One-pot production of 2,5-dimethylfuran from fructose over Ru/C and a Lewis-Br?nsted acid mixture in: N, N -dimethylformamide

Article abstract of DOI:10.1039/c6cy00275g

An efficient catalysis system composed of a Lewis-Br?nsted acid mixture and Ru/C using N,N-dimethylformamide as a solvent was developed for the one-pot conversion of fructose to 2,5-dimethylfuran (2,5-DMF) via the dehydration/hydrogenolysis sequence. The effects of various reaction parameters, such as solvent, catalyst type, catalyst loading, reaction pressure, temperature and time, on single fructose dehydration, 5-hydroxymethylfurfural (5-HMF) hydrogenolysis and the one-pot conversion of fructose to 2,5-DMF were systematically investigated. The results showed that 2,5-DMF could be successfully produced with a yield as high as 66.3 mol% by using a one-pot method directly from fructose under the optimized reaction conditions, which is by far the highest yield ever reported for the production of 2,5-DMF from fructose through a one-pot strategy. The Ru/C catalyst could be reused at least three times with a slight decrease in 2,5-DMF yield.

Full text of DOI:10.1039/c6cy00275g

Catalysis
Science &
Technology
View Article Online
View Journal
PAPER
One-pot production of 2,5-dimethylfuran from
fructose over Ru/C and a Lewis–Brønsted acid
mixture in N,N-dimethylformamide†
Cite this: DOI: 10.1039/c6cy00275g
a
a
b
b
*
Zuojun Wei, Jiongtao Lou, Zhenbin Li and Yingxin Liu
An efficient catalysis system composed of
a Lewis–Brønsted acid mixture and Ru/C using N,N-
dimethylformamide as solvent was developed for the one-pot conversion of fructose to 2,5-
a
dimethylfuran (2,5-DMF) via the dehydration/hydrogenolysis sequence. The effects of various reaction pa-
rameters, such as solvent, catalyst type, catalyst loading, reaction pressure, temperature and time, on single
fructose dehydration, 5-hydroxymethylfurfural (5-HMF) hydrogenolysis and the one-pot conversion of
fructose to 2,5-DMF were systematically investigated. The results showed that 2,5-DMF could be success-
fully produced with a yield as high as 66.3 mol% by using a one-pot method directly from fructose under
the optimized reaction conditions, which is by far the highest yield ever reported for the production of
2,5-DMF from fructose through a one-pot strategy. The Ru/C catalyst could be reused at least three times
with a slight decrease in 2,5-DMF yield.
Received 4th February 2016,
Accepted 11th May 2016
DOI: 10.1039/c6cy00275g
www.rsc.org/catalysis
Saha et al.10 have recently determined that Pd/C modified
with zinc could achieve a 2,5-DMF yield as high as 85 mol%
Introduction
2,5-Dimethylfuran (2,5-DMF) is regarded as one of the most
promising renewable biofuels.1–8 Compared with traditional
bio-derived ethanol, 2,5-DMF possesses a higher boiling point
(92–94 °C vs. 78.4 °C), a higher research octane number (119
vs. 111), and a higher energy density (31.5 kJ cm−3 vs. 23 kJ
cm−3) and is easier to blend with gasoline. It also has a lower
latent heat of vaporization (0.30 kJ cm−3 vs. 0.71 kJ cm−3),
which means lower energy consumption during purification
through distillation.1–4,9–12
The main strategy for 2,5-DMF production is the hydro-
genolysis of the biomass based platform chemical
5-hydroxymethylfurfural (5-HMF) in the presence of a metal-
based catalyst, which was first developed by Dumesic et al.13
in 2007, who hydrogenated 5-HMF over a CuRu/C catalyst in
1-butanol at 0.68 MPa of H2 and 220 °C for 10 h to obtain
2,5-DMF in 76 mol%–79 mol% yield. Hu et al.12 achieved a
much higher 2,5-DMF yield of 94.7 mol% by replacing the
solvent with tetrahydrofuran (THF) and using a commercial
Ru/C catalyst. A similar yield (93.4 mol%) was obtained by Zu
et al.,14 who optimized Co3O4 as a support for the Ru catalyst.
at 150 °C and 0.8 MPa of H2 in THF. Wang et al.15 fabricated
PtCo bimetallic nanoparticles in hollow carbon spheres and
obtained the highest 2,5-DMF yield ever reported, i.e. 98
mol%. Although ionic liquids, which act as both a solvent
and a catalyst, are very effective for carbohydrate dehydration
to 5-HMF,16–21 they failed to catalyze 5-HMF hydrogenolysis,
and a 2,5-DMF yield of only less than 20 mol% was
obtained.10,22 A process called transfer hydrogenation, which
uses small organic molecules instead of pressurized H2 as a
hydrogen source, has been developed to avoid pressurized H2
depletion. For example, Thananatthanachon and Rauchfuss23
obtained a 2,5-DMF yield of up to 95 mol% at 70 °C for 15 h
by using formic acid as both an acid catalyst and a hydrogen
source over the Pd/C catalyst. Jae et al.7 determined that iso-
propanol was also an effective hydrogen source for the trans-
fer hydrogenolysis of 5-HMF over Ru/C. However, this
method seems neither green nor economical.7,23
Considering the efficiency of biomass resource transfor-
mation, scientists have focused more on carbohydrates than
on 5-HMF as a starting material for 2,5-DMF production, be-
cause the former is more easily available, up-streaming, and
inexpensive. In general, carbohydrate conversion to 2,5-DMF
requires two steps. In the first step, carbohydrates are
dehydrated to 5-HMF in the presence of a Lewis or Brønsted
acid;17,24–29 in the second step, 5-HMF is converted into
2,5-DMF through hydrogenolysis over a metal catalyst.30,31
Although the aforementioned strategies can achieve
2,5-DMF from 5-HMF in considerably high yields, the
a Key Laboratory of Biomass Chemical Engineering of the Ministry of Education,
College of Chemical and Biological Engineering, Zhejiang University, Hangzhou
310027, PR China
b Research and Development Base of Catalytic Hydrogenation, College of
Pharmaceutical Science, Zhejiang University of Technology, Hangzhou 310014, PR
China. E-mail: yxliu@zjut.edu.cn
† Electronic supplementary information (ESI) available. See DOI: 10.1039/
c6cy00275g
This journal is © The Royal Society of Chemistry 2016
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
difficulty in 5-HMF purification from the dehydration prod-
ucts of carbohydrates13,17,32–34 casts a shadow on these pro-
cesses. Unfortunately, when carbohydrates are used as
starting feedstocks based on the aforementioned hydro-
genolysis catalysts, the selectivity toward 2,5-DMF becomes
very poor. For example, Dumesic13 and Binder1 initially devel-
oped a two-step process to obtain 2,5-DMF from untreated
corn stover. In the first step, CrCl3 was used as a catalyst to
dehydrate cellulose to 5-HMF in a mixture of HCl, N,N-
dimethylacetylamine and LiCl, followed by hydrogenolysis of
5-HMF over a Cu–Ru catalyst. A final 2,5-DMF yield of only 9
mol% was obtained. Hu et al.35 and Zu et al.14 claimed that
Ru/C and Ru/Co3O4 catalysts led to very high 2,5-DMF yields
of 75.1 mol% and 92.7 mol% from fructose, respectively.
