The Heteropolyacid-Catalyzed Conversion of Biomass Saccharides into High-Added-Value Products and Biofuels

Abstract

The industrial processes used to produce paper and cellulose generate many lignocellulosic residues. These residues are usually burned to produce heat to supply the energy demands of other processes, increasing greenhouse gas emissions and resulting in a high environmental impact. Instead of burning these lignocellulosic residues, they can be converted into saccharides, which are feedstock for high-value products and biofuels. Keggin heteropolyacids are efficient catalysts for obtaining saccharides from cellulose and hemicellulose and converting them into bioproducts or biofuel. Furfural, 5-hydroxymethylfurfural, levulinic acid, and alkyl levulinates are important platform molecules obtained from saccharides and raw materials in the biorefinery processes used to produce fine chemicals and biofuels. This review discusses the significant progress achieved in the development of the processes based on heteropolyacid-catalyzed reactions to convert biomass and their residues into furfural, 5-hydroxymethylfurfural, levulinic acid, and alkyl levulinates in homogeneous and heterogeneous reaction conditions. The different modifications that can be performed to a Keggin HPA structure, such as the replacement of the central atom (P or Si) with B or Al, the doping of the heteropolyanion with metal cations, and a proton exchange with metal or organic cations, as well as their impact on the catalytic activity of HPAs, are detailed and discussed herein.

1. Introduction

The development of strategies to reduce dependence on fossil resources has become important in today’s world due to energy, environmental, and economic reasons. In this context, exploring biomass as a source of chemicals and alternative energy has emerged as an attractive option [1,2]. Lignocellulosic biomass is a sustainable feedstock because of its renewable nature, abundance, and wide availability [3,4,5]. Among the various types of natural raw materials, lignocellulosic biomass has been converted to energy and biofuels through thermochemical processes, including combustion, liquefaction, gasification, pyrolysis, hydrothermal processing, and biochemical routes [4,5,6].

Lignocellulosic biomass is a rich and renewable source of carbon for biorefinery processes. It consists of lignin, cellulose, and hemicellulose fractions, as shown in Figure 1 [3]. Lignin is primarily made up of aromatic alcohols and phenols, while hemicellulose and cellulose fractions mainly consist of saccharides. This type of biomass can provide feedstocks with varying contents of lignin, cellulose, and pentosan (hemicellulose). Figure 2 displays some examples of various types of biomass with their respective compositions [4,5,6,7,8,9,10,11,12,13,14,15,16].

The proportions of these components depend on the type of biomass and local agronomic conditions [17]. Cellulose and hemicellulose are the most valuable fractions of lignocellulosic biomass because they contain important saccharides such as fructose, xylose, and glucose, which are platform molecules in the production of biofuels and industrial chemicals [16,17,18]. These saccharides are essential for the production of biofuels and industrial chemicals. However, separating them from lignocellulosic biomass is challenging due to the presence of recalcitrant lignin, which is bonded to these fractions through cross-linkages [18]. Lignin is primarily made up of the alcohols aromatic coniferyl, p-coumaryl, and sinapyl, which form hydrogen bonds with cellulose and hemicellulose fractions (see Figure 3) [19].

To convert cellulosic biomass into glucose and fructose, the lignin should first be removed due to its strong interactions with the cellulose and hemicellulose fractions. This can be achieved through various processes, such as acid hydrolysis. The hemicellulose fraction consists of a linear chain of xylans, which provides high stability and does not need to be isolated like cellulose. However, its hydrolysis produces saccharides such as mannose, galactose, xylose, and arabinose [20] (Figure 4).

The food and forest-derived industries produce a significant amount of waste and cellulose residues. Usually, these residues are burned to generate heat and electricity. However, when chemically processed, these biomass sources can produce platform molecules that are useful for making fine chemicals or biofuels [21]. This also helps in reducing the emission of greenhouse gases [22]. This review focuses on demonstrating the effectiveness of Keggin heteropolyacids and their salts as catalysts in converting cellulose and hemicellulose into saccharide monomers, as well as converting these saccharides into platform molecules.

The structure of lignocellulosic biomass comprises polymers of pentoses and hexoses that have fructose and glucose as the main monomers. The hydrolysis of these polymers in an acidic medium generates xylose and glucose, which can be converted into valuable derivatives of platform molecules. Figure 5 and Figure 6 illustrate the conversion routes and the main intermediates involved.

When biomass is broken down to produce xylose as the main sugar, it yields five-carbon chain products such as furfural, 5-HMF, and levulinic acid (see Figure 6). On the other hand, if the breakdown of lignocellulosic biomass produces hexose sugars like glucose, they can be converted to levoglucosan or isomerized into fructose, which, in turn, yield levulinic acid and 5-HMF.

Figure 7 describes a reaction pathway where the initial biomass contains a high amount of polymeric cellulose. In this scenario, the first acid hydrolysis produces a cellobiose disaccharide, which is then hydrolyzed to glucose. Glucose can also be obtained from another disaccharide, sucrose. Following isomerization in an acidic medium, fructose is obtained and converted to 5-HMF [23].

