Review

[Link]/acscatalysis

Heterogeneous Catalytic Transfer Hydrogenation as an Effective

Pathway in Biomass Upgrading
Matthew J. Gilkey and Bingjun Xu*
Catalysis Center for Energy Innovation, Department of Chemical and Biomolecular Engineering, University of Delaware, Newark,
Delaware 19716, United States

ABSTRACT: Reducing oxygen content in biomass-derived feedstocks via hydrodeoxygenation (HDO) is a key step in their
upgrading to fuels and valuable chemicals. Organic molecules, e.g., alcohols and formic acid, can donate hydrogen to reduce the
substrate in a process called catalytic transfer hydrogenation (CTH). Although it is practiced far less frequently than molecular-
hydrogen-based HDO processes, CTH has been proven to be an efficient and selective strategy in biomass upgrading in the last
two decades. In this paper, we present a selective review of recent progress made in the upgrade of biomass-derived feedstocks
through heterogeneous CTH, with a focus on the mechanistic interpretation. Hydrogenation and cleavage of CO and C−O
bonds, respectively, are the two main categories of reactions discussed, owing to their importance in the HDO of biomass-derived
feedstocks. On acid−base catalysts, Lewis acid−base pair sites, rather than a single acid or base site, mediate hydrogenation of
carbonyl groups with alcohols as the hydrogen donor. While acid−base catalysts typically only catalyze the hydrogenation of
carbonyl groups with alcohols as the hydrogen donor, metal-based catalysts are able to mediate both hydrogenation and
hydrogenolysis reactions with either alcohols or formic acid. Several model reactions involving platform chemicals in biomass
upgrading, e.g., 5-hydroxymethylfurfural, levulinic acid, and glycerol, are used in the discussion to illustrate general trends.
Because alcohols are typically both the hydrogen donor and the solvent, the donor and solvent effects are intertwined. Therefore,
solvent effects are discussed primarily in the context of sugar isomerization and reactions with formic acid as the hydrogen donor,
in which the solvent and hydrogen donor are two separate species. Current challenges and opportunities of future research to
develop CTH into a competitive and complementary strategy of the conventional molecular-hydrogen-based processes are also
discussed.
KEYWORDS: catalytic transfer hydrogenation, hydrogen donor, hydride transfer, hydrodeoxygenation, biomass, lignocellulose

1. INTRODUCTION weight. While the oxygen content of lignin monomers is 30−40

As the only abundant source of non-food-based renewable wt %, the exact structure of lignin is still not well established.1
carbon, the efficient upgrading of lignocellulosic biomass to Dehydration and hydrodeoxygenation (HDO) are key
fuels and chemicals has the potential of substantially reducing processes to partially or fully remove the oxygen-containing
the anthropogenic carbon footprint if implemented on a global functional groups in feedstocks. Dehydration removes oxygen
scale. The main components of lignocellulosic biomass, i.e., atoms by eliminating water, which does not involve changes in
lignin, cellulose, and hemicellulose, and their derived platform the oxidation state of carbon atoms and typically proceeds via
compounds possess many desired chemical linkages, e.g., acid−base chemistry. In contrast, HDO lowers the oxygen
furanic and phenyl rings, which provide an excellent starting content of the molecule through reduction with a hydrogen
point. However, the oxygen content of biomass-derived donor. Most HDO processes employ molecular hydrogen as
feedstocks is often significantly higher than that of the desired
products, especially for fuels or fuel additives (Scheme 1): C5 Received: September 26, 2015
(C5H10O5) and C6 (C6H12O6) sugars derived from hemi- Revised: January 12, 2016
cellulose and cellulose contain more than 50% oxygen by Published: January 19, 2016

© 2016 American Chemical Society 1420 DOI: 10.1021/acscatal.5b02171

ACS Catal. 2016, 6, 1420−1436
ACS Catalysis Review

Scheme 1. Removing Oxygen-Containing Functional Groups reduces the complexity and cost of the experimental setup.
Is Key in the Upgrading of Lignocellulosic Biomass to Fuel Furthermore, the lower hydrogenating capability of most
and Chemicals organic hydrogen donors, in comparison to molecular hydro-
gen, enhances the degree of control in selective hydrogenation
and/or hydrogenolysis, when partially hydrogenated molecules
are targeted. Meanwhile, substantial modifications of the
existing molecular-hydrogen-based hydrogenation facilities are
likely needed when organic hydrogen donors are employed. For
example, additional separation and recycling units are needed
to separate the unconverted hydrogen donor from the reaction
mixture, which can then be recycled into the feed. The spent
organic hydrogen donor can also be separated and integrated
into the other steps of biomass upgrading: e.g., aldehydes and
ketones produced when alcohols are used as hydrogen donors
could be used in carbon chain growth reactions through aldol
condensation.2 Another option is to rehydrogenate the spent
the hydrogen donor, owing to its wide availability and easy hydrogen donor; however, molecular hydrogen is likely needed.
activation on many metal surfaces. However, liquid-phase Some reactions involve two sequential steps: hydrogen
hydrogen donors could be advantageous because high-pressure production from organic molecules, e.g., aqueous-phase
hydrogen gas presents a considerable safety hazard. Moreover, reforming (APR), followed by hydrogen-based HDO. These
the highly oxygenated biomass-derived molecules generally reactions are not typically viewed as CTH even though all steps
have high boiling points and are prone to decomposition upon occur in the same reactor. CTH provides extra dimensions in
vaporization, and thus many biomass-upgrading processes are the design and optimization of HDO processes. For example,
conducted in the liquid phase in the presence of solvents. The organic hydrogen donors could transfer hydrogen through
low solubility of molecular hydrogen in most solvents makes different mechanistic pathways to the substrate from molecular
high H2 pressure necessary to achieve desired conversions and hydrogen, e.g. intermolecular hydride transfer, which could be
yields. In addition to the safety concerns, handling high- exploited for selectivity control. Further, the competitive
pressure hydrogen gas incurs a hefty infrastructure cost on the adsorption of organic hydrogen donors with reactants and
industrial scale, which poses an economic barrier for developing reaction intermediates could also have a significant effect on the
a sustainable biomass-upgrading industry, especially in the product distribution.
initial stages. Organic molecules offer a renewable alternative to Although CTH was introduced more than a century ago by
molecular hydrogen, acting as hydrogen donors in the the seminal report of Pd black catalyzed disproportionation of
reduction of chemical bonds with the assistance of catalysts methyl terephthalate by Knoevenagel,3 it has been largely
through a pathway known as catalytic transfer hydrogenation overshadowed by the success of molecular-hydrogen-based
(CTH). The employment of liquid organic hydrogen donors HDO processes until the last few decades. The development of
alleviates the safety concern of handling high-pressure, both homogeneous and heterogeneous catalysts has made
flammable hydrogen gas, enhances the solubility of the significant progress in broadening the scope of CTH chemistry
hydrogen donor in liquid-phase reactions, and substantially and in addressing the drawbacks of low reaction rates and yields

Scheme 2. Common Mechanisms for Homogeneous CTH Reactions: (a) Direct Hydrogen Transfer; (b) Monohydride
Mechanism; (c) Dihydride Mechanisma

a
Both (b) and (c) belong to the metal hydride route.

1421 DOI: 10.1021/acscatal.5b02171

ACS Catal. 2016, 6, 1420−1436
ACS Catalysis Review

Scheme 3. Common Mechanisms for Heterogeneous CTH Reactions: (a) Direct Hydrogen Transfer; (b) Metal Hydride Route

that have long plagued CTH in its early years. CTH is effective direct hydrogen transfer pathway, the α-H is transferred from
in reducing functional groups by adding hydrogen to either the α-C of the alcohol to the carbonyl carbon in a concerted
unsaturated bonds (hydrogenation), e.g., CC,4,5 CC,6,7 step via a six-membered-ring intermediate, without forming a
CO,8−10 NO,11−13 NN,14 and CN,13 or single bonds metal hydride. This is also commonly referred to as the
leading to bond cleavage (hydrogenolysis), e.g., C−O,15−18 C− Meerwein−Ponndorf−Verley (MPV) mechanism. The use of
N,7 C−S,19 and C−X (halogen).20 deuterated hydrogen donors proves to be a helpful diagnostic
This paper aims to selectively review recent progress in the tool in differentiating the underlying reaction pathways. If the
upgrade of biomass-derived feedstocks through CTH and α-C is deuterated in the hydrogen donor, as in Scheme 2, the
discuss challenges and opportunities in establishing CTH as an isotopic purity of D will be preserved on the α-C of the newly
efficient, selective, and potentially economically viable pathway formed alcohol. It should be noted that the presence of a base,
in the production of biofuels and renewable chemicals. Owing which could be a ligand, a solvent molecule, or a dissolved basic
to the vast amount of literature on the general topic of CTH, species, can extract the hydroxyl proton in the hydrogen donor
the scope of this article is limited to heterogeneously catalyzed enhances the reactivity. A strong parallel can be drawn in
CTH reactions relevant to the conversion of biomass-derived heterogeneous CTH reactions regarding the direct hydrogen
molecules, though mechanistic insights from homogeneous transfer pathway: solid catalysts with an electron-deficient
systems will be leveraged to rationalize experimental observa- Lewis acid site and a neighboring base site can also catalyze
tions. Brieger et al.3 and Johnstone et al.21 have reviewed earlier CTH reactions following the MPV mechanism (Scheme 3a).
CTH work in the context of both homogeneous and The formation of metal hydride is the signature of the metal
heterogeneous catalysis. There are also several excellent review hydride route (Scheme 2b,c), which could be grouped into two
articles dedicated to more recent work on homogeneously subcategories: i.e., the monohydride and dihydride mecha-
catalyzed CTH reactions.22−24 We will focus our effort on two nisms. In the monohydride route, only the hydrogen on the α-
categories of reactions, i.e., hydrogenation of carbonyl groups C of the hydrogen donor, e.g., alcohols and formic acid, is
and hydrogenolysis of C−O bonds, owing to their central transferred to the metal. In contrast, the dihydride mechanism
importance in the field of biomass research. entails that both the hydrogen atom in the hydroxyl group and
that bonded to the α-C are transferred to the metal. The key
2. GENERAL STRATEGIES FOR HYDROGEN ADDITION difference between these two pathways is whether the hydrogen
To facilitate the discussion of the effect of catalysts, hydrogen atoms in O−H and C−H maintain their identity in the product.
donors, and solvents on reaction rates and product distributions To demonstrate this difference, the α-C of isopropyl alcohol is
in the following sections, we will first introduce several deuterated while the hydroxyl is not in Scheme 2. In the
common mechanisms of CTH. The identification of active monohydride mechanism, only the D is transferred to the
sites on solid catalysts in CTH is challenging, in that the surface metal, which is subsequently added to the carbonyl carbon.
structure of the catalyst under reaction conditions is typically Thus, the D is always bonded to the α-C in the product. In
dynamic in nature. The presence of solvents, reactants, and contrast, upon formation of the dihydride, the H and D become
products in bulk liquid phase could mask the features from equivalent and are equally likely to end up bonding to the O or
adsorbed reaction intermediates in many conventional in situ C. Interested readers are referred to a recent review article for a
spectroscopies. In this regard, mechanistic insights into CTH detailed discussion of further subcategorization of the
reactions catalyzed by structurally well-defined homogeneous monohydride mechanism into inner-sphere and outer-sphere
catalysts will be instrumental. hydrogen transfers.24 Adsorbed hydrogen atoms are important
There are two main mechanisms for homogeneous CTH reaction intermediates in heterogeneous CTH reactions on
reactionsdirect hydrogen transfer (Scheme 2a) and the metal metal surfaces, which resemble the dihydride mechanism
hydride route (Scheme 2b,c).22,24 When an alcohol acts as the because adsorbed hydrogen atoms are typically considered as
hydrogen donor in reducing a carbonyl group following the chemically equivalent. However, this assumption could at least
1422 DOI: 10.1021/acscatal.5b02171
ACS Catal. 2016, 6, 1420−1436
ACS Catalysis Review

