Oxidative Esterification of Methacrolein to Methyl Methacrylate over Supported Gold Catalysts Prepared by Colloid Deposition

Article abstract of DOI:10.1002/cctc.201601688

Esterification is one of the most pivotal organic transformations. Au catalysts were prepared by using a colloid deposition method with poly(vinyl alcohol) (PVA) as a protective agent. The catalyst was used for the oxidative esterification of methacrylate (MAL) to methyl methacrylate (MMA). Three pre-treatments were used to remove the PVA, which is unfavorable for catalytic activity. It is found that the catalyst pre-treated at 300 °C showed substantially improved activity owing to the lower PVA loading compared with catalysts treated with hot water washing and water reflux. Surprisingly, it is also found that the distribution and loading content of Au particles was closely related to the pH of the colloid solution. We demonstrate that the deposition process is controlled by the different charge of the support surface at different colloid solution pH. Further, the catalysts with similar size Au particles loaded on TiO₂, SiO₂, Al₂O₃, CeO₂, ZrO₂, and ZnO were successfully prepared by controlling the colloid solution pH. The Au/ZnO catalyst presented the best performance, which may be a result of the strong basic surface sites that improved the formation of the intermediate and the strong interaction between Au and ZnO. This interaction caused an anchoring effect and changed the geometries of Au particles, which could enhance the stability of catalysts and promote the mobility of oxygen, respectively.

Full text of DOI:10.1002/cctc.201601688

Accepted Article
Title: Oxidative esterification of methacrolein to methyl methacrylate
over supported gold catalysts prepared by colloid deposition
Authors: Yuchao Li, Yanxia Zheng, Lei Wang, and Zhongjun Fu
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemCatChem 10.1002/cctc.201601688
Link to VoR: http://dx.doi.org/10.1002/cctc.201601688
A Journal of
www.chemcatchem.org

ChemCatChem
10.1002/cctc.201601688
FULL PAPER
Oxidative esterification of methacrolein to methyl methacrylate
over supported gold catalysts prepared by colloid deposition
Yuchao Li, * [a] Yanxia Zheng, [a] Lei Wang and Zhongjun Fu
[b]
[a]
Abstract: Esterification is one of the most pivotal organic
transformations. Au catalysts were prepared by colloid deposition
method using poly vinyl alcohol (PVA) as protective agent for the
oxidative esterification of methacrylate (MAL) to methyl methacrylate
route less green and more expensive. Others employed MgO as
promoter or support to prepare MMA with low conversion[4] or high
content of methanol[5] which goes against the commercial
applications due to the energy-extensive consumption for
separation. Au nanoparticles supported on lots of metal oxides
have been investigated on oxidative esterification of primary
aldehyde or alcohol with alcohol to corresponding ester with a
(MMA). Three pre-treatments were used to remove the PVA, which is
unfavorable for catalytic activity. It is found that the catalyst pre-
o
treated at 300 C substantially improved activity due to the reduced
[6]
PVA loading compared with hot water washing and water reflux.
Surprisingly, it is also found that the distribution and loading content
of Au particles was closely related to the pH of the colloid solution. We
demonstrate that the deposition process is controlled by the different
charge of the support surface at different colloid solution pH. Further,
high activity. Although it is reported that the acidic sites or basic
sites[ of support was in favor of the activity of catalysts, it is
unknown that how the acidic sites or basic sites patticipate in the
reaction process and whether there is other prior factor that
affects the performance of catalysts. The intrinsic effect of the
support still need to be investigated in order to further understand
the roles of the support.
5]
the catalysts with the similar size of Au particles loaded on TiO
2 2
, SiO ,
Al , CeO , ZrO and ZnO were successfully prepared by controlling
O
2 3
2
2
the colloid solution pH. The Au/ZnO presented the best performance,
which may be due to the strong basic surface sites that improved the
formation of the intermediate and the strong interaction between Au
and ZnO. This interaction caused the anchoring effect and the change
of geometries of Au particles which could enhance the stability of
catalysts and promote the mobility of oxygen respectively.
Scheme 1. Oxidative esterification of MAL with methanol to produce MMA
Introduction
In addition, the Au particle size and designed morphologies also
have an important effect on the activity of catalysts. Among the
reported catalysts, the main preparation method is deposition-
precipitation (DP) method, which is not fit for all support especially
for support with low isoelectric point.[7] Moreover, it is hard to
precisely control the loading content, the particle size and the
morphologies of Au by DP method. In this paper, we focus on the
colloid deposition method since it is possible to precisely tune the
particle size distribution and morphology before deposition on the
support. Thus, the effect of the intrinsic support could be easy to
investigate. However, the polymer ligands or surfactants are
typically required to stabilize the Au nanoparticles during the
colloid deposition method. The stabilizing ligands associated with
the Au particles as well as the support, which leads to the
decrease of active sites. It is expected that those ligands will
reduce the activity of catalysts. Hence, it is necessary to remove
these stabilizing ligands, nevertheless, it is difficult to remove by
only traditional washing. Recently, post-synthesis heat or
extraction treatment methodologies have been developed for
removing the stabilizing ligands after the immobilization step. For
instance, it is found that using an appropriate calcination
procedure these ligands could be effectively removed and the
catalysts are active for CO oxidation without causing a significant
MMA is an important industrial material, which can be used to
produce acrylic plastics and other polymer dispersions used in
[
1]
paints and coatings. Currently, MMA is mainly produced by the
acetone cyanohydrins process (ACH), but there are challenges to
deal with the byproduct ammonium bisulfate waste and toxic raw
material hydrogen cyanide. Hence, some environmentally friendly
starting alternatives have been developed such as isobutene and
syngas, which are firstly converted to MAL and then oxidative
esterification to MMA. The oxidative esterification of MAL with
methanol to MMA is the key step in these attractive routes for the
production of MMA (Scheme 1).