However, they also followed a two-step method, in which sep-
aration of 5-HMF from the dehydration products of fruc-
tose was performed prior to the hydrogenolysis step. Very
recently, Upare et al.36 have successfully produced 2,5-DMF
in 92% overall yield from fructose in 1-butanol through a
combination of dehydration over Amberlyst-15 and hydro-
genolysis over the Ru–Sn/ZnO catalyst. They also followed
a two-step process, in which 5-HMF was separated in the
dehydration step. Dutta and Mascal37 reported the use of
5-chloromethylfurfural, which could be easily separated
from cellulosic biomass dehydration products, as a feed-
stock for 2,5-DMF production under mild conditions and
predicted that an overall 2,5-DMF yield of up to 65 mol%
could be obtained from carbohydrates. In Saha's work,10
one-pot production of 2,5-DMF from fructose by using
ZnCl2–Pd/C as a catalyst only achieved 22 mol% yield. In
contrast, high 2,5-DMF yields (71–85 mol%) were achieved
when 5-HMF was used as a substrate. A summary of the
available reports describing the production of 2,5-DMF
from carbohydrates is provided in Table 1.
As reported, in all the two-step processes, purification
of 5-HMF is inevitable and the overall yield of 2,5-DMF
from carbohydrates is unsatisfactory. We have recently de-
veloped an effective method to synthesize 5-HMF from
fructose in the presence of a Lewis–Brønsted acid mixture
in N,N-dimethylformamide.27 In order to simplify the reac-
tion steps and avoid the complex purification process for
5-HMF, in the present article, we therefore attempt to de-
velop a one-pot process for producing 2,5-DMF from fruc-
tose in high yield.
Results and discussion
Determination of a suitable solvent for 2,5-DMF production
A combination of fructose dehydration and 5-HMF hydro-
genolysis to obtain 2,5-DMF by using a one-pot process
Table 1 Recent progress on the two-step process for the production of 2,5-DMF from carbohydrates
Step 1 of the two-step process: dehydration
Step 2 of the two-step process: hydrogenolysis
Con. of
carbohydr. yield
5-HMF
Con. of 2,5-DMF
5-HMF yield
P
t
Catalyst
Substrate Solvent
(mol%)
(mol%) Catalyst
Solvent
T (°C)
(MPa) (h)
(mol%) (mol%) Ref.
HCl
CrCl3
Fructose
Corn
1-Butanol
DMA +
88
—
72
—
CuRu/C
CuRu/C
1-Butanol 220
1-Butanol 220
0.68
0.68
—
10
100
—
71
49
13
1
stover
Glucose
[EMIM]Cl
[EMIMCl] +
CH3CN
12-MPA
98
97
Pd/C
[EMIM]Cl 120
6.2
1
47
32
22
HCOOH + H2SO4 Fructose
HCOOH Fructose
DMSO
HCOOH
100
—
93
—
Pd/C
THF
THF
70
75
—
—
15
1
—
—
>95
32
23
2
Ru/C +
H2SO4
Ru/C +
H2SO4
Ru/C
Ru/C
Ru/C
Ru/C
Ru/C
[DMA]+[CH3SO3]− Cellulose DMA/LiCl
—
—
THF
150
—
1
16
2
SO42−/ZrO2–TiO2 Glucose
SO42−/ZrO2–TiO2 Glucose
SO42−/ZrO2–TiO2 Glucose
SO42−/ZrO2–TiO2 Sucrose
SO42−/ZrO2–TiO2 Maltose
n-Butanol
[BMIM]Cl
[BMIM]Cl
[BMIM]Cl
[BMIM]Cl
96.5
—
—
—
—
—
—
—
—
—
96
26.0
82.5
47.4
63.2
46.8
45.1
43.9
81
n-Butanol 260
—
2
2
2
2
2
2
1
0.8
0.8
1
2
2
2
2
2
2
24
8
8
99.8
—
—
—
—
—
—
—
60.3
3
35
THF
THF
THF
THF
THF
THF
THF
200
200
200
200
200
200
150
150
150
92.7
88.1
90.4
87.3
86.2
84.6
75.1
85
SO42−/ZrO2–TiO2 Cellobiose [BMIM]Cl
Ru/C
Ru/C
Ru/Co3O4
SO42−/ZrO2–TiO2 Starch
[BMIM]Cl
THF
—
C–SO3H
Fructose
14
10
C–SO3H
—
—
93
ZnCl2–Pd/C THF
ZnCl2–Pd/C THF
C–SO3H
Amberlyst-15
Fructose
Fructose
—
22a
99
1-Butanol +
H2O
Ru–Sn/ZnO 1-Butanol 180–240 0.1–1 0.2 h−1 100
3
DMA: dimethylamine; MPA: 3-mercaptopropionic acid; DMSO: dimethylsulfoxide; THF: tetrahydrofuran; [EMIM]Cl: 1-ethyl-3-methylimidazolium
chloride; [BMIM]Cl: 1-butyl-3-methylimidazolium chloride.a One-pot production of 2,5-DMF from fructose.
