This is an open access article published under an ACS AuthorChoice License, which permits copying and redistribution of the article or any adaptations for non-commercial purposes.
Article pubs.acs.org/accounts
Cite This: Acc. Chem. Res. XXXX, XXX, XXX−XXX
Cellulose Depolymerization over Heterogeneous Catalysts Abhijit Shrotri, Hirokazu Kobayashi, and Atsushi Fukuoka* Institute for Catalysis, Hokkaido University, Kita 21 Nishi 10, Kita-ku, Sapporo, Hokkaido 001-0021, Japan ABSTRACT: Cellulosic biomass is the largest source of renewable organic carbon on our planet. Cellulose accounts for 40−50 wt % of this lignocellulose, and it is a feedstock for industrially important chemicals and fuels. The first step in cellulose conversion involves its depolymerization to glucose or to its hydrogenated product sorbitol. The hydrolysis of cellulose to glucose by homogeneous mineral acids was the subject of research for almost a century. However, homogeneous acids have significant drawbacks and are neither economical nor environmentally friendly. In 2006, our group reported for the first time the ability of heterogeneous catalysts to depolymerize cellulose through hydrolytic hydrogenation to produce sorbitol. Later, we reported the hydrolysis of cellulose to glucose using carbon catalyst containing weakly acidic functional groups. Understanding the reaction between cellulose and heterogeneous catalyst is a challenge as the reaction occurs between a solid substrate and a solid catalyst. In this Account, we describe our efforts for the conversion of cellulose to sorbitol and glucose using heterogeneous catalysts. Sorbitol is produced by sequential hydrolysis and hydrogenation of cellulose in one pot. We reported sorbitol synthesis from cellulose in the presence of supported metal catalysts and H2 gas. The reducing environment of the reaction prevents byproduct formation, and harsh reaction conditions can be used to achieve sorbitol yield of up to 90%. Glucose is produced by acid catalyzed hydrolysis of cellulose, a more challenging reaction owing to the tendency of glucose to rapidly decompose in hot water. Sulfonated carbons were first reported as active catalysts for cellulose hydrolysis, but they were hydrothermally unstable under the reaction conditions. We found that carbon catalysts bearing weakly acidic functional groups such as hydroxyl and carboxylic acids are also active. Weakly acidic functional groups are hydrothermally stable, and a soluble sugar yield of 90% was achieved in a 20 min reaction. We clarified that the polycyclic aromatic surface of the carbon adsorbs cellulose molecules on its surface by CH−π and hydrophobic interactions driven by a positive change in entropy of the system. The adsorbed molecules are rapidly hydrolyzed by active sites containing vicinal functional groups that recognize the hydroxyl groups on cellulose to achieve a high frequency factor. This phenomenon is analogous to the hydrolysis of cellulose by enzymes that use CH−π and hydrophobic interactions along with weakly acidic carboxylic acid and carboxylate pair to catalyze the reaction. However, in comparison with enzymes, carbon catalyst is functional over a wide range of pH and temperatures. We also developed a continuous flow slurry process to demonstrate the feasibility for commercial application of carbon-catalyzed cellulose hydrolysis to glucose using inexpensive catalyst prepared by air oxidation. We believe that further efforts in this field should be directed toward eliminating roadblocks for the commercialization of cellulose conversion reactions. conformation and are linked together via β-1,4-glycosidic bonds (Figure 1). The hydroxyl groups are arranged in equatorial position and are linked with neighboring anhydroglucose units through hydrogen bonds. The cellulose chains are packed to form a crystal structure that is impermeable to mild chemical attacks. In addition, the structure of cellulose makes it insoluble in water and common organic solvents. Therefore, reactions involving untreated cellulose are heterogeneous in nature, and the products are dissolved in the reaction media. Glucose is the target product of cellulose conversion as it is a precursor to many valuable chemicals (Figure 1).5 Glucose is obtained by depolymerization of cellulose through acid catalyzed hydrolysis of β-1,4-glycosidic bonds in the presence of water.6 Cellulose is initially hydrolyzed to soluble fragments of β-1,4-glucans7 that undergo further hydrolysis to yield
1. INTRODUCTION Cellulose is a naturally occurring polymer found in the structural matrix of lignocellulose in plants. As a major component, comprising 40−50% of lignocellulose, cellulose is the most abundant biopolymer on our planet.1 The utilization of cellulose for synthesis of chemicals and fuels can reduce the dependence of chemical industries on fossil fuels.2 Cellulose, unlike starch, is not a source of food for humans, and its use will not divert edible resources for chemical synthesis.3 Furthermore, cellulose is a carbon neutral feedstock as the CO2 generated from chemicals at the end of their use is recaptured in producing lignocellulosic biomass. Therefore, use of cellulose as an alternative feedstock will reduce the fossil fuel demand and contribute toward the mitigation of CO2 driven climate change. Cellulose is a homopolymer of anhydroglucose units with degree of polymerization ranging from 300 to 10000.4 The anhydroglucose monomers in cellulose are present in chair © XXXX American Chemical Society
Received: December 8, 2017
A
DOI: 10.1021/acs.accounts.7b00614 Acc. Chem. Res. XXXX, XXX, XXX−XXX
Article
Accounts of Chemical Research
Figure 1. Cellulose derived from lignocellulose is a feedstock for glucose and sorbitol, which are precursors to value added industrial chemicals.
