Production of dimethylfuran from hydroxymethylfurfural through catalytic transfer hydrogenation with ruthenium supported on carbon

Article abstract of DOI:10.1002/cssc.201300288

RuCees' transfer: Transfer hydrogenation using alcohols as hydrogen donors and supported ruthenium catalysts results in the selective conversion of hydroxymethylfurfural to dimethylfuran (>80 % yield). During transfer hydrogenation, the hydrogen produced from alcohols is utilized in the hydrogenation of hydroxymethylfurfural. Copyright

Full text of DOI:10.1002/cssc.201300288

CHEMSUSCHEM
COMMUNICATIONS
DOI: 10.1002/cssc.201300288
Production of Dimethylfuran from Hydroxymethylfurfural
through Catalytic Transfer Hydrogenation with Ruthenium
Supported on Carbon
Jungho Jae, Weiqing Zheng, Raul F. Lobo,* and Dionisios G. Vlachos*[a]
The hydrogenation of biomass-derived 5-(hydroxymethyl) fur-
fural (HMF) to 2,5-dimethylfuran (DMF) is a key reaction in up-
grading biomass-derived furanic compounds into platform
molecules toward the production of chemicals and liquid
transportation fuels. Several groups have studied the hydroge-
nation of HMF to DMF over various metal catalysts (Cu, Pd, Ni,
and Ru) using molecular hydro-
CTH of HMF using formic acid results in a good yield of DMF
(70%) over Pd/C catalyst and sulfuric acid.[2b] Hansen et al. re-
ported that the CTH of HMF using supercritical methanol over
a Cu-doped metal oxide catalyst yielded 48% DMF and 10%
to 2,5-dimethyltetrahydrofuran (DMTHF).[2d] These CTH routes
have, however, several disadvantages: they require the use of
gen.[1] It has been shown that
copper-based catalysts, carbon–
supported
copper–ruthenium
and copper chromite, are selec-
tive for DMF production (71%
yield).[1g] However, hydrogena-
tion utilizing petroleum-derived
H2 is neither entirely green nor
economic because it requires
high-pressure H2, which increas-
es process costs. Herein, we
report an alternative route for
the conversion of HMF to DMF
through catalytic transfer hydro-
genation (CTH), in which a hydro-
gen donor is used as a hydrogen
source, instead of molecular H2,
without the need for a Brønsted
acid catalyst commonly em-
ployed in hydrogenolysis reac-
tions.
Several investigators have
studied the CTH of HMF to DMF
using hydrogen donors over
metal catalysts.[2] Morikawa et al.
reported CTH of HMF using cy-
clohexane over Pd/C catalyst
and AlCl3 with a DMF yield up
to 60%.[2a] Thananatthanachon
et al. recently showed that the
Scheme 1. Reaction network of the hydrogenation of HMF into DMF using 2-propanol and Ru/C. Compounds:
5-(hydroxymethyl)furfural (HMF); 2,5-dimethylfuran (DMF); 2,5-bis(hydroxymethyl)furan (BHMF); 5-methyl furfuryl
alcohol (MFA); furfuryl alcohol (FA); 2-methyl furan (MF); 5-[(1-methylethoxy) methyl] furfural (1), 5-hydroxymeth-
yl-2-[(1-methylethoxy) methyl] furan (2), 2-methyl-5-[(1-methylethoxy)methyl]furan (3), 2,5-[bis(1-methylethoxy)-
methyl] furan (4).
homogeneous acid co-catalysts to facilitate the hydrogena-
[a] Dr. J. Jae, Dr. W. Zheng, Prof. R. F. Lobo, Prof. D. G. Vlachos
Catalysis Center for Energy Innovation
Department of Chemical and Biomolecular Engineering
University of Delaware, Newark, DE, 19716 (USA)
Fax: (+1)302-831-1048
tion[2b,c] that are difficult to separate from the reaction mixture
and lead to undesired furan-ring saturation due to the harsh
reaction conditions.[2d]
We have investigated the CTH of HMF using secondary alco-
hols and have shown that DMF is selectively produced with up
to 80% yield over a carbon-supported Ru catalyst (Scheme 1).
