Oxidative esterification of ethylene glycol in methanol to form methyl glycolate over supported Au catalysts

Article abstract of DOI:10.1039/c4cy00556b

Au/ZnO and Au/Al₂O₃ catalysts with various mean Au particle diameters (2.0-7.4 nm) were prepared by the deposition of pre-formed Au colloids. These catalysts were evaluated in the oxidative esterification of ethylene glycol to methyl glycolate. The results show that the catalytic activity per surface Au atom is independent of Au particle diameter in the range of 3-7.4 nm, whereas smaller Au particles (~2.0 nm) show an inferior activity. This behavior was observed on both Au/ZnO and Au/Al₂O₃ catalysts. This observed correlation between activity and Au particle diameter confirms the assertion that only exposed atoms are catalytically active. We prepared gold nanoparticles with a uniform mean diameter of ~3 nm loaded on various supports, i.e. ZnO, Al₂O₃, SiO₂, TiO₂ and CeO₂. Among these five catalysts, Au/ZnO gave the best catalytic activity in the reaction followed by Au/Al₂O ₃. Au/SiO₂, Au/TiO₂ and Au/CeO₂ gave significantly lower activities. The variation in catalytic behavior of these gold catalysts on different supports originates from differences in the anchoring of the supported Au particles, the gold oxidation state, the gold-support interaction, and the acidity of the support. the Partner Organisations 2014.

Full text of DOI:10.1039/c4cy00556b

Catalysis
Science &
Technology
View Article Online
View Journal | View Issue
PAPER
Oxidative esterification of ethylene glycol in
methanol to form methyl glycolate over
Cite this: Catal. Sci. Technol., 2014,
supported Au catalysts†
4, 3141
Yi-Hu Ke, Xiao-Xia Qin, Chun-Ling Liu, Rong-Zhen Yang and Wen-Sheng Dong*
2 3
Au/ZnO and Au/Al O catalysts with various mean Au particle diameters (2.0–7.4 nm) were prepared by
the deposition of pre-formed Au colloids. These catalysts were evaluated in the oxidative esterification of
ethylene glycol to methyl glycolate. The results show that the catalytic activity per surface Au atom is
independent of Au particle diameter in the range of 3–7.4 nm, whereas smaller Au particles (~2.0 nm)
2 3
show an inferior activity. This behavior was observed on both Au/ZnO and Au/Al O catalysts. This
observed correlation between activity and Au particle diameter confirms the assertion that only exposed
atoms are catalytically active. We prepared gold nanoparticles with a uniform mean diameter of ~3 nm
loaded on various supports, i.e. ZnO, Al O , SiO , TiO and CeO . Among these five catalysts, Au/ZnO
Received 30th April 2014,
Accepted 27th May 2014
2
3
2
2
2
2 3 2 2 2
gave the best catalytic activity in the reaction followed by Au/Al O . Au/SiO , Au/TiO and Au/CeO gave
significantly lower activities. The variation in catalytic behavior of these gold catalysts on different
supports originates from differences in the anchoring of the supported Au particles, the gold oxidation
state, the gold–support interaction, and the acidity of the support.
DOI: 10.1039/c4cy00556b
www.rsc.org/catalysis
towards a greener commercial process for the clean and
efficient production of carboxylic esters. Since then, the
1
. Introduction
Many methyl esters are industrially important chemicals,
which can be used as solvents, extractants, diluents, and
intermediates or as products in the perfume industry. The
oxidative esterification of various alcohols, e.g. primary alcohols,
aromatic and aliphatic primary alcohols, 1,2-propanediol,
glycerol, etc. or aldehydes has been reported by several
1
8
–14
conventional route for the production of methyl esters involves
a two-step synthetic procedure in which the synthesis of
carboxylic acids or activated carboxylic acid derivatives such
research groups.
Au catalysts supported on various
supports, e.g. β-Ga O , TiO , CeO , and Fe O , have been
2
3
2
2
2
3
found to exhibit catalytic activity in the process. Although a
2
,3
as acid anhydrides or chlorides is first required. The direct
formation of methyl esters from the oxidation of primary
alcohols in methanol is also well known. However, these
methods generally employ stoichiometric oxidants such as
iodine, trichloroisocyanuric acid, (diacetoxy)-iodobenzene or
previous study found that β-Ga O containing increased
2
3
acid sites was much more active and selective than
8
conventional oxide-supported Au catalysts, the exact nature
of the support effect still needs to be explored further,
because a deep understanding of the role of the support may
lead to significant improvements in supported Au catalysts
for the oxidative esterification of alcohols.
4–6
hypochlorite reagents. Hence, the use of facile, cost-effective,
and environmentally friendly procedures with air as the
oxidant together with heterogeneous catalysts will avoid the
use of erosive acid catalysts and a large excess of reagents.
This has attracted substantial interest.
Supported Au nanoparticles have shown unique catalytic
properties in a variety of reactions such as CO oxidation, the
aerobic oxidation of alcohols, propylene epoxidation, hydrogen
peroxide synthesis, hydrogenation reactions, C–C coupling
The oxidative esterification of ethylene glycol to methyl
glycolate using air as the oxidant over heterogeneous Au
catalysts was first reported by Hayashi et al. from the Nippon
1
5–17
reactions, etc.
It is generally accepted that the catalytic
behavior of Au depends to a large extent on the size of the Au
particles and the nature of the support, although many other
issues such as the Au–support interaction, the Au particle
shape and the chemical state of Au may also affect the catalytic
performance. For example, for the oxidation of CO it has
been shown both experimentally and theoretically that the
activity of Au particles increases with a decrease in the
particle size and the activity is strongly affected by the
7
Shokubai company. This route represents significant progress
Key Laboratory of Applied Surface and Colloid Chemistry (SNNU), MOE, School of
Chemistry and Chemical Engineering, Shaanxi Normal University, Xi'an 710062,
China. E-mail: wsdong@snnu.edu.cn; Fax: +86 029 81530806;
Tel: +86 029 81530806
†
Electronic supplementary information (ESI) available. See DOI: 10.1039/
c4cy00556b
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 3141–3150 | 3141

View Article Online
Paper
Catalysis Science & Technology
1
8–23
support.