Scheme 4. Possible Mechanisms of CTH Reactions via Direct Hydrogen Transfer on (a) General Acid−Base Pair, (b) Sn-β-
Zeolite (Open Sn Site), (c, d) MgO, and (e) Hydrous Zirconia

be partially attributed to the difficulty in experimentally electron-deficient metal center, is to bond with the electron-rich
differentiating hydrogen adsorbed on different types of sites oxygen in the hydroxyl and carbonyl groups in the hydrogen
on the metal particles. Adsorbed hydrogen atoms can be donor and acceptor, respectively, whereas the adjacent base site
produced by either the dissociative adsorption of molecular attracts the proton in the hydroxyl and weakens the O−H
hydrogen or dehydrogenation of organic hydrogen donors. The bond. The stronger the interaction between the hydroxyl
assumption that all adsorbed hydrogen atoms are chemically oxygen and the Lewis acid site, the more acidic the hydroxyl
equivalent leads to the hypothesis that the reaction mechanisms hydrogen in the hydrogen donor becomes, which facilitates the
and product distributions when using either hydrogen gas or an abstraction of the hydrogen by the base site. Conversely,
organic hydrogen donor should in principle be equivalent. strongly basic sites can effectively abstract the hydrogen from
Important differences in the surface coverage of atomic the hydroxyl group of the alcohol, leading to the formation of
hydrogen and other adsorbed species do exist, and their effect an alkoxide adsorbed on the adjacent Lewis acid site, thus
will be discussed in the next several sections. promoting the hydride transfer (Scheme 4a). Thus, both strong
acids and strong bases facilitate CTH, but it is unlikely that
3. SOLID CATALYSTS FOR CTH strong Lewis acid and base sites can coexist on the same
3.1. Acid−Base Catalysts. Acid−base pair sites can catalyst. Although either the acid or base property of a catalyst
facilitate the CTH between an alcohol and a carbonyl group, could dominate, an acid−base pair is needed to complete the
mainly through direct hydrogen transfer, since the formation of catalytic cycle.
hydrides on most acid and base sites is difficult. In addition, Alcohols are frequently used as both the hydrogen donor and
there is generally no change in the oxidation state of the active the solvent in the acid−base pair catalyzed CTH reactions.
sites during the course of the reaction in acid−base chemistry. Their low cost, renewable nature, and unique ability to
Although both Lewis acidic and basic materials have been participate in intra- and intermolecular hydride transfers make
reported to be active in mediating CTH processes following the alcohols an attractive option for HDO reactions. Secondary
MPV mechanism, it is important to recognize that both Lewis alcohols are more efficient hydrogen donors than primary
acid and base sites are necessary to catalyze the reaction alcohols, which can be rationalized by the enhanced
(Scheme 4a). The role of the Lewis acid site, typically an stabilization effect on the carbocation from two alkyl groups
1423 DOI: 10.1021/acscatal.5b02171
ACS Catal. 2016, 6, 1420−1436
ACS Catalysis Review

Scheme 5. General Pathway for Upgrading Cellulose-Derived Feedstocks to Fuel and Chemicals

(secondary carbocation) during the hydride transfer,21 in Scheme 6. (a) Formation of a Closed Sn Site through
comparison to a single alkyl group in primary carbocations. Hydrolysis of an Open Sn Sitea and (b) No Formation of
Tertiary alcohols cannot serve as hydrogen donors, due to the Na+ on Closed Sn Sites
lack of an α-H. Self-etherification of alcohols is a common side
reaction, especially in the presence of acid sites; however, it is
frequently neglected in the literature because the conversion,
yield, and carbon balance are calculated on the basis of the
substrate of the CTH reaction. Quantification of products
formed from the hydrogen donor provides information on the
total hydrogen that is produced from the hydrogen donor
versus hydrogen that participates in the hydrogenation
reactions. In other words, the hydrogen balance provides
quantitative information on the efficiency of an organic
hydrogen donor in the reaction.
a
Lewis acid sites on solid catalysts can effectively mediate Na+ exchange with an open Sn site occurs on the silanol group and
direct hydrogen transfers, in strong parallel with homogeneous potentially also on the stannanol group.
CTH systems.24,25 In this regard, zeolites containing tetravalent
metal dopants, e.g., Ti, Sn, Zr, and Hf, have shown remarkable calculations confirm that the open Sn site mediated pathway for
activity toward intra- and intermolecular hydride transfer glucose isomerization is favored over the closed Sn site.35
following the MPV mechanism. Replacing a small fraction of Recent computational results suggest that the bidentate
tetrahedrally coordinated framework Si atoms with tetravalent interaction of oxygen atoms on C1 and C2 in glucose in the six-
dopants, e.g., Ti, Sn, Zr, and Hf, does not disrupt the charge membered-ring intermediate may be less energetically favored
neutrality of the zeolite framework but introduces Lewis acid than pathways with glucose bonded to Sn in a monodentate
sites at the heteroatoms.26−28 In particular, Sn-containing Beta configuration.36,37 Three potential pathways have been
zeolite, or Sn-Beta, is remarkably active and selective in proposed that involve hydrogen-bonding-assisted C1−C2
catalyzing the isomerization of glucose to fructose (Scheme hydride transfer. (1) Rai et al. identified a pathway involving
5).29,30 It is a key step in converting cellulose-derived feedstock the neighboring silanol formed with the open Sn site via the
to a versatile platform chemical, 5-hydroxymethylfurfural hydrolysis of a Si−O−Sn bond (Scheme 7a). Glucose initially
(HMF), which can only be produced from dehydration of loses a proton to the Sn−OH group and bonds to the Sn center
fructose (five-membered ring) but not from glucose or through C2−O, which is followed by concerted C1−C2 hydride
mannose (six-membered ring). Isomerization of glucose to transfer and silanol-mediated proton transfer.37 (2) An
fructose is proposed to occur on the Lewis acidic Sn site, which adsorbed water molecule mediates the concerted C1−C2
bonds to the electron-rich oxygen atoms of the carbonyl and hydride transfer and proton transfer from water to the carbonyl
hydroxyl groups on the C1 and C2, respectively, and facilitates oxygen (Scheme 7b).36 (3) A vicinal silanol group mediates the
intramolecular hydride transfer from C2 to C1 (Scheme 4b).31 concerted C1−C2 hydride and proton transfer from the silanol
Furthermore, Davis and co-workers identified that the active group to the carbonyl oxygen (Scheme 7c).36 Conrad et al.
sites are “open Sn sites”, i.e., Sn atoms that bond to bridging attempted to mimic the closed Sn sites by grafting Sn(O-
oxygen atoms and a hydroxyl group, rather than the “closed Sn TMS)3O− and Sn(O-TMS)2(O−)2 species (TMS = trime-
sites”, in which Sn atoms bond to four bridging oxygen thylsilane) on SiO2 and found that the turnover number of
atoms.32−34 An open Sn site can be created by hydrolyzing a those sites is lower than that of Sn-Beta by more than a factor
Sn−O−Si bond of a closed Sn site, leading to a stannanol (Sn− of 6.38 This result shows that Sn(−OSi)4 sites are reasonably
OH) and a silanol group (Scheme 6a). Ion exchange with Na+ active in the glucose isomerization reaction. Meanwhile, the
leads to the formation of Si−O−Na+ and likely Sn−O−Na+ lack of pore structure and hydrophobic environment for the
(Na/Sn ratios are greater than unity in Na-exchanged Sn-Beta), grafted Sn(−OSi)4 site could lead to significant differences in
which are potential base sites for extracting the proton from a its interaction with the reactant in comparison to that on the Sn
hydroxyl during the hydride transfer.34 The absence of sites in Sn-Beta, which introduces uncertainty in the
isomerization activity of Na-exchanged Sn-Beta in NaCl/H2O, interpretation of the activity difference.
which prevents the loss of Na+ during the isomerization The high solubility of glucose in water makes water an
reaction, provides concrete evidence supporting the hypothesis excellent solvent for the isomerization reaction; however, the
that open Sn sites are the active sites in the isomerization of presence of bulk water severely deactivates Lewis acid sites by
glucose, because closed Sn sites should be unaffected by ion strong adsorption. Defect-free Sn-Beta, synthesized in the
exchange (Scheme 6b). Density functional theory (DFT) based fluoride medium, has hydrophobic pores and excellent water
1424 DOI: 10.1021/acscatal.5b02171
ACS Catal. 2016, 6, 1420−1436
ACS Catalysis Review

Scheme 7. Proposed Mechanisms of Glucose Isomerization on an Open Sn Site of Sn-Beta on the Basis of DFT Calculations:
(a) Neighboring Silanol-Assisted Pathway; (b) Adsorbed Water-Assisted Pathway; (c) Vicinal Silanol-Assisted Pathway

resistance, as evidenced by its more than 2-fold higher glucose ization,28,30,39 most likely because the rates of both reactions are
conversion in comparison to that on the relatively hydrophilic limited by the hydride transfer step.
Sn-MCM-41.30 However, Sn-Beta is more active than Ti-Beta Metal oxides, e.g. ZrO2,43 ZrO(OH)2,44 MgO,43 etc., have
in converting glucose to fructose, though both catalysts are also shown high activity in performing the MPV reaction to
hydrophobic.30 The activation barrier of glucose isomerization GVL. In situ generated HCl and ZrO(OH)2 catalysts from a
on tetravalent-metal-doped Beta zeolites has been correlated salt precursor (ZrOCl2·8H2O) form GVL with high selectivity
(84.5%) in 2-BuOH;44 however, the role of in situ generated
with the polarizability of the heteroatom and the basicity of the
HCl is unclear. One possibility is that HCl, a strong acid,
neighboring oxygen atom, with the activation barrier increasing
catalyzes the esterification reaction of LA to levulinates, from
in the sequence Sn ≈ Zr < Nb < Ti ≈ Ge < V.39 A similar which the reaction proceeds much more quickly over
intramolecular hydride transfer mechanism is responsible for amphoteric oxides such as ZrO2. This is consistent with the
the isomerization of pentoses40,41 and glyceraldehyde40 on Sn- work of Chia et al., which showed that levulinates, rather than
Beta. LA, reacted more quickly to form GVL over ZrO2.43 A higher
Lewis acid catalyzed CTH can be integrated into cascade concentration of basic sites in amphoteric oxides induces strong
reactions in the production of γ-valerolactone (GVL) and 2,5- binding of acidic functional groups on LA, thereby suppressing
bis(alkoxymethyl)furans. GVL, a versatile platform chemical, the rate of reaction. Thus, controlling the ratio of acid and base
can be produced in a two-step cascade reaction starting with sites is crucial in optimizing the rate of hydrogenation. Lewis
levulinates on Ti-, Sn-, Zr-, and Hf-Beta.28,42 Levulinate is acid mediated CTH over metal oxides can also be used to
initially hydrogenated to the corresponding 4-hydroxypenta- selectively produce 2,5-bis(hydroxymethyl)furan (∼94%) from
noate with an alcohol as the hydrogen donor, which is followed HMF using ZrO(OH)x, where surface hydroxyl groups act as
by lactonization to GVL (Scheme 8). The lower activation active sites for alkoxide formation, thereby facilitating the MPV
mechanism.45 Zeolites have also been employed to reduce
energy on Sn-, Zr-, and Hf-Beta in comparison to that on Ti-
furfurals into furfuryl alcohols.9,46−48 Vlachos et al. coupled
Beta in GVL formation parallels that in glucose isomer- hydrogenation to 2,5-bis(alkoxymethyl)furans with ether-
ification reactions with an alcohol to produce high-molecular-
Scheme 8. Reaction Mechanism for γ-Valerolactone weight ethers,9 which proceeds via the sequential CTH and
Production via CTH of Levulinate on Tetravalent-Metal- etherification steps on Lewis acid sites. Similarly, furfural can be
Doped Beta Zeolites hydrogenated to furfuryl alcohol via zeolite-mediated CTH
using a variety of Sn/Al combinations.46,47 Moreover, Sn/Al-
containing Beta zeolites can further catalyze tandem reactions
(hydrogenation and hydrolysis) to form levulinates and,
subsequently, GVL in one pot.46,47
Base-mediated CTH reactions were proposed to proceed
through a two-step process (Scheme 4c):8,49−53 the alcohol
(hydrogen donor) dissociates to form the corresponding
alkoxide on a weakly acidic metal cation, e.g., Mg2+, while the
hydroxyl proton resides on the basic O2− anion. The α-H of the
alkoxide and the proton adsorbed on the base site are then
transferred to the carbon and oxygen atoms in the carbonyl
group, respectively. This pathway is quite different from the
typical MPV mechanism because the six-membered-ring
intermediate does not involve the metal site. An alternative
1425 DOI: 10.1021/acscatal.5b02171
ACS Catal. 2016, 6, 1420−1436
ACS Catalysis Review

