Mitsuru Abe and Hiroyuki Ohno

Abstract Ionic liquids (ILs) are collecting keen interest as novel solvents for plant biomass, especially for cellulose. ILs have several unique properties and they dissolve cellulose under milder condition than existing procedures. Here, we give an outline of the development of biomass dissolving ILs together with their physico-chemical properties. Dissolution and/or extraction of not only cellulose but also lignin with ILs are overviewed. The extracted biomass is expected to be converted into other energies. For this purpose, energy-saving biomass treatment is inevitable, and ILs are one of the most potential media for this. This chapter will deliver further ideas on the design of ILs for cellulose dissolution or plant biomass treatment in the near future.

Keywords Ionic liquid • Polarity • Hydrogen bond • Cellulose • Lignin • Dissolution • Design of ions • Save energy

2.1 Introduction

Ionic liquids (ILs) are organic salts with melting point below 100 °C, and especially those with the melting point at and below room temperature are called “room temperature Ionic liquids” [1]. There are a few important properties required for solvents such as non-volatility, non-flammability, and thermal stability in a wide temperature range. Although there are many solvents that have some of these properties, there are few solvents that have all of the above-mentioned properties. ILs have unique properties different from molecular solvents. Many ILs have non-volatility, non-flammability, and stability in a wide temperature range. Furthermore, there is a potential chance to design new ILs through unlimited

M. Abe • H. Ohno (*)

Department of Biotechnology, Tokyo University of Agriculture and Technology,

1- 24-16 Naka-cho, Koganei, Tokyo 184-8588, Japan e-mail: ohnoh@cc. tuat. ac. jp

Z. Fang et al. (eds.), Production of Biofuels and Chemicals with Ionic Liquids, Biofuels and Biorefineries 1, DOI 10.1007/978-94-007-7711-8_2,

© Springer Science+Business Media Dordrecht 2014 possibility of combination of ion pairs. ILs are accordingly known as “designer solvents”. It is easy to change their physico-chemical properties by the selection of suitable ions.

One of successful examples on the design of ILs is cellulose dissolution. Plant biomass is one example of a renewable and abundant natural material. These materials can be considered to be the embodied energy of sunlight and so is one possible method to produce energy on earth. Considering the limit of fossil fuels, there are increasing trials to convert plant biomass into user-friendly energy. There are many industrial plants for bioethanol production from corn starch or sugar cane in US and other countries. There are established methods to convert starch into sugar in our human life. However, since these processes compete with food industry, there are ethical concerns about the use of edible plant biomass as raw materials for fuel production [2]. Cellulosic biomass therefore is attracting attention as energy sources because they are inedible materials for human beings.

Cellulosic biomass essentially consists of cellulose, hemicellulose, and lignin. To obtain energy from cellulosic biomass with minimum given energy, following three steps are required, namely (1) extraction of cellulose from biomass, (2) hydro­lysis of the cellulose into glucose or other oligosaccharides, and (3) oxidation or fermentation. However, cellulosic biomass is scarcely used for bioenergy produc­tion because of its very poor solubility in common molecular solvents. The chem­ical and physical stability of cellulose are known to be derived from many intra- and inter-molecular hydrogen bonds [3,4]. Since ordinary molecular solvents have not enough power to dissolve cellulose, it is required to heat the mixture or stir it for a long time which is inefficient for energy conversion. The energy cost for dissolution and extraction processes for cellulose should be very low.

Many scientists recognize that ILs have great potential as solvents for cellulose and are paying particular attention to ILs as novel solvents for cellulose under mild conditions. Design of ILs to dissolve cellulose with low energy cost is therefore indispensable for energy conversion. Without this step, it is difficult to use cellu — losic biomass as valuable materials as well as fossil fuel substitutes. The discussion in this chapter concentrates on the dissolution of biomass in ILs.

It should be pointed out that it is unlikely to find one particular IL that is the most suited cellulose solvent for all kinds of chemical derivatization reactions. Instead, the reaction media should be adapted to the specific requirements of a particular reaction. The synthesis of novel ILs for the use in cellulose chemistry involves the design of new cations and anions as well as the combination of both to an IL. Thereby, two major limitations need to be considered. It is self-evident that the new ILs (or low melting salts) must be able to dissolve reasonable amounts of cellulose (at least 5-10 %) without serve degradation of the polysaccharide. More­over, dissolution should be possible at moderate temperatures (<130 °C), which implies that the polysaccharide solvents needs to be liquid in that area. Strictly speaking, the term IL describes compounds with a melting point below 100 °C. However, it has been demonstrated that also several organic salts with slightly higher melting points in the range of 100-140 °C, henceforward designated as low-melting salts, can either directly or in combination with certain co-solvents be useful solvents for derivatization of cellulose.

The dissolution mechanism of cellulose in ILs is still not understood yet and is matter of controversial discussions; as is the question why cellulose is actually insoluble in most molecular solvents (see [8, 110] and references therein). How­ever, it has been demonstrated that in particular the anion has a very important role. It was found by means of solvatochromic experiments that cellulose is only soluble in ILs with a strong hydrogen acceptor capacity (в > 0.8) [111, 112]. This param­eter is mainly determined by the nature of the anion. Thus variation and structure design of the IL’s anion is restricted to a certain extent if the ability to dissolve cellulose needs to be retained. Most cellulose dissolving ILs applied so far for chemical derivatization of the polysaccharide contained either chloride or acetate as relatively strong hydrogen bond acceptor ions. Although chloride base ILs proofed to be efficient solvents for cellulose, their use is limited by the high melting points and viscosities. ILs bearing acetate anions are usually less viscous and liquid at room temperature but show some specific side reactions (see Sect. 5.3.2). Following the initial reports on this class of cellulose solvents, several task-specific ILs with alternative anions and beneficial properties have been reported that could be used for dissolution of cellulose. However, up to now only the minority has been studied as homogeneous reaction media.

