Читать книгу Biomass Valorization - Группа авторов - Страница 24

The valorization of lignocellulose is an intricate chemical problem. The challenge relates largely to the rigid supramolecular structure of native plant cell walls, formed by inter‐ and intramolecular bonding of polysaccharides, lignins, and sometimes other macromolecules [31]. A common strategy to improve reactivity is to separate carbohydrate polymers and lignin [32]. Fractionation of biomass is performed on large scale during commercialized pulping processes and may also be conducted by the selective catalytic transformation of one particular class of biomacromolecules [16,32–35]. For example, a carbohydrate fraction can be recovered by catalytic hydrogenolysis of lignin into low‐molecular‐weight phenolic substances [33]. It is also possible to retain the lignin fraction by the direct acid‐catalyzed processing of polysaccharides present in native biomass [34,35]. To complicate matters, the acid‐catalyzed transformation of individual macromolecules is a complex cascade of reactions that require both Brønsted acid and Lewis acid catalysts (Scheme 2.2) [4,7]. In addition, acidic catalysts may simultaneously promote side reactions of the products into undesirable by‐products, compromising selectivity of the targeted processes [4,7]. Consequently, judicious selection of the catalysts and processing conditions are essential to ensure the efficient processing of biomass into targeted value‐added products. This section will deal with major developments in the processing of cellulosic substances.

Cellulose is the most naturally abundant macromolecule on the Earth [36]. Even if this view is contested, lignocellulose is the only large volume biomass to which we have ready access on large industrial scale and where there is a globally distributed large volume industry and supply chain [4]. Cellulose consists of β(1 → 4) linearly linked glucose units and is a principal portion of plant cell walls. Hemicellulose is another polysaccharide present in lignocellulose and is often made of structurally branched xylose units and sometimes other moieties [6,16,31]. The past few decades have witnessed a significant interest in the acid‐catalyzed processing of cellulosic substances into organic building block chemicals (platform molecules), such as 5‐(hydroxymethyl)furfural (HMF), furfural (FF), levulinic acid (LevA), LacA, and their many derivatives [4].

A key first step in the valorization of polysaccharides is the depolymerization thereof into low‐molecular‐weight sugars, from which other value‐added chemicals are generated (Scheme 2.2) [6,37]. In their own right, low‐molecular‐weight saccharides are valuable ingredients in the food manufactory and also used as substrates for the fermentative production of ethanol or LacA [38–40]. Polysaccharides are commonly hydrolyzed in aqueous mineral acids in industrial processes [41]. Most technologies are based on thermal hydrolysis of cellulosic biomass in dilute aqueous sulfuric acid solution in batch or percolation reactors, with a typical glucose yield of 55–70% [41]. The glucose yield may be improved to 85%, when conducting the hydrolytic processing of biomass in a countercurrent shrinking bed reactor system at elevated temperatures (up to 225 °C) [42]. This technology has been engineered at National Renewable Energy Laboratory (NREL) by designing a cascade of percolation reactors simulating countercurrent. The available information suggests that this specific process has been demonstrated to bench scale at present [42]. The sustainability of such methods at industrial scale may be compromised by the need for forcing processing parameters [40,43]. On the other hand, complexities arise because of the formation of large amounts of acidic wastewater and solid waste (typically, calcium sulfate after neutralization of sulfuric acid) and also because of the requirement for corrosion‐resistant manufacturing equipment [4,40]. These ongoing challenges have led to the exploration of new efficient methods for the hydrolytic processing of polysaccharides into low‐molecular‐weight derivatives, where high yields and selectivities, accompanied by low levels of waste production, are key parameters.

Solid acid‐catalyzed reactions of polysaccharides in aqueous media have drawn particular interest [40]. The use of solid catalysts is attractive because of the ease of their recovery and sometimes recyclability. An interesting strategy has been devised using a cellulase mimetic catalyst [44], containing both cellulose binding and cellulose hydrolyzing sites, similar to cellulase enzymes. The catalyst is a sulfonated chloromethyl polystyrene resin (CP‐SO₃H) bearing chloride groups (–Cl) with which saccharides and sulfonic groups (–SO₃H) are coordinated to generate the Brønsted acidity requisite for the hydrolysis of the glycosidic linkages (Scheme 2.3) [44]. The catalyst CP‐SO₃H demonstrated impeccable activity during the processing of microcrystalline cellulose (MCC), or starch (a mixture of α(1 → 4) linearly linked glucose polymer amylose and branched α‐glucose polymer amylopectin), in high yields of glucose (up to 100%). CP‐SO₃H is significantly more efficient than a diluted solution of sulfuric acid under identical processing conditions, or many other reported solid acid catalysts, such as acidic resins, zeolites, carbonaceous acids, or functionalized silicas [44,45]. This success has spread the adoption of cellulase mimetic catalysts based on varied sulfonated chloroorganic materials [46]. Table 2.1 provides a summary of the outcomes of several acid‐catalyzed processes.

Scheme 2.3 Proposed structure and mechanism of catalytic action of the cellulase mimetic catalyst. Source: Based on Shuai and Pan [44].

