UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Pretreatment and hydrolysis of recovered fibre for ethanol production Ruffell, John 2008

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

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata


24-ubc_2008_fall_ruffell_john.pdf [ 1.34MB ]
JSON: 24-1.0058544.json
JSON-LD: 24-1.0058544-ld.json
RDF/XML (Pretty): 24-1.0058544-rdf.xml
RDF/JSON: 24-1.0058544-rdf.json
Turtle: 24-1.0058544-turtle.txt
N-Triples: 24-1.0058544-rdf-ntriples.txt
Original Record: 24-1.0058544-source.json
Full Text

Full Text

  PRETREATMENT AND HYDROLYSIS OF RECOVERED FIBRE FOR ETHANOL PRODUCTION  by JOHN RUFFELL B.A.Sc., The University of British Columbia, 2006  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF APPLIED SCIENCE  in  THE FACULTY OF GRADUATE STUDIES (Chemical and Biological Engineering)    THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)   August 2008 © John Ruffell, 2008 ii  ABSTRACT  Energy utilization is a determining factor for the standards of living around the world, and the current primary source of energy is fossil fuels. A potential source of liquid fuels that could ease the strain caused by diminishing petroleum resources is bioethanol. Effective exploitation of biomass materials requires a pretreatment to disrupt the lignin and cellulose matrix. The pretreatment utilized for this research was oxygen delignification, which is a standard process stage in the production of bleached chemical pulp. The model substrate utilized as a feedstock for bioethanol was recovered fibre. An analysis of the substrates digestibility resulted in a hexose yield of approximately 23%, which justified the need for an effective pretreatment. An experimental design was performed to optimize the delignification conditions by performing experiments over a range of temperature, caustic loadings, and reaction times. Equations were developed that outline the dependence of various response parameters on the experimental variables. An empirical model that can predict sugar concentrations from enzymatic hydrolysis based on the Kappa number, enzyme loading, and initial fibre concentration was also developed.        ( )             −+ −= −1 1 11 11 n Conkt CoS Where: Co = starting cellulose concentration (g/L)      S = sugar concentration (g/L)  2.0CoE5.0k ⋅⋅=       63.0Co08.0K05.0 Eee6.4n −×−⋅ ⋅⋅⋅=     K = kappa number E = enzyme concentration (FPU/g)  A study of hydrolysis feeding regimes for untreated recovered fibre (87 Kappa), pretreated recovered fibre (17 Kappa), and bleached pulp (6 Kappa) showed that the batch feeding regime offers reduced complexity and high sugar yields for lower Kappa substrates. iii  In order to evaluate the possibility of lignin recovery, the pH of delignification liquor was reduced by the addition of CO2 and H2SO4, resulting in up to 25% lignin yield. An experiment that looked at effect of post-delignification fibre washing on downstream hydrolysis found that a washing efficiency of approximately 90% is required in order to achieve a hexose sugar yield of 85%.                  iv  TABLE OF CONTENTS  Abstract ..................................................................................................................................... ii Table of Contents…………………………………………………………………………………iv List of Tables ........................................................................................................................... vii List of Figures......................................................................................................................... viii Nomenclature.............................................................................................................................xi Acknowledgements................................................................................................................. xiii 1. INTRODUCTION...................................................................................................................1 1.1  Background ..................................................................................................................... 1 1.1.1  Composition of lignocellulose................................................................................5 1.1.2  Investing in lignocellulosic ethanol ........................................................................7 1.1.3 Transportation of ethanol .......................................................................................7 1.1.4 Energy content of biofuels .....................................................................................8 1.1.5 Converting plant biomass into energy ....................................................................9 1.2 Chemistry of lignocellulose ........................................................................................... 11 1.2.1 Cellulose..............................................................................................................12 1.2.2 Hemicellulose ......................................................................................................13 1.2.3 Lignin ..................................................................................................................13 1.2.4 Extractives ...........................................................................................................17 1.3 Enzymatic hydrolysis .................................................................................................... 17 1.3.1 Cellulases.............................................................................................................19 1.3.2  Factors that limit enzymatic hydrolysis ................................................................21 1.3.3 Models of enzymatic hydrolysis ...........................................................................23 1.4 Biomass pretreatment .................................................................................................... 24 1.4.1 Physical pretreatments .........................................................................................25 1.4.2 Chemical pretreatments........................................................................................25 1.4.3 Biological pretreatments ......................................................................................28 1.5 Oxygen delignification process...................................................................................... 28 1.5.1 Medium and high consistency oxygen delignification systems .............................31 v  1.5.2 Oxygen delignification chemistry.........................................................................35 1.5.3 Cellulose reactions during oxygen delignification ................................................38 2. OBJECTIVES.......................................................................................................................40 3. MATERIALS AND METHODS...........................................................................................41 3.1 Oxygen delignification setup and procedure ....................................................................... 41 3.1.1 Laboratory scale oxygen delignification apparatus.....................................................41 3.1.2 Procedure for oxygen delignification experiments .....................................................42 3.2 Procedure for lab-scale enzymatic hydrolysis ..................................................................... 43 4 ANALYSES ..........................................................................................................................44 4.1 Procedure for determining fibre composition ...................................................................... 44 4.2 Procedure for determining sugar composition from a hydrolysate sample ........................... 45 4.3 Procedure for determining kappa number ........................................................................... 46 4.4 Procedure for the determining of cellulase activity ............................................................. 47 4.5 Procedure for determining β-Glucosidase activity............................................................... 48 4.5 Fed-batch and batch hydrolysis of multiple pulp substrate .................................................. 49 4.6 Procedure for lignin precipitation from oxygen delignification liquor ................................. 49 5. RESULTS AND DISCUSSION.........................................................................................51 5.1 Analysis of recovered fibre ............................................................................................ 51 5.1.1 Compositional analysis and kappa number of recovered fibre ..............................51 5.1.2 Hydrolyzability of short, long, and mixed recovered fibre ....................................52 5.2 Assessment of the hydrolyzability of multiple pulp substrates........................................ 56 5.2.1 Assessment of the digestibility of multiple brownstock press and bleached pulps.58 5.3 Oxygen delignification pretreatment .............................................................................. 59 5.3.1 Preliminary oxygen delignification runs ...............................................................59 5.3.2 Effect of temperature, caustic load, and time on the oxygen delignification yield,  kappa number, and hydrolyzability ......................................................................61 5.3.3      Development of empirical models for oxygen delignification yield, kappa number,  and hydrolyzability ..............................................................................................62 5.3.4 Empirical model for the enzymatic hydrolysis reaction ........................................68 5.4 Practical issues .............................................................................................................. 75 5.4.1 Hydrolysis feeding regimes..................................................................................75 vi  5.4.2 Fed-batch hydrolysis of micro-crystalline cellulose and pretreated recovered fibre  ……………………………………………………………………………………..77 5.4.3 A comparison of fed-batch and batch hydrolysis of multiple substrates ................80 5.4.4 Precipitation of lignin...........................................................................................85 5.4.5 Effect of post-delignification fibre washing on downstream hydrolysis ................88 6. CONCLUSIONS ...............................................................................................................91 7. FUTURE WORK...............................................................................................................93 8.     REFERENCES................................................................................................................. 95                        vii  List of Tables  Table 1.1 Sources, types, and current uses of lignocellulosic materials ......………………...2 Table 1.2 Breakdown of lignocellulosic material …...………………………………………6 Table 1.3 Energy content of fuels ….....……………………………………………………..8 Table 1.4 Hydrolysis limiting structural properties of cellulosic fibres …....……………...21 Table 1.5 Delignification: Pulping and bleaching processes …………..…………………..29 Table 3.1 Experimental design: enzymatic hydrolysis conditions …………………………44 Table 5.1  Chemical composition of recovered fibre samples ……………………………...51 Table 5.2 Hydrolyzability of short and long recovered fibre ……………………………....53 Table 5.3 Initial hydrolysis rates of short and long recovered fibre ……………………….54 Table 5.4 Maximum digestibility of short and long recovered fibre ………………………55 Table 5.5 Digestibility of recovered fibre at 5, 10, and 20 FPU/g …………………………56 Table 5.6 Hexose yield of brownstock press and bleached pulps ………………………….58 Table 5.7 Delignification yield, kappa number, and hydrolyzability of pretreated recovered    fibre ……………………………………………………………………………..61 Table 5.8 Empirical constant`s for enzymatic hydrolysis runs …………………………….69 Table 5.9 Concentration of lignin in oxygen delignification liquors ………………………86     viii  List of Figures  Figure 1.1 Process steps for the production of ethanol ……………………………………....4 Figure 1.2 Four different processes involved in gasification ……………………………….11 Figure 1.3 Structure of cellulose …………………………………………………………....13 Figure 1.4 Three common phenyl-propane precursors of lignin ……………………………15 Figure 1.5 Possible lignin structure ………………………………………………………....16 Figure 1.6 Enzymatic action during hydrolysis ……………………………………………..20 Figure 1.7 Process flow diagram for the steam-explosion of softwood …………………….26 Figure 1.8 Flowsheet of a typical high consistency oxygen delignification system ………..33 Figure 1.9 Flowsheet of a typical medium consistency oxygen delignification system …....34 Figure 1.10 Reduction of molecular oxygen in four consecutive steps ……………………...35 Figure 1.11 Single unit of complex lignin structure (Arylpropane) ………………………….36 Figure 1.12 High and low electron density centers on lignin units …………………………..37 Figure 1.13 Condensed and non-condensed phenolic subunits of lignin …………………….37 Figure 1.14 Oxidation and cleavage of the cellulose chain …………………………………..38 Figure 1.15 Features of cellulose degradation in alkali ………………………………………39 Figure 3.1 PARR Series 4520 oxygen delignification reactor ……………………………...41 Figure 4.1 Typical curve for quantifying the cellulase enzyme activity in FPU ……………48 Figure 4.2 β-glucosidase activity (CBU) over a range of enzyme dilutions ………………..49 Figure 5.1  Schematic image of a rotor foil passing a section of a screen cylinder ………....53 Figure 5.2  Hexose sugar yield of bleached and unbleached kraft pulp……………………...57 ix  Figure 5.3 Influence of caustic addition on the hydrolyzability of recovered fibre ………...60  Figure 5.4 Kappa number as a function of temperature, caustic load, and time ……………64  Figure 5.5 Oxygen delignification yield: function of temperature, caustic load, and time …65  Figure 5.6 Sugar loss versus Kappa number for pretreated recovered fibre ………………..65  Figure 5.7 Hydrolyzability of pretreated recovered fibre as a function of temperature,    caustic load, and time (Enzyme loading: 20 FPU/g, 2% Solids) ……………….66  Figure 5.8 Hydrolyzability of pretreated recovered fibre as a function of temperature,    caustic load, and time (Enzyme loading: 40 FPU/g, 2% Solids) ……………….67  Figure 5.9 Hydrolyzability of pretreated recovered fibre as a function of temperature,    caustic load, and time (Enzyme loading: 16 FPU/g, 10% Solids) ……………...68 Figure 5.10 Evaluation of empirical hydrolysis model fit on untreated recovered fibre, 87 Kappa, 10/40 FPU/g, 20 g/L …………………………………………………….71 Figure 5.11 Evaluation of empirical hydrolysis model fit on pretreated recovered fibre, 50 Kappa, 10/40 FPU/g, 20 g/L …………………………………………………….71 Figure 5.12 Evaluation of empirical hydrolysis model fit on pretreated recovered fibre, 40 Kappa, 10/40 FPU/g, 20 g/L …………………………………………………….72 Figure 5.13 Evaluation of empirical hydrolysis model fit on pretreated recovered fibre, 40 Kappa, 20 FPU/g, 50, 70, 100 g/L ...…………………………………………….72 Figure 5.14 Evaluation of empirical hydrolysis model fit on untreated recovered fibre, 87 Kappa, 10/40 FPU/g, 20 g/L …………………………………...………………..73  x  Figure 5.15 Evaluation of empirical hydrolysis model fit on untreated recovered fibre, 50 Kappa, 10/40 FPU/g, 20 g/L ………………...…………………………………..73 Figure 5.16 Evaluation of empirical hydrolysis model fit on untreated recovered fibre, 40 Kappa, 10/40 FPU/g, 20 g/L ……………………...……………………………..74 Figure 5.17 Evaluation of empirical hydrolysis model fit on pretreated recovered fibre, 40 Kappa, 20 FPU/g, 50, 70, 100 g/L ...……………...……………………………..74 Figure 5.18 Hexose sugar yield for the fed-batch hydrolysis of pretreated recovered fibre and microcrystalline cellulose (20 FPU/g)………...…………………………………78 Figure 5.19 Hexose sugar concentration for the fed-batch hydrolysis of pretreated recovered fibre and microcrystalline cellulose (80 FPU/g) ………………………………...79 Figure 5.20 Comparison between batch, substrate fed-batch, and substrate /enzyme fed-batch feeding regimes for untreated recovered fibre (Kappa 87) ……………………...80 Figure 5.21 Comparison between batch, substrate fed-batch, and substrate /enzyme fed-batch feeding regimes for untreated recovered fibre (Kappa 17) ……………………...82 Figure 5.22 Comparison between batch, substrate fed-batch, and substrate /enzyme fed-batch    feeding regimes for bleached fibre ……………………………………………..84 Figure 5.23 Two-stage lignin precipitation: (1) CO2 sparging (2) Sulphuric acid addition … 87 Figure 5.24 Hexose sugar yield of hydrolysis flasks utilizing a range of oxygen   delignification liquor volumes …………………………………………………..90     xi  NOMENCLATURE  α  Alpha β  Beta µL  microliter(s) µmol  micromole(s) AFEX  Ammonia fibre explosion BOD  Biochemical oxygen demand BTU  British thermal unit oC  degree Celsius cm  centimeter(s) CBU  Cellobiose units COD  Chemical oxygen demand DDGS  Distiller’s dried grains with solubles DNS  Dinitrosalicylic acid DP  Degree of polymerization EU  European Union FPU  Filter paper units g  gram(s) h  hour HPLC  High pressure liquid chromatography IU  International unit L  liter(s) m  meters mg  milligram(s) xii  mJ  mega joule mL  milliliter(s) mM  millimolar min  minute(s) M  molar MTBE  methyl tertiary butyl ether nm  nanometers NREL  National Renewable Energy Laboratory OCPA  Ontario Corn Producers Association psig  pounds per square inch (gauge) PCS  Primary clarifier sludge rpm  revolutions per minute s  second(s) SSF  Simultaneous saccharification and fermentation USD  United States dollar(s) USDA  United States Department of Agriculture w/w  weight per weight WTC  Weyerhaeuser Technology Centre         xiii  Acknowledgements  The research work presented in this thesis was funded by Dr. Sheldon Duff of the Department of Chemical and Biological Engineering at the University of British Columbia. A sincere thank you to Dr. Duff for his mentorship, and valuable contributions to every level my thesis. I would also like to thank Dr. Benjamin Levie of Weyerhaeuser, who provided the substrates used during experimentation, as well as helpful feedback during the course of my work. I also appreciate all of the assistance that I received from the Chemical and Biological Engineering office and workshop staff. Appreciation is extended to all of my peers in the Biofuels Research Lab, with a particular thanks to Dr. Steve Helle for his patience and assistance in training me in laboratory procedures. Finally, thanks to my family and Diana for their support while working towards my graduate degree.                1  1. INTRODUCTION  1.1  Background  The current driving force towards renewable energy is the price and availability of oil, despite the fact that future decision making will likely be guided by the environmental impact of our energy choices. In 2001 the average price of crude oil was 25.6 USD/barrel, but as of April 2008 the price has risen substantially, to 117 USD/barrel (1). The cost of oil has been steadily increasing due to demand from emerging economies like China and India, as well as price instability being created from global conflicts. The high price of oil is continually pushing countries towards reducing expensive oil imports while seeking innovative solutions to lower domestic dependence on fossil fuels.  1.1.1  Bioethanol resources  One source of renewable energy that shows promise in reducing the environmental impacts of the world’s energy demand is bioethanol. Bioethanol can simply be described as an ethyl alcohol that is obtained from a biological feedstock. A substrate that has enormous biotechnological value as a renewable substrate for bioethanol is lignocellulose. Lignocellulose is an important structural component of woody plants and consists of cellulose, hemicellulose, and lignin. The cellulose molecules, which are polymers with six carbon sugars linked in long chains, and five carbon sugar chains called hemicelluloses, are reinforced by cross linked organic molecules called lignin. This structure is very difficult for microbes to break down into simple sugars. In order to exploit the structural sugars from plant fibers for bioethanol, the recalcitrance of biomass must be overcome in a way that is cost competitive in current markets. The kind of cost- reducing measures that bio-refineries are working towards are efficient de-polymerization of cellulose and hemicelluloses, efficient fermentation (including pentose and hexose sugars) that can handle inhibitory compounds, optimizing process integration, and a cost efficient use of lignin (2) (3).  2  Traditionally, the primary sources for ethanol production have been from sugarcane, corn, and sugar beets. In addition to these primary sources of lignocellulose there are additional potential bioethanol sources in forest residues, agricultural waste, municipal/industrial wastes, and crops grown specifically for energy use. A list of current and potential sources of lignocellulose, gathered by Howard et al. (4), is shown in Table 1.1.  Table 1.1 Sources, types, and current uses of lignocellulosic materials Lignocellulosic substrate Residues Alternate use Grain harvesting Wheat, rice, oats, barley and corn Straw, cobs, stalks, husks Animal feed, burnt as fuel, compost, soil conditioner Processed grains Corn, wheat, rice, soybean Waste water, bran Animal feed Fruit and vegetable harvesting Seeds, peels, husks, stones, rejected whole fruit and juice Animal and fish feed, some seeds for soil extraction Fruit and vegetable processing Seeds, peels, waste water, husks, shells, stones, rejected whole fruit and juice Animal and fish feed, some seeds for oil extraction Sugar cane other sugar products Bagasse Burnt as fuel Oils and oilseed plants, nuts, cotton seeds, olives, soybean, etc. Shells, husks, lint, fibre, sludge, presscake, wastewater Animal feed, fertilizer, burnt fuel Animal waste Manure, other waste Soil conditioners Forestry-paper and pulp Harvesting of logs Wood residuals, barks, leaves Soil conditioners, burnt Saw-and plywood waste Woodchips, wood shavings, saw dust Pulp and paper industries, chip and fibre board Pulp and paper mills Fibre waste, sulphite liquor Reused in pulp and board industry as fuel Lignocellulose waste from communities Old newspapers, paper, cardboard, old boards, disused furniture Small percentage recycled, others burnt Grass Unutilized grass  Burnt  3  Brazil is currently producing over 19,000 million liters of ethanol annually, and in 2007 was the world’s largest exporter of ethanol at over 3,500 million liters (5). Brazil’s primary ethanol fuel source is sugar cane, which is well suited to grow in South America’s tropical climate. Sugar cane has a high concentration of sucrose that it stores in its sap for energy, and this sugar can then be concentrated in an industrial process by the evaporation of water. A study in 2004 by the University of Campinas looked at all of the energy inputs to the sugar cane industry, including agriculture and fertilizer manufacturing, and determined that ethanol produced from sugar cane yielded an eight-fold energy return (6). One reason for the high energy return of sugarcane ethanol is the burning of excess bagasse, which is a fibrous pulp that remains once the sugar has been squeezed from the cane. The energy return for sugarcane ethanol (8 to 9 units of energy for each unit expended) is far higher than corn ethanol-producers (1.34 units of energy for each unit expended) have been able to achieve in North America thus far (7) (8).  Despite the positive energy balance, some scientists view Brazil’s ethanol industry as having a negative environmental impact. This is due to 3-5% of the nitrogen in fertilizers being converted to nitrous oxide and emitted (through volatisation on application as well as interaction with soil compounds), doses of required pesticides, volumes of wastewater created (13L of wastewater for every 1L of bioethanol made, with a biochemical oxygen demand (BOD) between 18,000 and 37,000 mg/L compared to around 300 mg/L for raw sewage), and the formation of air pollutants caused from the burning of sugarcane fields (9) (10). The industry has been responding to these claims by seeking greener solutions, like using the leftovers from distillation as potassium fertilizer, and seeking legislation aimed at increasing the use of mechanized harvesters to reduce field burning, which has traditionally been done to make the cane easier to harvest (11).  An additional environmental concern in Brazil is the clearing of forest land, which releases stored carbon and reduces biodiversity. It can be argued that sugarcane production does not affect rainforest land in that sugar cane requires a dry season, since its roots cannot be waterlogged, and therefore would not be a suitable crop to plant in cleared rainforest land (12). The destruction of rainforest land is, however, occurring, with cane crops displacing other crops that thrive in the rainforest climate. This practice will continue until there are strong agricultural policies to preserve native forest land. 4  In the US the primary substrate for ethanol production is corn (maize). The conventional bioethanol production process is shown in Figure 1.1, with feedstocks including tapioca, wheat, corn, and milo (a species of grass within the genus Sorghum). In the liquefaction step the thermo-stable alpha amylase breaks down the large starch molecules into short sections, followed by the saccharification step where the glucoamylase breaks the shorter sections into glucose molecules for fermentation to ethanol. A valuable co-product produced from this process is dried distillers grains with solubles (DDGS), which are the remaining nutrients (protein, fiber, and oil) after fermentation that are used to create livestock feed. In contrast to Brazil’s sugar cane production, the US is the largest producer of corn in the world and harvests corn from a total area approximately the size of Arizona (13).     