Table 1. Metal-Mediated CTH Results with a Variety of Substrates Using Alcohols as the Hydrogen Sourcea
substrate desired product catalyst solvent temp (K) time (h) conversn (%) yield (%) ref
EL GVL Ru(OH)x/TiO2 2-propanol 363 24 >99 89 107
LA GVL Raney Ni 2-propanol 353 9 >99 >99 118
LA GVL Raney Ni 2-propanol room temp 9 >99 >99 118
LA GVL Ru/C 2-propanol 353 9 >99 93 79
furfural 2-MF Pd/Fe2O3 2-propanol 453 7.5 >99 13 17
furfural 2-MF Cu/Fe2O3 2-propanol 453 7.5 37 0 17
furfural 2-MF Ru/RuOx/C 2-propanol 453 10 95 61 10
furfural 2-MF Ru/RuOx/C 2-BuOH 453 10 >99 74 78
HMF DMF Cu-PMO MeOH 260 3 >99 48 74
HMF DMF Cu-PMO MeOH 320 0.75 >99 32 74
HMF DMF Pd/Fe2O3 2-propanol 453 24 >99 71 17
HMF DMF Ru/RuOx/C 2-propanol 463 6 >99 80 9
glycerol 1,2-PDO Cu:Al 2-propanol 493 3 70 38 102
glycerol 1,2-PDO Ni−Cu/Al2O3 2-propanol 493 24 60 39 18
glycerol 1,2-PDO Pd/Fe2O3 2-propanol 453 24 100 94 18
a
Abbreviations: EL = ethyl levulinate, GVL = γ-valerolactone, LA = levulinic acid, 2-MF = 2-methylfuran, HMF = 5-hydroxymethylfurfural, DMF =
2,5-dimethylfuran, 1,2-PDL = 1,2-propanediol.

mechanism similar to that of Lewis acid catalysts (MPV of hydrous zirconia is weaker than that of MgO,67 the
mechanism) could be drawn, which is consistent with all dissociation of the alcohol to an alkoxide is likely not favored.
existing results (Scheme 4d). In this case, the reaction is likely The condensation reaction between Zr−OH and the alcohol is
driven primarily by the basic sites through the abstraction of the key to the irreversible formation of the adsorbed alkoxide upon
proton in the hydroxyl group, rather than the interaction of desorption of water, which helps initiate the catalytic cycle.
reactants with the weak acid sites. However, as in the acid- Acid/base-catalyzed CTH processes are remarkably efficient
mediated case, the presence of both acid and base sites is and selective in reducing carbonyl groups. However, they are
necessary to achieve high catalytic performance.54,55 The rate- ineffective in cleaving C−O single bonds, which could be
limiting step in base-mediated CTH reactions remains unclear. attributed to (1) the inability to activate C−O bonds via the
Reduced electron density on the carbonyl group facilitates Lewis acid mediated MPV mechanism and (2) the inability to
hydride transfer, as evidenced by the enhanced reaction rate produce active hydrogen species, e.g., Had, on metal oxides from
when electron-withdrawing groups neighboring the carbonyl organic hydrogen donors to complete the catalytic cycle. The
group are present,52,56 which is an indication that the hydride cleavage of the C−O bond is an indispensible step in the
transfer is a kinetically relevant step. This hypothesis can be upgrading of cellulose-, hemicellulose-, and lignin-derived
tested by measuring the kinetic isotopic effect with an feedstocks (Scheme 1), which can be accomplished by metal-
undeuterated and a deuterated α-C in the alcohol. MgO is mediated CTH processes via the metal hydride mechanism
the most frequently used basic CTH catalyst,49,53,57−61 but it discussed in the next section.
deactivates rapidly. Szöllösi et al. reported that treating the 3.2. Metal Catalysts. The ability to efficiently activate H−
MgO catalyst with chloromethanes significantly enhances H, CO, C−O, and C−H bonds makes metalsparticularly
catalyst stability by blocking the Lewis acid site by adsorbed noble metalsthe most widely used hydrogenation and
Cl.53,60,61 Moreover, Mg/Al mixed oxides,50,52,55 including hydrogenolysis catalysts. The most common form of metal
hydrotalcites54 and potassium phosphate,56 are also active in catalysts is metal (nano)particles on a high-surface-area
catalyzing the MPV reactions. support. Although some support materials, e.g., activated
Hydrous zirconia has been shown to be a very effective MPV carbon,68−70 are inert, many supports contain acid or base
catalyst,62−65 and its catalytic performance is sensitive to the sites, e.g., Al2O3,68,71 zeolites,72,73 and porous metal oxides.74
preparation method and calcination temperature. As an Unsupported porous metal catalysts may also contain
amphoteric oxide, ZrO2 possesses acid−base pair sites, though heteroatoms due to the preparation procedure, e.g. the residual
neither is very strong. The high activity of hydrous zirconia has Al species in Raney Ni produced from the alkali corrosion of
been attributed to the presence of surface hydroxyl groups, Ni−Al alloy.75 The coexistence of multiple types of active sites
which diminish in number as the calcination temperature poses challenges in achieving a molecular level mechanistic
exceeds 300 °C.66 The surface hydroxyl group reacts with the understanding due to the structural complexity. At the same
alcohol (hydrogen donor) to form an adsorbed alkoxide and time, it provides flexibility in tailoring catalysts for reactions
water (Scheme 4e).62,63 The α-H of the alkoxide is then wherein different reaction steps require different types of active
transferred to the carbonyl carbon of the hydrogen acceptor via sites. The dissociative adsorption of H2 to adsorbed atomic
the formation of the six-membered-ring intermediate, similar to hydrogen is generally the first mechanistic step in metal-
that in the classic MPV mechanism. The aldehyde or ketone mediated HDO processes with hydrogen gas, whereas metal-
formed from the dehydrogenation of the alcohol (R1R2CO) catalyzed CTH processes start by activating the organic
desorbs from the Zr site, leaving behind the alkoxide hydrogen donor. It is generally assumed that the mechanistic
(R3R4CHO−) derived from the hydrogen acceptor. In the pathways of HDO processes with H2 and organic hydrogen
final step, the hydrogenated product (R3R4CHOH) is formed donors converge after adsorbed atomic hydrogen is formed
via the displacement of the alkoxide by an incoming alcohol (Scheme 3b); however, formation of negatively charged
molecule, thus completing the catalytic cycle. Since the basicity hydride species has been proposed as a surface intermediate
1426 DOI: 10.1021/acscatal.5b02171
ACS Catal. 2016, 6, 1420−1436
ACS Catalysis Review

with organic hydrogen donors,16,76 indicating more substantial Scheme 9. Reaction Pathway for Converting HMF to DMF
mechanistic differences may exist. In the following section, via Sequential Hydrogen of the Carbonyl Group and
metal-catalyzed CTH processes are broken down first by the Hydrogenolysis of the C−O Bonda
type of hydrogen donor and then by substrates and reactions.
3.2.1. Alcohols as Hydrogen Donors. Alcohols are widely
used hydrogen donors for metal-catalyzed CTH reactions
(Table 1), with many parallels with acid−base chemistry
discussed in section 3.1. Secondary alcohols generally show
higher activity than primary alcohols in dehydrogenation over
metal surfaces, facilitating hydrogen transfer to the substrate.
This can be readily attributed to the enhanced stabilizing effect
of two, rather than one, alkyl groups via inductive electron
donation to the α-C of the alcohol in the dehydrogenation
process.8,17,77,78 In particular, Vlachos and co-workers recently a
Abbreviations: HMF = 5-hydroxymethylfurfural, MFUR = 5-
systematically studied the effect of the structure of alcohols on methylfurfural, MFA = 5-methyl-2-furyl(methanol), BHMF = 2,5-
the conversion of furfural to 2-methylfuran (2-MF) on Ru/C.77 (bishydroxymethyl)furan.
Aside from confirming secondary alcohols being superior to
primary alcohols, this work shows that a longer alkyl chain in furfuryl alcohol to 2-MF. Significant amounts of ring-saturated
the alcohol is beneficial for CTH activity, which holds for both (tetrahydromethylfuran, 36%) and decarbonylated (furan,
primary and secondary alcohols. However, this effect 24%) products were formed along with 2-MF (26%) on Pd/
diminishes when the side chain contains more than two carbon Fe2O3 under continuous flow conditions. Pd/Fe2O3 shows
atoms: i.e., the CTH activity of alcohols increases in the much higher CTH activity than Pd/C, which is attributed to
sequence ethanol < 1-propanol ≈ 1-butanol < 2-propanol < 2- the oxophilic nature of Fe in facilitating the activation of the
butanol ≈ 2-pentanol. The reduced enhancement effect for O−H bond in 2-propanol. HDO of HMF to DMF is
longer side chains could be attributed to the diminished added mechanistically similar to that of furfural but requires one
stabilizing effect, site blocking caused by the larger footprint of more hydrogenolysis step. A 72% yield of DMF from HMF was
the adsorbed alcohol, or a combination of the two. achieved on Pd/Fe2O3 under conditions similar to those for
Interestingly, methanol has been employed as a hydrogen furfural.17 The lack of unprotected α-carbon makes BHMF and
donor in the hydrogenolysis of glycerol.18 When methanol DMF less susceptible to dimer or oligomer formation via
dehydrogenates into formaldehyde and H, the formaldehyde alkylation,77 which leads to the higher yield of DMF than 2-MF.
can react with water to form formic acid. The decomposition of Riisager and co-workers recorded ∼32% yield toward DMF
formic acid to CO2 and H is more energetically favorable than when using supercritical methanol as a hydrogen source (>99%
methanol dehydrogenation; thus, CO2 was the only product conversion of HMF) over Cu-doped hydrotalcite, with
aside from hydrogen when methanol was used as the hydrogen significant overhydrogenation to 2,5-dimethyltetrahydrofuran.74
donor on Ni−Cu/Al2O3 at 220 °C. Thus, two molecules of At high temperatures (>300 °C), fast dehydrogenation of
hydrogen are produced for each methanol molecule, twice as MeOH to H2 and CO2 is likely to create a hydrogen-rich
much as that from 2-propanol. Comparable glycerol con- environment that favors the saturation of the furan ring.
versions and product distributions were observed with 2- High yields for 2-MF (∼70%) and DMF (∼80%) were
propanol and methanol as the hydrogen donors.18 Moreover, obtained on Ru-based catalysts from furfural and HMF,
supercritical methanol has been demonstrated as an effective respectively, with secondary alcohols.10,77 This could in part
environment for conversion of HMF to DMF.74 be attributed to the oxophilic nature of Ru that leads to strong
[Link]. HDO of Oxygenated Furanics. HDO of oxygenated bonding to oxygen-containing species, such as furfural and
furanic compounds, e.g. HMF and furfural, to reduced furans, furfuryl alcohol, which accelerates the HDO of furfural and
e.g. DMF and 2-MF, involves both hydrogenation of the suppresses side reactions such as the hydrogenation of the furan
carbonyl bond and hydrogenolysis of the C−O bond (Scheme ring. In a subsequent study, Jae et al. discovered that mildly
9). The acid−base pair mediated CTH cannot accomplish the oxidized Ru/C catalysts (referred to as Ru/RuOx/C), which
latter, and therefore, 2,5-bis(alkoxymethyl)furans produced contain both metallic (Ru) and Lewis acidic (RuOx) sites, are
from the etherification between 2,5-bis(hydroxymethyl)furfural most active and selective in the HDO of HMF to DMF.69 The
(BHMF) and the alcohol is the final product.28 In contrast, benefit of the coexistence of metal and Lewis acid sites is
high yields of DMF and 2-MF can be achieved on metal evidenced by the meager DMF yields on fully reduced Ru/C
catalysts from HMF and furfural, respectively. Co-feeding (∼30%) and RuO2 (∼10%), in contrast with the ∼70% yield on
furfural and 1,4-butanediol (1,4-BDO) on a Cu−Ni catalyst the physical mixture of Ru/C and RuO2.9 Higher DMF yield
produces 2-MF and γ-butyrolactone (γ-BL) with close to 100% (>80%) on Ru/RuOx/C, which is prepared by mildly oxidizing
yield in the vapor phase at temperatures above 190 °C.79 In fully reduced Ru/C under a diluted O2 atmosphere, in
addition, similar conversions and product distributions were comparison to that on a physical mixture of Ru/C and RuO2
observed when the dehydrogenation of 1,4-BDO and the HDO emphasizes the importance of atomically mixed metal and
of furfural with hydrogen gas were carried out separately, Lewis acid sites. Isotopic labeling investigations with detailed
indicating that the dehydrogenation and HDO pathways are mass fragmentation pattern analysis reveal that the hydro-
independent. Hermans and co-workers investigated the CTH genation of the carbonyl group in furfural proceeds through a
of furfural with 2-propanol over Fe2O3-supported metals (Pd, Lewis acid mediated MPV mechanism, as evidenced by the
Ni, and Cu) in the liquid phase, showing that Pd was far more transfer of the deuterium atom from the α-carbon of 2-
active toward the HDO reaction than Cu and Ni.17 Moreover, propanol-d8 to the α-carbon of the formed furfuryl alcohol
only Pd/Fe2O3 is capable of catalyzing hydrogenolysis of (Scheme 10). In addition, the incorporation of deuterium into
1427 DOI: 10.1021/acscatal.5b02171
ACS Catal. 2016, 6, 1420−1436
ACS Catalysis Review

