Hydrogenation of 5-hydroxymethylfurfural in supercritical carbon dioxide-water: A tunable approach to dimethylfuran selectivity

Article abstract of DOI:10.1039/c3gc42145g

The use of supercritical carbon dioxide-water on the hydrogenation of 5-hydroxymethylfurfural (HMF) was investigated over a Pd/C catalyst. It was possible to achieve a very high yield (100%) of DMF within the reaction time of 2 hours at 80 °C. A significant effect of CO₂ pressure was observed on the product distribution. Simply by tuning the CO₂ pressure it was possible to achieve various key compound, such as tetrahydro-5-methyl-2-furanmethanol (MTHFM) (<10 MPa), 2,5-dimethylfuran (DMF) (10 MPa) and 2,5-dimethyltetrahydrofuran (DMTHF) (>10 MPa) with very high selectivity. Optimization of other reaction parameters revealed that H ₂ pressure, temperature, as well as the CO₂-water mole ratio, played an important role in the selectivity to the targeted DMF. It is interesting to note that a very high yield of DMF was achieved when a combination of CO₂ and water was used. For instance, in the absence of water or CO₂, the selectivity of DMF was low; similarly, an excess of water against the fixed pressure of CO₂ reduced the selectivity to DMF. Hence, an optimized amount of water was mandatory in the presence of CO₂ for the formation of DMF with high selectivity. This method was successfully extended to the hydrogenation of furfural, which could afford 100% selectivity to 2-methylfuran with complete conversion within a very short reaction time of 10 min. The studied catalyst could be recycled successfully without significant loss of catalytic activity.

Full text of DOI:10.1039/c3gc42145g

Green Chemistry
View Article Online
View Journal | View Issue
PAPER
Hydrogenation of 5-hydroxymethylfurfural in
supercritical carbon dioxide–water: a tunable
approach to dimethylfuran selectivity†
Cite this: Green Chem., 2014, 16,
1543
Maya Chatterjee,*a Takayuki Ishizakaa and Hajime Kawanami*a,b
The use of supercritical carbon dioxide–water on the hydrogenation of 5-hydroxymethylfurfural (HMF)
was investigated over a Pd/C catalyst. It was possible to achieve a very high yield (100%) of DMF within the
reaction time of 2 hours at 80 °C. A significant effect of CO2 pressure was observed on the product distri-
bution. Simply by tuning the CO2 pressure it was possible to achieve various key compound, such as
tetrahydro-5-methyl-2-furanmethanol (MTHFM) (<10 MPa), 2,5-dimethylfuran (DMF) (10 MPa) and 2,5-
dimethyltetrahydrofuran (DMTHF) (>10 MPa) with very high selectivity. Optimization of other reaction
parameters revealed that H2 pressure, temperature, as well as the CO2–water mole ratio, played an impor-
tant role in the selectivity to the targeted DMF. It is interesting to note that a very high yield of DMF was
achieved when a combination of CO2 and water was used. For instance, in the absence of water or CO2,
the selectivity of DMF was low; similarly, an excess of water against the fixed pressure of CO2 reduced the
selectivity to DMF. Hence, an optimized amount of water was mandatory in the presence of CO2 for the
formation of DMF with high selectivity. This method was successfully extended to the hydrogenation of
furfural, which could afford 100% selectivity to 2-methylfuran with complete conversion within a very
short reaction time of 10 min. The studied catalyst could be recycled successfully without significant loss
of catalytic activity.
Received 17th October 2013,
Accepted 19th November 2013
DOI: 10.1039/c3gc42145g
www.rsc.org/greenchem
ring, makes it an appealing starting material for various
chemical transformations. Comprehensive reviews of the syn-
Introduction
Depletion of fossil fuel resources is a great concern for the thetic chemistry of HMF and its derivatives have been provided
modern world since the availability of inexpensive crude oil is by Kunz and Lewkowski.4 Recently, Rosatella et al. also
diminishing and environmental pressure is escalating.1 To reviewed the synthesis and application of HMF in different
meet the growing demand of energy, due to the rise of the fields of importance.5 As mentioned before, HMF can serve as
automobile industry, sustainable and renewable sources are a precursor for a variety of important chemical intermediates
inevitable in the near future.2 A substantial amount of relevant to the fuel, polymer and pharmaceutical industries.6–8
research activity is currently undertaken for the development Among them 2,5-dimethylfuran (DMF) is one of the most
of strategies for transforming abundant lignocellulosic attractive compounds, which can directly serve as a suitable
biomass into liquid fuel suitable for the transportation sector candidate for liquid fuels. Compared to ethanol (the only
or platform chemicals, and to develop economically feasible sugar based liquid fuel), DMF has a low boiling point (92 °C),
processes on a commercial scale.
is stable in storage, insoluble in water, and has the highest
5-Hydroxymethylfurfural (HMF) is considered as one of the research octane number of 119.9 More attractively, DMF pos-
most versatile platform chemicals derived from biomass.3 The sesses a high volumetric energy density of 31.5 MJ l−1, which
presence of two functional groups, combined with the furan is 40% greater than ethanol, making it comparable to gasoline,
and consumes one-third of the energy in the evaporation stage
of its production in comparison to that required by the fer-
mentation for ethanol.10 The break-through of deriving DMF
from biomass derived fructose was first reported by Roman-
Leshkov et al. using a bimetallic CuRu/C catalyst with a yield
of 71% after 10 hours of reaction in 1-butanol at 120 °C and
0.68 MPa of hydrogen pressure.11 On the other hand, Binder
and Raines reported a DMF yield of 49% under similar
aOrganic Synthesis Team, Research Center for Compact Chemical System,
AIST Tohoku, 4-2-1, Nigatake, Miyagino-ku, Sendai 983-8551, Japan.
