Anaerobic Digestion of Submerged Macrophytes

DISSERTATION

ANAEROBIC DIGESTION OF SUBMERGED MACROPHYTES

—BIOCHEMICAL APPROACH FOR ENHANCING THE METHANE

PRODUCTION—

2016

SOKA UNIVERSITY GRADUATE SCHOOL OF ENGINEERING

MITSUHIKO KOYAMA

ANAEROBIC DIGESTION OF SUBMERGED MACROPHYTES

—BIOCHEMICAL APPROACH FOR ENHANCING THE

METHANE PRODUCTION—

March 2016

MITSUHIKO KOYAMA

SOKA UNIVERSITY

Author: Mitsuhiko Koyama Title: Anaerobic digestion of submerged macrophytes

—Biochemical approach for enhancing the methane production—

Department: Environmental Engineering for Symbiosis

Faculty: Engineering

Degree: Ph.D.

Convocation: March 2016

Permission is herewith granted to Soka University to circulate and copy for non- commercial purposes, at its discretion, the above title request of individual or institutions.

We certify that we have read this dissertation and that, in our opinion, it is satisfactory in scope and quality as a dissertation for the degree of Doctor of Philosophy in Engineering

March 2016

DISSERTATION COMMITTEE

Prof. Dr. Tatsuki Toda

Prof. Dr. Shuichi Yamamoto

Prof. Dr. Kiyohiko Nakasaki

Contents

page ACKNOWLEDGMENTS i

ABSTRACT iii

Chapter I GENERAL INTRODUCTION 1.1. Emerging issues of aquatic macrophytes 1 1.2. Harvesting of submerged macrophytes and treatment of the harvested phytomass 2

1.3. Anaerobic digestion of lignocellulosic biomass and enhancement of CH4 recovery 5 1.4. Objectives 7 Tables 9 Figures 10

ChapterⅡ CHEMICAL COMPOSITION AND ANAEROBIC DIGESTIBILITY OF FIVE SUBMERGED MACROPHYTE SPECIES DOMINANT IN LAKE BIWA

2.1. Introduction 16 2.2. Materials and methods 17 2.2.1. Substrates and inoculum 17 2.2.2. Batch anaerobic digestion of submerged macrophytes 18 2.2.3. Analytical parameters 19 2.2.4. Calculation 20 2.3. Results and discussion 21 2.3.1. Chemical composition of submerged macrophytes 21

2.3.2. CH4 production rate and CH4 yield 23

2.3.3. Relationship between CH4 recovery and lignocellulose content 25 2.4. Conclusions 27 Tables 29 Figures 33

ChapterⅢ EFFECT OF ALKALINE PRE-TREATMENT ON ANAEROBIC DIGESTIBILITY OF LIGNIN-RICH SUBMERGED MACROPHYTE

3.1. Introduction 39 3.2. Materials and methods 43 3.2.1. Substrates and inoculum 43 3.2.2. Hydrolysis test of submerged macrophytes using alkaline pre-treatment 43 3.2.3. Batch anaerobic digestion test of pre-treated macrophytes 44 3.2.4. Methanogenic and hydrolysis toxicity test of dissolved lignin 45

3.2.4.1. Lignin extraction and purification 45 3.2.4.2. Methanogenic and hydrolysis toxicity test 45 3.2.5. Analytical parameters 46 3.2.6. Calculations 48 3.2.7. Kinetic models 50 3.2.8. Statistical analysis 51 3.3. Results and discussion 51 3.3.1. Chemical component change of submerged macrophytes by alkaline pre-treatment 51 3.3.1.1. Hydrolysis efficiency 51 3.3.1.2. Lignocellulose composition 52 3.3.1.3. Lignin composition 55 3.3.2. Anaerobic digestion of thermochemically pre-treated submerged macrophytes and possible inhibitor 56 3.3.3. Influence of dissolved lignin on anaerobic digestion process 59 3.3.3.1. Chemical characteristics of purified dissolved lignin 59 3.3.3.2. Influence on methanogenesis 59 3.3.3.3. Influence on acidogenesis and hydrolysis 61 3.4. Conclusions 66 Tables 67 Figures 76

ChapterⅣ EFFECT OF ALKALINE PRE-TREATMENT ON SEMI-CONTINUOUS ANAEROBIC DIGESTION OF SUBMERGED MACROPHYTE

4.1. Introduction 92 4.2. Materials and methods 94 4.2.1. Substrates and inoculum 94 4.2.2. Semi-continuous anaerobic digestion of alkali pre-treated or un-treated P. maackianus 94 4.2.3. Analytical parameters 95 4.2.4. Fractionation of soluble organics and estimation of dissolved lignin/ humic substances in the digestate 96 4.2.5. Calculations 96 4.3. Results and discussion 97 4.3.1. pH and salinity variation in the digestate 97 4.3.2. Accumulation of dissolved lignin in the digestate 98 4.3.3. Effect of dissolved lignin on methane recovery in semi-continuous operation 99 4.3.3.1. Performance of semi-continuous anaerobic digestion of alkali pre-treated and un-treated P. maackianus 99

4.3.3.2. Inhibition and acclimatization against dissolved lignin 100 4.4. Conclusions 103 Table 105 Figures 106

ChapterⅤ GENERAL DISCUSSION

5.1. Economic and thermal balance assessment of alkaline pre-treatment 112 5.1.1. Scenario design 113

5.1.2. Recovery of CH4, electricity and thermal energy 115 5.1.3. Economic and thermal balance 116 5.1.3. Effect of electricity tariff of economic balance 116 5.1.3. Effect of species composition on economic and thermal balance 117 5.2. Further studies for more efficient anaerobic digestion of submerged macrophytes 118 5.2.1. Lignin recovery from alkali pre-treatment liquor prior to anaerobic digestion 118 5.2.2. Effect of seasons on the digestibility of submerged macrophytes 119 Tables 121 Figures 123

REFERENCES 127

ACKNOWLEDGEMENTS

I would like to express my deepest gratitude to my supervisor Professor Tatsuki Toda for all the guidance throughout my Ph.D. study and for his strict but warm hearted encouragement rendered in the completion of this dissertation. I sincerely appreciate for his tremendous academic support, as well as for giving me a lot of wonderful opportunities throughout this six years in his laboratory, especially Lake Biwa Project.

I am also deeply thankful to my co-supervisor Professor Shuichi Yamamoto and Professor

Kiyohiko Nakasaki for taking so much time for my dissertation and for their insightful comments and suggestions regarding the present study.

I am grateful to Professor Syuhei Ban for his advice and guidance, especially during the application of research grant and implementation of research project. I appreciate to Dr. Kanako

Ishikawa for the knowledge about submerged macrophytes and for heling harvest of macrophytes in

Lake Biwa. I deeply appreciate to Mr. Chiaki Niwa for his professional advices and helpful suggestions, especially during experimental set-up and data analysis. I am thankful to Professor

Satoru Taguchi, Professor Norio Kurosawa, Professor Tatsushi Matsuyama, Professor Junichi Ida,

Associate Professor Shinji Sato, Dr. Norio Nagao, Dr. Keiko Watanabe, Dr. Minako Kawai, Dr. Kenji

Tsuchiya and Dr. Shinichi Akizuki for their invaluable advice and support.

I am sincerely grateful to Mr. Yuichi Ezawa for his guidance, support and invaluable discussion throughout the first three years of the laboratory life, especially during my experiments, writings and presentations. I also appreciate to Mr. Koichi Izumi and Mr. Nobuyuki Nakatomi for advices during writings and presentations. I deeply thank to all my seniors and friends belong to Laboratory of

Restoration Ecology for their support, encouragement and friendship. Special thanks goes to Ms.

Hiroyo Takee, Ms. Keiko Mizubayashi, Mr. Kiyoshi Nakamura, Mr. Masatoshi Kishi and Ms. Yukiko

Sasakawa, for their encouragement and sharing the invaluable time throughout three to six years. I

am also grateful to Ms. Hiromi Joo, Mr. Nobuo Nakahashi, Mr. Kazuhiro Kaneda, Mr. Hideyuki Taira,

Mr. Masaaki Fujiwara, Ms. Midori Goto, Mr. Naoki Mizuno and Mr. Tang Jau Jiun for assistance of my experiments.

Many thanks to “University-Industry Joint Research” Project for Private Universities: matching fund subsidy from MEXT (Ministry of Education, Culture, Sports, Science and Technology) 2009-

2013, “Environment Research and Technology Development Fund” from the Ministry of the

Environment, Japan (4-1406), and Grant-in-Aid for Japan Society for the Promotion of Science

(JSPS) Fellows (15J12677) for funding this research. I also appreciate to Hokubu Sludge Treatment

Center, Lake Biwa Policy Division and Ohmi Environment Conservation Foundation for the donation of materials for the present study.

I would love to thank my parents and sisters and friends for their support, trust and understanding. I could never have done this work without their support, understanding and love.

Finally, I am sincerely grateful to the founder of Soka University, Dr. Daisaku Ikeda for his sincere trust, great expectation, heartwarming guidance and tremendous encouragement.

ABSTRACT

In recent decades, aquatic macrophytes including floating, emergent and submerged macrophytes have been excessively propagated and causing various environmental problems in lakes, dams and reservoirs worldwide. In the case of Japan, submerged macrophytes have been propagated and causing water stagnation, foul odor, fishing interference and landscape fouling. The treatment of harvested macrophytes is one of the largest concern, but the effective and low-cost treatment has not been established yet. Anaerobic digestion enables the recovery of bioenergy as methane gas from biomass with high moisture content. From this point of view, the application of anaerobic digestion has been receiving its attention for the effective alternative treatment. Anaerobic digestion has been employed to the treatment of wide variety of bio-wastes such as wastewater, food waste, animal manure, agriculture waste, and sewage sludge in commercial scale. In recent years, lignocellulosic biomass has been expected for the prospective feedstocks for biomethane recovery. Substrates with slow degradation rate and/or low methane recovery are generally regarded as unsuitable for anaerobic digestion. Hence, the improvement of biomass biodegradability is required for lignocellulosic biomass such as submerged macrophytes, in order to successfully apply anaerobic digestion treatment.

The overall research aim of this Ph.D. thesis is to examine anaerobic digestion process of submerged macrophytes and to enhance the anaerobic digestibility by chemical pre-treatment. First, anaerobic digestibility of five dominant submerged macrophyte species in relation to the chemical composition was investigated. The lignin content of the submerged macrophyte widely ranged depending on species from 3 to 21%-total solid. The total methane yield of submerged macrophytes greatly varied from 161 to 361 mL g-VS-1 depending on species (El. nuttallii > Eg. densa > P. malaianus > C. demersum > P. maackianus). The present study revealed that the methane recovery of submerged macrophytes is regulated by the lignin content, as well as other lignocellulosic biomass. For the lignin-rich macrophyte (i.e. P. maackianus), the application of delignification pre-treatments is iii

recommended to make this species feasible for anaerobic digestion. Lignin composition analysis revealed that submerged macrophytes contained alkali-labile hydroxycinnamic acids for 27.2 to

59.4% of the total lignin, suggesting alkaline pre-treatment can easily delignify and improve the anaerobic digestibility of submerged macrophytes. Next, the effect of alkaline pre-treatment on anaerobic digestibility of submerged macrophyte was investigated in batch mode. Alkaline pre- treatment greatly increased the SCOD/TCOD and reduced lignin of submerged macrophytes, clearly indicating the enhancement of the hydrolysis. Ferulic acid content greatly reduced with the increase of NaOH dose, implying the removal of lignin-ferulate complex from the lignocellulose. The cumulative CH4 yield of pre-treated P. maackianus and Eg. densa was 50.6% and 23.9% higher than un-treated, indicating that alkaline pre-treatment could be effective for anaerobic digestion of submerged macrophytes. The present study suggested that the pre-treatment with high NaOH dose to lignin-rich substrate could cause the inhibition of anaerobic digestion, possibly by the release of dissolved lignin from lignocellulose during pre-treatment. Anaerobic toxicity test of dissolved lignin revealed that methanogenesis and acidogenesis steps were slightly inhibited (7-15% decline from the control), while hydrolysis step was inhibited by 35%, suggesting that extracellular enzymatic hydrolysis may be the most susceptible step against dissolved lignin in anaerobic digestion process.

Finally, semi-continuous anaerobic digestion of alkali pre-treated and un-treated P. maackianus was conducted in order to evaluate the process stability in terms of CH4 recovery and acclimatization to accumulating dissolved lignin. Alkaline pre-treatment enhanced the CH4 production from P. maackianus by 65% from un-treated biomass. In the digestate of alkali pre-treated condition, accumulation of dissolved lignin was observed and the CH4 production was inhibited. However, the

CH4 production was recovered from day 42, which suggested the acclimatization of microorganisms and/or alleviation of toxicity of dissolved lignin. Overall, the present study demonstrated that alkaline pre-treatment is highly beneficial for enhancing anaerobic digestibility of lignin-rich submerged macrophyte.

Chapter I

General Introduction

1.1. Emerging issues of aquatic macrophytes

Generally, aquatic macrophytes (emergent, floating and submerged) play important roles in freshwater ecosystem by maintaining good water quality and providing habitats to fish juveniles or invertebrates. However, for the last several decades, aquatic macrophytes has been excessively propagating and causing various problems in lakes, dams and reservoirs worldwide. In many African countries, Food and Agriculture Organization of the United Nations (FAO) reported that the exotic floating macrophytes such as water hyacinth Eichhornia crassipes, water lettuce Pistia stratiotes and water fern Salvinia molesta have been invaded and infested dramatically since early 1950s (Food and agriculture organization of the United Nations, 2002). These floating macrophytes develops dense mats on the water surface, which have been enormously damaged to the environment and the economy by causing water loss through evapotranspiration, fishing interference, dam clogging, flooding, displacement of native plants and animals and interference of irrigation. In Europe, invasion of 96 exotic species of aquatic macrophyte such as Elodea Canadensis, Azolla filiculoides, Vallisneria spiralis and Elodea nuttallii has been reported in 46 out of 50 European countries (Hussner, 2012).

One of the major issues of infested macrophyte is the release of nutrients from decayed macrophyte during autumn in short period of time (van Donk et al., 1993). In Australia, exotic macrophytes such as Ei. crassipes, Cabomba caroliniana and S. molesta have been infested in South East Queensland, and being problematic in increase of mosquito breeding sites, interference with recreational activities and water loss by increased evaporation rate (O’Sullivan et al., 2010). In USA, invaded submerged macrophytes including Egeria densa, Ceratophylum demersum, Potamogeton crispus and

Myriophyllum spicatum have been causing similar problems mentioned above (Santos et al., 2011).

Many Asian countries also have reported the nuisance of aquatic macrophytes in many water bodies such as rice paddies and drainage systems (Abbasi et al., 1990).

Japan also has been suffering the problem of aquatic macrophytes, especially in Lake Biwa, where is the largest lake in Japan (total area: 674 km2). As shown in Figure 1-1, Lake Biwa consists of the Northern Basin (surface area: 617.6 km2; average depth: 45.5 m) and the Southern Basin

(surface area: 56.8 km2; average depth: 3.5 m). Since 1994, submerged macrophytes have been excessively infested mainly in the Southern Basin of Lake Biwa (Haga and Ishikawa, 2011). It has been proposed that the expansion of submerged macrophytes from 1994 was caused mainly by the serious water shortage during summer of that year, followed by the decline of water level up to -1.2m and increase of light penetration to the lake bottom of the Southern Basin (Hamabata et al., 2012).

Simultaneously, the water transparency gradually improved, which also supplied sufficient light to the macrophytes. Currently, the submerged macrophytes cover approximately 90% of the Southern

Basin (Fig. 1-2). The large quantity of phytomass has been causing water stagnation, foul odor, fishing interference, ecosystem change and landscape fouling (Haga et al., 2006a,b; Fig. 1-3 (A)). In addition, filamentous blue green algae such as Anabena spiroides, often attaches on submerged macrophytes is known to produce geosmin, which occurs nuisance musty odor in the water (Sugiura et al., 2004).

Lake Biwa is a very important source of drinking water for 14 million people in its watershed and downstream areas: Shiga, Osaka, Kyoto, Nara and Mie Prefectures. In recent years, it has been reported that drifted and stranded submerged macrophyte and water stagnation caused by the macrophyte infestation can increases geosmin concentration in Lake Biwa-derived drinking water

(Sugiura et al., 2004). Thus there is an urgent need to reduce the submerged macrophytes from the lake.

1.2. Harvesting of submerged macrophytes and treatment of the harvested phytomass

The harvesting of submerged macrophytes in Lake Biwa has been strongly propelled by

government of Japan and the local government. In order to promote the harvesting and treatment of submerged macrophytes, the Ministry of the Environment and Shiga Prefecture recently revised “Plan for Conservation of Lake Water Quality” and “Mother Lake 21 Plan”, respectively (Ministry of the

Environment, 2010; Scientific Committee for Comprehensive Conservation of Lake Biwa, 2010).

Currently, the submerged macrophytes have been extensively harvested by the local government, fishermen and citizens. There are mainly two harvesting methods: (1) trawling by using shellfish trawling equipment Mangan aimed for completely removing macrophytes from their roots (Fig. 1-3

B), and (2) cutting of macrophytes by harvesters, which harvests down to the depth of 1.5 m (Fig. 1-

3 C). The harvesting yield and the harvesting cost reach approximately 4,000 ton and JPY 180 million per annum, respectively (Hamabata et al., 2012)(Fig. 1-4). The harvested phytomass is piled up on the fishing port for several days (Fig. 1-3 D), transferred and has been either incinerated or composted.

Incineration is considered to be unfeasible due to the high moisture content (80-95%wwt) of submerged macrophytes. Currently, composting has been the main treatment method of submerged macrophytes in Lake Biwa. Composting is advantageous in nutrient recycling in low treatment cost, but it is still not competitive against chemical fertilizer mainly due to the long treatment time and the difference between the fertilizer application time and the growth period of the submerged macrophytes (Shiga Prefecture, 2012). In addition, harvested macrophytes often contains harmful contaminants such as fish hook, which can be a risk for farmers to get injury or contamination to the crops by using macrophyte compost. With the same reason, utilization of submerged macrophytes to animal feed is limited. On the other hand, bioethanol fermentation has recently been attempted for renewable biofuel production from submerged macrophytes. Bioethanol is advantageous in transportation by its liquid form. However, it is widely known that bioethanol fermentation of lignocellulosic biomass has significant drawback: high cost for addition of enzymes. Differ from bioethanol fermentation of sugar-rich or starch-rich biomass such as wheat, corn and sugar cane, cellulose has to be broken down to fermentable sugars such as glucose, prior to ethanol fermentation

(Taherzadeh and Karimi, 2008). Matsumoto (2010) conducted bioethanol fermentation using raw

Egeria densa harvested from Lake Biwa, and reported that the addition of both cellulase and pectinase were required for the saccharification. Thus, the alternative treatments to incineration/landfilling, composting, animal feed and bioethanol production should be established, in order to effectively utilize the submerged macrophytes.

Anaerobic digestion enables the recovery of bioenergy as methane gas from biomass with high moisture content. From this point of view, the application of anaerobic digestion has been receiving its attention for the effective alternative treatment of submerged macrophytes. Anaerobic digestion involves a series of metabolic reactions in which complex components of the substrate are sequentially converted to methane and carbon dioxide as the main end products (Figure 1-5). Since a wide variety of anaerobic microorganisms are involved in the digestion process, the majority of organic polymers can be degraded during anaerobic digestion. The process operation is relatively simple and the treatment cost is low as compared to bioethanol fermentation, since application of expensive enzymes/yeasts is not required in anaerobic digestion. The solid and/or liquid residue of digestion process can be used as fertilizer owing to their high nutrients content (Liedl et al., 2006;

Uysal et al., 2010). In recent years, microalgae cultivation by using anaerobic digestion effluent has been attempted for low-cost nutrient removal and/or production of high-value products such as biofuel, animal feed and health supplements (Park et al., 2010; Levine et al., 2011; Bahr et al., 2014).

Anaerobic digestion of solid waste has been mature and well developed in both research and industrial levels. In laboratory scale, substantial number of literatures regarding anaerobic digestion of solid waste has been elucidated the process performance such as maximum organic loading rate (OLR)

(e.g. Nagao et al., 2012), operating temperatures (e.g. Ferrer et al., 2010), inhibitors (e.g. Chen et al.,

2008), and fundamentals such as kinetics and modelling (e.g. Lay et al., 1997). Anaerobic digestion has been employed to the treatment of wide variety of bio-wastes such as wastewater, food waste, animal manure, agriculture waste, and sewage sludge in commercial scale (Sawatdeenarunat et al.,

2014). Moreover, “Act on Special Measures Concerning Procurement of Renewable Electric Energy by Operators of Electric Utilities” came into effect in Japan from July 2012 (Ministry of Economy

Trade and Industry (METI), 2011). The act obligates electricity companies to purchase the electricity generated from biomass with the fixed price. Currently, purchase price of electricity generated from anaerobic digestion is 39 JPY (+ tax) per kWh, which is up to 44% higher than that of photovoltaic power (Ministry of Economy, Trade and Industry, 2015). Thus the application of anaerobic digestion for the treatment of submerged macrophytes can possibly be highly beneficial, in terms of energy recovery and nutrient recycling.

1.3. Anaerobic digestion of lignocellulosic biomass and enhancement of CH4 recovery

In recent years, lignocellulosic biomass has been given its attention as prospective feedstocks for biomethane recovery. Lignocellulose, which is the main component of the plant cell wall, is mainly composed of cellulose, hemicellulose and lignin, and the each biopolymer has different characteristics. Cellulose is a polysaccharide consists of a linier chain of D-glucose. In a plant biomass, the cellulose chains are bundled by hydrogen bond and forming “cellulose microfibril”. The cellulose microfibrils have both crystalline and amorphous regions, and the major part is the crystalline form

(Chum, 1985). Although both types of cellulose have high methane recovery, the degradation rate of high-crystalline cellulose is lower than amorphous one (Jeihanipour et al., 2011). Hemicellulose is also a polysaccharide consists of different biopolymers such as pentose, hexose and sugar acids.

Contrary to cellulose, hemicellulose has amorphous structure so that it is easily hydrolyzed (Fengel and Wegener, 1984). Accordingly, both cellulose and hemicellulose are considered as the main carbon source for CH4 recovery in anaerobic digestion, due to the relatively high biodegradability. On the other hand, lignin is the most recalcitrant polymer among three component of lignocellulose. Lignin is amorphous polyphenol consists of three different phenylpropane units (p-coumaryl, coniferyl and sinapyl alcohol), forming extremely complex three-dimensional structure. The main role of lignin is

to give the structural support to plants and protect cellulose and hemicellulose against microbial/enzymatic attack by coating the polysaccharides (Figure 1-6). It is known that anaerobic microorganisms are unable to degrade lignin (Benner et al., 1984). Accordingly, the lignocellulose component of the inherent nature of the substrate greatly influences the biodegradability and the methane recovery.

Previous studies have investigated the anaerobic digestibility of wide variety of terrestrial plant biomass and lignocellulosic residue derived from terrestrial plants such as grass 94-432 mL-

-1 -1 CH4 g-VS (Triolo et al., 2011; Xie et al., 2011; Frigon et al., 2012); fruit 191-523 mL-CH4 g-VS

-1 (Gunaseelan, 2007; Zhang et al., 2013b); vegetable 175-400 mL-CH4 g-VS (Gunaseelan, 2007;

-1 Chen et al., 2014b); paper residue 170-238 mL-CH4 g-VS (Lin et al., 2009; Teghammar et al., 2010);

-1 wood processing residue 129 mL-CH4 g-VS (Yao et al., 2013). From these previous researches, it has been reported that the methane recovery of terrestrial wood, herbaceous plants, fruits and vegetables is limited mainly by the lignin content (Gunaseelan, 2007; Triolo et al., 2011; Frigon et al., 2012). However, the chemical composition in relation to the methane recovery of aquatic macrophytes has never been investigated yet. Previous studies have investigated mainly on anaerobic digestion of floating macrophytes such as water hyacinth and emergent macrophytes such as bulrush, while the knowledge about anaerobic digestion of submerged macrophytes is limited (Table 1-1).

Among aquatic macrophytes, only submerged macrophytes are “true water plant” since their whole body including leaves, stems and roots are submersed under water. In order to adapt to the aquatic environment, submerged macrophytes lacks or has little mechanical strength, cuticle and stomata on the surface, and has internal air-filled cavities to provide oxygen to themselves (Herdendorf et al.,

2006). Thus anaerobic digestibility of submerged macrophyte could be significantly different from other plant biomass.