The sustainability of solid acids is somewhat undermined by the insolubility of both the cellulose and the catalyst, the latter by design, in aqueous reaction mixtures [4,40]. The heterogeneity of the system limits the number of effective substrate–catalyst interactions during the processing, leading to several technical and economical hurdles. Firstly, there is a need for high loadings of the catalyst [4,44–46]. This may lead to the production of solid deactivated catalyst waste, if it is not fully recyclable. Another downside relates to the use of pretreated cellulose, instead of largely available lignocellulosic biomass. Pretreatment helps to ameliorate issues pertaining to mass transfer in heterogeneous systems, owing to the reduced physical size and sometimes lower molecular weight of pretreated polysaccharides. Examples of pretreated substrates are MCC, a medium value product obtained by the treatment of wood pulp in diluted aqueous acids, and ball‐milled cellulose [16,58]. Ball milling is arguably the most commonly used method for the pretreatment of cellulosic biomass. Although this approach is efficient at the bench scale and quite effectively depolymerizes cellulose to some extent, it becomes highly energy demanding at the larger scale and remains mostly unemployed in the industry [4]. Some studies, however, claim that some of the significant energy implications associated with this method of pretreatment may be avoided for acid‐assisted mechanochemical depolymerization [47,59]. For instance, mechanochemical depolymerization of beechwood and poplar wood may be realized at kilogram scale by ball milling of biomass in the presence of sulfuric acid [47]. Nevertheless, these processes require subsequent dilution of the pretreated acidified systems with water and hydrolytic treatment under forcing reaction conditions, which does not significantly differ from other processes in aqueous solvents (Table 2.1). Considering the energy costs accompanied by a need for corrosion‐resistant equipment, such means of depolymerization cannot be accepted as sustainable, unless the energy demands of ball milling can somehow be reduced.

Table 2.1 Conditions and results of the acid‐catalyzed processing of cellulosic biomass into carbohydrates and furansa.

Substrate	Reaction media	Catalyst	T (°C)	t (h)	Yield glucans (%)	Yield glucose (%)	Yield xylose (%)	Yield HMF (%)	Yield FF (%)	References
MCC	Water	CP‐SO3H	120	10	—	93	—	—	—	[44]
Starch				2	—	100	—	—	—
MCC		H2SO4			—	1	—	—	—
Ball‐milled MCC	Water	Carbonaceous Cl−, SO3H‐containing material	120	24	—	85	—	—	—	[46]
Ball‐milled MCC	Water	Si33C66‐823‐SO3H	150	24	4	50	—	0	—	[45]
		Sulfonated pyrolyzed sucrose			6	27	—	1	—
Beechwoodb	Water	H2SO4	145	1	2	76	74	2	6	[47]
Poplar woodb					2	78	82	2	8
Cotton linters	[C2mim]Clc	HCl	105	3d	—	87	—	6	—	[34]
Corn stover	[C2mim]Clc				—	70	79	—	—
Sigmacell	[C4mim]Cl	H2SO4	100	0.75d	—	38	—	—	—	[48]
	[C4mim]Cl	HCl	100	0.18d	—	21	—	—	—
Eucalyptus cellulose	[C4mim]Cl/ChCl/oxalicc	—	120	10	19	48	—	4	—	[49]
Pinus cellulose			120	10	21	50	—	6	—
Corncob			100	2	0	54	35	10	—
120	4
P. cruentum			100	2	0	55	40	0	—
120	4
Cellulose	[C2mim]Cl/DMA/LiCl	CrCl2/HCl	140	2d	—	—	—	54	—	[50]
Corn stover	CrCl3/HCl						47	37
MCC	[C2mim]Clc	CuCl2/CrCl2	120	8	—	—	—	58	—	[51]
MCC	[C2mim]OAc/[C4SO3Hmim]‐CH3SO3	CuCl2	160	3.5	—	—	—	70	—	[52]
Wood chipse	[C4mim]Cl	CrCl3	120	2	—	—	—	79	—	[53]
Rice strawe					—	—	—	76	—
Corn husk	ZnCl2·3.0 H2O/anisole	—	120	1d	—	—	—	—	26	[54]
P. cruentum				—	—	—	—	42
Corncob	ZnCl2·4.25 H2O	HCl	120	1d	—	61	—	30	22	[35]
P. cruentum					—	49	—	35	29
Softwood					—	26	—	22	11
Ulva lactuca					—	52	—	25	15
Inulin	ChCl/oxalic/EtOAc	—	80	2	—	—	—	64	—	[55]
Xylans	ChCl/citric/MIBK	AlCl3	140	0.58	—	—	—	—	69	[56]
P. cruentum	ChCl/oxalic	—	80	2	0	42	73	1	25	[57]
P. cruentum	ChCl/oxalic/MIBK	—	100	2	0	68	13	9	44
			4	0	32	0	5	72
Softwood	ChCl/oxalic/MIBK	—	100	5	0	0	0	1	55

^a We note that this table show yields of products either in mol% or in wt%; it is recommended to refer to the given references if the accurate evaluation of yields is sought. “—” = not specified; “0” = not detected, or detected in trace amounts; T, reaction temperature; t, reaction time; HMF, 5‐(hydroxymethyl)furfural; FF, furfural; MCC, microcrystalline cellulose; [C₂mim]Cl, 1‐ethyl‐3‐methylimidazolium chloride; [C₂mim]OAc, 1‐ethyl‐3‐methylimidazolium acetate; [C₄mim]Cl, 1‐butyl‐3‐methylimidazolium chloride; [C₄SO₃Hmim]CH₃SO₃, 1‐(4‐sulfobutyl)‐3‐methylimidazolium methanesulfonate; ChCl, choline chloride; DMA, dimethylacetamide; EtOAc, ethyl acetate; and MIBK, methyl isobutyl ketone.