Recently, Au based catalysts have been reported for the direct
oxidative esterification of MAL efficiently due to their high activity
in many reactions to replace the traditional Pd-Pb catalysts, which
generally showed a selectivity for MMA below 90% and the Pb is
quite toxic.[2] In addition, in most case, the alkali Mg(OH)
required to obtain high yield of ester,[3] which makes the whole
2
is
[
[
a]
b]
Dr. Y. Li, Y. Zheng, prof. Z. Fu
School of Chemical Engineering
Shandong University of Technology
Zibo 255049, PR China
E-mail:cyulee@126.com
Dr. L. Wang
[8]
increase in the mean particle size. It is also reported that using
hot water washing or hot water reflux could be efficient to remove
the ligands and greatly improve the activity of catalysts.[9] And
State Key Laboratory of Multiphase Complex Systems
Institute of Process Engineering, Chinese Academy of Sciences
Beijing 100190, PR China
2
both found that the Au supported on TiO presented the best
performance. However, the effect of textural structure of support
to the preparation process and the activity of catalyst was not
discussed in detail. Ke et al. prepared the Au catalysts by the
colloidal deposition method using PVA as protecting agent and
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201601688
FULL PAPER
investigated the effect of support on oxidative esterification of
ethylene glycol.[10] But, the effect of PVA was not involved.
Recently, we have aimed at developing efficient Au catalysts
using colloid deposition method applied for the oxidative
esterification of MAL with methanol to MMA in the absence of any
liquid bases. To obtain the best performance, we firstly examine
the ways of removing PVA included the calcination and hot water
pretreatment, as well the effects on the structure and activity of
those catalysts. The effects of pH during the immobilization step
were investigated. The effects of the support and the reason for
the high activity of ZnO supported Au catalyst were discussed.
Results and Discussion
Effective PVA removing method to minimize its negative
effect on the catalytic performance
The stabilizers are still associated with the Au particles as well as
the support after the deposition, which could obviously affect the
activity of catalysts. Therefore, it is necessary to remove the PVA
ligands to activate the Au particles and support after the
immobilization. However, one of the challenges is the loss of
activity due to the sintering of the Au particles during the removal
process. We investigated three methods to remove stabilizer PVA
including the thermal treatment, hot water washing and hot water
reflux. The evaluation of the size of Au particles supported on ZnO
after different treatments was carried out by TEM technology and
the representative micrographs and corresponding particle size
distributions are shown in Fig. 1. The mean Au particle size of
Figure 1. TEM micrographs of Au/ZnO catalysts after different methods of
removing PVA and the corresponding particle size distributions data. (a) dry at
100 oC, (b) calcined at 200 C, (c) calcined at 300 C, (d) 90 C water washing,
(
o
o
o
e) 90 oC water reflux for 2h, (f) gold colloid before deposition
o
o
samples treated only dry at 100 C or only washing by 90 C hot
water presented 4.16 nm or 4.09 nm. After the heat treatment at
o
Table 1. Catalytic performance of catalysts with different PVA removal
methods[ and corresponding catalysts weight loss and surface atomic ratio
of Au
300 C for 4 hours or water reflux, the mean size of Au particles
a]
increased slightly to 4.4 nm. Thus, these three treatments did not
affect the Au particle size distributions and they basically
presented similar size around 4.2 nm, which is close to the Au
colloid size (Figure 1f).
Selectivity (%)[b]
Surface
atomic
ratio of
Au [%]
PVA
removal
method
WL10
Conversio
n of MAL
[%]
Further, the DT-TGA analysis was used to investigate the PVA
content after treatments. The spectra of DT-TGA are shown in
Figure 2 and the analysis data are presented in Table 1. For all of
the samples, an obvious exothermic peak could be observed
0-450
[%]
Acet
MMA
al
Others
Calcinatio
o
around 250 C, which was assigned to the decomposition of PVA
n
C
at 100
0.38
0.22
6.6
6.8
7.3
56.4
97.9
99.9
46.5
74.4
85.9
4.1
0.8
0.6
8.2
o
o
stabilizer. Besides, the dramatic weight loss from 100 C to 300
oC could be also observed and the weight of catalysts mostly kept
Calcinatio
n
at 200
8.0
the same after 450 oC. These findings suggest that the PVA
o
C
o
o
started to decompose around 100 C and at 250 C decomposed
the most quickly, and finally, it could be totally removed above 450
Calcinatio
n
at 300
0.14
0.28
7.63
o
o
o
C
C. With the increase of treatment temperature from 100 C to 450
Water
reflux
Hot water
washing
oC, the weight loss (WL100-450) of samples (Table 1) decreased
from 0.38% to 0.14%. Hence, the heat treatment is an effective
method to remove the PVA, although little PVA still stayed on the
6.7
6.6
90.5
52.8
70.9
32.9
3.7
3.7
5.3
9.5
0.33
o
catalyst after calcination at 300 C. The thermal treatment at 400
[
a] Reaction conditions: catalyst (Au loading, 1 wt %), 0.50 g; CH
3
OH/MAL
oC was not adopted due to the dramatic increase of Au particle
size leading to the deactivation of catalyst. The WL100-450 for the
sample treated by hot water washing presented 0.33%, which is
=
30:1 (molar ratio); CH OH, 15 mL; P (O ) = 0.2 MPa; T = 343 K; t = 2 h.
