Microbial Pretreatment of Lignocellulosic Biomass with Ceriporiopsis. Subvermispora for Enzymatic Hydrolysis and Ethanol Production.

Transcription

1 Microbial Pretreatment of Lignocellulosic Biomass with Ceriporiopsis Subvermispora for Enzymatic Hydrolysis and Ethanol Production Dissertation Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University By Caixia Wan, M.S. Graduate Program in Food, Agricultural and Biological Engineering The Ohio State University 2011 Dissertation Committee: Dr. Yebo Li, Advisor Dr. Ann D. Christy Dr. Thaddeus Ezeji Dr. Frederick C. Michel

2 Copyright by Caixia Wan 2011

3 Abstract Pretreatment is a crucial step in the conversion of lignocellulosic biomass to fermentable sugars. Microbial pretreatment with white rot fungi can delignify lignocellulosic biomass under atmospheric conditions and, thus, has the potential to be applied to on-farm wet storage for cost-effective cellulosic ethanol production. The present research investigated the feasibility of solid-state microbial pretreatment of corn stover and other types of lignocellulosic biomass feedstocks with the white rot fungus Ceriporiopsis subvermispora. Pretreatment parameters, including moisture content, temperature, time, and particle size of corn stover, were studied for their effects on fungal pretreatment and the subsequent cellulose digestibility. The results showed that up to 31.59% lignin was degraded during 18-d pretreatment of corn stover (5-10 mm) with 75% initial moisture content and at 28 ºC. The glucose and ethanol yields, based on the theoretical yield of untreated corn stover, reached 66.61% and 57.80%, respectively, after a 35-d pretreatment, which were more than three times of that obtained with the untreated corn stover. Washing of fungal-pretreated corn stover did not cause significant improvement in ethanol yield, suggesting no water insoluble inhibitory compounds were formed during fungal pretreatment. A high correlation was obtained between the remaining lignin content in the treated corn stover and its cellulose digestibility, indicating that lignin removal facilitated enzymatic hydrolysis. The highest glucose yield of 71.81% was ii

4 obtained when the enzymatic hydrolysis process was supplemented with a commercial xylanase/cellulase enzyme complex. The delignification process of corn stover was investigated throughout a 42-d pretreatment by monitoring changes in composition and enzyme production. Lignin degradation increased with pretreatment time, reaching 39.20% at the end of pretreatment while cellulose degradation was less than 4.52% throughout the process, indicating a high selectivity of the fungus for lignin in the corn stover. However, hemicellulose degradation, mainly xylose loss, was substantial with up to 27.52% hemicellulose simultaneously degraded with lignin. Manganese peroxidase (MnP) and laccase were two detected oxidative enzymes associated with lignin degradation although their activities were not necessarily correlated with lignin degradation. For hydrolytic enzymes, xylanase was detected at a significant level of activity while cellulase, revealed as FPU and CMC activity, was not detectable. Observation by scanning electron microscopy showed invasive colonization of mycelial mass on the exterior surface and numerous holes and erosion troughs formed on corn stover, indicating that fungal pretreatment changed the microstructure of corn stover significantly. At the later stage of pretreatment, corn stover became light, soft, and spongy and the color changed to whitish-yellow. To evaluate the robustness of the fungal pretreatment process with C. subvermispora, other feedstocks, including wheat straw, soybean straw, switchgrass, and hardwood, were also examined. After pretreatment for 18 d, switchgrass and hardwood were effectively delignified with significant production of manganese peroxidase and laccase. The corresponding glucose yields reached 37.15% and 24.21% (of theoretical iii

5 glucose yield of raw materials), for switchgrass and hardwood, respectively, which was a 1-2 fold increase over untreated materials. An additional 10-30% increase was obtained with an extension of pretreatment time to 35 d. Wheat straw was greatly resistant to fungal pretreatment unless glucose and malt extract were added to the substrate. In contrast, no fungal degradation occurred in soybean straw even with addition of external carbon sources and enzyme inducers (Mn 2+, H 2 O 2 ). In order to investigate recalcitrance of wheat straw and soybean straw, hot water extraction (HWE) at 85 ºC and/or liquid hot water (LHW) pretreatment at 170 ºC were conducted prior to fungal pretreatment. HWE pretreatment removed extractives in substrates, which facilitated fungal degradation of wheat straw but not of soybean straw. Wheat straw pretreated by combined HWE and C. subvermispora doubled the glucose yield compared to the untreated straw. Soybean straw was further studied by applying a LHW pretreatment. LHW alone resulted in a 42.27% glucose yield, which was about 9% higher than that of the untreated soybean straw. Fungal pretreatment of LHW-pretreated soybean straw led to 36.70% lignin removal and 64.25% glucose yield. Empirical models satisfactorily predicted fungal growth (R 2 =0.98), oxygen uptake (R 2 =0.99), holocellulose consumption (R 2 =0.95), and lignin degradation (R 2 =0.94); however, the prediction of enzyme production was poor (R 2 =0.68). Nevertheless, these models adequately described the interrelationship between fungal growth and feedstock degradation and provided useful information for optimization and scale-up of a fungal pretreatment process being developed. iv

6 The knowledge obtained from this study is important for the development of concurrent wet storage and microbial pretreatment with white rot fungi for lignocellulosic biomass and of a combined fungal and thermal/physical pretreatment process that can potentially overcome the problems associated with existing pretreatment methods. v

7 Dedication This document is dedicated to my family vi

8 Acknowledgments I would like to express my deepest appreciation to my advisor, Dr. Yebo Li, for his infinite support, encouragement, patience as well as invaluable mentorship he provided me throughout my graduate study. I also sincerely thank him for his precious help and enthusiastic support on my professional development. I would like to gratefully thank professors serving as my dissertation committee: Dr. Ann D. Christy, Dr. Thaddeus Ezeji, and Dr. Frederick C. Michel for their suggestions, advice on various topics, reviewing dissertation, and giving important comments. Special thanks go to Mrs. Mary Wicks for her carefully proofreading my dissertation and providing useful suggestions. I also wish to thank Mike Klingman for his countless support on instrument maintenance and experimental setup throughout my doctoral project. Dr. Harold Keener, Dr. Jay Martin, Mrs. Peggy Christman, and Mrs. Candy McBride are also acknowledged for their administrative support. For those faculty and staff members in the department of FABE, who gave me help and encouragement during my stay in OARDC/OSU, Dr. Sukhbir Grewal, Dr. Robert Hansen, Dr. Peter Lin, and Dr. Heping Zhu are also acknowledged. The colleagues in our research group, Mr. Stephen Y. Park, Ms. Lo Niee Liew, Dr. Zhifang Cui, and Mr. Shengjun Hu, are especially acknowledged for their offering so vii

9 much support and help to me. The former colleagues, Dr. Jiying Zhu and Dr. Guiming Fu are also acknowledged for their help. I feel greatly indebted to my parents and my family for their care, consistent love, encouragement, understanding and warm support. Lastly and most importantly, I would like to thank the love of my life, my husband Hai Huang, for his devotion and love, endless support, patience, and encouragement. Without his love and encouragement, I would not have been able to accomplish this dissertation. viii

10 Vita September 11, Born Ningbo, China B.S. Food Science and Technology, Jiangnan University M.S. Fermentation Engineering, Jiangnan University 2007 to present...graduate Research Associate, Department of Food, Agricultural and Biological Engineering, The Ohio State University Publications Wan, C., and Li, Y Microbial pretreatment of corn stover with Ceriporiopsis subvermispora for enzymatic hydrolysis and ethanol production. Bioresource Technology 101(16): Wan, C., and Li, Y Microbial delignification of corn stover by Ceriporiopsis subvermispora for improving cellulose digestibility. Enzyme and Microbial technology 47: Zhu, J., Wan *, C., and Li, Y Enhanced solid-state anaerobic digestion of corn stover by alkaline pretreatment. Bioresource Technology 101: ( * equal contribution with the first author) Wan, C., Li, Y., Shahbazi, A., and Xiu, S Succinic acid production from cheese whey using Actinobacillus succinogenes 130 Z. Applied Biochemistry and Biotechnology 145: Fields of Study Major Field: Food, Agricultural and Biological Engineering Study in: Biological Engineering ix

11 Table of Contents Abstract Dedication... ii... vi Acknowledgments... vii Vita... ix Table of Contents... x List of Tables... xiv List of Figures... xv Chapter 1 Introduction Background Research objectives Contribution of the dissertation... 7 Chapter 2 Literature Review Introduction Lignin and biodelignification Solid-state fungal pretreatment process Inoculum Moisture content Particle size Supplements Temperature Aeration Decontamination Time Degrading enzymes x

12 Ligninolytic enzymes Hydrolytic enzymes Enzymatic hydrolysis and ethanol fermentation Non-selective degradation Selective degradation Combination of microbial pretreatment with other pretreatment methods Enzymatic treatment Modeling and scale up Concluding remarks Chapter 3 Effect of Pretreatment Parameters on Fungal Pretreatment of Corn Stover Introduction Materials and Methods Materials Fungus and inoculum preparation Fungal pretreatment Enzymatic hydrolysis Simultaneous saccharification and fermentation (SSF) Analytical methods Statistical analysis Results and Discussion Degradation of corn stover Enzymatic hydrolysis Ethanol production Conclusions Chapter 4 Degradation Mechanism of Fungal Pretreatment of Corn Stover Introduction Material and Methods Feedstock Fungus and inoculum preparation Pretreatment by solid-state fermentation Compositional analysis of corn stover samples Enzyme extraction and assays Enzymatic hydrolysis xi

13 4.3. Results and discussion Degradation of corn stover Lignocellulolytic enzymes and their effects on corn stover degradation Morphology of fungal-treated corn stover Correlation between fungal degradation and enzymatic digestibility Conclusions Chapter 5 Effectiveness of Fungal Pretreatment on Different Biomass Feedstocks Introduction Materials and methods Feedstocks Fungus and inoculum preparation Fungal pretreatment Enzymatic hydrolysis Enzyme assay Analytical methods Results and discussion Composition of biomass Degradation of biomass Ligninolytic enzymes Enzymatic hydrolysis Effect of nutrients and inducers Conclusions Chapter 6 Enhanced Enzymatic Hydrolysis of Biomass Feedstocks by Combined Hot Water Extraction/Hydrothermal and Fungal Pretreatment Introduction Materials and methods Raw materials Fungus and inoculum preparation Hot water extraction Liquid hot water pretreatment Fungal pretreatment Enzymatic hydrolysis Analytic method Results and discussion Effect of hot water extraction (HWE) xii

14 Effect of liquid hot water (LHW) pretreatment Conclusions Chapter 7 Modeling of Fungal Pretreatment of Corn Stover Introduction Materials and methods Materials Fungus and inoculum preparation Fungal pretreatment Analytical methods Mathematical modeling Fungal growth Oxygen consumption Substrate consumption Production model Lignin degradation kinetics Simulation method Results and discussion Growth kinetics Oxygen consumption Substrate consumption Enzyme production Lignin degradation Conclusions Chapter 8 Conclusions and Suggestions for Future Research Conclusions Suggestions for future research References xiii

15 List of Tables Table 2.1. The effect of white rot fungi on enzymatic hydrolysis and ethanol production Table 2.2. Combined white-rot fungal pretreatment and physical/chemical pretreatment Table 3.1. Degradation of corn stover pretreated by C. subvermispora at 28 ºC for 18 d Table 5.1. Composition of raw feedstocks Table 7.1. Parameter estimation results for fungal pretreatment models xiv

16 List of Figures Figure 3.1. Interactive effects of moisture content, pretreatment time, and particle size on enzymatic hydrolysis of corn stover pretreated by C. subvermispora Figure 3.2. Effect of pretreatment temperature on enzymatic hydrolysis of corn stover pretreated by C. subvermispora Figure 3.3. Effect of enzyme loading on enzymatic hydrolysis of corn stover pretreated by C. subvermispora Figure 3.4. Effect of supplemental enzymes on enzymatic hydrolysis of corn stover pretreated by C. subvermispora Figure 3.5. Ethanol yield of corn stover pretreated by C. subvermispora Figure 4.1. Degradation of corn stover during 42-d pretreatment by C. subvermispora Figure 4.2. Enzyme production during 42-d pretreatment of corn stover by C. subvermispora Figure 4.3. Scanning electron micrograph of corn stover pretreated by C. subvermispora for 18 d Figure 4.4. Enzymatic hydrolysis of corn stover pretreated by C. subvermispora for up to 42 d xv

17 Figure 4.5. Correlation between the sugar yield and remaining lignin content in corn stover pretreated by C. subvermispora Figure 5.1. Degradation of feedstocks after 18-d fungal pretreatment by C. subvermispora Figure 5.2. Ligninolytic enzyme production during 18-d fungal pretreatment by C. subvermispora Figure 5.3. Enzymatic hydrolysis of feedstocks after 18-d pretreatment by C. subvermispora Figure 5.4. Effect of pretreatment time on enzymatic hydrolysis of feedstocks Figure 5.5. Effect of addition of nutrients and additives to fungal pretreatment on enzymatic hydrolysis of wheat straw (a) and soybean stalk (b) Figure 6.1. Synergistic effects of hot water extraction and fungal pretreatment on degradation of biomass feedstocks Figure 6.2. Synergistic effects of hot water extraction and fungal pretreatment on enzymatic hydrolysis of biomass feedstocks Figure 6.3. Synergistic effects of liquid hot water pretreatment and fungal pretreatment on degradation of biomass feedstocks Figure 6.4. Synergistic effects of liquid hot water pretreatment and fungal pretreatment on enzymatic hydrolysis of biomass feedstocks Figure 7.1. Validation of fungal growth model for 35-d fungal pretreatment Figure 7.2. Validation of oxygen uptake model for 35-d fungal pretreatment xvi

18 Figure 7.3. Validation of substrate (holocellulose) consumption model for 35-d fungal pretreatment Figure 7.4. Validation of enzyme production model for 35-d fungal pretreatment Figure 7.5. Validation of lignin degradation model for 35-d fungal pretreatment xvii

19 Chapter 1 Introduction 1.1. Background Worldwide addiction to petroleum fuels has led to serious problems, such as environmental pollution, trade deficits, shortages and uncertainty of petroleum sources. In order to reduce dependence on petroleum fuels, biofuels derived from renewable and local sources have received extensive interest for displacement of fossil transportation fuels in many countries. The Biomass Program of the U.S. Department of Energy has projected production of 36 billion gallons of ethanol by 2030, which will supply one third of current gasoline demand (DOE, 2006; Hess et al., 2009). Currently, the majority of ethanol production in the U.S is based on corn grain, which supplied about 10 billion gallons in 2009 (DOE, 2011). Corn ethanol supplies will be limited in the near future due to its competition with feed and food production (DOE, 2011). Cellulosic ethanol production is becoming feasible through research and development because lignocellulosic biomass is rich in carbohydrates (55-75% dry matter) and widely available, including, but not limited to, agricultural residues (e.g., corn stover, wheat straw), dedicated energy crops (e.g., switchgrass, miscanthus), and forestry residues (e.g., poplar, pine) (Mosier et al., 2005a). According to the Billon-Ton Annual Supply report, more than1.3 billion tons of biomass could be available annually from agricultural and forestry industries in the U.S., which is enough to produce fuel ethanol to meet 30% of 1

20 current gasoline consumption (Perlack et al., 2005). Despite great potential as a feedstock for ethanol production, native lignocellulosic biomass is highly resistant to enzymatic hydrolysis due to its structural characteristics (Chandra et al., 2007; Himmel et al., 2007; Wright et al., 1988). Cellulose is tightly packed and highly crystallized, showing recalcitrance to depolymerization. Furthermore, hemicellulose is covalently linked with cellulose, providing a protective sheath to cellulose. Lignin, a complex phenolpropanoid polymer, is the major barrier to enzymatic hydrolysis by impeding penetration of enzymes into polysaccharides. Therefore, these substrate-related factors contribute to low enzymatic digestibility (<20% of theoretical yield) of lignocellulosic biomass (Mosier et al., 2005a). In order to harness carbohydrates in lignocellulosic biomass, pretreatment is a crucial step to reduce recalcitrance for ethanol production. The main effects of pretreatment include increased accessible surface area, cellulose decrystallization, removal of hemicellulose, defragmentation of lignin, and alteration of lignin structure (Mosier et al., 2005a). Pretreatment methods, including steam explosion, liquid hot water, lime, ammonia explosion, and dilute acid, have been regarded as the current leading pretreatments (Galbe and Zacchi, 2008). However, disadvantages of these pretreatments offset, to some extent, their economic and environmental feasibility. In general, they require reactors and equipment that can tolerate corrosion or high pressure/temperature. Generation of waste streams imposes potential environmental pollution and requires specific treatments before disposal. Severe pretreatment generally has resulted not only in remarkably improved polysaccharide digestibility but in formation of fermentation- 2

21 inhibiting compounds (e.g., weak acids, furan derivatives, phenolic compounds) which require detoxification prior to fermentation (Palmqvist and Hahn-Hagerdal, 2000). Thus, pretreatment still remains one of the most costly steps in cellulosic ethanol production and is a significant barrier to its commercialization (Mosier et al., 2005a). White rot fungi are the most effective lignin-degrading microorganisms and have been traditionally used for a wide range of applications such as biopulping, forage upgrading, and bioremediation of soil and wastewater (Sanchez, 2009). Recently, pretreatment with white rot fungi under solid state fermentation has been studied for improving enzymatic digestibility of various lignocellulosic biomass such as corn stover (Keller et al., 2003; Xu et al., 2010), wheat straw (Dias et al., 2010), cotton stalks (Shi et al., 2007), and woody biomass (Yu et al, 2009). This technology has advantages over thermochemical pretreatments, including, but not limited to, simple techniques, low energy requirements, no or reduced output of waste stream, reduced downstream processing costs, and no or reduced inhibitors to ethanol fermentation (Keller et al., 2003; Niagam and Pandey, 2009). The effectiveness of microbial pretreatment largely depends on fungal species and substrate sources. White rot fungi with a high selectivity for lignin degradation over cellulose consumption possess the most important factor for pretreatment. Selective degradation ensures a cellulose-rich but highly delignified fiber for enzymatic hydrolysis and ethanol fermentation although it often results in considerable hemicellulose loss (Eriksson et al., 1990). Among commercially available white rot fungi, Ceriporiopsis subvermispora is highly lignin selective due to its lack of a complete cellulase system and 3

22 is regarded as one of the most effective white rot fungi for biopulping (Akhtar et al., 1998; Ferraz et al., 2003). However, there is no information in regards to its application to microbial pretreatment for cellulosic ethanol production. Irrespective of selective patterns, the degradation capability of white rot fungi varies among substrate types due to different chemical structures (Akin et al., 1995; Anderson and Akin, 2008). Modification of lignocellulosic biomass with mild chemical or physical pretreatment could facilitate the fungal degradation performance and improve delignification efficiency to a great extent (Kadimaliev et al., 2003; Yu et al., 2008). The addition of carbon/nitrogen sources, mineral solutions, or enzyme inducers could also improve fungal delignification processes (Reid, 1989a; Shrestha et al., 2008). Solid-state microbial pretreatment generally requires a few weeks to months to achieve significant delignification. Low degradation efficiency makes this technology less feasible for direct industrial use. However, emerging on-farm in-storage pretreatment processes generally incorporate long pretreatment times that can provide year-round pretreated feedstock to a biorefinery (Digman et al., 2010; Shinner et al., 2007). Fungal pretreatment applied as an on-farm wet storage pretreatment would provide a pretreated biomass residue with much higher cellulose digestibility than traditional ensilages that harness lactic acid bacteria for fermentation. In addition, similar to methods used in the biopulping industry, fungal pretreatment could be used as the first step for other pretreatment methods (Galbe and Zacchi, 2008). Synergism arising from pretreatment combinations could overcome potential weaknesses of a single pretreatment by shortening the pretreatment time, saving energy in mechanical pretreatment, and reducing 4

