UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Isolation and characterization of lignin from populus Stewart, Jaclyn Jeanette 2005

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

Item Metadata

Download

Media
831-ubc_2005-0325.pdf [ 6.52MB ]
Metadata
JSON: 831-1.0075069.json
JSON-LD: 831-1.0075069-ld.json
RDF/XML (Pretty): 831-1.0075069-rdf.xml
RDF/JSON: 831-1.0075069-rdf.json
Turtle: 831-1.0075069-turtle.txt
N-Triples: 831-1.0075069-rdf-ntriples.txt
Original Record: 831-1.0075069-source.json
Full Text
831-1.0075069-fulltext.txt
Citation
831-1.0075069.ris

Full Text

ISOLATION AND CHARACTERIZATION OF LIGNIN FROM POPULUS  by  Jaclyn Jeanette Stewart B.Sc, The University of British Columbia  A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES (Forestry)  THE UNIVERSITY OF BRITISH COLUMBIA April 2005  © Jaclyn Jeanette Stewart, 2005  Abstract Wood chemistry, and in particular lignin chemistry, impacts chemical pulp yield and quality. In this research, a thorough lignin chemical structure was elucidated and related to pulp properties of natural aspen clones and genetically modified hybrid poplar. differences  The chemical  among phenotypically distinct naturally grown aspen clones indicated that  considerable variability exists in the concentration of carbohydrates, lignin, and extractive. Lignin monomer ratio was determined using thioacidolysis and nitrobenzene oxidation, and varied from 74.4 to 80.3% syringyl units. The ease of chemical pulping was very strongly correlated with the mol % S derived from nitrobenzene oxidation. Two clones (10-1 and 16-2) that revealed the greatest differences in Kraft pulping efficiencies (difference of yield and Kappa) were subject to extensive lignin analyses. The peak molecular weight of sample 10-1 lignin was shown to be significantly higher than sample 16-2 lignin. These two samples were also employed for in-depth NMR analysis in order to determine side chain linkages, functional group differences, and aromatic carbon nature. It was shown that the structure of the residual lignin generated by ball milling was consistent with Kraft pulp efficacy; the phenolic hydroxyl content and methoxyl content were both higher for sample 16-2 than for sample 10-1. A similar study, employing highly selective and controlled genetically modified lignin, involved in-depth lignin analysis. Hybrid poplar expressing the Arabidopsis ferulate 5-hydroxylase gene under the control of a cinnamate 4-hydroxylase promoter was the model sample.  As expected, the  modified lignin had an increase in syringyl content as measured by thioacidolysis and methoxyl analysis. In addition to this monomer ratio shift, the molecular weight of the lignin was reduced in the modified lignin, which may further enhanced chemical pulping efficiency. This is the first report of a dramatic molecular weight decrease due to up-regulation of a gene in the lignin biosynthetic pathway.  ii  Table of Contents Abstract  ii  Table of Contents  •  iii  List of Tables  vi  List of Figures  vii  List of Abbreviations  x  Acknowledgements  xii  Chapter 1 - Introduction  1  1.1  Plantation Forestry: Rationale and Tree Selection  1  1.2  Wood Chemistry and Cell Wall Structure  5  1.2.1  Lignin  5  1.2.2  Cellulose  8  1.2.3  Hemicellulose  8  1.2.4  Cell Wall Ultrastructure  9  1.2.5  Cell Types  10  Kraft Pulping  11  1.3.1  Kraft Pulping Reactions  13  1.3.2  Pulp Conventions  16  1.3.3  Factors Affecting the Rate of Delignification  17  1.3.4  Structural Effects  17  1.3.5  Effects of Lignin Composition  19  Selection of Trees for Improved Pulping Performance  21  1.3  1.4 1.5  Lignin Structural Analysis  1.6  22  Wet Chemistry Methods for Lignin Studies  24  1.6.1  Degradation Methods  24  1.6.2  Methoxyl Analysis  26  Spectroscopic Methods  27  Nuclear Magnetic Resonance Spectroscopy  27  1.8  Overview and Objectives  31  1.9  Bibliography  33  Chapter 2  Isolation and Characterization of Lignin from Natural Aspen Clones  37  2.1  Introduction  37  2.2  Methods and Materials  40  2.3  Results and Discussion  46  1.7 1.7.1  iii  2.3.1  Wood Composition  46  2.3.2  Kraft Pulping  47  2.2.3  Lignin Monomer Ratio  48  2.3.4  Molecular Weight Determination  50  2.3.5  Nuclear Magnetic Resonance Spectroscopy  52  2.4  Conclusion  59  2.5  Tables and Figures  60  2.5  Bibliography  81  Chapter 3 Elucidating the Effects of Lignin Monomer Composition on Chemical Pulping Efficacy in C4H:F5H Transformed Hybrid Poplar  83  3.1  Introduction  83  3.2  Materials and Methods  86  3.3  Results and Discussion  90  3.3.1  Wet Chemistry Methods  90  3.3.2  Molecular Weight Determination  91  3.3.3  Nuclear Magnetic Resonance Spectroscopy  93  3.4  Conclusion  100  3.5  Tables and Figures  101  3.6  Bibliography  116  Chapter 4  118  4.1  Conclusions and Future Work  118  4.2  Bibliography  121  Appendix A - Supplementary Tables and Figures for Chapter 2  122  Appendix B - Supplementary Tables and Figures for Chapter 3  135  Appendix C - Detailed Materials and Methods  138  C.l  Harvesting Aspen Clones  138  C.2  Modification of C4H-F5H Hybrid Poplar  139  C.3  Determination of Moisture Content  140  C.4  Quantitative Determination of Extractives  140  C.5  Determination of Methoxyl Content  142  C.6  Determination of Monolignol Composition by Degradation - Thioacidolysis  143  C.7 Determination of Monolignol Composition by Degradation - Nitrobenzene Oxidation  145  C.8  Carbohydrate and Acid Soluble & Insoluble Lignin Analysis  146  C.9  Kraft Pulping Trials  148 iv  C. 10  Determination of Pulp Residual Lignin  148  C.ll  Ball Milled Wood  150  C. 12  Isolation of Lignin from Ball Milled Wood  151  C. 13  Molecular Weight Measurements  152  C. 14  Elemental Analysis  152  C.15  Acetylation of Ball Milled Wood  153  C.16  Acetylation of Lignin  153  C. 17  Traditional Acetylation Method  153  C. 18  Nuclear Magnetic Resonance Spectroscopy Experiments  154  C. 19  Lignin Hydroxyl Content  155  C.20  Bibliography  156  v  List of Tables Table 1.1. NMR chemical shift ranges for common functional groups present in lignin  29  Table 2.1. Mole percent syringyl (% S) units in various substrates, wood, MWL, REL, and Kraft pulp, based on thioacidolysis 66 Table 2.2. Comparison of mole % syringyl units as determined by thioacidolysis and nitrobenzene oxidation of extractive-free ground wood samples  67  Table 2.3. Elemental analysis and expanded C9 formula for MWL for 10-1 and 16-2  68  Table 2.4a. Signal assignment in the NMR spectrum of non-acetylated MWL  69  Table 2.4b. Signal assignment in the NMR spectrum of acetylated MWL  70  Table 2.4c. Lignin structural unit calculations  71  Table 2.5. Summary of NMR results relevant to the Kraft pulping mechanism  77  Table 3.1. Thioacidolysis yield and S:G monomer content for isolated lignin fractions  102  Table 3.2. Elemental analysis and expanded C 9 formula for MWL for wild-type and C4H-F5H lignin 103 Table 3.3a. Signal assignment in the NMR spectrum of non-acetylated MWL  107  Table 3.3b. Signal assignment in the NMR spectrum of acetylated MWL  108  Table 3.3c. Lignin substructures determined by NMR  109  Table A l . Kraft pulp yield and Kappa numbers  122  Table A2. Elemental Analysis of wood and lignin fractions  123  Table B1. Elemental Analysis of wood and lignin fractions  137  vi  List of Figures Figure 1.1. Species yield variation in short-rotation species planted in Canada  2  Figure 1.2. A schematic of a portion of hardwood lignin showing the most frequent monomers and important linkages. See Figure 2.12 for more lignin subunits 7 Figure 1.3. Stylized drawing of wood cell wall layers depicting the middle lamella (ML), primary cell wall (P), secondary cell wall layers (Si, S2, S3), and the warty layer (W)  10  Figure 1.4. Major cell types in softwoods and hardwoods  11  Figure 1.5. Kraft pulping mechanism of a P-O-4 ether linkage in lignin  14  Figure 1.6. The formation of a stilbene unit from a p-1 lignin substructure under alkaline conditions  15  Figure 1.7. Precursors to lignin monomers. Coniferyl alcohol gives rise to the guaiacyl structures, and sinapyl alcohol gives rise to the syringyl monomers  19  Figure 1.8. Proposed mechanism of P-ether cleavage from thioacidolysis  25  Figure 2.1. Isolation, purification, and acetylation of various lignin fractions  60  Figure 2.2. Total wood composition of the six natural aspen clones from northern BC, Canada. 61 Figure 2.3. Individual neutral wood sugar composition of the six natural aspen clones from northern BC, Canada  62  Figure 2.4. Total lignin content of the six natural aspen clones from northern BC, Canada  63  Figure 2.5. Percent individual extractives classes of the total extractives of six natural aspen from northern BC, Canada. Error bars represent the standard deviation of three replicates  64  Figure 2.6. Pulp yield and residual Kappa number of pulp (H-factor 1100) from six natural aspen clones from northern BC, Canada  65  Figure 2.7a. Quantitative C NMR spectrum for MWL samples  72  Figure 2.7b. Quantitative C NMR of AcMWL  73  Figure 2.8. Quantitative C NMR spectra of aspen clonal acetylated REL samples  75  Figure 2.9. Molecular weight distributions for acetylated aspen MWL samples  76  Figure 2.10. Molecular weight distributions for acetylated aspen CEL  77  Figure 2.11. Molecular weight distributions for acetylated aspen REL  78  13  13  13  vii  Figure 2.12. Lignin substructures (Capanema et al. 2004)  79  Figure 2.13. HMQC spectrum of sample 10-1 and 16-2 MWL  80  Figure 3.1. Isolation, purification, and acetylation of various lignin fractions  101  Figure 3.2. Gel-permeation chromatograph of acetylated MWL (AcMWL) from wild-type and C4H-F5H poplar 104 Figure 3.3. Gel-permeation chromatograph of acetylated CEL (AcCEL) lignin from wild-type and C4H-F5H poplar 105 Figure 3.4. Gel-permeation chromatograph for acetylated REL (AcREL) lignin from wild-type and C4H-F5H 106 Figure 3.5a. Quantitative C NMR spectra of non-acetylated MWL. Expanded region is shown in Figure 3.5b 110 1 3  Figure 3.5b. Quantitative C NMR spectra of non-acetylated MWL in the region from 157 to 140 ppm : Ill 1 3  Figure 3.6. Quantitative C NMR spectra of acetylated REL  113  Figure 3.7. HMBC of acetylated cell wall wild-type and C4H-F5H sample  114  Figure 3.8. Lignin substructures  115  Figure A l . HMQC spectrum of sample 10-1  124  Figure A2. HMQC spectrum of sample 16-2  125  Figure A3. HMQC spectrum of sample 18-3  126  Figure A4. HMQC spectrum of sample 19-5  127  Figure A5. HMQC spectrum of sample 21-5  128  Figure A6. HMQC spectrum of sample 26-1  129  Figure A7. HMBC spectrum of sample 10-1  130  Figure A8. HMQC spectrum of sample 16-2  131  Figure A9. HMQC spectrum of sample 18-3  132  Figure A10. HMQC spectrum of sample 19-5  133  Figure A l l . HMQC spectrum of sample 26-1  134  l 3  viii  Figure B l . HMQC spectrum of wild-type and C4H-F5H acetylated cell wall  135  Figure B2. HMBC spectrum of wild-type and C4H-F5H acetylated cell wall  136  Figure CI. Isolation, purification, and acetylation of various lignin fractions  152  ix  List of Abbreviations a B y u ~ Ac AC2O Alk Ar BMW bp C C4H CAD CEL CH2CI2 cm Co. D Dl DMSO DNA DP DX e  alpha beta gamma micro approximately acetylated acetic anhydride alkyl aryl ball milled wood base pairs carbon cinnamate 4-hydroxylase cinnamyl alcohol dehydrogenase cellulase lignin methylene chloride centimetre company dextrorotary deionized dimethylsulfoxide deoxyribonucleic acid degree of polymerization Dionex erythro, natural logarithm  et al.  et alia  eV ETDA F5H FID ft G g GC GCMS GM ha HMBC HMQC HP HPLC HSQC K K L m M  electron volts ethylenediaminetetraacetic acid ferulate 5-hydroxylase flame ionization detector feet gelatinous layer, guaiacyl gram gas chromatography gas chromatography mass spectrometry genetically modified hectare heteronuclear multiple bond coherence heteronuclear multiple quantum coherence Hewlett Packard high performance liquid chromatography heteronuclear single quantum coherence Kelvin equilibrium constant litre, levorotary metre molar  eq  x  m Me mg min ML mL mm MS MWL N NAA NBO NMI nOe °C 3  P  P pH pm ppm REL rpm s S Si, S2, S3 t T T, TOCSY um W  cubic metre methyl milligram minute middle lamella millilitre millimetre Murashige and Skoog milled wood lignin normal naphthaleneacetic acid nitrobenzene oxidation N-methylimidazole nuclear Overhauser effect degrees Celsius para Populus  potential of hydrogen provisional method (Tappi) parts per million residual lignin revolutions per minute second syringyl secondary cell wall layers threo, ton time longitudinal (relaxation time) total correlation spectroscopy useful method (Tappi) warty layer  Acknowledgements I would like to thank my thesis supervisor, Dr. Shawn Mansfield, for all his effort and helpfulness during this degree. Shawn is dedicated to his students' success, and this is evident by his willingness to talk and his countless hours of research advice and editing. Also, thanks to Dr. Simon Ellis and Dr. John Kadla, my committee members.  I  appreciate your helpful insight into this research project, and your knowledge helped me to focus on what was most important and attainable. I am also grateful for the help of the members of the Mansfield, Kadla, and Saddler lab groups. Thank you to Kubo Satoshi for your expertise and guidance in lignin analysis. I would like to also give a huge thanks to Fadi Asfour for helping with the final NMR analyses. I also want to express my gratitude to Nick Burlinson, Liane Darge and Marietta Austria of the UBC Chemistry Department NMR facilities. When starting this degree, I had no idea I would meet such great friends in the lab. Tom, Rob, Vicki, Ian, Heather, Jeff, Nick, Lisa and Jenny, thanks for always listening and for your support. My family has also been a great support network. Dr. James Hebden, my high school chemistry teacher, is still a great inspiration to me, six years after high school. Most days I think about his classes and how relevant and interesting he made chemistry to me. Bringing to life atoms and molecules, and actually wanting and helping us to actually understand. I truly am grateful for such a wonderful educational opportunity early on in my life. I hope that one day I will as much of an inspiration to others as he has been.  xii  Chapter 1 - Introduction 1.1  Plantation Forestry: Rationale and Tree Selection As the global population increases at an exponential rate, so will the consumption of •  wood and wood products. At the same time, the public is becoming increasingly concerned with the preservation of our native forests. One way to balance the needs of wood consumers and environmental constraints is to actively adopt plantation forestry; which involves harvesting short-rotation, fast-growing trees with favourable properties. These practices have the potential to supply the desired wood needs and concurrently reduce harvesting pressures on old growth ecosystems. Forest plantations vary from reforested areas to intensively managed monocultures or mixed cultured tree "crops." According to the Food and Agriculture Organization of the United Nations (FAO), plantation forests are those that result from planting and/or seeding or are regularly spaced planting dispersed in an existing indigenous forest. To qualify as a plantation, stands must be intensively managed; those without management are deemed "semi-natural." In Canada, the number of plantation species is somewhat limited due to our harsh winter climate. Plantations are most productive when employing fast-rotation trees; common fast-growing species and their growth yields are shown in Figure 1.1.  1  Hybrid Poplar Populus spp. Norway Spruce Picea abies (L) Karst. Red Pine Pinus resinosa Ait. Larch (Larix spp.) White Spruce Picea glauca (Moench) Voss Trembling aspen Populus tremuloides Michx. Canadian forests average 0  5  10  15  20  25  30  35  Yield (rrf/ha/year)  Figure 1.1. Species yield variation in short-rotation species planted in Canada (CCFM 2001).  Hybrid poplars display very high growth rates (-35 m /ha/year), which is one of the 3  reasons they are suited for plantation forestry for fibre production. Hybrid poplars are crosses of various cottonwood varieties and occupy a range of climatic zones. They are found in every province in Canada, and throughout North America.  In 1992, the International Poplar  Commission reported that over 100,000 hectares of poplar plantations were found in several countries (Commission 1992). In general, the success of poplar plantations can be attributed to the species' fast growth rate, high disease resistance, and exceptional versatility. In Canada, several companies including the former MacMillan Bloedel, Scott Paper Ltd. and Alberta Pacific (AlPac) all have poplar plantations; MacMillan Bloedel began its 300 ha/yr plantation throughout Vancouver Island, the Sunshine Coast, and Washington state in 1995 (Cates 1998). In addition to its many desirable attributes, poplar is relatively easy to plant. It has been suggested that under ideal conditions one tree planter can plant 5,000 cuttings per day (Heilman 1999). Rotations are short for pulp feedstock - minimum 4 years, and 6-7 years considered an 2  average rotation time. For solid wood products, the time to harvest has been suggested to be between 10 and 20 years (Heilman et al. 1995). Poplar can be used for a wide-variety of products, including general-purpose pulp specialty paper, and oriented strand board (OSB).  In addition, it is suitable for producing  composites, especially laminated veneer lumber and laminated strand lumber, due to the uniform nature of the bond it produces. Currently, it is not widely used for lumber products due to its relatively small diameter and susceptibility to heart-rot decay (Balatinecz et al. 2001). In our natural forest, significant variations exist among hybrid poplars and aspen, which allows for the selection of desirable qualities to target specific end uses. Selecting from the natural population can be immediate, provided that there are variations within or between species, and that desired traits exist.  If the desired traits do not exist in nature, breeding  strategies may be implemented in order to produce trees with desired or improved traits. The most extreme and controversial method of selection is modifying a tree through targeted genetic modifications. An example of selecting from natural sources comes from a study on the variations of natural resistance of Populus deltoides to a defoliator.  Of 45 clones examined, 12 were  identified as especially resistant to the pest by measuring the insect's pupal weight, larval period, and leaf area consumed (Singh and Pandey 2002). Furthermore, genetically modifying trees with genes coding for specific traits may produce trees that are even less susceptible to the pest or completely unaffected. For example, an antimicrobial peptide gene has been inserted into a hybrid poplar via agrobacterium-mediated transformation that has been shown to confer enhanced resistance (Liang et al. 2002). Selecting trees with desired traits is economically important for pulp mills. AlbertaPacific Forest Industries Inc., a Kraft pulp mill, runs a tree-improvement program based on "selecting fast-growing, high quality fibre trees which are both insect and disease resistance"  3  (Thomas 2002/2003). This company produces hybrids from poplars and aspens, and its research has identified hybrids that exhibit better growth and performance under various silviculture prescriptions (Thomas 2002/2003). Traits for tree selection may include disease resistance, increased growth rate, fibre quality, and wood density, as well as improved chemical composition and appearance. Wood density influences lumber strength and pulp yield, lignin content correlates with pulp yield and brightness, and fibre length and coarseness correlate with paper properties. Within a plantation forest, growth and disease resistance are the most important traits for which trees are selected, and variation in wood characteristics often allows for later selection of a variety of qualities (Riemenschneider et al. 2001). A company that is capitalizing on being particularly selective and efficient is Aracruz Celulose, a leading producer of bleached eucalyptus pulp for tissue and paper production. Surrounded by 38,170 hectares of eucalyptus plantations, this technologically advanced company has selected clones that are disease and pest free, yield high volume, are well-shaped and self-pruning, have thin branches, and manifest the ability to coppice and produce rooted cuttings (Campinhos 1999). Campinhos et al. (1992) suggests that a large genetic base of clones is necessary for a successful plantation in order to ensure adaptation to different environments. In practice, Aracruz selects trees based on competitive tests at different sites.  Aracruz  established a seven-year rotation in 1989, and by 1995 the plantation's growth rate increased from 28 m hectares per year to 45 m hectares per year.  The economic impacts were  considerable, for example, the amount of wood required to produce 1 ton of air-dried bleached pulp dropped from 4.9 m wood/t pulp to 4.1 m wood/t pulp. The yearly productivity of the 3  3  land consequently grew from 5.9 t per hectare to 10.9 t per hectare.  These gains clearly  demonstrate that diligence in tree selection has both environmental and economic rewards.  4  In this project, natural aspen clones and transgenic hybrid poplar were examined with regards to pulping efficacy. This research study verifies the fact that selection can have dramatic positive economic and environmental effects on downstream wood processing.  1.2  Wood Chemistry and Cell Wall Structure The macroscopic and microscopic characteristics of wood result from the type,  abundance, and associations of chemicals in the wood cells. Cells are primarily composed of the high molecular weight macromolecules cellulose, hemicellulose, and lignin, which are organized into a highly ordered and complex structure that forms the cell wall. Cellulose is the main constituent in the cell wall, but depending on the species and the tree, cellulose, hemicellulose, and lignin can range in content and structure (Fengel and Wegener 1989). The relative amount and composition of each biopolymer determines the physical and chemical properties of wood, and thus the quality of product that can be obtainedfromit.  1.2.1  Lignin Lignin is an irregular three-dimensional biopolymer composed of phenylpropanoid  monomers. When isolated, lignin is involved in chemical reactions that change its composition, resulting in a loss of structural information related to the native structure. Lignin composes 1725% of the cell wall in hardwoods, conferring structural rigidity to the cell wall. Lignin is the substance that has been suggested to hold together hemicellulose and cellulose in regular arrangements called microfibrils.  The morphology of the cell wall is  determined by the arrangement of these microfibrils. Furthermore, the strength of wood has been correlated to the angles in which these microfibrils are arranged in the cell wall. Figure 1.2 shows a schematic of a portion of lignin, depicting important linkages found in the lignin biopolymer. 5  Lignin monomers can be linked together through carbon-carbon bonds or carbon-oxygen bonds that join atoms of the aromatic positions or the three-carbon side chains. These options result in significant variations in linkages, and this complexity is one of the reasons studying the structure of lignin is so difficult.  6  MeO.  OH  MeO'  1  OMe  f  \  (biphenyl ether)  1  OH  I 1  1.2. A schematic of a portion of hardwood lignin showing the most frequent monomers and important linkages. See Figure 2.12 for more lignin subunits. Figure  7  1.2.2 Cellulose Cellulose is a polymer comprised of B-D-glucopyranose linked together via repeating 1-4 glycosidic bonds (Sjostrom 1993). Mainly located in the secondary cell wall, it accounts for 4045% of dry weight of most wood. The exact molecular weight and polydispersity of native cellulose arranged into microfibrils is not known, however, it has been suggested that wood cellulose has approximately 10,000 glucose residues. It is possible that these molecules are of the same molecular weight in all species of wood (Sjostrom 1993). However, the degree of polymerization of cellulose has been shown to vary considerably after isolation (Goring and Timell 1962). Similar to lignin, isolating cellulose from wood risks altering its natural structure. Both crystalline and amorphous cellulose are organized into microfibril bundles that give rise to fibrils, and the fibrils are then organized to give rise to fibres. Hydrogen bonding and Van der Waals interactions between layers of cellulose chains are important adhering forces that keep the chains strongly associated, and result in the substantial strength that fibres possess. Since lignin and carbohydrates exist in the layers of the cell wall, covalent and/or intermolecular interactions may exist between these two biopolymers.  O-acetylated lignin  carbohydrate complexes (LCCs) were found to contain glucose, xylose, mannose, and uronic acids in Populus deltoides wood (Overend and Johnson 1991). The most well know LCCs are benzyl esters, benzyl ethers, and phenyl glycosidic linkages, with only the benzyl ester being reactive under alkaline conditions in the Kraft pulping process (Minor 1986).  1.2.3 Hemicellulose Functioning as support in the cell walls, hemicellulose is a heteropolysaccharide that varies according to the species. The individual sugars that comprise hemicelluloses include glucose,  galactose,  xylose,  methylgalacturonic acid.  arabinose, mannose, 4-O-methylglucuronic acid, and 4-0Hemicelluloses are branched and have a lower degree of 8  polymerization than cellulose,  ranging between 120 and 250  monomers.  Softwood  hemicellulose is principally galactoglucomannan, making up 20% of woody material, while arabinoglucuronoxylan makes up 5-10% of softwood by weight.  In contrast, hardwoods are  primarily composed of glucuronoxylan which accounts for 15-30% of woody material, and glucomannan is 2-5% by weight (Sjostrom 1993).  1.2.4 Cell Wall Ultrastructure The arrangement of biopolymers in the cell wall accounts for many wood properties. Lignin, cellulose, and hemicellulose combine to form cell walls that provide the ability for trees to grow tall and stand strong for many years. Clusters of cellulose are arranged into the unit cell configuration via intermolecular hydrogen bonding and dipole-dipole interactions. A microfibril is a "core crystalline region of cellulose surrounded by the paracrystalline cellulose and short chain hemicellulose" (Fujita and Harada 1991). Hemicellulose has been proposed to form the primary linkages between cellulose and lignin, through intermolecular interactions and/or covalent bonds (Fengel and Wegener 1989). Figure 1.3 depicts the various cell wall layers.  9  Warty layer  T e r t i a r y w a l l (T) la f e w l a m i n a e ) - I n n e r l a y e r ( S ) of secondary wall ,(no laminae except after pulping) 2  Thin lamina between T a n d S j layers  :  Outer layer (S,) o r s e c o n d a r y wall (many laminae SOr-QCP t o a x i s )  • S fibril a n g l e n o t o v e r 30° 2  Thin lamina between S, a n d S l a y e r s  P r i m a r y w a l l (P) (a f e w l a m i n a e )  2  Middle lamella  •Pit  Figure 1.3. Stylized drawing of wood cell wall layers depicting the middle lamella (ML), primary cell wall (P), secondary cell wall layers (Si, S2), tertiary wall, and the warty layer (W) (Clark 1978). 