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2016

View Article Online
Catalysis Science & Technology
Paper
requires a suitable solvent which is beneficial to both steps.
Different solvents were evaluated for the individual dehydra-
tion and hydrogenolysis steps, and the results are shown in
Table 2. It can be seen that the effects of the solvents on the
two steps were significantly different. For example, in the
case of the ionic liquid 1-butyl-3-methylimidazolium chloride
([BMIM]Cl) used as the solvent, fructose could be almost
completely dehydrated (entry 1), and the highest 5-HMF yield
of 58.5 mol% was achieved within 30 min under the reaction
conditions used. However, for the 5-HMF hydrogenolysis
step, a 2,5-DMF yield of only 9.0 mol% (entry 6) was obtained
in the presence of the same ionic liquid, which was similar
to that reported by Chidambaram and Bell.22 THF was con-
sidered as a favourable solvent for 5-HMF hydrogenolysis to
2,5-DMF.10,14,23,24,35 A 2,5-DMF yield as high as 90.1 mol%
(entry 10) was obtained in THF within our experimental
range. However, it was infeasible for fructose dehydration be-
cause the solubility of fructose in THF is rather low (0.15 g
per 100 g) and the solvent is unstable in acidic environ-
ments.38 Thus, the 5-HMF product was not detected in the
fructose dehydration step (entry 5) when THF was used as a
solvent. For the same reasons, C1–C4 alcohols were not ideal
solvent candidates for the integration of dehydration and
hydrogenolysis processes (entries 11–14). Moreover, when al-
cohols were used as solvents, 5-HMF tended to form
dihydroxymethylfuran or be deeply hydrogenated to 2,5-
dimethyltetrahydrofuran.39 As shown in Table 2, when N,N-
dimethylformamide was used as the solvent, a satisfactory
5-HMF yield in the fructose dehydration step (44.3 mol%, en-
try 4) and a satisfactory 2,5-DMF yield in the 5-HMF hydro-
genolysis step (73.8 mol%, entry 9) were achieved. Similar re-
sults were obtained in the case of dimethylsulfoxide (DMSO)
(48.3 mol% in entry 3 and 75.3 mol% in entry 8), indicating that
polar aprotic solvents such as THF, N,N-dimethylformamide
and DMSO having lone pair electrons may be beneficial to
the formation of 2,5-DMF. However, DMSO is one of the sul-
phur compounds, which can partially poison and deactivate
the Ru/C catalyst used in the hydrogenolysis step by chemi-
sorption on the catalyst surface.40 Moreover, even trace
amounts of DMSO in the products can introduce an unpleas-
ant smell that is difficult to remove.27 Therefore, we selected
N,N-dimethylformamide as the appropriate solvent for the
one-pot conversion of fructose to 2,5-DMF in the following
research.
To our surprise, when 1,4-dioxane was used as solvent for
hydrogenolysis of 5-HMF, the yield of 2,5-DMF was almost
not detectable, which is rather different from the highest
yield of 2,5-DMF obtained in THF, although these two sol-
vents have very similar molecular structures. Actually, this
phenomenon has already been observed,41,42 in which 5-HMF
could be switchably synthesized to different products more
than 2,5-DMF in the presence of 1,4-dioxane in different cata-
lytic systems.
Determination of a suitable metal catalyst for the 5-HMF
hydrogenolysis step
Although several metal catalysts, including Ni, Co, Cu, Pd, Pt,
Rh, Ru, and Au, have been applied to 5-HMF hydrogenolysis
in the presence of various solvents, such as THF, toluene,
1,4-dioxane, ionic liquids, and water (as shown in Table 1),
no results about N,N-dimethylformamide as a single solvent
for this reaction have been reported. We therefore tested the
performance of several commercially available metal catalysts
for 5-HMF hydrogenolysis in the presence of N,N-
dimethylformamide, as shown in Table 3. Among all of the
catalysts tested, Ru/C showed the highest catalytic perfor-
mance, achieving a 5-HMF conversion of 100 mol% and a
2,5-DMF yield of 73.8 mol% (entry 6 in Table 3). Raney Co
and Ni catalysts achieved good 5-HMF conversion, but their
selectivities toward 2,5-DMF were only 22.2 mol% and 9.2
mol% (entries 15 and 16 in Table 3), respectively. Other no-
ble metal catalysts unexpectedly gave unsatisfactory 5-HMF
Table 2 Dehydration of fructose and hydrogenolysis of 5-HMF in various solvents
Fructose to 5-HMF
5-HMF to 2,5-DMF
Fru./sol.
Fru. conversion
(mol%)
5-HMF yield
(mol%)
5-HMF conversion
(mol%)
2,5-DMF yield
(mol%)
Solvent
Entry
(g ml−1
)
Entry
[BMIM]Cl
1
2
3
4
5
0.05/2
1/100
5/100
5/100
>99
88
>99
>99
>99
—
—
—
—
58.5
26.6
48.3
44.3
N.D.c
—
—
—
—
6
7
8
30.0
76.7
99.2
100
9.0
1,4-Dioxanea
DMSO
N.D.b
75.3
73.8
90.1
44.2
38.0
13.8
47.6
N,N-Dimethylformamide
9
THFb
0.15/100
10
11
12
13
14
100
Methanol
Ethanol
Isopropanol
1-Butanol
—
—
—
—
97.4
96.8
100
87.3
Reaction conditions for fructose dehydration: catalyst, 26.9 mol% of AlCl3 (based on fructose); temperature, 120 °C; reaction time, 30 min.
Reaction conditions for hydrogenolysis of 5-HMF: 5-HMF, 1 g; solvent, 60 ml; catalyst, 0.2 g of Ru/C (5 wt%); temperature, 220 °C; H2 pressure,
0.7 MPa; reaction time, 3 h; agitation speed, 600 rpm.a The reaction was conducted at 101 °C. b The reaction was conducted at 66 °C. c N.D. de-
notes not detected.
