Facile synthesis of CeO2-supported gold nanoparticle catalysts for selective oxidation of glycerol into lactic acid

Article abstract of DOI:10.1016/j.apcata.2013.08.048

This study explores a new synthetic route toward CeO₂-supported gold nanoparticles catalysts (1, 3, and 5 wt.% Au) and their catalysis of the oxidation of glycerol at atmospheric pressure. Gold was loaded on a hydrothermally synthesized nanocrystalline CeO₂ support (～8 nm) via deposition-precipitation with urea (DPU). To obtain cerium oxide-supported gold nanoparticles, thermal reduction at 300 C under flowing H₂ is currently popular. However, this method invariably forms small gold particles (<5 nm) even with higher gold loading. In the present study, Au/CeO ₂ catalysts prepared by DPU were reduced via a new method: chemical reduction with glycerol (CRG). Interestingly, the reduction of a gold-urea complex supported on ceria occurs at 25 C, and the gold nanoparticles grow slowly. The resultant chemically reduced samples were compared with their hydrogen-reduced counterparts. The CRG method primarily generated gold particles with larger average sizes than the H₂ reduction method. Importantly, the CRG route enabled variation of the average gold nanoparticle size without requiring any other reagent, which is particularly advantageous for the Au/CeO₂ system. The efficiency of these catalysts toward the aerobic oxidation of glycerol was tested. Accordingly, the selective oxidation of glycerol into lactate can be effectively catalyzed by Au/CeO₂ catalysts at atmospheric pressure. The chemically reduced samples show increased lactate selectivity than the hydrogen-reduced samples due to the larger size of the gold particles.

Full text of DOI:10.1016/j.apcata.2013.08.048

Applied Catalysis A: General 468 (2013) 260–268
Contents lists available at ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Facile synthesis of CeO2-supported gold nanoparticle catalysts for
selective oxidation of glycerol into lactic acid
Pandian Lakshmanana, Pravin P. Uparea,b, Ngoc-Thuc Lea,b, Young Kyu Hwanga,b,∗
,
Dong Won Hwanga,b, U-Hwang Leea, Hyung Rok Kima,b, Jong-San Changa,b,c,∗∗
a Catalysis Center for Molecular Engineering, Korea Research Institute of Chemical Technology (KRICT), 141 Gajeong-Ro, Yuesong, Daejeon 305-600,
Republic of Korea
b Department of Green Chemistry, University of Science and Technology (UST), 217 Gajeong-Ro, Yuseong, Daejeon 305-350, Republic of Korea
c Department of Chemistry Sungkyunkwan University, Suwon 440-476, Republic of Korea
a r t i c l e i n f o
a b s t r a c t
Article history:
This study explores a new synthetic route toward CeO2-supported gold nanoparticles catalysts (1, 3, and
5 wt.% Au) and their catalysis of the oxidation of glycerol at atmospheric pressure. Gold was loaded on
a hydrothermally synthesized nanocrystalline CeO2 support (∼8 nm) via deposition-precipitation with
urea (DPU). To obtain cerium oxide-supported gold nanoparticles, thermal reduction at 300 ◦C under
flowing H2 is currently popular. However, this method invariably forms small gold particles (<5 nm)
even with higher gold loading. In the present study, Au/CeO2 catalysts prepared by DPU were reduced via
a new method: chemical reduction with glycerol (CRG). Interestingly, the reduction of a gold-urea com-
plex supported on ceria occurs at 25 ◦C, and the gold nanoparticles grow slowly. The resultant chemically
reduced samples were compared with their hydrogen-reduced counterparts. The CRG method primar-
ily generated gold particles with larger average sizes than the H2 reduction method. Importantly, the
CRG route enabled variation of the average gold nanoparticle size without requiring any other reagent,
which is particularly advantageous for the Au/CeO2 system. The efficiency of these catalysts toward the
aerobic oxidation of glycerol was tested. Accordingly, the selective oxidation of glycerol into lactate can
be effectively catalyzed by Au/CeO2 catalysts at atmospheric pressure. The chemically reduced samples
show increased lactate selectivity than the hydrogen-reduced samples due to the larger size of the gold
particles.
Received 15 April 2013
Received in revised form 7 August 2013
Accepted 28 August 2013
Available online xxx
Keywords:
Biomass
Glycerol oxidation
Lactic acid
Gold nanoparticles
Cerium oxide
Chemical-reduction
© 2013 Elsevier B.V. All rights reserved.
1. Introduction
perspective of the applications of lactic acid: Lactic acid is the
building block for the synthesis of polylactic acid (PLA) [1], which
The use of bio-renewable resources for the production of
chemicals continues to attract significant research attention to
support sustainable development. Glycerol is a particularly intrigu-
ing reagent since it is a major co-product of biodiesel production.
The oxidation of glycerol leads to the formation of a number of
valuable oxygenated compounds such as glyceric acid, lactic acid,
dihydroxyacetone, hydroxypyruvic acid, glycolic acid, oxalic acid,
and tartronic acid. Hence, control of product selectivity is crucial.
Conversion of glycerol into lactic acid is very significant from the
is involved in the production of biodegradable plastics and bio-
compatible medical materials [2–5]. Selective oxidation of glycerol
primarily using supported noble metal nanoparticles such as Pd, Pt,
and Au with molecular oxygen [6–14] has been extensively studied,
mainly for glyceric acid production. Previous studies showed that
lactic acid can be generated from glycerol through a heterogeneous
catalysis process. Different approaches including a hydrothermal
reaction [15–17], hydrogenolysis [18,19], and catalytic selective
oxidation [3,20–31] have been applied for the synthesis of lactic
acid from glycerol. Compared to the hydrothermal and hydrogenol-
ysis methods [15–19], catalytic selective oxidation allows the
reaction to be performed under milder reaction conditions [32]
with supported gold catalysts.
∗
Corresponding author at: Catalysis Center for Molecular Engineering, Korea
Research Institute of Chemical Technology (KRICT), 141 Gajeong-Ro, Yuesong,
Daejeon 305-600, Republic of Korea. Tel.: +82 42 860 7680; fax: +82 42 860 7676.