Substrates with slow degradation rate and/or low methane recovery are generally regarded as unsuitable for anaerobic digestion. Hence, pre-treatment should be applied to improve the methane

recovery to recalcitrant macrophyte species for stable anaerobic digestion. For improving feedstock degradability and microbial metabolic rate, application of pre-treatments have widely been attempted to various plant biomass. Amongst others, chemical pre-treatment is noted as the most promising methods in terms of rapid and highly-efficient disruption of rigid chemical structure of lignocellulose

(Taherzadeh and Karimi, 2008; Hendriks and Zeeman, 2009; Carrere et al., 2016). The enhancement of CH4 recovery by chemical pre-treatment has been reported for terrestrial herbaceous plants (Xie et al., 2011; Monlau et al., 2012; Chen et al., 2014b). On the other hand, it has been pointed out that toxic compounds such as dissolved lignin can be co-produced during thermal or chemical pre- treatment, and the inhibitory effect has been reported in several biological process such as hydrogen fermentation (Cao et al., 2010; Quéméneur et al., 2012) and methanogenesis step of anaerobic digestion (Field and Lettinga, 1987; Kayembe et al., 2013). To date, most studies have focused only on the enhancement of CH4 recovery by pre-treatment, and no studies have investigated the toxic effect of pre-treatment by-product directly derived from lignocellulosic biomass on anaerobic digestion process. In addition, most studies have been performed pre-treatment + anaerobic digestion of lignocellulosic biomass only in batch mode, although continuous (or semi-continuous) mode should be operated in order to truly predict the digestibility, process stability and inhibitory effect in full-scale anaerobic digesters. Therefore, in order to stably treat submerged macrophyte in anaerobic digestion, the effect of chemical pre-treatment on the digestion process should be elucidated.

1.4. Objectives

The overall research aim of this Ph.D. thesis is to examine anaerobic digestion process of submerged macrophytes and to enhance the anaerobic digestibility by chemical pre-treatment. The specific objectives were: (1) to reveal the relationship between chemical composition and anaerobic digestibility of five submerged macrophyte species which are excessively propagated and dominant in Lake Biwa (Chapter II); (2) to examine the effect of alkaline pre-treatment on anaerobic digestion

of recalcitrant submerged macrophytes (Chapter III); (3) to evaluate the effect of alkali pre-treatment on semi-continuous anaerobic digestion of recalcitrant submerged macrophyte (Chapter IV).

Thereafter, based on the obtained results in Chapter II to IV, the feasibility of alkaline pre-treatment with anaerobic digestion process of submerged macrophyte in full scale was evaluated in terms of cost and heat energy balance and the prospective process optimization were proposed as general discussion in Chapter V.

Table 1-1. CH4 yields from floating, emergent and submerged macrophytes. All data was obtained by batch anaerobic digestion.

CH conversion CH yield CH yield CH yield 4 Aquatic weeds Type 4 4 4 efficiency Country Literature (mL g-VS-1) (mL g-TS-1) (mL g-wwt-1) (%-COD)

Azolla pinnata Floating 132 107 6.3 - India Abassi et al. 1990 Cabomba caroliniana Floating 173 - - - Australia O’Sullivan et al. 2010 Ceratopteris sp. Floating 204 164 6.7 - India Abassi et al. 1990 Cyperus sp. Emergent 38 30 4.8 - India Abassi et al. 1990 9 Eichhornia crassipes Floating 209 - - - India Chanakya et al. 1993 140 - 180 120 - 154 8.4 - 10.8 - USA Moorhead and Nordstedt 1993 190 - - - Australia O’Sullivan et al. 2010 182 - 193 - - - India Patal et al. 1993 Elodea nuttallii Submerged 333 - - - Germany Muñoz Escobar et al. 2011 Hydrilla verticillata Submerged 81 66 5.3 - India Abassi et al. 1990 Salvinia molesta Floating 242 204 11.9 - India Abassi et al. 1990 Scirpas sp. Emergent 66 53 7.4 - India Abassi et al. 1990 Utricularia reticulata Emergent 132 108 4.2 - India Abassi et al. 1990

Some literature values were missing due to the lack of data (TS, VS, and/or COD values) from the literature. 0 500 km Northern Basin

Shiga Prefecture Southern Basin 0 20 km

Fig. 1-1. Map of Lake Biwa. The sampling stations were shown by ★.

10 1994 1997 2000 2001 2002 2003

2005 2006 2007 2008 2009

Fig. 1-2. Increase of submerged macrophytes in the southern basin of Lake Biwa (Haga 2010).

11 (A) (B)

12 (C) (D)

Fig. 1-3. The status quo of submerged macrophytes propagation and the treatment of harvested phytomass in the southern basin of Lake Biwa: (A) drifted macrophytes interferes the entry of fish boats, (B) Harvesting of macrophytes using shellfish trawling equipment Mangan, (C) Cutting macrophytes using harvester Gengoro, (D) piled macrophytes on the fishing port. 13

Fig. 1-4. Yearly trend of total amount of harvested submerged macrophytes (bar) and the cost (circle) in the southern basin of Lake Biwa (Kadono et al. 2012). The filled bars and filled circles show the amount harvested by the Natural Environment Conservation Division of Shiga Prefectural Government; the open bar and circle show the total of amount harvested by other divisions of Shiga prefectural office and other organizations. Complex organic polymer (Carbohydrate, protein, lipid)

Hydrolysis

Simple organic monomer (Sugar, amino acids, fatty acids)

Acidogenesis

Volatile fatty acids (Propionic acid, lactic acid)

Acetic acid H2, CO2 Methanogenesis

CH4, CO2

Fig. 1-5. The degradation process of organics during anaerobic digestion (McCarty and Smith 1986).

14 Plant cells

Plant biomass

Macrofibril

Cell wall

Cellulose microfibril Hemicellulose

っｒ

Lignin Microbes/ enzymes

Fig. 1-6. Lignocellulose structure of plant cell.

Chapter II

Chemical composition and anaerobic digestibility of

five submerged macrophyte species dominant in Lake Biwa

2.1. Introduction

Aquatic macrophyte can be classified into three categories depending on their vegetation type: floating, emergent and submerged. A number of previous studies reported that the methane yields of floating, emergent and submerged macrophytes greatly varied depending on the species from 38 to

333 mL g-VS-1 (Table 1-1). In general, hydrolysis of lignocellulose is a limiting step during anaerobic digestion of vascular plant biomass, since recalcitrant lignin protects cellulose and hemicellulose against microbial/enzymatic attack by coating them (Taherzadeh and Karimi, 2008). Lignin is scarcely degraded under anaerobic digestion, demonstrating only 2-17% of methane conversion efficiency even after 300 days of digestion (Benner et al., 1984; Tuomela et al., 2000; Barakat et al.,

2012). It is known that the lignin content of terrestrial gymnosperms, angiosperms and herbaceous plants are 27-30%, 20-25% and 5-20%, respectively (McKendry, 2002). Indeed, previous studies have already reported that the methane recovery of terrestrial woody and herbaceous plants is regulated by the lignin content (Gunaseelan, 2007; Triolo et al., 2011; Frigon et al., 2012). However, the chemical composition in relation to the methane recovery of aquatic macrophytes, especially submerged macrophytes, has not been investigated yet. The body structure of submerged macrophytes is

significantly different from other plants. Submerged macrophytes have more flexible and softer body structure in order to adapt to the water flow, while terrestrial herbaceous plants and floating and emergent macrophytes normally have more rigid body, since they are predominantly emerged from water (Asaeda et al., 2005). Indeed, total phenolic content of submerged macrophytes is lower than that of other aquatic macrophytes (Smolders et al., 2000). In addition, submerged macrophytes have much thinner leaves and waxy cuticle as compared to terrestrial plants, since they do not need to prevent water loss from their body surface (Herdendorf et al., 2006). Thus the methane yield from submerged macrophyte is expected to be different from other type of plants.

In this chapter, the relationship between chemical composition and anaerobic digestibility of five dominant submerged macrophyte species Ceratophyllum demersum, Egeria densa, Elodea nuttallii,

Potamogeton maackianus and Potamogeton malaianus collected from Lake Biwa was investigated.

Eg. densa and El. nuttallii are widely known as invasive species in many countries (Yarrow et al.,

2009; Muñoz Escobar et al., 2011) and accounts for 13% and 6% of the total phytomass in Lake Biwa, respectively (Table. 2-1) (Haga and Ishikawa, 2011). C. demersum, P. maackianus and P. malaianus are native species in Japan and accounts for 6%, 56% and 4% of the total phytomass, respectively. In the present study, batch anaerobic digestion tests of these five macrophyte species were conducted.

2.2. Materials and Methods

2.2.1. Substrates and Inoculum

For the substrate, five dominant submerged macrophyte species C. demersum, Eg. densa, El. nuttallii, P. maackianus and P. malaianus were harvested from the Southern Basin of Lake Biwa,

Shiga Prefecture, in Japan (Fig. 2-1). Fresh macrophyte samples were roughly shredded to the particle size of approximately 0.5 - 1.5 cm and preserved at -20ºC for the experiment. Before the batch anaerobic digestion and chemical component analysis, substrates were defrosted at room temperature.

For the inoculum of batch anaerobic digestion, mesophilic anaerobic sludge treating domestic sewage was used. The anaerobic sludge was obtained from full-scale biogas plant (6,800 m3) treating domestic sewage from Hokubu Sludge Treatment Center, Yokohama, Japan. The obtained sludge was preserved at 37ºC for 2 days before batch anaerobic digestion, in order to digest the sewage sludge left in the inoculum.

2.2.2. Batch Anaerobic Digestion of Submerged Macrophytes

Batch anaerobic digestion was performed at mesophilic temperature of 37 ± 1°C in a temperature controlled laboratory for 14 days. The mix ratio of the substrate to the inoculum was adjusted to 1:2 based on volatile solids (VS) contents, and the mixture was loaded to 500 mL

Erlenmeyer flask. The batch reactors were sealed by silicon stopper with two sampling ports to allow gas and slurry samples to be collected. 1-L aluminum gas bag (GL Sciences, AAK-2, Japan) was

attached for the biogas collection. The batch reactors were purged with N2 to make anaerobic environment in the reactor. pH was not controlled throughout the operational period. The reactors

were constantly agitated at 100 rpm using a shaker (Taitec, NR-150, Japan). All experiments were conducted in triplicate for El. nuttallii, Eg. densa, P. maackianus, P. malaianus and blanks, and duplicate for C. demersum. The methane yield of inoculum was also measured as blank, and subtracted from that of each reactor to determine the methane yield from submerged macrophytes.

2.2.3. Analysis Parameters

pH, total solids (TS), VS, chemical oxygen demand (COD), lignocellulose and biogas (CH4,

CO2) were measured. The pH of the samples was measured using a pH meter (Horiba, B-212, Japan).

Standard methods from American Public Health Association (1998) were applied to the analysis of

TS, VS and COD. Lignocellulose (cellulose, hemicellulose and lignin) content was measured by detergent system (Van Soest et al., 1991) using fiber analyzer (Ankom, A-200, USA).

Lignin composition was analyzed using GC/MS by following the method of Clifford et al.

(1995). Lignin phenols were derivatized from the dried samples using tetramethylammonium hydroxide (TMAH). The present study measured 11 lignin phenols: vanillin, acetovanillin, vanillic acid, syringaldehyde, acetosyringone, syringic acid, p-hydroxybenzaldehyde, p-hydroxyacetophenone, p-hydroxybenzoic acid, p-coumaric acid and ferulic acid (Fig. 2-2). For a standard solution, each lignin phenol was dissolved in methanol to be 200 μg mL-1. Standard solution or 3-5 mg of dried sample was mixed with 150 μL TMAH reagent (25% w/w, in methanol) in a 10 mL ampule. For an internal standard, 100 μg mL-1 of n-C19 FA (D) was added to each ampule. After removal of methanol

by N2 stream, the ampule was vacuumed and sealed, and heated at 300ºC for 30 min. Ampules were cooled to room temperature and the TMAH derivatives were extracted with 1 mL ethyl acetate. The ethyl acetate was evaporated under reduced pressure and dissolved in 100 μL ethyl acetate. The lignin phenol contents were identified by GC/MS (Agilent Technologies, 6890N GC/5973MS, USA). The

GC was fitted with a DB-5MS capillary column (30 m long, 0.25 mm I.D., 0.25 μm film thickness).

The temperature of injector and ion source were maintained at 310ºC and 230ºC, respectively. The oven temperature was gradually increased from 60ºC to 310ºC. The temperature of ion source and quadrupole in the mass spectrometry were maintained at 230ºC and 150ºC, respectively. Helium was used as carrier gas with the flow rate of 1.0 mL min-1.

Biogas collected in the gas bag was quantified using 50 mL disposable syringe under room temperature. Biogas was monitored using a gas chromatograph (Shimadzu, GC-2014, Japan) equipped with a packed column (Shimadzu, Shincarbon ST, 6.0 m long, 3 mm I.D., Japan) and a

thermal conductivity detector. For the calibration, standard gases (H2, O2, N2, CH4, CO2: purity

>99.9%, Taiyo Nippon Sanso Corp., Japan) were used. The temperature of injector and detector were maintained at 120ºC and 260ºC, respectively. The column temperature was gradually increased from

40ºC to 250ºC. Helium was used as carrier gas with the flow rate of 40 mL min-1.

2.2.4. Calculation

A COD-based methane conversion efficiency of the submerged macrophyte after batch

anaerobic digestion was calculated as follows;

Methane conversion efficiency (%-COD) = CODCH4 × 100 (1) CODtotal

where CODCH4 (g-COD) is the COD converted to methane from submerged macrophytes, and COD

total (g-COD) is the total COD loaded to the batch reactor as a substrate. The methane yield can be

expressed as the loss of COD, which is calculated as follows; 1 g-COD = 350 NmL-CH4 (Speece,

1996).

2.3. Results and Discussion

2.3.1. Chemical Composition of Submerged Macrophytes

The organic solid and lignocellulose content of submerged macrophyte significantly varied with species (Table 2-2). The lignin content of P. maackianus was the highest (20.7%-TS) among five species, which is 1.3 times, 1.7 times, 5.3 times and 6.5 times higher than that of C. demersum (15.8%-

TS), P. malaianus (12.2%-TS), Eg. densa (4.4%-TS) and El. nuttallii (3.2%-TS), respectively.

Interestingly, the lignin content of native species (P. maackianus, P. malaianus, C. demersum) was considerably higher than that of invasive species (Eg. densa, El. nuttallii). Eg. densa and El. nuttallii are widely known to have strong invasive capacity by its fast growth rate (Yarrow et al., 2009; Muñoz

Escobar et al., 2011). Poorter et al., (1992) investigated the chemical composition of 24 wild plant species and found that fast-growing species had lower lignin and hemicellulose content as compared

with slow-growing species. Since body rigidity of plant biomass is determined mainly by the lignin content, it was indicated that Eg. densa and El. nuttallii have more flexible body structure than three other native species possibly due to their fast growing characteristic.

The composition of lignin significantly varies with plant types and/or tissues (Hedges and Mann,

1979; del Río et al., 2007). Fig. 2-2 shows the classification of lignin types. Lignin is amorphous polyphenol consists of three different phenylpropane unit forming extremely complex three- dimensional structure; guaiacyl (G) lignin, syringyl (S) lignin, and p-hydroxyphenyl (H) lignin.

Generally, hardwoods are characterized by the dominance of S and G lignin, while softwoods contain relatively few S lignin. Within the hardwood, non-woody tissues such as leaf and pollen, and herbaceous plants are known to have abundant H lignin as well as G and S lignin. In addition to these, lignin in herbaceous plant and/or non-woody portion of softwood and hardwood contains hydroxycinnamic acids (ferulic acid and p-coumaric acid), which has ester bonds and are cross-linked to polysaccharides and lignin polymer (Sun et al., 1997; Buranov and Mazza, 2008; Sonoda et al.,

2010). del Rio et al. (2007) conducted GC/MS analysis for determination of lignin composition of five non-woody plants and reported that the lignin composition greatly varied with species. Table 2-

3 shows the TMAH-derivatized lignin composition of each submerged macrophyte species. All five macrophyte species contained G, S and H lignin and hydroxycinnamic acids, which are characteristic of herbaceous plant and/or non-woody tissue of vascular plants. Rabemanolontsoa and Saka (2012) also confirmed that submerged macrophytes contain these four types of lignin, although the

derivatization method and lignin classification is different. In the present study, the S/G lignin ratio of submerged macrophytes ranged from 0.2 to 0.9 depending on species. These S/G ratio are similar with flax (0.4) and hemp (0.8), but these values are considered to be fairly low, as compared with other plant biomass such as jute (1.7), abaca (2.9) and sisal (3.4) (del Río et al., 2007). In general, low S/G lignin ratio implies lower delignification rate and higher alkali consumption during alkaline delignification process, suggesting lignin polymers of submerged macrophytes are relatively resistant against chemicals. In contrast, the content of hydroxycinnamic acids was abundant in all submerged macrophyte species, accounting for 27.2 to 59.4% of all lignin phenols (Table 2-3). Ferulic acid is attached to lignin polymer with alkali-stable ether bonds and to hemicellulose with alkali-labile ester bonds (Ralph et al., 1995). In contrast, p-coumaric acid is ester-linked and ether-linked with lignin polymer. Since alkali can easily cleave ester linkage (Hartley and Morrison, 1991), application of alkali can induce the removal of lignin polymer-ferulic acid complex from the surface of polysaccharides. Accordingly, it is suggested that alkaline pre-treatment may have a great potential for effective delignification of lignin-rich submerged macrophyte species such as P. maackianus.

2.3.2. CH4 Production Rate and CH4 yield

The pH is an important parameter for stable operation of anaerobic digestion process, since the value strongly depends on the accumulation of volatile fatty acids (VFAs) and buffering capacity. The pH variation during the batch anaerobic digestion of submerged macrophytes was shown in Fig. 2-3.

In all submerged macrophyte species, the pH dropped from 7.7 – 7.9 on day 0 to 7.4 – 7.6 on day 1 or 2. This result suggests relatively rapid hydrolysis and acidogenesis occurred from cytoplasm and/or lignocellulose of submerged macrophytes. The pH started to recover from day 3 and reached 8.0 –

8.1 on day 10 in all macrophytes, indicating the bio-gasification of produced VFAs.

Fig. 2-4 (A) shows the methane production rate of roughly-shredded submerged macrophytes.

The methane production of C. demersum, El. nuttallii, Eg. densa, P. maackianus and P. malaianus rapidly occurred from day 1. The methane production rate peaked on day 3 for El. nuttallii (116.0 ±

2.8 mL g-VS-1 day-1), day 2 for Eg. densa (79.9 ± 13.3 mL g-VS-1 day-1) and day 1 for C. demersum

(80.4 mL g-VS-1 day-1), P. maackianus (37.2 ± 1.4 mL g-VS-1 day-1) and P. malaianus (73.6 ± 16.0 mL g-VS-1 day-1), respectively. After day 3, the methane production rate of all macrophytes rapidly decreased and the generation of methane mostly completed in 7 to 14 days. The rapid biogasification was demonstrated probably by the biodegradation of intracellular soluble organic matter (e.g. cytosols) and/or the accessible cellulose and hemicellulose exists in the cell wall.

The VS-based cumulative methane yield and COD-based methane conversion efficiency greatly varied with submerged macrophyte species (Fig. 2-4 (B), Table 2-4). The methane yield and methane conversion efficiency of El. nuttallii was 360.8 mL g-VS-1 and 61.4%-COD, indicating this species was relatively labile. Muñoz Escobar et al. (2011) obtained the similar results of approximately 333 mL g-VS-1. This result confirms El. nuttallii is highly feasible for anaerobic digestion. Eg. densa, P. malaianus and C. demersum were also labile, yielding 287.1 mL g-VS-1

(60.6%-COD), 277.5 mL g-VS-1 (72.2%-COD) and 249.2 mL g-VS-1 (57.1%-COD), respectively.

The VS-based methane yield of Eg. densa, C. demersum and P. malaianus was relatively high as compared to most of other aquatic macrophytes conducted in the previous studies. In addition, TS- based and wet weight based methane yields of these four macrophyte species were higher than the reported aquatic macrophytes (Table 2-4). These results indicated that the four macrophyte species were also labile and applicable to anaerobic digestion. Contrary to other four macrophytes, P. maackianus was relatively recalcitrant, showing the total methane yield and methane conversion efficiency of 161.2 mL g-VS-1 and 33.9%-COD, respectively. This result showed that P. maackianus is not preferable substrate for anaerobic digestion without the application of some pre-treatments.

2.3.3. Relationship between CH4 recovery and lignocellulose content

Fig. 2-5 demonstrated the relationship between lignocellulose component (TS or VS based) and

the cumulative CH4 yield of submerged macrophytes obtained in the present study. It was clearly

demonstrated that the higher CH4 yield was obtained from submerged macrophytes when the lignin

and hemicellulose content is low, and vice versa. In TS-base, the CH4 yield was significantly

correlated with lignin and hemicellulose content (p<0.05). In VS-base, the CH4 yield was significantly correlated with hemicellulose content (p<0.05), and the correlation was marginally significant with lignin content (p<0.10). On the other hand, cellulose content did not show any

significant relationship with the CH4 yield. In anaerobic digestion, hemicellulose is easily

hydrolysable (Hendriks and Zeeman, 2009; Barakat et al., 2012; Monlau et al., 2015), while lignin is recalcitrant (Benner et al., 1984). Therefore, it was suggested that the anaerobic digestibility of submerged macrophytes is regulated by the lignin content of the inherent nature of the substrate. Note that the number of samples (n=5) should be increased to validate this relationship, by examining different submerged macrophyte species, and/or macrophytes obtained from different location or season.

Fig. 2-6 summarized the CH4 yield of submerged macrophytes obtained in the present study and other biomass from literatures, in relation with TS or VS based lignocellulose component. Among

three lignocellulose components, the results of the present study was fitted well on the lignin-CH4 correlation of other plant biomass in both TS-base (Y= -13.0x + 445.3, n=44, R2= 0.584, p<0.001) and VS-base (Y=-11.3x+441.2, n=44, R2=0.562, p<0.001). Previous studies have reported that lignin content of the substrate influences the methane recovery (Gunaseelan, 2007; Triolo et al., 2011;

Frigon et al., 2012). Triolo et al. (2011) examined the correlation between total methane yield and several different chemical components of various energy crops and manure, and reported that lignin content was most significantly correlated with methane yield, as compared with cellulose, neutral detergent fiber (NDF: cellulose + hemicellulose + lignin + ash) and acid detergent fiber (ADF: cellulose + lignin + ash) contents. The present study confirmed that the lignin content of submerged macrophyte greatly varies with species and affects the anaerobic digestibility, as well as other terrestrial woody and herbaceous plants, fruits, and vegetables. Thus it was suggested that the lignin

content can be used for the prediction of anaerobic digestibility of submerged macrophytes.

The present study indicated that El. nuttallii, Eg. densa, P. malaianus and C. demersum are feasible for anaerobic digestion owing to the high methane recovery, whereas P. maackianus is not preferable for anaerobic digestion due to high content of lignin. A number of studies have reported that lignin is scarcely degraded under anaerobic digestion, demonstrating only 2-17% of methane conversion efficiency even after 300 days of digestion (Benner et al., 1984; Tuomela et al., 2000;

Barakat et al., 2012). Accordingly, it is assumed that delignification pre-treatment such as alkaline pre-treatment may be effective for P. maackianus in order to obtain the sufficient methane recovery.

2.4. Conclusions

In Chapter II, the chemical composition and the anaerobic digestibility of five submerged macrophytes species were investigated. The lignin content of the submerged macrophyte widely ranged depending on species from 3.2 to 20.7%-TS. All macrophytes contained G, S and H lignin and hydroxycinnamic acids, which indicates the lignin of submerged macrophytes was similar to herbaceous plant and/or non-woody tissue of vascular plants. The total methane yield of submerged macrophytes greatly varied from 161.2 to 360.8 mL g-VS-1 depending on species (El. nuttallii > Eg. densa > P. malaianus > C. demersum > P. maackianus). The present study revealed that the methane recovery of submerged macrophytes is regulated by the lignin content, as well as other lignocellulosic biomass. For the lignin-rich macrophyte (i.e. P. maackianus), the application of delignification pre-

treatments is recommended to make this species feasible for anaerobic digestion.

Table 2-1. Biomass and composition of submerged macrophytes in Southern basin of Lake Biwa in September, 2007 (cited from Haga and Ishikawa 2011).