However, for the gas-phase non-oxidative
the corresponding nitrates Zn(NO ) ·6H O and Ce(NO ) ·6H O,
3
2
2
3 3
2
dehydrogenation of ethanol, Hensen et al. found that gold
nanoparticles of about 6 nm showed a much higher activity
than smaller or larger particles. However, in the presence of
oxygen the intrinsic activity was constant at up to about 7 nm,
respectively. The precipitation was carried out by adding sodium
carbonate (0.25 M) to the Zn(NO ·6H O (0.3 M, 82 mL)
solution and an ammonia solution (25%) to the Ce(NO ) ·6H O
solution (0.3 M, 39 mL) at 70 °C under vigorous stirring until
the pH of the mixture was 8. The obtained solution was then
left at 70 °C while stirring for 90 min. The resulting
precipitates were filtered, washed with distilled water, dried
at 110 °C overnight, and subsequently calcined at 350 °C in
air for 4 h.
3
)
2
2
3
3
2
2
4
and it increased for larger particles. For liquid oxidation
2
5–36
reactions, the situation is even more complicated.
A
volcano-type relationship between gold particles and their
activity towards selective ethanol oxidation to acetic acid in
an aqueous solution was reported by Xu et al. They observed
that, among 3–30 nm Au particles, Au particles with an
average diameter of 5 nm showed an activity about 3 times
higher than that of smaller (~3 nm) Au particles and 15 times
The colloidal deposition method used in this study has been
2
1,37,38
extensively described elsewhere.
Gold nanoparticles with
different average diameters were prepared by the reduction of
2
5
higher than that of larger (10–30 nm) Au particles. In the
selective liquid phase oxidation of glycols, Bianchi et al.
found that the activity of Au catalysts supported on oxides
HAuCl (Sinopharm Chemicals (99.99%)) in aqueous solution
4
using polyvinyl alcohol (PVA, Mw 10 000 from Aldrich, 80%
hydrolyzed) as a protecting agent and NaBH4 (Sinopharm
Chemicals) as a reductant at different temperatures. In a
typical preparation, the protecting agent was added (Au/PVA =
(
Al O , TiO ) increased upon a decrease in the Au particle
2 3 2
diameter from 7.2 to 3.7 nm. However, on carbon the
maximum activity was achieved for gold particles with a mean
−
1
−1
2 : 1 mg mg ) to a 200 mg L aqueous gold solution at room
2
6
diameter of around 7–8 nm. For the aerobic oxidation of
benzyl alcohol, Haider et al. found optimal activity for CeO
and TiO supported Au catalysts with a medium particle size
temperature (25 °C) under vigorous stirring, followed by a rapid
2
4 4
injection of a 0.1 M aqueous solution of NaBH (Au/NaBH =
−
1
1 : 4 mol mol ) that led to the formation of a dark-orange–
brown solution, indicating the formation of gold colloids.
The immobilization of Au particles on the supports, i.e.
ZnO, Al O , TiO2, CeO2, was accomplished at room
2
(
ca. 6.9 nm), whereas smaller and bigger particles showed
2
7
inferior activity. For the oxidation of glucose, Comotti et al.
found that the catalytic activity was inversely proportional
to the diameter of colloidal Au particles in the range of
2
3
temperature (25 °C) by adding these supports to the colloidal
gold solution while stirring and they were kept in contact
until total adsorption (1.3 wt% gold on the support) occurred,
as indicated by the decolorization of the solution. Owing to the
different interactions between the metal-oxide particles and
the gold nanoparticles, it took 2 to 24 h for the full deposition
3
3
3
–6 nm. These results indicate that an Au size effect is
apparent in different reactions. An understanding of the effect
of Au particle size and the function of the supports in various
Au-catalyzed reactions will assist in designing highly efficient
Au catalysts.
To understand the particle size and support effects, it is
crucial that only one preparation method is used, as
important variables such as impurities, residues on the
catalyst surface, and particle morphology might be altered
when using different preparation techniques. Hence, in
this work gold nanoparticles with different average diameters
were prepared using polyvinyl alcohol as a protecting agent
of gold colloids onto these supports. For ZnO and TiO
~2 h. For Al , CeO , and SiO , it required ~24 h. In the case
of SiO , the pH of the colloidal solution needed to be
2
, it took
2
O
3
2
2
2
2
5
adjusted to 0.5 with nitric acid during the deposition. The
solids were collected by filtration followed by washing the
solid with 2 L of doubly-distilled water to remove all the
2
7
+
−
dissolved species (e.g. Na , Cl ). Finally, the solids were dried
under vacuum at 100 °C for 12 h.
and NaBH
was changed. The Au particles were deposited onto the ZnO
and Al supports by the colloidal deposition method to
4
as a reductant and only the reaction temperature
2
O
3
2.2 Catalyst characterization
determine the Au size effect. The Au catalysts were supported
on five different supports, ZnO, Al O , SiO , TiO , and CeO ,
The crystal structures of the as-prepared samples were
characterized by powder X-ray diffraction (XRD) using a
Rigaku D/MAX-III X-ray diffractometer (35 kV, 40 mA) with a
CuKα source. The diffraction patterns were obtained between
2
3
2
2
2
with a similar gold particle size to investigate the effect of
the support. From a combined study of these effects, we
identified important structural features of the gold particle
size and the support that control the efficiency of the gold
catalysts in the oxidative esterification of ethylene glycol.