Figure 2. Scheme showing single pot hydrolytic hydrogenation of cellulose to sugar alcohols using only heterogeneous catalysts.
glucose as the primary product. The formation of glucose is concomitant with its decomposition to form byproducts. Selective synthesis of glucose is challenging as the apparent activation energies are of the same order for the decomposition of glucose and the hydrolysis of cellulose in an acidic reaction medium.6 Sorbitol is a product of glucose hydrogenation, and sequential conversion of glucose during cellulose hydrolysis prevents the formation of byproducts (Figure 1).8 Sorbitol is used as an artificial sweetener, and it is a precursor to polycarbonate type polymers. Currently, sorbitol is commercially produced by hydrogenation of glucose derived from starch. Therefore, direct synthesis of sorbitol from cellulose is of high importance. Application of a heterogeneous catalyst for cellulose conversion has been a topic of intense research, as the use of homogeneous catalyst is not economical. Dilute and concentrated mineral acids have been used for hydrolysis of cellulose to glucose since the 1920s.9 Strong acids such as H2SO4 and HCl were used because of their high activity for cellulose hydrolysis. In the 1950s, Balandin and co-workers combined the hydrolysis of cellulose to sugars in dil. H2SO4 and the hydrogenation of soluble sugars over a supported Ru catalyst in a one pot reaction.10 Their work marked the first attempt for hydrolytic hydrogenation of cellulose to reduce the formation of byproducts by producing sorbitol. However, these processes were not commercially successful because of the cost of separation and neutralization of mineral acids from the product solution. Heterogeneous catalysts are widely used in process industry to overcome these issues. Therefore, use of heterogeneous catalysts for cellulose conversion is essential for commercial success of this reaction. This Account describes our
efforts in studying the conversion of cellulose to glucose and sorbitol using heterogeneous catalysts in the past decade.
2. HYDROLYTIC HYDROGENATION OF CELLULOSE 2.1. Discovery of the Reaction
The heterogeneously catalyzed conversion of cellulose is a solid−solid reaction as the cellulose is insoluble in water and other common reaction media. Early attempts in our group for hydrolysis of microcrystalline cellulose to glucose using zeolites as solid acid catalysts were unsuccessful. Glucose yield of only 4% was obtained after 24 h at 463 K owing to the rapid degradation of glucose during the reaction. When a supported metal catalyst (Pt/γ-Al2O3) was used under reducing reaction conditions (H2, 5 MPa at RT), sorbitol was obtained in 25% yield along with its isomer mannitol in 6% yield.11 A follow up study revealed that the role of solid catalyst was not limited to hydrogenation of glucose, and it also catalyzed cellulose hydrolysis as the conversion of cellulose increased in the presence of supported metal catalyst.12 Therefore, in the hydrolytic hydrogenation of cellulose using supported metal catalysts, cellulose was first hydrolyzed to soluble β-1,4-glucans and glucose, which were then rapidly hydrogenated to sugar alcohols with high yield and selectivity (Figure 2).12,13 2.2. Role of Catalysts in Hydrolytic Hydrogenation
We studied different solid catalysts for this reaction to understand the role of active sites and to increase the sugar alcohol yield. Durability of the catalysts during reaction was strongly influenced by the type of support used. γ-Al2O3 used in our initial study was hydrated in hot-compressed water to form AlO(OH) (boehmite) with a low specific surface area,12,14 which caused aggregation of Pt nanoparticles and reduced the B
DOI: 10.1021/acs.accounts.7b00614 Acc. Chem. Res. XXXX, XXX, XXX−XXX
Article
Accounts of Chemical Research Table 1. Representative Studies in Hydrolytic Hydrogenation of Cellulose yield, % catalyst 2.5 wt % Pt/Al2O3 2 wt % Pt/BP2000 2.5 wt % Ru/HUSY 4 wt % Ru/AC 2 wt % Ru/AC 2 wt % Ru/AC 5 wt % Ru/C, PO43−/ZrO2 3 wt % 60 nm-Ni/CNF 7.5 wt % 16 nm-Ni/oxidized CNF 70 wt % 16 nm-Ni/KB 40 wt % 21 nm-Ni/HZSM-5 16 wt % Ni2P/AC 7.3 wt % Ni−0.86 wt % Pt/H-beta 17 wt % Ni−1 wt % Pt/HZSM-5 1 wt % Ru/oxidized CNT 17 wt % 28 nm-Ni/meso-HZSM-5 5 wt % Ru/C, H4SiW12O40 (pH 1.9) 5 wt % Ru/C, H4SiW12O40 (pH 1.3) 0.2 wt % Ru/HUSY, HCl (pH 2.3) 5 wt % Ru/C, H2SO4a (pH 1) 10 wt % Ni−1 wt % Pt/Al2O3
substrate
temp, K
time, h
p(H2), MPa
Solid Catalysts 463 24 5 463 24 5 463 24 5 518 0.5 6 463 18 0.8 463 3 0.9 463 2.5 6 463 24 6 463 24 6 483 6 5 513 4 4 498 1.5 6 473 6 5 513 4 4 Solid Catalysts + Acid Pretreatment H3PO4-treated 458 24 5 H2SO4-treated 513 2.5 4 Solid Catalysts + Mineral Acids BCc 463 1 9.5 α-cellulose 433 7 5 BCc 463 24 5 Solid Catalysts + Mechanocatalytic Hydrolysis with Mineral Acids oligomersd 423 1 5 oligomersd,e 473 1 5 MCb BCc MCb MCb BCc mix-milled mix-milled BCc BCc BCc BCc MCb MCb MCb
sorbitol
25%
81%
58%
77%
26 49 in total 30 30 58 in total 50 64 61 in total 48 34 in total
69 75% in total
mannitol
ref
5 9
11 12 11 16 22 23 29 25 30 21 34 27 35 31
10 8 9 6 7 6 5 3
3
85% in total 53% in total 66% in total 91% in total 76
17 36 39 38 37
13
32 33
a
Present in substrate. bMicrocrystalline cellulose. cBall-milled cellulose. dProduced by mechanocatalytic hydrolysis of cellulose with H2SO4. Neutralized with Ba(OH)2.