3
2
[b] MMA, methyl methacrylate; others include methyl isobutyrate, isobutyric
acid, isobutyl aldehyde, dimers of the methacrolein, and some unknown
products.
o
similar with the sample only dry at 100 C. Hence, the hot water
washing method is not an effective way to remove the PVA
stabilizer. The water reflux is a better method than the hot water
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201601688
FULL PAPER
Figure 2. DT-TGA spretra of Au/ZnO catalysts after different methods of removing PVA. (a) dry at 100 oC, (b) calcined at 200 oC, (c) calcined at 300 oC, (d) 90 o
water washing, (e) 90 C water reflux for 2h.
C
o
washing due to that the WL100-450 presented 0.28%.
Comparatively, the hot treatment is the most efficient method.
XPS technology was employed to analyze the Au valence state
and surface atomic ratio. The binding energy of Au (Figure S1) for
catalysts with different PVA removal methods located the similar
position, hence, indicating that treatments did not change the Au
the sample calcined at 200 oC, it presented similar catalytic
activity. This relative enhancement of activity may be due to the
smaller average Au particle size (4.09 nm) than the sample
o
calcined at 200 C. Hence，the water reflux could remove PVA
without Au particles increasing obviously. However, the catalyst
after water reflux once as reported in literature could not be
efficient to catalyze the oxidative esterification of MAL. To improve
the removal of PVA, further measures should be carried out such
as multi water reflux, which is under study in our group.
In summary, we compared catalysts with three different pre-
treatment methods to remove the PVA stabilizer. The heat
thermal treatment presented the best activity with the slight
increase of mean Au particles size. The water reflux method could
improve the catalytic performance without the growth of Au size.
Besides, no matter which way was used to remove the PVA, the
valence state of Au was metal state with binding energy around
83.7 eV assigned to the strong interaction between Au particles
and ZnO support.
valence state. However, the surface atomic ratio of Au is different
o
(
Table 1). The sample calcined at 300 C possessed highest Au
atomic ratio at 7.3%, while the others presented around 6.7%.
This means that more Au sites exposed outside due to the
efficient removal of PVA. The catalysts with different PVA removal
methods were tested on the oxidative esterification of MAL to
MMA and the results are shown in Table 1. The catalysts
thermally treated at 100 oC or treated by hot water washing
presented bad activity with the conversion of MAL around 55%
and selectivity of MMA around 40%. Many side reactions occurred
and the main by-products included 1,1-dimethoxy-2-
methylpropylene (acetal), Methyl isobutyrate (MIB), isobutyl
2
aldehyde, dimers of the methacrolein, and CO , which are
products of over-oxidation and generally produced more with the
increased of conversion of MAL. The catalysts calcined at 200 C
Effects of colloid solution pH on the size of supported Au
particles
o
or treated by water reflux showed medium activity with conversion
of MAL about 95% and selectivity of MMA about 70%. The
Traditionally, it is expected that the colloid deposition process
should not alter the particle sizes and distributions, because the
gold clusters had been produced before deposition. It is also
reported that the Au nanoparticles would be completely adsorbed
on supports after enough time.[8] However, we found that the
supports with low isoelectric point could not completely adsorb the
Au particles and the size of Au particles grew up. The distribution
of Au nanoparticles and the loading content of Au were depended
on the point of zero charge of the support during the
o
catalyst calcined at 300
C provided the best catalytic
performance with the conversion of MAL above 99% and the
selectivity of MMA 85.9%. This activity of catalysts trends against
with the content of PVA. Therefore, the content of PVA
significantly affected the activity of Au particles and the thermal
treatment is an effective way to remove the PVA. The method of
pre-treatment by hot water washing could not be profitable to
remove the PVA stabilizer, which may due to that the time of
washing is too short to dissolve PVA by hot water. The water
extraction method is a relative good way to remove PVA. Although
the catalyst pre-treated by water reflux possessed more PVA than
immobilization step. To investigate this effect in detail, CeO
ZrO were selected due to their proper isoelectric point around 6.5,
since the Au colloid solution pH was around 9 without adjustment
2
and
2
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201601688
FULL PAPER
Figure 3. TEM micrographs of Au/CeO
2
catalysts prepared under different colloid solution pH and the corresponding particle size distributions. (a) pH=9, (b) pH=7,
(c) pH=5, (d) pH=3, (e) pH=1
by HCl solution. Thus, it is not easy to show the relationship
between the solution pH value and the isoelectric point of support
impacts the amount of Au deposition. Therefore, the Au particles
distribution and the loading content are jointly dependent on the
colloid solution pH and the surface area of support with the former
as major factor. These two factors may also explain the different
adsorption rate of the colloids with changing support.[8] The
support with high isoelectric point and big surface area presents
high adsorption rate of the Au particles.
due to the high isoelectric point of ZnO and Al
2 3
O around 10. The
Au particles were adsorbed on CeO with different colloid solution
2
pH and the typical TEM micrographs are shown in Figure 3.When
the pH value (pH=9) was above the isoelectric point, the Au
particles became bigger during deposition from 4.3 nm to 5.1 nm
(
Table S1), while the mean size of Au particles was around 4.4
Combing with the relationship between colloid solution and
isoelectric point of support, we deduce that the dependence on
colloid solution pH and isoelectric point may be due to the nature
of the protecting agent used with its negative charges, as shown
in Figure. 4. The negative charges of PVA may shield the charges
of the support and the gold clusters under the conditions of the
colloid solution pH much above isoelectric point of support. During
the deposition process, the size of Au particles would grow and
part of Au particles could not deposit on the support due to the
conflict between the negative PVA and the negative support and
Au particles. When the colloid solution pH was approximate the
isoelectric point of corresponding support, the strong interaction
between negative PVA and positive support immediately take
place leading to the quickly complete adsorption of Au colloid.