−
1
10 and 80° at a scan rate of 8° min . The specific surface
areas and the pore size distribution of the samples were
obtained by N (−196.2 °C) adsorption measurements using a
2
Micromeritics ASAP 2020M system. Before the measurement,
the samples were outgassed at 300 °C for 6 h. The surface area
2
. Experimental
2
.1 Catalyst preparation
The commercially available powder supports γ-Al O and SiO
2
(
SBET) was calculated using the BET method. The micropore
surface area Smi was obtained using the t-plot method. The
mesoporous surface area (Smeso) was determined by subtracting
Smi from SBET. The pore size distribution was derived from the
2
3
were purchased from Sinopharm Chemical Reagent Co. Ltd.
Shanghai, China). TiO was P25 from Evonik Industries. ZnO
(
2
and CeO were prepared by the precipitation method using
desorption branch of the N isotherms using the BJH method.
2
2
3142 | Catal. Sci. Technol., 2014, 4, 3141–3150
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
The total pore volumes (V ) were estimated considering
6820 GC with an INNOWAX capillary column (30 m × 0.32 mm ×
0.5 μm) and an FID.
p
nitrogen uptake at a relative pressure (P/P ) of ca. 0.99. The
0
average pore diameter (D
area and the total pore volume (D = 4V /SBET). The crystalline
p
) was estimated from the surface
The conversion of ethylene glycol (EG) was calculated
based on the following equation:
p
p
size and microstructure were characterized by transmission
electron microscopy (TEM, JEOL JEM-2100). The operating
voltage was 200 kV. The particle sizes and distribution of the
gold particles were obtained from TEM images by averaging a
minimum of 200 particles. By assuming a spherical model for
the Au nanoparticles, the dispersion of Au (D) was roughly
estimated by the following relationship, D = 1.17/d (nm) ×
moles of converted EG
EG conversion (%) 
100%
moles of initial EG
The selectivity of methyl glycolate (MGC) was calculated
based on the following equation:
2
8
moles of MGC formed
1
00%. X-ray photoelectron spectroscopy (XPS) measurements
MGC selectivity (%) 
100%
moles of converted EG
were performed on a PE PHI-5400 spectrometer with an Al Kα
source (1486.6 eV). The binding energy was calibrated relative
to the C1s peak (285.0 eV) of the contaminant carbon. The 3. Results
accuracy of the BE values was ±0.1 eV. The high-resolution
3.1 Characterization of the supports
scans were performed over the 78–96 eV (Au 4f spectrum)
for the tested samples. The peak area was estimated by
calculating the integral of each peak after subtracting an
S-shaped background and fitting the experimental peak to a
combination of Lorentzian/Gaussian lines of variable proportions.
The Au content was determined using an inductively coupled
plasma atomic emission spectrometer (ICP-AES, model Atom
Scan 16, TJA Corp.).
The surface area, pore volume and average pore diameters of
the five supports are shown in Table 1. XRD patterns of the
ZnO, Al
well-crystallized diffraction peaks while the diffraction peaks
of CeO are not sharp, indicating that this support is not
well crystallized.
2 3 2 2
O , SiO and TiO supports in Fig. S1 (ESI†) show
2
3
.2 Au particle size effect
Four different sizes of Au colloids were prepared by the
reduction of HAuCl in aqueous solution using polyvinyl
H temperature programmed reduction (H -TPR) experiments
2
2
were carried out using a Micromeritics AutoChem II-2920
instrument. In each trial, about 50 mg of sample was loaded
into a quartz reactor. TPR measurements were performed
4
alcohol as a protecting agent and NaBH as a reductant. We
4
−
1
using 10% hydrogen in argon (30 ml min ) at a heating
found that the reaction temperature was an important factor
in controlling the reduction rate of metal cations, and higher
temperatures gave a larger mean particle size. The reduction
−
1
rate of 10 °C min from room temperature to 900 °C.
Hydrogen consumption was monitored online using a thermal
conductivity detector.
of HAuCl in aqueous solution gave 1.8 nm Au particles at
4
The acidities of the samples were determined by ammonia
0 °C, 2.9 nm at 25 °C, 4.8 nm at 60 °C, and 7.4 nm at 90 °C
(see Fig. S2, ESI†). The Au particles in Fig. S2a–c (ESI†) have
a near-spherical shape with a homogeneous size distribution,
whereas the Au particles prepared at 90 °C (Fig. S2d, ESI†)
have an oval shape with a broad size distribution.
temperature-programmed desorption (NH -TPD) using a
3
Micromeritics AutoChemII-2920 system. A 100 mg sample
was loaded into a sample tube and pretreated in He at
00 °C for 60 min. The sample was cooled, and NH3 was
2
adsorbed at 120 °C for 30 min, and then the sample was
further purged with dry helium at the same temperature
The immobilization of these Au colloids onto the ZnO
support was also studied by HRTEM to verify the sizes of the
Au particles after immobilization. Typical TEM micrographs
and the corresponding Au particle size distributions for the
Au/ZnO catalysts with different mean Au particle sizes are
shown in Fig. 1. It is clear that after immobilization, only
minor amounts of agglomerated particles were detected and
the determination of mean particle size gave similar values to
those obtained for the Au colloids in solution.
to remove weakly adsorbed NH . Finally, the sample was
3
−
1
heated at a linear rate of 10 °C min under dry helium
30 ml min ) up to 700 °C. The effluent was monitored by a
−
1
(
quadrupole mass spectrometer.
