Obtención de furanos empleando Líquidos Iónicos

Resumo

SUMMARY Introduction In recent years, the scarce availability of fossil resources, the volatility of oil prices and the environmental impact of many processes based on the use of fossil resources have led to a critical reconsideration of certain chemical and energy production technologies. The “sustainable development” principles require the gradual utilization of renewable raw materials in advanced processes, which are expectec to be more efficient and with less environmental impact than the conventional ones. On this context, lignocellulosic materials (LCM) are expected to play a key role, as they are largest renewable source of organic carbon. Among the possible strategies suitable for LCM utilization, the strategies ones based on the biorefinery principle (which proposes the selective fractionation of raw materials and the individual processing of the different fractions) are of special interest for the objectives of this study. Following this general idea, the added value from LCM can be optimized by using processed based on green, efficient and cost-effective technologies causing a limited environmental impact (Budarin et al., 2011; Salazar y Cárdenas, 2013; Peleteiro et al., 2015b). The main goal of this study is focused on the production of value-added chemicals from hemicellulosic saccharides of Pinus pinaster and Eucalyptus globulus, in monophasic or biphasic media containing ionic liquids (ILs). As a starting point, and for comparative purposes, experiments were carried out on media containing commercial monosaccharides (which are model substrates for the processes considered here). This work has been developed in the framework of one of the research topics (development of chemical processes for the valorization of vegetal biomass) driving the research of the EQ-2 group (which belongs to the Chemical Engineering Department of the University of Vigo) at the Campus of Ourense. More specifically, the experimental work has been defined on the basis of the tasks included in two Research Projects (“Development and evaluation of processing methods for biorefineries”, reference CTQ2011-22972; and “Advanced processing technologies for biorefineries”, reference CTQ2014-53461-R) funded by the “Ministry of Economy and Competitivity” in specific calls of the Spanish National Research Plan. Lignocellulosic materials According to the F.A.O. (Food and Agriculture Organization of the United Nations), the existing vegetal biomass in the world in 2015 was distributed in 3999 million hectares (ha), with a total production of around 146 billion metric tons. More than 80% of this biomass corresponds to LCM (Hon, 2000). The different LCM types present a common feature: they are mainly composed of polysaccharides (including cellulose and hemicelluloses) and lignin (a three-dimensional, amorphous polymer of phenolic nature). LCM hold a great potential as feedstock for the chemical industry, due to their chemical composition and to other important features, such as their abundance, low cost, wide geographic distribution and versatility in terms of the type of applicable processes and the wide arrays of potential market products. The feasibility of processes for LCM biorefining depends on many factors, such as the efficiency of the fractionation (and eventually, of the subsequent processing of the fractionation products), the limitation of the environmental impact, and cost effectiveness, in agreement with the philosophy of "sustainable development" (Ligas et al., 2003, Zhao et al., 2009, Cheng et al., 2008). Sustainable development can be defined as “the development that meets the needs of the present without compromising the ability of future generations to meet their own needs” (Brundtland Report, 1987). Based on this statement, the consumption of natural resources should be limited to a level compatible with the capacity of the environment to generate them indefinitely. On he other hand, the pollution should be limited, and the production capacity shoud be adjusted to the demand. This involves interrelated factors influencing the economy, the environment and the society (Ghatak, 2011). In this study, P. pinaster and E. globulus woods were selected as starting materials for chemical processing due to their availability and accessibility. In the world, the total area covered by P. pinaster is estimated at around 4.4 million ha; whereas in Galicia, this species occupies 467351 ha (or 23% of the total Galician forests). Eucalyptus is distributed worldwide with 22 million ha approximately (not including more than 11 million ha of native forests in Australia), accounting for 12% of the world´s forest plantations. However, owing to productivity-related reasons, it is estimated that no more than 13 million ha of these plantations have real industrial interest. The plantation area in Spain is 760000 ha (which represents only 3% of the Spanish forest area) distributed mainly in Galicia (395000 ha, equivalent to approximately 52% of the total Galician forests) and Andalucía (155954 