A novel magnetic palladium catalyst for the mild aerobic oxidation of 5-hydroxymethylfurfural into 2,5-furandicarboxylic acid in water

Article abstract of DOI:10.1039/c4cy01407c

In this study, magnetically separable, graphene oxide-supported palladium nanoparticles (C-Fe₃O₄-Pd) were successfully prepared via a one-step solvothermal route. The C-Fe₃O₄-Pd catalyst showed excellent catalytic performance in the aerobic oxidation of 5-hydroxymethylfurfural (HMF) into 2,5-furandicarboxylic acid (FDCA). The base concentration and reaction temperature significantly affected both HMF conversion and FDCA selectivity. High HMF conversion (98.2%) and FDCA yield (91.8%) were obtained after 4 h at 80°C with a K₂CO₃/HMF molar ratio of 0.5. The C-Fe₃O₄-Pd catalyst was easily collected by an external magnet and reused without significant loss of its catalytic activity. The developed method is a green and sustainable process for the production of valuable FDCA from renewable, bio-based HMF in terms of the use of water as solvent, the use of stoichiometric amount of base, high catalytic activity under atmospheric oxygen pressure, and facile recyclability of the catalyst.

Full text of DOI:10.1039/c4cy01407c

Catalysis
Science &
Technology
View Article Online
View Journal
PAPER
A novel magnetic palladium catalyst for the mild
aerobic oxidation of 5-hydroxymethylfurfural into
Cite this: DOI: 10.1039/c4cy01407c
2,5-furandicarboxylic acid in water
Nan Mei, Bing Liu,* Judun Zheng, Kangle Lv, Dingguo Tang and Zehui Zhang*
3 4
In this study, magnetically separable, graphene oxide-supported palladium nanoparticles IJC–Fe O –Pd)
3 4
were successfully prepared via a one-step solvothermal route. The C–Fe O –Pd catalyst showed excellent
catalytic performance in the aerobic oxidation of 5-hydroxymethylfurfural (HMF) into 2,5-furandicarboxylic
acid (FDCA). The base concentration and reaction temperature significantly affected both HMF conversion
and FDCA selectivity. High HMF conversion (98.2%) and FDCA yield (91.8%) were obtained after 4 h at 80
°C with a K CO /HMF molar ratio of 0.5. The C–Fe O –Pd catalyst was easily collected by an external mag-
Received 27th October 2014,
Accepted 1st April 2015
2
3
3
4
net and reused without significant loss of its catalytic activity. The developed method is a green and sus-
tainable process for the production of valuable FDCA from renewable, bio-based HMF in terms of the use
of water as solvent, the use of stoichiometric amount of base, high catalytic activity under atmospheric
oxygen pressure, and facile recyclability of the catalyst.
DOI: 10.1039/c4cy01407c
www.rsc.org/catalysis
1
4
terephthalate) (PET) plastics. Therefore, there is growing
attention on the synthesis of FDCA by the oxidation of HMF.
The oxidation of HMF into FDCA has been performed
under different reaction conditions using both homogeneous
and heterogeneous catalysts. In an early work, the combined
Introduction
Currently, there is a growing concern stemming from the
gradual depletion of fossil oil reservoirs and the awareness of
1
,2
climate change. Much effort has been devoted to the search
for renewable resources as alternatives to fossil resources in
supplying chemicals and fuels. Biomass, which is the only
carbon-containing renewable resource, consists mainly of car-
bohydrate components such as starch, cellulose, and hemicel-
lulose. It is not only abundant on the earth, but also keeps
the carbon balance. Through biorefinery, biomass can be
2
+
2+
−
Co /Mn /Br catalysts were used for the oxidation of HMF
into FDCA in acetic acid at 125 °C under 70 bar air pres-
1
5
sure. As it is difficult to recycle homogeneous catalysts,
researchers mainly focused on the development of new
heterogeneous catalysts for the oxidation of HMF into FDCA.
In most cases, inorganic material-supported Pt, Pd, Ru and
Au nanoparticles have been used as heterogeneous catalysts
for the aerobic oxidation of HMF into FDCA. For example,
Davis et al. reported that Pd/C and Pt/C catalysts promoted
the full oxidation of HMF, with FDCA yields between 71%
3
–5
converted into useful chemicals and valuable fuel.
Among range of potential platform compounds,
-hydroxymethylfurfural (HMF) can be generated by acid-
a
5
6
–9
catalyzed dehydration of C6-sugar monomers.
It has been
identified as a versatile platform chemical with high potential
for the synthesis of valuable chemicals and liquid fuels.
and 79% and NaOH/HMF mole ratio of 2 under 690 kPa O
2
pressure. Supported Au nanoparticles have shown encour-
aging catalytic performance for the aerobic oxidation of HMF
1
0
16
Recently, selective oxidation of HMF has been of particular
interest, since it can produce several important chemicals,
such as 2,5-diformylfuran (DFF), 5-hydroxymethyl-2-furan-
carboxylic acid (HMFCA) and 2,5-furandicarboxylic acid
1
1–13
(FDCA) (Scheme 1).
FDCA has been identified as one of
the top 12 value-added chemicals from biomass by the U.S.
Department of Energy. It has a similar structure with
terephthalic acid; thus it has high potential as a substitute
for terephthalic acid in the manufacture of poly(ethylene
Department of Chemistry, Key Laboratory of Catalysis and Material Sciences of
the State Ethnic Affairs Commission & Ministry of Education, South-Central
University for Nationalities, Minyuan Road 182, Wuhan, PR China.
E-mail: liubing@mail.scuec.edu.cn, zehuizh@mail.ustc.edu.cn
Scheme 1 Major oxidation products from the oxidation of HMF.
This journal is © The Royal Society of Chemistry 2015
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
1
7–21
to FDCA in water and have received great interest.
