– BY CONTRIBUTION TO THE PRODUCTION AND USE OF BIOMASS-DERIVED SOLVENTS – A REVIEW

In this review key processes for the synthesis of greener or more sustainable solvents derived from renewable sources (saccharides, lignocellulose and triglycerides) are discussed. It is shown that a series of platform chemicals such as glycerol, levulinic acid and furans can be converted into a variety of solvents through catalytic transformations that include hydrolysis, esterification, reduction and etherification reactions. It was also considered several aspects of each class of solvent regarding performance within the context of the reactions or extractions for which it is employed.


Introduction
In the last decades, there has been an increasing effort to reduce the use of petroleum-derived chemicals and fuels in order to decrease environmental pollution and to counteract global warming [1,2]. Innovative strategies for the sustainable obtaining of these products has focused on using renewable raw materials [3]. Biomass is an ideal alternative to fossil resources, being triglycerides, lignocellulose and saccharides the main classes of feedstock that can be used for the production of greener biofuels and chemicals [3,4]. The conversion of these raw materials into valuable products is usually carried out by subsequent transformations of several biomass platform chemicals, such as saccharides (glucose and xylose), polyols (sorbitol, xylitol and glycerol), furans (furfural and 5-hydroxymethylfurfural) and organic acids (succinic, levulinic and lactic acids) [5,6]. The great advantage of biomass platform chemicals is that they are functionalized compounds, allowing its further conversion into more valuable chemicals through a lower number of steps when compared to compounds derived from fossil sources, which are essentially unfunctionalized alkanes [4]. The fine chemical and pharmaceutical industries produce different complex molecules, which usually require large amounts of solvent for their synthesis. In addition, extractions and purifications also depend on solvents, in which large excesses are needed to achieve suitable product purity. Nonetheless, solvents are essential for many chemical processes and have great effects in organic chemical reactions [7][8][9]. The exceeding consumption of non-renewable and toxic solvents poses risks to both human health and the environment. Therefore, biomassderived solvents are good candidates to overcome the aforementioned issues [10,11].
Biomass-derived solvents starting from platform chemicals could be obtained by several processes, such as fermentation, hydrolysis, reduction, etherification or esterification. Examples of these compounds are furfural, 5-hydroxymethylfurfural, levulinic acid and alkyl levulinates, γ-valerolactone, 1,4-pentanediol, 2methyltetrahydrofuran, as well as, glycerol and its derivatives [12]. Besides coming from renewable sources, a series of characteristics are required so that a compound can be called green. The concept of green solvents is strongly related to the principles of green chemistry that aims to minimize or eliminate the use and generation of hazardous substances, while reducing energy consumption and moving toward cleaner and sustainable production from renewable sources [1,10]. It is undeniable that a green solvent must reduce health and environmental damages; however, there are several evaluation criteria to be considered, including non-toxicity, low volatility, high boiling point, biodegradability, easily recycle and ability to dissolve a wide range of compounds [12]. In this review, the chemical transformation systems and the solvents properties are critically considered in order to call the compounds as green solvents.
Environmental and economic aspects The development of new biomass-derived solvent can be assessed by a guide created by Jin et al. [9]. The authors proposed a 10-step method, which are (1) identify the conventional solvent to be replaced; (2) select potential replacement candidates; (3) in silico modelling properties; (4) identify a green synthetic route; (5) optimize solvent production; (6) test physical properties; (7) assess performance and toxicology; (8) techno-economic assessment; (9) evaluate solvent greenness and (10) life cycle assessment. This framework is very helpful to qualify the new solvent candidate with the required properties of a green solvent and to potentially meet any required regulations. It is important to note that intermediate steps involve the careful optimization of synthetic pathways, making use of green chemistry principles, consideration of solvent toxicological testing, and final steps of more time-consuming life cycle assessment (LCA) studies. The LCA uses data acquired from secondary sources, such as databases, literature references and simulations, also, it considers the entire life of the products and raw materials [10]. The chemical industry is claimed to consider renewable sources and the valorization of wastes as the primary source of sustainable solvents toward reducing environmental impact [13]. However, some developed methods can sometimes be at odds with this goal, because researchers rarely consider the economically and environmental effects of scale up the manufacturing processes. There are a number of efficiency considerations that rule the viability of using a solvent in a given application, including both application-specific technical factors (yield, for example) and process-specific economic factors (solvent cost, for example) [14,15], along with more general considerations such as availability, scale, and disposal methods, as well as corrosion, thermal stability and toxicity concerns [10]. The choice of the compound to replace a "non-green" solvent should involve, besides the aforementioned factors, its properties in use. In some cases, the "green" solvent could lead to a lower yield of the product and create a large amount of waste that require more energy in recovery. That is why, one strategy proposed is to recovery and reuse the solvent so the process costs are minimized. Hence, the solvent stability needs to be considered in the selection of a substance that it is capable of supporting several consecutive reactions. In addition to environmental aspects, the sustainability of a process must involve economic aspects, since the cost of the implementation of a new technology can harms the adoption of the process. The requirement to change equipment is an issue for industries that aimed to adopt new greener technologies. Also, if it is a pharmaceutical industry the process should be able to attend to the regulation demands for final product quality [11]. The wide variety of biomass-derived solvents should be economically sustainable, supplied at competitive prices and industry-scale volumes. For these reasons, economic considerations can help to shape technical aspects of solvent production in order to make the chemical processes more sustainable. In addition, the solvent must have dual roles, as a reagent, lead to higher quality products, reduce the number of synthetic steps, reduce byproduct formation and improve product separation [11]. Therefore, the evolution from a traditional linear economy to a circular economy is necessary for sustainable production in all future chemical processes to advance. The circular economy is related with the recovery and regeneration of resources during process design, which consider the application of the biorefinery concept, also, recovery and reuse minimizes impact on ecology and health. Thus, these factors strongly guide the technical and economic decisions in the development of engineering projects and in the improvement of existing plants [10,16]. Glycerol Glycerol or 1,2,3-propanotriol ( Fig. 1) is the simplest of the triols, which was discovered in the year 1783 by the chemist Carl Wilhelm Scheele through his experiments that reacted naturally occurring oils with alkaline materials. From that date, this compound has been used in several applications, notably Alfred Nobel's discovery of nitroglycerin production and its adsorption in diatomite, a compound widely used as an explosive and popularly known as dynamite [17,18]. The physical and chemical features of this triol are that it is a sweet-tasting hygroscopic liquid that can form azeotropic mixtures and, when pure, it is odorless, colorless, viscous at room temperature and it has a high boiling point (290 °C) at atmospheric pressure. Still, glycerol has useful properties to be used as a solvent due to high miscibility in water and in short chain alcohols, but as expected its high polarity may have limited application since it is insoluble in hydrocarbons, halogenated solvents and aliphatic fatty alcohols [18][19][20][21]. The large applicability of glycerol as solvent drives the demand for this compound, which can be obtained naturally from oils and fats in amounts in the range of 8-14% w w -1 . In the case of its industrial production, the most well-known route is from a fossil propylene source. However, other methods are also used in its synthesis, such as sugar fermentation, high-pressure hydrolysis of fat triglycerides, saponification or transesterification of triglycerides [17,20]. Among the glycerol obtaining pathways aforementioned, Fig. 2 shows the scheme of transesterification of triglycerides, common in biodiesel industry, that has increased its production over the years. In 