Aerobic oxidation of 5-hydroxymethylfurfural into 2,5-furandicarboxylic acid in water under mild conditions

Article abstract of DOI:10.1039/c4gc02019g

In this study, Pd nanoparticles were immobilized on the core-shell structure C@Fe₃O₄ (carbon is the shell and Fe₃O₄ is the core) magnetic microspheres via in situ adsorption and reduction to generate the magnetically separable Pd/C@Fe₃O₄ catalyst. In this method, no excess reductant and capping reagents were required, and it is a clean, simple and green process for the preparation of the magnetic Pd nanocatalyst. The Pd/C@Fe₃O₄ catalyst showed high activity and extraordinary stability during the oxidation of biomass derived 5-hydroxymethylfurfural (HMF) into 2,5-furandicarboxylic acid (FDCA) under mild conditions. A study aimed at optimizing the reaction conditions such as reaction temperature, reaction solvent and base amount has been performed. Under optimal reaction conditions, HMF conversion of 98.4% and FDCA yield of 86.7% were achieved after 6 h at 80 °C. After reaction, the Pd/C@Fe₃O₄ catalyst could be easily recovered by an external magnet and reused without the loss of its activity. This journal is

Full text of DOI:10.1039/c4gc02019g

Green Chemistry
View Article Online
View Journal
PAPER
Aerobic oxidation of 5-hydroxymethylfurfural
into 2,5-furandicarboxylic acid in water under
mild conditions
Cite this: DOI: 10.1039/c4gc02019g
Bing Liu,* Yongshen Ren and Zehui Zhang*
In this study, Pd nanoparticles were immobilized on the core–shell structure C@Fe3O4 (carbon is the shell
and Fe3O4 is the core) magnetic microspheres via in situ adsorption and reduction to generate the mag-
netically separable Pd/C@Fe3O4 catalyst. In this method, no excess reductant and capping reagents were
required, and it is a clean, simple and green process for the preparation of the magnetic Pd nanocatalyst.
The Pd/C@Fe3O4 catalyst showed high activity and extraordinary stability during the oxidation of biomass
derived 5-hydroxymethylfurfural (HMF) into 2,5-furandicarboxylic acid (FDCA) under mild conditions. A
study aimed at optimizing the reaction conditions such as reaction temperature, reaction solvent and
base amount has been performed. Under optimal reaction conditions, HMF conversion of 98.4% and
FDCA yield of 86.7% were achieved after 6 h at 80 °C. After reaction, the Pd/C@Fe3O4 catalyst could be
easily recovered by an external magnet and reused without the loss of its activity.
Received 16th October 2014,
Accepted 27th November 2014
DOI: 10.1039/c4gc02019g
www.rsc.org/greenchem
been used as a value added intermediate for the production of
biomass-derived chemicals.16
Introduction
Nowadays, there has been a growing interest in the search for
In early reports, synthesis of FDCA was performed by the
renewable resources to replace non-renewable fossil resources oxidation of HMF using stoichiometric oxidants such as
for the production of energy and chemicals. Biomass is the KMnO4.17 However, the cost was high and large amounts of
only carbon-containing renewable resource, and it is abundant waste were released into the environment. Later, the oxidation
in quantity as a substitute for fossil fuels.1,2 Currently, con- of HMF into FDCA was catalyzed by homogeneous metal salts
siderable effort has been devoted to the conversion of biomass (Co2+/Mn2+/Br−) under high pressure (70 bar air).18 However,
into useful chemicals and valuable fuels.3–5 5-Hydroxymethyl- the use of the homogeneous catalysts showed some drawbacks
furfural (HMF) can be readily generated from the dehydration such as the difficulty in catalyst recycling and the toxic pol-
of various carbohydrates such as fructose, glucose, and lution of the environment. Therefore, great effort has been
cellulose.6–9 HMF can be used as a promising precursor for devoted to design heterogeneous catalysts for the oxidation of
the synthesis of bulk chemicals, fine chemicals, polymers and HMF into FDCA. A number of previous publications reported
liquid fuels. Therefore, HMF has been identified as one of the that the oxidation of HMF to FDCA was mainly catalyzed by Pt,
key platform chemicals to bridge biomass and fossil Pd, and Au-based catalysts.19–21 For example, Rass et al.
resources.10,11 Oxidation of HMF can generate several impor- demonstrated that HMF was quantitatively oxidized to FDCA
tant furan chemicals such as 2,5-diformylfuran (DFF), 5-hydro- over C/Pt catalyst at 100 °C under 40 bar air with Na2CO3/HMF
xymethyl-2-furancarboxylic
acid
(HMFCA)
and
2,5- molar ratio = 2.21 Davis et al. reported that full HMF conver-
furandicarboxylic acid (FDCA).12–14 The structure of FDCA is sion was achieved by the use of Pd/C and Pt/C catalysts with
similar to that of terephthalic acid, which is obtained from FDCA yields between 71% and 79% under 690 kPa O2
fossil resources and used as a monomer in the synthesis of pressure.22 Compared with other metal nanoparticles, sup-
polymers. Thus, FDCA can be used as a potential biorenewable ported Au catalysts showed encouraging catalytic performance
monomer to replace terephthalic acid in the production of and received considerably more attention. For example, Corma
polyethylene terephthalate (PET).15 In addition, FDCA has also et al. reported that oxidation of HMF over Au/CeO2 catalyst
afforded nearly quantitative FDCA yield at 130 °C under 1 MPa
air pressure and high concentration of NaOH (mole ratio of
Department of Chemistry, Key Laboratory of Catalysis and Material Sciences of
NaOH/HMF = 4).23 However, the Au/CeO2 catalyst was not
Ministry of Education, South-Central University for Nationalities, MinYuan Road
stable, and its catalytic activity decreased rapidly in the second
182, Wuhan, P.R. China. E-mail: liubing@mail.scuec.edu.cn,
run. Davis et al. reported that full HMF conversion and FDCA
zehuizh@mail.ustc.edu.cn
This journal is © The Royal Society of Chemistry 2014
Green Chem.

