /1.11 10.18 89.65 216.26 11.68 12.00 102.11 2.87 4.96 5.84 3.74 5.97 159.54 10.38 p-ethylphenol / 2.1 0.40 20.61 5.34 15.57 1.51 0.59 0.80 1.49 12.33 0.37 4.30 1.05 0.27 o-coumaric acid / 2.2 3.54 17.29 43.43 9.86 3.82 21.62 0.41 0.79 1.67 1.64 1.19 25.67 2.46 salicyl alcohol / 2.3 2.87 1.10 35.38 3.31 1.30 12.95 0.45 4.97 22.85 3.18 3.00 13.85 4.01 4-hydroxy.2-methylacetophenone / 2.4 1.67 11.91 9.04 7.18 2.65 7.86 1.06 1.50 1.79 1.89 0.97 6.08 1.94 9-oxononanoic acid / 2.5 161.33 56.10 68.82 156.02 146.31 244.41 229.66 166.20 125.70 144.27 237.16 96.83 388.87 p-hydroxvbenzoic acid / 2.6 31.46 64.23 31.47 57.12 55.40 63.51 39.35 45.26 45.22 23.93 71.20 60.61 25.35 p-hvdroxy-hvdrocinnamic acid / 2.7 1.81 8.65 7.45 1.91 3.55 6.49 0.95 1.05 5.42 1.73 0.76 2.21 0.78 conifervl alcohol / 2.8 96.20 72.17 0.50 206.99 99.28 22.46 120.74 136.16 108.58 124.90 245.83 14.35 143.10 coniferyl aldehyde / 2.9 9.47 24.12 6.70 17.52 12.62 10.80 12.95 12.77 13.81 16.40 12.62 10.52 12.24 p-hydroxycinnamic acid / 2.10 28.84 79.65 159.58 28.03 7.90 106.57 0.50 0.46 1.41 0.95 2.90 136.21 4.88 ferulic acid / 2.11 1.86 5.29 3.48 6.05 1.44 6.75 1.47 1.39 1.59 3.24 1.10 3.33 4.44 palmitic acid / 2.12 321.40 422.20 313.85 309.45 301.16 375.83 385.18 369.56 380.54 351.74 392.64 412.70 472.22 sinapyl alcohol / 2.13 35.33 13.36 0.70 56.34 46.72 2.31 88.55 102.69 72.44 82.27 149.94 1.23 58.90 sinapyl aldehyde / 2.14 9.56 14.56 4.37 15.25 13.51 7.61 14.29 16.20 15.39 15.00 14.55 8.42 8.23 linoleic acid / 2.15 1.208.81 1,373.63 538.52 1,078.78 1,645.01 998.77 1,404.46 1,226.04 1,575.03 2,209.93 1,943.65 458.80 858.53 stearic acid / 2.16 39.30 40.11 63.24 35.17 40.73 36.16 54.58 49.98 57.65 47.87 59.50 78.13 92.31 fatty acid 1 38.51 56.60 15.74 28.44 41.81 29.17 72.76 50.30 53.07 72.84 57.20 13.42 41.59 eicosanoic acid / 2.17 39.10 42.43 53.85 27.14 28.47 33.18 36.96 37.99 37.81 36.38 57.99 70.52 74.79 docosanoic acid / 2.18 72.33 86.76 84.00 64.97 30.27 75.14 45.00 49.12 50.32 70.03 66.49 79.85 95.96 fatty acid 2 0.00 5.58 31.15 160.30 34.01 622 0.73 66.49 14.41 10.35 16.91 0.00 0.00 unknown 10.78 44.73 18.32 24.56 11.67 15.77 8.14 9.64 11.83 14.09 13.06 12.25 11.70 flavonoid 1 18.44 65.88 29.27 67.58 5.26 60.76 0.56 1.92 12.24 0.65 3.57 29.09 5.89 lignoceric acid / 2.19 90.53 158.11 71.10 108.81 24.76 85.12 31.71 26.14 31.62 60.64 " 50.76 116.65 85.42 sakuranetin / 2.20 241.19 611.37 548.79 561.75 256.07 1,020.55 15.00 56.37 186.31 39.66 194.33 724.65 158.35 1-docosanol/2.21 50.29 65.54 35.36 88.43 29.46 32.71 36.91 38.20 34.68 40.65 27.69 27.29 65.45 flavonoid 2 79.76 148.09 103.26 205.36 107.91 359.36 10.21 29.44 46.63 21.38 69.79 126.84 78.17 naringenin / 