Several groups have used secondary alcohols for the CTH of
other biomass-derived molecules.[3] Kobayashi et al. reported
E-mail: lobo@udel.edu
vlachos@udel.edu
Homepage: http://www.efrc.udel.edu/
Supporting Information for this article is available on the WWW under
http://dx.doi.org/10.1002/cssc.201300288.
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2013, 6, 1158 – 1162 1158

CHEMSUSCHEM
COMMUNICATIONS
www.chemsuschem.org
the CTH of cellulose to sugar alcohols (sorbitol and mannitol)
with 45% yield using 2-propanol over Ru/C catalyst.[3a] Chia
and Dumesic reported the CTH of levulinic acid and its esters
to g-valerolactone with 92% yield using 2-butanol over ZrO2
catalyst.[3c] Gandarias et al. reported the hydrogenation of glyc-
erol to 1,2-propanediol with 65% selectivity using 2-propanol
over a Ni–Cu/Al2O3 catalyst.[3d]
the DMF increases with increasing reaction temperature.
At low temperatures, the primary product is 2,5-bis(hydroxy-
methyl)furan (BHMF) and upon increasing the reaction temper-
ature to 463 K, BHMF is completely converted to DMF with
a selectivity of up to 81%. Unlike a previous report in which
Ru/C was shown to be effective in producing 2,5-(dihydroxy-
methyl)tetrahydrofuran (DHMTHF),[4] we find that under our re-
action conditions it is possible to achieve the more difficult hy-
drogenation of HMF to DMF with the highest yield reported
without using a Brønsted catalyst. Other products observed at
403 and 433 K are hydrogenated furans: 5-methyl furfuryl alco-
hol (MFA), 5-methyl furfural, furfuryl alcohol (FA), and 2-methyl
furan (MF). The formation of 5-methyl furfural indicates that
hydrogenolysis of the alcohol group of HMF occurs in parallel
to the hydrogenation of the aldehyde group. The formation of
FA and MF indicates the occurrence of the decarbonylation re-
action; the selectivity to these products, however, is low
(<1.5%). In addition, two ethers, 2-methyl-5-[(1-methylethoxy)
methyl]furan (3) and 5-hydroxymethyl-2-[(1-methyl ethoxy)me-
thyl]furan (2), are observed at 403 and 433 K with a selectivity
of up to 9.0% and 9.2%, respectively. These ethers are formed
via etherification of 2-propanol and BHMF or MFA and can be
used as fuel additives. They are
Alcohols offer important advantages as hydrogen sources:
they can serve as reactants and solvents; they can be sustaina-
bly produced from biomass; and the aldehydes or ketones, de-
hydrogenated products of alcohols, can easily be recycled by
hydrogenation over base metals[3c] or used as chemicals. The
family of hydrogen donors we exploit herein provides the first
potentially fully green process of converting HMF to DMF by
using a single heterogeneous catalyst (without homogeneous
acids) at a high yield.
Initial reaction kinetic studies were performed using 2-prop-
anol (isopropyl alcohol (IPA)) as a hydrogen donor in a batch
reactor. Figure 1 shows HMF conversion and product distribu-
tion as well as the acetone yield as a function of temperature
over a 5 wt% Ru/carbon catalyst. The CTH reaction occurs at
reaction temperatures lower than 373 K and the selectivity of
not observed upon increasing
the temperature to 463 K.
During the transfer hydroge-
nation reaction, 2-propanol is
converted into acetone and H2.
The acetone yield increases from
0.1% to 1.6% with increasing re-
action temperature from 373 to
463 K (Figure 1). The molar ratio
of acetone produced to hydro-
gen consumed by the hydrogen-
ated products is between 1.1–
1.2 at each temperature (except
373 K), indicating that 2-propa-
nol is converted into acetone re-
leasing H2 equivalents that in
turn are utilized in the upgrad-
ing of HMF. At low temperatures,
isopropanol dehydrogenation is
faster than HMF conversion lead-
ing to hydrogen accumulation in
the system.