2. Compounds Obtained from the Hydrolysis of Biomass Waste

2.1. Main Platform Molecules Obtained from Lignocellulose

2.1.1. 5-Hydroxymethylfurfural (5-HMF) and Derivatives

5-hydroxymethyl furfural is one of the top ten platform molecules used in biorefinery processes. It is used to produce valuable commercial chemicals such as ethers, carboxylic acids, or amine furfural-derived compounds like 5-EMF (ethoxy), 2,5-DMF (dimethyl), 2,5 dicarboxylic acid (FDCA; a polyester precursor), and 5-HMFA (amine). The latter two compounds are particularly important chemicals [24]. 5-EMF and 2,5-DMF are considered potential bio additives for fuels, specifically diesel and gasoline, respectively [25].

Figure 8 illustrates the reaction pathway for the oxidation of 5-HMF to 2,5 dicarboxylic furan acid and the main intermediates involved. Noble metals such as Pt, Rh, or Pd are used as catalysts in the hydrogenation steps [26]. Additionally, the reductive transformation of 5-HMF can also result in the formation of products of interest, mainly the potential precursors for sustainable aviation fuels (SAFs) (see Figure 9).

The conversion of biomass to 5-HMF and levulinic acid involves various steps, including biomass pre-treatment, which makes cellulose a lignin-free polymer. Afterward, the saccharides obtained are hydrolyzed, isomerized, and dehydrated in successive steps that are normally promoted by the Lewis or Brønsted acid catalysts (Figure 10) [27].

The Keggin heteropolyacid salts containing metal are bifunctional catalysts that can promote various reactions. The main challenge lies in controlling the selectivity when the acid medium favors the formation of humins due to the polymerization of furan compounds.

2.1.2. Furfural: Production and Derivatives

Furfural is utilized in the production of lubricant oils, biofuels, plastics, resins, fine chemicals (such as agrochemicals and drugs), and flavoring agents. Figure 11 illustrates the most commonly used production routes, which normally use sulfuric acid or phosphoric acid as catalysts [28]. Furfural has been used in lubricant oils and biofuels to fabricate plastics, resins, fine chemicals such as agrochemicals and drugs, and flavoring agents, among several other processes [29]. In this review, the intention is to show that HPA catalysts can replace the mineral acids in such reactions.

The industrial processes are conducted under pressure in a homogeneous phase, resulting in waste products and effluents that need to be neutralized. Moreover, these processes are energy-intensive and involve sulfuric acid as catalysts, reaching commonly low yields [30]. Therefore, it is crucial to find reusable and more efficient catalysts to enable the conversion of biomass to furfural under more environmentally friendly and sustainable conditions [31].

Furfural is considered one of the top thirteen platform molecules due to its wide range of applications as a feedstock in various industrial processes (Figure 12).

Figure 13 shows some examples of valuable alcohols and nitrogen compounds obtained from furfural. For example, pyrrole can be used as pharmaceuticals, agrochemicals, and veterinary products, in addition to being precursors in the synthesis of important natural products. Proline has pharmaceutical and biological applications [32].

The hydrogenation and/or dehydration of these compounds provides furan derivatives that are potentially useful as precursors of jet biofuels (Figure 14) [32].

The furfural oxidation with molecular oxygen takes place through two different ways, without or with an opening ring, as shown in Figure 15.

The first one gives lactones such as furanone, furan 2,5-dione, and/or furoic acid. The second one preferentially gives carboxylic acids such as maleic, fumaric, and succinic acids, which are highly valuable to the food and pharmaceutical industries [27].

2.1.3. Levulinic Acid Production and Their Derivatives

Similar to 5-HMF and furfural, obtaining levulinic acid from cellulose requires a series of tandem acid-catalyzed reactions such as hydrolysis, dehydration, and isomerization. Most of the time, 5-HMF and furfural are primarily obtained and then converted to levulinic acid (Figure 16).

The processes are usually carried out under homogeneous catalysis conditions, which can cause corrosion in the reactors and make it difficult to purify and separate the products. Additionally, a large amount of effluent and neutralization residues are produced during these steps.

In the following sections, we will explore the evaluation of heterogeneous catalysts in such processes. Metal oxides, zeolites, and sulfonic resins are some examples of alternative solid catalysts that have been considered for converting cellulose to levulinic acid [34]. However, the main challenge in developing efficient heterogeneous catalysts is that they have little contact with insoluble cellulose, which remains stable due to the hydrogen bonds network [35].

Levulinic acid can also be produced from furanoses, which contain six-membered carbon chains; nonetheless, in this case, the combination of hydrogen pressure and a noble metal catalyst is required, resulting in furfuryl alcohol that has its ring cleaved and hydrated to give levulinic acid (Figure 17) [36].