E-mail: c-maya@aist.go.jp, h-kawanami@aist.go.jp; Fax: +81 22 237 5388;
Tel: +81 22 237 5213
bCREST, Japan Science and Technology (JST)
†Electronic supplementary information (ESI) available. See DOI: 10.1039/
c3gc42145g
This journal is © The Royal Society of Chemistry 2014
Green Chem., 2014, 16, 1543–1551 | 1543

View Article Online
Paper
Green Chemistry
reaction conditions.12 In a two-step approach, Chidambaram (do not contribute to increase the VOC’s level in the atmos-
and Bell obtained DMF with a yield of 16%, and a 47% conver- phere), health and safety (non toxic, non flammable) and
sion of HMF in 1 hour under a hydrogen pressure of 6.2 MPa process benefits, which is connected to the rare occurrence of
using Pd/C catalyst in an ionic liquid (1-ethyl-3-methylimidazo- by-products owing to side reactions, absence of solvent residue
lium chloride) medium along with acetonitrile.13 They attribu- and facile product separation (cost-effective; related to the sep-
ted the lower yield of DMF to the shorter reaction time, as well aration and purification steps). The most important advantage
as low availability of hydrogen in ionic liquid. Furthermore, of using scCO2 arises from the complete miscibility of reagents
several investigators used a catalytic transfer hydrogenation in the case of chemical reactions involving gasses. Further-
(CTH) process for the conversion of HMF to DMF. Instead of more, the yield and selectivity of a reaction can be enhanced
molecular hydrogen, cyclohexane, secondary alcohols or super- easily by tuning of the pressure and temperature.18 Recently,
critical methanol were used as reaction media, as well as a scCO2 has also been used in the conversion of biogenic
hydrogen source and the reported DMF yield was up to substrates.19
60–80% depending on the reaction conditions used.14
It has to be mentioned that at this stage HMF is an expen-
However, in the CTH process, sometimes the hydrogen sources sive starting material for bio-fuel production because of the
give rise to by-products or co-products, which requires basic difficulties related to the large scale formation of HMF.
additives and harsh reaction conditions.15 Thananatthana- However, to design suitable hydrogenation/hydrogenolysis cat-
chon and Rauchfuss reported a high yield of 95% after reflux- alysts and to establish potential reaction conditions for achiev-
ing a solution of HMF in THF, formic acid, H2SO4 and Pd/C ing the highest yield of DMF, the transformation of HMF to
for 15 hours at 120 °C,16 but the homogeneous acid co-catalyst DMF should be studied as model reaction using HMF as start-
is difficult to separate from the reaction mixture and leads to ing material.
the formation of undesired products. In addition, other strat-
egies were also reported to obtain high yields of DMF from the
direct conversion of HMF or via carbohydrates.17
Results and discussion
In this work, we attempted the hydrogenation of HMF in
A possible reaction profile of HMF hydrogenation is shown in
Scheme 1. All the reactions were conducted in scCO2–water
medium at 80 °C.
supercritical carbon dioxide (scCO2) using a Pd/C catalyst
under reasonably milder reaction conditions. Simple variation
of the reaction conditions results in different compounds with
comparatively higher selectivity.
Catalyst screening
Supercritical carbon dioxide (scCO2) could be introduced as
a potential medium for carrying out chemical reactions. This Different catalysts containing Pt, Pd, Rh and Ru metals on
solvent provides several advantages, including environmental various supports, such as activated carbon (C), mesoporous
Scheme 1 Possible reaction path of HMF hydrogenation in scCO2.
1544 | Green Chem., 2014, 16, 1543–1551
This journal is © The Royal Society of Chemistry 2014

View Article Online
Green Chemistry
Paper
Table 1 Catalyst screeninga
and 11). Despite the very low conversion of 2.2%, the Pd cata-
lyst results mainly in 5-methyl-2-furanmethanol (MFM)
(Table 1; entry 10), whereas the majority of the HMF was con-
verted to unidentified compounds along with polyols (∼20%)
when Ru supported on Al2O3 was used (Table 1; entry 12).
From our experimental findings it could be suggested that Pd
is the preferred choice for the hydrogenation of HMF under
the present reaction conditions. Among the support materials
studied, Al2O3 showed the lowest conversion in comparison to
C and MCM-41 for Pd. As the reaction was conducted in the
presence of water, the hydrophobic/hydrophilic nature of the
support material might influence the catalytic behavior. Based
on the TG/DTA of Pd/C, Pd/MCM-41 and Pd/Al2O3 (not shown),
the hydrophobicity follows the order of C > MCM-41 > Al2O3,
which might be a possible reason for the lowest activity of Pd/
Al2O3, as also detected previously.20 Based on the targeted
high conversion of HMF, as well as the high selectivity to DMF,
we have decided to continue the investigation further using
Pd/C catalyst.