1.2.5 Cell Types Many different cell types exist in trees, providing mechanical support, protection, storage, and conduction. Softwoods and hardwoods share some common cell types, but also have unique cell types, which are depicted in Figure 1.4.  10  softwood  hardwood perforation  wood  axial parenchyma cells  ray parenchyma^ceTT"* (B j ray tracheid  fiber  axial parenchyma cells  CD  ray initials  —<X ^ ray parenchyma eel 1 —'— "" _z-==—-  Figure 1.4. Major cell types in softwoods and hardwoods (Fujita and Harada 2001).  Softwoods are primarily comprised of longitudinal tracheids, which account for 90-95% of the total wood volume. Longitudinal tracheids are 100 times longer than they are wide, and have narrow lumens. The average dimension of a longitudinal tracheid is 3-4 mm in length and 25-45 micrometers in diameter. Longitudinal parenchyma cells comprise of 1-2% of softwood volume. Resin canals are both longitudinal and transverse, and are formed by epithelia cells. The purpose of these parenchyma cells is to secrete resin, which assists in wound healing and the prevention of pest attack. Ray tracheids have thicker walls and bordered pits than longitudinal tracheids. 11  In contrast, hardwoods contain vessel elements that conduct fluid throughout the xylem. Vessel elements have large diameters and form end-to-end connections via perforation plates. Another hardwood-specific cell type is the fibre tracheid. These tracheids are less than 1 mm long, are rounded in cross section, and have thick walls. Fibres confer strength to hardwoods, and provide support to the vessel elements. Longitudinal parenchyma cells have thin walls and are used for storage, similar to softwoods. Poplar, the hardwood investigated in this study, is composed of 53-60% fibres, 28-34% vessel elements, 11-14% ray cells, and 0.1-0.3% axial parenchyma (Panshin and de Zeeuw 1980). Although poplars are considered useful for many different products, they often possess deleterious characteristics that must be overcome. For example, poplar stems may have wet pockets, discoloration, and are prone to heart-rot decay. In order to preclude these undesirable traits, clones may be selected for high-value production (Balatinecz et al. 2001).  1.3  Kraft Pulping Chemical pulping is essentially removing lignin from wood to leave behind cellulose  fibres that can be used to make a variety of products. Chemical pulping involves breaking covalent bonds within lignin in order to solublize it, and consequently liberate a cellulose-rich fibre. In contrast, mechanical pulping does not separate lignin from cellulose, but instead uses physical force to break apart fibres. Therefore, high-yielding mechanical pulp is not used for fine paper products because of the high residual lignin content, producing pulp that is dark in colour and that forms weak paper. Chemical pulping accounted for approximately 80% of the pulping process worldwide (FAO 2002). Chemicals, heat, and pressure are necessary for chemical pulping to liberate lignin into smaller soluble units that are removed with the pulping (black) liquor waste. A variety of chemical pulping methods exist: aqueous alkaline, neutral, or acidic, all of which react at high 12  temperatures and pressures (Sjostrom 1993). The Kraft (sulphate) process, described below, is an alkaline pulping method that accounts for 73% of all chemical pulping worldwide (FAO 2002).  1.3.1  Kraft Pulping Reactions The Kraft pulping process uses aqueous sodium hydroxide and sodium sulphide. The  hydroxide ion breaks lignin into smaller soluble sodium salts, while sodium sulphide increases the activity of sodium hydroxide. The equilibrium between hydrogen sulphide and water serves to produce more hydroxide ions, as shown below (Sjostrom 1993).  S " + H 0 - HS" + OH"  Keq is approximately 10  HS" + H 0 - H S + OH"  Koq is approximately 10"  2  2  2  2  (Eq. 1) 7  (Eq. 2)  The mechanism of Kraft pulping involves alkaline attacks of various lignin side chains and is not entirely understood due to its complexity.  Figure 1.5 depicts the proposed Kraft  pulping mechanism showing a hydrogen sulphide ion acting to break a P-O-4 linkage. The P-O-4 linkage is depicted because it accounts for roughly 60% of all bonds in natural hardwood lignin.  13  Figure 1.5. Kraft pulping mechanism of a P-O-4 ether linkage in lignin. Hydrogen sulphide ion is the attacking nucleophile. The process releases soluble salts (Sjostrom 1993) and is initiated with the deprotonation of the phenolic hydroxyl group.  The active species in the reagent solution are formed through the complete dissociation of NaOH and Na2S. The S " ion produced goes on to hydrolyse a water molecule to produce SH" 2  and more OH", although the equilibrium constant for this reaction is very small. The presence of SH" does not significantly degrade cellulose, which is the key advantage of the Kraft pulping  14  process over basic alkaline pulping (Adams et al. 1989). Separation of the depolymerised lignin solution from the fibre leaves primarily cellulose and some residual lignin. Kraft pulping is usually carried out at or below 170°C, which is the maximum temperature desirable, because elevating temperatures beyond this point resulting in alkaline hydrolysis of cellulose. The p-aryl ether lignin substructure is not the only functional group susceptible to nucleophilic attack. The phenylcoumaran substructure (Figure 1.2), with its P-5 and a-ether can undergo the formation of a quinone methide under Kraft pulping conditions. Elimination or addition reactions may also take place, but because of the P-5 carbon-carbon bond the lignin substructure will not actually be fragmented.  Another structure that is reactive under these  conditions is the l,2-diarylpropane-l,3-diol which can undergo elimination via the a-hydroxy group (Gellerstedt 2001). Preferentially, these structures undergo elimination to form stilbene units, as shown in the Figure 1.6.  OH  Ri, R  OH 2  = H orOMe  Figure 1.6. The formation of a stilbene unit from a p-1 lignin substructure under alkaline conditions (Gellerstedt 2001).  Nonphenolic and phenolic biphenyl (5-5') and biphenyl ether (4-0-5) linkages do not react under Kraft pulping conditions. Also, hydrosulfide ions may attack aromatic methoxyl groups forming methyl mercaptan which acts as a nucleophile on other methoxyl groups forming dimethyl sulphide (Gellerstedt 2001). As Kraft pulping proceeds, delignification slows due to 15  the formation of condensation reaction products that result in new carbon-carbon bonds that are chemically resistant the pulping chemicals (Gellerstedt 2001). Due to the conjugated nature of lignin, chromophores are produced during the degradation process, and therefore pulp needs be bleached in order to produce lighter and sometimes stronger paper.  Furthermore, the residual lignin remaining after pulping has a  negative influence on paper properties, such as inter-fibre bonding (Clark 1978). Over time, even bleached residual lignin in paper yellows, which is the driving force to remove as much of lignin as possible.  1.3.2 Pulp Conventions Pulping convention dictates that the amount of sodium hydroxide and sodium sulphide are given as gram sodium oxide (Na20) equivalents. Active alkali is the concentration of NaOH and Na2S in g Na20/L, effective alkali is NaOH plus half of Na2S in g Na20/L. Sulfidity may also be cited, as a ratio of Na2S to total alkali. Because cooking temperature and time are so important to delignification, pulp properties can only be compared between samples that have been cooked under the same conditions for equal amounts of time. A sum of the energy used to produce pulp is called the H-factor, and it is found by calculating the area under the curve of the relative reaction rate as a function of time. The exponential function  43 e  - ( 2_  16113  ^W) j  s c a  i i t j c u  a  e c  over small time intervals to determine the total cooking time necessary to achieve a target Hfactor. In this equation, T is time in hours, and t is temperature in Kelvin. Another important pulping parameter is the liquor to wood ratio, which depends on the density and permeability of the substrate. Ideally, the pulping liquor fully covers the packed wood chips.  16  1.3.3 Factors Affecting the Rate of Delignification There are a variety of factors that influence the rate of delignification during pulping. These factors are both structural and chemical, including wood density, permeability, heat of wetting, lignin content, lignin structure, and lignin-carbohydrate interactions. Some or all of these traits may be involved in determining how to maximize yield from a given fibre supply.  1.3.4 Structural Effects In order for wood lignin to be effectively solubilized during chemical pulping it must maintain constant contact with the pulp liquor. The contact between wood and liquor occurs when cooking chemicals enter the wood via vessels and tracheids, allowing the pulping chemicals to diffuse and make physical contact with reaction sites. The ease of penetration of liquor may be impacted by the moisture content of the wood. For example, if water is occupying free space in the wood the pulping liquor cannot diffuse easily. One of the factors that will ensure optimal contact between pulping liquor and cell wall lignin is the size of the wood chip. Chips that are too thick generally will have uncooked centers (Adams et al. 1989) because the liquor does not fully penetrate the wood chips. Another factor that influences penetration of pulp liquor is the change in pH as pulping progresses - as the alkaline pulp liquor penetrates the wood chips, hydroxide ions are protonated, which consequently reduces the pH. If the pH decreases to below 12.5, the diffusion rate decreases and the rate of delignification is consequently decreased (Adams et al. 1989). Species that are of lower inherent wood density possess more "void volume" or empty space within the solid wood matrix, which decreases the amount of cellulose available to make pulp for a given volume of wood. In solid wood, density affects wood strength, shrinkage, swelling, and stiffness. Extractives content also has an impact on density, as heartwood is often  17  more dense than sapwood, and this as well as the extractives chemistry causes the two regions of wood to pulp differently. Wood density is .variable even among trees within the same species, which is under the direct influence of location of wood in the tree, site, and genetics. Choosing dense wood for pulping allows for more wood to be packed into a digester, and therefore a greater yield is obtained from each chemical cook in the digester.  Although high-density wood can lead to  higher yields, other factors should be taken into consideration as there may be negative impacts on pulp quality if the fibres are too coarse to make desirable paper (Clark 1978). At the beginning of the growing season, earlywood is formed with thinner cell walls than wood formed later in the growing season (latewood). The ratio of the new wood formed in the spring versus summer has an impact on overall wood density and fibre quality. Thinner walled cells are more easily compressed into a paper sheet, and therefore form stronger intermolecular bonds between adjoining fibres. This produces paper that has high tensile and burst strength. In contrast, summer wood produces cells with thicker walls, which decreases tensile and burst strength (Adams et al. 1989). If the pulp liquor cannot penetrate the woody substrate, the chemical reactions between the pulping chemicals and lignin are significantly limited. To ensure that the pulp liquor comes into contact with lignin, one must ensure that the chips are thin enough to allow full penetration. Hardwood penetration is achieved by vessels. In softwoods, tracheids are the main avenue of liquor transport, and rays enable transverse diffusion. Swelling occurs when Kraft liquor increases the reactive surface area, thus further enhancing delignification (Marton 1971). During the swelling stage, the overall wood porosity enlarges and consequently increases the size of the lignin molecules that can be liberated from a given pore, facilitating more efficient delignification. In addition, when the fibres begin to separate the lignin products may freely leave the cell wall (Marton 1971). 18  1.3.5 Effects of Lignin Composition Since lignin degradation reactions are fundamental to Kraft pulping, lignin content and structure also contribute to the overall efficiency of the process. One of the most basic and well documented relationships is the correlation between the relative amount of syringyl and guaiacyl monomer units in lignin, known as the S:G ratio. These monomers are based on coniferyl and sinapyl alcohol (Figure 1.7). It is clear that in coniferyl alcohol only the C3 is methoxylated, while in sinapyl alcohol both the C3 and C5 are methoxylated. As such, the C5 is free to form linkages with other lignin monomers to form guaiacyl lignin, while sinapyl units are restricted to linkages formed via only the C4 position on the benzene ring and the propane side chain.  H CO  O C H q  O C H o  3  O H  O H  coniferyl alcohol  sinapyl alcohol  Figure 1.7. Precursors to lignin monomers. Coniferyl alcohol gives rise to the guaiacyl structures, and sinapyl alcohol gives rise to the syringyl monomers.  An abundance of syringyl units is considered favourable in pulp substrates because of the reduced proportion of 5-5' carbon-carbon bonds formed between two guaiacyl monomers (Chang and Sarkanen 1973). The percentage of syringyl units in hardwoods can vary from 20% 19  to 70% of total lignin (Sarkanan et al. 1967). As the S:G ratio is under the control of lignin biosynthetic genes, it is thought that specific genotypes may have distinct limits for syringyl and total lignin content. It is possible to alter S:G ratio to produce wood that would consequently be more amenable to pulping through biotechnological means. Recently, poplar transformed with the C4H-F5H construct, which increases the concentration of syringyl monomer units in the lignin, was clearly shown to be pulped more easily than the wild-type, improving pulp efficiency by reducing the chemical and energetic load necessary for chemical pulping (Huntley et al. 2003). Both syringyl and guaiacyl units may participate in P-O-4 type linkages, and the kinetics of the P-O-4 ether cleavage has been studied of the various combinations of subunits (Tsutsumi et al. 1995).  For each of three model compounds, guaiacylglycerol-beta-guaiacyl ether,  syringylglycerol-beta-guaiacyl ether, syringylglycerol-beta-(methylsyringyl)ether, the yield of P-O-4 cleavage was measured to determine differences in reactivity in Kraft liquor. The cleavage of the syringyl type unit was higher yielding than that of the guaiacyl type unit. This indicated that the rate of cleavage is influenced by the structure of the ether-linked ring (Tsutsumi etal. 1995). Poplar with a 90% reduction in caffeic acid/5-hydroxyferulic acid O-methyltransferase (COMT) activity had the same total lignin content as wild-type and an decreased S:G ratio. During lignin biosynthesis, COMT methylates the hydroxyl group on the 5' position of either caffeic acid or 5-hydroxyferulic acid. Upon further analysis of COMT down-regulated poplar many chemical differences were observed in the lignin when compared to wild-type. An increase in the proportion of chemically resistant biphenyl linkages, a decrease in the frequency of P-O-4 linkages, as well as a decrease in phenolic hydroxyl groups was found in the modified lines. As a result of these lignin modifications, the COMT suppressed trees displayed higher residual lignin contents (Kappa) when subject to Kraft pulping. This resulted in lignin that was 20  more similar to softwood than hardwood. The resulting Kraft pulp results indicate that these lignin modifications had a negative impact on Kraft pulping (Lapierre et al. 1999).  1.4  Selection of Trees for Improved Pulping Performance Trees that produce relatively high yields of cellulose, possibly due to lower lignin  content, are also favourable for pulping. These trees require less harsh cooking conditions because their lignin takes less time to dissolve in the cooking liquor. In addition, these trees produce pulp with lower residual lignin after the Kraft pulping process. The combination of reduced chemical concentrations and decreased time and temperature required to generate pulp leads to a significant increase in mill profits.  In addition, lower amounts of residual lignin  reduce the cost of bleaching chemicals downstream (Adams et al. 1989). Substrate selection from among existing variation may enable higher productivity for the pulp and paper industry. A natural mutant with promising pulping results, such as the loblolly pine mutant "CAD null", is deficient in the cinnamyl alcohol dehydrogenase (CAD) enzyme. The enzymatic function of CAD is to reduce />-hydroxycinnamaldehydes to /?-hydroxycinnamyl alcohols, the monomeric lignin precursors.  Four-and six-year old pine trees displayed  differences in pulping efficacy, whereas older mutant pines did not show the same results. Six year old mutant pine samples had an average Kappa number 7 units lower than normal pine samples (Dimmel et al. 2002). These results show that natural mutants such as the CAD null are promising target pulp substrates. The occurrence of natural variation, such as the CAD activity, has significantly increased the interest of target genes suitable for directed gene manipulation, as well as the search for other natural mutants. In addition to natural mutants in CAD activity, anti-sense C A D tobacco plants confirmed the finding of the natural CAD deficient lines, and demonstrated woody tissues that was more easily pulped (O'Connell et al. 2002). 21  When considering trees for selection, whether they are natural or genetically modified, desired traits for selection should be present at the age of harvest. If trees grown in the field maintain phenotypic changes observed in the laboratory, the modification of lignin biosynthetic genes can be beneficial. A recent study by Pilate et al. (2002) showed that transgenic poplar trees responded well to field growing conditions and retained activity of modification (Pilate et al. 2002).  This is evidence that modifications made at an early developmental stage could be  retained through other stages of development. Genetic modifications of the lignin biosynthetic pathway have been extensively researched. There are two main reasons for altering the enzyme activity of trees or model species.  One is to elucidate the enzymatic mechanisms of the biosynthetic pathway. By  systematically up-regulating and/or down-regulating genes that encode for catalytic enzymes one may analyze the intermediate molecules for their location and abundance.  From this, the  function of the enzyme and its importance to the pathway can more conclusively be deduced. The second reason is the alteration of phenotypic properties of wood for specific purposes, ranging from increased pulp yield to strength of secondary wood processed products. For example, increasing syringyl content genetically has been shown to improve pulping efficacy (Huntley et al. 2003). When justifying the uses of genetically modified organisms (GM) for application to industrial processes, one must consider the plasticity of natural trees.  It seems that lignin  changes do not have a drastic effect on plant development, and natural mutants may possess more useful structural changes than G M ones (Boudet 1998; Sederoff et al. 1999).  1.5  Lignin Structural Analysis Since Kraft pulping success relies heavily on lignin structure, a detailed analysis of lignin  may provide a rationalization as to why yields and quality of pulp vary among samples. 22  Sometimes, lignin content is not correlated directly to pulp yield and lignin content, and other structural characteristics are of importance, such as S:G ratio (Huntley et al. 2003). Wood is complex in structure, and has been shown to be inherently difficult to study its components in the native form.  Thus, lignin is often extracted in order to characterize its  structure. The Bjorkman method is one way of extracting lignin from finely divided wood with a neutral solvent (Bjorkman 1956). Ground extractive free wood is milled into a fine powder, and is extracted with a 96:4 dioxane:water mixture, after which it may or may or may not be further purified. No matter what the method of extraction, lignin is changed from its native form, but in spite of this, the Bjorkman method produces lignin believed to be the closest to protolignin. The differences between Bjorkman lignin and natural lignin include increased content of highly oxidized carbonyl functional groups, phenolic hydroxyl groups, and lower molecular weight. Lai and Sarkanan (1971) proposed that ball milling breaks lignin and carbohydrate matrices into smaller components, as opposed to simply chipping the carbohydrate sheath away from a lignin core. Therefore, extracted lignin includes components of lignin-carbohydrate complexes as well as pure lignin (Lai and Sarkanen 1971).  Dioxane extraction in combination with enzyme  degradation have been used to circumvent this problem and obtain higher yields of lignin (Holtman and Kadla 2004). It has also been suggested that during ball-milling covalent bond breakage occurs homolytically, and both carbon-carbon and carbon-oxygen bonds are broken. This observation is partly due to the fact that radicals exist in MWL, which are thought not to form new bonds during the milling process (Lai and Sarkanen 1971). It has also been suggested that milled wood lignin is more representative of the secondary wall of wood cells, and may not be representative of the total lignin in wood (Whiting and Goring 1981). Ikeda et al. (2002) performed a study in which different fractions of lignin were milled and analyzed to determine structural variations. It was found that rotary ball milling had a 23  minimal effect on lignin structure, whereas rotary ball milling followed by vibratory milling decreased degradation yield possibly due to condensation. However, in the presence of toluene, yield and linkages were similar to those of wood (Ikeda et al. 2002).  1.6  Wet Chemistry Methods for Lignin Studies  1.6.1 Degradation Methods Since lignin isolation is complex and various isolation procedures alter its structure, many partial structure characterization methods have been developed.  Employing several  methods allows a more complete structural analysis of lignin. Identifying the frequency and type of monomers that constitute the macromolecule is of use to the pulp and paper industry. Thioacidolysis has been used since the mid-1980s as a method of depolymerising lignin into its syringyl and guaiacyl monomer constituents linked via p-O-4 bonds (Lapierre et al. 1984). Ground extracted wood, plant material, isolated lignin, and even pulp can be analyzed by this procedure, whereby the hard Lewis acid BF3 and the soft nucleophile HS" selectively cleaving the P-O-4 inter-unit bonds of lignin (Figure 1.8). Since only the P-O-4 bonds are cleaved, any monolignols attached through carbon-carbon bonds remain intact and are not quantified in the traditional thioacidolysis procedure.  24  H COH  H COH HCOR  1  HCOR , |0 R HCO.G BF  HCOR  1  1  2  Et Q-BF 2  R R R,R R 1  2  3  5  4  4  R  1I  EtSCH  2  = Ar = H, Ar = H, OMe = H, Alk  5  ee  Et 0-BF 2  V  3  OR  OR  1  V  3  R  5  R  3  OR  4  ^ R * 4  1* EtSCH  3  EtSCH  2  EtSCH  3  3  EtSH -R OH •R R  3  BF  EtSCH  2  R  5  1  I  2  e e  OR  2  2  X  HCOR  H COH . I . R  H COH  2  2  H COH  H C0H 2  2  e CH-s  EtSCH EtSCH  2  OH  x  EtS' I , HSEt CH-^  H co 3  Figure 1.8. Proposed mechanism of p-ether cleavage from thioacidolysis (Rolando et al. 1992).  In addition to thioacidolysis, methods such as nitrobenzene oxidation (NBO) and cupric oxide oxidation also depolymerise lignin for monomer analysis (Chen 1992). Nitrobenzene oxidation, which dates back to 1939 (Freudenberg 1939), reacts ground wood with nitrobenzene and sodium hydroxide at elevated temperatures to produce mainly vanillin and syringaldehyde. Minor amounts of guaiacol, vanillic acid, />-hydroxybenzaldehyde, and 5-carboxyvanillin may also be produced (Chen 1992). Employing model compounds, it has been determined that the oxidation reaction cleaves 4-hydroxypropane units and their ethers.  More specifically, 4-0-  alkylated and a-0-4 and P-O-4 linkages are broken during the reaction. Aryl-O-4-linked lignin units do not undergo reaction during nitrobenzene oxidation.  Most of the products of  nitrobenzene oxidation are aldehydes and carboxylic acids, and the three carbon side chain of lignin are cleaved to leave one or two carbons (Chen 1992). 25  The NBO mechanism is not as well understood as that for thioacidolysis, but may involve a one or two electron transfer in the oxidation (Chang and Allen 1971). A disadvantage of NBO compared to thioacidolysis is that there are more products formed, making the results more difficult to interpret, and requires more extensive analysis (Chen 1992).  However,  thioacidolysis is a relatively simple procedure with a higher level of precision. Due to selectivity for different lignin linkages, neither method should be used as the exclusive analysis of lignin. Generally, NBO results in higher degradation yields than thioacidolysis, and degrades diarylpropane units (Chang and Allen 1971) in addition to alky aryl ethers (Rolando et al. 1992). If thioacidolysis yields are low, it may mean that there is a lower portion of P-O-4 ether linkages in any given sample.  1.6.2 Methoxyl Analysis The methoxyl content of wood or lignin is often indicative of the syringyl content. Interference may occur in wood because methoxyl groups may exist in the hemicellulose and pectin.  An estimation of methoxyl content may provide insights not made possible by the  degradation methods, because a different portion of the lignin may react with methoxyl analysis reagents. Methoxyl composition is usually included in an elemental analysis formula calculated for each 9-carbon lignin subunit. Relative amounts of hydrogen, oxygen, and methoxyl content may be easily compared in a C9 formula. However, general C9 formulas do not contain information on lignin linkages and specific functional groups. For example, methoxyl content was determined in MWL and CEL extracted from loblolly pine as one of the chemical determinations used to compare these two isolated lignin fractions (Holtman and Kadla 2004). The modified Zeisel method has also been performed on eucalyptus black liquor and pulp residual lignin during the Kraft process. Using methoxyl as a measure of syringyl units, the 26  authors suggest that H (p-hydroxyphenyl) and G units are degraded early on in Kraft pulping, and hemicellulose and cellulose have more associations to lignin high in H and G (Pinto et al. 2002).  1.7  Spectroscopic Methods  1.7.1 Nuclear Magnetic Resonance Spectroscopy Nuclear Magnetic Resonance Spectroscopy (NMR) exploits the magnetic properties of the nucleus in order to provide insight into the molecular structure of wood. A magnetic field is applied to a sample, and the magnetic moment of each nuclei precesses around the magnetic field vector.  When a second magnetic field is applied with the same frequency as this  precessional motion, the energy is absorbed by the nucleus.  