Superior catalytic performance of Ce1-xBixO2-δ solid solution and Au/Ce1-xBixO2-δ for 5-hydroxymethylfurfural conversion in alkaline aqueous solution

Article abstract of DOI:10.1039/c4cy01060d

Porous Bi-doped ceria (Ce_1-xBi_xO_2-δ solid solution) was prepared by the easy citrate method and then used as a supporting material for Au nanoparticles (NPs) obtained by deposition-precipitation. In the presence of O₂, Ce_1-xBi_xO_2-δ (0.08 ≤ x ≤ 0.5) efficiently catalyzed the conversion of 5-hydroxymethylfurfural (HMF) to 5-hydroxymethyl-2-furancarboxylic acid (HFCA) and 2,5-bishydroxymethylfuran (BHMF) in alkaline aqueous solution without degradation of HMF. The excellent catalytic activity was attributed to the oxygen activation and hydride transfer enhanced by Bi doping and the large amount of oxygen vacancies. After Au NPs were supported on Ce_1-xBi_xO_2-δ (x ≤ 0.2), the presence of Au^δ+ facilitated the activation of the C-H bond in the hydroxymethyl group and then the production of 2,5-furandicarboxylic acid (FDCA) as an end product, inhibiting the generation of BHMF.

Full text of DOI:10.1039/c4cy01060d

Catalysis
Science &
Technology
View Article Online
View Journal
PAPER
Superior catalytic performance of Ce1−xBixO2−δ
solid solution and Au/Ce1−xBixO2−δ for
5-hydroxymethylfurfural conversion in alkaline
aqueous solution†
Cite this: DOI: 10.1039/c4cy01060d
ac
Zhenzhen Miao,ab Yibo Zhang,
Xiqiang Pan,d Tianxiao Wu,a Bin Zhang,a
*
Jingwei Li,ab Ting Yi,ab Zhendong Zhang and Xiangguang Yang
a
ac
*
Porous Bi-doped ceria IJCe1−xBixO2−δ solid solution) was prepared by the easy citrate method and then used
as a supporting material for Au nanoparticles (NPs) obtained by deposition–precipitation. In the presence
of O2, Ce1−xBixO2−δ (0.08 ≤ x ≤ 0.5) efficiently catalyzed the conversion of 5-hydroxymethylfurfural (HMF)
to 5-hydroxymethyl-2-furancarboxylic acid (HFCA) and 2,5-bishydroxymethylfuran (BHMF) in alkaline
aqueous solution without degradation of HMF. The excellent catalytic activity was attributed to the oxygen
activation and hydride transfer enhanced by Bi doping and the large amount of oxygen vacancies. After
Au NPs were supported on Ce1−xBixO2−δ (x ≤ 0.2), the presence of Auδ+ facilitated the activation of the
C–H bond in the hydroxymethyl group and then the production of 2,5-furandicarboxylic acid (FDCA) as an
end product, inhibiting the generation of BHMF.
Received 16th August 2014,
Accepted 13th November 2014
DOI: 10.1039/c4cy01060d
www.rsc.org/catalysis
could substitute for petro-based terephthalic, isophthalic, and
adipic acids.9
Introduction
With the continuous consumption of fossil resources such as
oil and natural gas, it is of great importance to exploit and
utilize renewable, abundant biomass resources for prepara-
tion of bulk and fine chemicals.1–3 5-Hydroxymethylfurfural
(HMF), which can be obtained from dehydration of glucose or
fructose,4,5 is a compelling platform chemical for the produc-
tion of 2,5-diformylfuran (DFF), 2,5-bishydroxymethylfuran
(BHMF), 5-hydroxymethyl-2-furancarboxylic acid (HFCA) and
2,5-furandicarboxylic acid (FDCA). DFF and BHMF have been
regarded as promising monomers for preparing resins, poly-
mers and fine chemicals.6,7 HFCA serves not only as a mono-
mer in various polyesters8 but also as a starting material for
synthesis of FDCA. The presence of two carboxylic groups in
FDCA makes it a potential polymer building block which
Many efforts have been made over the past two decades in
converting HMF to those more valuable derivatives.10–14
Vinke et al. first reported the oxidation of aqueous HMF to
FDCA in near-quantitative yield under basic reaction condi-
tions with a Pt/Al2O3 catalyst at 60 °C.10 Verdeguer et al.
demonstrated that high pH was needed to produce FDCA
over PtPb/C catalysts and hydroxide base was more effective
than carbonate base, which was applicable to the production
of HFCA.11 Davis et al. have shown that the activity of Au
catalysts for HMF oxidation was an order of magnitude
higher than that of either Pt or Pd catalysts, but the selectiv-
ity for FDCA was much lower than that over the Pt and Pd
catalysts.12
The oxidation of HMF comprises an aldehyde oxidation
and an alcohol oxidation (Scheme 1).15–18 Davis et al.17
recently elucidated the reaction mechanism that water, rather
than O2, was the source of oxygen atoms inserted into the
products by using isotopically labeled 18O2 and H218O. The
aldehyde side chain of HMF first underwent rapid reversible
hydration to a geminal diol intermediate in alkaline aqueous
a State Key Laboratory of Rare Earth Resource Utilization, Changchun Institute of
Applied Chemistry, Chinese Academy of Sciences, Changchun, Jilin, 130022, China.
E-mail: xgyang@ciac.ac.cn, yibozhang@ciac.ac.cn; Tel: 86 431 8526 2228
b University of Chinese Academy of Sciences, Beijing, 100039, China
c Laboratory of Green Chemistry and Process, Changchun Institute of Applied
Chemistry, Chinese Academy of Sciences, Changchun, Jilin, 130022, China
d The Northwest Research Institute of Chemistry Industry, Xi'an, Shaanxi,
710061, China
† Electronic supplementary information (ESI) available: Photogragh, N2
adsorption isotherms and pore size distributions of Ce1−xBixO2−δ samples, SEM,
1H NMR spectra, mass spectra, XRD, XPS and some control experiments. See
DOI: 10.1039/c4cy01060d
Scheme 1 Oxidation of HMF to FDCA.