In addition to furfural, since the 1990s, the “Biofine Process” has produced industrial levulinic acid from lignocellulosic biomass [37]. Levulinic acid has been used as an ingredient to produce resins, plasticizers, antifreeze agents, and various other products [38]. Because it is a functionalized substrate with carbonyl and carboxylic groups, it can be employed in various organic transformations, such as redox, esterification, and polymerization reactions (Figure 18) [38].

However, before its use as a high-added-value chemical (Figure 18), it should be upgraded, hampering its commercialization at competitive costs compared to other biomass derivatives [36]. Alternatively, products such as alkyl levulinates can be easily obtained from the esterification of levulinic acid in acid-catalyzed reactions, generating alkyl esters at competitive prices [39] (Figure 19).

3. Heteropolyacid Catalysts

Keggin heteropolyacids (HPAs) belong to the class of polyoxometalates that are metal–oxygen clusters, constituted of oxygen atoms coordinated octahedrally to a metal cation with a high oxidation number (Mo⁶⁺, W⁶⁺). These metal–oxygen octahedrals are tetrahedrally arranged around central heteroatoms such as P, B, or Si [40,41,42]. Figure 20 shows a typical Keggin anion.

Keggin HPAs are strong Brønsted acid catalysts, whose protons, like di-hydronium cation [H₅O₂⁺], are bonded to the oxygen atoms placed in the terminal position of the octahedral present in the Keggin anion. This is a key aspect that can make them efficient catalysts in reactions with biomass-containing saccharides (Figure 21).

Keggin HPAs are acid solids with a low surface area, hampering their use as heterogeneous catalysts. This drawback can be overcome by exchanging their protons with cations with a large ionic radius or supporting the HPAs over a matrix with a high surface area, such as silica, zeolites, and molecular sieves as well [43,44,45,46,47].

In this review, we aim to demonstrate how Keggin HPA catalysts (H₃PW₁₂O₄₀, H₃PMo₁₂O₄₀, and H₄SiW₁₂O₄₀) can act as homogeneous or heterogeneous catalysts after being converted to metal salts or supported on solid matrixes. Keggin HPAs and their salts have been highly efficient catalysts in biomass conversion in homogeneous, heterogeneous, or biphasic systems or in anchored ionic liquids.

4. Results

4.1. Soluble Keggin Heteropolyacid Catalysts

4.1.1. Heteropolyacid-Catalyzed Reactions to Produce Monosaccharides in the Homogeneous Phase

Keggin heteropolyacids are soluble in the polar solvents commonly used in biomass conversion processes and have acted as homogeneous catalysts in different reactions. For instance, Zhang et al. assessed the catalytic efficiency of tungsten heteropolyacids in converting cellulose to glucose (see Figure 22) in a 10:1 MIBK/water biphasic system [48].

The authors found that the cellulose hydrolysis reactions achieved high conversion, regardless of the tungsten catalyst. Among them, the phosphotungstic acid-catalyzed reaction reached the highest glucose yield. These authors determined the total number of acid sites of the catalysts—H₄SiW₁₂O₄₀ (2.32 mmol/g), H₃PW₁₂O₄₀ (1.78 mmol/g), and H₅BW₁₂O₄₀ (2.17 mmol/g)—as well as their strength of acidity—H₄SiW₁₂O₄₀ (699 mV), H₃PW₁₂O₄₀ (516 mV), and H₅BW₁₂O₄₀ (588 mV). These measurements were performed using potentiometric titration. When the initial electrode potential of the solution is higher than 100 mV, the catalyst has strong acid sites.

In another study, Zhang et al. synthesized a series of Lewis metal-substituted phosphotungstic acids with the general formula H_n+xPW₁₁L^x+O₃₉, where M^x+ = Cu²⁺, Sn⁴⁺, Cr³⁺, Zn²⁺, Fe³⁺, Ti⁴⁺, and Zr⁴⁺, and tested their activity in the reactions to convert cellulose to glucose [49]. The goal was to evaluate the catalytic activity of these salts in water/MIBK (methyl isobutyl ketone) biphasic systems (Figure 23).

Zhang et al. found that the total acid sites of catalysts influenced the reaction conversions in reactions catalyzed by heteropolyacids [49]. The catalyst with the highest acid site content (H₅PW₁₁TiO₃₉) was the most active. However, the glucose yield was lower in the presence of two catalysts with a higher number of acid sites (H₅PW₁₁TiO₃₉ and H₅PW₁₁CuO₃₉) [49]. Mizuno et al. studied the catalytic activity of Keggin heteropolyacids containing different heteroatoms as the central atoms in the saccharification of crystalline cellulose [50]. In this process, crystalline cellulose was immersed in the HPA solution at room temperature for 24 h and then heated to 333 K for 48 h. The yield was based on the glucose unit used in cellulose. Figure 24 provides a summary of the main results.

These authors concluded that Keggin anions with a higher negative charge achieved a higher glucose yield: the lowest yield was in the presence of H₃PW₁₂O₄₀ and the highest yield was reached by the H₅BW₁₂O₄₀ catalyst. All these Keggin HPAs were more efficient than mineral acids (HCl and H₂SO₄) [51].

4.1.2. Comparison of Heteropolyacid-Catalyzed Reactions to Produce Monosaccharides with Other Soluble Catalysts

Table 1. Glucose synthesis in homogenous conditions in the presence of soluble catalysts. ^a [51]; ^b [52].

Entry	Substrate	Catalyst	Temperature (K)	Time (h)	Glucose Yield (%)
1 a	Cellobiose	HClO4	393	24	42
2 a	Cellobiose	H2SO4	393	24	29
3 a	Cellobiose	H3PO4	393	24	13
4 a	Cellobiose	H3PW12O40	393	48	51
5 a	Cellobiose	H4SiW12O40	393	48	52
6 a	Cellulose	H3PW12O40	333	48	18
7 a	Cellulose	H4SiW12O40	333	48	61
8 a	Cellulose	HCl	333	48	<1
9 a	Cellulose	H2SO4	333	48	4
10 b	Cellulose	HCl	453	2	14
11 b	Cellulose	H3PW12O40	453	2	42

shows the glucose yields obtained in reactions in the presence of different soluble catalysts [51,52]. It is important to note that all of the acids were used at the same proton concentration. However, time and reaction temperatures also impacted the catalyst activity (Table 1).

These results can also be compared to those described in Figure 22, Figure 23 and Figure 24 (see Section 4.1.1). This comparison shows that HPA catalysts are more active than mineral acid catalysts.

4.1.3. Comparison of Solid Heteropolyacid-Catalyzed Reactions to Produce Monosaccharides with Other Solid Catalysts

The performance of solid catalysts and HPAs can be evaluated in

Table 2. Glucose synthesis in heterogeneous conditions in the presence of solid catalysts. ^a [53]; ^b [54].

Entry	Substrate	Catalyst	Time (h)	T (K)	Glucose Yield/%
1 a	Cellulose	HY	8	453	37
2 a	Cellulose	HZSM-5a	9.5	453	35.2
3 a	Cellulose	HZSM-5b	9.5	453	33.7
4 a	Cellulose	H beta	8.5	453	29.6
6 a	Cellulose	Fe3O4-SBA-SO3H	3	423	25
7 a	Cellulose	ACSO3H	3	423	22
8 a	Cellulose	Amberlyst	3	423	14
9 a	Cellulose	Al2O3	3	423	3
10 a	Cellulose	Fe3O4-SBA	3	423	3
11	Cellulose	HZSM	3	423	2
12 b	Cellulose	CsH2PW12O40	6	453	39
13 b	Cellulose	Cs2.2H0.8PW12O40	8	443	36
14 b	Cellulose	[MIMPSH]3PW12O40	5	423	30

. A direct comparison of the activity of solid catalysts is sometimes not easy. For this reason, solid-supported catalysts are not included in Table 2. Only acidic solids are mentioned. Time and reaction temperatures also impact the catalyst activity (Table 2).

4.1.4. Homogeneous Heteropolyacid-Catalyzed Reactions to Produce Levulinic Acid and Alkyl Levulinates

The Keggin heteropolyacids can undergo various transformations, resulting in the formation of soluble or insoluble heteropolyacids. Different salts can be synthesized by either fully or partially replacing the protons of the original heteropolyacid with organic or metal cations. Additionally, the anions of the Keggin can have their heteroatoms replaced, or tungsten and molybdenum atoms can be substituted with other metals. To better understand the catalysts, we will initially only discuss those containing protons (pristine heteropolyacids or those derived from partial exchange). The catalytic properties of heteropolyacids are influenced by the composition of the Keggin anion, particularly the addenda atoms (Mo or W) and the central atoms (P, Si, B, and Al), which can affect their activity. However, all heteropolyacids will remain soluble. Figure 24 illustrates the results of the conversion of cellulose to levulinic acid (LA) in the aqueous phase (Figure 25) [49,52].

The excellent catalytic performance of soluble Keggin HPAs can be explained by comparing the composition of the heteropolyanion and the nature of the cation, which affect the Lewis and Brønsted acidity of these compounds. Figure 24 illustrates the LA yield achieved in reactions using Keggin HPAs with different heteroatoms and Brønsted Lewis molar ratios. The reactions with catalysts exhibiting a higher Lewis acidity resulted in a higher LA yield, with ChH₄AlW₁₂O₄₀ being the most active catalyst (Figure 24). Zhang et al. synthesized metal-substituted phosphotungstic acids with Lewis acidity properties and assessed their performance in converting cellulose to levulinic acid (Figure 26) [49].

When one unit of MO (M = W⁶⁺) is removed from the Keggin anion of phosphotungstic acid, the PW₁₂O₄₀³⁻ anion becomes a lacunar anion PW₁₁O₃₉⁷⁻. The resulting vacancy can be filled with metal cations, and the acid’s electroneutrality will be achieved with a number of protons that depend on the positive charge of the substituent metal. Among the acids evaluated, only those containing Fe³⁺ or Cr³⁺ cations had four protons and, therefore, had anions with the lowest negative charge. These acids achieved the lowest LA yield with very similar results. The other catalysts with five protons achieved superior performance; however, the LA yields were different. Zhang et al. attributed this behavior to the higher strength of the Lewis acidity of the doping metals of the heteropolyanion [50].

Metal-exchanged heteropolyacid salts were also effective catalysts in the conversion of cellulose to alkyl levulinates. Tao et al. synthesized the salts of chromium, aluminum, and tin from the reaction of precursor metal chloride salts and phosphotungstic acid [54]. Figure 27 shows the ethyl levulinate yields obtained from cellulose using metal-exchanged salts as catalysts. For comparison, the reactions with synthesis precursors are also included.

The metal chloride salts are also Lewis acid catalysts, but they were not as effective in esterifying LA with ethyl alcohol as the heteropoly salts. Even though H₃PW₁₂O₄₀ yielded slightly higher, it was only marginally better than these salts. On the other hand, when the protons were partially exchanged with metal cations, the resulting salts were significantly more active, such as the Al_1/3H₂PW₁₂O₄₀ and Sn_1/4H₂PW₁₂O₄₀ salts. Tao et al. also evaluated the activity of silicotungstic acid and its metal-exchanged salts [55]. The main results are shown in Figure 28.

The efficiency of H₄SiW₁₂O₄₀ (Figure 28) was higher than H₃PW₁₂O₄₀ (Figure 27). These authors assigned this result to the higher negative charge of the silicotungstic anion compared to the phosphotungstic, which favors its interaction with the substrate (cellulose).

Among the three different metal silicotungstate salts, the Al_1/3H₃SiW₁₂O₄₀ was the most effective (Figure 28). For that reason, Tao et al. chose aluminum salts to evaluate the impact of the level of exchange of protons on their catalytic efficiency. The exchanged salt Al_4/3SiW₁₂O₄₀ presented the lowest efficiency, while the partially exchanged Al_2/3H₂SiW₁₂O₄₀ was the most effective. This suggests that both Lewis and Brønsted acid sites are important in this reaction.

These authors evaluated the efficiency of an aluminum silicotungstate catalyst in the conversion of various kinds of biomass to ethyl levulinate (Figure 29) [55].

These different ethyl levulinate yields can be a consequence of the varied composition of the lignocellulosic biomass present in these feedstocks. Tao et al. also assessed the efficiency of aluminum silicotungstate in reactions with different alcohols using cellulose as the raw material.

An increase in the carbon chain size of alcohol resulted in a lower alkyl levulinate yield. The same occurred when alcohols having hydroxyl groups with a higher steric hindrance were used, such as sec-propyl or sec-butyl (Figure 30) [55].

4.1.5. Homogeneous Heteropolyacid-Catalyzed Reactions to Produce 5-HMF

Soluble Keggin HPAs have been also used as catalysts in reactions to convert cellulose to 5-HMF. Figure 31 shows the results obtained by Zhang et al. [49].

Among the Keggin heteropolyacids, the H₄SiW₁₂O₄₀ catalyst presented the highest molar ratio of Brønsted to Lewis acid sites (B/L) and achieved the highest 5-HMF yield (Figure 31). In the same work, these authors evaluated the activity of metal-substituted phosphotungstic acid salts in the conversion of cellulose to 5-HMF (Figure 32) [49].

As was verified in the reactions of the conversion of cellulose to levulinic acid (Figure 26), the catalysts with anions having a lower negative charge (H₄PW₁₁FeO₃₉ and H₄PW₁₁CrO₃₉) were those that achieved the highest 5-HMF yields.

4.1.6. Ionic Liquid Heteropolyacid-Catalyzed Reactions to Produce Levulinic Acid, Furfural, and 5-HMF

In addition to the exchange with metal cations, another strategy used to improve the catalytic activity of Keggin HPAs in the reaction for LA production is the partial exchange of their protons with organic cations. This modification becomes even more attractive if the organic cations can be used as ionic liquids.

Heteropolyacid ionic liquids (ILs-POMs) have advantages compared to both homogeneous and heterogeneous catalysts. The most important is that, due to their structure, their solubility in polar solvents can be changed with temperature. This means that, although they are insoluble at room temperature in polar solvents, they can be solubilized if the temperature is suitably increased [56]. Moreover, ionic liquids are environmentally friendly compared to fossil origin solvents and present acidity that can be useful in the reactions of biomass conversion [54]. Zhang et al. synthesized a series of phosphotungstic acids with the protons totally or partially exchanged with choline cations (Ch⁺) (H_3-n(Ch)_nPW₁₂O₄₀), which were used as catalysts in the polysaccharide conversion to monosaccharides. These catalysts were used to convert cellulose to 5-HMF in a biphasic system (water/MIBK) (

Table 3. Yields of 5-HMF obtained from cellulose hydrolysis [48].