23 the severity and toxicity of chemical pretreatments (Kang et al., 2003; Messner et al., 1998). In this research, C. subvermispora was used to degrade lignocellulosic biomass under solid state fermentation. Corn stover was used as the model feedstock to develop a fungal pretreatment which was then applied to other feedstocks. Pretreatment conditions, including moisture content, temperature, time, and particle size of corn stover, were studied for their effects on fungal pretreatment and subsequent cellulose digestibility. Ethanol production was carried out by simultaneous saccharification and fermentation for evaluating the effectiveness of fungal pretreatment. Degradation mechanisms were investigated from several aspects, including secretion of ligninolytic and hydrolytic enzymes by the fungus and quantitative analysis of cell wall component, throughout the pretreatment process. In addition, cell wall deconstruction was determined qualitatively by observation of microstructure with scanning electron microscopy. The application of the fungal pretreatment process on various feedstocks provided an understanding of specific substrate-fungal interactions and of the robustness of fungal pretreatment processes for different sources of feedstocks. To improve degradation of recalcitrant feedstocks, amendments of culture media with external carbon sources and enzyme inducers were investigated. For the same purpose, a mild non-fungal pretreatment was also investigated for the potential synergistic effect on fungal pretreatment and resulting cellulose digestibility. Mathematical modeling contributed to our understanding of interrelationship between fungal growth and feedstock degradation and provided valuable information for the design, operation and scale-up of a solid-state fungal pretreatment 5

24 reactor. Therefore, the present research addressed problems associated with the current pretreatment processes (e.g., non-biological, biological) of lignocellulosic biomass, providing a feasible microbial pretreatment process with the potential to reduce environmental impacts and costs of cellulosic ethanol production Research objectives The goal of this project was to develop a cost-effective and environmentally friendly microbial pretreatment process to reduce the recalcitrance of lignocellulosic biomass for ethanol production. To attain this goal, the following specific objectives were accomplished: 1) Determine the effects of pretreatment parameters such as moisture content, temperature, time, and particle size on degradation of corn stover with C. subvermispora, and investigate the resulting enzymatic hydrolysis and ethanol production (Chapter 3). 2) Determine the degradation mechanism of corn stover during fungal pretreatment by investigating the production of ligninolytic and hydrolytic enzymes as well as morphology changes of the substrate (Chapter 4). 3) Evaluate the robustness of fungal pretreatment for other types of biomass feedstocks including agricultural residues (i.e., wheat straw and soybean straw), hardwood, and a dedicated energy crop (i.e., switchgrass) (Chapter 5). 4) Evaluate the synergistic effect of hydrothermal pretreatment (i.e., exhaustive hot water extraction and liquid hot water pretreatment) on fungal pretreatment of 6

25 feedstocks, especially those strongly resistant to fungal degradation (Chapter 6). 5) Investigate the interrelationship between fungal growth and substrate degradation during microbial pretreatment by modeling fungal growth, oxygen uptake, substrate consumption, enzyme production, and lignin degradation using empirical equations (Chapter 7) Contribution of the dissertation Two papers have been published in peer-reviewed journals, which are included in Chapter 3 and 4 in the dissertation: 1) Wan, C., and Li, Y Microbial pretreatment of corn stover with Ceriporiopsis subvermispora for enzymatic hydrolysis and ethanol production. Bioresource Technology 101(16): ) Wan, C., and Li, Y Microbial delignification of corn stover by Ceriporiopsis subvermispora for improving cellulose digestibility. Enzyme and Microbial technology 47 (1-2):

26 Chapter 2 Literature Review 2.1. Introduction To reduce dependence on petroleum and thus alleviate national security and global warming concerns, the Renewable Fuel Standard (RFS) program requires the use of 36 billion gallons per year (BGY) of biofuels by 2020 (EPA, 2007). Currently, ethanol in the U.S. is primarily produced from starch- or sugar- based crops (e.g., corn, sugar beet, sugar cane), which imposes pressure on arable land as well as food and feed supplies. Lignocellulosic biomass, mainly from agricultural residues and forestry sources, is the most promising alternative source for ethanol production as it is sufficiently available worldwide. Corn stover, as the most abundant agricultural residue in the U.S., has the greatest potential to be used for ethanol production. Based on projections by the U.S. Department of Agriculture (USDA) and the U.S. Department of Energy (USDOE), the available amount of corn stover is estimated to be million dry tons annually in the U.S. (Perlack et al., 2005). The complex structure of native lignocellulosic biomass is highly resistant to enzymatic digestion due to cellulose crystallinity, limited accessible surface area, and protective sheaths formed by lignin and hemicelluloses (Himmel et al., 2007). For example, enzymatic hydrolysis of native corn stover generally results in less than 20% glucose yield (Mosier et al., 2005a). Therefore, a pretreatment process is needed to 8

27 reduce the biomass recalcitrance for improved enzymatic hydrolysis by breaking lignin seals and disrupting the crystalline structure of cellulose. However, most of the current leading pretreatment technologies involve the use of chemicals (acid or alkali) at elevated temperature and pressure. The severe operational conditions of thermochemical pretreatment usually lead to at least partial degradation of fermentable sugars to furfural, 5-hydroxymethylfurfural (HMF), levulinic acid, and formic acid that inhibit enzymatic hydrolysis and ethanol fermentation (Galbe and Zacchi, 2007). Thermochemical pretreatments also have disadvantages in terms of requiring extensive washing, treatment of chemical waste, and pressurized and corrosion resistant reactors. Pretreatment, thus, has been regarded as one of the most expensive steps for cellulosic ethanol production as it accounts for one third of the total cost (Mosier et al., 2005a). Therefore, from both economic and environmental perspectives, effective pretreatment processes in terms of improved enzymatic digestibility, reduced formation of inhibitors, and simplified upstream and downstream processes are urgently needed for commercial cellulosic ethanol production. Microbial pretreatment of lignocellulosic biomass by solid state cultivation is considered to be a low cost and environmentally friendly process with advantages including no use of severe chemicals, reduced energy input, no requirement for pressurized and corrosion-resistant reactors, no waste stream generated, and reduced or no generation of inhibitors to fermentation (Keller et al., 2003). White rot fungi, which are the most effective lignin-degrading microorganisms, have been receiving extensive attention for biodelignification of lignocellulosic biomass and improvement of cellulose 9

28 digestibility. White rot fungi with high selectivity of lignin degradation over cellulose are important for successful microbial pretreatment. However, the patterns of cell wall deconstruction by white rot fungi vary among species and strains. Most white rot fungi, such as Phanerochaete chrysosposrium, simultaneously degrade holocellulose (cellulose and hemicellulose) and lignin, resulting in a low cellulose recovery (Eriksson et al., 1990). Some species preferentially degrade lignin and part of the hemicelluloses, leaving a cellulose-rich residue (Anderson and Akin, 2008; Eriksson et al., 1990). Ceriporiopsis subvermispora has high lignin-degrading selectivity, especially for removal of aromatics like ester-linked phenolic acids (Akin et al., 1993). Pretreatment of woody biomass by this fungus contributes to significant energy savings in mechanical pulping, low severity for chemical pulping, improved paper strength properties, and reduced toxicity of pulping waste (Kang et al., 2003; Akhtar et al., 1998). In addition to pretreatment of wood by C. subvermispora for biopulping, enhanced methane yield was also obtained with C. subvermispora-pretreated softwood (Amirta et al., 2006). In the study of combined pretreatment of beech wood with C. subvermispora and ethanolysis, fungal pretreatment caused a 60% improvement in ethanol yield compared to ethanolysis pretreatment only (Itoh et al., 2003). Taniguchi et al. (2005) examined degradation of rice straw by C. subvermispora and found that 21% of Klason lignin was removed after 24 d of pretreatment. Improved forage digestibility was also observed with C. subvermisporatreated sugarcane bagasse (Okano et al., 2006) and Bermuda grass (Akin et al., 1993). 10

29 2.2. Lignin and biodelignification Lignin, a complex phenolpropanoid polymer, plays important roles in plant growth and development by providing mechanical support to xylem, improving water transport through xylem tracheary elements, and resisting microorganism attack on plant tissues (Compbell and Sederoff, 1996). Lignin is mainly composed of three hydroxycinnamyl alcohol monomers that differ in their degree of methoxylation: p- coumary, coniferul and sinapyl alcohols (Campbell and Sederoff, 1996). As a heterogeneous polymer, lignin varies in subunit composition, intermolecular linkages, and quantities between different sources of lignocellulosic biomass and between tissues within the same plants. For example, softwood is rich in guaiacyl lignin (90%); hardwood consists of about equal amounts of guaiacyl and syringyl lignin; and herbaceous plants are made of guaiacyl and syringlyl lignin and significant amounts of P-hydroxyphenyl lignin (10-20%) (Boerjan et al., 2003). Even in the same plant, histochemical staining showed more coniferyl lignin in vascular tissues of leaves and stems of grasses than in other tissues (Akin, 2007). Vascular tissues exhibited the most recalcitrance in grasses. In leaf blade sclerenchyma (extensions of the vascular bundles) and the parenchyma cell walls of mature stems, more syringyl (dimethoxylated) lignin was found, which can be partially degraded and is more susceptible to some chemical treatments such as alkaline pretreatment. High contents of ester-linked p-coumeric and ferulic units (recalcitrance factor of grass) were also found in nonlignified cell walls. Therefore, various factors, such as the ester-linking of phenolic acid, co-polymerization of lignin subunits, and 11

30 cross-linking between lignin and hemicelluloses, restrict degradability of plant cell walls (Grabber, 2005). The ability of white rot fungi to delignify varies among genera and species as well as in various fungal-substrate combinations (Akin et al., 1995; Anderson and Akin, 2008). In delignification of Bermudagrass, C. subvermispora and Cythus stercoreus were capable of attacking both ester- and ether-linked phenolic acid from unlignified cell walls and had increased degradation of guaiacyl lignin over syringyl lignin (Akin et al., 1995). In contrast, in the degradation of wood by some white rot fungi, syringyl lignin was degraded more rapidly than guaiacyl lignin (Eriksson et al., 1990). Taniguchi et al. (2005) tested the degradation of rice straw by four white rot fungi (P. chrysosporium, Trametes vericolor, C. subvermispora, and Pleurotus ostreatus), and found that lignin removal by C. subvermispora was lower than by P. ostreatus, probably due to lignin components in rice straw, such as p-coumeric units, that were more resistant to C. subvermispora Solid-state fungal pretreatment process Inoculum Inoculum for solid state fermentation can be prepared by different methods, e.g., mycelium grown in liquid or agar medium, spawn grown in cereal grains, or fungalprecolonized substrate (Reid et al., 1989a). P. chrysosporium yields spores which is a convenient inoculum and can be mixed evenly with the substrate. In contrast, most white rot basidiomycetes do not produce spores. Instead, the precolonized lignocellulosic 12

31 biomass is generally used for inoculum. Similar to liquid fermentation, fermented materials in the reactor can also serve as inoculum and the fresh substrate can be fed to partially replace fermented materials. A minimum level of inoculum is generally required for effective colonization and subsequent delignification (Reid, 1989b). However, a further increase in inoculum level may only have a marginal effect on the fungal colonization rate and the subsequent reaction. Akhtar et al. (1998) tested different levels of inoculum using precolonized wood chips by P. chrysosporium for decay of aspen wood. The results showed that a 2-5% inoculation level saved a similar amount of energy for biomechanical pulping while a futher increase in the inoculation level to 20% did not correspondingly save energy for mechanical pulping Moisture content Initial moisture content of the substrate is important to the fungal growth and secondary metabolism in solid state fermentation (Reid, 1989a). Low moisture could prevent fungal growth and hamper substrate degradation and it has been observed that moisture content less than 50% had no significant improvement on lignin removal and in vitro digestibility of wheat straw (Zadrazil and Brunnert, 1982). An increase in moisture content can lead to increased lignin degradation. Many studies indicated that initial moisture ranging from 70-80% was the optimal level for lignin degradation and ligninolytic activities. When supplemented with nutrients, moisture content as low as 60% was also effective for degradation of rice straw (Taniguchi et al., 2005). Shi et al. (2008) observed that after 14-d cultivation of cotton stalks by P. chrysosposrium, lignin 13

32 degradation of 27.6% was obtained at a moisture content of 75% in the substrate, which was approximately 7% higher than that at a moisture content of 65% but was not significantly different from that at a moisture content of 80%. Asgher et al. (2006) investigated the solid state cultivation of P. chrysoporium on corn cobs with a moisture content ranging from 40% and 90%. The highest ligninase activity was obtained at 70% moisture content. Thus, the optimal moisture content appeared to be 75-85%. In general, high moisture content can provide adequate water for supporting vigorous fungal growth and active metabolic function. However, too much moisture reduces interparticle spaces and substrate porosity in solid state fermentation, which in turn decreases oxygen diffusion and inhibits aerobic cultures (Singhania et al., 2009) Particle size Particle size of the substrate is also a major factor affecting the performance of solid state fermentation. Large particle size can hamper accessibility of nutrients by fungi and also prevent the penetration of air and metabolite intermediates into the particles. However, the reduced particle size may adversely affect interparticle aeration. During the cultivation of Phlebia tremellosa with active aeration, hammer-milled aspen wood (150µm~2mm) showed the comparable delignification rate as finer particles (10 and 40 mesh), which was significantly faster than for larger sizes like chips ( mm ) (Reid, 1989b). Membrillo et al. (2008) reported that among three particle sizes tested (0.92, 1.68, and 2.9 mm), the 2.9 mm particle size led to the highest level of lignocellulytic enzymes. Zadrazil and Puniya (1995) tested fungal pretreatment of four 14

33 fractions (<1 mm, 1-3 mm, 3-5 mm, and 5-10 mm) of sugarcane bagasse with Pleurotus erynii and found that the fungal degradation rate was dependent on the particle size of the substrate. Sarikaya and Ladisch (1999) reported that too fine a particle size (<0.42mm) had less lignin degradation compared to that of mm during 32-d treatment of rapeseed with P. ostreatus Supplements Supplementation of nutrients and inducers can potentially improve fungal performance on substrate colonization and lignin degradation. Inducers, such as Mn 2+, H 2 O 2, veratryl alcohol, and aromatic compounds, have been shown to stimulate oxidative enzyme production and lignin degradation (Shrestha et al., 2008). Some studies, however, showed the addition of salts with or without Mn 2+ did not significantly improve lignin degradation (Shi et al., 2008). Buffer addition, especially for a liquid culture, is important to fungal growth and enzyme production (Michel et al., 1990). The addition of nutrients facilitates white rot fungi to colonize the deeper areas within the wood and also increases the fungal biomass at the surface of the wood. Corn steep liquor is a low cost carbon/nitrogen source, which has been used for biopulping to increase fungal biomass production as well as for mechanical pulping to reduce energy consumption (Messner et al., 1998). For some white rot fungi, the depolymerization of lignin is induced under nitrogen starvation. Ruttimann-Johnson et al. (1993) reported that nitrogen-sufficient media did not necessarily improve the delignification process of C. subvermispora and 15

34 the carbon source, rather than the nitrogen source, was the limiting factor of lignin mineralization of this fungus Temperature In general, white rot fungi (basidiomycetes) can grow well between 15 and 35 ºC but high delignification rate is generally obtained within an optimal temperature range between 25 and 30 ºC (Reid, 1985; Reid, 1989a). The metabolism of white rot fungus generates heat and causes temperature gradients in solid state cultivation. The accumulated heat can kill or inhibit the fungal growth and metabolism. Therefore, in the scale-up of solid state cultivation, heat dissipation is one of the key factors that needs to be taken into account in the bioreactor design Aeration Aeration is known to markedly affect solid-state fermentation. Since lignin degradation is an oxidative process, oxygen availability is important for ligninolytic enzyme activity of white rot fungi. As reviewed by Reid (1989a), oxygen could increase the delignification rate but it did not necessarily increase delignification selectivity. In flask reactors (< 500 ml), passive air diffusion through cotton plugs is as sufficient as forced air circulation for the delignification process (Reid, 1989a). However, for reactors containing packed feedstock, active aeration is necessary to provide uniform air diffusion throughout the substrate. Messner et al. (1998) tested air aeration (0.001, 0.022, 0.1 v v -1 min -1 ) for P. chrysosporium treatment of aspen chips for biopulping and concluded that a medium flow rate appeared to be enough to achieve good fungal performance. Hatakka 16

35 (1983) reported that flushing with oxygen 3 times per week shortened pretreatment time by approximately 1 week Decontamination Decontamination of feedstocks (e.g., gas, steam, chemicals) can effectively kill or inhibit indigenous microorganism in the feedstock and is generally required prior to fungal pretreatment, especially with basidiomycetes. P. chrysosporium was found to be efficiently competitive against fungal and bacterial infection while C. subvermispora required a higher degree of asepsis (Messer and Srebotnik et al., 1994). Akin et al. (1995) tested the influence of contamination on activity of two white rot fungi (C. subvermispora and C. steroreus). The results showed that abundant bacteria and unknown fungi were prevalent on both control and contaminated Bermuda grass stems. However, contamination did not affect either fungal performance or the resulting digestibility of contaminated and uncontaminated substrates. It indicated that activity of white rot fungi were not suppressed by fungal and bacterial infection. When fungal pretreatment is scaled up to larger reactors, decontamination is one of the major costs for the solid state fermentation. In the study of Akhtar et al. (1998), it was found that complete sterilization was not necessary and short atmospheric steaming (~15 seconds) was sufficient to allow white rot fungi to outcompete indigenous fungi. Instead of decontamination by atmospheric steam or autoclave, inexpensive chemicals such as sodium biosulfite, sodium meta-bisulfite, and sodium hydrosulfite, were reported to be effective at concentrations as low as 500 ppm (Akhtar et al., 1998). 17

36 Time Long pretreatment time due to low delignification rates is one of the major barriers to microbial pretreatment. Generally, several weeks to months are needed to obtain a high degree of lignin degradation. Locci et al. (2008) observed a decrease in carbon content with a concomitant increase in aliphatic and carboxylic content in the substrate after 62 d of fungal treatment by P. ostreatus. Of the four fungi, P. chrysosporium, T. versicolor, P. ostreatus, and C. subvermispora, studied by Itoh et al. (2003), lignin degradation of beech wood for 4 weeks was 21.7, 17.7, 10.3, and 13.0%, respectively. Lindenfelser et al. (1979) found that P. ostreatus degraded 51% of wheat straw lignin after 90 d of incubation. It was observed that after 72 d of cultivation by P. ostreatus, 32% of the carbohydrates in rice straw were converted to soluble sugars (Taniguchi et al., 2005). Pretreatment time of d were required for fungal pretreatment of bamboo to obtain significant improvement of sugar yield (Zhang et al., 2007a; Zhang et al., 2007b). P. chrysosporium grows fast but no enhancement of saccharification rate was observed due to simultaneous degradation of holocellulose and lignin by this fungus on various feedstocks (Keller et al., 2003; Sawada et al, 1995; Shi et al., 2009) Degrading enzymes Ligninolytic enzymes It is well-known that lignin peroxidase (LiP), manganese peroxidase (MnP), and laccase are three major oxidative enzymes secreted by white rot fungi (Eriksson et al., 18