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
solution. Then, the dehydrogenation of the geminal diol was
facilitated by the noble metal catalyst and hydroxide ions.
This process produced HFCA and deposited two electrons on
the catalyst. The following oxidation of the alcohol group to
form 5-formyl-2-furancarboxylic acid (FFCA) was the rate-
limiting step, as the activation of the C–H bond in the
hydroxymethyl group strongly depended on the abilities of
the catalyst, such as transferring hydride, depositing two
additional electrons and interacting with O2. O2 was essential
for these oxidation processes and played an indirect role via
removing the electrons deposited on the catalyst and being
reduced to peroxide and hydroxide ions. The aldehyde group
of FFCA underwent the same oxidation process in the conver-
sion of HMF to HFCA. Finally, FDCA was obtained.
Notably, although alkaline conditions are favorable for
the oxidation of HMF, other competitive reactions could
occur at the same time, such as degradation and Cannizzaro
reaction of HMF. The Cannizzaro reaction is base-induced
disproportionation reaction of an aldehyde lacking a hydro-
gen atom at an α-position to the carbonyl group. In alkaline
medium, one molecule of the aldehyde is hydrated to a gemi-
nal diol and acts as a hydride donor while the other func-
tions as an acceptor, resulting in a carboxylic acid salt and
an alcohol product, respectively. The hydride transfer was the
principal component of this redox process.19 The conversion
of HMF to HFCA and BHMF simultaneously via Cannizzaro
reaction is of great value (Scheme 2), circumventing the use
of an oxidant (for HFCA) or a reductor (for BHMF). Recently,
Subbiah et al.13 have investigated the reaction in NaOH
aqueous solution. High yields of HFCA and BHMF were
obtained when the reactions were conducted at 0 °C for 1 h
and then room temperature for certain hours. However, these
processes needed a quite long reaction time, such as 18 h,
without the presence of a catalyst. In addition, the yields
were very low at 60 °C even after reaction for 24 h, which was
due to serious degradation of HMF under basic conditions at
higher temperature.
centers (Lewis acid sites and stoichiometric oxidation sites of
CeO2) and Au+ species of Au/CeO2 could readily accept a
hydride from the C–H bond in alcohol or in the correspond-
ing alkoxide to form Ce–H and Au–H, with the formation of
a carbonyl compound at the same time. The oxygen vacancies
of ceria could activate O2 and form cerium-coordinated super-
oxide (Ce–OO˙) species which subsequently evolved to cerium
hydroperoxide by hydrogen abstraction from Au–H. The
cerium hydroperoxide then interacted with Ce–H, producing
H2O and recovering the Ce3+ centers. Au–H donated H and
changed back to the initial Au+ species. Without Au+, the
catalytic cycle could not be completed.
As discussed above, for ceria-based catalysts, increasing
the amount of oxygen vacancies would improve the catalytic
activity towards HMF conversion. It is favorable for O2 activa-
tion and hydride transfer from HMF (or the geminal diol
intermediate) to catalyst. The doping with low-value metal
cations to substitute CeIV in the CeO2 lattice would drive the
formation of a stable O-deficient bulk fluorite-type structure
to maintain charge conservation. Ce3+ centers would increase
with the creation of oxygen vacancies.26 Cubic δ-Bi2O3 has an
atomic configuration of C-type rare earth sesquioxide, which
is similar to CeO2. As an oxide ionic conductor, it could
release oxygen at low temperature, leading to Bi-doped CeO2
that has a higher low-temperature redox activity towards
CeO2.27,28 Glaeser et al.29 have also indicated that the lone
electron pair of Bi3+ was the active site for O2 activation, and
the oxide ion of Bi–O–M could help the hydrogen transfer
from reaction substrate to catalyst in selective oxidation.
Thus, the Bi-doped CeO2 may serve as an admirable catalyst
for Cannizzaro reaction of HMF or an excellent support of
Au NPs for HMF oxidation.
Herein, a series of Ce1−xBixO2−δ (x ≤ 0.5) were synthesized
by the easy citrate method and used as supports for gold NPs
obtained via a deposition–precipitation method. Ce1−xBixO2−δ
and Au/Ce1−xBixO2−δ were all examined in HMF conversion in
alkaline aqueous solution. Interestingly, in the presence of
O2, Ce1−xBixO2−δ (0.08 ≤ x ≤ 0.5) could catalyze HMF to trans-
form into HFCA and BHMF at 65 °C without degradation of
HMF. Au/Ce1−xBixO2−δ (0.08 ≤ x ≤ 0.2) catalysts showed
significantly improved catalytic activities towards HMF oxida-
tion to FDCA compared to Au/CeO2. To the best of our knowl-
edge, it is the first report about the Cannizzaro reaction of
HMF catalyzed by a mixed oxide of non-precious metals in
eco-friendly aqueous solution.
Nanoparticulate CeO2 with a high external-to-internal
atom ratio has a large number of surface oxygen vacancies
and defects, which could disperse and stabilize gold nano-
particles (NPs).20–24 To reach nearly 99% yield of FDCA, the
reaction time for Au NPs on nanoparticulate CeO2 was half of
that for the Au on non-nanometric CeO2.15 Reductive pre-
treatment of the Au/CeO2 was shown to efficiently increase
the catalytic activity because it increased the amount of
Ce3+ and oxygen vacancies. The increased Ce3+ and oxygen
vacancies have shown great importance in transferring
hydride and activating O2 during catalytic oxidation of alco-
hols in a former report.25 Corma et al.25 proposed that Ce3+
Experimental
Chemicals and characterization
All the reagents were used as received without further purifi-
cation. HMF (98%), HFCA, 5-formyl-2-furancarboxylic acid
(FFCA, 98%), and FDCA (97%) were obtained from J & K Co.
Ltd (Beijing, China). CeIJNO3)3Ĵ6H2O and BiIJNO3)3Ĵ5H2O were
purchased from Aladdin Chemicals Co. Ltd (Shanghai, China).
Concentrated HNO3, concentrated H2SO4, KOH, and citric acid
monohydrate were purchased from Beijing Chemicals Co. Ltd
Scheme 2 Cannizzaro reaction of HMF.