Scheme 10. Proposed Mechanism for 2-MF Formation from Furfural over Ru/RuOx/C: (a) Hydrogenation of Furfural; (b)
Hydrogenolysis of Furfuryl Alcohol

Scheme 11. Proposed Pathways for 1,2-PDO Formation from Glycerol, Proceeding by (a) Initial Dehydrogenation and (b)
Initial Dehydration

the furan ring in the product (2-MF) shows that the 1,2-PDO is currently produced industrially from propylene
hydrogenolysis of furfuryl alcohol occurs, at least in part, via oxide, which is derived from petroleum-based propylene.83
the activation of the furan ring,80 a phenomenon known to Hydrogen can be produced from glycerol from APR on Pt- and
occur facilely over Ru surfaces.81 Furan rings adsorb strongly Ni-based catalysts;83−86 thus, with a tailored combination of
with Ru surfaces, which promotes the hydrogenolysis of the C− metal and acid−base sites, hydrogen generated in situ can be
O bond in the hydroxymethyl group80 and the C−O bond used for the hydrogenolysis of glycerol.83,87−90 In this approach,
within the furan ring. Coadsorption of different solvent hydrogen production via APR and hydrogenolysis could be
molecules appears to play a significant role in the ability to viewed as sequential processes, and the latter typically occurs in
activate the ring.81 Furthermore, it is hypothesized that the a molecular-hydrogen-rich environment.83 Therefore, it is
Lewis acidic RuOx contributes to the hydrogenolysis by unclear whether CTH is a major pathway and is beyond the
accepting the −OH group from the furfuryl alcohol. These scope of this article.
findings were further confirmed by the significant kinetic Glycerol hydrogenolysis is proposed to occur through several
isotope effects observed in both hydrogenation and hydro- possible pathways. Many studies indicate that the initial step in
genolysis steps, indicating that hydrogen (or deuterium) glycerol hydrogenolysis to 1,2-PDO is dehydrogenation of the
transfer to the substrate is rate limiting. The synergistic effect terminal alcohol group to form glyceraldehyde (Scheme
of the metal and Lewis acid sites on Ru/RuOx/C is reinforced 11a),91−96 though this step is likely reversible in reducing
by the enhanced 2-MF yield on fully reduced Ru/C with environments. Glyceraldehyde can then be dehydrated and
homogeneous Lewis acids: e.g., AlCl3.82 Etherification occurs tautomerized, via an enol intermediate, to form pyruvaldehyde,
among the hydrogen donor molecules and between furfuryl which is hydrogenated to 1,2-PDO (Scheme 11a). The
alcohol and the hydrogen donor in the presence of Lewis acids; formation of glyceraldehyde through dehydrogenation at high
however, it is a reversible process.10 hydrogen pressure is still a topic of debate.83,91,97 It has been
[Link]. Hydrogenolysis of Glycerol to 1,2-Propanediol. proposed that this mechanism holds true when the rate of
Hydrogenolysis of glycerol, a byproduct of biodiesel dehydration from glyceraldehyde is higher than that of
production, is a promising renewable pathway to synthesize dehydration from glycerol,94 which is accelerated in the
1,2-propanediol (1,2-PDO), an important polymer precursor. presence of base.95,96 It should be noted that, in the absence
1428 DOI: 10.1021/acscatal.5b02171
ACS Catal. 2016, 6, 1420−1436
ACS Catalysis Review

Scheme 12. Proposed Reaction Pathways of Glycerol on Ni−Cu/Al2O3: (a) Hydrogenolysis to 1,2-PDO; (b) Dehydration to
Acetol

Scheme 13. Proposed Pathways of LA Hydrogenation to GVL and Further Hydrogenated Products

of hydrogen, glycerol can be converted to lactic acid through a 1,2-PDO with hydrogen gas, dehydration of glycerol to acetol is
similar mechanism over an iron pincer catalyst.98 However, a major pathway in CTH with 2-propanol.18,101 Higher yields
instead of hydrogenation of pyruvaldehyde, an intramolecular for 1,2-PDO were obtained on Ni−Cu/Al2O3 with 2-propanol,
Cannizzaro reaction occurs in the presence of a base to form which is still ∼20% lower than those with high-pressure
lactic acid (or the corresponding lactate salt). Alternatively, hydrogen (45 bar) under otherwise identical conditions.71
glycerol hydrogenolysis can begin through Brønsted acid Since hydrogenolysis and dehydration are parallel reaction
catalyzed dehydration (Scheme 11b), rather than dehydrogen- pathways (Scheme 12),71 the higher yield of acetol could be
ation, forming acetol as the only intermediate. 97,99,100 attributed to the lower atomic hydrogen coverage with the
Tomishige et al. showed that this pathway is accelerated in organic hydrogen donor (vs high-pressure hydrogen). More-
the presence of acid sites, e.g. Amberlyst.100 Acetol can then be over, Pd/Fe2O3 has also shown to be an effective catalyst for
hydrogenated over a metal, e.g. Ru, Cu, etc., to form 1,2-PDO. glycerol hydrogenolysis to 1,2-PDO.102 Similar to the case of
To eliminate the need for acids or bases, the use of bimetallic or furfural HDO,17 the presence of the Fe2O3 support is key to the
bifunctional catalysts, e.g. Ni−Cu,15,18,71 has been employed, catalytic performance, which most likely facilitates the
forming 1,2-PDO from glycerol through direct hydrogenolysis activation of the alcohol.
routes. [Link]. Levulinic Acid to γ-Valerolactone. The conversion
2-Propanol is an efficient hydrogen donor on coprecipitated of levulinic acid (LA) and levulinates to GVL can be mediated
Cu−Al catalysts for the hydrogenolysis of glycerol. Low not only by acid−base pairs, as discussed in the previous
conversion (<26%) and selectivity for 1,2-PDO (<10%) were section, but also by metal catalysts. Mechanistic studies by
obtained when the reaction was carried out in an aqueous Bond et al. using molecular hydrogen, rather than CTH,
solution of 20 wt % glycerol at 220 °C, indicating that not suggest that there are two possible pathways toward GVL over
much APR occurred. A more than 20-fold increase in 1,2-PDO metal catalysts: e.g., Ru/C.103 One possibility is that LA or
yield was observed when the reaction was conducted in 2- levulinates are first hydrogenated to form 4-hydroxypentanoic
propanol under otherwise identical conditions; however, the (HPA) acid or a corresponding ester; the resulting HPA can
dehydration pathway to acetol still dominated.101 The then undergo lactonization or ring closure to form GVL as the
beneficial effect of 2-propanol as a solvent, in comparison to final product (Scheme 13). Alternatively, LA or levulinates can
water, for the hydrogenolysis for glycerol was still appreciable in first undergo lactonization or ring closure to angelicalactones
the presence of 69 bar of H2,101 which indicates that different (α- or β-angelicalactone), followed by hydrogenation to GVL,
reaction pathways for H2 and 2-propanol could be at work. as proposed by Hasan et al. over Pd/C in the presence of
While Cu/Al2O3 and Ni/Al2O3 are selective in the formation of base.104 Bond et al. explored these two pathways in the
1429 DOI: 10.1021/acscatal.5b02171
ACS Catal. 2016, 6, 1420−1436
ACS Catalysis Review

Table 2. HDO of Various Biomass-Derived Molecules with FA as the Hydrogen Sourcea

substrate desired product catalyst solvent temp (K) time (h) conversn (%) yield (%) ref
HMF DMF Pd/C THF + H2SO4 383 15 99 95 132
fructose HMF Pd/C THF + H2SO4 393 1 65 132
fructose HMF Pd/C DMSO + H2SO4 373 1 93 132
glucose GVL Au/ZrO2 water + H2SO4 423−443 1−10 n.r. 51 127
fructose GVL Ru/C water + tetrafluoroacetic acid 453 16 >99 52 128
BL GVL Au/ZrO2 water 443 8 98 95 126
LA GVL Ag−Ni/ZrO2 water 493 5 99 99 131
LA GVL Ru NPs water + EtN3 423 24 99 99 129
LA GVL Au/ZrO2 water 423 6 >99 >99 127
LA GVL Cu/ZrO2 water 453 2 >99 >99 130
LA GVL Pd/ZrO2 water 423 6 n.r. trace 127
LA GVL Cu/ZrO2 water 473 2 100 100 130
glycerol 1,2-PDO Ni−Cu/Al2O3 water 493 10 34 29 18
1-HB ethylbenzene Pd/C water + MeOH + FoNH3 353 0.5 99 98 70
HMF HDL Pd/ZrP EtOH 413 21 97 43 143
a
Abbreviations: n.r. = not reported, HMF = 5-hydroxymethylfurfural, DMF = 2,5-dimethylfuran, THF = tetrahydrofuran, DMSO = dimethyl
sulfoxide, GVL = γ-valerolactone, BL = butyl levulinate, LA = levulinic acid, 1-HB = (1-hydroxyethyl)benzene, HDL = 1,6-hexanediol.

presence and absence of H2, subsequently concluding that is possible that the partial oxidation of Ru provides Lewis acidic
angelicalactones are observed as a result of proton-catalyzed sites that facilitate the lactonization step.
ring closure directly from LA only in the absence of molecular 3.2.2. Formic Acid and Formate as Hydrogen Donors.
H2. In the presence of H2, the rate of reaction toward HPA is Formic acid (FA) can be formed renewably from lignocellulosic
orders of magnitude higher than the rate toward angelicalac- biomass109 or from electrochemical reduction of CO2,110 which
tones. Thus, in similarity to acid−base catalysis, LA hydro- makes FA an environmentally friendly source for both high-
genation proceeds via HPA as an intermediate. Use of purity hydrogen production22,111,112 and a hydrogen donor for
molecular hydrogen over metal catalysts, however, can lead to CTH reactions (Table 2). Two surface species have been
a substantial level of overhydrogenation of GVL to 2- proposed to directly participate in hydrogen transfer (Scheme
methyltetrahydrofuran (2-MTHF) or linear alcohols such as 14): (a) adsorbed hydrogen and (b) adsorbed formate species.
1,4-pentanediol and 2-butanol (Scheme 13).105 Thus, CTH
promises better control of the degree of hydrogenation by Scheme 14. Proposed Surface Adsorbed Hydrogen Donor in
maintaining low effective H2 pressure on the catalyst. CTH Reaction with FA: (a) Adsorbed Hydrogen Atom; (b)
While metals easily reduce CO bonds, lactonization is an Adsorbed Formate
inherent challenge over metal catalysts. This is evidenced by the
work of Hasan, where Pd/C alone was not capable of forming
any GVL without the addition of base (KOH or NaOH) and a
maximum yield of 83% yield of GVL was obtained using 2-
propanol as the hydrogen source in the presence of KOH
under microwave conditions.104 Metals supported on acidic
supports catalyze the cascade reaction without the need for
homogeneous acid or base.78,106,107 Fujitani et al. reported a
GVL yield of up to 89% from levulinic esters on TiO2-
supported Ru(OH)x catalysts over 24 h with 2-propanol as the
hydrogen source.106 A higher GVL yield was obtained on Ru/C Adsorbed atomic hydrogen formed via stepwise hydrogen
in comparison to Ru(OH)x at lower reaction temperature (80 transfer from FA to the metal surface113 appears to be a natural
vs 90 °C) and shorter time (9 vs 24 h), indicating that metallic choice; however, it fails to explain some results obtained from
isotopic labeling studies on Pd/C,114,115 one of the most widely
Ru sites could be more active than Ru(OH)x sites. Remarkably,
used CTH catalysts with FA. Although the carboxylic acid
a close to 100% yield of GVL was obtained on Raney Ni at
hydrogen is more acidic than the formyl hydrogen in FA, they
room temperature in 9 h, in comparison to 27% on Ru/C become indistinguishable upon deposition on the metal surface
under otherwise identical conditions.108 In contrast, Ni (Scheme 14a) and thus should exhibit equivalent reactivity.
supported on TiO2, SiO2, CeO2, and C is inactive. The initial However, in the CTH of methyl phenylpropiolate, HCOOD
hydrogenation step to HPA occurs easily over oxophilic metal and DCOOH led to primarily undeuterated and deuterated
sites such as Ru; however, the lactonization step proceeds much products, respectively (Scheme 15a),114 which suggests that the
more slowly on metal surfaces.107 The high GVL yields formyl hydrogen, rather than the carboxylic acid hydrogen, is
obtained on Raney Ni, especially at low temperatures (Table transferred to the substrate in CTH. Moreover, isomerization
1), could be attributed to the residual Al species in the Raney of enol ether compound 1 (Scheme 15b) with both DCOOH
Ni from the alkali corrosion process of Ni−Al alloy,75 which is and DCOOD resulted in the D incorporation at the carbon
consistent with the low activity for supported Ni catalysts. In adjacent to the methoxy group, whereas no D incorporation
light of the reaction mechanism for the HDO of furanic was observed with HCOOD, which further supports the
compounds on Ru/RuOx/C and the oxophilic nature of Ru, it hypothesis that the formyl hydrogen is the active species in
1430 DOI: 10.1021/acscatal.5b02171
ACS Catal. 2016, 6, 1420−1436
ACS Catalysis Review

Scheme 15. (a) Hydrogenation of Methyl Phenylpropiolate on Pd/C with DCOOH or HCOOD, (b) Isomerization of
Compound 1 on Pd/C in with Deuterated FAs, (c) Proposed Mechanism for the Isomerization of Compound 1 on Pd/C, and
(d) Potential Reaction Mechanism for (i) Hydrogenolysis of C−O Bonds, (ii) Hydrogenolysis of CO Bonds, and (iii)
Hydrogenation of CC Bonds with FA via CTH

hydrogen transfer. The fact that H/D exchange only occurs at leading to the hydrogenation of the CC bond. The proposed
the electron-deficient carbon bonded to the methoxy group mechanisms outlined in Scheme 15d could be verified by
indicates that the transferred hydrogen is a negatively charged isotopic labeling studies similar to those conducted by Spencer
hydride species. Adsorbed atomic hydrogen does not have any and co-workers.114,115
appreciable preference for the carbon atom next to the methoxy The hypothesis that the adsorbed formate directly
or the phenyl group,114 as the Pd−H bond was proved to be participates in the hydride transfer mechanism is consistent
amphipolar.116 On the basis of these observations, the with the fact that the addition of bases, e.g., triethylamine
isomerization reaction is proposed to proceed via an (Et3N), are frequently added in the CTH system with
intermolecular hydride transfer mechanism with a six- FA,70,114,115,117−119 which likely promote the formation of the
membered-ring intermediate, similar to the MPV mechanism adsorbed formate species (Scheme 15d). Moreover, the
(Scheme 15c): the electron-deficient carbon atom in the CC hydrogen-donating ability of formates was shown to be
bond coordinates with the formyl hydrogen (D in Scheme dependent on the ionic radius of the countercations, with all
15c), while the electron-rich carbon bonds with the Pd atom. organic and metal cations showing better CTH activity than
The hydride transfer is accompanied by the release of CO2 and protons.76,118,120 Formate with larger metal cations showed
the reduction of the CC bond to the C−C bond. Cleavage of higher activity for the hydrogenolysis of benzyl acetate, which
C−H and C−Pd bonds occurs after the rotation of the newly was attributed to the ease of separating the ions due to the
formed C−C bond, converting the substrate from a trans to a longer initial distance of the charge centers.76 However, FA
cis configuration. Several recent works proposed the negatively exhibits higher activity than formate salts in the CTH of α-
charged adsorbed hydride species as an intermediate in CTH methylbenzyl alcohol (MBA) on Pd/C, which could be
reactions,70,76,117 which could occur following similar pathways attributed to the role the proton plays in the dehydration
(Scheme 15d): (i) the formyl hydrogen transfers to the step.121
electron-deficient carbon next to the ether bond, leading to the [Link]. Levulinic Acid to γ-Valerolactone. In contrast to the
hydrogenolysis of the ether bond, (ii) the formyl hydrogen case of alcohols as hydrogen donors, many metal catalysts can
transfers to the electron-deficient carbonyl carbon, leading to efficiently mediate the reaction cascade from LA to GVL with
the hydrogenation of the CO bond, (iii) the formyl FA,122 though many of the same principles apply. Fe(0)
hydrogen transfers to the electron-deficient carbon in the carbonyls have been shown to be an effective class of catalysts
CC bond, e.g., β-C in α,β-unsaturated carbonyl systems,115 for hydrogenation of organic molecules and are a promising
1431 DOI: 10.1021/acscatal.5b02171
ACS Catal. 2016, 6, 1420−1436
ACS Catalysis Review

Scheme 16. Proposed Transesterification-Assisted Hydrolysis Mechanism in the Conversion of HMF to DMF

class of catalysts due to economic feasibility, commercial propanol, MeOH, and H2 (Table 3). As a polyol, glycerol itself
availability, and facile synthesis.123 Burtoloso et al. used can serve as a hydrogen source via APR and provide molecular
Fe3(CO)12 to produce GVL very selectively from LA. However,
similarly to the work of Hasan et al. using Pd/C catalysts and Table 3. Comparison of H-Donors in Glycerol
alcohols as the hydrogen source, a base (pyridine or imidazole) Hydrogenolysisa,18
was needed in addition to Fe3(CO)12 to obtain a maximum
yield of 92% at 180 °C; without added base, only 12% yield was
observed under identical conditions.123 While Au-based
catalysts are generally unable to activate alcohols as hydrogen
donors, they are remarkably efficient in activating FA for both substrate H donor
hydrogen generation124 and the conversion of LA to GVL via conversn conversn
CTH with quantitative yields.125,126 One-pot synthesis of GVL H donor (%) S1,2‑PDO (%) Sacetol (%) (%) SH2 (%)
from glucose, fructose, sucrose, starch, and cellulose with 30− FA 34 86 7 100 100
60% yields has been achieved on Au/ZrO2 catalysts.126 Ru- 2-propanol 28 77 15 46 8
based catalysts are also active in reducing LA and C6 sugars to MeOH 26 51 44 43 100
GVL,118,119,127,128 though the activity is slightly lower than Au/ H2 (45 bar) 35 91 2 N/A N/A
ZrO2 under similar conditions.126 While Ru supported on inert none 16 49 39
a
materials such as activated carbon are active in converting LA to Reaction conditions: 45 bar of N2 or H2 pressure, 493 K for 10 h, 20
GVL, Au/C and Au/SiO2 show almost no activity.126 The mL of 20 wt % glycerol aqueous solution, 0.5 g of Ni−Cu catalyst, 0.02
difference could be attributed to the Lewis acid sites present on mL/min donor solution feed rate.
Ru for the lactonization step due to the almost unavoidable
presence of the RuOx phase, while on Au-based catalysts, acid hydrogen for the hydrogenolysis reaction. However, the low
sites must be introduced by the support. Interestingly, ZrO2- efficiency of this process is evidenced by the low glycerol
supported Cu129 and Ag−Ni130 catalysts also show very high conversion (16%) and selectivity for 1,2-PDO (49%), which is
activity toward GVL production, albeit at higher temperatures. likely limited by a slow hydrogen production rate. Higher
[Link]. HDO of Oxygenated Furanics. Like alcohols, FA glycerol conversions were observed in the presence of hydrogen
has been successfully implemented in the upgrading of donors, with FA and high-pressure hydrogen (45 bar) being
oxygenated furanics, i.e. furfural and HMF, to their reduced superior to the alcohols for both glycerol conversion and 1,2-
forms, i.e. 2-MF and DMF, respectively. In particular, PDO selectivity. The dehydrogenation of alcohols is likely
Thananatthanachon and Rauchfuss explored the conversion limiting the rate of glycerol conversion, due to the high
of HMF to DMF, acquiring 95% yield to the HDO product activation barrier of dehydrogenating a primary alcohol
with FA in a mixture of THF (or DMSO) and H2SO4 over Pd/ (MeOH) and the competing dehydration and hydrogenation
C.131 In combination with FA’s ability to dehydrate fructose to pathways for 2-propanol to propene and propane, respectively.
HMF, the same group demonstrated the possibility of a one- 1,2-PDO selectivities and yields are comparable with high-
pot synthesis of DMF directly from fructose. Fructose is pressure hydrogen and FA, indicating that CTH can be a
typically dehydrated in the aqueous phase in the presence of competitive route with a suitable choice of hydrogen donor.
Brønsted acids, an environment that accelerates the humin Hydrogenolysis of the C−O bond in cyclic ethers has
formation from HMF.132,133 By utilization of FA as the attracted considerable interest, owing to the prospect of
hydrogen source in organic solvents, fructose can be producing linear alcohols, alkenes, and alkanes from biomass-
dehydrated and subsequently reduced to DMF in one pot,131 derived oxygenated furanic compounds: e.g., furfural and
eliminating the need to use biphasic systems for HMF HMF.136,137 Ebitani et al. showed that FA is an effective
isolation.134 Esterification is proposed to be a key step prior hydrogen donor in the production of 1,6-hexanediol (HDL)
to the hydrogenolysis of the C−O bond in the HDO of HMF from HMF over Pd/ZrP, achieving a maximum HDL yield of
to DMF (Scheme 16) on both Pd/C and Ru/C.131,135 The 43% from HMF in ethanol. This is a promising result because
cleavage of the C−OH bond adjacent to the furan ring is shown ring-opening chemistry is typically affiliated with high H2 partial
to occur after the formation of the in situ generated formate pressures (>3 MPa).138−140 Mechanisms previously proposed
esters: e.g., the formation of FMF and FMMF (Scheme 14). for furan ring opening suggest that a protonation step is needed
Three distinct roles of formic acid in this process were for appreciable selectivity in ring opening.141,142 The strongly
identified: (1) hydrogen donor, (2) acid catalyst, and (3) acidic ZrP is likely a major cause of the enhanced ring-opening
deoxygenation agent for furanylmethanols.131 activity, because less acidic supports used in the work, i.e.
[Link]. Hydrogenolysis of C−OH and Ether Bonds. FA is Nb2O3 and ZSM-5, demonstrated a considerable drop in HDL
also an effective hydrogen donor for the production of 1,2- yield.
PDO via the hydrogenolysis of glycerol. Güemez et al. [Link]. Organic Hydrogen Donors in the Depolymeriza-
compared the activity and product distribution on the Ni− tion and HDO of Lignin Feedstocks. Depolymerization of
Cu/Al2O3 catalyst among four different hydrogen donors under lignin and HDO of lignin-derived monomeric and dimeric
otherwise similar reaction conditions (220 °C, 10 h):18 FA, 2- phenolics both need to cleave C−O bonds. Although high-
1432 DOI: 10.1021/acscatal.5b02171
ACS Catal. 2016, 6, 1420−1436
ACS Catalysis Review