The problem in the production of lactic acid from glycerol via
a selective oxidation process is that the glyceraldehyde interme-
diate and its isomer, dihydroxyacetone, are susceptible to further
oxidation [3,20,21]. This has been problematic in previous stud-
ies, resulting in the formation of primarily glyceric acid without
∗∗
Corresponding author at: Catalysis Center for Molecular Engineering, Korea
Research Institute of Chemical Technology (KRICT), 141 Gajeong-Ro, Yuesong,
Daejeon 305-600, Republic of Korea. Fax: +82 42 860 7676.
E-mail addresses: ykhwang@krict.re.kr (Y.K. Hwang),
jschang@krict.re.kr (J.-S. Chang).
0926-860X/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.apcata.2013.08.048

P. Lakshmanan et al. / Applied Catalysis A: General 468 (2013) 260–268
261
Scheme 1. Schematic illustration of the synthesis of Au/CeO2 catalysts by chemical reduction with glycerol.
any significant formation of lactic acid in most of the cases [9–16].
2. Experimental
Product selectivity during glycerol oxidation is strongly influenced
by the size of the gold particles and reaction conditions [26,33–35].
With Au/C catalysts, larger particles (>20 nm) imparted improved
selectivity to glyceric acid compared to gold particles with an
average size of 6 nm [26]. For gold particles, the activity toward
the formation of lactate from propylene glycol increased with
decreasing particle size (70 ◦C, 2 atm O2, 1 equiv. NaOH), whereas
with carbon-supported gold, the maximum activity (100% conver-
sion, selectivity was not mentioned) was achieved with a mean gold
particle diameter of 7–8 nm [33].
The higher catalytic activity of the carbon-supported gold par-
ticles toward the conversion of glycerol to glyceric acid was
attributed to the small size of the Au nanoparticles, which were
supported on anionic exchange resins (50 ◦C, 3 atm O2, 4 equiv.
NaOH) [34]. By tuning the metal particle size from 2 to 16 nm,
a progressive decrease in activity and simultaneous increase in
selectivity for sodium glycerate production was observed [35].
Baiker et al. studied the catalytic activity of Au/CeO2 with dif-
ferent sizes of gold particles (ranging from 1.5 to 12 nm) on the
aerobic oxidation of benzyl alcohol. Accordingly, optimal activity
was observed for catalysts with medium particle sizes (∼6.9 nm),
whereas both smaller and larger particles showed inferior activ-
ity (100 ◦C, 1 atm O2, no base, toluene/mesitylene as solvent)
[36].
2.1. Hydrothermal synthesis of nanocrystalline ceria support
The appropriate quantity of cerium ammonium nitrate salt
(Aldrich, 98.5%) to achieve a concentration of 0.1 M was dissolved
in deionized water and stirred vigorously for 30 min. This solution
was then poured into an autoclave and treated hydrothermally at
200 ◦C for 12 h (Scheme 1). The resultant precipitates were sepa-
rated by centrifugation, washed with deionized water, dried in an
oven at 80 ◦C for 2 h, and calcined at 450 ◦C for 3 h under flowing
air.
2.2. Au-loading by deposition precipitation using urea (DPU)
The calcined ceria was dispersed in distilled water and main-
tained at 80 ◦C. A solution of HAuCl4·4H2O was added followed by
the addition of urea (Scheme 1); the resulting mixture was stirred
at 80 ◦C for 12 h. Centrifugation was performed to isolate the solid,
which was washed thrice with distilled water to remove chlorides
and then dried in a desiccator.
2.3. Reduction of gold catalysts via chemical reduction with
glycerol (CRG)
Hydrogen-reduced Au/CeO2 samples invariably form smaller
gold nanoparticles (<5 nm) regardless of the gold loading [37,38].
Since smaller gold particles favor the formation of glyceric acid, not
much attention was paid to Au/CeO2 for the conversion of glycerol
to lactic acid. Herein, we report the facile synthesis of nanopar-
ticle Au/CeO2 catalysts with varying average gold particle sizes
and their efficiency toward the catalysis of the selective aerobic
oxidation of glycerol to lactic acid at atmospheric pressure. The
synthesis of Au/CeO2 catalysts with different average gold parti-
cles was accomplished using the glycerol-based chemical reduction
protocol (CRG). The efficiency of this method was examined with
three different gold loadings. The synthesized gold catalysts were
characterized via elemental analysis, UV–vis spectroscopy, X-ray
diffraction, and high-resolution transmission electron microscopy
(HR-TEM).
In a typical process, the Au/CeO2 samples obtained via the DPU
method were mixed in a beaker with the appropriate quantity of
glycerol and stirred at ambient temperature for 48 h (Scheme 1, sec-
ond step). The reduced solid catalysts were separated from glycerol
via centrifugation and then washed thrice with deionized water to
remove the glycerol. The sample was directly transferred into the
reaction mixture for glycerol oxidation.
2.4. Characterization
Chemical analyses were performed via inductively coupled
plasma atomic emission spectroscopy (ICP-AES) using a Thermo
Scientific ICAP 6500 instrument. The Au loadings are expressed
as wt.%. X-ray diffraction (XRD) analysis was performed on a
Rigaku D/MAX-2200V diffractometer (Cu K␣). TEM analysis was

262
P. Lakshmanan et al. / Applied Catalysis A: General 468 (2013) 260–268
performed using a TECNAI G2 microscope. UV–vis spectroscopic
measurements were carried out using a SHIMADZU UV-1800
instrument.
*
(c)
*
2.5. Catalytic testing
*
*
#
#
#
(vi)
(v)
(iv)
(iii)
(ii)
(i)
*
*
Reactions were carried out in a three-necked round-bottom
flask (100 mL) equipped with air flow at atmospheric pressure
under vigorous stirring at 90 ◦C. The products were analyzed via
high-pressure liquid chromatography (HPLC) using an Alltech OA-
1000 column (300 mm × 6.5 mm) with 0.1% H3PO4 solution as the
eluent. For the HPLC analyses, 0.5 mL of the reaction mixture was
diluted to a total volume of 5.0 mL with the eluent. The products
were identified by UV (210 nm) and refractive index detectors by
comparing them to the original samples. Glycerol appears only
in a refractive index detector giving a strong signal. For example,
glycerol was determined using the refractive index detector, while
lactic acid was determined using UV detector (Figs. S1 and S2).
The conversion of glycerol, selectivity to product, and carbon
balance were calculated according to Eqs. (1)–(3), respectively.