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
(Beijing, China). HAuCl4Ĵ4H2O was obtained from Shanghai
Chemical Reagent Co. Ltd (Shanghai, China).
was added and the pH value was readjusted to 10. The sus-
pension was kept under stirring for another 2 h, centrifuged
and washed four times with water, and dried at 40 °C for
2 days. The Au/Ce1−xBixO2−δ samples were calcined at 200 °C
for 2.5 h before use.
The powder X-ray diffraction (XRD) patterns were recorded
on a Bruker D8 Advance X-ray diffractometer using a Cu Kα
radiation source (λ = 1.5406 Å). Scanning electron microscopy
(SEM) was performed using a field emission Hitachi S-4800
instrument. Transmission electron microscopy (TEM) was
performed using a FEI Tecnai G2 S-Twin instrument with a
field emission gun operating at 200 kV. Raman spectra were
collected using a Renishaw 2000 model confocal microscopy
Raman spectrometer with a CCD detector and a holographic
notch filter (Renishaw Ltd., Gloucestershire, UK) under ambi-
ent conditions using radiation at 514.5 nm. The H2-TPR
experiment was conducted with a self-made chemisorption
instrument equipped with a thermal conductivity detector
(TCD).30 50 mg of fresh samples were loaded into a quartz
reactor at room temperature and reduced in a flowing 5%
H2/Ar mixed gas with a heating rate of 10 °C min−1. The efflu-
ent gas was analyzed with the TCD. The nitrogen adsorption–
desorption isotherms of the samples were measured with an
Autosorb-iQ-Station-1 instrument at 77 K after pretreatment
at 473 K for 10 h under high vacuum. The specific surface
areas were determined by the Brunauer–Emmett–Teller (BET)
equation. Elemental contents were measured by ICP-OES
(iCAP 6300 Thermo Scientific USA). X-ray photoelectron spectro-
HMF conversion in alkaline aqueous solution
The aqueous-phase catalytic conversion of HMF was carried
out in a 50 ml autoclave with a PTFE liner and equipped with
a magnetic stirrer. In a typical experiment, 6 ml of the reac-
tant solution (0.15 M HMF and 0.6 M NaOH) was added to
the reactor along with the appropriate amount of catalyst.
The reactor was purged 3 times with O2 and then pressurized
at 1.0 MPa. Timing was started once the reactor was placed
in a preheated oil bath pot under stirring. After reacting for a
certain time, the samples were centrifuged, diluted 50 times
with deionized H2O and acidified with sulfuric acid. The re-
sulting solutions were filtered by using a syringe filter (0.2 μm)
before analysis using a high-performance liquid chromato-
graph (HPLC) equipped with a Waters 2487 Dual λ absorbance
detector. The HPLC utilized a Bio-Rad Aminex HPX-87H
column at 35 °C and 5 mM H2SO4 flowing at 0.5 ml min−1
to perform the separation. Identification and calibration of
FDCA, HFCA, FFCA and HMF were performed by injecting
known concentrations.
scopy (XPS) spectra were obtained on
a VG Thermo
ESCALAB 250 spectrometer operated at 120 W. All the bind-
ing energies were calibrated using the Ce4+IJ3d104f0Vn) (u‴)
signal at 916.6 eV to remove surface charging effects.31
Liquid-state 1H NMR spectra were recorded at 298 K on a
Bruker Avance 300 NMR spectrometer.
Results and discussion
The as-prepared samples were proved to be porous material.
SEM images shown in Fig. S2† illustrate the presence of
macropores larger than 50 nm. The nitrogen physisorption
analysis results of the Ce1−xBixO2−δ (x = 0, 0.05, 0.1, 0.2, 0.5)
samples are summarized in Table 1. The adsorption iso-
therms and pore size distribution curves are presented in
Fig. S3.† The adsorption–desorption isotherms of all the
samples showed a type IV isotherm with a hysteresis loop,
indicating that these samples were mainly in the mesoporous
range of 2–50 nm. The pore size was centered at about
3.8 nm. In addition, the specific surface area of CeO2 was
calculated to be 117.90 m2 g−1 and the average pore volume
was 0.35 cm3 g−1 which both decreased upon increasing the
amount of Bi doping.
Catalyst preparation
A series of Ce1−xBixO2−δ mixed oxides with x = 0, 0.05, 0.08,
0.1, 0.125, 0.15, 0.2, and 0.5 were prepared by the citrate acid
method.32 In a typical synthesis, stoichiometric amounts of
BiIJNO3)3Ĵ5H2O and CeIJNO3)3Ĵ6H2O (Bi + Ce = 0.1 mol) were
mixed in an appropriate concentration of nitric acid solution
to obtain a clear solution. Citric acid monohydrate was then
added to the mixture (metal cation-to-ligand molar ratio was
1 : 2). The as-prepared solution was heated on a hot plate at
200 °C under stirring. The continuous heating process turned
this solution into a white (without Bi) or yellow gel, which
was afterwards combusted with flame and yielded highly
spongy and uneven color bulk particles. Finally, the precur-
sors were calcined at 500 °C for 2 h to obtain Ce1−xBixO2−δ
mixed oxides. The color of Bi-doped ceria samples was
yellow and deepened upon increasing the amount of Bi
doping (Fig. S1†).
The powder XRD patterns of Ce1−xBixO2−δ (x ≤ 0.2) sam-
ples showed in Fig. 1 demonstrate that only peaks attributed
Table 1 Textural properties of the typical Ce1−xBixO2−δ samples
Surface area Pore diametera
Average pore
Sample
(m2 g−1
)
(nm)
volumeb (cm3 g−1
)
CeO2
117.90
93.72
81.11
46.25
24.34
3.83
3.81
3.82
3.82
3.82
0.35
0.32
0.29
0.24
0.15
Gold NPs were deposited on Ce1−xBixO2−δ via a deposition–
precipitation method.33 15 ml of 8 mM HAuCl4 aqueous
solution was sampled into a round-bottomed flask, and
1.0 M KOH was then added dropwise to the stirred solution
until the pH value was adjusted to 10. The resulting solution
was heated at 80 °C in an oil bath. 2.0 g of Ce1−xBixO2−δ
Ce0.95Bi0.05O2−δ
Ce0.9Bi0.1O2−δ
Ce0.8Bi0.2O2−δ
Ce0.5Bi0.5O2−δ
a
Calculated by the BJH method from the desorption branch.