2
.3 Reaction and analysis
All the reactions were conducted in a 35 mL stainless steel
autoclave equipped with a magnetic stirrer. In a typical
experiment, 0.9 g of catalyst, 17.4 g of methanol and 3.43 g
of ethylene glycol were added to the autoclave. The autoclave
was then sealed tightly, purged and pressurized to 3.0 MPa
with O2 at room temperature. The reaction proceeded at
Table 2 shows the dependence of catalytic performance on
the mean size of the Au nanoparticles in Au/ZnO. The
loading of Au onto these catalysts was consistent (~1.25 wt%).
High selectivity towards methyl glycolate (93.1–95.3%) was
obtained for all these catalysts, which indicates that the Au
particle size did not affect product selectivity. The product
selectivity was mainly determined by the nature of the
support, which will be explained in section 3.3. Table 2 shows
that the conversion of ethylene glycol depends strongly on the
mean size of the Au nanoparticles. With a decrease in the
1
00 °C for 20 min. After the reaction, the reactor was cooled
to less than 5 °C with an ice–water mixture and depressurized.
An internal standard of n-amyl alcohol was added to the liquid
products. Quantitative analysis was performed using an Agilent
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 3141–3150 | 3143

View Article Online
Paper
Catalysis Science & Technology
Table 1 The physico-chemical properties of various supports
2
−1
2
−1
3
−1
3
−1
Supports
ZnO
S
BET (m g
)
S
meso (m g
)
V
p
(cm g
)
V
meso (cm g
)
p
D (nm)
31.8
25.0
88.4
293.0
46.3
23.0
0.20
0.16
0.93
0.19
0.11
0.197
0.11
0.922
0.19
24.7
5.8
8.7
15.9
16.6
γ-Al
SiO
TiO
CeO
2
O
3
185.9
309.8
48.5
2
2
2
27.9
0.108
Fig. 1 TEM micrographs and the corresponding Au particle size distributions for Au/ZnO catalysts: a) Au(2.0)/ZnO; b) Au(2.9)/ZnO; c) Au(4.9)/ZnO;
d) Au(7.4)/ZnO.
3144 | Catal. Sci. Technol., 2014, 4, 3141–3150
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
Table 2 Catalytic performance of Au/ZnO catalysts with varying mean Au particle sizesa
b
−1
Catalysts
Au loading (wt%)
Mean Au size (nm)
Au dispersion (%)
EG conversion (%)
MGC selectivity (%)
TOF (s
)
Au(2.0)/ZnO
Au(2.9)/ZnO
Au(4.9)/ZnO
Au(7.4)/ZnO
1.25
1.25
1.26
1.25
2.0
2.9
4.9
7.4
58.5
40.3
23.8
15.8
17.8
21.6
12.3
7.6
93.9
94.8
95.3
93.1
0.245
0.432
0.417
0.388
a
b
Reaction conditions: catalyst 0.9 g, methanol 17.4 g, EG = 3.43 g, 3.0 MPa, 100 °C, 20 min. Moles of converted ethylene glycol per mole of
surface gold atoms per second.
mean Au particle diameter from 7.4 to 2.9 nm, the conversion
of ethylene glycol over the catalysts increased. However, the
catalytic activity per surface Au atom is roughly independent
of the diameter of the Au particles in the range of 2.9–7.4 nm
−
1
(
TOF: ~0.4 s ). Au/ZnO with a 2.0 nm mean Au particle
−
1
diameter had the lowest TOF of 0.245 s .
We also prepared four Au/Al O catalysts with different Au
2
3
particle diameters, i.e. 2.0, 3.0, 5.0, and 7.6 nm (see Fig. S3, ESI†),
and determined their catalytic activities in the oxidative
esterification of ethylene glycol to examine whether the Au
size effect varies with the support. The same activity change
trend was observed (see Table S1, ESI†). The TOF value was
−
1
~
0.2 s for the 2.9–7.4 nm Au particles, which is lower than
−
1
2 3
that of Au/ZnO (~0.4 s ). The Au/Al O catalyst with a 2.0 nm
−
1
Fig. 2 Recycling of Au(2.9)/ZnO for the oxidative esterification of
ethylene glycol (reaction conditions: catalyst 0.9 g, methanol 17.4 g,
EG = 3.43 g, 3.0 MPa, 100 °C, 2 h).
mean Au particle diameter gave the lowest TOF of 0.133 s .
To verify whether the oxidative esterification of ethylene
glycol catalyzed by Au/ZnO is heterogeneous or homogeneous,
we investigated Au leaching during the reaction. The
heterogeneous Au(2.9)/ZnO catalyst was filtered while hot
after a 30 min reaction (ca. 28% ethylene glycol conversion)
and the filtrate was allowed to react for a further 1.5 h under
the same conditions. The conversion of ethylene glycol was
of ethylene glycol is shown in Table 3. The activities for the
pure supports were also tested before catalyst evaluation and
showed that under the conditions used no product was
2
9% after that time (i.e. no change). In addition, ICP analysis
detected. The Au nanoparticles supported on TiO
gave much lower activities with 3.4% and 4.3% ethylene
glycol conversion, respectively. The Au/SiO also gave a lower
2 2
and CeO
confirmed that none of the Au species was detected in the
reaction mixtures, confirming that the reaction is heterogeneous.
The recyclability of Au(2.9)/ZnO in the oxidative esterification
of ethylene glycol at 100 °C was also examined. As shown in
Fig. 2, the conversion of ethylene glycol and the selectivity
toward methyl glycolate did not change significantly during
three successful cycles. Additionally, we also examined the
catalytic performance of the Au(2.9)/ZnO catalyst after storage
in a refrigerator for 7 months. We found similar catalytic
activity to that of the fresh catalyst. HRTEM observation
2
ethylene glycol conversion (12.2%). A medium ethylene glycol
conversion (~60.7%) was obtained over the Au/Al O catalyst.
2
3
Among all the supported catalysts, Au(2.9)/ZnO gave the
best catalytic activity with 97.4% ethylene glycol conversion.