How-
DFF (98%) and FDCA (97.0%) were purchased from the J&K
Chemical Co. Ltd., (Beijing, China). Acetonitrile (HPLC grade)
was purchased from Tedia Co. (Fairfield, USA). All the sol-
vents were purchased from Sinopharm Chemical Reagent
Co., Ltd. (Shanghai, China).
ever, the stability of the supported Au catalysts in the aerobic
oxidation of HMF remains a problem. For example, Corma
et al. reported that CeO -supported Au nanoparticles (CeO /
Au) showed high catalytic activity for the oxidation of HMF to
FDCA with a high yield of 99% at 130 °C under 1 MPa air
pressure and high concentration of NaOH (4 equiv of HMF),
2
2
Preparation of graphite oxide
but the activity of CeO /Au in the second run declined
2
Graphite oxide was prepared by a modified Hummer's
method. Graphite (3.0 g) and NaNO3 (2.25 g) were added
1
7
sharply.
In addition, most of the currently reported
28
methods require the use of high amounts of base (with 2–20
mole ratio of HMF) and high oxygen pressure. Acknowledg-
ing these important achievements, the quest for milder and
greener methodologies for selective oxidation of HMF into
FDCA still remains a great challenge.
into a 1000 mL beaker in an ice bath. Then, 98.0 wt.% H
225 mL) was added into the beaker and vigorously stirred
for 30 min. Then, KMnO (13.5 g) was added into the mixture
slowly over 2 h. After the addition of KMnO , the mixture was
4
stirred for 2 h at 0 °C and subsequently at 25 °C for 5 days.
After that, the temperature was raised to 98 °C, and 5 wt.%
2 4
SO
(
4
In recent years, graphene-based materials such as
graphene and graphene oxide (GO) have shown important
applications in various fields such as electrochemistry,
2 4
H SO (450 mL) was added dropwise over 1 h. Then, the mix-
ture was stirred at 98 °C for 2 h. The temperature was then
reduced to 25 °C, and the mixture was stirred for another 2
h. Finally, the solid product was collected by centrifugation
and washed 15 times with 3 wt.% H SO , followed by wash-
2 4
ing 5 times with 3 wt.% HCl. Finally, the solid product was
2
2
electronics, biochemistry, and hydrogen storage. Graphene-
based materials have also received extensive attention as a
promising support to stabilize nanoparticles, including
metals and metal oxides, by strong π-interactions between
2
3
the nanoparticles and the supporting carbon nanosheets.
dried in a vacuum oven at 40 °C to obtain graphite oxide.
Among them, graphene oxide–Fe
formed by the decoration of Fe
3
O
O
4
nanoparticles, which are
on the graphene oxide,
3
4
Synthesis of C–Fe O –Pd composite
3
4
have particularly been attractive as an excellent carbonaceous
support in catalyst and separation fields due to their facile
30 mg of graphite oxide was added to 10 mL of water, and
2
4
the mixture was subjected to sonication for 30 min in order
to homogeneously disperse graphite oxide into water. 20 mL
of ethylene glycol was then added to the above mixture,
followed by sonication for additional 30 min. Then,
recovery using an external magnet.
Recently, the use of palladium nanoparticles has attracted
strong interest owing to their high catalytic activity compared
2
5
to the bulk phase. In order to prevent the aggregation of
palladium nanoparticles, several kinds of supports have been
used to immobilize the palladium nanoparticles. However,
the recyclability of supported palladium nanoparticles suffers
from the need for cost- and time-intensive filtration methods.
One effective method to solve this problem is the application
of magnetic supports, which allow the recovery of Pd nano-
particles by simple decantation in the presence of an external
3 2
FeCl ·6H O (0.1 g) and NaOAc (1.2 g) were added to a mixed
solvent of water (5 mL) and ethylene glycol (10 mL); then the
mixture was stirred at room temperature for 1 h to obtain a
clear solution. The as-prepared clear solution was added
dropwise to the graphite oxide solution, and then the mixture
was stirred for 30 min. Then, Na PdCl (4 mg) dissolved in 5
2
4
mL of N,N-dimethylformamide was added dropwise, and the
mixture was stirred for another 30 min. Finally, the mixture
was transferred into an autoclave and kept still at 130 °C for
2
6,27
magnet.
In an attempt to develop a green and sustainable
method for the oxidation of HMF into FDCA under mild con-
ditions, we used the abovementioned graphene–magnetite
nanocomposite as a supporting material for the immobiliza-
tion of palladium nanoparticles, which can act as a magneti-
cally separable catalyst for the oxidation of HMF into FDCA.
14 h. After cooling to room temperature, the catalyst was col-
lected by an external magnet, washed with water and ethanol,
and dried in a vacuum oven at 40 °C. Finally, 37 mg of
the black catalyst was obtained, which was abbreviated as
3 4
C–Fe O –Pd.
Experimental section
Catalyst characterization
Materials
Transmission electron microscope (TEM) images were
obtained using an FEI Tecnai G -20 instrument. The sample
2
Ethylene glycol (99.5%), FeCl
3
·6H
2
O (99.5%), sodium acetate
(
(
NaOAc, 99.5%), NaOH (99.5%), NaNO (99.5%) and KMnO
powder was firstly dispersed in ethanol and dropped onto
copper grids for observation. X-ray powder diffraction (XRD)
patterns of samples were determined with a Bruker advanced
D8 powder diffractometer (Cu Kα). All XRD patterns were col-
lected in the 2θ range of 10–80° with a scanning rate of
0.016° s . X-ray photoelectron spectroscopy (XPS) was
conducted on a Thermo VG scientific ESCA MultiLab-2000
spectrometer with a monochromatized Al Kα source (1486.6 eV)
3
4
99.5%) were purchased from Sinopharm Chemical Reagent
SO (98.0%)
were purchased from Kaifeng Chemical Reagent Co., Ltd.
Kaifeng, China). Graphite powder and sodium
tetrachloropalladateIJII) IJNa PdCl , 98%) were purchased from
Co., Ltd. (Shanghai, China). HCl (36.5%) and H
2
4
(
−
1
2
4
Aladdin Chemicals Co. Ltd. (Beijing, China). HMF (98%) was
purchased from Beijing Chemicals Co. Ltd. (Beijing, China).