^b The substrate was subjected to sulfuric acid‐assisted ball milling.

^c ILs were diluted with water during processing to a total water up to 10–43 wt% based on the reaction system.

^d t does not include the time for the dissolution of the substrate.

^e Substrate was treated with aqueous sodium hydroxide.

The rates and (often) selectivities of the chemical process improve if the process can be performed under homogeneous conditions. As mentioned above, cellulosic materials are essentially insoluble in aqueous systems and most common organic solvents. However, ionic liquids (ILs) in their many manifestations are potentially key to improved chemical transformations of cellulosic materials. ILs are a class of green solvents that consist solely of ions and, under certain conditions, are able to fully dissolve cellulosic polysaccharides [48,60,61]. This ability enables significant progress toward milder reaction conditions and better yields of some targeted products, making ILs outstanding reaction systems for the acid‐catalyzed valorization of native carbohydrates [4,61]. For example, quaternary ammonium salts, especially imidazolium derivatives such as 1‐alkyl‐3‐methylimidazolium chloride ([C_nmim]Cl, n = integer), have been effectively employed in the catalytic hydrolysis of cellulose and cellulosic biomass of diverse origin [34,49,62,63]. A seminal study [34] explored transformations of polysaccharides into monomer sugars (glucose and xylose) in a range of ILs in the presence of hydrochloric acid, a Brønsted acid catalyst. The study identifies that the molecular formula of ILs substantially influences the reaction outcomes, mostly related to the solubility of the substrate. There is an apparent need for the chloride anion to coordinate hydroxyl groups and to disrupt the extensive hydrogen bonding between polysaccharides, thus promoting their dissolution. In contrast, non‐coordinating, or weakly coordinating, anions, such as tetrafluoroborate (BF₄⁻), nitrate (NO₃⁻), bromide (Br⁻), and trifluoromethanesulfonate (triflate, OTf⁻), dissolve cellulose only poorly, slowing or preventing reactions from taking place [34,64]. The cation component of the ILs also influences the chemical reactivity of the cellulose in ILs: imidazolium salts or alkylpyridinium‐based solvents with longer alkyl chains reduce the solubility of cellulose in such ILs, hampering the reactivity [58,64]. Interestingly, imidazolium ILs with acetate (OAc⁻) and dimethylphosphate ((MeO)₂P(O)O⁻) counterions, which are excellent solvents for cellulose [64], provide zero yield of glucose after exposure to intended hydrochloric acid‐catalyzed processing of cellulose [34]. This is likely because of the reaction between the strong Brønsted acidic hydrochloric acid and the conjugate base of weaker acids, leading to formation of imidazolium chloride and a weaker Brønsted acid (acetic acid or dimethylphosphoric acid) that is unable to promote the hydrolysis under the selected processing conditions [62].

Another important finding is that acid‐catalyzed conversion of cellulosic biomass in IL media can be improved and effectively promoted by the gradual addition of water [34]. The optimized method provides impressive yields of glucose (up to 87 mol% based on the glucose content in the substrate) and xylose (up to 79 mol% based on the xylose component of the substrate) in [C₂mim]Cl solvent in the presence of hydrochloric acid catalyst at 105 °C, whose mixture was gradually diluted with 43 wt% of water (Table 2.1). Previous methods, which were based on the processing of cellulose in [C₄mim]Cl without addition of water, provided significantly lower yields of monosaccharides [48]. Although glucose is the terminal product of hydrolysis chemistry of cellulose, the hydrolysis of cellulose proceeds predominantly into glucose oligomers (cellotetraose, cellotriose, and cellobiose) from which glucose emerges (Scheme 2.4) [49]. There is also evidence that water suppresses the subsequent acid‐catalyzed conversion of saccharides into furanoids and by‐products in ionic solvents [34,49,62], thereby enhancing the yields of the desirable monosaccharides. The addition of water enhances the high yielding and selective transformation of polysaccharides into monomer sugars in ILs: it helps to promote the formation of monosaccharides and suppresses unwanted processes. However, care must be taken as to the timing of the addition of water. If this is done at the start of the process, then the substrate can remain undissolved because of the negative influence of the added water on the ability of ILs to dissolve biomass [34,62]. Finally, the processing of lignocellulose in imidazolium ILs leaves a lignin‐rich residue as the unreacted portion that can be potentially employed for subsequent valorization [34]. Table 2.1 provides processing conditions and outcomes of several instances, as described above.