This journal is © The Royal Society of Chemistry 2016
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
Table 3 Catalytic performance of different commercially available cata-
lysts for the hydrogenolysis of 5-HMF to 2,5-DMF
tested in the range of 140 °C to 220 °C. As shown in Fig. 2,
5-HMF conversion increased with temperature, although the
2,5-DMF yield first increased sharply from 140 °C to 160 °C
and then gradually increased when the reaction temperature
was higher than 180 °C. Conversion of 5-HMF reached 100
mol% at 180 °C, and the corresponding 2,5-DMF yield was
74.3 mol%. At 200 °C, the 2,5-DMF yield plateaued, and fur-
ther increasing the reaction temperature led to a slight de-
crease in 2,5-DMF yield. Besides, lower temperature is benefi-
cial to obtain a higher 5-HMF yield in the dehydration step
and inhibit unwanted side reactions. We therefore consid-
ered 200 °C as the appropriate reaction temperature for sub-
sequent studies.
5-HMF conversion
(mol%)
2,5-DMF yield
(mol%)
Entry
Catalyst
15
16
17
18
19
20
21
Raney Co
Raney Ni
100
76.9
100
92.0
28.9
100
82.0
22.2
9.2
5 wt% Pt/C
5 wt% Ir/C
5 wt% Pd/C
5 wt% Ru/C
5 wt% Rh/C
10.5
13.5
16.3
73.8
16.8
Reaction conditions: 5-HMF, 1 g; N,N-dimethylformamide, 60 ml; cat-
alyst, 0.2 g; temperature, 220 °C; H2 pressure, 0.7 MPa; reaction time,
3 h; agitation speed, 600 rpm.
The reaction time for hydrogenolysis of 5-HMF to 2,5-DMF
was optimized at 200 °C. As shown in Fig. 3, 5-HMF hydro-
genolysis proceeded quickly within the first 3 h to achieve a
5-HMF conversion and a 2,5-DMF yield of 80.2 mol% and
74.8 mol%, respectively. Afterward, the reaction rate gradu-
ally slowed down. After 12 h, a 5-HMF conversion of 100
mol% and a 2,5-DMF yield of 85.5 mol% were obtained. Fur-
ther increasing the reaction time might result in decomposi-
tion of 2,5-DMF to other by-products, as the 2,5-DMF yield
decreased when the reaction time was extended from 12 h to
15 h. Thus, we considered 12 h as the optimal reaction time.
Hydrogen pressure was optimized in the range of 0.5–2
MPa at 180 °C and 3 h as shown in Fig. 4. It can be seen that
both the 5-HMF conversion and the 2,5-DMF yield consis-
tently increased with hydrogen pressure from 0.5 MPa to 1.5
MPa and plateaued when the hydrogen pressure is larger
than 1.5 MPa. We therefore considered 1.5 MPa as the opti-
mal hydrogen pressure.
conversion and 2,5-DMF yields. Therefore, we selected Ru/C
as the hydrogenolysis catalyst for subsequent experiments.
Optimization of reaction conditions for single dehydration
and hydrogenolysis steps
Fructose
dehydration
in
the
presence
of
N,N-
dimethylformamide has previously been studied by using a
mixture of Lewis–Brønsted acids as the catalyst, in which a
maximum 5-HMF yield of 92.6 mol% with nearly 100% con-
version of fructose could be achieved at 120 °C within 20 min
when the mole ratio of AlCl3, H2SO4 and H3PO4 was 1 : 2.7 :
4.27 Here, we further studied the effect of reaction tempera-
tures ranging from 120 °C to 220 °C on the 5-HMF yield in
order to adapt the followed hydrogenolysis step. As shown in
Fig. 1, the reaction rates increased with temperature, while
the time needed to achieve the maximum 5-HMF yield de-
creased from 5 min at 220 °C to 20 min at 120 °C. However,
the maximum 5-HMF yield tended to decrease by ca. 20
mol% as the temperature increased.
Ru/C catalyst loadings of 5 wt%, 10 wt%, 20 wt%, and 30
wt% (mass ratio of wet Ru/C to 5-HMF) were then investi-
gated (Fig. 5). When the Ru/C loading was less than 20 wt%,
the 2,5-DMF yield increased with increasing Ru/C loading.
Further increasing the Ru/C loading more than 20 wt%, how-
ever, led to a slight decrease in 2,5-DMF yield. Thus, we se-
lected 20 wt% as the optimal catalyst loading.
Hydrogenolysis of 5-HMF over Ru/C was further optimized
by varying several reaction parameters, such as reaction tem-
perature, time, hydrogen pressure, catalyst loading, and ini-
tial substrate concentration. The reaction temperature was
Fig. 2 Effect of reaction temperature on the production of 2,5-DMF
Fig. 1 Effect of reaction temperature on the production of 5-HMF
from fructose over Lewis–Brønsted acids. Reaction conditions: fruc-
tose, 5 g; N,N-dimethylformamide, 60 ml; catalyst, the mole ratio of
AlCl3 : H2SO4 : H3PO4 = 1 : 2.7 : 4; reaction time, 20 min; agitation speed,
600 rpm. (□) 120 °C, (○) 150 °C, (△) 180 °C and (◇) 220 °C.
from 5-HMF over the
5 wt% Ru/C catalyst. Reaction conditions:
5-HMF, 1 g; N,N-dimethylformamide, 60 ml; catalyst, 0.2 g; H2 pres-
sure, 0.7 MPa; reaction time, 12 h; agitation speed, 600 rpm; (○) con-
version and (□) yield.
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2016

View Article Online
Catalysis Science & Technology
Paper
of 12.5–25 mg ml−1, and finally, we optimized the 5-HMF con-
centration to 16.7 mg ml−1 (Fig. 6).
In summary of the 5-HMF hydrogenolysis step, a maxi-
mum 2,5-DMF yield of 93.9 mol% might be obtained in N,N-
dimethylformamide solvent at a temperature of 200 °C, a H2
pressure of 1.5 MPa, a 5 wt% Ru/C catalyst loading of 20 wt%
(relative to 5-HMF), an initial 5-HMF concentration of 16.7
mg ml−1, and a reaction time of 12 h. The results obtained
were fairly similar to those obtained when Ru/C was used with
THF as the solvent,35 indicating that N,N-dimethylformamide
is another competitive solvent for 5-HMF hydrogenolysis.
