Hydrogenation of methyl acetate to ethanol over a highly stable Cu/SiO2 catalyst: Reaction mechanism and structural evolution

Article abstract of DOI:10.1016/j.apcata.2016.12.006

A Cu/SiO₂ catalyst with a core-shell structure was found to be highly active and stable for the hydrogenation of methyl acetate to ethanol during 96?h reaction test at 523?K; MA conversion reached at 95% while the selectivity towards the major products (ethanol and methanol) approached ～95%. Structural analysis revealed that most copper particles in the hydrogen-reduced Cu/SiO₂ catalyst had sizes of 3–7?nm and were coated with a thin layer of silica, forming a core-shell structure. The average size of copper particles and the core-shell structure kept unchanged in the spent catalyst, but the Cu⁺/(Cu⁰?+?Cu⁺) ratio increased during the course of reaction, probably because of the electronic interaction between copper surface and methyl acetate under the reaction conditions. Mechanistic and kinetic studies have identified that ethyl acetate was formed through trans-esterification of initially produced ethanol with methyl acetate, while the subsequent hydrogenation of ethyl acetate to ethanol proceeded much faster. The Cu⁺/(Cu⁰?+?Cu⁺) ratio played a crucial role in reaction network; MA could adsorb equally on Cu⁰ and Cu⁺ sites while the activation of molecular hydrogen occurs only on the Cu⁰ site. Therefore, a rather stable performance could be attained with the increase in the Cu⁺/(Cu⁰?+?Cu⁺) ratio during the course of reaction as long as the reaction is kinetically controlled by MA adsorption. The outstanding stability of the Cu/SiO₂ catalyst was ascribed to a combination of geometric and electronic effect.

Full text of DOI:10.1016/j.apcata.2016.12.006

Accepted Manuscript
Title: Hydrogenation of Methyl Acetate to Ethanol over a
Highly Stable Cu/SiO Catalyst: Reaction Mechanism and
2
Structural Evolution
Author: Xiumin Huang Meng Ma Shu Miao Yanping Zheng
Mingshu Chen Wenjie Shen
PII:
S0926-860X(16)30592-0
DOI:
Reference:
http://dx.doi.org/doi:10.1016/j.apcata.2016.12.006
APCATA 16091
To appear in:
Applied Catalysis A: General
Received date:
Revised date:
Accepted date:
22-6-2016
22-11-2016
7-12-2016
Pleasecitethisarticleas:XiuminHuang, MengMa, ShuMiao, YanpingZheng, Mingshu
Chen, Wenjie Shen, Hydrogenation of Methyl Acetate to Ethanol over a Highly Stable
Cu/SiO2 Catalyst: Reaction Mechanism and Structural Evolution, Applied Catalysis
A, General http://dx.doi.org/10.1016/j.apcata.2016.12.006
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

Hydrogenation of Methyl Acetate to Ethanol over a Highly Stable Cu/SiO
2
Catalyst: Reaction Mechanism and Structural Evolution
a,
a
a
b
b
a,
Xiumin Huang *, Meng Ma , Shu Miao , Yanping Zheng , Mingshu Chen , Wenjie Shen *
aState Key Laboratory of Catalysis, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian
16023, China.
1
bState Key Laboratory of Physical Chemistry of Solid Surfaces, Department of Chemistry, Xiamen University,
Xiamen 361005, China.
*
Corresponding authors. Email addresses: hxiumin@dicp.ac.cn; shen98@dicp.ac.cn
1

Graphical Abstract
+
H
2
CH COOCH
CH CH OH
3 2
3
3
Research Highlights


A core-shell structured Cu/SiO
of 1 nm was developed.
2
catalyst with a copper particle of 4-7 nm and a SiO shell
2
The catalyst was highly active and stable for methyl acetate hydrogenation at 523 K.
Methyl acetate conversion reached at 95% while the selectivity towards ethanol and
methanol approached ~95%.
+


The number of surface Cu sites increased during the course of reaction due to the
competitive adsorption of methyl acetate and hydrogen.
+
0
+
A proper Cu /(Cu +Cu ) ratio is the prerequisite for achieving a higher catalytic
performance of durable copper catalysts.
2

Abstracts
2
A Cu/SiO catalyst with a core-shell structure was found to be highly active and stable for the
hydrogenation of methyl acetate to ethanol during 96 hours reaction test at 523 K; MA conversion
reached at 95% while the selectivity towards the major products (ethanol and methanol) approached
~
2
95%. Structural analysis revealed that most copper particles in the hydrogen-reduced Cu/SiO catalyst
had sizes of 3-7 nm and were coated with a thin layer of silica, forming a core-shell structure. The
average size of copper particles and the core-shell structure kept unchanged in the spent catalyst, but
+
0
+
the Cu /(Cu +Cu ) ratio increased during the course of reaction, probably because of the electronic
interaction between copper surface and methyl acetate under the reaction conditions. Accordingly, the
2
outstanding stability of the Cu/SiO was ascribed to the combination of geometric effect and electronic
interaction. Mechanistic and kinetic studies have identified that ethyl acetate was formed through
trans-esterification of initially produced ethanol with methyl acetate, while the subsequent
+
0
+
hydrogenation of ethyl acetate to ethanol proceeded very fast. The Cu /(Cu +Cu ) ratio played a crucial
0
+
role in reaction network; MA could adsorb equally on Cu and Cu sites while the activation of
0
molecular hydrogen occurs only on the Cu site. Therefore, a rather stable performance could be
+
0
+
attained with the increase in the Cu /(Cu +Cu ) ratio during the course of reaction as long as the
reaction is kinetically controlled by MA adsorption.
2
Keywords: Cu/SiO catalyst; Core-shell structure; Methyl acetate; Hydrogenation; Stability.
1
. Introduction
Ethanol is an important industrial feedstock for the synthesis of fine chemicals and is also used as fuels
either directly or blended with gasoline. Recently, an integrated process for the production of ethanol
from dimethyl ether (DME) was proposed, which is characterized by the use of cheap catalysts and the
high reaction efficiency [1-3]. It consists of DME carbonylation to methyl acetate (MA) on zeolites and
methyl acetate hydrogenation to ethanol over copper catalysts. DME carbonylation to MA occurred
over mordenites with rather high selectivity and stability [4-9]; while MA hydrogenation to ethanol
could be attained by using copper catalysts [1, 10, 11], primarily because they are highly active for
C=O bond hydrogenation but much less active for C-C bond cleavage [12]. Mechanistic and kinetic
3