20
30
40
50
60
70
*
(b)
*
*
*
*
*
#
#
#
*
(vi)
*
Moles of glycerol converted
Moles of glycerol initial
Conversion (% mole) =
× 100
(1)
(v)
(iv)
(iii)
(ii)
(i)
Selectivity (% mole)
Moles of product formed
Total moles of C1 + C2 + C3 products formed
=
× 100
(2)
(3)
20
30
40
50
60
70
3MC + 2MC + MC
*
3
2
1
(a)
Carbon balance (% mole) =
× 100
3MCC
where MCC is total the moles of glycerol converted, MC is the moles
3
*
of the C3 products (lactic, glyceric and tartronic acids), MC is the
2
moles of C2 products (glycolic and oxalic acids), and MC is the
1
#
#
#
(vi)
*
*
moles of C1 product (formic acid), respectively.
(v)
(iv)
(iii)
(ii)
3. Results and discussion
3.1. Characterization of CeO2-supported gold catalysts
The nanocrystalline ceria support synthesized via the
(i)
20
30
40
60
70
2 /de5g0ree
hydrothermal method exhibited
a
surface area of 120 m2/g
determined from the N2 isotherm. Elemental analyses via ICP-AES
showed gold loadings of 0.95, 2.74, and 4.59 wt.%, which are
close to the nominal loadings of 1, 3, and 5 wt.%, respectively.
Fig. 1 shows the presence of CeO2 with distinguishable peaks that
indicate dominant (1 1 1), (2 0 0), (2 2 0), and (3 1 1) diffraction
patterns, suggesting a cubic structure of calcium fluorite (CaF2;
JCPDS 35-0816). The average particle size, which was calculated
using the Scherrer equation, suggests that the ceria nanoparticles
are ∼8 nm. Powder XRD patterns of the 1% Au/CeO2 samples
after chemical reduction with varying amounts of glycerol are
presented in Fig. 1a. The first line (i) in Fig. 1a corresponds to the
sample reduced by thermal treatment at 300 ◦C under a hydro-
gen flow. As the gold particles are very small, the XRD pattern
does not exhibit the diffraction pattern of gold. The other lines
in Fig. 1a correspond to the samples reduced by glycerol with
catalyst/glycerol weight ratios of 0.002, 0.005, 0.01, 0.02, and 0.04.
As is evident in Fig. 1a, these XRD patterns closely resemble that of
the hydrogen-reduced sample and the gold diffraction patterns are
not visible. The hydrogen-reduced sample is dark brown, while the
glycerol-reduced samples are violet-brown; this color difference
is indicative of the differences in the average size of the gold
particles. However, the XRD patterns are only slightly different
at the 2ꢀ value of 38.2◦, which corresponds to the (1 1 1) Bragg
reflection of the gold nanoparticles. To display the changes of the
Fig. 1. XRD patterns of Au/CeO2 catalyst at different concentration of Au. (a)
0.95 wt.% Au/CeO2, (b) 2.74 wt.% Au/CeO2 and (c) 4.59 wt.% Au/CeO2 by H2-reduction
((i) line) and glycerol reduction with catalyst to glycerol weight ratios of 0.002, 0.005,
0.01, 0.02, and 0.04, respectively (from (ii) to (vi) lines). * and # denoted CeO2 and
Au phases, respectively.
sizes of the gold nanoparticles, an expansion of the 2ꢀ range of
35–45◦ is shown in Fig. 2. It is evident that the widths of the Bragg
reflections for the glycerol-reduced samples gradually decrease
as the intensity increases. By comparing the broadening of the
curves, it can be inferred that the glycerol-reduced samples ((ii)
to (vi) lines) contain larger gold nanoparticles than the hydrogen-
reduced sample ((i) line) and that the diameter of the gold particles
is dependent on the catalyst/glycerol ratio.
The XRD results for 3% Au/CeO2 are depicted in Fig. 1b. Again,
the hydrogen-reduced 3% Au/CeO2 ((i) line) does not exhibit gold
peaks. The (ii) line of Fig. 1b shows a significant increase in the
intensity of Au (1 1 1) peak in comparison to that of the hydrogen-
reduced sample. This clearly shows that the average gold particle
size of the samples prepared via CRG is larger than that of the
hydrogen-reduced samples. With increasing catalyst/glycerol ratio,
a progressive sharpening of the Au (1 1 1) peak is evident. In con-
trast to the hydrogen-reduced sample, the (1 1 1) Bragg reflection
from the gold nanoparticles at the 2ꢀ value of 38.2◦ is clearly

P. Lakshmanan et al. / Applied Catalysis A: General 468 (2013) 260–268
263
5% Au/CeO2 catalysts (Fig. 1c). The hydrogen-reduced sample is
represented by the first line (i) in Fig. 1c: Compared with 1% and
3% Au/CeO2, the (1 1 1) Bragg reflection from the gold nanoparti-
cles in 5% Au/CeO2 is more intense, but still very broad. On the
other hand, the XRD results for glycerol-reduced 5% Au/CeO2 clearly
show a gold nanoparticle (111) Bragg reflection peak at the 2ꢀ
value of 38.2◦. The estimated average gold particle sizes are plot-
ted against the catalyst/glycerol ratio in Fig. 3b. These results show
that the average gold particle size increases with increasing cat-
alyst/glycerol ratio. In fact, the colors of the 5% Au/CeO2 catalysts
vary significantly: With increasing catalyst/glycerol ratio, the col-
ors varied from light-brown, dark brown, violet, magenta, pink, to
gray over time, which reflect the growth of the gold nanoparticles.
During the chemical reduction process, we observed changes
in the colors of the Au/CeO2 samples. UV–visible spectroscopy is
one of the few effective methods for investigating modifications of
the physical or chemical properties of supported gold catalysts. A
strong absorbance in the visible region around 540 nm is due to
excitation of the surface plasmon vibrations of Au nanoparticles
[39–42]. Fig. 4a shows the spectra recorded for 1% Au/CeO2 sam-
ples reduced by either hydrogen or different amounts of glycerol.