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Catalysis Science & Technology
Paper
at constant analyzer pass energy of 25 eV. The binding energy
was estimated to be accurate within 0.2 eV. All binding
energies (BEs) were corrected referencing the C 1s (284.6 eV)
peak of the contamination carbon as an internal standard.
FT-IR measurements were recorded on a Nicolet NEXUS-6700
FDCA yield = moles of FDCA/moles of starting HMF × 100%
Catalyst recycling
3 4
After the reaction, the C–Fe O –Pd catalyst was collected
−
1
FTIR spectrometer with a spectral resolution of 4 cm in the
using a permanent magnet, and the liquid solution was
decanted. Then, the spent catalyst was washed with water
three times and ethanol twice. Finally, it was dried under vac-
uum at 50 °C overnight. The spent catalyst was reused for the
next run, and the other steps remained the same.
−
1
wave number range of 500–4000 cm . The Pd content in the
C–Fe O –Pd catalyst was quantitatively determined by induc-
3
4
tively coupled atomic emission spectrometer (ICP-AES) on an
IRIS Intrepid II XSP instrument (Thermo Electron Corpora-
tion). Magnetization measurement was performed by using a
physical property measurement system (PPMS-9T) with VSM
option from Quantum Design. Applied magnetic fields H
between −30 and 30 kOe and temperature 300 K were used in
the experiments.
Results and discussion
Catalyst preparation and characterization
Scheme 2 shows the procedure for the preparation of the
C–Fe O –Pd catalyst. Firstly, graphite oxide was prepared by
3
4
the oxidation of pristine graphite by a modified Hummer's
Determination of Pd loading by ICP-AES
2
+
3+
method. Both Pd and Fe ions were then simultaneously
The palladium content in the C–Fe
3
O
4
–Pd catalyst was quan-
29
anchored on the surface of graphene oxide by π-bonding or
titatively determined by inductively coupled atomic emission
spectrometer (ICP-AES) on an IRIS Intrepid II XSP instrument
ionic interaction between the oxy functional groups of
graphene oxide and the corresponding positive charge of the
(Thermo Electron Corporation). The acid digestion solution
30
metal ions. Then, the metal cations immobilized on graphene
was composed of concentrated HNO (7.5 mL), 48 wt.% HF
3
oxide were solvothermally reduced to graphene–Fe O –Pd
3
31
4
(1.5 mL) and 2.5 mL concentrated HCl (2.5 mL). 0.5 g sample
nanocomposite by ethylene glycol at 130 °C for 14 h. The
C–Fe O –Pd catalyst was readily purified using an external
was digested in the above solution on the flat electric fur-
nace. After digestion, the sample was diluted with deionized
water and centrifuged at 3000 rpm. The supernatant liquid
was subjected to ICP-AES analysis. The ICP-AES operation
conditions were as follows: incident power: 1.1 KW, carrier
3
4
magnet. The Pd content was determined to be 1.95 wt.% by
ICP-AES analysis, which meant 1 g of the catalyst contained
19.5 mg of Pd.
The heterostructure of the C–Fe O –Pd catalyst can be ver-
3
4
−
1
−1
gas flow rate: 0.8 L min , auxiliary gas flow rate: 0.4 L min ,
ified by the morphological analyses. Fig. 1 shows the typical
−
1
coolant gas flow rate: 16 L min , observation height: 10 mm.
TEM images of graphite oxide and the C–Fe O –Pd catalyst.
3
4
In Fig. 1(a), the folding nature of graphite oxide is clearly visi-
ble. On the other hand, graphene oxide has a layer structure,
General procedure for the aerobic oxidation of HMF
32
which is believed to be composed of many individual sheets.
As shown in Fig. 1(b), the surface of graphene oxide is deco-
rated by the palladium and Fe nanoparticles, and almost
Typically, HMF (50.4 mg, 0.4 mmol) was firstly dissolved in
water (8 mL); then K
40 mg, Pd/HMF mol ratio = 1.8%) were added to the above
reaction solution. Then oxygen was flushed at a rate of 20 mL
2 3 3 4
CO (27.6 mg, 0.2 mmol) and C–Fe O –Pd
3 4
O
(
no particles are scattered out of the supports, which indicates
that there was a strong interaction between nanoparticles
−1
min , and the reaction was carried out at 80 °C with stirring
at a constant rate of 600 revolutions per minute (rpm) using
a condenser. After the reaction, the catalyst was removed,
and the reaction solution was diluted with water to a certain
volume in each case.
2
9
and the support. The morphologies of those nanoparticles
Analytic methods
Furan compounds were analyzed by a ProStar 210 HPLC sys-
tem equipped with a UV detector. Furan compounds were
well separated by a reversed-phase C18 column (200 × 4.6
mm) at the wavelength of 280 nm. Acetonitrile and 0.1 wt.%
acetic acid aqueous solution with a volume ratio of 30 : 70
−
1
were used as mobile phase at a flow rate of 1.0 mL min .
The amounts of HMF and FDCA in samples were obtained
directly by interpolation from calibration curves. HMF con-
version and FDCA yield are defined as follows:
HMF conversion = moles of HMF/moles of starting HMF × 100%
Scheme 2 Schematic illustration of the catalyst preparation.
This journal is © The Royal Society of Chemistry 2015
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
3
4,35
crystalline graphite (JCPDS no. 41–1487).
As shown in
Fig. 2(b), after chemical oxidation of graphite by KMnO , the
4
strong peak at 2θ = 26.4° disappeared in the spectrum of the
C–Fe O –Pd catalyst, revealing the successful oxidation of the
3
4
3
0
starting graphite. However, no obvious diffraction peaks of
graphite oxide were observed in the XRD patterns of the
C–Fe O –Pd catalyst. It is reported that when the regular
3
4
stacks of graphite oxide were destroyed, for example, by exfo-
3
6
liation, their diffraction peaks disappeared. The diffraction
peaks at 2θ = 30.1°, 35.4°, 43.1°, 56.9°, and 63.2° were assigned
to the reflections of Fe O , which matched well with the stan-
3 4
Fig. 1 TEM images of graphite oxide (a) and C–Fe O –Pd (b).