studies on MA hydrogenation have suggested that MA dissociated into methoxy and acetyl species
upon adsorption on the copper catalysts; the CH O* species were readily converted into CH OH while
3
3
the acetyl species adsorbed strongly on the catalyst surface, leaving hydrogenation of acetyls as the
rate-determining step [10].
Cu/SiO
wt.%) in order to promote the overall reaction efficiency. However, sintering of copper particles at
elevated temperatures and under hydrogen atmosphere caused rapid deactivation of the Cu/SiO
2
catalysts are typically used for MA hydrogenation at relatively higher copper loadings (> 30
2
catalysts. With this regard, enhancing copper-silica bonding becomes the crucial issue for maintaining
their higher dispersion at high copper loadings and to alleviate their sintering under hydrogenation
reaction conditions. For example, highly dispersed Cu particles with an average diameter of about 6 nm
not only exhibited a much higher activity but also were rather resistance to sintering during glycerol
hydrogenolysis [13, 14]. This was due to a strong interaction between Cu2+ and silica in the as-calcined
sample through the formation of copper phyllosilicates. Surface modification of SiO
commonly adapted to enhance its chemical bonding with copper particles. Doping of SiO surface with
[15-17], La [14, 18], NiO [19] and CeO [20] all alleviated Cu sintering to some extents in
Cu/SiO catalysts for the hydrolysis of dimethyl oxalate or methyl acetate. However, the interaction
2
has been
2
B
2
O
3
2
O
3
2
2
mechanism between copper and silica and their structural variation under the hydrogenation reaction
conditions are rarely examined.
In this work, we developed a core-shell structured Cu/SiO
and examined the structural evolution of the copper particles during MA hydrogenation. The Cu/SiO
catalyst showed excellent activity and stability towards methyl acetate hydrogenation to ethanol at 523
2
catalyst with a copper loading of 38 wt.%
2
+
0
+
K, and the surface Cu /(Cu +Cu ) ratio played an essential role in the adsorption and activation of
methyl acetate and hydrogen.
2
. Experimental
2
.1. Catalyst preparation
The Cu/SiO
2
catalyst was prepared by a precipitation-gel method, similar to a previous procedure [13].
·6H O was dissolved into 57 ml deionized water; 25 mL 5 M NaOH aqueous solution
6
.8 g Cu(NO
3
)
2
2
was then added at a rate of 0.7 ml/min under vigorous stirring at room temperature with a final pH
value of ~12.0, forming a blue precipitate. 4.18 g silica (Degussa A 200) was mixed with the slurry at
4

room temperature, and the mixture was further aged at 363 K for 4 h. The resulting solid was collected
by filtration, thoroughly washed with hot water, dried at 393 K overnight, and finally calcined at 723 K
for 3 h in air. The actual loading of copper was 38.3 wt. % measured by ICP analysis.
2
.2. Catalytic tests
Hydrogenation of methyl acetate was tested with a fixed-bed stainless steel tubular reactor with an
inner diameter of 8 mm. 300 mg catalysts (40-60 mesh) were loaded and reduced with pure H (18
mL/min) at 573 K for 5 h. After the temperature was lowered to 423 K, the reactor system was
pressurized to 1.0 MPa at a H flow rate of 34 mL/min. Liquid methyl-acetate was injected through a
2
2
micro-pump at a flow rate of 4 µl/min, corresponding to a MA concentration of ～3.0 vol.% in the
feed stream. The reaction was then tested at desired temperatures (493-528 K). All the pipelines were
heated at 353 K to avoid the condensation of the reactants and products. The exit stream from the
reactor was analyzed by an online gas chromatograph (Agilent 6890N) equipped with a thermal
2
conductivity detector (TCD) and a flame ionization detector (FID). An inert N was adapted to
maintain a constant flowrate in the reaction order tests. Conversion of MA is calculated via the amount
of methyl acetate reacted divided by the amount of methyl acetate in the feed stream. The selectivity of
a product is calculated as the stoichiometric number multiplies the mole amount of this product and
then divided by the total number of products formed (the sum of the stoichiometric number multiplies
the mole amount of each product). All the reaction data in this work was carefully checked by making
carbon balance. The hydrogen-reduced and spent samples were collected after passivation in a 1%
O
2
/N
.3. Catalyst characterizations
The copper loading was measured by inductively coupled plasma atomic emission spectroscopy
ICP-AES) on a Plasma-Spec-II instrument. 20 mg as-calcined CuO/SiO sample was dissolved in
2
flow at room temperature for further characterizations.
2
(
2
concentrated hydrochloric acid and hydrogen fluoride, and the mixture was diluted with water to meet
the detection range of the instrument.
The specific surface area and pore size of the as-calcined CuO/SiO
adsorption-desorption isotherms using an Autosorb-1 instrument (Quantachrome, USA) at 77 K. Before
measurement, the sample was degassed in vacuum at 573 K for 6 h. The specific surface area (SBET
2
sample was measured by nitrogen
)
was estimated by the Brunauer-Emmett-Teller (BET) method, while the mesopore size distribution was
5