View Article Online
Paper
Green Chemistry
yield of 65% were obtained by the use of a high NaOH/HMF solvothermal reaction of FeCl3·6H2O in ethylene glycol in the
mole ratio of 20 over Au/TiO2 under 345 kPa oxygen.24 Most of presence of polyvinylpyrrolidone and sodium acetate at 200 °C
the previously reported methods required harsh reaction con- for 8 h generated black Fe3O4 microspheres. The as-prepared
ditions such as high pressure, high temperature or high base Fe3O4 microspheres were homogeneously dispersed in an
concentration. Acknowledging these important achievements, aqueous solution of glucose by sonication, and then, the
it is still a challenge to develop new catalytic methods for the mixture was heated at 200 °C for 12 h to produce C@Fe3O4
oxidation of HMF to FDCA under comparatively mild reaction microspheres.
conditions.
Preparation of Pd/C@Fe3O4 catalyst
The growing demand for environmentally friendly chemical
processes has impelled many researchers to develop hetero- C@Fe3O4 microspheres (80 mg) were homogeneously dis-
geneous catalysts, which can be facilely recovered and reused. persed in ethylene glycol (50 mL) with an assist of ultrasonic
Recently, magnetically recyclable catalysts have proved to be wave for 30 minutes. Then, Na2PdCl4 (12 mg) in ethylene glycol
promising, by combining the advantages of high activity and (5 mL) was added dropwise. The mixture was further heated at
facile recovery.25 Among the various magnetic materials, core– 105 °C for another 3 h under vigorous stirring. After cooling to
shell structure C@Fe3O4 composites in particular, have room temperature, the catalyst was collected by an external
attracted great interest.26,27 In comparison to the common magnet and washed several times with deionized water to
silica shells, carbon shells exhibit superior properties such as remove the physically adsorbed ions. Finally, the purified cata-
considerably higher stability in acid or base media, as well as lyst was dried in a vacuum oven overnight. The palladium de-
at high temperatures and pressures.28 More importantly, large posited C@Fe3O4 microspheres are abbreviated as Pd/C@Fe3O4
amounts of functional groups, such as hydroxyl groups and in the following sections.
carboxylic groups, can be easily generated on the surface of
carbon, which could make a contribution to the adsorption of
Catalyst characterization
heavy metal ions and stabilize metal nanoparticles.29 For Nitrogen adsorption isotherms were determined using an
example, Makowski et al. reported that hydrophilic carbon Autosorb-1 (Quantachrome, USA) at 77 K. Prior to the measure-
sphere supported Pd nanoparticles showed high catalytic ment, all the samples were degassed at 200 °C for 6 h in a
activity in the hydrogenation of phenol to cyclohexanone.30
vacuum line. The BET surface area was determined by a multi-
Considering the evident difficulty in the recycling of the point BET method using the adsorption data in the relative
small size carbon particles, herein, the core–shell structure pressure (P/P0) range of 0.05–0.3. The mesoporous pore size
C@Fe3O4 microspheres were prepared by a hydrothermal distributions were derived from desorption branches of the
method and were used to immobilize Pd nanocatalysts isotherms using the BJH model. Transmission electron
(Pd/C@Fe3O4). The as-prepared Pd/C@Fe3O4 catalyst was used microscopy (TEM) images were obtained using an FEI Tecnai
for the aerobic oxidation of HMF into FDCA in water under G2-20 instrument. The sample powders were firstly dispersed
mild reaction conditions.
in ethanol and then dropped onto copper grids for obser-
vation. X-ray powder diffraction (XRD) patterns of samples
were determined with a Bruker advanced D8 powder diffract-
ometer (Cu Kα). All XRD patterns were collected in the 2θ
range of 10°–80° with a scanning rate of 0.016° s−1. X-ray
photoelectron spectroscopy (XPS) was conducted using a
Thermo VG scientific ESCA MultiLab-2000 spectrometer with a
monochromatized Al Kα source (1486.6 eV) 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 cali-
brated by C 1s (284.6 eV). FT-IR measurements were recorded
using a Nicolet NEXUS-6700 FTIR spectrometer with a spectral
Experimental section
Materials
FeCl3·6H2O (99.5%), polyvinylpyrrolidone (99.0%), ethylene
glycol (99.0%), sodium acetate (99.0%) and glucose (99.5%)
were purchased from Sinopharm Chemical Reagent Co., Ltd
(Shanghai, China). Acetonitrile (HPLC grade) was purchased
from Tedia Co. (Fairfield, USA). Sodium tetrachloropalladate(II)
(98.0%) (Na2PdCl4) was purchased from Aladdin Chemicals
Co. Ltd (Beijing, China). HMF (98%) was purchased from
Beijing Chemicals Co. Ltd (Beijing, China). DFF and FDCA
were purchased from the J&K Chemical Co. Ltd, (Beijing,
China). All the solvents were purchased from Sinopharm
Chemical Reagent Co., Ltd (Shanghai, China). All the chemi-
cals were of analytical grade and used without further purifi-
cation. Ultrapure water was used for the catalyst preparation
and catalytic reactions.