2.22 239.39 784.71 915.20 1,210.42 209.05 944.19 3.72 48.91 276.27 57.39 182.92 1,410.90 133.32 sterol / triterpene 1 64.53 86.53 50.17 29.69 61.11 31.99 78.10 81.22 53.09 41.82 73.29 92.48 101.55 sterol / triterpene 2 249.74 473.39 285.20 337.11 180.63 102.70 293.80 327.35 145.86 126.33 218.37 432.29 429.02 beta-sitosterol acetate / 2.23 211.85 205.62 190.57 127.83 140.39 236.42 181.88 183.50 230.67 301.47 215.28 106.59 32926 kaempferide / 2.24 83.66 123.99 79.11 274.29 89.98 239.04 9.08 30.97 63.33 7.06 63.19 67.10 96.64 vitamin E / 2.25 31.12 40.73 19.97 29.38 19.37 26.55 16.34 25.75 38.24 92.29 23.61 20.65 19.52 kaempferol / 2.26 64.01 216.19 172.06 165.94 14.83 351.39 0.23 0.78 9.08 2.51 7.90 129.04 15.86 beta-sitosterol / 2.27 1,373.03 1,862.36 888.49 1,611.77 1,353.99 1,199.24 1,355.32 1,560.64 1,388.61 1,258.32 1,630.47 868.32 1,465.28 sterol / triterpene 3 76.88 77.85 96.25 67.23 64.79 68.97 44.28 98.54 116.55 76.36 73.05 94.55 76.45 beta-amyrin / 2.28 81.31 94.66 35.05 76.99 117.45 62.28 102.12 123.64 102.69 60.47 138.56 70.70 59.24 butyrospermol / 2.29 90.23 119.05 63.09 125.91 97.59 49.88 62.14 113.50 124.49 66.58 128.64 87.63 81.10 cvcloartenol / 2.30 155.65 163.58 64.42 158.55 142.18 41.92 81.28 107.70 136.07 138.69 169.25 80.53 139.80 alpha-amyrin / 2.31 130.26 159.85 64.52 136.80 175.10 115.00 137.86 202.30 154.15 191.57 220.57 94.95 147.74 sterol / triterpene 4 193.49 244.67 324.46 235.33 212.52 255.47 291.93 237.51 174.17 168.23 173.01 396.11 384.10 24-methylenecycloartanol / 2.32 156.13 225.91 103.05 162.39 192.46 103.68 117.65 149.05 148.33 230.36 147.00 177.43 189.59 citrostandlenol / 2.33 86.65 136.77 21.41 48.18 46.01 7.88 52.03 46.67 71.71 77.05 73.04 28.68 53.71 sterol / triterpene 5 46.02 49.01 28.69 45.59 36.11 41.19 59.47 63.57 81.88 79.67 69.97 33.79 66.52 sterol / triterpene 6 54.66 59.09 34.31 34.64 39.94 66.62 37.36 31.80 37.61 76.95 57.74 50.41 21.13 sterol / triterpene 7 270.56 249.22 177.92 187.29 113.80 138.30 148.68 133.30 147.70 76.03 159.13 213.77 192.47 diglyceride 1 0.00 45.26 34.31 148.58 37.17 0.00 1.74 89.38 14.40 0.00 56.40 13.10 0.00 sterol / triterpene 8 250.38 648.94 362.05 870.15 615.66 385.16 512.92 629.97 664.00 505.20 710.66 334.58 445.52 diglyceride 2 144.03 31.76 194.59 29322 121.86 128.02 111.09 89.71 98.80 340.70 106.42 167.36 279.47 waxes 151.84 268.76 22.56 63.31 181.01 28.60 123.73 90.26 253.59 333.21 276.15 24.41 78.30 diglyceride 3 68.28 79.66 40.93 165.01 64.46 57.81 51.78 49.89 74.89 80.39 37.01 79.15 120.12 diglyceride 4 181.97 241.81 126.02 125.96 193.07 145.06 272.25 236.73 . 