Moreover, Au(2.9)/ZnO gave the highest selectivity towards
methyl glycolate followed by Au/Al
2 3 2 2
O , Au/SiO , Au/TiO and
Au/CeO . The main by-products of the reaction over these
2
supported Au catalysts, as determined by GC-MS, are methyl
formate, 1-ethoxyethane-1,2-diol, 1-methoxyethane-1,2-diol,
2,2-dimethoxyethanol, etc. Because it is very difficult to
obtain pure reagents, the amounts of these by-products were
not determined.
(
not shown) confirmed that this catalyst had a mean Au
particle diameter of 3.0 ± 0.51 nm, which is similar to that
of the fresh catalyst, confirming that the Au/ZnO catalyst is
very stable.
Typical TEM micrographs and the corresponding Au
particle size distributions for the fresh Au catalysts are shown
in Fig. 3. The mean diameters of the Au nanoparticles for all
these samples range from 2.9 to 3.3 nm. Fig. S4 (ESI†) shows
TEM micrographs and the corresponding Au particle size
distributions for these Au catalysts after the reaction. By
comparison with fresh Au catalysts, the mean Au particle
diameters and distributions of the corresponding used Au
catalysts increase to some extent. The mean Au particle
3
.3 Support effect
A rigorous study into the effect of the support requires the
use of catalysts with uniform Au particle sizes. Hence, in this
study the supported Au catalysts were prepared by the
deposition of Au colloids with a mean particle diameter of
2
.9 ± 0.3 nm onto ZnO, Al O , SiO , TiO , and CeO supports.
2 3 2 2 2
The performance of these catalysts in the oxidative esterification
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 3141–3150 | 3145

View Article Online
Paper
Catalysis Science & Technology
Table 3 Catalytic performance of Au catalysts supported on different supports
a
b
c
−1
Catalysts
Au loading (wt%)
Au size (nm)
EG conversion (%)
MGC selectivity (%)
TOF (s
)
Au(2.9)/ZnO
1.25 (1.26)
1.35 (1.37)
1.15 (1.13)
1.23 (1.25)
1.31 (1.33)
2.9 (3.0)
3.0 (4.4)
3.2 (5.0)
3.3 (5.9)
3.3 (6.8)
97.4
60.7
12.2
3.4
95.4
91.7
89.7
56.2
33.7
—
—
0.012
0.003
0.0037
Au/Al
Au/SiO
Au/TiO
Au/CeO
2
O
3
2
2
2
4.3
a
Reaction conditions: catalyst 0.5 g, methanol 4.83 g, EG = 0.935 g, 3.0 MPa, 100 °C, 4 h. The values in parentheses are Au loading of the used
b
c
catalysts. The values in parentheses are Au particle sizes of the used catalysts. The moles of converted ethylene glycol per mole of surface
gold atoms per second.
diameter for the fresh Au(2.9)/ZnO catalyst was 2.9 ± 0.5 nm.
After the reaction, it increased slightly to 3.0 ± 0.8 nm. For
not have a reduction peak below 900 °C. Au/ZnO had two
small reduction peaks at 450 and 575 °C. Au/ZnO-300 had a
similar reduction pattern to that of Au/ZnO, but the peak
areas were slightly smaller. XPS analysis confirmed that the
Au in these two Au/ZnO samples was in the metallic state,
and hence the peaks came from the reduction of lattice
oxygen in the catalyst. These results suggest the presence of a
strong interaction between gold and ZnO. Gold may increase
the oxygen mobility and lattice oxygen accessibility of ZnO,
and as a result, the reduction temperature of ZnO would
decrease. Previously, Muhler et al. also found that the
presence of Au particles enhanced the number of exposed
oxygen vacancies in ZnO, presumably located at the interface
Au/Al
was 3.0 ± 0.8 nm and after the reaction it increased to 4.4 ±
.0 nm. For Au/SiO , Au/TiO , and Au/CeO , after the reaction
2 3
O , before the reaction the mean Au particle diameter
1
2
2
2
the mean Au particle diameters increased significantly. These
results suggest that in this study the anchoring of Au
particles strongly depends on the type of support. ZnO has
the strongest anchoring effect toward Au particles followed
by Al O , SiO , TiO and CeO .
2
3
2
2
2
Au loading before and after the reaction, as measured by
inductively coupled plasma atomic emission spectrometry
(
ICP-AES), is listed in Table 3. These results confirm that no
3
9
Au leaching occurred during the reaction.
region in Au/ZnO.
Au/Al had a small reduction peak at 400 °C with a
main reduction peak at 500 °C, which could be attributed to
The surface gold content and the electronic state of the
catalysts were characterized by XPS and the results are shown
in Table 4 and Fig. 4. The binding energies of Au(0) 4f7/2 are 83.5,
2 3
O
+
2
the reduction of Au species. CeO had a small reduction
8
3.5, and 83.4 eV for Au/TiO
2
, Au(2.9)/ZnO, and Au/ZnO-300
peak around 600 °C and a large peak above 850 °C, which
(
prepared by calcination of Au(2.9)/ZnO at 300 °C in air for 3 h),
could be attributed to the reduction of the surface shell and
4
0
respectively. These values are lower than that of bulk
gold at 84.0 eV, indicating the presence of a gold–support
interaction. For Au/SiO , the binding energy of Au(0) 4f was
bulk CeO
2 2
, respectively. Au/CeO had four reduction peaks
around 120, 300, 500, and 750 °C. The peak at 120 °C may be
3
+
+
related to the reduction of Au to Au while the peak at
2
7/2
+
0
8
4.3 eV, indicating that the interaction between Au and the
SiO support is very weak. For Au/Al , about 27.5% of the
surface gold was in the oxidized state because of the
300 °C could be assigned to the reduction of Au to Au . By
comparison with CeO , the peaks attributed to the reduction
2
2
O
3
2
of the surface shell and bulk of the support in Au/CeO2
shifted to 500 and 750 °C, respectively. This may be due to
the presence of a H spillover effect on the Au/CeO .