calculated from the desorption branch of the isotherm using the BJH method.
Power X-ray diffraction (XRD) patterns were recorded on a D/Max-2500/PC diffractometer (Rigaku,
Japan) using a Cu Kα (λ = 0.154 nm) radiation source that operated at 40 kV and 200 mA. In situ XRD
measurements were tested with the same instrument at 40 kV and 300 mA. 200 mg CuO/SiO
2
sample
was pressed into a flake, mounted in a chamber, exposed to a 5% H /N mixture (30 mL/min) and
2
2
gradually heated to 573 K for 5 h at a rate of 5 K/min. XRD patterns were recorded at desired
0
temperatures. The average crystallite sizes of CuO and Cu in the as-calcined and hydrogen-reduced
samples was estimated from the full widths at half maxima of their typical diffractions according to the
Scherrer’s equation.
Transmission electron microscopy (TEM) images were recorded on a FEI Tecnai G2 Spirit microscope
operated at 120 kV; High-resolution TEM (HRTEM) images were recorded on a JEM-2100 microscope;
HAADF-STEM images were recorded on a JEM-ARM200F microscope. The specimen was prepared
by ultrasonically dispersing the solid sample into ethanol, and droplets of the suspensions were
deposited on a carbon-coated copper or molybdenum grid and then dried in air.
Temperature-programmed reduction with hydrogen (H
performed with an AutoChem II 2920 instrument (Micromeritics) equipped with a TCD. 30 mg
CuO/SiO sample was pretreated at 723 K (10 K/min) for 2 h with a 20% O /N mixture (30 mL/min)
to remove the surface impurities. After cooling down to room temperature and purging with Ar, the
sample was reduced with a 5% H /N mixture (30 mL/min) at 573 K at a rate of 5 K/min for 1 h. The
resulting Cu/SiO catalyst was purged with Ar flow and exposed to a 10% N O/He mixture (30 mL/min)
at 363 K for 30 min. A second TPR run was then conducted by heating the sample to 573 K at a rate of
K/min and under the flow of a 5% H /N mixture (30 mL/min). The dispersion (D) and average size
d) of copper particles were estimated according to the following equations:
D = (2·A /A ) ×100%, d = 0.5 A /A =1/D (nm)
Where A and A represented the amounts of hydrogen consumed during the first and second TPR runs,
respectively. The average diameter (d) of copper particles was calculated from A
2 2
-TPR) and dissociative N O titration was
2
2
2
2
2
2
2
5
2
2
(
2
1
1
2
1
2
1
and A
2
by assuming
.46 × 1019 copper atoms per m and a molar stoichiometry of N
copper atoms on the surface [21, 22].
Temperature-programmed desorption of methyl acetate (MA-TPD) on the Cu/SiO
2
O/Cu = 0.5, in which Cu means the
s
s
1
2
2
catalysts was also
conducted on the AutoChem II 2920 instrument. The generation of Cu(0) sample was achieved by
6

reducing the as-calcined CuO/SiO
the Cu(+) sample was produced by further exposing the above Cu(0) sample to a mixture of 10%
O/He at 363 K for 0.5 h. The obtained Cu(0) and Cu(+) samples were cooled down to 313 K, purged
2 2 2
sample (30 mg) with a 5% H /N mixture at 573 K for 1 h; while
N
2
with Ar (30 mL/min) for 0.5 h; and then exposed to a 2% MA/He mixture (20 mL/min) for 40 min.
After sweeping with He (30 mL/min) for 30 min to remove the physically adsorbed species, the
samples were heated to 823 K at a rate of 5 K/min and the desorbed species were monitored with a
TCD.
Temperature-programmed oxidation (TPO) of the spent catalysts was conducted with a U-type quartz
tubular reactor connected to an online mass spectrometer (OmniStar 200, Balzers). 22 mg spent sample
was heated to 873 K at a rate of 10 K/min under He flow (30 mL/min) and kept at that temperature for
1
h. After cooling down to room temperature, the sample was exposed to a 20% O
2
/He mixture (30
mL/min) and heated to 873 K at a rate of 10 K/min. The evolution of CO
and CO (m/e=28) was monitored by the mass spectrometer.
2
(m/e=44), H O (m/e=18),
2
X-ray photoelectron spectroscopy (XPS) and Auger electron spectroscopy measurements were carried
out in a UHV system using an Al Kα radiation source (hν=1486.6 eV) and an Omicron Sphera II
hemispherical electron energy analyzer. Both the as-calcined and spent samples were pressed into a
2
thin disk and reduced with pure H at 483 and/or 573 K for 2 h; the sample was then transferred into
the XPS analysis chamber without exposing to air. The binding energies were calibrated using the Si2p
peak at 103.6 eV as the reference. Deconvolution of Cu LMM Auger spectra was conducted by
keeping the Lorentzian/Gaussian ratio of the baseline constant at 1:99 and the binding energy of Cu+
being 4 eV higher than that of Cu0.
3
. Results and discussion
3
2
.1 Physical and Chemical properties of the Cu/SiO catalysts
Fig. 1 show the N
sample and the pore size distribution. The surface area was about 186 m /g and 196 m /g for bare SiO
and CuO/SiO
accumulation of SiO
value of the CuO/SiO
at the starting point (inset Fig. 1b), suggesting that pores below 3 nm were generated on the CuO/SiO
2
adsorption-desorption isotherm of the SiO
2
support, the as-calcined CuO/SiO
2
2
2
2
2
sample respectively. Besides the original mesopore of 26.1 nm caused by the
particles, no apparent additional mesopores appeared. However, the dV(logD)
sample was not equal to zero and was much higher than that of the SiO support
2
2
2
2
7