This absorption of energy is  resonance, and the resonance frequency depends on the chemical environment of the nucleus. NMR active interactions include through bond coupling (J coupling), through space interactions (nOe), and chemical exchanges. For lignin, the interactions are secondary to full structural determination. Due to the complexity of the lignin macromolecule, types of signals (chemical shifts) are the main focus of decoupled carbon spectra. It is a common myth that carbon NMR spectra cannot be integrated to give meaningful results. However, by carefully selecting experimental parameters, quantitative results can be obtained for lignin and other polymers (Adriaensens et al. 2003; Holtman et al. 2003; Capanema et al. 2004; Holtman and Kadla 2004; Rego et al. 2004). There are three important qualities to these quantitative results: the first is the absence of differential saturation effects, the second is lineshape characterization, and the third is the avoidance of nOe signal enhancements (Claridge 1999). In practice, during quantitative NMR the relaxation period must be five times longer than the longest Tf (longitudinal) relaxation time. The user also benefits from using 90° pulse widths 27  to ensure the maximum signal possible. Instead of waiting for 5 x Ti to ensure that all nuclei are fully relaxed when obtaining quantitative results for lignin, a relaxation agent may be used. The paramagnetic nature of chromium (III) acetylacetonate shortens the Ti relaxation times (often 1.7 seconds is adequate for lignin). Information obtained from quantitative C spectra is mostly 1 3  contained in the chemical shift (8) of the carbon nuclei. Carbonyl carbons, aromatic, or olefinic carbons, and aliphatic carbons are in three separate regions along the chemical shift axis (200165 ppm, 165-100 ppm, and 100-10 ppm, respectively).  1 3  C NMR spectroscopy is inherently  less sensitive than its proton counterpart, due mainly to the meagre 1.1% abundance of the C 1 3  12  nuclei. The  C nucleus has a spin quantum number of zero, and therefore does not have a  magnetic moment, a requirement for a NMR active nucleus. In addition to understanding how nuclei are determined to be NMR active, two nuclei relaxation phenomenon are of utmost importance. The first, as mentioned above, is longitudinal relaxation (also known as spin-lattice relaxation), occurs when random motion of nearby nuclei gives rise to changes in the magnetic field around the nuclei. When energy is transferred to the nucleus, the magnetic vector is restored along the z axis. The second, transverse relaxation (also known as spin-spin relaxation), is responsible for the x and y elimination of the magnetic vector. It arises from dipole-dipole interactions between spins, and the subsequent de-phasing of the vector of x and y components, which eventually become completely random in the transverse orientation (Claridge 1999). NMR has been applied to lignaceous material since the late 1970s (Robert 1992). Approximately 40 signals could be distinguished in the early proton decoupled carbon spectra of lignin. Chemical shifts of lignin parameters represent different categories (Table 1.1).  28  Table 1.1. NMR chemical shift ranges for common 1992). Structure or Functional Group Aromatic C total Aromatic-CH Aliphatic-COR Methoxyl C3/C5 in p-O-4 etherified syringyl CyH OH C2/C6 in syringyl Cp in p-p and p-5 Acetyl carboxyl carbons Primary hydroxyl groups Secondary hydroxyl groups Phenolic groups 2  functional groups present in lignin (Robert Frequency range (ppm) 102-162 102-127 58-95 54.5-56.5 150.8-153.3 58.4-64 102.3-107.5 53.4-54 170.4-171.6 169.5-170.4 168.5-169.5  Proton NMR spectroscopy was used long before carbon NMR to determine functional groups in lignin (Lundquist 1992). Due to its small frequency range and high sensitivity, this is also a useful tool for structure determination. As with carbon NMR, model compounds are the basis of lignin peak assignments. In addition, two-dimensional NMR experiments are becoming routine in lignin analysis, and new linkages have been discovered through correlation experiments such as the spiro-dienone (Zhang and Gellersted 2001) and dibenzodioxocin (Karhunen et al. 1995). The syringyl to guaiacyl nature of lignin is determined by long-range C - ' H correlation 13  experiments such as the heteronuclear multiple bond coherence (HMBC).  The correlations  between alpha protons and beta and gamma carbons on the lignin side chains are between carbons 1, 2, and 6 of the aromatic ring. The C2/C6 carbons in the syringyl correlation resonate at 105 ppm, and the guaiacyl C2/C6 carbons resonate at 113 and 120 ppm, which is sufficient for separation. Although most HMBC experiments are not quantified, the absence of a correlation is due to either syringyl or guaiacyl units. Marita et al. (1999) studied the HMBC spectra of acetylated Arabidopsis lignin from wild-type, and the fahl-2 mutant, 35S-F5H, and C4H-F5H genotypes. The cauliflower mosaic 29  virus 35S promoter coupled to the ferulate 5-hydroxylase (F5H) gene produced lignin similar to wild-type, however, using the cinnamate 4-hydroxylase (C4H) promoter instead of 35S resulted in significantly increased the syringyl content of the lignin. The HMBC spectrum of the C4HF5H did not show the guaiacyl units (Marita et al. 1999). The HMBC technique has also been applied to investigate aspen lignin S:G in the same manner as Arabidopsis. Like the Arabidopsis transgenic aspen upregulated in CAld5H (coniferyl aldehyde 5-hydroxylase, same enzyme as F5H) showed a decrease in guaiacyl lignin monomers (Li et al. 2003). Lignin linkages have been extensively examined in the literature through short-range C 1 3  'H correlation experiments. There are many types of short-range experiments, and heteronuclear single-quantum coherence (HSQC), heteronuclear multiple-quantum coherence (HMQC), and HSQC-TOCSY have also been shown to work well with isolated lignin. The TOCSY sequence (Total Correlation Spectroscopy) is able to correlate protons in the same spin system by relaying magnetization along the protons in a chain, resulting in a decrease in resonance overlap. TOCSY can also increase sensitivity, because antiphase peaks are not cancelled out. A TOCSY sequence following a HSQC or HMQC sequence results in the transfer of magnetization from a proton to its neighbouring J-coupled protons. The relayed proton magnetization again serves to disperse the proton signals (Claridge 1999). An even greater advantage is using HSQC-TOCSY and regular HSQC together, is that side chain structures are more easily elucidated (Marita, 1999). For example, one would expect a decreased frequency of certain lignin monomer linkages in the C4H-F5H modified lignin. In Arabidopsis this can be seen as a decrease in the number of p-5 and dibenzodioxocin units, and a concurrent increase in the number B-P linkages as compared to wild-type lignin. Furthermore, NMR can often be used to determine unusual components, such as a syringyl oc-keto-P-aryl ether, which can be formed during lignin ball milling due to oxidation (Marita et al. 1999). 30  The syringyl and guaiacyl nature of lignin can also be determined through onedimensional  1 3  C NMR experiments. As in correlation spectroscopy, syringyl C2/C6 carbons  have a chemical shift of 105 ppm, while C3/C5 resonate at 154 ppm. In guaiacyl units, carbons 2, 5, and 6 are between 110 and 125 ppm, while carbons 3 and 4 range from 145-154 ppm. Therefore, it is easy to determine which resonances come from which type of monomer unit (Marita et al. 1999).  1.8  Overview and Objectives Since wood and fibre quality impact how we use wood, by studying the physical,  morphological, and chemical properties of the natural wood we can gain important knowledge that may be used in both tree selection for industrial use and in reforestation efforts. The focus of this thesis was to investigate the chemical structure of lignin and attempt to relate differences in this complex biomolecule with Kraft pulping efficiency. Genetically modified trees can be used as models for how lignin differs depending on the genotype of wood. In this project, C4H-F5H up-regulated hybrid poplar was studied via wet chemistry and spectroscopic methods. In previous studies it was found that modified poplar had a dramatic increase in syringyl unit content (93% up from 66% in wild-type), which reduced the H-factor required to pulp these trees to a target Kappa number. An in-depth analysis of the structural differences of these genetically modified lignin lines will be presented in this thesis. As a direct extension of the hybrid poplar work, we surveyed the natural clonal variation among six natural aspen clones selected from Northern British Columbia for their efficiency in Kraft pulping.  In a previous study, differences in physical properties (specific gravity,  permeability, heat of wetting) were not able to be correlated with pulp efficiency (Avramidis and Mansfield 2004), so it was hypothesized that the differences in wood chemistry, in particular, lignin structure may directly influence the rates and efficacy of pulping. In order to determine if 31  the structure of the biopolymers of wood and their interactions with each other result in pulping differences, both wet chemistry and spectroscopic techniques were employed. This thorough characterization of lignin functional groups and inter-unit linkages was used to suggest a mechanism for improved rates or ease of chemical pulping.  32  1.9  Bibliography  Adams, T., B. Cowan, D. Clayton, D. Easty, D. Einspahr, D. E. Fletcher, T. M . Malcolm, T. McDonough, L. Nordman, J. K. Perkins, D. W. Reeve, L . Schroeder, G. A. Smook, N. Thompson, S. V. Fleet and B. Wikdahl, Eds. (1989). Alkaline Pulping. Pulp and Paper Manufacture. Atlanta, The Joint Textbook Committee of the Paper Industry. Adriaensens, P. J., F. G. Karssenberg, J. M. Gelan and V. B. F. Mathot (2003). "Improved quantitative solution state C NMR analysis of ethylene-1-octene copolymers." Polymer 44(12): 3483-3489. 1 3  Avramidis, S. and S. D. Mansfield (2004). "On Some Physical Properties of Six Aspen Clones." Holzforschung (in press). Balatinecz, J. J., D. E. Kretschmann and A. Leclercq (2001). "Achievements in the utilization of poplar wood - guideposts for the future." The Forestry Chronicle 77(2): 265-269. Bjorkman, A. (1956). "Studies on Finely Divided Wood Part 1. Extraction of Lignin with Neutral Solvents." Svensk Papperstidning 13: 477-485. Boudet, A. M . (1998). "A new view of lignification." Trends in Plant Science 3(2): 67-71. Campinhos, E. (1999). "Sustainable plantations of high-yield Eucalyptus trees for production of fiber: the Aracruz case." New Forests 17: 129-143. Capanema, E. A., M . Y. Balakshin and J. F. Kadla (2004). "A comprehensive approach for quantitative lignin characterization by NMR spectroscopy." Journal of Agricultural and Food Chemistry 52(7): 1850-1860. Cates, J. (1998). Hybrid Poplar for BC's Future Pulp. MillWatch Special Report: Hybrid Poplar. CCFM  (2001). Forest 2020, Canadian Council http://www.ccfm.org/forest2020/plantationscanada_e.html.  of  Forest  Ministers.  Chang, H.-m. and G. G. Allen (1971). Oxidation. Lignins. Occurence, formation, structure, and reactions. K. V. Sarkanan and C. H. Ludwig. New York, Wiley-Interscience: 433-485. Chang, H.-m. and K. V. Sarkanen (1973). "Species Variation in Lignin: Effects of Species on the Rate of Kraft Delignification." Tappi Journal 56(3): 132-134. Chen, C.-L. (1992). Nitrobenzene and Cupric Oxide Oxidations. Methods in Lignin Chemistry. S. D. Lin, C.W. Berlin, Springer-Verlag. Claridge, T. D. W. (1999). High-Resolution NMR Techniques in Organic Chemistry. Oxford, Pergamon. Clark, J. d. A. (1978). Pulp Technology and Treatment for Paper. San Francisco, Miller Freeman Publications, Inc.  33  Dimmel, D. R., J. J. MacKay, C. E. Courchene, J. F. Kadla, J. T. Scott, D. M. O'Malley and S. E. McKeand (2002). "Pulping and bleaching of partially CAD-deficient wood." Journal of Wood Chemistry and Technology 22(4): 235-248. FAO (2002). FAOSTAT. http://faostat.fao.org/faostat/default.jsp. Fengel, D. and G. Wegener (1989). Wood: chemistry, ultrastructure, reactions. Berlin, Walter de Gruyter. Freudenberg, K. (1939). "Uber Lignin." Angew Chem 52: 362-363. Fujita, M . and H. Harada (1991). Ultrastructure and formation of wood cell wall. Wood and Cellulosic Chemistry. D. N.-S. Hong and N. Shiraishi. New York, Marcel Dekker: 3-57. Fujita, M . and H. Harada (2001). Ultrastructure and formation of wood cell wall. Wood and Cellulosic Chemistry. New York, Marcel Dekker. Gellerstedt, G. (2001). Pulping Chemistry. Wood and Cellulosic Chemistry. D. H.-S. Hon and N. Shiraishi. New York, Marcel Dekker, Inc.: 914. Goring, D. A. I. and T. E. Timell (1962). "Molecular Weight of Native Cellulose." Tappi Journal 45(454-460). Heilman, P. E. (1999). "Planted forests: poplars." New Forests 17: 89-93. Heilman, P. E., R. F. Stettler, D. P. Hanley and R. W. Carkner (1995). High-yield Hybrid Poplar Plantations in the Pacific Northwest (revision). Pacific Northwest Regional Extension Bulletin PNW 356. Pullman, Washington, Washington State University Cooperative Extension. Holtman, K. and J. Kadla (2004). "Solution-state Nuclear Magnetic Resonance Study of the Similarities between Milled Wood Lignin and Cellulolytic Enzyme Lignin." Journal of Agricultural and Food Chemistry 52(4): 720-726. Holtman, K. M . , H.-M. Chang, H. Jameel and J. F. Kadla (2003). "Elucidation of Lignin Structure through Degradative Methods: Comparison of Modified DFRC and Thioacidolysis." Journal of Agricultural and Food Chemistry 51: 3535-3540. Huntley, S. K., D. Ellis, M . Gilbert, C. Chappie and S. D. Mansfield (2003). "Significant Increases in Pulping Efficiency in C4H-F5H-Transformed Poplars: Improved Chemical Savings and Reduced Environmental Toxins." Journal of Agricultural and Food Chemistry 51: 6178-6183. Ikeda, T., K. Holtman, J. Kadla, H.-m. Chang and H. Jameel (2002). "Studies on the Effect of Ball Milling on Lignin Structure Using a Modified DFRC Method." Journal of Agricultural and Food Chemistry 50: 129-135. International Poplar Commission. (1992). Synthesis of national reports on activities related to poplar and willow areas, production, consumption and the functioning of national poplar commissions. Note from the Secretariat. Zaragoza, Spain. FO:CIP:Misc./92/l. 34  Karhunen, P. T., P. Rummakko, J. Sipila, G. Brunow and I. Kilpelainen (1995). "Dibenzodioxocins - A novel type of linkage in softwood lignins." Tetrahedron Letters 36(1): 169-170. Lai, Y. Z. and K. V. Sarkanen (1971). Isolation and Structural Studies. Lignins: Occurence. Formation. Structure and Reactions. K. V. Sarkanen and C. H. Ludwig. New York, John Wiley and Sons, Inc. Lapierre, C , B. Monties and C. Rolando (1984). "Thioacidolyis of Poplar Lignins: Comparison with acidolysis." Journal of Wood Chemistry and Technology 5: 277-292. Lapierre, C , B. Pollet, M. Petit-Conil, G. Toval, J. Romero, G. Pilate, J. Leple, W. Boerjan, V. Ferret, V. De Nadai and L. Jouanin (1999). "Structural Alterations of Lignins in Transgenic Poplars with Depressed Cinnamyl Alcohol Dehydrogenase or Caffeic Acid O-Methyltransferase Activity Have an Opposite Impact on the Efficiency of Industrical Kraft Pulping." Plant Physiology 119: 153-163. Li, L., Y. Zhou, X. Cheng, J. Sun, J. Marita, J. Ralph and V. Chiang (2003). "Combinatorial modification of multiple lignin traits in trees through multigene cotransformation." Proceedings of the National Academy of Sciences of the United States of America 100(8): 4939-4944. Liang, H. Y., C. M. Catranis, C. A. Maynard and W. A. Powell (2002). "Enhanced resistance to the poplar fungal pathogen, Septoria musiva, in hybrid poplar clones transformed with genes encoding antimicrobial peptides." Biotechnology Letters 24(5): 383-389. Lundquist, K. (1992). Proton (*H) NMR Spectroscopy. Methods in Lignin Chemistry. S. Y. Lin and C. W. Dence. New York, Springer-Verlag: 578. Marita, J. M., J. Ralph, R. D. Hatfield and C. Chappie (1999). "NMR characterization of lignins in Arabidopsis altered in the activity of ferulate 5-hydroxylase." Proceedings of the National Academy of Sciences of the United States of America 96(22): 12328-12332. Marton, J. (1971). Reactions in Alkaline Pulping. Lignins: Occurrence. Formation, Structure, and Reactions. C. H. Ludwig. New York, John Wiley and Sons. Minor, J. L. (1986). "Chemical linkage of polysaccharides to residual lignin in loblolly pine Kraft pulps." J. Wood. Chem. Technol. 6(2): 185-201. O'Connell, A., K. Holt, J. Piquemal, J. Grima-Pettenati, A. Boudet, B. Pollet, C. Lapierre, M . Petit-Conil, W. Schuch and C. Halpin (2002). "Improved paper pulp from plants with suppressed cinnamoyl-CoA reductase or cinnamyl alcohol dehydrogenase." Transgenic Research 11(5): 495-503. Overend, R. P. and K. G. Johnson (1991). "Lignin-Carbohydrate Complexes from Poplar Wood Isolation and Enzymatic Degradation." Acs Symposium Series 460: 270-287. Panshin, A. J. and C. de Zeeuw (1980). Textbook of Wood Technology. New York, McGrawHill. 35  Pilate, G., E. Guiney, K. Holt, M. Petit-Coriil, C. Lapierre, J. C. Leple, B. Pollet, I. Mila, E. A. Webster, H. G. Marstorp, D. W. Hopkins, L. Jouanin, W. Boerjan, W. Schuch, D. Cornu and C. Halpin (2002). "Field and pulping performances of transgenic trees with altered lignification." Nature Biotechnology 20(6): 607-612. Pinto, P. C , D. V. Evtuguin, C. P. Neto and A. J. D. Silvestre (2002). "Behavior of Eucalyptus Globulus Lignin During Kraft Pulping 1. Analysis by Chemical Degradation Methods." Journal of Wood Chemistry and Technology 22(2&3): 93-108. Rego, R., P. J. Adriaensens, R. A. Carleer and J. M. Gelan (2004). "Fully quantitative carbon-13 NMR characterization of resol phenol-formaldehyde prepolymer resins." Polymer 45(1): 33-38. Riemenschneider, D. E., B. J. Stanton, G. Vallee and P. Perinet (2001). Poplar Breeding Strategies. Part A, Chapter 2. Poplar Culture in North America. D. I. Dickmann, J. G. Isebrands, J. E. Eckenwalter and J. Richardson. Ottawa, NRC Research Press: 43-76. Robert, D. (1992). Carbon-13 Nuclear Magnetic Resonance Spectrometry. Methods in Lignin Chemistry. S. Y. Lin and C. W. Dence. New York, Springer-Verlag: 578. Rolando, C , B. Monties and C. Lapierre (1992). Thioacidolysis. Methods in Lignin Chemistry. S. D. Lin, C.W. Berlin, Springer-Verlag. Sarkanan, K. V., H. M . Chang and G. G. Allen (1967). "Species Variation in Lignins III. Hardwood Lignins." Tappi Journal 50: 587-590. Sederoff, R. R., J. J. MacKay, J. Ralph and R. D. Hatfield (1999). "Unexpected variation in lignin." Current Opinion in Plant Biology 2(2): 145-152. Singh, A. P. and R. Pandey (2002). "Natural resistance of Populus deltoides clones selection to defoliator Clostera cupreata (Lep., Notodontidae) in northern India: relative pupal weight as an easier criteria for accurate evaluation." Journal of Applied Entomology-Zeitschrift Fur Angewandte Entomologie 126(9): 475-480. Sjostrom, E. (1993). Wood Chemistry: Fundamentals and Applications. San Diego, Academic Press, Inc. Thomas, B. (2002/2003). Fibre Enhancement Research Tour Booklet, Alberta-Pacific Forest Industries, Inc. Tsutsumi, Y., R. Kondo, K. Sakai and H. Imamura (1995). "The Difference of Reactivity between Syringyl Lignin and Guaiacyl Lignin in Alkaline Systems." Holzforschung 49(5): 423-428. Whiting, P. and D. A. I. Goring (1981). "The morphological origin of milled wood lignin." Svensk Papperstidning 84(15): R120-R122. Zhang, L. and G. Gellersted (2001). "NMR observation of a new lignin structure, a spirodienone." Chemical Communications: 2744-2745.  36  Chapter 2 Isolation and Characterization of Lignin from Natural Aspen Clones  2.1  Introduction The need to conserve natural forest ecosystems while providing adequate wood and wood  products for the growing population may be reconciled with harvesting trees with targeted properties. For example, the amount of pulp produced worldwide is projected to increase to 240 million tons over the next 50 years, which represents an increase of 50%. Harvesting wood from intensively managed plantations may reduce the environmental impact of this increased pulp production. Currently, tree plantations comprise approximately one quarter of the global pulp resource, and it has been estimated that this resource will need to increase to approximately 40% by 2030 to meet future fibre requirements (Brown 2000). Generally, species planted to reforest areas of harvest are selected to maximize fibre production, however, equally important are traits such as disease resistance, fibre quality and wood density, as well as improved chemical composition and appearance (Adams et al. 1989). Structural characteristics such as density and microfibril angle impact the performance of solid wood in structural products (Downes et al. 2003), while chemical attributes including lignin and carbohydrate content, type of lignin, distribution of lignin, and extractives content very much influence pulp feedstocks. A company that has capitalized on being particularly selective and efficient at maximizing plantation forestry is Aracruz Celulose, a leading producer of bleached eucalyptus pulp for tissue and paper manufacture. Surrounded by 38,170 hectares of eucalyptus plantations, this technologically advanced company has selected clones that are disease and pest free, yield high wood volumes, are well-shaped and self-pruning, have thin branches, and manifest the ability to coppice and produce rooted cuttings (Campinhos 1999). Furthermore, a large genetic 37  base of clones has been employed for the establishment of a successful plantation in order to ensure adaptation to different environments.  In practice, Aracruz selects trees based on  competitive tests at different sites. A seven-year rotation was established in 1989, and by 1995 the plantation growth rate increased from 28 m /hectare per year to 45 m /hectare per year. 3  3  Perhaps more striking is that the amount of wood required to produce 1 ton of air-dried bleached pulp was reduced from 4.9 m per ton pulp to 4.1 m per ton pulp. The yearly productivity of the land thus grew from 5.9 tons per hectare to 10.9 tons per hectare.  These gains clearly  demonstrate that diligence in tree selection has both environmental and economic rewards (Campinhos 1999). The incredible efficiency gain resulting from proper selection of clones can clearly improve yield and the health of a plantation to ensure success and profitability. Hybrid poplar plantations are becoming increasingly common in China, South Korea, and the United States. The success of these plantations is due to hybrid poplar's fast growth rate, their ability to establish in a range of variable sites, and a wide variety of end uses. Poplars are also extremely easy to plant, and it has been suggested that under ideal conditions one person could plant 5,000 cuttings per day of short rotation wood (Heilman et al. 1995). In the continuing search for high-yield, high-quality pulp, attention has been given to improving pulp substrates with reduced chemical and energetic requirements.  In particular,  lignin structure and content have been shown to impact pulp yield and residual lignin, which affect pulp bleachability (Adams et al. 1989; Huntley et al. 2003). For example, it has been suggested that feedstocks with a higher syringyl to guaiacyl lignin monomer ratio (S:G) are more easily delignified during the pulping process than substrate with lower S:G (Chang and Sarkanen 1973). Furthermore, it is also known that pulp yield is proportional to the syringyl content of the substrate. In order to explain such findings, it has been proposed that lignin containing fewer methoxyl groups (guaiacyl units) may have more extensive carbon-carbon bonding between 38  phenylpropane units, and these linkages are more chemically resistant to pulping chemicals and energy than the majority of linkages between syringyl units (Chang and Sarkanen 1973). Other chemical factors that may account for altered rates of delignification, including the 3-dimensional structure of the lignin polymer, its interactions with cellulose and hemicellulose in lignin-carbohydrate complexes (LCC), and the degree of polymerization of the lignin. Therefore, syringyl to guaiacyl content may simply be one of the more obvious and readily obtained delignification parameters, with other inherent differences in the lignin contributing to pulping efficacy. Furthermore, physical factors may also affect pulping efficacy, including the ability of pulping liquor to penetrate and diffuse through the wood, the specific gravity of the wood, as well as heat of wetting. In this study, natural aspen clones were obtained from the Fort Nelson region of northern British Columbia, Canada in September 2001. The wood chemistry and lignin structure were analyzed in order to rationalize the significant difference in pulp yields and residual lignin content of the selected clones. It is evident that these clones, although of the same genus and species and from the same environmental site, have vastly differently pulping efficacies and fibre qualities.  39  2.2  Methods and Materials  Wood Sampling. The individual clones were distinguished from one another based on bark colour and markings, branch angle, leaf shape, and the time of bud flush. Age, size, and degree of decay were the bases for selection of individual trees within each clonal line. The entire butt log (lower 2 metres) was chipped and mixed thoroughly before analysis.  The clonal trees  employed in this study ranged in age from 95 to 124 years old.  Material Preparation. Extractive-free ground wood was prepared by grinding wood in a Wiley mill to pass through a 40 mesh sieve and extracting with acetone for 24 hours. The extract free wood was dried over P2O5 desiccant until analyzed.  Quantification of Extractives. Extractives were quantified for each sample via a modified TAPPI Method T 280 pm-99. Each vial was oven-dried at 105°C overnight and allowed to cool over desiccant. Total extractives mass was then determined gravimetrically. In order to classify the extractives into its main components (fatty acids, resin acids, sterols, steryl esters, waxes, triglycerides) response factors were determined for representative compounds from each of these categories.  Tetracosane (Aldrich), betulin, cholesteryl palmitate (Sigma), and palmitin were  used as standards for fatty acids and resin acids, sterols and waxes, steryl esters, and triglycerides, respectively. A Hewlett Packard (HP) 5890 Series II G C fit with a HP 6890 Series injector, FID detector, and a 10 m D B - X L B column (J&W Scientific) was used to perform the extractives analysis. The G C method employed a 1.0 /LtL injection volume, with an injector temperature of 320°C and a detector temperature of 330°C. The initial oven temperature was set at 50°C for 3 min., and then ramped at a rate of 10°C min" until 240°C, where it was held for 3 min. The 1  temperature was again increased by 10°C min" . to 310°C and held for 3 min., then increased 1  once more by 10°C min" to 350°C and held for 25 min. 1  40  Wood Composition Analysis. A modified Klason method was employed to determine total carbohydrate and lignin content. In brief, 0.2 g of extractive-free wood was treated with 3 mL of 72% H2SO4 for 2 hours at 20°C with mixing every 10 min. This mixture was then diluted with 112 mL of de-ionized water to achieve a final acid concentration of 4% H 2 S O 4 and transferred to a serum bottle.  The solution was autoclaved at 121°C  for 1 hour before filtering through a  medium coarseness sintered glass filter for the gravimetric determination of acid-insoluble lignin.  