Biomass in Southern basin Composition (ton-dwt) (%-total)

Potamogeton maackianus 5370 55.8 Egeria densa 1211 12.6 Ceratophyllum demersum 1194 12.4 Elodea nuttallii 555 5.8 Potamogeton malaianus 342 3.6 Hydrilla verticillata 547 5.7 Vallisneria asiatica 202 2.1 Myriophyllum spicatum 200 2.1 Others 1 0.0 Filamentous algae 726 7.5

29 Table. 2-2. Chemical composition of submerged macrophytes used in the experiment.

Parameter Unit Ceratophyllum demersum Elodea nuttallii Egeria densa Potamogeton maackianus Potamogeton malaianus

Total solids (TS) %-wwt 6.4 7.0 4.9 9.7 9.1

Volatile solids (VS) %-wwt 4.9 5.8 4.0 8.2 5.1

30 VS/TS % 76.6 82.9 81.6 84.5 56.0

Total COD g kg-wwt-1 90.8 116.7 50.4 110.3 113.9

Cellulose %-TS 22.3 35.9 36.2 36.2 22.3

Hemicellulose %-TS 6.9 N. D. 1.9 11.4 0.4

Lignin %-TS 15.8 3.2 4.4 20.7 12.2

N. D. : not detected malaianus 5.4 5.5 5.5 2.7 1.4 0.9 0.9 0.9 3.0 3.0 0.2 21.9 32.0 40.8 13.6 13.6 14.7 16.4 16.4 16.4 16.4 24.4 24.4 Potamogeton 8.6 9.8 1.9 1.9 2.6 5.3 0.8 4.1 0.3 0.3 3.8 3.8 3.7 3.7 5.7 5.7 0.9 50.3 50.3 21.5 21.5 27.2 54.4 Potamogeton maackianus Potamogeton 2.6 2.6 5.0 6.6 0.9 1.1 0.7 0.7 1.8 1.8 3.9 0.3 11.0 11.0 10.8 10.8 17.5 17.5 41.9 41.9 59.4 22.5 14.2 Elodea nuttallii 5.2 0.5 1.7 1.7 2.9 6.2 0.0 2.3 0.9 0.9 3.0 3.0 44.8 39.2 10.8 25.6 25.6 12.7 12.7 25.0 25.0 19.8 19.8 Egeria densa densa Egeria demersum 6.2 6.2 4.1 2.2 2.2 4.7 0.9 0.9 8.5 8.5 0.7 28.5 35.2 15.5 20.9 20.4 20.4 10.6 10.6 13.9 13.9 15.4 15.4 13.1 13.1 Ceratophyllum lignin lignin lignin lignin lignin lignin lignin lignin lignin lignin lignin lignin lignin lignin lignin ------Unit % % % % % % % % % % % % % % % (H) (H) (H) type) total ) (hydroxycinnamic) (S) (S) G ( (S) (G) (hydroxycinnamic) ic acid ic 3. Lignin 3. Lignin phenol compositionmacrophytes. submerged of (G) r - phenol (lignin hydroxyacetophenone acid hydroxybenzoic hydroxybenzaldehyde couma /G ratio - - - - S Hydroxycinnamic H total S totalS G total p Vanillin Acetovanillin Vanillic acid Syringaldehyde Acetosyringone acid Syringic Lignin p p p Ferulic acid Ferulic Table 2 Table

31 Table 2-4. Cumulative CH4 yields and COD-based CH4 conversion efficiency from submerged macrophytes. All data was obtained by batch anaerobic digestion.

CH conversion CH yield CH yield CH yield 4 Submerged macrophyte 4 4 4 efficiency (mL g-VS-1) (mL g-TS-1) (mL g-wwt-1) (%-COD) Ceratophyllum demersum 249 191 12.2 57.1 Elodea nuttallii 361 299 20.9 61.4 Egeria densa 287 234 11.5 60.6 Potamogeton maackianus 161 136 13.2 33.9 Potamogeton malaianus 278 156 14.2 72.2

32 Elodea nuttallii Egeria densa

http://plants.minibird.jp/

Potamogeton malaianus Ceratophyllum demersum

Potamogeton maackianus

Fig. 2-1. Submerged macrophytes used in the present study.

33 Lignin polymer

Guaiacyl (G) lignin CHO COCH3 COOH

OCH3 OCH3 OCH3 OH OH OH Vanillin Acetovanillin Vanillic acid

Syringyl (S) lignin CHO COCH3 COOH

H3CO OCH3 H3CO OCH3 H3CO OCH3 OH OH OH Syringaldehyde Acetosyringone Syringic acid

p-hydroxybenzoic (H) CHO COCH3 COOH lignin

OH OH OH p-hydroxybenzaldehyde p- hydroxyacetophenone p- hydroxybenzoic acid

Hydroxycinnamic CH=CHCOOH CH=CHCOOH acids

OCH3 OH OH p-coumaric acid Ferulic acid

Fig. 2-2. Lignin polymer structure and lignin phenols exist in plant biomass.

34 8.5

8.0

7.5 pH Elodea nuttallii Egeria densa 7.0 Potamogeton malaianus Ceratopyullum demersum Potamogeton maackianus 6.5 0 2 4 6 8 10 12 14 Operation time (days)

Fig. 2-3. pH variation during batch mesophilic anaerobic digestion of submerged macrophytes.

35 140 (A) Elodea nuttallii ) 1 - 120 Egeria densa day 1 - Potamogeton malaianus VS 100 - Ceratophyllum demersum

mL g 80 Potamogeton maackianus rate ( 60

40 production production 4 20 CH

0 0 2 4 6 8 10 12 14 Operation time (day)

400

) (B) - 1 350 - VS 300 mL g 250 yield ( yield

4 200

150

100

50 Cumulative CH Cumulative

0 0 2 4 6 8 10 12 14 Operation time (day)

Fig. 2-4. CH4 production rate (A) and cumulative CH4 yield (B) of submerged macrophytes during batch mesophilic anaerobic digestion.

36 ) 1

- 400 400 (A) (D) VS - g 300 300 (mL 200 200 yield 4 100 100

0 0 0 10 20 30 40 50 0 20 40 60

Cumulative CH Cumulative （） Cellulose content %-TS Cellulose content （%-VS） ) 1 - 400 400

VS (B) (E) - g 300 300 (mL

yield 200 200 4

100 100

0 0

Cumulative CH Cumulative 0 5 10 15 20 25 0 5 10 15 20 25 Hemicellulose content （%-TS） Hemicellulose content （%-VS） ) 1 - 400 400 VS

- (C) (F) g 300 300 (mL yield

4 200 200

100 100

0 0

Cumulative CH Cumulative 0 5 10 15 20 25 0 5 10 15 20 25 30 Lignin content （%-TS） Lignin content （%-VS）

Fig. 2-5. Relationship between lignocellulose component and the CH4 yield of submerged macrophyte in mesophilic batch anaerobic digestion. (A) cellulose (TS-based, Y= 0.3x + 257.4, n= 5, R2= 0.001, p=0.953), (B) hemicellulose (TS-based, Y= -813.3x +321.8, n= 5, R2= 0.831, p=0.031), (C) lignin (TS-based, Y= -8.8x +365.8, n=5, R2= 0.833, p=0.031), (D) cellulose (VS-based, Y= 1.2x +218.0, n=5, R2= 0.012, p= 0.862), (E) hemicellulose (Y= -10.9x + 323.1, n=5, R2= 0.825, p=0.033), (F) lignin (Y= -5.9x + 360.0, n=5, R2=0.659, p= 0.095).

37 ) 1 - 600 600 (A) (D) VS -

g 500 500

(mL 400 400 300 300 yield 4 200 200 100 100 0 0 0 20 40 60 0 20 40 60 Cumulative CH Cumulative Cellulose content （%-TS） Cellulose content （%-VS） ) 1 - 600 (B) 600 (E) VS -

g 500 500

(mL 400 400 300 300 yield 4 200 200 100 100 0 0 0 20 40 60 0 20 40 60 80 Cumulative CH Cumulative Hemicellulose content （%-TS） Hemicellulose content （%-VS） ) 1 - 600 (C) 600 (F) VS -

g 500 500

(mL 400 400 300 300 yield 4 200 200 100 100 0 0 0 10 20 30 0 10 20 30 Cumulative CH Cumulative Lignin content （%-TS） Lignin content （%-VS）

Fig. 2-6. Relationship between lignocellulose component and the CH4 yield of submerged macrophyte other plant biomass in mesophilic batch anaerobic digestion. (A) cellulose (TS- based, Y= = -3.5x + 427.0, n= 44, R2= 0.146, p=0.011), (B) hemicellulose (TS-based, Y= 1.1x + 281.0, n=19, p= 0.563), (C) lignin (TS-based, Y= -13.0x + 445.3, n=44, R2= 0.584, p<0.001), (D) cellulose (VS-based, Y= -3.1x + 430.3, n=44, R2= 0.173, p= 0.006), (E) hemicellulose (Y= 1.2x + 270.8, n=19, R2= 0.047, p=0.375), (F) lignin (Y= -11.3x + 441.1, n=44, R2=0.562, p<0.001). (■) = Present study, (○) = Aquatic weeds (Chen et al. 2010; Wang et al. 2010), ( ) = Herbaceous plants (Gunaseelan 2007; Triolo et al. 2011; Xie et al. 2011; Frigon et al. 2012), (△) = Fruits and vegetable waste (Gunaseelan 2007). 38

Chapter III

Effect of alkaline pre-treatment on anaerobic digestibility of

lignin-rich submerged macrophyte

3.1. Introduction

Lignocellulosic biomass is considerably recalcitrant for biological degradation mainly due to the lignin barrier on cellulose and hemicellulose. The methane production potential and biodegradability of submerged macrophytes were investigated in Chapter II of the present study. It was found that one of the most dominant submerged macrophyte P. maackianus was the most recalcitrant for anaerobic digestion due to the high content of lignin. Accordingly, pre-treatment to effectively destruct the lignin coating should be applied to obtain the sufficient methane recovery.

A number of pre-treatment methods have been applied to improve the hydrolysis of lignocellulose (Taherzadeh and Karimi, 2008). These methods are classified into three categories: physical pretreatment, biological pretreatment and chemical pretreatment. Physical pre-treatment is known as the effective method in terms of short treatment time. Different types of physical processes such as mechanical shredding and grinding, steam explosion and irradiation (e.g. microwaves) has been used (Taherzadeh and Karimi, 2008; Hendriks and Zeeman, 2009; Carrere et al., 2016). Most physical pre-treatment aims to disrupt plant cells for enhancing the accessible surface area and to alter cellulose crystallinity. However, physical pre-treatment does not enhance lignocellulose digestibility, since lignin-carbohydrate chemical bonding is difficult to cleave only by mechanical forces. Zhang

(1999) conducted anaerobic digestion of rice straw and recommended the combination of mechanical

size reduction with thermochemical pre-treatment (heating with ammonia), since size reduction alone did not improve the methane recovery. On the other hand, biological pre-treatment can degrade lignin and hemicellulose by using microorganisms such as white-rot fungi (Srilatha et al., 1995; Dhouib et al., 2006). The main advantages of the biological pre-treatment are low energy consumption and no chemical requirement. Taniguchi et al. (2005) reported selective lignin degradation of rice straw by a white-rot fungi Pleurotus ostreatus, followed by the enhancement of cellulose enzymatic hydrolysis.

However, the treatment time is very long, taking up to several months (Srilatha et al., 1995; Sun and

Cheng, 2002).

Chemical pre-treatment can degrade the component of lignocellulose (lignin, hemicellulose and cellulose) by cleaving the physical/chemical bonding of each component. A variety of chemicals such as acids (e.g. sulfuric acid, hydrochloric acid) and alkali (e.g. sodium hydroxide, calcium hydroxide) have been applied to lignocellulosic biomass (Lavarack et al., 2002; Fernandes et al., 2009; Xie et al.,

2011; Badshah et al., 2012). Alkali pre-treatment is noted as an effective pre-treatment method for anaerobic digestion of lignocellulosic biomass, mainly due to the strong delignification ability (Gierer,

1985; Hendriks and Zeeman, 2009; Zhu et al., 2010). Previous studies about alkaline pre-treatment coupled with anaerobic digestion has investigated terrestrial herbaceous plants (He et al., 2009; Lin et al., 2009; Nieves et al., 2011; Sambusiti et al., 2013b; Chen et al., 2014b; Costa et al., 2014;

Taherdanak and Zilouei, 2014; Chufo et al., 2015) (Table 3-1). Xie et al. (2011) reported that

-1 alkaline pre-treatment enhanced the CH4 yield of grass silage from 325.8 to 452.5 mL g-VS at 100ºC,

NaOH loading rate of 7.5% (w/w) against the VS of the substrate. For aquatic macrophytes, alkaline pre-treatment was applied to co-digestion of water hyacinth (floating macrophyte) and cow manure

(Patel et al., 1993), and two-stage (H2 - CH4 fermentation) anaerobic digestion process of water hyacinth (Cheng et al., 2010). However, no studies have investigated the effect of alkaline pre- treatment on lignocellulose degradation and CH4 recovery of aquatic macrophytes.

It is known that the degradability of lignin by chemicals depends on the composition of lignin

(Buranov and Mazza, 2008). As described in the previous chapter, lignin polymer consists of four different phenylpropane unit (G, S and H lignin and hydroxycinnamic acids) and only hydroxycinnamic acids have alkali-labile ester bonds, and other bonds such as ether bond are relatively registrant against chemicals (Buranov and Mazza, 2008, Fig. 3-1). In Chapter II, it was revealed that submerged macrophytes in Lake Biwa contained substantial amount of hydroxycinnamic acids, which accounts for 27.2 to 59.4% of total lignin phenols. Accordingly, submerged macrophytes are expected to be easily delignified by using alkali under relatively mild pre-treatment conditions (e.g. low temperature, low alkali dose and short treatment time), as compared with woods. Thus the effect of alkali on lignocellulose degradability and CH4 recovery of submerged macrophytes may be significantly different from other plants.

On the other hand, it is known that chemical and/or physical pre-treatment of lignocellulosic biomass not only improves biomass degradability by altering the rigid lignocellulosic structure, but also can produce pre-treatment by-product, which influence the microbial and enzymatic activity.

Therefore, alkaline pre-treatment may cause adverse effect on anaerobic digestion of submerged macrophytes. Furans (furfural and 5-hydroxymethylfurfural) and phenolics (e.g. dissolved lignin and tannin) are known as possible inhibitor for several biological process such as hydrogen fermentation

(Cao et al., 2010; Quéméneur et al., 2012). For anaerobic digestion, it has been reported that furans

have no inhibitory effect (Barakat et al., 2012), while phenolic monomers including lignin model compound inhibited methanogenesis step (Sierra-Alvarez and Lettinga, 1991a; Sierra-alvarez and

Lettinga, 1991b; Kayembe et al., 2013). The inhibition mechanism of dissolved lignin on methanogenesis has not been fully elucidated, but it has been proposed that phenolics such as dissolved lignin can directly inhibit microbial cells by its high hydrophobicity (Kayembe et al., 2013), type of substituents on aromatic ring (Sierra-Alvarez and Lettinga, 1991a) and molecular weight

(Sierra-Alvarez and Lettinga, 1991b; Yin et al., 2000). However, these knowledges are obviously not adequate for understanding the toxic effect of dissolved lignin on anaerobic digestion of lignocellulosic biomass, since the inhibitory effect on hydrolysis and acidogenesis step are not taken into account, although it is crucial to concern. To date, the toxicity of dissolved lignin on each anaerobic digestion steps has never been investigated.

The objective of Chapter III was to investigate the effect of alkaline pre-treatment on anaerobic digestion of submerged macrophytes. The present study used two dominant submerged macrophyte species in Lake Biwa, which have remarkably different lignin content. Potamogeton maackianus is a lignin-rich indigenous macrophyte species which is dominated approximately 56% in Lake Biwa

(Haga and Ishikawa, 2011). An invasive species Egeria densa (dominance 11%) was also used to compare/contrast the results with P. maackianus, since the lignin content of Eg. densa is much lower than P. maackianus (Koyama et al., 2014). In the present study, chemical hydrolysis test and batch anaerobic digestion test of submerged macrophytes after alkaline pre-treatment were conducted. The toxic effect of dissolved lignin on anaerobic digestion process (methanogenesis, acidogenesis and hydrolysis) was also investigated.

3.2. Materials and Methods

3.2.1. Substrates and inoculum

P. maackianus and Eg. densa were harvested from the Southern Basin of Lake Biwa, Shiga prefecture, Japan. Fresh macrophytes were shredded to the particle size of 0.5-1.5 cm and preserved at -20ºC for the experiment. The chemical components were shown in Table 3-2. The lignin content of P. maackianus was 20.7%-TS, which was more than fourfold as compared with Eg. densa. For the inoculum of batch anaerobic digestion test, the mesophilic anaerobic sludge was obtained from

Hokubu Sludge Treatment Center, Yokohama, Japan. The totals solids (TS) and volatile solids (VS) contents of the sludge were 3.0%-wwt and 2.1%-wwt, respectively. The anaerobic sludge was degassed at 37ºC for 2 days before batch anaerobic digestion test, to digest the residual organics left in the inoculum.

3.2.2. Hydrolysis test of submerged macrophytes using alkaline pre-treatment

Alkaline pre-treatment was conducted on P. maackianus and Eg. densa in order to investigate the optimum pre-treatment conditions. The NaOH loading rate over total solids (TS) of the substrate

-1 was 0, 0.025, 0.05, 0.1, 0.2 g g-TSsubstrate , respectively. 10.0 g-wwt of substrate was added with 10.0 mL of NaOH solution to 50 mL centrifuging tubes. Previous studies have investigated a wide range of pre-treatment temperature of 50 to 150ºC (Xie et al., 2011; Monlau et al., 2012). However, low temperature below 100ºC should be more preferable since production of inhibitory substances may occur at high temperature (Ferrer et al., 2010). In the present study, the substrates with the NaOH solution were heated in a convection oven (EYELA, WFO-700, Japan) at two different temperature

(60 and 80ºC), for four different treatment time (0.5, 1.0, 2.0 and 3.0 h), respectively. For P.

-1 maackianus at NaOH 0.2 g g-TSsubstrate and 80°C, 4.0 and 5.0 h was also conducted. After heating, substrates were immediately centrifuged at 25ºC, 5,000 rpm for 10 minutes by using high-speed centrifugation (KUBOTA, Compact high speed refrigerated centrifuge 6500, Japan). The supernatant was filtered through 0.45 µm glass-fiber filter paper (ADVANTEC, GC-50, 47 mm, Japan) and preserved at -20°C in the freezer until the chemical analysis. The solid residues were rinsed with 1 M

H2SO4 aq. and distilled water for neutralization and dried in the convection oven at 60°C for more than 24 hours. The dried residue was finely milled to pass the sieve pore of 1.0 mm and used for lignocellulose analysis. All experiments were conducted in triplicate.

3.2.3. Batch anaerobic digestion test of pre-treated macrophytes

Pre-treated and un-treated submerged macrophytes were anaerobically digested in a batch mode, by using the same method described in Chapter II. P. maackianus and Eg. densa were pre-treated at

-1 80ºC, NaOH 0.1 and 0.2 g g-TSsubstrate for 3.0 h, followed by neutralization to around pH 7 using

HCl. The CH4 yield of inoculum was also measured and subtracted from the results. Substrate to inoculum ratio was adjusted to 1 to 2 based on VS. 500 mL flask was used for the batch reactors, and the reactors were purged with N2 to make anaerobic environment and sealed by silicon stopper.

Produced biogas was collected using 1-L aluminum gas bag (GL Sciences, AAK-2, Japan). A shaker

(TAITEC, NR-150, Japan) was used to agitate the reactors at 100 rpm. Batch anaerobic digestion test was performed at 37 ± 1°C in triplicate for 14 days.

3.2.4. Methanogenic and acidogenic/hydrolysis toxicity test of dissolved lignin

3.2.4.1. Lignin extraction/purification

Dissolved lignin was extracted from alkali pre-treatment liquor of P. maackianus at pre-

-1 treatment condition of NaOH 0.2 g g-TSsubstrate , 80°C for 3.0 h, and was purified by Dioxane extraction method (Björkman, 1957). For the detailed protocol, the method described by Lei et al.

(2013) and Zhang et al. (2013) was consulted in the present study. The purification steps of dissolved lignin were described in Fig. 3-2. Pre-treatment liquor was passed through the sieve (aperture: 125

µm) to get rid of large particles. Then lignin in the pre-treatment liquor was precipitated at pH 2 by

HCl. The precipitate was centrifuged at 8,000 rpm, 25°C for 10 minutes and dried at 80°C. Then lipid was extracted from crude lignin with hexane by using Soxhlet method for 24 hours and the hexane extractive was discharged. Next, the crude lignin was suspended in 96% dioxane (dioxane: water =

96:4 (v/v)) with a solid to liquid ratio of approximately 1:10 (g mL-1) and refluxed for 3.0 hours at

80°C for lignin extraction. For dioxane-insoluble fraction, 96% dioxane was added and refluxed again to obtain remained lignin fraction. The combined dioxane-soluble fraction (i.e. dioxane lignin) was then concentrated to approximately 1/10 of the initial volume, by using rotary evaporator under reduced pressure. Dioxane lignin was added to 0.1M HCl solution to remove polysaccharides and to precipitate lignin. The precipitate was then washed with diethyl ether several times and was dried to obtain purified lignin as end-product.

3.2.4.2. Methanogenic and acidogenic/hydrolysis toxicity test

Methanogenic and acidogenic/hydrolysis toxicity test were performed in order to elucidate

the inhibition of anaerobic digestion process by extracted dissolved lignin. In both experiments, dissolved lignin was added in different concentration of 0.5, 1.0, 2.5, 5.0 g L -1 and 0 g L -1 as control.

Dried purified lignin was milled and dissolved into 15 mL of 0.1 M NaOH solution. Then the alkali- dissolved lignin solution was neutralized by using 1.0 M HCl solution. As a result of neutralization, up to 87.6 mg of NaCl was produced, of which salinity in anaerobic digestion reactors was 0.6 g L-1.

In order to make the similar salinity level in dissolved lignin 0 g L-1 and blank (no substrate) conditions, 15 mL of 0.1 M NaOH solution was neutralized with 1.0 M HCl solution and added to the reactors. For methanogenic toxicity test, 2.0 g L-1 acetate was used for the substrate and added to

500 mL medium bottle with 20 mL dissolved lignin solution or neutralized water and 130 mL mesophilic anaerobic digestion sludge to make the effective volume of 150 mL. CH4 yield was monitored online by using automatic methane potential test system (AMPTS) II (Bioprocess Control

AB, Sweden. Fig. 3-3). Dissolved lignin 0.0, 0.5, 1.0, 5.0 g L-1 conditions were run in triplicate, while dissolved lignin 2.5 g L-1 and blank (=inoculum without substrate) were operated in duplicate and monoplicate, respectively. For hydrolysis toxicity test, 10 g L-1 cellulose was added with 20 mL dissolved lignin solution or neutralized water to 500 mL flask with 230 mL anaerobic digestion sludge to make the final effective volume of 250 mL. The biogas was collected in aluminum gas bag.

Dissolved lignin 0.0, 0.5, 2.5, 5.0 g L-1 conditions and blank were run in triplicate, while dissolved lignin 1.0 g L-1 were operated in duplicate.

3.2.5. Analytical methods

pH, TS, VS, chemical oxygen demand (COD), lignocellulose and biogas were measured. The

pH of the samples was measured using a pH meter (HORIBA, B-212, Japan). Standard methods from

APHA (American Public Health Asspciation, 1998) were applied to the analysis of TS, VS and COD.

Lignocellulose (cellulose, hemicellulose and lignin) content was measured by detergent system (Van

Soest et al., 1991) by using fiber analyzer (ANKOM Technology, A-200, USA). Lignin composition was analyzed by using the tetramethylammonium hydroxide (TMAH) derivatization method proposed by Clifford et al. (1995). Detailed lignin phenol contents were identified by GC/MS (Agilent

Technologies, 6890N GC/5973MS, USA). DB-5MS capillary column (30 m long, 0.25 mm I.D., 0.25

μm film thickness) was used for the GC column. The temperature of injector and ion source were

310ºC and 230ºC, respectively. The oven temperature was gradually increased from 60ºC to 310ºC.

In the mass spectrometry, the temperature of ion source and quadrupole were 230ºC and 150ºC, respectively. Helium was used as carrier gas with the flow rate of 1.0 mL min-1.

Biogas (CH4, CO2) composition was monitored using a gas chromatograph (SHIMADZU, GC-

2014, Japan) equipped with a packed column (Shimadzu, Shincarbon ST, 6.0 m long, 3 mm I.D.,

Japan) and a thermal conductivity detector. The temperature of injector and detector were maintained at 120ºC and 260ºC, respectively. The column temperature was gradually increased from 40ºC to

250ºC. Helium was used as carrier gas with the flow rate of 40 mL min-1. VFAs composition (acetic, propionic, n-butyric, i-butyric, n-valeric and n-valeric acid) was monitored using a gas chromatograph

(Hitachi, G-3000, Japan) equipped with a capillary column (Stabilwax, size×I.D. 30 m×0.53 mm,

1.00 µm film thickness) and a flame ionized detector (FID). The temperature of injector and detector were maintained at 240ºC. The column temperature was gradually increased from 100ºC to 160ºC at

4 ºC minutes-1. Nitrogen was used as carrier gas with the flow rate of 40 mL min-1.

3.2.6. Calculations

The solubilization efficiency is often evaluated by soluble COD over total COD (SCOD/TCOD) in many studies (e.g. Fernandes et al., 2009; Xie et al., 2011). The SCOD/TCOD can be calculated by using the equation as follows:

SCOD/TCOD (%) = SCOD treated ×100 (2) TCOD initial

where SCODtreated (g-COD) is the soluble COD released from the substrate after pre-treatment and

TCODinitial (g-COD) is the initial total COD of the substrate before pre-treatment.

A COD-based CH4 conversion efficiency of the submerged macrophyte after batch anaerobic digestion was calculated as follows;

COD CH4 conversion efficiency (%-COD) = ×100 (3) COD CHtotal4

where CODCH4 is the COD converted to CH4 from submerged macrophytes (g-COD), and CODtotal is the total COD loaded to the batch reactor as a substrate (g-COD).