1995, the amount of glycerol obtained by the oil transesterification process for biodiesel production was responsible for abandoning of the traditional production routes of this compound, such as from polypropylene. This is because glycerol production increased per year, which resulted in a decrease in its prices. Besides that, its use can be part of the circular economy issue through its in natura use or obtaining other products with higher added value [22]. In the case of glycerol without modification, some studies have shown that it can be efficiently used as solvent to accelerate some synthesis processes, because in addition to being biodegradable and highly hydrophilic, glycerol is widely available, inexpensive and do not require special manipulation and storage. It is also considered a green solvent that has low vapor pressure, polarity similar to dimethylsulfoxide (DMSO) and dimethylformamide (DMF) and dielectric constant of 42.5 (at 25 °C), which is compatible with most inorganic, acids, bases, organic compounds and enzymes that are poorly soluble in water [23][24][25]. In contrast, the purification of compound is easiest by simple extraction because different hydrophobic solvents, such as ethers and hydrocarbons, which are immiscible in glycerol [23][24][25][26]. Since glycerol forms crystals at <17.8 °C, its application as solvent is limited in low temperatures. Thus, depending on the reaction temperature, the choice of glycerol as a solvent may not be suitable, because at temperatures below 60 °C, this liquid is viscous and may hinder the phase transfer in the system [27]. Between of advantages, this green solvent could be used at high reaction temperatures without enhances the system pressure. Moreover, the products separation process can be feasible by distillation due to the high boiling point of glycerol or by liquid extraction when hydrophobic compounds are applied in the reaction. Due to its high versatility, glycerol is used in over 2000 industrial applications, the most important being in the pharmaceutical sector as an additive for the manufacture of personal care products (e.g. toothpaste) and cosmetics. Furthermore, because of its safety LD 50 (oral rat) = 12600 mg kg -1 , it is a good choice as solvent for application in product carefully controlled. In addition, it can be widely used as a food sweetening additive, as a tobacco wetting agent, in the production of fuel and soap additives, pulp and paper manufactures and through its functionalization can be obtained alkyl esters, polyethers and resins [17,25]. After Wolfson's group reported the first work that efficiently explored reactions like Suzuki, Heck and hydrogenation using glycerol as solvent in 2006, many researches are involved in the use of this compound [24]. Jérome et al. [28] applied glycerol as a solvent in aza-Michael reactions between amines and β-unsaturated compounds to obtain β-amino acids. The authors observed that glycerol can act at the system interface by accelerating the rate of reaction possibly by transition state solvation and tolerating more hydrophobic compounds in relation to reactions that employ water as a solvent. For instance, the reaction of p-anisidine with butyl acrylate at 100 °C during 20 h showed a yield increase from 5% to 82% by replacing water by glycerol as solvent. Radatz et al. [29] carried out condensation reaction of o-phenylenediamine with several ketones and aldehydes to produce benzodiazepines and benzimidazoles using glycerol as solvent. Initially, the reactions performed in the absence of glycerol were unsuccessful and using glycerol at room temperature achieved only 20% yield in 24 h. The problem not mentioned by authors was the poor mass transfer of glycerol in room temperature that probably limited the contact between reactants. However, by employing temperatures of 60 °C and 90 °C, in which this problem is negligible, yields of up to 97% of the isolated product was attained in 4 h. In addition, product recovery was performed with simple liquid-liquid extraction using ethyl acetate/hexane and glycerol was only vacuum dried and reused for 4 cycles without significant yield losses. Several other studies demonstrate the application of glycerol as a viable solvent in synthesis processes [24,30]. A reaction aqueous medium containing glycerol was used by Meyer et al. [31] for metallaelectro-catalyzed C-H activation. Usually, this reaction is made by applying costly and toxic oxidants, besides that fossil-derived compounds as showed in C-H/N-H alkyne annulation [32]. So, in the Meyer et al. [31], the authors performed the same reaction using glycerol that has a better conductivity in relation to other usual solvents such as of this reaction. In Fig. 3 it is demonstrated the scheme of this process, where glycerol/H2O proved to be a good solvent system in organic electrochemistry achieving 96% yield in the best condition. That was the first work reporting electro-enabled C-H activation using biomass-derived glycerol. Deep eutectic solvents (DES) using glycerol were reported by Li et al. [33] that studied the cellulose extraction from Okara by the use of three different complexing agents, choline salts in oxalic acid, in glycerol and in urea, via one-pot system. In this study, the authors investigated the structure and the properties of cellulose nanofibers obtained. Under the same homogenization conditions on a solid to liquid ratio of 1:20, at 100 °C for 30-120 min, with a stirring speed of 40 r min -1 , the deep eutectic solvents prepared with glycerol and Choline Chloride was more favorable for homogenization treatment. Therefore, it was suggested that this route has potential to replace the traditional pretreatment process, mainly because it is fast and improves the pretreatment efficiency, concomitant with the decrease of environmental pollution.
Due to the properties of glycerol, which evaporates without decomposition [34], reactions have been carried out in microwaves and ultrasound to convert it into other compounds or using directly it as a solvent [23]. Balaraman and Rathnasamy [35] performed the extraction of quail eggs immunoglobulins by chromatography with specific deep eutectic solvents. Such solvents were synthesized using quaternary ammonium salts (bond hydrogen acceptor) and glycerol (bond hydrogen donor). After this, an ultrasound assisted liquid-liquid microextraction was performed and higher density solvents presented the highest extraction capacity. The authors achieved through the response surface graph built that 65 mg mL -1 (85% yield) could be recovered in the optimal condition of 12 min three-dimensional ultrasound at 35 °C with DES concentration of 85% (v v -1 ) and sample load of 75 mg mL -1 . Despite using glycerol for the synthesis of DES, the work involved the use of irritable and toxic compounds (benzyltrimethylammonium chloride), in addition to do not mention the solvent reuse, which is important in the green chemistry concept. Kulturba et al. [36] performed the DES synthesis using glycerol or ethylene glycol as hydrogen bond donor and citric acid as hydrogen bond acceptor in the molar ratio of 4:1, respectively. In this case, 9.36 mg of anthocyanin recovery from Hibiscus sabdariffa was achieved in 180 s and 90 °C when using DES from ethylene glycol in relation to only 5.44 mg of anthocyanin total founded to DES from glycerol. This could be attributed to the lower viscosity of ethylene glycol in comparison to glycerol, which probably allowed for better sample diffusion. In this context, more investigations are need to found better conditions to use DES from glycerol to extraction processes in microwave, because it is a solution eco-friendlier. Another great applicability of glycerol was recently related to lignin fractionation. Usually, many solvents have been used to fractionate lignin such as ethyl ether, methanol, methane chloride, dichloromethane, n-butyl alcohol and other. However, the extraction process with several organic solvents is costly and harmful to the environment, despite being efficient. Glycerol was used in lignin fractionation by Liu et al. [37]. In the first step, corn straw was treated by steam explosion, and the residue went through a process of enzymatic hydrolysis. The enzymatic hydrolysis residue (EHR) was exposed to the alkaline extraction, followed by separation of soluble (EHL) and insoluble parts. After this, 20.00 g glycerol-ethanol solution (3:1, w/w) was added to 1,.00 g de EHL, magnetic stirrer at ambient temperature until dissolution, centrifuged to move the insoluble lignin and obtained the sequential fractions. In this work was possible recovered lignin with different molecular weight and while decrease in its heterogeneity. The authors carried out the extraction of lignin as showed. Moreover, the mixture of 3:1 glycerol:ethanol (v v -1 ) was recycled and reused, but the color of these compounds changed during the process due to contaminants from the extraction. Although the literature points out that glycerol can be used in the presence of acids and bases, the presence of three hydroxyls in its structure ( Fig. 1) may be important reactivity points, in addition to their coordinating properties, which can disable organometallic compounds limiting its application as inert substance [25,38]. Therefore, several modifications of glycerol have been studied in order to increase the applicability, as shown in the following section.