In addition to acetate, other carboxylates have been proposed as anionic species in cellulose dissolving ILs (Fig. 5.8). In particular imidazolium formates have been reported to possess lower viscosities compared to the corresponding chloride and acetate analogues [113]. With respect to the specific side reactions of the acetate anion, these low viscous ILs might be useful reaction media, e. g., for derivatization reactions that do not tolerate high temperatures. Several low melting ammonium formates have been prepared that could dissolve some polysaccharides including cellulose [71]. These low melting salts could successfully be applied also for the preparation of carboxymethyl cellulose with a rather high DS of about 1.6 and a non-statistic functionalization pattern.

Imidazolium- and ammonium salts with alkylphosphonate or dialkylphopshate as anions possess high hydrogen bond acceptor capacities and could be utilized as cellulose solvents [114, 115]. These ILs are liquid at room temperatures and their viscosities are in the range of 100-500 mPa s, which is comparable to the corresponding acetates. It has been proposed that substitution of oxygen by sulfur will result in a decreased viscosity due to the reduced symmetry of the anion [116]. First results on the use of phosphate ILs as homogeneous reaction media indicate that choice of the cation is crucial for the derivatization reaction. Whereas tetraalkylammonium dialkylphosphates could be used as solvents for the homoge­neous preparation of cellulose esters and mixed ester, the corresponding imidazolium salts appear to be less suitable for the derivatization of cellulose because gelation of the reaction mixture occurs shortly after the addition of derivatization reagents [35, 54].

Tailoring the nature of the IL cation, e. g., by choosing among different general types of cations or by modifying length, degree of branching, and flexibility of side chains, offers access to a very broad structural diversity (Fig. 5.8). Nevertheless, up to now cellulose research has been focused mainly on 1-alkyl-3-methylimi — dazolium based ILs, bearing mostly ethyl, butyl, or allyl in the side chain. The cation has a significant effect on the ILs physical properties, e. g., melting point and viscosity, as well as its chemical reactivity. If and to what extent it is also directly involved in the cellulose dissolution mechanism is matter of current scientific discussion [8]. However, it appears that the choice of potential cations is less restricted compared to the limitation in terms of the possible anions.

Regarding commonly applied 1,3-dialkylimidazolium IL, task-specific solvents with interesting properties can be generated be substituting also the other positions of the aromatic ring [117]. The possibility to suppress deprotonation and carbene formation by methylation at position 2 has already been described (see Sect. 5.3.2). Another possibility is to increase flexibility of the side chains by introducing oxy-alkyl groups that also facilitate additional hydrogen bond interactions [56]. Replacing the butyl group in BMIMCl by a 2-methoxyethyl group seems to improve the solubility of cellulose and also resulted in an increased reactivity for acetylation reactions performed in this IL. Further studies are required to fully understand the reasons for these improvements, in particular because the opposite effect was observed when changing from a heptyl — to a 2-(2-ethoxyethoxy)ethyl substituted IL.

Taking into account the specific side reactions of dialkylimidazolium ILs, quaternary ammonium salts can be expected to become more important as solvents for dissolution and chemical modification of cellulose. A certain degree of cation asymmetry is required to obtain ILs with low melting points [118]. Despite that limitation, quaternary ammonium salts exhibit a much broader structural diversity than imidazolium ILs because in principle four side chains can be tailored.

Moreover, they can easily be prepared by complete alkylation of primary, second­ary, and tertiary amines, precursors that are inexpensive and readily available.

Several pyrolidinium, piperidinium, and morpholinium ILs with different anions have been reported for the use as cellulose solvents [115, 119]. However, compre­hensive comparison of their dissolution power as well as chemical and physical properties in comparison to ILs derived from aromatic or acyclic amines are scarcely found in literature. Especially morpholinium salts are of considerable scientific interest due to their relationship with NMMO. Low melting triethyl — and tributylmethylammonium formates could be obtained from the commercially available methyl carbonates by conversion with formic acid [71]. In the presence of a small excess of acid, these salts could be used as solvents for the homogeneous carboxymethylation of cellulose. Cellulose dissolving ammonium ILs with rather low viscosities (30-220 mPa s at 25 °C) have been obtained by introducing a cyclohexyl side chain to decrease symmetry of the cation [120]. The ILs carrying either acetate or alkoxyacetates could dissolve 1-9 % cellulose, which might be sufficient for specific applications. The solubility could be increased by adding DMSO.

Ether functionalized IL cations are expected to be better biodegradable, less toxic, and less viscous than their aliphatic analogues [121]. Several quaternary di — and trimethylammonium acetates bearing one or two alkoxy side chains have been prepared that showed melting points of 30-40 °C. Compared to imidazolium acetate, the viscosities of these ILs were slightly higher but they could dissolve similar amounts of cellulose and might consequently be interesting alternative reaction media for the derivatization of the polysaccharide [122]. Mono — and bicationic ammonium ILs with longer alkoxy groups and acetate as counter ion could be derived from poly(ethylene glycols) (PEG) and PEG monomethylethers [123]. According to the same approach, imidazolium and piperidinium acetates could be prepared as well. Depending on the number of C2H4O-units, some of these PEG functionalized ILs could dissolve approximately 8-12 % cellulose.