Estimated by single point adsorption at a relative pressure of 0.99.
b
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
process. It was obvious that the oxygen vacancy peak was
enhanced significantly after Bi doping, implying the increase
in the amount of oxygen vacancies. When the relative content
of Bi reached 20 mol%, a characteristic peak of Bi2O3 at
ca. 325 cm−1 appeared.37,38 The Bi2O3 phase could not be
detected by XRD analysis, suggesting that Bi2O3 species are
of low content and dispersed well in Ce0.8Bi0.2O2−δ
.
In order to figure out the low-temperature redox activity
of Ce1−xBixO2−δ, the reduction profiles of the samples were
taken and the results are shown in Fig. 3. The H2 reduction
maximum of Bi-doped samples occurred at 200–400 °C asso-
ciated with the reduction of Ce4+ and Bi3+. The temperature
was much lower than that of pure ceria, indicating the more
reducible properties of the Bi-doped samples. Higher oxygen
release capacity of the Bi-doped samples could also be found,
as the peak area increased with the amount of Bi doping.
The results also indicated that the amount of the stoichio-
metric oxidation sites showed a similar behavior.
Fig. 1 XRD patterns of (a) CeO2, (b) Ce0.95Bi0.05O2−δ, (c) Ce0.9Bi0.1O2−δ
(d) Ce0.8Bi0.2O2−δ and (e) Ce0.5Bi0.5O2−δ (inset: XRD of samples scanned
at low speed).
,
to a cubic fluorite structure of CeO2 (JCPDS 34-0394) were
observed and no other phase was detected. Compared with
the pure CeO2 sample, the diffraction peaks of the Bi-doped
samples slightly shifted to lower angles (inset of Fig. 1), indi-
cating the expansion of the unit cell of CeO2. This variation
could be intelligibly ascribed to Ce4+ with an ionic radius of
0.97 Å partially replaced with Bi3+ (1.17 Å) and the formation
of solid solution. For the Ce0.5Bi0.5O2−δ sample, in addition to
the low value-shifted peaks of CeO2, peaks located at around
28.0° and 46.2° appeared, which were ascribed to Bi2O3
(JCPDS 65-1209).
Raman spectra of Ce1−xBixO2−δ (x = 0, 0.05, 0.1, 0.2, 0.5)
samples are presented in Fig. 2. The main peak at
ca. 460 cm−1 was assigned to the F2g mode vibration of CeO2
cubic fluorite structure.34 The Bi-doped samples also showed
two peaks at ca. 520 and 580 cm−1, which could be indexed
to the presence of Bi3+ ions35 and oxygen vacancies generated
as charge-compensating defects,36 respectively. For CeO2, a
very weak peak at ca. 580 cm−1 was also observed, suggesting
the presence of a small number of oxygen vacancies in CeO2
possibly caused in the violent and rough citrate-method
HMF conversion in alkaline aqueous solution was carried
out with or without Ce1−xBixO2−δ samples. The conversion of
HMF and the yields of FDCA and HFCA determined by HPLC
analysis are listed in Table 2. It is known that HMF is prone
to degradation in alkaline aqueous solution, especially at
high temperature. The red-brown solution was obtained with
the formation of by-products such as levulinic and formic
acids.18 When Ce1−xBixO2−δ (x = 0, 0.05) or no catalyst was
added into the reaction system, the resulting reaction solu-
tions after 1 h at 65 °C were red-brown with low yields of
HFCA (Table 2, entries 1–3). However, with Ce1−xBixO2−δ
(0.08 ≤ x ≤ 0.5) in the reaction system, the resultant solu-
tions were nearly colorless and the yields of HFCA were high
(Table 2, entries 4–9). The results may be ascribed to the
presence of abundant stoichiometric oxidation sites in
Ce1−xBixO2−δ (0.08 ≤ x ≤ 0.5).
To further explain the results, the following experiments
were performed. After reaction over CeO2 or Ce0.875Bi0.125O2−δ
for 1 h, the reaction solution was centrifuged. The obtained
supernatant was reacted for another hour by using the
Au/Ce0.9Bi0.1O2−δ catalyst under the same conditions.
Fig. 2 Raman spectra of (a) CeO2, (b) Ce0.95Bi0.05O2−δ, (c) Ce0.9Bi0.1O2−δ
,
Fig. 3 H2-TPR profiles of (a) CeO2, (b) Ce0.95Bi0.05O2−δ, (c) Ce0.9Bi0.1O2−δ
,
(d) Ce0.8Bi0.2O2−δ and (e) Ce0.5Bi0.5O2−δ
.
(d) Ce0.8Bi0.2O2−δ and (e) Ce0.5Bi0.5O2−δ
.
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
Table 2 HMF reaction in alkaline aqueous solution over Ce1−xBixO2−δ catalystsa
HMF Conversionb (%)
YFDCA (%)
YHFCA (%)
YBHMF (%)
b
b
d
Entry
Catalyst
1
2
3
4
5
6
7
8
9
—
CeO2
93
—
4
—
9
97 (100)c
99
— (9)c
15 (19)c
26
Ce0.95Bi0.05O2−δ
Ce0.92Bi0.08O2−δ
Ce0.9Bi0.1O2−δ
Ce0.875Bi0.125O2−δ
Ce0.85Bi0.15O2−δ
Ce0.8Bi0.2O2−δ
Ce0.5Bi0.5O2−δ
2
5
5
21
46
42
33
32
36
33
99
48
53
100
100 (100)c
100
2 (90)c
64 (10)c
65
2
3
8
100
100
60
58
a
b
General conditions: 0.15 M HMF solution in 0.6 M NaOH (6 ml), 0.1 g catalyst, 65 °C, 1.0 MPa O2. Conversion or product yield determined
c
by HPLC after reaction for 1 h. Yield after the reaction supernatant was reacted for another 1 h under the same conditions by using the
Au/Ce0.9Bi0.1O2−δ (Au: HMF = 6.67 × 10−3 mol mol−1) catalyst. d Yield determined by GC-MS after reaction for 1 h.