The dissolved lignin concentration in the digestate was estimated from the amount of lignin removed by alkaline pre-treatment as follows;

Dissolved lignin concentration (g L-1) = S × L ×R (4) V

where S is the amount of substrate loaded to the biomethane reactor (g-TSsubstrate), L is lignin content

-1 of un-treated biomass (g g-TSsubstrate ), R is lignin removal efficiency by the pre-treatment (%) and V is the working volume of the biomethane reactor (L).

COD-based acidogenesis and hydrolysis efficiency of cellulose during hydrolysis toxicity test was calculated as follows;

Acidogenesis efficiency (%-COD) on day n = Mn+Vn−Ln−En ×100 (5) Mn+Dn−Ln−En

Hydrolysis efficiency (%-COD) on day n = Mn+Dn−Ln−En ×100 (6)

where T is total COD of added cellulose (g-COD), Mn is CH4 yield on day n (g-COD), Vn is VFA content in the digestate on day n (g-COD), Dn is SCOD content in the digestate on day n (g-COD),

Ln is total COD of added dissolved lignin (g-COD), En is CH4 yield from dissolved lignin on day n

(g-COD).

Normalized methanogenic activity (NMA), normalized acidogenesis activity (NAA) and normalized hydrolytic activity (NHA) was calculated by consulting Gonzalez-Estrella et al. (2013) as follows;

NMA (%) = CH4 measured ×100 (7) CH4 theoretical

NAA (%) = A tested ×100 (8)

A control

NHA (%) = H tested ×100 (9) H control where CH4 theoretical is theoretical CH4 yield from CH4 + dissolved lignin and CH4 measured is the

measured CH4 yield in each dissolved lignin condition. A control is the acidogenesis efficiency in

-1 control (dissolved lignin 0 g L ) and A tested is the acidogenesis efficiency in each dissolved lignin condition. H control is the hydrolysis efficiency in control and H tested is the hydrolysis efficiency in each dissolved lignin condition.

3.2.7. Kinetic models

The time course of cumulative CH4 yield, acidogenic efficiency and hydrolysis efficiency in the lignin toxicity test was simulated by the modified Gompertz equation (Lay et al., 1997):

Rmax× e (10) M = P × �−exp� P (λ − t) + 1��

where M is cumulative CH4 yield (NmL), P is methane production potential (NmL), and Rmax is maximum methane production rate (NmL day-1) for methanogenic activity test. For acidogenic activity test, M is VFA production efficiency (%-COD), P is VFA production potential (%-COD),

-1 Rmax is maximum VFA production rate (%-COD day ). For hydrolysis activity test, M is hydrolysis efficiency (%-COD), P is hydrolysis potential (%-COD), Rmax is maximum hydrolysis rate (%-COD day-1). In all toxicity tests, λ is digestion lag phase (day) and t is digestion time (day). The constants

P, Rmax, λ were estimated by non-linear fitting approach with the aid of the SOLVER function of

Microsoft Excel, which employs an iterative least squares fitting routine to produce the optimal goodness of fit between data and function.

3.2.8. Statistical analysis

For the statistical analysis, the data were analyzed using the Tukey-Kramer’s multiple comparisons. Difference with p < 0.05 was considered significant.

3.3. Results and Discussion

3.3.1. Chemical component change of submerged macrophytes by alkaline pre-treatment

3.3.1.1. Hydrolysis efficiency

The efficiency of hydrolysis by pre-treatment can be expressed by SCOD/TCOD of the pre- treated substrate (Cheng et al., 2010; Xie et al., 2011). Fig. 3-4 exhibits the SCOD/TCOD of two submerged macrophyte species after alkaline pre-treatment. The SCOD/TCOD of both species was greatly enhanced with the increase of NaOH loading rate, pre-treatment temperature and treatment

- time. Amongst others, the pre-treatment temperature of 80˚C, NaOH loading rate of 0.2 g g-TSsubstrate

1 and treatment time of 3.0 h performed the highest SCOD/TCOD for both two species, and longer treatment (~5.0 h) did not further enhance the hydrolysis efficiency. This result is confirmed by Xie et al. (2011), that SCOD/TCOD of NaOH-treated biomass reached equilibrium at 2-4 hours at 60-

150°C, and longer treatment up to 24 hours did not further enhance the SCOD/TCOD under any

NaOH condition. Thus it is indicated that the treatment time of 3.0 hours is considered to be proper for alkaline pre-treatment of submerged macrophytes.

Although P. maackianus is recalcitrant lignin-rich macrophyte, the maximum SCOD/TCOD reached 45.4%-COD, while that of labile lignin-poor macrophyte (Eg. densa) was 56.8%-COD.

-1 Contrary, the maximum SCOD/TCOD of both species at NaOH 0 g g-TSsubstrate was considerably

low, demonstrating 12.3%-COD (P. maackianus) and 22.4%-COD (Eg. densa), respectively. These results suggested that alkali takes a key role in hydrolysis of submerged macrophytes. In all NaOH conditions, the SCOD/TCOD of both macrophytes was significantly higher at 80°C as compared with

60°C (p>0.05). Xie et al. (2011) also reported that SCOD/TCOD of thermochemically pre-treated grass silage was enhanced with the increase of NaOH loading rate and treatment temperature. Indeed, the application of higher NaOH loading would improve the SCOD/TCOD. However, it is concerned that higher NaOH loading may cause inhibition of anaerobic digestion. Since pH of pre-treatment

-1 -1 liquor is very high, neutralization is required. At NaOH 0.2 g g-TSsubstrate , up to 14.5 g L of salt is produced in the pretreatment liquor. Anaerobic digestion process is known to be inhibited under NaCl

>20-30 g L-1 (Feijoo et al., 1995; Kimata-Kino et al., 2011). In the present study, the produced salt is diluted in the digestate up to 3.1g L-1, so that the NaCl inhibition is no need for concern.

The application of higher treatment temperature also would improve the SCOD/TCOD, but it requires excessive cost for external energy input. Low-temperature heating below 100ºC should be more preferable since energy expenses for low-temperature heating of the reactor can be eliminated by applying waste heat from a power generation unit fueled with biogas. Thus alkaline pre-treatment

-1 at NaOH 0.2 g g-TSsubstrate , 80°C, 3.0 hours is suggested to be optimum pre-treatment condition feasible for applying anaerobic digestion, by enhancing chemical solubilization of submerged macrophytes.

3.3.1.2. Lignocellulose composition

Changes of lignocellulose composition in the solid fraction of pre-treated macrophytes after 3.0

h of alkaline pre-treatment were demonstrated in Fig. 3-5. Lignin, hemicellulose and cellulose content of P. maackianus and Eg. densa were significantly reduced with the increase of NaOH and pre-

-1 treatment temperature. At the pre-treatment condition of 80˚C and NaOH 0.2 g g-TSsubstrate , the lignin content of P. maackianus was 84.7 mg g-TS-1, which is a remarkable decline from the raw material of 207.0 mg g-TS-1. In contrast, Eg. densa showed slight decline of lignin content, from 44

-1 -1 -1 mg g-TS (untreated) to 31 mg g-TS (NaOH 0.2 g g-TSsubstrate ). Lignin content of the biomass is important parameter for evaluating the anaerobic digestibility, since lignin makes biomass recalcitrant and regulates CH4 recovery from lignocellulosic biomass (Hendriks and Zeeman, 2009; Triolo et al.,

2011; Koyama et al., 2014). The present study clearly indicated that lignin-rich macrophyte was delignified and became feasible substrate for anaerobic digestion treatment.

The reduction of TS and lignocellulose by alkaline pre-treatment (80°C, 3.0 h) of two macrophytes was shown in Table 3-3. Both TS and lignocellulose reduction was improved with increase of NaOH dose for both macrophyte species. The maximum removal efficiencies of cellulose, hemicellulose and lignin of submerged macrophytes were 60-72%, 61-85%, 58-79%, respectively. It is relatively high removal as compared with other plant biomass of 35%, 53%, 70% (grass silage)

(Xie et al., 2011), 17%, 72%, 70% (sorghum) (Sambusiti et al., 2012), 9%, 24%, 46% (asparagus stem) (Chen et al., 2014b). From these results, it was suggested that alkaline pre-treatment is very effective for submerged macrophytes to destruct the complex lignocellulose structure.

Among two submerged macrophyte species, P. maackianus demonstrated significantly higher removal efficiency of cellulose and lignin as compared with Eg. densa (p<0.01) (Table 3-3). In the

-1 pre-treatment condition of NaOH 0.2 g g-TSsubstrate - 80°C - 3.0 h, 79.2% of lignin was removed

from P. maackianus, which is significantly higher than that of Eg. densa (57.8%, p<0.05). Since lignin takes a key role in binding and coating cellulose and hemicellulose, the removal of lignin results in swelling of lignocellulose fibril (Taherzadeh and Karimi, 2008). Indeed, removal of cellulose by chemical pre-treatment is greatly depends on loss of physical protection from lignin and hemicellulose (Hendriks and Zeeman, 2009) Thus it was suggested that alkaline pre-treatment of lignin-rich macrophyte (i.e. P. maackianus) enhances the delignification from the lignocellulose surface, followed by the increase in the surface area of accessible polysaccharides by swelling, which also enhances the solubilization of cellulose and hemicellulose.

During chemical pre-treatment of plant biomass, lignocellulose (cellulose, hemicellulose and lignin) and non-lignocellulose component (cytosol such as sugar, free amino acids, lipids and protein) becomes soluble organics (i.e. SCOD) including soluble sugars, furans and dissolved lignin/phenolics

(Barakat et al., 2012; Sambusiti et al., 2012; Chen et al., 2014b; Di Girolamo et al., 2014). In the present study, SCOD/TCOD of pre-treated P. maackianus was lower than that of Eg. densa, yet P. maackianus showed higher reduction efficiency of lignocellulose (Fig. 3-4, Table 3-3). It is assumed that Eg. densa contains larger proportion of non-lignocellulose component such as cytosol in the dry weight than P. maackianus, owing to the lower lignocellulose content (especially lignin) as shown in

Table 3-2. Cytosols are considered to be more readily hydrolyzed by chemicals as compared with lignocellulose, since a variety of chemical bonds are complexly involved to form very rigid structure of lignocellulose. Therefore, higher cytosol content may be the cause of higher SCOD/TCOD of pre- treated Eg. densa as compare to P. maackianus. Contrary, it was suggested that SCOD of pre-treated

P. maackianus contains relatively larger proportion of lignocellulose derivatives such as dissolved

lignin, and smaller proportion of cytosol derivatives, as compared with Eg. densa.

3.3.1.3. Lignin composition

Table 3-4 demonstrates the lignin phenol contents in solid fraction of submerged macrophytes before and after the pre-treatment. Both macrophyte species contained all of four lignin types (G, S,

H and hydroxycinnamic acids), indicating they have similar lignin structure with herbaceous plants and/or non-woody tissues of angiosperms. The H lignin and hydroxycinnamic acid (ferulic acid and p-coumaric acid) contents of both macrophytes considerably declined with the increase of NaOH loading rate (Table 3-4). The ferulic acid content of P. maackianus decreased from 3913.6 µg g-1 (un-

-1 -1 treated) to 1135.7 µg g (NaOH 0.2 g g-TSsubstrate ), while ferulic acid content of Eg. densa decreased

-1 -1 -1 from 484.7 µg g (un-treated) to 216.8 µg g (NaOH 0.2 g g-TSsubstrate ). Lignin is associated with lignin polymers and/or polysaccharides by a number of different linkages such as α-ether bonds, phenyl glycosidic bonds, acetal linkages and ester bonds. Ferulic acid is attached to lignin with alkali- stable ether bonds and to hemicellulose with alkali-labile ester bonds, and p-coumaric acid is ester- linked and ether-linked with lignin polymer (Ralph et al., 1995). The present study exhibited the reduction of ferulic acid by alkaline pre-treatment, implies the removal of lignin-ferulate complex from lignocellulose. For P. maackianus, ferulic acid accounts for 21.5% of total lignin phenols, which is considerably high as compared with Eg. densa (13.2%). Since P. maackianus contains large amount and proportion of ferulic acid, the amount of alkali-labile ester bond between polysaccharides and lignin polymer in P. maackianus should be much higher than that of Eg. densa. Therefore, it was suggested that the high ferulic acid content is attributed to the high lignin removal efficiency of P.

maackianus by alkaline pre-treatment.

3.3.2. Anaerobic digestion of thermochemically pre-treated submerged macrophytes and

possible inhibitor

The variation of pH during batch anaerobic digestion of alkali pre-treated and un-treated submerged macrophytes was shown in Figure 3-6. In all conditions, the pH slightly dropped from 7.6

- 7.9 on day 0 to 7.2 - 7.6 on day 1 - 3. The lowest pH value of untreated P. maackianus was observed

-1 on day 3, while those of NaOH 0.1 and 0.2 g g-TSsubstrate were day 2 and day 1, respectively. Eg.

-1 densa demonstrated the similar trend, yet the lowest pH of both NaOH 0.1 and 0.2 g g-TSsubstrate were on day 1. The lowest pH value of Eg. densa was slightly lower than that of P. maackianus. The pH quickly recovered to 7.8 – 8.0 by day 5 in all conditions. Overall, the obtained pH values were in the optimum range for methanogenesis of 6.5 to 8.2 (Speece, 1996).

Fig. 3-7 exhibits the CH4 production rate of thermochemically pre-treated and un-treated submerged macrophytes. The initial CH4 production rate (day 1-3) of macrophytes were significantly enhanced by the increase of NaOH loading rate for both species. The maximum CH4 production rates of Eg. densa were significantly higher than that of P. maackianus in all pre-treatment conditions. The

CH4 production rates of thermochemically pre-treated and untreated submerged macrophytes reached maximum on day 1-2. After day 3, the CH4 production rate rapidly decreased in all conditions and demonstrated that the generation of CH4 was mostly completed in 14 days.

The cumulative CH4 yield of both macrophyte species considerably increased with the increase of NaOH loading rate (Fig. 3-8). The cumulative CH4 yields of untreated, NaOH 0.1 and 0.2 g g-

-1 -1 TSsubstrate were 161.3, 231.3 and 242.8 mL g-VS (P. maackianus) and 299.7, 312.8 and 371.2 mL

-1 g-VS (Eg. densa), respectively. The COD-based CH4 conversion efficiency of P. maackianus

-1 increased from 33.6%-COD (untreated) to 50.6%-COD (NaOH 0.2 g g-TSsubstrate ), which is approximately 50% higher than untreated substrate (Table 3-5). Meanwhile, the enhancement of CH4 conversion efficiency of Eg. densa was approximately 24% increase from the untreated (63.2%-

COD). These results can be explained by the difference in the initial lignin content of the substrate and extent of lignin removal after alkaline pre-treatment depending on the macrophyte species. The

CH4 recovery of P. maackianus was more effectively enhanced by alkaline pre-treatment as compared with Eg. densa, probably due to the high delignification efficiency caused by the cleavage of alkali- labile ester linkage between polysaccharides and ferulic acid coupled with lignin polymer. From these results, it was indicated that the alkaline pre-treatment at 80˚C is highly effective for enhancing anaerobic digestibility of lignin-rich macrophytes, and moderately effective for lignin-poor macrophytes.

-1 The cumulative CH4 yield of P. maackianus pre-treated at NaOH 0.1 g g-TSsubstrate was significantly higher than that of un-treated (p<0.01), but the cumulative CH4 yield at NaOH 0.2 g g-

-1 -1 TSsubstrate was similar to NaOH 0.1 g g-TSsubstrate (p>0.05), despite the lignin content of the substrate

-1 -1 -1 - declined from 126.8 mg g-TS (NaOH 0.1 g g-TSsubstrate ) to 84.7 mg g-TS (NaOH 0.2 g g-TSsubstrate

1) (Fig. 3-5, Fig. 3-8). A number of studies have pointed out that the pre-treatment of lignocellulosic biomass sometimes lead the formation of toxic compounds for CH4 recovery (He et al., 2009; Xie et al., 2011; Monlau et al., 2012; Chen et al., 2014b). During heating and/or chemical pre-treatment of lignocellulose, some chemical compounds such as furans (furfural and 5-hydroxymethylfurfural) and

phenolics (e.g. dissolved lignin and tannin) can be produced, which influence the microbial and enzymatic activity. Recent study revealed that furans have no inhibitory effect on anaerobic digestion

(Barakat et al., 2012). Other literatures pointed out the inhibition of methane recovery by phenolic compounds derived from lignin during chemical pre-treatment, but no proof has been confirmed yet

(He et al., 2009; Xie et al., 2011; Monlau et al., 2012; Chen et al., 2014b).

The present study estimated the influence of delignification by-product on the CH4 recovery of submerged macrophytes obtained in the present study and various lignocellulosic biomass from other literatures. Lignin removed from lignocellulose complex is dissolved to liquid fraction during pre- treatment and remained in the digestate during anaerobic digestion, since lignin polymer is scarcely degraded (Benner et al., 1984; Tuomela et al., 2000; Barakat et al., 2012). Then it was assumed that the amount of lignin removed from the solid fraction is equivalent to the dissolved lignin content in the digestate, and the concentration of dissolved lignin in the digestate during batch anaerobic digestion test was calculated. Fig. 3-9 showed that the increase of CH4 yield of all biomass decreased or ceased with the increase of dissolved lignin concentration in the digestate. These results suggest that anaerobic digestion can be inhibited at the condition of high lignin removal efficiency due to the increase of dissolved lignin in the digestate. The inhibitory level of dissolved lignin concentration was fluctuated with substrate, ranging from >0.9 g L-1 to >1.2 g L-1. The range is probably owing to the difference in the inoculum-substrate ratio in each literature and/or physical-chemical property of actual dissolved lignin in the digestate (e.g. molecular weight distribution, chemical composition of lignin phenols), which may vary with substrate particle size or pre-treatment condition. On the other hand, the CH4 yield of Eg. densa did not drop by alkaline pre-treatment, probably due to the low

lignin content of the substrate. From these results, it was suggested that alkaline pre-treatment remarkably enhanced the CH4 yield from submerged macrophytes, but the high NaOH dose to lignin- rich macrophyte P. maackianus would cause the inhibition of the digestion process by dissolved lignin derived from breakdown of lignocellulose, which is increased during the alkaline pre-treatment.

3.3.3. Influence of dissolved lignin on anaerobic digestion process

3.3.3.1. Chemical characteristics of purified dissolved lignin

Table 3-6 shows the chemical composition of purified dissolved lignin derived from alkaline pre-treatment of P. maackianus. The VS/TS was 97.3%, indicating inorganic content is very little.

Total sugar accounted for 3.6% of the purified lignin. The lignin phenol composition of purified dissolved lignin was closely resemble to that of raw P. maackianus (Table 2-2). From these results, it was suggested that the dissolved lignin was extracted in relatively high purity with little compositional alternation.

3.3.3.2. Influence on methanogenesis

Fig. 3-10 (A) shows the time course of cumulative CH4 yield during methanogenic activity

-1 test from 2 g L acetate under different dissolved lignin concentration. The cumulative CH4 yield gradually increased with the increase of dissolved lignin concentration. The cumulative CH4 yield at

-1 dissolved lignin 0 g L was 73.2 NmL, while the cumulative CH4 yield rose to 79.8, 91.4, 101.1 and

116.5 NmL at dissolved lignin concentration of 0.5, 1.0, 2.5 and 5.0 g L-1, respectively. This is probably due to the CH4 recovery from polysaccharides combined to the dissolved lignin, or from

dissolved lignin itself. The small portion of residual polysaccharides remains even after a series of purification steps, due to the widely accepted fact that lignin-polysaccharide network makes lignin purification difficult (Lei et al., 2013). In order to access the characteristic of digestion kinetics, the

CH4 production during methanogenic toxicity test was simulated with the modified Gompertz model

(Table 3-7). Interestingly, digestion kinetic assessment exhibited the deterioration of methanogenesis

-1 in the present study. The maximum CH4 production rate (Rmax) apparently declined in 2.5 g L , and significantly decreased in 5.0 g L-1 (p<0.01), as compared with 0, 0.5, and 1.0 g L-1. From these results, it was suspected that methanogenesis was inhibited at high dissolved lignin concentration.

In order to evaluate the influence of dissolved lignin on methanogenesis from acetate, the methane production from dissolved lignin was excluded from the measured CH4 yields. The methane yield from dissolved lignin was calculated by subtracting the maximum methane yield of dissolved lignin 0.0 g L-1 (i.e. only acetate) from dissolved lignin 0.5 g L-1. Dissolved lignin 0.5 g L-1 is considered to have no influence on acetoclastic methanogenesis, since the digestion kinetics (Rmax, λ, and T80) were not significantly different with dissolved lignin 0.0 g L-1 (Table 3-7, p>0.05). With this calculation, theoretical methane yields, which are the sum of CH4 yield from dissolved lignin and acetate in each dissolved lignin conditions, were estimated and compared with the measured CH4 yields. Fig. 3-10 (B) shows normalized methanogenic activity (NMA) in each dissolved lignin conditions, in order to evaluate the toxicity of dissolved lignin. NMA at dissolved lignin 0.5 and 1.0 g L-1 did not show the inhibition, while that of 2.5 and 5.0 g L-1 was dropped for 7.3% and 15.2%- control, respectively. Therefore, it was suggested that dissolved lignin negatively influences acetoclastic methanogenesis at high lignin concentration of ≧2.5 g L-1. Sierra-Alvarez and Lettinga

(1991a) used lignin isolated from forest industry wastewaters for methanogenic inhibition test and reported 50% inhibition (50% IC) of methanogenesis at lignin concentration of 3.3 to 6.0 g-COD L-

1, which is considerably higher than the present study. The inhibition mechanism of dissolved lignin has not been fully elucidated, but it has been proposed that phenolics such as dissolved lignin can directly inhibit microbial cells by its high hydrophobicity (Kayembe et al., 2013), type of substituents on aromatic ring (Sierra-Alvarez and Lettinga, 1991) and molecular weight (Sierra-Alvarez and

Lettinga, 1991; Yin et al., 2000a). Kayembe et al. (2013) reported that methanogenic activity was inhibited with the increase of hydrophobicity of phenolics, which leads to the high permeability of phenolics in microbial cell membrane exhibiting greater toxicity. Sierra-Alvarez and Lettinga (1991b) found that lignin model compound with aldehyde group (e.g. vanillin) apparently showed higher toxicity on methanogenic activity of 50% IC 1.8 g L-1 as compared with aromatic carboxylic acids

(e.g. vanillic acid: 50% IC >9.2 g L-1). In the present study, aldehyde-substituting lignin (i.e. vanillin, syringaldehyde and p-hydroxibenzaldehyde) accounted only for 4.7% of the dissolved lignin isolated from pre-treated P. maackianus, while lignin with carboxylic acid (i.e. vanillic acid, syringic acid, and p-hydroxibenzoic acid) accounted for that of 53.7% (Table 3-6). This could explain the lower toxicity of dissolved lignin on methanogenesis observed in the present study.

3.3.3.3. Influence on acidogenesis and hydrolysis

The variation of pH during batch anaerobic digestion of cellulose under various dissolved lignin concentration was shown in Figure 3-11. Both variation patterns and the values of pH were similar in all lignin conditions. The pH started to drop from day 4, probably due to the digestion lag

phase until day 2. The pH reached to the lowest values of 7.0 to 7.2 in all conditions on day 4, suggesting rapid acidogenesis. Thereafter, the pH gradually recovered with biogasification of organic acids and reached to the initial level by day 14 in all conditions.

Fig. 3-12 shows the time course of cumulative CH4 yield and SCOD concentration in digestate during batch anaerobic digestion of cellulose under various dissolved lignin concentration. Differ from methanogenic toxicity test, the cumulative CH4 yield declined with the increase of dissolved lignin (p<0.01). The production of CH4 and SCOD simultaneously started to increase on day 4, suggesting hydrolysis, acidogenesis and methanogenesis started to occur after day 2. Thereafter, the

SCOD concentration declined in all lignin conditions to reach to the initial level by day 10 to 14, indicating biogasification of hydrolyzed cellulose. As shown in Fig. 3-12 (B), higher SCOD concentration was seen in lignin-added conditions on day 0, and the SCOD values did not significantly change after 23 days of batch anaerobic digestion. It can be explained by the poor degradability of lignin under anaerobic condition (Benner et al., 1984). Fig. 3-13 shows the time course of VFA concentration in digestate. Corresponding to the variation of pH, CH4 production and

SCOD, the total VFA concentration peaked on day 4 and diminished thereafter in all lignin conditions except at lignin 5.0 g L-1, of which total VFA concentration peaked on day 6. The peak total VFA concentrations in day 4-6 were 273-375 mg L-1, which is considered as non-inhibiting level of anaerobic digestion system, as compared with the inhibiting concentration of acetic acid >6,200 mg

L-1 or propionic acid >3,600 mg L-1 (Ahring et al., 1995). Accumulation of propionic acid occurs due to (1) low microbial activity of propionate utilizing bacteria, 2) inhibition of propionate degradation by high H2 partial pressure, and 3) low methanogenic activity (Speece, 1996). In the present study,

propionic acid dominated only in day 4-6 and diminished afterwards, while acetic acid was dominant throughout the whole experimental period. The diminution of VFA and neutral pH in the digestate of all dissolved lignin condition indicated that the drop of CH4 recovery with the increase of dissolved lignin is obviously not caused by excessive acidification of anaerobic digestion system, which is known as the one of the most common inhibition factor.