3
4
37
3 4
dards of Fe O (JCPDS 65–3107). In addition, the peak at 2θ =
are almost regular in nature, and most of them are quite
spherical, with an average diameter of 10 nm. In addition, as
40.1° is also clearly observed, which is attributed to the inter-
3
3
planar spacing for the (111) plane of Pd(0) nanoparticles.
From the XRD results, it is confirmed that Fe and Pd(0)
the density and color of Fe
3
O
4
are close to those of Pd nano-
3 4
O
particles, it was difficult to discriminate Fe O and Pd nano-
nanoparticles were successfully immobilized on the surface
of the graphene oxide.
3
4
particles in the C–Fe O –Pd catalyst. The C–Fe O –Pd catalyst
3
4
3 4
was well dispersed in water, similar to that reported by Yu
et al., which is beneficial to the contact between the sub-
FT-IR is one of the most important methods to study the
oxidation of graphite to graphite oxide. Fig. 3 shows the
3
3
strate and the catalyst sites for chemical reactions in water.
3 4
FT-IR spectrum of the C–Fe O –Pd catalyst. As shown in Fig. 3,
3 4
XRD patterns of the graphite and the C–Fe O –Pd catalyst
several functional groups appeared after the chemical oxida-
−
1
are shown in Fig. 2. A strong XRD peak at 2θ = 26.4° was
observed in graphite (Fig. 2(a)), which is the characteristic
XRD peak of graphite and can be assigned to hexagonal
tion of graphite, such as –OH (3396 cm ), –COOH (1725
−
1
−1
−1
cm ), C–O (1058 cm ) and CO (1622 cm ). These func-
tional groups clearly indicate that the graphite was success-
fully oxidized by KMnO . The functional groups in graphene
4
oxide have a strong tendency to interact with metal cations,
with its enhanced hydroxyl and carboxyl groups. In addition,
−
1
a peak at 582 cm was also observed, which is assigned to
3
4
Fe–O vibrations of Fe
3 4
O .
3 4
The XPS survey scan spectra of the C–Fe O –Pd catalyst
and Pd 3d region is shown in Fig. 4. These peaks, with a
binding energy of about 285, 531, and 711 eV, are attributed
to C 1s, O 1s, and Fe 2p, respectively (Fig. 4(a)). In addition,
a weak peak with the centre around 335 eV was also
observed, which was assigned to Pd 3d. In the case of Fe 2p,
two peaks of Fe 2p3/2 and Fe 2p1/2 are located at 711.9 and
7
25.3 eV, respectively, which are the characteristic XPS peaks
2
+
38
of Fe in Fe
3
O
4
.
In order to give clear information on Pd
valence, high-resolution XPS spectra of Pd 3d was also col-
lected (Fig. 4(b)). The binding energy of Pd 3d3/2 and Pd 3d5/2
3 4
Fig. 2 XRD diffraction patterns of (a) graphite and (b) the C–Fe O –Pd
catalyst.
3 4
Fig. 3 FT-IR spectra of the C–Fe O –Pd catalyst.
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Catalysis Science & Technology
Paper
3 4
Fig. 5 Hysteresis loops for the C–Fe O –Pd catalyst at 300 K.
catalyst is strong enough for magnetic separation by a perma-
nent magnet, as shown in Fig. 6.
Aerobic oxidation of HMF over C–Fe O –Pd catalyst in
3
4
various solvents
In previous reports, some methods used organic solvents for
4
0
the oxidation of alcohols using Pd-based catalysts, while
others used water as the solvent for the oxidation of alco-
4
1
hols. Those reports indicated that the solvent played a cru-
cial role in the chemical reactions. Indeed, different solvents
have different properties, such as polarity, dielectric constant,
steric hindrance, and pH, which affect the efficiency of chem-
4
2
ical reactions. In addition, it is reported that the supported
metal nanoparticles may be deactivated by the formed
organic products, such as carboxylic acid. Thus, a base is
usually required to quickly neutralize the produced carboxylic
4
3
acid and keep the catalyst active. Therefore, the oxidation
3 4
Fig. 4 XPS spectra of the samples: (a) survey scan of the C–Fe O –Pd
catalyst, (b) Pd 3d region.
of HMF over C–Fe O –Pd catalyst was initially carried out in
3
4
2 3
some common solvents using 0.5 equiv of K CO as additive.
As shown in Table 1, the reaction solvent greatly affected
both HMF conversion and product selectivity. Generally
speaking, HMF conversion and the total selectivity of the
furan products were higher in polar solvents than those
obtained in solvents with low polarity, such as toluene and
MIBK. One of the main reasons should be that the base
K CO and the product FDCA showed poor solubility in tolu-
were 335.2 eV and 340.4 eV, respectively. The two peaks are
the characteristic peaks of Pd(0), which is in agreement with
the previous report, suggesting that the absorbed PdIJII) in
the graphene oxide was successfully reduced to Pd(0) nano-
particles under solvothermal reduction. Taking the XPS
results together with XRD information into consideration,
3
9
2
3
ene and MIBK. Thus, the formed FDCA could not be quickly
3 4
Fe O and Pd(0) were successfully decorated on the surface of
graphene oxide.
A magnetic catalyst should possess sufficient paramag-
netic properties for practical applications. Therefore, vibrat-
ing sample magnetometer (VSM) analysis was used to test
3 4
the magnetic property of the C–Fe O –Pd catalyst, and its
magnetic curve is shown in Fig. 5. The isothermal magnetiza-
tion curve of the C–Fe O –Pd catalyst at 300 K with the field
3
4
sweeping from −30 000 to +30 000 Oe displayed a rapid
increase with increasing applied magnetic field due to super-
paramagnetic relaxation. The saturation magnetization was
Fig.