Conversion of neat 2-propanol
over Ru/C catalyst was investi-
gated to determine the amount
of H2 gas that can be produced
under the same reaction condi-
tions. Figure S1 shows that
2-propanol is readily converted
into acetone and H2 along with
Figure 1. a) Effect of temperature on HMF conversion and product selectivity [HMF conversion (&),
{65,0,100,42}DMF (&), BHMF (&), MFA (&), 5-methyl furfural (&), 3 (&), 2 (&), FA (&), and MF (&)]. b) Acetone
yield (&) and molar ratio (&) of acetone produced to hydrogen consumed by hydrogenated products. Reaction
conditions: batch reactor, 1.2 wt% HMF in isopropanol solution with 5 wt% Ru/C, molar ratio of HMF to Ru of 35,
2.04 MPa N2, and reaction time 6 h at the defined temperatures.
diisopropyl ether (<3.8%) and
small amounts of propane
(<1%). Propane arises from iso-
propanol dehydration to propyl-
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2013, 6, 1158 – 1162 1159

CHEMSUSCHEM
COMMUNICATIONS
www.chemsuschem.org
ene followed by hydrogenation to propane. The acetone yield
reaches chemical equilibrium with a carbon yield of 4.1% at
423 K after 1 h whereas the corresponding H2 partial pressure
is 0.3 MPa in the closed reactor. At 463 K the H2 partial pres-
sure increases up to 0.95 MPa. However, when 2-propanol is
converted into HMF, as shown in Figure 1, the acetone yield is
significantly lower than that from pure 2-propanol. It is likely
that 2-propanol and HMF compete for active sites and 2-prop-
anol dehydrogenates to acetone and H2 at a slower rate
during the CTH of HMF than in neat conditions. Overall, 2-
propanol dehydrogenation is faster than either its etherifica-
tion or HMF hydrogenation and supplies the necessary H2 to
drive the HMF upgrade. Importantly, the hydrogen transfer hy-
drogenation of HMF to DMF can be performed by using other
hydrogen donors (Figure S2).
During the CTH of HMF, it is possible that adsorbed hydro-
gen atoms formed from 2-propanol decomposition can directly
hydrogenate HMF adsorbed on Ru instead of being released as
H2. To evaluate for this possibility, we performed the CTH reac-
tion of HMF at a H2 pressure of 0.7 MPa using tetrahydrofuan
(THF) as solvent. The conversion of HMF and the selectivity for
DMF was 92% and 60%, respectively, which were lower than
those using 2-propanol. This result suggests that hydrogena-
tion of HMF using hydrogen donors can be more effective
than that using external H2 and that the CTH reaction could
occur through direct hydride transfer. Further work will be
needed to delineate the mechanism of hydrogenolysis using
2-propanol and the effect of H2 partial pressure.
Figure 2. HMF conversion and product yield as a function of reaction time
at 423 K. Reaction conditions: batch reactor, 1.2 wt% HMF in isopropanol so-
lution with 5wt% Ru/C, molar ratio of HMF to Ru of 35, and 2.04 MPa N2.
Legend: HMF conversion (&), DMF (~), BHMF (*), 3 (*), 2 (~), MFA (&),
^
5-methyl furfural (*), and FA ( ).
A typical reaction profile for the CTH of HMF at 423 K over
Ru/C is shown in Figure 2. The yield of DMF increases mono-
tonically throughout the reaction time, whereas the yield of
BHMF reaches a plateau at 38% carbon yield at 360 min and
then decreases with reaction time. The yield of MFA exhibits
a similar profile to that of BHMF, passing through the maxi-
mum at 600 min. The temporal evolution of these products is
consistent with the series reaction network of HMF, BHMF,
MFA, and DMF shown in Scheme 1. However, BHMF is not fully
converted into DMF at prolonged reaction times (1200 min),
indicating that the hydrogenation of BHMF to DMF is a slow
reaction. BHMF hydrogenation is favorable at higher tempera-
tures and complete conversion of BHMF is achieved at 463 K
after 300 min (Figure 1). Both BHMF and MFA preferentially un-
dergo etherification with 2-propanol to produce 2 and 3, re-
spectively (Scheme 1). The yield of these products first increas-
es and then reaches a plateau at 1100 min. 5-methyl furfural
and FA are also observed, but the yields of both products are
low over the entire reaction time. The low yield of 5-methyl
furfural may be due to the fact that it does not form fast
enough or is consumed rapidly.