2 2
+
stabilization of Au by Al
2
O
3
. For Au/CeO
2
, about 42.4% of the
3
+
surface gold was in the oxidized state (Au ). We also found
that Au(2.9)/ZnO had the highest gold surface content (15.5%).
Au/ZnO-300 had a lower gold surface content (11.4%) than
Au(2.9)/ZnO, which may be due to the growth of gold particles
or diffusion into the pores of ZnO during calcination at 300 °C
Temperature programmed desorption (TPD) of NH over
3
these Au catalysts was conducted to examine the acid
properties of the catalysts and the results are shown in Fig. 6.
As shown in Fig. 6, the desorption of NH was examined for
3
for 3 h. Au/CeO
Au/Al , Au/SiO
catalysts that were supported on these five different
2
gave a higher surface gold content than
, and Au/TiO . TEM confirmed that the Au
2 3
Au/Al O and Au(2.9)/ZnO, and we found that these two
2
O
3
2
2
catalysts contained acidic sites. From the temperature
and intensity of the NH3 desorption peak, we found that
Au(2.9)/ZnO was more strongly acidic although the total
supports had similar Au particle sizes (2.9–3.3 nm); hence,
the lower surface gold content in Au/SiO
2
and Au/TiO
2
may
amount of acid was lower than that in Au/Al
hand, no desorption of NH was detected from Au/SiO , Au/TiO
2
2 3
O . On the other
suggest that most of the Au particles in these two catalysts
entered the pores of the supports, which could not be
detected by XPS analysis.
3
2
and Au/CeO
2
, indicating that these catalysts have no
significant acidity.
H temperature programmed reduction curves for the Au
2
catalysts, ZnO, and CeO
consumption peaks were observed for both Au/TiO
2
are shown in Fig. 5. No H
and Au/
2
4. Discussion
2
SiO , indicating that Au was present in the metallic state.
These results are consistent with the XPS analysis. ZnO did
This study confirms that the catalytic activity per surface Au
atom (TOF) in the oxidative esterification of ethylene glycol is
2
3146 | Catal. Sci. Technol., 2014, 4, 3141–3150
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
2 3 2 2
Fig. 3 TEM micrographs and the corresponding Au particle size distributions for fresh Au catalysts: a) Au(2.9)/ZnO; b) Au/Al O ; c) Au/SiO ; d) Au/TiO ;
2
e) Au/CeO .
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 3141–3150 | 3147

View Article Online
Paper
Catalysis Science & Technology
Table 4 Surface content and relative surface distribution of gold species in Au catalysts
a
Catalysts
Ausurf (%)
B.E. of Au4f7/2 (eV)
FWHM (eV)
Au species (%)
0
Au(2.9)/ZnO
Au/ZnO-300
15.5
11.4
11.0
83.5
83.4
84.7
1.16
1.19
2.21
1.63
2.31
1.23
1.64
1.51
Au (100)
0
Au (100)
δ+
Au/Al O
2 3
Au (72.5)
+
85.7
Au (27.5)
0
Au/SiO
Au/TiO
2
3.9
3.7
13.7
84.3
83.5
84.0
Au (100)
0
2
Au (100)
0
Au/CeO
2
Au (57.6)
3+
87.7
Au (42.4)
a
The surface gold weight content is calculated from the ratio of Au/M
x
O
y
obtained from the surface analysis by XPS.
Fig. 5
H
2
-TPR curves of Au catalysts and supports: a) ZnO; b) Au(2.9)/ZnO;
; e) Au/SiO ; f) Au/TiO ; g) CeO ; h) Au/CeO
c) Au/ZnO-300; d) Au/Al
2
O
3
2
2
2
2
.
Fig. 4 Au 4f XPS spectra of the supported Au catalysts.
roughly independent of the diameter of the Au particles in
the range of 2.9–7.4 nm. This fact is verified in both Au/ZnO
and Au/Al O catalysts and independent of the support.
3
Fig. 6 NH -TPD spectra of the supported Au catalysts.
2
3
Previously, Comotti et al. also found that the catalytic
activity of Au nanoparticles during glucose oxidation was
inversely proportional to the diameter of the Au particles in
both Au/ZnO and Au/Al O had lower activities than those of
around 3.0 nm Au.
2
3
33
the range of 3–6 nm, which means that the catalytic activity
per surface Au atom is independent of the diameter of the Au
particles. Our present result is consistent with the result
reported by Comotti et al., and this observed correlation
between activity and diameter agrees with the assertion that
It has been reported that a metal-to-insulator transition
occurs as the particle size falls below 3.5 nm in
2
0,41,42
diameter.
When the mean diameter of the Au particles
is around 2 nm, the electronic properties are thus closer to
that of insulator Au than bulk Au because of the quantum size
effect. Hence, variations in the overall electronic properties of
3
4
only exposed atoms are catalytically active. However, for a
Au mean particle diameter of around 2.0 nm, we found that
3148 | Catal. Sci. Technol., 2014, 4, 3141–3150
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
Scheme 1
the Au particles could be responsible for the lower activity
of Au/ZnO with ~2.0 nm Au particles compared to that with
5
. Conclusion
~
2.9 nm Au particles. In fact, the low-temperature oxidation of
We prepared three series of supported Au catalysts, namely,
ZnO and Al O supported Au nanoparticles with controlled
CO on supported Au particles also shows a marked decrease in
the turnover frequency (TOF, reaction rate per surface Au site
per second) as the diameter of the Au particles is reduced to
2
3
mean diameters that range between 2.0 and 7.4 nm, and Au
nanoparticles with uniform mean diameters (2.9–3.3 nm)
that were loaded onto various supports. These catalysts were
evaluated for the oxidative esterification of ethylene glycol to
methyl glycolate. The results show that the catalytic activity
per surface Au atom is roughly independent of the diameter
of the Au particles in the range of 3–7.4 nm, whereas smaller
Au particles (~2.0 nm) have inferior activity. This behavior
1
8,43
less than ~3 nm.