Catalyst	Conversion	5-HMF Yield
No catalyst	2	-
ChCl	4	2
HCl	41	2
H3PW12O40	88	50
Ch1H2PW12O40	87	78
Ch2HPW12O40	48	30
Ch3PW12O40	49	17

The reaction conditions were as follows: cellulose, 0.1 g; MIBK, 5.0 mL; water, 0.5 mL; catalyst, 0.11 mmol; temperature, 413 K; and time, 8 h.

) [48].

Although the yields reached in the reactions with the H₃PW₁₂O₄₀ or Ch₁H₂PW₁₂O₄₀ catalysts were almost the same, the latter had the great advantage of remaining insoluble in the reaction, having been easily recovered and reused without loss of activity [48].

Previous works demonstrated that water/polar organic solvents biphasic systems can positively impact cellulose conversion, not only helping in its degradation but also extracting the products, enhancing the reaction selectivity [53,56]. Thus, these authors also verified the effect of the composition of the biphasic system in the cellulose conversion and the 5-HMF yield on the reactions catalyzed by choline phosphotungstate (Figure 33).

Among the biphasic systems tested, the MIBK/water was the most efficient due to the higher affinity between the product and the organic solvent. The literature describes that, while the hydrolysis of cellulose to glucose and its conversion to 5-HMF preferably happens in the aqueous phase, the possible conversion of 5-HMF to LA occurs in the organic phases [53,56].

Sun et al. demonstrated that the heteropolyacid ionic liquid [C₄H₆N₂(CH₂)₃SO₃H]_(3−n)H_nPW₁₂O₄₀ ([MIMPSH]_(n)H_(3−n)PW₁₂O₄₀, n = 1, 2, 3] directly catalyzed the preparation of LA from starch in a biphasic system (water/MIBK), achieving a 48.7% yield [53]. In this sense, Figure 34 presents some results of reactions carried out in different solvents (water and MIBK) with ionic liquid cations-exchanged tungsten heteropolyacids and a titanium-doped choline phosphotungstate salt.

Several research groups have synthesized partially exchanged HPAs, which are efficient catalysts at high temperatures and more active than the precursor heteropolyacids [49,52,53]. Zhang et al. first evaluated choline-exchanged phosphotungstic acids [48]. Posteriorly, they synthesized a series of metal-substituted phosphotungstic acid salts that were soluble in the reaction medium [49]. Among these salts, titanium-substituted tungstic acid was the most effective. They then solidified this acid by exchanging its protons with choline cations. As shown in Figure 34, even at lower temperatures, this salt was the most efficient catalyst [49]. Zhang et al. also assessed the activity of phosphotungstic acid in the conversion of cellulose to furfural in the presence of 1-butyl-3-methylimidazolium chloride ionic liquid [BMIN]Cl (Figure 35). These authors verified that high yields of furfural were achieved within a short reaction time at moderate temperatures. Swatloski et al. were the first to describe that 1-butyl-3-methylimidazolium chloride ionic liquid could dissolve cellulose [56].

Although synthesized by different research groups, all of these partially exchanged heteropolyacids are catalysts at a responsible temperature and are more active than precursor heteropolyacids [49,52,53].

It is noteworthy that, firstly, Zhang et al. evaluated choline-exchanged phosphotungstic acids [48]. In a second work, these authors synthesized a series of metal-substituted phosphotungstic acid salts, which were soluble in the reaction medium [49]. Among those salts, titanium-substituted tungstic acid was the most effective. Afterward, they chose this acid to be solidified, exchanging its protons with choline cation. Figure 34 shows that, even at lower temperatures, this salt was the most efficient catalyst [49].

The 1-butyl-3-methylimidazolium chloride ionic liquid was used as a solvent in reactions catalyzed by phosphotungstic acid or its silver salt to convert cellulose to 5-HMF (Figure 36).

Figure 36 shows the 5-HMF yields achieved in different solvents and catalysts: Ag₃PW₁₂O₄₀/MIBK, H₃PW₁₂O₄₀/BIMMCl, H₃PW₁₂O₄₀ + B(OH)₃/BIMMCl, and (CH₁₆)H₄PW₁₁TiO₃₉. Among them, the latter reached the highest conversion and 5-HMF yield, operating at the lowest temperature and shortest reaction time [57,58,59]. This result can be assigned to the large organic cation (hexadecyltrimethylammonium) in the heteropoly salt and to the biphasic system that contributed to the cellulose hydrolysis and 5-HMF extraction.

4.2. Solid-Supported Keggin Heteropolyacid Catalysts: Production of Levulinic Acid, Alkyl Levulinates, 5-HMF, and Furfural

In addition to converting Keggin HPAs to insoluble salts, supporting them over a high surface area solid is also a strategy frequently used in reactions of biorefinery to obtain platform molecules from saccharides [60]. Saghandali et al. synthesized halloysite-supported Keggin HPAs and investigated their activity in the dehydration of fructose to 5-hydroxymethylfurfural (Figure 37) [61].