e
activity of catalyst.15 Porous carbon materials were watertolerant supports that were durable during the reaction.12,16,17 A catalyst made using carbon black as support, Pt/BP2000, produced sorbitol in ca. 50% yields from ball-milled cellulose after repeated use.12 This catalyst could also convert cellulose and hemicellulose in real biomass to sugar alcohols.18−20 Basic support materials were not suitable for synthesis of sugar alcohols, as the base sites caused retro-aldol reactions to form ethylene glycol and propylene glycol as final products.21 Different metal species supported on catalyst showed different activity for hydrolytic hydrogenation. Ru showed higher activity than Pt for the hydrogenation reactions. Pt/ BP2000 required more than 2 MPa of H2 pressure to obtain good selectivity for sugar alcohols, but Ru/AC (an activated carbon support, Norit SX Ultra) worked under 0.8 MPa of H2.22 The catalyst possessed small Ru particles of 1.5 nm in average, which were important for the high activity. Ru particles with 9 nm diameter were inactive under the same conditions. Ball milling of cellulose and Ru/AC together to promote contact between substrate and catalyst increased the yield of sorbitol to 58% along with 9% mannitol under 0.9 MPa H2 pressure.23 The reason for higher catalytic activity of Ru relative to Pt is unclear, but a possible factor is the reduction in activation barrier because of formation of hydrogen bond between the water adsorbed on oxophilic Ru surface and the carbonyl group on sugar molecules.24 Base metal catalysts are preferred over noble metals for the production of bulk chemicals because of their abundance and low cost. However, typical supported base metal catalysts such as 1 wt % Ni/C and Co/C were inactive for the hydrolytic hydrogenation of cellulose.19 Sels et al. first reported an active
Ni catalyst for the production of sorbitol from cellulose.25 They prepared a carbon nanofiber (CNF) supported Ni catalyst by methane decomposition over Ni/SiO2. This method produced large crystalline Ni particles on the tip of CNF.26 Ni/CNF converted ball-milled cellulose to sorbitol in 50% yield. Zhang and co-workers reported supported Ni2P as an active catalyst,27 although the catalyst was not very stable in hot-compressed water owing to the leaching of P.28 We studied supported Ni catalysts and found that good activity and durability are possible by increasing the loading of Ni.21 Ni (70 wt %)/KB (Ketjen black EC-600JD, Lion) gave more than 50% yield of sorbitol, and it was durable for reuse up to seven times. However, 10 wt % Ni/KB produced sorbitol in only 3.8% yield in the first reaction. Particle size of Ni was the deciding factor for the activity and durability of Ni catalyst. Low Ni loading resulted in small Ni particles (4 nm for 10 wt % Ni/KB) containing a large fraction of NiO. During reaction, Ni particles are severely sintered, and hydration of NiO produces Ni(OH)2 that covers the Ni particles. In contrast, high Ni loading catalysts have large Ni particles (16 nm for 70 wt % Ni/KB), which are more resistant to sintering and inhibit the complete coverage by Ni(OH)2. Table 1 summarizes representative results for the hydrolytic hydrogenation of cellulose. The yield of sugar alcohols has steadily increased over time and has surpassed 70%.29−31 Multiple studies, including our own, clarified the importance of large particle size for the activity and stability of Ni catalyst, a departure from the conventional notion that small metal particles are better catalysts. The yield can be further increased to 90% by using soluble oligomers produced by mechanocatalytic depolymerization of cellulose, as substrate.32,33 Metal C
DOI: 10.1021/acs.accounts.7b00614 Acc. Chem. Res. XXXX, XXX, XXX−XXX
Article
Accounts of Chemical Research
catalytic hydrolysis with H2SO4. After partial neutralization of the oligomer/2-propanol/water solution to pH 2.7, it was continuously converted to sorbitol by Ru/AC in a fixed bed reactor.
species (Pt, Ru, Ni), cellulose pretreatment, and presence of soluble acids dramatically influence the yield of sugar alcohols. 2.3. Kinetic Analysis
We performed kinetic analysis for hydrolytic hydrogenation of ball-milled cellulose in the presence of Pt/BP2000 catalyst.12 The amount of solid cellulose gradually decreased in a convex downward curve during the reaction. Glucose was detected as a major intermediate. Yield of sorbitol followed a second-order curve in the initial stage, indicating that sorbitol is the product of a consecutive reaction. Therefore, we assumed pseudo-firstorder reactions for the hydrolysis of cellulose to glucose and hydrogenation of glucose to sorbitol. The calculated rate constant for the hydrogenation (k2 = 2.6 h−1) was an order higher than that for the hydrolysis (k1 = 0.21 h−1), which resulted in the high yield of sorbitol by stabilizing the hydrolysis products. Significant amounts of cello-oligosaccharide alcohols were found as intermediates in some cases, especially when the hydrolysis of cellulose was accelerated by mix-milling pretreatment.23 The conversion of cellulose in the absence of catalyst provided k1 of 0.082 h−1.12 In this case, the hydrolysis may be caused by direct attack of water on the β-1,4-glycosidic bonds and catalyzed by possible acid species produced by the decomposition of cellulose.40 It is notable that the rate constant was lower than that in the presence of Pt/BP2000. This result shows that Pt/BP2000 catalyzes not only hydrogenation but also hydrolysis of cellulose.
2.5. Utilization of Sugar Alcohols
Sorbitol and its derivatives have many applications in the food industry, pharmacy, and chemical manufacturing.2 Polyisosorbide carbonate (Durabio, Mitsubishi Chemical), a polymer made from isosorbide, is an engineering plastic commercially used in smart phones and cars. Isosorbide for this polymer is produced by dehydration of sorbitol in the presence of an acid catalyst. The current method uses H2SO4 that produces a large amount of sulfuric acid pitch. Our group and Yokoi et al. independently reported that Hβ zeolite with a high Si/Al ratio (75) is specifically active for this reaction to give isosorbide in 76% yield.43,44 Presence of acid sites and hydrophobicity is necessary to obtain high reactivity using zeolites, and these two parameters are balanced at a Si/Al ratio of 75. Moreover, the confinement of molecules in the micropore of Hβ also provided selective conversion of mannitol to isomannide, a potential precursor to plastics.45 The micropore confinement in Hβ is significant because the mannitol dehydration in open space yields 2,5-sorbitan as the main product.