resolution of
4
cm−1 in the wave number range of
500–4000 cm−1. The Pd content in the Fe3O4@C-Pd catalyst was
quantitatively determined by inductively coupled atomic emis-
sion spectrometry (ICP-AES) using an IRIS Intrepid II XSP
instrument (Thermo Electron Corporation).
General procedure for the oxidation of HMF over Fe3O4@C-Pd
catalyst
Typically, HMF (50.4 mg, 0.4 mmol) and K2CO3 (0.5 equiv.,
27.6 mg) were firstly dissolved in water (8 mL). Then,
Preparation of core–shell structure C@Fe3O4 composites
Core–shell structure C@Fe3O4 composites were prepared Fe3O4@C-Pd (40 mg) was added and dispersed in the above-
according to a known method with slight modification.31 The mentioned solution by sonication. After that, oxygen was
Green Chem.
This journal is © The Royal Society of Chemistry 2014

View Article Online
Green Chemistry
Paper
allowed to flow at a rate of 20 mL min−1 into the bottom of the complex.32,33 The immobilized Pd2+ was solvothermally
reactor, and the reaction mixture was stirred at 80 °C at a con- reduced to Pd nanoparticles by ethylene glycol34 and stabilized
stant rate of 600 revolutions per minute (rpm) for a certain by oxygen-containing groups on the surface of C@Fe3O4 micro-
period of time. Time zero was recorded when the oxygen was spheres. The Pd content was determined to be 3.85 wt% by
allowed to flow into the reaction mixture. The reaction was ICP-AEC. It is worth noting that the Pd/C@Fe3O4 catalyst can
monitored and analyzed by characterizing the reaction mixture be well dispersed in water, which facilitates the substrate to
using the HPLC method.
contact the active sites of the catalyst adequately for chemical
reactions in water. The texture of the prepared catalyst was
determined by the N2 adsorption–desorption isotherms.
The BET surface area and pore size were determined to be
18 m2 g−1 and 0.6 m3 g−1, respectively.
Analytical methods
Product analysis was performed using a ProStar 210 HPLC
system. Furan compounds were separated by a reversed-phase
C18 column (200 × 4.6 mm) and detected by a UV detector at a
wavelength of 280 nm. Acetonitrile and 0.1 wt% acetic acid
aqueous solution with the volume ratio of 30 : 70 was used as
the mobile phase, and the flow rate was 1.0 mL min−1. The
content of the furan compounds in the samples was obtained
directly by interpolation from calibration curves.
The morphologies of C@Fe3O4 and Pd/C@Fe3O4 were
characterized by TEM technology. Fig. 1(a) presents the typical
TEM images of the as-synthesized C@Fe3O4 composite. The
obtained C@Fe3O4 microspheres exhibited a relatively uniform
size of 600 nm. However, some C@Fe3O4 microspheres with
small diameters were also observed. As shown in Fig. 3(a),
core–shell structure C@Fe3O4 microspheres were clearly
observed after the hydrothermal reaction of glucose in the
presence of Fe3O4 microparticles. The dark Fe3O4 micro-
spheres were coated with a grey layer of carbon with the shell
thickness of about 20 nm. The TEM image of a typical
Pd/C@Fe3O4 microsphere is presented in Fig. 1(b). Due to the
presence of numerous functional groups such as carbonyl and
hydroxyl moieties on the surface of the C@Fe3O4 micro-
spheres, Pd nanoparticles were successfully anchored to the
carbon shell of the C@Fe3O4 microspheres. As shown in
Fig. 1(b), Pd nanoparticles were not uniform in size, and some
aggregation of Pd nanoparticles was also observed. As far as
for the Pd nanoparticles size, it can be estimated from the full
width at half maximum of a peak in the diffraction pattern of
the nanoparticle according to Scherrer’s equation (as shown in
the following), and the average crystallite size of the Pd nano-
particles was determined to be 10 nm.
HMF conversion and FDCA yield are defined as follows:
HMF conversion ¼ moles of HMF=moles of starting HMF
ꢀ 100%
FDCA yield ¼ moles of FDCA=moles of starting HMF ꢀ 100%
Recycling of the catalyst
After the reaction, the Pd/C@Fe3O4 catalyst was collected using
a permanent magnet and washed several times with deionized
water and three times with ethanol. The spent catalyst was
then dried under vacuum at 50 °C overnight, and then, it was
reused for the next cycle under the same reaction conditions.