Fig. 3 Effect of reaction time on the production of 2,5-DMF from
5-HMF over the 5 wt% Ru/C catalyst. Reaction conditions: 5-HMF, 1 g;
N,N-dimethylformamide, 60 ml; catalyst, 0.2 g; H2 pressure, 0.7 MPa;
temperature, 200 °C; agitation speed, 600 rpm; (○) conversion and (□)
yield.
One-pot production of 2,5-DMF from fructose
After having obtained the optimized reaction conditions for
the individual dehydration step and hydrogenolysis step, we
could build a one-pot process for the conversion of fructose
to 2,5-DMF through a simple and effective way: a mixture of
fructose (10 mmol), AlCl3 (0.25 mmol), H2SO4 (0.7 mmol),
H3PO4 (1.0 mmol), N,N-dimethylformamide (60 ml) and 5
wt% Ru/C (0.36 g) was added to the reactor and hydroge-
nated at 200 °C and 1.5 MPa H2 pressure for 12 h. The
highest 2,5-DMF yield of 66.3 mol% was finally obtained (en-
try 22, in Table 4). Further changing the amounts and ratio
of acids led to a slight decrease in 2,5-DMF yield (entries 23–
26 in Table 4). We believe that the 2,5-DMF yield of 66.3
mol% obtained in this work is the highest yield ever reported
for the direct one-pot conversion of fructose to 2,5-DMF. A
similar work10 that used a ZnCl2–Pd/C bifunctional catalyst
for the one-pot production of 2,5-DMF from fructose achieved
a 2,5-DMF yield of only 22 mol%. Such a low yield might be
due to the fact that THF used in that study was neither stable
in Lewis acid nor favorable for fructose dehydration because
of its low solubility.
Fig. 4 Effect of H2 pressure on the production of 2,5-DMF from
5-HMF over the 5 wt% Ru/C catalyst. Reaction conditions: 5-HMF, 1 g;
N,N-dimethylformamide, 60 ml; catalyst, 0.2 g; temperature, 180 °C;
reaction time, 3 h; agitation speed, 600 rpm; (○) conversion and (□)
yield.
In general, researchers prefer a two-step process,12,13,35 i.e.
fructose dehydration → 5-HMF purification → 5-HMF hydro-
genolysis, instead of a one-pot reaction, because the reaction
conditions for these two consecutive reactions do not always
match. Although increased yields of the target product might
Fig. 5 Effect of Ru/C catalyst loading on the production of 2,5-DMF
from 5-HMF over the
5 wt% Ru/C catalyst. Reaction conditions:
5-HMF, 1 g; N,N-dimethylformamide, 60 ml; temperature, 180 °C; H2
pressure, 1.5 MPa; reaction time, 3 h; agitation speed, 600 rpm; (○)
conversion and (□) yield.
Fig. 6 Effect of initial substrate concentration on the production of
2,5-DMF from 5-HMF over the 5 wt% Ru/C catalyst. Reaction condi-
tions: 5-HMF, 1 g; catalyst, 0.2 g; temperature, 200 °C; H2 pressure, 1.5
MPa; reaction time, 12 h; agitation speed, 600 rpm. (○) conversion and
(□) yield.
The effect of the initial 5-HMF concentration on hydro-
genolysis of 5-HMF to 2,5-DMF was investigated in the range
This journal is © The Royal Society of Chemistry 2016
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
Table 4 Strategies for the one-pot conversion of fructose to 2,5-DMF
Entry
Fructose (mmol)
AlCl3 (mmol)
H2SO4 (mmol)
H3PO4 (mmol)
Con. (mol%)
Yield (mol%)
22
23
24
25
10
10
10
10
10
10
10
0.25
0.42
0.25
0.25
0.25
0.25
—
0.7
1
—
1
1.5
2
—
—
1
>99
>99
>99
>99
>99
>99
>99
66.3
58.5
47.3
53.1
47.6
40.2
35.1
2
26
—
0.7
—
27a
28b
—
Reaction conditions: catalyst, 0.36 g of Ru/C (5 wt%); N,N-dimethylformamide, 60 ml; temperature, 200 °C; H2 pressure, 1.5 MPa; reaction time,
b
12 h; agitation speed, 600 rpm.a Acids were removed between the dehydration and hydrogenolysis processes. Hydrogenolysis of fructose in
the absence of acids.
be achieved at each step, the difficulty in 5-HMF purification
may lead to a very low overall 2,5-DMF yield. For comparison,
a similar two-step process wherein the acid catalyst was re-
moved by sodium carbonate was performed instead of the
full purification of 5-HMF in the present work, and a final
2,5-DMF yield of 40.2 mol% was obtained (entry 27, Table 4).
Given that fructose dehydration has already been optimized
to produce 92.6 mol% of 5-HMF, the 2,5-DMF yield from
5-HMF in the second step was then estimated to be 43.6
mol%, which was much lower than that in the case when
pure 5-HMF was used as the substrate for the hydrogenolysis
step (93.9 mol%). This finding could be mainly attributed to
polymerization reactions between 5-HMF and other by-prod-
ucts, which were frequently observed during fructose dehy-
dration.17,24,41 Moreover, the overall 2,5-DMF yield in our
two-step process was far lower than that obtained in our one-
pot process (40.2 mol% vs. 66.3 mol%) because in the latter
case the acids can stabilize 5-HMF against thermal decompo-
sition and polymerization with other active by-products, as
verified by our previous studies16,17 and several other
laboratories.6,27
Intermediates and a possible mechanism for the 5-HMF
hydrogenolysis step
We analyzed the reaction mixtures by using gas chromatogra-
phy–mass spectrometry (GC–MS) to better understand the path-
way of 5-HMF hydrogenolysis to 2,5-DMF. Besides 2,5-DMF and
four other by-products, including 5-methylfurfural (5-MF),
5-methylfurfuryl alcohol (MFA), furfuryl alcohol (FA), and
2-methylfuran (2-MF), were detected in the reaction system. The
time course of the reaction components is shown in Fig. 8.