219.95 116.57 300.63 300.19 357.12 steryl / triterpene ester 1 12.61 13.37 16.72 8.29 32.88 23.67 27.28 36.59 34.74 32.68 16.82 41.87 33.82 steryl / triterpene ester 2 123.32 151.14 67.29 73.45 142.83 82.73 183.91 167.93 124.83 75.81 145.70 126.24 209.19 diglyceride 5 1,163.53 2,318.88 336.47 515.53 932.53 624.88 831.95 1,024.72 1,464.44 1,311.89 1.103.52 435.93 625.95 triglyceride 1 29.81 45.18 28.59 23.17 52.88 46.86 65.82 54.74 39.05 32.40 45.19 70.02 45.69 butyrospermol palmitate / 2.34 1,685.35 2,964.20 561.89 933.47 1.601.94 1.024.44 1,449.06 1,652.19 2,304.73 1,983.02 1,829.40 843.05 1,413.28 beta-amyrin palmitate / 2.35 270.68 468.52 53.57 127.23 364.25 231.21 346.15 351.89 457.28 418.48 421.20 213.95 258.63 triglyceride 2 27.41 138.80 39.82 57.50 97.43 64.69 94.60 90.29 116.16 127.09 77.83 58.62 44.39 lupeol palmitate / 2.36 395.79 513.00 345.12 204.44 433.88 392.84 427.46 407.53 511.58 435.81 544.92 380.72 334.88 alpha-amyrin palmitate / 2.37 566.97 931.10 262.96 386.02 682.75 447.47 634.58 711.60 852.16 1,292.56 906.28 447.47 726.24 triglyceride 3 35.53 21.30 15.03 39.50 28.09 45.09 56.31 71.51 42.73 64.29 76.27 54.09 66.06 butyrospermol stearate / 2.38 718.05 1.414.88 142.98 322.87 406.99 211.87 534.47 694.31 1,086.26 1,256.27 686.17 365.86 438.17 beta-amyrin stearate / 2.39 35.64 188.82 0.00 8.29 97.01 15.17 85.80 52.07 186.36 248.41 71.93 13.36 19.24 lupeol stearate / 2.40 21.49 22.10 19.97 10.35 33.83 13.63 38.31 43.01 26.70 28.65 27.24 41.25 55.19 alpha-amyrin stearate / 2.41 87.28 171.16 37.65 33.53 124.42 53.20 68.05 79.19 152.18 356.56 131.87 34.78 119.97 steryl / triterpene ester 3 124.96 234.11 40.64 73.30 165.73 65.86 98.12 119.51 272.89 452.38 170.92 57.45 93.51 triglyceride 4 95.42 112.27 13.16 40.54 44.38 25.63 30.05 68.66 137.12 167.12 55.53 19.23 46.72 steryl / triterpene ester 4 140.61 165.97 130.95 79.19 164.98 108.89 250.53 190.93 167.87 152.13 166.30 320.32 213.27 butyrospermol eicosanoate / 2.42 135.49 257.44 38.47 47.10 111.29 41.96 118.78 146.02 205.57 203.87 160.79 73.75 100.05 trilinolein / 2.43 699.72 1,504.54 16.98 98.52 605.59 30.85 479.10 357.85 1,417.84 3,467.21 853.86 15.40 111.96 147 APPENDIX IX (CONT'D) Compound D1-05 D1-20 D1-18 D1-06 D1-17 D1-16 D1-19 D1-04 D1-03 D1-09 •1-15 D1-14 benzoic acid /1.11 4.93 8.91 6.46 5.86 22.14 5.56 7.42 4.85 9.57 21.72 3.14 14.34 p-ethylphenol / 2.1 2.45 2.58 0.62 1.78 1.77 0.40 0.33 0.71 0.86 3.15 3.75 1.05 o-coumaric acid / 2.2 0.35 1.95 2.25 0.75 15.22 1.10 0.57 0.67 1.10 3.38 0.23 1.75 salicyl alcohol / 2.3 3.44 4.49 2.91 2.20 4.36 2.84 2.23 7.14 2.25 7.55 4.12 5.62 4-hydroxy-2-methylacetophenone / 2.4 0.73 1.44 1.57 1.09 5.14 1.80 0.79 0.99 1.25 2.19 0.77 1.45 9-oxononanoic acid / 2.5 32471 256.77 337.66 301.41 430.04 458.45 377.47 