The toxic effect of dissolved lignin on acidogenesis and hydrolysis step of anaerobic digestion process were summarized in Fig. 3-14 and 15, and Table 3-8 and 9. Acidogenesis efficiency slightly declined with the increase of dissolved lignin concentration (Fig. 3-14 (A)). The kinetics characterization by fitting with the modified Gompertz model exhibited that the lowering of Rmax and delay of T80 of acidogenesis occurred from dissolved lignin ≥2.5 g L-1 (Table 3-8). The drop of normalized acidogenic activity (NAA) with the increase of dissolved lignin was only 2.5 to 5.2% in dissolved lignin 0.5 to 2.5 g L-1 against control (=lignin 0 g L-1), while that of lignin 5.0 g L-1 was

10.9%-control (Fig. 3-14 (B)). This result indicates that dissolved lignin slightly inhibits acidogenesis step in similar extent to that of methanogenic step under high dissolved lignin condition. In contrast, the hydrolysis efficiency clearly deteriorated from dissolved lignin 1 g L-1 (Fig. 3-15 (A)). On day 23, the hydrolysis efficiency was 94.6% at dissolved lignin 0.0 g L-1, and did not show significant difference with dissolved lignin 0.5 g L-1. On the other hand, the hydrolysis efficiencies at dissolved lignin 1.0, 2.5 and 5.0 g L-1 significantly dropped to 69.6%, 72.5% and 63.9%, respectively (p<0.01, statistical analysis was not applied on dissolved lignin 1.0 g L-1, which was run in duplicate).

Moreover, fitting with the modified Gompertz model demonstrated that the maximum solubilization

-1 -1 -1 rate (Rmax) declined from 30.1 %-COD day at 0 g L to 24.0, 24.5 and 17.6 %-COD day at

dissolved lignin 1.0, 2.5 and 5.0 g L-1, respectively (Table 3-8). On the other hand, digestion lag phase

(λ) and time needed to produce 80% of the maximum solubilization efficiency (T80) did not show the significant difference among all dissolved lignin concentration.

Fig. 3-15 (B) summarizes the toxic effect of dissolved lignin on hydrolysis by normalized hydrolysis activity (NHA). The NHA of cellulose declined for 6.8%, 24.4%, 25.4% and 34.8% at dissolved lignin 0.5, 1.0, 2.5 and 5.0 g L-1, respectively. To date, the toxic effect of dissolved lignin on anaerobic hydrolytic bacteria has not been investigated, but it is expected that the damage to the microbial cell membrane can occur by dissolved lignin, as well as methanogens. In addition, a number of previous researches have reported that lignin adsorbs cellulase and inhibits the enzymatic hydrolysis of cellulose (Palonen et al., 2004; Ximenes et al., 2010; Rahikainen et al., 2013). It is known that cellulase adsorption by lignin is mainly irreversible (Palonen et al., 2004). Cellulase is suggested to be adsorbed on lignin via hydrophobic interactions, electrostatic interactions and hydrogen bonding interactions (Palonen et al., 2004; Berlin et al., 2006; Nakagame et al., 2010), although the exact mechanism of cellulase adsorption on lignin has not been elucidated yet. In addition, cellulase is most strongly inhibited by lignin, as compared with other enzymes such as xylanase and glucosidase (Berlin et al., 2006). The maximum cellulase adsorption capacity of lignin has been reported as 75 to 103 mg per gram of lignin (Nakagame et al., 2011; Nonaka et al., 2011).

Thus in the present study, the considerable amount of cellulase was probably adsorbed on the dissolved lignin, which may contributed to the inhibition of cellulose hydrolysis. During biochemical hydrolysis of organic compound, microorganisms generally either secretes extracellular enzymes to the bulk liquid (Jain et al., 1992), or attaches to a particle and produce intracellular enzymes in its

vicinity (Vavilin et al., 1996). Endo et al. (1982) reported that cellulolytic microorganisms secrete extracellular cellulase to the bulk liquid when the SRT is sufficiently long (<13 days). The present study was conducted in batch mode for 23 days, indicating the SRT is longer than at least 23 days. In addition, it is known that a number of anaerobic cellulolytic bacteria secretes extracellular multi- enzyme complex called cellulosome (Schwarz, 2001). From these perspectives, cellulolytic bacteria mainly secreted extracellular cellulase for cellulose hydrolysis and the substantial amount of cellulase and/or cellulosome was adsorbed on dissolved lignin in the present study.

The metabolic pathway of anaerobic hydrolysis and acidogenesis of cellulose was summarized in Fig. 3-16, based on Endo et al. (1982) and McCarty and Smith (1986). In general, anaerobic cellulose degradation (excluding methanogenesis) consists of the following multiple steps: secretion of cellulase (1), enzymatic hydrolysis of cellulose into glucose (2), conversion of glucose into pyruvate in glycolysis for cellular energy production (3), excretion of VFAs and gases (H2 and

CO2) as waste product of pyruvate metabolism (4), production of acetic acid, H2 and CO2 from VFAs

(5). Pathway (1) to (4) are taken place by cellulolytic bacteria, while pathway (5) is governed by acetogenic bacteria, respectively. The present study exhibited acidogenic activity was inhibited in only small extent at high dissolved lignin conditions, while hydrolysis activity was more strongly inhibited at even lower dissolved lignin condition. The low toxic effect on acidogenesis implies that the pathway (3) to (5) were only slightly inhibited. Therefore, it was speculated that pathway (1) and

(2), i.e. secretion of cellulase and extracellular enzymatic hydrolysis, may be the most susceptible steps against dissolved lignin in anaerobic digestion process.

4. Conclusion

Alkaline pre-treatment greatly increased the SCOD/TCOD and reduced lignin of submerged macrophytes, clearly indicating the enhancement of the hydrolysis. Ferulic acid content greatly reduced with the increase of NaOH dose, implying the removal of lignin-ferulate complex from the lignocellulose. The cumulative CH4 yield of pre-treated P. maackianus and Eg. densa was 50.6% and

23.9% higher than un-treated, indicating that alkaline pre-treatment could be effective for anaerobic digestion of submerged macrophytes. The present study suggested that the pre-treatment with high

NaOH dose to lignin-rich substrate could cause the inhibition of anaerobic digestion, possibly by the release of dissolved lignin from lignocellulose during pre-treatment. Dissolved lignin inhibited methanogenesis, acidogenesis and hydrolysis steps, and extracellular enzymatic hydrolysis may be the most susceptible step against dissolved lignin in anaerobic digestion process.

2014 2013b 2013b 2013a & Zilouei et al. 2015 al. et et al. al. 2011 et et al. 2009 al. et Reference 2009 al. et Zheng Lin al. He2009 et Xie 2012 al. et Monlau Sambusiti et al. Sambusiti et al. Chen et al. 2014 2014 al. et Costa Taherdanak Chufo Sambusiti et al. al. Jin2014 et 2015 al. et Monlau 2010 al. et Teghammar 2011 al. Nieves et yield yield 4 25.3 12.2 25.7 31.9 90.5 36.9 38.7 34.9 67.2 34.2 38.2 48.1 70.2 38.1 CH increase (%) ) -1 VS . - treated 297 332 191 313 125.2 404 102.0 211 322 456 452 259 341 361 369 405 243 - yield yield 4 Pre (mL g ) -1 lignocellulosics VS - treated treated 139 200 160 169 333 326 192 204 269 267 168 238 249 176 - Cumulative CH Un (mL g treated Batch Batch Batch Batch Batch Batch Batch Batch Batch Batch Batch Batch Batch -continuous-continuous-continuous 237 296 152 Semi Semi Semi Batch/Continuous 35 55 35 37 35 35 35 35 35 37 35 35 35 35 55 35 AD temp. bunches stem bagasse -1. Research on anaerobic digestion of Alkali (NaOH) Alkali (NaOH) Research pre- -1. on anaerobic digestion of Table 3 Table Substrate Corn stover sludge paper and Pulp Rice straw Grass silage Sunflower stalk straw Wheat sorghum forage Ensiled Asparagus Sugarcane plant Wheat Teff straw sorghum forage Ensiled Switchgrass Sunflower stalk Paper tube residual Oil palm empty

67 Table 3-2. Chemical composition of submerged macrophytes used in the experiments of thermochemical pre-treatment.

Parameter Unit Potamogeton maackianus Egeria densa

Total solids (TS) %-wwt 9.7 4.9 Volatile solids (VS) %-wwt 8.2 4.0 VS/TS % 84.5 81.6 Total COD g kg-wwt-1 110.3 50.4 Cellulose %-dwt 36.2 36.2 Hemicellulose %-dwt 11.4 1.9 Lignin %-dwt 20.7 4.4 Ash %-dwt 15.5 18.4

68 Table 3-3. TS and lignocellulose reduction of Potamogeton maackianus and Egeria densa thermochemically pre-treated at 80ºC, 3.0 h.

Removal efficiency of lignocellulose Submerged NaOH loading rate TS reduction macrophytes (g g-TS -1) (%-TS) Cellulose Hemicellulose Lignin substrate (%) (%) (%) 69 P. maackianus Un-treated －－－－ 0.100 34.4 ± 7.1 55.0 ± 4.6 29.5 ± 18.7 59.7 ± 0.9 0.200 49.3 ± 1.7 71.9 ± 1.0 84.8 ± 4.9 79.2 ± 2.7

E. densa Un-treated －－－－ 0.100 36.9 ± 3.6 50.6 ± 1.8 8.4 ± 9.7 43.1 ± 12.6 0.200 41.6 ± 1.5 60.1 ± 1.4 60.7 ± 15.6 57.8 ± 3.6 Table 3-4. Compositional change of lignin phenols in the solid fraction of submerged macrophytes before and after thermochemical pre-treatment at 80ºC, 3.0h.

-1 Pre-treatment condition (NaOH loading rate: g g-TSsubstrate ) Lignin phenol Unit P. maackianus E. densa (lignin type) Un-treated NaOH 0.100 NaOH 0.200 Un-treated NaOH 0.100 NaOH 0.200

p-hydroxybenzaldehyde (H) μg/g 694.0 651.6 706.6 607.7 432.8 442.9 p-hydroxyacetophenone (H) μg/g 51.8 63.9 92.0 58.7 44.1 64.3 p-hydroxybenzoic acid (H) μg/g 9140.4 6529.5 2246.8 729.1 668.7 542.9 Vanillin (G) μg/g 337.2 502.7 1551.0 109.0 124.4 227.4 Acetovanillin (G) μg/g 477.5 609.6 632.7 111.6 92.3 74.3 70 Vanillic acid (G) μg/g 971.5 816.5 947.3 260.4 205.8 220.0 Syringaldehyde (S) μg/g 146.1 199.4 705.7 N. D. 54.9 89.3 Acetosyringone (S) μg/g 750.5 750.7 656.4 91.4 98.2 77.3 Syringic acid (S) μg/g 665.6 605.8 721.5 N. D. N. D. N. D. p-coumalic acid (Hydroxycinnamic) μg/g 1035.0 656.5 718.1 1211.3 769.1 649.2 Ferulic acid (Hydroxycinnamic) μg/g 3913.6 1628.1 1135.7 484.7 253.6 216.8

G total μg/g 1786.2 1928.8 3131.0 480.9 422.5 521.7 S total μg/g 1562.3 1555.9 2083.7 91.4 153.0 166.6 H total μg/g 9886.2 7245.0 3045.5 1395.5 1145.5 1050.1 Hydroxycinnamyl total μg/g 4948.6 2284.6 1853.8 1696.0 1022.8 865.9

N. D. = not detected. Table 3-5. Cumulative methane yield and methane conversion efficiency of alkali pre- treated and untreated Potamogeton maackianus and Egeria densa.

Cumulative CH conversion Submerged NaOH loading rate 4 CH yield efficiency macrophytes (g g-TS -1) 4 substrate (mL g-VS-1) (%-COD)

P. maackianus Un-treated 161.3±23.0 33.6±4.8 0.100 231.1±12.3 48.2±2.6 0.200 242.8± 8.9 50.6±1.9

E. densa Un-treated 299.7±23.8 63.2±5.0 0.100 312.8± 3.6 66.0±0.8 0.200 371.2± 2.0 78.3±0.4

71 Table 3-6. Chemical characteristics of dissolved lignin purified from alkali pre-treatment liquor of Potamogeton maackianus.

Parameter Unit Values

Total solids (TS) %-wwt 100.0* Volatile solids (VS) %-dwt 97.3 Total sugar %-dwt 3.6 Ash %-dwt 2.7

Lignin phenol (lignin type)

Vanillin (G) %-lignin 1.2 Acetovanillin (G) %-lignin 2.7 Vanillic acid (G) %-lignin 5.5 Syringaldehyde (S) %-lignin 0.3 Acetosyringone (S) %-lignin 6.6 Syringic acid (S) %-lignin 3.6 p-hydroxybenzaldehyde (H) %-lignin 3.2 p- hydroxyacetophenone (H) %-lignin 0.4 p- hydroxybenzoic acid (H) %-lignin 44.6 p-coumaric acid (hydroxycinnamic) %-lignin 6.7 Ferulic acid (hydroxycinnamic) %-lignin 25.2

G total 9.4 S total 10.6 H total 48.1 Hydroxycinnamic total 31.9

S/G ratio 1.1

* Dried purified lignin

72 Table 3-7. Simulation of methane production from acetate + dissolved lignin by using modified Gompertz equation.

0.0 g L-1 0.5 g L-1 1.0 g L-1 2.5 g L-1 5.0 g L-1

λ (days) 0.32 ± 0.04 0.28 ± 0.06 0.21 ± 0.03 0.27 0.28 ± 0.13 -1 -1 ± ± ± Rmax (NmL g-VS day ) 102.6 ± 3.6 102.9 1.4 101.5 4.8 95.4 70.3 0.4 73 P (NmL g-VS-1) 73.2 ± 2.6 79.8 ± 4.5 91.4 ± 2.3 101.1 124.4 ± 8.8 T80 (days) 0.98 ± 0.01 0.99 ± 0.03 1.03 ± 0.04 1.24 1.91 ± 0.02 R2 0.994 0.994 0.994 0.996 0.984

λ: digestion lag phase

Rmax: maximum CH4 production rate P: theoretical CH4 production potential T80: time needed to produce 80% of the maximum CH4 production Table 3-8. Simulation of acidogenesis efficiency of cellulose under various dissolved lignin concentration by using modified Gompertz equation.

0.0 g L-1 0.5 g L-1 1.0 g L-1 2.5 g L-1 5.0 g L-1

λ (days) 2.3 ± 0.2 2.1 ± 0.1 2.2 2.3 ± 0.1 2.3 ± 0.4 -1 ± ± ± Rmax (%-COD day ) 22.9 ± 1.7 22.2 1.3 22.2 17.9 1.1 16.9 1.6 74 P (%-COD) 96.3± 1.1 94.9 ± 0.2 91.4 92.5 ± 0.8 86.8 ± 1.0 T80 (days) 6.2 ± 0.1 6.0 ± 0.1 5.9 7.0 ± 0.2 7.0 ± 0.3 R2 0.997 0.998 0.996 0.998 0.996

λ : digestion lag phase

Rmax: maximum acidogenesis rate P: theoretical acidogenesis efficiency T80: time needed to produce 80% of the maximum acidogenesis efficiency Table 3-9. Simulation of hydrolysis efficiency of cellulose under various dissolved lignin concentration by using modified Gompertz equation.

0.0 g L-1 0.5 g L-1 1.0 g L-1 2.5 g L-1 5.0 g L-1

λ (days) 2.2 ± 0.2 2.1 ± 0.1 2.4 2.1 ± 0.3 1.8 ± 0.4 -1 ± ± ± Rmax (%-COD day ) 30.1 ± 4.6 27.4 1.0 24.0 24.5 0.9 17.6 4.6 75 P (%-COD) 93.3± 5.9 85.4 ± 3.4 68.6 69.1 ± 8.7 65.0 ± 3.4 T80 (days) 5.1 ± 0.1 4.9 ± 0.1 5.0 4.6 ± 0.2 5.2 ± 0.4 R2 0.999 0.998 0.998 0.996 0.999

λ : digestion lag phase

Rmax: maximum solubilization rate P: theoretical solubilization efficiency T80: time needed to produce 80% of the maximum solubilization efficiency (A)

Arabinoxylan Ferulic acid Lignin unit

OCH O 3 Xyl O OCH3 HO γ α O O O HO β Ester bond Ether bond OCH (alkali labile) 3 (alkali stable)

(B) p-Coumaric acid Lignin unit

OCH O 3 γ α HO O β O

Ester bond OCH3 (alkali labile)

Fig. 3-1. Formation of (A) lignin-carbohydrate complex and (B) lignin- hydroxycinnamic acid complex by ferulic and p-coumaric acid (adapted from Buranov and Mazza 2008).

76 Alkaline pre-treatment of P. maackianus (NaOH0.2g g-TS-1, 80ºC, 3.0 hrs)

Pass the pre-treatment liquor through the sieve (125µm)

Precipitation of lignin at pH 2 by HCl

Centrifugation (8,000 rpm, 25ºC, 15 min)

Supernatant Extraction of lipids with hexane (soluble sugar etc.) by using Soxhlet method

Hexane extractive Extraction of lignin with dioxane:water (lipid etc.) (96:4 v/v) at 80ºC, 3.0 hrs

Dioxane insoluble residue Concentration of dioxane lignin by (add dioxane and extract rotary evaporator lignin again)

Precipitation/washing of dioxane lignin with 0.1M HCl

Soluble sugar etc. Washing of lignin with diethyl ether

Drying Soluble sugar etc. Purified lignin

Fig. 3-2. Purification steps of dissolved lignin by dioxane extraction method (Björkman 1956; Chin et al. 2013; Zhang et al. 2013).

77 78

CH4 gas volume CO2 fixing unit Incubation unit measure device (3M NaOH aq.)

Fig. 3-3. Diagram of methanogenic activity test by using Automatic Methane Potential Test System (AMPTS II). Potamogeton maackianus 70 70 60ºC 80ºC 60 60

50 50

40 40

30 30

SCOD/TCOD (%) 20 20

10 10

0 0 0 1 2 3 0 1 2 3 4 5 Treatment time (h) Treatment time (h)

Egeria densa 70 70 60ºC 80ºC 60 60

50 50

40 40

30 30

20 20 SCOD/TCOD (%) 10 10

0 0 0 1 2 3 0 1 2 3 Treatment time (h) Treatment time (h)

Fig. 3-4. SCOD/TCOD of Potamogeton maackianus and Egeria densa after thermochemical pre-treatment under different NaOH loading rate of 0 ( ), 0.025 ( ), 0.050 ( ), 0.100 ( ) -1 and 0.200 ( ) g g-TSsubstrate .

79 Potamogeton maackianus

800 800 60ºC 80ºC ) 1

- 700 700 TS - 600 600 500 500 400 400 300 300 200 200

Lignocellulose g (mg Lignocellulose 100 100 0 0 0.00 0.05 0.10 0.20 0.00 0.05 0.10 0.20 -1 NaOH loading rate (g g-TSsubstrate )

Egeria densa

800 800 60ºC 80ºC ) 1

- 700 700 TS - 600 600 500 500 400 400 300 300 200 200

Lignocellulose g (mg Lignocellulose 100 100 0 0 0.00 0.05 0.10 0.20 0.00 0.05 0.10 0.20

-1 NaOH loading rate (g g-TSsubstrate )

Fig. 3-5. Changes of lignocellulose composition in the solid fraction of Potamogeton maackianus and Egeria densa after thermochemical pre-treatment of 3.0 h: cellulose (□), hemicellulose ( ), lignin (■).

80 8.4 (A) 8.2

8.0

7.8 pH 7.6

7.4 NaOH 0.2 g g-TS-1 NaOH 0.1 g g-TS-1 7.2 Un-treated

7.0 0 2 4 6 8 10 12 14 Operation time (day)

8.4 (B) 8.2

8.0

7.8 pH 7.6

7.4 NaOH 0.2 g g-TS-1 NaOH 0.1 g g-TS-1 7.2 Un-treated

7.0 0 2 4 6 8 10 12 14 Operation time (day)

Fig. 3-6. pH variation during batch anaerobic digestion of submerged macrophytes after thermochemical pre-treatment.

81 )

1 140 - (A) day 1 - 120 VS - g 100

NaOH 0.2 g g-TS-1

rate (mL rate (mL 80 NaOH 0.1 g g-TS-1 60 Un-treated production production

4 40 CH 20

0 0 2 4 6 8 10 12 14

Operation time (day)

) 140 1 - (B) day

1 120 - VS - g 100

-1 80 NaOH 0.2 g g-TS rate (mL rate (mL NaOH 0.1 g g-TS-1 60 Un-treated

production production 40 4 CH 20

0 0 2 4 6 8 10 12 14

Operation time (day)

Fig. 3-7. Methane production rates of pre-treated (80ºC, NaOH loading rate -1 0.1 and 0.2 g g-TS substrate) and untreated Potamogeton maackianus and Egeria densa (B).

82 400 (A) 350 ) -1 300 VS - 250 mL g ( 200 yield yield 4 150 CH

100 NaOH 0.2 g g-TS-1 NaOH 0.1 g g-TS-1 50

Cumulative Cumulative Un-treated 0 0 2 4 6 8 10 12 14 Operation time (day)

400 (B) ) -1 350 VS - 300 mL g ( 250 yield yield 4 200 CH

150 NaOH 0.2 g g-TS-1 100 NaOH 0.1 g g-TS-1 Cumulative Cumulative Un-treated 50

0 0 2 4 6 8 10 12 14 Operation time (day)

Fig. 3-8. Cumulative methane yields of pre-treated (80ºC, NaOH loading -1 rate 0.1 and 0.2 g g-TS substrate) and untreated Potamogeton maackianus (A) and Egeria densa (B).

83 60

40 yield (%) yield 4 30

20 Grass silage a

Increase Increase of CH Asparagus stem b 10 E. densa ★ P. maackianus ★ 0 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8

-1 Estimated dissolved lignin concentration (g Ldigestate )

Fig. 3-9. Relationship between the increase of CH4 yield of various lignocellulosic biomass by NaOH pre-treatment and the estimated dissolved lignin concentration in the digestate. The increase of CH4 yield by NaOH pre- treatment is determined by the comparison with CH4 yield of controls (no pre-treatment). a Xie et al. 2012; b Chen et al. 2014; ★ The present study.

84 140 (A)

120

100 yield (NmL) (NmL) yield

4 80

60 0.0 g/L 40 0.5 g/L Cumulative CH Cumulative 1.0 g/L 20 2.5 g/L 5.0 g/L 0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 Operation time (day)

110 (B)

control) 100 (% -

70 Methanogenic activity

Normalized Normalized 50 0.0 1.0 2.0 3.0 4.0 5.0 6.0 Dissolved lignin (g L-1)

Fig. 3-10. Effect of dissolved lignin on methanogenesis under various dissolved lignin conditions during methanogenic toxicity test: (A) methane production from acetate + dissolved lignin, (B) normalized methanogenic activity.

85 9.0

8.0

pH 0 g/L 0.5 g/L 7.0 1.0 g/L 2.5 g/L 5.0 g/L Control

6.0 0 5 10 15 20 25 Operation time (day)

Fig. 3-11. Time course of pH during hydrolysis and acidogenesis toxicity test of dissolved lignin.

86 450 (A) ) 1 - 400 VS -

g 350

300

yield (mL (mL yield 250 4 0.0g/L 200 0.5g/L 150 1.0g/L 100 2.5g/L Cumulative CH Cumulative 50 5.0g/L

0 0 5 10 15 20 25 Operation time (day)

9.0 (B) 8.0

7.0

6.0 ) 1 - 5.0

4.0 SCOD (g L 3.0

2.0

1.0

0.0 0 5 10 15 20 25 Operation time (days)

Fig. 3-12. Hydrolysis of cellulose under various dissolved lignin conditions: (A) cumulative CH4 yield, (B) SCOD concentration in digestate.

87 400 600 (A) 0g/L (D) 350 ) 1 ) 0.5g/L

500 - 1 - 300 1.0g/L 400 250 2.5g/L 300 200 5.0g/L 150 VFAs (mg (mg L VFAs

VFAs (mg (mg L VFAs 200 100 100 50 0 0 0 5 10 15 20 25 0 5 10 15 20 25 Operation time (day) Operation time (day) 400 400 (B) (E) 350 Acetic acid 350 ) ) 1 - 1 - 300 Propionic acid 300 250 n-Butyric acid 250 200 i-Butyric acid 200 n-Valeric acid 150 150 VFAs (mg (mg L VFAs

VFAs (mg (mg L VFAs i-Valeric acid 100 100 50 50 0 0 0 5 10 15 20 25 0 5 10 15 20 25 Operation time (day) Operation time (day)

400 (C) 400 (F)

350 ) 350 ) 1 - 1 - 300 300 250 250 200 200

150 (mg L VFAs 150 VFAs (mg (mg L VFAs 100 100 50 50 0 0 0 5 10 15 20 25 0 5 10 15 20 25 Operation time (day) Operation time (day)

Fig. 3-13. Time course of VFA concentration during hydrolysis and acidogenesis toxicity test of dissolved lignin at various dissolved lignin concentration: (A) total VFAs, (B) lignin 0 g L-1, (C) lignin 0.5g L-1, (D) lignin 1.0 g L-1, (E) lignin 2.5 g L-1, (F) lignin 5.0 g L-1.