ure S3). Their concentration reaches a maximum after 60 min
and then decreases while the yield of DMF increases monoton-
ically as the reaction time increases to 360 min, indicating that
at higher temperatures de-etherification of ethers followed by
hydrogenation into DMF occurs. 5-methyl furfural is also reac-
tive and is converted into DMF and MFA along with 3 (see Fig-
ure S4). Based on the results outlined above, the proposed re-
action network of CTH of HMF to DMF has been developed
(Scheme 1).
To determine if acid sites on the carbon support or protons
from 2-propanol can contribute to the reaction chemistry, we
investigated the reactions without Ru/C catalyst (reactions 1
and 2) and over activated carbon (reaction 3) as shown in
Table 1. Both HMF and BHMF show very little reactivity (8%
and 4% conversion, respectively) without Ru/C catalyst, indi-
cating that the alcohol protons do not play an active role on
transfer hydrogenation. However, the reaction of HMF over ac-
tivated carbon shows 27% conversion with moderate selectivi-
ty toward ethers (37%) and partially hydrogenated products
(30% to BHMF and 5-methyl furfural) and a larger fraction of
unidentified species. This result suggests that acidic functional
groups on activated carbon are largely responsible for the ob-
served etherification reactions and heavy byproducts; some
partial hydrogenation of HMF via alcohol is possible. However,
complete hydrogenation of HMF to DMF is not feasible with-
out Ru under the same reaction conditions, indicating that the
CTH of HMF to DMF is mainly catalyzed by Ru.
To understand the reaction network, we investigated the re-
action of BHMF and 5-methyl furfural over Ru/C (Figure S3 and
S4). Figure S3 shows that BHMF is converted primarily into
DMF and MFA during the course of the reaction. The etherified
2 and 3 are also produced, indicating that etherification always
occurs in parallel to the hydrogenolysis of the alcohol group in
BHMF and MFA. However, these ethers are slowly converted
into DMF, especially at higher temperatures (e.g., 463 K in Fig-
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2013, 6, 1158 – 1162 1160

CHEMSUSCHEM
COMMUNICATIONS
www.chemsuschem.org
Table 1. HMF and BHMF conversion without Ru/C catalyst and over activated carbon.[a]
Reaction
Feed
Catalyst
Conversion
[%]
Acetone
yield [%]
Selectivity [%]
HMF
BHMF
5-methyl furfural
1
2
3
4
1
2
3
HMF
BHMF
HMF
none
none
carbon
7.7
3.4
27.3
0.08
0.07
0.12
–
18.5
–
45.3
0
12.1
0
0
18.3
6.6
0
5.3
7.9
27.4
12.7
0
0
11.8
0
0
6.7
[a] Reaction conditions: 1.2 wt% HMF or BHMF in isopropanol solution, temperature of 463 K, 2.04 MPa N2, and reaction time of 5 h. Activated charcoal
was pretreated at 673 K for 2 h under He to decompose the acidic functional groups at the surface (10 KminÀ1 ramp and 2 h at 673 K, He flow rate of
40 cm3 minÀ1 for 0.5 g of activated carbon).
Figure 3. HMF conversion over spent Ru/C catalysts. Reaction conditions: batch reactor, 1.2 wt% HMF in isopropanol solution, 463 K, molar ratio of HMF to
Ru of 35, 2.04 MPa N2, and reaction time 6 h. Legend: fresh Ru/C (&), spent Ru/C (&; the spent Ru/C catalyst was recovered from the reaction mixture by fil-
tration after reaction and dried at 353 K), and regenerated Ru/C (&; 0.1 g of the spent Ru/C catalyst was treated for 2 h at 573 K in flowing H2 at
40 cm3 minÀ1).