The first line (i) from the bottom shows that the spectrum of the
hydrogen-reduced sample is featureless, while the (ii) and (iii) lines
also do not exhibit significant absorbance. However, the samples
with higher catalyst/glycerol ratios exhibit a gold plasmon reso-
nance (PR) band at 550 nm; the intensity of the band increases from
the (iv) to the (vi) line. These results correlate with the XRD results
of the corresponding samples. If the gold particles are significantly
smaller (<10 nm) than the wavelength of light (300–800 nm), the
plasmon band is not dependent on the particle size because it is
dipolar-type plasmon resonance [41–45]. Hence, the differences
in the PR peak shape could be related to the shape/surface mor-
phology of the formed gold nanoparticles. It has been mentioned
that, based on the broadening of the Bragg reflection at the 2ꢀ
value of 38.2◦, the sizes of the gold particles vary with the cat-
alyst/glycerol ratio. Therefore, the slight differences in the shape
of the plasmon resonance band of the glycerol-reduced samples
((ii) to (vi) lines in Fig. 4a) suggest that the catalyst/glycerol ratio
could affect the shapes of the gold nanoparticles formed on the
ceria surface. During synthesis of the catalyst, an increased cata-
lyst/glycerol ratio results in less separation of the catalyst particles,
which increases the potential for the growth of the nanocrystals by
an aggregative mechanism. In contrast, if there is a large amount
of glycerol, the frequency of collisions between the ceria nanopar-
ticles should decrease due to their increased separation thereby
reducing the growth of gold nanoparticles. To verify this hypoth-
esis, we performed the CRG process by thoroughly mixing the 5%
Au/CeO2 sample with glycerol (30 min vigorous stirring) and then
stopping the stirring and letting the reaction stand for the remain-
der of 48 h. The resultant sample was violet (DAv. Au ∼6 nm) in
contrast to the sample obtained with continuous stirring for 48 h,
which appears gray (DAv. Au >15 nm). This indicates that effective
collision between ceria particles is important for the growth of gold
nanoparticles supported on ceria during the CRG process. The shape
of the prepared gold nanoparticles also varies, as is evident from the
changes in the PR bands of the gold nanoparticles in their UV–vis
spectra.
(vi)
(v)
(iv)
(iii)
(ii)
(i)
36
38
40
42
44
2 /degree
Fig. 2. Focused view of XRD patterns of 0.95 wt.% Au/CeO2 sample in the two theta
range 35–45◦ by H2-reduction ((i) line) and glycerol reduction with catalyst to glyc-
erol weight ratios of 0.002, 0.005, 0.01, 0.02, and 0.04, respectively (from (ii) to (vi)
lines).
observed for all glycerol-reduced samples. As determined from
the broadening of the (1 1 1) reflection in the XRD patterns, the
estimated gold particle sizes are plotted in Fig. 3a. As the cata-
lyst/glycerol ratio increases, the FWHM values gradually decrease
(not shown) and the estimated average gold particle sizes gradually
increase. Accordingly, the hydrogen-reduced sample is dark brown,
while the glycerol-reduced samples are violet-brown.
The observed variation of the average gold particle size with
the catalyst/glycerol ratio was further confirmed by reducing the
12
(a)
11
10
9
8
7
6
5
4
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
-2
Catalyst / Glycerol (x10
)
14
12
10
8
(b)
The hydrogen-reduced 3% Au/CeO2 sample also exhibits a fea-
tureless spectrum ((i) line, Fig. 4b). In contrast, the glycerol-reduced
sample (catalyst/glycerol weight ratio of 0.002, (ii) line) shows a
clear absorbance at ∼550 nm. Upon increasing the catalyst/glycerol
weight ratio, a red shift is noted up to line (iv). Further increasing
the catalyst/glycerol ratio leads to broadening of this band and gains
in the absorbance at longer wavelengths (∼700 nm), which implies
further changes in the sizes and shapes of the gold nanoparticles.
An additional band at ∼710 nm has been used as an indicator of the
6
4
0
1
2
3
4
5
-2
Catalyst / Glycerol (x10
)
Fig. 3. Variation of average gold particle size with catalyst to glycerol ratio: (a) 2.74%
Au/CeO2 and (b) 4.59% Au/CeO2.

264
P. Lakshmanan et al. / Applied Catalysis A: General 468 (2013) 260–268
(c)
(vi)
(v)
48h
40h
32h
24h
(iv)
20h
12h
8h
0h
(iii)
(ii)
(i)
300
400
500
600
700
800
450 500 550 600 650 700 750 800
Wave Length (n.m)
(b)
Fig. 5. Time-dependant UV–vis spectroscopy study during the chemical reduction
process with 4.59% Au/CeO2.
catalyst/glycerol ratio of 0.002 ((ii) line in Fig. 4c) displays a band
centered at ∼550 nm. With increasing catalyst/glycerol ratio, the
absorbance increases significantly and undergoes a red shift ((iii)
and (iv) lines). Further increasing the catalyst/glycerol ratio leads
to drastic changes in the line shape including broadening and
increased absorption in the longer wavelength region. However,
no significant band at ∼700 nm is evident, which suggests that,
despite the increased average size, very large aggregated gold par-
ticles do not form. As mentioned earlier, the significant differences
in the absorbance of the (ii) to (iv) lines are not related to size
increases since the wavelength of light used in the present study
(i.e., 450–800 nm) is larger than the sizes of the gold particles.
Hence, such significant differences could be related to significant
shape variations of the gold nanoparticles since the gold loading is
significantly higher. This conclusion is strongly supported by the
XRD results. Within the limitations of the XRD results, there is a
significant difference in the average sizes of the glycerol-reduced
samples, which could explain the shape variation and hence the
different shapes of the gold plasmon resonance band for each sam-
ple. The 5% Au/CeO2 samples clearly show that the CRG technique
can form larger gold particles than the hydrogen-reduction tech-
nique at 300 ◦C. The significant differences between the (ii) to (iv)
lines also suggest that the amount of glycerol affects the shape and
size of the gold particles with a catalyst with fixed gold loading.
The results obtained via UV–vis spectroscopy correlate well with
the XRD results and the colors of the catalysts.
(vi)
(v)
(iv)
(iii)
(ii)
(i)
300
400
500
600
700
800
(a)
(vi)
(v)
(iv)
(iii)
(ii)
(i)
300
400
500
600
700
800
Wave Length (nm)
To optimize the reaction time of the chemical-reduction proce-
dure, time-dependant UV–vis spectroscopy was performed (Fig. 5).