The efficiency of the solid-supported heteropolyacid catalysts was the following: H₄SiW₁₂O₄₀/halloysite > H₃PMo₁₂O₄₀/halloysite > H₃PW₁₂O₄₀/halloysite. These authors attributed this result to the higher strength of the acid of the supported silicotungstic acid. This catalyst was also evaluated in the dehydration of different monosaccharides (Figure 38) [61].

The dehydration of disaccharides containing hexoses (cellulose) to fructose was not very efficient, and only a 2% yield was reached. For disaccharides containing hexose + pentose (sucrose), the catalyst was a little bit more efficient achieving a 12% yield. The catalyst was more efficient when monosaccharides were the substrates: glucose (64%) and fructose (99%). These authors attributed the lower yield obtained from glucose to the required isomerization to provide fructose and consequently the five-membered product (5-HMF) [61].

The H₄SiW₁₂O₄₀/halloysite-catalyzed reaction was compared to the other supported HPAs (Figure 39).

Wu et al. supported phosphotungstic acid on the graphitic carbon nitride and compared the performance of the precursors, support, and solid-supported catalysts (

Table 4. Load effects of H₃PW₁₂O₄₀/g-C₃N₄ (g-C₃N₄ = graphitic carbon nitride; n wt.%) catalyst on cellulose hydrolysis conversion and product yields [65].

Catalyst	Acid Loading mmol/g−1	Conversion (%)	Levulinic Acid Yield (%)	HMF Yield (%)
No catalyst	0	0	0	0
a g-C3N4	0	8	0	0
H3PW12O40	1.780	89	20	42
H3PW12O40/ g-C3N4(20 wt. %)	0.750	73	11	38
H3PW12O40/ g-C3N4(25 wt. %)	1.090	83	13	43
H3PW12O40/ g-C3N4(30 wt. %)	1.360	100	21	58
H3PW12O40/ g-C3N4(35 wt. %)	1.470	100	25	47

^a The reaction conditions were as follows: cellulose, 0.1 g; H₃PW₁₂O₄₀/g-C₃N₄ (n wt%), 0.2 g; water, 0.5 mL; MIBK, 5 mL; temperature, 413 K; and time, 10 h.

) [65].

In the absence of an acid catalyst, no reaction progress was observed. When the H₃PW₁₂O₄₀ soluble catalyst was present, the conversion was 89%, and 5-MF was the main reaction product formed. An increase in the dopant load on the support led to an increase in the conversion and 5-HMF yield (Table 4) [65].

Wu et al. studied the effect of temperature and reaction time on the LA and 5-HMF yields (Figure 40A,B, respectively).

The reactions carried out at 313 K provided the highest product yields. Regardless of the time or reaction temperatures, the 5-HMF yields were always higher than the LA yields. Temperatures greater than 413 K led to the degradation of both products, diminishing their yields. Lv et al. assessed the activity of silica-supported silicotungstic acid in reactions performed in a DMSO/water biphasic system to convert cellulose to 5-HMF [62].

Soluble phosphotungstic acid achieved a 5-HMF yield similar to that obtained in the reaction over the H₃PW₁₂O₄₀/SiO₂ solid-supported catalyst. Nonetheless, when the silica was treated with ATS (aminopropyltrimethoxysilane), the yield jumped to 78.31% [66]. This yield was superior to those achieved with the H₄SiW₁₂O₄₀/SiO₂ (Figure 41) or H₃PW₁₂O₄₀/g-C₃N₄ catalysts (Figure 40), both in an MIBK/water biphasic system.

In comparison to the halloysite-supported catalyst, this was a superior result, although herein the reaction was longer (Figure 42). Huang et al. investigated the ATS (aminopropyltrimethoxysilane)-silica solid-supported phosphotungstic acid catalyst activity on the glucose conversion to 5-HMF in an acetone/water biphasic system [66].

The highest conversion of glucose to furfural, as well as the highest furfural yield, was achieved in the DMSO solution containing dissolved phosphotungstic acid. When using DMSO with the silica-supported acid catalyst, the conversion was significantly lower and the furfural yield was only 10% lower. The biphasic system of water and toluene was less efficient, probably due to the lower interaction between the substrate and catalyst and the more difficult product extraction [67,68]. The same group evaluated the efficiency of the toluene/water system in the conversion of glucose to furfural using (MCM-41) mesoporous silica-supported cesium phosphotungstate salt at different temperatures (Figure 43) [69].

Dias et al. studied the effect of the silica support porosity, and the solvent system strongly impacted the conversion of xylose dehydration reactions and furfural yield. Comparing the performance of MP or LP silica-supported cesium phosphotungstate salts, the highest conversions were achieved with MP silica-supported catalysts, regardless of the solvent (Figure 44) [69].

Similarly, independently of the silica porosity, the greatest results were obtained in reactions carried out in the DMSO solvent [69]. This can be attributed to the higher extractive efficiency of the DMSO solvent and the higher interaction of the substrate with the MP silica support.