Results and discussion
Catalyst preparation and characterization
In order to provide more information about the surface
functional groups of the prepared Pd/C@Fe3O4 catalyst, it was
further characterized by FT-IR technology. As shown Fig. 2, a
strong absorption peak at 586 cm−1 was assigned to the
characteristic absorption band of Fe–O,35 suggesting Fe3O4
was present in the catalyst. The band observed at 1640 cm−1
was assigned to the stretching vibration of CvC.36 In addition,
these peaks of 1400, 1450 and 1500 cm−1 were the character-
Scheme 1 illustrates the procedure for the synthesis of the
magnetic Pd/C@Fe3O4 catalyst. Fe3O4 microspheres were pre-
pared by the solvothermal method using FeCl3 as the precur-
sor in ethylene glycol. The core–shell structure C@Fe3O4
microspheres were prepared by the in situ carbonization of
glucose on the surface of the Fe3O4 microspheres. Under
hydrothermal conditions, the surface of the carbon layer was
enriched with multiple functional groups such as hydroxyl and
carboxylic acid groups. Thus, C@Fe3O4 microspheres were
used to anchor Pd2+ by the interaction between the oxygen-
containing groups with Pd2+ to form the ligand-substituted
Scheme 1 Schematic illustration of the synthesis of the Pd/C@Fe3O4
Fig. 1 TEM images of the C@Fe3O4 microspheres (a) and the Pd/
catalyst.
C@Fe3O4 catalyst (b).
This journal is © The Royal Society of Chemistry 2014
Green Chem.

View Article Online
Paper
Green Chemistry
at 2θ = 40°, 47° and 68° were the characteristic diffraction
peaks of the face centered cubic crystalline phase of palladium
nanoparticles. The intensity of the peak at 2θ = 40° was con-
siderably stronger than the other two peaks.38 As shown in
Fig. 3, a peak at 2θ = 40.1° was observed, which was attributed
to the interplanar-spacing (111) of Pd (0). The disappearance
of the peaks at 2θ = 47° and 68° was mainly due to the lower
weight percentage as compared with Fe3O4.
XPS is a powerful tool for studying the external surface of
catalysts. XPS is able to determine the chemical composition
and oxidation degrees of components present on the catalyst
surface. Therefore, XPS technology was employed to investigate
the surface chemical composition and valence state of the
element Pd. Fig. 4(a) shows the survey scan spectrum of the
Pd/C@Fe3O4 catalyst. These peaks with binding energies of
Fig. 2 FTIR spectra of the Pd/C@Fe3O4 catalyst.
istic vibrations of the aromatic ring, which supported an aro-
matization of glucose during the hydrothermal treatment.36
The absorption peak at 1700 cm−1 was assigned to the stretch-
ing vibration of CvO and it verified the presence of the carb-
oxylic group (–COOH).36 The band at 3400 cm−1 was attributed
to the stretching vibration of the O–H group. The band at
1040 cm−1 was assigned to the C–OH stretching vibration.
Using FT-IR analysis it was found out that after the hydrother-
mal reaction there were large numbers of hydrophilic func-
tional groups (mainly including –COOH, and –C-OH) on the
surface of C@Fe3O4, which facilitated the adsorption of Pd2+
onto the C@Fe3O4 surface and then immobilized the Pd nano-
particles. The surface atomic ratio of oxygen and Pd in the
catalyst was determined to be 15.8 by XPS. This result clearly
indicates that the surface oxygen groups were enough to
anchor the Pd nanoparticles.
XRD pattern of the Pd/C@Fe3O4 catalyst is shown in Fig. 3.
Diffraction peaks at 2θ = 30.1°, 35.5°, 43.1°, 53.6°, 57.0°, and
63.3° were assigned to (220), (311), (400), (422), (511) and (440)
Bragg reflections, respectively, of the face-centered cubic
lattice of Fe3O4, which matched well with the standards of
Fe3O4 (JCPDS 65-3107).37 For the Pd nanoparticles, three peaks
Fig. 4 XPS spectra of the Fe3O4@C-Pd microspheres. (a) Survey scan of
Pd/C@Fe3O4; (b) Pd 3d region in the fresh catalyst; (c) Pd 3d region in
the spent catalyst.
Fig. 3 XRD diffraction patterns of the Pd/C@Fe3O4 microspheres.
Green Chem.
This journal is © The Royal Society of Chemistry 2014

View Article Online
Green Chemistry
Paper
about 285, 531, and 338 eV were attributed to C 1s, O 1s, and illustrated in Fig. 5, the isothermal magnetization curve
Pd 3d, respectively, which confirmed that the elements C, O showed a rapid increase with the increase in the applied mag-
and Pd were present on the surface of the catalyst. Moreover, netic field. The saturated magnetization value of the catalyst
abundant C and O species indicated the presence of versatile was determined to be 42.7 emu g−1. The superparamagnetic
oxygen-containing functional groups on the surface after the property of the Fe3O4@C-Pd catalyst not only benefited their
hydrothermal reaction of glucose. The confirmation of Fe3O4 dispersibility and redispersibility in the solution, but also
by XPS technology was usually by the observation of the guaranteed the quick response of these nanoparticles to exter-
characteristic Fe 2p3/2 and Fe 2p1/2 of Fe2+, which had high nal magnetic fields, which could be promising in the practical
binding energies at 711.9 and 725.3 eV, respectively.39 As applications.