When the pH is lower such as around 1, the protective agent PVA
could be broken resulting in the grown up of the Au particles. Thus,
the colloid solution pH value should be adjusted to below the
isoelectric point of support and above the 1 to obtain the best Au
particles distribution.
nm at the colloid solution pH in the range of 3-7. The size and
distribution of Au particles did not alter during the deposition
process, when the colloid solution pH was around the isoelectric
point of support. However, the size of Au particles supported on
CeO
The similar trend could also been observed for the Au particles
loaded on ZrO with the change of pH value (Figure. S2). Thus,
these results allow us to propose that the size of Au particles
decreased at first then increased for other supports with the
decrease of pH around the isoelectric point of corresponding
support.
2
increased to 8.1 nm with the pH value decreasing around 1.
2
On the other hand, the surface area of support also affects the
deposition process. The size of Au particles supported on ZrO
changed smaller from 4.5 nm to 5.5 nm than the samples
supported on CeO from 4.3 nm to 8.1 nm. Besides, the Au colloid
could completely deposit (Table S1) on the ZrO varying different
pH, while part of Au particles still stayed in solution after
deposition on CeO at the pH 7-9 above the isoelectric point of
CeO . This means that other factors of the support also play
important role on the Au deposition process eliminating the
solution pH and the isoelectric point. The ZrO with bigger surface
8.9 m g presented the complete
2
2
2
2
2
The different mean size of Au particles supported on CeO
2
and
ZrO were prepared. As we all know, the size of Au particles has
2
a great important role on the activity for many reactions.[ To
investigate this role on oxidative esterification of methacrolein, the
catalytic performances of these Au catalysts were tested and the
results are shown in Table S1. The catalyst possessing similar
size of Au particles provided close catalytic performance
11]
2
2
-1
2
-1
area 54 m g than the CeO
2
deposition even at high pH. Thus, we speculate that the surface
area of support has an obvious influence on the adsorbing
capacity. Nevertheless, Au particles could not completely deposit
at pH 9 for the sample Au/SiO
area around 179 m g . This means that the colloid solution also
2
, as the SiO
2
possessed big surface
indicating that other factors could be excluded. For both Au/CeO
and Au/ZrO catalysts, the catalyst with smaller Au particles
2
2
-1
2
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201601688
FULL PAPER
Figure 4. The possible deposition process at different colloid solution pH
presented higher MAL conversion and selectivity for MMA, which
could be obtained at medium colloid solution pH. The catalysts
with big Au particles prepared at low or high colloid solution pH
showed low activity. Thus, the smaller Au particles presented
higher activity, which may be due to that these particles could
Figure 5. TEM images and size distribution histogram of Au nanoparticles
loaded on different supports. (a) Au/TiO
e) Au/ZrO , and (f) Au/ZnO.
2 2 3 2 2
, (b) Au/Al O , (c) Au/SiO , (d) Au/CeO ,
(
2
activated the O
selectivity to MMA.
2
resulting in the high conversion of MAL and
[
11b, 12]
produced by hemiacetal as important intermediate. The Au
particles loaded on CeO and ZrO had a low selectivity of MMA
with a higher selectivity of acetal and the selectivity of other by-
products was blow 7%. Au/Al and Au/ZnO also presented
higher selectivity of MMA and lower selectivity of acetal than other
catalyst. Among of all the catalysts, the sample Au nanoparticles
supported on ZnO possessed the best performance with the yield
of MMA 85.9%.
Further, the method was applied on other supports by adjusting
the pH of colloid solution according the isoelectric point of
corresponding support to obtain the similar Au distribution and Au
loading and the results were shown in Table 2. The Au loading of
all the catalysts were all around 1%, close to the theoretical value
2
2
2
O
3
1% according the results of ICP. Only the sample Au/CeO
2
presented much lower Au content than target value, which might
be due to the small surface area and the polyhedral structure of
2
the support CeO . The TEM technology was employed to
determine the distribution of Au particles. As one can see from
Fig. 5, Au particles were well dispersed and the mean size was
almost identical for the catalysts with the mean size slight bigger
than the size of gold colloid. As summarized in Table 2, the mean
size of Au particles for all catalysts was similar around 4.4 nm
Table 2. Properties of Au catalysts loaded on different supports
Catalyst
Au
Au
particle
size
PH
Surface
area[c]
Au 4f/2[d]
loading
value
[
a]
[b]
2 -1
[%]
[m g ]
[nm]
2
except the Au particles supported on TiO with a slightly bigger
4
. 6
mean size around 4.6 nm. These indicated that well distributed
Au particles with similar size on different supports were
successfully prepared by the simple adjustment of colloid solution
pH. Thus, the effect of the size of Au particles on the difference in
activities of catalysts could be neglected. The property of support
might be responsible to the different activity of catalysts.
Au/TiO
2
1.08
1.17
1.04
0.86
0.97
1.14
5
3
8
3
5
8
55
183
154
8
83.3
(
5.4)[e]
Au/SiO
2
4.5 (6.7)
4.2 (5.6)
4.4 (5.8)
4.5 (5.8)
4.4 (5.1)
84.0
84.0
84.0
83.8
83.3
2 3
Au/Al O
Au/CeO
2
Effects of support on the Au supported catalysts for
oxidative esterification of methacrolein
Au/ZrO
2
18
4
The Au catalysts loaded on different supports were tested for the
oxidative esterification of MAL with methanol in the absence of
base and the results are shown in Figure 6. All samples presented
good activities with the conversion all above 69% especially for
Au/ZnO
[a] Calculated from ICP-OES data. [b] Evaluated from TEM. [c] Determined
by N2-BET. [d] Obtained by XPS data.[e] The value in parentheses is Au
particle size of catalysts after reaction.