37 1990). They are responsible for oxidation of lignin and a wide range of lignin analogous compounds (Winquist et al., 2008). However, not all these enzymes are detected from a specific fungal culture. For example, P. chrysoporium produces LiP and MnP, but not laccase (Ruttimann-Johnson et al., 1993). C. subvermispora only produces MnP and laccase whereas LiP has not been detected, although a LiP-like gene was revealed in this fungus by a southern-blot hybridization technique using the same probe coding as LiP (Rajakumar et al., 1996). Although ligninolytic enzymes may be non-specific on different substrate sources, they may have similar degradation reactions. For example, both LiP and MnP were found to degrade nonphenolic lignin by one-electron oxidation of the aromatic ring (Kirk et al., 1986; Srebotnik et al., 1997). However, the role of ligninolytic enzymes on the early stage decay of biomass feedstock still remains unclear. A study conducted by Guerra et al. (2002) indicated that there was no evidence for a correlation between oxidative enzymes and lignin degradation. Moreover, the mineralization of lignin corresponding to lignin degradation occurred after the lignin modification, which was indicative of a serial reaction involved in lignin depolymerization. Studies in which an immunocytochemical technique was used revealed that lignin-degrading enzymes did not diffuse into sound and unaltered cell walls due to their large molecular weight (Blanchette et al., 1997; Srebotnik et al., 1988; Flournoy et al., 1991). However, the oxidative enzymes could be active at the surface of the cell wall and induce the formation of low molecular mass agents such as radicals and oxalic acids (Kapich at al., 1999; Enoki et al., 1999). These low molecular compounds are diffusible and have been suggested as a mechanism to initiate wood decay and thus facilitate the 19

38 penetration of lignin-degrading enzymes (Galkin et al., 1998). Liquid or solid culture of C. subvermispora was reported to produce a variety of low molecular compounds including glyoxylic acids (Urzua et al., 1998), oxalic acid (Galkin et al., 1998; Urzua et al., 1998), and several unsaturated fatty acids (Enoki et al., 1999; Gutierrez et al., 2002). Radical agents (e.g., peroxyl and acyl radicals), generated from MnP-dependent lipid peroxidation reaction of white rot fungi, have been suggested as playing an important role in non-phenolic lignin oxidation (Kapich et al., 1999; Watanabe et al., 2000) Hydrolytic enzymes Hydrolytic enzymes play important roles in fungal growth and metabolism by providing an easily digestible carbon source. However, non-selective white rot fungi, due to their high cellulytic activity, cause substantial cellulose loss. In contrast, selective white rot fungi mainly secrete hemicellulytic enzymes and utilize hemicellulose-derived sugars as the main carbons for self-growth and metabolism. Due to low permeability of sound wood cells, it is difficult for such large molecule enzymes to penetrate. Similar to ligninolytic enzymes, low molecular mass agents have also been suggested as a mechanism to induce cellulose depolymerization. The Fenton system (Fe 2+ and H 2 O 2 ), known to depolymerize cellulose, can be generated by the action of some enzymes such as cellobiohydranase and MnP in white rot fungal cultures (Dumonceaux et al., 2001; Henriksson et al., 2000; Xu et al., 2009). In contrast to cellulase and peroxidase, hemicellulases, such as xylanase, did not act in the synergism of low molecular compounds (Ferraz et al., 2003). Therefore, diffusion of xylanase into the plant cell wall 20

39 may be facilitated by lignin degradation which increases cell wall permeability (Machuca and Ferraz, 2001; Vicentim and Ferraz, 2007). This suggestion may be supported by a similar pattern of hemicellulose loss to lignin degradation observed from the solid culture of C. subvermispora (Aguiar et al., 2006; Machuca and Ferraz, 2001). Due to its lack of a complete cellulytic enzyme complex, C. subvermispora has become well-known as a selective white rot fungus that preserves most cellulose during fungal decay (Ferraz et al., 2003). In the study of Ferraz et al. (2003), no significant glucose loss was detected from C. subvermispora culture on the wood species E. grandis until 60 d. Maximum glucan loss of 7.3% was observed after 90 d. Low cellulase activity revealed by filter paper activity explained the low glucan loss caused by this fungus. Another wood species P. taeda cultured with this fungus was observed with only 2% glucan loss after 90 d (Guerra et al., 2003). Cellulose degradation up to 26% was observed during 200 d of degradation of Pinus radiate by C. subvermispora, probably due to slowly induced cellulytic activity (Ferraz et al., 2001). The study reported by Sethuraman et al. (1998) showed that the expression of cellulolytic activities of C. subvermispora strongly depended on carbon source. The authors detected a considerable amount of endo-1,4-β-glucanase in culture solutions of avicel, cellulose filter paper, cellobiose, flax, or pectin while 1,4- β-glucosidase in most of culture carbon solutions. Interestingly, no significant amount of exo-1,4- β-glucosidase activity was found in any carbon solution. 21

40 2.5. Enzymatic hydrolysis and ethanol fermentation Non-selective degradation Recently, fungal pretreatment by white rot fungi has received renewed interest as an alternative pretreatment method to thermal/chemical pretreatment for enzymatic hydrolysis and cellulosic ethanol production. As summarized in Table 2.1, various biomass feedstocks, including agricultural residues and woody biomass, have been studied for their cellulose digestibility following fungal pretreatment with different white rot species. P. chrysosporium, well known to be a non-selective lignin-degrading fungus, had little or no effect on improvement of enzymatic hydrolysis. The fungus itself consumed a large amount of readily accessible carbohydrates due to the simultaneous degradation of holocellulose and lignin (Bak et al., 2009). The remaining cellulose might be less digestible and thus resistant to the subsequent hydrolysis. Moreover, reduced cellulose content after fungal pretreatment also contributes to the reduced glucose yield. The longer time the biomass feedstock is pretreated by P. chrysosporium, the lower the expected saccharification yield compared to that from the non-treated feedstock. This expectation was observed during a 100-d pretreatment of aspen wood by this fungus, where the maximum saccharification yield was reached in 28 d and thereafter the saccharification yield decreased (Sawada et al., 1995). The lignin degradation was increased from 42% to 50% as pretreatment time increased from 28 to 100 d while the corresponding holocellulose loss was increased from 17% to 50%. Although the degradation selectivity was high on day 28, the fungal pretreatment was not sufficient to increase the saccharification yield of aspen wood. Similar results were reported by 22

41 Shrestha et al. (2008), who observed that pretreatment time beyond 2 weeks resulted in a reduced saccharification yield compared to that of the control. In their study, 40% lignin degradation may not be effective to reduce the recalcitrance of corn fiber as lignin content in the raw fiber was only 2%. Keller et al. (2003) reported that P. chrysosporium pretreatment of corn stover for 29 d did not significantly increase saccharification yield compared to controls, probably due to too long a pretreatment time. Even for a 14-d fungal pretreatment of cotton stalk by P. chrysosporium, a reduced saccharification yield was obtained with both submerged and solid-state fungal-treated cotton stalks (Shi et al., 2009). The cellulose loss from both methods was high, which was up to 40.49% loss, while lignin degradation was in the range of %. Hot water washing, for removal of fungal biomass/protein and lignin derivatives, improved the saccharification yield of solid-state treated stalks but not the submerged-treated stalks. Nevertheless, the saccharification yield was not significantly different from that of non-treated. There was an exception for P.chrysosporium with respect to the degree to which saccharification yield was improved. Bak et al. (2009) serially optimized the culture media for submerged pretreatment of rice straw with P. chrysosporium in order to achieve high ligninolytic enzymes. It was found that glucose yield was improved by at least 2 times during a 30-d pretreatment, reaching a maximum glucose yield of 64.9% (equivalent to 50% of glucose yield of the raw feedstock). The maximum ethanol yield of 63% was obtained from cellulose in treated rice straw, which was about 2 times that of the untreated straw. However, compared with solid state fungal pretreatment, limitations of submerged fungal pretreatment were apparent due to a considerably low substrate 23

42 loading (2.5% in their case) as well as other issues associated with submerged cultivation. Bak et al. (2009) also found that the saccharification yield of rice straw was decreased with the extension of P. chrysosporium pretreatment time from 15 to 30 d. This result is mainly caused by substantial cellulose degradation during fungal pretreatment. It was observed that cellulose content in rice straw decreased from 38.0% to 21.9% after 30 d of fungal pretreatment while it decreased from 19.3% to 13% for lignin and from 10.0% to 7.3% for xylan. Considering the considerable cellulose loss with long pretreatment times and the limitations of submerged cultivation, such non-selective fungus may not be practically used to improve the saccharification yield of biomass feedstocks Selective degradation Selective lignin-degrading white rot fungi preferentially partially remove lignin and keep the majority of carbohydrates. Due to native recalcitrance of biomass feedstocks, fungal performance on degradation and the resulting digestibility varies with different feedstocks. For agricultural residue, Keller et al. (2003) reported that the saccharification yield of 36% was obtained from corn stover pretreated with Cyathus stercoreus for 29 d, which was about 4 times that of the untreated. In contrast, the saccharification yield of 66.4% was obtained with 25-d Irpex lacteus CD2 treated corn stover (Xu et al., 2010). In this study, the saccharification yield of fungal-treated corn stover decreased slightly during d of pretreatment and then decreased dramatically due to holocellulose loss. It was found that holocellulose degradation, especially hemicellulose degradation, was dominant during the early stage (0-5 d) while no lignin 24

43 degradation was observed during this period. Thereafter, active lignin degrading was observed on day 5 to 10 with a rate higher than that of holocellulose degradation. These studies indicated that delignification of corn stover was fungus-specific. P. ostreatus was reported to be more effective with straw materials than other fungi (Taniguchi et al., 2005). Wheat straw pretreated with this fungus converted up to 27-33% cellulose (Hatakka, 1983; Taniguchi et al., 2005). It should be noted that this fungus was less selective in lignin and cellulose degradation when pretreatment time was longer than several weeks. Thus, the cellulose digestibility tended to level off during the later stage of cultivation (Taniguchi et al., 2005; Yu et al., 2008). Fungal pretreatment also greatly improved the digestibility of hardwoods such as aspen and birch. A desirable polysaccharide digestibility of aspen wood (50-55%) was obtained after pretreatment with Polyporus giganteus, Polyporus berkeleyi, or Polyporus resinosus for d while the digestibility of birch wood was not substantially improved with these fungi (Kirk and Moore, 1972). For Chinese willow, about 37% percent of the polysaccharides were converted after pretreatment with Echinodontium taxodii 2538 for 120 d. Although the digestibility of pretreated conifer wood (softwood) was not higher, the improvement by fungal pretreatment was obvious and saccharification yield was generally several times higher than that of non-treated. The saccharification yield of 37% was observed with bamboo residues treated with E. taxodii 2538 for 120 d, which was 8.76 fold of the yield of the untreated (Zhang et al., 2007b). However, the saccharification yield of China fir was only 17% after the same pretreatment condition as the Chinese willow (Yu et al., 25

44 2009). Lee et al. (2007) also reported a 21% reduced sugar yield from Japanese red pine pretreated with Stereum hirsutum for 8 weeks, of which 14% was from glucose Combination of microbial pretreatment with other pretreatment methods Microbial pretreatment of wood with white rote fungi has been widely applied in the pulp industry. This so-called biopulping process can potentially overcome the problems associated with mechanical and chemical pulping methods. The advantages include significant energy savings in mechanical pulping, low severity for subsequent chemical pulping, improved paper strength properties, and reduced toxicity of pulping waste (Akhtar et al., 1998; Kang et al., 2003). Messner et al. (1998) observed that a relatively short incubation time substantially favored the chemical pulping process due to appreciable pre-modification of lignin. Similar to benefits of biopulping, fungal pretreatment combined with mild mechanical or physical pretreatment has been of interest to improve degradation and the subsequent digestibility of lignocellulosic biomass. Fungal pretreatment followed by physical/chemical pretreatment is summarized in Table 2.2. Similar to biopulping, fungal pretreatment improved the performance of the subsequent non-fungal pretreatment. Yu et al. (2010) reported that the 15-d fungal pretreatment of corn stalks with I. lacteus modified the lignin structure and significantly facilitated lignin degradation and xylan removal during mild alkaline pretreatment (1.5% NaOH, ºC for min). The synergic effect largely depended on the severity of alkaline pretreatment. The less the severity of the alkaline pretreatment, the more the 26

45 cellulose digestibility was improved by fungal pretreatment. In other words, the fungal pretreatment reduced the required severity of alkaline pretreatment. For example, an 80% glucose yield was obtained with raw corn stalk pretreated by alkaline solution at 60 C for 120 min, while, with fungal pretreatment, the similar yield was obtained by less severe alkaline pretreatment conditions (60 ºC for 30 min or 30 ºC for 120 min). Similar results were also reported by Ma et al. (2010), who observed that acid pretreatment (0.25% H 2 SO 4, ºC for min) following fungal pretreatment improved enzymatic hydrolysis and ethanol yields of water hyacinth by 1-2 times over acid pretreatment alone. Ethanol yields of woody biomass such as Japanese cedar and beech wood resulting from combined C. subvermispora pretreatment and ethanolysis were much higher than that without fungal pretreatment (Baba et al., 2010). Fungal pretreatment of beech wood meal with P. chrysosporium and then steam explosion also increased saccharification yield compared to a single pretreatment (Sawada et al., 1995). However, the effect of combined pretreatment depended on the operational conditions of steam explosion. A severe steam explosion at a higher pretreatment temperature and time could cause condensation of Klason lignin with carbohydrate oligomers and methanol soluble lignin (Sawada et al., 1995). As a result, these newly formed compounds can lower the susceptibility of the co-treated substrate to hydrolytic enzymes. Therefore, regardless of chemical or physical pretreatment following fungal pretreatment, a moderate subsequent pretreatment method is suggested to achieve a more synergic effect exerted by fungal pretreatment. 27

46 Fungal pretreatment as the second step of combined pretreatment is also summarized in Table 2.2. One physical pretreatment method, ultrasonic pretreatment on lignocellulose modification, has been proposed to degrade α-o-4 and β-o-4 linkages in lignin (Seino et al., 2001), oxidize hydroxyl groups by radicals and H 2 O 2 formed during ultrasonic cavitation (Tan et al., 1985), and increase fiber wall porosity (Laine et al., 1977). The synergistic effect of ultrasonic pretreatment on the subsequent fungal pretreatment was evident. As reported in the study of Kadimaliev et al. (2003), prior ultrasonic pretreatment (22 khz, 10 min) improved fungal delignification of beech and pine sawdust with Panus (Lentinus) tigrinus. Yu et al. (2008) combined more powerful ultrasonic pretreatment (250 w, 40 khz, 30 min) with fungal pretreatment and found that during 18-d fungal pretreatment with P. ostreatus, lignin degradation on ultrasonicmodified rice hull was much higher than that of raw hull. The substrate resulting from combined pretreatment had a 20% higher glucose yield than the sole fungal-pretreated. Moreover, this glucose yield was comparable to that obtained from the substrate fungalpretreated for 42 d, which indicated that pre-modification of the feedstock by ultrasound led to a shortened fungal pretreatment time. In the same study, H 2 O 2 pretreatment (2% w/v, 48 hours) performed better on the following fungal delignification than ultrasound. In contrast to the effectiveness of mild chemical pretreatment, severe prior chemical pretreatment could mask the effect of fungal pretreatment and result in no or less improvement on lignin degradation and saccharification of the combined-pretreated substrate. A similar study conducted by Kadimaliev et al. (2003) showed that ammonia (5%, 165 ºC for 10 min) and sulfuric acid (2.5%, 165 ºC for 10 min) resulted in negative 28

47 effects on lignin degradation of both beech and pine sawdust substrates when applied before fungal pretreatment. This problem was presumed to be due to partial degradation of polysaccharides caused by chemical pretreatment, which provided a more easily accessible compounds for the fungal growth; thus, more carbohydrates were consumed during fungal pretreatment while fungal lignin degradation activity was probably also depressed. Similarly, prior NaOH pretreatment (0.4g NaOH/g straw, 115 ºC for 10 min) did not improve sugar yields of fungal-pretreated wheat straw (Hatakka, 1983) due to degradation caused by severe alkaline pretreatment. Instead, the sugar yield from the sole fungal pretreatment was comparable to that from the alkaline pretreatment. Therefore, it can be stated that the severity of physical or chemical pretreatment strongly affects the performance of fungal pretreatment and a mild prior pretreatment resulting in sufficient modification on biomass could favor fungal pretreatment Enzymatic treatment As mentioned in the above section, ligninolytic enzymes, mainly lignin peroxidase (LiP), manganese pexidase (MnP), and laccase, are responsible for delignification by white rot fungi. However, the slow fungal growth (several weeks to months) and holocellulose (cellulose and hemicelluloses) loss are the major problems related to fungal pretreatment. The use of ligninolytic enzymes for direct treatment of lignocellulose, only taking hours, has the potential to overcome the problems of fungal pretreatment. 29

48 Ligninolytic enzymes were shown to cleave lignin compound model oxidatively in vitro (Tuor et al., 1992). The oxidation by peroxidase is highly dependent on oxygen as the final electron acceptor (Ibarra et al., 2006) and H 2 O 2 is usually used as a source of O 2 (Khazaal et al., 1993). For oxidation of lignin by laccase, chemical mediators, such as 1- hydroxybenzotrizole (HBT) and N-hydroxy-N-phenylacetamide (NHA), are extensively used for radical formation during enzyme oxidation ( Palonen and Viikari, 2004). These radicals can react with aromatic acid compounds of lignin. Therefore, the ligninolytic enzyme-based treatment can be considered as an enzyme-catalyzed oxidative treatment (Ibarra et al., 2006). Temperature, ph and enzyme concentration are important factors for enzyme treatment. Ramos et al. (2004) tested the effect of treatment of defibrated sugarcane bagasse with crude enzyme extract from P. chrysosporium on the production of mechanical pulps. It was found that 36-hour enzymatic treatment with the addition of H 2 O 2 resulted in higher pulp yield than the 2-week fungal pretreatment. The treatment with enzymatic extract isolated from white-rot fungi also improved in vitro digestibility of wheat straw as reported by Rodrigues et al. (2008). However, this effect seemed strongly dependent on synergistic effects between extracted enzymes. For enzymatic treatment of diluted acid-pretreated wheat straw, laccase had a synergistic effect on cellulase for enhancing glucose yield but was less effective than xylanse and feruloyl esterase (Tabka et al., 2006). A maximal increase of 21% in the saccharification rate of steam-treated softwood was obtained from sequential combination of laccase-mediator treatment and commercial cellulase hydrolysis (Palonen and Viikari, 2004). 30