Xinhua Qi, Richard L. Smith Jr., and Zhen Fang

Abstract The furan derivative, 5-hydroxymethylfurfural (5-HMF), can replace many petroleum-derived monomers and intermediates presently used in the man­ufacture of plastics and fine chemicals. Ionic liquid solvents provide a sustainable path for 5-HMF production since they can dissolve crude biomass and allow conversion of polysaccharide fractions to 5-HMF with high selectivity. This chapter presents current progress in the synthesis of 5-HMF with ionic liquid solvents and considers conversion of saccharide substrates such as fructose, glucose, inulin and cellulose under catalytic reaction conditions. Challenges for 5-HMF production with ionic liquids are addressed and interesting aspects that still need to be explored for developing practical systems are highlighted.

Keywords Biomass • Ionic liquid • Cellulose • Hydroxymethylfurfural • Glucose • Fructose • Catalysis

9.1 Introduction

Biomass represents a possible sustainable resource for production of fuels and valuable chemicals [1]. However, the overabundance of oxygen within the molec­ular structures of carbohydrates limit the application of biomass as a feedstock for

X. Qi (*)

College of Environmental Science and Engineering, Nankai University, Tianjin, China e-mail: qixinhua@nankai. edu. cn

R. L. Smith Jr.

Graduate School of Environmental Studies, Research Center of Supercritical Fluid Technology, Tohoku University, Sendai, Japan e-mail: smith@scf. che. tohoku. ac. jp

Z. Fang

Xishuangbanna Tropical Botanical Garden, Chinese Academy of Sciences, Kunming, China e-mail: zhenfang@xtbg. ac. cn

Z. Fang et al. (eds.), Production of Biofuels and Chemicals with Ionic Liquids, Biofuels and Biorefineries 1, DOI 10.1007/978-94-007-7711-8_9,

© Springer Science+Business Media Dordrecht 2014

substitutes of traditional fossil fuels [2]. Removal of water from carbohydrates by dehydration is one of the main ways for reducing their oxygen content and obtaining commodity compounds [2]. Among the possible compounds that can be derived from dehydration of carbohydrates, 5-hydroxymethylfurfural (5-HMF) is a highly versatile intermediate, since it can be used for furan-based fine chemicals and polymers [3—5] (see Fig. 9.1). In the past few years, water [6—11], organic solvents [9, 12—15] and organic-water mixtures [1, 16—18] have been broadly investigated for the production of 5-HMF from a variety of biomass-derived carbohydrates such as fructose, glucose, and other di-poly-saccharides (Fig. 9.2).

Water is a good solvent for both monosaccharides and the products and it is favored as a reaction solvent from its environmental aspects. However, as a reaction solvent, water leads to the formation of undesired side-products, especially levulinic, formic acids, and humins result from either polymerization of 5-HMF or cross-polymerization of 5-HMF and carbohydrates that is attributed to water’s ionization [20—22]. When water is used as a solvent for conversion of saccharides to

5- HMF, by-product formation leads to 5-HMF yields of below 40 % and the conditions required are harsh [8 , 22—24]. Therefore, it is necessary to prevent rehydration of 5-HMF in the reaction chemistry so that humin formation can be suppressed. Suppression of humin formation can be accomplished by carrying out the reactions in nonaqueous systems. Dimethylsulfoxide (DMSO) was identified early as an efficient solvent for the preparation of 5-HMF from fructose [12,13, 25], since the furanoid form of fructose is favored in DMSO [4]. However, DMSO suffers from difficult product recovery and environmental problems in its use. Acetone is a good alternative solvent for DMSO due to some similarities in its

chemical properties, however, it has to be used with water or DMSO as cosolvent since monosaccharides have only limited solubility in pure acetone [4, 15].

Ionic liquids have favorable properties, such as nonvolatility, high thermal and chemical stability and adjustable solvent power for organic substances [26]. ILs have good attributes as solvent for producing 5-HMF from carbohydrates since they can allow reaction under relatively mild conditions and have the possibility of being used in one-pot reactions with biomass as the feedstock. This chapter provides an overview on research works that have studied the catalytic transformation of 5-HMF from a variety of carbohydrates in ionic liquids.

As smartly recently stated by M. Francisco et al. [14] finding an eco-efficient solvent for the dissolution of cellulose and more largely lignocellulosic biomass is becoming the Achille’s heel of renewable chemicals and biofuels processing.

In this context, few groups have attempted the design of bio-inspired ILs from choline chloride (ChCl) with the aim of dissolving (and thus decreasing the crystallinity index of cellulose) in a more sustainable way. ChCl is a very cheap (<2€/kg), biodegradable and non-toxic quaternary ammonium salt which can be either extracted from biomass or readily synthesized from fossil reserves (million metric tons per year) through a very high atom economy process. Its ionic structure makes of this organic salt a suitable candidate for the design of safer solvents that are particularly promising for biomass processing. In this context, two strategies are employed to produce media from ChCl (1) a chloride metathesis to produce the so-called bio-inspired ILs or (2) its combination with safe hydrogen bond donors such as urea, renewable carboxylic acids (e. g. oxalic, citric, succinic or aminoacids) or renewable polyols (e. g. glycerol, carbohydrates) to produce a deep eutectic solvent (DES). More information regarding DES is provided later in the chapter. To date the number of examples involving bio-renewable ChCl-derived solvents for the dissolution of cellulose and more largely lignocellulosic biomass is rather scarce mostly due to the novelty of these systems.

ChCl being a solid with a high melting point, one of the main strategies reported in the current literature consists in properly exchanging the chloride anion by acetate or an amino acid affording ILs that are liquid at a temperature below than 70 °C. All these so-called bio-inspired ILs were conveniently prepared by neutral­ization of the commercially available choline hydroxide derivative with the corresponding acid (yield >95 %). Main advantages of these ILs stem from their convenient synthesis, biodegradability, low toxicity and low price. To date, these bio-inspired ILs were essentially used in various catalytic reactions such as aldol [15] and Knoevenagel [16] reactions for instance but their use for biomass processing remained scarce.