shown in Fig. 4, the two peaks were not observed in the XPS
survey spectrum, suggesting that Fe3O4 was successfully coated
with a carbon shell. High resolution XPS spectrum of Pd 3d is
shown in Fig. 4(b). Two peaks with binding energies of 335.8
and 341.2 eV were observed in Fig. 4(b), which were assigned
to Pd 3d5/2 and Pd 3d3/2, respectively. The two peaks were the
characteristic peaks of metallic Pd, similar to those seen in
Pd/C or Pd powder.40
The valence state of the Pd nanoparticles in the spent cata-
lyst was also analyzed by XPS technology. As shown in Fig. 4(c),
there was no difference in the Pd 3d between the fresh catalyst
and the spent catalyst. After the reaction, Pd was still present
in the metallic state; therefore, our prepared catalyst showed a
high stability during the catalyst recycling experiments. Our
results were consistent with previous results, in which the sup-
ported Pd nanoparticles were stable for the oxidation of alco-
hols.41 Because Pd nanoparticles are exposed to oxygen during
the oxidation reactions, they have the risk to be oxidized to
high valence state Pd2+. Even though some Pd(0) was oxidized
into Pd2+, the Pd2+ could be reduced by solvent or substrate
(alcohol) to Pd(0). For example, Zhang et al. recently reported
that the organic ligands stabilized Pd2+ showed high catalytic
activity for the oxidation of alcohol with molecular oxygen and
they found that some Pd2+ was reduced to metallic Pd(0) after
reaction.42
Catalytic aerobic oxidation of HMF in various solvents
To evaluate the catalytic activity of the as-prepraed
Pd/C@Fe3O4 catalyst, the aerobic oxidation of HMF was used
as a model reaction. It is known that various solvents have
different properties, such as polarity, dielectric constant, steric
hindrance, acid–base properties, which would affect chemical
reactions.43 Therefore, experiments were initially carried out in
various solvents to screen the best reaction medium. In
addition, 0.5 equiv. of K2CO3 were also used as an additive, as
the oxidation of alcohol over Pd based catalysts generally
requires the use of a base. As shown in Table 1, HMF conver-
sion was low in toluene and acetonitrile. (Table 1, entries 1–2).
One of the main reasons was that K2CO3 was insoluble in
toluene and acetonitrile, which blocked the catalytic acitivity
of Pd/C@Fe3O4. DFF was obtained in a relatively high selecti-
vity without the formation of FDCA, indicating that other side
reactions, such as the degradation of HMF, were not evident.
Although HMF conversion was slightly improved to 15.5% in
dimethyl sulfoxide (Table 1, entry 3), DFF was the only product
with a low selectivity of 27.1%. Interestingly, it should be
noted that FDCA was the major product when reactions were
carried out in the protic solvents (Table 1, entries 4 & 5).
However, the catalytic activity of Pd/C@Fe3O4 in ethanol and
H2O was quite different. Low HMF conversion of 19.1% and
FDCA yield of 16.1% were obtained in ethanol (Table 1, entry
4), whereas the reaction in water produced high HMF conver-
sion of 86.5% and FDCA yield of 76.9% (Table 1, entry 5). It is
quite appealing to use H2O as the solvent for the synthesis of
important chemicals from biomass based chemicals due to its
low cost and the absence of toxic pollution in line with green
chemistry. In addition, a control experiment was also carried
Magnetic measurement of the Pd/C@Fe3O4 catalyst was
carried out by a vibrating sample magnetometer (VSM) and the
isothermal magnetization curve of the Pd/C@Fe3O4 catalyst is
shown in Fig. 5. The measurement was conducted at 300 K in
the applied magnetic field from −30 000 to +30 000 Oe. As
Table 1 The results of HMF aerobic oxidation in different solventsa
HMF
conversion
(%)
FDCA
yield
(%)
DFF
yield
(%)
Entry
Solvent
1
2
3
4
Toluene
5.9
9.7
15.5
19.1
86.5
20.6
0
0
0
16.1
76.9
11.9
4.7
8.6
4.2
1.2
3.2
7.1
Acetonitrile
Dimethyl sulfoxide
Ethanol
H2O
H2O
5
6b
a Reaction conditions: HMF (50.4 mg, 0.4 mmol), H2O (8 mL),
Pd/C@Fe3O4 (40 mg), K2CO3 (27.6 mg, 0.2 mmol), oxygen flow rate
(30 mL min−1), 80 °C, 4 h. b The reaction was carried out in the
absence of K2CO3.
Fig. 5 Hysteresis loops for the Pd/C@Fe3O4 catalyst at 300 K.
This journal is © The Royal Society of Chemistry 2014
Green Chem.

View Article Online
Paper
Green Chemistry
out in water without K2CO3. Compared with the results in the tralize the resultant FDCA in theory. These results clearly indi-
presence of K2CO3, HMF conversion and FDCA selectivity cate that K2CO3 indeed promoted the oxidation of HMF into
greatly decreased without K2CO3 (Table 1, entries 5 vs. 6). FDCA in water over the Pd/C@Fe3O4 catalyst. As reported by
These results clearly indicate that the base played a crucial others on the role of base in the oxidation of HMF over a sup-
role in the aerobic oxidation of HMF into FDCA over the ported Au nanocatalyst,23,24 the base was used to neutralize
Pd/C@Fe3O4 catalyst.