The 5% Au/CeO2 sample is initially yellow, which indicates the pres-
ence of unreduced gold in the as-prepared DPU sample. At this
stage, a UV–vis spectrum was recorded (0 h line). Spectra were then
recorded at specific time intervals until a stable gold plasmon res-
onance band was obtained. At 8 h, the yellow color disappears and
the sample shows a featureless UV spectrum. Note that the spectra
recorded at 8 h and 12 h resemble those of the hydrogen-reduced
samples; this suggests the formation of small gold nanoparticles,
which correlates with the light-brown color of the sample. The
spectrum recorded at 20 h shows a PR band centered at ∼550 nm.
Over time, the color of the sample changes from brown to violet.
A corresponding red shift in the UV–vis spectra reveals that the
shape of the gold nanoparticles changes. After a stable PR band
was attained at 48 h, stirring was stopped and the sample was
used for the catalytic run. During the chemical-reduction process,
aliquots of this sample were removed to examine the gold in the
glycerol. The aliquots were centrifuged to separate the liquid and
tested for the presence of gold via ICP: Gold was not detected
in any of the samples removed at different time intervals. The
Fig. 4. UV–vis absorption spectra of CeO2-supported gold nanoparticles: (a) 0.95%
Au/CeO2, (b) 2.74% Au/CeO2 and (c) 4.59% Au/CeO2. The first line (i) from the bottom
corresponds to the samples reduced by H2 and all other lines (ii–vi) correspond to
the samples reduced by glycerol with catalyst to glycerol weight ratios of 0.002,
0.005, 0.01, 0.02, and 0.04, respectively.
extent of aggregation of nanoparticles and can be attributed to the
excitation of longitudinal surface plasmon vibrations due to close-
packing of gold nanoparticles [46–48]. Careful comparison of the
spectra of the glycerol-reduced 3% Au/CeO2 samples show that the
absorbance is slightly higher around 700 nm (indicated by arrow)
for samples with a higher catalyst/glycerol ratio ((v) and (vi) lines);
however, there is no significant band. This indicates that varying
the amount of glycerol may slightly change the average particle
size, but there is no significant aggregation. It is worth noting the
color differences: The sample pertaining to the (iv) line appears
gray, which indicates that it contains larger gold nanoparticles.
Fig. 4c shows the UV–vis spectra of the 5% Au/CeO2 samples.
The hydrogen-reduced sample exhibits a very broad line shape,
while the glycerol-reduced samples clearly show absorbance due
to gold surface plasmon resonance. The sample reduced with a

P. Lakshmanan et al. / Applied Catalysis A: General 468 (2013) 260–268
265
combined results of UV–vis spectroscopy, XRD, and colors of the
samples clearly show that glycerol acts as a mild reducing agent
for gold-ceria DPU samples, resulting in the preparation of ceria-
supported gold nanoparticles in an environmentally friendly and
facile manner. The use of glycerol to reduce gold provides a single-
reagent method for the synthesis of supported gold nanoparticles.
A schematic illustration of chemical reduction process is shown in
Scheme 1. In the first step of the DPU process, gold is loaded onto the
ceria nanoparticles as a gold-urea complex. This as-prepared yellow
powder is dispersed in glycerol and stirred for 48 h. After nucleation
and growth, the reaction mixture containing the finished catalyst
is diluted with excess water and centrifuged to remove glycerol (3
times) and then mixed with the reaction mixture.
A representative HRTEM image of pure ceria support is shown
in Fig. 6a. The inter-planar spacing of ∼0.3 nm corresponds to the
(1 1 1) plane of the fluorite cubic structure of cerium oxide. From
the TEM analyses, we determined that the cerium oxide particles
have an average size of ∼8 nm, which is in good agreement with the
particle size measured by XRD. Fig. 6b shows the HRTEM image of
glycerol-reduced 5% Au/CeO2, which reveals the presence of Au par-
ticles in close contact with CeO2. The inter-planar spacing of CeO2
(∼0.3 nm), which corresponds to the (1 1 1) plane of the cerium
oxide, could be distinguished from that of Au (∼0.21 nm), which
corresponds to the (2 0 0) plane of gold. It is known that conven-
tional TEM cannot reveal small gold particles on ceria that has a
high surface area [48,49] due to the poor contrast between Au and
CeO2 [50,51]. Despite the difficulty of TEM measurement in high
magnification, we conducted the TEM analysis in a low magnifica-
tion to confirm the particle size of Au nanoparticle on CeO2. Fig. 7
shows the low magnification of TEM images and particle size dis-
tributions of gold nanoparticles corresponding to the 3% Au/CeO2
(a) and 5% Au/CeO2 (b) samples reduced by glycerol with a catalyst
to glycerol ratio of 0.04. The individual particles of Au have spher-
ical shapes and are well dispersed with average sizes of 6–7 nm in
3% Au/CeO2, and 7–8 nm in 5% Au/CeO2, respectively. This result
matches well with the particle size measured by XRD in Table 1.
From histograms of samples, even in the presence of gold particles
Fig. 6. HR-TEM images of (a) pure ceria support and (b) 5% Au/CeO2.
Fig. 7. TEM images and particle size distribution of (a) 3% Au/CeO2 and (b) 5% Au/CeO2 samples reduced by glycerol with the catalyst to glycerol ratio of 0.04.

266
P. Lakshmanan et al. / Applied Catalysis A: General 468 (2013) 260–268
Table 1
Product distribution in the oxidation of glycerol using Au/CeO2 catalysts.
Catalysts
Conv. (%)
Selectivity (%)
LAC
Particle sizes (nm)g
GLY
TAR
OXA
GLYC
FOR
a
b
c
d
a
e
a
f
1% Au/CeO2
1% Au/CeO2
1% Au/CeO2
1% Au/CeO2
3% Au/CeO2
3% Au/CeO2
5% Au/CeO2
5% Au/CeO2
99.1
97.8
98.5
98.0
98.0
98.0
98.0
98.0
73.1
75.8
77.1
83.0
73.5
79.7
72.2
79.0
24.8
21.4
19.8
15.7
22.7
18.4
24.5
18.3
1.4
2.1
1.3
1.1
2.07
0.9
2.0
1.0
0.3
–
0.2
–
0.3
–
0.25
0.5
0.6
0.6
0.2
–
0.8
–
0.26
0.3
–
n.d.