In 2010, C. S. Pereira and his co-workers reported the efficient use of cholinium ethanoate and lactate for the dissolution of refined cork, an insoluble residue from the cork manufactures, composed of ~20 wt% of polysaccharides, ~30 wt% of poly (phenolics) (“lignin-like”) and ~50 wt% suberine [17]. Ability of choline-derived ILs to dissolve refined cork has been investigated at 100 °C. The residue (not soluble) was then analyzed by ATR-FTIR to identify which polymer of refined cork has been extracted. Choline ethanoate was able to dissolve a larger quantity of refined cork than the reference 1,3-dialkylimidazolium ionic liquid, especially of the aromatic suberin component (Table 3.1).

Other anions such as butanoate, hexanoate, methylpropanoate and lactate were also tested. Among them, lactate was found the less efficient (extraction efficiency of 20.7 and 39.7 % for lactate and ethanoate, respectively). An increase of the chain length of the anion led to an improvement of the extraction efficiency and best results were obtained with the hexanoate anion for which the extraction efficiency reached 64.9 %. The ability of choline-derived ILs to dissolve refined cork follow the pKa value of the conjugate acid of the anion i. e. an increase of the basicity led to an increase of the extraction efficiency. ATR-FTIR analyses revealed a drastic reduction in the aliphatic and aromatic bands of refined cork after extraction suggesting that tested choline-derived ILs mostly extract suberin.

In 2012, Zhang et al. investigated the dissolution of microcrystalline cellulose (PH AVICEL 106) in choline acetate [18]. Although choline acetate ([Ch]OAc) was not capable of dissolving microcrystalline cellulose in a large extent (solubility <0.2 and ~0.5 wt% after 5 min and 12 h, respectively), it is noteworthy that microcrystal­line cellulose started to swell after immersion in [Ch]OAc for 12 h at 110 °C. After filtration, cellulose looked like a flocky precipitate rather than a powder, suggesting that this media does affect the supramolecular organization of cellulose. To ensure a complete dissolution of cellulose, effect of additives was investigated. Among various tested additives, it was shown that addition of 5-15 wt% of tributylmethyl ammonium chloride ([TBMA]Cl) in [Ch]OAc dramatically enhanced the dissolution of microcrystalline cellulose. In particular, in the presence of 15 wt% of ([TBMA] Cl), 6 wt% of cellulose was dissolved within only 10 min at 110 °C. For comparison, when the dissolution of microcrystalline cellulose was performed in neat 1-butyl-3- methylimidazolium chloride (commonly used for the dissolution of cellulose), only 4 wt% of MCC were dissolved after 8 h of reaction, further demonstrating the effectiveness of the [Ch]OAc/[TBMA]Cl mixture.

As commonly performed in the case of imidazolium-based ILs, cellulose was then regenerated by addition of ethanol and recovered. XRD analyses performed on regenerated cellulose revealed that the cellulose structure was successfully changed from a high to a low crystalline structure indicating that the [Ch]OAc/[TBMA]Cl mixture was capable of disrupting the hydrogen bond network of cellulose. During the dissolution process, the glucose units were not damaged (checked by infra-red) while viscosimetry analyses revealed that the degree of polymerization of cellulose remained unchanged before and after the dissolution process. After washing of regenerated cellulose, no nitrogen was detected by elemental analysis suggesting that the [Ch]OAc/[TBMA]Cl mixture can be conveniently separated from cellulose.

After filtration of regenerated cellulose and removal of ethanol, the [Ch]OAc/ [TBMA]Cl mixture was recycled three times without appreciable decrease of its dissolution abilities. After the third run, the ability of the [Ch]OAc/[TBMA]Cl

Scheme 3.3 Dissolution/regeneration of microcrystalline cellulose in Choline acetate

mixture to dissolve cellulose however started to drop mainly due to the accumula­tion of impurities in the solution that hampered the long term recycling of this system. A similar trend is observed with imidazolium-based ILs. The whole dissolution/regeneration process is summarized in Scheme 3.3.

In the same year, Zong and his co-workers reported the design of ILs by combining cholinium as a cation and amino acids as anions affording the so-called [Ch][AA] with AA = glycine, alanine, serine, proline, among many others. Liu et al. [19] Ability of these bio-inspired ILs to dissolve biopolymers such as lignin, xylan (a model of hemicellulose) and cellulose was investigated. All prepared [Ch][AA] were fully characterized in term of viscosity, stability (TGA analysis), melting point (DSC analysis), alkalinity and optical rotation in order to rationalize the efficiency of [Ch][AA] in the dissolution of biopolymers. In a first approximation, authors observed that [Ch][AA] with a high alkalinity and low viscosity are more favorable for the dissolution of lignin. In particular, [Ch][gly — cine] was found to be the best ILs with a dissolution of up to 220 mg of lignin per gram of ILs at 90 °C. Although alkalinity and viscosity of [Ch][AA] also exerted an influence on the dissolution of xylan, this biopolymer was found much less soluble than lignin. Cellulose, a recalcitrant biopolymer, was however insoluble.

Despite the low solubility of xylan and cellulose in [Ch][glycine], it can be used for the pre-treatment of rice straw. In particular, a pre-treatment of rice straw in [Ch][glycine] at 90 °C for 24 h prior to enzymatic hydrolysis, led to an enhance­ment of the glucose production. For instance, after pre-treatment of rice straw in [Ch][glycine], the concentration of glucose obtained after enzymatic hydrolysis was improved from 0.31 to 2.05 g L-1 which was attributed to the ability of the [Ch]

[glycine] to partly dissolve lignin, thus making more accessible the hemicellulosic and cellulosic fraction to enzymes. Optimization has been recently reported by same authors [20] (Table 3.2).