6
Simple separation of the catalyst by
a
magnet: (a) after
−
1
2
3 4
8.75 emu g . Therefore, the magnetization of C–Fe O –Pd
reaction, (b) after separation by an external magnet.
This journal is © The Royal Society of Chemistry 2015
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
Table 1 Results of the aerobic oxidation of HMF in different solventsa
d
Entry
Solvent
HMF con. (%)
FDCA yield (%)
DFF yield (%)
HMFCA yield (%)
Total sel. of furans (%)
1
2
3
Toluene
MIBK
DMSO
DMSO
Ethanol
9.2
21.9
50.6
52.7
61.2
87.2
12.1
2.0
16.2
0
45.8
50.4
82.3
11.1
3.5
2.3
30.0
1.4
3.8
1.2
0.9
0.4
14.1
0.9
1.2
0.9
75.0
86.3
87.2
91.3
90.5
96.8
99.2
b
4
5
6
7
H
H
2
O
O
c
2
0.5
0.4
a
Reaction conditions: HMF (50.4 mg, 0.4 mmol), solvent (8 mL), C–Fe
3
O
4
–Pd (40 mg, 1.8 mol.%), K
2
CO
3
(27.6 mg, 0.2 mmol), oxygen flow rate
−1
b
c
(30 mL min ), 80 °C, 4 h. The reaction temperature was 100 °C. Otherwise, reaction conditions were the same as described above. The
d
reaction was carried out without K
HMFCA to the consumed HMF.
2
CO
3
.
Total selectivity of furans is defined as the mole ratio of furan products including FDCA, DFF and
a
2 3
Table 2 Results of HMF oxidation with different amounts of K CO
2 3
neutralized by K CO , but is absorbed on the catalyst surface,
leading to the low catalytic activity. It is interesting to note
that different products were produced in DMSO at different
reaction temperatures (Table 1, entries 3 vs. 4). DFF yield of
Mole ratio of
HMF
FDCA
FDCA
Entry
K
2
CO to HMF conversion (%) yield (%) selectivity (%)
3
1
2
3
4
5
6
0.5
1.0
1.5
0.25
0.125
0
87.2
98.7
100
45.4
23.8
12.1
82.3
87.6
62.7
40.1
21.9
11.1
94.4
88.8
62.7
88.4
92.0
91.7
3
8
0.0% and HMFCA yield of 10.4% were obtained in DMSO at
0 °C (Table 1, entry 3). However, FDCA was the major prod-
uct, with a yield of 45.8% in DMSO when the reaction tem-
perature was increased to 100 °C (Table 1, entry 4). Seeing
the results in entries 3 & 4, DFF and HMFCA should be the
intermediates during the oxidation of HMF into FDCA at
a
Reaction conditions: HMF (50.4 mg, 0.4 mmol), H
2
O (8 mL),
C–Fe
3
O
4
–Pd (40 mg, 1.8 mol.%), a set amount of K
2
CO
3
, oxygen flow
1
00 °C in DMSO. Among all of the organic solvents tested,
−1
rate (30 mL min ), 80 °C, 4 h.
C–Fe –Pd showed the best catalytic performance in the
3 4
O
protic solvent ethanol, with HMF conversion of 61.2% and
FDCA yield of 50.4% (Table 1, entry 5). To our pleasure, water
proved to the best solvent for the oxidation of HMF into
equiv of K
equiv of K
2
CO
CO
3
was a little lower than that obtained with 0.5
(Table 2, entries 1 vs. 2). It should be noted
FDCA over the C–Fe
HMF conversion of 87.2% and FDCA yield of 82.3% were
achieved in water in the presence of 0.5 equiv of K CO . The
3
O
4
–Pd catalyst (Table 1, entry 6). High
2
3
that K CO also played a negative effect on this reaction, as
2
3
2
3
HMF is not stable under acidic and alkaline conditions. For
example, Rass et al. found that the treatment of HMF in
Na CO (2 equiv) aqueous solution at 100 °C yielded 50%
oxidation of HMF using water, a green solvent, and molecular
oxygen as the oxidant appears very appealing due to its
low cost and nontoxicity. A control experiment was also
conducted with the oxidation of HMF in water without
K CO . As expected, the C–Fe O –Pd catalyst showed poor
2
3
4
4
HMF degradation after 2 h. This trend was much more
apparent with the further increase of K CO amount to 1.5
equiv. Full HMF conversion was obtained with FDCA yield of
62.7% by the use of 1.5 equiv of K CO (Table 2, entry 3). As
0.5 equiv of K CO is the stoichiometric dosage to neutralize
the resultant FDCA in theory, the final reaction solution with
1 and 1.5 equiv of K CO was alkaline. When the amount of
CO was below 0.5 equiv, HMF conversion and FDCA yield
gradually decreased with the decrease of K CO (Table 2,
2
3
2
3
3 4
catalytic performance in water without K
2
CO
3
(Table 1,
2
3
entries 6 vs. 7), confirming the crucial role of the base in
obtaining excellent catalytic activity of the C–Fe O –Pd cata-
2
3
3
4
lyst for the aerobic oxidation of HMF into FDCA in water. It
is also interesting to note that each solvent gave not only dif-
ferent product distributions, but also different total furan
selectivity. The lowest total selectivity was observed in tolu-
ene, and the highest total selectivity was obtained in water.
2
3
K
2
3
2
3
entries 4–6). For example, HMF conversions were 45.4% and
23.8% using 0.25 and 0.125 equiv of K CO , respectively, and
the corresponding FDCA yields were 40.1% and 21.9%,
respectively. The reason was that K CO was not sufficient to
neutralize the FDCA product, resulting in the loss of catalytic
activity during the reaction process.