Au/Al
However, the Au nanoparticles supported on TiO
showed low selectivity for MMA and acetal, which were both
2
O
3
and Au/ZnO with the conversion 99.6% and 99.9%.
2
and SiO
2
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201601688
FULL PAPER
Figure 7. XPS spectra of Au 4f in catalysts with different supports.
Figure 6. Catalytic performances of Au catalysts loaded on different supports.
Reaction conditions: catalyst, 0.50 g; CH
3 3
OH/MAL = 30:1 (molar ratio); CH OH,
1
5 mL; P (O ) = 0.2 MPa; T = 343 K; t = 2 h
2
facilitate the oxidative esterification transformation.
3
On the other hand, the support ZnO and Al O bearing the
2
Generally, the surface area is also an important factor which
could affect the diffusion of the substrates leading to the variation
of reaction rate and the BET surface areas of catalysts are
highest isoelectric point around 9 (Table S2) also presented the
best activity. The trend of the isoelectric point of the support also
is well associated with the variation of catalytic performance. This
interesting phenomenon is not reported before. However, the
differences between the supports with respect to the activity of the
catalysts can presently not be explained by the isoelectric point,
and this effect is under research. We preliminarily speculate that
the isoelectric point of the support may change the pH of the
reaction solution leading to the variation of the activity.
2
presented in Table 2. The sample Au/SiO presented the biggest
surface area, but its activity is very low with the yield of MMA 43%.
On the contrary, the catalyst Au/ZnO possessed the best
performance with the smallest surface area. Obviously, the
surface area of the catalyst mainly determined by the support is
not the determinant factor affecting the catalytic activity. Other
properties of the support need to be investigated.
The surface valence state of gold on different supports was
analyzed by the XPS technology and the results are shown in Fig.
The chemical adsorption method can be used to determine the
strength and amount of acidity, basicity and reducibility for solid
7. As we can see, all the catalysts showed the spectra with two
3 2
catalysts. NH -TPD, CO -TPD and H
2
-TPR were used to analyze
distinct peaks around 83.9 eV and 87.6 eV for Au 4f7/2 and Au 4f5/2
the Au catalysts, since it is reported that strong acidity, basicity
and reducibility of support are beneficial to the oxidative
lines, respectively, which are assigned to the Au 4f binding
energies of metallic Au.[14] The binding energy of Au 4f for all
catalysts differed a little and the binding energies of Au 4f7/2 of
different catalysts are shown in Table 2. The binding energy of Au
esterification of aldehyde.[5-6, 13] The NH
-TPD, CO
TPR profiles for catalysts loaded on different supports are shown
in Figure. S3. The Au/SiO presented the largerst amount of acidic
sites, while the Al and ZnO showed nearly no acidity. The
sample Au/ZnO presented the largest amounts of basic sites and
3
2 2
-TPD and H -
2
4
f
7/2 loaded on SiO
similar to the bulk Au at 84.0 eV. A slight shift to low binding
energies of Au 4f7/2 for the sample Au/ZrO could be observed.
Simultaneously, it is noteworthy that a negative shift toward a
lower binding energy region of 83.4 of the Au 4f7/2 for Au/TiO and
Au/ZnO is observed. This means that there is a strong interaction
between gold particles and TiO or ZnO and efficient electron
2 2 3 2
, Al O and CeO located at 83.9 eV, which is
2
O
3
2
the Au/Al
the catalysts, no obvious peak of H
2
O
3
showed second most amount of basic sites. For all
o
2
reduction below 500 C was
2
observed. This means Au particles reduction by NaBH
4
presented
no adsorbed oxygen, which is different from the Au particles
prepared by deposition-precipitation (DP) method. Obviously, no
correlation between the activity of catalyst and the acidity or
reducibility of support could be observed, since the Au/ZnO with
the highest activity possessed low acidity and reducibility. ZnO
2
transfer from the conduction band of TiO2 or ZnO to the gold
particles.[15]
Nevertheless, the effect of this strong interaction with respect
to the shape of Au particles is different and the HRTEM images
of these catalysts are shown in Figure 8. The Au particles
and Al
for the oxidative esterification of MAL to MMA, while the use of
SiO or TiO as the support provided much poorer catalytic
performances. Meantime, the Au/ZnO and Au/Al possessed
high densities of basic sites and the SiO or TiO as the support
2 3
O supported Au catalysts were the most efficient catalysts
supported on SiO
and are the same with the native ones. However, the
immobilization on the TiO and ZnO does seem to change the
2 2 3 2 2
, Al O , CeO and ZrO are almost spherical
2
2
2
O
3
2
2
2
shape of these Au particles in this study. The particles clearly
appear to be sunk and faceted under the surface if adsorbed on
the TiO and ZnO. This anchoring effect to Au clusters is benefit
2
showed low basic sites (Table S2). The trend of basic sites
correlates well with that for the change of catalytic performance
with different supports (Figure 6). These results allow us to
propose that the densities of basic sites of the support may
to the stability of the clusters. As shown in Table 2, the size of Au
particles supported on TiO and ZnO after reaction increased
2
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201601688
FULL PAPER
enhance the formation of intermediate hemiacetal leading to the
improvement of activity. That may be one reason that the Au/ZnO
presented the best activity. On the other hand, the active oxygen
species could not be observed by H -TPR technology, which may
2
be due to that our Au catalysts were prepared by colloid
deposition method. Compared with the Au particles prepared by
traditional DP method, our Au particles reduced by NaBH
presented totally metallic state as confirmed by XPS spectra,
even for the Au particles supported on TiO , ZrO and ZnO
negatively charged. The charging of Au particles has less
pronounced effect on the activation for O adsorption and
4
2
2
2
dissociation than the size and shape effects, when the Au size
goes beyond about 40 atoms (about 1 nm).[20] This charge
transformation indicates there is strong interaction between Au
2
and TiO or ZnO, which enhanced the stability and changed the
geometries of Au particles particular for the ZnO. The cubo-
octahedron truncated morphology for Au/ZnO owned more
perimeter sites, which may activated the methoxy species[21] and
promote the active oxygen through oxygen spillover.[22] In
summary, the Au/ZnO exhibited the best performance due to the
strong basic surface sites and the strong interaction between Au
and ZnO leading to the anchoring effect and the change of
geometries of Au particles which could enhance the stability of
catalysts and promote the oxygen mobility and activate the
methanol respectively.