In 2012, Itoh and co-workers reported the use of what they have called “ionic liquids inspired by Nature” for the dissolution of cellulose [21]. Although this work does not deal with the use of choline, it opens key data for improving the ability of ChCl-derived ILs to dissolve cellulose. After examination of the protein sequences of several cellulase (enzymes responsible for the hydrolysis of cellulose), authors hypothesized that amino-acids might be suitable anions to disrupt the hydrogen bond network of cellulose and thus enable its dissolution. As a cation, authors have highlighted the superior performances of N, N-diethyl-N-(2-methoxyethyl)-N — methylammonium, a cation with a structure close to that of an etherified cholinium. In particular, after combination of this cation with amino-acids such as tryptophan,

Scheme 3.4 Dissolution of microcrystalline cellulose in amino acid derived ILs

alanine, cysteine, lysine, among many others, cellulose was efficiently dissolved at 100 °C. Among tested amino-acids, alanine was found to be the best achieving up to 6 wt% of dissolution of cellulose at 60 °C. All other tested anions such as Cl~, AcO~, HO~ were found inefficient further supporting the pivotal role played by amino acid in the dissolution of cellulose. More importantly, these amino acid — derived ILs were tolerant to the presence of up to 2 wt% of water which is an important point considering that biomass always contained water. At higher tem­perature (150 °C), the dissolution rate of cellulose as well as the solubility of cellulose was greatly improved but a degradation of ILs was noticed at such temperature and authors recommended to not proceeding dissolution experiments at a temperature higher than 120 °C. Once dissolved, cellulose can be regenerated after addition of water as an anti-solvent. Precipitated cellulose was of the Type II supporting that such ILs are able to induce a change of the crystalline structure of cellulose. Impact of this change of crystallinity on the recalcitrance of cellulose to deconstruction was then evaluated in the enzymatic hydrolysis of cellulose. After regeneration, 88 % of cellulose was converted after 10 h of reaction at 50 °C and pH5 which has to be compared to the 40 % obtained without pre-treatment in ILs. In our views, this work is of prime importance within the scope of this chapter since it indirectly suggests that combination of amino acid with an etherified cholinium cation should provide competitive bio-inspired ILs for the dissolution of cellulose. Such aspect is the topic of current investigations in our group (Scheme 3.4).

Preparation of fractions 1-4 was carried out as follows; Crude fraction 1 was precipitated from IL using acetonitrile as non-solvent. The crude fraction was separated using centrifuge and washed with water, so that filtrate was retained. The solid residue was dried to give fraction 1. For preparation of fraction 2 addi­tional acetonitrile was added into IL-solution followed by similar separation and washing procedure. Fraction 3 was precipitated from IL-solution by water addition, after the acetonitrile had been concentrated by evaporation. Fraction 4 was prepared from the aqueous filtrates retained from the purification of fractions 1 and 2. Fil­trates were concentrated down to a volume ca. 2 mL, which was followed by addition of MeOH. Formed fluffy precipitate was filtered and purified with MeOH. Fraction 5 was prepared as follows. The aqueous IL-solution left from the precipitation of fraction 3 was concentrated so that nearly all water had removed. The remaining IL-solution was precipitated with MeOH (1:10 v/v). Formed precipitation was filtered in grade-3 sinter, washed with MeOH and the filtrand was dried in a vacuum over for 18 h at 40 °C to give fraction 5.

Reaction kinetic studies are not only important for uncovering the mechanism of 5-HMF formation on a molecular level, but also useful for process development studies to optimize process conditions and reactor configurations to obtain high 5-HMF yields. A number of kinetic studies have been carried out on the transfor­mation of various mono — or poly-saccharides into 5-HMF in different reaction medias, but herein we will mainly overview the kinetics studies reported in the ionic liquids system.

Moreau et al. [29] investigated the reaction kinetic of the conversion of fructose into 5-HMF in ionic liquid, 3-methylimidazolium chloride acting as both solvent and catalyst. The reactions were carried out at the temperature range of 90-120 °C with the initial fructose concentration of 0.01-2.5 M. A reaction order of one was applied for the kinetic analysis of the dehydration of fructose, obtaining an activa­tion energy of 143 kJ/mol for the reaction of fructose to 5-HMF. Wei et al. [96] reported the conversion of fructose to 5-HMF in 1-butyl-3-methyl imidazolium chloride using IrCl3 as the catalyst. The reactions were performed at temperatures ranging from 80 to 120 °C and a fructose concentration of 10 wt%. A kinetic network involving fructose conversion to 5-HMF and byproducts was proposed to model the experimental data using first order reaction kinetics. The activation energies for fructose conversion were estimated to be 165 and 124 kJ/mol for the formation of 5-HMF and formation of byproducts, respectively. Based on a reaction order of one, Qi et al. [20] reported an activation energy of 65 kJ/mol for the formation of 5-HMF from fructose, which was performed in 1-butyl-3-methyl imidazolium chloride in the presence of a strong cation exchange resin Amberlyst 15A. The authors contributed the lower activation energy to higher acidity of the strong sulfonic ion-exchange resin over that of the other catalysts.