Product selectivity (%)
Conv. (%) MFM DMF DMTHF MTHFM
Entry Catalyst
1
2
Pt/C
Pd/C
4.6
100
0.0 100
0.0 100
0.0
0.0
2.0
0.0
0.0
0.0
0.0
0.0
0.0
2.2
0.0
0.0
2.1
0.0
0.0
0.0
9.1
13.4
0.0
0.0
0.0
95.7
0.0
0.0
0.0
0.0
6.9
0.0
0.0
3
4
Rh/C
Ru/C
23.9
1.5
0.0
0.0
0.0
100
0.0
2.6
0.0
92.8
0.0
0.0
63.9
65.0
86.6
0.0
0.0
0.0
1.7
0.0
5.0
0.0
0.0
26.1
5
6
7
8
Pt/MCM-41
Pd/MCM-41
Rh-MCM-41
Ru/MCM-41
Pt/Al2O3
Pd/Al2O3
Rh/Al2O3
Ru/Al2O3
Pd/C
0.0
41.0
0.0
47.0
6.3
2.2
14.1
13.0
100
94.5
75.8
9b
10
11b
12c
13d
14e
15f
Pd/C
Pd/C
0.0 100
0.0 0.0 100
a Reaction conditions: catalyst : substrate = 1 : 5; temperature = 80 °C,
reaction time = 2 hours, PCO = 10 MPa, PH = 1 MPa. MFM = 5-methyl-
2-furanmethanol, DMF = 2,5-dimethylfuran, DMTHF = 2,5-dimethyl-
2
2
tetrahydro-furan; MTHFM
= tetrahydro-5-methyl-2-furanmethanol.
b Unidentified products. c ∼20% was identified as polyols. d In acidic
condition. e After 4th recycle. f DMF as substrate, PCO = 16 MPa.
Effect of CO2 pressure
2
The effect of CO2 pressure on the hydrogenation of HMF was
studied from 4 to 16 MPa at 80 °C and a fixed hydrogen
pressure of 1 MPa (Fig. 1). Independent to the CO2 pressure,
complete conversion was achieved within the reaction time of
2 hours. However, CO2 pressure exhibited a strong effect on
the product distribution. The selectivity of DMF was increased
from 42.2 to 100% as the pressure increased from 4 to 10 MPa
and then remains constant until 11 MPa. At the lower pressure
region (4–6 MPa), MTHFM was formed with a comparatively
higher selectivity of 57.8% and then decreased with pressure.
Moreover, at higher CO2 pressures (>12 MPa), complete hydro-
genation of DMF was observed and the selectivity abruptly
dropped to 27%, with an increase in the selectivity of DMTHF.
Therefore, for the sake of simplicity, 10 MPa was considered as
the standard and the product distribution was divided into
silica (MCM-41) and alumina (Al2O3), were screened for the
hydrogenation of HMF and the results are shown in Table 1.
In all cases, the catalyst to substrate ratio was taken as 1 : 5
and the reaction was studied using a CO2–water medium.
Under the studied reaction conditions (temperature = 80 °C,
reaction time = 2 hours, PCO = 10 MPa and PH = 1 MPa)
mainly three quantified products 2,5-dimethylfuran (DMF),
2,5-dimethyltetrahydrofuran (DMTHF) and tetrahydro-5-
methyl-2-furanmethanol (MTHFM) were observed. Notably, no
evidence for the decomposition products of HMF was
detected.
When activated carbon (C) was used as the support, the
conversion of HMF varied from 1.5 to 100%, and followed the
reactivity order of Pd > Rh > Pt > Ru. When using either Pt or
Pd, the highest selectivity of DMF (100%) was obtained,
(Table 1; entries 1 and 2), however, Pt/C provides only 4.6%
conversion of HMF (Table 1; entry 2). On the other hand, DMF
was also formed in the presence of the Rh and Ru catalysts
with moderately high selectivities of 65% and 86%, respect-
ively, but the conversion was relatively low [23.0 (Rh/C) and
1.5% (Ru/C)] for both materials (Table 1; entries 3 and 4).
By changing the catalyst support from C to MCM-41, the
conversion of HMF as well as the product distribution changed
significantly. For Pd/MCM-41, the conversion dropped to 41%
(Table 1; entry 6) and no peak corresponding to DMF was
observed in the GC-MS trace. The Pt and Rh/MCM-41 catalysts
were completely inactive (Table 1; entries 5 and 7), whereas for
Ru/MCM-41, a conversion of 47% and a very low selectivity to
DMF (∼2%) was detected (Table 1; entry 8).
2
2
Fig. 1 Effect of CO2 pressure on the conversion, and product profile.
Reaction conditions: catalyst : substrate = 1 : 5; temperature = 80 °C,
As shown in Table 1 (entries 9–12), when the reaction was
conducted on Al2O3 supported catalysts, Pt and Rh exhibited
6.3 and 14.1% conversion, respectively, and the product
mixture did not reveal any significant peaks (Table 1; entries 9
reaction time = 2 hours, PH = 1 MPa, water = 1 ml; DMF = 2,5-dimethyl-
2
furan, DMTHF = 2,5-dimethyl-tetrahydro-furan; MTHFM = tetrahydro-5-
methyl-2-furanmethanol.