Previous work has identified two pathways to reach the
target 2,5-DMF, i.e. (1) hydrogenation of the aldehyde group
in 5-HMF to form dihydroxymethylfuran (DHMF), followed by
individual hydrogenolysis of the two hydroxyl groups,10,23,35
and (2) hydrogenolysis of the hydroxyl group in 5-HMF to
form 5-MF, the aldehyde group.41 Some researchers consider
that these two pathways occur in the same reaction bulk,
while the DHMF pathway is processed much faster than the
5-MF pathway4,22,43 and tends to occur at lower tempera-
tures.15,39,41 Our GC–MS analysis detected no DHMF in the
The mixture of Lewis and Brønsted acids is inevitable for
the one-pot process. When the acid catalysts were absent, a
decreased 2,5-DMF yield of 35.1 mol% (entry 28) was
obtained. Given that a 2,5-DMF yield of 93.9 mol% was
obtained from 5-HMF hydrogenolysis, the overall yield of
5-HMF from dehydration of fructose was estimated to be
around 37.4 mol%, which is even lower by 7 mol% compared
to the yield in the single dehydration step without any acid
catalyst (44.3 mol%, entry 4).
Recyclability of the catalyst
The recyclability of the Ru/C catalyst was determined for both
5-HMF hydrogenolysis and the one-pot conversion of fructose
to 2,5-DMF. After reaction, the Ru/C catalyst was centrifuged,
washed with N,N-dimethylformamide, and returned to the re-
actor for reuse. The relevant results are shown in Fig. 7. It
can be seen that Ru/C could be reused at least three times in
both cases with only an approximately 5–10 mol% decrease
in 2,5-DMF yield. A further study on improving the stability
of Ru-based catalysts is ongoing in our laboratory.
Fig. 7 Recyclability of the 5 wt% Ru/C catalyst in the hydrogenolysis
of 5-HMF and the one-pot conversion of fructose. Reaction condi-
tions: 5-HMF, 1 g; N,N-dimethylformamide, 60 ml; catalyst, 0.36 g; H2
pressure, 1.5 MPa; temperature, 200 °C; reaction time, 12 h; agitation
speed, 600 rpm. White bar, 2,5-DMF yield in the hydrogenolysis of
5-HMF; light grey bar, conversion of 5-HMF; dark grey bar, 2,5-DMF
yield in the one-pot conversion of fructose; black bar, conversion of
fructose.
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2016

View Article Online
Catalysis Science & Technology
Paper
during our experiment, which indicates that deep hydrogena-
tion does not occur in our reaction process.
Conclusions
The production of 2,5-DMF from hydrogenolysis of 5-HMF is a
highly attractive route. In order to avoid the difficulty in the
purification of 5-HMF from the dehydration products of fruc-
tose, we have developed a one-pot process for direct conver-
sion of fructose to 2,5-DMF, in which N,N-dimethylformamide
was demonstrated as the optimal solvent.
At first, the reaction conditions for the 5-HMF hydro-
genolysis step were optimized as follows: a reaction tempera-
ture of 200 °C, a hydrogen pressure of 1.5 MPa, a 5 wt% Ru/C
catalyst loading of 20 wt% (relative to 5-HMF), an initial
5-HMF concentration of 16.7 mg ml−1, and a reaction time of
12 h. Under these conditions, a 2,5-DMF yield up to 93.9
mol% was achieved.
Fig.
8 Time course of components during the hydrogenolysis of
5-HMF over the 5 wt% Ru/C catalyst. Reaction conditions: 5-HMF, 1 g;
N,N-dimethylformamide, 60 ml; catalyst, 0.2 g; H2 pressure, 1.5 MPa;
temperature, 200 °C; agitation speed, 600 rpm. The intermediate com-
ponents were quantified through GC analysis by assuming that their
correction factors to 2,5-DMF are equal to 1.
A catalysis system with a Lewis–Brønsted acid mixture and
Ru/C in N,N-dimethylformamide solvent was then developed
for the one-pot conversion of fructose to 2,5-DMF. A maxi-
mum yield of 66.3 mol% was achieved under the following
reaction conditions: 10 mmol of fructose, 0.25 mmol of AlCl3,
0.7 mmol of H2SO4, 1.0 mmol of H3PO4, 0.2 g of 5 wt% Ru/C,
60 ml of N,N-dimethylformamide, 200 °C of reaction tempera-
ture and 12 h of reaction time. This is the highest 2,5-DMF
yield ever reported on the one-pot strategy from carbohy-
drates. The Ru/C catalyst could be reused at least three times
with a slight decrease in 2,5-DMF yield for both the individ-
ual hydrogenolysis step of 5-HMF and the one-pot conversion
of fructose. We suggest that this simple, efficient, and renew-
able process for the production of 2,5-DMF directly from the
one-pot conversion of fructose will promote the wide adop-
tion of 2,5-DMF as a fossil fuel alternative. In addition, the
method developed here is very useful for the development of
effective conversion of biomass into important platform
chemicals and biofuels.
Scheme 1 Plausible mechanism for the hydrogenolysis of 5-HMF to
2,5-DMF.
system over the entire reaction process. In contrast, large
amounts of 5-MF and MFA were detected during the first
hour of reaction. We therefore deduce that the 5-MF pathway
(solid line in Scheme 1) might be followed in our newly devel-
oped reaction bulk where N,N-dimethylformamide was used
as the solvent, in accordance with the work of Kong et al.41
Theoretically, current experimental data cannot fully rule out
the DHMF pathway since there is still the possibility that
DHMF converts immediately to the downstreaming interme-
diates once it is generated from 5-HMF by hydrogenation of
the carbonyl group at such high temperature. A further kinet-
ics study on the hydrogenolysis of those intermediates such
as DHMF, 5-MF and MFA is needed to obtain a more sound
reaction mechanism.
Two other by-products, namely, FA and 2-MF, might be
generated from 5-HMF decarbonylation and the subsequent
FA hydrogenolysis.7,42 No furan ring-hydrogenated by-prod-
ucts,9,12,22 i.e. 2,5-dihydroxymethyl tetrahydrofuran or 2,5-
dimethyltetrahydrofuran, or ring-opened by-products,19,44 e.g.