88 100 (A)

COD) 80 -

40 0.0 g/L 0.0 g/L simulated 0.5 g/L 0.5 g/L simulated 20 1.0 g/L 1.0 g/L simulated Acidogenesis efficiency (% 2.5 g/L 2.5 g/L simulated 5.0 g/L 5.0 g/L simulated 0 0 5 10 15 20 25 Operation time (day)

110 (B)

control) 100 -

60 Normalized acidogenic activity (% activity acidogenic Normalized 50 0.0 1.0 2.0 3.0 4.0 5.0 6.0 Dissolved lignin (g L-1)

Fig. 3-14. Acidogenesis efficiency (A) and normalized acidogenic activity (B) during hydrolysis and acidogenesis toxicity test of dissolved lignin.

89 120 (A)

100 COD) - 80

40 0.0 g/L 0.0 g/L simulated 0.5 g/L 0.5 g/L simulated 1.0 g/L 1.0 g/L simulated Hydrolysis efficiency (% efficiency Hydrolysis 20 2.5 g/L 2.5 g/L simulated 5.0 g/L 5.0 g/L simulated

0 0 5 10 15 20 25 Operation time (day)

110

） (B)

100 control -

90 activity (% activity 80

60 Normalized hydrolytic hydrolytic Normalized 50 0.0 1.0 2.0 3.0 4.0 5.0 6.0 Dissolved lignin (g L-1)

Fig. 3-15. Effect of dissolved lignin on hydrolysis process: (A) hydrolysis efficiency, (B) normalized hydrolysis activity.

90 Hydrolysis step

Cellulose

Cellulase (2)

(1) Glucose

(glycolysis) (3)

Pyruvate

[Cellulolytic bacteria] Propionic acid (4) Butyric acid Valeric acid

(4)

(5) Acetic acid (Reductive acetyl- CO2 CoA pathway) H 2 [Acetogenic bacteria]

Acidogenesis step

Fig. 3-16. Hydrolysis and acidogenesis of cellulose in anaerobic digestion process (modified from Endo et al. 1982; McCarty and Smith 1986)

Chapter IV

Effect of alkaline pre-treatment on semi-continuous anaerobic

digestion of submerged macrophyte

4.1. Introduction

Although most full-scale anaerobic digesters are operated either in continuous or semi- continuous mode, the effect of pre-treatment on anaerobic digestion has been investigated mostly in batch mode (Taherzadeh and Karimi, 2008; Hendriks and Zeeman, 2009; Carrere et al., 2016). Batch tests are useful for evaluating methane production potential under specific parameters such as different pre-treatment and the pre-treatment conditions (Fernandes et al., 2009; Xie et al., 2011; Chen et al., 2014b), digestion kinetics (Lay et al., 1997; Li et al., 2015) and inhibitory levels of toxic substances (Vidal and Diez, 2005; Vavilin et al., 2008; Gonzalez-Estrella et al., 2013) in a short period of time. However, continuous (or semi-continuous) operation should be conducted even in lab-scale experiment, in order to estimate the digestibility, process stability and inhibitory effect in full-scale anaerobic digesters. As summarized in Fig. 3-1, only a few researches on anaerobic digestion of alkali pre-treated lignocellulosic biomass have operated in continuous mode. Sambusiti et al. (2013) reported NaOH pre-treatment of ensiled sorghum forage improved the CH4 production by 25% as compared with un-treated sorghum in continuously stirred tank reactor (CSTR). Monlau et al. (2015) also confirmed that the CH4 recovery of sunflower stalk in semi-continuous operation increased by

26-29% due to NaOH pre-treatment. From these point of view, Carrere et al. (2015) suggested in their review paper that alkaline pre-treatment may be the most suitable method for enhancing the anaerobic

digestion process for lignocellulosic biomass. However, these studies focused only on changes in chemical characteristics of substrates (e.g. lignocellulose content), improvement of CH4 recovery, and accumulation of VFA. Alkali pre-treated lignocellulosic biomass produces not only readily degradable soluble sugars and proteins, but also dissolved lignin-derived compounds (Taherzadeh and Karimi, 2008; Barakat et al., 2012; Quéméneur et al., 2012). As confirmed in Chapter III, dissolved lignin inhibits anaerobic digestion process especially hydrolysis step. In addition, lignin is scarcely degraded in anaerobic environment (Tuomela et al., 2000). Accordingly, it is expected that dissolved lignin accumulates in digestate under continuous operation of alkali pre-treated submerged macrophyte and may cause the adverse effect on the anaerobic digestion process.

On the other hand, it is known that anaerobic microbes can be adapted to the inhibitory substances during long term operation. The microorganisms can retain viability at concentrations far exceeding the initial inhibitory levels, once they acclimatized to toxic substances (Chen et al., 2008).

A number of studies have reported the acclimatization of anaerobic microorganisms on phenolic compounds during anaerobic digestion treatment (Healy and Young, 1979; McCarty and Smith, 1986;

Carbajo et al., 2010; Rosenkranz et al., 2013). Carbajo et al. (2010) treated the highly concentrated phenol synthetic wastewater in both continuous and batch modes. Although VFAs degradation was inhibited in batch mode, no inhibition was observed at steady state in continuous operation, probably due to the acclimatization of microorganisms. Rosenkranz et al. (2013) also reported the acclimatization of anaerobic microbes to phenol, and observed the shift of microbial community structure at higher phenol concentration. Accordingly, the methane recovery of alkali pre-treated P. maackianus may be enhanced by microbial acclimatization during continuous operation.

In this Chapter, the effect of alkaline pre-treatment on anaerobic digestion of lignin-rich submerged macrophyte was investigated in semi-continuous mode. The process stability in terms of

CH4 recovery and acclimatization to accumulating dissolved lignin were evaluated.

4.2. Materials and method

4.2.1. Substrate and inoculum

P. maackianus was used as a sole substrate, of which sample collection and preservation were same with Chapter II and III. The pre-treatment condition of P. maackianus was 80ºC, NaOH 0.2 g

-1 g-TSsubstrate for 3.0 h, followed by neutralization to pH 7 using HCl, according to the results of

Chapter III. For the inoculum of semi-continuous experiment, the mesophilic anaerobic sludge was obtained from Hokubu Sludge Treatment Center, Yokohama, Japan.

4.2.2. Semi-continuous anaerobic digestion of alkali pre-treated or un-treated P. maackianus

Alkali-treated and un-treated P. maackianus was anaerobically digested in semi-continuous mode at mesophilic (37°C) condition in a temperature controlled laboratory. Trace element solution

-1 -1 -1 -1 (FeCl3.6H2O 4.46 mg L , ZnSO4.7H2O 0.15 mg L , CoCl2.6H2O 0.17 mg L , H3BO3 0.06 mg L ,

-1 -1 -1 MnCl2.4H2O 0.5 mg L , NiCl.6H2O 0.04 mg L , CuCl2.2H2O 0.03 mg L , NaMoO4.2H2O 0.03 mg

L-1) was added with substrate according to Sekiguchi et al. (1998). For anaerobic digestion bioreactor, completely stirred tank reactor (CSTR) of the working volume of 4.5 L was used for alkali pre-treated macrophytes. Organic loading rate (OLR) was 1.0 g-VS L-1 day-1. Alkali pre-treated P. maackianus was fed to CSTR every two days for the period of 120 days. CSTR was operated at 37°C, HRT 40

days, and continuous mixing of 100 rpm. Un-treated P. maackianus was also digested at mesophilic condition as a control, with same OLR and HRT conditions in 6.0 L CSTR. The headspace of CSTR was filled with nitrogen gas to make anaerobic environment. Produced biogas was collected using

10-L aluminum gas bag (GL Sciences, AAK-10, Japan), and the gas volume was measured every two days by using water displacement method.

4.2.3. Analytical parameters

pH, salinity, TS, VS, soluble COD (SCOD), volatile fatty acids (VFA), lignocellulose and biogas were measured. The pH of the samples was measured using a pH meter (HORIBA, B-212,

Japan). Salinity was measured by using salinity meter (ATAGO, ES421, Japan). Standard methods from APHA (American Public Health Asspciation, 1998) were applied to the analysis of TS, VS and

COD. Lignocellulose (cellulose, hemicellulose and lignin) content was measured by detergent system

(Van Soest et al., 1991) by using fiber analyzer (ANKOM Technology, A-200, USA).

Biogas (CH4, CO2) composition was monitored using a gas chromatograph (SHIMADZU,

GC-2014, Japan) equipped with a packed column (Shimadzu, Shincarbon ST, 6.0 m long, 3 mm I.D.,

Japan) and a thermal conductivity detector (TCD). The temperature of injector and detector were maintained at 120ºC and 260ºC, respectively. The column temperature was gradually increased from

40ºC to 250ºC. Helium was used as carrier gas with the flow rate of 40 mL min-1. VFAs composition

(acetic, propionic, n-butyric, i-butyric, n-valeric and n-valeric acid) was monitored using a gas chromatograph (Hitachi, G-3000, Japan) equipped with a capillary column (Stabilwax, size×I.D. 30 m×0.53 mm, 1.00 µm film thickness) and a flame ionized detector (FID). The temperature of injector

and detector were maintained at 240ºC. The column temperature was gradually increased from 100ºC to 160ºC at 4 ºC minutes-1. Nitrogen was used as carrier gas with the flow rate of 40 mL min-1.

4.2.4. Fractionation of soluble organics and estimation of dissolved lignin/humic substances in

the digestate

In order to clarify the composition of SCOD, acid precipitation was conducted for fractionation of SCOD. Digestate was firstly filtered through 0.45µm glass fiber filter (ADVANTEC,

GC-50, 47 mm, Japan) to remove particle, and SCOD and VFAs of the filtrate was measured. Next,

3% H2SO4 solution was added to the filtered digestate, in order to separate acid precipitate fraction from total SCOD. Acid precipitate is assumed to be composed mainly of lignin or humic substances.

After acid precipitation, the supernatant was filtered through 0.45µm glass fiber filter and SCOD was measured. The amount of acid precipitate was calculated by subtracting supernatant SCOD from total

SCOD.

4.2.5. Calculation

The time course of theoretical dissolved lignin concentration in the digestate was calculated based on dissolved lignin concentration of alkali pre-treated P. maackianus by consulting Usack et al.

(2012). When a chemical species with aqueous solution, slurry or solid is continuously fed to the reactor, the concentration of the chemical species increases with operation time, and eventually reaches to the steady state. The concentration of chemical species in time n (Cn) can be calculated by using the following equation.

Cn = −𝑘𝑘𝑒𝑒× 𝑡𝑡 (11) Csub × �1 − e �

where Csub is the concentration of chemical species in the substrate and ke is the elimination rate constant of the chemical species.

Hypothesizing dissolved lignin does not degrade in anaerobic digestion, the elimination rate of dissolved lignin in anaerobic digester relies on the HRT (i.e. ke = 1/HRT). Thus the theoretical concentration of dissolved lignin in digestate can be calculated by using the following equation.

-1 1 Theoretical dissolved lignin in day n (g L ) = −HRT× 𝑡𝑡 (12) Csub �1 − e �

where Csub is the concentration of dissolved lignin in alkali pre-treatment liquor (i.e. pre-treated substrate) (g L-1), t is the operational time (day) and HRT is the hydraulic retention time of CSTR

(day).

4.3. Results and discussion

4.3.1. pH and salinity variation in the digestate

In both alkali pre-treated and un-treated conditions, pH in the digestate fluctuated from 7.0 to 8.2

(Fig. 4-1). The obtained pH values were in the optimum range for methanogenesis of 6.5 to 8.2

(Speece, 1996). Overall, alkali pre-treated condition exhibited slightly higher pH values as compared with that of un-treated. Sambusiti et al. (2013) also reported higher pH at alkali pre-treated sorghum than un-treated one, due to the higher alkalinity in the system. This result would be a benefit of anaerobic digestion process coupled with alkaline pre-treatment, because pH drop by the

accumulation of organic acid would be less likely to occur.

As described in Chapter III, up to 15 g L-1 of NaCl is theoretically produced in alkaline pre- treatment liquor due to the neutralization of NaOH by HCl. In semi-continuous operation, salinity level in the digestate of alkali pre-treated condition gradually increased with operation time and reached steady state of approximately 17 g L-1 (Fig. 4-2). Since anaerobic digestion process is known to be inhibited at NaCl >20-30 g L-1 (Feijoo et al., 1995; Kimata-Kino et al., 2011), this salinity level does not cause the inhibition of methane recovery in the present study. For un-treated condition, the inhibition by salinity is no need to concern because no NaCl is produced.

4.3.2. Accumulation of dissolved lignin in the digestate

Fig. 4-3 shows the fractionation of SCOD in the digestate of un-treated and alkali pre-treated

P. maackianus. In alkali pre-treated condition, the total SCOD concentration, which is the sum of acid precipitate and others, kept increase from the startup and fluctuated at high SCOD value of 13 to 20 g-COD L-1 at day 38 to 120. The acid precipitate was the main component of SCOD in alkali pre- treated condition, which accounted for 79.4±5.0% of the total SCOD during day 38-120. In un- treated condition, total SCOD concentration was about a half as compared with alkali-treated condition, but still high total SCOD concentration of 9.6 g-COD L-1 was observed on day 50. Since dissolved lignin is much less likely to leach out in un-treated condition, the increase of acid precipitate was speculated to be hydrolysable tannin in the macrophyte cells and/or humic substances produced during anaerobic digestion.

Theoretical dissolved lignin concentration in the digestate was calculated based on dissolved

lignin concentration of alkali pre-treated P. maackianus. In order to compare the concentration of theoretical dissolved lignin and acid precipitate, theoretical dissolved lignin concentration was converted from gravimetric base (g L-1) into COD base (g-COD L-1). The COD values of dissolved lignin was calculated based on lignin phenol composition of alkali pre-treated P. maackianus (Table

3-4), i.e. 1.714 g-COD g-lignin-1. Since lignin is undigested in anaerobic environment (Benner et al.,

1984), the theoretical dissolved lignin concentration in the digestate would eventually become constant at the same level in the inlet alkaline pre-treatment liquor. Fig. 4-4 shows the acid precipitate concentration and theoretical dissolved lignin concentration in the digestate of pre-treated P. maackianus. The present study demonstrated that the theoretical dissolved lignin concentration and acid precipitate concentration showed relatively similar values. Therefore, it was suggested that the acid precipitate accumulated in pre-treated condition is mostly composed of dissolved lignin produced by alkali pre-treatment.

4.3.3. Effect of dissolved lignin on methane recovery in semi-continuous operation

4.3.3.1. Performance of semi-continuous anaerobic digestion of alkali pre-treated and un-

treated P. maackianus

During semi-continuous anaerobic digestion of alkali pre-treated and un-treated P. maackianus, pre-treated condition exhibited higher CH4 recovery (Table 4-1). The cumulative CH4

-1 yield and the CH4 conversion efficiency was 219.2 mL g-VS and 45.9%-COD for alkali pre-treated

P. maackianus, and 132.6 mL g-VS-1 and 27.8%-COD for un-treated P. maackianus, respectively.

Approximately 65% improvement of the CH4 production by alkali pre-treatment was probably

attributed to the increase of biodegradable SCOD fraction (i.e. dissolved polysaccharides and cytosol derivatives) and accessible cellulose and hemicellulose in the solid fraction of P. maackianus by alkaline delignification, as described in Chapter III. Previous studies of semi-continuous anaerobic digestion of alkali pre-treated lignocellulosic biomass have reported that increase of CH4 recovery after alkaline pre-treatment is 12% to 26% as compared with un-treated biomass (Sambusiti et al.,

2013; Jin et al., 2014; Monlau et al., 2015). Accordingly, the improvement of CH4 recovery obtained in the present study was considerably higher than the reported herbaceous plants. It is probably due to the lower lignin content of the substrate (4.1%-VS, Sambusiti et al., 2013) or milder pre-treatment

-1 condition (NaOH loading rate 0.04 g g-TSsubstrate , 55ºC, 24 h, Monlau et al., 2015) as compared with the present study. With these results, it was indicated that alkaline pre-treatment is preferable anaerobic digestion process for lignin-rich submerged macrophyte, in terms of higher CH4 recovery.

4.3.3.2. Inhibition and acclimatization against dissolved lignin

Fig. 4-5 exhibits the time course of CH4 production rate of alkali pre-treated and un-treated P. maackianus and theoretical dissolved lignin concentration in the digestate of pre-treated P. maackianus. The theoretical dissolved lignin concentration was shown in gravimetric base (i.e. g L-

1), in order to compare the lignin inhibiting concentration obtained in Chapter III. In alkali pre-treated condition, the CH4 production rate started to drop from day 16 and showed lowest CH4 production rate of 59.4 mL L-1 day-1 in day 30, suggesting the inhibition of anaerobic digestion process by dissolved lignin (Fig. 4-5). The possible mechanism of the inhibition is probably the direct toxicity of dissolved lignin on bacteria and archaea (Vidal and Diez, 2005) and/or irreversible adsorption of

100

cellulolytic enzymes on dissolved lignin (Palonen et al., 2004), as speculated in Chapter III.

Thereafter, the CH4 production rate started to recover from day 42 and high CH4 production rate of

241.0 ± 32.1 mL L-1 day-1 was maintained throughout the rest of the operational period (day 46-120).

This result indicates that the acclimatization against the inhibitors seems to occur in this phase. The deterioration of CH4 production rate was observed when theoretical dissolved lignin concentration exceeded 3.3 g L-1 (day 16). In Chapter III, it was found that the methanogenic and hydrolysis activity was inhibited by dissolved lignin at ≧2.5 g L-1 and ≧1.0 g L-1 respectively. In semi-continuous operation, it seems that the anaerobic digestion process was inhibited at similar level of dissolved lignin as batch toxicity test. Thereafter, the CH4 production was recovered even at high acid precipitate concentration of ≧7 g L-1, which clearly indicates the acclimatization of anaerobic microorganisms to dissolved lignin. The inhibition and acclimatization on anaerobic digestion against dissolved lignin has never been investigated, but those against phenol has been studied in a number of researches (Sierra-Alvarez and Lettinga, 1991; Veeresh et al., 2005; Kayembe et al., 2013;

Rosenkranz et al., 2013). In anaerobic digestion of phenol, the duration of the acclimatization period is reported to ranges from 6 weeks to 10 months, depending on the nature of inoculum and the operational strategy (Veeresh et al., 2005). Accordingly, the acclimatization period of 46 days obtained in the present study is comparable to the previous studies of anaerobic digestion of phenols.

For the previous studies of semi-continuous anaerobic digestion of alkali pre-treated lignocellulosic biomass, only Sambusiti et al. (2013) showed the time course of CH4 production rate. According to their results, the CH4 production rate gradually declined from startup but no abrupt decline was observed throughout the operational period. It was presumably due to the low lignin content (4.1%-

101

VS) of the inherent nature of the substrate (ensiled sorghum forage), which is fivefold lower than P. maackianus, should have less inhibitory effect of dissolved lignin on the anaerobic digestion process as compared to the present study.

The inhibitory effect and the following acclimatization on anaerobic digestion was observed not only in CH4 recovery but also in the digestate characteristics, i.e. VFAs accumulation. The time course of VFAs concentration in the digestate of alkali pre-treated P. maackianus was shown in Fig.

4-6. From day 20, VFAs concentration dramatically increased and reached 5.1 g L-1 in day 36. During

“inhibition phase of methane recovery (i.e. day 20-36)”, acetic acid was most abundant, showing 0.7 to 4.2 g L-1, which accounts for 78.4 to 93.2% of the total VFAs. The accumulation of acetic acid clearly indicates the inhibition of acetoclastic methanogenesis. Methanogenesis inhibition usually occurs when the pH drops to <6.0-6.5 with VFA accumulation (Speece, 1996). However, in the present study, pH was slightly lowered but maintained around 7.2 to 7.5 during the inhibition phase

(Fig. 4-1). Accordingly, it was suggested that the methanogenesis inhibition was caused not by pH drop, but by the other factors such as accumulation dissolved lignin. On the other hand, the VFAs concentration started to decline after day 36, and stabilized to 0.1 to 0.4 g L-1 after day 70, indicating the anaerobic digestion process was fully recovered. It is probably due to the acclimatization of anaerobic microbes to dissolved lignin in the digestate. The acclimatization of anaerobic microorganisms on phenol has been investigated by Rosenkranz et al. (2013). They reported the elimination capacity of phenol in anaerobic sequencing batch reactor (ASBR) greatly improved from around 5 mg VSS-1 day-1 to 40 - 60 mg VSS-1 day-1 after 4 to 6 cycles of fed-batch operation.

Simultaneously, they reported the drastic change of microbial community structure when phenol

102

concentration increased. Their results cannot simply be compared with the present study since lignin monomers and phenol are biodegradable but lignin polymers are highly recalcitrant (Healy and Young,

1979; Tuomela et al., 2000; Barakat et al., 2012). Nevertheless, it can be speculated that acclimatization against dissolved lignin (polymers) by the specialization of microbial community can take place in the present study. In addition, since the majority of adsorbed cellulolytic enzymes on lignin does not leach into the bulk liquid and forms strongly-bonded “lignin-enzyme complex” on the surface of lignin (Funaoka, 1998), the adsorption capacity of lignin should be saturated at certain amount of enzymes. Palonen et al. (2004) reported that the binding affinity of added enzyme on lignin is decreased due to the high amount of cellulolytic enzymes already attached on lignin. Decrease of binding affinity, i.e. the decrease of hydrophobicity, indicates that the toxicity of lignin on microorganism could be decreased. Thus the toxic effect of dissolved lignin on anaerobic digestion process could be alleviated once adsorption of cellulolytic enzymes is saturated. Therefore, dissolved lignin can “temporary” inhibit anaerobic digestion especially during startup phase, but acclimatization to dissolved lignin, possibly due to the adaptation of microorganisms and/or saturation of enzyme adsorption on dissolved lignin, can recovers the stable anaerobic digestion process. In conclusion, the present study indicated that alkaline pre-treatment is highly beneficial for enhancing anaerobic digestibility of lignin-rich submerged macrophyte.

4.4. Conclusion

In Chapter IV, semi-continuous anaerobic digestion of alkali pre-treated and un-treated P. maackianus was conducted. By continuous feeding of alkali pre-treated macrophyte, dissolved lignin

103

-1 accumulated in the digestate up to 13-20 g-COD L . Alkaline pre-treatment enhanced the CH4 production of P. maackianus from 28%-COD (un-treated) to 46%-COD, i.e. 65% improvement. In the digestate of alkali pre-treated condition, accumulation of dissolved lignin was observed and the

CH4 production was inhibited. However, the CH4 production was recovered from day 42, which suggested the acclimatization of microorganisms and/or alleviation of toxicity of dissolved lignin.

Overall, alkaline pre-treatment was highly beneficial for enhancing anaerobic digestibility of lignin- rich submerged macrophyte.

104

Table 4-1. Cumulative CH4 yield and CH4 conversion efficiency of untreated and NaOH pre-treated P. maackianus in semi-continuous treatment.

Cumulative CH yield CH conversion Pretreatment 4 4 （mL g-VS-1 ） (%-COD)

Untreated 132.6 27.8

Pre-treated 219.2 45.9

105 9.0

8.5 Pre-treated Un-treated

8.0

pH 7.5

7.0

6.5

6.0 0 20 40 60 80 100 120 Operation time (days)

Fig. 4-1. pH variation in the digestate of alkali pre-treated and un-treated P. maackianus in semi-continuous treatment.

106 20 18 16 14 ) 1

- 12 10 8

Salinity (g L (g Salinity 6 4 2 0 0 20 40 60 80 100 120 Operation time (day)

Fig. 4-2. Variation of salinity in the digestate of alkali pre-treated P. maackianus in semi-continuous treatment.

107 25 Others (A) Acid precipitate

20 ) 1 - 15

10 SCOD (g L

0 0 10 20 30 40 50 60 70 80 90 100 110 120 Operation time (days)

25 (B)

20 Others Acid precipitate ) 1

- 15

10 SCOD (g L

0 0 10 20 30 40 50 60 70 80 90 100 110 120 Operation time (day)

Fig. 4-3. Fractionation of SCOD in the digestate of (A) alkali pre-treated and (B) un-treated P. maackianus in semi-continuous treatment.