Scheme 17. Proposed Mechanism for Sorbose Formation from Glucose via a C1−C5 Intramolecular Hydride Transfer on Ti-
Betaa

a
Adapted with permission from ref 146. Copyright ©2015 American Chemical Society.

pressure hydrogen is by far the most frequently used hydrogen elegantly demonstrated that sorbose is produced through a C1−
donor in these processes, Song et al. showed that active C5 intramolecular hydride transfer (Scheme 17), while a C1−C2
hydrogen species formed from methanol, ethanol, and ethylene hydride transfer leads to the formation of fructose (Scheme
glycol on Ni/C was more effective than high-pressure hydrogen 4b). More interestingly, Sn-Beta almost exclusively catalyzes the
(5 MPa) in the depolymerization of birch sawdust.143 In C1−C2 hydride transfer of glucose to fructose in water, while a
addition, alcohols can stabilize the species formed upon 15% selectivity for mannose, formed via a C1−C2 intra-
cleaving C−O bonds. Furthermore, CTH of supercritical molecular carbon shift of glucose, was observed when the
methanol on Cu-doped porous metal oxides has also been reaction was carried out in methanol. Moreover, almost a factor
reported to be effective in lignin depolymerization through the of 4 increase in glucose conversion on Sn-Beta was observed in
organosolv process.144 More recently, CTH of vanillin, a methanol, in comparison to that in water. Increased solvent
derivative from lignin upgrade, with FA has been reported on a uptake into the hydrophobic pores in methanol, which results
supported Pd/Ag bimetallic catalyst, which involves the in better stabilization of the transition state, has been proposed
reduction of both CO and C−O bonds.145 However, as a possible factor contributing to the enhanced glucose
applications of CTH in lignin upgrading are still relatively rare. conversion in methanol.147
The choice of solvent has a substantial impact on the
4. SOLVENT EFFECTS catalytic activity and product distribution when formic acid or
In homogeneous CTH processes, solvent molecules often formate serves as the hydrogen donor; however, the mechanism
coordinate with the metal center of the soluble catalyst, thus through which solvents exert their influence is far from clear. In
potentially restricting the access of the reactants to the metal the hydrogenolysis from α-methylbenzyl alcohol to ethyl-
center and altering the electronic structure. Similar site- benzene with FA in EtOH on supported Pd catalysts, the
blocking effects are expected for heterogeneous CTH reactions presence of 20 vol % water as a cosolvent increases both the
when solvent molecules interact more strongly with the catalyst conversion of α-methylbenzyl alcohol and selectivity for
surface than the reactant molecules do. However, the effects of ethylbenzene.121 Enhanced ionizability of FA is speculated to
the solvent and the hydrogen donor are frequently intertwined be the reason for the beneficial effect of the presence of water as
when an alcohol is used as the hydrogen donor, because it a cosolvent. However, the selectivity of ethylbenzene decreases
typically also serves as the solvent. In this regard, Lewis acid when the volumetric fraction of water exceeds 20%, and the
catalyzed glucose isomerization via intramolecular hydride dehydrogenation pathway to acetophenone becomes more
transfer provides the rare opportunity of independently varying prominent. The product distribution of CTH of HMF with FA
the solvent, since the hydrogen donor and acceptor are two as the hydrogen donor on supported Pd catalysts could be
functional groups in the substrate. Davis and co-workers drastically different with different solvents:131,142 the formate
reported that fructose was the favored product when the ester of HMF (FMF in Scheme 16) is converted quantitatively
isomerization of glucose on Ti-Beta was carried out in water, to DMF in tetrahydrofuran on Pd/C with added sulfuric acid,
while sorbose became the major product when the reaction was while ring-opening products, e.g., HDL and 2,5-hexanedione,
conducted in methanol under otherwise identical conditions.146 account for almost half of the products on Pd/ZrP in EtOH.
In addition, a 15−125% increase in the turnover number was Since both reaction systems have protons (homogeneous and
observed in methanol in comparison to water as the solvent. A heterogeneous) and Pd sites, the role solvents play could be an
combined isotopic labeling, 13C NMR, and kinetic study important factor leading to the difference in product
1433 DOI: 10.1021/acscatal.5b02171
ACS Catal. 2016, 6, 1420−1436
ACS Catalysis Review

distributions. However, more rigorous comparative studies are conditions is highly desirable, because the signals from bulk
needed to elucidate the solvent effect. solvents often overwhelm the spectral features from the
interface. Furthermore, a detailed knowledge of the similarity
5. PERSPECTIVES and differences in mechanisms of CTH and hydrogen gas based
The wide range of reactions and intricate mechanistic pathways HDO reactions will aid the choice of a suitable approach for
discussed in sections 2−4 have demonstrated the breadth and target reactions.
depth of CTH as an effective approach to upgrade biomass- 5.4. Mechanisms through Which Solvents Interact
derived feedstocks. The employment of organic hydrogen with Catalytic Sites and Affect Reaction Rates. Properties
donors, e.g., secondary alcohols and FA, introduces alternative of active sites at the solid−liquid interface could be drastically
mechanistic pathways other than those based on hydrogen gas, different from those measured with ex situ vapor-phase
which adds extra dimensions in manipulating the product techniques: e.g., CO2/NH3-TPD and CO chemisorption. In
distribution. Moreover, the prospect of using renewable organic situ characterization tools and computational methods are
hydrogen donors could further reduce the carbon footprint of needed to accurately probe how the presence of solvent affects
fuel and chemicals produced from biomass by removing the the interaction between reactants and active sites. The influence
dependence on natural-gas-derived hydrogen. In order for of the coordination of solvent molecules on the active sites
CTH to make a real impact in biomass conversion, it is critical should be correlated with rates and product distributions to
that CTH-based processes be either competitive with or identify predictive descriptors for the solvent choices.
complementary to the conventional high-pressure hydrogen gas 5.5. Upgrading Lignin-Derived Feedstocks. While CTH
based HDO processes, which will stay dominant in the has been increasingly regarded as an alternative in the upgrade
foreseeable future. The versatility of hydrogen donors, solvents, of (hemi)cellulose-based biomass to molecular-hydrogen-based
and mechanistic pathways of CTH afford the possibility of routes, its applications in the HDO of lignin-derived feedstocks
developing highly efficient and selective processes; meanwhile, remain relatively unexplored. CTH has the potential to
a better understanding of the following five aspects through selectively activate C−O bonds in the depolymerization of
continued research efforts is needed to fully realize CTH’s lignin and HDO of phenolic monomers and dimers.
potential.
5.1. Key Descriptors for Metal Sites’ Ability to
Activate Organic Hydrogen Donors in the Absence of
■ AUTHOR INFORMATION
Corresponding Author
Coexisting Acid or Base Sites. Activation of organic *E-mail for B.X.: bxu@[Link].
hydrogen donors on metallic sites typically involves adsorption,
Notes
C−H/O−H bond cleavage, and metal−H bond formation , and
The authors declare no competing financial interest.

■
metal−H bond cleavage on reaction with the substrate. The
ability of metal to facilitate each of these steps will affect the
overall efficiency of CTH reactions. Thus, descriptors such as ACKNOWLEDGMENTS
adsorption energy of common hydrogen donors, e.g., alcohols We acknowledge support from the Catalysis Center for Energy
and FA, hydrogen binding energy, and oxophilicity have the Innovation, an Energy Frontier Research Center funded by the
potential to predict metal effectiveness in CTH. U.S. Department of Energy, Office of Science, Office of Basic
5.2. Synergy between Metal and Acid−Base Sites. The Energy Sciences, under Award No. DE-SC0001004.
roles of metal and acid−base sites in CTH need to be
elucidated. A detailed knowledge of the nature (metal, Brønsted
or Lewis acid, or base sites), strength, and density of sites is
■ REFERENCES
(1) Zakzeski, J.; Bruijnincx, P. C. A.; Jongerius, A. L.; Weckhuysen, B.
essential to establish structure−activity relationships. Most M. Chem. Rev. 2010, 110, 3552−3599.
heterogeneous CTH reactions are carried out in the liquid (2) Chheda, J. N.; Dumesic, J. A. Catal. Today 2007, 123, 59−70.
phase, and the properties of catalytic sites in the presence of (3) Brieger, G.; Nestrick, T. J. Chem. Rev. 1974, 74, 567−580.
solvents would be drastically different from those in the absence (4) Brieger, G.; Nestrick, T. J.; Fu, T.-H. J. Org. Chem. 1979, 44,
of solvents. Therefore, characterization capable of taking 1876−1878.
solvent effects on catalytic sites into account will be highly (5) Cortese, N. A.; Heck, R. F. J. Org. Chem. 1978, 43, 3985−3987.
desirable. In addition, the systematic employment of catalysts (6) Johnstone, R. A. W.; Wilby, A. H. Tetrahedron 1981, 37, 3667−
3670.
with tailored site compositions and well-defined structures, e.g.,
(7) Weir, J. R.; Patel, B. A.; Heck, R. F. J. Org. Chem. 1980, 45, 4926−
size-controlled metal nanoparticles supported on Lewis/ 4931.
Brønsted acidic supports, in diagnostic model reactions will (8) Aramendía, M. a. A.; Borau, V.; Jiménez, C.; Marinas, J. M.; Ruiz,
be helpful in elucidating the synergistic effects among different J. R.; Urbano, F. J. J. Mol. Catal. A: Chem. 2001, 171, 153−158.
types of sites. (9) Jae, J.; Mahmoud, E.; Lobo, R. F.; Vlachos, D. G. ChemCatChem
5.3. Molecular Level Reaction Mechanism. Mechanisms 2014, 6, 508−513.
of CTH reactions are typically the fusion of two catalytic cycles, (10) Panagiotopoulou, P.; Vlachos, D. G. Appl. Catal., A 2014, 480,
i.e. the activation of the organic hydrogen donor and the 17−24.
reactant, which can be either sequential or concerted steps. (11) Entwistle, I. D.; Johnstone, R. A. W.; Povall, T. J. J. Chem. Soc.,
Combined kinetic, in situ spectroscopic, isotopic labeling, and Perkin Trans. 1 1975, 1300−1301.
(12) Gowda, D. C.; Mahesh, B. Synth. Commun. 2000, 30, 3639−
computational investigations are needed to identify the active
3644.
hydrogen species involved in the hydrogen transfer step and (13) Gowda, S.; Gowda, D. C. Tetrahedron 2002, 58, 2211−2213.
achieve a comprehensive understanding of the reaction pathway (14) Mohapatra, S. K.; Sonavane, S. U.; Jayaram, R. V.; Selvam, P.
on the molecular level. In particular, the development of novel Org. Lett. 2002, 4, 4297−4300.
interfacial sensitive spectroscopic tools capable of identifying (15) Gandarias, I.; Requies, J.; Arias, P. L.; Armbruster, U.; Martin, A.
reaction intermediates at solid/liquid interfaces under reaction J. Catal. 2012, 290, 79−89.