Acid-soluble lignin was quantified using absorption spectroscopy at 205 nm (Tappi  Useful Method UM-250). Carbohydrate concentration was determined by high-performance liquid chromatography (Dionex BioLC, Dionex, CA) equipped with an ion exchange PA1 (Dionex) column, a pulsed amperometric detector with a gold electrode, and a Dionex AS50 auto-injector (Dionex). (Millipore, Bedford, MA).  Prior to injection, samples were filtered through 0.45 um filters A volume of 20 uL was loaded on the column equilibrated with 250  mM NaOH and eluted with de-ionized water at a flow rate of 1.0 mL min" , followed by a post 1  column addition of 200 mM NaOH at a flow rate of 0.5 mL min" . Each sample was determined 1  in triplicate.  Ball Milled Wood. Extractive-free ground wood (-15 g) was ball-milled using a Uni Ball Mill II (Australian Scientific Instruments) set to 85% maximum rotation speed. The ball mill was fit with two stainless steel drums containing 5 stainless steel ball bearings. After one week, 400 mL toluene (Fisher) was added to minimize lignin oxidation reactions.  Following an additional  week of ball milling, centrifuging the sample at 4,000 rpm for 10 minutes and decanting the toluene removed the toluene. This process generated ball milled wood (BMW).  Isolation of Lignin from Ball Milled Wood. Lignin was extracted from the B M W (Figure 2.1) using a mixture of dioxane and water (96:4) according to the method of Bjorkman (Bjorkman 1956). Approximately 200 mL of dioxane solution was used to extract the lignin from 15 g of BMW.  After one day of constant stirring, the solvent was recovered by centrifugation, and the 41  dioxane:water replenished. The extraction was allowed to proceed for an additional 3 days, after which time lignin extraction was shown to be minimal, and the extractions pooled. The crude lignin resulting from the extraction of B M W was subject to the purification scheme via the Bjorkman method. This purification involved dissolving the liberated lignin in 90% of acetic acid and precipitating into de-ionized water. After washing with double distilled water twice, the sample was dried under vacuum.  The lignin was then dissolved in 1:2  ethanol:l,2-dichloroethane and precipitated it into diethylether. The purified lignin was washed twice with diethylether and dried. The resulting lignin was classified as milled wood lignin (MWL). The insoluble material left after dioxane extraction, consisting of lignin more closely associated with carbohydrates, was subject to a carbohydrate digestion by hydrolytic enzymes. Xylanase (Iogen, Ottawa, ON), Gammanase (Novozyme,  Franklinton, NC), Novo-188  (Novozyme), and Fibrilase H D L 160 (Iogen) were mixed together in equal amounts, and 1 mL of this enzyme cocktail was added to each sample suspension in a 50 m M acetate buffer solution (pH 4.5). The cocktail was supplemented with tetracycline (Sigma, 40 ug mL" ), cyclohexamide 1  (Sigma, 30 ug mL" ) and sodium azide (Fisher, 0.02%) to prevent microbial growth, and 1  incubated at 45°C and 115 rpm. After two days the solution was regenerated, the solublized carbohydrates removed, and the digestion repeated two more times.  The material was again  subjected to a dioxane:water extraction and purified in the same manner as the M W L . The term C E L applies to purified lignin from this fraction, and R E L refers to residual lignin that was remaining and found to be insoluble in the dioxane:water solution.  Molecular Weight Measurements.  M W L , C E L , and R E L samples were acetylated by  dissolving mg quantities in N-methylimidazole and dimethylsulfoxide in a ratio of 1:2 and adding acetic anhydride (0.6 x NMI). Samples were left for 3 hours, precipitated into 6 m M E D T A (pH 8), filtered through a 0.2 um membrane filter, and dried under vacuum. Samples of 1 42  mg mL" in THF were filtered through a 0.45 um filter. 1  A Dionex Summit HPLC fit with  Waters Styragel H5 column was employed to obtain the molecular weight at 50°C with a 0.5 mL min" flow rate. The total run time was 60 minutes. 1  Determination of Monolignol Composition by Degradation - Nitrobenzene Oxidation (NBO). Accurately weighed dry extractive free wood (200 mg) or lignin (50 mg) was weighted into a Pyrex test tube to which 7 mL of 2 M NaOH (diluted from 10 N NaOH, B D H , Inc.) and 0.4 mL nitrobenzene (Fisher Scientific) were added. The tube was then placed into a 170°C oil bath for 2.5 hours and shaken regularly. Placing the test tubes in an ice water bath for at least 5 minutes terminated the reaction. A n internal standard, 3-ethoxy-4-methoxy-benzaldehyde (200 uL of 14.09 mg mL" in 2 M NaOH) was then added, and the tubes vigorously shaken to mix the 1  internal standard with the contents. The contents were then transferred to a separatory funnel, to which chloroform (3 x 25 mL) was added and used to remove any nitrobenzene reduction products. The aqueous layer was acidified with concentrated HC1 to a pH of 2 and transferred to a liquid-liquid extractor for extracted with CHCI3 for 48 hours. After which the solvent was reduced to dryness through rotary evaporation. The sample was then re-dissolved in 5 mL CH2CI2 and made up to exactly 10 mL in a volumetric flask. GC analysis was performed after filtration through a 0.45  um filter.  The oven  temperature profile for G C analysis consisted of 1 minute at 60°C, a 10°C min" ramp from 60 to 1  250°C and a 10 minute hold at 255°C, ending at 260°C. The injector set to 200°C and the FID was set at 270°C.  One microlitre injections were separated on a 30 m (0.25 um ID) DB-5  column (J&W Scientific) using helium as a carrier gas at 1 mL min" . 1  Kraft Pulping Trials. Small-scale pulping was performed on wood samples at 25% sulfidity, 13% effective alkali, and at a 4.5:1 liquor to wood ratio. Each sample was pulped 4 times.  43  The reactors were placed into an oil bath and the temperature was ramped from room temperature (23°C) to 170°C over approximately 1 hour. The vessels were maintained at 170°C for set times to achieve the desired H-factor. Following the cook, the reactors were removed from the oil bath and placed in a large container of cool water to stop the pulping reaction. The cooked wood chips were washed with warm tap water over a metal screen, separated in a standard British disintegrator, and then filtered through a Buchner funnel lined with a plastic mesh liner. The pulp was washed until the filtrate was colourless, and then dried for least 2 days in a 50°C oven.  Determination of Pulp Residual Lignin.  Pulp micro Kappa numbers were determined  according to TAPPI Useful Method U M 246.  Elemental Analysis. Elemental analysis (C, H , N and O by difference) was performed on a PerkinElmer Series II CFTNS/O Analyzer 2400 (Boston, MA), as per manufacturer instruction.  Methoxyl analysis. Methoxyl content was determined by a modified Ziesel method (Browning 1967).  Thioacidolysis. The standard degradation procedure was performed on wood, M W L , C E L and R E L to determine differences in lignin fractions (Rolando et al. 1992).  Nuclear Magnetic Resonance Spectroscopy Experiments. Samples of acetylated cell wall material were prepared at a concentration of 150 mg mL" in CDCI3 (Cambridge Isotope 1  Laboratories). Ten minutes of sonnication was usually required to dissolve the sample, which was then filtered through a Kimwipe plugged disposable pipette into a 5 mm N M R tube. Heteronuclear multiple quantum coherence (HMQC) and heteronuclear multiple bond coherence (HMBC) spectra were recorded on a Bruker A V A N C E 400 M H z BBi-x probe spectrometer. A l l runs were performed at room temperature, approximately 293 K. M W L or A c - M W L was dissolved in d-DMSO (Cambridge Isotope Laboratories) at the approximate concentration of 60 mg in 0.25 mL. The sample was allowed to fully dissolve 44  overnight, stored over dessicant. The next day, the sample was filtered through a Kimwipe into a 5 mm Shigemi N M R microtube (with the magnetic susceptibility of DMSO, Shigemi Co.). H M Q C and H M B C spectra were obtained as above. For quantitative C N M R spectrum acquisition, 10 m M final concentration of chromium 1 3  (III) acetylacetonate, Cr(acac)3 (Aldrich), was added to shorten the T i relaxation times of the carbon nuclei to allow signals to be quantified. For M W L , a Bruker AM-400 spectrometer (dual 13  C / ' H probe) was used, and for the R E L , a Bruker AV-300 QNP probe was used. A 90° C 1 3  pulse, a 1.4 s acquisition time and a 1.7 s relaxation delay were the parameters for the inverse gated decoupling sequence (decoupler on during acquisition only).  The number of scans for  M W L (20,000) and R E L (60,000) were different due to different sample concentrations. The d 6 DMSO peak at 39.5 ppm or the CDCI3 peak at 77 ppm was used as the reference peak.  Hydroxyl Content. Aliphatic and phenolic hydroxyl content was determined via proton N M R spectroscopy. Para-nitrobenzaldehyde was used as the internal standard (stock solution of 1 mg mL" in CDCI3). A c - M W L was dissolved into this solution at a concentration of 30 mg mL" of 1  internal standard solution.  A proton N M R experiment was performed with 256 scans on a  Bruker AV-300 M H z spectrometer.  Integrated aliphatic and phenolic hydroxyl proton signals  were quantified by comparison to the benzaldehyde proton signal from the internal standard.  45  2.3  Results and Discussion  2.3.1 Wood Composition The overall wood composition (extractive, carbohydrate, and lignin content) was determined and found to vary among the six naturally occurring aspen clones. There is a large variation in carbohydrate and lignin content, ranging from 18.6 to 20.3% total lignin and 66.3 to 71.0% total carbohydrate (Figure 2.2). This large range in chemical properties is one example of a very important property that can be potentially selected for to improved processing. Carbohydrate composition was determined using the Klason method, which involved the hydrolysis of polymeric carbohydrates to their associated monomers, and HPLC was employed to quantify these sugars (Figure 2.3).  It is interesting to note that sample 16-2, which had the  highest total carbohydrate content, also show a correspondingly high level of glucose.  This  implies that the elevated level of carbohydrate is primarily due to cellulose and not hemicellulose, as the associated hemicellulose-derived monomers did not concurrently increase. Furthermore, a detailed analysis of lignin indicated that not only does sample 16-2 have much lower total lignin content than other clones, but that difference results mostly from the acidsoluble portion of lignin which was consistently lower than all other clones (Figure 2.4).  A  lower amount of acid-soluble lignin may suggest that its lignin is more reactive to acid hydrolysis. In contrast, the acid insoluble fraction is only slightly lower for this wood sample. The clones have a considerable spread in both soluble and insoluble lignin.  Soluble lignin  content ranged from 2.2% (sample 16-2) to 4.2% (sample 10-1) and insoluble lignin ranged from 16.0% (sample 19-5) to 18.0% (sample 21-5). The range of total lignin content was therefore, shown to be between 18.6% and 21.3% for sample 16-2 and 10-1, respectively. A thorough analysis of the wood extractives indicated that among the different clones of aspen there was little difference in the total amount of extractives (Figure 2.2). Further analyses of the major classes (resin and fatty acid, sterols, sterol esters and triglycerides) of the extractives 46  by gas chromatography also showed that there were only minor differences in the class composition, with an inverse relationship between content of resin and fatty acids and triglycerides (Figure 2.5). Some classes of extractives influence pulping (i.e. plicatic acid in cedar), while others can have a major effect on the environment (i.e. resin acid in soft pines on aquatic ecosystems).  Furthermore, pitch deposits, which can have a significant impact on  product quality, results mainly from resin acids and neutral acids containing terpenoids and fats (Chang and Kondo 1971).  2.3.2 Kraft Pulping The results of pulping analyses of the six aspen clones are shown in Figure 2.6. At an H factor of 1100, samples 19-5 and 16-2 had the highest yields (61.4 and 61.3%, respectively) while sample 10-1 was consistently more difficult to pulp and showed an average yield of 56.9%. Sample 10-1 also consistently had a higher concentration of shives in the resulting pulp samples. The pulp yield ranged from 56.9% to 61.6%, and a difference of 4.7 percentage points from natural aspen is a considerable difference that could lead to substantial economic and environmental benefits to any mill. The residual lignin content of the pulp, as determined by the Kappa number, was also lowest for the highest yielding pulp samples, 16-2 and 19-5, with Kappa numbers of 16.6 and 17.1, respectively, while among the clones evaluated, a range from 16.6 to 25 was observed (Figure 2.6). It has been shown that pulp with lower Kappa values is more easily bleached, with increases of up to 20 ISO brightness units in transgenic poplar from wild-type (Huntley et al. 2003). The difference in pulping efficacy found in these data may be partially due to the higher total lignin content of sample 10-1, however chemical factors such as lignin inter-unit linkages may also play a role. A detailed analysis of these linkages and their frequency may result in  47  additional explanations as to why these aspen samples behave differently when subjected to the same Kraft pulping conditions.  2.2.3 Lignin Monomer Ratio In order to elucidate possible reasons for differences in pulping efficacy, the S:G ratio was determined by thioacidolysis for ground wood, fractions of isolated lignin, and Kraft pulp (Table 2.1). The initial wood samples have very similar lignin monomer ratios as determined by thioacidolysis, whereas isolated lignin and pulp fractions display differences.  Nitrobenzene  oxidation was employed as a secondary means to determine S:G of wood and liberated more than the /3-0-4 linkages degraded by thioacidolysis. Nitrobenzene oxidation results show a larger variation in S:G ratio than thioacidolysis. Interestingly, the S:G ratio resulting from nitrobenzene oxidation corresponded very well pulping efficacy, while thioacidolysis did not follow a similar trend (see Table 2.2). Sample 101 had 74.4 mol % S content and was shown to be recalcitrant to chemical pulping, while sample 16-2 had 79.0 mol % S and proved to be the most effective sample for pulping.  Similarly,  sample 19-5 had a syringyl content of 80.3 mol % S and resulted in a similar yield to sample 162, supporting this trend. There is a definite and direct correlation between pulp yield and mol % S as determined by NBO; mol % S and pulp yield decrease in the order 19-5, 16-2, 26-1, 18-3, 21-5, 10-1.  These results demonstrate that although the thioacidolysis results do not directly  correlate with pulp yield, the information gathered from this technique can give insight into the ultrastructure of lignin. Conventionally, N B O is used for the determination of S:G ratio, whereas thioacidolysis is a good indicator of the linkages in lignin which are associated via P-O-4 bonds. Thioacidolysis only breaks monomers apart that are bonded via P-O-4 linkages, leaving some syringyl and guaiacyl units undetected. This is because both guaiacyl and syringyl units may be involved in 48  other linkages, such as other ether linkages and carbon-carbon bonds.  The S:G values from  NBO are all higher than found from thioacidolysis, indicating that there may be syringyl units that are present in the lignin not connected via a P-O-4 bonds. This ratio is significant due to the increase in pulping efficacy suggested to be associated with an increase in P-O-4 bonds (Chang and Sarkanen 1973). However, the yields of the two procedures tend to differ among similar wood species. For aspen, a N B O total yield of 42.4% of vanillin and syringaldehyde is considered normal (Pepper et al. 1967). Poplar M W L yielded 37% and 60% w/w from thioacidolysis and NBO, respectively. It is usually the case that N B O has a higher yield than thioacidolysis due to two reasons; N B O degrades other linkages in addition to alkyl-aryl ethers, and N B O may degrade non-lignin phenolics in addition to lignin (Chang and Allen 1971). Thioacidolysis yield is a measure of the portion of lignin monomers depolymerised during the wet chemistry procedure. For ground wood samples, 16-2 had a yield much higher than 10-1, at 1898 umol g" versus 1583 umol g" (Table 2.1). For M W L and pulp, the trend was 1  1  reversed, but for R E L sample 16-2 was higher than 10-1, as in ground wood. Sample 16-2 is more easily pulped because of a higher yield, and therefore may possess a higher percentage of pulping chemical susceptible bonds than sample 10-1. R E L had extremely low thioacidolysis yields (between 471 and 68 umol g") compared to ground wood or extracted lignin, indicating 1  that many of the labile /3-0-4 linkages had been degraded before the thioacidolysis procedure. Methoxyl content was determined by a modified Zeisel method (Browning 1967). In addition to the information gathered from thioacidolysis and nitrobenzene oxidation, methoxyl content was employed to confirm the ability of lignin to participate in unreactive carbon-carbon bonds (Chang and Sarkanen 1973). The methoxyl content is show in Table 2.3, as part of the expanded Cg formula for 10-1 and 16-2 M W L . Sample 10-1, with 1.33 methoxyl groups per Cg  49  is lower than the 1.40 for sample 16-2.  This further supports the increased syringyl content  determined for sample 16-2, and concurs with the nitrobenzene oxidation results.  2.3.4 Molecular Weight Determination Chemical differences are not the only property that accounts for difference in pulping performance, however.  The degree of polymerization of the lignin may also account for  differences in residual lignin content in Kraft pulp. Since Kraft pulping involves alkaline attack of lignin moieties to break the polymer into smaller charged species, inherently smaller lignin molecules  require less energy and chemical to  facilitate breakage, and consequently  solubilization. In order to evaluate this characteristic, gel-permeation chromatography was used to determine comparative molecular weight distributions for M W L , C E L and R E L fractions of lignin of the aspen clones. For M W L samples, both the elution time and distribution were different. Sample 10-1 and 16-2, representing extremes in pulp efficacy, had similar distributions, but 16-2 eluted slightly earlier than sample 10-1, indicating the slightly higher weight average molecular weight. Sample 19-5 stands out from the four remaining samples, as its distribution is clearly bimodal, with the second peak (molecular weight of approximately 400 g mol" based on polystyrene 1  standards) much larger than the first (molecular weight of 2200 g mol" for sample 10-1 and 1  7400 g mol" for sample 16-2 relative to polystyrene standards). 1  The five other samples also have this second molecular weight fraction, but it is seen mostly as a small "shoulder" and not as distinct as that observed in sample 19-5 (Figure 2.9). The chromatograph of sample 18-3 M W L shows an additional very low molecular weight (relative to polystyrene) signal (Figure 2.9).  Sample 18-3 also has a bimodal distribution,  however, it is not as pronounced as that of sample 19-5 and the peak at lower retention time is larger that the peak at the higher retention time. 50  Sample 19-5 had a relatively high yield in pulping trials, similar to sample 16-2. Sample 19-5 also had a relatively high S:G ratio as determined by nitrobenzene oxidation, 80.3%. There are a few possible explanations for the different molecular weight distribution in sample 19-5. One is that the increased syringyl units consequently result in a greater concentration of the more labile B-O-4 ether linkages that are more easily broken during lignin isolation procedures. However, since the exact same procedure was performed for all samples, including sample 16-2, this explanation is not likely. Another explanation may be that sample 19-5 naturally has lignin with lower molecular weight, and thus aids in improving its pulping efficacy and, as such displays a slightly higher pulp yield than 16-2 despite having higher total lignin content. The acetylated C E L molecular weight distributions look similar among all samples (Figure 2.10), each displaying three distinct fractions. The peak molecular weights for sample 10-1 were calculated as 13 700 g mol" , 400 g mol" and less than 100 g mol" relative to 1  1  1  polystyrene standards. For sample 16-2, the same peaks were at 14 300 g mol" , 400 g mol" , and 1  1  less than 100 g mol" . Analogous to the M W L fraction, sample 19-5 was unique with the two 1  higher molecular weight components eluting closer together than the other samples, and demonstrate a large portion of lower molecular weight moieties. Sample 16-2 has all three of its components at slightly higher molecular weight.  Samples 18-3, 21-5, and 26-1 have relative  distributions that are alike. Comparing M W L and C E L samples to each other shows that indeed, as expected, the C E L fraction has a higher molecular weight than M W L for most samples as determined by retention times.  This is expected since the C E L required the addition of enzyme treatment in  order to liberate it from the associated carbohydrates, and is not readily dissolved in dioxane:water due to the higher molecular weight. The M W L and C E L fractions make up a small part of the overall lignin, and therefore may not significantly influence the depolymerization during pulping as much as REL. 51  In  addition, because the R E L fractions were difficult to solubilize, even with acetylation, it is probable that the M W of these samples is actually higher, if the higher molecular weight material was filtered out before the GPC analysis. The peak molecular weights for the three main peaks of the R E L spectra were at approximately 120 000 g mol" , 400 g mol" , and less than 100 g 1  1  mol" , as calculated from polystyrene standards (Figure 2.11).  The highest molecular weight  1  peak for R E L is not surprisingly much higher than for M W L and C E L . The R E L molecular weight distribution for sample  10-1  has a large peak at  approximately 40 minutes (Figure 2.11). This peak is much higher than the lower and higher molecular weight peaks in this chromatogram. For sample 16-2, the largest signal is from the higher molecular weight peak at approximately 33 minutes retention time. This indicates that sample 16-2 has a higher proportion of its R E L in the higher molecular weight form than sample 10-1.  This indicates that if Kraft pulping efficacy depends on lignin molecular weight, other  factors are outweighing this difference in R E L molecular weight distribution between samples 10-1 and 16-2.  2.3.5 Nuclear Magnetic Resonance Spectroscopy Two of the six aspen samples were chosen for a more in-depth N M R analysis of isolated lignin. These samples were chosen because of their extreme differences in pulp efficiencies and lignin content (Figures 2.4 and 2.6).  One and two-dimensional N M R experiments were  performed on aspen M W L samples 10-1 and 16-2. Quantitative C N M R spectra were obtained 1 3  for both non-acetylated and acetylated 10-1 and 16-2 M W L as well as acetylated R E L (Figures 2.7a, 2.7b, 2.8, respectively).  The M W L samples chosen for these analyses had a slight  difference in the S:G ratio as determined by thioacidolysis (65.1 and 67.7 mol % S for 10-1 and 16-2, respectively) and N B O (74.4 mol % S for 10-1 and 79.0 mol % S for 16-2).  52  The aromatic region from 162.5 to 102 ppm was set equal to the integer 6, representing the 6 carbons in each aromatic ring. This may be slightly inaccurate as there are other types of carbon nuclei that display chemical shifts in this region that are not aromatics, such as some other sp hybridized carbons (Claridge 1999). For example, in softwoods, the presence of 0.12 2  vinylic carbons per aromatic ring (Chen 1998) necessitates using the integral of 160-100 normalized to 6.12 (Capanema et al. 2004). However, hardwoods are usually assumed to be free of anything other than aromatic carbons in this region (Yang et al. 2002): In addition, the H M Q C spectra (Figure 2.13) show separate aliphatic and aromatic regions. The anomeric carbon in carbohydrates usually has a chemical shift of 102 ppm, and this shift does not show a correlation with any aliphatic hydrogen atoms in the H M Q C spectrum. The samples were therefore considered to have insignificant carbohydrate content, and calculations will reflect this. The spectra were analyzed in order to quantify the lignin attributes associated with the Kraft pulping mechanism, and additional differences between samples were also noted.  Side Chain Moieties P-O-4 Moieties (Fig. 2.12, A-D) There are many different types of functional groups that surround (3-0-4 ether linkages. N M R is usually able to distinguish between structures with - O H , - C H 2 - , -C=0, -O-Aryl, and -OAlk groups at the a position, dibenzodioxocin, and other moieties (Ralph et al. 2001).  During  Kraft pulping, /3-0-4 linkages are primarily broken, and as such the study of the variations of this linkage is important in rationalizing pulping efficacy differences between samples. In the N M R spectra of 10-1 and 16-2 M W L , carbonyl groups above 170 ppm were absent, indicating that if present, non-conjugated CO, a-CO, structure L , structure M (vanillin), and I of Figure 2.12 are undetectable in these samples. If the region from 61-58 ppm is taken to estimate the amount of (3-0-4 linkages excluding those with a-CO, for sample 10-1, the result is 53  0.56/Ar while 16-2 is 0.68/Ar. This region contains two peaks, corresponding to erythro and threo isomers. Since no structure P is evident in the H M Q C spectrum shown in Figure 2.13 and the B-l (H content is 0.10/Ar and 0.09/Ar), the (3-0-4 content can be corrected to 0.54/Ar and 0.59/Ar for samples 10-1 and 16-2, respectively (Figure 2.12). It has previously been shown that the 8-0-4 content of beech 0.65/Ar (Nimz 1974) and spruce is 0.49-0.51/Ar (Erickson et al. 1973) . The N M R calculated values for these aspen samples are slightly lower than beech, but higher than the softwood spruce, as thus are good estimates of this bond. Since thioacidolysis yield is proportional to the syringyl units bound by 8-0-4 linkages, the results from N M R and wet chemistry seem to agree very closely (Table 2.1). In both cases, sample 16-2 has more units bonded together with Kraft pulp labile linkages, consistent with its better pulping characteristics than sample 10-1.  Phenylcoumaran,  Pinoresinol, and 0-1 Structures (Fig. 2.12, E, F, and H)  The absence of signal at 87 ppm in both 10-1 and 16-2 A c M W L suggests that there are no detectable phenylcoumaran subunits in the aspen lignin. Since the integral from 50-48 ppm (AcMWL) represents phenylcoumaran and B-\ structures, the value for 10-1 and 16-2 samples for 8-1 is simply this integral, 0.10/Ar and 0.09/Ar, respectively.  Beech 8-1 frequency was  found to be 0.15/Ar which is slightly higher than the values found for the aspen samples (Nimz 1974) . The pinoresinol/syringoresinol structure (Fig. 2.12, F) was calculated by subtracting the phenylcoumaran amount from the integral from 54 to 51.5 ppm in the non-acetylated spectrum. The frequency is very similar between the two samples, with 10-1 having 0.08/Ar and 16-2 having 0.09/Ar since one pinoresinol/syringoresinol structure contains two C-8 atoms.  