2 2
catalyst. Fig. 2 shows the XRD patterns of the Cu/SiO catalysts. The CuO/SiO sample showed
characteristic diffraction lines of CuO with a mean crystalline size of ~ 5.5 nm. After hydrogen
reduction at 573 K for 5 h, only the metallic Cu crystallite was detected. Fig. 3 displays the TEM
2
images of the as-calcined CuO/SiO sample. Both rounded particles of about 20 nm and nanorods with
lengths up to several hundreds of nanometers (200-500 nm) were clearly observed. HRTEM analysis
on the nanorods further identified that the CuO particles were arranged in a pearl-threaded pattern and
2
surrounded by a thin layer of amorphous SiO . The amorphous layer was clearly seen as the incident
o
angle of electrons was rotated for 30 , suggesting the formation of a core-shell structure on the rod-like
section. The formation of such a rod-like structure might be assigned to the arrangement of CuO
particles, which originated from the Cu(OH)
structure under highly basic solutions during the preparation. Fig. 4 shows the TEM images of the
Cu/SiO catalyst that was obtained by hydrogen reduction at 573 K of the as-calcined CuO/SiO sample.
Apparently, the Cu/SiO catalyst inherited the morphology of the precursor. Highly dispersed Cu
particles with a diameter of ~5 nm were clearly observed; most copper particle had sizes in the range of
-7 nm; small particles of 1-3 nm and large particles of ~10 nm were also observed occasionally. More
importantly, most copper particles were uniformly coated with a thin layer of silica with a thickness of
1 nm, confirming the core-shell structure on the highly dispersed copper particles. The exact
formation mechanism of the core-shell structure was not clear, but it could be assumed that Cu(OH)
precipitate was initially formed in the strong basic solution, and the later added SiO particles were
partly dissolved and negatively charged under such a high pH solution [18, 23]. The strong electronic
interaction between the positively charged Cu(OH) particles and the negatively charged SiO species
resulted in the enwrap of the Cu(OH) precipitate during the aging process and therefore the core-shell
structure in the CuO/SiO and Cu/SiO samples. To get a clear point of view whether there are
pathways to the core particles through the SiO layer, the CuO/SiO powder sample was immersed in
aqueous H SO solution with a PH value of ~ 1 at room temperature, the CuO particles can be almost
dissolved after 2 h, suggesting that the CuO particles are not totally covered by SiO . Some cracks or
2
precipitate that are prone to form one-dimensional
2
2
2
4
~
2
2
2
2
2
2
2
2
2
2
4
2
micropores should be presented in the thin silica layers and were responsible for the reactant molecules
to access the Cu nanoparticles, in consistent with the adsorption-desorption results that pores below 3
2
nm were exhibited on the CuO/SiO catalyst.
Fig. 5 shows the H -TPR profile of the as-calcined CuO/SiO
2
2
sample. The intense reduction peak at
8

5
2
21 K was ascribed to the reduction of CuO. However, the amount of H consumed (4.67 mmol/g) was
lower than the stoichiometric value (5.46 mmol/g) required for the complete reduction of CuO to
metallic copper. This indicates there presents hard-reducible Cu2+ species, which are most likely those
that strongly interacted with the silica support at the CuO-SiO
2
interface. Therefore, the broad
reduction is actually a collective contribution from the transformation of CuO to Cu and Cu2+ species
0
+
at the CuO-SiO
2
interface to Cu . Estimated from the amount of H
2
consumed, 71.1% copper species
0
+
could be fully reduced into Cu while 28.9% copper species were only partially reduced into Cu . N
2
O
2
titration on the hydrogen-reduced Cu/SiO sample indicated that the dispersion of Cu was 16.0%,
equaling to a mean particle size of 6.2 nm.
Fig. 6 shows the XPS spectra recorded during hydrogen reduction of the as-calcined CuO/SiO
2
sample.
The binding energy of Cu 2p3/2 at 934.7 eV, together with the characteristic shakeup 2p → 3d satellite
2
+
peaks at 942-944 eV [21], evidenced the presence of Cu . Upon hydrogen reduction at 573 K, the
satellite peaks vanished while new binding energies at 952.9 and 932.9 eV appeared, evidencing that
the Cu2+ species were converted into Cu and/or Cu species [15, 18, 24, 25]. Because of the identical
+
0
+
0
binding energies between Cu and Cu , Cu LMM X-ray excited Auger spectroscopy was adapted to
+
distinguish these copper species. The binding energy at 572.2 eV was assigned to Cu while that at
0
+
+
0
+
5
68.2 eV was ascribed to Cu [18, 25]. Accordingly, the remaining Cu species, with a Cu /(Cu +Cu )
ratio of 28.1%, could be attributed to those strongly interacted or incorporated into silica at the Cu-SiO
2
+
interface, which is in good agreement with the H
hydrogen-reduced Cu/SiO
2
-TPR result that 28.9% Cu species presented on the
sample. The absence of Cu+ species by XRD technique for the
2
2
hydrogen-reduced sample might be due to that the Cu species strongly interacted with SiO layer was
+
amorphous, or it was located on the catalyst surface or the Cu particle was too small to be detected,
because XRD technique detects only the crystalline phase in the bulk while XPS probes the surface
species both in crystalline and amorphous form. The surface Cu/Si atomic ratio was estimated to be
0
2
.12 on the as-calcined CuO/SiO catalyst by XPS, far below the theoretical Cu/Si ratio of 0.58 based
on ICP analysis. Considering the relatively small copper size of 5-6 nm, the prominent difference in
Cu/Si ratio as measured by XPS and ICP methods inferred that silica was dominated on the catalyst
surface, an auxiliary evidence for the formation of a core-shell structure.
3
2
.2. Hydrogenation of methyl acetate over the Cu/SiO catalyst
9