1434 DOI: 10.1021/acscatal.5b02171

ACS Catal. 2016, 6, 1420−1436
ACS Catalysis Review

(16) Liu, X.; Lu, G.; Guo, Y.; Guo, Y.; Wang, Y.; Wang, X. J. Mol. (50) Aramendía, M. a. A.; Borau, V.; Jiménez, C.; Marinas, J. M.;
Catal. A: Chem. 2006, 252, 176−180. Ruiz, J. R.; Urbano, F. J. Appl. Catal., A 2003, 255, 301−308.
(17) Scholz, D.; Aellig, C.; Hermans, I. ChemSusChem 2014, 7, 268− (51) Ivanov, V. A.; Bachelier, J.; Audry, F.; Lavalley, J. C. J. Mol.
75. Catal. 1994, 91, 45−59.
(18) Gandarias, I.; Arias, P. L.; Fernández, S. G.; Requies, J.; El (52) Ruiz, J. R.; Jiménez-Sanchidrián, C.; Hidalgo, J. M. Catal.
Doukkali, M.; Güemez, M. B. Catal. Today 2012, 195, 22−31. Commun. 2007, 8, 1036−1040.
(19) Kibby, C. L.; Swift, H. E. J. Catal. 1976, 45, 231−241. (53) Szöllösi, G.; Bartók, M. Appl. Catal., A 1998, 169, 263−269.
(20) Cortese, N. A.; Heck, R. F. J. Org. Chem. 1977, 42, 3491−3494. (54) Jyothi, T. M.; Raja, T.; Rao, B. S. J. Mol. Catal. A: Chem. 2001,
(21) Johnstone, R. A. W.; Wilby, A. H.; Entwistle, I. D. Chem. Rev. 168, 187−191.
1985, 85, 129−170. (55) Ruiz, J. R.; Jiménez-Sanchidrián, C.; Hidalgo, J. M.; Marinas, J.
(22) Johnson, T. C.; Morris, D. J.; Wills, M. Chem. Soc. Rev. 2010, 39, M. J. Mol. Catal. A: Chem. 2006, 246, 190−194.
81−88. (56) Radhakrishan, R.; Do, D. M.; Jaenicke, S.; Sasson, Y.; Chuah, G.-
(23) Robertson, A.; Matsumoto, T.; Ogo, S. Dalton Trans. 2011, 40, K. ACS Catal. 2011, 1, 1631−1636.
10304−10310. (57) Gliński, M. Appl. Catal., A 2008, 349, 133−139.
(24) Samec, J. S. M.; Bäckvall, J.-E.; Andersson, P. G.; Brandt, P. (58) Heidari, H.; Abedini, M.; Nemati, A.; Amini, M. M. Catal. Lett.
Chem. Soc. Rev. 2006, 35, 237−248. 2009, 130, 266−270.
(25) Clapham, S. E.; Hadzovic, A.; Morris, R. H. Coord. Chem. Rev. (59) Kijeński, J.; Glińksi, M.; Czarnecki, J. J. Chem. Soc., Perkin Trans.
2004, 248, 2201−2237. 2 1991, 1695−1698.
(26) Boronat, M.; Corma, A.; Renz, M. J. Phys. Chem. B 2006, 110, (60) Szöllösi, G.; Bartók, M. J. Mol. Catal. A: Chem. 1999, 148, 265−
21168−21174. 273.
(27) Corma, A.; Domine, M. E.; Nemeth, L.; Valencia, S. J. Am. (61) Szöllösi, G.; Bartók, M. Catal. Lett. 1999, 59, 179−185.
Chem. Soc. 2002, 124, 3194−3195. (62) Chuah, G. K.; Jaenicke, S.; Zhu, Y. Z. Curr. Org. Chem. 2006, 10,
(28) Luo, H. Y.; Consoli, D. F.; Gunther, W. R.; Román-Leshkov, Y. 1639−1654.
J. Catal. 2014, 320, 198−207. (63) Liu, S. H.; Jaenicke, S.; Chuah, G. K. J. Catal. 2002, 206, 321−
(29) Moliner, M. Dalton Trans. 2014, 43, 4197−4208. 330.
(30) Moliner, M.; Román-Leshkov, Y.; Davis, M. E. Proc. Natl. Acad. (64) Matsushita, H.; Ishiguro, S.; Ichinose, H.; Izumi, A.; Mizusaki, S.
Sci. U. S. A. 2010, 107, 6164−6168. Chem. Lett. 1985, 14, 731−734.
(31) Román-Leshkov, Y.; Moliner, M.; Labinger, J. A.; Davis, M. E. (65) Zhu, Y.; Liu, S.; Jaenicke, S.; Chuah, G. Catal. Today 2004, 97,
Angew. Chem., Int. Ed. 2010, 49, 8954−8957. 249−255.
(32) Bermejo-Deval, R.; Assary, R. S.; Nikolla, E.; Moliner, M.; (66) Chuah, G.; Jaenicke, S.; Cheong, S.; Chan, K. Appl. Catal., A
Román-Leshkov, Y.; Hwang, S.-J.; Palsdottir, A.; Silverman, D.; Lobo,
1996, 145, 267−284.
R. F.; Curtiss, L. A.; Davis, M. E. Proc. Natl. Acad. Sci. U. S. A. 2012, (67) Ono, Y. H.; Hattori, H. Solid Base Catalysis. In Springer Series in
109, 9727−9732.
Chemical Physics; Castleman, A. W., Toennies, J. P., Yamanouchi, K.,
(33) Bermejo-Deval, R.; Gounder, R.; Davis, M. E. ACS Catal. 2012,
Zinth, W., Eds.; Springer: New York, 2011; Vol. 101.
2, 2705−2713.
(68) Huang, L.; Zhu, Y.; Huo, C.; Zheng, H.; Feng, G.; Zhang, C.; Li,
(34) Bermejo-Deval, R.; Orazov, M.; Gounder, R.; Hwang, S.-J.;
Y. J. Mol. Catal. A: Chem. 2008, 288, 109−115.
Davis, M. E. ACS Catal. 2014, 4, 2288−2297.
(69) Jae, J.; Zheng, W.; Lobo, R. F.; Vlachos, D. G. ChemSusChem
(35) Yang, G.; Pidko, E. A.; Hensen, E. J. M. ChemSusChem 2013, 6,
1688−1696. 2013, 6, 1158−1162.
(36) Li, G.; Pidko, E. A.; Hensen, E. J. M. Catal. Sci. Technol. 2014, 4, (70) Sawadjoon, S.; Lundstedt, A.; Samec, J. S. M. ACS Catal. 2013,
2241−2250. 3, 635−642.
(37) Rai, N.; Caratzoulas, S.; Vlachos, D. G. ACS Catal. 2013, 3, (71) Gandarias, I.; Arias, P. L.; Requies, J.; El Doukkali, M.; Güemez,
2294−2298. M. B. J. Catal. 2011, 282, 237−247.
(38) Conrad, S.; Verel, R.; Hammond, C.; Wolf, P.; Göltl, F.; (72) Pellet, R. J. J. Catal. 1998, 177, 40−52.
Hermans, I. ChemCatChem 2015, 7, 3270−3278. (73) Subramanian, T.; Pitchumani, K. ChemCatChem 2012, 4, 1917−
(39) Li, Y.-P.; Head-Gordon, M.; Bell, A. T. ACS Catal. 2014, 4, 1921.
1537−1545. (74) Hansen, T. S.; Barta, K.; Anastas, P. T.; Ford, P. C.; Riisager, A.
(40) Assary, R. S.; Curtiss, L. A. J. Phys. Chem. A 2011, 115, 8754− Green Chem. 2012, 14, 2457−2461.
8760. (75) Geboers, J.; Wang, X.; de Carvalho, A. B.; Rinaldi, R. J. Mol.
(41) Choudhary, V.; Pinar, A. B.; Sandler, S. I.; Vlachos, D. G. ACS Catal. A: Chem. 2014, 388−389, 106−115.
Catal. 2011, 1, 1724−1728. (76) Rajagopal, S.; Spatola, A. F. Appl. Catal., A 1997, 152, 69−81.
(42) Wang, J.; Jaenicke, S.; Chuah, G.-K. RSC Adv. 2014, 4, 13481− (77) Panagiotopoulou, P.; Martin, N.; Vlachos, D. G. J. Mol. Catal. A:
13489. Chem. 2014, 392, 223−228.
(43) Chia, M.; Dumesic, J. A. Chem. Commun. 2011, 47, 12233− (78) Yang, Z.; Huang, Y.-B.; Guo, Q.-X.; Fu, Y. Chem. Commun.
12235. 2013, 49, 5328−5330.
(44) Tang, X.; Zeng, X.; Li, Z.; Li, W.; Jiang, Y.; Hu, L.; Liu, S.; Sun, (79) Zhu, Y.-L.; Xiang, H.-W.; Li, Y.-W.; Jiao, H.; Wu, G.-S.; Zhong,
Y.; Lin, L. ChemCatChem 2015, 7, 1372−1379. B.; Guo, G.-Q. New J. Chem. 2003, 27, 208−210.
(45) Hao, W.; Li, W.; Tang, X.; Zeng, X.; Sun, Y.; Liu, S.; Lin, L. (80) Gilkey, M. J.; Panagiotopoulou, P.; Mironenko, A. V.; Jenness,
Green Chem. 2016, DOI: 10.1039/C5GC01221J. G. R.; Vlachos, D. G.; Xu, B. ACS Catal. 2015, 5, 3988−3994.
(46) Antunes, M. M.; Lima, S.; Neves, P.; Magalhães, A. L.; Fazio, E.; (81) Mironenko, A. V.; Gilkey, M. J.; Panagiotopoulou, P.; Facas, G.;
Fernandes, A.; Neri, F.; Silva, C. M.; Rocha, S. M.; Ribeiro, M. F.; Vlachos, D. G.; Xu, B. J. Phys. Chem. C 2015, 119, 6075−6085.
Pillinger, M.; Urakawa, A.; Valente, A. A. J. Catal. 2015, 329, 522−537. (82) Panagiotopoulou, P.; Martin, N.; Vlachos, D. G. ChemSusChem
(47) Bui, L.; Luo, H.; Gunther, W. R.; Roman-Leshkov, Y. Angew. 2015, 8, 2046−2054.
Chem., Int. Ed. 2013, 52, 8022−8025. (83) Martin, A.; Armbruster, U.; Gandarias, I.; Arias, P. L. Eur. J.
(48) Lewis, J. D.; Van de Vyver, S.; Crisci, A. J.; Gunther, W. R.; Lipid Sci. Technol. 2013, 115, 9−27.
Michaelis, V. K.; Griffin, R. G.; Román-Leshkov, Y. ChemSusChem (84) Huber, G. W.; S, J. W.; Dumesic, J. A. Science 2003, 300, 2075−
2014, 7, 2255−2265. 2077.
(49) Aramendía, M. A.; Borau, V.; Jiménez, C.; Marinas, J. M.; Ruiz, (85) Wawrzetz, A.; Peng, B.; Hrabar, A.; Jentys, A.; Lemonidou, A.
J. R.; Urbano, F. J. J. Colloid Interface Sci. 2001, 238, 385−389. A.; Lercher, J. A. J. Catal. 2010, 269, 411−420.