Modification of Glycerol
Glycerol applications in other areas include its use as a wetting agent, because its high viscosity gives greater consistency to the products and it is rapid heat transfer medium, also; it can be lubricant as it remains fluid at temperatures above 17.8 °C and is relatively resistant to oxidation. However, its properties may also restrict its application, since this triol is highly viscous at room temperature and has low solubility in hydrophobic compounds and gases. On the other hand, microwaves or ultrasound may be employed in an attempt to mitigate these issues, because in relation to conventional heating (heating plates, blankets) the use of microwaves enables higher heating rates and greater energy absorption by reactants or solvents, which results in improved mass transfer in the system. Thus, when using heating sources like microwaves, no temperature gradients or localized overheating occur and, therefore, there are advantages in the chemical modification of compounds, for example in the glycerol functionalization [23,[39][40][41]. The chemical modification of glycerol aiming its absorption by the industry has been increasing and this compound is denominated a key raw material of many industrial chemical processes. Therefore, as can be seen in the Fig. 4, glycerol can provide a wide range of products, which are resulting of dehydration, oxidation, esterification, acetylation and etherification reactions [17,23,25,42,43]. As aforementioned, glycerol can be transformed into many molecules due to the possibility of functionalization in its three hydroxyls, which are important reactivity points. Among such processes, the transformation of glycerol into ethers is being studied by some authors [17,18,42,[45][46][47][48][49][50][51]. Glycerol ethers have applications as fuel additives, pharmaceutical intermediates, agrochemicals, solvents and nonionic surfactants. Among these molecules, glyceryl monoethers obtained using C4-C12 aliphatic chain alcohols have been investigated as they can be used as alternatives for petroleum-derived surfactants, as antiseptic and antimicrobial agents, as precursors for polymers and as green solvents. In the case of their interesting features for application, the use of glycerol ethers as a solvent in chemical reactions have attracted great interest currently [26,[52][53][54]. The glycerol alkyl ethers can be naturally occurring with an alkyl or alkenyl chain linked by an ether bound to the glycerol at the C1 position and the remaining two hydroxyls can be free or acetylated. The compounds most known of this class are bathylic, chimylic and selaquilic alcohol, which are found in lipid membranes of marine animals (e.g. whales, starfish, squids and corals). These molecules receive several nomenclatures, such as glycerol ethers, lipid ethers, alkoxyglycerols, among others. They also can have substituents in only one hydroxyl position (monoalkyl ether of glycerol -MAEG), two (DAEG) or in all of their hydroxyls (TAEG) [44,[55][56][57]. The etherification of glycerol to produce its alkyl derivative ethers can be performed via Williamson synthesis in a basic medium, via telomerization with butadiene derivatives, via reduction of compounds, in addition to condensation in an acid medium [53]. When dealing with routes that have a more environmentally friendly appeal, one of the most acceptable is acid condensation because it can use glycerol directly. In another way, other routes could reach a best yield in lower reaction time with higher selectivity, while using toxic and dangerous reactants, for instance in the case of the epichlorohydrin route [58], making it difficult to characterize them as green solvents. Some authors have studied the synthesis of glycerol ethers through condensation for several purposes and achieved different selectivities in glycerol mono-, di-and triether by using homogeneous or heterogeneous phase catalysts, as well as the most varied sizes of hydrocarbon chains of alcohols or alkenes for the glycerol functionalization [17,44,55,59,60]. In addition, the alkyl chains of the compounds used in the etherification can be differentiated into linear or branched, long or short, which modifies the characteristics of the resulting ethers that became difficult to predict all the properties of these substances as a solvent, given the infinite possibilities of groups that can be used for the synthesis of the glycerol alkyl ethers. Table 1 shows some studies involving the synthesis of glycerol ethers that have been published. Among the methods of glycerol alkyl ethers synthesis, the Williamson's route, the reaction with alkenes or the dehydration of alcohols are some frequently used. However, from the green chemistry point of view, a conventional synthesis (via Williamson ethers synthesis) of alkyl glycerol ethers is disliked for generating waste and employing toxic reagents. In this way, despite of the need to improve some issues, the condensation between different alcohols (that can be obtained from renewable sources) could be a better route to produce these compounds. Indeed, glycerol properties can result in low conversions due to its high viscosity and hydrophilicity, or the reproducibility of the conversions are not appropriate when using very hydrophobic organic compounds. In addition, other problems hinder the synthesis of EAG by condensation, such as the low selectivity towards the product of interest, due to the very similar reactivity of glycerol's three hydroxyls, affording the formation of mono, di and triethers mixtures in different proportions. Glycerol alkyl-ethers are stable in the presence of water and chemically more inert than other glycerol-derived solvents, what make them attractive in the green chemistry concept. So, from the choice of oxygen-linked groups, characteristics with high boiling point, low flash point and volatility can be properties of this solvent. On the other hand, although they have a lower boiling point, for example, greater polarity can be observed in some glycerol alkyl ethers, such as 2,3-trimethoxypropane, 1,3-dimethoxy-2-propanol and 2,3-dimethoxypropan -1-ol, being able to replace polar solvents as dimethoxyethane in reactions [53,58,61,62]. Also, in the case of tunable properties, when trifluoroethanol is used as reactant in the synthesis of ethers, the obtained solvent 1,3-bis (2,2,2-trifluoroethoxy) propan-2-ol can present similar properties to some ionic liquids [63]. The toxicological issues of the glycerol alkyl ethers were evaluated by some studies and, in many cases, the toxicity of these compounds was almost negligible. However, by increasing the size of the hydrophobic part and the number of substituents on the compound, an increase in its toxicity can be observed by some biomodels, such as Dapnhia magna, Aliivibrio fischeri, Chlamydomonas reinhardtii and Danio rerio [12,53]. Therefore, several questions must be taken into account for the synthesis of glycerol alkyl ethers in order to obtain a green solvent and their properties deserve more attention as well as the study of the application of them in reaction medium. Butanol [52] Hybrid composites from Aquivionsilica The Aquivion-silica composite exhibited a high catalytic activity (91% n-butanol conversion) producing 45% MAEG and 6% dibutyl ether.
Dodecanol [26] DBSA, DMSO, sulfolane and 1,4dioxane solvents The use of DBSA resulted in the formation of 25% of MAEG in the conditions of 3 h of reaction, 20 mol% of catalyst, 160 °C in a water removal system.
Ethanol [48] A-15 The optimized conditions that produced 56% of MAEG were: 238 °C, MR of ethanol: glycerol 16:1 and 0.61 g of catalyst. The temperature, pressure and the amount of catalyst had a statistically significant effect on the MAEG response.
Octanol [66] USY-550, USY-550-L, USY-650, USY-650-L-2, Hβ e HZSM-5 Hydrophobic zeolites were active in the etherification of glycols and alcohols. The conversion of glycols was closely related to the hydrophobicity index of the material and the structure of the zeolite was crucial for the reaction. Levulinic acid Levulinic acid (LA) (Fig. 5), also known as 4-oxopentanoic acid or as 3-acetylpropionic acid is a γ-keto acid and non-volatile that has a boiling point of 245 °C. This acid presents great solubility in water as well as in polar solvents, such as diethyl ether and ethanol. The presence of a keto group in its structure make the LA has higher dissociation constants than a common saturated acid, with a pKa of 4.59 (in water at 25 °C). Furthermore, the two functional groups of this compound, carbonyl and carboxyl, afford to it a large variety of application in synthesis, as a consequence, LA together with glycerol were considered by the US Department of Energy's one of the "Top 12" most important bio-based chemicals in the world [68][69][70][71]. For instance, LA has potential to be used in the synthesis of several products that nowadays is provided mostly by the petrochemical industry, such as succinic acid, resins, polymers, herbicides, pharmaceuticals, flavoring agents, plasticizers, antifreeze agents, biofuels/oxygenated fuel additives and solvents [72]. Currently, the widely known methods of LA production involve the complex conversion of maleic anhydride, the hydrolysis of furfuryl alcohol or the oxidation of ketones with ozone. However, these processes are based on petrochemical and expensive starting feedstocks, which results in the relatively high market price of LA (~US$10 kg -1 ). In addition, large amounts of side products and residues are produced in these vias. An alternative to reduce the cost of LA production is the conversion of sugars from biomass through acid-catalyzed dehydration and hydrolysis (Fig. 6) [73,74]. During the conversion of biomass, in acid medium, the hexose sugars are converted in LA and formic acid (FA), therefore, the