203.33 396.55 340.38 466.55 503.99 p-hydroxybenzoic acid / 2.6 19.38 26.06 21.91 30.97 46.54 49.61 34.76 13.62 33.00 35.54 34.53 49.49 p-hydroxy-hydrocinnamic acid / 2.7 0.55 0.89 4.28 1.83 21.70 2.64 0.88 0.86 2.15 1.42 0.91 2.03 coniferyl alcohol / 2.8 87.75 120.73 101.85 82.21 93.39 100.47 100.68 84.44 111.25 121.44 77.54 114.86 coniferyl aldehyde / 2.9 9.44 11.13 12.17 9.70 12.33 13.17 9.93 6.51 11.17 18.05 8.99 14.97 p-hydroxycinnamic acid / 2.10 0.73 7.73 6.34 0.93 11.97 2.59 2.04 1.12 2.47 22.40 0.29 13.36 ferulic acid / 2.11 2.59 2.50 3.14 1.45 4.57 3.35 2.69 1.95 3.34 4.41 1.21 6.42 palmitic acid / 2.12 497.62 398.50 469.28 439.05 480.19 542.89 482.07 419.77 528.88 480.37 420.86 706.98 sinapyl alcohol / 2.13 58.43 79.10 56.83 59.05 44.36 93.10 73.95 52.68 66.13 37.66 63.58 71.14 sinapyl aldehyde / 2.14 8.90 11.66 14.47 11.82 12.39 18.01 12.32 7.74 11.79 13.09 10.76 13.02 linoleic acid / 2.15 1,302.37 899.01 1,310.29 1,712.63 1,341.19 996.55 1,287.82 880.95 1,356.09 943.70 1,716.36 1,314.50 stearic acid / 2.16 82.90 64.99 72.87 70.54 73.77 103.08 70.39 71.48 99.45 83.48 64.92 150.51 fatty acid 1 55.12 26.53 39.42 67.60 34.17 38.55 43.00 29.15 51.88 27.25 39.15 52.51 eicosanoic acid / 2.17 70.86 44.25 54.98 52.97 62.74 82.34 59.18 57.53 70.12 60.86 42.82 102.04 docosanoic acid / 2.18 72.24 73.70 65.93 50.93 86.42 80.23 72.88 65.41 86.02 95.06 71.03 122.90 fatty acid 2 9.01 22.80 19.42 16.58 14.58 10.02 15.66 10.54 21.26 17.55 111.82 0.00 unknown 7.82 11.99 12.52 8.49 18.07 11.87 9.95 6.84 11.72 16.23 6.67 12.11 flavonoid 1 1.08 4.64 5.16 3.26 13.77 1.08 1.87 2.78 2.72 11.72 3.01 2.63 lignoceric acid / 2.19 73.63 127.37 53.14 27.04 88.61 86.45 53.97 62.29 70.97 103.86 63.99 124.42 sakuranetin / 2.20 18.43 119.88 47.21 76.69 352.19 40.95 32.19 29.09 53.18 173.73 23.53 90.38 1-docosanol/2.21 57.61 122.25 88.14 29.89 101.50 96.61 105.89 81.71 55.24 62.46 58.05 68.69 flavonoid 2 32.32 40.43 28.36 48.71 169.72 11.07 18.02 18.30 19.73 75.89 17.63 38.13 naringenin / 2.22 16.69 232.48 78.08 63.40 606.71 74.99 39.28 29.13 50.10 133.68 63.13 73.97 sterol / triterpene 1 78.85 79.35 72.98 113.83 108.62 112.47 118.69 103.67 79.48 83.52 80.90 147.61 sterol / triterpene 2 435.38 299.29 447.12 107.09 395.83 678.25 248.25 278.31 625.90 222.61 49.21 585.36 beta-sitosterol acetate / 2.23 320.31 246.41 377.22 523.89 266.47 349.95 469.18 496.27 529.40 461.83 365.33 840.70 kaempferide / 2.24 15.44 52.86 29.61 31.80 167.71 14.35 18.24 14.06 19.21 66.18 10.75 29.58 vitamin El 2.25 20.32 22.07 27.17 20.75 23.28 27.59 23.53 14.22 19.73 16.27 19.59 13.48 kaempferol / 2.26 0.00 33.84 7.13 1.70 126.98 8.78 5.49 0.86 8.19 22.77 3.27 7.52 beta-sitosterol / 2.27 1,393.20 1,386.36 1,430.21 1,403.87 1,491.33 1,505.72 1,273.17 1,182.75 1,462.01 1,438.33 1,120.92 1,581.24 sterol / triterpene 3 116.47 134.82 114.81 110.00 72.64 131.54 80.26 125.57 124.83 150.46 103.53 152.71 beta-amyrin / 2.28 51.63 103.83 88.19 58.25 95.39 99.78 98.44 78.98 62.51 60.22 82.53 66.58 butyrospermol / 2.29 115.87 129.22 81.69 115.22 112.12 136.97 121.01 95.55 103.27 121.54 101.53 73.48 cycloartenol / 2.30 135.90 71.66 133.04 173.15 62.82 140.76 108.20 60.81 99.90 146.03 67.10 141.19 alpha-amyrin / 2.31 161.36 143.38 147.60 168.07 140.95 161.31 139.72 199.56 142.10 134.63 120.08 178.33 sterol / triterpene 4 520.55 333.95 333.04 373.22 349.15 491.98 441.98 412.76 407.71 415.96 244.70 759.22 24-methylenecyeloartanol / 2.32 227.89 255.59 167.49 267.92 275.87 201.06 263.53 150.57 175.46 213.82 259.40 227.05 citrostandienol/2.33 52.57 37.03 26.46 57.19 67.32 32.71 65.60 30.83 48.38 35.95 38.81 54.00 sterol / triterpene 5 69.44 59.39 75.38 93.31 73.85 55.86 45.73 71.32 69.82 56.92 74.91 90.38 sterol / triterpene 6 20.72 51.53 36.53 26.97 44.40 60.88 41.47 17.35 24.62 32.71 61.60 24.34 sterol / triterpene 7 197.05 267.48 181.84 139.15 198.87 251.47 206.33 217.59 200.01 204.11 160.10 307.83 diglyceride 1 0.00 34.77 27.96 0.00 10.26 1.84 0.00 17.15 0.00 14.87 97.16 0.00 sterol / triterpene 8 321.52 390.68 583.08 335.43 626.61 603.01 446.88 253.60 402.56 562.83 414.04 572.26 diglyceride 2 219.59 216.29 153.06 57.93 372.85 333.07 286.77 298.43 267.71 304.74 129.89 91.92 waxes 90.43 72.54 100.06 162.08 90.86 89.69 44.14 96.88 93.48 58.95 86.36 125.76 diglyceride 3 93.32 150.00 125.66 42.16 184.56 192.95 114.97 108.50 64.27 184.43 39.57 111.92 diglyceride 4 350.69 282.09 256.03 489.23 306.95 425.42 388.64 211.45 454.22 370.54 358.60 688.72 steryl / triterpene ester 1 52.78 22.35 47.97 65.10 78.91 72.31 60.06 30.66 57.70 66.59 76.48 151.21 steryl / triterpene ester 2 216.54 173.08 153.11 196.79 157.87 231.20 177.23 174.59 190.24 169.99 140.86 314.71 diglyceride 5 768.72 627.75 658.96 1,190.40 723.11 711.11 581.62 576.07 838.08 622.03 635.48 789.19 triglyceride 1 56.20 47.93 55.50 43.98 54.96 78.21 66.64 54.49 46.72 40.50 45.04 57.83 butyrospermol palmitate / 2.34 1,601.14 1,190.68 1,224.66 2.124.67 1,315.01 1,396.83 1,109.18 1,298.22 1,555.83 1,250.33 1,310.27 1,894.33 beta-amyrin palmitate / 2.35 229.10 242.10 267.28 318.65 263.57 346.22 220.60 187.05 296.61 230.91 219.71 370.04 triglyceride 2 86.99 36.47 36.94 123.40 16.40 19.06 51.20 74.17 59.09 68.11 39.97 52.17 lupeol palmitate/ 2.36 367.82 318.88 341.58 408.61 380.20 427.90 368.35 283.86 375.95 315.46 313.69 471.66 alpha-amyrin palmitate / 2.37 759.69 392.41 