Catalyst recycling experiments (with spent Ru/C catalyst)
were conducted to examine catalyst stability. The catalyst was
recovered after reaction by filtration and reused after drying at
353 K. As shown in Figure 3, the conversion of HMF and the
yield of DMF significantly decreased to 47% and 13%, respec-
tively, compared to 100% conversion and 81% DMF yield of
the fresh catalyst, showing considerable deactivation of the
catalyst even after its first use. It is possible that the deactiva-
tion is due to the formation of high molecular weight byprod-
ucts on ruthenium surfaces given that we could not identify
approximately 18% carbon in the liquid product. The fresh
and spent Ru/C catalysts were subjected to elemental analysis
to measure its ruthenium and carbon content, and it was
found that the amount of carbon in the Ru/C catalyst in-
creased from 75% to 81% after the reaction, whereas the
amount of ruthenium decreased from 4.6% to 3.8% (Table 2).
The increase of carbon content after the reaction could be due
to solid products recovered with the spent Ru/C or the loss of
ruthenium during the reaction. Inductively Coupled Plasma
(ICP-OES) analysis of the product solution confirmed that Ru
leaching is negligible. In addition, CO chemisorption measure-
ments after reduction of the catalysts at 573 K under H2 flow
showed that the fresh and spent catalysts have a similar
number of surface sites, 220 and 196 mmolgÀ1, respectively.
Therefore, it is likely that the heavy byproducts are present on
the catalyst causing catalyst deactivation. Thus, the spent cata-
lyst was regenerated by heating at 573 K under H2 flow for 2 h
and then reused for the reaction. The initial activity of the cata-
lyst was almost completely regained after regeneration, show-
ing 96% conversion and 68% DMF selectivity along with
16.1% selectivity for the hydrogenated products. Thus, the cat-
alyst can be reused under our reaction conditions after
regeneration.
Table 2. Ruthenium and carbon content for fresh and spent Ru/C based
on elemental analysis.
Catalyst
Ru
C
Ratio
[wt%]
[wt%]
(C/Ru)
fresh Ru/C
spent Ru/C
4.63
3.76
74.69
80.65
16.1
21.4
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2013, 6, 1158 – 1162 1161

CHEMSUSCHEM
COMMUNICATIONS
www.chemsuschem.org
sealed, purged of air with N2, and charged with N2 to a pressure of
2.04 MPa. The reactor was heated to reaction temperature by
using a silicon oil bath, and mixing was accomplished by using
a magnetic stirrer. After the reaction, the reactor was quenched in
ice-water and the liquid products were analyzed by means of
GC-FID (Agilent 7890A; FID=flame ionization detector).
The fresh and spent Ru/C samples were investigated using
XRD, TEM, CO chemisorption, and temperature-programmed
reduction (TPR) of H2. The diffraction pattern of the fresh Ru/C
(Figure S5) and the lattice index of the fresh Ru/C (Figure S6) in
the TEM diffractograms clearly show the presence of Ru metal
particles. The size of the Ru metal particles measured by TEM
was in the range of 3–5 nm. CO chemisorption of Ru/C was
performed after reduction of the catalyst for 2 h at 573 K in H2
flow. The Ru metal dispersion is 46% with 220 mmolgÀ1 of CO
uptake at 298 K, indicating that the Ru metal is highly dis-
persed on the carbon support as corroborated by TEM. The
H2-TPR analysis of the Ru/C catalyst showed two peaks at
403 K, attributable to the reduction of the Ru species, and at
873 K, which is assigned to the reduction of the surface func-
tional groups on the carbon support, after the fresh Ru/C cata-
lyst was preoxidized for 1.5 h at 403 K in 5% O2 and 95% He
flow (Figure S7). These characterization studies indicate that
highly dispersed ruthenium nanoparticles are present in the
catalyst samples and are responsible for the selective conver-
sion of HMF to DMF.
Acknowledgements
This work was supported from the Catalysis Center for Energy
Innovation, an Energy Frontier Research Center funded by the
U.S. Department of Energy, Office of Science, Office of Basic
Energy Sciences under Award Number DE-SC0001004.
Keywords: alcohols · biofuels · biomass · hydrogenolysis ·
transfer hydrogenation
[1] a) J. G. M. Bremner, R. K. F. Keeys, J. Chem. Soc. 1947, 1068–1080; b) D. G.