This journal is © The Royal Society of Chemistry 2014
Green Chem., 2014, 16, 1543–1551 | 1545

View Article Online
Paper
Green Chemistry
three different regions depending on the CO2 pressure: <10 DMF was successfully converted to DMTHF (Table 1; entry 15)
MPa, 10 MPa and >10 MPa. As such it is difficult to explain the under the working conditions. Therefore, from the results of
observed results; hence, the reaction was conducted at 4, 10 phase observation it might be suggested that depending on
and 16 MPa of CO2 pressure in a view cell to obtain insight the CO2 pressure, three completely different situations arise
into the phase behavior during reaction (Fig. 2a–c), because during the reaction and result in three different products. It
the reaction was performed in scCO2; one factor that might has to be mentioned that the DMTHF is also an important
contribute to the observed results is phase behavior. At 4 MPa, compound with a higher carbon to hydrogen ratio, which
two phases (i) the aqueous phase containing dispersed catalyst translates into a higher energy content. In addition, it may
and HMF (bottom phase) and (ii) CO2–H2 in gaseous state provide additional stability on storage because of the fully
(upper phase) were visible (Fig. 2a). The scenario is different at hydrogenated furan ring.11
10 MPa; where one can see two immiscible liquid phases of (i)
substrate–water/catalyst (bottom phase) and (ii) liquid part
(middle phase) along with the (iii) CO2–H2 in gaseous state
Effect of hydrogen pressure
(upper phase) (Fig. 2b). Accumulated experimental results
indicated that the observed situation was favorable to the for-
mation of DMF, which is a hydrophobic product and can be
separated spontaneously from water.11 At the higher pressure
of 16 MPa, in comparison with the substrate–catalyst/water
phase (bottom phase), a large portion of the view cell was
occupied by the CO2–H2 liquid phase (upper phase) (Fig. 2c),
which is a favorable situation for the complete hydrogenation
of DMF to DMTHF. As observed from the phase behavior of
DMF, it was completely soluble in CO2 (ESI, Fig. 1s†) at 80 °C
and a CO2 pressure of 16 MPa. From the results given above,
the formation of DMTHF was observed in the single phase
system where CO2 homogenizes the mixture of DMF and
hydrogen, leading to high concentrations of each reactant and
hence high DMTHF selectivity. To confirm the conversion of
DMF to DMTHF, a separate experiment was conducted on the
hydrogenation of DMF under similar reaction conditions.
Hydrogen pressure is one of the important reaction parameters
to be optimized, as the reaction was conducted in scCO2. Fig. 3
presents the effect of changing the hydrogen pressure from 0.2
to 2.5 MPa on the conversion and selectivity of HMF hydrogen-
ation. The temperature and CO2 pressure were kept constant at
80 °C and 10 MPa, respectively, in accordance with the
observed highest activity and selectivity. As expected, the con-
version of HMF was increased from 47.1 to 100% with the
change in hydrogen pressure from 0.2 to 1 MPa, and then
remained constant. The highest selectivity of the targeted DMF
(100%) was maintained until 1 MPa, the selectivity then
sharply decreased to 19% as the pressure was enhanced to 2.5
MPa. The increase in hydrogen pressure suppressed the
selectivity of DMF due to the formation of over hydrogenated
DMTHF, by-products and hexanediol. Therefore, a lower
pressure of hydrogen was most suitable for the selective for-
mation of DMF. An optimum hydrogen pressure of 1 MPa was
Fig. 2 Phase observation through the view cell during reaction (a) initial (b) 4 MPa, (c) 10 MPa and (d) 16 MPa. Reaction conditions: temperature =
80 °C, reaction time = 2 hours, PH = 1 MPa.
2
1546 | Green Chem., 2014, 16, 1543–1551
This journal is © The Royal Society of Chemistry 2014

View Article Online
Green Chemistry
Paper
(ESI†). The GC-MS traces of the product mixture displayed a
few signals, allowing the detection of 2,5-bis(hydroxymethyl)-
furan (MF) as the main intermediate. In addition to that, some
other compounds (<10%) were also identified in the GC-MS
traces (Table 1s, ESI†). Thus, it could be suggested that when a
1 MPa hydrogen pressure was used, HMF was directly con-
verted to DMF, but the presence of an intermediate (MF) was
evident at a very low pressure of hydrogen. Based on this
observation, we hypothesized that the reaction proceeds by
rapid reduction of the –CHO group in HMF, followed by hydro-
genolysis of the hydroxyl functionalities in scCO2–water.
However, additional research is necessary to determine the
exact reaction path of HMF hydrogenation in the studied reac-
tion conditions.
In addition, DMF was fully converted to DMTHF due to the
reduction of the furan ring after a reaction time of 6 hours.
With further extension of the reaction time to 18 hours,
DMTHF along with 1,2-hexandiol was formed.
Fig. 3 Variation of hydrogen pressure. Reaction conditions: catalyst :
substrate = 1 : 5; PCO = 10 MPa, temperature = 80 °C, reaction time =
2
2
hours, water
=
1
ml; DMF
=
2,5-dimethylfuran, DMTHF
= 2,5-
dimethyl-tetrahydro-furan; MTHFM
methanol.
=
tetrahydro-5-methyl-2-furan-
Effect of temperature
chosen to achieve a high yield of the desired DMF in order to
optimize the other reaction parameters.
The effect of temperature on the hydrogenation of HMF was
studied at 35, 50, 80, 100 and 130 °C at constant pressures of
CO2 and H2, 10 and 1 MPa, respectively (Fig. 5). When the reac-
tion was performed at a temperature of 35 °C the conversion
was low (28.2%), but with an increase in temperature the con-
version increased and reached a maxima of 100% at 80 °C.