2
3
2
3
Effect of base amount on the aerobic oxidation of HMF into
FDCA in water
As discussed above, K CO plays a crucial role in the aerobic
2
3
3 4
oxidation of HMF over C–Fe O –Pd catalyst; therefore, the
Effect of reaction temperature on the aerobic
oxidation of HMF
effect of base amount on this reaction were studied. Com-
pared with the results of using 0.5 equiv of K CO (Table 2,
2
3
entry 1), HMF conversion and FDCA yield increased to 98.7%
and 87.6%, respectively, when 1 equiv of K CO was used
The effect of reaction temperature was also studied for the
oxidation of HMF over C–Fe O –Pd catalyst. Experiments
2
3
3 4
(Table 2, entry 2). It is calculated that FDCA selectivity with 1
were carried out at four different reaction temperatures in
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Catalysis Science & Technology
Paper
the range of 298 K to 373 K. As shown in Table 3, the aerobic
oxidation of HMF is sensitive to the reaction temperature.
HMF conversion and FDCA yield increased with an increase
of reaction temperature from 298 K to 353 K (Table 3, entries
It is noted that increasing the catalyst dosage to 40 mg
resulted in a sharp increase in HMF conversion (87.2%) and
FDCA yield (82.3%) (Table 4, entry 2). The increase in HMF
conversion with an increase of catalyst dosage at the same
reaction time could be attributed to an increase in the avail-
ability and number of catalytically active sites. With further
increase of the catalyst amount to 60 mg, HMF conversion
continually increased from 87.2% to 98.2%, and the corre-
sponding FDCA yield was 91.8% (Table 4, entry 3). It is worth
noting that the selectivity of FDCA was almost the same in
each case, suggesting that it had no direct relation with the
catalyst amount. As discussed above, the main side reaction
was caused by the base-promoted degradation of HMF, which
mainly depended on the base concentration and reaction
1
, 3 & 4). For example, HMF conversion of 41.3% and FDCA
yield of 38.3% were obtained at the reaction temperature of
5 °C (Table 3, entry 1). Increasing the reaction temperature
2
to 60 °C greatly enhanced the reaction efficiency, leading to
HMF conversion of 77.6% and FDCA yield of 71.1% (Table 3,
entry 3). Further increasing the reaction temperature to
8
8
0 °C, HMF conversion and FDCA yield still improved to
7.2% and 82.3%, respectively (Table 3, entry 4). It is interest-
ing to note that the selectivity of FDCA was almost the same,
at around 90%, indicating that the side reaction, especially
base-promoted degradation of HMF, was not serious in the
reaction temperature range from 298 K to 353 K. HMF con-
version further increased from 87.2% at 80 °C to 97.1% at
3 4
temperature. The high catalytic activity of C–Fe O –Pd
inspired us to carry out the reaction in air, without oxygen
flush. To our delight, good results were also observed for the
oxidation of HMF in air. HMF conversion of 47.7% and FDCA
yield of 43.9% were obtained after 4 h in air, which were
lower than those with oxygen flush (Table 4, entries 2 vs. 4).
The reason could be that the concentration of oxygen in the
reaction solution in air was lower than that with oxygen
flush. After prolonging the reaction time to 12 h, oxidation of
HMF in air also afforded an excellent FDCA yield of 85.7%
(Table 4, entry 5). The excellent results of the aerobic oxida-
tion of HMF in air make this method easily handled and eco-
nomic in practical application. Control experiments were also
carried out using graphene oxide–Fe O as the catalyst, which
1
00 °C (Table 3, entries 4 & 5). However, FDCA yield of 80.5%
at 100 °C was a little lower than that at 80 °C (Table 3, entries
vs. 5). The main reason should be that the base-promoted
4
degradation of HMF became more serious at the higher reac-
tion temperature. For instance, Rass et al. reported that up to
5
1
0% HMF was degraded into other byproducts after 2 h at
00 °C in Na CO (2 equiv) aqueous solution. It is worth
2 3
4
4
noting that high HMF conversion of 95.4% and FDCA yield
of 87.9% was achieved at room temperature (25 °C) after
prolonging the reaction time from 4 h to 12 h (Table 3, entry 2).
To the best of our knowledge, there is no report on the
oxidation of HMF into FDCA under such mild reaction condi-
tions (e.g. room temperature, atmospheric oxygen pressure,
low dosage of base).
3
4
was prepared by the same method as the C–Fe
without the addition of Na PdCl during the preparative pro-
cedure. No FDCA was determined in the oxidation of HMF
over graphene oxide–Fe catalyst (Table 4, entry 6), which
3 4
O –Pd catalyst
2
4
3 4
O
indicated that Pd nanoparticles were the active sites for the
oxidation of HMF into FDCA. The conversion of HMF in entry
Aerobic oxidation of HMF with different catalyst amounts
6
was caused by the base-promoted degradation of HMF into
Experiments were also carried out with different amounts of
the catalyst to investigate the effect of catalyst loading.
Table 4 shows the results of HMF conversion, FDCA yield
and selectivity. Generally, the larger the catalyst dosage was,
the higher the HMF conversion and FDCA yield were (Table 4,
entries 1–3). HMF conversion of 57.8% and FDCA yield of
other byproducts.
Catalyst recycling experiments and large-scale reaction
Facile recycling and high stability are important characteris-
tics of magnetic catalysts. Therefore, experiments on catalyst
recycling were also studied. As shown in Fig. 6, after reaction,
the C–Fe O –Pd catalyst was easily collected from the reac-
5
2.9% were obtained after 4 h using 20 mg of C–Fe
3 4
O –Pd
(Table 4, entry 1).