Compared with fructose, fewer studies have been carried out for the kinetic study on the dehydration of glucose to 5-HMF, and these studies reported so far have been performed in aqueous systems, except for one work [19], which was performed in the ionic liquid 1-butyl-3-methyl imidazolium chloride using CrCl3 as catalyst under microwave irradiation. In that reaction system, an activation energy of 115 kJ/mol was reported using first-order reaction approach to model the experimental data [19]. The value of the activation energy of this work was comparable with the values (108-137 kJ/mol) reported for the dehydration of glucose in water with sulfuric acid as catalyst. However, the pre-exponential factor determined in that work was 3.5 x 1014 min-1, which was 3-8 orders of magnitude larger than those of previous works. The difference in the pre-exponential factors should be a result of the enhanced effective collision among reactant molecules in the homogeneous phase [19, 97]. It has been found that the activation energy for the side reactions that glucose decomposed to undesired humins is higher than for the desired reaction of glucose to 5-HMF, indicating that lower temperatures favor 5-HMF formation [98]. Therefore, the reaction temperature should be optimized for 5-HMF production.

In many kinetic studies of carbohydrates, 5-HMF is involved as an intermediate or decomposition product in reaction pathway to obtain sugars or target compounds and so the kinetics of 5-HMF is not well known. The decomposition of cellulose in aqueous systems is often modeled with first-order approaches that give activation energies for the decomposition of cellulose in water to be between 140 and 190 kJ/ mol [2]. Vanoye et al. [99] conducted a kinetic study on the acid-catalyzed hydrolysis of cellobiose in [EMIM][Cl] with 3.5 mM methanesulfonic acid catalyst in the presence of small amounts of water (3.5 mM) as the cosolvent. Activation energies of 111 and 102 kJ/mol were obtained for cellobiose hydrolysis and glucose degradation, respectively. Dee and Bell [100] performed kinetic studies on the cellulose hydrolysis in a batch set up in ionic liquids ([EMIM][Cl] and [BMIM] [Cl]) with mineral acid catalysts, and glucose, 5-HMF and cellobiose were the primary reaction products. The reaction kinetics for glucose formation were fit with first-order reaction rate equation and an activation energy of 96 kJ/mol was determined.

For 5-HMF production, the undesired decomposition of 5-HMF to levulinic acid, formic acid, and humins should be suppressed as much as possible. To gain insight in the reactivity of 5-HMF, kinetic studies using 5-HMF as the starting material should be investigated. However, kinetic studies on the decomposition of 5-HMF in ionic liquids system thus far are not available. Instead, there are many works that focus on the reaction kinetic of 5-HMF decomposition with 5-HMF as starting material in aqueous systems and organic-aqueous mixtures. For the rehy­dration of 5-HMF to levulinic acid, activation energies vary from 47 to 210 kJ/mol, and the activation energies range between 100 and 125 kJ/mol for the decomposi­tion of 5-HMF to humins. Asghari et al. [21] reported a higher activation energy for the formation of humins from 5-HMF (122 kJ/mol) than that for the decomposition of 5-HMF to levulinic acid (94 kJ/mol) catalyzed by HCl in subcritical water. On the contrary, Girisuta et al. [101] found similar activation energies of about 111 ± 2 kJ/mol for both the rehydration of 5-HMF to levulinic acid and the side reaction of decomposition of 5-HMF to humins.

Ionic liquids (ILs), are organic compounds containing salts with many attractive properties like extremely low vaporization pressure and melting point, excellent thermal stability and wide liquid ranges [2—5]. ILs have widely been used in many areas, for example, chemical synthesis, catalysis, biocatalysts and electrochemical devices [6—10]. ILs can be chosen to have different anions and cations so that one can form IL with the desired properties. Especially, some kinds of ILs with special functional groups have been designed for application in many industrial processes, such as imidazolium-based ILs, phosphonium-based ILs, amino-based ILs, acid-based ILs, and biodegradable ILs [11].

Lignocellulosic biomass is an abundant plant material and widely available, so that it has attracted much attention for conversion to fuels and chemicals [12,

13] . The main components of biomass are cellulose, hemicelluloses, lignin and other extractives. However, the complex structure of biomass makes its chemical degradation and biological conversion difficult to realize [14]. Pretreatment to disrupt the structures is necessary and a key procedure for biomass utilization.

Cellulose, as an important component of biomass, is composed of thousands of P — (1-4) — linked glucose units [15], which form many intermolecular or intramo­lecular hydrogen bonds [16]. Cellulose is widely treated with several organic solvents, such as N, N-dimethylformamide/nitrous tetroxide (DMF/N2O4) [17], N, N-dimethylacetamide lithium chloride (DMAc/LICl) [18], N-methylmorpholine (NMMO) [19] and dimethyl sulfoxide (DMSO)/tetrabutylammonium fluoride (TBAF) [20]. These traditional solvents suffer from volatility, toxicity, and solvent recovery issues [21, 22], so novel solvents, such as ILs, have received attention for cellulose dissolution in recent years. Cellulose dissolution with present ILs dates back to 2002 [23], before which, a solvent system for dissolving cellulose was discovered by Graenacher [24].

5.2.1.1 Cellulose Esters of Short C2 to C6 Carboxylic Acids

Cellulose esters of short chain carboxylic acids, in particular acetates, acetate- propionates, and acetate-butyrates, are of huge commercial importance and found in many applications in the form of fibers, films, coatings, and additives [50, 51]. Nowadays, these derivatives are exclusively prepared by heterogeneous pro­cesses. However, homogeneous esterification in ILs is considered as a commer­cially attractive alternative. Various patents have been published, e. g., by Eastman Chemical Company, that describe the preparation of cellulose esters and mixed esters in imidazolium — and ammonium based ILs, including the synthesis of the solvents and their recycling subsequent to the esterification [52—55].