Regarding the product selectivity, at the lower temperature of
35 °C, the main product was MTHFM (68%) and the selectivity
of the targeted DMF was substantially low (11%), which was
improved significantly with an increase in temperature, where
the highest selectivity of 100% was achieved at 80 °C. Interest-
ingly, when the temperature was further increased to 130 °C, a
ring hydrogenated product (DMTHF) constitutes up to 80% of
the total products and consequently the selectivity to DMF
Variation of reaction time
To understand the reaction path of HMF hydrogenation it is
necessary to evaluate the intermediate product/products
formed during the reaction. The hydrogenation of HMF was
studied as a function of time and the results are shown in
Fig. 4. Under the investigated conditions, a very short reaction
time of 5 min provided low conversion of <10% and then
gradually improved to 100% as the reaction time progressed to
2 hours. Surprisingly, independent to the reaction time, DMF
was the sole product detected, even in the shortest reaction
time of 5 min, which might be attributed to the rapid hydro-
genation/hydrogenolysis of HMF, possibly because of the easy
availability of hydrogen in the reaction system. Thus, to slow
down the reaction rate, the reaction was conducted at a very
low pressure of hydrogen (0.1 MPa) within the shortest reac-
tion time of 5 min and the results are described in Table 1s
Fig. 5 Effect of temperature on the conversion and product selectivity.
Reaction conditions: catalyst : substrate = 1 : 5; PCO = 10 MPa, PH
=
2
2
1 MPa, reaction time = 2 hours, water = 1 ml; DMF = 2,5-dimethylfuran,
DMTHF 2,5-dimethyl-tetrahydro-furan; MTHFM tetrahydro-5-
Fig. 4 Time dependent conversion of HMF hydrogenation in scCO2–
=
=
water. Reaction conditions: catalyst : substrate = 1 : 5; PCO = 10 MPa,
temperature = 80 °C, reaction time = 2 hours, water = 1 ml.
methyl-2-furanmethanol. The circled section shows the results of
10 min reaction time.
2
This journal is © The Royal Society of Chemistry 2014
Green Chem., 2014, 16, 1543–1551 | 1547

View Article Online
Paper
Green Chemistry
dropped. This result might be attributed to the increase in the (Scheme 1: Path B). From the accumulated results one can
reaction rate, facilitated by the formation of DMF and conse- understand that in the absence of water or CO2, the reaction
quently DMTHF, which is in good agreement with the results proceeds through the path of HMF → MFM → MTHFM.
of the previous section on the hydrogen pressure effect. To A combined effect of CO2 and water forced the reaction to
explain this observation, the reaction time was shortened from move in the direction of HMF → DMF (Scheme 1: Path C). We
2 hours to 10 min at 130 °C and the results are shown in Fig. 5 hypothesized that the acidity of the CO2–water system might
(circled). The reaction rate (turnover frequency; TOF) was com- be one of the responsible factors.22 A truly biphasic system
pared at 80 and 130 °C under a fixed CO2 pressure of 10 MPa consisting of CO2–H2 and the aqueous solution of HMF exists
and reaction time of 10 min. As expected, the TOF was very during the course of the reaction (neglecting the solid catalyst
high (3167 h−1) at 130 °C, in comparison with 351 h−1 at phase) (Fig. 2s, ESI†). In the presence of CO2, water cannot
80 °C. The acceleration of the reaction rate with increased expand; therefore, the properties of the water remain intact,
temperature at a given CO2 pressure was also observed for the except the acidity.23 Thus, when CO2 was introduced into the
hydrogenation that occurred in water–CO2 biphasic media.21 reaction system, the reaction medium became acidic. Accord-
Regarding the product profile, the selectivity of DMF and ing to the literature, the pH value of 3 MPa CO2 in water at
DMTHF was changed from 100 to 28.9% and 0 to 66%, 50 °C was reported to be about 3.4, and there was no signifi-
respectively (Fig. 5, circled) within the reaction time of 10 min, cant change observed upon increasing the CO2 pressure.20
as the temperature increased from 80 to 130 °C. From the Namely, the weak acidic conditions might be one of the
results it seems that very low temperatures were clearly unsui- factors behind the hydrogenation of HMF to DMF. To confirm
table for the generation of DMF, and a higher temperature of this observation, the reaction was conducted separately at pH
130 °C boosted the reaction rate and consequently the selecti- 3.0 using acetic acid and without addition of any CO2. Interest-
vity of DMTHF. Therefore, a mild temperature of 80 °C, which ingly, the selectivity of DMF was increased to 26.1% in com-
provides complete conversion of HMF to DMF, was used parison with only water or the CO2 system (Table 1; entry 13),
throughout the rest of the optimization process.
but considerably lower than for the CO2–water combined
system, which emphasizes the beneficial effect of CO2 for
achieving excellent catalytic activity. Previously, a similar effect
was also detected during the hydrogenation/hydrogenolysis of
diphenylether in scCO2–water medium.24
Variation of the water to CO2 mole ratio
Fig. 6 presents the results of the hydrogenation of HMF in
CO2–water, only water and only CO2 systems under similar
reaction conditions. Complete conversion of HMF was
achieved within the reaction time of 2 hours in each case.