Dibenzodioxocin  Structures (Fig. 2.12, G)  The signal at 83-81.5 ppm in the spectrum of A c M W L has been ascribed to be exclusively dibenzodioxocin structures (Fig. 2.12, G), which in this analysis gives identical estimations (0.11/Ar) for both samples.  However, since this region contains a large signal 54  overlap from a peak upfield, this is a large overestimation, and is not likely associated with dibenzodioxocin. Furthermore, hardwood samples are not expected to possess dibenzodioxocin units. If the value obtained for pinoresinol structures is subtracted from the integral 86-83 ppm, a more likely estimation of dibenzodioxocin subunits of 0.003/Ar and 0.01/Ar for 10-1 and 16-2, respectively are obtained, and can be considered negligible.  Functional Groups Methoxyl Groups The integral from 58-54 ppm is attributed to methoxyl groups plus a few very minor moieties. To correct for these, the values for 8-1 (Fig. 2.12, H), spirodienone (Fig. 2.12, I), and structure R (Fig. 2.12) were subtracted from the integral of 58-54 ppm. Since there was no spirodienone or R (60-59 ppm in AcMWL) found in the samples, the integral is simply the methoxyl minus the (3-1 content. For samples 10-1 and 16-2, this results in a methoxyl content of (1.43-0.10)/Ar = 1.33/Ar and (1.51-0.09)/Ar = 1.42/Ar. This small difference may suggest that more of 10-1 lignin methoxyl groups are taking part in linkages than the 16-2 lignin. The N M R calculation of methoxyl content translates directly into S:G ratio per aromatic ring. Methoxyl analysis was also carried out using a modified Zeisel method (Browning 1967) and the results are shown in the C9 formula of Table 2.3.  The percent OMe by weight was  19.6% in sample 10-1 and 20.4% in sample 16-2. Since both the wet chemistry method and the N M R were performed on the same M W L , it can be inferred that the two methods may be measuring the methoxyl content in slightly different ways.  Hydroxyl Groups Phenolic hydroxyl content is of interest due to the belief that the general Kraft pulping delignification reaction begins with the deprotonation of the phenolic proton (Sjostrom 1993). 13  From the  C N M R studies, phenolic hydroxyl content and conjugated ester functionalities can  55  be determined from integrating the signal at 168.8-167.5 ppm (AcMWL) (Capanema et al. 2004). The acetylated M W L values for this region are the same, 0.21/Ar. From a proton N M R analysis of the hydroxyl groups in M W L , both phenolic and aliphatic hydroxyl content can be quantified as a weight percentage.  For sample 10-1 M W L ,  phenolic and aliphatic hydroxyl content was 12.5 and 17.4%, respectively, and for sample 16-2 the values were calculated to be 12.1 and 16.6%. The slightly higher phenolic hydroxyl content for sample 10-1 M W L is not consistent with a higher phenolic hydroxyl content enabling more efficient pulping. The cluster at 77-65 ppm includes secondary O H groups and 7-O-Alk ethers.  In both  acetylated and non-acetylated, this value was higher for sample 16-2 than for sample 10-1. For sample 10-1, M W L was 0.95/Ar and A c M W L was 0.94/Ar and for sample 16-2 M W L was 1.04/Ar and A c M W L was 1.06/Ar. acetylated and non-acetylated spectra.  These values indicate very good agreement between Since the only substructure identified with 7-O-Alk  linkages is pinoresinol/syringoresinol, the remaining signal at 77-65 ppm is due to secondary O H groups.  Balance of Side Chain Structures The total amount of oxygenated carbon atoms (90-58 ppm) was 2.69/Ar and 2.70/Ar for samples 10-1 and 16-2 A c M W L , respectively.  These values include ether linked side chain  structures such as /3-0-4 linkages. No lignin subunits with short side chains were observed in the M W L . The carbonyl group of vanillin and syringaldehyde (Fig. 2.12, M) usually appears at 193191 ppm, but was absent in the aspen spectra. Other short lignin structures include the structures N and O in Figure 2.12. Since both of these structures contain carbonyl groups, it is likely they are not present in the M W L of these aspen samples. Structure N contains a conjugated carbonyl, which is expected to have a chemical shift above 190 ppm, and structure O contains a carboxylic acid, which is expected to have a chemical shift of approximately 178 ppm. 56  Based on N M R analyses, it appears that there are no structures with shortened side chains, and therefore it is expected that the number of side chain carbons for each aromatic ring is 3. The methoxyl content was subtracted from the integral of the aliphatic region from 90-45 ppm and the aliphatic esters from 170-164 ppm were added to that value.  This side chain  estimation gave 3.06/Ar for sample 10-1 and 3.04/Ar for sample 16-2. This could either mean that in this region the calculations are slight overestimations, or that the error for these calculations due to spectral processing is approximately 2%.  Aromatic Carbons and Degree of Condensation The frequency of various aromatic carbon types can be estimated.  The methine (C-H)  carbons in the range 125-103 ppm indicate how many positions on the ring are substituted with carbon and oxygen. The methine carbon range for sample 10-1 was 2.06/Ar whereas for sample 16-2 it was 2.02/Ar. A small amount of/>-hydroxyphenol units were detected in the spectra at 162.5-160 ppm. Since aspen lignin has both syringyl and guaiacyl lignin monomers, the chemical shift range of 155-151 ppm cannot be used in the calculation of degree of condensation without correction (Capanema et al. 2004). The C4 of non-conjugated 5-5' phenolics has a chemical shift of 144142 ppm, along with C3 from phenylcoumaran. However, since there is no phenylcoumaran a-C signal at 88-86 ppm, this region can be attributed to 5-5' non-conjugated phenolics. As expected for hardwood lignin, the amount of this 5-5' is very low, 0.06/Ar and 0.07/Ar in non-acetylated M W L and negligible and 0.05/Ar for A c M W L . Oxygenated aromatic carbons resonate between 160 and 141 ppm and the quantification of these is an estimation of syringyl content.  For sample 10-1 the integral of 2.01/Ar and  2.00/Ar for sample 16-2 are not different, although this fits with the wet chemistry results. Sample 16-2 has a higher S:G ratio based on thioacidolysis and N B O , and the values oxygenated aromatic carbons is consistent with this trend. 57  The degree of condensation of lignin is calculated by taking the difference between the theoretical value for non-condensed guaiacyl units and the integral cluster from 125-103 ppm. The S:G ratio from M W L thioacidolysis (Table 2.1) was taken and the theoretical degree of condensation was found for each sample (2.35/Ar for sample 10-1 and 2.32/Ar for sample 16-2). Corrections for the amount of j7-hydroxyphenyl (h units) moieties made as these compounds have carbons that resonate similarly in this region. In this case, samples 10-1 and 16-2 have 0.11/Ar and 0.10/Ar of h units, and the degree of condensation calculation gives 0.07/Ar for sample 10-1 and 0.02/Ar for sample 16-2. Hardwoods are not expected to have a high degree of condensation due to their high syringyl content. For these aspen samples, sample 10-1 is about three times as condensed as sample 16-2.  This relationship between samples correlates to  pulping efficacy, with sample 10-1 being more resistant to pulping which may come from a higher degree of condensation than sample 16-2. The literature value for beech 5-5' linkage frequency is 0.023/Ar which is very low, and expected since beech has a higher methoxyl content to block the formation of 5-5' carbon carbon bonds (Nimz 1974).  Residual Lignin The A c R E L sample spectra were obtained, however, sample 10-1 did not have enough S/N to integrate and give accurate results. The region from 162.5 to 102 ppm did not have as much signal as one would expect for a primarily aromatic sample.  A high sample viscosity  made it difficult to have a dilute enough solution where the molecules could tumble freely yet remain in high enough abundance to give a large enough signal. The spectra are shown in Figure 2.8 to display the low signal in the aromatic region. The extraction and characterization of R E L is on-going, and since R E L makes up such a large portion of the total lignin, may contain useful structural information.  58  2.4  Conclusion The variation in wood chemistry among the six aspen clones evaluated is very clear.  Both carbohydrate and lignin content play key roles in Kraft pulp yield and quality. However, this research has shown that in addition to total lignin content, there are structural lignin differences that can influence pulping efficacy. The most likely difference is the distribution in molecular weight and methoxyl content, which vary between the different fractions of lignin. Nitrobenzene oxidation supports the differences observed in pulping, whereas the thioacidolysis results were more subtle. However, it is important to look at the whole wood and the individual lignin fractions, M W L , C E L , and R E L to fully understand the differences in lignin chemistry and its association with various carbohydrates. Table 2.5 contains a comparison of the main lignin characteristics determined by N M R and how these attributes may impact the Kraft pulping of these samples. Samples 10-1 and 16-2 were chosen based on their dramatic differences in pulping characteristics, despite being of the same aspen species. Although M W L is often taken to represent the total lignin in wood, it can only be extracted with a low yield. What is left over, the R E L , has a greater impact on how the total lignin behaves.  The relative molecular weight of the R E L samples correlates with Kraft  pulp yield and Kappa number results, even though M W L and C E L fractions do not.  59  2.5  Tables and Figures  aspen  i ball milled wood  crude lignin  acetylated milled wood lignin (AcMWL)  insoluble  acetylated cellulase lignin (AcCEL)  acetylated residual lignin (AcREL)  Figure 2.1. Isolation, purification, and acetylation of various lignin fractions. 1) 96:4 dioxane:water 2) washed with 3 x D l H2O 3) 90% acetic acid, ppt into water, centrifuge, wash 2x with water, 1:2 ethanol:l,2-dichloroethane, ppt into ether wash 2x with ether dry 4) enzyme treatment 3x2 days 5) same purification as for M W L 6) rinsed 3x with water 7) N M I / D M S O / A c 0 acetylation 2  60  mm  10-1  16-2  18-3  19-5  Extractives  21-5  26-1  Sample Number  Figure 2.2. Total wood composition of the six natural aspen clones from northern BC, Canada. Error bars represent the standard deviation of three replicates.  61  10-1  16-2  18-3  19-5  21-5  26-1  Sample Number  Figure 2.3. Individual neutral wood sugar composition of the six natural aspen clones from northern BC, Canada. Error bars represent the standard deviation of three replicates.  62  Figure 2.4. Total lignin content of the six natural aspen clones from northern B C , Canada. Error bars represent the standard deviation of three replicates.  63  V///A  Fatty Acids and Resin Acids Sterols Steryl Esters  10-1  16-2  18-3  19-5  21-5  26-1  Sample Number  Figure 2.5. Percent individual extractives classes of the total extractives of six natural aspen from northern B C , Canada. Error bars represent the standard deviation of three replicates.  64  Y////A  ^  10-1  16-2  18-3  19-5  Kappa Number Pulp Yield  21-5  26-1  Sample Number  Figure 2.6. Pulp yield and residual Kappa number of pulp (H-factor 1100) from six natural aspen clones from northern B C , Canada. Error bars represent the standard deviation of at least three replicates (10-1 pulp yield is based on six replicates).  65  Table 2.1. Mole percent syringyl (% S) units in various substrates, wood, M W L , REL, and Kraft pulp, based on thioacidolysis.  Sample  Yield pmol/g 1583  MWL %S  10-1  Wood %S 73.7  16-2  73.6  18-3  65.1  Yield pmol/g 1102  1898  67.7  73.4  1795  19-5 21-5  73.6  26-1  Pulp %S  76.4  Yield pmol/g 61.8  54.4  Yield pmol/g 258.2  1041  67.7  85.2  49.9  162.4  61.7  1112  78.0  266  59.1  178.5  66.3  1364  75.0  302  64.4  135.7  71.3  2151 1897  62.5  1329  74.0  471  58.0  188.5  75.6  1910  64.9  1348  76.9  229  65.2  182.5  66  REL %S  Table 2.2. Comparison of mole % syringyl units as determined by thioacidolysis and nitrobenzene oxidation of extractive-free ground wood samples. Error represents the range of two samples.  Sample 10-1 16-2 18-3 19-5 21-5 26-1  % S thio 73.7 ± 0.5 73.6 ± 0.7 73.4 ± 0 . 1 73.6 ± 0.6 71.3 ± 0 . 0 2 75.6 ± 0.4  67  % S 74.4 79.0 77.2 80.3 74.6 78.4  NBO ± 0.5 ± 3.2 ± 1.4 ± 1.2 ± 0.7 ±1.0  Table 2.3. Elemental analysis and expanded C g formula for M W L for 10-1 and 16-2. Methoxyl content was determined by a modified Zeisel wet chemistry method and hydroxyl content was determined by ' H NMR. MWL Sample 10-1 16-2  Elemental Analysis C (%) H (%) 0 (%) 58.90 6.41 34.63 58.65 6.91 34.35  OMe (wt %) 19.56 20.38  68  OH (wt %) 17.36 16.59  C  g  formula  C9H7.26O-i.08OME-i.33OH2.15 C9H8.33O1  ioOMei  .40OH2.08  Table 2.4a. Signal assignment in the N M R spectrum of non-acetylated M W L (sample 10-1 16-2). Peaks correspond to Figure 2.7a.  spectral region C-4 in h units C-3 in T  e t >  C3,5 in Set, C-a in L, C-3,6 in 1, C-4 in  chemical shift range (ppm) 162.5-160.2 155-151  conjugated S , unknown 145.6-144.9 C3 in E, C-4 in conjugated Set C-3 in E , C-4 in T , C-4 in conjugated S , 144.5-142.5 unknown 61-58 B-O-4 58-54 OMe, C-p in H, C-1 in I, C-y in R 54-51.5 C-p in E and F n e  10-1 (perAr) 0.11 1.08  16-2 (per Ar) 0.10 1.10  Peak Label 1 2  0.06 0.06  0.05 0.06  3 4  0.56 1.54 0.16  0.68 1.49 0.19  5 6 7  2.01 1.61 2.27 2.76 0.94 0.94 0.89  2.00 1.60 2.20 3.05 1.06 1.06 0.92  clusters CAr-0 CAr-C CAr-H  Alk-OAlk-O-Ar, a-O-Alk y-O-Alk, OHsec OH prim  160-141 141-125 125-103 90-58 90-77 77-65 65-58  69  Table 2.4b. Signal assignment in the N M R spectrum of acetylated M W L . Peaks correspond to Figure 2.7b.  chemical shift range (ppm) primary aliphatic OH 171.4-169.7 secondary aliphatic OH 169.7-168.8 phenolic OH, conjugated COOR 168.8-167.5 all C-3 (except E and h-units), C-5 in S, C-a in L, C-6 160-148 in 1, C-4 in conjugated CO/COOR etherified and hunits C-3 in E, C-4 in conjugated Set, unknown 144.5-142.5 85.55-84.3 C-a in F, G 80.84-78.55 C-p in G 77-72.5 C-a in A, H, P, carbohydrates 57-54 OMe, C-1,and C-p in I 50-48 F./3-1 clusters 125-103 CAr-H Alk-O90-58 90-77 Alk-O-Ar, a-O-Alk 77-65 y-O-Alk, OH ec 65-58 OH pnm spectral region  S  70  10-1 (per Ar) 0.68 0.54 0.21 1.63  16-2 (per Ar) 0.73 0.56 0.21 1.72  Peak Label 1 2 3  0.05 0.08 0.35 0.56 1.43 0.10  0.04 0.07 0.41 0.58 1.51 0.09  4 5 6 7 8 9  2.08 2.67 0.88 0.95 0.83  2.00 2.70 0.86 1.04 0.79  Table 2.4c. Lignin structural unit calculations as determined by N M R spectroscopy. Legend for substructures is in Figure 2.12. .  structure  calculation  6-1 (H) pinoresinol (F) dibenzodioxocin (G) methoxyl side chain degree of condensation p-hydroxyphenyl (h units)  (50-48) - peak at 87 if present (54-52) -E(na) (86- 83)-F(ac) (58-54) - H (90-45) -OMe + (170-164) theoretical from S:G - (125-103)na (162.5-160)  71  10-1 (per Ar) 0.10 0.08 0 1.33 3.06 0.07 0.11  16-2 (per Ar) 0.09 0.09 0.01 1.42 3.04 0.02 0.10  10-1 MWL  V  S-  1  2  1  YM 34  5 6 7  16-2 MWL  -r-  1  2  VV 3 4  5 ~i  ppm (t1)  1  150  Figure 2.7a. Quantitative Table 2.4a.  1  7  r  ~i  100  , 3  6  1  1  50  C N M R spectrum for M W L samples. Peak labels correspond to  72  1-3  8  9  8  9  16-2 AcMWL  1-3  5 ~i  ppm (t1)  1  150  Figure 2.7b. Quantitative  7  1-  100  1 3  6  50  C N M R of AcMWL. Peak labels correspond to Table 2.4b.  73  Table 2.5. Summary of M W L N M R results relevant to the Kraft pulping mechanism.  lignin characteristic p-O-4  phenolic hydroxyl content methoxyl content degree of condensation  function in Kraft pulping  10-1 (perAr) 0.54  this alkyl-O-aryl ether bond is cleaved during Kraft pulping, higher frequency of this linkages gives rise to pulp efficacy phenolic OH needed for initiation of Kraft pulping 0.21 mechanism methoxyl "block" the formation of 5-5' carbon-carbon 1.33 bonds, which are unreactive in pulping condensed units are relatively chemically resistant 0.07 carbon-carbon bonds  74  16-2 (per Ar) 0.59  0.21 1.42 0.02  10-1 AcREL  150  100  50  ppm (f1) Figure 2.8. Quantitative C N M R spectra of aspen clonal acetylated R E L samples. 1 3  75  I  I  I  I  I  1  I  25  30  35  40  45  50  55  Elution Time (min) i  i 100 000  i  i  10 000  i  I  1 000 100  Molecular Weight Relative to Polystyrene (g/mol)  Figure 2.9. Molecular weight distributions for acetylated aspen M W L samples.  76  Figure 2.10. Molecular weight distributions for acetylated aspen C E L .  77  I—I—I—I—I—I—I—I—I—I—I—I—I—I—I—I—L_J  25  30  35  I  I  I  40  45  I  I  I  •  !  »  50  Elution Time (min) i  i  i  100 000 10 0 0 0  i 1 000  i 100  Molecular Weight Relative to Polystyrene (g/mol)  Figure 2.11. Molecular weight distributions for acetylated aspen R E L .  78  •  •  •  !  55  79  10-1 MWL  HMQC ©  F®  OMe  co.©0"  i  i  r  i  |  i  i  I  i  |  i  I  I  r  |  i  i  1  I  |—I—I  I — i — | — i — i — i — i — | — i — i — i — i — | — i — i — i — r  ppm(t2) Figure 2.13. H M Q C spectrum of sample 10-1 and 16-2 70  6  0  5  0  4  0  3  0  2  80  0  1  0  MWL.  2.5  Bibliography  Adams, T., B. Cowan, D. Clayton, D. Easty, D. Einspahr, D. E . Fletcher, T. M . Malcolm, T. McDonough, L . Nordman, J. K . Perkins, D. W. Reeve, L . Schroeder, G. A . Smook, N . Thompson, S. V . Fleet and B. Wikdahl, Eds. (1989). Alkaline Pulping. Pulp and Paper Manufacture. Atlanta, The Joint Textbook Committee of the Paper Industry. Bjorkman, A . (1956). "Studies on Finely Divided Wood Part 1. Extraction of Lignin with Neutral Solvents." Svensk Papperstidning 13: 477-485. Brown, C. (2000). The Global Outlook for Future Wood Supply from Forest Plantations. Rome, Italy, Food and Agriculture Organization of the United Nations. Browning, B. L . (1967). Acetyl and Methoxyl Groups. Methods in Wood Chemistry. B. L . Browning. Appleton, Wisconsin. II: 660-664. Campinhos, E . (1999). "Sustainable plantations of high-yield Eucalyptus trees for production of fiber: the Aracruz case." New Forests 17: 129-143. Capanema, E . A . , M . Y . Balakshin and J. F. Kadla (2004). "A comprehensive approach for quantitative lignin characterization by N M R spectroscopy." Journal of Agricultural and Food Chemistry 52(7): 1850-1860. Chang, F.-J. and T. Kondo (1971). "Studies on Brightness Reversion of Pulps. II. Chemical Constituents of Bleached Pulps on Brightness Reversion." Mokuzai Gakkaishi 17: 160166. Chang, H.-m. and G. G. Allen (1971). Oxidation. Lignins. Occurence, formation, structure, and reactions. K. V . Sarkanan and C. H. Ludwig. New York, Wiley-Interscience: 433-485. Chang, H.-m. and K. V . Sarkanen (1973). "Species Variation in Lignin: Effects of Species on the Rate of Kraft Delignification." Tappi Journal 56(3): 132-134. Chen, C. L . (1998). Characterization of milled wood lignins and dehydrogenated polymerates from monolignols by carbon-13 N M R spectroscopy. A C S Symposium Series 697. N. G. Lewis and S. Sarkanan. Washington, D C , American Chemical Society: 255-275. Claridge, T. D. W. (1999). High-Resolution N M R Techniques in Organic Chemistry. Oxford; Pergamon. Downes, G., R. Evans, R. Wimmer, J. French, A . Farrington and P. Lock (2003). "Wood, pulp and handsheet relationships in plantation grown Eucalyptus globulus." Appita Journal 56(3): 221-228. Erickson, M . , S. Larsson and G. E . Miksche (1973). "Gas-chromatographische Analyse von Ligninoxydationsprodukten. VII. Zur Struktur des Lignins der Fichte." Acta Chem Scand 27: 903-904.  81  Heilman, P. E., R. F. Stettler, D. P. Hanley and R. W. Carkner (1995). High-yield Hybrid Poplar Plantations in the Pacific Northwest (revision). Pacific Northwest Regional Extension Bulletin PNW 356. Pullman, Washington, Washington State University Cooperative Extension. Huntley, S. K., D. Ellis, M . Gilbert, C. Chappie and S. D. Mansfield (2003). "Significant Increases in Pulping Efficiency in C4H-F5H-Transformed Poplars: Improved Chemical Savings and Reduced Environmental Toxins." Journal of Agricultural and Food Chemistry 51: 6178-6183. Nimz, H . H . (1974). "Beech lignin - proposal of a constitutional scheme." Agnew Chem Int Ed 13: 313-321. Pepper, J. M . , B. W. Casselman and J. C. Karapally (1967). "Lignin oxidation. Preferential use of cupric oxide." Can. J. Chem. 45: 3009-3012. Ralph, J., S. A . Ralph and L. L . Landucci (2001). N M R database of lignin and cell wall model compounds. 2004. Rolando, C , B. Monties and C. Lapierre (1992). Thioacidolysis. Methods in Lignin Chemistry. S. D. Lin, C.W. Berlin, Springer-Verlag. Sjostrom, E . (1993). Wood Chemistry: Fundamentals and Applications. San Diego, Academic Press, Inc. Yang, R., L . Lucia, A . J. Ragauskas and H . Jameel (2002). "Oxygen Degradation and Spectroscopic Characterization of Hardwood Kraft Lignin." Industrial and Engineering Chemistry Research 41: 5941-5948.  82  Chapter 3 Elucidating the Effects of Lignin Monomer Composition on Chemical Pulping Efficacy in C4H:F5H Transformed Hybrid Poplar 3.1  Introduction In the continuing search for high-yield, high-quality pulp, a great deal of attention has  recently been focused on improving woody substrate through biotechnological techniques to both reduce chemical load and energy requirements for processing. In particular, lignin structure and content has been shown to impact both pulp yield and residual lignin content (Kappa), which subsequently affects pulp bleachability and delignification efficiency (Huntley et al. 2003). For example, it has been accepted that substrates that display a higher syringyl to guaiacyl lignin monomer ratio (S:G) are more easily delignified during the chemical pulping process than substrate with lower S:G (Chang and Sarkanen 1973). Furthermore, it was shown that pulp yield is proportional to syringyl content of the substrate. In order to explain such findings, the authors proposed that lignin containing fewer methoxyl groups {i.e., more guaiacyl units) may participate in more extensive carbon-carbon bonding between phenylpropane units, and these linkages are more chemically resistant to pulping chemicals (Chang and Sarkanen 1973). Other factors may also account for altered rates of delignification, including the 3dimensional structure of the lignin polymer, its interactions with cellulose and hemicellulose in lignin-carbohydrate complexes (LCC), and the extent of polymerization of the lignin. Therefore, total lignin content and syringyl to guaiacyl ratio may only represent the more obvious and readily obtained delignification parameters, while other differences in the lignin are likely to contribute to pulping efficacy.  It should be noted that not only chemical limitations exist, but  also physical and morphological parameters likely may influence rates of delignification, including the ability of pulping liquor to permeate the sample, heat of wetting, and specific gravity.  83  Genetically modified trees offer the ability to design model substrates for pulping studies. For example, the over-expression of the ferulate 5-hydroxylase gene under the regulation of the cinnamate 4-hydroxylase promoter (C4H-F5H) demonstrated substantial increases in the mole percent syringyl units in several lines of hybrid Poplar (Huntley et al. 2003). The production of sinapyl alcohol and subsequently syringyl units is controlled by the F5H enzyme (Franke et al. 2000). As a result of the increased syringyl units incorporated into the lignin biopolymer, these transformed lines were delignified at a significantly lower chemical loading as compared to wild-type trees grown under similar conditions to the same age. In addition, these modifications were not associated with a change in total lignin content or observed phenotypic differences (Huntley et al. 2003). To expand on these very interesting, and significant findings, we chose to investigate the nature of the chemical modification to the lignin in these up-regulated C4H-F5H hybrid Poplar lines in order to support the notion that increased S:G improves pulping efficacy, and to determine the effects of modification on the nature of the lignin macromolecule. Alteration in lignin degree of polymerization, linkage, and functionality may also play a part in changes in pulp delignification. Elucidating carbon-carbon skeletal and functional group information of wood with N M R spectroscopy has become routine, and a combination of 1- and 2-dimensional N M R techniques have been shown to give a comprehensive evaluation of milled wood lignin (Capanema et al. 2004). These N M R techniques have been used to study lignin from transgenic trees and plants, isolated or cell wall lignin, and pulp residual lignin (Froass et al. 1998; Ralph et al. 1998; Kim et al. 2003; Lu and Ralph 2003). The advantage of using N M R over traditional wet chemistry methods, and especially degradation methods, is that the "whole" picture can be obtained with no chemical alteration. However, until cell wall dissolution and acetylation was used on ground wood .samples, isolation of lignin produced chemical alterations despite the lack of reactive substances added to the lignin (Ikeda et al. 2002). 84  In addition to different N M R spectroscopy techniques, gel permeation chromatography (GPC) was used to characterize the molecular weight properties of the isolated lignin, as well as traditional wet chemistry methods.  As more information is obtained regarding the effects of  lignin ultrastructure on pulping efficacy,  further insights  will direct targeted genetic  manipulations. Thus, the main purpose of this research was to comprehensively determine the nature of the difference in the chemistry and substructure of lignin between wild-type and the C4H-F5H transformed hybrid poplar.  85  3.2  Materials and Methods  Plant material. The generation of nine lines of C4H-F5H transgenic hybrid poplar (P. tremula x P. alba) was previously described (Meyer et al. 1998; Franke et al. 2000). Both wild-type and the transgenic line containing 93.4% syringyl lignin content (as determined by thioacidolysis) was selected based on previous Kraft pulping studies (Huntley et al. 2003) as the substrates for detailed analysis.  Material Preparation. Extractive-free ground wood was prepared by grinding wood in a Wiley mill to pass through a 40 mesh sieve and extracting with acetone for 24 hours. The extract free wood was dried over P2O5 desiccant until analyzed.  Methoxyl Analysis. Methoxyl content was determined by a modified Ziesel method (Browning 1967).  Thioacidolysis. The standard degradation procedure was performed on extract free wood, M W L , C E L and R E L to determine differences in lignin monomer content in the various wood and lignin fractions (Rolando et al. 1992).  Ball Milled Wood. Extractive-free ground wood (-15 g) was ball-milled using a Uni Ball Mill II (Australian Scientific Instruments) set to 85% maximum rotation speed. The ball mill was fit with two stainless steel drums containing 5 stainless steel ball bearings. After one week, 400 mL toluene (Fisher) was added to minimize lignin oxidation reactions.  Following an additional  week of ball milling, the toluene was removed by centrifuging the sample at 4,000 rpm for 10 minutes and decanting the toluene. This process generated ball milled wood (BMW).  