3. HYDROLYSIS OF CELLULOSE Shortly after our initial report for the hydrolytic hydrogenation of cellulose,11 the groups of Hara and Onda reported the hydrolysis of cellulose to glucose using sulfonated carbon catalysts.46,47 The catalysts were prepared by sulfonating carbon materials with sulfuric acid at elevated temperature. The treatment introduced phenolic and carboxyl groups along with sulfonic groups. The catalyst hydrolyzed cellulose to glucose in 16% yield after 27 h reaction at 373 K.46 The glucose yield increased to 40% when cellulose was pretreated by ball milling.47 Many researchers improved upon these ideas and reported cellulose hydrolysis using carbon catalysts with sulfonic groups.48−50
2.4. Hydrolytic Transfer Hydrogenation of Cellulose and Its Application for Flow Process
Transfer hydrogenation is an alternative for hydrolytic hydrogenation of cellulose to avoid use of high pressure H2. Transfer hydrogenation with 2-propanol is widely used in organic synthesis, where a base is often required for the activation of the alcohol.41 However, presence of base is detrimental for the hydrolytic hydrogenation of cellulose, because hydrolysis is an acid catalyzed reaction. Furthermore, base catalyzes the retro-aldol reaction of sugars to produce smaller molecules. We found that Ru/AC converts cellulose to sorbitol using 2-propanol in the absence of a base (Figure 3).22 Acetone is produced as a dehydrogenated product of 2propanol, which can be hydrogenated back to 2-propanol in a separate process. Using transfer hydrogenation, the hydrolytic hydrogenation can be performed in a fixed bed reactor.42 In this process, cellulose is first converted to soluble oligomers by mechano-
3.1. Weakly Acidic Carbon Catalysts
We studied the use of weakly acidic functional groups present on carbon surface for the hydrolysis of cellulose. In our initial study, CMK-3, a mesoporous carbon material, was active for hydrolysis of amorphous cellulose to glucose in 14% yield with 24% yield of β-1,4-glucans.7,51 The activity of CMK-3 was attributed to the favorable interaction between cellulose and carbon surface along with the presence of small number of weakly acidic groups such as carboxyl and phenolic groups (0.41 mmol g−1 in total). The activity of carbon materials was increased by introducing oxygenated functional groups through oxidation. K26, an alkali activated carbon containing 0.88 mmol g−1 of weakly acidic groups, hydrolyzed cellulose with 36% glucose yield.52 A screening of carbon materials revealed a strong correlation between the number of weakly acidic groups and the yield of glucose (Figure 4). Furthermore, removing the functional groups from K26 by heat treatment at 673−1273 K reduced the catalytic activity, confirming that the weakly acidic groups were the active sites. In these reactions, the lack of efficient contact between catalyst and cellulose was a major limiting factor. To overcome this challenge, cellulose and carbon catalyst were ball-milled together in a pretreatment step denoted as mix-milling. This method forced the adsorption of cellulose molecules on the carbon surface in a solvent free condition. Hydrolysis of cellulose and K26 mix-milled substrate
Figure 3. Transfer hydrogenation of cellulose to sorbitol and mannitol over Ru/AC catalyst with 2-propanol as hydrogen donor. D
DOI: 10.1021/acs.accounts.7b00614 Acc. Chem. Res. XXXX, XXX, XXX−XXX
Article
Accounts of Chemical Research
Carboxyl groups are preferred over phenolic and lactone groups at higher oxidation temperature. 3.2. Adsorption of Cellulose on Carbon
Functionalized carbon catalysts show high activity for cellulose hydrolysis because of their tendency to adsorb cellulose molecules on their surface. In a solid−solid reaction, the contact between substrate and catalyst particles is limited to their interaction by collision. Mix-milling removes the probability of collision as a limiting factor by adsorbing cellulose on carbon surface prior to reaction. The rate of cellulose hydrolysis is one order higher after mix-milling than the reaction when catalyst and cellulose were milled separately.55 The influence of mix-milling is limited to physical adsorption of cellulose as it does not affect the rate of hydrolysis of β-1,4-glucans that are soluble in water. The ability of carbon catalysts to adsorb cellulose was analyzed by studying the equilibrium adsorption of glucose and soluble small β-1,4-glucans in water.56 The surface of carbon adsorbs cellulose through CH−π interaction, mainly consisting of dispersion forces working between CH groups of cellulose and π electrons on the polyaromatic surface of carbons (Figure 6A). The CH−π interaction provides a negative change in enthalpy (ΔHads ° = −14 kJ mol−1 for cellobiose). Furthermore, the axial face of cellulose molecule, which lacks hydroxyl groups, is hydrophobic, and its interaction with the surface of carbon reduces the exposed hydrophobic surface area in aqueous phase. Consequently, the adsorption of cellulose on carbon reduces the number of restricted water molecules near the hydrophobic surface resulting in a positive change in entropy (ΔSads ° = +24 J K−1 mol−1 for cellobiose). The CH−π and hydrophobic interactions are similar to those of cellulose hydrolyzing enzymes. The adsorption is stronger for large molecules of β-1,4-glucans as evidenced by an increase in adsorption coefficient with the degree of polymerization. The porosity of carbon catalysts plays an important role in cellulose adsorption and hydrolysis. Zeolite templated carbon, having a pore radius of 0.6 nm, adsorbs cellulose molecules dissolved in HCl having a radius of gyration of 2.9 nm within 2 min.57 The carbon material pulls the cellulose molecule within its pore causing a strain on the cellulose chain that promotes the hydrolysis of β-1,4-glycosidic bonds to relieve the strain.