However, the product distribution was different, depending on
the presence or absence of the either of CO2 or water. In the
absence of water (only CO2), the major product was MTHFM,
which was obtained with a selectivity of 95.3% (Scheme 1: Path
A). On the other hand, when only water was used, the reaction
ended with the formation of MFM (95.8%) and DMF (4.2%)
To optimize the amount of water, the hydrogenation of
HMF was also studied with the variation of CO2 to water mole
ratio (Fig. 6). Independent to the amount of water used, the
conversion remain constant, however the product distribution
was strongly affected. It can be seen that when the water–CO2
ratio was changed from 0.16 : 1 to 1.3 : 1 the hydrogenation of
HMF followed a different reaction path of HMF to MTHFM
and the selectivity of DMF was dropped significantly. For
instance, the selectivity of DMF changes from 100 to 32% as
the water–CO2 ratio changes from 0.17 : 1 to 1.3 : 1. Based on
our observations, we might conclude that the formation of
DMF prefers low water–CO2 ratio. When the amount of water
increased at the fixed concentration of CO2, the system
behaves like an only water system and shifts the reaction
towards the formation of MTHFM except that DMF was
formed with very low selectivity because of the presence of
CO2. This result agreed well with the observation of the reac-
tion at lower CO2 pressure (4 MPa), where MTHFM was
detected with comparatively higher selectivity due to the
water–CO2 ratio of 0.64 : 1. Depending on the highest conver-
sion and selectivity an optimum ratio of water–CO2 = 0.32 : 1
was used throughout the experiment.
Instead of CO2, different organic solvents such as hexane,
1-butanol and tetrahydrofuran (THF) were used along with the
water (Fig. 7). According to the results, highest and lowest con-
version of HMF was detected in 1-butanol (69.2%) and hexane
(23.3%), respectively. In hexane, no DMF was detected,
whereas THF shows highest selectivity of 38.3%. The selectivity
of DMF follows the order of THF > 1-butanol > hexane. It has
Fig. 6 Variation of the water–CO2 mole ratio. Reaction conditions:
catalyst : substrate = 1 : 5; PCO = 10 MPa, PH = 1 MPa, reaction time =
2
2
2 hours, temperature = 80 °C; DMF = 2,5-dimethylfuran, DMTHF = 2,5-
dimethyl-tetrahydro-furan; MTHFM tetrahydro-5-methyl-2-furan-
=
methanol, MF = 2,5-bis(hydroxymethyl)furan (MF), MFM = 5-methylfura-
nyl methanol.
1548 | Green Chem., 2014, 16, 1543–1551
This journal is © The Royal Society of Chemistry 2014

View Article Online
Green Chemistry
Paper
Hydrogenation of furfural
This method was also been extended to the hydrogenation of
furfural and the results are shown in Table 2. Under similar
reaction conditions, furfural was completely hydrogenated to
2-methyltetrahydrofuran (Table 2; entry 1) within the reaction
time of 2 hours. However, when the reaction time was changed
from 2 hours to 10 min, furfural was completely converted to
2-methylfuran (Table 2; entry 3). Similarly, tuning of the
different reaction parameters, such as hydrogen pressure,
changes the product profile and a mixture of products was
obtained (Table 2; entries 4–6).
Conclusion
In conclusion, it has been demonstrated that scCO2–water is a
potential medium for facilitating the hydrogenation/hydroge-
nolysis of HMF to DMF. The activity of Pt, Pd, Rh and Ru sup-
ported on C, MCM-41 and Al2O3 was compared. Complete
conversion of HMF to DMF was achieved under mild reaction
conditions within the reaction time of 2 hours using Pd/C cata-
lyst at 80 °C. The product distribution was found to be depen-
dent on the different reaction parameters, such as CO2 and H2
pressure, temperature, and the amount of water used. A com-
bined effect of CO2 and water played a critical role in the for-
mation of DMF with highest selectivity. A promotional effect of
Fig. 7 Different organic solvents along with water; Reaction conditions:
catalyst : substrate = 1 : 5; PCO = 10 MPa, PH = 1 MPa, reaction time =
2 hours, temperature = 80 °C, water = 1 ml; DMF = 2,5-dimethylfuran.
2
2
to be mentioned that although scCO2 is considered as a non-
polar solvents, it is superior to similar organic solvents
(hexane) as it possesses substantially different characteristics,
which can direct the activity and selectivity of a reaction.
Recycling of catalyst
After reaction, the catalyst was separated simply by filtration, the CO2–water system was discovered when different organic
washed with acetone and then ready to recycle. Recycling solvents were used along with water. Tuning of the reaction
experiments showed that it is possible to recycle the catalyst conditions offers the ability to achieve either DMF, DMTHF or
four times. The result after the 4th recycle suggested a moder- MTHFM with very high selectivity. This study suggested that it
ate decrease in the activity (Table 1; entry 14). The TEM image is possible to obtain a very high yield of DMF by the hydrogen-
before and after the 4th recycle indicated a slight increase in ation/hydrogenolysis of HMF under reasonably mild reaction
the Pd particle size, which might be the probable reason conditions, which is an example of a green and sustainable
behind the decrease in the catalytic activity (ESI, Fig. 3s†).
process. Further studies are necessary to improve the
Table 2 Hydrogenation of furfurala
Product selectivity (%)
Entry
PH (MPa)
Time (min)
Conv. (%)
2
1
2
3
4
5
6
1
1
1
0.2
0.2
0.2
120
20
10
15
10
5
100
100
100
65.5
21.1
13.0
0.0
0.0
0.0
0.0
0.0
0.0
74.9
100
59.9
100
8.1
100.0
25.1
0.0
40.1
0.0
91.9
0.0
a Reaction conditions: catalyst : substrate = 1 : 5; temperature = 80 °C and PCO = 10 MPa.
2
This journal is © The Royal Society of Chemistry 2014
Green Chem., 2014, 16, 1543–1551 | 1549

View Article Online
Paper
Green Chemistry
reusability of the catalyst and also to understand the reaction
path in the CO2–water system.
Acknowledgements
H. Kawanami is funded by the Japan–U.S. cooperation
project for research and standardization of Clean Energy
Technologies, The Ministry of Economy, Trade and Industry
(METI), Japan and CREST, Japan Science and Technology
(JST).