460.10 1,003.45 447.48 616.70 412.28 583.02 833.55 627.35 471.28 1,000.12 triglyceride 3 83.83 45.92 74.73 95.20 71.78 99.83 76.71 72.93 81.05 76.51 89.47 44.44 butyrospermol stearate / 2.38 501.11 366.15 424.05 831.36 370.42 343.51 347.49 438.61 559.60 368.45 496.46 673.03 beta-amvrin stearate / 2.39 13.64 20.41 22.83 38.23 1652 11.05 25.82 22.49 61.06 16.51 20.95 20.62 lupeol stearate / 2.40 57.69 38.66 36.15 48.74 46.08 44.08 59.72 38.47 56.94 53.66 42.75 81.16 alpha-amvrin stearate / 2.41 66.30 43.73 38.90 278.29 83.61 33.58 5.80 30.30 35.78 101.35 26.10 181.82 steryl / triterpene ester 3 131.54 62.54 86.23 186.16 101.72 65.42 59.64 58.75 144.80 86.35 107.34 172.64 triglyceride 4 79.40 38.10 42.03 52.08 49.38 27.29 32.77 57.05 48.38 45.25 48.08 64.35 steryl / triterpene ester 4 281.96 120.66 151.01 237.44 147.02 235.28 182.46 247.67 227.69 209.99 108.18 398.54 butyrospermol eicosanoate / 2.42 135.83 97.29 115.23 204.85 95.68 124.72 112.11 116.11 95.97 113.48 122.33 163.67 trilinolein / 2.43 288.47 58.70 128.67 448.64 80.86 91.60 54.12 191.60 232.51 75.27 129.95 256 62 148 APPENDIX X ANOVA RESULTS FOR THE QUANTITY OF EXTRACTIVE COMPONENTS FOUND IN NATURAL ASPEN CLONES 149 APPENDIX XI DECAY CORRELATION PLOTS FOR BENZOIC ACID AND O-COUMARIC ACID (Note: R or correlation coefficient is listed on every correlation plot) •a 2 § 3 * 2 = -* <U — CQ oi E 250, 200 150 100 50 — i — i — i — i — i — i — i — i — r R = 0.37 ° o o o o 0 f i l l J L 1 1.5 2 2.5 3 3.5 4 4.5 5 Decay Index 2 -o CS O 6 1 1 1.5 2 2.5 3 3.5 4 4.5 5 Decay Index 150 A P P E N D I X XII P E A R S O N S C O R R E L A T I O N M A T R I C E S F O R B I O S Y N T H E T I C A L L Y S I M I L A R C O M P O U N D S F O U N D I N A S P E N (Note: R or correlation coefficient is listed on every correlation plot) 0.91850 0.91170 Fatty Acids Lt ELTZL JZL 0.94130 III r—i 0.83290, 0.7905]) 0.85482 1 r r - m palmitic acid stearic acid eicosanoic acid docosanoic acid 151 A P P E N D I X X I I ( C O N T ' D ) 0.82879 0.91770 0.82622 o o Flavonoids Q.75182 " 0.77160 > T b - r r u 0.76191 Efca. 0.90585 0.78417 0.89745 0.82964 m - m sakuranetin flavonoid 2 naringenin kaempferide kaempferol 152 APPENDIX XII (CONT'D) tin 0.39897 0.22993 Sterols/ Triterpenes IL n 0.47552 0.56870 0.62030 0.30687 Led ta P-sitosterol cycloartenol 24-methylene- citrostandienol cycloartanol 153 APPENDIX XIII CHEMICAL STRUCTURES O O i.e. Flavanone O - O H 1.11 ~Y fin 1.12 Q H 1.13 O H O 154 APPENDIX XIII (CONT'D) 155 APPENDIX XIII (CONT'D) 156 APPENDIX XIII (CONT'D) APPENDIX XIII (CONT'D) APPENDIX XIII (CONT'D) 

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.831.1-0089333/manifest

Comment

Related Items