Conversions of cellulose with carboxylic acid chlorides and anhydrides proved to be very efficient when performed homogeneously in an IL. The reactions are usually performed at elevated temperature (>80 °C), thus the adverse effect of high viscosity of cellulose/IL solutions is less pronounced. The esterification of cellulose proceeds completely homogeneous even up to a complete derivatization of all hydroxyl groups. Using relatively small amounts of acetic anhydride or chloride (3-5 mol equivalents, Table 5.2), highly functionalized cellulose acetates with DS up to 3 can be obtained within short reaction times (0.5-8 h) [12—14]. Using BMIMCl as homogeneous reaction medium, acetylated derivatives could even be obtained from bacterial cellulose that is usually difficult to dissolve and to chem­ically modify in other cellulose solvents due to its high degrees of crystallinity and polymerization [15].

In analogue to the homogeneous acetylation in ILs, homologues cellulose esters of higher carboxylic acids, from propionates up to hexanoates, have been prepared in various ILs using the corresponding anhydrides [16—18, 56]. The DS values of cellulose esters were found to decrease successively when increasing the number of carbon atoms in the acyl moiety from 2 to 4 but to increase again with further prolongation of the alkyl chain up to 6 carbon atoms [17]. Cellulose pentanoates

Reaction conditions Product

aAMIMCl: 1-allyl-3-methylimidazolium chloride, BMIMCl: 1-butyl-3-methylimidazolium chloride bBC: bacterial cellulose (DP: 6,493), CH: cellulose from corn husk (DP: 530), DIP: dissolving pulp (DP: и 650), MC: microcrystalline cellulose (DP: 286) c+: soluble, —: insoluble, n. a.: no information available dAcetic acid derivative used

eMolar ration of acetylation reagent to anhydroglucose units fAdditionally, 2.5 mol equivalent pyridine

and hexanoates slightly exceeded the DS values of cellulose acetates, prepared under identical reaction conditions. In contrast, the corresponding propionates and butyrate had slightly lower DS values. Cooperative interaction of the long chain anhydrides with the partially substituted, i. e., lipophilic, cellulose chain have been postulated to explain this unexpected finding.

The efficiency of the esterification of cellulose in ILs can be increased by adding pyridine (stoichiometric amounts) or 4-dimethylaminoaminopyridine (DMAP; cat­alytic amounts) [14, 18]. Moreover, microwave assisted esterification can yield products with an increased DS in comparison to products prepared under conven­tional heating [17, 56]. Within a microwave field, ILs rapidly heat due to their ionic nature [57]. Thus, the irradiation must be controlled by monitoring power input, pulse length and — interval, and maximum temperature. Efficient mixing is also

crucial, especially in case of highly viscous cellulose/IL solutions, in order to avoid local ‘hot-spots’ of extremely high temperature. Otherwise, degradation of cellu­lose and carbonization might occur. In addition to microwave assisted cellulose derivatization in ILs, esterification with the aid of ultrasound irradiation has been reported [32].

Parallel conversion of cellulose, dissolved in an IL, with two different carboxylic acid anhydrides yields mixed cellulose esters (Table 5.3) [16, 17, 58]. The product properties (e. g., hydrophobic/hydrophilic character) can be tailored by variation of ester moieties, their partial DS values, and the overall amount of substituents, attached to the cellulose backbone. As already pointed out, the reactivity of carboxylic acid anhydrides is dependent on the length of the alkyl chain [17]. In addition to reaction temperature, time, and amount of acylation reagent, the sequence of adding the two different anhydrides (simultaneous vs. step-wise) is consequently of huge importance. Considering the huge commercial importance of mixed cellulose esters, in particular acetate/propionates, acetate/butyrates, and propionate/butyrates, further comprehensive studies are required in order to eval­uate the individual effect of reaction parameters on the product composition, including the distribution of ester moieties within the AGU and along the polymer chain. Moreover, choice of the IL and its recycling subsequent to the reaction (see Sect. 5.3.3) are going to be important issues.

The most common biomass source, lignocellulose, is the principal component of plant matter and is the largest renewable resource available [1]. Lignocellulose is composed primarily of three biopolymers: cellulose, hemicellulose, and lignin. Cellulose, which is the most abundant biopolymer in lignocellulosic biomass, is composed of glucose monomers linked together through 1-4 glycosidic linkages. As shown in Fig. 8.1, these chains of glucose hydrogen bond with the hydroxyl groups of neighboring cellulose molecules, providing a stable, crystalline structure to cellulose fibers in cell walls [23]. Because cellulose is the single most abundant renewable resource available, there has been significant work in its utilization across a wide range of applications. Cellulose can be somewhat difficult to break into its component glucose units and there are a number of methods, such as enzymatic or acid catalyzed hydrolysis, that will convert cellulose into monosac­charides or short carbohydrate chains [24, 25].

Hemicellulose, which, like cellulose, is a polymer composed of monosaccha­rides, makes up 20-30 % of plant biomass. Unlike cellulose, however, hemicellu — lose is a branched carbohydrate that can be made up of multiple monosaccharides, bonded through a number of different glycosidic linkages. The structure of hemi- cellulose is composed of a polysaccharide backbone made from glucose, xylose, or mannose units connected through p—(1—3) or p-(1 —4) glycosidic bonds. From these backbones, there are side chains of glucose, glucuronic acid, 4-O-methyl- glucuronic acid, mannose, xylose, arabinose, or galactose [26]. The composition of the hemicellulose is dependent on the plant species that produced it [26, 27]. Compared to cellulose, hemicellulose is significantly easier to hydrolyze into small carbohydrate chains and monosaccharides. Currently, there are a number of methods for hemicellulose extraction and degradation, including steam explosion, dilute acid treatment, and ammonia explosion [9].