Experimental
Materials
5-Hydroxymethylfurfural (HMF) (Aldrich) was used as received.
Carbon dioxide (>99.99%) was supplied by Nippon Sanso Co.
Ltd. Pd/C, Pd/Al2O3, Pt/C, Pt/Al2O3, Rh/C, Ru/C and Ru/Al2O3
were from Aldrich, except Rh/Al2O3, which was from Wako
Pure Chemicals. In each case, the metal content was 5%.
Metal catalysts supported on MCM-41 were synthesized in our
laboratory using a modified method.25
References
1 (a) C. J. Campbell and J. H. Laherrere, Sci. Am., 1998, 278,
60–65; (b) D. H. Klass, Biomass for renewable energy, fuels
and chemicals, Acad. Press, San Diego, California, 1998,
pp. 10–19; (c) C. Okkerse and H. van Bekkum, Green Chem.,
1999, 1, 107–114; (d) L. A. Lucia, D. S. Argyropoulos,
L. Adamopoulos and A. R. Gaspar, Can. J. Chem., 2006, 84,
960–970.
2 (a) B. Kamm, Angew. Chem., Int. Ed., 2007, 46, 5056–5058;
(b) J. J. Bozell, Science, 2010, 329, 522–523; (c) J. Q. Bond,
D. A. Alonso, D. Wang, R. M. West and J. A. Dumesic,
Science, 2010, 327, 1110–1114.
3 Pacific Northwest National Laboratory (PNNL) and National
Renewable Energy Laboratory (NREL), Top Value Added
Chemicals From Biomass, ed. T. Werpy and G. Petersen, U.S.
Department of Energy, August 2004, http://www.nrel.gov/
docs/fy04osti/35523.pdf
4 (a) M. Kunz, Hydroxymethylfurfural a possible basic chemi-
cal for industry intermediates, in Inulin and Inulin Contain-
ing Crops, ed. A. Fuchs, Elsevier Science Publisher B, 1993;
(b) V. J. Lewkowski, ARKIVOC, 2001, 17–54.
Catalytic activity
The HMF hydrogenation studies were carried out in a 50 ml
batch reactor placed in a hot air circulating oven, full details
are given elsewhere.26 In a typical experiment, a specified
amount of catalyst and substrate dissolved in water was intro-
duced into the reactor. The reactor was heated for a specified
amount of time. After the required temperature of 80 °C was
attained; hydrogen, followed by CO2, was charged into the
reactor using a high-pressure liquid pump and then com-
pressed to the desired pressure. The content of the reactor was
stirred with a magnetic stirrer bar during the reaction. After
the reaction was complete, the reactor was quenched using an
ice bath followed by the separation of solid catalyst from the
liquid product by filtration. The products were identified by
GC-MS against a standard, which was also used for qualitative
analysis. For all results reported, the selectivity is follows:
5 A. A. Rosatella, S. P. Simeonov, R. F. M. Frade and
C. A. M. Afonso, Green Chem., 2011, 13, 754–793.
concentration of the product
total concentration of products
% selectivity ¼
ꢀ 100
6 (a) J. N. Chheda, G. W. Huber and J. A. Dumesic, Angew.
Chem., Int. Ed., 2007, 46, 7164–7183; (b) G. W. Huber,
S. Iborra and A. Corma, Chem. Rev., 2006, 106, 4044–4098;
(c) J. N. Chheda and J. A. Dumesic, Catal. Today, 2007, 123,
59–70; (d) M. Mascal and E. B. Nikitin, Angew. Chem., Int.
Ed., 2008, 47, 7924–7926; (e) G. W. Huber, J. N. Chheda,
C. J. Barret and J. A. Dumesic, Science, 2005, 308, 1446–
1450; (f) E. L. Kunkes, D. A. Simonetti, R. M. West,
J. C. Serrano-Ruiz, C. A. Gartner and J. A. Dumesic, Science,
2008, 322, 417–421.
7 (a) G. W. Huber and A. Corma, Angew. Chem., Int. Ed., 2007,
46, 7184–7201; (b) J. O. Metzger, Angew. Chem., Int. Ed.,
2006, 45, 696–698; (c) C. Moreau, R. Durand, S. Razigade,
J. Duhamet, P. Faugeras, P. Rivalier, P. Ros and G. Avignon,
Appl. Catal., A, 1996, 145, 211–224.
Phase observation
To determine the phase behavior of the hydrogenation/hydro-
genolysis of HMF during the reaction, an aqueous solution of
HMF and DMF was studied separately in a 10 ml high pressure
view cell fitted with a sapphire window. The cell was placed
over a magnetic stirrer for stirring of the contents and was con-
nected to a pressure controller in order to regulate the pressure
inside the view cell. In addition, a temperature controller was
also used to maintain the desired temperature of 80 °C.
Required materials were introduced into the view cell at a con-
stant hydrogen pressure of 1 MPa, while the CO2 pressure was
varied and the phase was observed. For phase observation
during the reaction, the catalyst dispersed in an aqueous solu-
tion of HMF was introduced into the view cell and the content
was stirred continuously. The image was recorded after stop-
ping the stirrer. The phase behavior of HMF and DMF was also
checked separately. Independent to the CO2 pressure, an
aqueous solution of HMF and CO2 always existed as a biphasic
8 M. Chatterjee, K. Matsushima, Y. Ikushima, T. Yokoyama,
M. Sato, H. Kawanami and T. Suzuki, Green Chem., 2010,
12, 779–782.