Manly, A. P. Dunlop, J. Org. Chem. 1958, 23, 1093–1095; c) R. S. Rao,
R. T. K. Baker, M. A. Vannice, Catal. Lett. 1999, 60, 51–57; d) H. Y. Zheng,
Y. L. Zhu, B. T. Teng, Z. Q. Bai, C. H. Zhang, H. W. Xiang, Y. W. Li, J. Mol.
Catal. A: Chem. 2006, 246, 18–23; e) S. Sitthisa, D. E. Resasco, Catal. Lett.
2011, 141, 784–791; f) J. P. Lange, E. van der Heide, J. van Buijtenen, R.
Price, ChemSusChem 2012, 5, 150–166; g) Y. Romꢁn-Leshkov, C. J. Barrett,
Z. Y. Liu, J. A. Dumesic, Nature 2007, 447, 982–985; h) M. Chidambaram,
A. T. Bell, Green Chem. 2010, 12, 1253–1262; i) J. B. Binder, R. T. Raines, J.
Am. Chem. Soc. 2009, 131, 1979–1985; j) J. M. R. Gallo, D. M. Alonso,
M. A. Mellmer, J. A. Dumesic, Green Chem. 2013, 15, 85–90; k) S. Sitthisa,
T. Sooknoi, Y. G. Ma, P. B. Balbuena, D. E. Resasco, J. Catal. 2011, 277, 1–
13; l) G. C. A. Luijkx, N. P. M. Huck, F. van Rantwijk, L. Maat, H. van Bek-
kum, Heterocycles 2009, 77, 1037–1044.
[2] a) S. Morikawa, Noguchi Kenkyusho Jiho 1980, 23, 39–44; b) T. Thananat-
thanachon, T. B. Rauchfuss, Angew. Chem. 2010, 122, 6766–6768; Angew.
Chem. Int. Ed. 2010, 49, 6616–6618; c) S. De, S. Dutta, B. Saha, Chem-
SusChem 2012, 5, 1826–1833; d) T. S. Hansen, K. Barta, P. T. Anastas, P. C.
Ford, A. Riisager, Green Chem. 2012, 14, 2457–2461.
[3] a) H. Kobayashi, H. Matsuhashi, T. Komanoya, K. Hara, A. Fukuoka, Chem.
Commun. 2011, 47, 2366–2368; b) G. S. Yi, Y. G. Zhang, ChemSusChem
2012, 5, 1383–1387; c) M. Chia, J. A. Dumesic, Chem. Commun. 2011, 47,
12233–12235; d) I. Gandarias, P. L. Arias, J. Requies, M. El Doukkali, M. B.
Guemez, J. Catal. 2011, 282, 237–247.
In conclusion, we have discovered a new route for the selec-
tive production of DMF from HMF via CTH by using secondary
alcohols as the source of hydrogen and Ru/C as the catalyst.
During the CTH of HMF, H2 is readily transferred from alcohols
to the aldehyde group of HMF followed by further hydrogenol-
ysis of hydroxyl groups over Ru/C catalyst to produce DMF.
Ongoing investigations focus on the deactivation kinetics in
a flow reactor system for improving the lifetime of the catalyst.
Hydrogenation and hydrogenolysis are key reactions in up-
grading biomass-derived molecules into usable fuels and
chemicals. Our results suggest that the CTH can be a viable
route to upgrade other biomass-derived molecules in an effi-
cient and sustainable way without utilizing petroleum-derived
H2.
Experimental Section
Experimental details are given in the Supporting Information.
In brief, the CTH of HMF was performed in a stainless steel Parr re-
actor (100 mL). For a typical reaction, the reactor was charged with
HMF (99%, Sigma–Aldrich; 240 mg), Ru/C catalyst (5 wt% metal,
Sigma–Aldrich; 100 mg), and 2-propanol (24 mL). The reactor was
[4] R. Alamillo, M. Tucker, M. Chia, Y. Pagan-Torres, J. Dumesic, Green Chem.
2012, 14, 1413–1419.
Received: April 2, 2013
Published online on June 10, 2013
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2013, 6, 1158 – 1162 1162