Interestingly, the Ce0.875Bi0.125O2−δ system obtained 90%
yield of FDCA and 10% yield of HFCA (Table 2, entry 6) while
with CeO2 a total 28% yield of FDCA and HFCA was obtained
(Table 2, entry 2). Obviously, no degradation of HMF
occurred in the presence of Ce0.875Bi0.125O2−δ in alkaline
aqueous solution. The product undetected by HPLC analysis
after 1 h over Ce0.875Bi0.125O2−δ was thought to be BHMF.
BHMF was the Cannizzaro reaction product13,17 and did not
respond to UV light39 used in the HPLC analysis (Table 2,
contained a large amount of stoichiometric oxidation sites
and acted as an oxygen donor.40–42 When gaseous oxygen was
introduced into the reaction system, Ce0.875Bi0.125O2−δ could
act as an oxygen pump, releasing and adsorbing oxygen
through a redox process. The redox process involved the vari-
ation of the two oxidation states of Ce (III and IV) and related
to the exotic highly reductive Bi3+. The lone electron pair of
Bi3+ and the large amount of oxygen vacancies facilitated O2
adsorption and dissociation during this process.27–29 As a
result, the O2 present in the reaction system replenished the
consumed lattice oxygen (mainly oxide ions of the Bi–O–Ce
linkages) in Ce0.875Bi0.125O2−δ and prevented Ce0.875Bi0.125O2−δ
from deactivating, leading to the transformation of HMF via
Cannizzaro reaction without degradation.43,44
1
entry 6). H NMR analysis also proved it (Fig. S4†). Then, the
BHMF yields of these reactions were determined by gas
chromatography-mass spectrometry (GC-MS) analysis after
extracting the resulting reaction solutions with ethyl ether
(Fig. S5†). The results are also listed in Table 2.
Under air or 1.0 MPa Ar atmosphere and otherwise the
same conditions, reactions with Ce0.875Bi0.125O2−δ all resulted
in sepia solutions and low yields of HFCA (Table 3, entries 2
and 3). These results indicated that O2 was essential for HMF
conversion catalyzed by Ce0.875Bi0.125O2−δ and there was
strong interaction between Ce1−xBixO2−δ (0.08 ≤ x ≤ 0.5) and
O2. Other M-doped Ce0.9M0.1O2−δ (M = Y, La, Mn, Fe) samples
were also examined in HMF conversion in the presence of O2
(1.0 MPa). They all resulted in serious degradation of HMF
with low yields of HFCA (Table S1†). It implied that the
Bi-doped CeO2 was unique for O2 activation during HMF
conversion in alkaline aqueous solution.
Subbiah et al.13 have shown that Cannizzaro reaction of
HMF could be performed at room temperature without a
catalyst in NaOH aqueous solution. The reaction time was up
to 18 h. The 43% yield of BHMF and 44% yield of HFCA
obtained under the optimized reaction conditions suggested
more than 10% degradation of HMF. Moreover, serious
decomposition of HMF was found at higher temperature
(60 °C), producing only 22% total yield of BHMF and HFCA
even after 24 h. In our system, however, the degradation of
HMF was inhibited even at 65 °C, and the reaction time for
100% conversion of HMF was reduced greatly to less than
1 h in the presence of Ce1−xBixO2−δ (0.08 ≤ x ≤ 0.5) and
1.0 MPa O2 in the reaction system simultaneously (Table 2,
entries 4–9). The reaction efficiency was improved greatly.
A 66% total yield of carboxylic acid products (more than
50%) was obtained under 1.0 MPa O2 pressure over the
Ce0.875Bi0.125O2−δ catalyst (Table 2, entry 6). It implied that
Cannizzaro reaction and oxidation of HMF proceeded
simultaneously. The geminal diol intermediate was first
formed via rapid reversible hydration of HMF.17,40 Then, it
partly reacted with the aldehyde moiety of HMF and obtained
the Cannizzaro reaction products HFCA and BHMF
equivalently,13,19 while others underwent the oxidation of the
geminal diol or the hydroxymethyl group and yielded HFCA
or FDCA. The electrons generated during the oxidation reac-
tion were removed by molecular oxygen, and the catalytic
16% yield of HFCA was obtained under a 1.0 MPa Ar
atmosphere (Table 3, entry 3). It meant that Ce0.875Bi0.125O2−δ
a
Table 3 HMF reaction at different atmospheres over Ce0.875Bi0.125O2−δ
HMF
Conversionb (%)
YFDCA (%)
YHFCA (%)
b
b
Entry
Atmosphere
1
1.0 MPa O2
Air
1.0 MPa Ar
100
97
98
2
1
1
64
24
16
2c
3
a
General conditions: 0.15 M HMF solution in 0.6 M NaOH (6 ml),
b
0.1
g
Ce0.875Bi0.125O2−δ
,
65 °C.
Conversion or product yield
c
determined by HPLC after reaction for 1 h. Reaction autoclave was
sealed after all the reactants were added without purging of any gas.
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
However, Au/Ce1−xBixO2−δ (x = 0, 0.05) catalysts resulted in
not only low yields of FDCA but also the production of some
degradation products (Table 4, entries 1 and 2). The favorable
catalytic activity of Au/Ce1−xBixO2−δ (0.08 ≤ x ≤ 0.2) towards
HMF oxidation was attributed to the presence of Auδ+ and Bi
doping. Auδ+ in the Au/Ce1−xBixO2−δ catalysts assisted the
oxidation of alcohol, including transformation of alcohol into
a metal–alkoxide intermediate and activation of the C–H
bond. Besides, in the whole oxidation process, the hydride
transferred from the C–H bond could combine not only
with Ce3+ and Auδ+ but also with oxygen ions of Bi–O–Ce link-
ages and Ce3+–O2 centers.16,25,45 The electrons generated
during the hydride transfer were captured by Au clusters and
the reducible Ce1−xBixO2−δ support and then released when
the Au/Ce1−xBixO2−δ interacted with molecular O2.17,46 As the
geminal diol intermediate was more easily oxidized, the
Cannizzaro reaction was inhibited over Au/Ce1−xBixO2−δ
(x ≤ 0.2) catalysts. Oxidation products HFCA and FDCA were
obtained without the formation of BHMF. Based on the
above analysis, the first two reaction processes are listed in
Scheme 3.