(5-hydroxy-2,5-hexanedione) or 2,5-hexanedione, were detected
Experimental section
Materials
5-HMF (>99.9%) was purchased from Tengzhou Runlong
Flavors and Fragrances Co., Ltd. 2,5-DMF standard was pur-
chased from Aladdin Reagent Co., Ltd. Metal catalysts, in-
cluding Raney Co and Ni, Pt/C, Ir/C, Pd/C, Ru/C and Rh/C,
were purchased from Shanghai Eybridge Co., Ltd. The load-
ings of the noble metals were 5 wt%. Other chemicals used,
including fructose (purity, 99%), AlCl3, inorganic acids, and
organic solvents, were analytically pure and purchased from
Sino-pharm Chemical Reagent Co., Ltd. All of the solvents
used in the reactions were passed through a 4A molecular
sieve and reboiled to remove water.
Production of 5-HMF
The production of 5-HMF from fructose dehydration was
conducted in a 100 ml stainless autoclave equipped with a
This journal is © The Royal Society of Chemistry 2016
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
mechanical stirring impeller. In a typical experiment, 1 g of
fructose, 0.2 g of AlCl3, 40 mg of concentrated sulfuric acid
(98 wt%), 60 mg of concentrated H3PO4 (98 wt%) and 60 ml
of N,N-dimethylformamide were successively added into the
autoclave. The mixture was stirred and incubated in an oil
bath that had been preheated to designated temperatures.
The samples were withdrawn at specified intervals. Each reac-
tion was repeated more than twice, and the results were
presented as average data.
n-dodecane as an internal standard. GC–MS analysis was
performed using an Agilent 6890 GC system coupled to a
mass spectrometer equipped with an Agilent 5973 quadru-
pole mass analyzer. Chromatographic analysis was conducted
at an injector temperature of 280 °C and using an HP-5 capil-
lary column (30.0 m × 0.25 mm × 0.32 μm) with a helium
(99.999%) flow rate of 2.0 ml min−1 and a 1 : 10 split ratio.
The oven was heated using the following temperature pro-
gram: the initial temperature of 100 °C was increased to 250
°C at a heating rate of 5 °C min−1 and maintained at that
temperature for 10 min. The mass spectrometer was operated
in electron ionization mode at an energy value of 70 eV.
Production of 2,5-DMF
The production of 2,5-DMF from 5-HMF hydrogenolysis was
conducted in the same reactor as that used for the fructose
dehydration process. In a typical experiment, the reactor was Acknowledgements
charged with catalyst (0.2 g of 5 wt% Ru/C), substrate (1 g of
This work was supported by the National Natural Science
5-HMF), and 60 ml of N,N-dimethylformamide solvent. After-
wards, the reactor was sealed, purged, and pressurized with
H2 (0.7 MPa). The reaction mixture was stirred at 600 rpm
and incubated in an oil bath at a designated temperature. Af-
ter reaction, the reactor was cooled to room temperature, and
the reaction solution was centrifuged and sampled for analy-
Foundation of China (21276230, 20936005 and 21476211), the
Zhejiang Provincial Natural Science Foundation of China
(LY14B060003 and LY16B060004), and the Program for Zhe-
jiang Leading Team of S&T Innovation (2011R50002).
sis. For the one-pot production of 2,5-DMF from fructose, a Notes and references
mixture of fructose (10 mmol), AlCl3 (0.25 mmol), H2SO4 (0.7
1 J. B. Binder and R. T. Raines, J. Am. Chem. Soc., 2009, 131,
mmol), H3PO4 (1.0 mmol), N,N-dimethylformamide (60 ml)
and 5 wt% Ru/C (0.2 g) was added into the reactor, and the
reaction was conducted at 200 °C and 1.5 MPa H2 pressure
for 12 h.
1979–1985.
2 S. De, S. Dutta and B. Saha, ChemSusChem, 2012, 5,
1826–1833.
3 J. H. Zhang, L. Lin and S. J. Liu, Energy Fuels, 2012, 26,
4560–4567.
4 P. Nilges and U. Schroder, Energy Environ. Sci., 2013, 6,
2925–2931.
5 N. Lucas, N. R. Kanna, A. S. Nagpure, G. Kokate and S.
Chilukuri, J. Chem. Sci., 2014, 126, 403–413.
6 S. Nishimura, N. Ikeda and K. Ebitani, Catal. Today,
2014, 232, 89–98.
7 J. Jae, W. Zheng, R. F. Lobo and D. G. Vlachos,
ChemSusChem, 2013, 6, 1158–1162.
8 J. Kang, A. Vonderheide and V. V. Guliants, ChemSusChem,
2015, 8, 3044–3047.
9 M. Chatterjee, T. Ishizaka and H. Kawanami, Green Chem.,
2014, 16, 4734–4739.
10 B. Saha, C. M. Bohn and M. M. Abu-Omar, ChemSusChem,
2014, 7, 3095–3101.
11 Y. Kwon, E. de Jong, S. Raoufmoghaddam and M. T. M.
Koper, ChemSusChem, 2013, 6, 1659–1667.
12 L. Hu, L. Lin and S. J. Liu, Ind. Eng. Chem. Res., 2014, 53,
9969–9978.
13 Y. Román-Leshkov, C. J. Barrett, Z. Y. Liu and J. A. Dumesic,
Nature, 2007, 447, 982–985.
14 Y. H. Zu, P. P. Yang, J. J. Wang, X. H. Liu, J. W. Ren, G. Z. Lu
and Y. Q. Wang, Appl. Catal., B, 2014, 146, 244–248.
15 G. H. Wang, J. Hilgert, F. H. Richter, F. Wang, H. J. Bongard,
B. Spliethoff, C. Weidenthaler and F. Schuth, Nat. Mater.,
2014, 13, 294–301.