the produced FDCA in a timely manner. Thus, the use of base
One important reason for the use of base in the oxidation can avoid FDCA to adsorb onto the surface of the Pd nano-
of HMF to FDCA is that it was used to neutralize the dicarb- particles, keeping the catalytic activity of Pd/C@Fe3O4 stable.
oxylic acid (FDCA), as FDCA can be adsorbed onto the surface
When the mole ratio of K2CO3/HMF was further increased
of metal nanoparticles. The salt of FDCA was dissolved in the to 1, complete HMF conversion was obtained. However, FDCA
reaction solution and kept the catalyst activity stable. In fact, yield in this case was close to that with 0.5 equiv. of K2CO3. It
most of the reported methods required the use of excessive should be noted that the K2CO3 had a dual effect on this reac-
NaOH (2–20 times the molar amount of HMF). For example tion. As is known to us, HMF is stable under acidic or alkaline
Davis et al. reported that full HMF conversion and FDCA yield conditions due to the co-presence of the aldehyde and
of 65% were obtained by the use of a high NaOH/HMF mole hydroxyl groups in its structure. For instance, Rass et al. also
ratio of 20.24 In our developed method, we only required the found that the treatment of Na2CO3 (2 equiv.) with aqueous
stoichiometric amount of K2CO3 to neutralize FDCA, which solution of HMF at 100 °C resulted in HMF degradation up to
greatly decreased the amount of waste released into the 50% after 2 h.44 Therefore, the use of too much K2CO3 could
envrioment.
also accelerate the degradation of HMF, in addition to the pro-
motion of the oxidation reaction, and the K2CO3/HMF mole
ratio of 0.5 was the appropriate amount (the stoichiometric
amount to neutralize the FDCA).
Effect of base amount on the aeroboic oxidation of HMF
In order to give more insight into the role of K2CO3 in the oxi-
dation of HMF into FDCA over Pd/C@Fe3O4 catalyst, expri-
ments were carried out in water at 80 °C with different
amounts of the catalyst. As shown in Fig. 6, HMF conversion
and FDCA yield simultaneously increased with the increase in
the base loading. HMF conversion greatly improved from
34.9% with the K2CO3/HMF mole ratio of 0.125, to 63.1% with
K2CO3/HMF mole ratio of 0.25. Furthermore, by increasing the
K2CO3/HMF mole ratio to 0.5, HMF conversion further
increased to 86.5%. The change trend in FDCA yield was
similar to that of HMF conversion. In addition to the major
product FDCA, DFF was also detected as one of the oxidation
products. It is interesting to note that the ratio of FDCA and
DFF (also FDCA selectivity) gradually increased with the
increase in the base amount. 0.5 equiv. of K2CO3 (K2CO3/HMF
mole ratio = 0.5) is the stoichiometric amount required to neu-
Effect of reaction temperature on the oxidation of HMF
With the aim to optimize the reaction conditions to get the
highest FDCA yield, oxidation of HMF was also carried out at
different reaction temperatures in the range from 298 K to
373 K, and the results are show in Table 2. It is worth noting
that the aerobic oxidation of HMF was sensitive to the reaction
temperature. The higher the reaction temperature was, the
higher the HMF conversion was. Low HMF conversion of
28.6% was obtained at 25 °C. HMF conversion was greatly
enhanced from 28.6% to 60.1% by the increase of the reaction
temperature from 25 to 50 °C. Further inceasing the reaction
temperature to 80 °C, still greatly increased HMF conversion to
86.5%. Complete HMF conversion was achieved at a 100 °C.
Significantly more molecules were in contact with catalyst
active sites at higher reaction temperatures. Therefore, HMF
conversion increased with an increase of reaction temperature.
In addition to HMF conversion, FDCA selectivity was also
greatly affected by the reaction temperaure. FDCA and DFF
Table 2 The results of HMF oxidation at different reaction
temperaturesa
Reaction
temperature
(°C)
HMF
conversion
(%)
FDCA
yield
(%)
DFF
yield
(%)
FDCA
selectivity
(%)
Entry
1
2
25
50
80
80
100
28.6
60.1
86.5
98.4
100
11.9
36.6
76.9
86.7
83.6
15.1
18.9
3.2
1.3
0
41.6
60.9
88.9
88.1
83.6
3
4b
5
a Reaction conditions: HMF (50.4 mg, 0.4 mmol), H2O (8 mL), K2CO3
(27.6 mg, 0.2 mmol). Pd/C@Fe3O4 (40 mg), oxygen flow rate (30 mL
min−1), 4 h. b The reaction time was 6 h and other reaction conditions
were same as above.
Fig. 6 The results of HMF oxidation with different amounts of K2CO3.
Reaction conditions: HMF (50.4 mg, 0.4 mmol), H2O (8 mL),
Pd/C@Fe3O4 (40 mg), oxygen flow rate (30 mL min−1), 80 °C, 4 h.
Green Chem.
This journal is © The Royal Society of Chemistry 2014

View Article Online
Green Chemistry
Paper
were both formed during the oxidation of HMF. It was found after the fifth run and calculated that the total weight loss per-
that the ratio of FDCA to DFF (FDCA selectivty) increased with centage of Pd was about 2.3% after the fifth run.
the increase of reaction temperature from 298 K to 353 K
Reaction conditions: HMF (50.4 mg, 0.4 mmol), H2O
(Table 2, entries 1–3). These results indicated that high reac- (8 mL), K2CO3 (27.6 mg, 0.2 mmol). Pd/C@Fe3O4 (40 mg),
tion temperatures not only promoted the oxidation of HMF oxygen flow rate (30 mL min−1), 80 °C, 6 h.
into DFF, but also benefited the following oxidation of DFF
into FDCA. However, the selectivity of FDCA at 100 °C was a
little lower than that obtained at 80 °C (Table 2, entries 3 vs. 5).