Table 1 summarizes the reaction results of methyl acetate hydrogenation over the Cu/SiO
2
catalyst at
4
93-528 K. The main products were ethanol, methanol and ethyl acetate (EtOAc), among which ethyl
acetate was formed through trans-esterification of the initially formed ethanol with methyl acetate; the
by-products included acetaldehyde and alkanes in traceable amounts. The conversion of MA increased
dramatically from 26.5% at 493 K to 96.2% at 528 K; the corresponding selectivity of EtOH increased
prominently from 30.8% to 61.7% while the selectivity of EtOAc dramatically decreased from 32.7%
to 2.5%, suggesting that hydrogenation of EtOAc occurred as the major reaction. As a consequence, the
total selectivity of EtOH and EtOAc kept nearly constant at ~ 64% while the selectivity of MeOH only
marginally decreased from 36.4% to 33.6% with raising the temperature. Concerning the by-products,
CH
acetaldehyde were formed at 518 K and their total selectivity slightly increased to 2.2% at 528 K.
Reaction kinetic of MA hydrogenation on the Cu/SiO catalyst was then examined. The effect of
4
was the only detectable byproduct at 493 K with a selectivity of 0.2%, while ethane and
2
internal diffusion is usually negligible at the used submillimeter particles of 0.25-0.35 mm (40-60
mesh), while the external mass-transfer resistance on the reaction was precluded through primary tests
(
Table 2). Fig. 7 shows the variation in product distribution at different MA conversions at 498 K. The
conversion of MA showed an almost linear correlation with the contact time (W/F) passing through the
origin, indicating that MA hydrogenation is a typical heterogeneous catalytic surface reaction over the
2
Cu/SiO catalyst. At MA conversions less than 28%, the selectivity of EtOH continuously decreased
while the selectivity of ethyl acetate increased with increasing the MA conversion. This suggests that
EtOH is the primary product (Equation 1) while ethyl acetate is the secondary product, which was
produced through trans-esterification of methyl acetate with the initially produced EtOH (Equation 2).
As the conversion of MA was above 28%, the selectivity of EtOH increased while that of EtOAc
decreased with increasing MA conversion. This means that hydrogenation of ethyl acetate occurred
predominantly and ethanol was produced as the secondary product (Equation 3) at this stage. Control
experiments for hydrogenation of methyl acetate and ethyl acetate further confirmed that
hydrogenation of ethyl acetate proceeded much faster than that of methyl acetate over the Cu/SiO
2
catalyst (Table 3), as previously observed [11]. Under identical reaction conditions, the conversion of
MA was only 26.5% with an ethanol selectivity of 30.8% at 493 K, whereas the conversion of EA
approached 77.3% with an EtOH selectivity of 99.4%. At 513 K, MA conversion increased to 81.1%
with a moderate increase in ethanol selectivity (52.9%), while EA conversion attained 91.8% with an
10

ethanol selectivity of 98-99%. Therefore, it is reasonable to conclude that the faster hydrogenation of
ethyl acetate to ethanol was responsible for the greatly enhanced EtOH selectivity at higher MA
conversions. It is worth noting that the selectivity of CH
3
OH was essentially independent of MA
conversion. CH OH can be produced through hydrogenation of MA and trans-esterification of MA with
3
EtOH, but its amount is only correlated with the amount of MA converted, keeping a constant
selectivity in the product slate.
Concerning the minor byproducts, the selectivity of acetaldehyde decreased at lower MA conversions,
3
indicating that it was a primary product through hydrogenation of the CH CO* fragment adsorbed on
the catalyst surface (equation 1-1), as previously proposed [10], and the subsequent hydrogenation of
acetaldehyde to EtOH (equation 1-2) proceeded very fast [26]. At higher MA conversions, however, the
selectivity of acetaldehyde increased, suggesting that acetaldehyde was also formed as a secondary
product through EtOH dehydrogenation (Equation 1-2). Although in traceable amounts, alkanes were
formed at 493 K and their selectivity showed an inverse volcano pattern with MA conversion. Since
CH
decreasing MA conversion, it could be inferred that CH
by direct hydrogenation of the CH * fragments in MA (equation 4). Meanwhile, the formation of
traceable AcOH further evidenced the occurrence of this reaction route. The selectivity of C among
alkanes increased with MA conversion in the whole range, suggesting that ethane was a secondary
product. To identify the formation route of C , hydrolysis of ethanol was tested on the Cu/SiO
4
was the only detected alkane at lower MA conversions and its selectivity decreased with
4
acted as the primary product probably formed
3
2
H
6
2
H
6
2
catalyst (Table 3). Ethane was formed at 498 K although the conversion of ethanol was only 0.6%, and
its amount increased as the reaction temperature further ramped to 518 K. This clearly evidences that
4
ethane was produced by ethanol hydrolysis (equation 5). The fact that there was no detectable CH in
the outlet stream during ethanol hydrolysis excluded the occurrence of ethanol decomposition under the
current reaction conditions. However, acetaldehyde was formed as a major product with a selectivity of
~
65%, suggesting that EtOH dehydrogenation occurred under the current reaction conditions. Besides,
blank tests for MA hydrogenation, MA decomposition and hydrolysis of ethanol revealed the absence
of any products below 523 K, indicating that all the involved reactions were took place on the catalyst
surface. On the basis of these kinetic results, the surface reaction network for MA hydrogenation was
illustrated in Scheme 1.
11