On the other hand, the support also plays important roles
in the oxidative esterification of ethylene glycol. The investigation
of five Au catalysts, i.e. Au(2.9)/ZnO, Au/Al O , Au/SiO , Au/TiO ,
and Au/CeO
Au loadings show that Au(2.9)/ZnO has the highest activity
followed by Au/Al , Au/SiO , Au/TiO , and Au/CeO
2
3
2
2
2
, with similar Au mean particle diameters and
2
O
3
2
2
2
.
2 3
was observed on both Au/ZnO and Au/Al O catalysts. Au/ZnO
From the characterization of these catalysts using TEM,
XPS, H -TPR and NH -TPD, we believe that the main reason
that Au/ZnO exhibits the highest catalytic activity can be
explained as follows: first, ZnO has the strongest anchoring
effect toward metallic Au particles. TEM confirmed that after
the reaction the diameter of the Au particles in Au/ZnO
with a mean Au particle diameter of ~2.9 nm could be reused
at least three times without an obvious loss of activity.
After storage in a refrigerator for 7 months, we obtained
comparable catalytic activity to that of the fresh catalyst.
Among the five supported Au catalysts, Au/ZnO gave the
2
3
best catalytic performance in the reaction followed by Au/Al
2 3
O .
increased slightly from 2.9 to 3.0 nm. Although Al
2
O
3
also
Au/SiO , Au/TiO , and Au/CeO gave far less catalytic activity.
2
2
2
anchors Au particles well, around 27.5% of the Au in Au/Al O
The high catalytic activity of Au/ZnO is likely due to the
following: first, ZnO has the strongest anchoring effect toward
metallic Au particles, which effectively suppresses the
growth of Au particles during the reaction; second, the
strong interaction between Au and ZnO increases the
mobility of oxygen and lattice oxygen accessibility, which
may promote the oxidative esterification reaction; third,
cooperation between Au nanoparticles and the acidic ZnO
support efficiently promotes the generation and subsequent
oxidation of intermediates to the corresponding esters.
2
3
is in the oxidized state, which is inactive in this reaction. The
other supports, i.e. SiO , TiO and CeO , have a far poorer
2
2
2
anchoring effect toward Au particles. Second, the strong
interaction between Au and ZnO increases the mobility of
oxygen and lattice oxygen accessibility and this may promote
the oxidative esterification reaction because it is generally
considered that lattice oxygen plays an important role in
alcohol oxidation, and the reaction undergoes the Mars–van
4
4
Krevelen process. Third, the acidity of ZnO may also play
an important role in the oxidative esterification of
ethylene glycol.
Acknowledgements
It is generally believed that the oxidative esterification of
primary alcohols to methyl esters in the presence of gold
catalysts involves three steps (see Scheme 1): first, the
selective oxidation of primary alcohols in the presence of
gold catalysts results in the formation of an aldehyde via
the abstraction of β-H from the alcohol. A condensation
reaction then occurs between the aldehyde and the alcohol
leading to the formation of a hemiacetal species (step 2).
Ester products are formed by the direct oxidation of
The authors gratefully acknowledge financial support from
the National Natural Science Foundation of China (grant no.
2
0976101), the Program for Key Science and Technology
Innovation Team of Shaanxi Province (2012KCT-21) and the
Fundamental Research Funds for the Central Universities
(
GK201305011).
7
–12,45
hemiacetals (step 3).
References
It has been reported that inorganic oxides contain
appropriate acid sites that are able to facilitate hemiacetal
1 I. S. Nielsen, E. Taarning, K. Egeblad, R. Madsen and
C. H. Christensen, Catal. Lett., 2007, 116, 35–40.
2 K. Ishihara, M. Kubota, H. Kurihara and H. Yamamoto,
J. Am. Chem. Soc., 1995, 117, 4413–4414.
4
6
formation and thus promote the formation of esters.
Previously, Su et al. also found that cooperation between Au
nanoparticles and an acidic β-Ga support efficiently
2
O
3
promoted the generation and subsequent oxidation of
intermediates to the corresponding esters.
3 G. W. Breton, J. Org. Chem., 1997, 62, 8952–8954.
4 N. Mori and H. Togo, Tetrahedron, 2005, 61, 5915–5925.
8
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 3141–3150 | 3149

View Article Online
Paper
Catalysis Science & Technology
5
6
7
8
9
N. N. Karade, G. B. Tiwari and D. B. Huple, Synlett, 2005, 13,
039–2042.
C. E. McDonald, L. E. Nice, A. W. Shaw and N. B. Nestor,
Tetrahedron Lett., 1993, 34, 2741–2744.
T. Hayashi, T. Inagaki, N. Itayama and H. Baba, Catal.
Today, 2006, 117, 210–213.
F.-Z. Su, J. Ni, H. Sun, Y. Cao, H.-Y. He and K.-N. Fan,
Chem. – Eur. J., 2008, 14, 7131–7135.
26 C. Bianchi, F. Porta, L. Prati and M. Rossi, Top. Catal.,
2000, 13, 231–236.
27 P. Haider, B. Kimmerle, F. Krumerich, W. Kleist,
J.-D. Grunwaldt and A. Baiker, Catal. Lett., 2008, 125,
169–176.
28 W. Fang, J. Chen, Q. Zhang, W. Deng and Y. Wang,
Chem. – Eur. J., 2011, 17, 1247–1256.