Figure 8. HRTEM images and size distribution histogram of Au nanoparticles
loaded on different supports. (a) TiO
and (f) ZnO
2
, (b) Al
2 3
O , (c) SiO
2
, (d) CeO
2
, (e) ZrO
2
,
There are many important factors that affect the catalytic
performance. However, it is unclear that which factor is the
determining factor. It is necessary and significant to study the
reaction mechanism. Although Xu et al[19] reported the possible
reaction process on vapour-phase gold-surface-mediated
coupling of aldehydes with methanol, the reaction mechanism in
gas-liquid-solid phase particularly under actual reaction
conditions is unknown. These factors studied in our paper is the
foundation of research for the reaction mechanism. In addition,
more technologies in situ are needed.
smaller than other catalysts, indicating that the catalysts Au/TiO
2
and Au/ZnO shown better stability. No leaching of Au after
reaction for these two catalysts also confirmed this. On the other
hand, the change of geometries could generate the formation of
defect on the gold clusters, leading to a change in the number of
active corner, edge atoms and the perimeter atoms as previously
reported.[16] In our study, the Au particles supported on TiO
and
2
ZnO resembled to cuboctahedral and cubo-octahedron truncated
morphology respectively. The latter structure presented more
amounts of active perimeter and step sites.[17] The Au particles
Conclusions
2
supported on ZnO presented more flat than on TiO , indicating the
stronger interaction between Au and ZnO.[18] Moreover, combined
with the results of XPS and the change of geometries, the
We prepared Au catalysts by colloid deposition method using PVA
as protective agent for the oxidative esterification of MAL to MMA
in the absence of base. Three different pre-treatment: thermal
treatment, hot water washing and water reflux were compared to
remove the PVA stabilizer and no obvious increase of the Au
particles were observed after treated by these three methods. The
2
interaction between Au and TiO or ZnO could be assigned to the
reductive strong metal−support interactions (SMSI) as
reported in the literature[15a] and these negative charge Au
particles may be benefit to the hydrogenation. That may be why
the content of byproduct MIB increased catalyzed by Au/TiO and
2
Au/ZnO than other catalysts prepared by deposition-precipitation
method, which the valence state of Au mainly presented part of
oxidation state.
o
catalyst heated treatment at 300 C presented the best activity
due to the minimum content of PVA confirmed by DTA and XPS
o
analysis. The heat treatment below 300 C or other two methods
could not efficiently remove the PVA and the residual PVA
decreased the catalytic performance. On the other hand, we
found that the colloid solution pH strongly influenced the
distribution and loading content of Au particles. The catalyst
prepared at proper colloid solution pH around isoelectric point of
corresponding support possessed small Au particles and high
content Au leading to a high catalytic performance. The effects of
the support were also investigated and it is found that the Au/ZnO
presented the best activity with the conversion of MAL 99.9% and
The reaction mechanism for the oxidative esterification of
aldehydes with alcohols over Au-based catalysts has been
investigated recently,[
5, 12-13, 19]
and two steps are involved in the
mechanism. Initially, hemiacetal is formed by condensation
reaction between aldehyde and alcohol. Then, the ester is
produced by the elimination of H from hemiacetal. During these
two steps, the atomic oxygen and the basicity of support play key
roles in the oxidative esterification. In our case, the significance of
the basicity of support demonstrated obviously, which could
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201601688
FULL PAPER
the selectivity to MMA 85.9% followed by Au/Al
2
O
3
and the
2
Hydrogen temperature programmed reduction (H -TPR) was conducted
on Autochem II 2920 (Micromeritics, USA). A total of 50 mg catalyst was
filled in a quartz reactor and pretreated under Ar flow (20 mL/min) at 120
oC for 2 h. After the sample cooled to 50 oC, the powder was performed
2 2
Au/TiO and Au/CeO had the far less catalytic performance. The
best performance may be due to the strong basic surface sites
that improved the formation of the intermediate and the strong
interaction between Au and ZnO. This interaction caused the
anchoring effect and the change of geometries of Au particles
which could enhance the stability of catalysts and promote the
oxygen mobility and activate the methanol respectively.
2
with a 10% H /Ar gas mixture (20 mL/min) from 50 to 850 °C (20 °C /min).