ILs are highly tunable systems, and their physical and chemical characteristics can be modulated for specific tasks, providing high levels of flexibility for acid‐catalyzed processing [65–67]. It is possible to design acidic ILs, which can simultaneously act as a solvent and catalyst, enabling the dissolution of substrates (solvent effect) and their subsequent conversions (catalyst effect) [49,61,63,66,67]. Such approaches potentially eliminate the need for highly corrosive mineral acid catalysts, such as hydrochloric or sulfuric acid, potentially avoiding the associated technological downsides accompanying their use. In an interesting study involving mixed solvent systems [63], the use of ammonium salts functionalized with Brønsted acidic sulfonic acid groups and hydrogen sulfate anion was investigated. N,N,N‐Triethyl‐N‐(3‐sulfopropyl)ammonium hydrogen sulfate showed exceptional activity as a cosolvent in [C₄mim]Cl. The hydrogen sulfate system provided the Brønsted acid catalyst for the hydrolytic processing of MCC, affording very high yields of low‐molecular‐weight carbohydrates (yields of total reducing sugars 99%). The yields were determined using a colorimetric method based on the interaction of reducing carbohydrates with dinitrosalicylic acid [68], which may be subject to errors because of the reaction of dinitrosalicylic acid with other cellulose‐derived reducing substances. Nonetheless, the reported yields are impressive. More recently, our group has probed the use of a mixed ionic solvent system formed by [C₄mim]Cl and biorenewable acidic deep eutectic solvents (DESs; DESs are eutectic mixtures of Brønsted and Lewis acids and bases, often forming ILs) formed from choline chloride (ChCl) and oxalic acid dihydrate [49]. This mixed ionic system afforded high yields of low‐molecular‐weight saccharides (glucose yield up to 55 wt% and xylose yield up to 40 wt%, based on the substrate; yields were determined by liquid chromatography–mass spectrometry analysis, providing unambiguous detection of targeted products [49]) after processing of non‐pretreated cellulose (eucalyptus and Pinus) and cellulosic materials of terrestrial (corncobs) and marine origin (micro‐ and macroalgae, Table 2.1), avoiding the use of corrosive acids. We proposed that the catalytic activity of the DES is generated by Lewis acid‐assisted Brønsted acidity, owing to the complexation between Lewis acidic ChCl and Brønsted acidic oxalic acid [49]. It is worth mentioning that reactions in the cosolvent system required the addition of water after dissolution of the substrate to promote the hydrolysis into monomer sugars, similar to the previous instances [34].

Scheme 2.4 Acid‐catalyzed hydrolysis of cellulose into low‐molecular‐weight carbohydrates in ILs via oligosaccharides. n, integer; m, 0, 1, and 2 for cellobiose, cellotriose, and cellotetraose, respectively.

Acid‐catalyzed depolymerization of cellulose in ILs opens numerous avenues to generate significantly value‐added chemicals from low‐molecular‐weight sugars. Pertinent examples include the fermentative production of ethanol for biofuel applications, or transition metal‐catalyzed synthesis of hexitols for food and medical uses [34,69,70]. Another interesting application is found in their transformation into alkyl glycosides, a class of biodegradable surfactants with widespread employment in differentiated products such as cosmetics, body care, and cleaning formulations [71]. The Corma research group from the Polytechnic University of Valencia published a series of papers dedicated to the synthesis of alkyl glycosides through the hydrolysis of cellulose into low‐molecular‐weight carbohydrates in an ionic solvent, followed by glycosidation into alkyl glycosides and alkyl polyglycosides [72–74]. This was accomplished by the addition of long‐chain alcohols, such as 1‐octanol, 1‐decanol, or 1‐dodecanol, after depolymerization of cellulose in [C₄mim]Cl in the presence of acidic resin Amberlyst® 15. The method permitted the production of surfactants in high yield (up to 82 mol%), under mild processing conditions (temperatures around 100 °C) [72]. It is worth noting that alkyl glucosides are commercially synthesized from low‐molecular‐weight carbohydrates or structural polysaccharides in two steps, namely, glycosidation of the substrate with low‐molecular‐weight alcohols, followed by transacetalization with long‐chain alcohols [71]. The production of alkyl glycosides from cellulosic biomass in ILs is therefore a promising alternative pathway to generate bio‐based surfactants. The ongoing challenge relates to the separation of glycosides from ILs and is currently based on conventional chromatographic methods, which are difficult to sustain at larger scale, especially with mid‐priced performance chemicals [74]. Nevertheless, the problem may be potentially solved by the use of simulated moving bed chromatography that has already been applied to the simultaneous recovery of monosaccharides and ionic solvents [C₄mim]Cl at the multigram scale [70]. Simulated moving bed chromatography is largely employed in the fractionation of carbohydrates, and this method demonstrates significant commercial potential [70].

Among the various biorefinery processes, there is a particular focus on the transformation of cellulosic saccharides into furan derivatives [4,75]. These are useful targets because HMF (a cellulose‐derived product) and FF (a hemicellulose‐derived product) are raw materials for the production of biofuels, bioplastics, food additives, and pharmaceuticals [75,76]. The synthesis of furans is well investigated in aqueous media, mostly based on the acid‐catalyzed transformation of low‐molecular‐weight sugars such as fructose, sucrose, and xylose [75,76]. The direct conversion of undervalued polysaccharides into furans in aqueous solvents is difficult, mostly as a consequence of the chemical reactivity of aldehydes in the acidic reaction media [4]. As a case in point, furaldehydes are convertible into LevA and its derivatives (Scheme 2.2) and also to high‐molecular‐weight by‐products such as humins (condensation products of saccharides and aldehydes), which have a limited scope of applications [75,77,78]. Additional complexities arise in the planning and execution of these reactions because the conversion of polysaccharides into furans involves aldose–ketose isomerization promoted by Lewis acids (Scheme 2.2), whose activity is often compromised in aqueous reaction media [4,7].