Figure 4. Correlation of total number of weakly acidic functional groups with the yield of glucose after the hydrolysis of amorphous cellulose.
at 453 K for 20 min converted 93% of cellulose to soluble products. β-1,4-Glucans with degree of polymerization 2−6 units were the primary product (70% yield) and glucose yield was 20%. Addition of a small amount of mineral acid (0.012 wt % HCl) rapidly converted the β-1,4-glucans to glucose (total yield 88%) under the same conditions. It should be noted that homogeneous acid played only a minor role in the hydrolysis of cellulose as a reaction in the presence of acid but without carbon catalyst resulted in only 39% conversion of cellulose. Therefore, the role of homogeneous catalyst was limited to the rapid hydrolysis of soluble β-1,4-glucans to glucose and increase in selectivity by reducing the byproduct formation. The cost of catalyst synthesis is high when chemicals such as HNO3 and KOH are used for oxidation. We used air as an inexpensive and environmentally friendly oxidant for the synthesis of cellulose hydrolyzing catalyst. Eucalyptus, a hardwood lignocellulosic biomass, was heated in air at 573 K to produce a catalyst containing aromatic structure and carboxylic groups (2.1 mmol g−1).53 This material can catalyze the hydrolysis of Eucalyptus, the source of catalyst, to glucose (78%) and xylose (94%) after mix-milling and in the presence of trace HCl (Figure 5). After hydrolysis, the solid residue contains the catalyst and the unreacted lignin, which was converted back to active catalyst via oxidation under the same conditions. This process eliminates the need for fractionation of lignocellulose prior to hydrolysis. Air oxidation can also functionalize activated carbon to produce cellulose hydrolyzing catalyst. Interestingly, the selectivity of functional groups introduced is controlled by the oxidation temperature.54
3.3. Synergy of Weak Acid Functional Groups
Weakly acidic functional groups located on adjacent carbon atoms show synergistic effect for hydrolysis of β-1,4-glycosidic bonds. Phthalic acid and salicylic acid, having adjacent
Figure 5. Reaction cycle for carbon catalyst prepared by air-oxidation of Eucalyptus and used for the hydrolysis of Eucalyptus itself. HC denotes hemicellulose fraction in Eucalyptus. Reproduced with permission from ref 53. Copyright 2016 Royal Society of Chemistry. E
DOI: 10.1021/acs.accounts.7b00614 Acc. Chem. Res. XXXX, XXX, XXX−XXX
Article
Accounts of Chemical Research
3.4. Continuous Flow Process for Cellulose Hydrolysis
The ambition that hydrolysis of cellulose serves as an entry point for chemical industry requires commercialization of the reaction. The heterogeneous hydrolysis of cellulose is difficult to scale-up as it is a slurry phase reaction requiring high pressure and temperature. Currently, there are no examples of this process being operated above laboratory scale. All reports use a batch reactor that cannot be scaled up for commercial application. Therefore, we designed a continuous flow slurry reactor to demonstrate the feasibility for scale-up of this reaction. Our process, shown in Figure 7, uses a pneumatic pump to continuously feed the slurry mixture into the reactor.59 Mixmilled slurry rapidly hydrolyzes to soluble β-1,4-glucans (57%) and glucose (25%) in the slurry reactor maintained at 493 K. Addition of a small amount of H3PO4 to the slurry before reaction (4.5 mM H3PO4) hydrolyzes the β-1,4-glucans to glucose and increases the glucose yield to 67%. Alternatively, if homogeneous acids are to be avoided, a fixed bed reactor with Amberlyst-70 catalyst can be used for hydrolysis of the β-1,4glucans after separation of solid residue. The process is robust, and it can hydrolyze mix-milled Eucalyptus to yield a mixture of glucose and xylose. The space time yield for glucose production in our process was as high as 472 kg m3 h−1, which shows the high productivity of a continuous process for cellulose hydrolysis.
4. CONCLUSIONS The unreactive nature of cellulose and a lack of understanding in solid−solid reaction systems make the heterogeneous catalytic conversion of cellulose a challenging task. In our research, porous carbon materials have shown remarkable ability to perform as catalyst and catalyst support for cellulose conversion. Carbon catalysts show high affinity for cellulose molecules because of CH−π and hydrophobic interactions between them. The adsorption of cellulose molecules on carbon prevents hydrogen bond formation among the molecules, which sterically hinders the conformational change required for hydrolysis. Using mainly carbon based heterogeneous catalysts, we have developed catalytic processes for hydrolysis and hydrolytic hydrogenation of cellulose to glucose and sorbitol, both highly valuable products for chemical industry. Ni and Ru supported carbon materials are inexpensive catalysts that show high activity for sorbitol synthesis from cellulose. Ru/AC can catalyze the transfer hydrogenation reaction to produce sorbitol without the use of molecular
Figure 6. Proposed mechanisms of cellulose adsorption on carbon surface by CH−π interaction (A) and hydrolysis of β-1,4-glycosidic bonds by vicinal functional groups on carbon surface (B).
carboxyl−carboxyl and carboxyl−phenolic groups, show higher frequency factor for hydrolysis of cellobiose than their corresponding meta and para isomers.58 One of the functional groups makes a hydrogen bond with the hydroxyl group of cellobiose and increases the probability of the neighboring carboxyl group attacking the β-1,4-glycosidic bond (Figure 6B). o-Chlorobenzoic acid, which has a similar pKa but lacks a hydrogen bond forming hydroxyl group, shows much lower frequency factor for cellobiose hydrolysis than salicylic acid and phthalic acid. Therefore, it can be inferred that proximity of functional groups may be more important for cellulose hydrolysis than their acidic strength.