Isolation of Lignin from Ball Milled Wood. Lignin was extracted from the B M W (Figure 3.1) using a mixture of dioxane and water (96:4) according to the method of Bjorkman (Bjorkman 1956). Approximately 200 mL of dioxane solution was used to extract the lignin from 15 g of BMW. After one day of constant stirring, the solvent was recovered through centrifugation, and the dioxane:water solution replenished. The extraction was allowed to proceed for an additional 86  three days, after which time lignin extraction was shown to be minimal, and the extractions pooled. The crude lignin resulting from the extraction of B M W was subject to the purification scheme described by Bjorkman.  This purification involved dissolving the liberated lignin in  90% acetic acid and precipitating into de-ionized water.  After washing with double distilled  water twice, the sample was dried under vacuum. The lignin was then dissolved in 1:2 ratio of ethanol: 1,2-dichloroethane and precipitated into diethylether.  The purified lignin was then  washed twice with diethylether and dried. The resulting lignin was classified as milled wood lignin (MWL). The insoluble material left after dioxane extraction, consisting of lignin more closely associated with carbohydrates, was subject to a carbohydrate digestion by hydrolytic enzymes. Xylanase (Iogen, Ottawa, ON), Gammanase (Novozyme,  Franklinton, NC), Novo-188  (Novozyme), and Fibrilase H D L 160 (Iogen) were mixed together in equal volumes, and 1 mL of the resulting enzyme cocktail added to each sample suspension in a 50 m M acetate buffer solution (pH 4.5). The cocktail was supplemented with tetracycline (Sigma, 40 ug mL" ), 1  cyclohexamide (Sigma, 30 ug mL" ) and sodium azide (Fisher, 0.02%) to prevent microbial 1  growth, and incubated at 45°C and 115 rpm.  After two days the enzyme solution was  regenerated and the solublized carbohydrates removed, and the digestion repeated two more times. The material was again subjected to a dioxane:water extraction and purified in the same manner as the M W L . The term C E L applies to purified lignin from this fraction, while R E L refers to residual lignin that was remaining and found to be insoluble in dioxane:water solution.  Molecular Weight Determination.  M W L , C E L , and R E L samples were acetylated by  dissolving mg quantities in N-methylimidazole and dimethylsulfoxide in a ratio of 1:2 and adding acetic anhydride (0.6 x NMI). Samples were left for 3 hours, precipitated into 6 m M E D T A (pH 8), filtered through a 0.2 urn membrane filter, and dried under vacuum. Samples of 87  ~1 mg mL" in THF were filtered through a 0.45 um filter. A Dionex Summit HPLC fit with 1  Waters Styragel H5 column was employed to obtain the molecular weight distribution at 50°C with a 0.5 mL min" flow rate using a photodiode array detector with the wavelength set to 280 1  nm. The total run time was 60 minutes.  Nuclear Magnetic Resonance Spectroscopy Experiments. Samples of acetylated cell wall material were prepared at a concentration of 150 mg mL" in CDCI3 (Cambridge Isotope 1  Laboratories). Ten minutes of sonnication was usually required to dissolve the sample. The sample was then filtered through a Kimwipe plugged disposable pipette into a 5 mm N M R tube. Heteronuclear multiple quantum coherence (HMQC) and heteronuclear multiple bond coherence (HMBC) spectra were recorded on a Bruker A V A N C E 400 MHz (BBi-x probe) spectrometer. All runs were performed at room temperature, approximately 293 K. M W L or A c - M W L was dissolved in d 6 - D M S O (Cambridge Isotope Laboratories) at the approximate concentration of 60 mg in 0.25 mL. The sample was allowed to fully dissolve overnight, and stored over dessicant. The next day, the sample was filtered through a Kimwipe into a 5 mm Shigemi N M R microtube (with the magnetic susceptibility of DMSO, Shigemi Co.). H M Q C and H M B C spectra were obtained as previously described. 1^  For quantitative  C N M R spectrum acquisition, 10 m M final concentration of chromium  (III) acetylacetonate,. Cr(acac)3 (Aldrich), was added to shorten the T i relaxation times of the carbon nuclei to allow signals to be quantified. For M W L , a Bruker AM-400 spectrometer was used, while for the R E L , a Bruker AV-300 was employed.  A pulse width of 90°, a 1.4 s  acquisition time and a 1.7 s relaxation delay were the parameters for the inverse gated sequence which means the decoupler was on during acquisition.  Approximately 20,000 scans were  collected for M W L samples. The d 6 - D M S O peak at 39.5 ppm or the CDCI3 peak at 77 ppm was used as the reference peak. 88  Hydroxyl Content. Aliphatic and phenolic hydroxyl contents were determined via proton N M R spectroscopy. Para-nitrobenzaldehyde was used as the internal standard (stock solution of 1 mg mL" in CDCI3). A c - M W L was dissolved into this solution at a concentration of 30 mg mL" of 1  1  internal standard solution. A standard ID proton N M R experiment was used with 256 scans on a Bruker AV-300 M H z spectrometer.  Integrated aliphatic and phenolic hydroxyl proton signals  were quantified by comparison to the benzaldehyde proton signal from the internal standard.  Elemental Analysis.  Elemental analysis (C,H,N,0 by difference)  was performed on a  PerkinElmer Series II CHNS/O Analyzer 2400 (Boston, MA) as per manufacturers instruction.  89  3.3  Results and Discussion  3.3.1 Wet Chemistry Methods The transgenic C4H-F5H poplar lines employed in this study were previously shown to exhibit a significant range of mole percent syringyl lignin, as high as 93.5% (Rolando et al. 1992; Huntley et al. 2003). Table 3.1 shows the extreme differences in lignin monomer ratio and thioacidolysis yield as seen in both wood and various lignin fractions when comparing wildtype and the most extreme C4H-F5H samples. Clearly, large phenotypic differences result from selective gene manipulation. The  S:G ratio of the modified lignin is consistent with previously determined  nitrobenzene oxidation (Franke et al. 2000).  The R E L , which is least soluble, has a lower  thioacidolysis yield than the more soluble M W L and C E L fractions at similar concentrations, and this is likely related to a combination of higher molecular weight and differences in chemical structure. Although M W L has been previous compared and thought to be similar in structure to CEL, there is a difference in the thioacidolysis yield between the two samples (Holtman and Kadla 2004). The higher yield for C E L during the thioacidolysis reaction for both the wild-type and C4H-F5H samples  suggests that the actions of the enzyme cocktail has liberate  carbohydrates both free and linked to the lignin, and consequently more lignin is available for degradation during the thioacidolysis procedure. In addition to thioacidolysis, methoxyl, hydroxyl and elemental analyses were performed to determine an expanded C9 formula for both samples of M W L . The variation in C, H , and O composition is indicative of varying functionality within the lignin polymer (Table 3.2).  The  increased methoxyl and hydroxyl content in the C4H-F5H M W L indicates that not only is more syringyl monomer being produced via the up-regulation of the F5H gene, but that other functional groups are affected as well, such as hydroxyl content. These structural changes will be put into context with other structural changes through N M R spectroscopy. 90  The extracted M W L makes up 12% of wild-type and 10% of C4H-F5H dry wood by weight. The wild-type C E L fraction is 5% of dry wood weight and the C4H-F5H only 2%. Since the total lignin content of these samples, as determined by the Klason method, is approximately 30%, most of the lignin is not extracted as M W L or C E L , but left as REL.  3.3.2 Molecular Weight Determination The improved delignification efficiency during chemical pulping of the C4H-F5H poplar lines is undoubtedly linked to the increased S:G ratio, however, other factors such as a decrease in the degree of polymerization may also facilitate the ease of delignification. As Kraft pulping proceeds, linkages between lignin monomers are broken, and as such the fewer the linkages the more intensive the chemical degradation. Gel-permeation chromatography (GPC) was used to investigate the qualitative molecular weight distribution of the isolated lignin fractions. The GPC data of three fractions from both wild-type and C4H-F5H poplar lignin illustrates that the genetically modified lignin has a much lower molecular weight, and thus is likely to take less time and chemicals for delignification to occur (Figures 3.2, 3.3, and 3.4). This is possibly due to the reduced number of linkage sites for the syringyl units, as the C5 is blocked from bonding by an additional methoxyl group. Perhaps the reduction in binding sites reduces the degree of polymerization of lignin. It should be noted that all samples were acetylated to ensure comparability across the samples types, even though M W L and C E L were soluble without being acetylated. The GPC trace of M W L demonstrates that the distribution of molecular weight is slightly different between samples (Figure 3.2), with the wild-type having a lower retention time and therefore a higher molecular weight. The width of the peak for the C4H-F5H M W L is broader than that of wild-type, which indicates that the polydispersity of the molecules is higher in the modified sample. 91  In contrast, the C E L fractions have almost identical distributions except for a small lower molecular weight component around 45 min present in the wild-type sample (Figure 3.3). This retention time corresponds to a molecular weight of less than 100 g mol" calculated from the 1  retention times of polystyrene standards. This molecular weight is less than that of one lignin monomer unit; however, the calculated molecular weight may not reflect the lignin molecular weight exactly since polystyrene is not chemically and structurally like lignin. Despite this, the small peak below 100 g mol" is likely due to very low molecular weight material. The residual 1  lignin fractions were measured and compared, and the distribution shows that the wild-type has a substantially greater portion of its lignin as high molecular weight, and a lower portion of its lignin as low molecular weight (Figure 3.4). The opposite is true for modified residual lignin, which undoubtedly contributes to its ease of pulping and higher yield in thioacidolysis. The R E L samples were both less soluble in the GPC solvents than were M W L and C E L fractions, and with filtering there is a chance that even larger molecular weight portions of R E L are not accounted for. The samples are bimodal, whereas the M W L and C E L are not. The wildtype R E L has a higher portion of its lignin at a higher molecular weight, whereas for the C4HF5H the higher portion of its lignin is in the lower molecular weight fraction. Since the R E L makes up the highest proportion of the total lignin by weight, differences in this fraction will have a higher impact on overall lignin chemistry and pulp performance. It is interesting to note that the C E L fraction has a higher molecular weight (based on peak retention times) than the M W L for the modified C4H-F5H lignin. This may indicate that the genetically modified lignin is more susceptible to the cellulase treatment than the corresponding wild-type trees. The wild-type M W L is actually lower in molecular weight than its C E L fraction, indicating that higher molecular weight lignin is liberated from the carbohydrate matrix during the enzyme treatment. Although it is thought that C E L is similar in  92  chemical structure than M W L , the thioacidolysis yield (a measure of B-O-4 bonds) is higher for C E L and the molecular weight of the C E L samples is different than the M W L . These dramatic changes in the molecular weight of the three lignin fractions are the first reported for a genetic modification of the lignin biosynthetic pathway.  The C4H-F5H  modification not only leads to lignin that has been altered in S.G ratio, but has a lower molecular weight as well. It is strongly conceivable that this large change in lignin substructure is a key modification that results in the modified wood being more easily delignified during Kraft pulping.  3.3.3 Nuclear Magnetic Resonance Spectroscopy The aromatic region from 162.5 to 102 ppm was set equal to the integer 6, representing the 6 carbons in each aromatic ring. Although this may be slightly inaccurate as there are other types of carbon nuclei that display chemical shifts in this region that are not aromatics, such as some other sp hybridized carbons (Claridge 1999), is does serve as a means for comparative 2  analyses between different samples.  For example, in softwoods, the presence of 0.12 vinylic  carbons per aromatic ring (Chen 1998) necessitates using the integral of 162.5-102 normalized to 6.12 (Capanema et al. 2004). However, hardwoods are usually assumed to be free of anything other than aromatic carbons in this region (Yang et al. 2002). *  In order to determine functional group frequency, quantitative  1 "\  C N M R spectra were  obtained on both the M W L and R E L since these are the two most different ligninfractions,and that previous research (Holtman and Kadla 2004) has shown that the C E L lignin closely resembles the M W L in functionality and frequency. The results of quantitative  C N M R spectra  acquisition are tabulated in Tables 3.3a, 3.3b, and 3.3c. Both individual moieties and clusters of similar carbon signals are reported and individual integral ranges are shown in Figures 3.5a, 3.5b, 3.5c. Non-acetylated and acetylated M W L spectra were both obtained because of signal 93  overlap, and as such, some structural information can be elucidated by taking the integral difference between the two spectra (Capanema 2004). As detected by thioacidolysis, the C4H-F5H lignin is almost exclusively composed of syringyl units, and as such N M R signals arising from guaiacyl units are not seen in the carbon spectrum. Figure 3.8 contains a legend for the naming of lignin substructures, following the conventions previously described by Capanema et al. (2004). Relevant H M B C portions are shown in Figure 3.8.  The H M Q C spectra obtained were  identical in all segments of the spectrum, both aromatic (lignin) and aliphatic (carbohydrate), and are shown in Appendix B.  Side Chain Moieties 8-0-4 Moieties (Fig. 3.8, A-D) There are many different types of functional groups that can potentially surround any given (3-0-4 ether linkage. N M R is usually able to distinguish between structures with - O H , C H - , -C=0, -O-Aryl, and -O-Alk groups at the a position, dibenzodioxocin, and other side 2  chain moieties (Ralph et al. 2001).  During Kraft pulping the 8-0-4 linkages are primarily  broken, and because of this the study of the variations of this linkage is important in rationalizing pulping efficacy differences between samples. In the N M R spectra of wild-type and C4H-F5H M W L , carbonyl groups above 180 ppm were absent, indicating that if present, non-conjugated C O , a-CO, structure L , structure M (vanillin), and I (Figure 3.8) are undetectable in these samples. The region from 77-72.5 ppm in the A c M W L spectrum was corrected for the amount of 8-1, 0.12/Ar for wild-type and 0.03/Ar for C4H-F5H (Fig. 3.8, H). Since the H M Q C spectrum of A c M W L was not collected we were unable to determine if there were any arylglycerol structures present in the samples (Fig. 3.8, P) as was done for the calculation for other lignin (Capanema et al. 2004). The resulting frequency of a-OH//3-0-4 linkages (Fig. 3.8, A) in the wild-type and C4H-F5H were similar, 0.43/Ar and 94  0.42/Ar. This is surprisingly low, as this is of a similar concentration that was determined for loblolly pine M W L , and hardwoods with a lower guaiacyl content would be expected to have a higher amount of this type of linkage (Capanema et al. 2004). The total amount of jS-O-4 moieties is the combination of subunits A , D, and G (Fig. 3.8). Dibenzodioxocin, subunit G, was not found in either sample. Since dibenzodioxocin linkages also contain a 5-5' linkage that are possible only through guaiacyl-derived linkages, the frequency of G is not expected to be high in aspen lignin. No subunit D was found, as there were no conjugated carbonyl carbons in the M W L spectra between 196 and 193 ppm. A study of wild-type, fahl-2 mutant, and C4H-F5H Arabidopsis demonstrated that compared to wild-type, the fahl-2 mutant contained a higher frequency of dibenzodioxocin units (Fig. 3.8, G) and fewer resinols (Fig. 3.8, E). The transformed lignin used in this experiment contained almost no guaiacyl units and thus no dibenzodioxocin units, which is consistent with the literature (Marita et al. 1999).  Phenylcoumarin,  Pinoresinol/syringoresinol,  and 6-1 Structures (Fig. 3.8, E, F, and H)  Subunits E , F, and H (Fig. 3.8) can be formed from both syringyl and guaiacyl lignin monomers. These subunits were not found in abundance, but are present. Phenylcoumaran is present 0.05/Ar for wild-type and 0.06/Ar for C4H-F5H based on the integration of 88-86 ppm in the A c M W L spectra. Pinoresinol/syringoresinol (F) were found with a frequency of 0.11/Ar for wild-type and 0.06/Ar for C4H-F5H based on the integration of 54-51 ppm corrected for the amount of phenylcoumaran. The B-l structure (H) was found in the wild-type (0.12/Ar) and C4H-F5H (0.03/Ar), albeit in small amounts in the modified lignin.  Functional Groups Methoxyl Groups After correction for B-l units, the methoxyl content of the wild-type and C4H-F5H samples was found to be 1.28/Ar and 2.32/Ar, respectively. These results concur with the wet 95  chemistry findings (modified Zeisel method) which showed a methoxyl per C9 unit of 1.08 and 2.20 for wild-type and C4H-F5H, respectively. The calculations from N M R are higher, and are not within reason because even 100% syringyl lignin would have a OMe/Ar of 2.00. suggests that both methods slightly overestimating the methoxyl content.  This  The increase in  methoxyl content in the C4H-F5H lignin clearly results from the up-regulation of the F5H gene as measured by both wet chemistry and spectroscopic methods.  Hydroxyl Groups Phenolic hydroxyl groups (168.8-167.5 ppm AcMWL) were present in similar amounts for both wild-type and C4H-F5H (0.24/Ar). For both primary aliphatic and secondary aliphatic hydroxyl groups, the wild-type was found to be slightly higher than the C4H-F5H lignin (Table 3.3b).  Since many subunits contain hydroxyl groups, it is difficult to infer individual  substructure frequency from these values, since the subunits cannot be distinguished from each other with these groups. The primary hydroxyl groups are calculated from 171.4-169.7 ppm in the acetylated M W L , and were found in higher frequency in the wild-type (Table 3.4b). Secondary hydroxyl frequency was calculated by subtracting the non-acetylated from the acetylated integral values at 169.7-168.8 ppm. This resulted in a frequency of 0.45/Ar and 0.40/Ar for wild-type and C4HF5H, respectively. The total hydroxyl content was also determined by proton NMR, and the results are displayed in the C9 formula (Table 3.2).  Based on this method, the total O H content was  calculated to be much higher for the modified lignin, 1.89/C9 compared to 0.90/C9 for the wildtype. This is quite different than the total O H content calculated from the 1.42/Ar and 1.14/Ar for the wild-type and C4H-F5H, respectively.  1 3  C spectra of M W L ,  These conflicting results  make it difficult to determine if the modification to increase syringyl units of lignin also has an impact on hydroxyl content. 96  Balance of Side Chain Structures The total amount of oxygenated carbon atoms (90-58 ppm) was 2.90/Ar and 2.80/Ar for the wild-type and C4H-F5H M W L , respectively. The non-aromatic carbons are considered to be part of the lignin side chain, with the exception of methoxyl carbons and carbonyls.  This  calculation, shown in Table 3.3c gave 3.37/Ar for wild-type and 2.90/Ar for C4H-F5H. Although the theoretical value for phenylpropane is 3.00/Ar, the value for the wild-type is high; however, this does suggest that perhaps side chain length and monomer type is altered through the up-regulation of the F5H gene.  Aromatic Carbons and Degree of Condensation The quantitative  C N M R spectrum of wild-type M W L has many signals that are  associated with guaiacyl lignin units from 150 ppm to about 110 ppm (Robert 1992). The signals in this region are represented more in the wild-type spectrum (Table 3.4a). The S:G ratio can be calculated from the  1 3  C spectra by taking the integral from one of the guaiacyl structures  and one of the syringyl structures.  For example, the ratio of the signal at 105.3-103.4 ppm  divided by two, since it represents two carbons and the result divided by the signal at 110.7109.7 ppm is essentially taking the ratio of S2/6 and G2. For the wild-type, the S:G ratio calculated is 1.4:1 and for the C4H-F5H, the ratio is 13.7:1. This is very close to the results obtained via the wet chemistry thioacidolysis reaction (Tables 3.1 and 3.3a).  HMBC Spectral Indication of S:G Ratio The two and three bond correlations are particularly useful for determining S:G ratio. Figure 3.7 contains the aromatic region displaying the different chemical shifts of the hydrogen atoms two and three bonds away from the aromatic carbons. The correlations between a protons and B and y carbons on the lignin side chains are between carbons 1, 2, and 6 of the aromatic ring.  The C2/C6 carbons in the syringyl correlation resonate at 105 ppm, and the guaiacyl  C2/C6 carbons resonate at 113 and 120 ppm, which is sufficient for separation. Although most 97  H M B C experiments are not quantifiable, the absence of a correlation may indicate a dramatically lower content of that subunit. From Figure 3.7, it is not possible to determine S:G ratio, as both monomers are represented.  Since the modified lignin is expected to have very little guaiacyl  lignin, it is not consistent with the literature that the guaiacyl signals show up at all in the C4HF5H spectrum (Marita et al. 1999). Since the ratio of S:G is approximately 12:1 for the modified M W L , it is expected that the detection level of the H M B C spectrum should also show the guaiacyl lignin signals. The complete absence would indicate a poor detection limit. The H:S:G ratio from  1 3  C N M R was determined from the h content (0.10/Ar and  0.11/Ar), S content (0.28/Ar and 0.55/Ar from C2/6), and G content (0.21/Ar and 0.04/Ar). This translates to be 1:2.9:2.1 for wild-type and 1:4.95:0.63 for C4H:F5H M W L . calculated from these integrals is 1.38 for wild-type and 7.85 for C4H-F5H.  The S:G ratio However, in  previous studies H lignin was not reported (Huntley et al. 2003). This is substantially lower than that from thioacidolysis. Thioacidolysis measures those lignin monomers that are joined with 80-4 linkages. The N M R results suggest that there are differences in the proportion of syringyl and guaiacyl units that are joined by other types of linkages.  Residual NMR Analysis The recent development of an acetylation technique that facilitates the whole wood to be acetylate was employed with the R E L in order to enable it to dissolve in N M R solvents. The technique worked best when the wood was not only ball milled for a week, but also further pulverized in a vibratory shaker mill for 24 hours. The lignin had a medium brown colouration after the shaking, indicating some change in the lignin had taken place. NMI and D M S O were used to dissolve the lignin, and after 3 hours of reacting with acetic anhydride, the solution turned very dark brown (Lu and Ralph 2003).  After precipitation into aqueous E D T A and  filtration, the resulting acetylated cell wall sample was a red-brown colour, indicating the possible formation of an unidentified chromophore. 98  Unfortunately, the R E L solutions were extremely viscous and needed to be diluted substantially in order to get a good lock on the sample (20 mg in 1.5 mL). This is a substantially lower concentration than is usually used for lignin (60 mg in 0.25 mL). The S/N was not high enough to integrate the spectra and compare them to one another. The spectra are shown in Figure 3.6 in order to demonstrate that it is possible to get good quality R E L , but the acquisition times are long and the solubility of the R E L is low, even after acetylation.  99  3.4  Conclusion The relationship between lignin monomer composition and Kraft pulping efficacy is well  known and accepted.  However, it is clear from this research that there may be other lignin  structural changes other than simply an increase in syringyl content from the up-regulation of the F5H gene that contribute to the altered pulping efficiency. In particular, molecular weight studies have revealed that in the R E L , the molecular weight of the wild-type lignin is much higher than in the C4H-F5H lignin. This reduction in lignin molecular weight is another opportunity to produce wood with targeted characteristics. The trees had normal growth even with these lignin molecular weight changes (Huntley et al. 2003). The alteration of molecular weight due to a change in lignin biosynthesis is the first to be reported. It is likely that the decreased H-factor required to attain the target Kappa is not just due to the increase in syringyl units, but the lower molecular weight as well. These findings with the model, genetically modified substrate, support our findings with the natural aspen clones that suggest molecular weight of the lignin contributes to the efficacy of delignification during chemical pulping. The N M R analysis of the M W L samples did not show much variation, except in methoxyl content, which is expected due to the higher syringyl content of the modified lignin. Since the M W L only makes up approximately one third of the overall lignin, this portion of the lignin may be pulped through very quickly at harsh pulping conditions. The R E L seemed to have carbohydrates associated with the lignin, which led to  1 3  C N M R that were high in S:G in  the aromatic region. The samples of A c R E L for N M R analysis were extremely viscous, and high concentrations were therefore not possible to achieve in the C4H-F5H sample. There is a lot of potential for R E L NMR, as the wild-type spectrum has an acceptable level of signal.  100  3.5  Figures and Tables  hybrid poplar  ball milled wood 1  2  crude lignin  insoluble  4 residual lignin (REL)  7 acetylated milled wood lignin (AcMWL)  7  7 acetylated cellulase lignin (AcCEL)  acetylated residual lignin (AcREL)  Figure 3.1. Isolation, purification, and acetylation of various lignin fractions.  1) 96:4 dioxane:water 2) washed with 3 x DI H 0 3) 90% acetic acid, ppt into water, centrifuge wash 2x with water, 1:2 ethanol: 1,2-dichloroethane ppt into ether wash 2x with ether dry 4) enzyme treatment 3x2 days 5) same purification as for M W L 6) rinsed 3x with water 7) N M I / D M S O / A c 0 acetylation 2  2  101  Table 3.1. Thioacidolysis yield and S:G monomer content for the various isolated lignin samples from wild-type and modified lignin.  Sample  umol g-1 S:G  wood Wild-type 739 MWL CEL REL  1.90  C4H-F5H 895  14.17  Wild-type  1120  1.51  C4H-F5H  1530  12.04  Wild-type 2053 C4H-F5H 1948 Wild-type 848 C4H-F5H 846  2.24 12.37 1.20 8.76  102  Table 3.2. Elemental analysis and expanded C9 formula for M W L for wild-type and C4H-F5H lignin.  MWL Sample Wild-type C4H-F5H  Elemental Analysis C(%) H(%) 0(%) 58.8 5.35 35.6 58.9 5.72 35.1  OMe (wt%) 16.3 29.9  103  OH (wt%) 6.62 14.1  C 9 formula C9H6.89O2.69OMe1.08OH0.80 C9H4.43O0.92OMe2.20OH1.89  —  1.0  Wild-type AcMWL C4H-F5H AcMWL  /7C\  0.8  1\  0.6  0.4  0.2  //  -  \ \ I  0.0 I  .  I  30  .  I  40  50  Elution Time (min) i 100 000 10 000  i 1 000  i 100  Molecular Weight Relative to Polystyrene (g/mol)  Figure 3.2. Gel-permeation chromatograph of acetylated M W L (AcMWL) from wild-type and C4H-F5H poplar  104  Wild-type AcCEL C4H-F5H AcCEL  1.0  0.8 CD  (/}  c o  Q.  0.6  W  0)  a: o o 0  Q  0.4  0.2  0.0 30  40  50  Elution Time (min) 100 000 10 000  1 000  100  Molecular Weight Relative to Polystyrene (g/mol)  Figure 3.3. Gel-permeation chromatograph of acetylated C E L (AcCEL) lignin from wild-type and C4H-F5H poplar.  105  Wild-type AcREL C4H:F5H A c R E L  1.00  2b  30  40  50  Retention Time (min) _L  100 000 10 000  1 000  100  Molecular Weight Relative to Polystyrene (g/mol)  Figure 3.4. Gel-permeation chromatograph for acetylated R E L (AcREL) lignin from wild-type and C4H-F5H.  106  Table 3.3a. Signal assignment in the N M R spectrum of non-acetylated M W L . Peaks correspond to Figure 3.5a.  spectral region aliphatic COOR C-4 in h units C-3 in Tet. C3,5 in Set, C-a in L, C-3,6 in 1, C-4 in conjugated S , unknown C3/C3' in etherified 5-5' units, C a in Ar-CH=CH-CHO units C3/C5 in B-ring of 4-0-5' units, C4 in etherified G units with a-C=0 C3 in etherified G C3 in nonetherified G units, C4 in etherified G units C3 in E, C-4 in conjugated Set C-3 in E, C-4 in T C-4 in conjugated S , unknown C4/C4' in nonetherified 5-5' units, C-3 in B rings of (3-5 units C4 in nonetherified G units,, C4 in B-ring of p-ring of p-5 units C4/C4' in nonetherified 5-5' units C1 in etherified G units C5/C5' in etherified 5-5' units Ca and Cp in Ar-CH=CH-CH OH C6 in G units C5 in G units C2 in G units C2 in G, G in stilbene units C2/6 in S units OMe, C-p in H, C-1 in I, C-y in R C-p in E and F clusters ne>  2  CAr-0 CAr-C CAr-H  Alk-OAlk-O-Ar, a-O-Alk y-O-Alk, OHsec OHprim  107  chemical shift range (ppm) 169.7-168.8 162.6-161.2 155-151  WT (per Ar)  C4H-F5H (per Ar)  Peak Label  0.07 0.10 1.01  0.04 0.11 1.19  1 2 3  155.1-152.4 152.4-150.9  0.47 0.50  0.49 0.65  4 5  150.2-149.5 149.5-148.5 148.5-147.2 145.6-144.9 144.5-142.5  0.06 0.13 0.21 0.07 0.07  -  6  -  7 8 9  147.3-146.6  0.14  0.02  10  146.6-145.7 136.4-134.5 132.0-130.7 128.6-127.5 119.8-117.1 116.2-114.9 112.2-110.7 110.7-109.7 105.3-103.4 58-54 54-51.5  0.09 0.35 0.20 0.08 0.25 0.27 0.21 0.06 0.58 1.50 0.16  -  11 12 13 14 15 16 17 18 19 20 21  160-141 141-125 125-103 90-58 90-77 77-65 65-58  2.02 1.65 2.13 2.84 0.87 1.06 0.90  1.83 1.69 2.26 2.78 0.88 0.96 0.93  0.02 -  0.02  0.29 0.25 0.06 0.08 0.21 0.04 0.02 1.09 2.35 0.12  Table 3.3b. Signal assignment in the N M R spectrum of acetylated M W L . Peaks correspond to Figure 3.5c.  spectral region non-conjugated CO (acetone impurity) primary aliphatic OH secondary aliphatic OH phenolic OH, conjugated COOR all C-3 (except E and h-units), C-5 in S , C-a in L, C-6 in 1, C-4 in conjugated CO/COOR etherified and hunits -  C-3 in E, C-4 in conjugated Set, unknown C-a in E C-a in F, G C-p in G C-a in A, H, P, carbohydrates OMe, C-1,and C-p in I C-p in H, E clusters CAr-H  Alk-OAlk-O-Ar, a-O-Alk y-O-Alk, OH ec S  OH prim  108  chemical shift range (ppm) 207.4-205.1 171.4-169.7 169.7-168.8 168.8-167.5 162-148  WT (per Ar) 0.66 0.52 0.24 1.53  C4H-F5H (per Ar) 0.33 0.45 0.44 0.24 1.70  162.5-160 144.5-142.5 88-86 85.55-84.3 80.84-79.34 77-72.5 58-54 50-48  0.04 0.06 0.05 0.08 0.30 0.55 1.44 0.17  0.14 0.06 0.06 0.08 0.29 0.45 1.27 0.09  125-103 90-58 90-77 77-65 65-58  2.07 2.86 0.91 1.06 0.88  2.01 2.30 0.81 0.83 0.63  -  Peak Label 1 2 3  4 5 6 7 8 9 10 11  Table 3.3c. Lignin substructures determined by NMR. Substructures are shown in Figure  structure B-O-4/a-OH (A) Phenylcoumaran (E) dibenzodioxocin (G) pinoresinol (F) p-1 (H) OMe OHph Alk-O-Ar total etherified y-O-alk maximum a-O-alk minimum a-O-alk side chain degree of condensation  calculation if necessary (77-72.5, ac)-H (88-86, ac) (86-83, ac) - F (54-51)-E (50-48, ac) - E (58-54, na) - H (168.6-166, ac)-(168.6-166, na) A + E + 2xG 1.00-OHph (77-65)-OH (90-77)-Alk-O-Ar (90-77)-(1.00-OHph) (90-52)-OMe+ (175-165, na) theoretical - h units - (125-103, na) sec  109  WT (perAr) 0.43 0.05 0 0.11 0.12 1.28 0.22 0.48 0.78 0.54 0.43 0.09 3.37. 0.17. •  C4H-F5H (perAr) 0.42 0.06 0.02 0.06 0.03 2.32 0.24 0.48 0.76 0.52 0.40 0.12 2.90 0  Wild-type MWL  1  2  121314  15161718 19  20 21  C4H-F5H MWL  1  V  2  Y  Y  W  121314  4V  V  15161718 19  20 21 "i  ppm (t1)  150  100  1  r  50  Figure 3.5a. Quantitative C N M R spectra of non-acetylated M W L . Expanded region is shown in Figure 3.5b. 1 3  110  Wild-type MWL  Figure 3.5b. Quantitative 140 ppm.  C N M R spectra of non-acetylated M W L in the region from 157 to  Ill  C4H-F5H AcMWL  V 1-3 ~i  ppm(t1)  r  4  S-  67 8 9  5 1  1  1  1  150  1  1  1  1  1  10  V 11  r  50  100  Figure 3.5c. Quantitative C N M R spectrum of acetylated M W L . 1 3  112  Wild-type AcREL  C4H-F5H AcREL  H  ppm (t1)  1  -i  [ —  1  100  150  Figure 3.6. Quantitative C N M R spectra of acetylated R E L . 1 3  113  1  1  50  1  r  —100  150  ppm (t1)  6.70 6.60 6.50 6.40 6.30 ppm (t2)  6.70 6.60 6.50 6.40 6.30 ppm (t2)  Figure 3.7. H M B C of acetylated cell wall wild-type and C4H-F5H sample. This spectral region shows the correlations of the oc-protons of P-aryl ether units in lignin. The a-proton correlate with aromatic carbons 1, 2, and 6 and side chain p and y carbons. Guaiacyl and syringyl carbons have slightly different C2 and C6 resonance frequencies therefore can be distinguished by NMR.  114  115  3.6  Bibliography  Bjorkman, A . (1956). "Studies on Finely Divided Wood Part 1. Extraction of Lignin with Neutral Solvents." Svensk Papperstidning 13: 477-485. Browning, B. L . (1967). Acetyl and Methoxyl Groups. Methods in Wood Chemistry. B. L . Browning. Appleton, Wisconsin. II: 660-664. Capanema, E . A . , M . Y . Balakshin and J. F. Kadla (2004). "A comprehensive approach for quantitative lignin characterization by N M R spectroscopy." Journal of Agricultural and Food Chemistry 52(7): 1850-1860. Chang, H.-m. and K. V . Sarkanen (1973). "Species Variation in Lignin: Effects of Species on the Rate of Kraft Delignification." Tappi Journal 56(3): 132-134. Chen, C. L. (1998). Characterization of milled wood lignins and dehydrogenated polymerates from monolignols by carbon-13 N M R spectroscopy. A C S Symposium Series 697. N. G. Lewis and S. Sarkanan. Washington, DC, American Chemical Society: 255-275. Claridge, T. D. W. (1999). High-Resolution N M R Techniques in Organic Chemistry. Oxford, Pergamon. Franke, R., C. McMichael, K. Meyer, A . M . Shirley, J. C. Cusumano and C. Chappie (2000). "Modified lignin in tobacco and poplar plants over-expressing the Arabidopsis gene encoding ferulate 5-hyroxylase." The Plant Journal 22(3): 222-234. Froass, P. M . , A . J. Ragauskas and J.-e. Jiang (1998). "Nuclear Magnetic Resonance Studies. 4. Analysis of Residual Lignin after Kraft Pulping." Industrial and Engineering Chemistry Research 37: 3388-3394. Holtman, K. and J. Kadla (2004). "Solution-state Nuclear Magnetic Resonance Study of the Similarities between Milled Wood Lignin and Cellulolytic Enzyme Lignin." Journal of Agricultural and Food Chemistry 52(4): 720-726. Huntley, S. K., D. Ellis, M . Gilbert, C. Chappie and S. D. Mansfield (2003). "Significant Increases in Pulping Efficiency in C4H-F5H-Transformed Poplars: Improved Chemical Savings and Reduced Environmental Toxins." Journal of Agricultural and Food Chemistry 51: 6178-6183. Ikeda, T., K. Holtman, J. Kadla, H.-m. Chang and H. Jameel (2002). "Studies on the Effect of Ball Milling on Lignin Structure Using a Modified D F R C Method." Journal of Agricultural and Food Chemistry 50: 129-135. Kim, H., J. Ralph, F. C. Lu, S. A . Ralph, A. M . Boudet, J. J. MacKay, R. R. Sederoff, T. Ito, S. Kawai, H . Ohashi and T. Higuchi (2003). "NMR analysis of lignins in CAD-deficient plants. Part 1. Incorporation of hydroxycinnamaldehydes and hydroxybenzaldehydes into lignins." Organic & Biomolecular Chemistry 1(2): 268-281. Lu, F. and J. Ralph (2003). "Non-degrative dissolution and acetylation of ball-milled plant cell walls: high resolution solution state NMR." The Plant Journal 35: 535-544. 116  Marita, J. M . , J. Ralph, R. D. Hatfield and C. Chappie (1999). "NMR characterization of lignins in Arabidopsis altered in the activity of ferulate 5-hydroxylase." Proceedings of the National Academy of Sciences of the United States of America 96(22): 12328-12332. Meyer, K., A. M . Shirle, J. C. Cusomano, D. A. Bell-Lelong and C. Chappie (1998). "Lignin monomer composition is determined by the expression of a cytochrome P450-dependent monooxygenase in Arabidopsis." Proceedings of the National Academy of Sciences of the United States of America 95: 6619-6623. Ralph, J., R. D. Hatfield, J. Piquemal, N . Yahiaoui, M . Pean, C. Lapierre and A. M . Boudet (1998). "NMR characterization of altered lignins extracted from tobacco plants downregulated for lignification enzymes cinnamyl- alcohol dehydrogenase and cinnamoylCoA reductase." Proceedings of the National Academy of Sciences of the United States of America 95(22): 12803-12808. Ralph, J., S. A. Ralph and L. L . Landucci (2001). N M R database of lignin and cell wall model compounds. 2004. Robert, D. (1992). Carbon-13 Nuclear Magnetic Resonance Spectrometry. Methods in Lignin Chemistry. S. Y . Lin and C. W. Dence. New York, Springer-Verlag: 578. Rolando, C , B. Monties and C. Lapierre (1992). Thioacidolysis. Methods in Lignin Chemistry. S. D. Lin, C.W. Berlin, Springer-Verlag. Yang, R., L . Lucia, A . J. Ragauskas and H . J A M E E L (2002). "Oxygen Degradation and Spectroscopic Characterization of Hardwood Kraft Lignin." Industrial and Engineering Chemistry Research 41: 5941-5948.  117  Chapter 4 4.1  Conclusions and Future Work The results of this research indicate that the chemical structure and molecular weight of  lignin varies significantly among natural clones and genetically modified wood. Furthermore, it suggests that the R E L has a greater impact on the chemical properties and efficiency (yield and residual lignin) of chemical pulping, while the M W L or C E L are not as influential. Since the pulping conditions used in this study were fairly harsh (H-factor of 1100 for aspen and H-factors 800-1400 for hybrid poplar) the delignifying chemicals attack the more soluble lignin first, and subsequently proceed until substantive delignification occurs; it is likely that the recalcitrant lignin resembles the R E L lignin that is produced during ball milling. Thus, the Kraft pulp lignin is likely to be most similar in structure to REL, and this lignin should be extracted and analyzed from the wood/pulp to determine if any inferences can be made relating lignin structure and functionality to processing efficiency. The results of quantitative  1 3  C N M R indicate that although changes are evident between  samples, much of the lignin is very similar. Many of the lignin subunits and carbon types were identified and quantified using calculations from the literature (Robert 1992; Marita et al. 1999; Capanema et al. 2004). Wet chemistry techniques give partial chemical structural information, and are usually not time consuming nor expensive. N M R analysis is very much more powerful as it measures all the carbon nuclei in a sample, not just those that are liberated by the variety of wet chemistry methods. This is the first time that a difference in the molecular weight of the lignin has been identified in a genetically modified sample.  These findings indicate that by genetically  engineering trees for altered levels of (increased) mol % syringyl units, the molecular weight of the resultant lignin is reduced. This suggests that the shift to produce more syringyl units affects more than simply the monomer ratio. A possible explanation for this is that the syringyl units 118  are limited with respect to polymerization at carbon 5, which affects the degree of polymerization of the lignin. This study paves the way for other studies directed at changing molecular weight via alteration in the lignin biosynthetic pathway. Understanding and being able to predict the natural variation in wood is key to maintaining global competitiveness of the Canadian industry in the pulp and paper marketplace. Aspen and the hybrid poplars represent a huge untapped natural resource in Canada, and can potentially yield high quality targeted characteristic pulps, if the natural clones are well characterized. Understanding how wood chemistry, especially lignin chemistry affects pulping characteristics, is a step towards being able to fully utilize our largely untapped, genetically diverse poplar inventory. There is also much support for vigorous research into genetically modified poplars to meet the demands for wood products and environmental protection (Strauss et al. 2001). In addition, the ecological effects of genetically modified trees are thought to be similar to those of conventional plantations and, as long as regulatory practices are upheld should pose few, if any, problems (Strauss et al. 2001). Even if genetically modified trees are not utilized in plantation forestry in the near future, these innovative tools will elucidate several important, previously unknown facts about wood structure, its function, its biosynthesis, and as such may direct researchers to fully utilize the variety of material that naturally exists among our native population. Additional future work should include attempting to generate R E L samples with higher lignin content for both the wild-type and C4H-F5H poplar in order to improve the signal to noise ratio for comprehensive N M R analyses.  The carbohydrates in the R E L samples were not  quantified in this study due to sample size constraints, and as such it would be useful to determine what carbohydrates and how much are associated with the R E L fraction. Since the REL  1 3  C N M R results from aspen samples 10-1 and 16-2 were able to give useful structural 119  information, the key to understanding more about lignin chemistry may be in analyzing the R E L in addition to the M W L . As identified, the residual Kraft pulp lignin is likely similar in structure to R E L , and a comparison of lignin structures between the two validating the assumption that the R E L fraction has a greater total affect on the pulping process that M W L and C E L is warranted in the future. If the lignin remaining in Kraft pulp were similar in structure to the isolated R E L fractions, this would offer some insight as to what characteristics should be assessed in selecting trees from natural populations, or what characteristics should be targeted for modification in order to improve pulping efficacy. The apparent relationship between molecular weight of lignin fractions to pulp yield and Kappa number may extend to species other than Poplar. A study to confirm this relationship should be conducted with other hardwood populations, and extended to softwood species. For example, trees could be examined that have similar genotypes and different pulping characteristics, and the molecular weight of extracted lignin measured.  Additionally, clonal  hybrids from common garden plots or different geographic regions should be studied in order to further support this finding. Genetically modified samples should also be analyzed with respect to lignin molecular weight. For example, trees with reduced C A D activity (Pilate et al. 2002) have been shown to exhibit superior pulping performance, and the lignin from such trees, which display the improved pulping efficiencies should be extracted and the molecular weight determination quantified in an attempt to validate the presented theory.  120  4.2  Bibliography  Capanema, E . A . , M . Y . Balakshin and J. F. Kadla (2004). "A comprehensive approach for quantitative lignin characterization by N M R spectroscopy." Journal of Agricultural and Food Chemistry 52(7): 1850-1860. Marita, J. M . , J. Ralph, R. D. Hatfield and C. Chappie (1999). "NMR characterization of lignins in Arabidopsis altered in the activity of ferulate 5-hydroxylase." Proceedings of the National Academy of Sciences of the United States of America 96(22): 12328-12332. Pilate, G., E . Guiney, K. Holt, M . Petit-Conil, C. Lapierre, J. C. Leple, B. Pollet, I. Mila, E . A . Webster, H . G. Marstorp, D. W. Hopkins, L. Jouanin, W. Boerjan, W. Schuch, D. Cornu and C. Halpin (2002). "Field and pulping performances of transgenic trees with altered lignification." Nature Biotechnology 20(6): 607-612. Robert, D. (1992). Carbon-13 Nuclear Magnetic Resonance Spectrometry. Methods in Lignin Chemistry. S. Y . Lin and C. W. Dence. New York, Springer-Verlag: 578. Strauss, S. H . , S. P. DiFazio and R. Meilan (2001). "Genetically modified poplars in context." The Forestry Chronicle 77(2): 271-279.  121  Appendix A Supplementary Tables and Figures for Chapter 2  Table A l . Kraft pulp yield calculated by dividing grams dry pulp by grams dry wood chips, each pulp was performed at least four times. Kappa number is measure of the residual lignin in pulp, each pulp was determined in triplicate and the results averaged across all samples. Error is standard deviation. Sample 10-1 16-2 18-3 19-5 21-5 26-1  Pulp 56.9 61.3 59.2 61.4 57.5 59.8  Yield ±3.1 ±0.1 ± 0.7 ±0.2 ±0.7 ±0.1  Kappa 22.2 16.6 20.9 17.1 25.0 19.9  Number ± 2.3 ± 1.4 ± 0.6 ±0.8 ± 1.0 ±0.3  122  Table A 2 . Elemental Analysis Data. Nitrogen has been omitted. S.D. is standard deviation based on at least three samples. These values were used to calculated M W L "C9" formula values. Ball Milled Wood Sample 10-1 B M W 16-2 BMW 18-3 B M W 19-5 BMW 21-5 BMW 26-1 B M W  %C 46.1 46.1 46.4 46.5 45.2 46.0  s.d. 0.9 0.9 0.4 1.1 1.2 0.2  % H 7.5 7.4 7.7 6.9 7.3 7.5  s.d. 0.3 0.4 0.1 0.1 0.4 0.2  %0 46.4 46.1 45.9 47.2 47.4 46.5  s.d. 1.1 0.9 0.5 1.1 1.6 0.1  s.d. 1.2 0.9 0.9 0.7 1.4 0.8  % H 6.4 6.9 6.1 6.9 5.5 6.9  s.d. 0.3 0.4 0.4 0.3 0.4 0.3  %0 34.6 34.4 35.8 34.2 35.5 35.1  s.d. 1.4 1.1 1.1 0.7 1.7 1.1  Milled Wood lignin Sample 10-1 M W L 16-2 M W L 18-3 M W L 19-5 M W L 21-5 M W L 26-1 M W L  %C 58.9 58.7 58.1 58.9 59.0 58.0  Cellulase Lignin Sample %C 10-1 C E L 57.6 16-2 C E L 57.3 18-3 C E L 55.7 19-5 C E L 58.2 21-5 C E L 58.0 26-1 C E L 56.1  s.d. 0.6 0.9 0.6 0.6 0.0 1.0  % H 5.9 5.7 6.9 5.6 5.8 6.8  s.d. 0.3 0.2 0.2 0.1 0.6 0.1  %0 36.4 36.9 37.1 36.1 36.9 36.9  s.d. 0.8 1.2 0.7 0.5 0.7 1.1  s.d. 0.5 1.0 0.5 0.7 1.2 6.0  %H 6.9 6.7 6.5 7.0 5.9 6.1  s.d. 0.1 0.1 0.1 0.1 0.4 0.7  %0 40.0 38.6 39.4 40.5 44.2 44.3  s.d. 0.6 1.0 0.5 0.8 1.6 6.7  Residual Lignin Sample 10-1 R E L 16-2 R E L 18-3 R E L 19-5 R E L 21-5 R E L 26-1 R E L  %C 52.8 54.3 53.7 52.2 59.7 49.3  123  10-1 M W L H M Q C 0 0 o  A  ^  a  P.a  F  a°°  o  F  o " q  ©  Ob  w.OO'  |—i—i—i—i—|—i—i—i—i—|—i—i—i—i—|—i—i—i—i—|—i—i—i—i—|—i—r  ppm(t2)  7  -  0  6  0  5  0  4  -°  3  0  Figure A l . H M Q C spectrum of sample 10-1 M W L .  124  2.0  1.0  16-2 MWL HMQC o  Q *>  h-50 OMeA© A O  F °  a  y  p @  °  Moo  o.. 0® c© 0 N  a  6  ppm(t1) T  1  1  1  1  ppm(t2) 7  1 0  1  1  1 6  1 -°  1  1  1  1 -°  1 5  1  1  1  1 -°  1 4  1  1  1 1  1 3  0  Figure A2. H M Q C spectrum of sample 16-2 M W L .  125  1  1  1  1 2  0  1  1  1  1 1  1  -°  1  1  1 1  1  1  Figure A3. H M Q C spectrum of sample 18-3 M W L .  126  0 8  19-5 M W L H M Q C  0  J  V v  8  h50  ° M e J |  F o <o? r  F a t t  8 0 0 0  ^ °c&SS  0  0  e ® 0  •> ®  N  M O O  8  8  9  6  •o. <g5a o  0  8 0  ( I  I  I  ppm (12)  I 7  | 0  I  I  I  I  | 6  0  I  I  1  1  1  1  5.0  1  1  1 1  4.0  11  ppm(t1) 1  |  3.0  Figure A4. H M Q C spectrum of sample 19-5 M W L .  127  1  1  1  1  |  2.0  1  1  1  1  |  1.0  1  1  1 1  21-5 M W L H M Q C  8aS)  0  o  °Q„ o  of  5  h50  F  a®  30>  3  h100 COO  o O  »  ppm (t1) I  I  I  ppm 02)  I 7  I 0  I  I  I  I  I 6  0  I  1  1  |  5.0  1  1  1  1  |  1  1  1  4.0  1  |  3.0  Figure A5. H M Q C spectrum of sample 21-5 M W L .  128  1  1  1  1  |  2.0  1  1  1  1  |  1.0  1  1  1 1  26-1 M W L H M Q C ©  0 Q„ @  t  °a R  po  ^>  U>  0Me^^4^  F  a«  h50  P a  €2b  ppm (t1) i  ppm(t2)  7  0  i  i  i 6  | 0  i  i  i  i 5  | 0  i  i  i  i 4  |—i—i—i—i—|—i—i—i—i—|—i—i—i—i—|—i—i—i—i—| 0  3  -°  Figure A6. H M Q C spectrum of sample 26-1 M W L .  129  2  0  1-0  10-1 MWL HMBC  ®<a®3 gamma ® beta <*0 °o°  0 o  10.0  S2/6 G2 G6 G1 and S1  5.0  ppm (t2)  Figure A7. H M B C spectrum of sample 10-1 M W L .  130  16-2 MWL HMBC  53  ^  gamma  ® beta «*0 S2/6  °; G2 G6 G1/S1  T  PrJmfe)  5.0  Figure A8. H M B C spectrum of sample 16-2 M W L .  131  18-3 MWL HMBC ®  D O  0 V  •» eia gamma • beta „  G  S2/6 SO QO q. ^-J^-  G  G6 ®#> G1/S1  o  0  1 10.0 ppm (t2)  5.0  Figure A8. H M B C spectrum of sample 18-3 M W L .  132  0  19-5 MWL HMBC  0  h50  •• gamma • beta M O O  * S2/6 G2 G6 * G1/S1 o o  hl50  <©>  o  -200 ppm 10.0  5.0  ppm (f2)  Figure A9. H M B C spectrum of sample 19-5 M W L .  133  26-1 MWL HMBC o o  0  h50 *  gamma ® beta  0 0  Moo  •<*> S2/6 r G2 °-®> G6 G1/S1  ©o  M50 < 3 >o o  -200 ppm (t1) 10.0 ppm (t2)  5.0  Figure A l l . H M B C spectrum of sample 26-1 M W L .  134  Appendix B Supplementary Tables and Figures for Chapter 3 Wild-type Acetylated Cell Wall HMQC  <  * "  ^  s  -  -50  O M e g ^ ^  0  -  -100  a <© ©>  ,  J>pm (t1)  1 1 1 1 [ ' 1 1 1 | 1 1 1 1 | 1 1 1 1 | 1 1 1 1 | 1 1 i1 p p m ( t 2 )  7.0  6.0  4.0  5.0  3.0  €  C4H-F5H Acetylated Cell Wall HMQC  K  ©  -50  O M e g ^ A  °  °  -  A.  F cr>  -  p  -100  0 0 J>pm (t1) < < < 1 1 : . 1 1 1 1 1 1 1 : 1 1 1 1 , < < , 1 , , , , ppm(t2)  7  0  6  0  5  -°  4  -°  3  0  Figure B l . H M Q C spectra of wild-type and C4H-F5H acetylated cell wall. 135  Wild-type Acetylated Cell Wall HMBC  °  p  -50 -  ©=>  o  -100  o  -150  o  •  T#«>(t1) I  i  i  t  ppm(t2)  | 7  I  I  1  0  1 6  |  1  1  1  0  1 5  j  1  1  1  1  0  4  |  1  1  1  0  1  1 3.0  1  2.0  1.0  C4H-F5H Acetylated Cell Wall HMBC  -50  <£  •  &  -  -100  O  CD  @  @  -150  -ppW(f1 I  I  I  ppm(f2)  I  | 7  0  1  1  1  1 6  | 0  1  1  1  I 5  | 0  I  I  I  I 4  -°  j  I  I  I  ' 1  3.0  1  I  2.0  I  I  I  I  I  1.0  Figure B2. H M B C spectra for wild-type and C4H-F5H acetylated cell wall samples. H M B C spectra show 2 and 3 bond couplings between C and H . Peaks are labelled in Figure 3.7.  136  Table B l . Elemental composition of hybrid poplar wood and lignin fractions. Standard deviation of triplicate samples is labelled as s.d.  WT BMW F5H B M W W T MWL F5HMWL WT CEL F5HCEL WT REL F5H R E L  %C 45.8 46.8 58.8 58.9 60.2 58.4 55.0 52.6  s.d. 0.5 0.7 0.0 0.6 2.8 1.4 0.3 0.7  %H 5.76 5.69 5.35 5.72 5.24 6.27 5.98 5.54  137  s.d. 0.22 0.01 0.25 0.10 0.26 0.31 0.07 0.33  %0 48.0 47.1 35.6 35.1 34.3 35.0 38.2 40.9  s.d. 0.7 0.7 0.4 0.7 2.7 1.4 0.4 0.7  Appendix C - Detailed Materials and Methods C.1  Harvesting Aspen Clones Natural aspen clones were obtained from a common site near Fort Nelson British  Columbia, Canada in September 2001.  The clones were distinguished based on bark colour,  bark markings, branch angle, timing of bud flush and leaf shape. Age, fibre length, size, and degree of decay were the bases for selection of certain trees within each line of clones. The trees in this study ranged in age from 95 to 124 years old. The entire butt log (lower 2 metres) was chipped and used for analysis. The following pictures are of the bark of the six clones.  10-1  19-5  16-2  21-5  138  18-3  26-1  C.2  Modification of C4H-F5H Hybrid Poplar The generation of nine lines of C4H-F5H transgenic hybrid poplar (P. tremula * P. alba)  involved generating over 40 transgenic lines by Agrobacterium-mediated transformation (Rochus, 2000). The C4H-F5H transcriptional fusion construct was generated by using a 2,897bp fragment of the C4H promoter and a 2,719-bp fragment of the F5H genomic sequence fused 50-bp upstream of the inferred F5H A T G start codon. As a result, the C4H promoter drives the expression of the F5H gene using the C4H transcription start site and the termination signal present on the F5H genomic sequence. A small fragment of pGEM-7Zf(+) (Promega) polylinker sequence remains in this construct at the C4H:F5H fusion junction. This expression cassette was inserted into the T - D N A of the binary vector pGA482 to give pGA482-C4H:F5H.  The  generation of the pGA482-35S:F5H construct has been described (Meyer et al. 1998). Plant transformation constructs were introduced into Agrobacterium tumefaciens C58 pGV3850 by electroporation. Stability of constructs was confirmed by restriction analysis of plasmid D N A isolated from A. tumefaciens. Cultures harbouring the binary vectors were used to transform the fahl-2 mutant by vacuum infiltration. Kanamycin-resistant seedlings derived from independent infiltration experiments were grown in soil and permitted to set seed. Plants from seed stocks that segregated 3:1 for kanamycin-resistant progeny were again permitted to set seed, and homozygous transgenic lines were reselected in the next generation (Meyer et al. 1998). These, together with wild-type control plants, were maintained as shoot cultures on MS medium with a 16 hour photoperiod. Shoots were multiplied from each line by excising nodal segments and allowing axillary buds to elongate. Prior to planting into the greenhouse, 5-8 cm long tips from actively growing shoots were excised and placed on MS medium supplemented with 0.01 u M a-naphthaleneacetic acid (NAA) for 2 weeks to initiate root formation. Shoots were then transplanted directly into potting soil, acclimated for 2 weeks in a high-humidity environment, and grown and transplanted into successively larger pots over the next 2 years. 139  One-year-old plants were harvested prior to dormancy in the fall. Two-year-old plants were top-pruned at 4 ft above the soil and allowed to over-winter in a non-heated greenhouse. This two-year-old material was then harvested ~2 months after flushing in the spring. Wood designated as 2 years old was, therefore, the bottom 4 ft section of each 2-year-old tree. Leaves and bark were removed from the harvested stems, and these stems were then left to air-dry at ambient temperatures in the laboratory (Huntley et al. 2003).  C.3  Determination of Moisture Content Glass beakers or aluminium weighing dishes were dried overnight in a 105°C oven  (Thelco, Precision Scientific Co.), placed in a desiccator to cool, and weighed on an analytical balance to obtain value A in the formula below. The sample (at ambient moisture) was added to the weighing dish and the new mass recorded as B.  The sample was then put in the oven  overnight, cooled in a desiccator, and weighed again giving value C. The masses were used to calculate the percent moisture in the sample via the following equation. Percent Moisture = 100 [(B-A) - (C-A)]/(B-A) The moisture content values were frequently used to back-calculate the dry mass of a sample used for other analyses.  C.4  Quantitative Determination of Extractives Total extractives were quantified for each sample via a modified TAPPI Method T 280  pm-99. In short, wood was ground in a Wiley mill through a 30 or 40 mesh screen the exact dry weight recorded (in extraction thimble). Each thimble was arranged in a Soxhlet extraction apparatus connected to a flat-bottomed round bottom flask containing 150 mL of acetone. Glass spherical boiling chips were added to the flask in order to prevent bumping. The wood was extracted for approximately 24 hours, after which time the extractives solution was reduced to 140  less than 100 mL using a Btichi R E III (Switzerland) rotary evaporator. The extractives were transferred into a 100 mL volumetric flask and made up to the mark with acetone. Once made up to the mark, 20 mL was transferred by pipette into an oven dry glass scintillation vial, which was then warmed slightly on a hotplate and simultaneously blown down with a light stream of nitrogen gas.  