49 White rot fungi may serve as a good producer of extracellular enzymes including oxidative enzymes and polysaccharides-degrading enzymes. The isolated enzyme complex has potential to replace fungal pretreatment to avoid long pretreatment and the possible concomitant degradation of lignin and holocellulose on lignocellulosic biomass. However, enzymatic treatment is immature and the efficiency of oxidative enzymes largely relies on chemical mediators. On the other hand, the fungal-pretreated substrate containing the enzyme complex could be directly used for the subsequent digestion and fermentation if it acts in the synergism between commercial enzymes Modeling and scale up Solid state fermentation involves the degradation of the substrate taking place in absence (or near absence) of free water, due to release of extracellular enzymes or cell bound enzymes to the external environment. Both interparticle and intraparticle mass transfers occur in solid state fermentation. In aerobic digestion, the transfer of oxygen to the growing microorganisms is the major interparticle mass transfer. The intraparticle mass transfer refers to the transfer of nutrients and enzymes within the substrate solids. Modeling of microscale phenomena, such as the microorganism growth behavior, mass diffusion, and particle size reductions, can describe how microscale processes influence growth kinetics of microorganisms. The growth kinetics models, such as linear, exponential, logistic, and Monod equation, have been used to describe microorganism growth. The logistic model, which is most commonly used, assumes the growth rate (µ) is independent of substrate concentration (Saucedo-Castaneda et al., 1990). Some studies 31

50 proposed to express the growth rate as a function of temperature, which varies during growth (Fanaei and Vaziri, 2009; Saucedo-Castaneda et al., 1990). Due to the difficulty in separating the filamentous fungal biomass from the substrate, various indirect measurements, such as dry mass loss, oxygen consumption/co 2 production, and cell specific compounds, have been proposed for fungal biomass predications (Mitchell et al., 2004). Cell specific compounds such as glutamate or ergosterol content have been better correlated to dry weight of fungal biomass compared to other kinds of indirect measurements. The determination of white rot fungi biomass is commonly based on ergosterol content (Joergensen, 2000). Due to lack of free water and low conductivity of solid particles, heat generation related to metabolic activities of the microorganism growth causes temperature gradients in solid state fermentation; thus, modeling of heat removal is the major consideration in the design of bioreactors. Various bioreactors have developed for solid state fermentation, including tray reactors, packed-bed, rotating drums, and stirred bioreactors (Mitchell et al., 2006). Although mixing can improve heat dissipation, it is only useful for solid state fermentation in which the fungus does not bind the solid particles together. For fungal fermentation with the inter-particle hyphal bridges across the substrate, mixing is deleterious to bioreactor performance because of disruption of the hyphae between particles and the shear force resulting from mixing (Fanaei and Vaziri, 2009; Mitchell et al., 2000). Instead, static operation is preferred for binding the substrate bed with fungi that does not tolerate mixing. Two types of reactors, tray and packed-bed bioreactors, are commonly used for this operation. The tray reactors are simple but suitable for low 32

51 volume production due to a limited loading capacity. In packed-bed reactors, modeling is mostly focused on temperature gradients and heat transfer while neglecting oxygen transfer because heat transfer has a greater impact on microorganism growth than oxygen transfer (Mitchell et al., 2000). Heat transfer with a water jacket is impractical for a large bioreactor as it requires a large diameter water jacket. Instead, convective heat transfer, such as forced humid air, is more practical to remove the heat and also prevents the substrate bed from drying out. The traditional packed-bed bioreactor has the problem of axial temperature gradients. It causes significant water evaporation and high temperature at the outlet of the bed even if saturated air is supplied. In contrast, the Zymoties packedbed bioreactor has internal heat transfer plates, which has advantages over traditional packed-bed bioreactors, including a decrease in axial temperature gradients and evaporation rates (Mitchell et al., 2006). However, the substrate loading and unloading is difficult in the Zymotis bioreactor due to the presence of the heat transfer plates. Also, the water condensation on the exposed surface limits oxygen on the top of the bed. Scott et al. (1998) tested both the tubular reactor and chip piles for the scale-up of fungal pretreatment of wood chips for biopulping. The tubular reactors such as PVC tubes and silos were used as packed-bed reactors. The larger scale-up trial of up to 40 tons was done by an outdoor wood chip pile; however, the heat gradient was still a problem. Compared to the tubular reactors, the chip piles are more difficult to control due to two or three dimensional airflow. In addition, the surface of the chip piles is more exposed to undesirable fungal species, and air flow and weather changes. Less controlled 33

52 chip pile could also lead to reduction and variation of pulp quality due to more degradation of cellulose which was caused by undesirable fungi and bacteria Concluding remarks White rot fungi can degrade lignin and a wide range of analogous compounds through lignin-degrading enzymes; thus, they have been studied for a wide range of applications including bioremediation of toxic compounds in soils and waste waters, biopulping, and forage upgrading. Recently, white rot fungi are of interest for their delignification of agricultural and forestry residues for ethanol production. Compared to current leading thermal or chemical pretreatment processes, microbial pretreatment with white rot fungi by solid-state cultivation is an environmental-friendly, energy efficient and inhibitor-free process to depolymerize lignocellulosic biomass and subsequently improve polysaccharide digestibility. White rot fungi with a high selectivity of lignin degradation over cellulose loss are important for fungal pretreatment because conversion of biomass feedstocks is highly dependent on the degree to which lignin is removed and cellulose is preserved. C. subvermispora is well-known for its high degradation selectivity in lignocellulosic biomass. Prior studies of this fungus mostly focused on the biopulping industry, applying this fungus for wood pretreatment, and have shown its superiority over some other fungi. Therefore, C. subvermispora can be a good candidate for solid-state fungal pretreatment of lignocellulosic biomass for ethanol production. The cultivation parameters, such as moisture and particle size of the substrate, supplements, and aeration and pretreatment 34

53 time, are critical for fungal growth and metabolism. Complete decontamination may not be necessary since white rot fungi survive in contamination and actively act on degradation. Therefore, decontamination and its associated costs can be largely saved. Modeling and bioreactor design in regards to solid state fermentation have been studied for process optimization and scale-up of biobased chemical or enzyme production. However, these studies lack simulation of the delignification process for lignocellulosic biomass. The feedstock itself essentially affects fungal performance due to specific lignin subunits and cross-linking between lignin and hemicelluloses. The long pretreatment times, considered the major limitation of fungal pretreatment, are largely due to specific fungal-substrate interactions. Currently, it is not feasible to apply lignin-degrading enzymes isolated from white rot fungi for effective degradation of lignocellulosic biomass due to various problems associated with in vitro oxidative-enzymatic pretreatment. Therefore, fungal pretreatment is still the dominant technology for biodelignification regardless of its long pretreatment time. Digestibility resulting from fungal pretreatment is closely related to degradation pattern. Non-selective degradation is not desirable with respect to cellulose loss during pretreatment, resulting in no or less improvement on saccharification yield. The selective degradation has resulted in greatly improved saccharification yields that are comparable to that of the current leading pretreatment methods. However, hemicellulose loss is inevitable for selective degradation since no fungus can grow in lignin without a carbon source. 35

54 In order to further improve the delignification process and the resulting digestibility, combined fungal pretreatment with mild physical or chemical pretreatment has been proposed. Prior fungal pretreatment for improvement of cellulose digestibility has shown synergism with advantages similar to that of biopulping process. It can also increase the robustness of fungal pretreatment to handle a wide range of feedstock sources. On the other hand, fungal pretreatment alone can be applied to on-farm wet storage by taking advantage of long pretreatment times while improving cellulose digestibility. In conclusion, solid state fungal-pretreatment with or without involvement of mild thermo-chemical pretreatment has great potential to be an alternative to thermochemical pretreatment. 36

55 Table 2.1. The effect of white rot fungi on enzymatic hydrolysis and ethanol production Fungus Substrate Sugar/ethanol yield * Reference Phaerochaete chrysosporium Cotton stalk Reduced glucose yield Shi et al., 2009 Phaerochaete chrysosporium Corn stover No improvement on Keller et al., 2003 glucose yield Phaerochaete chrysosporium Rice straw 50% glucose yield Bak et al., % ethanol yield Phaerochaete chrysosporium Beechwood 9.5% total sugar yield a Sawada et al., 1993 Phaerochaete chrysosporium Corn fiber Reduced sugar yield or no Shrestha et al., 2008 significant improvement Pleurotus ostreatus Rice straw 33% glucose yield Taniguchi et al., 2005 Pleurotus ostreatus Rice hull 38.9% glucose yield Yu et al., 2008 Pleurotus ostreatus, Wheat straw 27-28% glucose yield b Hatakka, 1983 Pycnoporus cinnabarinus 115 Euc-1 Wheat straw 22.5% total sugar yield a Dias et al., 2010 Cyathus stercoeus Corn stover 36% glucose yield Keller et al., 2003 Irpex lactues Corn stover 66.4% total sugar yield c Xu et al., 2010 Pheblia tremellosus Aspenwood 11.6% glucose yield a Mes-Hartree et al.,1987 Polyporus giganteus Aspenwood 55.2% glucose yield Kirk and Moore, 1972 Stereum hirsutum Echinodontium taxodii Echinodontium taxodii 2538, Coriolus versicolor Japanese red pine 13.56% glucose yield a Lee et al., 2007 Chinese willow, 5-35% glucose yield Yu et al., 2009 China-fir Bamboo culm 37% total sugar yield d Zhang et al., 2007a, 2007b results were obtained from submerged fungal pretreatment. * % of theoretical yield of glucan in the original material, unless stated otherwise. a % of dry mass of the treated material. b % of dry mass of the original material. c % of theoretical yield of holocellulose in the original material./ d % of theoretical yield of holocellulose in the treated material. 37

56 Table 2.2. Combined white-rot fungal pretreatment and physical/chemical pretreatment Fungal pretreatment Prior a Physical /chemical pretreatment Alkaline (NaOH) Substrate Effectiveness References Corn stalk Prior Diluted acid Water hyacinth Prior Ethanolysis Beech wood chips Prior Ethanolysis Japanese cedar wood Prior Post b Steam explosion Alkaline (NaOH) Beech wood meal Wheat straw Fungal pretreatment improved delignification and xylose removal during mild alkaline pretreatment Combined pretreatment increased sugar and ethanol yields by 1-2 folds over a single acid pretreatment Combined pretreatment saved 15% electricity and increased the ethanol yield by up to 1.6 times over ethanolysis alone. Combined pretreatment increased the sugar yield by 7 times over ethanolysis alone Combined pretreatment improved overall sugar yield Strong alkaline pretreatment masked the synergistic effect of fungal pretreatment on the combined process Post H 2 O 2 Rice hull H 2 O 2 enhanced fungal delignification, resulting in a sugar yield that is comparable to that obtained from a long-term sole fungal pretreatment. Post Post Acid (Sulfuric acid) Alkaline (Ammonia) Beechwood and pine sawdust Beechwood and pine sawdust Post Ultrasound Beechwood and pine sawdust Acid pretreatment reduced lignin degradation on fungal pretreatment Alkaline pretreatment reduced lignin degradation on fungal pretreatment Ultrasound accelerated fungal lignin degradation Post Ultrasound Rice hull Ultrasound accelerated fungal delignification while slightly increased cellulose and hemicellulose loss, resulting in a higher sugar yield compared to the fungal pretreatment alone. Yu et al., 2010 Ma et al., 2010 Itoh et al., 2003 Baba et al., 2010 Sawada et al., 1995 Hatakka, 1983 Yu et al., 2008 Kadimaliev et al., 2003 Kadimaliev et al., 2003 Kadimaliev et al., 2003 Yu et al., 2008 a Prior indicates that fungal pretreatment is the first step of combined pretreatment. b Post indicates that fungal pretreatment is the subsequent step of combined pretreatment. 38

57 Chapter 3 Effect of Pretreatment Parameters on Fungal Pretreatment of Corn Stover The feasibility of concurrent wet storage and microbial pretreatment of corn stover with C. subvermispora for ethanol production was investigated in this study. The effects of particle size (5-15 mm), moisture content (45-85%), pretreatment time (18-35 d), and temperature (4-37 ºC) on lignin degradation and enzymatic hydrolysis yield were studied. The results showed that C. subvermispora selectively degraded lignin up to 31.59% but cellulose degradation was less than 5% during an 18-d pretreatment. When 5 mm corn stover was pretreated at 28 ºC with 75% moisture content, overall glucose yields of 57.67%, 62.21% and 66.61% were obtained with 18-, 28-, and 35-d microbial pretreated corn stover, respectively. For the above condition, the highest ethanol yield of 57.80% was obtained with 35-d pretreated corn stover. A higher glucose yield of about 72% was obtained when the enzymatic hydrolysis of the 18 d-treated corn stover was supplemented with an accessory xylanase/cellulase enzyme complex Introduction White rot fungi, which are the most effective lignin-degrading microorganisms, have been receiving extensive attention for biodelignification of lignocellulosic biomass. Taniguchi et al. (2005) found that 33% of cellulose in wheat straw was converted to glucose after a 60-d pretreatment with P. ostreatus. Cellulose conversion of 36% was 39

58 obtained from corn stover pretreated with C. stercoreus for 29 d (Keller et al., 2003). Combined chemical and P. ostreatus pretreatment of rice hulls resulted in a maximum sugar yield of 49.6% (Yu et al., 2008). Zhang et al. (2007a) obtained a maximum saccharification yield of 37% from bamboo residues pretreated by Coriolus versicolor. To study the feasibility of applying microbial pretreatment of lignocellulosic biomass by C. subvermispora to wet storage, the effects of moisture content, particle size, cultivation time, and temperature on lignin degradation, enzymatic digestibility, and ethanol yield of corn stover were investigated. The effects of washing and no-washing of fungal-pretreated corn stover on ethanol yield were also compared to examine the potential fermentation inhibitors resulting from fungal pretreatment Materials and Methods Materials Corn stover (shredded and baled) was harvested from the Ohio Agricultural Research and Development Center (OARDC) farm. The composition of corn stover was 38.42% cellulose, 22.95% hemicellulose, 20.18% lignin, and 3.82% ash, of which the hemicellulose was composed of % xylan, 1.60% galactan, and 2.82% arabinan, and the lignin was composed of 18.65% acid- insoluble lignin (AISL) and 1.53% acid-soluble lignin (ASL). Corn stover was dried in a 40 ºC convention oven to obtain a moisture content less than 10% and then ground to pass through 5, 10, and 15 mm screens, respectively, using a Thomas Wiley Mill. The particle size tested was simply denoted as the heterogeneous sizes of 5, 10, and 15 mm, respectively, without direct measurement of 40

59 particle size distribution within each batch. The ground corn stover was stored in air tight containers at room temperature before use. Cellulase (Spezyme CP), MUTIFACT xylanase, and Accellerase XC were obtained from Genencor (Palo Alto, CA). The activity of cellulase was 3434 carboxymethyl cellulase (CMCU)/ml (equivalent to 50 FPU /ml) and the activity of MUTIFACT xylanase was acid birch xylan units (ABXU)/ml. Accellerase XC is an accessory xylanase/cellulase enzyme complex with cellulase activity of 1201 CMCU/ml and xylanase activity of 3003 ABXU/ml Fungus and inoculum preparation The white rot fungus C. subvermispora (ATCC 96608) was obtained from American Type Culture Collection (Manassas, VA) and maintained on 2% (w/v) malt extract agar (MEA) plates at 4 ºC. The fungus was cultured on 2% MEA at 28 ºC for 5-7 d. Ten discs (10 mm in diameter) of the plate culture were inoculated into 50 ml of 2% (w/v) malt extract medium in a 500 ml cotton-plugged Erlenmeyer flask and then incubated at 28 ºC in static conditions for 7 d. The liquid culture was aseptically homogenized in a blender for three 15-second cycles and then used for inoculum Fungal pretreatment Ten grams of ground corn stover for each of the three particle sizes were placed in 250 ml Erlenmeyer flasks and then conditioned with different amounts of Deionized (DI) water to obtain moisture contents of 45%, 60%, 75%, and 85 % (wet basis), respectively. Flasks containing wet corn stover were autoclaved (121 ºC, 30 min) and cooled prior to 41

60 inoculation. A homogeneous liquid culture of 2 ml was inoculated on the top of the substrate in each flask. Pretreatment was carried out in a refrigerated incubator at 28 ºC in static conditions for 35 d. The subsamples were taken at predetermined time periods (18, 28, and 35 d) for compositional analysis. Corn stover (5 mm) with a moisture content of 75 % was pretreated by C. subvermispora at seven levels of temperature (4, 19, 24, 28, 32, 34, and 37 ºC) for 18 d. All samples were subjected to enzymatic hydrolysis to evaluate digestibility. Samples without fungal inoculation were used as controls. All tests were performed in triplicate Enzymatic hydrolysis Enzymatic hydrolysis was carried out following NREL Laboratory Analytical Procedure (LAP)-008 (Dowe and McMillan, 2001). Cellulase (Spezyme CP) was obtained from Genencor (Palo Alto, CA). Corn stover was loaded at a solid concentration of 2.5 % (w/w) in a 50 mm citrate buffer (ph 4.8) and cellulase loaded at 10 FPU/g solid. The hydrolysis was carried out at 50 ºC and 130 rpm on a rotary shaker for 72 h. Supplementation of xylanase and Accellerase XC were also tested. The hydrolyzed slurry was boiled to inactivate cellulase. After cooling to room temperature, the slurry was filtered through 0.2 µm nylon membrane filter for sugar analysis by high performance liquid chromatography (HPLC) Simultaneous saccharification and fermentation (SSF) SSF was carried out using Spezyme CP and Saccharomyces cerevisiae (ATCC ) following NREL LAP-008 (Dowe and McMillan, 2001). Corn stover (5 mm) 42

61 pretreated for 75% moisture content and 28 ºC was used for the substrate. The fungalpretreated corn stover after washing with 250 ml DI water twice was also tested. The sample and Spezyme CP were loaded at 5.0 % (w/w) and 5 FPU/g solid, respectively in a 50 mm citrate buffer (ph 4.8). S. cerevisiae stock culture was revived in YM medium (yeast extract 3 g/l, malts extract 3 g/l, peptone 3 g/l and glucose 50 g/l) at 37 ºC and 120 rpm on a rotary shaker for 12 h. The cells were harvested, washed with DI water twice by centrifugation (5000 rpm for 5 min) and then used for inoculating SSF flasks (the starting optical density was around 0.5 in a SSF flask). The flasks were equipped with water traps to vent CO 2. The SSF was conducted at 37 ºC and 120 rpm for 72 h. The samples were taken at predetermined intervals (0, 6, 12, 24, 48, and 72 h) and filtered through a 0.2 um nylon membrane filter for ethanol analysis by HPLC Analytical methods The structural carbohydrate (cellulose and hemicellulose) and lignin contents of corn stover were determined following NREL LAP (Sluiter et al., 2008) using two-step acid hydrolysis. Lignin is the sum of acid-soluble and acid-insoluble lignin. The acid insoluble lignin was measured by gravimetric analysis and the acid soluble lignin was measured by UV-Vis spectroscopy. Cellulose and hemicellulose were acid-hydrolyzed into monomeric sugars, which were measured by HPLC. The content of cellulose was calculated from glucose while hemicellulose was calculated from the sum of xylose, arabinose, galactose, and mannose, using an anhydro correction of 0.90 for C-6 sugars and 0.88 for C-5 sugars, respectively (Sluiter et al., 2008). 43