3.0
4.0
6.0
<5
6.2
<5
8.2
0.7
0.2
1.22
0.2
0.7
0.2
0.8
Reaction conditions: temperature of 90 ◦C, pressure of 1 atm under air with a flow rate of 0.15 L min−1, and NaOH/glycerol = 4:1 (mole ratio). n.d.: not determined. LAC:
lactic acid; GLY: glyceric acid; TAR: tartronic acid; OXA: oxalic acid; GLYC: glycolic acid; FOR: formic acid. The carbon balance: a98%, b94%, c99%, d94%, e98%, f97%, for
hydrogen-reduced 3% Au/CeO2 and 5% Au/CeO2, 93% and 95%, respectively.
a
H2 reduced samples.
b–f
Glycerol reduced samples with catalyst to glycerol weight ratios of 0.005 (b), 0.01 (c), 0.02 (d) and 0.04 (e and f), respectively.
Particle size of Au was confirmed by XRD.
g
with size below 4 nm, it can be seen that majority of the particles
lies between 5 nm and 8 nm in particles size. It demonstrates that
glycerol reduction can lead to the formation of larger gold parti-
cles above 5 nm, which is not possible with the hydrogen-reduced
samples.
observations are consistent with literature reports [15,32] that gold
catalyst, base, and oxygen are required for the oxidation reaction to
occur. According to literature, several factors such as gold particle
size, nature of the support, and the reaction conditions (i.e., pres-
ence/absence of base, metal/glycerol ratio, reaction temperature,
and oxygen pressure) influence the catalytic performance (i.e., con-
version and selectivity) toward glycerol oxidation reactions. In the
present study, the nature of the support and the reaction conditions
were held constant. All reactions were carefully performed under
identical conditions to validate the comparison. The average sizes
of the gold particles were changed by reducing the as-prepared cat-
alysts in different manners. Therefore, we attempt to correlate the
catalytic results with changes in the sizes of the gold particles.
In the case of hydrogen-reduced 1% Au/CeO2, the selectivities
to lactic acid and glyceric acid were found to be 73.1% and 24.8%
respectively (Table 1). In the case of hydrogen-reduced 3% and
5% Au/CeO2 samples, nearly the same values of selectivities were
obtained (72.5%, 22.8% for 3% Au/CeO2 and 72.2%, 24.5% for 5%
3.2. Catalytic activity toward the selective oxidation of glycerol
The oxidation of glycerol leads to the formation of a mixture
of acid products (Table 1). In the present study, lactic (LAC) and
glyceric (GLY) acids are the main products, while oxalic (OXA),
tartronic (TAR), glycolic (GLYC), formic (FOR), and acetic acids are
produced in lesser amounts (see Supporting information for the
HPLC chromatographs of a typical reaction mixture (Fig. S1) and
the mixtures of authentic compound (Fig. S2)). In the presence of
excess base, the reaction proceeds well, while in the absence of
base, no reaction occurs. Without the gold catalyst, neither NaOH
nor pure ceria support could catalyze the glycerol conversion. These
Scheme 2. Proposed reaction pathways of glycerol oxidation leading to glyceric and lactic acids.

P. Lakshmanan et al. / Applied Catalysis A: General 468 (2013) 260–268
267
Fig. 8. Variation of lactic acid to glyceric acid mole ratio with catalyst to glycerol
weight ratio during catalyst reduction. (a) 0.95% Au/CeO2 (b) 2.74% Au/CeO2 and
(c) 4.59% Au/CeO2–triangles. Reaction conditions: temperature of 90 ◦C, pressure
of 1 atm under air with a flow rate of 0.15 L min−1, and NaOH/glycerol = 4:1 (mole
ratio).
Au/CeO2). These results are consistent with the fact that, the gold
particle size is roughly the same, whatever the gold loading (see
the first lines (i) in the XRD patterns of Fig. 1a–c). Samples reduced
under flowing hydrogen (300 ◦C/2 h) invariably produced small
gold particles (<5 nm, invisible in the XRD analyses). In contrast, the
chemically reduced samples exhibited gradually increasing aver-
age gold particle sizes due to the increasing catalyst/glycerol ratio
during the reduction of the as-prepared catalysts (as seen by com-
paring the 1st XRD line in Fig. 2 with the other lines), which may
influence the product distribution.
To elucidate the efficiency of gold catalysts toward the selective
oxidation of glycerol to lactic acid, the product mole ratio of lactic
acid to glyceric acid is plotted against the catalyst/glycerol weight
ratio during the chemical reduction step (Fig. 8). As is evident from
Fig. 8, increasing catalyst/glycerol ratio during the reduction of the
gold catalysts leads to a progressive increase in the ratio of lactic
acid to glyceric acid obtained with 1% Au/CeO2 catalyst. In other
words, as the average gold particle size increases, the selectivity to
glyceric acid decreases and the selectivity to lactic acid increases.
However, 3% and 5% Au/CeO2 catalysts exhibit a volcano-shaped
curve with maximum selectivity to lactic acid with medium-sized
gold particles; both smaller and larger gold particles show infe-
rior selectivity. Interestingly, the fall in selectivity to glyceric acid
does not increase the selectivity to the other acid products that are
derived from glyceric acid oxidation except lactic acid [26]. This
observation correlates with literature reports [26,33–36] and sug-
gests that the increased selectivity to lactic acid could be related
to decreased activity of catalysts with larger gold particle sizes
toward glycerol to glyceric acid oxidation. The catalytic results of
the present study closely resemble those of Baiker et al. [36]. As the
Au particle size of 1% Au/CeO2 increases, the selectivity to lactic
acid increases to a maximum level. In 3% and 5% Au/CeO2 samples,
optimum selectivity was obtained with a medium particle size, and
smaller and larger gold particles lead to lower lactic acid selec-
tivity. According to Davis et al., the selective oxidation of glycerol
in the aqueous phase (with base) with supported gold catalysts
proceeds through a reaction path that involves both solution-
mediated (basic) and metal- (i.e., gold-) catalyzed steps [53]. In
the present study, the metal-catalyzed aspect is considered sep-
arately (by maintaining the same basicity of the solution), which
enables the effect of the average gold particle size to be determined.