ha, around 21% of total forest land). These data suggest that the increasing the production of Eucalyptus wood could result in opportunities for a sustainable economic and social development. Fractionation of lignocellulosic materials LCM valorization can be achieved by means of two different philosophies: they can be used as a "whole" (without separation of components); or after separating "fractions" or arrays of compounds (which can be further used for specific purposes) (Peleteiro et al., 2015b). Regarding the separation of fractions, it has to be hihglighted that conventional methos for physical separation (for example, crystallization, extraction or precipitation) are not suitable for achieving the individual and simultaneous isolation of the different polymeric constituents from the raw materials, due to the close chemical and physico-chemical connections between them. An efficient fractionation requires the selective depolymerization of one (or several) of the LCM polymeric constituents (cellulose, hemicellulose or lignin). The resulting fractions from these treatments can be processed individually following the biorefinery philosophy, to achieve specific goals (FitzPatrick et al., 2010; Liu et al., 2012). The use of LCM based on fractionation requires an efficient and separation of their polymeric componentes, in a way that each fraction is rich in a given compound or in compounds that come from a specific precursor present in from the raw material. At the same time, each fraction should not contain significant amounts of compounds derived from the rest of polymeric components. In addition, raw materials must be exploited in the most complete way possible, in order to achieve high yields in the target products, using the minimum number of stages and applying technologies ensuring profitability (Liu et al., 2012). Solubilization of hemicelluloses In the last years, special attention has been paid to the selective separation of hemicelluloses from native MLC (Peng et al., 2012). Hemicelluloses can be separated from the raw materials as polymers (for example, as the soluble fraction in alkaline solutions), or after depolymerization (for example, by treatments in aqueous media in the presence or absence of externally-added acidic catalysts). Treatments with hot, compressed water (autohydrolysis or hydrothermal processing) or steam followed by a sudden decompression (steam explosion) take benefit of the susceptibility of hemicelluloses to hydrolysis reactions, allowing their selective solubilization at high yields. As a result of these treatments, solutions rich in oligosaccharides or low molecular weight polymers (which also contain monosaccharides and products derived from them, such as HMF and furfural) are obtained, together with solids enriched in cellulose and lignin. The components of the solid phase may be further fractionated by suitable reactions (for example cellulose hydrolysis by acid or enzymatic processing; or lignin solubilization treatments, for example using sulphite, sulphate or organosolv methods). The study of the composition and the use of the solids obtained in the autohydrolysis stage is not considered in this work, which is focused on the use of the hemicellulosic fraction. For this purpose, P. pinaster and E. globulus wood were subjected to autohydrolysis treatments, in order to solubilize its hemicellulosic fractions selectively. The raw materials were treated in aqueous media at elevated temperatures (at 175 °C for 26 min in the case of P. pinaster, and at 196 °C in the case of E. globulus with immediate cooling after reaching this temperature). In the reaction media took place the partial hydrolysis of the glycosidic bonds from the hemicellulosic polymers, due to the catalytic action of the hydronium ions (from the autoionization of water and the organic acids generated in situ, including acetic acid, uronic acids and phenolic acids). The partial hydrolysis of hemicelluloses results in the production of oligomers, sugars and acetic acid, which can be in turn decomposed to yield reaction byproducts in proportions that depend on the type of raw material and on the operational conditions (Garrote et al., 1999 y 2002; Kabel et al., 2002; González Muñoz et al., 2012). The hemicellulosic fraction of softwoods (such as P. pinaster) is mainly made up of glucomannan, in which the most abundant structural units are anhydromannose (accounting for around 40-45% of the total). Additionally, this type of woods also contain hemicellulosic polymers (xylans) made up of anhydroxylose, which account for about 25-30% of the total hemicelluloses. Wood glucomannans present two types of substituents: esterified acetyl groups in positions 2 and 3, and galactose units linked by α-(1,6) bonds (Ebringerová et al., 2005). On the other hand, hardwoods (such as E. globulus) have hemicelluloses consisting mainly of xylan (accounting usually for more than 70% of the total hemicelluloses), which may also contain esterified acetyl groups and uronic substituents. Finally, E. globulus hemicelluloses have limited amounts of polymers containing anhydromannose units. Hydrothermolysis is considered in literature considers as a "green" technology, because the reaction medium contains only the lignocellulosic raw material and water. This fact allows the design of simple processes with limited environmental impact, avoiding problems related to corrosion and to the formation of neutralization sludges, typical of the acid prehydrolysis stages (Gullón et al., 2012). Nevertheless, an exhaustive study of the autohydrolysis kinetics is complex, since it depends on a large number of variables. Among them, the most influential are the composition and the distribution of particle sizes of the solid substrate, the water/solid mass ratio, the temperature and the reaction time. Other influential variables related to the type of substrate include the polymerization degree of the structural components, and their with each other or with other constituent fractions from the native raw material, such as proteins and mineral elements; particularly if they have the capacity to neutralize the acids generated in the reaction media that possess catalytic activity. Production of furans from hemicelluloses The production of furans (such as HMF and furfural) and certain organic acids (such as levulinic and formic acid) is carried out from sugars (hexoses or pentoses, depending on the target product). All these compounds have a huge potential as intermediates for the manufacture of fuels and chemicals (Van Putten et al., 2013a). Three main possibilites have been described for the production of HMF: dehydration of hexoses in presence of an acid acting as a catalyst (a widely studied approach), Maillard reactions between saccharides and nitrogen compounds (including amino acids, peptides, proteins or amines), or aldol condensation of molecules containing three carbon atoms. Due to the focus of this work, special interest is paid to acid treatments leading to the formation of HMF (from hexoses or hexose precursors such as oligosaccharides or polymers), or to the formation of furfural (from pentoses or pentose precursors). When the sugars or higher saccharides are heated in acid media, a number of simultaneous reactions takes place (including hydrolysis of polymers or oligosaccharides, isomerization of sugars produced by hydrolysis or already present in the media, dehydration of intermediates into furans, rehydration of these latter, and side-reactions leading to the formation of insoluble substances known as humins), according to a complex overall mechanism (Kuster, 1990; Lewkowski, 2001; Rackemann and Doherty, 2011; Mukherjee et al., 2015). Since the furans act as intermediates in this set of reactions (being generated and consumed in the medium), their yields are often significantly below stoichiometric ones. Owing to economic reasons, the production of furans from polysaccharide-containing raw materials has been considered en literature, which they can be employed as cheap substrates to give rise to the set of reactions described in the previous paragraph. Nowadays, neither of the tested routes for the production of HMF from hexosans and/or hexoses is considered to be definitive (Van Putten et al., 2013a). When a ketose (as fructose) is employed as a substrate, it has been proposed that the formation of HMF takes place through a direct reaction, which could justify why the furan yields obtained with this type of sugars are higher than in the cases when other types of sugars were used as a substrate (Rackemann y Doherty, 2011). For example, the formation of HMF from aldoses (such as glucose, galactose or mannose) (Binder et al., 2010b) involves a previous step of isomerization, with the formation of an enediol as an intermediate which is subsequently transformed into HMF (Rackemann and Doherty, 2011, Peleteiro et al., 2014a). Similarly to the situation described for hexoses, the pentose dehydration into furfural (a well-known reaction, in which the industrial process for furfural manufacture is based) may proceed according two mechanisms (which can be alternative or simultaneous): the first one accepts that the reaction proceeds via a cyclic intermediate (Antal et al., 1991); while the second one postulates the formation of acyclic intermediates, either begining with the isomerization of the pentose and its subsequent enolization (Zeitsch, 2000), or by a direct conversion into xylulose through hydride transfer (Enslow y Bell, 2012). The overall kinetics of furfural generation and decomposition is complicated, and it is still a subject of discussion (Peleteiro et al., 2016a). In the same way described from HMF production from hexoses or anhydrohexose units present in higher saccharides, a complex set of series and parallel reactions take place when pentosans in aqueous acidic media. The overal mecanism includes hydrolysis of polysaccharides or higher saccharides, formation of intermediates fom the sugars generated in the previous stage, dehydration of intermediates into furfural and side-reactions consuming furfural. Since furfural is an intermediate in this set of reactions, its yields can be limited when operating under practical conditions (Weingarten et al., 2010; Cai et al., 2014). The furfural-consuming reactions lead to soluble and insoluble unwanted byproducts, including humins (Antal et al., 1991; Nimlos et al., 2006; Aida et al., 2010; Kumar et al., 2013; Rasmussen et al., 2014; Peleteiro et al., 2015a,c). Production technologies Literature has been reported on the technologies of HMF and furfural production, and their comparative advantages and disadvantages have been assessed (Rosatella et al., 2011; Van Putten et al., 2013; Teong et al., 2014; Mukherjee et al., 2015). As an alternative to the conventional aqueous process, methods based on the use of other reaction media have been proposed (including ILs, eventually in the presence or an insoluble organic co-solvent). The studies reported in media containin ILs include operation in the presence of homogeneous or heterogeneous catalysts, or based on the utilization of acidic ILs able to perform simultaneously as a solvent and as a catalyst. The choice of one of these alternatives limits the yield in the final product, among other factors. The most usual homogeneous catalysts employed for manufacturing furans from suitable substrates (sugars, polysaccharides or polysacchride-containing biomass) behave as Brönsted acids; even if some studies deal with the utilization of Lewis acids (for example, AlCl3, SnCl4, VCl3, InCl3, GaCl3, LaCl3 or YbCl3), operating individually or in combination with Brönsted acids (Pagán-Torres et al., 2012; Choudhary et al., 2013). Alternatively, the production of furans has been carried out using heterogeneous catalysts (named “solid acid catalysts”, for example polymeric resins, zeolites, natural clays, metal oxides as Al2O3 and TiO2, or supports supplemented with ionic forms of Pt, Ni, Cu, Co or Pd). The heterogeneous catalysts present advantages over the homogeneous catalysis, including the easy separation and reuse of the catalyst, the ability to operate at high temperatures (which reduces time and improves selectivity) and the limited equipment corrosion (Tong et al., 2010a; Dias et al., 2010). The next section describes the production of furans from model sugars or hemicellulosic saccharides of P. pinaster and E. globulus in reaction media containing ILs, topics in which the objectives of this study are focused. Ionic liquids Ionic liquids (ILs) are salts made up of organic cations and inorganic or organic anions (asymmetrical and bulky), which differ from molecular solvents by their chemical nature, structure, organization, and properties (Casas, 2013; Peleteiro et al., 2016a). Many ILs melt at temperatures below 100 °C, and most of these latter are liquids at room temperature. These properties, together with their very low vapor pressures (which limit their losses by evaporation and facilitate the recovery and reuse, allowing ILs to be considered as “green” solvents), define important advantages in both the technological and environmental fields. There are other aspects of the ILs that increase their potential in the chemical technology, for example the possibility of designing their synthesis according to the specific needs of the process in which they are to be used (which is known as “tailor-made synthesis” or “customized synthesis”) (Montón, 2011; Casas, 2013). The possibility of adjusting the physical and chemical properties of the ILs by acting on the chemical nature of cations and/or anions has allowed applications in several sectors (as electrochemistry, biocatalysts, separation, synthesis of materials, etc.). In particular, ILs can play different roles in the production processes of furans, including: - Acidic catalysts for the dehydration of hexoses and/or pentoses in aqueous media, eventually in the presence of organic co-solvents (Qi et al., 2010; Tao et al., 2011; Serrano-Ruíz et al., 2012; Peleteiro et al. 2016a; Amarasekara, 2016). - Additives to improve the furan yields in reaction media made up of suitable substrates (hexosans, hexoses, pentosans or pentoses) (Binder y Raines, 2009; Binder et al., 2010a) in the presence of organic solvents, externally-added catalysts and (eventually), co-catalysts. - Reaction media for manufacturing furans from monosaccharides (hexoses and/or pentoses) or higher saccharides (made up of hexoses and/or pentoses, either pure or present in native LCM); eventually in the presence of co-solvents and externally-added catalysts (Van Putten et al., 2013a; Peleteiro et al., 2016a; Guang-Way et al., 2016). The reaction can be carried out in acidic ionic liquids (AILs), which perform simultaneously as solvents and catalysts, allowing the design of simple and efficient processes with easy recovery of solvent and product (Mukherjee et al., 2015; Peleteiro et al., 2016a; Amarasekara, 2016). The above information justifies why the scientific community is paying a growing attention to the application of IL in a number of fields, including the development of processes for LMC biorrefineries (Rogers y Seddon, 2003; Xie et al., 2007; Liebert y Heinze, 2008; Plechkova y Seddon, 2008; Zhu, 2008; Torr et al., 2012; Wang et al., 2012; Yang et al., 2013). The ILs can be employed in biorefineries, for example, as separation agents, as reagents capable of dissolving the cellulose (with or without further transformation in their derivatives), as agents for LCM fractionation (Peleteiro et al., 2015b), and as reaction media and/or catalysts to obtain furans from poly-, oligo- or monosaccharides (Peleteiro et al., 2016a,b,c). In this study, three types IL-containing media have been considered: homogeneous mixtures of 1-butyl-3-methylimidazolium chloride [(bmim]Cl) and a Lewis acid catalyst; the AIL, 1-butyl-3-methylimidazolium hydrogen sulfate ([bmim]HSO4) acting simultaneously as a reaction media and a catalyst suitable to carry out the hydrolysis-dehydration of higher saccharides (without adding other catalyst); or the same AIL acting as an acidic catalyst for aqueous media. - Production of HMF in IL-containing media The available literature includes studies on the HMF production from hexoses, disaccharides, oligosaccharides or polysaccharides in different reaction media, including ILs (Tong y Li, 2010; Binder et al., 2010b; Rosatella et al., 2011; Van Putten et al., 2013a; Mukherjee et al., 2015). A comparison of the results published on the production of HMF from model compounds (mainly, fructose and glucose) confirms the possibility of reaching high substrate consumptions, but the HMF yields are strongly influenced by the operating conditions. For example, when the substrate concentrations are comparatively high, the HMF yields decrease significantly due to the enhanced formation of by-products (including levulinic acid, formic acid and humins). The situation is even more complex when native lignocellulosic raw materials or solutions from their fractionation are used, since different compounds present in the reaction medium (such as sugars, intermediates or HMF) may participate in reactions consuming reactive species. In general, obtaining satisfactory results (high consumption of substrate with a favorable selectivity toward the formation of the desired product) requieres the selection of adequate operating conditions (including the nature of the medium and the type of catalyst) (Guang-Way et al., 2016). In recent years, different ILs suitable to be employed for converting suitable sustrates (monosaccharides, oligosaccharides and/or polysaccharides) into HMF in the presence of external catalysts have been identified (Zakrzewska et al., 2011; Ståhlberg et al., 2011a). The most common cations in this type of ILs are 1-ethyl-3-methylimidazolium ([emim]) and 1-butyl-3-methylimidazolium ([bmim]), while the most frequent anions are chloride and bromide (Guang-Way et al., 2016). Despite the huge potential of HMF in the framework of sustainable development, it is not currently manufactured industrially. Their large-scale utilization depends on the development of new, efficient and cost-effective processes able to extract the whole commercial potential from this compound. - Furfural production using ionic liquids According to a report commissioned by the US Department of Energy (Werpy and Petersen, 2004), which was further updated by Bozell and Petersen (2010), furfural is one of the 30 “green” chemicals with greatest potential that can be produced from plant biomass. Xylose and xylans (polymers composed of anhydroxylose units) are the most studied model substrates for the furfural production. In practice, owing to economic and technological factors, there is a comparative advantage when the feedstock employed is either a native raw material containing xylan (for example, almond shells, bamboo, wheat and rice straws, cotton seeds and woods), or a biorefinery stream containing higher saccharides made up of anhydroxylose units. In media containing polymers or oligosaccharides formed by anhydroxylose units, the first reaction is the hydrolysis of the potential substrates into xylose, which is followed by the dehydratation of this sugar, together with productive reactions (dehydration to furfural) and reactions where the desired product and/or reactive species are consumed (Lange et al., 2012; Peleteiro et al., 2015a). The furfural production at the commercial scale began in 1920, due to the interest of the Quaker Oats Company to get added value from xylan-rich agricultural residues. Currently, furfural is commercially produced in different parts of the world (typically from corn cobs, a material wtih a high xylan content), using a batch process for heating the raw material in an aqueous medium catalyzed with sulfuric acid. This technology has technological and environmental disadvantages, including limited furfural yields (usually, in the range 45-55% of the stoichiometric one), high energy consumption, equipment corrosion, catalyst recovery and environmental hazards (mainly related to the solid residue from the process). From the above ideas it can be concluded the development of new processes for furfural production, which should be cost-effective, “green” and sustainable, is of interest. Some improvements have been proposed to conventional technologies, including the utilization of distillation (conventional, stripping or flash) in order to remove furfural from the reaction medium when it is generated, or the utilization of biphasic systems, new catalysts or catalyst mixtures, or alternative reaction media (including the ones containing ILs). The most relevant results of this study have been reported in seven articles (all of them published in journals included in the “Science Citation Index”) and one patent. These documents are included as Annexes in this report. Initially, for comparative purposes, the influence of the reaction conditions on the consumption of model substrates (hexoses and pentoses present in hemicelluloses of hardwoods and softwoods) and on the generation of furans (HMF and furfural) was studied, using reaction media containing [bmim]Cl as a solvent and CrCl3 as an external catalyst. The effects caused by the most influential operational variables were assessed from experiments performed under diverse conditions (see Annex B for detailed information). The results confirmed that glucose and mannose behaved as suitable substrates for HMF production. The maximal conversions of substrates into furfural were 66.8% for glucose and 62.4% for mannose. However, the production of furfural from xylose and arabinose proceeded at lower yields, with maximum molar conversions in the target product in the vicinity of 50% and 16%, respectively. The co-production of HMF and furfural from hemicellulosic saccharides of P. pinaster (containing hexoses, higher saccharides made up of hexoses, pentoses, and higher saccharides made up of pentoses) was assessed in experiments performed with reaction media containing the same solvent and catalyst cited above. The results (see Annex D for detailed information) showed that the substrates reacted according to a complex kinetic mechanism, with multiple series- or parallel- reactions. Among them, the hydrolysis of hemicellulosic saccharides (including low molecular weight polymers and oligomers) into their constituent sugars, the dehydration of sugars into furans, and reactions in which these latter are consumed to yield undesired compounds were the most important ones. Operating under selected conditions, the maximum conversion of potential substrates (hexoses and anhydrohexoses making part of higher saccharides) into HMF was 36.1% (corresponding to 57.2% of the result obtained with the model substrate). The consumption of pentoses and anhydropentoses in xylan-derived saccharides proceeded with a comparatively fast kinetics, in a way that the maximum conversion of potential substrates into furfural (37.7%) was reached under milder operating conditions than in the case of the hexoses. Interestingly, the conversion of pentoses and hemicellulosic saccharides composed of anhydropentoses into furfural was in the range of the one observed for commercial xylose. Alternatively, simple and efficient operational protocols were developed for obtaining furans in AIL ([bmim]HSO4) media, since this compounds shows ability to perform as a reaction media and a catalyst at once. The substrates employed in these experiments were commercial xylose (see Annexes E and F for detailed information) or hemicellulosic saccharides from E. globulus wood (see Annex G for detailed information). In selected experiments, an extracting agent (toluene, MIBK or dioxane) was used, in order to assess if this type of biphasic systems, in which furfural is transferred to the organic phase to limit its consumption through undesired side reactions (particularly, resinification and condensation with formation of humins), could improve the product yields. The optimum conversion of xylose to furfural was above 80%, whereas a maximum yield close to 60% of the stoichiometric value was obtained in experiments using hemicellulosic saccharides E. globulus as substrates. An alternative way, consisting on the utilization of [bmim]HSO4 as a catalyst for aqueous media, was explored to convert xylose or hemicellulosic saccharides from E. globulus into furfural, operating in the presence of an organic co-solvent (toluene or MIBK). The experimental results (see Annex H for detailed information) allowed the identification of operational conditions under which the conversion of xylose into furfural reached 71.2% yield. In comparison, the maximum conversion of the potential substrates in hemicellulosic saccharides of E. globulus into furfural was 61.6%. Annex H also includes data on the recovery of the [bmim]HSO4 contained in reaction media from the processing of commercial xylose or hemicellulosic saccharides, and its reutilization as a catalyst for further reaction stages. The separation and recycling of the AIL entailed limited losses of catalytic activity, in a way that the average conversion of the potential substrates from E. globulus wood after 8 consecutives stages of reaction with recycled catalyst accounted for 94.2% of the average conversion determined for solutions containing commercial xylose.