62 HPLC (Agilent 1200 series, MN, USA) equipped with an online degasser, isocratic pump, autosampler, column compartment, and refractive index detector (RID) was used in this study. Biorad Aminex HPX-87P column was used for sugar analysis and the mobile phase was HPLC grade water eluting at 0.6 ml/min. The temperature of the column and RID was maintained at 80 ºC and 55 ºC, respectively. Phenomenex Rezex RFQ-Fast Fruit H + column (Phenomenex Inc., CA) was used for the analysis of ethanol and sugar-degraded products (e.g., HMF, furfural). The mobile phase for this column was 5 mm H 2 SO 4 at a flow rate of 0.6 ml/min. The temperature of the column and the RID was maintained at 55 ºC and 45 ºC, respectively. Guard columns (Micro-Guard Refill Cartridge, Biorad, CA, USA) were installed with the analytical columns. Dry matter loss was calculated as the percentage of total solid loss during pretreatment. Degradation of lignin, cellulose, and hemicellulose was defined as the percentage of the corresponding component decreased during pretreatment, respectively. Overall glucose and xylose yields were defined as the percentage of theoretical glucose and xylose yield of untreated corn stover, respectively. Ethanol yield was calculated as the percentage of theoretical ethanol yield on the basis of total cellulose in untreated corn stover Statistical analysis The software SAS 9.1 (SAS Inc, Cary, NC, USA) was used for statistical analysis. All component degradation data, sugar yield, and ethanol yield were subjected to analysis of variance (ANOVA) using PROC GLM. Multiple comparison tests were performed 44

63 with Tukey's Test Results and Discussion Degradation of corn stover The effects of moisture content and particle size on degradation of corn stover after 18 d of cultivation are shown in Table 3.1. Initial moisture content of the substrate is important to fungal growth and secondary metabolism in solid state fermentation (Reid, 1989a). Low moisture prevents fungal growth and hampers substrate degradation as indicated by slow growth of C. subvermispora at 45% moisture and minimal lignin degradation (less than 1.50%) of corn stover. An increase in moisture content led to a significant increase in lignin degradation (P < 0.05). After 18 d of fungal pretreatment, the lignin degradation reached 19.48%, 29.54%, and 31.33% for 5 mm corn stover with moisture contents of 60%, 75%, and 85%, respectively. The optimal moisture content for fungal pretreatment of corn stover with C. subvermispora appeared to be around 75%. These results were in agreement with findings from other studies indicating that an initial moisture ranging from 70-80% was the optimal level for lignin degradation and ligninolytic activities. Shi et al. (2008) reported that maximum lignin degradation of 27.70% was obtained at a moisture content of 75% when moisture content ranged from 65-80% was tested during lignin degradation of cotton stalk by Phanerochaete chrysoporium. Asgher et al. (2006) investigated the solid state cultivation of P. chrysoporium on corn cob with moisture content ranging from 40-90%. The highest ligninase activity was obtained at 70% moisture content. In general, high moisture 45

64 content can provide adequate water for supporting vigorous fungal growth and active metabolic function. However, too much moisture reduces interparticle spaces and substrate porosity in solid state fermentation, which in turn decreases oxygen diffusion and inhibits aerobic cultures (Singhania et al., 2009). It can possibly explain why no significant improvement of lignin degradation of corn stover was achieved when the moisture content was increased from 75% to 85%. In addition to moisture content, particle size of the substrate is also a major factor affecting the performance of solid state fermentation. Large particle size can hamper accessibility of nutrients by fungi and also prevent the penetration of air and metabolite intermediates into the particles. However, the reduced particle size may adversely affect interparticle aeration. During the cultivation of Phlebia tremellosa with active aeration, hammer-milled aspen wood ( 150 µm-2 mm) showed the similar delignification rate as finer particle size (10-40 mesh), but significantly faster than the larger sizes such as chips (Reid, 1989b). As shown in Table 3.1, substantially lower lignin degradation ( %) was obtained with 15 mm corn stover than was obtained with 5 and 10 mm corn stover ( %) at moisture content ranging from 60-85%. The highest lignin degradation of 31.59% was obtained with 10 mm corn stover at 85% moisture content. However, there was no significant difference between the lignin degradation of 5 and 10 mm corn stover (P > 0.05). The interactive effect between moisture content and particle size on lignin degradation of corn stover was not significant (P > 0.05). Selectivity value, defined as degradation of lignin over cellulose, is important for evaluating the selective lignin-degrading ability of white rot fungi. The highest selectivity 46

65 value of 6.97 with lignin degradation of 31.33% was obtained from 5 mm corn stover at 85% moisture content. For all the tests in this study, the cellulose degradation caused by C. subvermispora was less than 6.16%, and was not significantly affected by moisture content (60-85%) or particle size (5-15 mm) (P > 0.05). However, there was up to 22.45% hemicellulose reduction along with lignin degradation, due to utilization of hemicellulose as the carbon source by C. subvermispora. For corn stover with the particle size of 5 and 10 mm, an increase in hemicellulose degradation was observed with an increase in moisture content from 45% to 85%. For 15 mm particle size, hemicellulose degradation increased from 0.42% to 18.29% when moisture content increased from 45% to 75%. However, hemicellulose degradation for corn stover at 85% moisture content was not significantly different from that at 75% moisture content (P > 0.05) Enzymatic hydrolysis Effect of moisture content Fungal pretreatment of corn stover at 45% moisture content did not significantly improve enzymatic hydrolysis compared to that of untreated corn stover (Fig.3.1), which indicated that low moisture content had unfavorable effects on cellulose digestibility. Substantial increases in glucose and xylose yields were obtained when the moisture content of corn stover increased from 45% to 75% (Fig.3.1); however, no increase in glucose and xylose yields was observed when the moisture content further increased from 75% to 85%. The highest glucose and xylose yields of 66.61% and 38.30%, respectively, were obtained with 5 mm corn stover at 75% moisture content after 35-d pretreatment. 47

66 Similar to corn stover, the optimal moisture content for degradation of straw materials by several white rot fungi also appears to be about 75% (Reid, 1989a). When supplemented with nutrients, moisture content as low as 60% was also effective for degradation of rice straw and an overall sugar yield of 32% was obtained (Taniguchi et al., 2005) Effects of particle size As shown in Fig.3.1, the effect of particle size on the glucose and xylose yields was not significantly different between corn stover with a moisture content of 45% and untreated corn stover (P > 0.05). Increases in glucose and xylose yields were obtained with a decrease in particle size of corn stover at moisture content 60-85%. The overall glucose yields of 45.76%, 54.18%, and 57.67% were obtained with 15, 10, and 5 mm corn stover, respectively, with fungal-pretreatment for 18 d at 75% moisture content. The xylose yield was 30.40%, 36.88%, and 38.21% accordingly. Increases in glucose and xylose yields were also obtained with a decrease in particle size for 28 d- (data not shown) and 35 d-pretreated corn stover when the moisture content was in the range of 60-85%. Particle size can affect both pretreatment and enzymatic hydrolysis as particle size reduction can increase the ratio of accessible surface area to volume. The particle size being reduced to submicron scale caused significant increase in digestibility of native cellulose (Yeh et al., 2010). In this study, size reduction from 15 to 10 mm caused an increase of about 5-10% in the glucose yield of corn stover at a moisture content of 60-85%. Further reduction of particle size from 10 to 5 mm did not increase the sugar yield remarkably as only a 2-5% increase was observed. 48

67 Effect of pretreatment time Pretreatment time also plays an important role in fungal pretreatment. As fungal pretreatment was not effective for degradation of corn stover at 45% moisture, no improvement of sugar yields were observed at this moisture level during 35 d of pretreatment compared to that of untreated corn stover (Fig.3.1). The glucose yield increased with prolonged pretreatment time at high moisture contents of 75-85% while it changed slightly with pretreatment time at a moisture content of 60%. For 5 mm corn stover with 75% moisture, the glucose yields of 57.67% and 62.21% (data not shown), and 66.61% were obtained after 18, 28, and 35 d of fungal pretreatment, respectively. For the same pretreatment periods, xylose yields of 38.42%, 38.19% (data not shown), 38.30% were obtained respectively, which were not significantly different from each other (P > 0.05). Prolonged pretreatment caused more degradation of lignin which facilitated cellulose conversion, but the hemicellulose conversion was not significantly affected by pretreatment time (P > 0.05). It was observed that after 72 d of cultivation by P. ostreatus, 32% of carbohydrate in rice straw were converted to soluble sugars (Taniguchi et al., 2005). For corn stover pretreated by C. stercoreus, a maximum of 36% glucose yield was obtained after 29-d cultivation (Keller et al., 2003). Pretreatment time of d were required for fungal pretreatment of bamboo to obtain significant improvement of sugar yield (Zhang et al., 2007a; Zhang et al., 2007b). Phanerochaete chrysosporium is a fast-growing white rot fungus but no enhancement of saccharification rate was observed due to simultaneous degradation of holocellulose and lignin by this fungus (Keller et al., 2003; Sawada et al., 1995; Shi et al., 2009). In submerged 49

68 cultivation of P. chrysosporium, a glucose yield of 50%, corresponding to 64.9% based on the treated, was obtained from rice straw after 15-d pretreatment (Bak et al., 2009). In contrast, solid state pretreatment of corn stover by C. subvermispora resulted in a high glucose yield during a relatively short degradation period. The interactive effects between moisture content and particle size, and moisture content and pretreatment time on overall glucose yield were significant (P < 0.05). However, the interaction between particle size and time as well as among these three factors was not significant (P > 0.05). For xylose yield, only the interactive effect between particle size and moisture content was significant (P < 0.05) Effect of pretreatment temperature Temperature is important for on-farm, in-storage pretreatment by white rot fungi as the concurrent storage and pretreatment needs to be carried out throughout seasons with fluctuating temperatures instead of with controlled temperatures. Therefore, corn stover with 75% moisture content was pretreated at temperatures between 4 and 37 ºC for 18 d to investigate the effect of fungal pretreatment temperature on the subsequent sugar yield of corn stover. As indicated in Fig.3.2, the optimal temperature for fungal pretreatment appeared to be between 28 and 32 ºC. The highest glucose and xylose yields of 57.67% and 38.21% were obtained at 28 ºC. Temperature out of the optimal range may inhibit or kill the fungus. Glucose and xylose yields obtained at 4 and 37 ºC were close to that of untreated corn stover and no fungal growth was observed at these two temperatures. A glucose yield as low as 26.69% was obtained from corn stover pretreated 50

69 at 34 ºC. Glucose yield decreased from 57.67% to 30.01% when pretreatment temperature decreased from 28 to 19 ºC. As discussed above, the sugar yield increases with the increase of lignin degradation by the white rot fungus; thus, it can be concluded that the reduced delignification ability of C. subvermispora at temperatures outside the optimal range resulted in reduced sugar yield. In general, white rot fungus (basidiomycetes) can grow well between 15 and 35 ºC but high delignification rate is generally obtained within an optimal temperature range between 25 and 30 ºC (Reid, 1985; Reid, 1989a) Effect of enzyme loading and supplemental enzymes on enzymatic hydrolysis Non-cellulolytic enzymes such as xylanase and pectinase can break down heterogeneous hemicellulose polymers that interlink with cellulose microfibrills (Murnen et al., 2007). Therefore, supplementation of these enzymes to cellulase may benefit cellulose conversion. In this study, the effect of cellulase loading and supplemental enzymes on saccharification was studied using 18 d-pretreated corn stover as the substrate. As shown in Fig.3.3, the glucose yields of 57.67%, 62.57% and 65.83% were obtained at enzyme loadings of 10, 20, 30 FPU/g solid, respectively. The xylose yield was 38.21%, 42.20%, 45.28%, respectively, with the above enzyme loading. No further increase in glucose and xylose yields was observed when the enzyme loading increased to 40 FPU/g solid. The saccharification tests with enzyme supplementation were carried out at the cellulase loading of 10 FPU/g solid. The glucose yield increased with the addition of supplemental enzymes (Fig.3.4), while the yield was dependent on the type of 51

70 supplemental enzymes and enzyme loading. The addition of xylanase at loadings of 240 and 600 ABXU/g solid caused about 8 and 10% increase in the glucose yield, respectively. Correspondingly, about 6 and 8% increase in the xylose yield was obtained, respectively. Accellerase XC is an accessory xylanase/cellulase enzyme complex designed for blending with cellulase for biomass hydrolysis. When XC was added at 40 µl /g solid (equivalent to 120 ABXU /g solid & 48 CMCU /g solid), the glucose yield of 71.81% and xylose yield of 51.39% were obtained, which was about 14% and 13% higher, respectively, than that obtained without accessory enzymes (only cellulase was applied). In contrast, the glucose yield increased by about 8-10% at the XC loading of 20 µl /g solid. Although the xylanase activity of XC added was only one fifth of that of single xylanase added, the results indicated that XC had a synergistic effect on cellulase to enhance polysaccharide conversion, probably due to the cellulase activity of XC. However, an increase in cellulase loading did not cause such noticeable effect on saccharification (Fig.3.4). This suggested cellulase and xylanase in XC worked synergistically on saccharification. A possible explanation of less effect of single xylanase supplementation is that concomitant lignin and hemicelluloses degradation by C. subvermispora could mask the hydrolysis of heterogeneous xylan polymers by external xylanase (Ko et al., 2009) Ethanol production Simultaneous saccharification and fermentation (SSF) was conducted to evaluate ethanol production from fungal-treated samples. As shown in Fig.3.5, rapid increases in 52

71 ethanol yields were obtained within the first 24 h of saccharification and fermentation. The effect of pretreatment time on the ethanol yield during the first 24 h is not significant (P < 0.05). After 24 h, the ethanol yields increased slowly and higher ethanol yields were obtained with extended pretreatment time. Untreated corn stover resulted in an ethanol yield around 15.91% and ethanol fermentation was completed within 12 h. After 72 h, ethanol yields of 50.68%, 54.26% and 57.80% were obtained from the sample treated for 18, 28, 35d, respectively. This result is comparable to that obtained with combined pretreatment of wood chips with C. subvermispora and ethanolysis, where an ethanol yield of 62% was achieved (Itoh et al., 2003). An ethanol yield of 63% based on the treated solids, corresponding to 50% based on the untreated, was obtained with rice straw when it was pretreated by P. chrysosporium under submerged cultivation (Bak et al., 2009). Unlike thermochemical pretreatment, no inhibitory compounds to yeast fermentation were observed during fungal pretreatment. Ethanol yields of % and % were obtained from unwashed samples pretreated for 18 and 35 d, respectively (Fig.3.5). Thus, there was no significant difference between the ethanol yields of washed and unwashed samples (P < 0.05). It suggested that washing or detoxification of fungal-pretreated lignocellulosic biomass is not necessary for ethanol production, which simplifies the upstream process of ethanol production. 53

72 3.4. Conclusions C. subvermispora can effectively reduce recalcitrance of corn stover with high selective delignification. Lignin degradation and enzymatic hydrolysis yield were significantly affected by the particle size and moisture content of corn stover, as well as pretreatment time and temperature. This solid-state fungal pretreatment shows great potential to be a cost-effective pretreatment process for ethanol production from corn stover. Considering the ambient temperature, the relatively long pretreatment time, and the limited amount of cellulose loss, it is feasible to apply fungal treatment concurrently with on-farm wet storage for ethanol production. 54

73 Table 3.1. Degradation of corn stover pretreated by C. subvermispora at 28 ºC for 18 d PZ a MC b Selectivity Degradation (%) d (mm) (%) Value c Dry mass Cellulose Hemicellulose Lignin ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±0.21 a PZ is particle size; b MC is moisture content; c defined as degradation of lignin over cellulose; d data are mean values ± Standard error of three replicates. 55

74 (a) (b) Figure 3.1. Interactive effects of moisture content, pretreatment time, and particle size on enzymatic hydrolysis of corn stover (Pretreatment condition: 28 ºC): (a) glucose yield and (b) xylose yield. 56

75 Figure 3.2. Effect of pretreatment temperature on enzymatic hydrolysis of corn stover pretreated by C. subvermispora (pretreatment conditions: 5 mm corn stover, 75% moisture content, and 18 d). 57

76 Figure 3.3. Effect of enzyme loading on enzymatic hydrolysis of corn stover pretreated by C. subvermispora (pretreatment conditions: 5 mm corn stover, 28 ºC, 75% moisture content, and 18 d). 58

77 Figure 3.4. Effect of supplemental enzymes on enzymatic hydrolysis of corn stover pretreated by C. subvermispora (pretreatment conditions: 5 mm corn stover, 28 ºC, 75% moisture content, and 18 d). Cellulase: 10 FPU/g solid; xylanase +: cellulase (10 FPU/g solid) and xylanase (240 ABXU/g solid); xylanase ++: cellulase (10 FPU/g solid) and xylanase (600 ABXU/g solid); XC+: cellulase (10 FPU/g solid) and XC (60 ABXU/g solid & 24 CMCU/g solid)); XC++: cellulase (10 FPU/g solid) and XC (120 ABXU/g solid & 48 CMC/g solid). 59

78 Figure 3.5. Ethanol yield of corn stover pretreated by C. subvermispora (pretreatment conditions: 5 mm corn stover, 28 ºC, 75% moisture content, and d). 60

79 Chapter 4 Degradation Mechanism of Fungal Pretreatment of Corn Stover Delignification of corn stover by the white rot fungus C. subvermispora under solid-state fermentation was evaluated for improving subsequent enzymatic hydrolysis. The results showed that C. subvermispora selectively degraded lignin by as much as 39.20% while the cellulose degradation was less than 5% during 42 d of pretreatment. However, hemicellulose loss of up to 27.0% was concomitant with the lignin degradation. MnP and laccase were two ligninolytic enzymes detected during degradation of corn stover by C. subvermispora. For major hydrolytic enzymes, xylanase was the only enzyme detected. The enzymatic hydrolysis yield of corn stover delignified by C. subvermispora was remarkably improved, reaching about 57-67% glucose yield after d of pretreatment, while the glucose yield of the untreated was only 22% Introduction Biological delignification by white rot fungi is carried out through ligninolytic enzymes secreted into the substrate. Two peroxidases, namely lignin peroxidase (LiP) and manganese peroxidase (MnP), and a copper-containing phenol oxidase, laccase, are considered the main enzymes involving lignin depolymerization. However, not all of these enzymes are produced by a specific fungal culture. For example, Phanerochaete chrysoporium produces LiP and MnP that are associated with lignin oxidation, but lacks laccase (Ruttimann-Johnson et al., 1993). C. subvermispora produces two extracellular 61