The decreasing selectivity to glyceric acid with increasing average
gold particle size is also in agreement with the literature reports
Fig. 9. Reusability results of (a) 0.95% Au/CeO2, (b) 2.74% Au/CeO2, (c) 4.59% Au/CeO2
(catalyst to glycerol weight ratios of 0.005, 0.02 and 0.02, respectively). Reaction
conditions: temperature of 90 ◦C, pressure of 1 atm under air with a flow rate of
0.15 L min−1, and NaOH/glycerol = 4:1 (mole ratio).
[26,33–36,49–52]. Accordingly, the increase in the selectivity to
lactic acid may be explained by the solution-mediated aspect of
the catalysis, i.e., lactic acid selectivity could be more related to the
base-catalyzed steps of the reaction (Scheme 2).
The glycerol oxidation pathways that lead to glyceric and lac-
tic acids are different [32]. According to Liu et al., the formation of
glyceric acid involves the gold-catalyzed transformation of glycerol
into glyceraldehyde and dihydroxyacetone as intermediates and
further gold-catalyzed oxidation of these intermediates to glyceric
acid [32]. In contrast, the formation of lactic acid from the glycer-
aldehyde and dihydroxyacetone intermediates proceeds through
base-catalyzed benzilic acid rearrangement and does not appear to
involve the gold catalyst. Hence, the gold particle size is expected
to have a more significant effect on glyceric acid formation than
on lactic acid formation. Literature reports have shown that the
catalytic activity decreases with increasing gold particle size for
glycerol to glyceric acid oxidation. The present study shows that
the glyceric acid selectivity also decreases with increasing gold par-
ticle size. In other words, as the decreased activity of catalysts with
larger gold particles toward the conversion of glyceraldehydes and
dihydroxyacetone into glyceric acid may facilitate base-catalyzed
benzilic acid rearrangement to lactic acid since the reaction mix-
ture contains four equivalents of base (in the present study). Thus,

268
P. Lakshmanan et al. / Applied Catalysis A: General 468 (2013) 260–268
these catalytic results show that lactic acid can be efficiently pro-
duced from glycerol using Au/CeO2 catalysts containing suitably
sized gold particles. Furthermore, the capacity of glycerol to act
as a reducing agent enables the synthesis of Au/CeO2 catalysts
with larger gold particles; this presents a single-reagent method of
controlling the size of ceria-supported gold particles. Thus, the pre-
pared gold nanoparticles exhibit unprecedented catalytic activity
toward the selective oxidation of glycerol at atmospheric pressure.
The performance of glycerol-reduced catalysts in reusability
tests is presented in Fig. 9. The 1% Au/CeO2 catalyst (reduced with a
catalyst/glycerol ratio of 0.005, Fig. 9a) shows quite stable conver-
sion and lactate selectivity during the reusability tests. The results
obtained for 3% Au/CeO2 (Fig. 9b) shows that the lactate selectivity
decreases gradually with the number of cycles: A fall in the lac-
tate selectivity from 74% to 58% was accompanied with decreasing
conversion from 98% to 80% after four successive runs. To assess
gold leaching, aliquots of the reaction mixture were analyzed with
inductively coupled plasma atomic emission spectroscopy (ICPAES)
after each run. The results showed no detectable leaching of gold,
which demonstrates the stability of Au/CeO2 samples against leach-
ing during the catalytic reactions.
[3] R.M. West, M.S. Holm, S. Saravanamurugan, J. Xiong, Z. Beversdorf, E. Taarning,
C.H. Christensen, J. Catal. 269 (2010) 122–130.
[4] J. Lunt, Polym. Degrad. Stab. 59 (1998) 145–152.
[5] K. Fukushima, Y. Kimura, Polym. Int. 55 (2006) 626–642.
[6] H. Kimura, K. Tsuto, T. Wakisaka, Y. Kazumi, Y. Inaya, Appl. Catal. A: Gen. 96
(1993) 217–228.
[7] A. Abbadi, H.V. Bekkum, Appl. Catal. A: Gen. 148 (1996) 113–122.
[8] R. Garcia, M. Besson, P. Gallezot, Appl. Catal. A: Gen. 127 (1995) 165–176.
[9] D.I. Enache, D. Barker, J.K. Edwards, S.H. Taylor, D.W. Knight, A.F. Carley, G.J.
Hutchings, Catal. Today 122 (2007) 407–411.
[10] A. Abad, C. Almela, A. Corma, H. Garcia, Chem. Commun. (2006) 3178–3180.
[11] T. Mallat, A. Baiker, Chem. Rev. 104 (2004) 3037–3058.
[12] S. Carrettin, P. McMorn, P. Johnston, K. Griffin, G.J. Hutchings, Chem. Commun.
(2002) 696–697.
[13] S. Carrettin, P. McMorn, P. Johnston, K. Griffin, C.J. Kiely, G.A. Attard, G.J. Hutch-
ings, Top. Catal. 27 (2004) 131–136.
[14] S. Carrettin, P. McMorn, P. Johnston, K. Griffin, C.J. Kiely, G.J. Hutchings, Phys.
Chem. Chem. Phys. 5 (2003) 1329–1336.
[15] I.H. Kishida, F. Jin, Z. Zhou, T. Moriya, H. Enomoto, Chem. Lett. 34 (2005)
1560–1562.
[16] D. Roy, B. Subramaniam, R.V. Chaudhari, ACS Catal. 1 (2011) 548–551.
[17] C.A. Ramirez-Lopez, J.R. Ochoa-Gomez, M. Fernandez-Santos, O. Gomez-
Jimenez-Aberasturi, A. Alonso-Vicario, J. Torrecilla-Sori, Ind. Eng. Chem. Res.
49 (2010) 6270–6278.
[18] E.P. Maris, R.J. Davis, J. Catal. 249 (2007) 237–328.
[19] E.P. Maris, W.C. Ketchie, M. Murayama, R.J. Davis, J. Catal. 251 (2007) 281–294.
[20] J.C. Sowden, E.K. Pohlen, J. Am. Chem. Soc. 80 (1958) 242–244.
[21] G. Machell, G.N. Richards, J. Chem. Soc. (1960) 1924–1931.
[22] W.C. Ketchie, M. Murayama, R.J. Davis, J. Catal. 250 (2007) 264–273.
[23] P. Gallezot, Catal. Today 37 (1997) 405–418.
4. Conclusions
[24] M. Besson, P. Gallezot, Catal. Today 57 (2000) 127–141.