9 M. T. Barlow, D. J. Smith and D. G. Steward, Fuel compo-
sition, European patent, EP0082689, 1983.
system and shown in Fig. 2s,† whereas the solubility of DMF 10 H. B. Nisbet, J. Inst. Petrol., 1946, 32, 162–166.
was increased with increasing CO2 pressure. A homogeneous 11 Y. Roman-Leshkov, C. J. Barrett, Z. Y. Liu and
phase was observed at 12 MPa of CO2 (Fig. 1s-d†).
J. A. Dumesic, Nature, 2007, 447, 982–986.
1550 | Green Chem., 2014, 16, 1543–1551
This journal is © The Royal Society of Chemistry 2014

View Article Online
Green Chemistry
Paper
12 J. B. Binder and R. T. Raines, J. Am. Chem. Soc., 2009, 131,
A. Suzuki and H. Kawanami, Green Chem., 2012, 14,
3415–3422; (c) M. Chatterjee, T. Ishizaka, T. Suzuki,
A. Suzuki and H. Kawanami, Catal. Sci. Technol., 2011, 1,
1466–1471.
1979–1985.
13 M. Chidambaram and A. T. Bell, Green Chem., 2010, 12,
1253–1262.
14 (a) S. Morikawa, Noguchi Kenkyusho Jiho, 1980, 23, 39–44; 19 (a) S. X. Wu, H. I. Fan, Y. Xie, Y. Cheng, Q. Wang,
(b) J. Jae, W. Zheng, R. F. Lobo and D. G. Vlachos, Chem-
SusChem, 2013, 6, 1158–1162; (c) T. S. Hansen, K. Barta,
P. T. Anastas, P. C. Ford and A. Riisager, Green Chem., 2012,
14, 2457–2461.
Z. F. Zhang and B. X. Han, Green Chem., 2010, 12, 1215–
1219; (b) R. A. Bourne, J. G. Stevens, J. Ke and M. Poliakoff,
Chem. Commun., 2007, 4632–4634; (c) J. G. Stevens,
R. A. Bourne, M. V. Twigg and M. Poliakoff, Angew. Chem.,
Int. Ed., 2010, 49, 8856–8859.
15 (a) G. Briger and T. J. Nestrick, Chem. Rev., 1974, 74, 567–
580; (b) R. C. Mebane and A. J. Mansfield, Synth. Commun., 20 (a) H. Yamada and S. Goto, J. Chem. Eng. Jpn., 2003, 36,
2005, 35, 3083–3086; (c) F. Alonso, P. Riente and M. Yus,
Acc. Chem. Res., 2011, 44, 379–391; (d) F. Alonso, P. Riente
105–109; (b) F. Omota, A. C. Dimian and A. Blick, Appl.
Catal., A, 2005, 294, 121–130.
and M. Yus, Tetrahedron, 2008, 64, 1847–1852; 21 K. Burgemeister, G. Franci, V. H. Gego Lasse Greiner,
(e) G. P. Boldrini, D. Savoia, E. Tagliavini, C. Trombini and
A. U. Ronchi, J. Org. Chem., 1985, 50, 3082–3086;
H. Hugl and W. Leitner, Chem.–Eur. J., 2007, 13, 2798–
2804.
(f) M. Kidwai, V. Bansal, A. Saxena, R. Shankar and 22 (a) C. Roosen, M. Ansorge-Schumacher, T. Mang,
S. Mozumdar, Tetrahedron Lett., 2006, 47, 4161–4165.
16 T. Thananatthanachon and T. B. Rauchfuss, Angew. Chem.,
2010, 122, 6766–6768, (Angew. Chem., Int. Ed., 2010, 49,
6616–6618).
17 (a) G. C. A. Luijkx, N. P. M. Huck, F. Van Rantwijk, L. Maat
and H. Van Bekkum, Heterocycles, 2009, 77, 1037–1044;
W. Leitner and L. Greiner, Green Chem., 2007, 9, 455–458;
(b) X.-M. Qi, S.-J. Yao and Y.-X. Guan, Biotechnol. Prog.,
2004, 20, 1176–1182.
23 (a) P. G. Jessop and B. Subramaniam, Chem. Rev., 2007,
107, 2666–2694; (b) H.-W. Lin, C. H. Yen and C.-S. Tan,
Green Chem., 2012, 14, 682–687.
(b) J. Zhang, L. Lin and S. Liu, Energy Fuels, 2012, 26, 4560– 24 M. Chatterjee, T. Ishizaka, A. Suzuki and H. Kawanami,
4567; (c) Y. Kwon, Ed. de Jong, S. Raoufmoghaddam and
M. T. M. Koper, ChemSusChem, 2013, 6, 1659–1667.
18 (a) M. Chatterjee, A. Chatterjee, T. Ishizaka, T. Suzuki,
Chem. Commun., 2013, 49, 4567–4569.
25 M. Chatterjee, T. Iwasaki, Y. Onodera and T. Nagase, Catal.
Lett., 1999, 61, 199–202.
A. Suzuki and H. Kawanami, Adv. Synth. Catal., 2012, 26 M. Chatterjee, F.-Y. Zhao and Y. Ikushima, New J. Chem.,
354, 2009–2018; (b) M. Chatterjee, T. Ishizaka, T. Suzuki,
2003, 27, 510–513.
This journal is © The Royal Society of Chemistry 2014
Green Chem., 2014, 16, 1543–1551 | 1551