108 18

) 12 1 - 10

8 SCOD (g L 6

4 Dissolved lignin (theoretical) 2 Acid precipitate

0 0 10 20 30 40 50 60 70 80 90 100 110 120

Operation time (days)

Fig. 4-4. Acid precipitate concentration and theoretical dissolved lignin concentration in the digestate of alkali pre-treated P. maackianus.

109 Pre-treated Un-treated 400 Dissolved lignin (theoretical) 10 9

） 350 1 - 8 day 300 1 - 7 Dissolved 250 6 （ mL L 200 5 lignin rate

150 4 （ g L 3 -

100 1 ） production

4 2

CH 50 1 0 0 0 10 20 30 40 50 60 70 80 90 100 110 120 Operation time (days)

Fig. 4-5. Time course of CH4 production rate of alkali pre-treated and un-treated P. maackianus in semi-continuous treatment and theoretical dissolved lignin concentration in the digestate of alkali pre-treated condition.

110 6

Acetic acid 5 Propionic acid n-Butyric acid 4 i-Butyric acid ) 1 - n-Valeric acid 3 i-Valeric acid

VFAs (g L (g VFAs 2

0 0 10 20 30 40 50 60 70 80 90 100 110 120 Operation time (days)

Fig. 4-6. VFAs concentration of alkali pre-treated P. maackianus in semi- continuous treatment.

111

Chapter V

General discussion

Among five submerged macrophyte species dominant in Lake Biwa (all of these species are also commonly found in all over the world), four macrophyte species exhibited relatively high CH4 recovery as compared with other lignocellulosic biomass (Chapter II). For recalcitrant macrophyte P. maackianus, alkaline pre-treatment significantly improved its CH4 recovery both in batch (Chapter

III) and semi-continuous modes (Chapter IV), indicated lignin-rich macrophytes can be degraded well in anaerobic digestion. In order to confirm the feasibility of alkaline pre-treatment, the economic and energy balance should be concerned. In this chapter, the economic and technical feasibility of alkaline pre-treatment on anaerobic digestion of submerged macrophytes was discussed. Furthermore, future studies for improving anaerobic digestion of submerged macrophytes were discussed.

5.1. Economic and thermal balance assessment of alkaline pre-treatment

In the thermal alkaline pre-treatment conducted in the present study, alkali (NaOH) and acid

(HCl for neutralization) are added, which increases the extra cost. In addition, the heating at 80ºC requires the external energy consumption. Therefore, the preliminary assessment of economic and energy balance was conducted in order to evaluate the techno-economic feasibility of alkaline pre- treatment on anaerobic digestion of submerged macrophytes by assuming full-scale treatment in Lake

Biwa, Japan. In the economic and thermal balance assessment of alkaline pre-treatment on anaerobic digestion of submerged macrophytes, (1) additional costs for chemicals (NaOH and HCl) versus

112

increase of electricity sales (i.e. benefit), and (2) extra energy input for heating versus increase of gained heat energy by biogas combustion, were compared with un-treated condition, by consulting the method described by Monlau et al. (2015).

5.1.1. Scenarios design

Daily throughput and operational period of submerged macrophytes in anaerobic digestion treatment was determined based on the harvested yield and season in Lake Biwa. Approximately

4,000 tons-wwt of submerged macrophytes have been harvested per annum in Lake Biwa (Fig. 1-4).

The harvesting season is usually between May to October, since submerged macrophytes start to grow from May and become most abundant in September, and decline from October and eliminates in winter to early spring (K. Ishikawa, personal communication, June 26, 2015). Accordingly, the operation period was presumed as 6 months (May to October), of which average temperature is assumed to be 25ºC. In order to treat 4,000 tons-wwt of submerged macrophyte in 6 months for 5 days per week (≒ 22 days per month), the daily throughput of submerged macrophytes was determined as 30 tons-wwt per day (= 4,000 tons year-1 ÷ 6 months ÷ 22 day month-1). Species composition of submerged macrophytes was determined according to the previous work of Haga and

Ishikawa (2011), which surveyed the total biomass and the species composition of submerged macrophytes in South Basin of Lake Biwa in September 2007, and found that P. maackianus accounts for 55% of all biomass, followed by Eg. densa 12.6%, C. demersum 12.4%, El. nuttallii 5.8% and P. malaianus 3.6% (Table 2-1). In the present study, it was presumed that the only five macrophyte species mentioned above existed throughout the year.

113

Operating condition of the pre-treatment was determined by the results of the present study

(Table 5-1). Alkaline pre-treatment was applied only on P. maackianus, since alkaline pre-treatment

is effective in improving CH4 yield of lignin-rich macrophyte, as described in Chapter III. The

alkaline pre-treatment condition was NaOH 0.1 g g-TS-1, 80ºC and 3.0 hours, since the cumulative

-1 CH4 yield of P. maackianus in NaOH 0.1 and 0.2 g g-TS did not show the significant difference.

During alkaline pre-treatment, macrophytes have to be soaked in NaOH solution, so that up to 15.0

tons-wwt (or 15.0 m3) of NaOH + HCl solution should be added to P. maackianus. Accordingly, the

total daily throughput to simulated full-scale anaerobic digester is 45.0 tons-wwt day-1.

In Japan, the tariff of generated electricity from biogas is currently 39 JPY (tax exclusive) per

kWh in 2015 (Ministry of Economy Trade and Industry (METI), 2015). It was assumed that 10% of

generated electricity was utilized for the operation of anaerobic digester and the surrounding

equipment, and the rest of generated electricity was sold. On the other hand, the costs of chemicals

for alkaline pre-treatment were 18 JPY per kg for NaOH (24%) (National Institute for Land and

Infrastructure Management, 2015) and 18 JPY per kg for HCl (35-37%) (CMC, 2015), respectively.

Thermal balance was calculated by the difference in the external energy consumption by alkaline pre-

treatment (i.e. heating) and the increase of thermal energy owing to the increase of CH4 recovery by

alkaline pre-treatment. In order to evaluate the external energy consumption, the heat energy

requirement (HER) was calculated according to Monlau et al. (2015) as follows:

m×Cp×(Tfinal-Tinitial) HER (kWh ton-VS-1) = × (1.0 - 0.8) (12) 3600 � �

where m is the mass of substrate + NaOH solution (kg-wwt), Cp is the water-specific heat (4.18 kJ

114

-1 -1 kg ºC ), Tinitial is the initial temperature of the substrate + NaOH solution before pre-treatment

(=25ºC) and Tfinal is the temperature after pre-treatment (=80ºC). Heat losses during alkaline pre- treatment was considered negligible. Dhar et al. (2012) conducted the techno-economic evaluation of thermal pre-treatment for enhancing anaerobic digestion of waste activated sludge and reported that up to 80% of heat energy can be recovered from pre-treated sludge. Accordingly, it was presumed that 80% of heat energy required for the pre-treatment was recovered in the present study.

5.1.2. Recovery of CH4, electricity and thermal energy

Monthly CH4 yield from submerged macrophyte under mesophilic condition (37ºC) was roughly estimated based on the species composition of submerged macrophytes in Lake Biwa and the data of the cumulative CH4 yields obtained in the present study. Generally, CH4 recovery in semi- continuous mode is slightly lower than that in batch mode, since digestate with certain amount of undigested feedstock is discharged according to the HRT of the anaerobic digester when adding new feedstock. Accordingly, the CH4 yields of each macrophyte were calculated based on the wet-weight

-1 based cumulative CH4 yield (mL g-wwt ) obtained in batch test (Chapter II), and the CH4 yields were calibrated with CH4 continuous/CH4 batch ratio of alkali pre-treated P. maackianus. The monthly CH4 yield of alkali pre-treated macrophytes was 32.4% higher than that of un-treated macrophytes (Fig. 5-1).

Generated biogas is converted to heat and electricity by using combined heat and power (CHP) system, of which electricity efficiency is 35% and thermal efficiency is 50%. Accordingly, the yield of electricity and thermal energy from submerged macrophyte in simulated full-scale treatment was determined as shown in Table 5-2.

115

5.1.3. Economic and thermal balance

Economic balance was calculated based on the difference in the cost for the chemicals required in alkaline pre-treatment and the increase of the electricity sale owing to the increase of CH4 recovery by alkaline pre-treatment. Fig. 5-2 shows the economic and thermal balance of alkali pre-treated and un-treated macrophytes. As clearly presented in Fig. 5-2 (A), the economic balance of alkaline pre- treatment became positive. This result indicates that the cost for chemicals is completely compensated by the increase of electricity sale owing to alkaline pre-treatment. The thermal energy balance of alkaline pre-treatment in simulated full-scale anaerobic digestion of submerged macrophytes were summarized in Figure 5-2 (B). By applying 80% heat recovery, the net thermal energy of alkaline pre-treatment became positive. This result indicates that no extra thermal energy input is required for alkaline pre-treatment.

5.1.4. Effect of electricity tariff of economic balance

METI has been planning to reduce the electricity tariff from renewable energy producing facilities, with the increases of prevalence rate (METI 2015), so that it is important to note that the electricity tariff generated from macrophyte biogasification would be reduced to some extent in near future. Accordingly, the effect of electricity tariff on the economic balance of alkaline pre-treatment

+ anaerobic digestion of submerged macrophytes was roughly evaluated by varying the tariff from

39 to 30 JPY kWh-1 (Fig. 5-3). It was demonstrated that the net benefit (i.e. [benefit from electricity sale] – [chemical cost]) of pre-treated macrophytes became lower than that of un-treated macrophytes from 37 JPY kWh-1. It should be noted that the price of chemicals widely varies depending on the

116

year, which implies that the net benefit of pre-treated macrophytes can be lowered below un-treated macrophytes even in the current electricity tariff, depending on the price of the chemicals in the year.

5.1.5. Effect of species composition on economic and thermal balance

The effect of species composition on economic and thermal balance was evaluated. To date, no study has quantified the seasonal variation of the species composition, but it has been empirically known that each species have different annual growth pattern. A number of studies have stated the biomass of El. nuttallii reaches the maximum in late spring to early summer, while that of other macrophytes (C. demersum, Eg. densa, P. maackianus and P. malaianus) are in summer to early autumn (Hamabata, 2005). In order to estimate the effect of species composition, the composition of

El. nuttallii was roughly changed from 10% to 70% in wet weight basis, as shown in Fig. 5-4 (A). In this scenario, the relative composition of other macrophytes were unchanged from Table 2-1.

The simulated results of economic and thermal balance under different species composition was shown in Fig. 5-4 (B) and (C). With the increase of El. nuttallii, the net benefit and heat gain of pre- treated macrophyte became closer to that of un-treated macrophyte. This result suggests that alkaline pre-treatment has advantage in the seasons when readily degradable macrophytes such as El. nuttallii declines and recalcitrant macrophytes such as P. maackianus increases. In the future, it could be economically unfeasible to apply alkaline pre-treatment when recalcitrant macrophyte is not abundant, if the electricity tariff drops. To conclude, the present study suggested that alkaline pre-treatment on anaerobic digestion of submerged macrophytes could be economically feasible in late summer to autumn in case of Japan.

117

5.2. Future studies for more efficient anaerobic digestion of submerged macrophytes

5.2.1. Lignin recovery from alkali pre-treatment liquor prior to anaerobic digestion

The present study demonstrated that the application of high NaOH loading rate to lignin-rich macrophyte not only enhances lignin removal and CH4 recovery, but also increases dissolved lignin concentration in the bulk liquid. Inhibitory effect of the dissolved lignin on anaerobic digestion process was presented in Chapter III, but Chapter IV revealed anaerobic microbes was acclimatized to the dissolved lignin and the methane production rate was recovered under long term operation (>50 days). On the other hand, another problem of dissolved lignin on anaerobic digestion process, the requirement of post-treatment of lignin-containing digestate, is still a large concern when considering the practical use. If cost for post-treatment rises up by dissolved lignin post-treatment, the techno- economic feasibility of alkaline pre-treatment would be lowered. Then, is there any strategies for treatment of dissolved lignin?

In recent years, the recovery of lignin from lignin-rich wastewater such as black liquor, pulp industry wastewater and pre-treatment liquor of plant biomass has been given its attention. The potential use of lignin is, for instance, automobile brakes, wood panel products, bio-dispersants, polyurethane foams and epoxy resins (Lora and Glasser, 2002). Variety of lignin recovery techniques including chemical coagulation, filtration, acid precipitation and electrocoagulation has been investigated for development of energy-efficient and cost-effective lignin biorefinery process (Das and Patnaik, 2000; de Pinho et al., 2000; Ibrahim et al., 2004; Uğurlu et al., 2008; Nagy et al., 2010).

Accordingly, the combination of lignin recovery process and anaerobic digestion of alkali pre-treated biomass could be more beneficial not only in bioenergy recovery but also in additional benefit from

118

recovered lignin. However, dissolved lignin and anaerobic digestion effluent (ADE) should not be mixed when considering lignin refinery, due to the reduction of lignin recovery efficiency by adsorption of lignin on sludge (Hernandez and Edyvean, 2008) and contamination with unwanted microorganism or pathogens (Chen et al., 2012). Accordingly, in order to reduce the risk of contamination of lignin with ADE, lignin recovery process should be placed prior to anaerobic digestion process.

5.2.2. Co-digestion of submerged macrophytes with other biomass

Generally, the propagation of submerged macrophytes in temperate zone including Japan is limited by season. As stated in 5.1.1. of this Chapter, the harvesting season of submerged macrophyte is usually limited to May to October. Thus the abundant macrophyte biomass can be supplied to the full-scale anaerobic digester during spring summer to autumn, while the macrophyte can be scarcely loaded to the digester during winter to early spring. The insufficient supply of organic substrate to the digester can result in the decline of microbial activity (Hwang et al., 2010). Therefore, it is important to continuously fulfill the sufficient organic loading to anaerobic digester, in order to maintain the microbial activity and the methane yield.

Anaerobic co-digestion with different substrate can be attractive in the process, for hold-up of the organic loading to anaerobic digester during wintertime. Food waste is often applied to the co- substrate in the process due to its abundance and the high biodegradability/methane yield of 364-489 mL g-VS-1 (Verrier et al., 1987; Heo et al., 2004). Accordingly, anaerobic co-digestion of submerged macrophytes with food waste would definitely increase biogas-derived electricity sale and heat energy.

119

Co-digestion has been gaining momentum in many aspects including optimization of carbon to nitrogen (C/N) ratio (Yen and Brune, 2007), increase of buffering capacity (Mshandete et al., 2004) and provision of micro/macronutrient (Alatriste-Mondragón et al., 2006). Previous studies have reported that submerged macrophytes contain relatively abundant nitrogen, of which C/N ratio of 8.5-

12.9 (Kobayashi et al., 2015). It is known that C/N ratio of 20-30 is preferable for the growth of anaerobic microorganisms (Gu et al., 2014). From these perspectives, co-digestion with carbon-rich food waste would be suitable for year-round stable anaerobic digestion of macrophytes in optimum

C/N ratio condition.

The application of food waste on thermochemically pre-treated submerged macrophytes may have another advantage: the reduction of neutralizing agent. After the alkaline pre-treatment, the pH reaches 12 or above, which is obviously too high for microorganisms. Usually, acids such as HCl is used for neutralizing the thermochemically pre-treated substrate, but the addition of the neutralizing agent rises the operation cost and high salinity in digestate, which can makes this pre-treatment less competitive. Food waste is usually easily degradable, so that it is quickly broken down into organic acids. Thus the idea of mixing with acidified food waste may have a potential to reduce the cost for neutralizing agent. Nakahashi (2013) conducted the co-digestion of P. maackianus after alkaline pre- treatment with food waste and reported that the addition of food waste achieved the decline of pH of the pre-treated macrophytes, suggesting 51% of the neutralizing agent can be reduced when treating

-1 P. maackianus at NaOH 0.1 g g-TSsubstrate , 80˚C for 3 h. Accordingly, the addition of food waste can be attractive not only due to the increase of the organic loading rate during wintertime, but also due to the reduction of the cost for alkaline pre-treatment.

120

Table 5-1. Daily throughput in simulated full-scale anaerobic digestion of submerged macrophyte.

Daily throughput (tons day-1) Un-treated Alkali pre-treated P. maackianus 15.0 15.0 NaOH aq. * - 14.9 HCl aq. ** - 0.1 Other macrophytes 15.0 15.0 Water 15.0 0.0 Total 45.0 45.0

* NaOH load is 0.100 ton-TS against 1 ton-TS of P. maackianus ** HCl load is 0.112 ton-wwt (12N) against 1 ton-TS of P. maackianus

121 Table 5-2. Thermal and electricity energy produced in anaerobic digestion of submerged macrophytes in presumed full-scale treatment in semi-continuous mode (30 tons-wwt day-1).

Parameters Unit Untreated Pre-treated

3 -1 Monthly CH4 yield m month 7,195 9,523 -1 Total energy potential of CH4 kWh month 79,478 105,196 Electricity energy produced kWh month-1 13,909 18,409 Thermal energy produced kWh month-1 39,739 52,598

122 ) 10,000 1 -

month 8,000 3

6,000 Potamogeton malaianus yield (m yield

4 Elodea nuttallii Ceratophyllum demersum 4,000 Egeria densa Potamogeton maackianus 2,000 Cumulative CH Cumulative

0 Un-treated Pre-treated

Fig. 5-1. Monthly cumulative CH4 yield of submerged macrophytes of simulated full-scale anaerobic digestion treatment in semi-continuous mode.

123 Untreated Pre-treated 1,400 (A)

) 1,200 1 - 1,000

800 Electricity sale

JPY month JPY NaOH cost 3 600 HCl cost Net benefit 400

200

Cost balance (10 0

-200

-400

Untreated Pre-treated 60,000 (B) 50,000 ) 1 - 40,000

30,000 Thermal energy produced HER for pre-treatment 20,000 Net heat energy gain 10,000

0 Heat balance (kWh month

-10,000

-20,000

Fig. 5-2. Economic balance (A) and thermal balance (B) of simulated full-scale anaerobic digestion of submerged macrophytes in semi-continuous mode.

124 1,100 ) 1 - 1,000 Un-treated Pre-treated

JPY month JPY 900 3

800 Net benefit (10 700

6000 30 31 32 33 34 35 36 37 38 39 Electricity tariff of biogas (JPY kWh-1)

Fig. 5-3. Effect of electricity tariff on net benefit from un-treated and pre-treated submerged macrophytes. The net benefit of pre-treated macrophytes was determined by subtraction of chemical cost from the benefit from electricity sale.

125 30

) (A) 1 - 25

20 Potamogeton malaianus wwt day Ceratophyllum demersum 15 Egeria densa 10 Potamogeton maackianus 5 Elodea nuttallii

Daily - (ton throughput Daily 10 20 30 40 50 60 70 Proportion of El. nuttallii (%-wwt)

) 10

1 (B) - 9 8 7

JPY month JPY 6 3 5 4 3 2 1 Net benefit (10 0 10 20 30 40 50 60 70 proportion of El. nuttallii (%-wwt) )

1 5000 - (C)

4000

3000

2000

1000

Net heat balance (kWhmonth 0 10 20 30 40 50 60 70 proportion of El. nuttallii (%-wwt)

Fig. 5-4. Effect of species composition on the difference of net benefit and net heat balance between un-treated and pre-treated submerged macrophytes under various composition of El. nuttallii. Species composition of daily throughput (A); net benefit (B); net heat balance (C). (B) and (C) were obtained by the subtraction of un-treated from pre-treated macrophyte.

126

References

Abbasi SA, Nipaney PC and Schaumberg GD (1990) Bioenergy potential of eight common aquatic weeds. Biol. Wastes 34(4): 359–366. Ahring BK, Sandberg M and Angelidaki I (1995) Volatile fatty acids as indicators of process imbalance in anaerobic digestors. Appl. Microbiol. Biotechnol. 43(3): 559–565. Alatriste-Mondragón F, Samar P, Cox HHJ, Ahring KB and Iranpour R (2006) Anaerobic co- digestion of municipal, farm, and industrial organic wastes: a survey of recent literature. Water Environ. Res. 78(6): 607–636. American Public Health Asspciation (1998) Standard Methods for the Examination of Water and Wastewater. 20th editi Washington DC, USA. Asaeda T, Fujino T and Manatunge J (2005) Morphological adaptations of emergent plants to water flow: a case study with Typha angustifolia , Zizania latifolia and Phragmites australis. Freshw. Biol. 50(12): 1991–2001. Badshah M, Lam DM, Liu J and Mattiasson B (2012) Use of an Automatic Methane Potential Test System for evaluating the biomethane potential of sugarcane bagasse after different treatments. Bioresour. Technol. 114: 262–269. Bahr M, Díaz I, Dominguez A, Sánchez AG and Muñoz R (2014) Microalgal-biotechnology as a platform for an integral biogas upgrading and nutrient removal from anaerobic effluents. Environ. Sci. Technol. 48(1): 573–581. Barakat A, Monlau F, Steyer J-P and Carrere H (2012) Effect of lignin-derived and furan compounds found in lignocellulosic hydrolysates on biomethane production. Bioresour. Technol. 104: 90–99. Benner R, Maccubbin AE and Hodson RE (1984) Anaerobic biodegradation of the lignin and polysaccharide components of lignocellulose and synthetic lignin by sediment microflora. Appl. Environ. Microbiol. 47(5): 998–1004. Berlin A, Balakshin M, Gilkes N, Kadla J, Maximenko V, Kubo S and Saddler J (2006) Inhibition of cellulase, xylanase and beta-glucosidase activities by softwood lignin preparations. J. Biotechnol. 125(2): 198–209. Björkman A (1957) Lignin and lignin-carbohydrate complexes. Ind. Eng. Chem. Res. 49(9): 1395– 1398. Buranov AU and Mazza G (2008) Lignin in straw of herbaceous crops. Ind. Crops Prod. 28(3):

127

237–259. Cao GL, Ren NQ, Wang AJ, Guo WQ, Xu JF and Liu B-F (2010) Effect of lignocellulose-derived inhibitors on growth and hydrogen production by Thermoanaerobacterium thermosaccharolyticum W16. Int. J. Hydrogen Energy 35(24): 13475–13480. Carbajo JB, Boltes K and Leton P (2010) Treatment of phenol in an anaerobic fluidized bed reactor (AFBR): continuous and batch regime. Biodegradation 21(4): 603–613. Carrere H, Antonopoulou G, Affes R, Passos F, Battimelli A, Lyberatos G and Ferrer I (2016) Review of feedstock pretreatment strategies for improved anaerobic digestion: From lab-scale research to full-scale application. Bioresour. Technol. 199: 386–397. Chen X, Gu Y, Zhou X and Zhang Y (2014) (a) Asparagus stem as a new lignocellulosic biomass feedstock for anaerobic digestion: Increasing hydrolysis rate, methane production and biodegradability by alkaline pretreatment. Bioresour. Technol. 164: 78–85. Chen X, Wang Z, Fu Y, Li Z and Qin M (2014) (b) Specific lignin precipitation for oligosaccharides recovery from hot water wood extract. Bioresour. Technol. 152: 31–37. Chen Y, Cheng JJ and Creamer KS (2008) Inhibition of anaerobic digestion process: a review. Bioresour. Technol. 99(10): 4044–4064. Chen Y, Fu B, Wang Y, Jiang Q and Liu H (2012) Reactor performance and bacterial pathogen removal in response to sludge retention time in a mesophilic anaerobic digester treating sewage sludge. Bioresour. Technol. 106: 20–26.