maximum theoretical yield possible of LA is 64.5 wt %. In spite of that, the FA can be used in the production of formaldehyde, rubber, plasticizers, pharmaceuticals and textiles or it can be hydrogen source for other reactions, which is important to the valorization of the whole biomass processing [73,77]. Other factor that affects the LA yield is the instability under acid conditions of the intermediate formed in the biomass conversion, HO O O 5-hydroxymethylfurfural (HMF), which can be transformed in other compounds, such as 2,5furandicarbaldehyde and 2,5-furandicarboxylic acid (FDCA). Also, several reports found side reactions harming the obtaining of LA that led to the formation of black insoluble-materials, called humins. Therefore, the reaction conditions should be optimized in order to increase LA yields [73,74]. A useful tool to improve LA yield and selectivity is the use of solvents in the reaction medium during the conversion of the sugars. The ideal solvent to the reaction mixture would be water, however, side reactions are favored in this medium, such as polymerization and humins formation. Thus, commonly high boiling point solvents are employed, for example, DMSO, N,N-dimethylformide (DMF) and sulfolane. These compounds can reduce side reactions, but in contrast, they add to the process energy costs for their separation from the products and to overcome this issue, the reuse of the solvent should be performed [78,79]. For instance, Wang et al. [79]reported the use of sulfolane:water mixture as solvent in the conversion of cellulose into LA and recycled three time the organic solvent. In this study, the mass ratio used was of 9:1 (sulfolane:water) and it was achieved 72.5% of LA yield LA (in relation to the total cellulose), even after three solvent recycles. Although the production cost could be reduced, sulfolane is a toxic non-renewable solvent that should be avoided in environmentally friendly processes. The use of greener solvents was investigated by other authors, such as Han et al. [80] that converted cellulose into LA using a solvent derived from biomass, γ-valerolactone (GVL), and the best result obtained was 35.6% of LA yield, at 185 °C in 120 min of reaction by using a lignin-based catalyst. The efficiency of the reaction was related to the ability of GVL to enhance the adsorption of cellulose in the solid catalyst, so the interaction of the substrate with acid sites was more effective. In contrast, at the end of the reaction it was found several byproducts such as FA, glucose, furfural and fructose, which indicates that the catalyst promoted the isomerization of glucose into fructose and the selectivity to LA decreases. Other biomass-derived solvent that was already employed to the conversion of glucose into LA is 2methyltetrahydrofuran (2-MTHF). Jiang et al. [78] investigated the transformation of glucose into LA in water/2-MTHF biphasic solvent system under microwave heating. The selectivity of LA achieved was of 88% with 100% glucose conversion, at 200 ºC by 60 min (pH = 1) using FeCl3 as catalyst. The conversion of Poplar WT 717 in the same conditions gave 53% of LA (based on glucose content in biomass). Despite the less LA production in the biomass conversion, it was demonstrated that 2-MTHF avoided side reaction of C5 and C6 sugar to humins and increases the yields of HMF and LA in relation to the use of water only. On the other hand, 2-MTHF was decomposed into (Z)-3-penten-1-ol and 1,4-pentanediol due to acid and high temperatures, wherefore the solvent could not be reused in order to decrease the process costs. By contrast, ionic liquids (ILs), in spite of being expensive, can be more stable at the reaction conditions, so the literature have reported their recycling in some cases of the synthesis of LA [81,82]. In the work of Kumar et al. [83], LA selectivity of 56% was attained through glucose conversion, that involved the use of the 1-(4-sulfonic acid)butyl-3-methylimidazolium chloride ionic liquid (IL) (IL-SO3H) and NiSO4 as catalyst. Besides that, the process cost was diminished, since the IL was recycled three times remaining the LA yields. In another study [82], higher LA selectivity (66%) was achieved from the reaction of cellulose using the IL [C3SO3Hmim][HSO4] under microwave heating. The selectivity was even improved to 86% by the reduction of the mass proportion of 0.55:1:6 to 0.02:1:6 (cellulose: IL: H2O) and the IL was reused for five cycles by its recovery using methylisobutylketone (MIBK). Surprisingly, the authors found an increase of LA yield from 58% (first cycle) to 66% (fifth cycle), which was attributed to the residual oligomers of cellulose dissolved in the IL. Therefore, the spread of this system to use of lignocellulosic biomass feedstock would be hampered, whereas the high affinity of it toward ILs would make the economical reuse of IL unfeasible as well as the contamination of the products with ILs of unknown toxicity [81]. The disadvantages aforementioned related to the high costs of LA production through different methods could be minimized by the direct use of lignocellulosic biomass, which also contributes to the sustainability of the process by the use of renewable materials [81,84]. For instance, the hydrolysis of cellulosic food waste under microwave irradiation was studied by Chen et al. [84], whom achieved 17% of LA yield in only 5 min of reaction in a medium containing water and Amberlyst-36. When DMSO was added to the system, 40% of LA yield was obtained. The authors attributed the efficiency of the reaction to the high content of amorphous cellulose in the raw material and its solubility in DMSO, which affords good interaction between the substrate and the catalyst. The major disadvantage of this system was the coverage of the solid catalyst with the humins formed in the medium that hindered the improvement of LA yields. Moreover, the results attained cannot be reproducible, since the material used is variable due to be an urban waste. Therefore, the use of other biomass sources that could consist by less amorphous cellulose could give lesser LA yields due to biomass recalcitrance [81,85].
In the study of Kumar et al. [85], it was used rice straw as substrate, which is an agriculture residue of known high recalcitrant nature. The approach used to convert the lignocellulosic material into LA was a co-solvent system consisting by dichloromethane (DCM) and HCl that allowed the continuous extraction of furans from reaction medium and made it possible the achievement of 15% of LA yield (57% of theoretical yield), at 180 °C and 3 h. Contrasting the idea of using a residue (rice straw) to avoid environmental issues, the authors applied DCM and HCl, which are hazardous and require recovery and treatment, making the greenness of the overall process dubious. In addition, energy saving was not performed since a high temperature was used. The high temperature (180 °C) was crucial to the LA accumulation, but it reduced the partition coefficient of the products, thus, the conversion of LA required large amounts of solvent, in the volume ratio of 5:1 (DCM:Water). However, fortunately, the system allowed the recovery of 90% of the solvent and its recycling in 5 runs, maintaining the LA yield. Other authors have reported the LA obtaining by the conversion of acid-pretreated eucalyptus wood. The pretreated material was composed by approximately 63.8% of glucans and 34.6% of lignin. After this stage, LA was obtained by the reaction of the substrate with H2SO4 0.2 mol L -1 in repeated batch reactions, at 170 °C by 5 h. The study showed that, for the pretreated material, there was need 3 reaction cycles to achieve about 70% of LA yield and for the untreated material the same yield was obtained with 6 batches of reaction [86]. In both investigations [85,86], the issue related to LA recovery was related to the temperature used that led to the formation of humins, which have affinity to water and dragged LA to the aqueous phase. Also, the residual solids (humins and lignin) hindered LA synthesis, not only by physical interaction with it, but also, by reactions between those compounds that were not very well elucidated. Therefore, the production of humins should be avoided to facilitate the LA processing. When using biomass as substrate for LA synthesis, the pentoses present in the material are easily transformed in humins under the conventional harsh reaction conditions. Although, the reaction conditions used to convert hexoses into LA can transform pentoses into the furfural, which can result in LA too and this procedure involves more reaction steps. To overcome these questions, an alternative could be the pentose extraction prior to the LA production [74]. Runge and Zhang [74] performed a two-stage synthesis of LA from hybrid poplar, which is a material with 14-18% of pentosans. The first step of the process consisted of an acid extraction (1% H2SO4), under 160 °C, 60 min, and a 6:1 liquor to wood mass ratio, which removed 85% of pentosans and kept a solid with 92% of hexosans. The second step was the LA production by using the solid material concentrated in hexosans in harsh conditions (190 °C, 5% H2SO4, liquor to wood ratio 10:1 and 50 min), which provided 66% of LA yield in a 2L reactor. When using the material non-extracted, under same conditions, the LA yield produced was about 50%. In addition, the furfural obtained from the non-extracted material was very low (<0.1% of theoretical), which confirms that optimal conditions to LA production are too harsh to produce furfural that is further transformed into LA. In short, the process described is quite interesting, however, it aggregates one more stage on biomass conversion that requires time and energy. Additionally, the use of H2SO4, a non-recyclable catalyst, results in costs of waste treatment. Therefore, although the