Lignin is the third kind of structural biopolymer, which composes 15-30 % lignocellulosic biomass by weight [28]. In the structure of the cell wall, lignin fills the space between cellulose/hemicellulose fibers. Unlike cellulose and hemicellu — lose, lignin is not made from carbohydrates but from phenylpropane units that are linked through enzymatic radical polymerization [29]. The monomers that plants employ to create lignin are cinnaminyl alcohol, sinapyl alcohol, and p-coumaryl alcohol. These monomers are bonded together through a number of different linkages that form a complex, amorphous structure. The most common of these bonds are the p-O-4, 5-5, P-5, P-1, and a-O-4 linkages (Fig. 8.2), which represent 45-50, 18-25, 9-12, 7-10, and 6-8 % of the linkages in softwood lignin, respec­tively [30]. Lignin is a major inhibitor of biological degradation of lignocellulose, as there are only a few species in nature that can effectively metabolize it [31, 32]. While lignin benefits living plants, it presents a significant challenge to the successful utilization of biomass in the production of fuels or other chemicals.

Currently, lignin is processed through a number of different techniques. The pulp and paper industry generally uses a process called kraft pulping, in which a strong bases and sulfur compounds depolymerize and extract lignin from wood pulp [30, 33]. Other methods, such as acid pulping, organosolve pulping, and high temperature ethanol/water have been used to degrade lignin have also been employed on an industrial scale [34—38]. In other cases, lignin depolymerization, along with disruption of cellulose crystallinity, has been used for biomass pretreatment for further processing [9, 14].

Other sources of biomass can be used as substrates. Starches and simple sugars are currently used in the production of fuel ethanol. These carbohydrates can be sourced from corn, sugar cane, beets, or as a product from the depolymerization of longer chain polysaccharides [20, 39]. Algae have also received significant atten­tion as a source of renewable energy feed stocks [15]. In research on catalysis of biomass, simple sugars are often used as a model or substitute for more complicated carbohydrates and biomass in general. Other sources of biomass may be similar to common lignocellulose, but have unique characteristics that warrant special

attention. Rice hulls, for example, are coated in a layer of silica that makes effective catalytic conversion difficult [40].

In recent years, the application of ionic liquids (ILs) as (co)solvents and/or reagents in enzymatic catalysis for the production of biofuels is an emerging research area (Fig. 11.1).

ILs are composed entirely of ions and are liquids at ambient or far below ambient temperature, and have been extensively used as a potential alternative to toxic, hazardous, flammable and highly volatile organic solvents. ILs, particularly, have been shown to be exceptionally interesting non-aqueous reaction media for enzy­matic transformations [8, 9]. Typical ILs are based on organic cations, e. g. 1,3- dialkylimidazolium, tetraalkylammonium, etc., paired with a variety of anions that have a strongly delocalized negative charge (e. g. [BF4], [PF6], etc.), which results in colourless and easily handled materials with very interesting properties as solvents [10]. Their interest as green solvents resides in their negligible vapour pressure, excellent thermal stability, high ability to dissolve a wide range of organic and inorganic compounds, and their non-flammable nature, which avoids the problem of the emission of volatile organic solvents to the atmosphere. Moreover, their solvent properties, such as miscibility or immiscibility with water or some

organic solvents (e. g. hexane), can be tuned by selecting the appropriate cation and anion, which increases their usefulness for recovering products from the reaction mixture [8]. So, an important aspect for an IL is that it able to partially or completely solubilize the reactants, as well as low solubility for the reaction products. Solubility with reactants improves the reaction by allowing reactants to come into contact with each other, while the product can be separated by simple decantation as it is insoluble with ionic liquids, which can therefore be recycled.

The use of ILs as a reaction medium for enzymes, the application of such green compounds as cosolvents and/or reagents for biotransformation is well recognized. Studies on enzymatic reactions in ILs over the last 8-9 years have revealed not only that ILs are environmentally friendly alternatives to volatile organic solvents, but also that they have excellent selectivity, including substrate, regio — and enantios — electivity. Besides, many enzymes maintain very high thermal and operational stability in ILs. In this strategy, enzymes are simply suspended in the ILs, and then the resulting mixture can be used for biocatalytic reactions. It has been reported that lipase from Candida antarctica immobilized with IL is active at very high temperatures (95 °C) in hexane — and solvent-free conditions [11].

The most interesting biotransformations in ILs were observed at low water contents or in nearly anhydrous conditions because of the ability of hydrolases to carry out synthetic reactions. Furthermore, it is possible to design two-phase reaction systems that easily permit product recovery [12—14]. In anhydrous condi­tions, most of the ILs miscible with water clearly act as enzyme deactivating agents (e. g. [BMIM][Cl] or [BMIM][NO3]), with a few exceptions (e. g. [BMIM][BF4]) [8, 15, 16]. However, all the water-immiscible ILs assayed (e. g. [BMIM][PF6] or [BMIM][NTf2]) were shown to be suitable reaction media for biocatalytic reactions at low water content (<2 %, v/v) or in anhydrous conditions. In this context, lipases
are enzymes that have been most widely studied because of the high level of activity and stereoselectivity displayed in synthesizing many different compounds, e. g. aliphatic and aminoacid esters [17], chiral esters by kinetic resolution of sec-alcohols [12—14, 18, 19], flavonoid derivatives [20], polymers [15, 21], etc. Numerous other enzymatic reactions have also been reported in ILs [22]. Further­more, water-immiscible ILs also have an important stabilizing effect on hydrolases (lipases, esterases, proteases, etc.) in nearly anhydrous conditions [23—26].

We will focus on the use of enzymes in ILs for biofuels production.