29 J. Strunk, K. Kähler, X. Xia, M. Comotti, F. Schüth,
T. Reinecke and M. Muhler, Appl. Catal., A, 2009, 359,
121–128.
2
E. Taarning, A. T. Madsen, J. M. Marchetti, K. Egeblad and
C. H. Christensen, Green Chem., 2008, 10, 408–414.
1
0 B. Jørgensen, S. E. Christiansen, M. L. D. Thomsen and
C. H. Christensen, J. Catal., 2007, 251, 332–337.
30 M. B. Boucher, N. Yi, F. Gittleson, B. Zugic, H. Saltsburg and
M. Flytzani-Stephanopoulos, J. Phys. Chem. C, 2011, 115,
1261–1268.
31 N. S. Phala, G. Klatt, E. van Steen, S. A. French, A. A. Sokol
and C. R. A. Catlow, Phys. Chem. Chem. Phys., 2005, 7,
2440–2445.
32 P. Bera and J. M. Vohs, J. Phys. Chem. C, 2007, 111,
7049–7057.
33 M. Comotti, C. D. Pina, R. Matarrese and M. Rossi, Angew.
Chem., Int. Ed., 2004, 43, 5812–5815.
1
1
1 G. Liu, G. Li and H. Song, Catal. Lett., 2009, 128, 493–501.
2 G. L. Brett, P. J. Miedziak, N. Dimitratos, J. A. Lopez-
Sanchez, N. F. Dummer, R. Tiruvalam, C. J. Kiely,
D. W. Knight, S. H. Taylor, D. J. Morgan, A. F. Carley and
G. J. Hutchings, Catal. Sci. Technol., 2012, 2, 97–104.
1
3 R. K. P. Purushothaman, J. van Haveren, D. S. van Es,
I. Melian-Cabrera and H. J. Heeeres, Green Chem., 2012, 14,
2
031–2037.
1
1
1
4 K. Suzuki, T. Yamaguchi, K. Matsushita, C. Iitsuka, J. Miura,
T. Akaogi and H. Ishida, ACS Catal., 2013, 3, 1845–1849.
5 A. S. K. Hashmi and G. J. Hutchings, Angew. Chem., Int. Ed.,
34 H. Tsunoyama, H. Sakurai, Y. Negishi and T. Tsukuda,
J. Am. Chem. Soc., 2005, 127, 9374–9375.
35 H. Tsunoyama, H. Sakurai and T. Tsukuda, Chem. Phys.
Lett., 2006, 429, 528–532.
36 T. Ishida, N. Kinoshita, H. Okatsu, T. Akita, T. Takei and
M. Haruta, Angew. Chem., Int. Ed., 2008, 47, 9265–9268.
37 S. Biella, F. Porta, L. Prati and M. Rossi, Catal. Lett.,
2003, 90, 23–29.
38 M. Comotti, C. Pina, R. Matarrese, M. Rossi and A. Siani,
Appl. Catal., A, 2005, 291, 204–209.
39 J. Strunk, K. Kähler, X. Xia, M. Comotti, F. Schüth,
T. Reinecke and M. Muhler, Appl. Catal., A, 2009, 359,
121–128.
40 G. Jacobs, P. M. Patterson, L. Williams, D. Sparks and
B. H. Davis, Appl. Catal., A, 2004, 258, 203–214.
41 M. Valden, X. Lai and D. W. Goodman, Science, 1998, 281,
1647–1650.
42 D. W. Goodman, Catal. Lett., 2005, 99, 1–4.
43 G. R. Bamwenda, S. Tsubota, T. Nakamura and M. Haruta,
Catal. Lett., 1997, 44, 83–87.
44 J. Huang, W.-L. Dai and K. Fan, J. Catal., 2009, 266, 228–235.
45 A. Abad, P. Concepcion, A. Corma and H. Garcia, Angew.
Chem., Int. Ed., 2005, 44, 4066–4068.
2
006, 45, 7896–7936.
6 D. I. Enache, J. K. Edwards, P. Landon, B. Solsona-Espriu,
A. F. Carley, A. A. Herzing, M. Watanabe, C. J. Kiely,
D. W. Knight and G. J. Hutchings, Science, 2006, 311, 362–365.
7 A. K. Sinha, S. Seedlan, S. Tsubota and M. Haruta,
Angew. Chem., Int. Ed., 2004, 43, 1546–1578.
8 M. Haruta, S. Tsubota, T. Kobayashi, H. Kageyama,
M. J. Genet and B. Delmon, J. Catal., 1993, 144, 175–192.
9 M. M. Schubert, S. Hackenberg, A. C. van Veen, M. Muhler,
V. Plzak and R. J. Beham, J. Catal., 2001, 197, 113–122.
0 M. S. Chen and D. W. Goodman, Catal. Today, 2006, 111,
1
1
1
2
2
2
2
2
2
2
2–33.
1 M. Comotti, W.-C. Li, B. Spliethoff and F. Schüth, J. Am.
Chem. Soc., 2006, 128, 917–924.
2 D. Widmann, Y. Liu, F. Schüth and R. J. Beham, J. Catal.,
2
010, 276, 292–305.
3 S. Ivanova, V. Pitchon, C. Petit and V. Caps, ChemCatChem,
010, 2, 556–563.
4 Y. Guan and E. J. M. Hensen, Appl. Catal., A, 2009, 361,
9–56.
5 K.-Q. Sun, S.-W. Luo, N. Xu and B.-Q. Xu, Catal. Lett.,
008, 124, 238–242.
2
4
46 D. I. Enache, D. W. Knight and G. J. Hutchings, Catal. Lett.,
2005, 103, 43–52.
2
3150 | Catal. Sci. Technol., 2014, 4, 3141–3150
This journal is © The Royal Society of Chemistry 2014