common method of LA production by homogeneous acid catalysis usually affords yields up to 70% (in harsh conditions), there are unfeasible steps of separation and treatment of the liquid catalyst, since its distillation is expensive and promotes reactor corrosion. Besides that, side reactions involving humins formation could be also due to the non-selective nature of the homogeneous catalysts [87]. Based on that, it is proposed over the literature the use of heterogeneous catalysts that can be more selective, since there are of possibility of tune its properties, like acidity, porosity and specific area, and they are easily recovered and potentially reused [88][89][90]. Together with the advantages already mentioned, heterogeneous catalysts reduce the problem of reactor corrosion. However, the use of this type of catalysts, that are generally solids, implies in limited mass transfer, which is worsen depending on the substrate used to LA production. In this context, the conversion of the liquid hydrolysate from Quercus mongólica treated with H2SO4 towards LA synthesis was investigated by Jeong et al. [88], whom showed the efficient use of the commercial solid zeolite Y modified with NaOH as catalyst. The LA production was assigned to the mesopores and strong Lewis acid sites of modified-zeolite. There was demonstrated that the catalyst preferable converts C5 than C6 sugars and it would be suitable for a biorefinery concept, which consists of a multi-step treatment of biomass and uses separated glucose and xylose. However, the maximum LA yield could not be increased, since temperatures higher than 190 °C afforded humin production. Also, a large portion of the C5 sugar remained as furfural and was not transformed in furfuryl alcohol and, then, LA. That was attributed to the insufficient number of strong acid sites in the catalyst and to the glucose or xylose molecules hindering the access of reactant to the active sites. Moreover, no catalyst reuse was investigated, which do not justify the use of a commercial catalyst in the process. Another example of the use of liquid sugar source and a solid catalyst to the production of LA was reported by Kang and Yu [89], whom used sugar beet molasses as substrate and acidic cation exchange resin (Amberlyst-36) as catalyst. The yield of LA achieved was of 53.2% and it was improved to 78% with the removal of non-sugar components from the molasses. In despite of the catalyst be active in a complex substrate (not only composed by C5 or C6 sugars), its reuse was not efficient, since the regeneration required harsh conditions of sulfonation, which resulted in a gradual activity loss. Solid catalysts have been successful active in transformation of liquid sugar sources, however, many of the challenges associated with them are related to hydrolysis of solid feedstocks, such as biomass and long-chain cellulose. Most of the studies that reports the use solid biomass conversion have showed yields ranging from 30-60%mol of LA in detriment of the use of organic solvents. Furthermore, generally the catalysts used are commercial or of expensive method of production, such as zeolites, ionic exchange resins, zirconium dioxide and Gallium salt of molybdophosphoric acid [83,90,91]. Indeed, it is increasing the development of solid catalysts based on renewable sources for the greener obtaining of LA. Li et al. [92] produced magnetic ferric oxide/SO4 2biomass-based solid acid from corn straw carbonized and sulfonate (MIO/SO4 2− B-BSACs). Such catalyst was used in the pyrolysis of corn straw and at optimized conditions (250 °C, 70 min) the yield of LA was 23%. The conversion was attributed to the similar structures of corn straw substrate and that of the catalyst that afforded good interaction between them. Meanwhile, the authors did not mention the catalyst lixiviation that could happen, then the LA yield could be not only from substrate but also from MIO/SO4 2− B-BSACs structure. The magnetic particles of the catalyst enable its separation from reaction medium, which avoids issues related to outgoing in product purification and residues treatment, but still, the use of 250 °C as reaction temperature is an expensive energy input. In lower temperature (185 °C) Han et al. [80] attained higher yield of LA (36%) by using a solid catalyst base on alkaline lignin carbonized and treated with a ferrous sulfide solution. The catalyst was used in the synthesis of LA from microcrystalline cellulose in a system of GVL:water as solvent. Those green solvents were used, but the authors did not study the possible recovery and recycling of them, which are expense and environmental effects must be taken into account. In 2011, Lomba and co-workers studied several thermophysical properties of five compounds classified as green solvents derived from biomass (furfural, furfuryl alcohol, levulinic acid, ethyl levulinate and butyl levulinate). Remarkably, levulinic acid showed low vapor pressure, so it can be considered a suitable candidate for substitution of solvents that are volatile organic compounds (VOCs) and interesting for industrial applications, since its recovery could be facilitated [93]. In relation to the use of LA as solvent, there are reports of it in mixtures to solubilize diesel and in the composition of some degreasing agents. Also, LA and its esters are solvents in polymers, textiles and coatings [93]. Indeed, given LA reactivity, its use to produce other molecules that are widely used as solvents have increasing over the years. Well known examples of levulinic acid derivative that can be used as solvent are the alkyl levulinates, which will be better explored in the next section of this article.
Alkyl levulinates Alkyl levulinates (AL) (Fig. 7), particularly those with hydrocarbon chain ranging from C1 to C6, are biomassderivatives that have valuable properties, such as consistent permittivity, high lubricity, high boiling point, suitable flash point and low vapor pressure. These compounds can be used as building blocks in organic synthesis, as fragrance, as flavoring agents, as well as fuel additives and solvents. Along the properties mentioned, some features make AL more interesting to the use as green solvents, such as adequate cytotoxicity and mutagenicity. Besides that, the possibility of varying the alkyl chain can afford properties tuning to specific solvency requirement [94][95][96]. The common routes used to the AL obtaining are direct alcoholysis of carbohydrate biomass (Fig. 8) or conversion of biomass-derived chemical, such as LA, furfuryl alcohol, HMF or furfural. The first report of AL synthesis is over 150 years old, when the authors used LA for the synthesis of methyl and propyl esters, using HCl as catalyst. Ever since, the direct esterification of levulinic acid over acidic catalysts or enzymes (lipases) is widely studied and high selectivity values can be attained (up to 99%) [94,95]. Authors, adapted from [72].
A great number of studies have reported the production of AL using homogeneous acid catalysts, though some drawbacks, mainly with the rise of the concerns related to environmental issues, have encouraged the use of greener catalyst [69]. For instance, Di et al. [97] showed the use of a biocatalyst (lipase), which is used at mild reaction conditions, it is non-toxic, non-corrosive and, in this case, the use of the immobilized enzyme (Novozym 435) gives the advantages of its recovery and reuse. Then, the authors carried out the reaction between LA and methanol in the presence of some solvents, including the  43.2%, which limits the dilution of the high cost of enzymatic process. Additionally, the process lasted 24 h, being too long in relation to the conventional homogeneously-catalyzed (up to 6 h). The alternative use of chemical heterogeneous catalysts was reported, such as the commercial heteropolyacids, sulfated metal oxides, zeolite molecular sieves and hydrotalcite-like compounds [69], whose also present the disadvantage of high cost. Recently, Bosilj et al. [98] reported the use of a cheaper biomass-derived catalyst in the conversion of glucose into ethyl levulinate (EL). In this study, an acid-functionalized hydrothermal catalyst derived from glucose was produced by hydrothermal carbonization of glucose that was first treated with sodium borate (borax) to generate carbon nanoparticle size and, then, it was sulfonated with H2SO4. By using this material and ecofriendly solvents (ethanol, glycerol and GVL), the EL yields attained were as high as 25-37 mol%, depending on the solvent system used. In 6 h of reaction at 200 °C, the yields achieved were of 37, 35 and 25 mol% of EL, when using ethanol:glycerol, etanol:GVL and etanol by itself as a solvent, respectively. One inconvenience of using GVL is its probable dehydration to α-angelica lactone and hydration to levulinic acid, which could further form EL with ethanol, giving an EL yield that account not only to the original substrate used (glucose). On the other hand, the use of glycerol was advantageous, since it promoted rapidly conversion of glucose into 5-HMF and reduced humin deposition on the reactor wall. Yang et al. [99] prepared a magnetic carbonaceous solid acid (SMWP) catalyst from waste paper scraps via Feimpregnation, carbonization and sulfonation process, which could easily be recovered from reaction medium with external magnetic fields. The SMWP catalyzed the alcoholysis of furfuryl alcohol with n-butanol, reaching to 90.6% yield of n-butyllevulinate in 5 h at 120 °C, such catalytic activity was assigned to a strong acidity and the affinity towards furfuryl alcohol, due to the presence of the SO3H, COOH and phenolic OH groups on the catalyst surface. Moreover, the catalyst was reused for 7 times with a slightly loss of activity that was attributed to the adsorption of some oligomeric by-products on the surface of SMWP.
The results aforementioned are quite interesting from the synthetic point of view, however, it was reported the use of pure substrates (furfuryl alcohol and glucose) that are costly products. Therefore, the investigation of the direct synthesis of alkyl levulinates from lignocellulosic materials have increased, since the reaction can be performed in a one-step way, which reduces stages of products purification and the amounts of wastewater, besides that, this method could save time and be cost effective [69,94,95]. On the other hand, recalcitrance and insolubility of the raw biomass require drastic reaction conditions for the alcoholysis/hydrolysis of the structural carbohydrates, which hinders the obtaining of alkyl levulinates selectively with the yields obtained about 10-30% [100]. Therefore, some authors described the fractionation of the lignocellulosic material prior to its conversion into AL. For instance, Liang et al. [101] carried out the preparation of bifunctional solid-acid catalyst from the hydrothermal hydrolysate of dewaxed wood powder of eucalyptus globulus residue that was sulfonated and impregnated with Zr 4+ . The other portion of the biomass hydrolyzed had the lignin removed with DES and, then, it was used as substrate to the production of methyl levulinate (ML) and the highest yield reached was 38.7%. This study showed the improvement of the ML yields by using a pretreated lignocellulosic material. Nevertheless, the proposal is a multistage process, which can result in exceeding steps of recovery and residues treatment. The one-pot synthesis of methyl levulinates was carried out by Feng  The use of solid catalyst is very limited in the processes of direct biomass conversion, thus, ILs seems to be a good option, since they can increase the dissolution of lignocellulosic materials and act as catalysts in the alcoholysis of the substrates. Therefore, the recovery and the reuse of ILs should be optimized to the reduction of processes costs. Also, one-pot reaction is advantageous in relation to biomass fractionation, because it reduces process steps. Some authors have highlighted that direct obtaining of alkyl levulinates instead of levulinic acid is advantageous, because the levulinate esters, mainly those of short chain (C1-C4), present lower viscosity, acidity and boiling point. Additionally, ALs are non-corrosive and more stable under reaction conditions, which requires lower energy consumption in the stage of their separation from the medium. One issue in the one-pot reaction may be the low selectivity achieved of AL, but, in comparison to dehydration/hydration reaction to obtain biomass derivatives and further convert them into AL, reactions of biomass performed directly in alcohols can protect highly reactive intermediates and prevent unwanted polymerization reactions, improving the yields of the required products [69,100,102]. About the potential use of AL as solvents, The group of Sah and that of Schuette published some physical and chemical properties of methyl to hexyl levulinates, whose had boiling point in the range of 190−270 °C, they were all soluble in classical solvents (e.g., alcohols, ethers and chloroform), but insoluble in water (except methyl levulinate) [103,104]. Lomba et al. [93] measured the vapor pressures of alkyl levulinates and the values obtained at 100 °C were 5.96, 4.69, and 1.56 kPa for methyl, ethyl and butyl levulinate, respectively. These results show their easily handle in relation to chlorinated solvents that present vapor pressures more than 400 kPa at the same temperature. Other property evaluated was the partition coefficient in octanol:water (Log P) that showed the highest solubility in water of ML among the esters, whose water solubility decreased as the alkyl chain increased. It is important, because the solubility in water affects the potential for biodegradation of the compounds. In addition, Ferrer et al [105] presented in their patent of the use of alkyl levulinates for metallic surface degreasing some data showing that butyl, iso-butyl and pentyl levulinates present excellent environment, health and safety properties, such as not being cytotoxic neither mutagenic. Marcel et al. [96] performed the use of alkyl levulinates (methyl, ethyl and n-butyl) as solvents for the heterogeneously Pd-catalyzed Heck coupling, as far as it is known, for the first time. The reaction of Heck coupling between 2-iodoanisole and n-butylacrylate giving 2-methoxycinnamic acid butyl ester using alkyl levulinates as solvents showed yields ranging from 88 to 98%. The efficiency of the process was attributed to the catalysis occurring through the soluble Pd species, since, those species leach from the support and form complexes with reactants/solvents in solution and, after, Pd is redeposited on the support. Methyl and ethyl levulinates afforded 96% and 98% of product yields, while in n-butyl levulinate the yield was 88%, however, the non-solubility in water of n-butyl levulinate is an advantage for its recovery and treatment before environmental disposure. This solvent was then used in the reactions of other aryl halides, taking to conversions between 83 and 100%. These results were different from that usually observed for C-C couplings, when normally lower yields are attained for deactivated aryl halides, including iodides, which was attributed to a specific interaction between the solvent and the catalyst under the conditions, due to different degrees of reduction of Pd species. Additionally, n-butyl levulinate was recovered by simple distillation and its integrity was asserted by NMR. Afterwards, n-butyl levulinate was reused in one run that presented the same result of the first. More studies are need to fully explain the reactivity improvement in AL. Besides that, these compounds revealed potential that should be explored in other reaction. With respect to the reliable of the use of alkyl levulinates as green solvents it also requires further investigation towards their toxicity.
γ-Valerolactone (GVL) γ-Valerolactone (GVL) has attracted great attention in the last years especially due to its exceptional chemical and physical properties, such as low melting (-31 °C) and high boiling (207 °C) points, remarkably low vapor pressure even at higher temperatures (3.5 kPa at 80 °C), miscibility with water without azeotrope formation, no peroxides formation in the presence of air and high stability at neutral pH, making it a safe material for largescale use [1,106]. The chemical structure of GVL is shown in Fig. 9. Fig. 9. Chemical structure of γ-valerolactone (GVL). Source: Authors.
Owing to the aforementioned properties GVL is seen as a renewable solvent and as a potential biofuel additive. For instance, Strappaveccia et al. [107] identified GVL as a sustainable alternative to classic and toxic dipolar aprotic solvents, such as acetonitrile, dimethylformamide (DMF), or N,N-dimethylacetamide (DMA) in crosscoupling reactions. The authors observed a successful clean use of GVL in palladium-catalyzed Heck reaction of several small molecules (iodobenzene with acrylic esters and iodobenzene with styrenes). Horvath et al. [106] carried out a comparative evaluation of GVL and ethanol as fuel additives by mixing 10% (v v -1 ) GVL or ethanol with 90% (v v -1 ) 95-octane gasoline. The results showed that most of the data for GVL are comparable with ethanol and its lower vapor pressure leads to improved combustion at similar octane numbers. Besides that, GVL can also be precursor to high-grade liquid alkene transportation fuels and fine chemicals, including 2methyltetrahydrofuran (2-MTHF), 1,4 pentanediol, alkyl pentenoates and α-methylene-γ-valerolactone, which may find application in the production of biobased polymers [106,[108][109][110][111][112][113]. Nevertheless, the maximum economic and environmental benefits associated with utilization of GVL will only be achieved if a greener process is employed during its production [112]. The synthesis of GVL can follow two different pathways that entail in a sequence of dehydrogenation/hydrogenation of LA or vice versa (Fig. 10). The first pathway is a catalytic hydrogenation of LA to yield 4-hydroxypentanoic acid, followed by acid-catalyzed intramolecular esterification to form GVL. The second pathway is an endothermic acid-catalyzed dehydration of LA to angelica lactones and under H2 atmospheres, the preferred lactone GVL is thermodynamically obtained. Depending on the reaction conditions such as solvent, metal catalyst and presence of residual acid-impurities, one pathway may be favored [114]. Abdelrahman et al. [108] observed that Ru-catalyzed LA hydrogenation in the aqueous phase at temperatures below 150 °C, it forms exclusively via intramolecular esterification of 4-hydroxypentanoic acid. It is also important to highlight that two hydrogen sources are possible, molecular hydrogen (H2) or catalytic transfer hydrogenation using alcohols or formic acid [77,[115][116][117].  Heterogeneous catalysis may provide an efficient methodology for LA conversion allowing high reaction rates and high selectivity of the target products. The ability of a range of precious metals to catalyze LA hydrogenation has already been extensively explored, with good GVL yields (in the range of 80 to 99%) obtained from ruthenium-supported catalysts [108,116,117]. For instance, Al-Shaal et al. [112] examined the ability of Ru/C to catalyze LA hydrogenation using 12 bar H2 in methanol, ethanol and 1-butanol at 130 °C for 160 min. Among the alcohols screened, methanol facilitated the highest GVL yield, 84.4%, while ethanol and 1-butanol resulted in yields of 61.1% and 39.7%, respectively. When mixtures comprised of water (10% v v -1 ) and a given alcohol (90% v v -1 ), were utilized as solvents, ethanol/water and 1-butanol/water systems resulted in a substantial increase in GVL yields. The result can be explained by the lower H2 solubility in less polar alcohols [118] and the high capacity of water to dissolve H2 [119]. In fact, the authors obtained 86.2% yield of GVL using only water as a reaction solvent in the LA hydrogenation. This is very important because the commercial LA production via hydrolysis of lignocellulosic biomass is generally performed in water-contained medium that results in LA product streams containing water [110], which facilitates closer process integration and increasing efficiency, by removing the costly need to separate water from LA feeds prior to hydrogenation. However, the presence of water in LA product streams could create additional challenges with regard to catalyst stability. When alcohols are employed as reaction solvents, both LA and GVL can undergo esterification. Thus, Al-Shaal et al. [112] also showed the hydrogenation of alkyl-levulinates (methyl-levulinate and butyl-levulinate) in methanol, using identical conditions to those employed for LA (12 bar H2, at 130 °C for 160 min). GVL yields were similar to those obtained from LA, establishing that esters of LA can be readily hydrogenated to GVL as well. The conversion of lignocellulosic biomass to levulinic esters instead of LA could enables higher yields and easier product separation [87], suggesting an alternative route to GVL synthesis. To achieve an economical and sustainable production of GVL, the development of noble metal-free heterogeneous catalysts is an important research target, as noble metal-based catalysts are costly. An effective non-noble-metal catalyst was studied by Shimizu et al. [120]. The hydrogenation of LA to GVL under solvent-free conditions in 8 bar H2 at 140 °C for 5 h in the presence of 1 mol% of Ni-MoOx/C catalyst resulted in 97% yield. However, the catalyst recyclability was lower, achieving only 50% yield of GLV after the first cycle. Yi et al. [121] carried out a comparative study using Ni and Ru catalyst, both supported on HZSM-5, at 220 °C for 10 h in 30 bar H2 and using 1,4-dioxane as solvent. The catalytic experimental results showed that Ni/HZSM-5 catalyst achieved 93.1% yield of GVL, while Ru/HZSM-5 catalyst exhibited 85.7% yield of pentanoic esters (PE) and pentanoic acid (PA) under identical conditions (Fig. 11). Therefore, it was found that Ru/HZSM-5 increased the strong acidic sites and provided the ring opening of GVL, promoting the formation of PE and PA. In comparison, Ni/HZSM-5 catalyst was much more effective to produce GVL, once it showed relative lower acidic sites and negligible GVL ringopening ability.

Ru catalyst Acid condition
A greener process to convert LA to GVL was proposed by Liu et al. [122], which used a cobalt-based catalytic system in order to avoid the use of noble metal catalysts. The highly efficient catalytic system was composed of commercially available cobalt salt, Co(BF4)2·6H2O and a tetradentate phosphine ligand P(CH2CH2PPh2)3, affording 95% GVL yield at 100 °C and with atmospheric H2 pressure. Cobalt was also used as catalyst for GVL synthesis by Murugesan et al. [123], which applied LA for the preparation of cobalt-based nanocatalysts. These nanoparticles created reusable catalysts (up to 4 times) for the hydrogenation of LA, achieving 97% GVL yield in the following reaction conditions: 30 bar H2, 1,4-dioxane as solvent, 120 °C during 24 h.
Another approach to produce GVL is using a one-step conversion of carbohydrates. Both Brønsted acid catalyst and hydrogenation catalyst are required for the conversion of carbohydrates into GVL in the one-step method [124,125]. The acidity of Brønsted acid catalysts is determined by its dissociation constant (pKa), but also dependent on the solvent. Particularly water-contained solvents that are common in the biomass conversion process, usually negatively affect the acidity of Brønsted acid catalysts [126]. Cui et al. [109] studied several carbohydrates conversion into GVL combining a strong Brønsted acid, H3PW12O40 and Ru/TiO2 catalysts. The onestep conversion in a monophasic γ-butyrolactone/water (80/20, v v -1 ) solvent under a mild reaction condition (150 °C and 40 bar H2) resulted in GVL yields of 70.5 and 58.5 mol% from inulin and sucrose, respectively. As the most abundant biomass, cellulose is considered as an ideal resource to produce GVL. However, a direct conversion of cellulose into GVL is more challenging due its recalcitrance structure. The direct conversion of cellulose to GVL was investigated by the authors using the same conditions above and a yield of 40.5% of GVL was obtained. Most works concerning GVL production from carbohydrates involves the selective dehydration of carbohydrates to LA followed by hydrogenation of LA to give GVL by using an external H2 supply. A different route to convert carbohydrates into GVL without use of any external H2 supply was reported by Deng et al. [77]. The authors related a hydrogenation process using ruthenium-based catalytic systems accomplished only in the presence of the formic acid produced from the original acidic dehydration step of glucose. The route improved the atom economy of the process and avoided the energy-costly separation of LA from the mixture of LA and formic acid in aqueous solution. However, it is important to note that the proposed process presents some drawbacks, such as temperatures as high as 220 °C and reagents that are corrosive and toxic (hydrochloric acid and pyridine). Son et al. [117] synthesized GVL from one-pot dehydration/hydrogenation reaction of fructose in water solvent over supported metal catalysts. In this reaction, formic acid played two roles, an acid catalyst for dehydration of fructose to LA, and a hydrogen source for hydrogenation of the obtained LA. The Au/ZrO2 was the best catalyst showing an overall GVL yield of 48% and could be reused for several times, though it was partly deactivated. On the other hand, Fabos et al. [116] showed the hydrogenation of LA to produce GVL with a small excess of formic acid in the presence of the ruthenium Shvo catalysts. The reactions were performed at 100 °C in an open vessel with yields higher than 99% after 5 h. The only byproducts were water and carbon dioxide, which were easily eliminated, by distillation in the case of water and the carbon dioxide was bubble out of the solution. Although the transformation of LA to GVL can be achieved using heterogeneous catalysis, homogeneous catalytic systems that operate under milder reaction conditions give higher selectivities and can be recycled continuously. By using ruthenium complexes with a related chelating triphosphine ligand, N-triphos (N(CH2PPh2)3), Phanopoulos et al. [127] obtained high yields of GVL (77-95%) by using [RuH2(CO)(N-triphos)] as catalyst with either NH4PF6 or para-toluenesulfonic acid additives under the following conditions, 160 °C and 65 bar H2. A range of palladium complexes as catalysts was proved to be effective for the GVL synthesis under transfer hydrogenation using formic acid as hydrogen source and carrying out the reaction at only 5 bar H2 at 80 °C for 5 h [128]. The best-performing catalyst [Pd(DTBPE)Cl2] (DTBPE=1,2-(bis-di-tert-butylphosphino)-ethane) displayed 98% GVL yield. In addition, the catalyst could be recycled several times, but showed loss of catalytic activity due to the in-situ formation of an inactive Pd-carbonyl and a Pd-hydride dimer complex. Nevertheless, this report demonstrates the first use of Pd complexes for these transformations under mild reaction conditions [128]. GVL is a promising building block in organic synthesis due to its functional groups and reactivity. Although it contains a chiral center, it is usually produced and used in the racemic form because major applications, such as fuels or solvents, do not require one specific enantiomer. However, enantiomeric purity is important in the fragrance, flavouring and pharmaceutical industries, where a specific enantiomer can present different properties [129,130]. Starodubtseva et al. [131] studied the conversion of LA and γ-ketoesters derived from LA to enantiopure lactones using homogeneous catalysis. LA could be converted to (S)-GVL in moderate yields of 66% with 98.5% ee after 5 h at 60 °C and 60 bar H2 in ethanol using a Ru-BINAP catalysts prepared in situ and activated with HCl. When ethyl levulinate was used as substrate in the same reaction conditions, 95% yield and 99% ee for (S)-GVL was obtained. Using RuCl3-BINAP-HCl catalytic system, 96% (R)-GVL yield and 99% ee was achieved from methyl levulinate under similar catalytic conditions [132]. Biotransformations can also address the enantiomeric purity request, showing the benefit of mild reaction conditions and remarkable chemo-, regio-, and stereoselectivity. Gotz et al. [111] presented a chemo-enzymatic reaction sequence for the synthesis of optically pure (S)-GVL (Fig. 12). Initially, LA was esterified with ethanol at 70 °C for 16 h in the presence of Amberlyst 15 and MgSO4, resulting in 95% yield of ethyl levulinate. Without further purification, isolated ethyl levulinate was reduced by (S)-specific carbonyl reductase from Candida parapsilosis (CPCR2) at 30 °C and using isopropanol as cosubstrate to produce (S)-ethyl-4-hydroxypentanoate (95% of conversion and >99% ee). A subsequent lactonization of (S)-ethyl-4-hydroxypentanoate catalyzed by immobilized CalB yielded the desired (S)-GVL (>99% ee) at a reaction temperature of 30 °C. The heterogeneous catalyst was easily filtered off, and no further downstream processing was needed. An overall yield of approximately 90% (S)-GVL (based on LA) was achieved in this chemo-enzymatic reaction sequence.  [7,76,129,133]. The chemical structure of 2-MTHF is shown in Fig. 13. Because of the low melting point and low viscosity (1.85 cp at -70 °C), 2-MTHF is increasingly used in organometallic reactions, such as Grignard reaction, low-temperature lithiation, lithium aluminum hydride reductions, Reformatsky reaction and metal-catalyzed coupling reactions [7]. For instance, Mondal and Mora [134] related a catalyzed Suzuki-Miyaura cross-coupling reaction of acid chlorides and aryl boronic acids to yield aryl ketones using 2-MTHF as solvent. The great benefit of the water-immiscible 2-MTHF use was the easy isolation of the crude reaction mixture just by separation of 2-MTHF and water layers, followed by the evaporation of 2-MTHF. Also, it is important to highlight that 2-MTHF performed better than acetone, toluene, DMF, THF, acetonitrile, dichloromethane, PEG-400, isopropanol and a 3:1 mix of 2-MTHF and water. The 2-MTHF use in chemical reactions have been extensively reviewed elsewhere [8,133,135]. Besides that, Antonucci et al. [136] revealed that 2-MTHF has low toxicity and with a maximum concentration of 2% would not be expected to contribute to any toxicity potentially exhibited by an active pharmaceutical ingredient containing this solvent, and that is why, it has been approved for use in pharmaceutical chemical processes. The synthesis of 2-MTHF from GVL involves a hydrogenation of the carbonyl group to yield the cyclic hemiacetal (5-methyltetrahydrofuran-2-ol), which is in equilibrium with the 4-hydroxylpentanal. Further reduction of the remaining carbonyl group affords 1,4-pentanediol and an acid-catalyzed dehydration leads to cyclization of the diol by etherification to form 2-MTHF (Fig. 14) [2]. Thermodynamic data showed that the GVL ring-opening to 1,4pentanediol is a highly endothermic process at 250 °C (ΔG 0 = 70 kJ mol −1 ) [137]. Therefore, harsher reaction conditions are needed to obtain 1,4-pentanediol and consequently, 2-MTHF from GVL, making the process more challenging when compared with the GVL synthesis from LA [114]. The production of 2-MTHF based on GVL as a substrate was investigated by Al-Shaal et al. [76] applying Ru/C as catalyst. A full conversion of GVL was obtained in a solvent-free reaction system after 24 h at 190 °C and using a H2 pressure of 100 bar with a maximum 2-MTHF yield of 43%. The hydrogenation of GVL to 2-MTHF was coupled with the formation of by-products such as 2-butanol, butane, 1,4-pentanediol, 2-pentanol, 1-pentanol, pentane and methane (Fig. 15). On the other hand, Zhang et al. [138] using Ni-MoOx/Al2O3 as catalyst found, in addition to the 1,4-pentanediol, others by-products such as pentenoic and valeric acids. The maximum 2-MTHF yield (31%) was achieved when the reaction conditions were 200 °C, 40 bar H2 during 4 h. It is important to highlight that 1,4-pentanediol yield was 68% in this same reaction condition. Du et al. [113] have demonstrated that tuning the acidic properties of the catalyst surface alternates the product distribution in the hydrogenation of GVL. The catalyst Cu/ZrO2 was significantly modified by calcination in air at different temperatures in the range of 300-700 °C for 4 h. It was found that the catalyst obtained by 400 °C-calcination can deliver a remarkable 2-MTHF yield as high as 91% within 6 h of reaction, using ethanol as solvent and 60 bar H2 at 240 °C. However, the same catalyst calcinated at 600 °C resulted in 1,4-pentanediol as the major product (73% yield), while 2-MTHF yield was only 20%. Fig. 15. Proposed reaction pathway for the formation of by-products during the synthesis of 2-MTHF from GVL over Ru/C catalyst. Source: Authors, adapted from [76].
A direct conversion of LA to 2-MTHF involves several reaction steps, including hydrogenation of LA into GVL; further ring-opening to 1,4-pentanediol and final dehydration to afford 2-MTHF. Al-Shaal et al. [76] using a twostep hydrogenation reaction over Ru/C produced 2-MTHF from LA. The first hydrogenation step, LA was totally converted to GVL under 12 bar H2 at 190 °C and 45 min [112]. At the end of this step the formed water was evaporated from the reaction mixture and the catalyst was collected, washed, and dried. This procedure was necessary because water resulting from the hydrogenation of LA may inhibit the later dehydration reaction of 1,4-pantanediol to 2-MTHF. Thus, the second hydrogenation step was conducted using the dried catalyst and the produced GVL applying 100  The role of the hydrogen source on the selective production of GVL and 2-MTHF from LA was investigated by Obregon et al. [114] using three different solvents (1,4-dioxane, 1-butanol and 2-propanol) under reactive H2 and inert N2 atmospheres. The applied reaction conditions (250 °C, 40 bar H2 initial pressure and 5 h reaction time) were combined with the use of the following catalysts: Ru/C, Ni/Al2O3 and Ni-Cu/Al2O3. Under N2 atmosphere, catalytic transfer hydrogenation reactions are the principal source of hydrogen for the transformation of LA into GVL or 2-MTHF. Low hydrogen availability provided by a poor hydrogen donor such as 1,4-dioxane resulted in relatively low LA conversions for the three tested catalysts. The performance of all the three catalysts was improved with 2-propanol as the solvent, which is a well-known hydrogen donor [142], achieving more than 70% GVL yield. However, these in situ-generated hydrogen sources alone were insufficient to convert the highly stable GVL into 2-MTHF (yields were lower than 3%). Under a H2 atmosphere, catalytic transfer hydrogenation and hydrogenation with molecular H2 were effective and very fast at producing high yields of GVL, up to 85.3% using Ru/C in 1,4-dioxane and 93% using Ru/C in 1-butanol. The combination of both sources of hydrogen was indispensable to achieve significant yields of 2-MTHF. For all the catalysts, the highest 2-MTHF yields were obtained when the best hydrogen donor, 2-propanol, was used as a solvent. Overall, the use of Ni-Cu/Al2O3 resulted in 2-MTHF yields of approximately 40% after 5 h and 80% after 20 h.

Impact
The 2030 Agenda for Sustainable Development, adopted by all United Nations Member States in 2015, provides a clear guideline towards the economic, social and environmental sustainability. The aforementioned agenda established the 17 Sustainable Development Goals (SDGs), which are an urgent call for action by all countriesdeveloped and developing -in a global partnership. Among them are those related to the economic growth as well as those focused on subjects tackling a climate change and working to preserve our oceans and forests. Some of those goals can be accomplished by the integration of Green Chemistry Principles and Biorefinery Concept [143]. This work clearly demonstrates that both can be easily integrated and their integration may have a clear positive impact on numerous areas of life. Among them are e.g. environment, human health, economics and business.
In terms of impact of the proposed solutions on environment, are aspects related to the fact that many chemicals end up in the environment by intentional release during use (e.g., pesticides), by unintended releases (including emissions during manufacturing), or by disposal. Green chemicals either degrade to innocuous products or are recovered for further use. Additionally, plants and animals suffer less harm from toxic chemicals in the environment. Also, lower potential for global warming, ozone depletion, and smog formation can be achieved by the use of green chemistry and biorefinery concept. Furthermore, use of biomass as source of novel chemicals allows less chemical disruption of ecosystems as well as contributes to lesser use of landfills, especially hazardous waste landfills.
Hence, this has a direct impact on human health, namely on cleaner air because lesser release of hazardous chemicals to air leading to less damage to lungs can be achieved. Also, cleaner water can be expected because lesser release of hazardous chemical wastes to water leads to cleaner drinking and recreational water. Furthermore, by the implementation of green chemistry principles, an increased safety for workers in the chemical industry can be expected since less toxic materials are in use, less personal protective equipment are required and less potential for accidents (e.g., fires or explosions) can be foreseen. Consequently, safer consumer products of all types can be obtained. In this sense, new, safer products will become available for purchase and some of those commodities will be made with less waste or some products (i.e., pesticides, cleaning products) will be replacements for less safe products. Finally, the above-mentioned aspect has a direct impact on economy and business because higher yields for chemical reactions can be attained due to consuming smaller amounts of feedstock to obtain the same amount of product. Due to less synthetic steps, often allowing faster manufacturing of products, increasing plant capacity, and saving energy and water, economy of the process is definitively more favorable. At the same time, waste reduction, elimination costly remediation, hazardous waste disposal, and end-of-the-pipe treatments is additional economic benefit for the process. The use of waste as new feedstock allow replacement of a purchased feedstock contributing to favorable economics of these new businesses. In addition, due to better performance, less substrates are needed to achieve the same function. Furthermore, reduced manufacturing plant size or footprint through increased throughput is additional impact on economics of the process. Therefore, the use of biomass processed according to the green chemistry principles demonstrates a great potential for different areas with a positive impact on society.

Conclusions
The development and usage of greener solvents are topics emerging in the recent years, due to the increasing concerns about pollution and climate changes. In this context, more sustainable solvents coming from natural sources can also reduce the dependence of petrochemicals that are harmful, toxic, and environmentally damaging. The literature is full of examples of methods for the biomass conversion into chemicals and, in many cases, the authors claimed for a synthesis of they as potential green solvents and building blocks. However, there are controversies in the use of hazardous and petrochemical derived solvents and catalysts during these compounds obtaining. Among the various technologies described, the ideal in the green chemistry concept would be the use of residual biomass combined with heterogeneous recyclable catalysts and renewable solvents. The successful use of economically and environmentally sustainable solvents depends not only on the source of them but also on their properties to facilitate product/catalyst isolation and reaction workups, improving in reaction yields and reduction of environmental issues.

Conflict of interest
There are no conflicts to declare.