80 ligninolytic enzymes, MnP and laccase, which contribute to lignin removal; however, LiP has not been detected, although LiP-like genes have been identified (Rajakumar et al., 1996). The specific mechanism by which ligninolytic systems contribute to the early stage decay is not clear. These enzymes, which have high molecular weights, cannot permeate a sound cell wall (Blanchette et al., 1997). However, oxidative enzymes may be active at the surface of the cell wall and induce the formation of low molecular mass agents, such as radicals, to diffuse into an unaltered cell wall (Kapich et al.; Watanabe et al., 2000). The expression of ligninolytic systems largely depends on the lignin structure and culture conditions (e.g., C/N ratio, cultivation time, and composition of medium) (Arora and Gill; Sethuraman et al., 1998). The purpose of this study was to evaluate biodelignification performance of C. subvermispora on corn stover under solid state fermentation. The compositional changes of corn stover during fungal degradation were investigated. The ligninolytic enzymes (LiP, MnP, and laccase) and hydrolytic enzymes (cellulase and xylanase) involved in the degradation of corn stover were also examined. The effect of fungal pretreatment on enzymatic hydrolysis of corn stover was also examined Material and Methods Feedstock Corn stover was harvested from the Ohio Agricultural Research and Development Center (OARDC) farm in Wooster, OH. The collected corn stover was dried at 40 ºC for 62

81 72 h and then ground to pass through 5 mm sieve using a Thomas Wiley Mill. The ground corn stover (about 5% moisture) was stored at room temperature prior to use Fungus and inoculum preparation The cultivation and inoculum preparation of C. subvermispora were the same as described in Section of Chapter Pretreatment by solid-state fermentation Ten grams of corn stover (dry basis) were mixed with DI water in a 250 ml Erlenmeyer flask to obtain 75% moisture and then autoclaved for 30 min at 121 ºC. After cooling to room temperature, the substrate in each flask was inoculated with 2 ml of blended liquid culture for fungal pretreatment. The substrate added with 2 ml sterile water instead of fungal inoculation was used as a control. Fungal pretreatment was carried out under static conditions in a refrigerated incubator at 28 ºC for up to 42 d. The samples were taken at intervals of 2-7 d for compositional analysis, enzymatic hydrolysis and lignocellulolytic enzyme assays. All tests were run in triplicates Compositional analysis of corn stover samples The procedures used for determination of dry mass, structural carbohydrate and lignin were the same as described in Section of Chapter Enzyme extraction and assays The flask culture of fungal pretreatment was suspended in 120 ml sodium acetate buffer (50 mm, PH 4.5). The mixture was extracted at 150 rpm in a rotary shaker at 28 ºC 63

82 for 4 h and subsequently filtrated to separate extracellular enzymes. The enzyme extract was assayed using a UV-Vis spectrophotometer (Biomate TM 3, Thermo-scientific, MA, USA). LiP activity was assayed based on Azure B oxidation (ε = M 1.cm 1 ) due to color-formed enzyme extract (Archibald, 1992). The reaction mixture contained 1 ml of 32 μm Azure B, 100 μm H 2 O 2, 50 mm sodium tartrate buffer (ph 4.5), and 500 μl enzyme extract. The reaction was initiated by adding H 2 O 2. The decrease in absorbance was monitored at 651nm. MnP activity was determined by phenol red oxidation (ε = M 1.cm 1 ) (Glenn and Gold, 1985). Reaction mixtures contained 5.5 ml of 25 mm sodium lactate (ph 4.5), 20 mm sodium succinate buffer (ph 4.5), 0.1 mm H 2 O 2, 0.1 mm MnSO4, 1% (w/v) bovine serum albumin, 0.1 % (w/v) phenol red, and 250 μl enzyme extract. The reaction was started by the addition of H 2 O 2. At 1-min interval, 1 ml of reaction mixture was taken and immediately added to 50 μl of 10% (w/v) NaOH to terminate the reaction. The absorbance was read at 610 nm. Laccase activity was determined by oxidation of 2, 2 -azino-bis-(3-ethyl benzthiazoline-6-sulphonate) (ABTS, ε = M 1.cm 1 ) (Bourbannais and Paice, 1990). The reaction mixture contained 1 ml of 0.5 mm ABTS, 20 mm sodium acetate buffer (ph 4.5), and 250 μl enzyme. Oxidation of ABTS was followed by absorbance increase at 420 nm. One unit of enzyme activity (LiP, MnP, and laccase) was defined as one µmol of substrate oxidized per minute. The activity of enzyme extract was determined as IU/ml. The total unit per flask culture was calculated by multiplying IU/ml by the total volume of enzyme extract (Ferraz et al., 64

83 2003). The enzyme activities reported here were expressed as IU/gram solid of untreated corn stover. Hydrolytic enzymes were assayed based on the dinitrosalicylic acid (DNS) method (Miller, 1959). Filter paper activity and endo-β-1, 4-glucanase were measured with Whatman No.1 filter paper and carboxymethyl cellulose (CMC) as the substrates, respectively (Ghose, 1987). Xylanase was assayed with acid birch xylan as the substrate (Bailey et al., 1992). One unit of activity was defined as the amount of enzyme required to produce 1µmol reducing sugars per min Enzymatic hydrolysis Spezyme CP obtained from Genencor (Palo Alto, CA) was determined to have an activity of 50 FPU /ml and used for enzymatic hydrolysis. The procedure was the same as described in Section of Chapter Results and discussion Degradation of corn stover The native corn stover used in this study was composed of 38.4% cellulose, 23.0% hemicellulose (mainly 18.5 % xylan), 20.2% lignin, and others (e.g., extractives, ash). Degradation of corn stover during 42-d cultivation is depicted in Fig.4.1. The dry mass, cellulose, hemicellulose and lignin degradation of the control were 3.4%, 1.8%, 3.9% and 1.4%, respectively (date not shown). For fungal-treated corn stover, total dry mass loss reached 18.8% after 42 d, which was higher than that resulting from the pretreatment of wood chips by C. subvermispora (Guerra et al., 2003), probably due to 65

84 more degradable compounds in corn stover than in wood. Lignin degradation increased rapidly from 2.9% to 25.4% between days 7 and 14 and then slowed down. A maximum lignin degradation of 39.2% was reached at the end of 42-d cultivation. Compared to the control, no significant lignin degradation was observed during the first 7-d cultivation. Hemicellulose degradation by C. subvermispora reached 27.0% after 42-d cultivation while cellulose degradation was less than 5% (Fig.4.1). These results indicate that C. subvermispora apparently utilized more easily accessible hemicellulose rather than cellulose to support its growth and metabolism. The hemicellulose was mainly converted to CO 2 and H 2 O as byproducts. C. subvermispora was able to degrade both ester- and ether-linked aromatics, particularly β-o-4 ether-linked phenolic acids in lignin complex (Akin et al., 1993). This probably contributed to cleavage of ester or ether linkages between aromatic acid and xylan (Anderson and Akin, 2008), facilitating xylan degradation by this fungus Lignocellulolytic enzymes and their effects on corn stover degradation Enzyme production during fungal pretreatment Extracellular enzymes produced during fungal pretreatment of corn stover were extracted and assayed to determine the activities of oxidative and hydrolytic enzymes. No LiP was detected during the corn stover decay process (Fig.4.2a), which is consistent with the observations of Ruttimann-Johnson et al. (1993) and Rajukumar et al. (1996) that C. subvermispora lacks LiP activity in either liquid culture or solid culture. MnP and laccase were two major lignin degradation enzymes detected in the solid culture on corn 66

85 stover (Fig.4.2a). MnP reached its maximum activity of 2.2 IU/g solid on day 7 and then decreased to approximately 0.3 IU/g solid after day 28. The MnP activity obtained in this study was higher than was obtained in pretreatment of sugar cane (Costa et al., 2005) and loblolly Pine (Guerra et al., 2003) by C. subvermispora which were reported to be 0.2 IU/g and 0.1 IU/ g solid, respectively. Laccase activity reached its first peak (1.5 IU/g solid) on day 7. The lowest level of laccase activity was observed on day 14. Thereafter, the laccase activity increased rapidly and reached 3.6 IU/g solid at the end of 42-d cultivation (Fig.4.2a). It was also observed that laccase became the dominant ligninolytic enzymes after 22-d cultivation. The result differed from the production of ligninolytic enzymes in wood culture of C. subvermispora that indicated MnP was the predominant enzyme during fungal cultivation with less or no laccase detected (de Souza-Cruz et al., 2004). Costa et al. (2005) also reported that no laccase was detected in sugarcane bagasse culture of C. subvermispora. However, laccase produced in pretreatment of switchgrass by C. subvermispora also indicated higher activity than MnP after 18-d cultivation (data not shown). Reducing sugars released from the assay mixture by incubation of filter paper strips with enzyme extract were not high enough to determine filter paper activity (Fig.4.2b). This result is consistent with several reports on cellulolytic activity of C. subvermispora cultured on woody biomass. Those studies indicated that cellulase determined as FPU was too low to attain an appropriate level of cellulase activity (Ferraz et al., 2003; de Souza-Cruz., 2004). Endo-glucanase as determined to be CMC activity was also negligible in this study (Fig.4.2b); although Sethuraman et al. (1998) reported that a 67

86 considerable amount of endo-glucanase was induced in the liquid culture of C. subvermispora grown in high fiber substrates such as kenaf and flax. A considerable amount of xylanase was produced during cultivation of corn stover by C. subvermispora (Fig.4.2b). The pattern of xylanase production was similar to that of laccase, reaching its peak value of 4.8 IU/g solid on day 7. The lowest level was observed on day 18, after which a slight increase in activity followed and the activity reached 4.1 IU/g solid at the end of the fungal pretreatment. The level of xylanase activity in this study was close to that observed from some other solid cultures of C. subvermispora. Xylanase activity of 4.5 and 4.3 IU/g solid was observed from 30-d solid fermentation of sugarcane bagasses and hardwood P. taeda, respectively (Guerra et al., 2003; Costa et al., 2005). In contrast, xylanase activity of the culture of the white rot fungus Irpex lacteus CD2 on corn stover showed two peaks during a cultivation period of 120 d and reached a maximum level of 8.9 IU/g solid on day 60 (Xu et al., 2009) Effect of lignocellulolytic enzymes on corn stover degradation Lignin degradation increased rapidly from day 7 to 14 and then increased slowly while no significant lignin degradation was observed during the first 7 d (Fig.4.1). However, both laccase and MnP activity reached their peaks on day 7 while the laccase activity increased again after 14 d until the end of the cultivation (Fig.4.2a). Our results indicated that significant lignin degradation occurred after the production of laccase and MnP. This is in agreement with the observation of Guerra et al. (2002) that the mineralization of lignin corresponding to lignin degradation occurred after the lignin 68

87 depolymerization. Ligninolytic enzymes produced during fungal cultivation are primarily responsible for the lignin depolymerization. On the other hand, high molecular-weight ligninolytic enzymes cannot penetrate the cell wall due to the low permeability of wood at the incipient stages of degradation, indicating that low molecular mass agents catalyze the lignin depolymerization at the incipient stages (Ferraz et al., 2003). Lipid radicals such as peroxyl and acyl radicals derived from MnP-dependent lipid peroxidation might serve as small, diffusible agents for initiating lignin depolymerization in sound wood, where ligninolytic enzymes cannot penetrate (Blanchette et al., 1997; Kapich et al., 1999; Watanabe et al., 2000). No significant levels of cellulase were detected in the solid culture of C. subvermispora on corn stover (Fig.4.2b), which is in agreement with the low glucan loss during the fungal pretreatment. A non-enzymatic process, such as the fenton system (Fe 2+ and H 2 O 2 ), might have caused cellulose depolymerization (Aguiar et al., 2006; Ferraz et al., 2003). Xylanase was the main hydrolytic enzyme detected in degradation of corn stover (Fig.4.2b). High levels of xylanase were detected before significant degradation of hemicellulose occurred (Fig.4.1 and Fig.4.2b), which indicates that xylanase was not sufficient to degrade xylan at the early stage probably due to low permeability of the cell wall (Vicentim and Ferraz, 2007). Lignin degradation could facilitate diffusion of xylanase into the plant cell wall to cause significant xylan degradation by increasing cell wall permeability (Vicentim and Ferraz, 2007; Machuca and Ferraz, 2001). It may explain why hemicellulose loss increased with an increase in lignin degradation, following a similar pattern of lignin degradation (Fig.4.1). In the study on the degradation of wood (Machuca 69

88 and Ferraz, 2001) and sugarcane straw (Costa et al., 2005) by C. subvermispora, xylanase was also found to be the major hydrolytic enzyme and the hemicellulose degradation was closely related to the extent of lignin degradation Morphology of fungal-treated corn stover The morphological change of corn stover was observed during 42-d cultivation. C. subvermispora colonized corn stover quickly with abundant mycelial mass. After 7-d cultivation, the color of the entire substrate became yellowish-orange and moisture started to accumulate, probably due to water generated from degradation of cell wall components during fungal metabolism. At the advanced stage of decay (beyond 35 d), the color changed into whitish-yellow and the substrate became light, soft, and spongy. The observation of 18 d-treated corn stover by scanning electron microscopy (SEM) clearly showed the modification of the cell wall by fungal attacks (Fig.4.3). Abundant hyphae colonized on the entire vessel of corn stover and penetrated through pit structure or cell wall (Fig.4.3b-d). Numerous holes and erosion troughs were formed on the vessel of corn stover beneath or around hyphae (Fig.4.3b). When erosion troughs enlarged, cavity or cracks developed on the cell wall (Fig.4.3b-c). In contrast, the untreated corn stover had a rigid and intact structure (Fig. 4.3a) Correlation between fungal degradation and enzymatic digestibility Enzymatic digestibility of corn stover treated by C. subvermispora for 7-42 d was evaluated by enzymatic hydrolysis. As shown in Fig.4.4, the glucose and xylose yields of 7-d pretreated corn stover were 23.32% and 11.38%, respectively, which was close to 70

89 that of the control. However, the overall glucose yield increased with the cultivation time beyond 7 d, reaching 66.91% at the cultivation time of 35 d but no further increase was observed when cultivation time was extended to 42 d. Similarly, the overall xylose yield increased with cultivation time, reaching the highest yield of about 38% after 18d. These results showed that sugar yields of corn stover were considerably affected by the cultivation time. In the study of fungal pretreatment of corn stover conducted by Keller et al. (2003), a maximum cellulose conversion of 35.7% was obtained with C. stercoreuspretreated substrate while P. chrysosporium caused a reduction of cellulose conversion. Xu et al. (2010) tested degradation of corn stover by Irpex lacteus CD2 for up to 150 d and obtained a maximum saccharification yield of 66.4% (based on theoretical yield of holocellulose in raw materials) with 25-d pretreated substrate. In contrast, C. subvermispora appeared to be comparable or more effective than other white rot fungi for improving enzymatic hydrolysis efficiency of corn stover. Lignin in biomass is the major barrier to enzymatic hydrolysis; thus, lignin removal can improve accessibility of the substrate to cellulase (Taniguchi et al., 2005; Yu et al., 2009). When plotting the remaining lignin content in the fungal-treated substrate against the sugar yield, there was a good linear correlation between the residual lignin content and cellulose digestibility (R 2 =0.97) (Fig.4.5), indicating the enzymatic hydrolysis is strongly affected by the residual lignin content. A slight change in residual lignin content from 35 to 42 d could explain a leveled-off glucose yield. The xylose yield was also negatively related to lignin content in corn stover, but not as significantly as the glucose yield (R 2 =0.92), which might be attributed to substantial hemicellulose 71

90 degradation during fungal pretreatment. The correlation between fermentable sugar yield and the lignin content of fungal-treated biomass was also reported by other researchers (Kirk and Moore, 1972; Sawada et al., 1995; Taniguchi et al., 2005; Zhang et al., 2007b), which indicates cellulose digestibility can be potentially enhanced by preferential degradation of lignin. In addition to the effect of lignin degradation, increased pore sizes in the substrate caused by fungal pretreatment also contributed to the improved cellulose conversion by creating more accessible surface area to cellulase (Taniguchi et al., 2005). An increase in pore size from 20 Ǻ or less to Ǻ was observed in fungal-treated wood, which allowed some enzymes to penetrate the cell wall at the advanced stage of decay (Blanchette et al., 1997; Arora and Gill, 2000). Degradation of hemicellulose by white rot fungi also removed the physical protective coat of cellulose and consequently improved cellulose digestibility (Ohgren et al., 2007) Conclusions C. subvermispora can selectively degrade lignin in corn stover over cellulose, but with concomitant hemicellulose degradation. MnP and laccase were two major oxidative enzymes detected in solid culture of C. subvermispora on corn stover. Xylanase was produced during degradation of corn stover but there was no detectable cellulase. Lignin degradation caused by C. subvermispora may have facilitated diffusion of enzymes into the cell wall of corn stover during the delignification process as well as during enzymatic 72

91 hydrolysis. As a result, cellulose digestibility of fungal-treated corn stover was remarkably improved compared to that of untreated corn stover. 73

92 Figure 4.1. Degradation of corn stover during 42-d pretreatment by C. subvermispora. 74

93 (a) (b) Figure 4.2. Enzyme production during 42-d pretreatment of corn stover by C. subvermispora: (a) ligninolytic enzymes and (b) hydrolytic enzymes. 75

94 a b c d Figure 4.3. Scanning electron micrograph of corn stover pretreated by C. subvermispora for 18 d. (a) untreated corn stover (at 450 magnification), (b) fungal-treated corn stover (at 250 magnification) showing numerous holes and erosion troughs on the vessel wall (arrows), (c) fungal-treated corn stover (at 1000 magnification) showing cracks and cavities when erosion troughs enlarged (arrows), and (d) fungal-treated corn stover (at 900 magnification) showing hyphae penetration through cell walls (arrows). 76

95 Figure 4.4. Enzymatic hydrolysis of corn stover pretreated by C. subvermispora for up to 42 d. CTL: control. 77

96 Figure 4.5.Correlation between the sugar yield and remaining lignin content in corn stover pretreated by C. subvermispora. 78

97 Chapter 5 Effectiveness of Fungal Pretreatment on Different Biomass Feedstocks Different types of feedstocks, including corn stover, wheat straw, soybean straw, switchgrass, and hardwood, were tested to evaluate the effectiveness of fungal pretreatment by Ceriporiopsis subvermispora. After 18-d pretreatment, corn stover, switchgrass, and hardwood were effectively delignified by the fungus. Correspondingly, glucose yields during enzymatic hydrolysis reached 56.50%, 37.15%, and 24.21%, respectively, which were 2-3 fold greater than those of the raw materials. A further 10-30% increase in glucose yields was observed when pretreatment time was extended to 35 d. In contrast, cellulose digestibility of wheat straw and soybean straw were not significantly improved by fungal pretreatment. When external carbon sources and enzyme inducers were added during fungal pretreatment of wheat straw and soybean straw, only glucose and malt extract addition improved cellulose digestibility of wheat straw while the cellulose digestibility of soybean straw was not improved. Further study is needed to improve specific fungal-substrate for the most recalcitrant feedstocks Introduction Due to differences in native recalcitrance of biomass feedstocks, pretreatment performance is feedstock-dependent. The effectiveness of chemical pretreatment may be less affected by feedstock sources than for biological pretreatment. Nevertheless, it is difficult to select a pretreatment method suitable for each substrate without 79