The abovementioned issues have been largely alleviated by the employment of ILs, often because of the stabilization of the reactive furanoids and the catalyst in the ionic media [4,61,79]. A commonly applied strategy is the processing of polysaccharides in imidazolium‐based solvents in the presence of metal chloride catalysts (MCl_n, M = metal, n = integer) [4,61]. The solvent–catalyst interaction presumably leads to the formation of acidic catalytic complexes, [C_nmim]⁺[MCl_n+1]^– in the case of 1‐alkyl‐3‐methylimidazolium chloride solvents (Scheme 2.5), and these complexes tend to promote the requisite Lewis acid‐catalyzed aldose–ketose isomerization [79]. Brønsted acidity is likely achievable by the hydrolysis of some of these species in the presence of water with concomitant formation of metal aquo complexes and (hydrated) hydrogen cations (Scheme 2.5), as is commonly observed in aqueous systems [80]. Decomposition of imidazolium salts into N‐heterocyclic carbenes and HCl may also be a source of Brønsted acid activity [81]. However, in many instances, the processing requires the addition of protic acids to the reaction media [4,61]. The direct conversion of native cellulose and lignocellulose has been conducted in a cosolvent system comprising [C₂mim]Cl (20–80 wt%, based on the reaction system) and dimethylacetamide (DMA)/LiCl mixture (LiCl, 10 wt%) [50]. Mixtures of DMA and LiCl (present in differing ratios depending on the targeted application) are ionic media that have already been extensively employed in cellulose refining technologies [58]. This solvent system enabled the direct processing of biomass into HMF in good yields (up to 54 mol%, based on the cellulose content, Table 2.1) in the presence of the combined acid catalyst chromium(II) or chromium(III) chloride and hydrochloric acid [50]. With lignocellulosic substrate (corn stover), FF was also obtained (in addition to HMF), owing to acid‐catalyzed reactions of xylans (yield up to 37 mol%, based on the xylan content, Table 2.1). The yields of HMF could be improved during the direct processing of MCC in [C₂mim]Cl in the presence of Lewis acid‐assisted Lewis acid catalyst CuCl₂/CrCl₂ (58%) [51], or in the cosolvent system [C₂mim]OAc (1‐ethyl‐3‐methylimidazolium acetate)/1‐(4‐sulfobutyl)‐3‐methylimidazolium methanesulfonate ([C₄SO₃Hmim]CH₃SO₃) in the presence of CuCl₂ (70 wt%) [52], but these methods have not been applied to the valorization of native biomass. Combined treatment of lignocellulosic biomass (wood chips and rice straw) in aqueous sodium hydroxide solution (3%), followed by the CrCl₃‐catalyzed transformation in [C₄mim]Cl, enabled inspiring yields of HMF (up to 79 mol%) under relatively mild processing conditions (120 °C, two hours, Table 2.1) [53]. Apparently, treatment in aqueous basic solution removes a substantial portion of lignin and hemicellulose from the biomass, facilitating to the rapid dissolution and hydrolytic processing of the substrate in the ionic solvent. This method may become industrially viable, once the efficient recovery of the reaction system and the targeted furaldehyde have been engineered.

Scheme 2.5 Proposed formation of catalytic species in [C_nmim]Cl. R, alkyl; n, integer.

Although ILs are excellent media for the valorization of carbohydrates, these systems suffer some drawbacks, mostly related to the high cost of common ionic solvents, and sometimes to intricacies relating to their recycling [4]. These downsides pose a barrier to their widespread industrial acceptance. In this regard, many researchers are currently investigating less‐expensive ionic solvents for the valorization of biomass [4]. Zinc chloride hydrate solvents, with the conventional formula ZnCl₂·nH₂O (these systems are true ILs with the molecular formula [Zn(OH₂)₆][ZnCl₄] in the case of n = 3), have proved to be suitable for some biorefinery applications [35,54,82–86]. Such ionic systems have been historically employed as solvents in cellulose refining technologies and have been found useful in the production of cellulose aerogels, low‐molecular‐weight saccharides, and their derivative sugar alcohols [82–85]. Our recent systematic studies [35,54,86] demonstrate that ZnCl₂·nH₂O possesses intrinsic catalytic activity, which promotes the conversion of polysaccharides into value‐added molecules (Scheme 2.2 and Scheme 2.6). Moreover, it is possible to adjust the activity of ZnCl₂·nH₂O by manipulating the hydration number n [35,54,86]. Less‐hydrated media, ZnCl₂·2.5–3.0H₂O, favor the transformation of cellulose into furans, namely, HMF, furyl hydroxymethyl ketone (FHK), and FF (Figure 2.1). FHK and FF are unusual major dehydration products, and their formation is accordingly seldom recorded for the reactions of cellulose. FHK is considered to originate by the dehydration of the intermediate ketohexose (isomerization product derived from fructose), while FF is thought to form from fructose via an intermediate pentose (e.g. arabinose), as shown in Scheme 2.6. These two furanoids are used as specialty solvents, pharmaceutical intermediates, and in the production of performance resins. Interestingly, the selectivity to these unusual furanoids may be significantly improved when performing reactions in the biphasic system ZnCl₂·3.0H₂O/anisole [54,86]. This biphasic system is especially useful for the production of FF from native biomass because of the simultaneous conversion of both cellulose and hemicellulose into the targeted furaldehyde (yield up to 42 wt% based on cellulose and hemicellulose content in biomass, Table 2.1) [54]. In distinct contrast to less‐hydrated solvents, highly hydrated molten salts ZnCl₂·4.0–4.5H₂O transform cellulose predominantly into HMF (yield up to 21 mol%) and low‐molecular‐weight saccharides (total yield up to 48 wt%, Figure 2.1) [35]. The correlation between selectivity of the products and hydration levels of ILs is presumed to be related to the acidity of the reaction media, which diminishes with rising n, as was shown by pH readings and NMR spectroscopy [35]; however, the exact nature of the catalytic action of ZnCl₂·nH₂O remains to be established. After optimizing the process, high yields of HMF (up to 35 mol%), FF (up to 29 wt%), and sugars (up to 61 wt%) are achievable by performing the conversion of native lignocellulose (corncob and softwood) and algal biomass (macroalga Ulva lactuca or microalga Porphyridium cruentum) in ZnCl₂·4.25H₂O under relatively mild conditions (Table 2.1) [35]. In addition, the transformation of lignocellulose in zinc chloride hydrate solvents enabled the recovery of a lignin‐containing residue [35,87]. However, not all types of biomass were found to transform efficiently in the inorganic solvent. For example, native softwood is less amenable for the catalytic conversion (Table 2.1). Additionally, economical methods to recover products and solvents demand further investigations.

Scheme 2.6 Unusual acid‐catalyzed transformations of cellulose in zinc chloride hydrate solvents into furan‐type molecules. n, integer. Source: Bodachivskyi et al. [86].

Figure 2.1 Acid‐catalyzed transformation of cellulose into low‐molecular‐weight molecules in ZnCl₂·nH₂O. The figure specifies combined yields of mono‐, di‐, tri‐, and tetrasaccharides in wt% and yields of furans in mol%. Reaction conditions: MCC (50 mg), solvent–catalyst (5.000 g), 80 °C, 2.5 hours, then 120 °C, 1 hour [35]. ■: saccharides; □: 5‐(hydroxymethyl)furfural; : furyl hydroxymethyl ketone; and : furfural.

DESs are another class of alternative ionic media that exhibit some advantages relatively to imidazolium‐based salts. These relate to the ease of the preparation of DESs, which often require a one‐step combination of less toxic and naturally renewable precursors at moderate temperatures [88]. DESs are useful for the pretreatment and fractionation of cellulosic biomass and for the transformation of bulk cellulose into micro‐ or nanocrystalline cellulosic materials [89–91]. Some DESs are able to dissolve biomacromolecules and enable their subsequent catalytic conversion [55–57,92]. For example, acidic eutectic systems composed of ChCl and organic acids (usually oxalic acid or citric acid) are suited to this task and are recoverable media for the acid‐catalyzed transformation of inulin, a β(1 → 2) linearly linked fructose polymer with occasional chain‐terminating glucose units, and of xylans, yielding HMF (64%, Table 2.1 [55]) and FF (69 mol%, Table 2.1 [56]), respectively. However, in most applications, such DESs fail to convert cellulose into low‐molecular‐weight derivative products, with the exception of the process in the cosolvent system [C₄mim]Cl/ChCl/oxalic acid, as discussed above (Table 2.1) [49]. A study of the reactivity of several polycarbohydrates in the neat ChCl/oxalic acid solvent showed that starch, hemicellulose, and inulin are all soluble and convertible into monosaccharides and furans in this DES, while the linear polymer cellulose was almost insoluble and unreactive under similar conditions (reaction temperature 60–100 °C, time one hour) [57]. Arguably, there is an apparent correlation between solubility and reactivity of carbohydrates in the acidic DES. The subsequent processing of lignocellulose (corn husk, corncobs, and softwood chips) and algal biomass (U. lactuca, P. cruentum, and Chlorella vulgaris) provided conversions of native starch, xylans, and fructans into monosaccharides (glucose yield up to 68 wt%, xylose yield up to 73 wt%, based on respective polysaccharide content in biomass), HMF (up to 13 mol%), and FF (up to 72 mol%) in the neat DES or in the biphasic system ChCl/oxalic acid/methyl isobutyl ketone (MIBK) [57]. Some of these instances are given in Table 2.1, demonstrating that the formation of specific product(s) is favorable under specific reaction conditions and for specified substrate types [57]. The polysaccharide component of the microalgae P. cruentum, comprising predominantly structurally branched glucans and xylans, may be transformed into the respective monomers at modest reaction temperature and into FF at elevated temperature and extended reaction time (Table 2.1). In contrast, the direct processing of softwood chips with high cellulose content converts only hemicellulose into FF (yield 55 mol%, Table 2.1) and leaves a fine cellulosic powder as an unreacted portion of the biomass [57]. The cellulose so formed was shown to be highly amenable for the subsequent acid‐catalyzed transformation into low‐molecular‐weight saccharides and platform chemicals (furans and alkyl levulinate), likely associated with reduced size of the particulate substrate [57]. Such combined technologies are valuable with a view to a multistage sustainable biorefinery.