The consumption of hydrogen was monitored online using a thermal
conductivity detector. The temperature-programmed desorption of
ammonia and carbon dioxide (NH - and CO -TPD) experiments were also
3
2
carried out on Autochem II 2920 (Micromeritics, USA). The samples were
pretreated under He flow (20 mL/min) at 120 °C for 1 h. The sample was
cooled to 50 °C, and then exposed to 10% NH
mixture flow (20 mL/min) for 1 h. Then, the sample was flushed under Ar
flow at 50 °C to remove physically adsorbed CO or NH until the baseline
is horizontal. A CO -TPD (or NH -TPD) profile of the sample was recorded
by increasing the temperature from 50 °C to 700 °C at a heating rate of
0 °C/min under 20 mL/min of Ar flow. The effluent was monitored online
3 2
/He mixture or 10% CO /He
Experimental Section
2
3
2
3
Catalyst preparation
1
The sol-immobilization method used here has been extensively described
elsewhere with slight modification.[23] The colloidal gold solutions were
prepared using PVA (Degree of Polymerization: 1700, 88% alcoholysis
by a thermal conductivity detector.
Catalytic reaction
4
from Aladdin) as stabilizing ligands and NaBH (Sinopharm Chemicals) as
reductant. In a typical preparation, a required (PVA/Au (wt/wt) = 1.2) the
protecting agent PVA solution (1 wt%) was added to a 1 mmol L-1 aqueous
The oxidative esterification of MAL with methanol to MMA was performed
in a 50 mL steel autoclave. In a typical experiment, the substrate (25 mmol)
and methanol (20 mL) were added into the reactor pre-charged with 0.50
gold solution (as HAuCl
temperature under vigorous stirring. The obtained solution was then left
under stirring for 5 min. A freshly prepared aqueous solution of NaBH (0.1
M, NaBH /Au (mol/mol) = 5) was then added to form a dark orange-brown
4
Sinopharm Chemicals, 99.99%) at room
g catalyst. After O
to 80 oC, and then the catalytic reaction was started by vigorous stirring
800 rpm). After 2 h, the stirring and introducing oxygen were immediately
2
charged with 0.3 MPa pressure, the mixture was heated
4
4
(
gold sol. After 30 min of sol generation, the colloid gold solution was
immobilized by adding the support under vigorous stirring. The amount of
support material required was calculated so as to have a total final metal
loading of 1 wt%. It is noteworthy that the pH of colloid solution needs to
be adjusted to the value below the isoelectric point of the support using the
stopped. Sequentially, the reactor was quickly cooled down to room
temperature. Next, the excess oxygen was depressurized slowly. The
products were analyzed using a gas chromatograph equipped with an FID
detector and a capillary column (DB-624, 30 m × 0.32 mm × 0.25 μm) using
ethanol as an internal standard for quantification.
0
.1 mol L-1 HCl solution. After 1 h, the slurry was collected by vacuum
filtration and the catalyst was washed thoroughly with 2L of doubly distilled
water to remove all the dissolved species. Finally, the solids were dried
o
under a vacuum condition at 80 C overnight. All supports (TiO
2
P25 from
Acknowledgements
Degussa, SiO
2
from Aldrich , γ-Al
were pre-treated at 400 C before used.
O
2 3
、CeO
2
、ZrO
2
and from Aladdin)
o
This work was supported by the National Basic Research
Program of China (2015CB251401), the "Strategic Priority
Research Program" of the Chinese Academy of Sciences
(XDA07070600), the National Instrumentation Grant Program
(2011YQ12003907) and Special Funds of the National Natural
Science Foundation of China (21127011).
To remove the PVA agent, three different ways were employed as
reported: The calcined catalysts were pre-treated at 100-300 oC under
o
[24]
static air for 4 h using a heating rate 5 C/min; For the hot water washing,
o
the catalysts was washed with 90 C hot water at the final step and then
dried at 80 oC overnight;[9b] For the reflux method, the prepared catalyst
was added into the round bottom flask connected to a reflux condenser
o
and placed in an water bath at 90 C, under vigorous stirring; The solution
Keywords: supported Au catalysts • oxidative esterification •
methyl methacrylate • PVA protecting agent • colloid deposition
was left to reflux for 2 h and then the solids were dried under a vacuum
condition at 80 oC overnight.[9a]
.
[
[
1]
2]
K. Nagai, Applied Catalysis A: General 2001, 221, 367-377.
W. W. Cooper, H. Hemphill, Z. Huang, S. Li, V. Lelas, D. W. Sullivan,
European Journal of Operational Research 1997, 96, 1-35.
J. Han, S. Zhang, J. Zhang, R. Yan, Rsc Advances 2014, 4, 58769-58772.
K. Suzuki, T. Yamaguchi, K. Matsushita, C. Iitsuka, J. Miura, T. Akaogi,
H. Ishida, Acs Catalysis 2013, 3, 1845-1849.
Catalyst characterization
Nitrogen adsorption/desorption isotherms at 77 K were measured using
Micromeritics ASAP 2460. Prior to the measurement, all samples were
degassed for 4 h at 573 K. The specific surface area, SBET was calculated
using BET equation. Transmission electron microscopy (TEM) was
performed using JEOL TEM 2011 operated at 200 kV. Samples were
deposited on the TEM grids after ultrasonic dispersion in ethanol. The Au
loading was determined by inductive couple plasma optical emission
spectrometer (ICP-OES) by IRIS Intrepid II XSP (Thermofisher, USA). X-
ray photoelectron spectroscopy (XPS) studies were conducted on a
PHI5700 spectrometer applying a monochromated Al Kα source. The
binding energy scale was corrected for surface charging by considering
the C1s peak of contaminant carbon as a reference at 285.0 eV.
[
[
3]
4]
[
[
5]
6]
X. Wan, W. Deng, Q. Zhang, Y. Wang, Catalysis Today 2014, 233, 147-
154.