Figure 1.1 has been removed because of copyright restrictions. The information removed is a visual representation of the process steps in the production of ethanol. The original source of the material is: Genencor Introduces New Protease Enzyme for Ethanol Industry. Greencar Congress. [Online] 2006. [Cited: February 12, 2008.] List of Figures (14).      Figure 1.1 Process steps for the production of ethanol (14)   The difficulty with corn as a fuel is that a significant portion of the sugar that the plant creates goes towards the formation of the stalk and stover, which is structural material that contains lignin, making the cellulose more difficult to access. However, the corn kernels themselves consist mainly of starch which can be broken down easily, a task that current bio-refineries can readily handle. Debate on the energy balance of ethanol from corn has led critics to conclude that 5  roughly 30% more energy is required to produce corn-ethanol than is returned (15), while proponents like the US Department of Agriculture (USDA) argue a net energy gain of 16,000 BTU/gallon (4.46 MJ/L), which is a gain of 1.3 units of fuel energy compared to a gain of 8 to 9 units of fuel energy for sugarcane (16). However, a bipartisan issue on whether maize is well suited for an industrial feedstock lies in the demand from multiple markets. Maize is used for animal and human consumption, in addition to serving as a sweetener for the food industry. The rise in price for animal feed translates into a higher cost for the consumer purchasing meat, eggs, and dairy products. In the US the competition for corn led to a dramatic price increase in 2007 of over 60%, with ethanol producers consuming close to 30% of the 12 billion-bushel harvested in 2007 according to the USDA (17). The Ontario Corn Producers Association (OCPA) states that the US is not alone in feeling the effects of high corn prices, in Canada the price per bushel rose to over $4 in 2007 (18).  There is also an ongoing debate about whether is it ethical to use human food resources like corn and sugarcane for ethanol production. The moral and ethical issues that are raised include the ever increasing number of malnourished people in the developing world (820-million, 2006), and stress the importance of increasing the exports of North American corn and grains (19) (20). In order to effectively lower the cost of fuel and secure new resources, there are existing alternatives to the use of food-based fuels, including increasing fuel efficiency standards for new model vehicles, and switching to highly efficient gas-electric hybrid and plug-in vehicles.  1.1.1  Composition of lignocellulose  Lignocellulose is the world’s most plentiful biopolymer and can be utilized for the production of bioethanol. Advantages of using lignocellulose as a fuel feedstock include: being able to use land that is currently marginal for agriculture for the production of lignocellulosic crops like switch grass, being non-competitive with the production of food, reduction of solid waste disposal through cellulosic conversion, and the process generates lignin which can be recovered and utilized as an energy-rich boiler fuel (21).  6  The composition of lignocellulosic material depends on the source, as well as any processing or pretreatment performed on the substrate. In British Columbia, softwood and wood waste are a potential plentiful source of renewable feedstock. In comparison to hardwood, softwood trees generally contain a higher percentage of lignin and a lower percentage of hemicelluloses (4). A summary of the cellulose, hemicellulose, and lignin content of various lignocellulosic substrates is shown in Table 1.2, data taken from Howard et al. (4). Additional compounds found in wood and some other plant sources are extractives, which include resin acids, fatty acids, turpenoid compounds and alcohols. These extractives are generally soluble in neutral organic solvents or water. Most North American wood species contain less than 1% extractives content based on oven-dried substrate weight (22). Table1.2 Breakdown of lignocellulosic material Lignocellulosic substrate Cellulose (%) Hemicellulose (%) Lignin (%) Hardwood stems 40-55 24-40 18-25 Softwood stems 45-50 25-35 25-35 Corn cobs 45 35 15 Corn stover Paper 38 85-99 32 0 15 0-15 Wheat straw 30 50 15 Rice straw 32.1 24 18 Leaves 15-20 80-85 0 Newspaper 40-55 25-40 18-30 Waste paper from chemical pulps 60-70 10-20 5-10 Primary wastewater solids 8-15 NA 24-29 Fresh bagasse 33.4 30 18.9 Solid cattle manure 1.6-4.7 1.4-3.3 2.7-5.7 Coastal Bermuda grass 25 35.7 6.4 Switch grass 45 31.4 12 Grasses (avg. for grasses) 25-40 25-50 10-30 7   1.1.2  Investing in lignocellulosic ethanol  The race is on to develop the most efficient conversion process for the many sources of lignocellulose, and the development is coming from both the public and private sector. In August 2006, the USDA announced a $250 million dollar commitment over five years to the development of two Bioenergy Research Centre’s that will work towards the development of cellulosic ethanol and other biofuels (23). The European Union (EU) has also committed over $100 million towards cellulosic ethanol research under its 7th EU Framework Programme (24). Industry has also been forming partnerships with University labs through corporate investments, for example British Petroleum’s $500 M investment over ten years for biofuel research at UC Berkeley, and a joint partnership between ExxonMobil, Toyota, Schlumberger, and General Electric for $225 M over ten years at Stanford University to research biofuels and renewable energy (25). A current partnership between Purdue University and the US Department of Energy (USDE) is looking at whether transgenic poplar trees could replace corn as an ethanol fuel source. The goal of this project is to determine whether cells could be altered to make various kinds or amounts of lignin, which serves as a barrier for enzymes attempting to access cellulose (26). Iogen, located in Ottawa, Canada, is one of the current leaders in the cellulosic ethanol market. Their demonstration plant has received $130 million in private investment, along with another $120 million from various investors like Royal Dutch/Shell Group, Goldman Sachs, PetroCanada, and Technology Partners Canada (27). The next step that the companies in the ethanol business need to make is taking the capital-investment risk to move beyond the pilot- scale stage, in order to take advantage of current and developing technology.  1.1.3 Transportation of ethanol  It is generally accepted that the most cost-effective and quickest method of transporting liquid fuels is via a pipeline. Pipelines are currently used to transport vast quantities of gasoline and natural gas from refineries to distribution centers. Utilizing pipelines for the transportation of ethanol is viewed as being more complicated because ethanol absorbs water and impurities, there 8  are limitations on the availability of existing pipelines, and insufficient volumes of ethanol that currently require transport (28). Since ethanol cannot currently be shipped through pipelines it must be “splash-blended”, which refers to the transportation of ethanol by tanker trucks to the terminal for blending, as opposed to at the refinery. Another challenge is the cost associated with the development of supply chains that are required for ethanol distribution, which will be required to transport ethanol from an agriculture region to a more heavily populated urban centre in the same manner that oil is currently distributed (29).  For vehicles with fuel systems that are currently not designed to E85-ready (85% ethanol, 15% gasoline) standards, the limitation on the percentage of ethanol that can be used in the fuel is 10% due to corrosion issues (30). Corrosion issues stem from the hygroscopic nature of ethanol, which is due to the hydrogen bonding of its hydroxyl group (31).  1.1.4 Energy content of biofuels  Both ethanol and biodiesel have a lower energy content than gasoline or diesel. The values of the energy content of various fuels gathered by the US Energy Information Centre is shown in Table 1.3, the basis for comparison is MJ per liter and liters of gasoline equivalent. The table displays both the high heating value (amount of heat released by the fuel including the latent heat of vaporization of water) and the low heating value (amount of heat released by the fuel, excluding the latent heat of vaporization of water). The reduced energy content of ethanol and biodiesel translates into a reduction in the miles driven per gallon of fuel for a vehicle (32).  Table 1.3 Energy content of fuels Fuel MJ per Liter (low heating value) MJ per Liter (high heating value) Liters of gasoline equivalent (high heating value) Conventional gasoline 32,340 35,020 1.00 Fuel ethanol (E100) 21,280 23,593 0.67 E85 (74% blend on average) - 26,564 0.76 Distillate fuel oil (diesel) 35,980 38,833 1.11 Biodiesel (B100) 33,123 35,986 1.03 9   As a liquid fuel, ethanol offers a higher octane value when compared to gasoline. The higher octane value of ethanol gives good antiknock properties, which allows for increased compression ratios and engine efficiencies (33). Another gasoline additive, methyl tertiary butyl ether (MTBE), was a previously popular oxygenate to increase the octane value of fuels, before environmental and health concerns has resulted in a decrease in its use. Ethanol has been presented as a safe alternative to MTBE, with proponents arguing that ethanol is the safest component of blended gasoline and offers a reduction in smog pollution (34). This has not been without counter-argument however; there are concerns that since lower volumes of ethanol are used compared to MTBE it provides less reduction in toxic air emissions, and also decreases the shelf-life of fuel from several years to a maximum of 90 – 100 days (35). 1.1.5 Converting plant biomass into energy  There are a number of different pathways by which plant biomass can be converted into energy, fuels, and chemicals. These include, direct combustion, gasification, pyrolysis, and hydrolysis and fermentation.  Direct combustion can be used to convert plant biomass into useful energy, by utilizing the heat and/or steam which is produced during the process. The heat can be used for domestic cooking and space heating, as well as for industrial purposes by making use of a boiler or kiln. Direct combustion can also be used to generate and deliver electricity to the grid, and/or satisfy the power demand of a stand-alone process. The net energy that is available from direct combustion of biomass ranges from approximately 8 MJ/kg for green wood, to 20 MJ/kg for dry plant matter (36).  A process that can be applied for the conversion of lignocellulosic substrates into combustion gases, instead of a liquid fuel, is called gasification. The gasification of biomass is a high temperature process, in the range of 600-1000oC, that decomposes the complex hydrocarbons into principally hydrogen, carbon monoxide, and carbon dioxide (37). Additional gasification products include char, tars, methane, water, and other constituents, the distribution of which is dependent on the substrate. The value of biomass gasification over combustion is that 10  gasification offers double the power generation efficiency by applying a combined cycle gas turbine, versus simply a steam cycle used in combustion (38).  Three methods for the gasification of biomass are: pyrolytic gasification, gasification with air, or gasification with oxygen, shown in Figure 1.2. Pyrolytic gasification involves heating the biomass to over 400oC, which yields roughly 25% char and significant amounts of condensibles like tars. When air is used for gasification “producer gas” is formed, which typically contains the following proportion of gases: 22% CO, 18% H2, 3% CH4, 6% CO2, and 51% N2. Gasification with oxygen produces synthesis gas which can be used to create ammonia, methanol, and diesel fuel. The typical gas composition of synthesis gas from biomass is: 40% CO, 40% H2, 3% CH4, 17% CO2 (39). In order to conceptualize the thermochemical processes occurring in a gasifier it is useful to separate the gasifier into distinct stages that are occurring at the same time. These stages include the drying, pyrolysis, oxidation, and reduction stages. The drying stage reduces the moisture content of the biomass without any decomposition. This is followed by a pyrolysis stage which is the thermal decomposition of the biomass in the absence of oxygen. The principle reactions that occur in the oxidation and reduction stages according to the ARTES Institute are outlined in Figure 1.2 (38).   11    Figure1.2 Four different processes involved in gasification  Another pathway that can be used for the conversion of plant biomass is hydrolysis and fermentation. An initial hydrolysis step is used to liberate hexose and pentose sugars from the lignocellulosic substrate, which is followed by a fermentation step to convert the sugars into ethanol. A more detailed explanation of enzymatic and acid hydrolysis can be found in Sections 1.3 and 1.4.2, respectively.   1.2 Chemistry of lignocellulose  The chemical structure of the three main components of biomass; cellulose, hemicelluloses, and lignin are all unique and present different challenges for their separation and utilization in the bioethanol process. 12  1.2.1 Cellulose  Cellulose is both a carbohydrate and polysaccharide, meaning that cellulose contains carbon, hydrogen, and oxygen (C6H10O5)n with multiple sugar units. The number of repeating sugar units is called the degree of polymerization (DP), with native softwood cellulose and paper making fibres having respective average DP values of 3500 and 600-1500 (40) (22).  unitecosgluoneofweightMolecular celluloseofweightMolecular DP =  Cellulose is the primary structural component of green plants, making it the most abundant organic polymer on earth (41) (42). Despite its abundance, the majority of animals are unable to utilize cellulose as an energy source because they lack the enzyme required to hydrolyze the β(1- 4) linkages that form cellulose (43). Humans are unable to digest cellulose, however the undigestible cellulose does aid in the smooth working of the intestinal tract. Animals such as sheep, cows, horses, and insects like termites, can digest cellulose because they have symbiotic bacteria that possess the necessary enzymes (44).  The structure of cellulose, shown in Figure 1.3, is composed of a recurring unit of two consecutive anhydride units, known as a cellobiose unit (22). Cellulose is a straight chain polymer that is formed by β(1-4)-glycosidic bonds, in contrast to the α(1-4-)-glycosidic bonds that are found in starch and other carbohydrates. The β(1-4) linkages gives cellulose a fibrous, tough, and water insoluble nature because this conformation ensures that each glucose residue is orientated in ‘trans’ with its neighbour. This orientation results in hydrogen bonding between hydroxyl groups and oxygen, producing a linear chain of high tensile strength. Additional hydrogen bonding occurs between cellulose molecules, resulting in parallel bundles of 40-70 cellulose chains called microfibrils. Cellulose is heterogeneous, with highly ordered crystalline regions where the orientation of the microfibrils is such that the cellulose reducing ends are at one terminus and the non-reducing chain ends are at the other. The reducing end of a cellulose molecule refers to the end with a free hemi-acetal (or aldehyde) group at the C-1 position, and the non-reducing end refers to the end with a free hydroxyl at the C-4 position (45). There are 13  also disordered sections of the cellulose chain that are intermixed with the crystalline regions called amorphous cellulose (46). The crystalline zones are difficult to penetrate by reagents or solvents, and the amorphous regions are readily penetrable, and consequently are more susceptible to hydrolysis (22). The microfibrils are assembled in the cell wall in layers (or lamellae) with spacing between each microfibril of 20-40 nm, and are linked by lignin and cross- linking glycans called hemicelluloses (47) (48) (49).   Figure 1.3 Structure of cellulose  1.2.2 Hemicellulose  Hemicellulose consists of numerous different sugar monomers. Possible sugar monomers in hemicellulose include pentoses (arabinose, xylose) and hexoses (galactose, glucose, and mannose). The sugar monomer that is always present in the largest quantities is xylose (50).  Hemicellulose is a polysaccharide, however it consists of shorter chain lengths than cellulose – approximately 200 sugar units. Hemicellulose is a branched polymer with a random, amorphous structure, in contrast to the un-branched, crystalline structure of cellulose.  1.2.3 Lignin  In addition to the carbohydrate content of plant fibers, there is an intracellular material that cements fibers together called lignin. The complex, random nature of lignin has evolved to provide plants with a strong defense against enzymatic and physical attack, by anchoring and 14  stiffening cellulose fibers (4).  Overcoming the natural defenses that lignin provides make degradation of lignin and access to the cellulose a challenge. However, even with this difficulty overcome, the chemical structure of lignin is such that intermediate degradation products include phenolic compounds which are highly toxic to fermentation organisms which convert cellulose into ethanol (51).  A general classification of lignins typically involves three major groups, those being hardwood lignin, softwood lignin, and grass lignin. The distinction between the groups is based on the chemical structure of their monomer units (52).  The aromatic ring content of lignin is high, resulting from the enzyme catalyzed polymerization of three hydroxyphenyl-propane units or lignin precursors, shown in Figure 1.4 (53). The relative amounts of the lignin precursors are dependent on the origin of the lignin. For example, softwoods consist primarily of coniferyl alcohol, while hardwoods possess a more even balance of coniferyl alcohol and sinapyl alcohol (54). Para-coumaryl alcohol is found almost exclusively in grasses (51). The biosynthesis of lignin is caused by a free radical coupling reaction, where enzymes break the covalent bond between the phenolic oxygen and its hydrogen. This broken bond results in a reactive free radical that attempts to stabilize itself by moving the electron throughout the molecule, in a process called resonance stabilization. When one free radical comes into contact with another they combine to form a new covalent bond. This dimer still has a free phenolic hydroxyl and an enzyme can remove the hydrogen and electron to make an additional free radical. This process of addition then continues causing the formation of a larger, randomly polymerized structure (55).  Lignin has found been found to be heterogeneous in structure, consisting of amorphous regions as well as more structured oblong particles and globules (56). Studies have also shown that the structure and chemical composition of lignin is strongly correlated to the nature of the polysaccharide matrix in which lignin is polymerized. Despite the fact that lignin is hydrophobic in nature, there have been molecular simulations that have shown that the hydroxyl and methoxyl groups in lignin precursors interact with cellulose microfibrils. Cellulose microfibrils can hold significant amounts of water in their amorphous and void fractions (51). 15                           p-Coumaryl alcohol         Coniferyl alcohol        Sinapyl alcohol Figure 1.4 Three common phenyl-propane precursors of lignin  The whole lignin molecule is a three-dimensional, complex framework, consisting of thousands of phenyl-propane units (57) (58), formed by the unsystematic polymerization of C6-C3 units, which leads to the strengthening of cell walls in plants, and a subsequent difficulty in degrading the polymer to low molecular weight fragments (59) (60). A possible structure of lignin is shown in Figure 1.5.  16   Figure 1.5 Possible lignin structure  Lignin is the second most abundant organic chemical on earth after cellulose, and therefore plays a key role in any process utilizing cellulosic fibres. In the pulp and paper industry, lignin is currently recovered from the black liquor and can either be burned in the recovery boiler or utilized as a source for specialty chemicals like detergents or biosorbents (61). A similar trend can be observed in cellulosic ethanol processing, with plans calling for the burning of lignin to generate steam and heat to run the process (51).  17  1.2.4 Extractives  A small portion of plant biomass consists of hydrophobic or hydrophilic compounds that are not an integral part of the cellular structure, which are called extractives (62). It is possible to extract these compounds using polar and non-polar solvents without degrading the structure of the biomass. The variety of extractives found in a biomass sample is dependent on the source of the sample (63) (51).  Different categories of extractives include: fats, waxes, fatty acids, alcohols, and resin (which include terpenes, lignans, and other aromatics). Extractives typically constitute 4-10% of the total dry weight of cellulosic material, and for tree species there are some extractives that provide a toxic barrier to attacks from termites and fungi (64).  1.3 Enzymatic hydrolysis  There are numerous factors that need to be in place for cellulosic ethanol to emerge commercially on an industrial scale. The traditional cost ‘road-block’ in the three step bio- ethanol process of pretreatment, enzymatic hydrolysis, and fermentation has been the hydrolysis step. This has been due to the high price of cellulase enzymes capable of hydrolyzing lignocellulose. The determination of this enzyme cost is further complicated by the need to fully understand the influence of various pretreatments and feedstocks on enzyme efficiency. A four year study undertaken by Novozymes and the National Renewable Energy Laboratory (NREL) between 2001-2005 was able to successfully reduce the cost of enzymes capable of hydrolyzing cellulosic waste by 30-fold, to a final price range of $0.10-$0.18 per gallon (65). Depending on the activity of the enzymes in question this translates into approximately 7.9E-7 $/FPU, based on a price and activity of $0.18 per gallon and 60 FPU/mL. However, there has been an acknowledgement by Novozymes that the realized enzyme cost will be dependent on the enzyme dose and the yield of ethanol obtained in the process, which is highly dependent on the specific pretreatment and fermentation process applied (66). In March 2007, Novozymes outlined a five- step approach that it is following to develop cost-effective cellulosic ethanol:  18  1. Continued funding of research and development (specifically in the areas of biomass conversion and the development of a commercial process technology). 2. Establishment of flexible configuration testing and development centers, geographically distributed to address multiple types of biomass feedstock and integrate processes (pretreatment, hydrolysis and fermentation). 3. Scientific advancement to increase cost efficiency by improving underlying agricultural practices (collection and harvest of biomass) and pre-treatment methods. 4. Scientific advancement in biotechnology (including enzyme technology, metabolic engineering and novel separation methods). 5. Continued bi-partisan support of a national infrastructure to support practical implementation (including funding, incentives and tax credits) (67)  A Canadian biotechnology company, Iogen, is currently involved in the production of bio- ethanol. Iogen is currently operating a demonstration-scale cellulosic ethanol facility in Ottawa that is capable of producing 3-million liters of ethanol annually from wheat straw. In March 2008, it was announced that Iogen Corp.’s application to receive funding for the country’s inaugural commercial-scale cellulosic ethanol plant has reached the due-diligence stage (68). This Saskatchewan-based cellulosic ethanol bio-refinery, partially funded by Sustainable Development Technology Canada, will consume about 750 tons of feedstock/day, and will be capable of producing approximately 75-million liters of ethanol annually (69).  Perhaps in the future the 3-step process for bio-ethanol will be reduced to 2-steps, with enzymes capable of degrading the recalcitrant cell walls of lignocellulose, thus eliminating the need for pretreatment. Researchers at the University of Maryland have been working on enzymes that can achieve this reduction in the number of process steps, and have a developed a bacterium called Saccharophagus degradans, which is touted as being capable of efficiently breaking down any form of biomass into sugars (70). However, until such enzymes are commercially available, a separate pretreatment to increase the “accessibility” of the cellulose to enzymes, followed by the use of a highly potent and efficient enzyme system will be necessary.  19  1.3.1 Cellulases  The properties of cellulase can vary depending on their source, however most of the microbial cellulases that have been researched have been found to be acidic proteins containing large quantities of carbohydrates (71) (72). Temperature and pH optima have been determined for most cellulases to fall in the range of 40 – 50oC and 4 – 5 on the pH scale (73).  A more detailed explanation of hydrolysis of native cellulose by microbial enzyme systems involves breaking the enzymes down into three categories: endoglucanase, exoglucanase, and β- glucosidase. Endoglucanases randomly target for hydrolysis the amorphous cellulose regions and the soluble derivatives of cellulose (74). The endoglucanase reaction cleaves β-1-4-glycosidic bonds, resulting in cellodextrins (glucose polymers of varying length) and reactive chain ends, with only a small release of reducing sugars (75) (76). Exoglucanases, also called cellobiohydrolases, then act on the reducing and non-reducing chain ends of the cellulose chain, cleaving cellobiose molecules and converting cellodextrins to cellobiose (76) (77). The hydrolytic process is then completed by β-Glucosidase, which hydrolyses the cellobiose into glucose. These general classifications of cellulases are still widely used, however, due to the varying complexity of different substrates the specific mechanisms of the hydrolytic process have been found to be more complex (78). This process, previously described by Gregg et al. (79), is outlined in Figure 1.6, which shows cellulase and β-glucosidase completely hydrolyzing an ideal cellulose substrate into monomeric glucose units.  20   Figure 1.6 Enzymatic action during hydrolysis  The activity of the endo- and exoglucanase, as well as β-glucosidase enzymes, are measured using independent assays. A common method for measuring endo- and exoglucanase activity is the filter paper test, which measures the enzyme’s ability to hydrolyze pure cellulose in the form of a Whatman No.1 filter paper strip. The unit of measured activity is the filter paper unit (FPU), and is defined by the International Unit (IU) to be 0.37 divided by the concentration of enzyme that releases 2.0 mg of glucose. The 0.37 is derived from the factor for converting the 2.0 mg of “glucose-equivalents” generated in the assay to mmoles of glucose (2.0 / 0.18016), from the volume of the enzyme being tested that is used in the assay (0.5 mL), and from the incubation time (60 minutes) required for generation of the reducing equivalents (80).  mLmin mol37.0 min)60dilutionenzymemL5.0( )mol/ecosglumg18016.0/ecosglumg0.2( ⋅ µ = × µ  21  The cellobiose unit (CBU) is used to measure the activity of β-glucosidase.  The CBU is based on the enzymes ability to release p-nitrophenol from a stock solution of p-nitrophenyl-β-D- glucoside (81).  1.3.2  Factors that limit enzymatic hydrolysis  Enzymatic hydrolysis involves the breaking down of the cellulose polysaccharide chain into individual sugar monomers. The cellulosic hydrolysis yields that can be obtained are primarily affected by the type of substrate, for instance softwood substrates have been found to be more recalcitrant to lignin removal and enzymatic hydrolysis than hardwood substrates (82). Research is continually searching for consensus on which factors have the greatest influence on the hydrolysis rate, and has currently grouped the limiting factors into two broad categories: structural substrate factors and enzymatic interaction and mechanistic factors. The structural factors that limit hydrolysis, according to Mansfield et al. (76), are listed in Table 1.4.  Table 1.4 Hydrolysis limiting structural properties of cellulosic fibres Structural Level Substrate Factor Microfibril DOP: degree of polymerization  Degree of crystallinity  Structure of cellulose lattice Fibril Lignin content and distribution  Particle size Fibre Available surface area  Degree of fibre swelling  Pore structure and distribution  As the conversion of lignocellulosic substrate increases, it has been found that the rate of hydrolysis decreases (83).  A contributing factor to this decline in hydrolysis rate is the decreasing concentration of the readily hydrolysable amorphous cellulose, and a corresponding increase in concentration of the more recalcitrant crystalline cellulose (84). During hydrolysis, 22  amorphous cellulose has been reported to be degraded quickly to cellobiose, leaving the rate of hydrolysis dependent on the crystallinity and DP (85) (86).  The enzyme-accessible surface area has been regarded as crucial in the rate of hydrolysis because enzyme adsorption to cellulose is a required reaction step (87). A high degree of crystallinity corresponds to cellulose molecules that are hydrogen bonded and tightly packed together, making them less accessible to enzymatic attack than the amorphous cellulose portion of the substrate (88) (89). However, it has been suggested that pretreatments that aim to reduce cellulose crystallinity also result in an increase in substrate specific surface area, making it difficult to separate these two limiting factors. There is some evidence that links decreasing particle size with increasing hydrolysis rates, including a study that found that smaller fractions of pulp were hydrolyzed first during the initial reaction stages (90). Additional studies have also shown that small particle sizes result in increased hydrolysis yields over a range of steam pretreatment conditions (91) (92).  As insoluble cellulosic substrates are hydrolyzed they form soluble molecules with a lower DP, eventually generating cellobiose and glucose. However, with lignocellulosic material a significant amount of cellulosic material remains recalcitrant (ie. meaning it cannot be degraded) and insoluble after hydrolysis. It has been shown that cellulose can be recalcitrant to hydrolysis above a definite molecular weight range, however it is not clear whether the DP of cellulose is itself a limiting factor, or if it is coupled with other factors like crystallinity and enzyme accessible surface area (93) (94).  The quantity of lignin plays a large role in how well enzymes can penetrate and hydrolyze a substrate. It has been found that the rate of enzymatic hydrolysis and lignin content have an inverse relationship (95) (96). The presence of hydrophobic groups and/or the charge on enzymes can lead to their adsorption and non-productive binding to lignin or cellulose (51). In order to achieve complete hydrolysis it is not necessary to completely remove the lignin, however, enough must be removed to account for enzyme adsorption to isolated lignin and recalcitrant cellulose residues formed during hydrolysis (97) (98). A reduction in the rate of hydrolysis of lignocellulosic substrates can occur at low enzyme levels from irreversible enzyme 23  adsorption to lignin, making those bound enzymes unavailable for any supplementary reactions with cellulose. Various pretreatments seek to either dissolve lignin (chemical) or separate lignin from the cellulose (physical), or a combination of the two, aiming to remove sufficient lignin to achieve a desired hydrolysis yield (99). The degree of fibre swelling and an increase in pore size have also been shown to increase following lignin removal, leaving the fibre more accessible to enzymatic attack (100) (101).  Enzyme factors that are capable of influencing hydrolysis include end-product inhibition, thermal inactivation, and irreversible enzyme adsorption (102) (103). Synergism between β- glucosidase and cellobiohydrolases can minimize the effects of end-product inhibition by the hydrolysis of cellobiose (104). The thermal inactivation of cellulase can be minimized by operating under hydrolysis conditions that have been determined to be optimal for a given enzyme. The hydrolysis of cellulose is a unique enzymatic reaction because it involves the soluble enzyme binding to an insoluble substrate before the hydrolysis can occur. Due to the required adsorption step a lot of research has focused on the adsorption reaction. Two domains have been shown to contribute to cellulase adsorption, the cellulose binding domain and the catalytic domain, which are joined by a linker region (105). Studies have shown that the removal of the cellulose binding domain lowers enzyme hydrolysis efficiency on crystalline cellulose, and the addition of this domain to an enzyme that does not possess one will increase its activity (106) (107). Two additional factors that have been shown to strongly influence the effectiveness of cellulases are the tightness of adsorption to the substrate (108), as well as the distribution of lignin in the substrate (109).  1.3.3 Models of enzymatic hydrolysis  The development of models to predict and understand enzymatic hydrolysis can be dated back to some of the very early work on cellulases (110). A distinction between quantitative models of enzymatic hydrolysis that has been proposed includes: non-mechanistic models, semi- mechanistic models, and functionally-based models (111).  24  Non-mechanistic models, alternatively referred to as empirical models, are based the correlation of experimental data. These models can be useful in predicting results where experimental data does not exist, however, they are unlikely to be reliable under different experimental conditions than those for which the correlation was developed. Non-mechanistic models in literature have been developed to predict the rate of substrate reaction and fractional conversion. Models that predict substrate conversion as a function of enzyme loading and substrate concentration have been developed by Sattler et al. (1989) (112). Additional non-mechanistic models that predict hydrolysis rate as a function of hydrolysis time and enzyme loading have been developed by Karrer et al. (1925) and Miyamoto et al. (1942) respectively (110) (113).  Semi-mechanistic models feature an enzyme adsorption model, and utilize concentration to describe the state of the substrate during hydrolysis (111). Holtzapple et al. (1984) developed a model to predict the initial rate of hydrolysis that was able to agree well with experimental data over various hydrolysis experiments with a 10-fold range in enzyme concentration, and a 30-fold range in cellulose concentration (114). Additional two substrate models have been developed based on the theory that cellulose can be partitioned into reactive amorphous fraction, and a less reactive crystalline fraction (Fan and Lee, 1983) (Scheidling et al. 1984) (115) (116).  The development of functionally based models has proven useful in understanding enzymatic hydrolysis at the level of structural substrate features, multiple enzyme activities, and accounting for rate-limiting factors (111). A model by Suga et al. (1975) predicts that substrate DP changes as a function of time in the presence of endoglucanase, and that for the degradation of longer chain cellulose molecules there is a synergy between exoglucanase and endoglucanase (117). Another functional based model developed by Converse et al. (1993) predicted competitive adsorption between exoglucanase and endoglucanase when cellulase is in considerable excess relative to the substrate (118).  1.4 Biomass pretreatment  There are numerous structural and compositional features of lignocellulose that have been cited as obstructions to enzymatic hydrolysis. Such features include the degree of cellulose 25  crystallinity, distribution of lignin and hemicellulose, and porosity or enzyme-accessible surface area (119) (120) (51). The structural complexity of lignocellulose varies between substrates, leading to the requirement of specific pretreatments that are able to optimize both the digestibility and fermentability of the substrate. These pretreatments can generally be classified as physical, chemical, or biological pretreatments.  1.4.1 Physical pretreatments  Pretreatment techniques that do not involve the use of any chemicals are called physical pretreatments, which aim to reduce particle size and reduce cellulose crystallinity.  The mechanical pretreatment technique of comminution involves the breaking down of the lignocellulosic material through various milling processes, including dry, wet, and vibratory ball milling (121). The power requirement of mechanical comminution of lignocellulosic substrates depends on the final particle size as well the initial substrate characteristics (122). The energy requirements of mechanical comminution are regarded as high for hardwood, which consumes 130 kWh/ton to reduce the particle size to 1.6 mm. To reduce the size of corn stover with mechanical comminution to 1.6 mm requires far less energy, consuming only 14 kWh/ton (123).  1.4.2 Chemical pretreatments  Pretreatment techniques that involve chemical application are classified as chemical pretreatments. Examples of chemical pretreatments include acid catalyzed steam-explosion, ammonia fibre explosion, CO2 explosion, solvent, dilute acid, and autohydrolysis.  Steam explosion has been researched as a pretreatment alternative to other existing chemical and mechanical techniques. A process flow diagram for the steam explosion of softwood is shown in Figure 1.7. This process utilizes high temperature steam (185 – 260oC) that enters a sealed vessel containing the lignocellulosic material, a subsequent explosive pressure release expands the lignocellulosic matrix resulting in fibre separation (124). Encouraging results have been obtained for steam explosion of hardwood chips and agricultural residues (125) (126), however it has not 26  appeared as effective as a pretreatment for softwoods, which have a more rigid structure and contain lignin that is more recalcitrant (127). It has been found that a catalyst is required to assist in the hydrolysis of the hemicellulose fraction of softwoods (128), which is directly related to the lower acetylated group content in softwoods compared to hardwoods, that assists in auto- hydrolysis of the substrate (127). Two catalysts that have been extensively studied to assist in this hydrolysis are sulfuric acid and sulfur dioxide (129) (130) (131) (132). Advantages attributed to sulfuric acid are its relatively inexpensive cost and effectiveness, however substrates treated with a high concentration of sulfuric acid are poorly fermentable due to the formation of inhibitory compounds. Steam explosion with a sulfur dioxide catalyst does not form as many inhibitory compounds that interfere with fermentation. Sulfur dioxide itself is toxic but not as corrosive, and still offers ease at introducing the catalyst to the substrate (127). Both acid catalyst’s have been studied in one and two stage pretreatments, both exclusively and combined. One stage pretreatment of spruce utilizing H2SO4 or SO2, has produced overall sugar yields (final/initial concentration of hexose and pentose sugars) of 67% and 66% respectively when hydrolyzed with a cellulase loading of 15 FPU/g, and an FPU:CBU ratio of 1:4. A two-stage pretreatment of spruce with SO2 produced an overall sugar yield of 80% of the theoretical value of sugar available, a 14% increase over single stage pretreatment under equal hydrolysis conditions. A two stage process with an H2SO4 catalyst was able to increase overall sugar yield by 10% under the same hydrolysis conditions mentioned above, and further studies by Stenberg and Tengborg have shown the advantages of combining the two acid catalysts in a two-stage pretreatment (133) (134).   Figure 1.7 Process flow diagram for the steam-explosion of softwood 27  Ammonia Fibre Explosion (AFEX) is similar to the process of steam explosion, where the lignocellulosic substrate is exposed to high pressure and temperature, followed by a quick reduction in pressure. However in AFEX pretreatment the lignocellulose is exposed to liquid ammonia rather than SO2 or H2SO4 in the reaction vessel. Typical operating conditions in the AFEX process are 90oC, 30 minute residence time, and a liquid ammonia dose of 1-2 kg ammonia / kg dry biomass (135). AFEX has been an effective pretreatment for the hydrolysis of low lignin content biomass like Bermuda grass (~5% lignin) and bagasse (~15% lignin), achieving over 90% cellulose and hemicellulose hydrolysis of these substrates with an enzyme loading of 5 IU cellulase/g substrate (136). AFEX pretreatment has not been as successful with high lignin content biomass like newspaper (18 – 30% lignin) and aspen chips (25% lignin), achieving only 40% and less than 50% cellulose conversion respectively, with an enzyme loading of 5 IU cellulase/g substrate (137). Ammonia is recovered from this pretreatment due to environmental concerns and as a cost saving measure. The ammonia recovery process strips the residual ammonia from the pretreated lignocellulosic substrate (138), and since no inhibitors are formed during ammonia pretreatment there is no water wash required for the biomass (139) (140). AFEX pretreatment also operates at a lower temperature, however difficulties in recovering all of the ammonia as a reusable gas stream, and its corrosive and toxic nature complicates the process when compared to CO2 explosion (141).  A third chemical pretreatment that utilizes fibre separation via a rapid pressure drop is CO2 explosion. Carbon dioxide molecules are similar in size to those of water and ammonia, thus are able to penetrate the small fibre pores that water and ammonia can. The interest in this pretreatment stemmed from the hope an increased hydrolysis rate could be achieved by the CO2 forming carbonic acid (142). Compared to steam explosion, CO2 explosion offers the benefit of lower operating temperatures (35 – 80oC) which reduces sugar decomposition.  To remove lignin from lignocellulosic material it is possible to utilize solvents (organosolv process), such as methanol and ethanol, to provide good separation of the hemicellulose, cellulose, and lignin. The organic solvents are often combined with inorganic acid catalysts, like sulfuric acid or hydrochloric acid, to break the lignin and hemicellulose bonds. At temperatures above 185oC it was found that no acid catalyst was required to achieve acceptable delignification 28  (143). In order to prevent inhibitory compounds from affecting any downstream processes, like enzymatic hydrolysis or fermentation, the solvents must be evaporated, condensed, and recovered. Although this process can remove large portions of lignin, the organosolv process has traditionally only been economically viable for the production of highly pure cellulose (144).  A common process for pretreatment of lignocellulosic material is dilute acid hydrolysis. Research has shown that dilute sulfuric acid pretreatment yields high reaction rates and can considerably improve cellulose hydrolysis (145). Two general dilute acid processes are high temperature (>160oC), for solids loadings of 5-10% by weight (substrate/total mixture), and low temperature (<160oC), for solids loadings of 10-40% by weight (146) (147).  Another chemical pretreatment is called autohydrolysis, which converts lignocellulose into fermentable sugars after exposure to high temperature steam. The steam is used to release acetylated hemicellulose in the form of acetic acid, which then carries out a partial hydrolysis of the sugars present in the substrate (148). The disadvantage of this pretreatment method is that the sugar yields that are obtained are generally low (149).  1.4.3 Biological pretreatments  The biological pretreatment of lignocellulosic material utilizes microorganisms like white, brown, and soft-rot fungi. The white and soft-rot fungi target both lignin and cellulose, whereas the brown-rot fungi only targets cellulose (150). When nitrogen or carbon is limiting, a white-rot fungus called P. chrysosporium produces enzymes (lignin peroxidases and manganese-dependent peroxidases) that degrade lignin (151). Utilizing biological methods would simplify pretreatment through low energy requirements and relatively mild conditions, however low yields and slow reaction rates are unfavorable for industrial scale-up (50). 1.5 Oxygen delignification process  The process of delignification involves the complete or partial removal of the amorphous, highly polymerized lignin from woody plant material. Delignification has traditionally been achieved in 29  the pulp and paper industry by a variety of bleaching and pulping processes under a diverse set of conditions (152) (153). Different delignification processes are outlined in Table 1.5.  Table 1.5 Delignification: Pulping and Bleaching Processes Pulping Processes Bleaching Processes Kraft (sulfate) pulping Chlorine Bleaching Soda (alkali) pulping Chlorine dioxide bleaching Neutral sulfite pulping Oxygen bleaching Acidic sulfite pulping Ozone bleaching   Alkali-treatment   Hydrogen peroxide bleaching   Hypochlorite bleaching   In a pulp and paper operation, the process of oxygen delignification can be considered partly as a continuation of the cooking process, and partly as the first stage of the bleaching process. Unbleached kraft pulp typically has a lignin content of 3-5%, which after oxygen delignification is usually decreased to about 1.5% (152). Oxygen delignification involves the removal of residual lignin in unbleached pulp through the application of alkali and oxygen. The application of oxygen delignification is compatible with the Kraft cooking process due to the fact that both the effluent and black liquor are caustic, and can consequently be combined and processed through the recovery boiler (154).  Increasing environmental pressure on the pulp and paper industry has led to the use of oxygen delignification in modern pulp bleaching operations (155). The driving force for increasing environmental regulations on bleach plants was due to the discharge of chlorinated organic compounds, thus focusing efforts on delignifying pulp to a higher degree prior to bleaching (154). Today, oxygen delignification is a standard stage in the production of bleached chemical pulp, and in addition to the reduction in organic chlorine compounds, it also reduces effluent BOD, colour, and chemical oxygen demand (COD) (156). The removal of lignin prior to bleaching also offers the potential of an energy source, as the dissolved material is not 30  contaminated by chlorine ions, which presents the option of redirecting this material to a recovery furnace. Further cost savings in bleaching chemicals is proportional to how much lignin is removed in the oxygen stage. Specific savings can be made in chlorine, caustic, and chlorine dioxide (157).  Previous work on oxygen delignification pretreatment by Draude et al. (2000), focused on a number of substrates including kraft pulp, pulp mill primary clarifier sludge (PCS), and steam- exploded Douglas fir chips. Oxygen delignification removed up to 67% of lignin from softwood pulp, and improved the hydrolysis yield by 174%. The hydrolysis yield of PCS was also improved by up to 90% after delignification, however a decrease in hydrolysis yield was found for steam-exploded Douglas fir chips (158). Another study, performed by Charles et al. (2003) on hemlock pulp, showed an increase in hexose sugar yield from 26 to 47% for pre and post oxygen delignified pulps (159).  Two different oxygen delignification units traditionally applied commercially are referred to as medium and high consistency systems. The key commercial parameters that are optimized for efficient delignification are temperature, caustic charge, and degree of mixing. It is important to note that there is a contrast between the industrial reactors being discussed and the lab-scale reactors typically used at a research scale. In a lab reactor there is constant mixing and an excess volume of oxygen around the cellulosic substrate, however, in industrial reactors there are oxygen bubbles of varying sizes dispersed through the reactor which makes mixing a critical parameter at this scale (160).  Commercial medium consistency processes operate with a pulp consistency in the 8-16% range. Application of intensive mixing in this consistency range offers an increased water volume compared to high consistency processes, which is able to dissolve the oxygen that is required for the delignification reaction.  The high consistency process operates with a pulp consistency of 20-28%, removing the majority of the free liquid phase. The high consistency process has the advantage of providing a large liquid-gas interfacial area and also reducing the liquid layer thickness, thus making it easier for 31  oxygen to diffuse and contact the fiber. Further advantages of the high consistency process include reduced amounts of dissolved oxidizable material to recycle, and a lower volume of water that needs to be heated to reaction temperature (156). Despite these advantages, the majority of new oxygen delignification units (1985 to present) are medium consistency, primarily due to a lower capital cost, since a press is not required prior to the reactor (161) (157).  Additional reasons for the trend towards medium consistency units include an increased simplicity in stock handling due to novel mixing and pumping equipment, and improved selectivity (ratio of attack on lignin to attack on carbohydrates). However, medium consistency processes have a relatively higher chemical consumption, and an extended retention time is required to reach an equal level of delignification (157) (162). The next section describes the basic process steps of these systems, which has previously been described by Tench et al. (157). The process flow diagrams and typical operating conditions of the medium and high consistency systems outlined in Figures 1.8 and 1.9.  1.5.1 Medium and high consistency oxygen delignification systems High Consistency System 1. A press is used to achieve a consistency increase from 3 – 30% 2. Fresh caustic or oxidized white liquor is added to the pulp at the press discharge 3. A thick stock pump transfers the pulp to a fluffer via a feed pipe in which a gas tight plug is formed 4. The fluffed pulp flows down the pressurized reactor as a loose bed and reacts with oxygen (tray-like high consistency reactors are also common) 5. Steam is injected into the top of the reactor to maintain temperature 6. Oxygen is added to the top or bottom of the reactor to maintain oxygen partial pressure 7. Reacted pulp is diluted to 6% consistency with post-oxygen filtrate and is discharged to a blow tank where dissolved gases are discharged 8. Pressure control in the reactor is achieved by releasing gas from the head space (157)    32  Medium Consistency System  1. Pulp is fed from brownstock washer or decker at 10-14% consistency, and is charged with caustic or oxidized white liquor 2. Pulp is then pumped through one or more medium consistency gas mixers to an upflow pressurized reactor 3. Oxygen and steam at medium-pressure are added upstream or directly into reactor 4. Pulp is discharged from the top of the reactor, possibly assisted by post-oxygen filtrate 5. Stock is depressurized and blown through a separator, where by-product gases, inert gases, oxygen, and possibly steam (depending on reactor temperature) are released 6. Magnesium compounds are normally used to preserve pulp quality, particularly when intensive delignification is required. The magnesium compounds inhibit degradation caused by reactive radicals that are formed by the decomposition of peroxides catalyzed by metal ions present in the pulp. The inhibition of degradation has been explained by a co-precipitation of transition metal ions with magnesium hydroxide which stabilizes hydrogen peroxide against decomposition (157) (163).  33   Fi gu re  4 : T yp ic al  H ig h- Co ns ist en cy  D el ig ni fic at io n Sy st em  w ith  T yp ic al  O pe ra tin g Pa ra m et er s ( 15 7)   M ag ne siu m Br ow n- st oc k Sh ow er Br ow n- st oc k Br ow n- st oc k di lu tio n Tw in  ro ll pr es s FI C To rq ue FI C FX LI CLX O xy ge n Bl ow  ta nk Po st -o xy ge n filt ra te Po st -o xy ge n w as he r FI C FI C Pr es sa te ta nk O xy ge n Al PX PI C Ve nt St ea m TX TI C FI C FX N aO H FX FI C FXFI C C sX C sI C Bl en d ch es t Oxygen  H ig h C on sis te nc y Pu lp  C on sis te nc y, %  25 -2 8 D el ig ni fic at io n,  %  45 -5 0 Re te nt io n tim e,  m in  30  In iti al  T , o C 10 0- 10 5 In le t P , k Pa  50 0- 60 0 O ut le t P , k Pa  50 0- 60 0 St ea m  C on su m pt io n,  k g/ m et ric  to n  Lo w  P re ss ur e (4 50  k Pa ) - M ed iu m  p re ss ur e (1 14 0 kP a)  75 -1 00  Ev ap or at or  (4 50  k Pa ) 30 -5 0 Po w er  C on su m pt io n,  k W h/ m et ric  to n 40 -5 0 A lk al i C on su m pt io n,  k g/ m et ric  to n 21 -2 3 O xy ge n C on su m pt io n,  k g/ m et ric  to n 20 -2 4 M ag ne siu m  io n,  k g/ m et ric  to n 0. 5  Fi gu re  1 .8  F lo w sh ee t o f a  ty pi ca l h ig h co ns ist en cy  o xy ge n de lig ni fic at io n sy st em  34   Fi gu re  4 : T yp ic al  M ed iu m -C on sis te nc y D el ig ni fic at io n Sy st em  w ith  T yp ic al  O pe ra tin g Pa ra m et er s ( 15 7)   O xy ge n m ixe r PI C O xy ge n R ea ct or M ag ne siu m FX FI CFX C oo le r TXTI C Br ow ns to ck w as he r Br ow ns to ck Se al  ta nk FX FI C NaOH Steam Oxygen FX F IC St ea m m ixe r M ed iu m C on sis te nc y pu m p FX FI C FX F IC O xy ge n re ac to r TX TI C St ea m PI C C W S C W S Post-oxygen washer Post-oxygen filtrate O xy ge n Bl ow  ta nk C sI C C sX  M ed iu m  C on sis te nc y Pu lp  C on sis te nc y, %  10 -1 2 D el ig ni fic at io n,  %  40 -4 5 Re te nt io n tim e,  m in  50 -6 0 In iti al T,  o C  10 0- 10 5 In le t P , k Pa  70 0- 80 0 O ut let  P , k Pa  45 0- 50 0 St ea m  C on su m pt io n,  k g/ m et ric  to n  Lo w  P re ss ur e (4 50  k Pa ) 70  M ed iu m  p re ss ur e (1 14 0 kP a)  20 0- 30 0 Ev ap or at or  (4 50  k Pa ) 90 -1 00  Po w er  C on su m pt io n,  k W h/ m et ric  to n 35 -4 5 A lk ali  C on su m pt io n,  k g/ m et ric  to n 25 -2 8 O xy ge n Co ns um pt io n,  k g/ m et ric  to n 20 -2 4 M ag ne siu m  io n,  k g/ m et ric  to n 0. 5  Fi gu re  1 .9  F lo w sh ee t o f a  ty pi ca l m ed iu m  c on sis te nc y ox yg en  d el ig ni fic at io n sy st em  35  1.5.2 Oxygen delignification chemistry  The chemical pretreatment that was utilized for lignin removal during this research work was oxygen delignification. The process of oxygen delignification involves the suspension of a partially delignified lignocellulosic fibre in an alkaline solution, then pressurizing this solution in a reactor where the lignin will take part in reactions with oxygen radicals until it dissolves.  The primary chemistry involved in oxygen delignification has been the focus of extensive study, with a general review provided by Gierer et al (53). Lignin degradation in an oxygen-alkaline environment occurs through the interaction of oxygen and hydroxide with the phenolic hydroxyl group on lignin. The phenolic hydroxyl group first reacts with alkali, which leads to the formation of a phenolate ion, which then goes on to react with oxygen, forming an intermediate called hydroperoxide. The fragmentation of this intermediate hydroperoxide then occurs through numerous possible pathways, including reactions that lead to the break-down of lignin’s polymeric structure, which develops more water-soluble reaction fragments and side chain scission. These degradation pathways lead to the formation of carboxylic acid in addition to the ultimate degradation of lignin (164) (161).  The four-electron reduction of molecular oxygen to water is the driving force behind oxygen delignification. The reaction is summarized below in Equation 1.  O2 + 4e- + 4H+ Æ 2H2O                                                    [1] As shown below in Figure 1.10, the reduction of oxygen proceeds in four consecutive electron steps.  Figure 1.10 Reduction of molecular oxygen in four consecutive steps 36  Due to the complexity of lignin’s structure, and the large number of associated degradation reactions, any attempt to describe the delignification process chemically requires simplification. A method proposed by Grier et al. (165) involved two simplifications; the first is to reduce lignin to a single unit (Figure 1.11), and the second is to divide the reactions into nucleophilic and electrophilic categories.  Figure 1.11 Single unit of complex lignin structure (Arylpropane), Aryl: Any functional group derived from a simple aromatic ring, Aroxyl: Includes all aromatics with a benzene oxygen linkage, Alkyl: Only C and H in a chain (not all valence e- are used in bonding)  In a nucleophilic reaction the delignifying reagent provides the electron pair that forms a bond to lignin, and in electrophilic reactions the delignifying reagent accepts the electron pair to form such a bond. It has been shown that the process of removing lignin involves coupled electrophilic and nucleophilic reactions (165). According to Gierer, the oxygen biradical and the hydroxyl radicals are electrophilic, and thus will react with the high electron density areas in lignin. In addition, the hydroperoxide anions and super-oxide anion radicals are classified as nucleophilic, and thus will react with low electron density centers in lignin (166) (167). The areas of high and low electron density in a lignin unit are outlined in Figure 1.12.      37   Figure 1.12 High (-) and low (+) electron density centers on lignin units  The importance of the phenolic structure and the hydroxyl group content on the extent of delignification has been studied. The phenolic structure of lignin has been shown to influence the degree to which lignin responds to oxidants. In NMR studies on free phenolics in residual lignin it was found that over 50% of the free phenolics were resistant to oxidation, while the condensed phenolic concentration remained comparatively constant displaying a complete resistance to degradation (57). Phenolic subunits of lignin in the condensed and non-condensed form are shown in Figure 1.13.   Figure 1.13 Condensed and non-condensed phenolic subunits of lignin, A: 5- Condensed phenolic subunits of lignin B: Non-condensed (free) phenolic subunits of lignin A B 38  1.5.3 Cellulose reactions during oxygen delignification  The reactions involved in oxygen delignification are still not fully understood, however it is known that the same reactions that remove lignin from cellulosic fibers also result in carbohydrate degradation, leading to a degradation of fibre strength. There are two main types of cellulose degradation reactions that occur during oxygen delignification, specifically random chain cleavage and endwise peeling.  Random chain cleavage, shown in Figure 1.14, occurs through the oxidation of a hydroxyl group to a carbonyl group, which is a carbon atom double bonded to an oxygen atom. The carbonyl group then undergoes a beta-elimination reaction which severs the glycosidic linkage (168).   Figure 1.14 Oxidation and cleavage of the cellulose chain  Cellulose endwise peeling occurs in alkaline media and is responsible for a loss in yield. However, less importance is placed on the peeling reaction because there needs to be excessive chain cleavage before it is responsible for excessive degradation. For endwise peeling to occur there must be a carbonyl group on the end group, and since this reaction will result in a new end unit containing a carbonyl group this process is self propagating (168). What prevents this reaction from completely dissolving all of the cellulose is a competing reaction, termed as the 39  stopping reaction, which converts the end unit into one without a carbonyl group (169) (170). The cellulose reactions that occur in an alkali media are summarized in Figure 1.15.     Figure 1.15 Features of cellulose degradation in alkali      40  2. OBJECTIVES  There has been extensive research focusing on the pretreatment and conversion of lignocellulosic substrates into fermentable sugars. Recovered fibre represents a model substrate that has not previously been evaluated with a combination of oxygen delignification pretreatment and subsequent enzymatic hydrolysis. The primary objective of this research was to evaluate the effectiveness of oxygen delignification as a pretreatment for recovered fibre. The independent variables that were studied included the reaction temperature, caustic concentration, and time. The impact that these variables had on the oxygen delignification yield and kappa number was then used to develop empirical equations to model the role of lignin. A secondary objective of this research was to utilize enzymatic hydrolysis for the conversion of the recovered fibre. A range of enzyme and solid loadings was studied, and the results allowed for mathematical equations to be developed that described the sugar conversion as a function of the independent oxygen delignification variables. These results can be used to optimize the pretreatment and enzymatic hydrolysis of recovered fibre under the range of conditions studied. The third objective of this research was to look at a series of practical issues involved in the lignocellulosic bioconversion process. These practical issues included the feedings regimes of substrate and enzymes during the hydrolysis reaction, precipitation of lignin from oxygen delignification liquor, and the washing of pretreated recovered fibre after oxygen delignification. It was hoped that general trends obtained from these practical issues could provide insight into the scale-up of lignocellulosic bioconversion process steps.       41  3. MATERIALS AND METHODS  3.1 Oxygen delignification setup and procedure  3.1.1 Laboratory scale oxygen delignification apparatus  A bench-top laboratory scale oxygen delignification reactor was used to decrease the lignin content of cellulosic fibres. The reactor used was a PARR Series 4520 HastAlloy C-2000, 1L reactor that was applied for the small-scale simulation of one-stage industrial oxygen delignification, shown in Figure 3.1. The bench-top reactor was equipped with a fixed head that included a thermowell, a gas inlet and outlet valve, and pressure gauge, and a rupture disc. A charged reactor had the temperature and mixing speed controlled by a 4843 PARR controller (171).     Figure 3.1 has been removed because of copyright restrictions. The information removed is a visual representation of the side-profile of the lab-scale oxygen delignification reactor with reaction vessel removed. The original source of the material is: PARR Series 4520 Operating Manual. PARR Instruments. [Online] March 2008. [Cited: April 25, 2008.] http://www.parrinst.com/doc_library/members/383M.pdf (171).        Figure 3.1 Side-profile of the lab-scale oxygen delignification reactor with reaction vessel removed (171)  42  3.1.2 Procedure for oxygen delignification experiments  The reaction vessel was charged with a solids loading of 2% so that the light duty magnetic drive could provide adequate mixing. The fibre intended for delignification was weighed and added to the reaction vessel along with the desired load of caustic and water to raise the total weight to 750 g. The reactor was then sealed by enclosing the split ring around the reaction vessel and tightening the cap screws, as well as sliding the safety drop ring around the split ring and tightening. The sealed reaction vessel was then mounted on the reaction stand and the heating element was raised and plugged into the temperature controller. The gas inlet was attached to a nitrogen gas feed line that was used to sparge the reaction vessel for five minutes to remove oxygen. The reaction vessel was then resealed and the temperature was set. The necessary time was then given for the vessel to reach the reaction temperature.  The reaction temperature was varied between 90 and 150oC for different delignification experiments. The reaction time and caustic concentration were also varied from 20 to 60 minutes and 2 to 20% per gram of oven dried fibre. The oxygen partial pressure and reactor mixing speed were kept constant at 120 psig and 150 rpm for all experiments.  Once the reaction temperature was reached, the gas inlet was attached to an oxygen gas feed line at 120 psig. The pressure relief valve was slightly opened to keep a steady flow of fresh oxygen to the reactor and to allow combustible byproducts to be released. Once the desired reaction time was completed the heating mantle was lowered, the reactor mixing was turned off, and the reaction vessel was quickly cooled in a bucket of ice and water. Once the reaction vessel had cooled to 90oC the outlet valve was opened to accelerate the pressure reduction.  Once the reaction vessel was cooled and the pressure reached zero the lid was removed and the fibre suspension was removed and filtered. The filtration of the fibre suspension was done under vacuum using a Buchner funnel and a Whatman No. 1 filter paper to retain the fibre. The pulp was washed with 4 L of water, and the suction continued until no more water could be withdrawn. The washed fibre was then removed and stored in a sealed plastic bag and placed in a 43  refrigerator at 4oC. Two samples of approximately 3 g wet fibre were removed from each washed fibre sample and dried overnight at 105oC in order to determine percent moisture. 3.2 Procedure for lab-scale enzymatic hydrolysis  The enzymatic hydrolysis of various fibre samples was carried out in 250 mL Erlenmeyer flasks at consistencies ranging from 2 to 10%. The buffer used for hydrolysis was 50mM sodium acetate with a pH of 4.8. Prior to hydrolysis the flasks containing fibre and buffer were incubated for 20 minutes at a speed of 200 rpm. For fed-batch hydrolysis experiments additional wet fibre, and in some instance fresh enzyme, was added to the flasks once the hydrolysis mixture was adequately liquefied. The commercial cellulase that was used during experimentation was Celluclast, which had an activity of 41.54 FPU/g and was applied during hydrolysis in loadings of 2 to 111 FPU/g. The Celluclast also had a β-glucosidase activity of 20.12 Cellobiose units (CBU) per milliliter. A commercial β-glucosidase enzyme, Novozym, with an activity of 640.5 CBU per milliliter was also added to hydrolysis mixture so that the FPU:CBU ratio was 1:5. The increased β- glucosidase loading was to ensure that there was no end-product inhibition caused by cellobiose and other oligosaccharides. The hydrolysis reaction was carried out for a total duration ranging from 48 to 104 h. During hydrolysis, 1 mL samples were withdrawn at various time points (typically 1, 4, 8, 24, and 48 h). The samples were immediately centrifuged at 9447×g (13,000 rpm) for five minutes in a micro- centrifuge tube, and then the supernatant was withdrawn and frozen at -20oC in fresh micro- centrifuge tubes. These frozen hydrolysis samples were later withdrawn for high pressure liquid chromatography (HPLC) analysis. The initial rate of hydrolysis was measured by determining the sugars liberated during the first hour of hydrolysis. The hexose sugar yield, which was used to represent the effectiveness of hydrolysis, was determined by measuring the hexose sugars present at the end of hydrolysis and comparing this to the initial fibre mass. 44   A large hydrolysis component of the experimental work was the factorial design that looked at the hexose yield of pretreated recovered fibre. The hydrolysis conditions for the factorial design are shown in Table 3.1. Table 3.1 Experimental design: enzymatic hydrolysis conditions Enzymatic Hydrolysis Conditions High Consistency (10%) Low Consistency (2%) Enzyme Loading (FPU/g) 16 20, 40 Total hydrolysis time (h) 48 48 Sample points (h) 48 1, 4, 8, 24, 48 Hydrolysis Temperature (oC) 50 50 Hydrolysis pH 4.8 4.8 Incubator Shaking Speed (rpm) 200 200  4 ANALYSES  4.1 Procedure for determining fibre composition  The composition of fibre refers to the fraction of hexose sugars, pentose sugars, and lignin present in the fibre sample. The composition of untreated recovered fibre was determined using TAPPI method T 222 om-98, for Acid Insoluble Lignin.  The lignin analysis began by weighing a fibre sample equivalent to 0.15 g of dry fibre. This fibre sample was added to a reaction flask along with 3 mL of 72% (w/w) sulfuric acid and constantly stirred for two minutes. For the next two hours the reaction mixture was stirred every ten minutes with a glass rod, after which the reaction mixture was transferred to a serum bottle. The reaction flask was then washed with deionised water and the contents were also transferred to the serum bottle so that the final acid concentration was 4% w/w. The serum bottle was sealed with a butyl rubber septum and crimped aluminum seal, and then autoclaved at 121oC for one hour.  45  After the autoclaved serum bottle cooled, the contents were suction-filtered through pre- weighed, medium coarseness glass crucibles. It was ensured that all of the solid content present in the serum bottles was recovered. The filtrate was collected to determine concentration of acid soluble lignin in the fibre sample based on Equation 2.  df ab A)L/g(estimated,ASL × × =                                            [2]  A = absorbance at 205 nm df = dilution factor b = cell path length, 1 cm a = absorptivity, equal to 110 L/g-cm unless experimentally determined for a given sample  HPLC analysis was used to determine the carbohydrate composition of the filtrate. The acid insoluble lignin was washed with deionised water to remove any traces of acid and the crucible with the lignin was dried overnight at 105oC in an oven. After cooling the in a dessicator, the crucible was weighed to determine the mass of acid insoluble lignin in the recovered fibre sample.  4.2 Procedure for determining sugar composition from a hydrolysate sample  The concentration of hexose and pentose sugars in a liquid hydrolysate sample was measured by HPLC, using a Dionex DX600 system.  Analysis of hydrolysates began by thawing the frozen samples, and then diluting them with deionised water to a concentration that fit within the standard calibration curve. A 5 mg/mL internal fucose solution was used as an internal standard. The diluted samples were then added to 1.5mL HPLC screw-top vials and sealed with septa-lined lids. An autosampler, Dionex AS50, was used to withdraw a 20µL sample from the HPLC vials and the sugars (fucose, arabinose, galactose, glucose, xylose, and mannose) were separated using a Dionex CarboPac PA1 column. 46  Deionised water than had also be degassed was used as the mobile phase, with a flow rate 1.0 mL/min.  The analysis time for each sample was 60 minutes, with the resulting sugar peaks appearing on a chromatogram that was recorded by a Dionex ED50 electrochemical detector. A software program called Chromeleon was then used to measure the area of each peak and determine the concentration of each sugar present in the standards. 4.3 Procedure for determining kappa number  The kappa number is used as an indication of the lignin content of a fibre sample. The procedure used was a TAPPI method for Micro Kappa Number (UM 246), that were carried out at the Weyerhaeuser Technology Centre (WTC), in Federal Way, WA.  The procedure began by the weighing of a dry fibre sample and dispersing it in 80 mL of deionised water. The fibre-water mixture was continuously stirred in order to keep the fibre in suspension. In a separate 50 mL Erlenmeyer flask, 10 mL of 4 N sulfuric acid was mixed with 10 mL of 0.1 N potassium permanganate, and then was added to the fibre-water mixture. The reaction temperature was measured after five minutes, and the reaction was stopped after ten minutes by the addition of 2 mL of potassium iodide. This mixture was then titrated using 0.1 N sodium thiosulphate, and the amount of thiosulphate consumed was recorded. A test blank was also done using the same procedure without the addition of fibre.  The target potassium permanganate consumption for most accurate results is 50%, with a reaction temperature of 25oC. However, the equation for determining the kappa number can be adjusted for permanganate consumption ranging from 10 to 70%, and temperatures between 20 and 30oC. The equation for kappa number is shown below in Equation 3.     47  ))T25(013.01( nw )f100(NumberKappa −⋅+ ⋅ ⋅ =  [3]   Where: f = correction factor for 50% permanganate consumption     w = fibre dry weight, mg     n = amount of 0.1 N potassium permanganate consumed, mL     T = reaction temperature measured after 5 minutes of reaction, oC  4.4 Procedure for the determining of cellulase activity  The cellulase activity was measured using the National Renewable Energy Lab (NREL) method for Measurement of Cellulase Activities (LAP-006), commonly referred to as the filter paper assay.  For each enzyme dilution, three test tubes were prepared containing 1 mL of 50 mM sodium acetate buffer with a pH of 4.8, and a 50 mg strip of Whatman No. 1 filter paper. The test tubes were then capped and placed in a water bath at 50oC for ten minutes to equilibrate. A 0.5 mL sample of the diluted Celluclast solution was then added to the appropriate test tubes and incubated for one hour at 50oC. Additional enzyme and substrate blanks were also included in this assay. After one hour, a 3 mL sample of Dinitrosalicylic Acid (DNS) reagent was added to each test tube and the tube contents were gently stirred to stop the reaction. All test tubes were then placed in a boiling water bath for exactly five minutes to allow the DNS to react with the glucose molecules, which causes colour formation. A 20 mL volume of deionised water was then added to each tube and the absorbance was measured at 540 nm to determine the concentration of glucose present in the solution. The reducing sugar concentration is determined from a standard graph relating the mass of glucose in solution to the absorbance. The definition of a filter paper unit (FPU) is the amount of enzyme required to liberate 2 mg of glucose in one minute from 50 mg Whatman No. 1 filter paper under the conditions of the assay. A typical curve showing the effect of enzyme concentration on the amount of glucose released is shown in Figure 4.1. 48    Figure 4.1 Typical curve for quantifying the cellulase enzyme activity in FPU 4.5 Procedure for determining β-Glucosidase activity  A test tube was prepared containing 1.8 mL of 50 mM sodium acetate buffer (pH 4.8) , and 1 mL of a 5 mM p-Nitrophenyl-β-D-glucoside in 50 mM acetate buffer. The test tube was then placed in a water bath at a temperature of 50oC and equilibrated for 10 minutes, after which 200 µL of a diluted enzyme solution was added. Enzyme and substrate blanks were also included in the assay. The reaction proceeded for 30 minutes, at which point it was stopped by the addition of 4 mL of glycine buffer. The glycine buffer addition causes the p-Nitrophenol to turn yellow. Samples from each test tube were then measured in a spectrophotometer at 430 nm, and the concentration of p-Nitrophenol was determined from a standard graph that relates micromoles of p-Nitrophenol in solution to absorbance. The definition of CBU is the amount of enzyme required to release 1 µmol p-Nitrophenol in one minute under the conditions of the assay. The CBU activity of a typical Novozymes β-glucosidase enzyme over a range of dilutions is shown below in Figure 4.2.  49   Figure 4.2 β-glucosidase activity (CBU) over a range of enzyme dilutions  4.5 Fed-batch and batch hydrolysis of multiple pulp substrate  The substrates chosen for evaluation were untreated recovered fibre (87 Kappa), a pretreated recovered fibre (17 Kappa), and a bleached Douglas-fir pulp (~6 Kappa). For each substrate there were three separate hydrolysis flasks, one containing all substrate and all enzyme at time zero (batch), another flask that was fed substrate during hydrolysis but contained all enzyme at time zero (substrate fed-batch), and finally a flask that was fed both substrate and enzyme during the course of hydrolysis (substrate/enzyme fed-batch).  4.6 Procedure for lignin precipitation from oxygen delignification liquor  The oxygen delignification liquor used for this experiment was derived from the delignification of recovered fibre, with a starting Kappa number of 87. The liquor was filtered in a Buchner funnel under vacuum in order to remove the separate liquid phase from the pretreated fibre.  To observe the effect of pH on the solubility of lignin in delignification liquor a precipitation experiment was performed. Recovered fibre with a Kappa number of 87, corresponding to 13.1% 50  w/w lignin, was delignified under the conditions under severe conditions (150oC, 10% caustic, 60 min) and moderate conditions (120oC, 6% caustic, 40 min). The tannin and lignin assay was used to determine the amount of hydroxylated aromatic compounds present in the delignification liquor before and after the pH was adjusted (172).  The method for precipitation of lignin that was used is analogous to a method for the precipitation of lignin from black liquor (173). The delignification liquor was placed in a 500 mL beaker, and the temperature was raised to 80oC under continuous mixing on a hotplate/mixer. Once the temperature had stabilized the liquor was sparged with carbon dioxide through a porous stone, until the pH reached the lower limit. This pH value was recorded, and then sulphuric acid was used to lower the pH value to approximately 2.5 – 3. Once the pH value had stabilized the value was recorded.  In order to quantify the precipitation of lignin, the liquor was then centrifuged at 1,400×g (5,000 rpm) for ten minutes to separate the liquid and solid phases. A tannin and lignin assay was then performed on the supernatant from the centrifuged liquor sample (174). A stock solution of 1 g/L lignin solution was prepared. Then 50 mL portions of clear samples and standards were brought to 20oC (+/- 2oC), and in rapid succession 1 mL folin phenol reagent and 10 mL carbonate- tartrate reagent were added. Thirty minutes were allowed for colour development, and the reagent blanks and sample were compared at 700 nm in a spectrophotometer.           51  5. RESULTS AND DISCUSSION  Previous work has shown that pulp mill fibre is a suitable substrate for enzymatic hydrolysis (159) (175) (176) (158). In order to gain insight into the effect that the fibre recycling process has on substrate digestibility the model substrate chosen for further analysis was a recovered fibre.   5.1 Analysis of recovered fibre  Recovered fibre was utilized as a substrate for lab-scale pretreatment and enzymatic hydrolysis experiments. Recovered fibre as a feedstock was evaluated based on the degree to which it could be delignified and subsequently hydrolyzed to monomeric sugars. The recovered fibre samples were analyzed to determine the proportions of both carbohydrates and lignin. The hydrolyzability of recovered fibre at enzyme loadings of 2, 5, and 10 FPU/g was determined in order to observe the effect of low enzyme loadings on untreated recovered fibre.  5.1.1 Compositional analysis and kappa number of recovered fibre  The chemical composition of the recovered fibre, shown in Table 5.1, was determined using the acid insoluble lignin assay outlined in Section 4.1. The percentage of lignin present in the recovered fibre is equivalent to a Kappa value of 86.5, based on the correlation shown in Equation 4 (177).  Percent Lignin (%) = (0.15) x (Kappa Number)                                    [4]  Table 5.1 Chemical composition of recovered fibre samples  Arabinose (%) Galactose (%) Glucose (%) Xylose (%) Mannose (%) Lignin (%) Recovered Fibre 0.7 0.6 72.2 8.9 5.4 13.0 52   The monomermic sugar composition of the recovered fibre consists predominantly of glucose, which primarily emanates from the straight chain polymer cellulose fraction. The hemicellulose fraction is predominantly xylose, with lesser amounts of mannose, arabinose, and galactose. The percentage of lignin in the recovered fibre is relatively low compared to values found in native softwoods and hardwoods, which are 20 and 28% respectively, and is more comparable to the 15% lignin content found in corncob, the central core of maize (178).  5.1.2 Hydrolyzability of short, long, and mixed recovered fibre  Recovered fibre represents a plentiful source of easily recyclable material for use in papermaking. The recovery of paper fibre diverts large volumes of waste from landfills, and as of 2006, Canada is recycling 46% of all paper and cardboard material (179). However, recycling fibre is not sufficient to displace virgin wood because the supply does not meet the demand, some fibre is too contaminated, and once fibre has been recycled five to seven times it becomes too short and weak and is subsequently washed out during the recycling process (180). Recycling of paper/boxboard requires fractionation of the individual fibres prior to introduction to the papermaking process. The process steps in converting waste fibre into a product vary by the desired grade and the type of contaminants present, for example fibre that has been printed will require a ‘de-inking’ process step (181). The recovered fibre contains two components that are distinctly different in their composition. The first component consists of long softwood fibres, and the second component consists of short hardwood fibres. The shorter hardwood fibres contain a higher fraction of lignin and hemicellulose sugars compared to the long softwood fibres. Although the ratio of the short and long fibres was not known for the samples studied, the ratio of long to short recovered fibres is generally accepted to be around 2:1 (182). The method that is most commonly used in industry to fractionate short and long fibres is the application of pressure screens (183). Although pressure screening was traditionally used for the removal of oversized contaminants, it can also be applied for fibre fractionation. In Figure 5.1, a cross-section of a rotor foil and screen cylinder 53  is shown, outlining the local flow patterns in a pressure screen. The operation of the pressure screen can be optimized to force the shorter fibre fractions into the outer chamber, leaving the longer softwood fibres in the inner chamber (184).   Figure 5.1 Schematic image of a rotor foil passing a section of a screen cylinder  Short and long fibre fractions were obtained from the WTC in Federal Way, WA. The hydrolyzability of each fraction was determined by enzymatic hydrolysis at enzyme loadings of 2, 5, and 10 FPU/g, the results are shown in Table 5.2.  Table 5.2 Hydrolyzability of untreated short and long recovered fibre (20 g/L initial fibre load)  Total Hexose Yield (%) Substrate 2 FPU/g 5 FPU/g 10 FPU/g Recovered short fibre 10.1 +/- 0.2 14.4 +/- 1.2 19.8 +/- 0.5 Recovered long fibre 9.8 +/- 0.1 17.3 +/- 0.6 22.1 +/- 0.6  At the selected low enzyme loadings, there were only minor differences in the conversions of hexose sugars, which are defined as per Equation 5. The percent hexose yield will be different for each pretreated substrate, since the starting sugar distribution will be influenced by the degree of pretreatment applied. However, since a full compositional analysis of each pretreated substrate was not performed, Equation 5 was used as a reasonable basis for comparison. 54                       [5]  At low enzyme loadings the quantity of enzyme is insufficient to completely titrate all of the lignin binding sites and thus only a portion of the readily hydrolysable amorphous cellulose appears to be converted into hexose sugars. Increasing the enzyme loading from 2 to 5, and then further to 10 FPU/g, did not produce a linear increase in hexose sugar yield. The fact that the doubling of enzyme load did not double the sugar yield is due to substrate factors that limit hydrolysis, like accessible surface area, pore volume, and the number of available active sites, as well as enzyme factors like steric hindrance and poor enzyme desorption (76) (102). A possible explanation for the slightly improved digestibility of recovered long fibre at 5 and 10 FPU/g is that the long softwood fibres were able to swell more readily than the short hardwood fibres. Two main factors could be considered when discussing the degree of swelling of the softwood and hardwood fibres, which are the relative proportions of the two major lignin groups: guaiacyl lignins and guaiacyl-syringyl lignins and the quantity of lignin (185). Hardwood fibre contains both syringyl and guaiacyl units, while softwood consists mainly of guaiacyl units. It has been suggested that guaiacyl lignin reduces the enzymatic accessibility more than syringyl lignin, which suggests that the hardwood fibres should have been hydrolyzed more readily. However, the shorter hardwood fibres have been found to contain a higher fraction of lignin (14.2% for recovered short fibres and 10.9% for recovered long fibres), which appears to be present in a large enough quantity to negate the advantage of containing the more favorable syringyl lignin groups (186). At the same enzyme loadings of 2, 5, and 10 FPU/g, the initial rates of hydrolysis were determined, and are shown in Table 5.3. Table 5.3 Initial hydrolysis rates of short and long recovered fibre (20 g/L initial fibre load)  Initial Rate (g/L/h) Substrate 2 FPU/g 5 FPU/g 10 FPU/g Recovered short fibre 1.125 1.930 2.450 Recovered long fibre 1.210 1.940 3.215 100 substrateg11.1 substrateg )substrateg( )sugarg((%)yieldsugarHexose ××= 55   The recovered long fibre may have a larger proportion of more readily hydrolysable amorphous cellulose in addition to the lower lignin content. This conclusion is based on the fact that the initial hydrolysis rates at the 2 and 5 FPU/g loadings are nearly identical, but when an enzyme loading of 10 FPU/g is applied, a higher initial rate is found. This result suggests that at enzyme loadings of 2 and 5 FPU/g, the amorphous cellulose is still in excess for both fibres, with a distinction being made once a 10 FPU/g enzyme loading was applied. To determine the highest possible sugar yields of the short and long recovered fibres, a set of maximum digestibility experiments were performed at an enzyme loading of 86 FPU/g. The results, outlined in Table 5.4, show a larger final hexose yield for the recovered long fibre. An average of two duplicate experiments produced a final 48 hour hexose yield for short and long recovered fibre of 42.9 +/- 5.1%, and 50.5 +/- 2.0% respectively. The long softwood fibre was determined to have both a higher initial rate of hydrolysis and maximum digestibility, suggesting that recovered fibre with a reduced fraction of short fibre is desirable for increased improved enzymatic hydrolysis of untreated recovered fibre.  Table 5.4 Maximum digestibility of short and long recovered fibre (86 FPU/g, 48h) Substrate Total Sugars (g/L) Total Hexose (g/L) Final Hexose Sugar Yield  (%) Short Fibre 11.48 +/- 1.51 9.53 +/- 1.13 42.87 +/- 5.09 Long Fibre 13.53 +/- 0.99 11.21 +/- 0.45 50.46 +/- 2.02  A blend of recovered fibre, containing both short and long fibres, was analyzed for digestibility at varying enzyme loadings. The hydrolyzability of the recovered fibre was determined at enzyme loadings of 5, 10, 20, and 86 FPU/g. The results are outlined in Table 5.5. At an enzyme loading of 5 FPU/g the hexose sugar yield of the recovered fibre was 14.3%, which is close to the 14.4% hexose sugar yield of the recovered short fibre. However, at an enzyme loading of 10 FPU/g the hexose sugar yield of the recovered fibre was 22.6%, which is close to the 22.1% hexose sugar yield of the recovered long fibre. These results suggest that there may variation in 56  the fibre lengths within the sample obtained from WTC, producing results for the mixed recovered fibre that are slightly variable. Table 5.5: Digestibility of recovered fibre at 5, 10, and 20 FPU/g  Hexose Sugar Yield (%) Substrate 5 FPU/g 10 FPU/g 20 FPU/g Recovered fibre 14.3 22.6 35.6  Having established a baseline of digestibility for recovered fibre, other pulp substrates were analyzed to determine their relative recalcitrance.  5.2 Assessment of the hydrolyzability of multiple pulp substrates  To observe the well known recalcitrant effect of residual lignin on enzymatic hydrolysis a series of experiments involving bleached and unbleached pulps were performed. A previous study by Charles et al. (2003) on hemlock pulp, showed an increase in hexose sugar yield from 26 to 47% for pre and post oxygen delignified pulps (159).  The kraft pulping process refers to the treatment of wood chips in a mixture of sodium hydroxide and sodium sulfide that works to break the bonds between cellulose and lignin (22). This pulp, which is called brownstock, can be further washed and delignified through a series of bleaching stages. A typical brownstock pulp has a Kappa number in the range of 20 - 53, and a typical bleached kraft pulp has a Kappa number of 6.6 or less (187).  A sample of bleached and unbleached kraft pulp was received from the WTC. An analysis of the digestibility of the two pulps was performed in order to observe the well documented effect of residual lignin on enzymatic hydrolysis. Each pulp was hydrolyzed at an enzyme loading of 5 and 20 FPU/g for 96 hours, the results are shown in Figure 5.2.  57   Figure 5.2 Hexose sugar yield of bleached and unbleached Douglas-fir kraft pulp at 5 and 20 FPU/g As expected, the high enzyme loading of 20 FPU/g produced a higher hexose sugar yield compared to the lower enzyme loading of 5 FPU/g for both bleached and unbleached kraft pulps. This result is due to the increased number of substrate active sites that can be attacked at higher enzyme loadings. To achieve increased hexose sugar yield it is also necessary to have a sufficient enzyme loading to titrate the lignin, as was already observed with the short and long recovered fibre. The effect of titrating the lignin was significant in the unbleached kraft pulp, with an increase in hexose sugar yield from 31 to 78% that was associated with an increase in enzyme load from 5 to 20 FPU/g. In a study by Mooney and Mansfield, a comparison of the effect of lignin on Douglas-fir refiner mechanical pulp and kraft pulp was made. The results showed that there is a correlation between the enzyme adsorption capacity of a substrate and the degree of fibre swelling and specific surface area, but this does not correlate to substrate digestibility. It was determined that the best indicator of substrate ease of digestibility is the amount of surface area that is present in the form of small pores. By removing lignin from a substrate the population of small pores 58  increases along with the fibre saturation point, a method for quantifying changes in fibre swelling, which results in a significant improvement in hydrolysis (175). 5.2.1 Assessment of the digestibility of multiple brownstock press and bleached pulps  Four softwood pulp substrates were received from the WTC, which are designated as fir - brownstock press and bleached, and pine – brownstock press and bleached. The analysis of the digestibility was performed at a low range of enzyme loadings (2, 5, 10 FPU/g), as well as at a high enzyme loading of 86 FPU/g in order to determine the maximum digestibility. These enzyme loadings of the two bleached and unbleached softwood pulps would allow for direct comparison to the short and long recovered fibre samples previously discussed. The results from the enzymatic hydrolysis of the brownstock press and bleached kraft pulps are shown below in Table 5.6.  Table 5.6: Hexose sugar yield of softwood fir and pine pulps (brownstock press / bleached)  Hexose sugar yield (%) Pulp Substrate 2 FPU/g 5 FPU/g 10 FPU/g 86 FPU/g Fir  - Brownstock Press 11 +/- 0.3 27 +/- 0.7 59 +/- 1.2  87 +/- 0.2  Pine - Brownstock Press 15 +/- 0.9 44 +/ 1.5 70 +/- 1.8 88 +/- 2.5 Fir - Bleached 33 +/- 0.2 63 +/- 0.3 80 +/- 0.9 100 +/- 2.8 Pine - Bleached 32 +/- 1.2 61 +/- 2.1 94 +/- 1.9 100 +/- 1.3  Limited information was known about the two softwood pulps prior to running the hydrolysis screening experiment. However, the cell wall thickness of a Douglas-fir fibre has been found to be nearly double the thickness of a pine fibres cell wall (188). The hydrolysis results correlate well with the more recalcitrant cell walls present in the Douglas-fir substrates, which produced lower hexose sugar yields compared to the pine substrates. For the brownstock press pulp samples, an increase in hexose sugar yield of 4, 17, and 11% was found for the softwood pine pulp in comparison to the softwood fir pulp. In an excess of enzyme, 86 FPU/g, the maximum digestibility experiment produced equal hexose sugar yield for both softwood pulps. The 59  bleached samples did not show a measurable difference in hexose  sugar yield  below a 10 FPU/g enzyme load, and both achieved a maximum yield at 86 FPU/g. Now that the effect of residual lignin on the ability of enzymes to saccharify lignocellulosic substrates was observed for multiple samples, an analysis of the impact of oxygen delignification pretreatment was performed to measure the improvement in recovered fibre digestibility. The recovered fibre substrate offers potential as a bio-ethanol feedstock because of its relatively low lignin content in addition to a bulk price of less than 80 $CDN/metric-ton at the time of project commencement. However, the price of recovered fibre has been fluctuating recently due to increased demand, and in 2007 the price reached peaks of over 150 $CDN/metric-ton (189). The volatility in the price of recovered fibre outlines the need for an effective and economical pretreatment. 5.3 Oxygen delignification pretreatment  Previous work has shown that oxygen delignification can significantly decrease the residual lignin present in lignocellulosic material (159) (190) (158). Three factors that significantly affect the removal of lignin are caustic load, temperature, and time. 5.3.1 Preliminary oxygen delignification runs  Initial oxygen delignification runs on recovered fiber were performed with a constant temperature (150oC), time (60 min), and oxygen partial pressure (120 psig). The caustic load applied to the charged reactor was varied from 0, 10, and 20% based on the percentage of oven dried fibre. The subsequent hydrolysis results at enzyme loadings of 5, 10, and 20 FPU/g are shown in Figure 5.3.  60   Figure 5.3 Influence of caustic addition on the hydrolyzability of recovered fibre at multiple enzyme loadings (20 g/L initial fibre load) There is a correlation between increasing the caustic concentration and the substrate site activation that occurs during oxygen delignification (166). In this experiment, increasing the caustic concentration from 0-10% resulted in the hexose sugar yield increasing from 36 to 80%, at an enzyme loading of 20 FPU/g. It was found that increasing the caustic concentration above 10% caused a decrease in sugar yield, an effect likely due to the extensive cellulose degradation that occurs at such a high caustic load. During the delignification process an optimized amount of caustic should be applied in order to minimize cellulose degradation and to make the process cost effective. In industrial scale delignification reactors, typical caustic charges in a two-stage delignification process are 2.5% and 3.5% on oven dried fibre for brownstock pulp (191). Softwood pulp from a conventional cooking process contains around 4.5% lignin, which is lower than the 13% in the untreated recovered fibre (192). However, the 10% and 20% caustic loads that were applied can still be considered “extreme” caustic environments for the recovered fibre, where significant cellulose degradation would occur. From a bioethanol process perspective, excessive base-induced attacks 61  during the oxygen delignification stage can lead to a decrease in the final ethanol yield. However, an advantage of oxygen delignification is the low formation of furfural, which is a microbial inhibitor often produced by other pretreatment methods (193).  5.3.2 Effect of temperature, caustic load, and time on the oxygen delignification yield, kappa number, and hydrolyzability  In order to take a more structured approach to the pretreatment of recovered fibre, an experimental design that looked at temperature, caustic charge, and reaction time as independent variables was performed. The dependent variables that were analyzed were delignification yield, kappa number, and hydrolyzability. The delignification yield refers to oven dried fibre that is present after delignification divided by the initial oven dried fibre. The kappa number, as has been previously mentioned, measures the fraction of lignin present in a fibre sample. The untreated recovered fibre has a kappa number of 87, various combinations of pretreatments will lower this value. The hydrolyzability refers to hexose sugar yield obtained during enzymatic hydrolysis of the pretreated recovered fibre. The oxygen delignification and enzymatic hydrolysis conditions that were applied for the experimental design are outlined in Table 5.7, along with the results of the tests. Table 5.7 Pretreated recovered fibre delignification yield, kappa number, and hydrolyzability Temperature Caustic Load Time Delignification Kappa Hexose Sugar Yield (%)   (oC) (% g dry fibre) (min) Yield (%) Number 2% Solids, 2% Solids,  10% Solids,             20 FPU/g 40 FPU/g 16 FPU/g 1 120 2 40 87 59.3 73.1 73.8 59.5 2 150 10 20 73 38 78.1 85.2 62.2 3 90 2 60 91 69.3 71.4 73.5 56.1 4 120 6 60 81 42 81.0 84.2 62.1 5 120 6 20 83 54 78.3 80.8 59.2 6 150 2 20 81 56 72.0 75.2 56.3 7 150 2 60 76 44.7 75.0 79.9 57.7 8 120 10 40 78 38 80.6 83.9 61.6 9 90 10 20 90 64.7 76.4 80.3 56.7 10 90 6 40 85 58.7 76.2 79.6 57.8 62   Hexose Sugar Yield (%)   Temperature (oC)  Caustic Load (% g dry fibre)  Time (min)  Delignification Yield (%)  Kappa Number  2% Solids, 20 FPU/g 2% Solids, 40 FPU/g 10% Solids, 16 FPU/g 11 90 10 20 89 65.3 75.9 80.9 54.6 12 90 10 60 84 48.7 78.2 83.4 58.2 13 120 2 20 88 67.3 71.4 78.6 52.9 14 150 10 60 72 20 89.7 92.0 64.4 15 150 6 40 76 35.3 83.2 86.6 61.8 16 90 2 20 93 76.7 61.9 66.9 49.9  5.3.3 Development of empirical models for oxygen delignification yield, kappa number, and hydrolyzability  Linear regression analysis was performed in Excel in order to develop empirical equations that show the effect of temperature, caustic load, and time on kappa number, oxygen delignification yield, and hydrolyzability. The parameters that were viewed as statistically viable for the empirical equations had a p-value of 0.1 or less, which means that the probability of obtaining a value more extreme than the experimental result is less than 10%. The resulting model equations are shown below in Equations 6 to 10, and had R2 values of 0.94, 0.99, 0.90, 0.90, and 0.93, respectively. Yield (%) = 116.47 - 0.22(T) – 0.70(C) – 0.088(t)                                                            [6] Kappa = 128.56 – 0.35(T) – 2.64(C) – 0.74(t) – 0.011(C)(T) + 0.21(C)2 + 0.0069(t)2 –     0.024(t)(C)                   [7] Hydrolyzability (2%, 20 FPU/g) = 45.31 – 0.24(C2) + 4.16(C) + 0.11(T) + 0.12(t)                    [8] Hydrolyzability (2%, 40 FPU/g) = 52.10 – 0.16(C2) + 3.22(C) + 0.11(T) + 0.092(t)                  [9] Hydrolyzability (10%, 16 FPU/g) = 36.95 – 0.16(C2) + 2.64(C) + 0.084(T) + 0.081(t)            [10] Where T: temperature (oC) 63  C: caustic load (% of dry fibre) T: time (minutes)  By utilizing Matlab software, 3-dimensional planar representations of equations 6 to 10 were generated and are shown in Figures 5.4 to 5.8. The only equation that was a first order polynomial with respect to all of its variables was Equation 6, which was first order with respect to temperature, caustic load and time. Kappa number was found to have a second-order relationship with caustic load and time, shown in Equation 7. All three hydrolyzability relations, Equations 8 to 10, produced first order polynomial relationships with temperature and time, and second order polynomial relationships with caustic load.  The effect of temperature, caustic load, and time on Kappa number is shown in Figure 5.4. The effect of caustic load appears to have an increased effect at higher temperatures, with the Kappa number ranging from 77.6 to 64.9 at 90oC and 20 minutes reaction time, versus 55.3 to 37.3 at 150oC. This result agrees with previous research that shows the extent of delignification increasing as the temperature is raised (194). The lower Kappa number at higher caustic loads can be explained by a greater portion of the lignin structure being activated by alkali, leading to an increased rate of delignification reactions (22). Time was shown to have a non-linear effect on Kappa number, with most of the delignification occurring in the first twenty minutes, and only a small decrease occurring between 40 and 60 minutes.             64     Figure 5.4 Kappa number as a function of temperature, caustic load, and time  The effect of temperature, caustic load, and time on oxygen delignification yield is shown in Figure 5.5. Increasing caustic load from 2 to 10% decreased yield from 93.5% to 87.9% at 90oC and 20 minutes reaction time, and 80.3% to 74.7% at 150oC. The effect of time was found to be most significant in the first twenty minutes, which correlates with the drop in Kappa number that was observed. From 20 to 60 minutes and 2% caustic load there was a decrease from 93.5% to 90.0% at 90oC, versus 80.3% to 76.8% at 150oC. A decrease in the oxygen delignification yield includes a loss of lignin, which is the primary goal, as well as the cellulose degradation of cellulose. In order to observe the impact of delignification conditions on cellulose degradation a comparison of sugar loss versus Kappa number is shown in Figure 5.6. At Kappa values above 60 the sugar loss ranged from 10% or less, with further decreases in Kappa number leading to further increases in sugar loss to a maximum of just over 20%.  65   Figure 5.5 Oxygen delignification yield as a function of temperature, caustic load, and time (2% Solids)   Figure 5.6 Sugar loss versus Kappa number for pretreated recovered fibre  The effect of temperature, caustic load, and time on hydrolyzability is shown in Figures 5.7 to 5.9. At the enzyme loading of 20 FPU/g, increasing the pretreatment caustic load from 2 to 10%, while keeping the temperature at 90oC and reaction time at 20 minutes, increased the hydrolyzability from 65.0% to 75.2%, versus 71.6% to 81.8% at 150oC during pretreatment. A 66  similar relationship was found for the enzyme loading of 40 FPU/g. The effect of time was found to be smaller at an enzyme loading of 40 FPU/g versus 20 FPU/g. This can be attributed to the larger volume of enzyme being able to overcome the unproductive binding with the lignin remaining after pretreatment, and thus being available for cellulose conversion.   Figure 5.7 Hydrolyzability of pretreated recovered fibre as a function of temperature, caustic load, and time (Enzyme loading: 20 FPU/g, 2% Solids) 67   Figure 5.8 Hydrolyzability of pretreated recovered fibre as a function of temperature, caustic load, and time (Enzyme loading: 40 FPU/g, 2% Solids) The effect of increasing the solids loading from 2% to 10%, and decreasing the enzyme loading to 16 FPU/g, are shown in Figure 5.9. Increasing the pretreatment caustic load from 2 to 10%, while keeping the temperature at 90oC and reaction time at 20 minutes, increased the hydrolyzability by 5.7%, versus 5.8% at 150oC during pretreatment. These improvements in hydrolyzability for the high solids loading as the caustic load is increased are lower compared to the 2% loading at 20 FPU/g. An increase in the effect of pretreatment time on hydrolyzability was observed for the higher solids loading. These results can be explained by the fact that higher solids concentrations reduce enzyme mobility, which occurs by limiting enzyme access to readily hydrolysable cellulose through the binding to recalcitrant lignaceous residues. Therefore, the impact of extending the oxygen delignification time from 20 to 60 minutes results in improved hydrolyzability compared to a lower solids loading (2%). These results suggest that under equal pretreatment conditions, an increased enzyme loading is required when the solids concentration is increased from 2% to 10% in order to achieve an equal sugar yield. 68   Figure 5.9 Hydrolyzability of pretreated recovered fibre as a function of temperature, caustic load, and time (Enzyme loading: 16 FPU/g, 10% Solids)  5.3.4 Empirical model for the enzymatic hydrolysis reaction  Previous work on the development of models to depict the enzymatic hydrolysis of cellulose and lignocellulose has been distinguished as non-mechanistic (empirical) (112), semi-mechanistic (114), and functionally based models (118). Empirical models cannot be used to explain a system, however, they are useful for predicting behavior where experimental data does not exist. The objective of developing an empirical model for cellulosic recovered fibre was to define an equation that could be used to determine sugar concentration as a function of kappa number, enzyme loading (FPU/g), and starting recovered fibre concentration. An n-order model, shown in Equation 11, was used to empirically account for all of the simultaneous reactions taking place during cellulose degradation. By integrating Equation 11 from the starting concentration of cellulose at time zero, to the cellulose concentration (g/L) at time t, and then subtracting this value from the starting cellulose 69  concentration, Equation 12 is formed. Since the sugar concentration is what was measured during the hydrolysis experiments, a conversion factor of 1.11 g sugar/g cellulose was used to account for the addition of water during the hydrolysis reaction, and is shown in Equation 12. n Co Ck dt dC    −=          [11]         ( ) celluloseg sugarg Conkt CoS n 11.1 11 11 1 1 ×            −+ −= −                 [12] Where:  C = cellulose concentration (g/L)       Co = starting cellulose concentration (g/L)       k, n = empirical constants S = sugar concentration (g/L)  Optimal values for the empirical constants n and k were determined for a set of experimental hydrolysis data, shown in Figures 5.10 to 5.13, in order to create an optimal fit between the experimental data and Equation 12. The set of empirical constants that were determined for each hydrolysis run are shown in Table 5.8, and when they are utilized in Equation 12 the experimental data agrees well with the theoretical curves. Table 5.8 Empirical constants for enzymatic hydrolysis runs (Figures 5.10 to 5.13) Solids concentration (g/L) Kappa Number Enzyme Loading (FPU/g) n k (g/L/min) 20 87 10 40 18.4 11.4 10.7 22.2 20 50 10 40 4.8 4.5 6.8 39.7 20 40 10 40 5.3 3.7 13.8 16.6 50 70 100 120 40 10 3.8 3.4 2.8 2.5 29.9 26.2 25.5 23.2 70  The empirical constants were then related to various dependent experimental variables such as kappa number, enzyme loading, and initial solids loading in order to determine their respective relationships. These relationships, shown in Equations 13 and 14, when substituted into Equation 12, result in an empirical formula that could be used to predict sugar concentration during the course of hydrolysis. The association of the experimental and empirical model data points is shown in Figures 5.14, 5.15, 5.16, and 5.17.  1.09.0 CoEk ⋅⋅=  [13] −− ⋅⋅⋅= CoEKn                                            [14]     K = kappa number E = enzyme concentration (FPU/g) An increase of the Kappa numbers resulted in an increased dependence between the starting substrate concentration and the rate of reaction. An increase in the reaction order (n) means that the hydrolysis reaction levels off sooner, so an increase in the Kappa number causes a faster reduction in enzyme activity, which can be explained by an increased loss of enzyme due to unproductive lignin binding. Increasing the enzyme load causes the reverse effect, a decrease in the dependence between starting substrate concentration and the rate of reaction. As enzyme load increases, there is a decreased fraction of enzymes lost to irreversible binding, which results in longer lasting hydrolysis reactions characterized by a smaller reaction order. As the starting substrate concentration is increased the reaction order was found to decrease. This can be explained by the higher substrate concentrations resulting in hydrolysis reactions that continue for an extended period of time. The relationship between kappa number and the reaction rate was found to be constant. A possible explanation for this is that lignin is mainly responsible for a reduction in active enzyme levels, but does not greatly affect the rate of hydrolysis. The impact of the substrate lignin content is accounted for in the reaction order for this semi-empirical model. Increasing the enzyme load increased the reaction rate, since the addition of more enzymes results in an increase in the number of active enzymes that are not bound unproductively. The breakdown of this semi-empirical model at substrate concentrations over 70 g/L can be explained by the mass transfer limitations that occur when there is a lack of free water in the hydrolysis mixture. This explains the power relationship of the starting cellulose concentration being less than one, since 71  the reaction rate will initially increase quickly with a higher starting concentration and then begin to level off.  Figure 5.10 Evaluation of empirical hydrolysis model fit on untreated recovered fibre (Kappa 87) at 10 FPU/g (R2 = 0.98) and 40 FPU/g (R2 = 0.97), 20 g/L solids loading  Figure 5.11 Evaluation of empirical hydrolysis model fit on untreated recovered fibre (Kappa 50) at 10 FPU/g (R2 = 0.99) and 40 FPU/g (R2 = 0.96), 20 g/L solids loading 72   Figure 5.12 Evaluation of empirical hydrolysis model fit on untreated recovered fibre (Kappa 40) at 10 FPU/g (R2 = 0.95) and 40 FPU/g (R2 = 0.96), 20 g/L solids loading  Figure 5.13 Evaluation of empirical hydrolysis model fit on pretreated recovered fibre (Kappa 40) at 50 g/L (R2 = 0.94), 70 g/L (R2 = 0.99), and 100 g/L (R2 = 0.94), 20 FPU/g enzyme loading 73   Figure 5.14 Evaluation of empirical hydrolysis model fit on untreated recovered fibre (Kappa 87) at 10 FPU/g (R2 = 0.86) and 40 FPU/g (R2 = 0.87), 20 g/L solids loading  Figure 5.15 Evaluation of empirical hydrolysis model fit on pretreated recovered fibre (Kappa 50) at 10 FPU/g (R2 = 0.88) and 40 FPU/g (R2 = 0.88), 20 g/L solids loading 74   Figure 5.16 Evaluation of empirical hydrolysis model fit on pretreated recovered fibre (Kappa 40) at 10 FPU/g (R2 = 0.89) and 40 FPU/g (R2 = 0.83), 20 g/L solids loading  Figure 5.17 Evaluation of empirical hydrolysis model fit on pretreated recovered fibre (Kappa 40) at 50 g/L (R2 = 0.94), 70 g/L (R2 = 0.96), and 100 g/L (R2 = 0.41), 20 FPU/g enzyme loading  75   5.4 Practical issues  In order to gain further understanding into some of the practical issues that have an impact on the bioconversion process, experiments looking at the hydrolysis feeding regime, lignin precipitation, and effect of pulp washing were performed.  The impact of adding fresh substrate to a hydrolysis mixture in order to utilize unbound enzymes has proven to be effective in achieving high rates of conversion while minimizing the costs associated with enzymes (79) (195).  5.4.1 Hydrolysis feeding regimes  In the development of a process for converting lignocellulosic substrates into ethanol, economics favor high-solids enzymatic hydrolysis due to the reduction in operating and capital costs for ethanol production (196). The primary objective of the hydrolysis step is to achieve high conversion of substrate into monomeric sugars while minimizing the residence time. Two experimental approaches were taken to investigate the digestibility of high-solids recovered fibre. The first approach was a batch hydrolysis where all substrate and enzyme was present at the start of the experiment, and the second approach was a fed-batch hydrolysis where small substrate and enzyme additions were made once the hydrolysis mixture was liquefied. In order to utilize high solids loading it is necessary to consider any heat and mass transfer limitations that occur during hydrolysis. The comparison between feeding regimes was performed at the shake- flask scale in an incubator/shaker that provided a constant temperature environment, and thus the impact of heat transfer was regarded as negligible. The mass transfer limitations, or mobility of the enzymes, was considered by comparing the enzymatic rates of conversion in a batch and fed- batch feeding regime. In addition to heat and mass transfer limitations, there are two important factors to consider that could have an impact on the final conversion of batch and fed-batch hydrolysis feeding regimes. The first factor is whether the substrate reactivity changes with time. It is known that cellulose consists of both amorphous and crystalline regions, and a current suggestion is that the 76  amorphous region is responsible for the high initial rate of hydrolysis followed by the slower conversion rate of the crystalline region (197). Previous studies on Avicel, a microcrystalline cellulose, and hardwood lignocellulose, measured the substrate reactivity as a function of conversion and found very little loss in substrate reactivity over time. By washing the substrate and applying fresh enzyme it was also found that the hydrolysis re-start rate was higher than for uninterrupted hydrolysis (198) (199). Therefore, the drop-off in reaction rate for uninterrupted enzymatic hydrolysis cannot be attributed to a reduction in the substrate reactivity, but could be related to other effects such as enzymes slowing down or becoming “tethered” to the substrate (200).  An additional factor that could have an impact on the final sugar conversion of batch and fed- batch hydrolysis feeding regimes is how the enzyme activity changes with varying substrates and time. When enzymes are added to the substrate mixture their activity is known, however this activity is based on the enzymes ability to saccharify a very pure form of cellulose instead of a complex lignocellulosic structure. A study by van Wyk (1997) illustrated this difference by hydrolyzing filter paper, microcrystalline cellulose, and newsprint. Over a five hour period the percentage increase in cellulase adsorption for the substrates was 5% for filter paper, 15% for microcrystalline cellulose and 23% for newsprint (201). The results clearly show that diverse cellulose substrates display varying adsorption tendencies which reflect their different structural compositions. In order to minimize the costs associated with the enzymes applied in hydrolysis, various recycling strategies have been proposed to reuse enzymes in multiple hydrolysis runs. The key to these recycling strategies is ensuring that the enzyme retains its activity over time. To maximize the recovery of enzymes, a process to recover the enzymes free in solution, as well the enzymes adsorbed to the recalcitrant hydrolysis residue is required. However, enzyme recovery at the industrial scale faces the difficulty of the chemical costs associated with enzyme desorption, and the diluted enzyme streams that result from using large amounts of buffer to wash the substrate (79). A more simple method of reusing enzymes involves taking advantage of the ability of enzymes to partition themselves from residual substrate to freshly added substrate (that could include 77  residual hydrolysis substrate). In a recycle experiment performed by Ramos et al. (1993) the reaction filtrate and the residual substrate were both recycled to account for activity of bound enzymes. Substrates with a low lignin content (0.7%) like steam-exploded, peroxide treated eucalyptus were able to achieve 95% of the original sugar yields after five rounds of recycling, at which point the lignin content was 17% of the substrate weight. High-lignin content (21-23%) substrates like wheat straw were only able to retain 71% of the original activity after one recycle round, and by the fifth round the lignin content was 70% of the substrate weight (79). The lignin content of untreated recovered fibre (13%) is roughly in the middle of the treated eucalyptus and wheat straw.  5.4.2 Fed-batch hydrolysis of micro-crystalline cellulose and pretreated recovered fibre  Microcrystalline cellulose and pretreated recovered fibre were hydrolyzed under a fed-batch feeding regime in order to observe differences in their final hexose sugar yields. The recovered fibre was pretreated with oxygen delignification at a temperature of 150oC, 10% caustic charge based on dry fibre weight, for 60 minutes. The fed-batch hydrolysis experiments were performed at enzyme loadings of 20 and 80 FPU/g. The microcrystalline cellulose is a highly crystalline substrate, whereas the pretreated recovered fibre contains both amorphous and crystalline regions. The difference in cellulose regions is expected to have some impact on enzyme binding, however the presence of lignin in the pretreated recovered fibre is expected to have a larger impact on sugar yield (175). The flasks with an enzyme loading of 20 FPU/g had the consistency of microcrystalline cellulose and pretreated recovered fibre increased from 20 to 70 g/L. The hexose sugar yields after 72 hours of hydrolysis of microcrystalline cellulose and pretreated recovered fibre were 44.6% and 42.1%, respectively. Substrate was added during the hydrolysis in 10 g/L increments when the hydrolysis mixture was liquefied to the point where there were no apparent solids. The hexose sugar yield results and the substrate feeding times can be seen in Figure 5.18. 78   Figure 5.18 Hexose sugar yield for pre-treated recovered fibre and micro-crystalline cellulose in a fed-batch feeding regime. Substrate additions: 20 g/L (1, 4, 7.3h), 10 g/L (49h) Enzyme loading: 20 FPU/g  If the hexose sugar yields found at 20 FPU/g are considered to be low for the two substrates, then it is reasonable to hypothesize that an increase in the enzyme loading will result in increased sugar concentrations. Thus, the enzyme loading was increased from 20 to 80 FPU/g, to see if an increased amount of enzyme would result in a higher saturation of substrate active sites and would therefore increase sugar yields. The flask with an enzyme loading of 80 FPU/g had the consistency of microcrystalline cellulose and pretreated recovered fibre increased from 20 to 100 g/L. The hexose sugar yields after 72 hours of microcrystalline cellulose and pretreated recovered fibre were 49.8% and 47.8% respectively. The hexose sugar yield results and the substrate feeding times are shown in Figure 5.19. By increasing the enzyme loading by a factor of four (20 to 80 FPU/g) and increasing the substrate loading from 70 to 100 g/L, there was a minor hexose sugar yield increase of 5.2% and 5.7% for the microcrystalline cellulose and pretreated recovered fibre, respectively. The small increase in hexose sugar yield for both substrates when increasing the enzyme loading by a 79  factor of four suggests that the number of active sites on the substrate were not in excess, but rather the recalcitrance of the crystalline cellulose and lack of enzyme desorption from lignaceous residues were the primary causes for the low yield. Adsorption profiles performed by Sutcliffe et al. (1986) showed that both protein and other cellulase components were adsorbed to residual carbohydrate and lignin fractions of the substrates, which result in reduced activity of all enzyme components. Additionally, it was found the beta-glucosidase component, which does not bind to the cellulose fraction of the substrate, has a higher affinity for lignin than the cellulase components (195).  Figure 5.19 Hexose sugar yield for pre-treated recovered fibre and micro-crystalline cellulose in a fed-batch feeding regime. Substrate additions: 20 g/L (1, 4, 7.3, h), 10 g/L (12.3, 24.5, 49, 72h). Enzyme loading: 80 FPU/g. In order to gain further understanding of the role that lignin plays in high solids loading hydrolysis, a set of substrates with varying lignin contents were hydrolyzed under various feeding regimes. The hydrolysis times were also extended to over four days, since time may have been a contributing factor to the low hexose sugar concentrations achieved with the microcrystalline cellulose and pretreated recovered fibre. 80  5.4.3 A comparison of fed-batch and batch hydrolysis of multiple substrates  The ability of enzymes to desorb from substrates during hydrolysis, as well as the recalcitrance of hydrolysis residues, have previously been discussed as having an unfavorable effect on hydrolysis yield. To further explore the impact of enzyme mobility during hydrolysis an investigation into the effect of adding fresh enzyme during the course of hydrolysis was explored. Additionally, the effect of unproductive binding of enzyme components was explored by utilizing substrates with a range of lignin content. The hydrolysis results for the untreated recovered fibre with a Kappa value of 87 are shown in Figure 5.20.  Figure 5.20 Comparison between batch, substrate fed-batch, and substrate /enzyme fed-batch feeding regimes for untreated recovered fibre, Kappa 87 (Concentration is expressed on initial batch hydrolysis volume) For the untreated recovered fibre the two feeding regimes that provided the highest hexose concentrations were the substrate fed-batch and substrate/enzyme fed-batch regimes, with 81  negligible difference observed between the two. For the substrate fed-batch regimes the recovered fibre concentration started at 20 g/L and was increased by 20 g/L additions to a final substrate concentration of 120 g/L. The cellulase enzyme loading was 111 FPU/g based on a substrate concentration of 20 g/L, and for the enzyme fed-batch regime the cellulase enzyme was added in equal 18.5 FPU/g batch additions. The beta-glucosidase to cellulase component was always kept at a ratio of 5:1. The low yield for the recovered fibre batch feeding regime was due to the high starting solids concentration of 120 g/L. At this solids concentration it took approximately 72 hours before the hydrolysis mixture had liquefied enough for a sample to be drawn. The lack of liquid in the hydrolysis mixture meant that enzyme mobility was severely reduced, and would likely lead to areas where cellobiose end-product inhibition was occurring due to the lack of beta-glucosidase. A high initial hydrolysis rate can be observed for the first eight hours for the substrate fed-batch feeding regime because the entire enzyme loading is present at the beginning of hydrolysis, and thus at the one and four hour sample points the enzyme loading was 111 FPU/g and 55.5 FPU/g respectively. An interesting observation is the lack of improved hexose concentrations when adding fresh enzyme throughout the hydrolysis period. This result is likely due to the increasing content of recalcitrant lignaceous residues as the hydrolysis proceeds. If the lignaceous residues had not been completely saturated with bound enzyme, which is likely based on the final cellulase enzyme concentration of only 18.5 FPU/g, then the additional enzyme supplemented to the supernatant could be bound to the excess lignin sites. The hydrolysis results, expressed as a concentration on initial volume, for the pretreated recovered fibre with a Kappa value of 17 are shown in Figure 5.21. By reducing the lignin content from a Kappa value of 87 to 17, the hexose sugar concentration after 104 h for the two fed-batch feeding regimes increased from 53 to 81 g/L, a percentage increase of 52%. During the initial stages of hydrolysis (< 24 h), the substrate fed-batch feeding regime resulted in the highest hexose sugar concentrations. By the 24 h sample point the substrate/enzyme fed-batch feeding regime produced hexose sugar concentrations equal to substrate fed-batch regime. This initial difference during the first 24h of hydrolysis could be attributed to the large excess of enzyme in the substrate-fed batch flask saturating the substrate active sites, resulting in a faster initial rate of 82  hydrolysis. The batch hydrolysis of the pretreated recovered fibre increased the hexose sugar concentration to 83 g/L compared to 30.7 g/L for the untreated recovered fibre, a percentage increase of 170%. The batch pretreated recovered sample was liquefied enough to sample after 24 h of hydrolysis, which was 48 h earlier than the batch untreated recovered fibre sample. A possible explanation for the hexose sugar concentration of the batch regime equaling the concentration of the two fed-batch regimes is the decreased lignin content and hydrolysis residue, which results in a reduction of unproductive enzyme binding.   Figure 5.21 Comparison between batch, substrate fed-batch, and substrate /enzyme fed-batch feeding regimes for pretreated recovered fibre, Kappa 17 (Concentration is expressed on initial batch hydrolysis volume) The hydrolysis results for the bleached fibre with a Kappa value of approximately 6 are shown in Figure 5.22. The bleached fibre was chosen to observe the effect of the experimental feeding regimes when the substrate had very low lignin content. In comparison to the 17 Kappa fibre, the bleached fibre batch and substrate/enzyme fed-batch produced hexose sugar concentrations that 83  were 21.7% and 37.5% higher respectively. The substrate fed-batch regime for the bleached fibre produced a hexose sugar concentration of 83.7%, which is a minor 2.8% increase compared to the 17 Kappa fibre. The hydrolysis of bleached fibre was the first run that showed any advantage to making enzyme additions during the course of hydrolysis, with the substrate/enzyme fed-batch regime reaching a hexose sugar concentration of 111 g/L after 104 h of hydrolysis. This improvement in hexose concentration could be understood by the bleached fibre having only minor quantities of hydrolysis residue which are a source of unproductive enzyme binding. In comparing the two fed-batch feeding regimes, they were close to having equal sugar concentrations at the time of the final substrate addition at 48h. After this sample point it can be observed that the hexose sugar yield of the two regimes started increasing, however the substrate/enzyme fed-batch regime starting increasing at a higher rate. The substrate fed-batch regime increased moderately between 48h and 72h, beyond which the hexose sugar concentration reached a maximum of 80%. The separation between the two fed-batch feeding regimes could be explained by the substrate fed-batch regime having a higher capacity of substrate binding sites being occupied by enzymes. It has previously been suggested that this increased capacity of sites bound by enzymes can lead to enzymes interfering with each other through steric hindrance, due to the large size of the cellulase enzymes (200). The improvement in the batch feeding regime for the bleached fibre can be attributed to the reduced amount of lignin in the substrate. A significant portion of the readily hydrolysable cellulose had been broken down by the 48 h sample point, beyond which a slower saccharification of the more recalcitrant cellulose caused a minor increase in the hexose sugar yield over the duration of the hydrolysis period. 84   Figure 5.22 Comparison between batch, substrate fed-batch, and substrate /enzyme fed-batch feeding regimes for bleached fibre (Concentration is expressed on initial batch hydrolysis volume) The hydrolysis step in an industrial process has certain criteria that it must meet in order to minimize the production cost of bioethanol. These criteria include maximizing the sugar yield at a high substrate loading (>10% w/v) over short residence time (<4 days) (202). At the enzyme loadings studied, it is clear that the untreated recovered fibre did not meet adequate hexose sugar concentrations.  The pretreated recovered fibre reached hexose sugar concentrations of close to 80% for all of the feeding regimes studied within 72 h, which is a reasonable threshold for an industrial process.  In 72 h, at the high solids loading of 12% w/v, the bleached fibre was able to achieve 111 g/L and 102 g/L hexose sugar concentrations for the substrate/enzyme and batch feeding regimes, respectively. Issues of scale up relating to heat and mass transfer limitations would need to be taken into account in an industrial process, however at the lab-scale the performance of the batch feeding regime offers reduced complexity and high hexose sugar yields for lower Kappa substrates. Based on the lower yields found for substrates that contain lignin, it appears that there are two main factors that will affect the efficiency of a fed-batch feeding regime. In order to minimize 85  the unproductive binding of beta-glucosidase and cellulase components, as much lignin should be removed as is economically possible, while also minimizing carbohydrate degradation. The carbohydrate fraction of the substrate should also be hydrolyzed as much as possible in order reduce the amount of adsorbed enzyme to the hydrolysis residue.  5.4.4 Precipitation of lignin  The presence of lignin in all lignocellulosic biomass creates an opportunity for lignin recovery in any bioethanol production plant. Current technology allows for lignin to be combusted to provide heat or power for the bioethanol plant, however, higher valued products like fuel additives or specialty grade chemicals could also increase the viability of the bioethanol technology (203). Little work has been done to examine lignin recovery from oxygen delignification liquor, since, in pulp mills, these liquors are usually sent to recovery. However, the recovery of kraft lignin from black liquor has been studied for use in reducing the load to the recovery boiler and as previously mentioned as a specialty chemical and fuel source (204). In a Kraft mill, the two kinds of precipitation techniques that are the most widely used are CO2 and H2SO4 precipitation (205). Precipitation using solely CO2 is the preferred method for precipitation because the quantity of H2SO4 required upsets the liquor cycle chemical balance with excess sulfur (204). An additional advantage to utilizing CO2 in a bioethanol production plant would utilizing the CO2 generated in the fermentation process stage. The recovery of lignin has been found to be greater than 90% when the pH has been lowered to 2-3 by the addition of CO2 and H2SO4. However, at pH levels this low the lignin is precipitated into a gel-like form that has been found to be difficult to filter and wash (205). Additionally, using H2SO4 to achieve such low pH levels, results in large quantities of H2S gas being produced. Therefore, an optimal pH for the carbonation stage of kraft black liquor that minimizes chemical costs, results in an easily filterable product, and produces lignin yields of 65-75%, has been determined to be approximately 8-9 (204). The oxygen delignification conditions, methods for the reduction of liquor pH, and subsequent lignin concentrations and percent lignin removed are shown in Table 5.9.   86  Table 5.9 Lignin concentrations of delignification liquors at various pH values Delignification Conditions Liquor pH Method for pH Reduction Absorbance  Lignin Concentration Lignin Removed (oC - % caustic - min)    (700 nm) (mg/L) (%) 9.24 - 0.3628 128.1 - 6.07 CO2 sparge 0.3225 111.6 13 4.04 CO2 sparge / H2SO4 0.3168 109.2 15 150-10-60 (Severe) 2.86 CO2 sparge / H2SO4 0.2865 96.7 25 10.26 - 0.3533 124.2 - 5.98 CO2 sparge 0.3351 116.7 6 3.81 CO2 sparge / H2SO4 0.3314 115.2 7 120-6-40 (Moderate) 2.53 CO2 sparge / H2SO4 0.3094 106.2 15  Two important factors that have been determined for lignin precipitation are the concentration of sodium and the pH of the solution. When the concentration of a metal cation, like sodium, exceeds a certain concentration limit, lignin will start to precipitate and the lignin concentration in solution will decrease (206). At a given temperature lignin has been found to be more stable against coagulation as the pH increases, which has been attributed to the ionization of phenol groups in the lignin, which leads to increased steric and electrostatic repulsion between lignin molecules. Studies have shown that only the fraction of lignin with a high molecular weight is precipitated, and have compared lignin precipitation behavior to that of the coagulation of a colloid (207). For this experiment the lignin precipitation was performed at 80oC, since experiments aimed at lignin precipitation from black liquor have shown that an increased temperature has a destabilizing effect on lignin (206). During lignin precipitation there are two phases that occur in the colloidal system: nucleation and crystal growth (206). Nucleation is the process by which nuclei form in solution, and crystal growth is the aggregation of these nuclei into larger structures (208). In Kraft lignin solutions the soluble macromolecular lignin self-associates into colloidal lignin particles, which then grow into fractal lignin networks up to 1 – 2 mm scale (209). The pH of the oxygen delignification liquors was lowered in a primary CO2 sparging stage, and a secondary sulphuric acid addition stage (Figure 5.23). Sparging CO2 through the delignification liquor provided a pH drop from 9.24 to 6.07 for the severe delignification liquor and from 10.26 87  to 2.53 for the moderate delignification liquor. The starting pH of the moderate delignification liquor is lower because a smaller amount of NaOH was consumed during delignification. In the second stage, an addition of sulphuric acid further dropped the pH to 4.04 and 2.86 for the severe delignification and 3.81 and 2.53 for the moderate delignification liquor. After CO2 and sulphuric acid addition the concentration of lignin in the liquor dropped by 31.4 mg lignin/L and 18.1 mg/L for the severe and moderate delignification conditions, respectively. The highest lignin recovery of 25% is low compared to the 65-75% that is achieved during carbonation of kraft black liquor. It has previously been determined that dilute kraft liquors require increased amounts of CO2 for carbonation and also generate low lignin yields (204), therefore, it is possible that by increasing the starting lignin concentration of the delignification liquor, a higher percentage lignin could be precipitated.   Figure 5.23 Lignin precipitation in two stages. 1: CO2 sparging 2: Sulphuric acid addition Both delignification liquors had similar concentrations of lignin before precipitation, however, it was observed that a greater amount of lignin precipitated from the more severe delignification liquor. This could be attributed to a larger fraction of high molecular weight lignin being removed under the more severe delignification conditions, which was then precipitated during the trial. Primary Stage Secondary 88  5.4.5 Effect of post-delignification fibre washing on downstream hydrolysis  For the hydrolysis experiments discussed in previous sections the delignified fibre was washed in order to lower the pH by removing caustic, as well as to remove lignin precipitated on the fibre surface. In an industrial process the washing of the delignified fibre should be adequate so as not to adversely affect enzymatic hydrolysis. The degree to which the fibre is washed is an important process consideration that can offer environmental and cost savings by minimizing water use and maximizing caustic chemical recovery. The purpose of a washing stage in between an oxygen delignification and enzymatic hydrolysis stage is similar in principle to washing stages that occur during pulp bleaching and brownstock washing stages in the pulp and paper process. Since the optimum chemical conditions in sequential bleaching stages varies, the washing of the pulp is intended to modify the pH, metal content of the pulp, and temperature. In a brownstock washing stage the filtrate is alkaline and rich in organic substances of a large molecular size, including the lignin in kraft black liquor that is considered to be of high molecular weight (210). The weight average molecular weight for industrial black liquor has been found to range from 1,700 to 20,000 g/mol (211). The free delignification liquor can be removed easily during washing, but to remove the liquor that is entrained in the fibre requires diffusion or capillary force (212). The entrained liquor refers to an immobilized liquor phase that is in close contact with the fibres. A typical brownstock washer utilizes a counter-current system with a series of washing stages that utilizes the cleanest wash water for the final washing stage, and the filtrate of the final washer is re-used in the previous washer. This was technique keeps wash water consumption to minimum while still achieving adequate washing (210). The effectiveness of the brown stock washing stage at kraft mills is conventionally expressed as saltcake (Na2SO4) loss per mass of pulp, and has traditionally been considered effective if the loss in the washing stage is less than 10 kg Na2SO4/kkg of pulp (213). For the pulp washing experiment, oxygen delignification was carried out on recovered fibre at conditions of 150oC, 10% caustic load, for 60 minutes. The delignification liquor was subsequently collected and used in varying quantities for enzymatic hydrolysis in order to simulate varying washing efficiencies that could occur during post-delignification. The enzymatic hydrolysis was carried out at 20g/L with an enzyme loading of 20 FPU/g. The 89  delignification liquor was added at 0, 20, 60, and 86% of the total hydrolysis volume, which translates into washing efficiencies of 100, 80, 40, and 14%. The hydrolysis flask with 86% delignification liquor did not include any buffer, since the remaining volume was made up of enzymes and water present in the fibre. Prior to enzymatic hydrolysis all flasks had their pH values adjusted to 4.8. The total hexose conversion for the hydrolysis flasks containing a range of delignification liquor is shown in Figure 5.24. The addition of 20% delignification liquor resulted in a drop in hexose sugar yield of 11%, to a value of 77%. The additions of 60% and 86% delignification liquor resulted in hexose sugar yield decreases of 13% and 18% respectively. If a desired hexose sugar yield greater than 80% is desired then the volume of delignification liquor that could be added to a hydrolysis flask is likely less than 10% of the total reaction volume. Based on these results it is likely that a washing stage that does not require any diffusion or capillary force would be required, as removal of the free delignification liquor would likely be adequate to achieve desirable hydrolysis results. The delignification conditions that were performed for this experiment can be considered severe, so this experiment could be repeated at conditions that are intended for process scale-up. Less severe oxygen delignification conditions may result in an increased volume of delignification liquor that could be utilized in the hydrolysis stage, while still achieving high sugar conversion.  90   Figure 5.24 Hexose sugar yield of hydrolysis flasks utilizing a range of oxygen delignification liquor volumes, 48h (Hydrolysis conditions: 20 g/L, 20 FPU/g)            20 60 86 91  6.  CONCLUSIONS  The oxygen delignification pretreatment of recovered fibre, with a Kappa number of 87, produced dependent variables that related to the temperature, reaction time, and caustic loading by the following equations: Yield (%) = 116.47 - 0.22(T) – 0.70(C) – 0.088(t) Kappa = 128.56 – 0.35(T) – 2.64(C) – 0.74(t) – 0.011(C)(T) + 0.21(C)2 + 0.0069(t)2 –     0.024(t)(C) Hydrolyzability (2%, 20 FPU/g) = 45.31 – 0.24(C2) + 4.16(C) + 0.11(T) + 0.12(t) Where T: temperature (oC) C: caustic load (% of dry fibre) t: time (minutes)  An increase in any of the independent variables resulted in a decrease in the delignification yield and kappa number, and an increase in the hydrolyzability of the pretreated recovered fibre. An empirical model was also constructed that could predict sugar concentrations from enzymatic hydrolysis based on the kappa number, enzyme loading, and starting recovered fibre concentration:        ( )             −+ −= −1 1 11 11 n Conkt CoS Where: Co = starting cellulose concentration (g/L)      S = sugar concentration (g/L)  2.0CoE5.0k ⋅⋅=       63.0Co08.0K05.0 Eee6.4n −×−⋅ ⋅⋅⋅=     K = kappa number E = enzyme concentration (FPU/g)   92  An analysis was made of three separate hydrolysis feeding regimes, including a batch regime where all enzyme and substrate were added at time zero, a substrate fed-batch regime where substrate was added in intervals, and a substrate/enzyme regime where both substrate and enzyme were added in intervals. For untreated recovered fibre (87 Kappa), the substrate fed- batch, and the substrate/enzyme fed-batch achieved 71% hexose concentrations than the batch feeding regime. By decreasing the Kappa number to 17 the batch feeding regime produced a hexose concentration 4% higher than the two fed-batch regimes. Finally, for a bleached fibre with a Kappa number of approximately 6 the batch feeding regime produced a hexose concentration that was 9% and 32% higher than the substrate fed-batch and substrate/enzyme fed-batch regimes respectively. The results show that the lab-scale performance of the batch feeding regime offers reduced complexity and high hexose yields for low Kappa substrate. By reducing the pH of oxygen delignification liquor it was found to be possible to precipitate lignin from the liquor. The pH was lowered by a two-step reduction of bubbling CO2 and sulphuric acid. By reducing the pH of the oxygen delignification liquor from approximately 9.5 to 3 it was possible to precipitate up to 25% of the lignin present in solution. Hydrolysis experiments were performed utilizing varying volumes of delignification liquor in order to understand the degree of washing necessary post-delignification. The effect of increasing the percentage of liquor from 0% to 20, 60, and 86% of the total hydrolysis volume reduced the hexose sugar yield from 89% to 77, 75, and 70% respectively.  If a hexose sugar yield of 85% or higher is desired, then a wash efficiency of approximately 90% is required.        93  7. FUTURE WORK  The research that has been presented in this thesis has shown that oxygen delignification is a promising technology for the pretreatment of lignocellulosic substrates. The ability of oxygen delignification to improve the enzymatic hydrolysis sugar yield of recovered fibre has been demonstrated. It has been proven that the addition of magnesium salts during oxygen delignification can act to protect the fibre and result in an increased delignification yield. An exploration in the degree to which a magnesium salt could increase the delignification yield over a range of temperatures, caustic charges, and time could offer an improvement to the empirical models developed in this thesis by reducing the negative effects of increasing these process variables. Additionally, since any promising results would need to be replicated at a larger scale in order to be considered for an industrial process, a more detailed look at the heat and mass transfer issues related to delignifying recovered fibre in a larger reactor would be useful. There is no current information on how oxygen delignification affects the porosity and crystallinity of the recovered fibre during pretreatment. A study that looked at a range of delignification conditions, followed by a study of these fibre characteristics, would allow for further optimization of pretreatment conditions in order to maximize hydrolysis results. The research work presented has given a general understanding of the effect of oxygen delignification on the enzymatic hydrolysis of recovered fibre. This understanding could be enhanced by taking a detailed look at how the specific activity of enzymes change during the course of hydrolysis, which would provide further details on the degree to which enzymes bind to the residual lignin of recovered fibre. For the hydrolysis experiments performed, β-glucosidase was always added in excess in order to ensure there was no cellobiose inhibition of the cellulase enzymes. The ratio of β-glucosidase to cellulase necessary to prevent any inhibition could be studied in order to minimize the use and associated cost of utilizing β-glucosidase. In order to enhance the enzymatic conversions of the pretreated recovered fibre it is also possible to utilize surfactants that reduce the unproductive binding of enzymes to lignin. This occurs by the surfactant binding to lignin and thus reducing the number of sites available for enzymes to 94  bind to. A study that looked at a range of surfactants on recovered fibre samples with varying Kappa numbers could result in less severe pretreatment conditions achieving optimal hydrolysis conversions by reducing this unproductive binding.                   95  8. REFERENCES  1. Energy Information Administration: Official Energy Statistics from the US Government website. U.S. Department of Energy. [Online] [Cited: June 14, 2007.] http://tonto.eia.doe.gov/dnav/pet/hist/wtotworldw.htm. 2. Bio-ethanol – The fuel of tomorrow from the residues of today. Hahn-Hagerdal, B., Galbe, M., Gorwa-Grauslund M.F., Liden, G., Zacchi, G. 12, s.l. : TRENDS in Biotechnology, 2006, Vol. 24. 3. Process Engineering Economics of Bioethanol Production. Galbe, M., Sassner, P., Wingren, A., Zacchi, G. 1, s.l. : Applied Biochemistry and Biotechnology, 2007, Vol. 124. 1101-1117. 4. Lignocellulose biotechnology: issues of bioconversion and enzyme production. Howard, R.L., Abotsi, E., van Rensburg, J.E.L., Howard, S. 12, s.l. : African Journal of Biotechnology, 2003, Vol. 2. 602-619. 5. U.S. Department of Energy. Energy Information Administration: Brazil Oil Overview. [Online] June 14, 2007. http://www.eia.doe.gov/emeu/cabs/Brazil/Oil.html. 6. Macedo, I. et al. Assessment of Greenhouse Gas Emissions in the Production and Use of Fuel Ethanol in Brazil. s.l. : Government of the State of Sao Paulo, 2004. 7. Fuel for Friendship. s.l. : The Economist, 2007. 8. Climate Change Policy: Measures to address - Agriculture Sector GHG Emissions. Ministry of Agriculture and Forestry. [Online] 2006. [Cited: June 2, 2008.] http://www.maf.govt.nz/climatechange/slm/bruce-white/page-01.htm. 9. Ethanol as Fuel: Energy, Carbon Dioxide Balances, and Ecological Footprint. Dias de Oliveira, M.E., Vaughan, B.E., Rykiel, E.J. s.l. : BioScience, 2005, Vol. 55. 593-602. 10. Biofuels - facts and fiction. Ecologist Online. [Online] 2007. [Cited: June 2, 2008.] http://www.theecologist.org/archive_detail.asp?content_id=755. 11. Drink the best and drive the rest. Marris, E. s.l. : Nature, 2006, Vol. 444. 670-672. 12. How Cane Sugar is Made: The Basic Story. [Online] [Cited: June 14, 2007.] http://www.sucrose.com/lcane.html. 13. Patsek, T, W. Thermodynamics of the Corn-Ethanol Biofuel Cycle. s.l. : Department of Civil and Environmental Engineering, University of California, Berkeley , 2006. 14. Genencor Introduces New Protease Enzyme for Ethanol Industry. Greencar Congress. [Online] 2006. [Cited: February 12, 2008.] List of Figures. 96  15. Ethanol Production Using Corn, Switchgrass, and Wood; Biodiesel Production Using Soybean and Sunflower. Pimentel, D. and Patzek, T. W. 1, s.l. : Natural Resource Research, 2005, Vol. 14. 65-76. 16. Shapouri, H., Duffield, J.A., Graboski, M.S. Estimating the Net Energy Balance of Corn Ethanol, . s.l. : United States Department of Agriculture, 1995. 17. Agriculture, (2007) United States Department of. USDA Agriculture Projections to 2016 . s.l. : Office of the Chief Economist, Long-term Projections Report: OCE-2007-1, 2007. 18. Shaw, P. Market Trends. Ontario Corn Producers Association. [Online] [Cited: June 14, 2007.] http://www.ontariocorn.org/magazine/Regular%20Features/Market%20Trends/2007/market0307 .htm. 19. State of Food Insecurity in the World 2006. s.l. : Food and Agriculture Organization of the United Nations, 2006. 20. Ethanol Production Using Corn, Switchgrass, and Wood; Biodiesel Production Using Soybean and Sunflower. Pimentel, D., Patzek, T.W. 1, s.l. : Natural Resources Research, 2005, Vol. 14. 21. Cellulosic Ethanol: Benefits and Challenges. Systems Biology for Energy and the Environment. [Online] 2008. [Cited: June 2, 2008.] http://genomicsgtl.energy.gov/biofuels/benefits.shtml. 22. Sjostrom, E. Wood Chemistry Fundamentals and Applications. San Diego, CA : Academic Press, Inc., 1993. 23. Bioenergy Research Center Funding Announcement. [Online] 2006. [Cited: June 14, 2007.] http://www.energy.gov/news/3878.htm. 24. Amon, A., Froggatt, A., Schneider, M. Stimulating A Democratic Debate About The EU’s Research Priorities: A Criteria Based Approach to the 7th EU Research Framework Programme for Energy and Nuclear, Greens-EFA Group in the European Parliament. 2005. 25. BP Bets Big on UC Berkeley for Novel Biofuels Center. Kintisch, E. 5813, s.l. : Science, 2007, Vol. 315. 747. 26. How to Make Biofuels Truly Poplar. Kintisch, E. 5813, s.l. : Science, 2007, Vol. 315. 789. 27. Can biofuels finally take center stage? Schubert, C. s.l. : Nature Biotechnology, 2006, Vol. 24. 777-784. 97  28. Whims, J. Pipeline Considerations for Ethanol. s.l. : Department of Agriculture Economics, Kansas State University, 2002. 29. Nalley, L., Hudson, D. The potential viability of biomass ethanol as a renewable fuel source: A discussion. s.l. : Missisipi State University, Department of Agriculture Economics, 2003. 30. Company, Orbital Engine. Market Barriers to the Uptake of Biofuels Study. s.l. : Report to Environment Australia, 2003. 31. Windholz, M. Merck Index of Chemicals and Drugs, 9th ed. 1976. 32. Biofuels in the U.S. Transportation Sector. Energy Information Administration: Official Energy Statistics from the U.S. Government. [Online] February 2007. [Cited: March 30, 2008.] http://www.eia.doe.gov/oiaf/analysispaper/biomass.html. 33. EFFECT OF ETHANOL ON FUEL PROPERTIES. Environment Canada. [Online] December 2002. [Cited: March 30, 2088.] http://www.ec.gc.ca/cleanair- airpur/CAOL/transport/publications/ethgas/ethgas4.htm. 34. Ethanol Facts: Environment. Renewable Fuels Association. [Online] 2005. [Cited: March 30, 2008.] http://www.ethanolrfa.org/resource/facts/environment/. 35. Ethanol E10, MTBE vs. Non-Alcohol Conventional Gasoline. MLR Solutions. [Online] 2008. [Cited: May 14, 2008.] http://www.fuel-testers.com/ethanol_mtbe_vs_non_alcohol_gas.html. 36. Biomass resource facilities and biomass conversion processing for fuels and chemicals. Demirbas, A. s.l. : Energy Conversion and Management, 2000, Vol. 42. 37. Hot Gas Conditioning: Recent Progress With Larger-Scale Biomass Gasification Systems. Stevens, D.J. s.l. : National Renewable Energy Laboratory , 2001. 38. Chandrakant, T. Biomass Technology - Technology and Utilisation. [Online] 1997. [Cited: March 22, 2008.] http://members.tripod.com/~cturare/bio.htm. 39. Biomass Gasification. Biomass Technology Group. [Online] 2004. [Cited: March 22, 2008.] http://www.btgworld.com/technologies/gasification.html. 40. Virginia Tech, Digital Library and Archives. Network Digital Library of Theses. [Online] http://scholar.lib.vt.edu/theses/available/etd-2998-114756/unrestricted/e-body1.pdf. 41. Semi-micro determination of cellulose in biological materials. Updegraff, D.M. s.l. : Analytical Biochemistry, 1969, Vol. 32. 420-424. 98  42. Structural and biochemical studies of GH family 12 cellulases: Improved thermal stability, and ligand complexes. Sandgren, M., Stahlberg, J., Mitchinson, C. 3, s.l. : Prog. Biophys. Mol. Biol., 2005, Vol. 89. 43. Carbohydrates. Penn State Education. [Online] 2003. [Cited: July 5, 2007.] http://www.bmb.psu.edu/courses/bmb401_spring2003/carbohydrates_2.pdf. 44. Ophardt, C.E. Cellulose. Elnhurst College. [Online] 2003. [Cited: June 2, 2008.] http://www.elmhurst.edu/~chm/vchembook/547cellulose.html. 45. Cellulose. Fiber Source. [Online] [Cited: May 12, 2008.] http://www.fibersource.com/f- tutor/cellulose.htm. 46. Cellulose hydrolysis – the role of monocomponent cellulases in crystalline cellulose degradation. 2, s.l. : Cellulose, Springer Netherlands, 2004, Vol. 10. 47. al., Alberts et. Molecular Biology of the Cell, 4th Edition. s.l. : Garland Sciences, 2002. 48. Voet, D., Voet, J. G. Biochemistry, 2nd edition. s.l. : John Wiley & Sons, Inc., 1995. 49. Colebrook, Michael. Green Spirit. Life Chemistry. [Online] [Cited: July 5, 2007.] www.greenspirit.org.uk/Resources/cellulose.gif. 50. Biomass Ethanol: Technical Progress, Opportunities and Commercial Challenges. Wyman, C.E. s.l. : Annual Review of Energy and the Environment, 1999, Vol. 24. 189-226. 51. Palonen, Hetti. Role of lignin in the enzymatic conversion of lignocellulose. s.l. : VIT Publications, 2004. 52. Look back over the studies of lignin biochemistry. Takayoshi, H. 1, s.l. : Springer Japan, 2006, Vol. 52. 53. Formation and involvement of superoxide and hydroxyl radicals in TCF bleaching process: a review. Gierer, J. s.l. : Holzforsch, 1997, Vol. 23. 34-46. 54. Lignins: Natural polymers from oxidative coupling of 4-hydroxyphenyl- propanoids. Ralph, J., Lundquist, K., Brunow, G. 1-2, s.l. : Phytochemistry Reviews, 2005, Vol. 3. 55. Helm, R. Lignin. Wood Chemistry, Production and Processes. [Online] 2000. [Cited: February 22, 2008.] http://dwb.unl.edu/Teacher/NSF/C06/C06Links/www.chem.vt.edu/chem- dept/helm/3434WOOD/notes1/lignin.html. 56. Changes in Macromolecular characterisitics and biological activity of hydrolytic lignin in the course of composting. Novikoca, L.N., Medvedeva, S.A., Volchatova, I.V., Bogatyreva, S.A. s.l. : Appl. Biochem. Microbial., 2002, Vol. 38. 181-185. 99  57. Ragnar, M. On the Importance of Radical Formation in Ozone Bleaching. s.l. : Hogskoletryckeriet, Stockholm, 2000. 58. Deacon, J.W. Modern Mycology. s.l. : Blackwell Scientific, Oxford, 1997. 59. A structural model of softwood lignin. Sakakibara, A. 2, s.l. : Wood Science and Technology, 1980, Vol. 14. 