TEM images of Au/Ce1−xBixO2−δ (x = 0.1, 0.5) after reaction
for 1 h are shown in Fig. 5. Because of the low loading
amount of Au and the low contrast between Au and ceria,
Au particles were hardly detectable by TEM.47 Only a few
nanocrystals with a size of about 4 nm were identified as the
gold phase on the Au/Ce0.9Bi0.1O2−δ catalyst (Fig. 5a, b), and
no Au particle was found on the Au/Ce1−xBixO2−δ (x = 0, 0.2,
0.5) catalysts. XRD patterns of all the samples are shown in
Fig. S6.† No characteristic diffraction peak corresponding to
gold (38.2°, 44.3°) was observed. The Bi2O3 phase was also
found in the Au/Ce0.5Bi0.5O2−δ sample by TEM (Fig. 5c).
Quantitative fitting of the Au (4f) peaks of Au/Ce1−xBixO2−δ
(x = 0, 0.1, 0.2, 0.5) catalysts after reaction for 1 h is shown in
Table 5 and Fig. S7.† It is well documented in the literature
that the electronic state of surface gold NPs depends on the
size of gold NPs and the nature of supports and always plays
an important role in the oxidation reaction.48 For alcohol
oxidation catalyzed by Au/CeO2, Auδ+ was considered as the
active site on Au clusters.23,25,46 There existed about 25% of
Auδ+ in the Au/Ce0.5Bi0.5O2−δ catalyst, which might be too low
to afford the activation of the C–H bond in alcohol, and
Fig. 4 HMF conversion and product yields as a function of reaction
time during HMF reaction in aqueous solution over the
Ce0.875Bi0.125O2−δ catalyst (0.1
g Ce0.875Bi0.125O2−δ, 4 equiv. NaOH,
1.0 MPa O2, 65 °C; HMF: ■, HFCA: ▲, FDCA: ●).
cycle was closed.17 The oxidation and Cannizzaro reaction
both involved hydride transfer from the geminal diol and the
Bi-doped CeO2 showed unique performance towards the
reaction system. Therefore, we summarized that the presence
of Bi–O–Ce linkages and Ce3+ centers, together with the lone
electron pair of Bi3+ and the large amount of oxygen
vacancies, assisted Ce0.875Bi0.125O2−δ in transferring hydride
and activating O2, thereby catalysing the reaction.25,29
As shown in Fig. 4, HMF was quickly transformed into
products in the presence of O2 and Ce0.875Bi0.125O2−δ. A 97%
conversion of HMF was achieved only after 20 min. It further
indicated the excellent properties of Ce0.875Bi0.125O2−δ towards
activating oxygen and transferring hydride. Ce0.875Bi0.125O2−δ
itself showed weak ability to activate the cleavage of C–H
bond in the hydroxymethyl group, resulting in the low yield
of FDCA and the large amount of BHMF.
All the Au/Ce1−xBixO2−δ samples showed similar Au loading
around 1 wt%. To compare their catalytic activities, they
were all applied in the oxidation of HMF under the same
reaction conditions for 1 h. The results are listed in Table 4.
Besides Au/Ce0.5Bi0.5O2−δ, all the Bi-doped catalysts displayed
a higher FDCA yield with respect to Au/CeO2. Au/Ce1−xBixO2−δ
(0.08 ≤ x ≤ 0.2) catalysts were active enough to catalyze the
oxidation of HMF to HFCA and FDCA (Table 4, entries 3–7).
Table 4 Catalytic activity of Au/Ce1−xBixO2−δ towards HMF oxidationa
Au loadingb (wt%)
HMF Conversionc (%)
YFDCA (%)
YHFCA (%)
c
c
Entry
Catalyst
1
2
3
4
5
6
7
8
Au/CeO2
1.02
1.04
1.05
1.11
1.07
1.09
1.10
1.10
100
100
100
100
100
100
100
99
39
46
73
74
75
77
74
4
57
49
27
26
25
23
26
56
Au/Ce0.95Bi0.05O2−δ
Au/Ce0.92Bi0.08O2−δ
Au/Ce0.9Bi0.1O2−δ
Au/Ce0.875Bi0.125O2−δ
Au/Ce0.85Bi0.15O2−δ
Au/Ce0.8Bi0.2O2−δ
Au/Ce0.5Bi0.5O2−δ
a
b
General conditions: 0.15 M HMF solution in 0.6 M NaOH (6 ml), Au: HMF = 6.67 × 10−3 mol mol−1, 65 °C, 1.0 MPa O2. From ICP-OES
c
analysis. Conversion or product yield determined by HPLC after reaction for 1 h.
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
Scheme 3 The first two reaction processes of HMF conversion in
alkaline aqueous solution. CeBi represents Ce1−xBixO2−δ (0.08 ≤ x ≤ 0.5)
and Au represents Au/Ce1−xBixO2−δ (x ≤ 0.2).
Fig. 6 HMF conversion and product yields as a function of reaction
time in the oxidation of HMF in aqueous solution over the
Au/Ce0.9Bi0.1O2−δ catalyst (Au: HMF = 6.67 × 10−3 mol mol−1, 4 equiv.
NaOH, 1.0 MPa O2, 65 °C; HMF: ■, FDCA: ●, HFCA: ▲, FFCA: ▼).
FFCA was detected within the first 45 min, indicating that
the oxidation of the alcohol group to aldehyde was the rate-
limiting step. It was identical to previous reports.15,18
Fig. 5 HRTEM of (a, b) Au/Ce0.9Bi0.1O2−δ and (c) Au/Ce0.5Bi0.5O2−δ
after reaction for 1 h.