CH
3
COOCH
COOCH
CHO + H
3
+ 2H
+ H
2
→ CH
3
CH
2
OH + CH
3
OH
(1)
(1-1)
(1-2)
(2)
CH
CH
CH
CH
CH
CH
3
3
2
→ CH
CH CH
+ CH CH
CH + 2H
+ H → CH
→ CH
3
2
CHO + CH
OH
3
OH
3
3
3
3
3
2
3
COOCH
COOCH
COOCH
3
2
3
3
2
OH → CH
→ 2CH CH
COOH
3
COOCH
2 3 3
CH + CH OH
3
2
3
2
OH
(3)
2
4
+ CH
3
(4)
CH
2
OH + H
2
3
-CH
3
+ H
2
O
(5)
2
Fig. 8a shows the effect of H partial pressure on the reaction rates and the product selectivity at 513 K.
As increasing hydrogen pressure from 0.4 to 1.0 MPa, the reaction rate of MA increased linearly from
.6 to 8.8 mmol/(gcat.h); the selectivity of ethanol increased remarkably while the selectivity of ethyl
3
acetate deceased dramatically. This indicates that hydrogenation of ethyl acetate to ethanol occurred
more significantly as raising H partial pressure. The reaction order of 0.9 with respect to hydrogen
2
suggests that hydrogen molecule was associatively adsorbed on copper surface, supposing the reaction
follows a Langmuir-Hinshelwood model [10]. Fig. 8b shows the effect of MA partial pressure on the
reaction rate of MA and the product selectivity at 523 K. The reaction rate showed a typical volcano
shape with respect to MA partial pressure. As MA partial pressure was less than 50 kPa (i.e. MA/H
2
<
0
.05), the conversion rate of MA increased continuously with increasing MA partial pressure,
indicating the adsorption of MA might be the rate-determining step. As such, increasing MA
concentration in the reaction stream would cause more MA species adsorbed on the Cu surface,
showing a reaction order of 0.65. This is in accordance with a previous observation that methyl acetate
was adsorbed dissociatively at lower partial pressures with an apparent reaction order of about 0.5 [10].
At MA partial pressures higher than 50 kPa (MA/H
2
> 0.05), however, the reaction rate of MA declined
significantly with increasing MA concentration in the feed stream. Later work on MA-TPD evidenced
0
+
that MA adsorption took place on both Cu and Cu sites of the hydrogen-reduced Cu/SiO
2
catalyst.
-TPA and
adsorption was a strongly activated process started at
desorption corresponding to chemisorbed H on the metallic
happened below 373 K [18, 27]. Therefore, it can be inferred that
in an atmosphere with both MA and H presented, competitive adsorption of MA and H happened on
0
The adsorption of H
2
on Cu surface is generally accepted and has been verified by H
2
H
2
-TPD methods. It was demonstrated that H
00 K on Cu/Al catalyst [27], while H
Cu surface of Cu/SiO and Cu/Al
2
2
2
O
3
2
2
2
2 3
O
2
2
12

0
0
2
the Cu sites of the Cu/SiO catalyst. Accordingly, too much MA adsorption on the active Cu site
decreased the adsorption or activation of hydrogen, thus lowering the overall reaction rate. The reaction
order of -0.89 with respect to MA suggested that MA was dominantly adsorbed on the catalyst surface
in a molecular form under the reaction conditions, differing from the dissociative adsorption at lower
MA partial pressure. The competitive adsorption of hydrogen and methyl acetate implies that hydrogen
molecule was activated on copper surface in the reaction network, instead of gaseous molecular
hydrogen.
3
2
.3. Structural evolution of the Cu/SiO catalyst during MA hydrogenation
2
Fig. 9 shows the result of stability test of the Cu/SiO catalyst for MA hydrogenation at 523 K for ~100
hours. The conversion of MA maintained at 95%; the selectivity for the major products (EtOH, MeOH
and EA) approached ~98% while the selectivity for the minor products (acetaldehyde and alkanes) was
about 2%. This demonstrates the rather high activity and stability of the current Cu/SiO
MA hydrogenation. Fig. 10 shows the temperature-programmed desorption profile of the spent Cu/SiO
catalyst. The intense desorption of CO and H O below 700 K under a He flow suggested that
carbonaceous species were adsorbed on the catalyst surface. The subsequent temperature-programmed
oxidation with a 20%O /N flow did not produce appreciable CO and no obvious consumption of O
2
2
catalyst for
2
2
2
2
2
2
was detected; this confirms that the carbon-containing deposits were completely removed during the
treatment with an inert gas. That is, no hard coke was presented over the spent sample.
2
Fig. 2d shows the XRD pattern of the spent Cu/SiO catalyst after passivation. The average crystalline
size of metallic copper (4.7 nm) was quite close to that in the hydrogen-reduced sample just before
reaction (5.0 nm). This suggests that there was no detectable sintering of copper particles during methyl
acetate hydrogenation. However, a prominent diffraction line at 2-theta value of 36.4 appeared,
2 2
corresponding to Cu O, indicating that part of the metallic copper was transformed into Cu O during
the course of reaction. TEM/STEM analyses on the spent catalyst suggested that there was no apparent
aggregation of Cu particles during MA hydrogenation (Fig. 11), in consistent with the XRD result.
Most copper particles were still in the size range of 4-7 nm, together with few small particles of 1-3 nm
and large ones of ~10 nm. Moreover, the core-shell structure of a thin layer of SiO
also observed. The H -TPR profile of the spent Cu/SiO catalyst was shown in Fig. 5. The passivated
sample after the stability test was treated with a 20% O /N mixture at 723 K for 2 h, followed by
2
on Cu particles was
2
2
2
2
13