The highest FDCA yield was obtained in 86.7% after 6 h at the
Conclusion
reaction temperature of 80 °C (Table 2, entry 4). Although In summary, the Pd/C@Fe3O4 catalyst was successfully pre-
HMF conversion at 100 °C was a little higher than that at pared by the deposition of Pd nanoparticles on the core–shell
80 °C, FDCA yield at 100 °C was a little lower than that at structure C@Fe3O4 composites and characterized by TEM,
80 °C. The possible reason could be that the degradation of XPS, FT-IR, XRD technologies. The Pd/C@Fe3O4 catalyst
HMF was considerarably more significant at the higher reac- showed high catalytic activity in the aerobic oxidation of HMF
tion temperature of 100 °C. As we mentioned above, HMF into FDCA in water under mild reaction conditions. It was
molecules contain three functional groups: the hydroxyl found that the base K2CO3 was crucial for the aerobic oxi-
group, aldehyde group, and furan ring. HMF was not stable in dation of HMF into FDCA. In addition, the base dosage, the
acidic or alkaline solution and was degraded into humins. reaction temperature and the solvent also showed a great effect
Vuyyuru et al.45 also found that HMF was degraded into on HMF conversion and FDCA yield. Under optimal reaction
humins (dark brown color) and the degradation became more conditions, HMF conversion of 100% and FDCA yield of 87.8%
significant with the increase of the reaction temperature.
were obtained after 6 h at 80 °C. Moreover, the Pd/C@Fe3O4
catalyst could be easily separated from the reaction mixture by
an external magnet. Compared with other reported methods,
our catalytic system shows some distinct advantages: (a) the
use of the Pd/C@Fe3O4 catalyst did not require a large amount
of base; (b) this method could be conducted under atmo-
spheric oxygen pressure, not requiring high oxygen pressure;
(c) the catalyst could be easily separated by an external magnet
and reused without the loss of catalytic activity. Therefore, the
developed method shows promising potential for the practical
production of FDCA from the renewable biomass-based plat-
form chemical.
Catalyst recycling experiments
Recyclability and stability are two important characteristics for
the use of heterogeneous catalysts. Therefore, the recycling
experiments of the Pd/C@Fe3O4 catalyst were studied. The oxi-
dation of HMF at 80 °C with the use of 0.5 equiv. of K2CO3 was
used as the model reaction. After reaction, the Fe3O4@C-Pd
catalyst was easily collected by an external magnet, washed
successively with water and ethanol, followed by drying under
vacuum at 50 °C. The magnetic recycling of the catalyst can
avoid weight loss of the catalyst, which usually occurs in fil-
tration processes. As shown in Fig. 7, FDCA yields were almost
stable after five consecutive cycles. However, there was still a
very small decrease in the FDCA yield. FDCA yield in the first
run was 86.7%, and that in the fifth run was 83.7%. In
addition, HMF conversion was also observed to show a very
small decrease from 98.4% in the first run to 94.6% in the
fifth run. We determined the Pd content in the spent catalyst
Acknowledgements
The Project was supported by the National Natural Science
Foundation of China (no. 21203252 & 21206200), and the
Chenguang Youth Science and Technology Project of Wuhan
City (no. 2014070404010212).
References
1 P. Gallezot, Chem. Soc. Rev., 2012, 41, 1538.
2 J. C. Serrano-Ruiz, R. Luque and A. Sepulveda-Escribano,
Chem. Soc. Rev., 2011, 40, 5266.
3 M. E. Zakrzewska, E. Bogel-Łukasik and R. Bogel-Łukasik,
Chem. Rev., 2010, 111, 397.
4 S. Kumar, E. Hablot, J. L. G. Moscoso, W. Obeid,
P. G. Hatcher, B. M. DuQuette, D. Graiver, R. Narayan and
V. Balan, J. Mater. Sci., 2014, 49, 7824.
5 P. Kangas, I. Hannula, P. Koukkari and M. Hupa, Fuel,
2014, 129, 86.
6 S. P. Simeonov, J. A. S. Coelho and C. A. M. Afonso, Chem-
SusChem, 2013, 6, 997.
Fig. 7 Recycle experiments of the Fe3O4@C-Pd catalyst.
This journal is © The Royal Society of Chemistry 2014
Green Chem.

View Article Online
Paper
Green Chemistry
7 Z. F. Fang, B. Liu, J. J. Luo, Y. S. Ren and Z. H. Zhang,
Biomass Bioenergy., 2014, 60, 171.
O. M. N. D. Teodoro and R. Zboril, ACS Sustainable Chem.
Eng., 2014, 2, 1699.
8 S. H. Xiao, B. Liu, Y. M. Wang, Z. F. Fang and Z. H. Zhang, 26 C. Lei, F. Han, Q. Sun, W. C. Li and A. H. Lu, Chem. – Eur.