3
4
tion mixture by a permanent magnet. Then, the liquid solu-
tion was decanted, and the spent catalyst was washed with
water three times and ethanol twice. Finally, the spent cata-
lyst was dried under vacuum at 50 °C overnight. The spent
catalyst was used for a second run under the same reaction
conditions as described for the first run. As shown in Fig. 7,
FDCA yields were almost the same, at about 91%. The results
Table 3 The results of HMF oxidation at different reaction temperaturesa
HMF
FDCA
FDCA
Entry Temperature (°C) conversion (%) yield (%) selectivity (%)
1
2
3
4
5
25
25
60
80
100
41.3
95.4
77.6
87.2
97.1
38.3
87.9
71.1
82.3
80.5
92.7
92.1
91.6
94.4
82.9
b
3 4
indicate that the C–Fe O –Pd catalyst was stable during the
reaction process without the loss of its catalytic activity. In
addition, the stability of the catalyst was also confirmed by
ICP-AES. No palladium was determined in the reaction solu-
tion, which indicated that there was no leach of palladium
from the catalyst to the reaction solution. The stability of the
a
Reaction conditions: HMF (50.4 mg, 0.4 mmol), H
2
O (8 mL),
C–Fe
3
O
4
–Pd (40 mg, 1.8 mol%), K CO (27.6 mg, 0.2 mmol), oxygen
2
3
−1
b
flow rate (30 mL min ), 4 h. The reaction time was 12 h under
otherwise the same reaction conditions as described above.
This journal is © The Royal Society of Chemistry 2015
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
Table 4 The results of HMF oxidation with different catalyst loadinga
Entry
Time (h)
Catalyst amount
HMF conversion (%)
FDCA yield (%)
FDCA selectivity (%)
1
2
3
4
4
4
4
12
4
20 mg, 0.9 mmol%
40 mg, 1.8 mol%
60 mg, 2.7 mol%
40 mg, 1.8 mol%
40 mg, 1.8 mol%
60 mg, 2.7 mol%
57.8
87.2
98.2
47.7
94.4
13.0
52.9
82.3
91.8
43.9
85.7
0
91.5
94.4
93.5
92.0
90.8
0
b
4
b
5
c
6
a
Reaction conditions: HMF (50.4 mg, 0.4 mmol), a certain amount of the C–Fe
3
O
4
–Pd catalyst, water (8 mL), K
2
CO
3
(27.6 mg, 0.2 mmol),
−1
b
oxygen flow rate (30 mL min ), 80 °C, 4 h. The reaction was carried out in air. Otherwise, reaction conditions were the same as above.
c
3 4
Graphene oxide–Fe O was used as the catalyst for the same reaction.
and it showed high catalytic activity in the aerobic oxidation
of HMF into FDCA in water under mild reaction conditions.
Results demonstrated that reaction temperature and base
concentration greatly affected not only the oxidation of HMF,
but also its degradation. Under optimal reaction conditions,
high HMF conversion of 98.2% and FDCA yield of 91.8%
2 3
were obtained after 4 h at 80 °C with a K CO /HMF molar
ratio of 0.5. Excellent results were also achieved by the oxida-
tion of HMF over C–Fe O –Pd catalyst in air and at room
3
4
temperature with an appropriate reaction time. More impor-
tantly, the catalyst was easily recovered by an external magnet
and reused without the loss of its catalytic activity. Compared
with other reported methods, the present catalytic system
showed three distinct advantages: the use of stoichiometric
amount of base, high activity under atmospheric oxygen pres-
sure, and facile catalyst recycling. It is believed that this find-
ing provides an efficient, green and sustainable method for
the production of other valuable chemicals by the oxidation
of various biomass-derived hydroxyl compounds.
Fig. 7 Recycling experiments of the C–Fe
conditions: HMF (50.4 mg, 0.4 mmol), H O (8 mL), K
mmol), oxygen flow rate (30 mL min ), C–Fe –Pd (60 mg, 2.7 mol%),
0 °C, 4 h.
3
O
4
–Pd catalyst. Reaction
2
2
CO (27.6 mg, 0.2
3
−1
3 4
O
8
prepared C–Fe O –Pd catalyst was the same as that reported
3
4
5
4
by Villa et al.
They used active carbon-supported, bi-
metallic nanoparticles IJAu Pd /AC) for the oxidation of HMF
into FDCA and found that it was stable during the recycling
experiment.
8
2
Acknowledgements
This project was supported by the National Natural Science
Foundation of China (no. 21203252 & no. 21206200), the
Chenguang Youth Science and Technology Project of Wuhan
City (no. 2014070404010212), and the Natural Science Foun-
dation of Hubei Province (no. 2014CFB180).
Finally, the large-scale oxidation of HMF into FDCA was
also carried out. 1 g of HMF was used as the starting mate-
rial. As the reaction required neutralization of the formed
FDCA, 0.5 equiv of Na CO (0.55 g) was used. 1 g of HMF and
2
3
0
.55 g of Na CO were added into 100 mL of water, and
2 3
then the reaction was carried out at 80 °C by the use of 0.5 g
−
1
of C–Fe O –Pd with oxygen flow rate at 30 mL min . Thin-
3
4
References
layer chromatography was used to track the reaction process.
After 12 h reaction, HMF completely disappeared. Then, the
catalyst was removed, and the reaction solution was concen-
trated by reducing evaporation. Then, the resulting residues
were purified on silica gel, eluting with methanol/ethyl ace-
tate. The amount of isolated FDCA was 1.05 g, equaling a
yield of 84.8%. Compared with the above microreaction, the
large-scale reaction yield was a little lower.
1
F. Rizzi, N. J. van Eck and M. Frey, Renewable Energy,
014, 62, 657–671.
2
2
A. J. Ragauskas, C. K. Williams, B. H. Davison, G. Britovsek,
J. Cairney, C. A. Eckert, W. J. Frederick, J. P. Hallett, D. J.
Leak, C. L. Liotta, J. R. Mielenz, R. Murphy, R. Templer and
T. Tschaplinski, Science, 2006, 311, 484–489.
3
4
Y. Xiong, Z. H. Zhang, X. Wang, B. Liu and J. T. Lin, Chem.
Eng. J., 2014, 235, 349–355.
S. Pickard, S. S. Daood, M. Pourkashanian and W. Nimmo,
Fuel, 2014, 134, 171–179.
Conclusion
In this study, the superparamagnetic C–Fe O –Pd catalyst
was successfully prepared by a one-pot solvothermal route,
5 I. Agirrezabal-Telleria, I. Gandarias and P. L. Arias, Catal.
Today, 2014, 234, 42–58.