2 2
H -TPR and N O titration measurements. The dispersion of Cu was 15.8%, equaling to a mean particle
size of 6.3 nm. This is comparable with the hydrogen-reduced sample prior to the reaction with a Cu
dispersion of 16.0%. All these results confirmed that copper particles kept rather stable under MA
hydrogenation conditions without obvious change in size.
2
The XPS spectra of the spent Cu/SiO catalyst were compared in Fig. 6. For the passivated spent
+
0
sample, the main binding energies of Cu2p at 932.9 and 952.9 eV represented Cu /Cu species, while
the minor satellite peak for Cu2+ indicated surface oxidation during the passivation treatment.
Hydrogen treatment at 483 K totally removed those surface oxidized species and fully regenerated the
+
0
+
0
+
catalyst surface, leaving Cu and Cu as the surface copper species. However, the Cu /(Cu +Cu ) ratio
on the regenerated catalyst (38.9%) is much higher than that on the hydrogen-reduced Cu/SiO catalyst
prior to reaction (28.1%). As the passivated spent sample was further reduced at 573 K with hydrogen,
2
+
0
+
the Cu /(Cu +Cu ) ratio decreased to 31.6%; which is still higher than that on the hydrogen-reduced
+
Cu/SiO
2
catalyst. This implies that a substantial amount of Cu species were generated during the
course of reaction, probably involving the surface oxidation of copper particles and/or copper species
at the Cu-SiO
2
interface by methyl acetate.
+
0
It is generally accepted that the Cu site stabilizes methoxy group while the Cu site catalyzes
hydrogenation of acyl species during MA hydrogenation [15, 24, 25, 28-30]. To elucidate the catalytic
0
+
roles of Cu and Cu species in MA hydrogenation, temperature-programmed desorption of MA on the
Cu/SiO
2
catalysts containing different copper species was examined (Fig. 12). It should be noted that
+
the Cu(0) sample contained about 28.9% Cu owing to the presence of copper strongly interacted with
+
the silica support, while the Cu(+) sample composed of nearly 100% Cu on the surface. The two
samples showed almost identical MA desorption patterns at 528 K, suggesting that MA adsorbed
0
+
equally on the Cu and Cu sites of the copper particles. Here, the similar adsorption behavior of MA
0
+
over Cu and Cu sites implies that their ability for H
As demonstrated by the kinetic measurements, MA and H
sites at the beginning of the reaction. It has been generally known that hydrogen molecules adsorbed
only on metallic copper with a weak interaction. Desorption of H took place in a temperature range of
70-350 K [18, 27]. Our experiment data showed MA adsorbed equally on Cu and Cu sites and MA
2
activation would control the overall activity.
2
were competitively adsorbed on the copper
2
+
0
2
0
desorption occurred at a temperature of 528 K. This indicates that the adsorption of MA on Cu site
0
was much stronger than that of H
2
. As a consequence, Cu site covered by MA was gradually oxidized
14

+
into Cu during the hydrogenation process through electronic chemical interaction between metallic
copper surface and the strongly adsorbed oxygen-containing groups of the esters.
In catalytic reactions, the promotion or degradation in activity is eventually controlled by the rate
determining step at the used reaction conditions. During the course of stability test shown in Fig. 9, the
+
0
+
surface Cu /(Cu +Cu ) ratio increased from 28.1% to 38.9%, but rather high stability of the reaction
was maintained. This can be explained by the lower MA/H molar ratio adopted here. At MA/H molar
ratio of 0.30, MA adsorption acted as the rate-determining step based on the kinetic study shown in Fig.
2
2
+
0
8
b. Considering the fact that MA adsorbed equally on the Cu and Cu sites, the catalytic activity
+
would be kept unchanged provided that newly generated Cu sites were still responsible for MA
0
activation and the remaining Cu sites were still sufficient for hydrogen activation, as shown in Fig. 9.
If the reaction is kinetically controlled by H
2
2
activation, for example, at higher MA/H ratios in the
+
feed stream, a gradual enhancement in the number of Cu sites on the catalyst surface would hinder the
activation of H
2
on metallic copper surface and consequently caused deactivation. Therefore, a proper
Cu /(Cu +Cu ) ratio is the prerequisite for achieving a higher catalytic performance, which is related
with the adopted reaction conditions, in particular the MA/H molar ratio. Moreover, the core-shell
structure of the Cu/SiO
catalyst efficiently prohibited the aggregation of copper particles. The Cu+
species formed during the course of reaction might also take some role on relieving the coagulation of
+
0
+
2
2
+
0
copper particles owing to the fact that Cu is more difficult to coagulation than that of Cu [31]. That is,
the outstanding stability of the Cu/SiO
effects.
2
catalyst is actually a combination of geometric and electronic
4
. Conclusions
A highly stable Cu/SiO
gel method. TEM/STEM analyses on the as-calcined CuO/SiO
particles were in the size range of 3-7 nm and coated with a thin layer of silica. 71% copper species in
the core was loosely interacted with silica while 29% copper species at the CuO-SiO interface was
2
catalyst with a core-shell structure was prepared using a simple precipitation-
2
sample identified that most CuO
2
strongly interacted with silica. Upon hydrogen reduction at 573 K, the core CuO particles were fully
reduced into metallic copper species while the Cu2+ species at the CuO-SiO
2
interface were only
+
+
0
+
partially reduced into Cu , giving a Cu /(Cu +Cu ) ratio of 28.1%. The Cu/SiO
2
catalyst showed
outstanding activity and stability for methyl acetate hydrogenation at 523 K for ~100 h on-stream; the
15

conversion of MA maintained at 95% with a total selectivity of 95% for ethanol and methanol. The size
of copper particles kept unchanged and the core-shell structure of copper particles was preserved.
+
0
+
However, the ratio of surface Cu /(Cu +Cu ) increased to 38.9% due to the electronic interaction of
copper particles and/or copper species at the Cu-SiO interface with methyl acetate under the reaction
conditions. Accordingly, the outstanding stability of the Cu/SiO catalyst was assigned to a
2
2
+
0
+
combination of geometric (the core-shell structure) and electronic (a proper Cu /(Cu +Cu ) ratio)
effects. This result offers a general strategy for the rational design of durable copper catalyst.
Acknowledgements
This work was supported by the National Key Basic Research Program of China (2013CB933100)
and the National Natural Science Foundation of China (21533009).
References
[
[
1] X.G. San, Y. Zhang, W.J. Shen, N. Tsubaki, Energy & Fuels 23 (2009) 2843-2844.
2] X.G. Li, X.G. San, Y. Zhang, T. Ichii, M. Meng, Y.S. Tan, N. Tsubaki, ChemSusChem 3 (2010)
1192-1199.
[3] P. Haro, P. Ollero, A.L. Villanueva Perales, C. Reyes Valle, Energy 44 (2012) 891-901.
[4] P. Cheung, A. Bhan, G. J. Sunley, E. Iglesia, Angew. Chem. Int. Ed. 45 (2006) 1617-1620.
[5] P. Cheung, A. Bhan, G. J. Sunley, D. J. Law, E. Iglesia, J. Catal. 245 (2007) 110-123.
[6] J.L. Liu, H.F. Xue, X.M. Huang, P.H. Wu, S.J. Huang, S.B. Liu, W.J. Shen, Chin. J. Catal. 31
(
2010) 729-738.
[
[
[
[
7] H. Zhou, W.L. Zhu, L. Shi, H.C. Liu, S.P. Liu, S.T. Xu, Y.M. Ni, Y. Liu, L.N. Li and Z.M. Liu,
Catal. Sci. Technol. 5 (2015) 1961-1968.
8] Y.H. Liu, N. Zhao, H. Xian, Q.P. Cheng, Y.S. Tan, N. Tsubaki, X.G. Li, ACS Appl. Mater.
Interfaces 7 (2015) 8398-8403.
9] H.F. Xue, X.M. Huang, E. Ditzel, E.S. Zhan, M. Ma, W.J. Shen, Ind. Eng. Chem. Res. 52 (2013)
11510-11515.
10] A.K. Agarwal, N.W. Cant, M.S. Wainwright, D.L. Trimm, J. Mol. Catal. 43 (1987) 79-92.
16