Catalyst recycling was also studied for the Au/Ce0.9Bi0.1O2−δ
catalyst. After the catalyst was recovered by filtration, washed
with water and dried at 60 °C, a new set of HMF oxidation
experiments were carried out (Fig. 7). The presence and
quantity of the Au element in the filtrate were analyzed by
ICP-OES analysis, thus excluding the hypothesis of Au
leaching. The catalyst could be used three times under
identical reaction conditions without much loss of activity.
The yield of FDCA was only 86% after the fourth use, but
the solution after reaction was also colorless and there was
almost no other by-product besides HFCA. We deemed that
increasing the reaction time would also result in a >99% yield
of FDCA. After using three times, we found that large cata-
lyst particles appeared. The decrease in the FDCA selectivity
may be due to the agglomeration of the catalyst or the
adsorption of organic molecules on partial active sites of
the catalyst.
resulted in the low yield of FDCA (Table 4, entry 8). Here, we
also noticed that after calcination at 200 °C for 2.5 h, the
Au/Ce1−xBixO2−δ (x ≤ 0.15) samples turned into purple, while
Au/Ce0.8Bi0.2O2−δ was sap green and Au/Ce0.5Bi0.5O2−δ was
emerald. It implied the presence of different Au species in
Au/Ce1−xBixO2−δ catalysts. In addition, the color of these
catalysts deepened after NaOH was added into the reaction
mixture, which was similar to a previous report.48
The above results demonstrated that the Au/Ce1−xBixO2−δ
(0.08 ≤ x ≤ 0.2) catalysts were preferable for HMF oxidation
to FDCA. Au/Ce0.9Bi0.1O2−δ was selected to further investigate
the effect of the reaction parameters, such as amount of
base, temperature, oxygen pressure, HMF/Au molar ratio and
reaction time. The results are shown in Fig. S8–S11.†
4 equiv. of NaOH, 1.0 MPa O2, 65 °C and 150 HMF/Au
molar ratio were chosen to study the effect of reaction time
over the Au/Ce0.9Bi0.1O2−δ catalyst. As shown in Fig. 6, the
aldehyde moiety of HMF was quickly oxidized to carboxylic
acid at the initial stage of the reaction. 100% conversion of
HMF and 95% yield of HFCA were achieved only after
10 min, which stressed the superiority of Ce1−xBixO2−δ for
HMF conversion. Subsequently, the amount of HFCA slowly
decreased while the yield of FDCA increased. Only a little
Table 5 XPS Au 4f peak analysis for Au/Ce1−xBixO2−δ catalysts after
reaction for 1 h
XPS Au 4f peak
Sample
Au0
Au+
Au3+
Fig. 7 Reusability study of the Au/Ce0.9Bi0.1O2−δ catalyst in the oxidation
Au/CeO2
44.0
54.4
62.2
75.4
34.9
45.6
37.8
24.6
21.1
—
—
of HMF in aqueous solution (Au: HMF = 6.67 × 10−3 mol mol−1
,
Au/Ce0.9Bi0.1O2−δ
Au/Ce0.8Bi0.2O2−δ
Au/Ce0.5Bi0.5O2−δ
4 equiv. NaOH, 1.0 MPa O2, 65 °C, 2 h; HMF: black, FDCA: red,
HFCA: blue).
—
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
13 S. Subbiah, S. P. Simeonov, J. Esperanca, L. P. N. Rebelo and
C. A. M. Afonso, Green Chem., 2013, 15, 2849–2853.
14 B. Saha, S. Dutta and M. M. Abu-Omar, Catal. Sci. Technol.,
2012, 2, 79–81.
15 O. Casanova, S. Iborra and A. Corma, ChemSusChem, 2009,
2, 1138–1144.
16 N. K. Gupta, S. Nishimura, A. Takagaki and K. Ebitani, Green
Chem., 2011, 13, 824–827.
17 S. E. Davis, B. N. Zope and R. J. Davis, Green Chem., 2012,
14, 143–147.
18 T. Pasini, M. Piccinini, M. Blosi, R. Bonelli, S. Albonetti,
N. Dimitratos, J. A. Lopez-Sanchez, M. Sankar, Q. He,
C. J. Kiely, G. J. Hutchings and F. Cavani, Green Chem., 2011,
13, 2091–2099.
19 Y. M. Shapiro, V. P. Smolyakov and V. G. Kulnevich,
Khim. Geterotsikl. Soedin., 1990, 741–744.
20 G. Hutchings, Gold Bull., 2009, 42, 260–266.
21 Y. Kuwauchi, H. Yoshida, T. Akita, M. Haruta and S. Takeda,
Angew. Chem., Int. Ed., 2012, 51, 7729–7733.
22 W. Y. Song and E. J. M. Hensen, Catal. Sci. Technol., 2013, 3,
3020–3029.
23 M. Wang, F. Wang, J. Ma, M. Li, Z. Zhang, Y. Wang,
X. Zhang and J. Xu, Chem. Commun., 2014, 50, 292–294.
24 N. Ta, J. Liu, S. Chenna, P. A. Crozier, Y. Li, A. Chen and
W. Shen, J. Am. Chem. Soc., 2012, 134, 20585–20588.
25 A. Abad, P. Concepcion, A. Corma and H. Garcia, Angew.
Chem., Int. Ed., 2005, 44, 4066–4069.
Conclusions
In this work, a series of porous Bi-doped ceria were pre-
pared by the easy citrate method. XRD and Raman analyses
indicated that Ce1−xBixO2−δ solid solutions were obtained
with a large amount of oxygen vacancy defects. In the pres-
ence of O2, Ce1−xBixO2−δ (0.08 ≤ x ≤ 0.5) efficiently catalyzed
HMF to BHMF and HFCA in alkaline aqueous solution.
Nearly complete conversion of HMF (97%) was reached only
in 20 min at 65 °C and 1.0 MPa O2 without degradation of
HMF. The superior catalytic properties of Ce1−xBixO2−δ
(0.08 ≤ x ≤ 0.5) were attributed to the specific Bi doping.