These vials were then oven dried in a 105°C oven overnight, cooled, and the  extractive mass was determined gravimetrically. Quantification of extractives was therefore achieved taking into account the mass difference of vial and vial with dried extractives, the dilution, and the starting dry mass of wood that was extracted. One must keep in mind that this will method not quantify any volatile extractives, such as mono-terpenes, which can be isolated by steam distillation (Fengel and Wegener 1989). Gas chromatography (GC) samples were prepared in order to quantify the classifications of wood extractives.  Approximately 5 mL solution was evaporated with a rotary evaporator  using a 50°C water bath. Once the sample appeared neat, the flask was removed from the heat and further evacuated to remove any water molecules present in the extractives. This was then dissolved in 1 mL of acetone and filtered through a Kimwipe plugged glass disposable pipette into a G C vial. In order to classify the extractives into the main components (fatty acids, resin acids, sterols, steryl esters, waxes, triglycerides) response factors were determined for representative compounds from these categories. Tetracosane (Aldrich), betulin (Sigma), cholesteryl palmitate (Sigma) and palmitin (Aldrich) were used as standards for fatty acids and resin acids, sterols and waxes, steryl esters, and triglycerides, respectively.  Five standard solutions ranging from  approximately 1 mg mL" to 0.0625 mg mL"' were analyzed by G C under similar conditions. 1  Response factors were determined for each class of compounds as a ratio of peak area to concentration of the sample.  141  A Hewlett Packard (HP) 5890 Series II G C fitted with a HP 6890 Series injector and a 10m D B - X L B column (J&W Scientific) was used for the analysis. The G C method employed an injection volume of 1 \xL of sample and Helium as a carrier gas (1 mL/ min" ) The injector 1  port was set to 320°C and the flame ionization detector (FID) was held at 330°C for the duration of the run. The initial oven temperature was 50°C, where it was held for 3 minutes and then ramped at 10°C per minute until it reached 240°C, held for 3 minutes, increased to 310°C at the same rate, and finally increased at the same rate to 350°C. The temperature was held at this final temperature for 25 minutes. .  C.5  Determination of Methoxyl Content Dried, extractive free wood sample (100 mg) or purified lignin (40 mg) was accurately  weighed into a gelatine capsule. A solution of 2.00 mL propionic anhydride (Aldrich) and 6.00 mL hydroiodic acid (55%, Sigma-Aldrich) was placed in a round bottomed reaction flask fitted with a condenser.  Nitrogen gas was bubbled through the reaction flask and up through the  condenser through a trap filled with an aqueous solution of red phosphorous (Sigma-Aldrich, 0.06 g per 100 mL H2O). The trap was connected to a receiver filled with approximately 15 mL of bromine & potassium acetate solution (Aldrich, 5 mL in 145 mL). The sample was added to the reaction flask which was subsequently heated in an oil bath at 150 °C for 40 minutes; the contents of the receiver were then emptied into a 250 mL Erlenmeyer flask containing 10 mL aqueous sodium acetate solution (220 g L" ). The receiver 1  was washed with water to ensure all of the contents were transferred to the flask. Deionized water was added to bring the contents to approximately 125 mL, and formic acid (90%, Fisher) was added to discharge the bromine. Formic acid was added until the solution turned colourless, and 6 extra drops were added. Approximately 3 g of KI was added 3 minutes later, along with 15 mL of dilute (1.8 M) sulphuric acid (Fisher Scientific). 142  The solution was titrated with a  0.1026 N standardized solution of sodium thiosulfate (Aldrich) using 1% soluble starch (Eastman Kodak Co.) solution as the indicator. The endpoint was colourless and clear. A blank determination was made with only a gelatine capsule. The methoxyl content of wood or lignin was calculated using the following equation. % methoxyl = [(A-B)N x 0.00517 x 100]/dry weight sample in grams where, A = endpoint for the sample B = endpoint thiosulfate for blank N = normality of titrant solution (sodium thiosulfate) This method is a modified Zeisel method (Browning 1967).  C.6  Determination of Monolignol Composition by Degradation - Thioacidolysis Dioxane (Fisher) used in this procedure was obtained as reagent grade, and further  purified by distillation under nitrogen. Depending on the number of samples analyzed, 50 - 150 mL of reaction solution was made up the day of or the day before the analysis. To make 50 mL of solution, ~10 mL of purified dioxane was poured into a 50.00 mL volumetric flask. To this, 1.25 mL of boron trifluoride diethyl etherate (in a SureSeal bottle, Aldrich) was transferred via syringe to the volumetric flask.  The volume removed from the SureSeal bottle was replaced  with an equal volume of nitrogen.  A pipette was used to transfer 5.00 mL of ethanethiol  (Aldrich) to the flask, and then filled to the mark with purified dioxane. If not used immediately, the flask was closed with a glass stopper and sealed with Parafilm. Dry samples (wood, lignin, or pulp) were weighed into dry test tubes fitted with Teflon lined caps (10 mg each). To each sample 10.00 mL of prepared solution was added via pipette. The head space was filled with nitrogen gas and the samples were capped and placed in a 100°C stirring glycerine bath. The vials were left for exactly 4 hours, during which time they were 143  removed and shaken approximately every half hour. After the 4 hour reaction time, the vials were removed and placed in an ice-water bath for at least 5 minutes to stop the reaction. A 250 mL separatory funnel was prepared with ~2 mL deionized water and 2.00 mL of tetracosane (Aldrich) standard (~0.25mg mL" ). The contents of the test tube were added to the 1  funnel, and the test tube was rinsed with 2 portions of 10 mL deionized water.  Sodium  bicarbonate solution (0.4 M) was used to adjust the pH of the upper aqueous layer to between 3 and 4, which usually required 1-3 mL of the bicarbonate solution.  Thirty mL of methylene  chloride (CH2CI2) (Fisher) was added to the separatory funnel, and after extraction the organic layer was drained into a 250 mL Erlenmeyer flask. Two more 30 mL portions of CH2CI2 were used to extract the degradation products, and the organic layers were combined. The organic solution was dried over anhydrous sodium sulphate for at least 10 min. After this time, the dried solution was gravity filtered through a phase separation filter paper (Whatman IPS).  The  Erlenmeyer was rinsed with two portions of 10 mL C H 2 C I 2 , which was also filtered, and finally 30 mL of CH2CI2 was used for the final rinse of drying agent and filter paper. The organic solution was then reduced via rotary evaporation at 50°C. Once only 3 or 4 mL of the solution remained, a small amount of methanol (Fisher Scientific) was added to co-evaporate the solvent to dryness. The product was re-dissolved in 2.0 mL CH2CI2 and transferred to a storage vial.  If stored, the headspace was filled with  nitrogen and the caps used were lined with Teflon.  The samples were then stored in a dark  refrigerator and analyzed in less than one month to ensure minimal degradation of analyte. Gas chromatography samples were made by combining 10 uL pyridine, 10 uL sample, and 50 uL of the silylating agent N , 0 bis(trimethylsilyl)acetamide (Sigma) into a GC vial insert. The sample was capped and run on a DB-5 column wall-coated open tubular column. Analysis was performed on the same G C as extractive analysis, but with a 15 m x 0.25 um DB-5 column (J&W Scientific). The sample was injected at 250°C and detector at 270°C. 144  The initial oven temperature was set at 130°C and held for 3 minutes. The oven temperature was then ramped at 3°C per minute and held at 260 °C for 5 minutes. One ul of samples was injected and carried through the column using Helium at 1 mL min" . 1  The response factors for both  syringyl and guaiacyl components were previously determined to be 1.5 (Rolando et al. 1992).  C.7 Determination of Monolignol Composition by Degradation - Nitrobenzene Oxidation Accurately weighed dry extractive free wood (200 mg) or lignin (50 mg) was weighted into a pyrex test tube. To this, 7 mL of 2 M NaOH (diluted from 10 N NaOH, B D H Inc.) was added along with 0.4 mL nitrobenzene (Fisher Scientific). A Teflon cap and Teflon tape were used to seal the tube and it was placed into a 170 °C oil bath for 2.5 hours, and occasionally shaken during this reaction period.  The test tubes were then cooled in an ice water bath for at  least 5 min. The internal standard, 3-ethoxy-4-methoxy-benzaldehyde (200 uL of 14.09 mg mL"  1  in 2 M NaOH) was added and the tube was shaken to mix the internal standard with the contents, and then transferred to a separatory funnel. Chloroform was used to remove any nitrobenzene reduction products (3 x ~ 25mL). The aqueous layer was acidified with concentrated HC1 to a pH of 2 and transferred to a liquid-liquid extractor to be extracted with CHCI3 for 48 hours. After this time, solvent was reduced by rotary evaporation to dryness, and the sample was redissolved in 5 mL C H 2 C I 2 and made up to 10.00 mL in a volumetric flask. G C analysis was performed on stock solution that had been filtered through a Whatman polyvinylidene fluoride 0.45 u.m filter. The temperature profile for G C M S analysis began with 1 min. at 60°C, a 10°C per minute ramp from 60 to 250°C, a 10 minute hold at 255°C, and the profile ended at 260°C. The MS detector had a source temperature of 280°C and a 300°C transfer line. The EI mode was run at 70eV. The column was a DB-5, as for thioacidolysis and the carrier gas was helium at 1 mL min" . 1  145  C.8  Carbohydrate and Acid Soluble & Insoluble Lignin Analysis Oven dried wood was weighed into small glass reaction cups (200 mg accurately  weighed).  To this, 3.00 mL 72%  H2SO4  was added to each sample in three minute intervals.  Following acid addition, the samples were mashed with glass rods to fully submerge the sample in the acid. The samples were stirred every 10 minutes for 2 hours. After 2 hours the samples were transferred into 250 mL serum vials with previously weighed 115.0 mL D l water. The reaction flasks were rinsed well with a portion of this water.  Once all the contents were  transferred to the serum bottles they were sealed with a rubber septum and metal crimp cap. The bottles were autoclaved at 121 °C for 1 hour to hydrolyze the polymeric sugars. The samples were allowed to cool and removed from the autoclave.  If necessary, the serum bottles were  stored in the fridge overnight. The next day, the samples were filtered through medium porosity glass frit crucibles of known dry weight (previously dried overnight at 105°C). The insoluble material (Klason acid insoluble lignin) was rinsed with warm D l water and allowed to dry in a 105°C oven overnight. Acid insoluble lignin was determined gravimetrically from the difference in mass of the crucible and lignin and dry crucible masses. The filtrate was removed before rinsing in order to maintain the concentration of soluble lignin and sugars for further analysis. The filtrate was stored in the refrigerator until analysis via ultraviolet optical spectroscopy (205 nm) and HPLC for acid soluble and sugar analysis, respectively. High, medium, and low sugar concentration standards were subjected to the same autoclaving as the samples, to estimate the loss via decomposition during the heated reaction.  The standards  contained L-(+)-arabinose, D-(+)-galactose, D-(+)-glucose, D-xylose, and D-(+)-mannose (all Sigma) and varied from 1 mg mL" to 0.0625 mg mL" . 1  1  146  Chromatography samples were prepared with L-fucose (Sigma) as the internal standard, and filtered through a 0.45 um polyvinylidene fluoride filter (Whatman). Determination of sugar monomer content after hydrolysis was achieved with a DX-600 BioLC high performance liquid chromatograph. After elution through a CarboPac PA1 column, analytes were detected using pulsed amperometry (gold electrode) after the addition of 200 m M NaOH wash. Each run was 60 minutes in length, the first 40 minutes the eluent was 100% nanopure water, minutes 40-50 was 250 m M NaOH, and the last 10 minutes were again water.  Each sample (20uL) was  injected and eluted with a flow rate of 1 mL min" . 1  Acid soluble lignin was determined spectroscopically by diluting the filtrate sample so that it absorbed between 0.2 and 0.7 absorbance units at 205 nm (Tappi Useful Method U M 250). Each sample was then analyzed and the soluble lignin concentrated with Beer's law. The blank used for calibration of the spectrometer was 4% H2SO4. The formula used to calculate acid soluble lignin was acid-soluble lignin % = B-vlOO/1000-w where, B = Absorbance • volume of diluted filtrate / 110 • volume of original filtrate v = Total volume of filtrate w = Oven-dry weight of sample This method is a modified version of the classical method, to allow 0.2 g sample to be of sufficient quantity (Dence 1992). The Klason method for determination of lignin was used to determine the carbohydrate content of various lignin preparations. In this case, a smaller portion of sample was used to conserve sample.  147  C.9  Kraft Pulping Trials Small scale pulping was performed with liquor containing 25% sulfidity, 13% effective  alkali and at a 4.5:1 liquor to wood ratio. This correlated to a 2.00 L aqueous solution of 65.88 g NaOH (pellets from B D H , Inc.) and 64.08 g Na S-9H 0 (Fisher). 2  2  Wood chips were efficiently packed into small stainless steel reactors to ensure pulp liquor would adequately cover the sample. The reactors were placed into an oil bath and the temperature was ramped from room temperature to 170°C over approximately one hour. After the cooking to an H-factor of 1100, the reactors were removed from the oil bath and placed in a large container of water to cool to working temperature, about 15 minutes.  Once cooled, the  cooked wood chips were washed with warm tap water over a metal screen. The rinsed chips were placed in standard British disintegrator for 15 minutes, after which time the frothy mixture of pulp and water was filtered through a Buchner funnel lined with a plastic mesh liner. The filtered pulp was washed until the filtrate was colourless and dried at least 2 days in a 50°C oven. The pulp was allowed to come to ambient moisture content and weight was determined. Dry weight was calculated by determining moisture content, which was used in subsequent analyses.  C.10  Determination of Pulp Residual Lignin Kappa number was determined according to TAPPI Useful Method U M 246. This is a  micro Kappa number, which allows very small masses of pulp to be analyzed. To start, the pulp sample (ambient moisture content) was peeled into fine pieces, and any visible shives were removed. The sample was weighed on an analytical balance, and placed into a small Waring blender with a blunt propeller. Deionized water was added (-50 mL) and the pulp was ground until homogenous. The mixture was then quantitatively transferred into a 250 mL beaker. The sample was stirred with a one inch magnetic stirrer throughout the duration of the reaction and titration. A combined solution of 10.00 mL 4 N sulphuric acid and 10.00 mL 0.987 N K M n 0 148  4  (Aldrich) was added to the pulp and the reaction was timed from this point.  At 5 min, the  temperature was recorded to the nearest half degree Celsius. The target temperature was 25°C, which was accomplished by using slightly warmed deionized water and heat transfer from the magnetic stirrer (motor was warm). Generally, the reaction temperature was between 24 and 25°C.  After exactly 10 min, 2 mL of 1 N KI solution was added to halt the reaction. At this  point the pulp mixture was titrated with 0.1 N sodium thiosulfate (VWR) until pale yellow. A few drops of 1% soluble starch solution were added as an indicator to adequately determine the endpoint of the titration. The endpoint was the disappearance of the dark blue starch colour to leave a white cloudy solution. The amount of pulp used in the reaction was altered so that between 30% and 60% of the permanganate was consumed in the reaction. In order to calculate this, a blank titration with no pulp was performed. Correction factors were used for any mass that did not consume 50% of the permanganate. In addition, correction factors for temperature were used to account for different rates of the reaction of lignin with permanganate.  149  In order to obtain the millilitres of 0.1 N KMn04 that reacts with 1 gram of moisture free pulp, the following calculation was carried out: K = [(p x f)/w][l + 0.0013(25-t)] P = (b-a)N 2S204/NKMn04 Na  where f = correction factor to 50% permanganate consumption (Tappi). w = dry weight of sample t = temperature in °C after 5 min. of reaction time p = amount of permanganate solution actually consumed by the sample b = endpoint of the blank titration (volume thiosulfate) a = endpoint of sample titration (volume thiosulfate) NN 2S204 3  NKMII04  = normality of thiosulfate solution  - normality of permanganate solution  Incorporating the ratio of reagent concentrations facilitates the conversion of units between thiosulfate and permanganate.  The temperature correction portion of the equation  corrects for deviations from 25°C, which were usually less than 1 degree.  C.11 Ball Milled Wood Ground extractive-free wood was ball milled using a Uni Ball Mill II (Australian Scientific Instruments) set to 85% maximum rotation speed.  The ball mill was fit with two  stainless steel drums containing 5 stainless steel ball bearings. After one week, 400 mL toluene (Fisher) was added to minimize lignin oxidation reactions. Following an additional week of ball milling, the toluene was removed by centrifuging the sample at 4000 rpm for 10 minutes and decanting the solvent. This process generated ball milled wood (BMW). In order to prepare wood samples suitable for dissolution into organic solvents for acetylation (section C.15), a vibratory shaker mill was employed to further pulverize the ball 150  milled wood. Wood was added to a small stainless steel cup with 3 ball bearings. The container with lid was shaken in a mill for 24 hours. This wood is known as shaker mill wood (SMW).  C.12  Isolation of Lignin from Ball Milled Wood Lignin was extracted from B M W using a mixture of dioxane and water (96:4).  Approximately 200 mL of solution was used to extract the lignin from 15-20 g BMW. After a day, the solvent was recovered through centrifugation, and the dioxane:water was replenished. The extraction was allowed to proceed for approximately 3 more days, after which time lignin extraction is minimal. The crude lignin resulting from the extraction of B M W was subject to the purification scheme via the Bjorkman method. The insoluble material, presumably consisting of lignin more closely associated with carbohydrates, was subject to a cellulase digestion.  Xylanase (Iogen,  Ottawa, ON), Gammanase (Novozyme) Novo-188 (Novozyme) and Fibrilase H D L 160 (Iogen) were mixed together in equal volumes, and 1 mL of this enzyme cocktail was added to each sample suspension in a 50 m M acetate buffer (pH 4.5). The enzyme cocktail was supplemented with tetracycline (Sigma, 40ug mL" ), cyclohexamide (Sigma, 30 ug mL" ) and sodium azide 1  1  (0.02%) to prevent microbial growth. After 2-3 regenerations of the cellulase solution, the sugars were discarded and the remaining lignin was once again subjected to a dioxane:water extraction, and purified according to Figure C I . The term C E L applies to the purified lignin from this fraction, and R E L applies to the residual lignin that is insoluble in dioxane:water. R E L lignin content was determined by the Klason method.  151  ball milled wood  insoluble 4  cellulase lignin (CEL)  7  residual lignin (REL)  7  acetylated milled wood lignin (Ac-MWL)  7  acetylated cellulase lignin (Ac-CEL)  acetylated residual lignin (Ac-REL)  Figure C I . Isolation, purification, and acetylation of various lignin fractions. 1) 96:4 dioxane:water 2) washed with 3x DI H 0 3) 90% acetic acid, ppt into water, centrifuge, wash 2x with water, 1:2 ethanol: 1,2-dichloroethane, ppt into ether wash 2x with ether dry 4) enzyme treatment 3x2 days 5) same purification as for M W L 6) rinsed 3x with water 7) N M I / D M S O / A c 0 acetylation 2  2  C.13  Molecular Weight Measurements Acetylated samples were made as 1 mg mL" in T H F and filtered through a 0.45 um 1  filter. A Dionex Summit HPLC fitted with Waters Styragel H5 column was used at 50°C with a 0.5 mL min" flow rate. The total run time was 60 minutes. Detection was via a deuterium U V 1  lamp set to 280 nm.  C.14  Elemental Analysis Elemental analysis (C, H , N , and O content by difference) was performed on a  PerkinElmer Series II CHNS/O Analyzer 2400 (Boston, M A ) according to manufacturer instructions. Samples were prepared by drying in a drying pistol (CH C1 ) overnight, and were 2  152  2  held in a desiccator until use.  Analyses were performed in triplicate and error calculated as  standard deviation from the mean.  C.15 Acetylation of Ball Milled Wood To 400 mg B M W and SMW, was added 5 mL 1-methylimidazole (Sigma). After stirring this mixture for at least 3 hours, 10 mL DMSO (Fisher) was added, and the mixture was allowed to stir overnight.  If the sample was pulverized adequately, a clear solution would form  overnight. Excess acetic anhydride, about 3 mL, was then added, and upon subsequent stirring the solution turned dark brown in colour.  This mixture was precipitated out into 2 L of  ethylenediaminetetraacetic acid (EDTA) (Sigma) solution at pH 8. The solution was allowed to settle out, filtered through a 2 urn membrane filter, and the precipitate was washed with 250 mL D l water. The resulting acetylated samples were dried over P2O5 (BDH Inc.).  C.16 Acetylation of Lignin In addition, acetylation of M W L , C E L , and R E L were performed with the above dissolution method.  This method was chosen as opposed to a more traditional acetylation  method in order to keep the treatment of samples as consistent as possible. A comparison of the yields from each method was performed and the yield for the N M I in D M S O method was not lower than the traditional method (detailed method section C.17).  C.17 Traditional Acetylation Method Milled wood lignin was acetylated using a pyridine and acetic anhydride in order to compare this "traditional" method with the cell wall dissolution method that uses NMI and DMSO as the solvent.  153  To mg quantity of M W L , pyridine was added so that the amount that was 10 times greater in volume. For example, if 0.05 g of M W L was used, 0.5 mL of pyridine was added to dissolve the sample. Once dissolved, an equivalent volume of acetic anhydride was added (0.5 mL with this example). This mixture was stirred for 2 days. After two days, the sample was precipitated into 100 mL of water. The precipitated mixture was stirred for A hour and then l  filtered through a 0.2 urn membrane filter in order to recover the highest possible yield while working with this small amount of material.  C.18  Nuclear Magnetic Resonance Spectroscopy Experiments Samples of acetylated cell wall material were prepared with a concentration of 150  mg mL" in C D C I 3 (Cambridge Isotope Laboratories). Ten minutes of sonnication was usually 1  required to dissolve the sample. The sample was then filtered through a Kimwipe plugged disposable pipette into a 5 mm N M R tube. Heteronuclear multiple quantum coherence (HMQC) and heteronuclear multiple bond coherence (HMBC) spectra were recorded on a Bruker A V A N C E 400 MHz BBi-x probe spectrometer. A l l runs were performed at room temperature (approximately 293 K). M W L or A c - M W L was dissolved in d6-DMSO (Cambridge Isotope Laboratories) at the approximate concentration of 60 mg in 0.25 mL. The sample was allowed to fully dissolve overnight in a desiccator. The next day, the sample was filtered through a Kimwipe into a 5 mm Shigemi N M R microtube (with the magnetic susceptibility of D M S O , Shigemi Co.). The microtube allows a smaller volume of sample to be analyzed, thereby facilitating higher concentrations of lignin even with a limited amount of sample material. H M Q C and H M B C spectra were obtained as above. For quantitative  1 3  C N M R spectra acquisition, 10 m M final concentration of chromium  (III) acetylacetonate, Cr(acac)3 (Aldrich), was added to shorten the T i relaxation times of the 154  carbon nuclei to allow the signals to be quantified. A Bruker AM-400 spectrometer (dual  1 3  C/'H  probe) set to a 90° pulse width, a 1.4 s acquisition time and a 1.7 s relaxation delay were used. At least 20,000 scans were collected over approximately 18 hours for M W L samples, and 60,000 scans for R E L samples due to low solubility and thus low sample concentration.  C.19  Lignin Hydroxyl Content Aliphatic and phenolic hydroxyl content was determined via N M R spectroscopy. Para-  nitrobenzaldehyde was used as the internal standard (stock solution of 1 mg mL" in CDCI3). 1  A c - M W L was dissolved into this solution at a concentration of 30 mg mL" of internal standard 1  solution. For each sample, 0.6 mL solution was prepared in order to conserve sample as much as possible. A proton N M R experiment was performed with 254 scans on a Bruker AV-300 MHz spectrometer. The integrated aliphatic and phenolic hydroxyl proton signals were quantified by comparison to the aldehyde peak from the internal standard. Peaks for aliphatic acetyl protons appear at 2.17-1.70 ppm, while phenolic acetyl appeared at 2.50-2.17 ppm. The internal standard aromatic protons appear at higher frequencies than 8.0 ppm. The molar content of hydroxide groups were estimated from the peak area of C H 3 C O - as the O H have all been replaced with acetyl groups in the acetylation reaction. For determination of a weight percent hydroxyl groups: A = moles internal standard in sample (calculated from volume, concentration, and molecular weight) B = grams lignin in sample C = integral value for hydroxyl region in question, divided by 3 weight percent OH = (A x C x 17.007 x 100VB where 17.007 g mol"  1  is the molecular weight of a OH group  155  C.20  Bibliography  Browning, B. L . (1967). Acetyl and Methoxyl Groups. Methods in Wood Chemistry. B. L . Browning. Appleton, Wisconsin. II: 660-664. Dence, C. W. (1992). Determination of Lignin. Methods in Lignin Chemistry. S. D. Lin, C.W. Berlin, Springer-Verlag. Fengel, D. and G. Wegener (1989). Wood: chemistry, ultrastructure. reactions. Berlin, Walter de Gruyter. Huntley, S. K., D. Ellis, M . Gilbert, C. Chappie and S. D. Mansfield (2003). "Significant Increases in Pulping Efficiency in C4H-F5H-Transformed Poplars: Improved Chemical Savings and Reduced Environmental Toxins." Journal of Agricultural and Food Chemistry 51: 6178-6183. Meyer, K., A. M . Shirle, J. C. Cusomano, D. A. Bell-Lelong and C. Chappie (1998). "Lignin monomer composition is determined by the expression of a cytochrome P450-dependent monooxygenase in Arabidopsis." Proceedings of the National Academy of Sciences of the United States of America 95: 6619-6623. Rolando, C , B. Monties and C. Lapierre (1992). Thioacidolysis. Methods in Lignin Chemistry. S. D. Lin, C.W. Berlin, Springer-Verlag.  156  

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.831.1-0075069/manifest

Comment

Related Items