Catalyst recyclability study
The recyclability of the Ru/C catalyst for both 5-HMF hydro-
genolysis and the one-pot conversion of fructose to 2,5-DMF
was determined. After reaction, the Ru/C catalyst was
centrifuged, washed several times with N,N-dimethyl-
formamide, and returned to the reactor for the next cycle.
The supernatant of each cycle was analyzed by gas chroma-
tography (GC) and high-performance liquid chromatography
(HPLC) for quantification of reaction components.
Product analysis
The analytical method described in our previous reports was
adopted.16,17 Fructose was analyzed by using HPLC (detector:
Waters 410 Differential Refractometer; Aminex HP X-87H col-
umn, 9 μm, 300 × 7.8 mm) with 5 mmol l−1 H2SO4 in
ultrapure water as the mobile phase flowing at a rate of 0.6
ml min−1 and a column temperature of 60 °C. 5-HMF and
other components were analyzed using an Agilent 7890 GC
equipped with an HP-5 capillary column (30.0 m × 0.32 mm ×
0.25 μm) and a flame ionization detector. The injector tem-
perature was set at 250 °C, and the sampling volume was 0.4
μl. The detector temperature was 260 °C. The split ratio was
1 : 10. The column temperature was increased from 100 °C to
250 °C at a heating rate of 5 °C min−1 and then maintained
at 250 °C for 10 min. 2,5-DMF was identified by its retention
time in comparison with a standard sample and through GC–
MS analysis. 5-HMF and 2,5-DMF were quantified by using
16 Z. J. Wei, Y. Li, D. Thushara, Y. X. Liu and Q. L. Ren,
J. Taiwan Inst. Chem. Eng., 2011, 42, 363–370.
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2016

View Article Online
Catalysis Science & Technology
Paper
17 Z. J. Wei, Y. X. Liu, T. Dilantha and Q. L. Ren, Green Chem.,
2012, 14, 1220–1226.
31 P. Yang, Q. Cui, Y. Zu, X. Liu, G. Lu and Y. Wang, Catal.
Commun., 2015, 66, 55–59.
18 F. Liu, M. Audemar, K. De Oliveira Vigier, J. M. Clacens, F. De
Campo and F. Jerome, ChemSusChem, 2014, 7, 2089–2093.
19 Z. W. Xu, P. F. Yan, W. J. Xu, X. M. Liu, Z. Xia, B. Chung,
S. Y. Jia and Z. C. Zhang, ACS Catal., 2015, 5, 788–792.
20 N. N. Wang, Y. Yao, W. Li, Y. Yang, Z. X. Song, W. T. Liu, H. J.
Wang, X. F. Xia and H. Y. Gao, RSC Adv., 2014, 4, 57164–57172.
21 S. P. Teong, G. Yi and Y. Zhang, Green Chem., 2014, 16,
2015–2026.
32 E. B.-Ł. Małgorzata, E. Zakrzewska and R. Bogel-Łukasik,
Chem. Rev., 2011, 111, 397–417.
33 H. Ma, F. Wang, Y. Yu, L. Wang and X. Li, Ind. Eng. Chem.
Res., 2015, 54, 2657–2666.
34 G. Yi, S. P. Teong, X. Li and Y. Zhang, ChemSusChem,
2014, 7, 2131–2135.
35 L. Hu, X. Tang, J. X. Xu, Z. Wu, L. Lin and S. J. Liu, Ind. Eng.
Chem. Res., 2014, 53, 3056–3064.
22 M. Chidambaram and A. T. Bell, Green Chem., 2010, 12,
1253–1262.
23 T. Thananatthanachon and T. B. Rauchfuss, Angew. Chem.,
Int. Ed., 2010, 49, 6616–6618.
24 B. R. Caes, M. J. Palte and R. T. Raines, Chem. Sci., 2013, 4,
196–199.
25 Y. Nakagawa, M. Tamura and K. Tomishige, ACS Catal.,
2013, 3, 2655–2668.
26 A. Chinnappan, A. H. Jadhav, H. Kim and W. J. Chung,
Chem. Eng. J., 2014, 237, 95–100.
27 Y. X. Liu, Z. B. Li, Y. Yang, Y. X. Hou and Z. J. Wei, RSC Adv.,
2014, 4, 42035–42038.
36 P. P. Upare, M. Lee, S.-K. Lee, J. W. Yoon, J. Bae, D. W.
Hwang, U. H. Lee, J.-S. Chang and Y. K. Hwang, Catal.
Today, 2016, 265, 174–183.
37 S. Dutta and M. Mascal, ChemSusChem, 2014, 7,
3028–3030.
38 W. K. Schumann, O. M. Kut, A. Baiker and G. Gut, J. Chem.
Technol. Biotechnol., 1989, 46, 303–312.
39 A. J. Kumalaputri, G. Bottari, P. M. Erne, H. J. Heeres and K.
Barta, ChemSusChem, 2014, 7, 2266–2275.
40 B. J. Arena, Appl. Catal., A, 1992, 87, 219–229.
41 X. Kong, Y. F. Zhu, H. Y. Zheng, F. Dong, Y. L. Zhu and
Y. W. Li, RSC Adv., 2014, 4, 60467–60472.
28 M. H. Tucker, A. J. Crisci, B. N. Wigington, N. Phadke, R.
Alamillo, J. P. Zhang, S. L. Scott and J. A. Dumesic, ACS
Catal., 2012, 2, 1865–1876.
29 T. D. Swift, C. Bagia, V. Choudhary, G. Peklaris, V. Nikolalds
and D. G. Vlachos, ACS Catal., 2014, 4, 259–267.
30 Y. W. Juanjuan Shi, Xini Yu, Weichen Du and Zhaoyin Hou,
Fuel, 2016, 163, 74–79.
42 J. Mitra, X. Y. Zhou and T. Rauchfuss, Green Chem.,
2015, 17, 307–313.
43 M. Chatterjee, T. Ishizaka and H. Kawanami, Green Chem.,
2014, 16, 1543–1551.
44 F. Liu, M. Audemar, K. De Oliveira Vigier, J.-M. Clacens,
F. De Campo and F. Jérôme, Green Chem., 2014, 16,
4110.
This journal is © The Royal Society of Chemistry 2016
Catal. Sci. Technol.