The chemistry covered to this stage shows that representative transformations of cellulosic biomass lead mostly to the formation of low‐molecular‐weight saccharides and furan derivatives. Evidently, the processing into these products requires somewhat similar reaction conditions (Table 2.1), albeit that there is a requirement for the presence of Lewis acid activity for efficient conversions into furanoids (Scheme 2.2). The production of functionalized organic acids or their derivatives via the rehydration of furans, or via retro‐aldol reaction of monosaccharides, is often favored by more forcing processing parameters (temperature 160–280 °C) [4,7]. Such cellulose‐derived products (e.g. LevA, formic acid (ForA), or LacA) are another group of platform chemicals with a broad scope for commercial applications [3,4]. Even though ILs are shown to be efficient for certain hydrolytic transformations, the need for more forcing conditions for the production of the named acids potentially causes issues with the ionic media: some of them may decompose at elevated temperatures, while others may increase the rates of side reactions, reducing the selectivity of the targeted product(s) [86,93,94]. Instead, water and alcohols become suitable media for the preparation of cellulose‐derived acids and esters, respectively. For example, the production of LevA and ForA, rehydration products derived from HMF, is commonly conducted in aqueous media. These platform chemicals are produced commercially from cellulosic materials by the Biofine process [95–97]. The technology involves the two‐step hydrolytic processing of low‐value cellulosic materials catalyzed by sulfuric acid (Figure 2.2). The first stage is a rapid depolymerization into low‐molecular‐weight sugars and their dehydration products in a plug flow reactor at a temperature of 210–220 °C. The second stage is the subsequent rehydration in a back‐mix reactor at temperature 170–200 °C [95–97]. The process ultimately yields LevA (70–80 wt% of theoretical maximum based on hexoses), ForA (70–80 wt% of theoretical maximum based on hexoses), and tars (biochar). Sometimes, FF arises if hemicellulose is present in the substrate (up to 70 wt% of theoretical maximum based on pentoses) [95–97]. Importantly, the ratio of products may be manipulated by changing the processing parameters at the first and at the second stages of the process [95–97]. Many works have attempted to outperform the Biofine process but fail to provide similar outputs and levels of flexibility to the original technology [4]. However, there remain problems to be solved. The most apparent downside of the Biofine process is the use of sulfuric acid catalyst that is typically consumed during the reaction and cannot be reused, along with the challenging recovery of the products from acidic aqueous media [98,99]. These identify a need for further developments to efficiently produce LevA and ForA from cellulosic biomass.

Figure 2.2 Acid‐catalyzed valorization of cellulose via the Biofine process. Source: From Hayes et al. [97]. © 2006, John Wiley & Sons.

Although the production of ForA has been substantially improved with commercialization of OxFA process (catalytic oxidation of carbohydrates using a catalyst H₅PV₂Mo₁₀O₄₀·35H₂O [100]), LevA and its derivatives remain in the center of the biorefinery research. Some issues that relate to the recovery of the catalyst and products may be resolved by conversions of saccharides in alcohol media instead of water. Such processes generate alkyl levulinates as principal products, which are relatively easy to recover by vacuum distillation [101]. These products may be converted into LevA or are directly used for the production of fuels and other specialty chemicals [101,102]. Lewis acidic metal trifluoromethanesulfonates (metal triflates) and their mixtures with sulfonic acids appeared to be effective and recoverable catalysts, providing high output of methyl levulinate (MLev) during the processing of MCC in methanol [101]. Sulfonic acids presumably improve the reaction rates of the Brønsted acid‐catalyzed cellulose solvolysis into low‐molecular‐weight saccharides, while the metal triflates promote further conversions of saccharides into MLev, likely via Lewis acid‐catalyzed isomerization (Scheme 2.2) [101]. In combination, these two acids (Brønsted and Lewis) favor the overall cascade of reactions and selectivity of MLev. In(OTf)₃ and aromatic sulfonic acids (1 : 5 molar ratio, respectively) provide the highest MLev yield (75 mol%, 180 °C, five hours, Table 2.2).

Our recent study employs an integrated technology comprising consecutive processing of unrefined low‐value cellulose in the DES ChCl/oxalic acid and then in ethanol to form ethyl levulinate (ELev) [57,103]. The first step generates fine cellulosic powder from bulk cellulose during the processing in the DES under mild conditions (80 °C, two hours). The product possesses a structure and properties consistent with MCC [103]. The second step is a high‐temperature transformation (160–180 °C) in ethanol under the action of metal triflates. We discovered that soft Lewis acidic metal triflates form synergistic Lewis acid‐assisted Brønsted acid complexes with phosphoric acid, among which a composition Y(OTf)₃/H₃PO₄ (1 : 1 molar ratio, respectively), affording for the highest ELev yield (up to 75 mol%, Table 2.2) [103]. Neither Y(OTf)₃ nor H₃PO₄ can separately catalyze the conversion of MCC into ELev, and only their combination generates the active catalyst (Table 2.2). Hard Lewis acids, including In(OTf)₃, show moderate activity in this process, which can be marginally improved in combination with p‐toluenesulfonic acid (TsOH), as was noted during the conversion of MCC in methanol (Table 2.2) [101]. Usefully, the conversion of wood‐derived cellulose, obtained after the processing of softwood chips in the biphasic system ChCl/oxalic acid/MIBK (Table 2.1), as disclosed earlier in the text, enabled similarly excellent conversion thereof into ELev in the presence of a combined acid catalyst Y(OTf)₃/H₃PO₄ (Table 2.2) [57]. Such integrated methods that involve different catalytic processes, leading to a range of value‐added chemicals, have significant potential to become commercially viable. Meanwhile, engineering of (preferably) continuous processes requires further laboratory‐ and pilot‐scale research.

Table 2.2 Conditions and results of the acid‐catalyzed processing of cellulosic biomass into organic acids or esters^a. Source: Bodachivskyi et al. [57,103].