Figure 7. Schematic of continuous flow slurry process for hydrolysis of mix-milled cellulose and Eucalyptus to soluble β-1,4-glucans and monomeric sugars. Reprinted with permission from ref 59. Copyright 2017 American Chemical Society. F
DOI: 10.1021/acs.accounts.7b00614 Acc. Chem. Res. XXXX, XXX, XXX−XXX
Article
Accounts of Chemical Research
(3) Mika, L. T.; Cséfalvay, E.; Németh, Á . Catalytic Conversion of Carbohydrates to Initial Platform Chemicals: Chemistry and Sustainability. Chem. Rev. 2018, 118, 505−613. (4) Krässig, H.; Schurz, J.; Steadman, R. G.; Schliefer, K.; Albrecht, W.; Mohring, M.; Schlosser, H. Cellulose. Ullmann’s Encyclopedia of Industrial Chemistry; Wiley-VCH Verlag GmbH & Co. KGaA: 2000. (5) Alonso, D. M.; Bond, J. Q.; Dumesic, J. A. Catalytic Conversion of Biomass to Biofuels. Green Chem. 2010, 12, 1493−1513. (6) Rinaldi, R.; Schüth, F. Acid Hydrolysis of Cellulose as the Entry Point into Biorefinery Schemes. ChemSusChem 2009, 2, 1096−1107. (7) Kobayashi, H.; Komanoya, T.; Hara, K.; Fukuoka, A. WaterTolerant Mesoporous-Carbon-Supported Ruthenium Catalysts for the Hydrolysis of Cellulose to Glucose. ChemSusChem 2010, 3, 440−443. (8) Kobayashi, H.; Komanoya, T.; Guha, S. K.; Hara, K.; Fukuoka, A. Conversion of Cellulose into Renewable Chemicals by Supported Metal Catalysis. Appl. Catal., A 2011, 409−410, 13−20. (9) Faith, W. L. Development of the Scholler Process in the United States. Ind. Eng. Chem. 1945, 37, 9−11. (10) Balandin, A. A.; Vasyunina, N. A.; Barysheva, G. S.; Chepigo, S. V.; Dubinin, M. M. Hydrogenation Catalysts for Polysaccharides. Bull. Acad. Sci. USSR, Div. Chem. Sci. (Engl. Transl.) 1957, 6, 403. (11) Fukuoka, A.; Dhepe, P. L. Catalytic Conversion of Cellulose into Sugar Alcohols. Angew. Chem., Int. Ed. 2006, 45, 5161−5163. (12) Kobayashi, H.; Ito, Y.; Komanoya, T.; Hosaka, Y.; Dhepe, P. L.; Kasai, K.; Hara, K.; Fukuoka, A. Synthesis of Sugar Alcohols by Hydrolytic Hydrogenation of Cellulose over Supported Metal Catalysts. Green Chem. 2011, 13, 326−333. (13) Negahdar, L.; Oltmanns, J. U.; Palkovits, S.; Palkovits, R. Kinetic Investigation of the Catalytic Conversion of Cellobiose to Sorbitol. Appl. Catal., B 2014, 147, 677−683. (14) Ravenelle, R. M.; Copeland, J. R.; Kim, W.-G.; Crittenden, J. C.; Sievers, C. Structural Changes of γ-Al2O3-Supported Catalysts in Hot Liquid Water. ACS Catal. 2011, 1, 552−561. (15) Ravenelle, R. M.; Copeland, J. R.; Van Pelt, A. H.; Crittenden, J. C.; Sievers, C. Stability of Pt/γ-Al2O3 Catalysts in Model Biomass Solutions. Top. Catal. 2012, 55, 162−174. (16) Luo, C.; Wang, S.; Liu, H. Cellulose Conversion into Polyols Catalyzed by Reversibly Formed Acids and Supported Ruthenium Clusters in Hot Water. Angew. Chem., Int. Ed. 2007, 46, 7636−7639. (17) Deng, W.; Tan, X.; Fang, W.; Zhang, Q.; Wang, Y. Conversion of Cellulose into Sorbitol over Carbon Nanotube-Supported Ruthenium Catalyst. Catal. Lett. 2009, 133, 167−174. (18) Kobayashi, H.; Yamakoshi, Y.; Hosaka, Y.; Yabushita, M.; Fukuoka, A. Production of Sugar Alcohols from Real Biomass by Supported Platinum Catalyst. Catal. Today 2014, 226, 204−209. (19) Yamaguchi, A.; Sato, O.; Mimura, N.; Hirosaki, Y.; Kobayashi, H.; Fukuoka, A.; Shirai, M. Direct Production of Sugar Alcohols from Wood Chips Using Supported Platinum Catalysts in Water. Catal. Commun. 2014, 54, 22−26. (20) Yamaguchi, A.; Sato, O.; Mimura, N.; Shirai, M. Catalytic Production of Sugar Alcohols from Lignocellulosic Biomass. Catal. Today 2016, 265, 199−202. (21) Kobayashi, H.; Hosaka, Y.; Hara, K.; Feng, B.; Hirosaki, Y.; Fukuoka, A. Control of Selectivity, Activity and Durability of Simple Supported Nickel Catalysts for Hydrolytic Hydrogenation of Cellulose. Green Chem. 2014, 16, 637−644. (22) Kobayashi, H.; Matsuhashi, H.; Komanoya, T.; Hara, K.; Fukuoka, A. Transfer Hydrogenation of Cellulose to Sugar Alcohols over Supported Ruthenium Catalysts. Chem. Commun. 2011, 47, 2366−2368. (23) Komanoya, T.; Kobayashi, H.; Hara, K.; Chun, W.-J.; Fukuoka, A. Kinetic Study of Catalytic Conversion of Cellulose to Sugar Alcohols under Low-Pressure Hydrogen. ChemCatChem 2014, 6, 230−236. (24) Michel, C.; Gallezot, P. Why Is Ruthenium an Efficient Catalyst for the Aqueous-Phase Hydrogenation of Biosourced Carbonyl Compounds? ACS Catal. 2015, 5, 4130−4132. (25) Van de Vyver, S.; Geboers, J.; Dusselier, M.; Schepers, H.; Vosch, T.; Zhang, L.; Van Tendeloo, G.; Jacobs, P. A.; Sels, B. F.
hydrogen. Hydrolysis of cellulose with carbon catalyst containing weakly acidic oxygenated functional groups eliminates the cost associated with use of mineral acids. Prior adsorption of cellulose on carbon catalyst increases the reaction rate and facilitates the use of a continuous flow slurry process. The yields of glucose and sorbitol from cellulose conversion have maximized, and further research is required to reduce the process costs. Fractionation of lignocellulose and pretreatment of cellulose are the most cost intensive steps in this process, and these issues must be overcome for the success of cellulose conversion technologies.