3
4
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Catalysis Science & Technology
Paper
6
7
8
9
Z. H. Zhang, Q. Wang, H. B. Xie, W. J. Liu and Z. K. Zhao,
ChemSusChem, 2011, 4, 131–138.
L. Hu, G. Zhao, X. Tang, Z. Wu, J. X. Xu, L. Lin and S. J. Liu,
Bioresour. Technol., 2013, 148, 501–507.
26 R. Linhardt, Q. M. Kainz, R. N. Grass, W. J. Stark and O.
Reiser, RSC Adv., 2014, 4, 8541–8549.
27 Q. M. Kainz, R. Linhardt, R. N. Grass, G. Vile, J.
Perez-Ramirez, W. J. Stark and O. Reiser, Adv. Funct. Mater.,
2014, 24, 2020–2027.
T. F. Wang, M. W. Nolte and B. H. Shanks, Green Chem.,
2
014, 16, 548–572.
28 Y. Xu, H. Bai, G. Lu, C. Li and G. Shi, J. Am. Chem. Soc.,
2008, 130, 5856–5857.
S. H. Xiao, B. Liu, Y. M. Wang, Z. F. Fang and Z. H. Zhang,
Bioresour. Technol., 2014, 151, 361–366.
0 R.-J. van Putten, J. C. van der Waal, E. de Jong, C. B.
29 K. S. Kim, P. Tarakeshwar and J. Y. Lee, Chem. Rev.,
2000, 100, 4145–4186.
1
Rasrendra, H. J. Heeres and J. G. de Vries, Chem. Rev.,
30 K. Jasuja and V. Berry, ACS Nano, 2009, 3, 2358–2366.
31 C. Xu, X. Wang and J. Zhu, J. Phys. Chem. C, 2008, 112,
19841–19845.
32 V. Georgakilas, D. Gournisb, V. Tzitziosa, L. Pasquato, D. M.
Guldie and M. Prato, J. Mater. Chem., 2007, 26, 2679–2694.
33 H. P. Cong, J. J. He, Y. Lu and S. H. Yu, Small, 2010, 6,
169–173.
34 Y. M. Li, L. H. Tang and J. H. Li, Electrochem. Commun.,
2009, 11, 846–849.
35 C. Xu, X. D. Wu, J. W. Zhu and X. Wang, Carbon, 2008, 46,
386–389.
2
013, 113, 1499–1597.
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
1 B. Liu, Z. H. Zhang, K. L. Lv, K. J. Deng and H. M. Duan,
Appl. Catal., A, 2014, 472, 64–71.
2 G. S. Yi, S. P. Teong, X. K. Li and Y. G. Zhang,
ChemSusChem, 2014, 7, 2131–2135.
3 Z. H. Zhang, Z. L. Yuan, D. G. Tang, Y. S. Ren, K. L. Lv and
B. Liu, ChemSusChem, 2014, 7, 3496–3504.
4 C. H. R. M. Wilsens, B. A. J. Noordover and S. Rastogi,
Polymer, 2014, 55, 2432–2439.
5 B. Saha, S. Dutta and M. M. Abu-Omar, Catal. Sci. Technol.,
2
012, 2, 79–81.
36 Joint Committee on Powder Diffraction Standards,
Diffraction Data File: JCPDS International Center for
Diffraction Data: Swarthmore, PA, 1991.
37 F. Y. Zhao, B. L. Zhang and L. Y. Feng, Mater. Lett., 2012, 68,
112–114.
38 X. Zhang, Y. G. Niu, X. D. Meng, Y. Li and J. P. Zhao,
CrystEngComm, 2013, 15, 8166–8172.
39 Z. C. Ma, H. Q. Yang, Y. Qin, Y. J. Hao and G. Li, J. Mol.
Catal. A: Chem., 2010, 331, 78–85.
40 A. S. Guram, X. H. Bei and H. W. Turner, Org. Lett., 2003, 5,
2485–2487.
41 B. Karimi, H. Behzadnia, M. Bostina and H. Vali, Chem. –
Eur. J., 2012, 18, 8634–8640.
42 R. R. Sever and T. W. Root, J. Phys. Chem. B, 2003, 107,
4080–4089.
43 N. Dimitratos, A. Villa, D. Wang, F. Porta, D. Su and L. Prati,
J. Catal., 2006, 244, 113–121.
6 S. E. Davis, L. R. Houk, E. C. Tamargo, A. K. Datye and R. J.
Davis, Catal. Today, 2011, 160, 55–60.
7 O. Casanova and A. Corma, ChemSusChem, 2009, 2,
1
138–1144.
8 N. Guppta, S. Nishimura, A. Takagaki and K. Ebitani, Green
Chem., 2011, 13, 824–827.
9 S. Davis, B. Zope and R. Davis, Green Chem., 2012, 14,
1
43–147.
0 J. Cai, H. Ma, J. Zhang, Q. Song, Z. Du, Y. Huang and J. Xu,
Chem. – Eur. J., 2013, 19, 14215–14223.
1 A. Villa, M. Schiavoni, S. Campisi, G. M. Veith and L. Prati,
ChemSusChem, 2013, 6, 609–612.
2 S. Navalon, A. Dhakshinamoorthy, M. Alvaro and H. Garcia,
Chem. Rev., 2014, 114, 6179–6212.
3 R. Hao, W. Qian, L. Zhang and Y. Hou, Chem. Commun.,
2
008, 6576–6578.
4 J. Huang, Q. Chang, Y. B. Ding, X. Y. Han and H. Q. Tang,
Chem. Eng. J., 2014, 254, 434–442.
5 Z. H. Zhang, J. D. Zhen, B. Liu, K. L. Lv and K. J. Deng,
Green Chem., 2015, 17, 1308–1317.
44 H. A. Rass, N. Essayem and M. Besson, Green Chem.,
2013, 15, 2240–2251.
45 A. Villa, M. Schiavoni, S. Campisi, G. M. Veith and L. Prati,
ChemSusChem, 2013, 6, 609–612.
This journal is © The Royal Society of Chemistry 2015
Catal. Sci. Technol.