[25] F. Porta, L. Prati, J. Catal. 224 (2004) 397–403.
For the first time, the catalytic activity of ceria-supported gold
nanoparticle catalysts prepared by chemical reduction with glyc-
erol toward the selective aerobic oxidation of glycerol to lactic
acid was explored. The results were compared with those of con-
ventional hydrogen-reduced samples. The glycerol-based chemical
reduction method not only enables the synthesis of larger gold par-
ticles (compared with hydrogen-reduced samples) on a nanosized
ceria support, but also allows variation of the average gold par-
ticle sizes with fixed gold loading and without the need for other
reagents. The catalyst/glycerol weight ratio affects the average gold
particle size and, hence, the product distribution. This study reveals
that Au/CeO2 catalysts can efficiently catalyze the selective oxida-
tion of glycerol to lactic acid at atmospheric pressure with higher
selectivity to lactic acid than hydrogen-reduced samples.
[26] C.L. Bianchi, P. Canton, N. Dimitratos, F. Porta, L. Prati, Catal. Today 102–103
(2005) 203–212.
[27] N. Dimitratos, C. Messi, F. Porta, L. Prati, A. Villa, J. Mol. Catal. A:Chem. 256
(2006) 21–28.
[28] S. Demirel-Glen, M. Lucas, P. Claus, Catal. Today 102–103 (2005) 166–172.
[29] S. Demirel, K. Lehnert, M. Lucas, P. Claus, Appl. Catal. B: Environ. 70 (2007)
637–643.
[30] S. Hirasawa, H. Watanabe, T. Kizuka, Y. Nakagawa, K. Tomishige, J. Catal. 300
(2013) 205–216.
[31] M. Checa, F. Auneau, J. Hidalgo-Carrillo, A. Marinas, J.M. Marinas, C. Pinel, F.J.
Urbano, Catal. Today 196 (2012) 91–100.
[32] Y. Shen, S. Zhang, H. Li, Y. Ren, H. Liu, Chem. Eur. J. 16 (2010) 7368–7371.
[33] C. Bianchi, F. Porta, L. Prati, M. Rossi, Top. Catal. 13 (2000) 231–236.
[34] A. Villa, C.E. Chan-Thaw, L. Prati, Appl. Catal. B: Environ. 96 (2010) 541–547.
[35] N. Dimitratos, J.A. Lopez-Sanchez, D. Lennon, F. Porta, L. Prati, A. Villa, Catal.
Lett. 108 (2006) 147–153.
[36] P. Haider, B. Kimmerle, F. Krumeich, W. Kleist, J.D. Grunwaldt, A. Baiker, Catal.
Lett. 125 (2008) 169–176.
[37] L. Delannoy, K. Fajerwerg, P. Lakshmanan, C. Potvin, C. Methivier, C. Louis, Appl.
Catal. B: Environ. 94 (2010) 117–124.
[38] Q. Fu, S. Kudriavtseva, H. Saltsburg, M.F. Stephanopoulos, Chem. Eng. J. 93
(2003) 41–53.
Acknowledgments
This work was supported by the Institutional Research Pro-
gram (SI-1301) and partially funded by Korea Research Council
for Industrial Science & Technology (ISTK, SK-1311). P. L. acknowl-
edges Korea Research Institute of Chemical Technology (KRICT)
for financial support through a post-doctoral program. Authors
would like to thank to Prof. Harold H. Kung for his wonderful
discussion.
[39] J.J. Storhoff, A.A. Lazarides, R.C. Mucic, C.A. Mirkin, R.L. Letsinger, G.C. Schats, J.
Am. Chem. Soc. 122 (2000) 4640–4650.
[40] R.R. Reynolds, C.A. Mirkin, R.L. Letsinger, J. Am. Chem. Soc. 122 (2000)
3795–3796.
[41] M. Sastry, N. Lala, V. Patil, S.P. Chavan, A.G. Chittiboyina, Langmuir 14 (1998)
4138–4242.
[42] R. Elghanian, J.J. Starhoff, R.C. Mucic, R.L. Letsinger, C.A. Mirkin, Science 277
(1997) 1078–1081.
[43] Y. Borensztein, L. Delannoy, A. Djedidi, R.G. Barrera, C. Louis, J. Phys. Chem. C
114 (2010) 9008–9021.
[44] L.D. Landau, E.M. Lifshitz, Electrodynamic of Continuous Media, Pergamon
Press, New York, 1960.
Appendix A. Supplementary data
[45] U. Kreibig, M. Vollmer, Optical Properties of Metal Clusters, Springer, New York,
1994.
Supplementary data associated with this article can be
found, in the online version, at http://dx.doi.org/10.1016/j.apcata.
2013.08.048.
[46] B. Palpant, B. Prevel, J. Lerme, E. Cottancin, M. Pellarin, M. Treilleux, A. Perez,
J.L. Vialle, M. Broyer, Phys. Rev. B 57 (1998) 1963–1970.
[47] V. Myroshnychenko, J. Rodriguez-Fernandez, I. Pastoriza-Santos, A. Funston,
M.C. Novo, P. Mulvaney, L.M. Liz-Marzan, F.J.G. De Abajo, Chem. Soc. Rev. 37
(2008) 1792–1805.
[48] C.G. Blatchford, J.R. Campbell, J.A. Creighton, Surf. Sci. 120 (1982) 435–455.
[49] N. Dimitratos, J.A. Lopez-Sanchez, G.J. Hutchings, Top. Catal. 52 (2009) 258–268.
[50] L. Prati, M. Rossi, J. Catal. 176 (1998) 552–560.
References
[1] S. Ebnesajjad, Handbook of Biopolymers and Biodegradable Plastics, Applied
Science Publishers, 2012.
[2] P.P. Upare, Y.K. Hwang, J.S. Chang, D.W. Hwang, Ind. Eng. Chem. Res. 51 (2012)
4837–4842;
[51] F. Porta, L. Prati, M. Rossi, S. Colluccia, G. Martra, Catal. Today 60 (2000) 165–172.
[52] L. Prati, G. Martra, Gold Bull. 32 (1999) 96–101.
[53] B.N. Zope, D.D. Hibbitts, M. Neurock, R.J. Davis, Science 330 (2010) 74–78.
Y. Fan, C. Zhou, X. Zhu, Catal. Rev. Sci. Eng. 51 (2009) 293–324.