98 comprehensive investigation. Prior work on fungal pretreatment of biomass has emphasized screening of white rot fungi with the best performance for one specific feedstock. Research on the effectiveness of white rot fungus on various feedstocks is scare. C. subvermispora has been shown to degrade herbaceous (i.e., Bermudagrass, corn stover, rice straw) (Akin et al, 1995; Taniguchi et al., 2005; Wan and Li, 2010a) and woody biomass (Messner et al., 1998) with the degradation extent depending on the cultivation conditions and substrate. Therefore, the feasibility of our current pretreatment process using C. subvermispora on different feedstocks will help optimize fungal pretreatment and ensure the process is robust enough to handle a wide range of feedstocks. In this study, various feedstocks, including agricultural residue (corn stover, wheat straw, and soybean straw), a dedicated energy crop (switchgrass), and forestry residue (hardwood chips) were pretreated with C. subvermispora and the degree of lignin removal and resulting cellulose digestibility of these feedstocks were investigated. The inducers (Mn 2+, H 2 O 2 and veratryl alcohol) and buffer solution were also tested in microbial pretreatment to evaluate their effects on sugar yields of enzymatic hydrolysis Materials and methods Feedstocks Agricultural residues including corn stover, wheat straw, and soybean straw were collected from the Ohio Agricultural Research and Development Center (OARDC) farm in Wooster, Ohio. Hardwood chips from campus tree trimmings were collected from the 80

99 OARDC Farm Operation Facility in Wooster, Ohio. Switchgrass was harvested from the OARDC farm in Jackson, Ohio. All feedstocks were dried at 40 ºC in a convection oven until the moisture content reached less than 10%. After drying, they were ground to pass through 5 mm screen and stored in air tight containers at room temperature prior to use. The feedstock compositions were analyzed using NREL Laboratory Analytical Procedure (LAP) (Sluiter et al., 2005, 2008) and the results are summarized in Table Fungus and inoculum preparation The white rot fungus C. subvermispora (ATCC 96608) was obtained from American Type Culture Collection (Manassas, VA). Its maintenance and inoculation preparation were the same as described in Sections of Chapter Fungal pretreatment Fungal pretreatment procedure was the same as described in Section of Chapter 4. Addition of nutrients and inducers in Section was only tested for wheat straw and soybean straw. Both feedstocks were conditioned with DI water or buffer solution in which chemicals were first dissolved to obtain desired concentrations Enzymatic hydrolysis Feedstocks (raw, control, fungal-treated) were subjected to enzymatic hydrolysis following NREL LAP-008 (Dowe and McMillan, 2001). Spezyme CP was used for hydrolytic enzymes. Detailed procedures were the same as described in Sections of 81

100 Chapter 3. The sugar yields were calculated by dividing the yield by the theoretical sugar yields of the raw feedstocks Enzyme assay Extracellular enzymes were collected by extracting the solid-state flask culture with sodium acetate buffer (50 mm, ph 4.5) in a rotary shaker (150rpm and 28ºC) for 4 h. The liquid fraction was collected by vacuum filtration and assayed for ligninolytic enzyme activities. MnP activity was determined by phenol red oxidation (Glenn and Gold, 1985) and Laccase activity by oxidation of 2,2-azino-bis-(3-ethyl benzthiazoline-6- sulphonate) (Bourbonnais, 1990). The detailed assay procedure was described in Section of Chapter 4. The total unit per flask culture was calculated by multiplying IU/ml by the total volume of enzyme extract (Ferraz et al., 2003) and expressed as IU/gram solid of raw feedstock Analytical methods Cellulose, hemicellulose, and lignin were determined based on NREL LAP for the structural carbohydrate and lignin measurement in biomass (Sluiter et al., 2008). Sugars were analyzed with HPLC (Shimadzu LC-20, Shimadzu, Columbia, MD) which was equipped with refractive index detector (RID) and Biorad Aminex HPX-87P column. Details are included in Section of Chapter 3. 82

101 5.3. Results and discussion Composition of biomass The chemical compositions of the 5 feedstocks used in this study are summarized in Table 1. The values are similar to those reported in the Biomass Feedstock Composition and Property Database (DOE, 2010). Holocellulose (cellulose and hemicelluloses), representing the carbohydrate portion of the plant cell wall, was % of total dry biomass. Cellulose, a polymer of glucose, was the main carbohydrates in the feedstocks and ranged from %. Xylan was the major hemicellulose component in the feedstocks ( %); galactan and arabinan were two minor hemicellulose constituents and accounted for about 4.5% of the biomass compositions. In addition, hardwood contained a small amount of mannan, which agreed with the report on compositions determined by others (Scheller and Ulvskov, 2010). It is interesting to note that 2.06% of mannan was also detected in soybean straw while no mannan was detected in other herbaceous species studied here. Xylan contents in soybean straw and hardwood were 12.68% and 13.73%, respectively, which were much lower than that in other feedstocks ( %). The major non-carbohydrate component was lignin, with acid-insoluble lignin of % and a minor fraction of acidsoluble lignin of % in the studied feedstocks. Water extractives were the dominant extractives of the feedstocks and the detectable soluble sugars were mainly glucose and xylose, but both only accounted for less than 1% of total biomass. This is due to the low amounts of extractable glucose sources (free glucose, sucrose and starch) in mature plant biomass (Digman et al., 2010). 83

102 Other unidentified compounds in the feedstocks included non-structural compounds (e.g., protein, pectin, resins, soil) and structural compounds (e.g., acetyl groups, glucuronic acid substitutes). The extractives (combined water and ethanol extractives) in wheat straw and switchgrass were higher than for the other feedstocks, probably due to more inorganic material and nitrogenous material as well as hydrophilic materials (e.g., chlorophyll, wax) associated with the biomass (Sluiter et al., 2005) Degradation of biomass The degradation results of all feedstocks pretreated with C. subvermispora under solid state fermentation for 18 d are shown in Fig.5.1. The degradation of lignocellulosic biomass by C. subvermispora was dependent on biomass species. The dry mass loss was correlated with the extent of degradation; the more degradation, the more dry mass loss. For corn stover and switchgrass, more than 26% lignin along with 15-19% hemicellulose were degraded after 18 d fungal pretreatment (Fig.5.1a). The degradation of control samples (without fungal inoculation), most likely resulting from autohydrolysis during autoclave, was less than 3% (Fig.5.1b). Fungal degradation of wheat straw and soybean straw was much lower than that of corn stover and switchgrass. Autohydrolysis degradation might have accounted for the majority of the total degradation observed in wheat straw and soybean straw. This finding was also supported by evidence of no or little fungal growth in the soybean straw or wheat straw. If present, the fungus typically appeared with white mycelia and a color change of biomass materials. As a wooddecaying fungus, C. subvermispora was able to effectively decompose the hardwood 84

103 material in this study, with hemicellulose removal comparable to that of corn stover and switchgrass but with lower lignin degradation (Fig.1a). Cellulose loss was less than 5% for all the feedstocks tested here as this fungus lacks a complete cellulolytic system (Fig.1.a) (Ruttiman-Johnson et al., 1993; Ferraz et al., 2003; Srebotnik et al., 1988). Various factors, such as lignin composition and structure and cross-linking between lignin and hemicelluloses, can potentially restrict degradability of plant cell walls (Grabber, 2005). C. subvermispora was reported to be capable of attacking both ester- and ether-linked phenolic acids from unlignified cell walls (Anderson and Akin, 2008). More guaiacyl lignin was removed over syringyl lignin for herbaceous biomass (Akin et al., 1995) whereas more syringyl lignin removal was reported for woody biomass (Eriksson et al., 1990). Therefore, the extent and rate to which the total lignin was removed may be relative to types of lignin subunits. Similarly, straw materials, such as rice straw rich in p-coumeric units, were reported to be more resistant to C. subvermispora (Taniguchi et al., 2005), which could explain the lower lignin removal of wheat straw compared to some other herbaceous biomass such as corn stover and switchgrass. However, the quantity of acid soluble lignin and acid insoluble lignin as summarized in Table 1 did not necessarily reflect the degree of recalcitrance of feedstocks against the fungus. Hemicellulose, which differs mainly with respect to chemical composition, is also considered to be one of the potential factors that can affect digestibility. Hemicellulose of hardwood and herbaceous biomass mostly contains xylan whereas softwood hemicellulose mostly contains glucomannans (Saha, 2003). As stated above in regards to 85

104 composition of feedstocks, hemicellulose in soybean straw had a different sugar profile from other herbaceous biomass, containing more glucuronoxylan rather than arabinoxylan as well as a small amount of (gluco)mannan or galactoglucomannan, which was very similar to the hemicellulose in hardwood (Table 1). This similarity, however, did not result in successful degradation of soybean straw by C. subvermispora, even with extended pretreatment time and addition of various nutrients as described later. The other components of hemicellulose matrix, such as acetylated group of ferulate, may also partially contribute to the rigid structure of soybean straw and make it resistant to fungal attack. Apart from structural material in biomass, extractives in some feedstock may impose a barrier to fungal attack. As shown in our later study, hot water extraction noticeably improved the fungal degradation of wheat straw but not that of soybean straw. Additional studies are required to elucidate the exact mechanism behind poor degradation of soybean straw and wheat straw Ligninolytic enzymes Oxidative enzyme systems, mainly lignin peroxidase (LiP) and manganese peroxidase (MnP), and laccase, are responsible for delignification of biomass feedstocks by white rot fungi (Eriksson et al., 1990). The findings that C. subvermispora lacks LiP activity in either liquid culture or solid culture were well-documented in prior studies (Rajukumar et al., 1996; Ruttimann-Johnson et al., 1993) as well as in our previous study as reported in Chapter 4. Thus, only MnP and laccase activity were investigated in this 86

105 study. As shown in Fig. 2, both MnP and laccase were detected during pretreatment of corn stover, switchgrass and wood. MnP was the predominant ligninolytic enzyme in wood, which was in agreement with other studies for C. subvermispora degradation on woody biomass (de Souza-Cruz et al., 2004; Vicentim and Ferraz, 2007). In contrast to wood, laccase with activity of 2.25 U/g solid was the predominant enzyme in switchgrass, which was different from C. subvermispora culture growing on sugarcane bagasse (Costa et al., 2005). Indeed, enzyme activity dynamically changed with cultivation time as observed during the 42-d decay process of corn stover (Wan and Li, 2010b), where MnP was prevailing in the early stage of cultivation while laccase became dominant after 22-d cultivation. There was a low level of laccase but no MnP detected in wheat straw, which was consistent with the less effectiveness of fungal degradation of wheat straw and indicated that lignin degradation is highly dependent on ligninolytic enzyme production. This result was further confirmed by the findings from the degradation process of soybean straw in which no lignin-degrading enzymes were detected after 18 d and, during that time, no significant degradation was caused by the fungus. The profile of ligninolytic enzyme secretion by C. subvermispora during cultivation on different substrates may be a reflection of specific fungal-substrate interactions. The enzyme production probably depends on the presence of nutritional sources, inducers, and intermediates available from the biomass materials under decomposition (Kapich et al., 1999). 87

106 Enzymatic hydrolysis After fungal pretreatment, feedstocks were subjected to enzymatic hydrolysis with the yields depicted in Fig.5.3. Control samples, which were incubated with fungal inoculation, gave similar sugar yields to that of raw materials, indicating that minor lignin degradation of control samples as a result of autohydrolysis or autoclaving did not lead to improved cellulose digestibility. In contrast, as a result of effective fungal pretreatment by C. subvermispora, the glucose yields of corn stover, switchgrass and hardwood were increased markedly, reaching 56.50, 37.15, and 24.21%, respectively, which were a fold increase over that obtained from the corresponding raw material. The more lignin that was degraded, the more cellulose digestibility was observed. Not surprisingly, the glucose yield of soybean straw was not improved after fungal pretreatment since no fungal growth was observed in soybean straw. Although wheat straw was slightly degraded by C. subvermispora during the 18-d pretreatment, the cellulose digestibility was not significantly improved (P > 0.05). It was interesting to note that raw soybean straw was less resistant to cellulase hydrolysis and reached the highest cellulose digestibility of all the raw feedstocks tested, which was close to the glucose yield of fungal-treated switchgrass and apparently higher than that of fungal-treated hardwood. The high cellulose digestibility of raw soybean straw may be due to more amorphous cellulose existing in cell walls (Bertran and Dale, 1985). Xylose yield, as shown in Fig.5.3b, followed a similar pattern to glucose yield. Although xylose yields of effectively treated feedstocks (corn stover, switchgrass, and hardwood) were increased by 88

107 more than 2 folds over that of raw materials, they were still lower compared to glucose yield, mostly due to a limited xylanase activity in Spezyme CP (Jensen et al., 2010). Fig.5.4 shows the effect of pretreatment time on sugar yields. The extension of pretreatment time to 35 d led to a 20-30% increase in glucose yields for hardwood and switchgrass, which reached 54.22% and 59.13%, respectively. Although corn stover had the highest glucose yield after 35-d pretreatment, only 8% of the increase was attributed to prolonged pretreatment time. Moreover, xylose yield of corn stover leveled off after 35-d pretreatment, probably due to less susceptibility of remaining xylan in the cell wall to enzymatic hydrolysis. For wheat straw, a small increase (about 2%) in glucose yield and about 10% increase in xylose yield were resulted from a prolonged pretreatment period of 35 d, which indicated that C. subvermispora activity might have increased slowly in wheat straw. The increased degradation seemed more favorable to xylose digestibility. Soybean straw appeared to be the most resistant feedstock as no improvement on glucose yields was observed even after 35 d of fungal pretreatment Effect of nutrients and inducers Culture media, such as nutrients, buffer, inducers (e.g., Mn 2+, H 2 O 2, and veratryl alcohol), have important roles in growth and ligninolytic enzyme production of white rot fungi (Michel et al., 1990; Ruttimann-Johnson et al., 1993; Messner et al., 1998). As mentioned above, raw wheat straw and soybean straw were resistant to fungal decay (Fig. 2-4). Thus, with an attempt to improve fungal performance, carbon/nitrogen sources, 89

108 enzyme inducers (Mn 2+ and H 2 O 2 ), and a buffer (sodium acetate buffer) were added to wheat straw and soybean straw. The results of the subsequent enzymatic hydrolysis of pretreated samples are shown in Fig.5.5. Compared to the control (without fungal inoculation), acetate buffer (50 mm, ph 4.8) addition did not result in improvement in glucose yield of wheat straw whereas water as the medium slightly, but not significantly (P > 0.05), improved glucose yield over that of the control (Fig.5a). These results indicated that the natural ph of wheat straw might be more important to fungal pretreatment. Inducer addition seemed to have no effect on improvement of sugar yields of wheat straw regardless of water and acetate buffer as solution. Glucose, similar to malt extract, as an external carbon, significantly improved fungal colonization and decay of wheat straw. The resulting sugar yields were increased by 10% over the control. However, corn steep liquor, a low cost carbon/nitrogen source which has been used for biopulping of woody biomass by C. subvermispora (Messner et al., 1998), only marginally increased the sugar yield. Therefore, corn steep liquor, originally containing about 6% carbohydrate and 25-40% protein, vitamins, and minerals (Anon, 1975), seemed less favorable to fungal growth compared to glucose or malt extract, indicating that the fungus preferred to assimilate all readily available carbon source for its growth and metabolism. On the other hand, it suggested that nitrogen-sufficient media did not necessarily improve the delignification process of C. subvermispora. This is consistent with the study by Ruttimann-Johnson et al. (1993) that the carbon source rather than the nitrogen source was the limiting factor of lignin mineralization of C. subvermispora. It can be concluded that carbon source and 90

109 inducer supplements were, to some extent, favorable to fungal pretreatment of wheat straw. In contrast, the sugar yields of soybean straw with all tests were not significantly improved (P>0.05), further approving a strong recalcitrance of soybean straw to C. subvermispora degradation Conclusions Among various feedstocks tested in this study, corn stover, switchgrass and wood were effectively degraded by C. subvermispora. The resulting sugar yields were 2-3 fold greater than those of raw materials and further increased with extended pretreatment time. In contrast, even with the addition of external carbon sources and enzyme inducers to the culture media, soybean straw was still most resistant to fungal pretreatment whereas wheat straw was less resistant. To increase the robustness of fungal pretreatment to handle a wide range of substrate sources, combined pretreatment processes would be required to facilitate the fungal degradation on the most resistant feedstocks. 91

110 Table 5.1. Composition of raw feedstocks. Composition (%) a Corn stover Switchgrass Wheat straw Soybean straw Hardwood Holocellulose 61.42± ± ± ± ±0.80 Cellulose 38.48± ± ± ± ±0.37 Hemicellulose 22.95± ± ± ± ± Xylan 18.52± ± ± ± ± Galactan 1.60± ± ± ± ± Arabinan 2.82± ± ± ± ± Mannan 0.00± ± ± ± ±0.14 Lignin 20.18± ± ± ± ±0.10 Acid-insoluble 18.65± ± ± ± ±0.17 Acid-soluble 1.53± ± ± ± ±0.07 Extractives 7.64± ± ± ± ±0.46 Water extractives 6.30± ± ± ± ±0.45 Ethanol extractives 1.31± ± ± ± ±0.01 Ash 3.82± ± ± ± ±0.29 Others 6.96± ± ± ± ±1.55 a data reported as mean ± S.D. of three measurements and based on dry matter. 92

111 (a) (b) Figure 5.1. Degradation of feedstocks after 18-d fungal pretreatment by C. subvermispora: (a) fungal pretreatment and (b) control (without fungal inoculation). 93

112 * * * Figure 5.2. Ligninolytic enzyme production during 18-d fungal pretreatment by C. subvermispora. *denotes that no enzyme was detected. 94

113 (a) (b) Figure 5.3. Enzymatic hydrolysis of feedstocks after 18-d pretreatment by C. subvermispora: (a) glucose yield and (b) xylose yield. 95

114 Figure 5.4. Effect of pretreatment time on enzymatic hydrolysis of feedstocks. 96

115 (a) (b) Figure 5.5. Effect of addition of nutrients and additives to fungal pretreatment on enzymatic hydrolysis of wheat straw (a) and soybean stalk (b). CTL: control; AA: acetate buffer (ph=4.8); ME: malt extract; CSL: corn steep liquor. 97

116 Chapter 6 Enhanced Enzymatic Hydrolysis of Biomass Feedstocks by Combined Hot Water Extraction/Hydrothermal and Fungal Pretreatment Exhaustive hot water extraction (HWE) and liquid hot water (LHW) pretreatment were studied for their effects on fungal degradation of biomass feedstocks (i.e., corn stover, wheat straw, and soybean straw). HWE (85 ºC for 10 min) partially removed water soluble extractives and significantly facilitated the fungal degradation of wheat straw. The lignin removal reached 24.99% and the resulting glucose yield was 43.43%. In contrast, for corn stover and soybean straw, HWE had less or no effect on improvement of cellulose digestibility. Further study was performed to explore the fungal biodegradability of soybean straw by combining it with LHW pretreatment (170 ºC for 3 min) using corn stover as a reference feedstock. It was observed that LHW pretreatment eventually improved the fungal degradation of soybean straw by C. subvermispora and the combined pretreatment led to 36.70% lignin removal and 64.25% glucose yield. However, corn stover was little affected by this combined pretreatment. The lignin structural and chemical changes during LHW pretreatment, rather than lignin removal, most likely contributed to reduced recalcitrance of soybean straw. Our studies indicated that combined LHW pretreatment and fungal pretreatment was effective for improving cellulose digestibility of soybean straw. 98

117 6.1. Introduction Microbial pretreatment with white rot fungi under solid state fermentation has been shown to effectively improve enzymatic hydrolysis of various biomass feedstocks (e.g., hardwood, corn stover, and switchgrass). Its advantages include low energy input, little or no use of chemicals, and no waste stream output. However, microbial pretreatment has inherent problems, such as long pretreatment time and simultaneous degradation of carbohydrate and lignin. Our previous study showed that some biomass feedstocks, such as soybean straw and wheat straw, were strongly resistant to fungal degradation, even with addition of nutrients and enzyme inducers. Therefore, a fungal pretreatment process combined with physical and chemical pretreatment methods is preferred to improve the performance of fungal pretreatment. In addition, the combined pretreatment can potentially overcome the problems associated with physical and chemical pretreatment, such as intensive energy use, severe pretreatment conditions, high chemical loading, and toxicity of the waste stream. These benefits have been reviewed in biopulping industries, where fungal pretreatment with white rot fungi is widely applied with mechanical or chemical pulping (Akhtar et al., 1998; Kang et al., 2003). Most importantly, the combined pretreatment process could result in a synergic effect, improving the yields of end products. Several studies have investigated a combination of fungal pretreatment with one of the other types of pretreatment methods for biomass. The results showed combined pretreatments improved enzymatic hydrolysis and biofuel production when compared to a single pretreatment process. Fungal pretreatment with C. subvermispora followed by 99

118 ethanolysis was found to remarkably increase ethanol yield of woody biomass over ethanolysis alone (Baba et al., 2010; Itoh et al., 2003). Similarly, fungal pretreatment followed by steam explosion further increased saccharification by about 10% (Sawada et al., 1995). Yu et al. (2008) tested fungal pretreatment of ultrasonic- or H 2 O 2 -pretreated rice hull and found that the combined pretreatment resulted in increased lignin degradation and enzymatic hydrolysis yield. Kadimaliev et al. (2003) found that modification of birch and pine sawdust by ultrasound accelerated lignin degradation with P. tigrinus. The combination of fungal pretreatment of corn stover with mild alkaline pretreatment was also shown to enhance delignification and xylan removal as well as glucose yield (Yu et al., 2010). The ethanol yield of water hyacinth pretreated with combined fungal and dilute acid pretreatment was 1-2 fold higher than that obtained from dilute acid pretreatment only (Ma et al., 2010). Liquid hot water (LHW) pretreatment has been proposed to be one of the current leading pretreatment methods for improving cellulose digestibility of lignocellulosic materials (Mosier et al., 2005b). This method, generally conducted at elevated temperatures (120 to 260 ºC), is regarded as environmental-friendly pretreatment process because no chemicals are used (Kim and Mosier, 2009; Weil et al., 1998). Water itself and acetyl groups in hemicelluloses, which act as acid at around 200 ºC, are believed to catalyze extensive hydrolysis of hemicellulose to its component sugars, primarily xylose. The effectiveness of LHW pretreatment on cellulose digestibility is strongly related to pretreatment severity. On the other hand, severe pretreatment conditions, resulting in accumulation of organic acids and subsequent acidic environment, cause degradation of 100

119 monomeric sugars present in the liquid fraction to compounds inhibitory to ethanol fermentation (e.g., hydroxymethyl furfural (HMF), furfural, formic acid, levulinic acid) (Weil et al., 1998). The ph controlled LHW pretreatment can prevent degradation of fermentable sugars but involves the use of a base (Mosier et al., 2005b). To eliminate use of chemicals while minimizing the degradation of fermentable sugars, LHW pretreatment conducted at less severe conditions can be followed by other pretreatment methods. Some synergistic effects were reported by Inoue et al. (2008) that ball milling combined with hot-compressed water treatment of Eucalyptus saved energy for pretreatment and enzyme loading for enzymatic hydrolysis. Combination of LHW pretreatment with fungal pretreatment has not been reported. It was the goal of this study to investigate the synergic effect of hydrothermal pretreatments and fungal pretreatment on degradation of biomass. Two uncatalyzed hydrothermal methods, exhaustive HWE and LHW pretreatment, were compared. Their effects on modification of the chemical compositions of plant cell walls and improvement on glucose yields were investigated Materials and methods Raw materials Crop residues including corn stover, wheat straw, and soybean straw were obtained from the Ohio Agricultural Research and Development Center (OARDC) farm in Wooster, Ohio. The feedstocks were oven-dried at 40 ºC and then ground in a Wiley 101

120 mill to pass through a 5 mm screen. The ground feedstocks were stored at room temperature prior to use Fungus and inoculum preparation C. subvermispora (ATCC 96608) was obtained from American Type Culture Collection (Manassas, VA) and maintained on malt extract agar (MEA, 2% w/v) at 4ºC. Inoculum preparation was the same as described in Section of Chapter Hot water extraction Biomass feedstocks were extracted with hot water at 85 ºC for 10 min with a solid to liquid ratio of 1:20. The mixture was stirred every 2 min. After extraction, the slurry was filtered and the solid fraction was collected for further use Liquid hot water pretreatment A 1-liter Parr reactor (Parr Instrument Company, Moline, IL) was used for LHW pretreatment of corn stover or soybean straw. Forty gram of biomass feedstock (dry mass) was mixed with DI water in the stainless vessel of the reactor to obtain the solid to liquid ratio of 1:10. The reactor was sealed and heated to 170 ºC within min with continuous agitation (approximately 400 rpm). After being held at 170 ºC for 3 min with pressure maintained at 110 psi, water was circulated around the tank to cool the reactor to room temperature within 30 min. The pretreated slurry was washed with 1000 ml DI water and separated by vacuum filtration using glass fiber filters (1.6 µm). The solid was collected for further use. 102

121 Fungal pretreatment Fungal pretreatment procedure was the same as described in Section of Chapter 4 and the pretreatment time was 18 d for all testes Enzymatic hydrolysis Spezyme CP was used for enzymatic hydrolysis and the procedure was the same as described in Section of Chapter Analytic method Compositional analysis and sugar measurement in hydrolysate followed the same methods as described in Section of Chapter 3. Dry mass loss, lignin degradation, cellulose loss, and hemicellulose loss were defined as the percentage of reduction of the corresponding fraction in biomass after pretreatment. The sugar yields were calculated by dividing the yield by the theoretical sugar yields of the raw feedstocks Results and discussion Effect of hot water extraction (HWE) Degradation Our previous study showed that fungal degradation on native wheat straw and soybean straw was lower compared to other feedstocks such as corn stover and switchgrass. Since extractives accounted for more than 10% in these two feedstocks (Table 5.1), it was our hypothesis that removal of extractives may facilitate fungal attack. 103

122 A simplified exhaustive HWE, therefore, was used to partially remove water soluble components from biomass. As expected, HWE did not modify the structure of biomass feedstocks. There was no cellulose and lignin degradation observed while hemicelluloses degradation was also negligible for all three feedstocks (Fig.6.1). The dry matter loss was attributed to the removal of water soluble materials. The dry matter loss (about 8%) from wheat straw and soybean straw was expected to be higher than for corn stover, which was similar to the extractives obtained by a Soxhlet extraction but with a lower amount due to incomplete extraction. Glucose and xylose in extractives were detected at a very low concentration and accounted for less than 1% total biomass. No attempt was made to distinguish other water soluble materials in extractives. The hot water extracted biomass feedstocks were subjected to fungal pretreatment. It was found that the HWE improved the fungal degradation on corn stover (Fig.6.1a). The dry matter loss caused by combined HWE and fungal pretreatment was 4% higher than that by the fungal pretreatment alone. Hemicellulose loss also increased by 9% as a result of combined pretreatment while there was no significant difference in lignin removal (P > 0.05). Compared to the strong resistance of native wheat straw to fungal pretreatment alone, hot water extracted wheat straw was much more degradable by C. subvermispora as more than 22% lignin and 10% hemicellulose was removed (Fig.6.1b). Apparently, HWE facilitated the fungal degradation of wheat straw, indicating that water soluble extractives in wheat straw, other than native structural components, may have imposed a protective barrier to fungal degradation. As the hot water only extracted hydrophilic compounds, the remaining hydrophilic extractives after water 104

123 extraction, such as wax, chlorophyll and resin, should not inhibit fungal growth, which is consistent with the findings that white rot fungi can efficiently degrade lipophilic extractives (Martinez-Inigo et al., 1999). C. subvermispora was reported to have a very little cellulase activity (Ferraz et al., 2003); thus, the cellulose loss, most likely due to non-enzymatic degradation, was limited for both pretreatments (fungal as well as HWE+ fungal). Less than 5% cellulose loss was observed with corn stover and was even lower with wheat straw. In contrast to wheat straw and corn stover, HWE did not improve the biodegradability of soybean straw (Fig.6.1c), suggesting that chemical-bound components were the most recalcitrant part of soybean straw. As mentioned in our previous study, hemicellulose in soybean straw that was similar to that in wood but quite different from that in other crop residues, may partially contribute to the recalcitrance Enzymatic hydrolysis Lignin is the major barrier in the plant cell wall to hydrolytic enzymes as it covalently links with hemicelluloses, providing a protective coat to protect carbohydrates from degradation (Anderson and Akin, 2008). Therefore, the more lignin degraded by the fungus, the more the susceptibility of cellulose is improved, which can be seen from Fig.6.1 and Fig.6.2. Since the lignin degradation of corn stover by fungal pretreatment and combined pretreatment was similar, the glucose yields were also similar. For HWE alone, it did not result in degradation of cell wall components and improvement on the glucose yield as it only removed water soluble extractives in feedstocks. Also, it was not surprising that the glucose yield of soybean straw was not improved because neither 105

124 fungal pretreatment nor combined pretreatment was effective to its lignin degradation. In contrast, the glucose yield of combined-pretreated wheat straw was 43.43% as a result of substantial lignin removal, which was 2-fold that of native wheat straw. However, the glucose yield of the fungal-pretreated wheat straw was only slightly higher than that of native wheat straw. The xylose yield of all native or extracted feedstocks followed a very similar pattern to their glucose yields. The xylose yield of extracted wheat straw was double that of native and fungal-pretreated wheat straw, which indicated that extractive removal facilitated lignin degradation by the fungus and subsequently increased the digestibility of cell wall carbohydrates. For a specific substrate, such as corn stover and wheat straw, fungal pretreatment or combined pretreatment might be able to improve the cellulose digestibility. Unfortunately, no improvement in glucose yield was observed since soybean straw was recalcitrant to fungal pretreatment regardless of HWE. Therefore, soybean straw was considered in need of further study to explore the mechanism behind the recalcitrance. Extracted wheat straw remarkably improved glucose yield; however, its yield was still lower than fungal-treated or combined-pretreated corn stover. As indicated in our previous study, the ability of C. subvermispora to depolymerize lignin in biomass feedstocks was very selective. Moreover, this fungus grew slowly in wheat straw and only improved about 4% cellulose digestibility after a 35 d pretreatment. A similar result was also reported in a study by Dorado et al. (1999), which showed that C. subvermispora had a slower lignin degradation rate on wheat straw during 30-d cultivation when compared to Pleurotus eryngii and Phanerochaete chrysosporium. Most 106

125 likely, the higher proportion of ester-linked ferulic and p-hyrolxycinanmic acid structure in straw materials (wheat straw or rice straw), rather than water extractives as discussed above, contributed to most of its recalcitrance to C. subvermispora (Sun et al., 2002; Taniguchi et al., 2005) Effect of liquid hot water (LHW) pretreatment Degradation LHW pretreatment (170ºC and 3 min) was tested to improve fungal pretreatment of soybean straw. Corn stover, which can be effectively degraded by C. subvermispora, was also tested in the same conditions as a reference. As shown in Fig.6.3, the effect of LHW by itself on degradation of both feedstocks was not significant. Only a minor hemicellulose loss (about 0.9%) was observed but no cellulose and lignin degradation was observed for both feedstocks, which was similar to HWE. It also suggested that the dry mass loss (about 12%) of LHW pretreatment was mainly from extractives and ashes. Differing from HWE, LHW pretreatment significantly facilitated the fungal pretreatment, in particular for lignin degradation, of both substrates. Cellulose degradation from combined pretreatment for both substrates was still limited. The color of LHW pretreated soybean straw and corn stover turned from dark to yellow, which also indicated an effective fungal pretreatment. For corn stover, the combined pretreatment caused 13% more lignin degradation than the fungal pretreatment only. However, hemicellulose degradation was also considerable and the fungus consumed more than 20% of the hemicellulose from the LHW pretreated corn stover than from the untreated corn stover. 107

126 The significant loss of lignin and hemicelluloses mostly accounted for the doubled dry mass loss of the combined pretreatment. For soybean straw, the lignin degradation of 36.70% along with the 41.42% hemicellulose degradation resulted from combined pretreatment while there was no degradation taking place during fungal pretreatment alone. The dry mass loss was also substantial, reaching 21.88% after the combined pretreatment. It is known that in hydrothermal or thermo-chemical pretreatment fragmented lignin reacts by itself or with carbohydrate oligomers to form larger molecules, called pseudo-lignin that precipitate in solid residues (Lora et al., 1978; Bobleter and Concin, 1979; Liu and Wyman, 2003). This phenomenon could explain why no degradation of lignin occurred during LHW pretreatment in this study. Regardless of actual lignin and hemicellulose degradation caused by LHW pretreatment, the cell wall structure of soybean straw may have been modified by LHW pretreatment, which noticeably reduced its recalcitrance to C. subvermispora decay. It can be concluded that LHW pretreatment under less severe conditions were sufficient to modify physical and chemical properties of corn stover and soybean straw to facilitate the fungal degradation Enzymatic hydrolysis Fig.6.4 shows sugar yields of native and pretreated feedstocks obtained by enzymatic hydrolysis. LHW pretreatment alone significantly increased glucose yield of soybean straw and corn stover by 8-12%. In contrast to a good correlation between lignin removal and cellulose digestibility observed from fungal pretreatment in our study 108

127 reported in the previous chapters, the glucose yield of LHW pretreated samples did not appear to be related to lignin removal as virtually no lignin was removed by LHW pretreatment. The poor relationship is common in hydrothermal or thermal chemical pretreatment mostly likely due to condensation and precipitation of dissolved lignin and carbohydrate oligomers as mentioned above (Teramoto et al., 2009; Liu and Wyman, 2003). Nevertheless, compared to native materials, the condensed and precipitated solids appear to be less resistant to hydrolytic enzymes (Teramoto et al., 2009). Our observations also suggested that rearrangement of the lignin structure occurred during LHW which in turn made the cellulose more susceptible to enzymatic hydrolysis. As LHW pretreatment facilitated fungal pretreatment to a great extent, the resulting cellulose digestibility of soybean straw was improved by about 30% over native soybean straw, reaching 64.25% based on theoretical yield of raw feedstock. Both LHW and fungal pretreatment did not involve the use of chemicals and thus are environmental friendly. This synergic effect of the combined pretreatment process was significant for soybean straw as this process may provide an alternative pretreatment method to such carbohydrate-rich feedstock. The glucose yield obtained in this study was 64.24%, which was higher than that obtained by ammonia soaking (Xu et al., 2007) and comparable to that obtained with sodium hydroxide soaking or a single LHW pretreatment under severe conditions (200 ºC and 10 min) (Wan et al., 2011). On the other hand, it is worth further study to investigate whether a LHW pretreatment condition less severe than the current condition (170 ºC and 3 min) is sufficient to facilitate fungal degradation. 109

128 Combined pretreatment of corn stover did not prove to be more efficient than fungal pretreatment alone. Unexpectedly, a reduced glucose yield was observed. Moreover, the glucose yield based on the theoretical yield of LHW pretreated corn stover was not much different from the yield based on raw material. Most likely, the more effective fungal pretreatment masked the effect of LHW pretreatment. Similar results have been reported by Hatakka (1983), who found that combined fungal and alkali pretreatment did not improve saccharification of wheat straw, probably because strong alkali pretreatment (0.4g NaOH/g straw) masked the effect of the fungi pretreatment. In contrast, a combination of biological pretreatment with mild chemical and physical pretreatments has proven to be more effective (Yu et al., 2008; Yu et al., 2010; Ma et al., 2010). On the other hand, those studies as well as the present study suggested that biomass feedstocks suitable for combined pretreatment should be resistant to fungi treatment to some extent. One of other potential reasons for reduced effectiveness of the combined pretreatments was the limited effect of lignin removal on improvement of cellulose digestibility. In other words, further lignin removal, as observed in Fig.6.3a, was not correlated to a further increase of cellulose digestibility, which suggests that the cellulose itself, after considerable lignin removal (41.99% in this case) became a limiting factor for enzymatic hydrolysis. Xylose yield had a similar pattern to glucose yield (Fig.6.4). The native corn stover and soybean straw had xylose yields of 12-13%. LHW pretreatment improved xylose yield about 13% for both feedstocks. However, combined pretreatment reduced the xylose yield of corn stover compared to the fungal pretreatment alone. Hemicellulose 110

129 in LHW pretreated corn stover was more readily accessible by the fungus as hemicellulose degradation of LHW pretreated corn stover by the fungus was 20% higher than of the untreated corn stover. On the other hand, the xylan residue of LHW pretreated corn stover after fungal pretreatment may be less digestible by hydrolytic enzymes. Unlike corn stover, LHW pretreated soybean straw had a further increased xylose yield which was 26% higher than that of the LHW pretreatment only. The xylan degraded during fungal pretreatment and hydrolyzed enzymatic hydrolysis accounted for 92.83% of total xylan in soybean straw while only 71.02% of total xylan in corn stover. This result indicated that xylan remaining in soybean straw after combined pretreatment was more digestible than in corn stover Conclusions HWE significantly improved the fungal degradation of wheat straw while it was less effective for corn stover and not effective for soybean straw. The resulting sugar yield of the combined HWE and fungal-treated wheat straw was doubled compared to the native wheat straw, which indicated that water extractives in wheat straw could partially contribute to recalcitrance to C. subvermispora. LHW pretreatment facilitated fungal degradation of soybean straw. The lignin degradation of the combined LHW and fungal pretreatment was comparable to that of corn stover. However, the resulting glucose and xylose yields were significantly higher than that of corn stover. Therefore, it can be stated that hydrothermal pretreatment had a synergistic effect on the fungal pretreatment of the most resistant feedstocks. 111

130 (a) (b) 112

131 (c) Figure 6.1. Synergistic effects of hot water extraction and fungal pretreatment on degradation of biomass feedstocks: (a) corn stover, (b) wheat straw, and (c) soybean straw. HWE: hot water extraction; FT: fungal pretreatment; HWE+FT: combined hot water extraction with fungal pretreatment. 113

132 (a) (b) Figure 6.2. Synergistic effects of hot water extraction and fungal pretreatment on enzymatic hydrolysis of biomass feedstocks: (a) glucose yield and (b) xylose yield. HWE: hot water extraction; FT: fungal pretreatment; HWE+FT: combined hot water extraction with fungal pretreatment. 114

133 (a) (b) Figure 6.3. Synergistic effects of liquid hot water pretreatment and fungal pretreatment on degradation of biomass feedstocks: (a) corn stover and (b) soybean straw. LHW: liquid hot water pretreatment; FT: fungal pretreatment; LHW+FT: combined liquid hot water pretreatment with fungal pretreatment. 115