1435 DOI: 10.1021/acscatal.5b02171

ACS Catal. 2016, 6, 1420−1436
ACS Catalysis Review

(86) Zhang, L.; Karim, A. M.; Engelhard, M. H.; Wei, Z.; King, D. L.; (122) Tang, X.; Chen, H.; Hu, L.; Hao, W.; Sun, Y.; Zeng, X.; Lin, L.;
Wang, Y. J. Catal. 2012, 287, 37−43. Liu, S. Appl. Catal., B 2014, 147, 827−834.
(87) Auneau, F.; Noel, S.; Aubert, G.; Besson, M.; Djakovitch, L.; (123) Metzker, G.; Burtoloso, A. C. Chem. Commun. 2015, 51,
Pinel, C. Catal. Commun. 2011, 16, 144−149. 14199−14202.
(88) Barbelli, M. L.; Santori, G. F.; Nichio, N. N. Bioresour. Technol. (124) Ojeda, M.; Iglesia, E. Angew. Chem. 2009, 121, 4894−4897.
2012, 111, 500−503. (125) Du, X.-L.; Bi, Q.-Y.; Liu, Y.-M.; Cao, Y.; Fan, K.-N.
(89) D’Hondt, E.; Van de Vyver, S.; Sels, B. F.; Jacobs, P. A. Chem. ChemSusChem 2011, 4, 1838−1843.
Commun. 2008, 6011−6012. (126) Du, X.-L.; He, L.; Zhao, S.; Liu, Y.-M.; Cao, Y.; He, H.-Y.; Fan,
(90) Roy, D.; Subramaniam, B.; Chaudhari, R. V. Catal. Today 2010, K.-N. Angew. Chem. 2011, 123, 7961−7965.
156, 31−37. (127) Heeres, H.; Handana, R.; Chunai, D.; Rasrendra, C. B.;
(91) Montassier, C.; Giraud, D.; Barbier, J.; Boitiaux, J. P. Bull. Soc. Girisuta, B.; Heeres, H. J. Green Chem. 2009, 11, 1247−1255.
Chim. Fr. 1989, 2, 148−155. (128) Ortiz-Cervantes, C.; García, J. J. Inorg. Chim. Acta 2013, 397,
(92) Montassier, C.; Ménézo, J. C.; Hoang, L. C.; Renaud, C.; 124−128.
Barbier, J. J. Mol. Catal. 1991, 70, 99−110. (129) Yuan, J.; Li, S.-S.; Yu, L.; Liu, Y.-M.; Cao, Y.; He, H.-Y.; Fan,
(93) Montassier, C.; Ménézo, J. C.; Moukolo, J.; Naja, J.; Hoang, L. K.-N. Energy Environ. Sci. 2013, 6, 3308−3313.
C.; Barbier, J.; Boitiaux, J. P. J. Mol. Catal. 1991, 70, 65−84. (130) Hengne, A. M.; Malawadkar, A. V.; Biradar, N. S.; Rode, C. V.
(94) Wang, S.; Zhang, Y.; Liu, H. Chem. - Asian J. 2010, 5, 1100− RSC Adv. 2014, 4, 9730−9736.
1111. (131) Thananatthanachon, T.; Rauchfuss, T. B. Angew. Chem., Int. Ed.
2010, 49, 6616−6618.
(95) Lahr, D. G.; Shanks, B. H. Ind. Eng. Chem. Res. 2003, 42, 5467−
(132) Mushrif, S. H.; Caratzoulas, S.; Vlachos, D. G. Phys. Chem.
5472.
Chem. Phys. 2012, 14, 2637−2644.
(96) Maris, E.; Ketchie, W.; Murayama, M.; Davis, R. J. Catal. 2007,
(133) Tsilomelekis, G.; Josephson, T. R.; Nikolakis, V.; Caratzoulas,
251, 281−294. S. ChemSusChem 2014, 7, 117−126.
(97) Dasari, M. A.; Kiatsimkul, P.-P.; Sutterlin, W. R.; Suppes, G. J. (134) Ordomsky, V. V.; van der Schaaf, J.; Schouten, J. C.; Nijhuis, T.
Appl. Catal., A 2005, 281, 225−231. A. ChemSusChem 2012, 5, 1812−1819.
(98) Sharninghausen, L. S.; Mercado, B. Q.; Crabtree, R. H.; Hazari, (135) De, S.; Dutta, S.; Saha, B. ChemSusChem 2012, 5, 1826−1833.
N. Chem. Commun. 2015, 51, 16201−16204. (136) Sutton, A. D.; Waldie, F. D.; Wu, R.; Schlaf, M.; Silks, L. A.,
(99) Kusunoki, Y.; Miyazawa, T.; Kunimori, K.; Tomishige, K. Catal. 3rd; Gordon, J. C. Nat. Chem. 2013, 5, 428−432.
Commun. 2005, 6, 645−649. (137) Chia, M.; Pagan-Torres, Y. J.; Hibbitts, D.; Tan, Q.; Pham, H.
(100) Miyazawa, T.; Kusunoki, Y.; Kunimori, K.; Tomishige, K. J. N.; Datye, A. K.; Neurock, M.; Davis, R. J.; Dumesic, J. A. J. Am. Chem.
Catal. 2006, 240, 213−221. Soc. 2011, 133, 12675−89.
(101) Mane, R. B.; Rode, C. V. Org. Process Res. Dev. 2012, 16, 1043− (138) Liu, S.; Amada, Y.; Tamura, M.; Nakagawa, Y.; Tomishige, K.
1052. Green Chem. 2014, 16, 617−626.
(102) Musolino, M. G.; Scarpino, L. A.; Mauriello, F.; Pietropaolo, R. (139) Mizugaki, T.; Yamakawa, T.; Nagatsu, Y.; Maeno, Z.;
Green Chem. 2009, 11, 1511−1513. Mitsudome, T.; Jitsukawa, K.; Kaneda, K. ACS Sustainable Chem.
(103) Abdelrahman, O. A.; Heyden, A.; Bond, J. Q. ACS Catal. 2014, Eng. 2014, 2, 2243−2247.
4, 1171−1181. (140) Yao, S.; Wang, X.; Jiang, Y.; Wu, F.; Chen, X.; Mu, X. ACS
(104) Amarasekara, A. S.; Hasan, M. A. Catal. Commun. 2015, 60, 5− Sustainable Chem. Eng. 2014, 2, 173−180.
7. (141) Chia, M.; Pagan-Torres, Y. J.; Hibbitts, D.; Tan, Q.; Pham, H.
(105) Al-Shaal, M. G.; Dzierbinski, A.; Palkovits, R. Green Chem. N.; Datye, A. K.; Neurock, M.; Davis, R. J.; Dumesic, J. A. J. Am. Chem.
2014, 16, 1358−1364. Soc. 2011, 133, 12675−12689.
(106) Kuwahara, Y.; Kaburagi, W.; Fujitani, T. RSC Adv. 2014, 4, (142) Tuteja, J.; Choudhary, H.; Nishimura, S.; Ebitani, K.
45848−45855. ChemSusChem 2014, 7, 96−100.
(107) Tang, X.; Zeng, X.; Li, Z.; Hu, L.; Sun, Y.; Liu, S.; Lei, T.; Lin, (143) Song, Q.; Wang, F.; Cai, J.; Wang, Y.; Zhang, J.; Yu, W.; Xu, J.
L. Renewable Sustainable Energy Rev. 2014, 40, 608−620. Energy Environ. Sci. 2013, 6, 994−1007.
(108) Yang, Z.; Huang, Y. B.; Guo, Q. X.; Fu, Y. Chem. Commun. (144) Barta, K.; Matson, T. D.; Fettig, M. L.; Scott, S. L.; Iretskii, A.
2013, 49, 5328−30. V.; Ford, P. C. Green Chem. 2010, 12, 1640−1647.
(109) Jin, F.; Yun, J.; Li, G.; Kishita, A.; Tohji, K.; Enomoto, H. Green (145) Singh, A. K.; Jang, S.; Kim, J. Y.; Sharma, S.; Basavaraju, K. C.;
Chem. 2008, 10, 612−615. Kim, M.-G.; Kim, K.-R.; Lee, J. S.; Lee, H. H.; Kim, D.-P. ACS Catal.
(110) Leitner, W. Angew. Chem., Int. Ed. Engl. 1995, 34, 2207−2221. 2015, 5, 6964−6972.
(111) Enthaler, S.; von Langermann, J.; Schmidt, T. Energy Environ. (146) Gounder, R.; Davis, M. E. ACS Catal. 2013, 3, 1469−1476.
Sci. 2010, 3, 1207−1217. (147) Christianson, J. R.; Caratzoulas, S.; Vlachos, D. G. ACS Catal.
(112) Grasemann, M.; Laurenczy, G. Energy Environ. Sci. 2012, 5, 2015, 5, 5256−5263.
8171−8181.
(113) Luo, Q.; Feng, G.; Beller, M.; Jiao, H. J. Phys. Chem. C 2012,
116, 4149−4156.
(114) Yu, J.; Spencer, J. B. Chem. Commun. 1998, 1935−1936.
(115) Yu, J.; Spencer, J. B. Chem. - Eur. J. 1999, 5, 2237−2240.
(116) Yu, J.; Spencer, J. B. J. Am. Chem. Soc. 1997, 119, 5257−5258.
(117) Bi, Q.-Y.; Du, X.-L.; Liu, Y.-M.; Cao, Y.; He, H.-Y.; Fan, K.-N.
J. Am. Chem. Soc. 2012, 134, 8926−8933.
(118) Deng, L.; Li, J.; Lai, D. M.; Fu, Y.; Guo, Q.-X. Angew. Chem.,
Int. Ed. 2009, 48, 6529−6532.
(119) Deng, L.; Zhao, Y.; Li, J.; Fu, Y.; Liao, B.; Guo, Q.-X.
ChemSusChem 2010, 3, 1172−1175.
(120) Rajagopal, S.; Spatola, A. F. J. Org. Chem. 1995, 60, 1347−
1355.
(121) Feng, J.; Yang, C.; Zhang, D.; Wang, J.; Fu, H.; Chen, H.; Li, X.
Appl. Catal., A 2009, 354, 38−43.

1436 DOI: 10.1021/acscatal.5b02171

ACS Catal. 2016, 6, 1420−1436