F.-Z. Su, J. Ni, H. Sun, Y. Cao, H.-Y. He, K.-N. Fan, Chemistry-a
European Journal 2008, 14, 7131-7135.
[
[
7]
8]
M. Haruta, CATTECH 2002, 6, 102-115.
M. Comotti, W.-C. Li, B. Spliethoff, F. Schüth, Journal of the American
Chemical Society 2006, 128, 917-924.
[
9]
a) J. A. Lopez-Sanchez, N. Dimitratos, C. Hammond, G. L. Brett, L.
Kesavan, S. White, P. Miedziak, R. Tiruvalam, R. L. Jenkins, A. F. Carley,
D. Knight, C. J. Kiely, G. J. Hutchings, Nat Chem 2011, 3, 551-556; b) X.
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201601688
FULL PAPER
Wang, B. Meng, X. Tan, X. Zhang, S. Zhuang, L. Liu, Industrial &
[18] N. Lopez, J. K. Nørskov, T. V. W. Janssens, A. Carlsson, A. Puig-Molina,
B. S. Clausen, J. D. Grunwaldt, Journal of Catalysis 2004, 225, 86-94.
[19] B. Xu, J. Haubrich, C. G. Freyschlag, R. J. Madix, C. M. Friend, Chemical
Science 2010, 1, 310.
Engineering Chemistry Research 2014, 53, 5636-5645.
[
[
10] Y.-H. Ke, X.-X. Qin, C.-L. Liu, R.-Z. Yang, W.-S. Dong, Catalysis Science
&
Technology 2014, 4, 3141.
11] a) M. Valden, X. Lai, D. W. Goodman, Science 1998, 281, 1647-1650; b)
F. Pinna, A. Olivo, V. Trevisan, F. Menegazzo, M. Signoretto, M. Manzoli,
F. Boccuzzi, Catalysis Today 2013, 203, 196-201.
[20] A. Roldan, J. M. Ricart, F. Illas, G. Pacchioni, Physical chemistry
chemical physics : PCCP 2010, 12, 10723-10729.
[21] F. Menegazzo, M. Signoretto, F. Pinna, M. Manzoli, V. Aina, G. Cerrato,
F. Boccuzzi, Journal of Catalysis 2014, 309, 241-247.
[
[
12] B. Xu, X. Liu, J. Haubrich, C. M. Friend, Nature Chemistry 2010, 2, 61-
65.
[22] M. Manzoli, F. Menegazzo, M. Signoretto, G. Cruciani, F. Pinna, Journal
of Catalysis 2015, 330, 465-473.
13] a) Y. Li, L. Wang, R. Yan, J. Han, S. Zhang, Catalysis Science &
Technology 2015; b) Y. Li, L. Wang, R. Yan, J. Han, S. Zhang, Catalysis
Science & Technology 2016, 6, 5453-5463.
[23] a) L. Kesavan, R. Tiruvalam, M. H. Ab Rahim, M. I. bin Saiman, D. I.
Enache, R. L. Jenkins, N. Dimitratos, J. A. Lopez-Sanchez, S. H. Taylor,
D. W. Knight, C. J. Kiely, G. J. Hutchings, Science 2011, 331, 195-199;
b) J. Pritchard, M. Piccinini, R. Tiruvalam, Q. He, N. Dimitratos, J. A.
Lopez-Sanchez, D. J. Morgan, A. F. Carley, J. K. Edwards, C. J. Kiely,
G. J. Hutchings, Catal. Sci. Technol. 2013, 3, 308-317.
[
[
14] E. D. Park, J. S. Lee, Journal of Catalysis 1999, 186, 1-11.
15] a) X. Liu, M.-H. Liu, Y.-C. Luo, C.-Y. Mou, S. D. Lin, H. Cheng, J.-M.
Chen, J.-F. Lee, T.-S. Lin, Journal of the American Chemical Society
2012, 134, 10251-10258; b) F. Wang, Y. Jiang, A. Gautam, Y. Li, R. Amal,
ACS Catalysis 2014, 4, 1451-1457; c) J. Zhang, X. Liu, S. Wu, B. Cao,
S. Zheng, Sensors and Actuators B: Chemical 2012, 169, 61-66.
[24] a) H. Yin, C. Wang, H. Zhu, S. H. Overbury, S. Sun, S. Dai, Chemical
Communications 2008, 4357-4359; b) L. Wen, J.-K. Fu, P.-Y. Gu, B.-X.
Yao, Z.-H. Lin, J.-Z. Zhou, Applied Catalysis B: Environmental 2008, 79,
402-409.
[
[
16] M. S. Chen, D. W. Goodman, Catalysis Today 2006, 111, 22-33.
17] a) M. Mavrikakis, P. Stoltze, J. K. Nørskov, Catalysis Letters 2000, 64,
1
01-106; b) M. Cargnello, V. V. T. Doan-Nguyen, T. R. Gordon, R. E.
Diaz, E. A. Stach, R. J. Gorte, P. Fornasiero, C. B. Murray, Science 2013,
41, 771-773.
3
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201601688
FULL PAPER
Entry for the Table of Contents (Please choose one layout)
Layout 1:
FULL PAPER
]
The Au supported on ZnO catalyst
prepared by colloid deposition presetned
high performance on the oxidative
esterification of methacrolein to methyl
methacrylate due to the strong
Yuchao Li, * Yanxia Zheng, Lei Wang
Zhongjun Fu
Page No. – Page No.
Oxidative esterification of
interaction between Au and ZnO.
methacrolein to methyl methacrylate
over supported gold catalysts
prepared by colloid deposition
This article is protected by copyright. All rights reserved.