Substrate	Catalyst	T (°C)	t (h)	Yield LevA or alkyl levulinates (%)	Yield ForA (%)	Yield LacA or α‐hydroxy acid derivatives (%)	References
Paper pulp	H2SO4	205 185	15 (s) 0.42	61 (LevA)	82	—	[96]
MCC	TsOH	180	5	20 (MLev)	—	—	[101]
	In(OTf)3			42 (MLev)	—	—
	In(OTf)3/TsOH			70 (MLev)	—	—
	In(OTf)3/NSA			75 (MLev)	—	—
MCC	H3PO4	160	4	0 (ELev)	—	—	[103]
	TsOH			0 (ELev)	—	—
	Y(OTf)3			0 (ELev)	—	—
	Y(OTf)3/H3PO4			68 (ELev)	—	—
	In(OTf)3			20 (ELev)	—	—
	In(OTf)3/TsOH			45 (ELev)	—	—
MCC	Y(OTf)3/H3PO4	180	2	75 (ELev)	—	—
Pinus cellulose (unbleached)b				75 (ELev)	—	—
Pinus celluloseb				73 (ELev)	—	—
Softwood				52 (ELev)	—	—
Softwood cellulosec				62 (ELev)	—	—	[57]
MCC	Er(OTf)3	240	0.5	4 (LevA)	3	90 (LacA)	[104]
MCC	Ga‐doped Zn/H‐nanozeolite Y	280	1	5 (MLev)	—	58 (MLac) 13 (MMP)	[105]
MCC	Zr‐SBA‐15	260	6	2 (ELev)	—	30 (ELac) 14 (EHB)	[106]
Scenedesmus	Sn‐beta	210	2	0	—	83 (LacA)	[107]

^a) We note that this table shows yields of products either in mol% or in wt%; it is recommended to refer to the given references if the accurate evaluation of yields is sought. “—”, not specified; “0”, not detected, or detected in trace amounts; T, reaction temperature; t, reaction time; MCC, microcrystalline cellulose; OTf, trifluoromethanesulfonate; TsOH, p‐toluenesulfonic acid; 2‐NSA, 2‐naphthalenesulfonic acid; LevA, levulinic acid; ForA, formic acid; MLev, methyl levulinate; ELev, ethyl levulinate; LacA, lactic acid; MLac, methyl lactate; MMP, methyl 2‐methoxypropanoate; ELac, ethyl lactate; and EHB, ethyl‐2‐hydroxybutanoate.

^b) Cellulose treated in the DES ChCl/oxalic acid [103].

^c) Cellulose treated in the DES ChCl/oxalic acid/MIBK [57].

Finally, we turn now to the valorization of cellulosic biomass into α‐hydroxy acids. The formation of these products occurs via Lewis acid‐catalyzed retro–aldol transformation of cellulose‐derived monosaccharides into C₂–C₄ sugars and sequential retro‐Michael dehydration into pyruvic aldehyde, rehydration, and isomerization into 2‐hydroxy carboxylic acids catalyzed by both Brønsted and Lewis acids (Scheme 2.2) [3,4,7]. LacA is usually the desired product of such transformations; however, other derivatives may also appear. The conversion of sugars into LacA requires hydrothermal conditions and Lewis acidic catalysts that remain stable in water at elevated temperature. Metal triflates are water‐tolerant Lewis acids and so are suitable catalysts for the task of hydrothermal conversion of cellulose into products [7,21]. In a study of a series of lanthanide triflates, it has been established that yields of LacA increase with a reduction of the ionic radius of the metal center La³⁺ < Ce³⁺ < Pr³⁺ < Nd³⁺ < Dy³⁺ < Ho³⁺ < Er³⁺ and is approximately equal for Er³⁺ ≈ Yb³⁺ ≈ Lu³⁺ [104]. Among these, Er(OTf)₃ was found to be the most active catalyst, providing an outstanding yield of 90 mol% yield of LacA during the reaction of MCC (Table 2.2) [104]. Importantly, the catalyst may be recovered by distillation of the product and be reused in the next cycle.

There is also a great deal of interest in zeolite‐catalyzed conversions of cellulosic saccharides into α‐hydroxy acids and esters [105–108]. For example, Ga‐doped Zn/H‐nanozeolite Y catalysts are active in the transformation of MCC in supercritical methanol, yielding methyl lactate (MLac) as the major product (yield 58%, Table 2.2) and small amounts of methyl‐2‐methoxypropionate (MMP, yield 13%, Table 2.2) and MLev (5%, Table 2.2) [105]. Mesoporous Zr‐SBA‐15 silicate was reported to catalyze the transformation of cellulose in supercritical aqueous ethanol (95%), providing moderate yields of ethyl lactate (ELac, yield 30%), ethyl‐2‐hydroxybutanoate (EHB, yield 14%), and ELev (yield 2%, Table 2.2) [106]. A Sn‐beta zeolite has been employed in the valorization of the carbohydrate‐rich microalga Scenedesmus sp. into LacA in an aqueous ForA solution (in this instance, formic acid is a catalyst, not a product) [107]. The catalytic system enabled selective and high‐yielding conversion of algal sugars into LacA in 83% yield under optimal conditions (210 °C, two hours, Table 2.2). It is proposed that ForA helps to degrade the algal cell walls, whereas the Sn‐beta catalyst promotes the retro–aldol reaction of glucose and other reactions leading to LacA [107]. Most likely, the reported excellent yields relate to the ease of the depolymerization of structurally branched algal sugars under acid‐catalyzed conditions.