Cheng J, Xie B, Zhou J, Song W and Cen K (2010) Cogeneration of H2 and CH4 from water hyacinth by two-step anaerobic fermentation. Int. J. Hydrogen Energy 35(7): 3029–3035. Chufo A, Yuan H, Zou D, Pang Y and Li X (2015) Biomethane production and physicochemical characterization of anaerobically digested teff (Eragrostis tef) straw pretreated by sodium hydroxide. Bioresour. Technol. 181: 214–219. Chum HL (1985) Evaluation of pretreatments of biomass for enzymatic hydrolysis of cellulose. Solar Energy Research Institute (ed.) Golden, CO (USA): Colorado State University. Clifford DJ, Carson DM, Mckinney DE, Bortiatynski JM and Hatcher PG (1995) A new rapid technique for the characterization of lignin in vascular plants : thermochemolysis with tetramethylammonium hydroxide. Org. Geochem. 23(2): 169–175. CMC (2015) Naigai Kagaku Yakuhin Shiryo (Domestic and international information for chemical materials) F file: inorganic chemical materials. CMC Inc. Costa a. G, Pinheiro GC, Pinheiro FGC, Dos Santos a. B, Santaella ST and Leitão RC (2014) 128

The use of thermochemical pretreatments to improve the anaerobic biodegradability and biochemical methane potential of the sugarcane bagasse. Chem. Eng. J. 248: 363–372. Das CP and Patnaik LN (2000) Removal of lignin by industrial solid wastes. Pract. Period. Hazardous, Toxic, Radioact. Waste Manag. 4(4): 156–161. de Pinho MN, Minhalma M, Rosa MJ and Taborda F (2000) Integration of flotation/ultrafiltration for treatment of bleached pulp effluent. Pulp Pap. Canada 101(4): 50–54. del Río JC, Gutiérrez A, Rodríguez IM, Ibarra D and Martínez ÁT (2007) Composition of non- woody plant lignins and cinnamic acids by Py-GC/MS, Py/TMAH and FT-IR. J. Anal. Appl. Pyrolysis 79(1-2): 39–46. Dhar BR, Nakhla G and Ray MB (2012) Techno-economic evaluation of ultrasound and thermal pretreatments for enhanced anaerobic digestion of municipal waste activated sludge. Waste Manag. 32(3): 542–549. Dhouib A, Ellouz M, Aloui F and Sayadi S (2006) Effect of bioaugmentation of activated sludge with white-rot fungi on olive mill wastewater detoxification. Lett. Appl. Microbiol. 42(4): 405– 411. Di Girolamo G, Bertin L, Capecchi L, Ciavatta C and Barbanti L (2014) Mild alkaline pretreatments loosen fibre structure enhancing methane production from biomass crops and residues. Biomass and Bioenergy 71: 318–329. Endo G, Noike T and Matsumoto J (1982) Characteristics of cellulose and glucose decomposition in acidogenic phase of anaerobic digestion. Proc. Japan Soc. Civ. Eng. 1982(325): 61–68. Feijoo G, Soto M, Méndez R and Lema JM (1995) Sodium inhibition in the anaerobic digestion process: Antagonism and adaptation phenomena. Enzyme Microb. Technol. 17(2): 180–188. Fengel D and Wegener G (1984) Wood: chemistry, ultrastructure, reactions. Walter de Gruyter 613: 1960–1982. Fernandes T V., Bos GJ, Zeeman G, Sanders JPM and van Lier JB (2009) Effects of thermochemical pre-treatment on anaerobic biodegradability and hydrolysis of lignocellulosic biomass. Bioresour. Technol. 100(9): 2575–2579. Ferrer I, Palatsi J, Campos E and Flotats X (2010) Mesophilic and thermophilic anaerobic biodegradability of water hyacinth pre-treated at 80 degrees C. Waste Manag. 30(10): 1763– 1767. Field JA and Lettinga G (1987) The methanogenic toxicity and anaerobic degradability of a hydrohyzable tannin. Water Res. 21(3): 367–374. 129

Food and agriculture organization of the United Nations (2002) Management of problematic aquatic Wweeds in Africa-FAO Efforts and Achievements During the Period 1991-2001. (consulted http://www.fao.org/docrep/006/y4270e/y4270e00.HTM) Frigon JC, Mehta P and Guiot SR (2012) Impact of mechanical, chemical and enzymatic pretreatments on the methane yield from the anaerobic digestion of switchgrass. Biomass and Bioenergy 36: 1–11. Funaoka M (1998) A new type of phenolic lignin-based network polymer with the structure- variable function composed of 1,1-diarylpropane units. Polym. Int. 47(3): 277–290. Gierer J (1985) Chemistry of delignification. Wood Sci. Technol. 19(4): 289–312. Gonzalez-Estrella J, Sierra-Alvarez R and Field JA (2013) Toxicity assessment of inorganic nanoparticles to acetoclastic and hydrogenotrophic methanogenic activity in anaerobic granular sludge. J. Hazard. Mater. 260: 278–285. Gu Y, Chen X, Liu Z, Zhou X and Zhang Y (2014) Effect of inoculum sources on the anaerobic digestion of rice straw. Bioresour. Technol. 158: 149–155. Gunaseelan VN (2007) Regression models of ultimate methane yields of fruits and vegetable solid wastes, sorghum and napiergrass on chemical composition. Bioresour. Technol. 98(6): 1270– 1277. Haga H, Ashiya M, Ohtsuka T, Matsuda M, Tuji A, Baba K, Numahata S and Yamane T (2006) (a) Relationship between dissolved oxygen concentration of bottom water and macrophyte biomass in the southern basin of Lake Biwa, Japan. Japanese J. Limnol. 67: 23–27. Haga H, Ohtsuka T, Matsuda M and Ashiya M (2006) (b) Spatial distribution of biomass and species composition in submerged macrophytes in the southern basin of Lake Biwa in summer of 2002. Japanese J. Limnol. 67: 69–79. Haga Y and Ishikawa K (2011) Spatial distribution of submerged macrophytes in the southern basin of Lake Biwa in the summer of 2007, in comparison with that in 2002. Japanese J. Limnol. 72: 81–88. Hamabata E (2005) Submerged macrophyte community in Lake Biwa. Lake Biwa Res. Inst. Bull. 22: 105–119. Hamabata E, Sugimura S and Ishikawa K (2012) 7.2.3 The explosive development and control of aquatic weeds. In Kawanabe H Nishino M and Maehata M (eds) Lake Biwa: Interactions between Nature and People. Springer Science and Business Media, 469–472. Hartley RD and Morrison WH (1991) Monomeric and dimeric phenolic acids released from cell 130

walls of grasses by sequential treatment with sodium hydroxide. J. Sci. Food Agric. 55(3): 365–375. He Y, Pang Y, Li X, Liu Y, Li R and Zheng M (2009) Investigation on the changes of main compositions and extractives of rice straw pretreated with sodium hydroxide for biogas production. Energy and Fuels 23(4): 2220–2224. Healy JB and Young LY (1979) Anaerobic biodegradation of eleven aromatic compounds to methane. Appl. Environ. Microbiol. 38(1): 84–89. Hedges JI and Mann DC (1979) The characterization of plant tissues by their lignin oxidation products. Geochim. Cosmochim. Acta 43(11): 1803–1807. Hendriks ATWM and Zeeman G (2009) Pretreatments to enhance the digestibility of lignocellulosic biomass. Bioresour. Technol. 100(1): 10–18. Heo NH, Park SC and Kang PH (2004) Effects of mixture ratio and hydraulic retention time on single-stage anaerobic co-digestion of food waste and waste activated sludge. J. Environ. Sci. Heal. Part A 39(7): 1739–1756. Herdendorf CE, David M. Klarer and Ricki C. Herdendorf (2006) Ecology of Old Woman Creek, Ohio: an estuarine and water shed profile. , 2nd ed Columbus, Ohio: Ohio Department of Natural Resources, Division of Wildlife. Hernandez JE and Edyvean RGJ (2008) Inhibition of biogas production and biodegradability by substituted phenolic compounds in anaerobic sludge. J. Hazard. Mater. 160(1): 20–28. Hussner A (2012) Alien aquatic plant species in European countries. Weed Res. 52(4): 297–306. Hwang K, Song M, Kim W, Kim N and Hwang S (2010) Effects of prolonged starvation on methanogenic population dynamics in anaerobic digestion of swine wastewater. Bioresour. Technol. 101 Suppl (1): S2–S6. Ibrahim MNM, Chuah SB and Rosli WDW (2004) Characterization of lignin precipitated from the soda black liquor of oil palm empty fruit bunch fibers by various mineral acids. ASEAN J. Sci. Technol. Dev. 21: 57–68. Jain S, Durand H and Tiraby G (1992) Development of a transformation system for the thermophilic fungus Talaromyces sp. CL240 based on the use of phleomycin resistance as a dominant selectable marker. Mol. Gen. Genet. MGG 234(3): 489–493. Jeihanipour A, Niklasson C and Taherzadeh MJ (2011) Enhancement of solubilization rate of cellulose in anaerobic digestion and its drawbacks. Process Biochem. 46(7): 1509–1514. Jin G, Bierma T and Walker PM (2014) Low-heat, mild alkaline pretreatment of switchgrass for 131

anaerobic digestion. J. Environ. Sci. Health. A. Tox. Hazard. Subst. Environ. Eng. 49(5): 565– 574. Kayembe K, Basosila L, Mpiana PT, Sikulisimwa PC, Mbuyu K and others (2013) Inhibitory effects of phenolic monomers on methanogenesis in anaerobic digestion. Br. Microbiol. Res. J. 3(1): 32–41. Kimata-Kino N, Ikeda S, Kurosawa N and Toda T (2011) Saline adaptation of granules in mesophilic UASB reactors. Int. Biodeterior. Biodegradation 65(1): 65–72. Kobayashi T, Wu Y, Lu Z and Xu K (2015) Characterization of anaerobic digestibility and kinetics of harvested submerged aquatic weeds used for nutrient phytoremediation. Energies 8: 304– 318. Koyama M, Yamamoto S, Ishikawa K, Ban S and Toda T (2014) Anaerobic digestion of submerged macrophytes: Chemical composition and anaerobic digestibility. Ecol. Eng. 69: 304–309. Lavarack BP, Griffin GJ and Rodman D (2002) The acid hydrolysis of sugarcane bagasse hemicellulose to produce xylose, arabinose, glucose and other products. Biomass and Bioenergy 23(5): 367–380. Lay JJ, Li YY and Noike T (1997) Influences of pH and moisture content on the methane production in high-solids sludge digestion. Water Res. 31(6): 1518–1524. Lei M, Zhang H, Zheng H, Li Y, Huang H and Xu R (2013) Characterization of lignins isolated from alkali treated prehydrolysate of corn stover. Chinese J. Chem. Eng. 21(4): 427–433. Levine RB, Costanza-Robinson MS and Spatafora GA (2011) Neochloris oleoabundans grown on anaerobically digested dairy manure for concomitant nutrient removal and biodiesel feedstock production. Biomass and Bioenergy 35(1): 40–49. Li Q, Qiao W, Wang X, Takayanagi K, Shofie M and Li Y-Y (2015) Kinetic characterization of thermophilic and mesophilic anaerobic digestion for coffee grounds and waste activated sludge. Waste Manag. 36: 77–85. Liedl B, Bombardiere J and Chatfield J (2006) Fertilizer potential of liquid and solid effluent from thermophilic anaerobic digestion of poultry waste. Water Sci. Technol. 53(8): 69–79. Lin Y, Wang D, Wu S and Wang C (2009) Alkali pretreatment enhances biogas production in the anaerobic digestion of pulp and paper sludge. J. Hazard. Mater. 170(1): 366–373. Lora J and Glasser W (2002) Recent industrial applications of lignin: a sustainable alternative to nonrenewable materials. J. Polym. Environ. 10(1-2): 39–48. Matsumoto T (2010) Bioethanol production using aquatic weeds taken from Lake Biwa, Japan: the 132

development of pre-treatment method for saccharification process (2nd report). Northeast. Ind. Res. Cent. Shiga Prefect. Res. Rep.: 9–20. McCarty PL and Smith DP (1986) Anaerobic wastewater treatment. Environ. Sci. Technol. 20(12): 1200–1206. McKendry P (2002) Energy production from biomass (part 1): overview of biomass. Bioresour. Technol. 83(1): 37–46. Ministry of Economy Trade and Industry (METI) (2011) Act on special measures concerning procurement of renewable electric energy by operators of electric etilities. (consulted http://www.kyuden.co.jp/library/pdf/company/liberal/elec/) Ministry of Economy Trade and Industry (METI) (2015) Feed-in tariff for electricity from renewable energy sources. (consulted http://www.enecho.meti.go.jp/category/saving_and_new/saiene/kaitori/kakaku.html) Ministry of the Environment J (2010) Plan for Conservation of Lake Water Quality. (consulted http://www.env.go.jp/water/kosyou/gaiyo_proj.pdf) Monlau F, Barakat A, Steyer JP and Carrere H (2012) Comparison of seven types of thermochemical pretreatments on the structural features and anaerobic digestion of sunflower stalks. Bioresour. Technol. 120: 241–247. Monlau F, Kaparaju P, Trably E, Steyer JP and Carrere H (2015) Alkaline pretreatment to enhance

one-stage CH4 and two-stage H2/CH4 production from sunflower stalks: Mass, energy and economical balances. Chem. Eng. J. 260: 377–385. Mshandete A, Kivaisi A, Rubindamayugi M and Mattiasson B (2004) Anaerobic batch co-digestion of sisal pulp and fish wastes. Bioresour. Technol. 95: 19–24. Muñoz Escobar M, Voyevoda M, Fühner C and Zehnsdorf A (2011) Potential uses of Elodea nuttallii-harvested biomass. Energy. Sustain. Soc. 1(1): 1–8. Nagao N, Tajima N, Kawai M, Niwa C, Kurosawa N, Matsuyama T, Yusoff FM and Toda T (2012) Maximum organic loading rate for the single-stage wet anaerobic digestion of food waste. Bioresour. Technol. 118: 210–218.

Nagy M, Kosa M, Theliander H and Ragauskas AJ (2010) Characterization of CO2 precipitated Kraft lignin to promote its utilization. Green Chem. 12(1): 31–34. Nakagame S, Chandra RP, Kadla JF and Saddler JN (2011) The isolation, characterization and effect of lignin isolated from steam pretreated Douglas-fir on the enzymatic hydrolysis of cellulose. Bioresour. Technol. 102(6): 4507–4517. 133

Nakagame S, Chandra RP and Saddler JN (2010) The effect of isolated lignins, obtained from a range of pretreated lignocellulosic substrates, on enzymatic hydrolysis. Biotechnol. Bioeng. 105(5): 871–879. Nakahashi N (2013) Effect of mix ratio on anaerobic co-digestion of alkali pre-treated submerged macrophyte and food waste. Soka University. National Institute for Land and Infrastructure Management (2015) B-DASH Project No. 10 Guideline for introducing an electricity generation system from sewage biomass source. Tech. note Natl. Inst. L. Infrastruct. Manag. Japan 860: 223. Nieves DC, Karimi K and Horvath IS (2011) Improvement of biogas production from oil palm empty fruit bunches (OPEFB). Ind. Crops Prod. 34(1): 1097–1101. Nonaka H, Tanaka H and Funaoka M (2011) Adsorption of Trichoderma reesei cellulase on softwood lignin-based lignophenol. J. Japan Inst. Energy 90(10): 977–981. O’Sullivan C, Rounsefell B, Grinham A, Clarke W and Udy J (2010) Anaerobic digestion of harvested aquatic weeds: water hyacinth (Eichhornia crassipes), cabomba (Cabomba Caroliniana) and salvinia (Salvinia molesta). Ecol. Eng. 36(10): 1459–1468. Palonen H, Tjerneld F, Zacchi G and Tenkanen M (2004) Adsorption of Trichoderma reesei CBH I and EG II and their catalytic domains on steam pretreated softwood and isolated lignin. J. Biotechnol. 107(1): 65–72. Park J, Jin H-F, Lim B-R, Park K-Y and Lee K (2010) Ammonia removal from anaerobic digestion effluent of livestock waste using green alga Scenedesmus sp. Bioresour. Technol. 101(22): 8649–8657. Patel V, Desai M and Madamwar D (1993) Thermochemical pretreatment of water hyacinth for improved biomethanation. Appl. Biochem. Biotechnol. 42(1): 67–74. Poorter H, Bergkotte M and others (1992) Chemical composition of 24 wild species differing in relative growth rate. Plant Cell Env. 15(2): 221–229. Quéméneur M, Hamelin J, Barakat A, Steyer J-P, Carrère H and Trably E (2012) Inhibition of fermentative hydrogen production by lignocellulose-derived compounds in mixed cultures. Int. J. Hydrogen Energy 37(4): 3150–3159. Rabemanolontsoa H and Saka S (2012) Characterization of Lake Biwa Macrophytes in their Chemical Composition. J. Japan Inst. Energy 91: 621–628. Rahikainen JL, Martin-Sampedro R, Heikkinen H, Rovio S, Marjamaa K, Tamminen T, Rojas OJ and Kruus K (2013) Inhibitory effect of lignin during cellulose bioconversion: the effect of 134

lignin chemistry on non-productive enzyme adsorption. Bioresour. Technol. 133: 270–278. Ralph J, Grabber JH and Hatfield RD (1995) Lignin-ferulate cross-links in grasses: active incorporation of ferulate polysaccharide esters into ryegrass lignins. Carbohydr. Res. 275(1): 167–178. Rosenkranz F, Cabrol L, Carballa M, Donoso-Bravo A, Cruz L, Ruiz-Filippi G, Chamy R and Lema JM (2013) Relationship between phenol degradation efficiency and microbial community structure in an anaerobic SBR. Water Res. 47(17): 6739–6749. Sambusiti C, Ficara E, Malpei F, Steyer JP and Carrère H (2012) Influence of alkaline pre-treatment conditions on structural features and methane production from ensiled sorghum forage. Chem. Eng. J. 211-212: 488–492. Sambusiti C, Ficara E, Malpei F, Steyer JP and Carrère H (2013) (a) Benefit of sodium hydroxide pretreatment of ensiled sorghum forage on the anaerobic reactor stability and methane production. Bioresour. Technol. 144: 149–155. Sambusiti C, Monlau F, Ficara E, Carrère H and Malpei F (2013) (b) A comparison of different pretreatments to increase methane production from two agricultural substrates. Appl. Energy 104: 62–70. Santos M, Anderson L and Ustin S (2011) Effects of invasive species on plant communities: an example using submersed aquatic plants at the regional scale. Biol. Invasions 13(2): 443–457. Sawatdeenarunat C, Surendra KC, Takara D, Oechsner H and Kumar Khanal S (2014) Anaerobic digestion of lignocellulosic biomass: Challenges and opportunities. Bioresour. Technol. 178: 178–186. Schwarz W (2001) The cellulosome and cellulose degradation by anaerobic bacteria. Appl. Microbiol. Biotechnol. 56(5-6): 634–649. Scientific Committee for Comprehensive Conservation of Lake Biwa (2010) Mother Lake 21 Plan: The evaluation of first phase and the proposal to the second phase. (consulted http://www.pref.shiga.jp/biwako/koai/gakujutuiinkaiteigen.pdf) Sekiguchi Y, Kamagata Y, Syutsubo K, Ohashi A, Harada H and Nakamura K (1998) Phylogenetic diversity of mesophilic and thermophilic granular sludges determined by 16S rRNA gene analysis. Microbiology 144: 2655–2665. Shiga Prefecture (2012) Attempts for effective use of aquatic weeds in Lake Biwa. (consulted http://www.pref.shiga.lg.jp/d/saisei/files/mizukusa/files/yuukouriyou.pdf) Sierra-alvarez R and Lettinga G (1991) The effect of aromatic structure on the inhibition of 135

acetoclastic methanogenesis in granular sludge. Appl. Microbiol. Biotechnol. 3: 544–550. Sierra-Alvarez R and Lettinga G (1991) The methanogenic toxicity of wastewater lignins and lignin related compounds. J. Chem. Technol. Biotechnol. 50(4): 443–455. Smolders AJP, Vergeer LHT, Velde G Van Der and Roelofs JGM (2000) Phenolic contents of submerged , emergent and floating leaves of aquatic and semi-aquatic macrophyte species : why do they differ ? Oikos(91): 307–310. Sonoda K, Takahashi M, Chikaraishi Y and Yamamoto S (2010) Compound-specific stable carbon isotope analysis of lignin phenols by gas chromatography/combustion/ isotope ratio mass spectrometry (GC/C/IRMS). Res. Org. Geochemistry 26: 115–122. Speece R. (1996) Anaerobic Biotechnology for Industrial Wastewaters. Nashville, Tennessee: Archae Press. Srilatha HR, Nand K, Babu KS and Madhukara K (1995) Fungal pretreatment of orange processing waste by solid-state fermentation for improved production of methane. Process Biochem. 30(4): 327–331. Sugiura N, Utsumi M, Wei B, Iwami N, Okano K, Kawauchi Y and Maekawa T (2004) Assessment for the complicated occurrence of nuisance odours from phytoplankton and environmental factors in a eutrophic lake. Lakes Reserv. Res. Manag. 9(3-4): 195–201. Sun R, Lawther JM and Banks WB (1997) A tentative chemical structure of wheat straw lignin. Ind. Crops Prod. 6(1): 1–8. Sun Y and Cheng J (2002) Hydrolysis of lignocellulosic materials for ethanol production: a review. Bioresour. Technol. 83(1): 1–11. Taherdanak M and Zilouei H (2014) Improving biogas production from wheat plant using alkaline pretreatment. Fuel 115: 714–719. Taherzadeh MJ and Karimi K (2008) Pretreatment of lignocellulosic wastes to improve ethanol and biogas production: a review. Int. J. Mol. Sci. 9(9): 1621–1651. Taniguchi M, Suzuki H, Watanabe D, Sakai K, Hoshino K and Tanaka T (2005) Evaluation of pretreatment with Pleurotus ostreatus for enzymatic hydrolysis of rice straw. J. Biosci. Bioeng. 100(6): 637–643. Teghammar A, Yngvesson J, Lundin M, Taherzadeh MJ and Horvath IS (2010) Pretreatment of paper tube residuals for improved biogas production. Bioresour. Technol. 101(4): 1206–1212. Triolo JM, Sommer SG, Møller HB, Weisbjerg MR and Jiang XY (2011) A new algorithm to characterize biodegradability of biomass during anaerobic digestion: influence of lignin 136

concentration on methane production potential. Bioresour. Technol. 102(20): 9395–9402. Tuomela M, Vikman M, Hatakka A and It M (2000) Biodegradation of lignin in a compost environment : a review. Bioresour. Technol. 72: 169–183. Uğurlu M, Gürses a, Doğar C and Yalçin M (2008) The removal of lignin and phenol from paper mill effluents by electrocoagulation. J. Environ. Manage. 87(3): 420–428. Usack JG, Spirito CM and Angenent LT (2012) Continuously-stirred Anaerobic Digester to Convert Organic Wastes into Biogas: System Setup and Basic Operation. J. Vis. Exp.(65): 3978. Uysal A, Yilmazel YD and Demirer GN (2010) The determination of fertilizer quality of the formed struvite from effluent of a sewage sludge anaerobic digester. J. Hazard. Mater. 181(1-3): 248– 254. van Donk E, Gulati RD, Iedema A and Meulemans JT (1993) Macrophyte-related shifts in the nitrogen and phosphorus contents of the different trophic levels in a biomanipulated shallow lake. Hydrobiologia 251(1-3): 19–26. Van Soest PJ, Robertson JB and Lewis BA (1991) Methods for dietary fiber, neutral detergent fiber, and nonstarch polysaccharides in relation to animal nutrition. J. Dairy Sci. 74(10): 3583–3597. Vavilin VA, Fernandez B, Palatsi J and Flotats X (2008) Hydrolysis kinetics in anaerobic degradation of particulate organic material: an overview. Waste Manag. 28(6): 939–951. Vavilin VA, Rytov SV and Lokshina LY (1996) A description of hydrolysis kinetics in anaerobic degradation of particulate organic matter. Bioresour. Technol. 56(2-3): 229–237. Veeresh GS, Kumar P and Mehrotra I (2005) Treatment of phenol and cresols in upflow anaerobic sludge blanket (UASB) process: a review. Water Res. 39(1): 154–170. Verrier D, Roy F and Albagnac G (1987) Two-phase methanization of solid vegetable wastes. Biol. Wastes 22(3): 163–177. Vidal G and Diez MC (2005) Methanogenic toxicity and continuous anaerobic treatment of wood processing effluents. J. Environ. Manage. 74(4): 317–325. Xie S, Frost JP, Lawlor PG, Wu G and Zhan X (2011) Effects of thermo-chemical pre-treatment of grass silage on methane production by anaerobic digestion. Bioresour. Technol. 102(19): 8748–8755. Ximenes E, Kim Y, Mosier N, Dien B and Ladisch M (2010) Inhibition of cellulases by phenols. Enzyme Microb. Technol. 46(3-4): 170–176. Yao Y, He M, Ren Y, Ma L, Luo Y, Sheng H, Xiang Y, Zhang H, Li Q and An L (2013) Anaerobic 137

digestion of poplar processing residues for methane production after alkaline treatment. Bioresour. Technol. 134: 347–352. Yarrow M, Marin VH, Finlayson M, Tironi A, Delgado LE, Fischer F and others (2009) The ecology of Egeria densa Planchon (Liliopsida: Alismatales): A wetland ecosystem engineer. Rev. Chil. Hist. Nat. 82(2): 299–313. Yen H and Brune DE (2007) Anaerobic co-digestion of algal sludge and waste paper to produce methane. Bioresour. Technol. 98: 130–134. Yin C, Seo D, Kim M and Lee S (2000) (a) Inhibitory effect of hardwood lignin on acetate-utilizing methanogens in anaerobic digester sludge. Biotechnol. Lett.: 1531–1535. Yin C-R, Seo D-I, Kim M-K and Lee S-T (2000) (b) Inhibitory effect of hardwood lignin on acetate-utilizing methanogens in anaerobic digester sludge. Biotechnol. Lett. 22(19): 1531– 1535. Zhang A, Liu C, Sun R and Xie J (2013) (a) Extraction, purification, and characterization of lignin fractions from sugarcane bagasse. BioResources 8(2): 1604–1614. Zhang C, Li J, Liu C, Liu X, Wang J, Li S, Fan G and Zhang L (2013) (b) Alkaline pretreatment for enhancement of biogas production from banana stem and swine manure by anaerobic codigestion. Bioresour. Technol. 149: 353–358. Zhang R (1999) Biogasification of rice straw with an anaerobic-phased solids digester system. Bioresour. Technol. 68(3): 235–245. Zhu J, Wan C and Li Y (2010) Enhanced solid-state anaerobic digestion of corn stover by alkaline pretreatment. Bioresour. Technol. 101(19): 7523–7528.

138