The reaction carried out at the highest temperature (423 K) provided the greatest levulinate ethyl yield (Figure 45) [70]. When the reactions were at temperatures near 350 K, the performance of the phosphotungstic acid/ZrO₂-Si-Ph-Si catalyst was much superior to the other catalysts [70,71,72,73].

Figure 46 shows that the esterification of levulinic acid with short- or long-carbon chain size alcohols can achieve high ester yields, since the temperature and time reactions can be optimized, regardless of the catalyst or support used [71,72].

The support played a key role in the esterification reaction of levulinic acid with methyl alcohol (Figure 47). The reaction over the H₃PW₁₂O₄₀/ZrO₂-Si-Ph-Si catalyst achieved the highest ester yield (99%) within the shortest reaction time. The greater surface area of the support can explain this result [73].

Replacing the protons of Keggin HPAs with large cations such as Cs can also allow the use of these salts as heterogeneous catalysts. Guo et al. synthesized the bimetallic salts of 12-tungstophosphoric acid with different Sn/Cs molar ratios using ultrasound-assisted coprecipitation and tested their activity on the xylose dehydration (Figure 48) [74].

The authors verified that the partial exchange of Cs⁺ with Sn⁴⁺ did not modify its Keggin structure but increased the particle sizes, improving its catalytic performance in D-xylose dehydration [74].

The protons of Keggin HPAs can also be exchanged with organic cations. Figure 49 displays the 5-HMF yields obtained from glucose dehydration carried out over the heterogeneous catalysts derived from phosphotungstic acid.

The phosphotungstate salt exchanged with the organic cation with the sulfonic group was the most effective catalyst, with ([MIM-PS]₃PW₁₂O₄₀ achieving the highest 5-HMF yield. Possibly, the sec-butyl alcohol was also the best solvent to interact with the product [77]. The use of Ag₃PW₁₂O₄₀ as the catalyst in an aqueous medium also gave a high 5-HMF yield (85.7%) [78]. Two different research groups also assessed the activity of silver silicotungstate or phosphotungstate salts; nonetheless, the reactions were performed in a water/MIBK biphasic system (Figure 50) [79,80].

Although the reaction over Ag₄SiW₁₂O₄₀ achieved the highest yield (86%), the catalyst load was two and a half times higher and the reaction time two times longer; therefore, it suggests that the Ag₃PW₁₂O₄₀ salt was the most efficient catalyst (Figure 50).

Likewise, observed in reactions of cellulose conversion to 5-HMF, the combination of titanium-doped phosphotungstate salt with an efficient solvent system (water/MIBK) provided the highest yield [49]. This biphasic system was also very efficient on the conversion of fructose to 5-HMF (Figure 51), achieving a 94% yield [80,81].

5. Conclusions

In this review, we discussed the processes involved in synthesizing furfural, 5-hydroxymethylfurfural, levulinic acid, and alkyl levulinates from saccharides like fructose, glucose, and xylose. We focused on the use of Keggin heteropolyacids in both homogeneous and heterogeneous reaction conditions. We began with a brief discussion about the components of lignocellulosic biomass (lignin, hemicellulose, and cellulose) and then explored the main pathways for converting cellulose and hemicellulose into high-value products. Next, we addressed the applications and production methods of biorefinery platform molecules such as 5-hydroxymethylfurfural, furfural, levulinic acid, and alkyl levulinates. Subsequently, we delved into the use of Keggin heteropolyacids in biorefinery reactions, describing their main properties and discussing their use as homogeneous catalysts, focusing on acids like H₃PW₁₂O₄₀, H₃PMo₁₂O₄₀, and H₄SiW₁₂O₄₀.

The focus was on demonstrating the use of heteropolyacids in the production of saccharides or their conversion into biorefinery platform molecules. The discussion also covered the performance of these heteropolyacids after modifications such as the partial or complete exchange of protons with metal cations or the replacement of Si and P heteroatoms with Al, B, and Ti, specifically in biorefinery reactions. Keggin heteropolyacids can be transformed into heterogeneous catalysts by replacing their protons with large organic or metal cations (such as choline, cesium, and imidazolium). The use of these salts in the conversion of saccharides to platform molecules was explored. Additionally, it was noted that Keggin heteropolyanions containing W can be substituted with metal cations to create efficient catalysts. The use of these salts in biorefinery reactions was also discussed. Special attention was given to the use of solid-supported heteropolyacids or heteropoly salts. Furthermore, the use of ionic liquids as solvents in heteropolyacid-catalyzed reactions or as a source of organic cations was also discussed. A comparison and discussion of a series of heterogeneous processes using solid-supported heteropolyacid catalysts or heteropolyacid salts was provided. The effects of the critical factors affecting the conversion and product yields, such as solvent, temperature, time, concentration, and type of heteropolyacid catalyst, were also discussed. The primary goal was to describe the recent advances in the development of heteropolyacid catalysts and their application in reactions to convert saccharides into biorefinery platform molecules.