■
AUTHOR INFORMATION
Corresponding Author
*
[email protected]. ORCID
Abhijit Shrotri: 0000-0001-9850-7325 Hirokazu Kobayashi: 0000-0001-8559-6509 Atsushi Fukuoka: 0000-0002-8468-7721 Funding
This work was supported by Japan Society for the Promotion of Science (JSPS) KAKENHI (Nos. 20226016 and 23760734) and by Japan Science and Technology Agency (JST) ALCA. Notes
The authors declare no competing financial interest. Biographies Abhijit Shrotri is an Assistant Professor at the Institute for Catalysis in Hokkaido University. He received his Ph.D. in Chemical Engineering from the University of Queensland under the supervision of Dr. Jorge Beltramini in 2014. He worked as a postdoctoral fellow at the Catalysis Research Center in Hokkaido University from 2014 to 2016. His research interest includes biomass conversion using heterogeneous catalysts and design of functional carbon catalysts. Hirokazu Kobayashi earned his Ph.D. in Tokyo Institute of Technology under the supervision of Prof. Ichiro Yamanaka. Then, he moved to Catalysis Research Center, Hokkaido University, as an Assistant Professor. He received Chemical Society of Japan Award for Young Chemists for 2017. He is now studying heterogeneous catalysis for the utilization of biomass and oxidation of alkanes. Atsushi Fukuoka is a Professor at the Institute for Catalysis (ICAT) in Hokkaido University. He studied homogeneous catalysis and received a Ph.D. from the University of Tokyo in 1989. Then he joined Catalysis Research Center (CRC) and started research on heterogeneous catalysis. Since 2010, he had served as Director of CRC, and now he is Advisor to the President of Hokkaido University. He received a Society Award from The Catalysis Society of Japan in 2015 and GSC Award from Ministry of Education, Culture, Sports, Science and Technology, Japan, in 2015. He is an executive council member, Officer, of the International Association of Catalysis Societies. His current research interests are in biomass conversion by heterogeneous catalysts and the catalysis of mesoporous materials.
■
REFERENCES
(1) Klemm, D.; Heublein, B.; Fink, H.-P.; Bohn, A. Cellulose: Fascinating Biopolymer and Sustainable Raw Material. Angew. Chem., Int. Ed. 2005, 44, 3358−3393. (2) Kobayashi, H.; Fukuoka, A. Synthesis and Utilisation of Sugar Compounds Derived from Lignocellulosic Biomass. Green Chem. 2013, 15, 1740−1763. G
DOI: 10.1021/acs.accounts.7b00614 Acc. Chem. Res. XXXX, XXX, XXX−XXX
Article
Accounts of Chemical Research Selective Bifunctional Catalytic Conversion of Cellulose over Reshaped Ni Particles at the Tip of Carbon Nanofibers. ChemSusChem 2010, 3, 698−701. (26) Takenaka, S.; Ogihara, H.; Otsuka, K. Structural Change of Ni Species in Ni/SiO2 Catalyst During Decomposition of Methane. J. Catal. 2002, 208, 54−63. (27) Ding, L. N.; Wang, A. Q.; Zheng, M. Y.; Zhang, T. Selective Transformation of Cellulose into Sorbitol by Using a Bifunctional Nickel Phosphide Catalyst. ChemSusChem 2010, 3, 818−821. (28) Yang, P.; Kobayashi, H.; Hara, K.; Fukuoka, A. Phase Change of Nickel Phosphide Catalysts in the Conversion of Cellulose into Sorbitol. ChemSusChem 2012, 5, 920−926. (29) Liao, Y.; Liu, Q.; Wang, T.; Long, J.; Zhang, Q.; Ma, L.; Liu, Y.; Li, Y. Promoting Hydrolytic Hydrogenation of Cellulose to Sugar Alcohols by Mixed Ball Milling of Cellulose and Solid Acid Catalyst. Energy Fuels 2014, 28, 5778−5784. (30) Van de Vyver, S.; Geboers, J.; Schutyser, W.; Dusselier, M.; Eloy, P.; Dornez, E.; Seo, J. W.; Courtin, C. M.; Gaigneaux, E. M.; Jacobs, P. A.; Sels, B. F. Tuning the Acid/Metal Balance of Carbon Nanofiber-Supported Nickel Catalysts for Hydrolytic Hydrogenation of Cellulose. ChemSusChem 2012, 5, 1549−1558. (31) Liang, G.; He, L.; Arai, M.; Zhao, F. The Pt-Enriched PtNi Alloy Surface and Its Excellent Catalytic Performance in Hydrolytic Hydrogenation of Cellulose. ChemSusChem 2014, 7, 1415−1421. (32) Hilgert, J.; Meine, N.; Rinaldi, R.; Schüth, F. Mechanocatalytic Depolymerization of Cellulose Combined with Hydrogenolysis as a Highly Efficient Pathway to Sugar Alcohols. Energy Environ. Sci. 2013, 6, 92−96. (33) Shrotri, A.; Lambert, L. K.; Tanksale, A.; Beltramini, J. Mechanical Depolymerisation of Acidulated Cellulose: Understanding the Solubility of High Molecular Weight Oligomers. Green Chem. 2013, 15, 2761−2768. (34) Liang, G.; He, L.; Cheng, H.; Li, W.; Li, X.; Zhang, C.; Yu, Y.; Zhao, F. The Hydrogenation/Dehydrogenation Activity of Supported Ni Catalysts and Their Effect on Hexitols Selectivity in Hydrolytic Hydrogenation of Cellulose. J. Catal. 2014, 309, 468−476. (35) Shrotri, A.; Tanksale, A.; Beltramini, J. N.; Gurav, H.; Chilukuri, S. V. Conversion of Cellulose to Polyols over Promoted Nickel Catalysts. Catal. Sci. Technol. 2012, 2, 1852−1858. (36) Zhang, B.; Li, X.; Wu, Q.; Zhang, C.; Yu, Y.; Lan, M.; Wei, X.; Ying, Z.; Liu, T.; Liang, G.; Zhao, F. Synthesis of Ni/Mesoporous ZSM-5 for Direct Catalytic Conversion of Cellulose to Hexitols: Modulating the Pore Structure and Acidic Sites Via a Nanocrystalline Cellulose Template. Green Chem. 2016, 18, 3315−3323. (37) Geboers, J.; Van de Vyver, S.; Carpentier, K.; de Blochouse, K.; Jacobs, P.; Sels, B. Efficient Catalytic Conversion of Concentrated Cellulose Feeds to Hexitols with Heteropoly Acids and Ru on Carbon. Chem. Commun. 2010, 46, 3577−3579. (38) Palkovits, R.; Tajvidi, K.; Procelewska, J.; Rinaldi, R.; Ruppert, A. Hydrogenolysis of Cellulose Combining Mineral Acids and Hydrogenation Catalysts. Green Chem. 2010, 12, 972−978. (39) Geboers, J.; Van de Vyver, S.; Carpentier, K.; Jacobs, P.; Sels, B. Efficient Hydrolytic Hydrogenation of Cellulose in the Presence of RuLoaded Zeolites and Trace Amounts of Mineral Acid. Chem. Commun. 2011, 47, 5590−5592. (40) Bobleter, O. Hydrothermal Degradation of Polymers Derived from Plants. Prog. Polym. Sci. 1994, 19, 797−841. (41) Wang, D.; Astruc, D. The Golden Age of Transfer Hydrogenation. Chem. Rev. 2015, 115, 6621−6686. (42) Shrotri, A.; Kobayashi, H.; Tanksale, A.; Fukuoka, A.; Beltramini, J. Transfer Hydrogenation of Cellulose-Based Oligomers over Carbon-Supported Ruthenium Catalyst in a Fixed-Bed Reactor. ChemCatChem 2014, 6, 1349−1356. (43) Kobayashi, H.; Yokoyama, H.; Feng, B.; Fukuoka, A. Dehydration of Sorbitol to Isosorbide over H-Beta Zeolites with High Si/Al Ratios. Green Chem. 2015, 17, 2732−2735. (44) Otomo, R.; Yokoi, T.; Tatsumi, T. Synthesis of Isosorbide from Sorbitol in Water over High-Silica Aluminosilicate Zeolites. Appl. Catal., A 2015, 505, 28−35.