Bioresour. Technol., 2014, 151, 361. J., 2014, 20, 139.
9 I. Agirrezabal-Telleria, I. Gandarias and P. L. Arias, Catal. 27 M. Y. Zhu, C. J. Wang, D. H. Meng and G. W. Diao, J. Mater.
Today, 2014, 234, 42. Chem. A, 2013, 1, 2118.
10 R. J. van Putten, J. C. van der Waal, E. de Jong, 28 N. Q. Zhao, S. Wu, C. N. He, Z. Y. Wang, C. S. Shi, E. Z. Liu
C. B. Rasrendra, H. J. Heeres and J. G. de Vries, Chem. Rev.,
2013, 113, 1499.
and J. J. Li, Carbon, 2013, 57, 130.
29 I. Do and L. T. Drzal, Carbon, 2014, 75, 43.
11 T. F. Wang, M. W. Nolte and B. H. Shanks, Green Chem., 30 P. Makowski, R. D. Cakan, M. Antonietti, F. Goettmann
2014, 16, 548. and M. M. Titirici, Chem. Commun., 2008, 999.
12 Z. H. Zhang, B. Liu, K. L. Lv, J. Sun and K. J. Deng, Green 31 H. G. Wang, L. Sun, Y. P. Li, X. L. Fei, M. D. Sun,
Chem., 2014, 16, 2762.
13 X. Y. Wan, C. M. Zhou, J. S. Chen, W. P. Deng, Q. H. Zhang,
Y. H. Yang and Y. Wang, ACS Catal., 2014, 7, 2175.
C. Q. Zhang, Y. X. Li and Q. B. Yang, Langmuir, 2011, 27,
11609.
32 K. Jasuja and V. Berry, ACS Nano, 2009, 3, 2358.
14 Z. H. Zhang, Z. L. Yuan, D. G. Tang, Y. S. Ren, K. L. Lv and 33 Y. H. Kim, M. N. Alam, Y. Marutani, T. Ogata, S. Morisasda
B. Liu, ChemSusChem, 2014, 7, 3496. and Y. Nakano, Chem. Lett., 2009, 10, 956.
15 B. S. Wu, Y. T. Xu, Z. Y. Bu, L. B. Wu, B. G. Li and 34 C. Xu, X. Wang and J. Zhu, J. Phys. Chem. C, 2008, 112,
P. Dubois, Polymer, 2014, 55, 3648. 19841.
16 A. J. J. E. Eerhart, A. P. C. Faaij and M. K. Patel, Energy 35 S. H. Xuan, L. Y. Hao, W. Q. Jiang, X. L. Gong, Y. Hu and
Environ. Sci., 2012, 5, 6407. Z. Y. Chen, J. Magn. Magn. Mater., 2007, 308, 210.
17 T. Miura, H. Kakinuma, T. Kawano and H. Matsuhisa, U.S 36 X. Sun and Y. D. Li, Angew. Chem., Int. Ed., 2004, 43, 597.
Patent, US-7411078, 2008. 37 F. Lin and R. Doong, J. Phys. Chem. C, 2011, 115, 6591.
18 W. Partenheimer and V. V. Grushin, Adv. Synth. Catal., 38 Y. M. Lu, H. Z. Zhu, W. G. Li, B. Hu and S. H. Yu, J. Mater.
2001, 343, 102. Chem. A, 2013, 1, 3783.
19 P. Vinke, W. Van der Poel and H. Van Bekkum, Stud. Surf. 39 G. Kataby, A. Ulman, M. Cojocarua and A. Gedanken,
Sci. Catal., 1991, 59, 385. J. Mater. Chem., 1999, 9, 1501.
20 M. Kroger, U. Prusse and K. D. Vorlop, Top. Catal., 2002, 40 K. McEleney, C. M. Crudden and J. H. Horton, J. Phys.
13, 237. Chem. C, 2009, 113, 1901.
21 H. A. Rass, N. Essayem and M. Besson, Green Chem., 2013, 41 E. V. Johnston, O. Verho, M. D. Kärkäs, M. Shakeri,
15, 2240.
C.-W. Tai, P. Palmgren, K. Eriksson, S. Oscarsson and
J. E. Bäckvall, Chem. – Eur. J., 2012, 18, 12202.
42 L. Zhang, P. H. Li, J. Yang, M. Wang and L. Wang, Chem-
PlusChem, 2014, 79, 217.
43 R. R. Sever and T. W. Root, J. Phys. Chem. B, 2003, 107, 4080.
44 H. A. Rass, N. Essayem and M. Besson, Green Chem., 2013,
15, 2240.
22 S. E. Davis, L. R. Houk, E. C. Tamargo, A. K. Datye and
R. J. Davis, Catal. Today, 2011, 160, 55.
23 O. Casanova, S. Iborra and A. Corma, ChemSusChem, 2009,
2, 1138.
24 S. Davis, B. Zope and R. Davis, Green Chem., 2012, 14, 143.
25 S. N. Shelke, S. R. Bankar, G. R. Mhaske, S. S. Kadam,
D. K. Murade, S. B. Bhorkade, A. K. Rathi, N. Bundaleski, 45 K. R. Vuyyuru and P. Strasser, Catal. Today, 2012, 195, 144.
Green Chem.
This journal is © The Royal Society of Chemistry 2014