60. Possible Lignin Structure. Ohio State University. [Online] [Cited: June 15, 2007.] http://hcs.osu.edu/hcs300/gif/LIGNIN.GIF. 61. Gardfeldt, K. Lignin as a raw material for chemicals. Nordisk Innovations Centre. [Online] 2008. [Cited: May 12, 2008.] http://www.nordicinnovation.net/prosjekt.cfm?Id=1-4415-202. 62. Forestry Commission. Extraction Technologies For Tree Metabolites. [Online] [Cited: March 30, 2008.] http://tree-chemicals.csl.gov.uk/review/extraction.cfm. 63. Fengel, D., Gerd, W. Wood Chemistry, Ultrastructure and Reactions. Berlin : Walter de Gruyter, 1989. 64. Mazlan, I. Clean fractionation of biomass - Steam explosion and extraction. s.l. : Wood Science and Forest Products, Virginia Tech., 1998. 65. Novozymes and NREL Reduce Cost of Enzymes for Biomass-to-Ethanol Production 30- Fold. Green Car Congress. [Online] March 29, 2005. http://www.greencarcongress.com/2005/04/novozymes_and_n.html. 66. Biomass FAQ. Novozymes. [Online] [Cited: May 13, 2008.] http://biomass.novozymes.com/faq/. 67. Clark, Giles. Novozymes introduces strategy for economically viable cellulosic ethanol. Biofuel Review. [Online] March 2007. [Cited: March 29, 2008.] http://www.biofuelreview.com/content/view/870/. 68. Ethanol Producer Magazine. Proposed Canadian cellulosic ethanol plant moves forward. [Online] March 26, 2008. [Cited: March 29, 2008.] http://www.ethanolproducer.com/article.jsp?article_id=3901. 69. All Headline News (AHN). Iogen Closer To Canadian Federal Funding for Saskatchewan Biofuel Plant. [Online] March 16, 2008. [Cited: March 29, 2008.] http://www.allheadlinenews.com/articles/7010348912. 70. Saccharophagus degradans gen. nov., sp. nov., A versatile marine bacterial degrader of complex polysaccharides. Ekborg, N.A., J.M. Gonzalez, M.B. Howard, L.E. Taylor, S. Hutcheson, R. Weiner. s.l. : Int J Syst Evolut Microbiol, 2005, Vol. 55. 1545-1549. 100  71. Whitaker, D.R. The Enzymes, 3rd edition. s.l. : Academic Press. , 1971. 273-290. 72. The cellulase of Trichoderma viride. Purification, characterization and comparison of all detectable endoglucanases, exoglucanases and beta-glucosidase. Beldman, G., Searle-van Leeuwen, M.F., Rombouts, F.M., Voragen, F.G.J. s.l. : Eur. J. Biochem., 1985, Vol. 146. 73. Some Characteristics of the Cellulase of Aspergillus Candidus. Ortega, J. 2, s.l. : Biotechnology Letters, 1985, Vol. 7. 109-112. 74. Biodegradation of cellulose. In Biosynthesis and Biodegradation of Wood Components. Eriksson, K.E., Wood, T.M., Higuchi, T. s.l. : Academic Press, 1985. 469-503. 75. Biotechnology of Biomass Conversion. Wayman, M., Parekh, S.R. Milton Keynes, UK : Open University Press, 1990. 76. Substrate and Enzyme Characteristics that Limit Cellulose Hydrolysis. Mansfield, S.D., Mooney, C., Saddler, J.N. s.l. : Biotechnology Progress, 1999, Vol. 15. 804-816. 77. Synergism of cellulases from Trichoderma reesei in the degradation of cellulose. Henrissat, B., Driguez, H., Viet, C., Schulein, M. s.l. : Bio. Tech., 1985, Vol. 3. 722-726. 78. Studies of the cellulolytic system of the filamentous fungus Trichoderma reesei. Substrate specificity and transfer activity of endoglucanase. Claeyssens, M., van Tilbeurgh, H., Kamerling, J.P., Berg, J., Vrsanska, M., Biley, P. I. s.l. : Biochem. J., 1990, Vol. 270. 79. Factors affecting cellulose hydrolysis and the potential of enzyme recycle to enhance the efficiency of an integrated wood to ethanol process. Gregg, J., Saddler, J.N. s.l. : Biotech. and Bioeng., 1996, Vol. 51. 80. Measurement of Cellulase Activities. Ghose, T.K. 2, s.l. : Oure & Appl. Chem., 1987, Vol. 59. 257-268. 81. Measurement of β-Glucosidase/Cellobiase Activity. Wood, W. s.l. : Methods in Enzymology, 1998, Vol. 160. 109-110. 82. On an available pretreatment for the enzymatic saccharification of lignocellulosic materials. Maekawa, E. s.l. : Wood Sci. Technol., 1996, Vol. 30. 133-139. 83. Kinetic studies of enzymatic hydrolysis of insoluble cellulose, Analysis of extended hydrolysis times. Lee, Y.H., Fan, L. T. s.l. : Biotechnol. Bioeng. , 1983, Vol. 25. 939-966. 84. Cellulase production technology: Evaluation of current status. Phillippidis, G. P. Washington, D.C. : American Chemical Society, 1994, Vol. 49. 85. Studies on the mechanism of enzymatic hydrolysis of cellulosic substances. Ghose, T. K. 39, s.l. : Adv. Biochem. Eng., 1977, Vol. 7. 101  86. Mode of action of cellulases. Nisizawa, K. J. s.l. : J. Ferment.Technol., 1973, Vol. 51. 267- 304. 87. Comparison of pretreatment methods onthe basis of available surface area. Thompson, D. N., Chen,H. C., Grethlein, H. C. s.l. : Bioresource Technol. , 1992, Vol. 39. 155-163. 88. Mechanisms of the enzymatic hydrolysis of cellulose: Effects of major structural features of cellulose on enzymatic hydrolysis. Fan, L. T., Lee, Y.H., Beardmore, D. H. s.l. : Biotechnol. Bioeng., 1980, Vol. 22. 177-199. 89. The influence of major structural features of cellulose on rate of enzymatic hydrolysis. Fan, L.T., Lee, Y.H., Beardmore, D.H. s.l. : Biotechnol. Bioeng. , 1981, Vol. 23. 419-424. 90. Enzymatic modifications of secondary fibre. Jackson, L. S., Heitmann, J., Joyce, T. s.l. : Tappi J. , Vol. 76. 147-154. 91. Effect of a process of explosion for the effective utilization of biomass. Sawada, I., Kuwahara, M., Nakamura, Y., Suda, H. s.l. : Int. Chem. Eng, 1987, Vol. 27. 686-693. 92. Effects of fungal pretreatment and steam explosion pretreatment on enzymatic saccharification of plant biomass. Sawada, T., Nakamura, Y., Kobayashi, F., Kuwahara, M., Watanabe, T. s.l. : Biotechnol. Bioeng. , 1995, Vol. 48. 719-724. 93. Hydrolysis and crystallization of cellulose. Battista, O. A. s.l. : Ind. Eng. Chem. , 1950, Vol. 42. 502-507. 94. Structure, pretreatment and hydrolysis of cellulose. Chang, M. M., Chou, T. Y. C. and Tsao, G. T. Berlin : Springer-Verlag, 1981, Vol. 20. 15-42. 95. Structural modification of lignocellulosics by pretreatments to enhance enzymatic hydrolysis. Gharpuray, M. M., Lee, Y.H., Fan, L. T. s.l. : Biotechnol. Bioeng., 1983, Vol. 25. 157-172. 96. Effect of structural and physicochemical features of cellulosic substrates on the efficiency of enzymatic hydrolysis. Sinitsyn, A. P., Gusakov, A. V., Vlasenko, E. Y. s.l. : Appl. Biochem. Biotechnol., 1991, Vol. 30. 43-59. 97. Adsorption of high purity endo-1-4-0-gucanases from Trichodermareesei on components of lignocellulosic materials: Cellulose, lignin and xylan. Chernaglazov, V. M., Ermolova, 0. V., Klyosov, A. A. s.l. : Enzyme Microb. Technol., 1988, Vol. 10. 503-507. 98. Reutilization of enzymes for saccharification of lignocellulosic materials. Deshpande, M. V., Eriksson, K.E. s.l. : Enzyme Microb. Technol., 1984, Vol. 6. 338-340. 99. Bioconversion of forest products industry waste cellulosics to fuel ethanol: a review. Duff, S. J. B., Murray, W.D. s.l. : Bioresource Technology, 1996, Vol. 55. 1-33. 102  100. A comparative study of the enzymatic hydrolysis of acid-pretreated white pine and mixed hardwood. Grethlein, H. E., Allen, D. C., Converse, A. O. s.l. : Biotechnol. Bioeng., 1984, Vol. 26. 1498-1505. 101. Chlorite delignification of spruce wood. Comparison of the molecular weight of the lignin dissolved with the size of the pores in the cell wall. Ahlgren, P. A., Yean, W. Q., Goring, D. A. I. s.l. : Tappi J., 1971, Vol. 54. 737-740. 102. Optimisation of temperature and enzyme concentration in the enzymatic saccharification of steam pretreated willow. Eklund, R., Galbe, M., Zacchi, G. s.l. : Enzyme Microb. Technol. , 1990, Vol. 12. 225-228. 103. Process considerations in the enzymatic hydrolysis of biomass. Ladisch, M. R., Lin, K. W., Voloch, M., Tsao, G. T. s.l. : EnzymeMicrob. Technol., 1983, Vol. 5. 104. Influence of â-glucosidase on the filter paper activity and hydrolysis of lignocellulosic substrates. Breuil, C., Chan, M., Gilbert, M., Saddler, J. N. s.l. : Bioresour. Technol., 1992, Vol. 39. 139-142. 105. Cellulose hydrolysis by bacteria and fungi. Tomme, P., Warren, A. J., Gilkes, N. R. s.l. : Adv. Microb. Physiol. , 1995, Vol. 37. 1-81. 106. Studies of the cellulolytic system of Trichoderma reesei QM 9414, Analysis of domain function in two cellobiohydrolases by limited proteolysis. Tomme, P., Van Tilbeurgh, H., Pettersson, G., Van Damme, J., Vandekerckhove, J., Knowles, J., Teeri, T., Claeyssens, M. s.l. : Eur. J. Biochem., 1988, Vol. 170. 575-581. 107. Properties of a genetically reconstructed Prevotella sunimicola endoglucanase. Maglione, G., Matsushita, O., Russell, J. B., Wilson, D. B. 58, s.l. : Appl. Environ. Microbiol. , Vol. 1992. 3593-3597. 108. Role of the activity and adsorption of cellulases in the efficiency of the enzymatic hydrolysis of amorphous and crystalline cellulose. Klyosov, A. A., Mitkevich, O. V., Sinitsyn, A. P. s.l. : Biochemistry , 1986, Vol. 25. 540-542. 109. Adsorption of cellulase from Trichoderma reesei on cellulose and lignacious residue in wood pretreated by dilute sulfuric acid with explosive decompression. Ooshima, H., Burns, D. S., Converse, A. O. s.l. : Biotechnol. Bioeng. , 1990, Vol. 36. 446-452. 110. Karrer, P., Schubert, P. s.l. : Helv. chim. acta, 1926, Vol. 9. 111. Toward an aggregated understanding of enzymatic hydrolysis of cellulose: noncomplexed cellulase systems. Zhang, Y.P., Lynd, L.R. s.l. : Wiley Perdiodicals, 2004. 103  112. The effect of enzyme concentration on the rate of the hydrolysis rate of cellulose. Sattler, W., Esterbauer, H., Glatter, O. s.l. : Biotechnol. Bioeng., 1989, Vol. 33. 113. Model for cellulose hydrolysis by cellulase. Miyamoto, S., Nisizawa, K. s.l. : Vet Soc. Arm Jpn, 1942, Vol. 396. 114. The HCH-1 model of enzymatic cellulose hydrolysis. Holtzapple, M.T., Caram, H.S., Humphrey, A.E. s.l. : Biotechnol. Bioeng., 1984, Vol. 26. 115. Kinetics studies of enzymatic hydrolysis of insoluble cellulose: derivation of a mechanistic kinetics model. Fan, L.T., Lee, Y.H. s.l. : Biotechnol. Bioeng., 1983, Vol. 25. 116. Modeling of the enzymatic hydrolysis of cellobiose and cellulose by a complex enzyme mixture of Trichoderma reeseri. Scheiding, W., Thoma, M., Ross, A. s.l. : Appl. Microbiol. Biotechnol., 1984, Vol. 20. 117. Degradation of polysaccharides by endo and exo enzymes: a theoretical analysis. Suga, K., van Dedem, G., Moo-Young, M. s.l. : Biotechnol. Bioeng., 1975, Vol. 17. 118. Converse, A.O. Substrate factors limiting enzymatic hydrolysis. [book auth.] J.N., editor Saddler. Bioconversion of forest and agricultural plant residues. 1993. 119. Howard R.L, Abotsi E., Jansen van Rensburg E.L., Howard S. Lignocellulose biotechnology: issues of bioconversion and enzyme production. 2003. 120. Broda, Paul. Biotechnology in the degradation and utilization of lignocellulose. Manchester : University of Manchester Institute of Science and Technology, 1992. 121. Lignocellulose Pretreatment: A comparison of wet and dry ball attrition. Rivers, D.D., Emert, G. H. 5, s.l. : Biotechnology Letters , 1987, Vol. 9. 365-368. 122. Structure, Pretreatment and Hydrolysis of Cellulose. Chang, M.M., T.Y.C. Chou, G.T. Tsao. s.l. : Advances in Biochem. Eng., 1981, Vol. 20. 123. Hydrolysis of lignocellulosic materials for ethanol production: a review. Sun, Y., Chang, J. s.l. : Bioresource Technology, 2002, Vol. 83. 124. Updates on Softwood-to-Ethanol Process Development. Warren E. Mabee, David J. Gregg, Claudio Arato, Alex Berlin, Renata Bura, Neil Gilkes, Olga Mirochnik, Xuejun Pan, E. Kendall Pye, John N. Saddler. 1-3, s.l. : Applied Biochemistry and Biotechnology, 2006, Vol. 129. 55-70. 125. Bioconversion of forest products industry waste cellulosics to fuel ethanol: a review. Duff, S. J. B., Murray, W.D. s.l. : Bioresource Technology, 1996, Vol. 55. 1-33. 104  126. Optimization of steam explosion to enhance hemicellulose recove. Michael M. Wu, Kevin Chang, David J. Gregg, Abdel Boussaid , Rodger P. Beatson, John N. Saddle. 1-3, s.l. : Applied Biochemistry and Biotechnology, 1999, Vol. 77. 127. A review of the production of ethanol from softwood. M. Galbe, G. Zacchi. s.l. : Appl Microbiol Biotechno, 2002, Vol. 59. 618-628. 128. The nature of lignin from steam explosion/enzymatic hydrolysis of softwood. Sergey M. Shevchenko, Rodger P. Beatson, John N. Saddler. 1-3, s.l. : Applied biochem. and bio., 1999, Vol. 79. 867-876. 129. Dilute acid pretreatment of short rotation woody and herbaceous crops. Torget R, Werdene P, Himmel M, Grohmann K. s.l. : Appl Biochem Biotechnol, 1990, Vol. 24. 115- 116. 130. Dilute-acid pretreatment of corn residues and short-rotation woody crops. Torget R, Walter P, Himmel M, Grohmann K. s.l. : Appl Biochem Biotechnol, 1991, Vol. 28. 75-86. 131. Steam explosion of the softwood Pinus radiata with sulphur dioxide addition. Clark TA, Mackie KL. s.l. : Wood Chem Technol, 1987, Vol. 7. 373-403. 132. Steam explosion of the softwood Pinus radiata with sulphur dioxide addition. Process characterisation. Clark TA, Mackie KL, Dare PH, McDonald AG. s.l. : Wood Chem Technol , 1989, Vol. 9. 135-166. 133. Optimisation of steam pretreatment of SO2-impregnated mixed softwoods for ethanol production. Stenberg K, Tengborg C, Galbe M, Zacchi G. s.l. : Chem Technol Biotechno, 1998, Vol. 71. 299-308. 134. Comparison of SO2 and H2SO4 impregnation of softwood prior to steam pre-treatment on ethanol production. Tengborg C, Stenberg K, Galbe M, Zacchi G, Larsson S, Palmqvist E, Hahn-Hägerdal, B Comparison of SO2 and H2SO4 impregnation of softwood prior to steam pre-treatment on ethanol production. 3, s.l. : Appl Biochem Biotechnol , 1998, Vol. 70. 135. Hydrolysis of lignocellulosic materials for ethanol production: a review. Sun, Y., Cheng, J. 1, s.l. : Bioresource Technology, 2002, Vol. 83. 1-11. 136. The ammonia freeze explosion (AFEX) process: a practical lignocellulose pretreatment. M.T. Holtzapple, J-H. Jun, G. Ashok, S.L. Patibandla and B.E. Dale. s.l. : Appl. Biochem. Biotechnol., 1991, Vol. 28. 59-74. 137. Enzymatic Conversion of Biomass for Fuels Production. J.D. McMillan, M.E. Himmel, J.O. Baker and R.P. s.l. : American Chemical Society, 1994. 292-324. 105  138. Holtzapple, M.T., Davison, R.R., Stuart, E.D. Biomass refining process. s.l. : US patents, 1992. 139. Fermentation of lignocellulosic materials treated by ammonia freeze-explosion. Dale et al., 1984B.E. Dale, L.L. Henk and M. Shiang. s.l. : Dev. Ind. Microbiol., 1984, Vol. 26. 223-233. 140. Comparison of steam and ammonia pretreatment for enzymatic hydrolysis of cellulose. Mes-Hartree et al., 1988M. Mes-Hartree, B.E. Dale, W.K. Craig. s.l. : Appl. Microbiol. Biotechnol., 1988, Vol. 29. 462-468. 141. Supercritical carbon dioxide explosion as a pretreatment for cellulose hydrolysis. Zheng, Y., Lin, H.M., Wen, J. 8, s.l. : Biotechnology Letters, 1995, Vol. 17. 142. Carbonic acid enhancement of hydrolysis in aqueous pre-treatment of corn stover. Walsum, P.G., Shi, H. 3, s.l. : Bioresource technology, 2004, Vol. 93. 217-226. 143. Acid-catalyzed delignification of lignocellulosics. Sarkanen, K.V. s.l. : Academic Press, 1980, Vol. 2. 129-144. 144. Pretreatment of lignocellulosic biomass. McMillan, J.D. ACS Symp. Ser. 566, ed. ME Himmel, JO Baker, RP Overend. Washington : Amer. Chem. Soc., 1994. 292-324. 145. Modeling and optimization of the dilute-sulfuric-acid pre-treatment of corn stover, poplar and switchhgrass. Esteghlalian, A., Hashimoto, A.G., Fenske, J.J., Penner, M.H. s.l. : Bioresour. Technol., 1997, Vol. 59. 129-136. 146. High temperature acid hydrolysis of biomass using an engineering-scale plug flow reactor: result of low solids testing. A.H. Brennan, W. Hoagland, D.J. Schell. s.l. : Biotechnol. Bioeng. Symp. 17, 1986. 53-70. 147. Modeling of percolation process in hemicellulose hydrolysis. Cahela et al., 1983D.R. Cahela, Y.Y. Lee, R.P. Chambers. s.l. : Biotechnol. Bioeng., 1983, Vol. 25. 3-17. 148. Biotech, Wisconsin Education. [Online] 1997. [Cited: July 25, 2008.] http://www2.biotech.wisc.edu/jeffries/bioprocessing/bioconversion.html. 149. Lee, S., Speight, J.G., Loyalka, S.K. Handbook of Alternative Fuel Technologies. s.l. : CRC Press, 2007. 150. Bioconversion of Cellulosic Substances into Energy Chemicals and Microbial Protein Symposium Proceedings. J. Schurz, T.K. Ghose. New Delhi  : s.n., 1978. 37. 151. cAMP-mediated differential regulation of lignin peroxidase and manganese-dependent peroxidase production in the white-rot basidiomycete Phanerochaete chrysosporium. Boominathan, C.A. Reddy. s.l. : Proc. Natl. Acad. Sci., 1992. 106  152. Chemistry of delignification. Gierer, J. s.l. : Wood Science and Technology, 1985, Vol. 19. 289-312. 153. Sarkanen, K.V., Ludwig, C.H. Classification and Distribution, Lignins: Occurrence, Formation, Structure and Reactions. West Sussex, England : John Wiley and Sons, 1971. 154. Smook, G.A. Handbook of Pulp and Paper Technologists, 2nd Edition. Vancouver, BC : Angus Wilde Publications, 1992. 155. Oxygen Delignification Chemistry and its Impact on Pulp Fibers. Rallming Y., Lucian, L. Ragauskas, A.J., Jameel, H. s.l. : Journal of Wood Chemistry and Technology, 2003, Vol. 23. 13-29. 156. Oxygen Delignification. McDonough, T.J. s.l. : The Institute of Paper Chemistry, IPC Technical Paper Series, 1989. 318. 157. Oxygen Bleaching Practices and Benefits: An Overview. Tench, L., Harper, S. s.l. : Tappi Journal, 1987, Vol. 70. 158. Effect of oxygen delignification on the rate and extent of enzymatic hydrolysis of lignocellulosic material. Draude, K.M., Kurniawan, C.B., Duff, S.J.B. s.l. : Bioresource Technology, 2001, Vol. 79. 159. Effect of Oxygen Delignification Operating Parameters on Downstream Enzymatic Hydrolysis of Softwood Substrates. Charles, N., Mansfield, S.D., Mirochnik, O., Duff, S.J.B. 5, s.l. : Biotech. Prog., 2003, Vol. 19. 160. Violette, S.M. Oxygen delignification kinetics and selectivity improvement. s.l. : University of Maine, 1994. 161. McDonough, T.J. Oxygen Delignification. Appleton, Wisconsin : IPC Technical Paper Series, 1989. 318. 162. Delignifying High Yield Pulps with Oxygen and Alakli. Kleppe, P.J., Storebraten, S. s.l. : Tappi Journal, 1985, Vol. 68. 163. Reitberger, T., Gierer, J., Erquan, Y. Involvement of oxygen-derived free radicals in chemical and biochemical degradation of lignin. Oxidative delignification chemistry, Fundamentals and catalysis. Washington : American Chemical Society, 2001. 164. Oxygen Delignification Chemistry and its Impact on Pulp Fibers. Rallming Y., Lucian, L. Ragauskas, A.J., Jameel, H. s.l. : Journal of Wood Chemistry and Technology, 2003, Vol. 23. 13-29. 107  165. Formation and Involvement of Radicals in Oxygen Delignification Studied by the Autooxidation of Lignin and Carbohydrate Model Compounds. Gierer, J., Reitberger, T., Yang, E., Yoon, B-H. s.l. : Journal of Wood Chemistry and Technology, 2001, Vol. 21. 313. 166. Chemistry of Delignification. Gierer, J. Paris : International Symposium on Wood and Pulp Chemistry, 1987. 279. 167. Asgari, F., Argyropoulos, D. Fundamentals of oxygen delignification. Part II. Functional group formation/elimination in residual kraft lignin. s.l. : NRC, Canada, 1998. 168. Dogan, I. Mass Transfer and Kinetics in Oxygen Delignification. s.l. : Middle East Technical University, 2004. 169. Pulp Bleaching Principles and Practice. Dence, C.W., Reeve, D.W. Atlanta : Tappi Press , 1996. 170. Lennart, F., Arbin, A. The effect of oxygen and anthraquinone on the alkaline depolymerization of amylose. s.l. : Chalmers University of Technology, 1980. 171. PARR Series 4520 Operating Manual. PARR Instruments. [Online] March 2008. [Cited: April 25, 2008.] http://www.parrinst.com/doc_library/members/383M.pdf. 172. Clesceri, L.S., Greenberg, A.E., Eaton, A.D. Standard methods for the examination of water and wastewater, 20th Edition. s.l. : Standard Methods Committe, 1998. 173. Washing lignin precipitated from kraft black liquor. Ohman, F., Theliander, H. 5, s.l. : Paperi ja puu, 2006, Vol. 88. 174. Determination of tannin and lignin. Kloster, M.B. s.l. : J. Amer. Water Works Ass., 1973, Vol. 66. 175. The effect of initial pore volume and lignin content on the enzymatic hydrolysis of softwoods. Mooney, C.A., Mansfield, S.D., Touhy, M.G., Saddler, J.N. 2, s.l. : Bioresource Tech., 1998, Vol. 64. 176. Influence of Enzyme Loading and Physical Parameters on the Enzymatic Hydrolysis of Steam-Pretreated Softwood. Tengborg, C., Glabe, C., Zacchi, G. s.l. : Biotech. Prog., 2001, Vol. 17. 177. Effect of lignin content and magnesium-to-magnesium ratio on the selectivity of oxygen delignification in softwood kraft pulp. Lucian, L.A., Smereck, R.S. s.l. : Pure Appl. Chem., 2001, Vol. 73. 178. Estimating of Structural Composition of Wood and Non-Wood Biomass Samples. Demirbas, A. s.l. : Energy Sources, 2005, Vol. 27. 108  179. Earth Day ... by the numbers. Statistics Canada. [Online] April 2006. [Cited: January 25, 2008.] 180. Paper University. TAPPI Association. [Online] 2001. [Cited: February 21, 2008.] 181. Jolley, C. Using Recycled Paper. Highbeam Encyclopedia. [Online] 1990. [Cited: March 3, 2008.] 182. Optimization of Recycled Fibre in Linerboard. Appleton, Wisconsin : The Institute of Paper Chemistry, 1978. 183. Fibre Length Fractionation Caused by Pulp Screening, Slotted Screen Plates. Olson, J.A. 8, s.l. : Journal of Pulp and Paper Science, 2001, Vol. 27. 184. Enhanced Pulp Screening Using High-performance Screen Components and Process Simulation. Wechroth, R., Marshall, L., Tuomela, P., Gooding, R. Durban : African Pulp and Paper Week, 2002. 185. Gibbs, R.D. The Maule reaction, lignins, and the relationship between woody plants. The Physiology of Forest Trees. 1958. 186. Comparison of steam pretreatment of Eucalytpus, Aspen, and spruce wood chips and their enzymatic hydrolysis. Ramos, L.P., Breuil, C., Saddler, J.N. s.l. : Appl. Biochem. Biotech., 1992, Vol. 35. 187. Pulping: Oxygen and Extended Delignification. Clean Air, Clean Water, Pulp Info Centre. [Online] 2004. [Cited: February 27, 2008.] http://www.rfu.org/cacw/basic4KraftPulp.htm. 188. Helping you solve the fibre puzzle. Canfor. [Online] 2006. http://www.temap.com/assets_main/documents/Morph_mini.pdf. 189. Dixie, J. The Price Sheet. CSR Composite Index. [Online] May 2007. 190. Upgrading of pulp from corrugated containers by oxygen delignification. De Ruvo, A., Farnstrand, P.A., Hagen, N. 12, s.l. : Tappi Journal, 2001, Vol. 71. 191. McDonough, T.J. Bleaching Agents - Pulp and Paper Industry. Atlanta : IPS Technical Paper Series, 1993. 192. Pulping. Reach for Unbleached Foundation. [Online] 2008. [Cited: June 5, 2008.] http://www.rfu.org/cacw/basic4KraftPulp.htm. 193. Cellulosic Biomass Chemical Pretreatment Technologies. [Online] September 2007. [Cited: June 5, 2008.] http://www.wvdo.org/community/Pauley.pdf. 109  194. Kinetics of oxygen delignification. Argarwal, S.B., Genco, J.M., Cole, B.J., Miller, W. 10, s.l. : Journal of Pulp and Paper Science, 1999, Vol. 25. 195. The role of lignin in the adsorption of cellulases during enzymatic treatment of lignocellulosic material. Sutcliffe, R., Saddler, J.N. s.l. : J. Wiley, 1986. 196. Enzyme recovery in high-solids enzymatic hydrolysis of steam-pretreated willow. Pristavka, A.A., Salovarova, V.P., Zacchi, G. 3, s.l. : MAIK Nauka/Interperiodica, 2000, Vol. 36. 197. The effect of fibre characteristics on hydrolysis and cellulase accesibility on softwood substrates. Mooney, C., Mansfield, S., Beatson, R.P., Saddler, J.N. s.l. : Enzyme and Microb. Tech., 1998, Vol. 25. 198. Enzyme activity of cellulase adsorbed on cellulose and its change during hydrolysis. Ooshima, H., Kurakake, M., Kato, J., Harano, Y. s.l. : Applied Biochem. Biotechnol., 1991, Vol. 3. 199. Substrate reactivity as a function of the extent of reaction in the enzymatic hydrolysis of lignocellulose. Desai, S.G., Converse, A.O. s.l. : Biotech. Bioeng., 1997, Vol. 56. 200. Changes in the Enzymatic Hydrolysis Rate of Avicel Celulose with Conversion. Yang, B., Willies, D.M., Wyman, C.E. s.l. : Wiley InterScience (www.interscience.wiley.com), 2006. 201. Cellulase adsorption-desorption and cellulose saccharification during enzymatic hydrolysis of cellulose materials. van Wyk, J.P.H. 8, s.l. : Biotech. Letters, 1997, Vol. 19. 202. Ching-Tsang, H. Handbook of industrial biocatalysis. s.l. : CRC Press, 2005. 203. Energy Efficiency and Renewable Energy. U.S. Department of Energy. [Online] 2001. [Cited: January 14, 2008.] http://www1.eere.energy.gov/biomass/lignin_derived.html. 204. Lignin recovery from kraft black liquor: preliminary process design. Loutfi, H., Blackwell, B. s.l. : Tappi Journal, 1991, Vol. 74. 205. Uloth, V.C., Wearing, J.T. 9, s.l. : Pulp and Paper Canada, 1989, Vol. 90. 206. Ohman, F. Precipitation and separation of lignin from kraft black liquor. Gotenberg, Sweden : Chalmers University of Technology, 2006. 207. Associative interactions between kraft lignin components. Sarkanen, S., Teller, D.C., Stevens, C.R., McCarthy, J.L. s.l. : Macromolecules, 1984, Vol. 17. 208. Analytical Compendium. International Union of Pure and Applied Chemistry. [Online] [Cited: February 17, 2008.] http://www.iupac.org/publications/analytical_compendium/Cha09sec50.pdf. 110  209. Vainio, U., Maximova, N., Laine, J. A small-angle x-ray scattering study on the morphology of kraft lignin. Helsinki : University of Helsinki, 2006. 210. Sillanpaa, M. Studies on washing in kraft pulp bleaching. s.l. : Department of Process and Environmental Engineering, University of Oulu, 2005. 211. Lignin separation from kraft black liquor by tangential ultrafiltration. Rojas, O.J., Song, J., Argyropoulos, D.S. s.l. : La Chimica e l'Industria, 2006, Vol. 1. 212. Ala-Kaila, K. Modeling of mass transfer phenomena in pulp-water suspensions. Helsinki : Helsinki University of Technology, 1998. 213. Water, Office of. Development document for effluent limitations guidelines and standards for the pulp, paper, and paperboard category. Washington : United States Environmental Protection Agency, 1997. 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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


Related Items