O2 activation and hydride transfer were enhanced by the lone
electron pair of Bi3+ and the presence of Bi–O–Ce linkages
together with large amounts of Ce3+ centers and oxygen
vacancies. The presence of Auδ+ in the Au/Ce1−xBixO2−δ (x ≤ 0.2)
catalysts assisted the oxidation of alcohol, especially the
geminal diol intermediate, yielding FDCA as an end product
without BHMF. A >99% yield of FDCA was obtained after 2 h
over the Au/Ce0.9Bi0.1O2−δ catalyst. Ce1−xBixO2−δ (0.08 ≤ x ≤ 0.5)
may also be applicable for other Cannizzaro reactions with
O2. Combined with other noble metals or metal oxides,
Ce1−xBixO2−δ could activate the C–H bond in alcohols, leading
to the formation of active catalysts for HMF or other alcohol
oxidation processes in alkaline aqueous solution.
Acknowledgements
26 X. Liu, K. Zhou, L. Wang, B. Wang and Y. Li, J. Am. Chem.
Soc., 2009, 131, 3140–3141.
27 N. Imanaka, T. Masui, K. Koyabu, K. Minami and T. Egawa,
Adv. Mater., 2007, 19, 1608–1611.
The authors thank the National Natural Science Foundation
of China for support (21273221).
28 D. Jiang, W. Wang, E. Gao, L. Zhang and S. Sun, J. Phys.
Chem. C, 2013, 117, 24242–24249.
Notes and references
1 T. Werpy and G. Petersen, Top Value Added Chemicals from
Biomass, Pacific Northwest National Laboratory, 2004, vol. 1,
p. 27.
29 L. C. Glaeser, J. F. Brazdil, M. A. Hazle, M. Mehicic and
R. K. Grasselli, J. Chem. Soc., Faraday Trans., 1985, 81,
2903–2912.
2 Y. Roman-Leshkov, C. J. Barrett, Z. Y. Liu and J. A. Dumesic,
Nature, 2007, 447, 982–U985.
30 J. Zhu, D. Xiao, J. Li and X. Yang, Catal. Lett., 2009, 129,
240–246.
3 A. Corma, S. Iborra and A. Velty, Chem. Rev., 2007, 107,
2411–2502.
4 A. Osatiashtiani, A. F. Lee, D. R. Brown, J. A. Melero, G. Morales
and K. Wilson, Catal. Sci. Technol., 2014, 4, 333–342.
5 D. A. Kotadia and S. S. Soni, Catal. Sci. Technol., 2013, 3,
469–474.
31 A. Karpenko, R. Leppelt, J. Cai, V. Plzak, A. Chuvilin,
U. Kaiser and R. J. Behm, J. Catal., 2007, 250, 139–150.
32 L. Xue, C. Zhang, H. He and Y. Teraoka, Catal. Today,
2007, 126, 449–455.
33 Z. Ma, H. Yin, S. Overbury and S. Dai, Catal. Lett., 2008, 126,
20–30.
6 A. S. Amarasekara, D. Green and L. D. Williams, Eur. Polym. J.,
2009, 45, 595–598.
34 L. Li, F. Chen, J.-Q. Lu and M.-F. Luo, J. Phys. Chem. A,
2011, 115, 7972–7977.
7 A. Gandini, ACS Symp. Ser., 1990, 433, 195–208.
8 H. Hirai, J. Macromol. Sci., Chem., 1984, 21, 1165–1179.
9 C. Moreau, M. N. Belgacem and A. Gandini, Top. Catal.,
2004, 27, 11–30.
35 I. Kosacki, T. Suzuki, H. U. Anderson and P. Colomban,
Solid State Ionics, 2002, 149, 99–105.
36 M. Guo, J. Lu, Y. Wu, Y. Wang and M. Luo, Langmuir,
2011, 27, 3872–3877.
10 P. Vinke, H. E. van Dam and H. van Bekkum, New Dev.
Select. Oxidat., 1990, 55, 147–157.
37 L. Kumari, J. H. Lin and Y. R. Ma, J. Phys.: Condens. Matter,
2007, 19, 406204.
11 P. Verdeguer, N. Merat and A. Gaset, J. Mol. Catal., 1993, 85,
327–344.
12 S. E. Davis, L. R. Houk, E. C. Tamargo, A. K. Datye and
R. J. Davis, Catal. Today, 2011, 160, 55–60.
38 V. N. Denisov, A. N. Ivlev, A. S. Lipin, B. N. Mavrin and
V. G. Orlov, J. Phys.: Condens. Matter, 1997, 9, 4967–4978.
39 S. Goswami, S. Dey and S. Jana, Tetrahedron, 2008, 64,
6358–6363.
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
40 O. Casanova, S. Iborra and A. Corma, J. Catal., 2009, 265,
109–116.
C. J. Kiely, D. Bethell, D. W. Knight and G. J. Hutchings,
Chem. – Eur. J., 2011, 17, 6524–6532.
41 J. Guzman, S. Carrettin, J. C. Fierro-Gonzalez, Y. L. Hao,
B. C. Gates and A. Corma, Angew. Chem., Int. Ed., 2005, 44,
4778–4781.
42 J. Beckers, A. F. Lee and G. Rothenberg, Adv. Synth. Catal.,
2009, 351, 1557–1566.
45 A. Abad, C. Almela, A. Corma and H. Garcia, Tetrahedron,
2006, 62, 6666–6672.
46 B. T. Teng, J. J. Lang, X. D. Wen, C. Zhang, M. H. Fan and
H. G. Harris, J. Phys. Chem. C, 2013, 117, 18986–18993.
47 R. Si and M. Flytzani-Stephanopoulos, Angew. Chem., 2008,
120, 2926–2929.
43 R.
Sumathi,
K.
Johnson,
B.
Viswanathan
and
T. K. Varadarajan, Appl. Catal., A, 1998, 172, 15–22.
44 M. Sankar, E. Nowicka, R. Tiruvalam, Q. He, S. H. Taylor,
48 J. Yang, Y. J. Guan, T. Verhoeven, R. van Santen, C. Li and
E. J. M. Hensen, Green Chem., 2009, 11, 322–325.
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol.