Biomass to Furanics: Renewable Routes to Chemicals and Fuels

Subscriber access provided by The University Library | University of Sheffield

Perspective

Biomass to furanics: Renewable routes to chemicals and fuels Benjamin R. Caes, Rodrigo E. Teixeira, Kurtis G. Knapp, and Ronald T Raines ACS Sustainable Chem. Eng., Just Accepted Manuscript • DOI: 10.1021/ acssuschemeng.5b00473 • Publication Date (Web): 20 Sep 2015 Downloaded from http://pubs.acs.org on September 28, 2015

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Sustainable Chemistry & Engineering is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 65

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sustainable Chemistry & Engineering

Biomass to furanics: Renewable routes to chemicals and fuels Benjamin R. Caes,†,‡,* Rodrigo E. Teixeira,‡,|| Kurtis G. Knapp,|| and Ronald T. Raines†,‡,ł,** †

Department of Chemistry, University of Wisconsin–Madison, 1101 University Avenue,

Madison, Wisconsin 53706-1322, United States ‡

DOE Great Lakes Bioenergy Research Center, Wisconsin Energy Institute, 1552 University

Avenue, Madison, Wisconsin 53726, United States ||

Hyrax Energy, Inc., 3475 Edison Way–Suite N, Menlo Park, California 94025, United States

łDepartment

of Biochemistry, University of Wisconsin–Madison, 433 Babcock Drive,

Madison, Wisconsin 55706-1544, United States

**R. T. Raines. E-mail: [email protected].

ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Caes et al.

Page 2 of 65

page 2

ABSTRACT: The quest to achieve a sustainable supply of both energy and chemicals is one of the great challenges of this century. 5-(Hydroxymethyl)furfural (HMF), the long-known dehydration product of hexose carbohydrates, has become an important nexus for access to both liquid fuels and chemicals. One such biofuel is 2,5-dimethylfuran (DMF), which is a product of HMF hydrogenolysis and contains an energy density 40% greater than that of ethanol. In recent years, much work has been done to effect the chemical conversion of fructose, glucose, cellulose, and even lignocellulosic biomass into HMF in high yield. Here, we provide an overview of known methods to access HMF from carbohydrates, along with a discussion of unmet technological needs.

KEYWORDS: Cellulose, Chemurgy, 2,5-Dimethylfuran, 5-(Hydroxymethyl)furfural, Ionic liquid, Lignocellulose


Page 3 of 65

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Caes et al.

page 3

INTRODUCTION In the early 20th century, William Jay Hale was a member of the faculty at the University of Michigan and later a scientist at Dow Chemical (Figure 1). During most of his professional life, Hale advocated the use of agricultural carbohydrates as industrial feedstocks for hundreds of commercial products. In the 1930s, he coined the term “chemurgy” from Greek expressions meaning “chemistry at work”. According to Hale,1 “Chemurgy brings out in relief the correct interpretation of agriculture. No longer a pursuit to supply man with food and raiment, but a pursuit that shall bring into existence a vast array of chemical compounds to fit a myriad of needs…Both the lignin and cellulose, out of wood, admirably fit into this picture.” Hale believed that ethanol would eventually become the greatest agricultural product.2 These prescient thoughts preceded by nearly a century alcohol reaching cost-parity with gasoline in Brazil. The use of chemistry to transform the farm gained early traction, but fell out of favor in the 1950s due to the emergence of inexpensive petroleum-derived products.3,4 The term “chemurgy” was never adopted broadly. Since the Neolithic Revolution, humans have been rudimentary chemurgists. We relied on animal power, agriculturally derived chemicals and materials, and biomass burning for heat. More recently, the Industrial Revolution was unleashed by tapping into vast deposits of coal and then petroleum. Now, two centuries later, the carbon cycle has been altered on a sufficiently short timescale to pose an imminent threat to micro- and macro-environments.5 Yet, the demand for petroleum continues to grow. Meanwhile, the Earth can generate enough biomass in 10 years to match the energy content of all existing petroleum reserves.6 Hence, technologies that convert biomass into fuels and chemicals on massive scales could play a leading role in the next chapter of human civilization. The cellulose in biomass is the most abundant organic molecule on Earth. Versatile derivatives of furan, such as 5-(hydroxymethyl)furfural (HMF), can be accessed from



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Caes et al.

Page 4 of 65

page 4

cellulose without breaking a single C–C bond (Figure 2). More specifically, HMF can be accessed from the furanose form of hexose sugars by three dehydration reactions (Figure 3),7,8 and forms spontaneously when sugars are heated, especially under acidic conditions. HMF has been detected in foods such as coffee and dried fruits,9 and is prevalent in baked goods;10 its daily intake is estimated to be 30–150 mg per person.11 Although HMF can be toxic to humans when ingested at high concentrations, multi-gram doses of HMF as “Aes-103” (Baxter International, Inc.) are being administered to sickle-cell disease patients in a Phase 2 clinical trial. Here, we focus on the potential of HMF to become a platform for industrial fuels and chemicals. HMF can be transformed into a large number of useful compounds through simple chemical reactions (Figure 4).12-16 Rehydration of HMF causes it to decompose into levulinic and formic acids,17-21 both valuable commodity chemicals. Levulinic acid is a precursor to the liquid fuel γ-valerolactone,22-24 which can also solubilize biomass.25-28 Oxidation of the hydroxymethyl group leads to 2,5-diformylfuran and 2,5-furandicarboxylic acid. The latter, which has been accessed by electrochemical oxidation of HMF,29 is of particular interest for its use in the manufacturing of polyamides, polyesters, and polyurethanes as a substitute for terephthalic and isophthalic acid polymers.30 Reduction of the formyl moiety accesses 2,5-bis(hydroxymethyl)furan, whereas hydrogenation forms 2,5-bis(hydroxymethyl)tetrahydrofuran and 2,5-diformyltetrahydrofuran. These products can undergo condensation to generate polymers that can ultimately become liquid alkanes.31-33 Finally, hydrogenolysis yields 2,5-dimethylfuran (DMF), 2,5-dimethyltetrahydrofuran, and 2-methyltetrahydrofuran. With its high energy density, low volatility, and immiscibility with water, DMF has evoked substantial interest as a fuel.34-36 HMF and its derivatives hold vast potential and could eventually match the array of compounds derived from petroleum, allowing biomass to become a viable source of energy


Page 5 of 65

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Caes et al.

page 5

and materials. To reach this goal, conversion technologies must reach high yields without cost-prohibitive catalysts or processes that are too complex. Many investigators have manifested the conversion of mono- and polysaccharides into HMF in high yields, but no single route has prevailed to date. Diverse aspects of HMF production and utilization have been reviewed elsewhere.15,16,36-45 Herein, we focus on chemical strategies for accessing HMF that we believe hold special promise. In addition, we highlight routes to the candidate biofuel: DMF.

SOLVENTS Carbohydrate polymers can form inter- and intrastrand hydrogen bonds, making them more recalcitrant to dissolution, and hence inaccessible for chemical reactions. A select few solvents are able to accomplish dissolution.46-49 Of these privileged solvents, ionic liquids are especially noteworthy. Ionic liquids are organic salts that melt near room temperature, typically below 100 °C.5052

As salts, they have a negligible vapor pressure. The binary combination, resulting from the

pairing of a cation (e.g., ammonium, imidazolium, phosphonium, or pyridinium) with an anion (e.g., acetate, bromide, chloride, or hexafluorophosphate), allows access to a staggering number of species.51 This variety allows for the customization of properties for specific tasks. One such task is the dissolution of polysaccharides.53-56 For a solvent to dissolve cellulose, it must out-compete the hydrogen-bond network. Ionic liquids do so by manifesting a high concentration of anionic species that form more stable hydrogen bonds with the hydroxyl groups of cellulose.53 Dialkylimidazolium and dialkylpyridinium chloride ionic liquids are especially well known for their capacity to dissolve concentrations of cellulose at significant concentrations.46,54,57 The reuse of an ionic liquid after a bioconversion process mitigates its cost,58,59 and the addition of a fluorous tag to



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Caes et al.

Page 6 of 65

page 6

an ionic liquid can facilitate separations.60 Moreover, ionic liquids themselves can catalyze the transformation of hexose sugars into HMF,61 but yields are enhanced by exogenous catalysts, as described below. In addition to serving as the solvent for biomass conversions, ionic liquids also dissolve and extract lipids from algae62 and serve as the basis for pretreatment strategies.63-65

CATALYSTS Mineral Acids. The use of common mineral acids (e.g., hydrochloric, sulfuric, and

phosphoric acids) to catalyze the conversion of sugars to furanics has a long history. This approach had its origins in the so-called “furfuraldehyde tests” for carbohydrates that were developed in the late 19th century. The first, Molisch’s test, is based on pentoses and hexoses being dehydrated by sulfuric acid to form furfural and HMF, respectively, which condense with α-naphthol to form a purple dye. Seliwanoff’s test for ketoses relies on ketoses (like fructose) undergoing hydrochloric acid-catalyzed dehydration to yield furfural derivatives more rapidly than do aldoses (like glucose). These derivatives react with resorcinol to form a xanthine dye with a deep red color. In Bial’s test, orcinal instead of resorcinol and added FeCl3 is used to distinguish between pentoses and hexoses. In 1895, HMF was synthesized with intent from inulin by Düll66 and from sugar cane by Kiermayer67 using oxalic acid as a catalyst. These initial reports led to more extensive work by Fenton and co-workers.68-70 In 1919, Middendorp reported a detailed study on the synthesis, physical characteristics, and chemical behavior of HMF.71 In subsequent years, Reichstein72 and Haworth and Jones73 made immense contributions by putting forth a detailed process and chemical mechanism for the acid-catalyzed formation of HMF from fructose.


Page 7 of 65

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Caes et al.

page 7

In recent years, numerous advancements have been made in the use of mineral acids to transform carbohydrates (Table 1). Using H2SO4, Antal and co-workers reported the dehydration of fructose in subcritical water at 250 °C to access HMF in 53% yield.7 Moreover, they were able to conclude that fructose dehydrates to HMF through closed-ring intermediates due to the ease of HMF formation from fructose and the fructosyl moiety of sucrose, as well as the facile conversion of 2,5-anhydro-D-mannose to HMF. Grin′ and coworkers observed a kinetic isotope effect using HCl and D2O for the dehydration of fructose to HMF, indicating that a proton was transferred in the rate-limiting step of the reaction.74 Kuster and co-workers used H2SO4 to access HMF from the fructose acetonide 1,2:4,5-di-oisopropylidene-β-D-fructopyranose (prepared via the condensation of fructose with acetone).75 As fructose has limited solubility in alcohols, transformation to the fructose acetonide enabled the volatile ethylene glycol dimethyl ether to be used as a solvent, providing a facile means for HMF recovery and solvent recycling. Dumesic and co-workers employed biphasic reaction systems, both to form HMF from fructose and to extract the HMF with an organic solvent. Using HCl as a catalyst, fructose was dehydrated at 85% selectivity in an aqueous phase containing dimethylsulfoxide (DMSO) and poly(1-vinyl-2-pyrrolidinone) (PVP), which were added to suppress unwanted side reactions.76 HMF was extracted continuously into an organic phase of methylisobutylketone (MIBK), which was phase-modified with 2-butanol to improve the partitioning of the HMF from the aqueous phase. These workers were also able to adjust the biphasic system conditions for monomeric sugars and thereby achieve HMF selectivities of 89% and 53% for fructose and glucose, respectively, and 91% selectivity for furfural from xylose.77 Furthermore, these optimal conditions could be applied toward disaccharides sucrose (a glucose–fructose dimer) and cellobiose (a glucose dimer), along with the polysaccharides inulin (a polyfructan), starch (a polyglucan), and xylan (a polyxylose). HMF



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Caes et al.

Page 8 of 65

page 8

selectivities were 77% from sucrose, 52% from cellobiose, 77% from inulin, and 43% from starch, with a furfural selectivity of 66% from xylan. Finally, the use of inorganic salts was shown to increase the partitioning of HMF in biphasic systems (NaCl being the most beneficial), and tetrahydrofuran (THF) was shown to have a superior extraction ability for HMF, with an attained selectivity of 83%.78 The Yoshida group investigated the use of a variety acids in subcritical water to dehydrate fructose.79 They found that an HMF yield of 65% was attained using H3PO4, although HMF was also formed even in the absence of acid catalysts at increased temperatures. They also used HCl to study the dehydration of fructose to HMF followed by a rehydration to form levulinic and formic acids in subcritical water, as well as the formation of decomposition products.80 They determined that soluble polymer byproducts were formed not only from fructose, but also from HMF, and no soluble polymers were formed from levulinic and formic acids. Vogel and co-workers used H2SO4 to catalyze the dehydration of fructose in sub- and supercritical acetone–water mixtures,81 accessing HMF in 77% yield. They also used their system to access HMF from glucose (48%), sucrose (56%), and inulin (78%). Riisager and co-workers used HCl to catalyze a microwave-assisted dehydration of concentrated aqueous fructose solutions to access HMF.82 Water, which normally results in low HMF yields when used as the reaction medium, performed substantially better with microwave irradiation to give HMF rapidly with selectivity of 63% and fructose conversion of 52%. A reaction time of 60 s gave 95% fructose conversion, but resulted in a decreased HMF selectivity of 55%. Loebbecke and co-workers also used HCl to dehydrate fructose in an aqueous solution at increased temperatures and pressures (185 °C and 17 bar) in a microreactor.83 Whereas an HMF yield of 54% was obtained in 1 min, addition of a DMSO as a co-solvent and extracting with a MIBK/2-butanol mixture increased the HMF yield to 83%. Unfortunately, HMF is difficult to separate from DMSO. Raines and co-workers


Page 9 of 65

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Caes et al.

page 9

showed that a versatile industrial solvent, sulfolane, supports the HBr-catalyzed conversion of fructose to HMF in 93% yield in 1 h.84 Acid catalysis, while serving as a common method to access HMF from carbohydrates, also hydrolyzes cellulosic polysaccharides to monomeric sugars. The Amarasekara group used the acidic ionic liquids 1-(1-propylsulfonic)-3-methylimidazolium chloride and 1-(1butylsulfonic)-3-methylimidazolium chloride to dissolve and hydrolyze cellulose to glucose and other reducing sugars in the presence of water.85 The highest yields for reducing sugars (62%) and, specifically, glucose (14%) were obtained with 1 h of preheating at 70 °C and continued heating for 30 min after water addition. Seddon and co-workers used H2SO4 and methanesulfonic acid to study the kinetics of cellobiose hydrolysis, and observed a linear dependence on acid concentration.86 The maximal amount of glucose obtained from cellobiose was 68%, though degradation of glucose occurred in the presence of the strong acids. Hydrolysis of lignocellulosic biomass (miscanthus grass) gave only 5% glucose; though if lignin were extracted first, 30% yields were obtainable, as were 25% yields of xylose. In another study, Binder and Raines used HCl in the ionic liquid [EMIM]Cl to accomplish the hydrolysis of corn stover, another lignocellulosic biomass source.87 When untreated biomass was hydrolyzed with 20 wt% HCl with gradual water addition to enhance glucose stability, the yields of glucose and xylose were 42% and 71%, respectively. A second stage hydrolysis of the precipitated, unhydrolyzed residues allowed more sugar to be accessed, with final yields of 70% for glucose and 79% for xylose. Those yields were increased later by the same group to 92% for glucose and 95% for xylose after reoptimization of HCl concentration, gradual water addition in [BMIM]Cl, and separation of the products by simulated moving bed chromatography.88 The Dumesic group achieved 44– 48% and 57–59% yields of glucose and xylose, respectively, in biomass-derived γ-valerolactone containing water and 0.05 wt% H2SO4.28



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Caes et al.

Page 10 of 65

page 10

Although HMF has great potential to serve as a platform chemical, a related compound, 5-(chloromethyl)furfural (CMF), is also valuable for its ability to form furanic ethers, which are used as diesel additives. The Mascal group discovered a method for its conversion and isolation using a two-phase reaction medium.89 By heating microcrystalline cellulose with 5 wt% LiCl in concentrated HCl at 65 °C, CMF was obtained as the major product. Continuous extraction in 1,2-dichloroethane gave a total recovered CMF yield of 71%. Amino Acids and Enzymes. Two recent contributions borrowed from biochemistry. Su,

Li, Cheng, and their co-workers showed that the conversion of sucrose to HMF in ionic liquids can be catalyzed by amino acids, with tyrosine being the best.90 A maximum yield of 76% HMF was obtained in [EMIM]Br after 4 h at 160 °C. The authors suggested that tyrosine lends a proton to catalyze the hydrolysis of sucrose, while the unprotonated amino group helps to isomerize glucose to fructose. The Afonso group converted glucose to HMF by first isomerizing to fructose in water using sweetzyme, an industrial enzyme used in the production of high-fructose corn syrup. Subsequent dehydration to HMF was investigated in the presence of a number of Brønsted acid catalysts. By using H2SO4 and HNO3 catalysts, yields reached 80% and 87%, respectively.91 Boric and Boronic Acids. Recently, researchers utilized boron-containing compounds to

facilitate the transformation of carbohydrates to HMF. Boric acid, which is known to dehydrate alcohols,92,93 was shown in the Riisager laboratory to convert glucose into HMF in a 41% yield in [EMIM]Cl.94 Computational modeling suggested the formation of a glucose– borate complex to isomerize glucose into fructose though an enediol mechanism prior to dehydration to HMF. The Raines group developed an boron-containing organocatalyst for HMF production.95 Their 2-alkoxycarbonylphenyl boronic acids mediate the transformation of fructose, glucose, and cellulose (including cotton, paper towels, and newspaper) into HMF. Although catalyst


Page 11 of 65

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Caes et al.

page 11

loadings for rapid (~2 h) conversions are high, this liability is overcome by the facility of recovering the catalyst or, potentially, by its immobilization. Metal Ions. Metal ion catalysts are also used for the transformation of carbohydrates into

HMF (Table 2). A study by the Chohan group used a variety of metal chlorides to test the conversion of glucose to HMF.96 All metal ions tested resulted in an enhancement in the rate of HMF production relative to acid catalysis, which became even more pronounced at increased catalyst concentrations and increased temperatures. Another study done by Tyrlik and co-workers used aluminum salts in various oxygenated solvents to serve as ligands to the metal center for the conversion of glucose to HMF.97 The ligands were seen to influence both the yield and the selectivity of HMF formation, with ethanol proving the most efficient. The Abu-Omar group used AlCl3·6H2O in a biphasic system of water and THF to transform glucose into HMF in 65% yield.98 They also obtained good yields from other sugar sources such as maltose, cellobiose, starch, and cellulose. Nevertheless, their best HMF yields were only 35% from pine wood biomass materials, although furfural was obtained in yields greater than 60%. Yugen Zhang and his group used another biphasic system of THF and [BMIM]Cl to transform fructose into HMF.99 Using tungsten salts, notably WCl6, they were able to achieve conversion to HMF with 72% yield at 50 °C; even at room temperature, they achieved a 60% yield of HMF. Further, they demonstrated the recyclability of the catalyst. Chou and co-workers used MnCl2 in the acidic ionic liquid 1-(4-sulfonic acid) butyl-3methylimidazolium hydrogen sulfate ([SA-BMIM]HSO4) to transform cellulose into HMF.100 They reported HMF yields of 37%, although some levulinic acid was produced from the rehydration of the HMF product. Han and co-workers utilized SnCl4 to transform carbohydrate materials into HMF in the ionic liquid 1-ethyl-3-methylimidazolium tetrafluoroborate ([EMIM]BF4).101 Good HMF yields were obtained from fructose (62%) and



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Caes et al.

Page 12 of 65

page 12

glucose (61%). They further accessed HMF by utilizing their system on polysaccharides sucrose (65%), inulin (40%), cellobiose (57%), and starch (47%). They postulated that the formation of a five-membered ring chelate complex of the Sn with C1 and C2 hydroxyls of α-glucose facilitated an enolization of glucose to fructose before dehydration to HMF. Conrad Zhang and his group screened many metal chloride catalysts, and CrCl2 and CrCl3 emerged as best for the conversion of monosaccharides to HMF.102 In [EMIM]Cl, glucose and fructose were transformed in yields near 70%. This pioneering work established Cr(II) and Cr(III) as privileged catalysts for the conversion of carbohydrates to HMF. Binder and Raines used CrCl2 and CrCl3 in a solvent mixture of N,N-dimethylacetamide (DMA)–LiCl and [EMIM]Cl.103 Importantly, they discovered that their system was not only amenable to cellulose dissolution, but also enabled an HMF yield of 48% from untreated corn stover, along with a 37% yield of furfural. This work was the first “one-pot” conversion of crude biomass to a furanic. Further work by the same group demonstrated that the catalytic system could be applied to convert other sugars such as mannose, galactose, lactose, and tagatose into HMF,104 and to transform xylose and xylan into furfural.105 Fructose is a ketose and prefers to exist in a furanose form,106 which is a ready precursor to HMF. In contrast, glucose is an aldose and prefers a pyranose form. The Zhang group proposed that chromium ions catalyzed this key aldose-to-ketose transformation through a chromium enolate intermediate (Figure 5A).102 Later work by Hensen and co-workers demonstrated that a chromium ion could form a complex with sugar hydroxyl groups.107 Another plausible mechanism enlists chromium ions to expedite a 1,2-hydride shift (Figure 5B). Interestingly, known enzymic catalysts of hexose isomerization use each pathway.108 The enolization mechanism is used by phosphoglucose isomerase, which does not employ a metal cofactor, whereas the hydride-shift mechanism is used by xylose isomerase, the manganese-dependent enzyme used for the commercial isomerization of


Page 13 of 65

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Caes et al.

page 13

glucose to fructose. Deuterium-labeling studies by the Raines group indicated that chromium ion mediated a 1,2-hydride shift (Figure 5B).104,105 The success of chromium chloride catalysts to convert carbohydrates to HMF inspired attempts to enhance its transformative properties by pairings with other metal chlorides. Further work by the Zhang group used CrCl2 and CuCl2 in [EMIM]Cl to convert cellulose into HMF with yields consistently near 57%.109 HMF was extracted and the system maintained catalytic performance while being recycled for repeated use. Furthermore, they demonstrated that the bi-catalyst system was able to depolymerize cellulose more rapidly than typical acid hydrolysis procedures. Cho and co-workers paired CrCl2 with RuCl3 in [EMIM]Cl to access HMF from cellulose in yields of 60%, and from reed biomass in a 41% yield with a 26% furfural yield.110 A pairing of CrCl3 with ammonium halides, notably NH4Br, in DMA by Zhang and co-workers was used to transform sucrose, glucose, and fructose into HMF at 87%, 74%, and 92% yields.111 They observed a halide effect by varying chloride, bromide, and iodide as the anion of the ammonium salt for glucose conversion to HMF, with bromide being the best. This halide effect is consistent to that observed previously in the Raines laboratory.103 The Zhang and Ying groups used N-heterocyclic carbenes (NHC) as ligands to complex with Cr(II) and Cr(III) in an effort to enhance their catalytic activity.112 They were able to convert fructose and glucose into HMF at 96% and 81% yields using 1,3-bis(2,6diisopropylphenyl)imidazolylidene as the NHC ligand. Both the catalysts and the ionic liquid in this system could be recycled for continued use. More recently, Dumesic, Rinaldi, and coworkers combined AlCl3 catalyst with an aqueous-organic biphasic reactor and microwave heating. Interestingly, they started from beechwood, sugarcane bagasse, or cellulose, and impregnated the biomass with HCl before comminution. This procedure created oligosaccharides that were hydrolyzed quickly (in 6 min) into monosaccharides and



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Caes et al.

Page 14 of 65

page 14

converted into HMF. Yields of 60–69% for HMF and 74–84% for furfural were obtained depending on the biomass.113 Lanthanide metals can serve as Lewis acids in their ionic forms to catalyze the conversion of saccharide materials to HMF. Ishida and Seri found that glucose could be dehydrated by lanthanide(III) ions, potentially due to the high affinity of lanthanide ions for oxygen atoms.114 Still, the nascent HMF also suffered decomposition. In further work, they demonstrated that high yields of HMF (>90%) could be obtained from fructose using the chloride salts of lanthanides in DMSO.115 Similarly, all the lanthanide(III) ions were shown to dehydrate hexoses in water without rehydration to levulinic and formic acids.116 A kinetic analysis indicated that the rate-determining step in the dehydration is the reaction of the hexose–catalyst complex rather than the association of the two to form the complex. The activity of lanthanide catalysts has also been investigated in ionic liquids. The Riisager group demonstrated the conversion of glucose to HMF in ionic liquids with ytterbium chloride (YbCl3) and ytterbium triflate (Yb(OTf)3).117 They achieved their best yield for YbCl3 (23%) in 1-octyl-3-methylimidazolium chloride ([OMIM]Cl) and for Yb(OTf)3 (24%) in [BMIM]Cl. Although these yields are modest in comparison to chromium catalysts, the ytterbium catalysts tended to give the higher HMF yields in less hydrophilic ionic liquids, whereas chromium preferred more hydrophilic ionic liquids such as [EMIM]Cl. A study in the Okuda laboratory compared lanthanum chloride (LaCl3) to the rare earth salts ytterium chloride (YCl3) and scandium chloride (ScCl3) on conversion of glucose and cellobiose to HMF in the organic solvent DMA.118 They concluded that the conversion and selectivity depend on the ionic radii of the metal center, with the smaller radius giving greater activity. Heterogeneous Catalysts. Heterogeneous catalysts are also used to convert

carbohydrates into HMF (Table 3). Their use in conversion reactions can be advantageous for


Page 15 of 65

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Caes et al.

page 15

several reasons: facile product separation, catalyst recyclability, high temperature-tolerance, and modulation of surface properties (acidity, basicity, and pore size) to achieve maximal HMF selectivities and yields. Here, we discuss ion-exchange resins, heteropolyacids, zirconia compounds, and H-form zeolites. Smith and co-workers investigated the dehydration of fructose using DOWEX 50WX8 ion-exchange resin with microwave heating.119 Whereas initial work had used pure DMSO as the solvent, they were interested in using acetone as a co-solvent to serve as a promotor for higher HMF selectivities. In only 10 min, HMF yields could be obtained at 82% with this system, and the catalytic activity of the resin was maintained over multiple runs. In another study by the same group, Amberlyst-15 ion-exchange resin was used to dehydrate fructose in [BMIM]Cl with the addition of various co-solvents.120 All of the co-solvents tested (acetone, DMSO, methanol, ethanol, ethyl acetate, and supercritical CO2) accessed HMF yields of 78– 82%. Dumesic and co-workers achieved somewhat lower yields by using biomass-derived solvents (tetrahydrofuran, γ-valerolactone, γ-hexalactone, or methyltetrahydrofuran) to access HMF from glucose with a combination of Lewis (Sn-β) and Brønsted (Amberlyst-70)25 or HBeta zeolite solid catalysts.27 Lansalot-Matras and Moreau compared the effect of two ionic liquids, the hydrophilic 1butyl-3-methylimidazolium tetrafluoroborate ([BMIM]BF4) and the hydrophobic 1-butyl-3methylimidazolium hexafluorophosphate ([BMIM]PF6), on the dehydration of fructose using Amberlyst-15.121 Using DMSO as a co-solvent to assist with sugar solubility, they observed HMF yields near 80% in both systems. But without the DMSO co-solvent, HMF yields dropped to 52% in [BMIM]BF4, and fructose was not soluble in pure [BMIM]PF6. The Smith group was able to access HMF in yields as high as 83% using the same catalyst in [BMIM]Cl with microwave heating at 80 °C in only 10 min.122 When the temperature was increased to



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Caes et al.

Page 16 of 65

page 16

120 °C, a similar yield could be accessed in only 1 min. Furthermore, they demonstrated that 5 wt% water content or lower in the ionic liquid had no adverse effects. As HMF can be difficult to separate from ionic liquids, Huang and co-workers used the heteropolyacid salt of an ionic liquid cation as a catalyst to convert fructose in sec-butanol.123 Using 1-(3-sulfonic acid)propyl-3-methylimidazolium phosphotungstate ([MIMPS]3PW12O40), they achieved an HMF yield as high as 99% in 2 h. Furthermore, simple precipitation of the catalyst enabled its recycling in further reactions with no loss of activity. Xiaochong Wang and co-workers used another heteropolyacid salt, Ag3PW12O40, to transform both fructose and glucose into HMF.124 Using a biphasic system of water and MIBK to extract the HMF product, HMF yields of 78% and 76% were obtained from fructose and glucose, respectively. The catalyst was recycled easily, demonstrated no loss of activity over multiple reactions, and was tolerant to high feedstock concentrations with minimal byproduct formation. Further, Wang and co-workers used the Brønsted–Lewis– surfactant-combined heteropolyacid Cr[(OSO3C12H25)H2PW12O40]3 to achieve both the depolymerization of cellulose and its conversion to HMF.125 A 53% yield of HMF was obtained in 2 h at 150 °C from pure cellulose, and yields of 31% and 36% were obtained from corn stover and Xanthoceras sorbifolia Bunge husk (which is native to northern China), respectively. The catalyst existed as an emulsion during extraction of HMF using MIBK, allowing facile recovery for reuse. The group attributed the high activity of the catalyst to its dual Brønsted and Lewis acidities. Heterogeneous zirconia (ZrO2) compounds can facilitate dehydration of carbohydrates. Qi and co-workers impregnated ZrO2 with H2SO4 to serve as a catalyst for conversion of fructose to HMF in [BMIM]Cl.126 An 88% yield was obtained in only 30 min at 100 °C. The solid catalyst and ionic liquid were able to be recycled with constant activity for multiple runs. Another study by Qi, Smith, and co-workers coupled the sulfated zirconia with


Page 17 of 65

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Caes et al.

page 17

microwave heating to dehydrate fructose in an acetone–DMSO mixture.127 At 180 °C, an HMF yield of 73% was attained in 20 min. When the reaction was carried out in water, however, the HMF yield was only 36%, demonstrating that ZrO2 has little activity in aqueous systems. Hu and co-workers used an ethylene dichloride solution of chlorosulfonic acid to impregnate Zr(OH)4 (CSZ) and Zr(OH)4–Al(OH)3 (CSZA) to serve as catalysts for fructose and glucose conversion to HMF with maximal yields of 68% and 48%, respectively.128 The zirconium contained both Brønsted and Lewis acid sites, and the aluminum contained basic sites. They found that increased basicity promoted isomerization of glucose to fructose, and an ideal mole ratio of 1:1 Zr:Al resulted in the highest yield of HMF. The catalyst was also recoverable with minimal loss of activity. Inomata and co-workers used both TiO2 (anatase and rutile) and ZrO2 to convert glucose and fructose into HMF in hot-compressed water.129 They found that rutile TiO2 was inactive in glucose conversion, but both anatase TiO2 and ZrO2 were able to isomerize glucose into fructose. Anatase TiO2 also dehydrated fructose, suggesting that ZrO2 is a base catalyst and anatase TiO2 has both basic and acidic sites. McNeff and co-workers again used anatase TiO2 and ZrO2 in a fixed bed reactor to access HMF from a variety of carbohydrates (i.e., fructose, glucose, starch, and cellulose).130 An HMF yield of 35% was obtained from cellulose after extraction with MIBK, and the catalysts were regenerated once their activity diminished by heating to 450 °C for 5 h. Mesoporous TiO2 nanoparticles were used by Bhaumik, Saha, and co-workers to convert carbohydrate materials with microwave heating in aqueous and organic solvents.131 The maximal HMF yields from fructose were 34% and 54% in water and DMSO, respectively, and 37% from glucose in DMSO. The nanoparticles were shown to retain their catalytic activity over 4 cycles. The Moreau group employed dealuminated H-form mordenites (DAHM) in a biphasic system of water and MIBK to transform fructose into HMF.132 They achieved an HMF yield



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Caes et al.

Page 18 of 65

page 18

of 69% with the maximal reaction rate attained using H-mordenites with an 11:1 Si:Al ratio. A correlation was observed between HMF selectivity and the bidimensional structure of the H-mordenites, particularly the absence of cavities within the structure that prevented secondary product formation. Continuous extraction of HMF using MIBK for countercurrent circulation in a continuous heterogeneous pulsed column reactor resulted in an increase in HMF selectivity. Heterogeneous catalysts can accomplish the depolymerization and hydrolysis of cellulose. Schüth and co-workers used Amberlyst 15DRY in [BMIM]Cl to produce reducing sugars and cellooligomers from microcrystalline cellulose and wood biomass after an induction period of 1 h.133 When they used p-toluenesulfonic acid, however, no induction period was apparent. The Shimizu group employed the heteropolyacids H3PW12O40 and H4SiW12O40 to access reducing sugars from cellulose in an aqueous phase.134 They found that a stronger Brønsted acidity resulted in more active reactions. Catalysis by sulfonated silica/carbon nanocomposites were investigated by Jacobs, Sels, and co-workers to access high yields of glucose from cellulose.135 They attributed the high yields to the hybrid surface facilitating the adsorption of the β-1,4-glucan. Hara and co-workers used amorphous carbon with sulfuric acid, hydroxyl, and carboxylic acid groups to hydrolyze solid cellulose in water directly.136 They too attributed the activity of the catalysts on their ability to adsorb the cellulose glucan onto the catalytic surface. 2,5-Dimethylfuran as a Product. In an effort to supplement petroleum-derived fuels

with a renewable resource, gasoline is blended with ethanol. Some chemical properties of ethanol, such as miscibility with water and low enthalpy of vaporization, are undesirable.137 Even with substantial progress being made on accessing ethanol from cellulosic materials, the question of its ultimate viability and sustainability as a biofuel remains unanswered.138 Among the alternatives, DMF, which is formed by the catalytic hydrogenolysis of HMF,


Page 19 of 65

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Caes et al.

page 19

stands out. It has an increased energy density compared to ethanol (31.5 MJ/L versus 23 MJ/L), lower volatility (bp 92–94 °C versus 78 °C), and immiscibility with water, making it a potential drop-in replacement to gasoline and diesel.34 Xu and co-workers compared DMF to both gasoline and ethanol as a fuel. One such study was to compare the combustion performance of DMF in a gasoline direct-injection (GDI) engine to gasoline and ethanol.139 They found that using pure DMF did not adversely affect the performance of the research engine. Different loads of DMF gave different initial combustion durations (though gasoline had a longer duration) and engine knock was induced at 7.1 bar indicated mean effective pressure. Furthermore, emissions of CO, hydrocarbons, NOx, and particulate matter were all similar to gasoline emissions (although ethanol emissions were lower). Similar conclusions were reached using a dual-injection of DMF or ethanol with gasoline in a spark-ignition engine.140 High performance gains were attained with increased direct injection fractions of both DMF and ethanol, and emissions were mostly reduced using the dual-injection strategy (although CO2 and NOx emissions increased for DMF blends with gasoline). They concluded that the dual-injection strategy was most effective at lower fractions of the biofuels for port fuel injection. In another study by Xu and co-workers, the effect of spark timing and load of DMF was analysed on a direct-injection spark-ignition engine.141 As compared to gasoline, they found that DMF was more resistant to engine knock, had a lower initial and total combustion duration (as did ethanol, indicating the rapid combustion rate of oxygenated fuels), and had a similar volumetric consumption rate (allowing for a similar driving range). Additionally, DMF had a greater loss of thermal energy due to higher combustion temperatures and similar engine-out emissions to gasoline. Ethanol combustion was not limited by engine knock, but it has a lower volumetric consumption rate than both gasoline and DMF. Unsurprisingly, they also found that ethanol had the highest laminar burning velocity, followed by gasoline and



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Caes et al.

Page 20 of 65

page 20

then DMF.142 Thus, a laminar flame propagates most quickly in ethanol, though the velocities of gasoline and DMF were most similar. As the advantages of DMF acquired a more solid foundation, researchers have sought to access it from biomass materials. The Dumesic group developed a two-step process to transform fructose into DMF.34 Acid-catalyzed dehydration of fructose in a biphasic reactor led to the formation of HMF, which was then extracted into 1-butanol. The HMF was subjected to hydrogenolysis using a 3:1 atomic ratio copper–ruthenium (Cu:Ru/C) catalyst to access DMF in a 71% yield. The Cu:Ru/C catalyst was more resistant to poisoning by chloride anions than were copper catalysts, and could be regenerated by flowing hydrogen at the reaction temperature of 220 °C. The Raines group used the same catalyst to access DMF from corn stover biomass in a two-step process.103 Using CrCl3 in a [EMIM]Cl/DMA–LiCl mixture, they converted corn stover into HMF, which was purified with ion-exclusion chromatography to remove chloride anions and prevent poisoning of the Cu:Ru/C catalyst. The recovered HMF was subjected to hydrogenolysis in 1-butanol to give DMF in 49% yield, with an overall 9% yield from corn stover. Luijkx, Maat, van Bekkum, and co-workers used a palladium on carbon (Pd/C) catalyst to access DMF from HMF in 1-propanol.143 By GC-MS, they observed formation of 5-hydroxymethyl-2-(propyloxymethyl)furan during the initial stages of hydrogenolysis. The alcohol bond was the first to undergo hydrogenolysis, followed by the ether bond to access DMF. Similar reactivity was observed in 2-propanol, but 2,5-bis(hydroxymethyl)furan became the major product in 1,4-dioxane. The Rauchfuss group developed a one-pot process to convert fructose into DMF using formic acid.144 The formic acid served to dehydrate the fructose, which upon addition of a Pd/C catalyst in tetrahydrofuran, underwent hydrogenolysis using the formic acid as a hydrogen source to access 2,5-bis(hydroxymethyl)furan before deoxygenation to DMF. The Chidambaram group


Page 21 of 65

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Caes et al.

page 21

likewise used a Pd/C catalyst to access DMF from glucose in an ionic liquid/acetonitrile mixture in a two-step process.145 Glucose was converted into HMF using the heterocatalyst 12-molybdophosphoric acid, after which the catalyst could be replaced with the Pd/C without isolation of HMF to access DMF in a 16% yield. The lower yield was attributed to decreased reaction temperature, duration, and solubility of H2 in ionic liquids. Catalyst combinations were used to access DMF from HMF at high yields. Yanqin Wang and co-workers used a Ru/Co3O4 catalyst to convert DMF at 93.4% yield at 130 °C and 0.7 MPa.146 Further, they showed that their catalyst can be reused up to 5 times without loss of activity. Saha, Bohn, and Abu-Omar demonstrated the synergy of ZnCl2 with Pd/C to achieve an 85% yield of DMF at 150 °C and 0.8 mPa.147 Their catalyst was recycled up to 4 times without loss of activity. The Chatterjee and Kawanami groups used supercritical CO2 and water to effect the hydrogenation of HMF in the presence of various metal catalysts.148 They reported 100% yields of DMF in 2 h with a Pd/C catalyst and relatively mild reaction conditions: 80 °C, 𝑝!"! = 10 MPa and 𝑝!! = 1 MPa. Selectivity towards other products was tuned by varying 𝑝!"! alone. When 𝑝!"! was

Biomass to Furanics: Renewable Routes to Chemicals and Fuels

Recommend Documents