(45) Yokoyama, H.; Kobayashi, H.; Hasegawa, J.; Fukuoka, A. Selective Dehydration of Mannitol to Isomannide over Hβ Zeolite. ACS Catal. 2017, 7, 4828−4834. (46) Suganuma, S.; Nakajima, K.; Kitano, M.; Yamaguchi, D.; Kato, H.; Hayashi, S.; Hara, M. Hydrolysis of Cellulose by Amorphous Carbon Bearing SO3H, COOH, and OH Groups. J. Am. Chem. Soc. 2008, 130, 12787−12793. (47) Onda, A.; Ochi, T.; Yanagisawa, K. Selective Hydrolysis of Cellulose into Glucose over Solid Acid Catalysts. Green Chem. 2008, 10, 1033−1037. (48) Pang, J.; Wang, A.; Zheng, M.; Zhang, T. Hydrolysis of Cellulose into Glucose over Carbons Sulfonated at Elevated Temperatures. Chem. Commun. 2010, 46, 6935−6937. (49) Van de Vyver, S.; Peng, L.; Geboers, J.; Schepers, H.; de Clippel, F.; Gommes, C. J.; Goderis, B.; Jacobs, P. A.; Sels, B. F. Sulfonated Silica/Carbon Nanocomposites as Novel Catalysts for Hydrolysis of Cellulose to Glucose. Green Chem. 2010, 12, 1560−1563. (50) Takagaki, A.; Nishimura, M.; Nishimura, S.; Ebitani, K. Hydrolysis of Sugars Using Magnetic Silica Nanoparticles with Sulfonic Acid Groups. Chem. Lett. 2011, 40, 1195−1197. (51) Komanoya, T.; Kobayashi, H.; Hara, K.; Chun, W.-J.; Fukuoka, A. Catalysis and Characterization of Carbon-Supported Ruthenium for Cellulose Hydrolysis. Appl. Catal., A 2011, 407, 188−194. (52) Kobayashi, H.; Yabushita, M.; Komanoya, T.; Hara, K.; Fujita, I.; Fukuoka, A. High-Yielding One-Pot Synthesis of Glucose from Cellulose Using Simple Activated Carbons and Trace Hydrochloric Acid. ACS Catal. 2013, 3, 581−587. (53) Kobayashi, H.; Kaiki, H.; Shrotri, A.; Techikawara, K.; Fukuoka, A. Hydrolysis of Woody Biomass by a Biomass-Derived Reusable Heterogeneous Catalyst. Chem. Sci. 2016, 7, 692−696. (54) Shrotri, A.; Kobayashi, H.; Fukuoka, A. Air Oxidation of Activated Carbon to Synthesize a Biomimetic Catalyst for Hydrolysis of Cellulose. ChemSusChem 2016, 9, 1299−1303. (55) Yabushita, M.; Kobayashi, H.; Hara, K.; Fukuoka, A. Quantitative Evaluation of Ball-Milling Effects on the Hydrolysis of Cellulose Catalysed by Activated Carbon. Catal. Sci. Technol. 2014, 4, 2312−2317. (56) Yabushita, M.; Kobayashi, H.; Hasegawa, J.; Hara, K.; Fukuoka, A. Entropically Favored Adsorption of Cellulosic Molecules onto Carbon Materials through Hydrophobic Functionalities. ChemSusChem 2014, 7, 1443−1450. (57) Chung, P.-W.; Yabushita, M.; To, A. T.; Bae, Y.; Jankolovits, J.; Kobayashi, H.; Fukuoka, A.; Katz, A. Long-Chain Glucan Adsorption and Depolymerization in Zeolite-Templated Carbon Catalysts. ACS Catal. 2015, 5, 6422−6425. (58) Kobayashi, H.; Yabushita, M.; Hasegawa, J.; Fukuoka, A. Synergy of Vicinal Oxygenated Groups of Catalysts for Hydrolysis of Cellulosic Molecules. J. Phys. Chem. C 2015, 119, 20993−20999. (59) Shrotri, A.; Kobayashi, H.; Kaiki, H.; Yabushita, M.; Fukuoka, A. Cellulose Hydrolysis Using Oxidized Carbon Catalyst in a Plug Flow Slurry Process. Ind. Eng. Chem. Res. 2017, 56, 14471−14478.
H
DOI: 10.1021/acs.accounts.7b00614 Acc. Chem. Res. XXXX, XXX, XXX−XXX