[
11] M. A. Nata Santiago, M. A. S´anchez-Castillo, R. D. Cortright, J. A. Dumesic, J. Catal. 193(2000)
6-28.
1
[
[
12] Z.W. Huang, F. Cui, J.J. Xue, J.L. Zuo, J. Chen, C.G. Xia, Catal. Today 183 (2012) 42-51.
13] Z.W. Huang, F. Cui, H.X. Kang, J. Chen, X.Z. Zhang, C.G. Xia, Chem. Mater. 20 (2008)
5
090-5099.
[
[
[
14] Z.W. Huang, H.L. Liu, F. Cui, J.L. Zuo, J. Chen, C.G. Xia, Catal. Today 234 (2014) 223-232.
15] Z. He, H.Q. Lin, P. He, Y.Z. Yuan, J. Catal. 277 (2011) 54-63.
16] S. Zhao, H.R. Yue, Y.J. Zhao, B. Wang, Y.C. Geng, J. Lv, S.P. Wang, J.L. Gong, X.B. Ma, J. Catal.
2
97 (2013) 142-150.
[
[
[
[
[
17] S.H. Zhu, X.Q. Gao, Y.L. Zhu, Y.F. Zhu, H.Y. Zheng, Y.W. Li, J. Catal. 303 (2013) 70-79.
18] X.L. Zheng, H.Q. Lin, J.W. Zheng, X.P. Duan, Y.Z. Yuan, ACS Catal. 3 (2013) 2738-2749.
19] A.Y. Yin, C. Wen, X.Y. Guo, W.L. Dai, K.N. Fan, J. Catal. 280 (2011) 77-88.
20] C.L. Ye, C.L. Guo, J.L. Zhang, Fuel Process. Technol. 143 (2016) 219-224.
21] C. J. G. Van Der Grift, A. F. H. Wielers, B. P. J. Joghi, J. Van Beijnum, M. De Boer, M.
Versluijs-Helder, J. W. Geus, J. Catal. 131 (1991) 178-189.
[
[
22] J. W. Evans, M. S. Wainwright, A. J. Bridgewater, D. J. Yong, Appl. Catal. 7 (1983) 75-83.
23] Z.W. Huang, F. Cui, J.J. Xue, J.L. Zuo, J. Chen, C.G. Xia, J. Phys. Chem. C 114 (2010)
1
6104-16113.
24] L.F. Chen, P.J. Guo, M.H. Qiao, S.R. Yan, H.X. Li, W. Shen, H.L. Xu, K.N. Fan, J. Catal. 257
2008) 172-180.
25] J.L. Gong, H.R. Yue, Y.J. Zhao, S. Zhao, L. Zhao, J. Lv, S.P. Wang, X.B. Ma, J. Am. Chem. Soc.
34 (2012) 13922-13925.
[
[
(
1
[
[
[
[
26] P. Claus, M. Lucas, B. Lucke, Appl. Catal. A: Gen. 79 (1991) 1-18.
27] H. Wilmer, T. Genger, O. Hinrichsen, J. Catal. 215 (2003) 188-198.
28] J.W. Evans, M.S. Wainwright, N.W. Cant, D.L. Trimm, J. Catal. 88 (1984) 203-213.
29] S.H. Zhu, X.Q. Gao, Y.L. Zhu, W.B. Fan, J.G. Wang, Y.W. Li, Catal. Sci. Technol. 5 (2015)
1169-1180.
[
[
30] E.K. Poels, D.S. Brands, Appl. Catal. A: Gen. 191 (2000) 83-96.
31] A.J. Marchi, J.L.G. Fierro, J. Santamaría, A. Mozón, Appl. Catal. A: Gen. 142 (1996) 375-386.
17

3
2
2
1
1
0
0
.0
.5
.0
.5
.0
.5
.0
0
.4
.3
a
26.2 nm
b
1
000
0
8
6
4
2
00
00
00
00
0
0.2
0
.1
.0
0
2
4
6
8
10
CuO/SiO2
SiO2
+
250
CuO/SiO2
SiO2
0
.0
0.2
0.4
0.6
0.8
1.0
1
10
100
P/P0
Pore diameter (nm)
Figure 1. N
2
adsorption-desorption isotherm and pore size distribution of the bare SiO
2
and calcined
CuO/SiO
2
sample. (a) N adsorption-desorption isotherm (b) mesopore size distribution by BJH
2
method.
18

▲
CuO
Cu2O
Cu
★
★
▲
d
c
b
a
3
0
35
40
45
50
55
2-theta(degree)
Figure 2. XRD patterns of (a) the as-calcined CuO/SiO
catalyst at 573 K, (c) the Cu/SiO sample obtained by passivation of (b), and (d) the spent Cu/SiO
catalyst sample obtained by passivation after stability test for MA hydrogenation.
2
sample, (b) the hydrogen-reduced Cu/SiO
2
2
2
19

0
.5 nm
Cu
Cu-SiO2 SiO2
interface
2
Figure 3. TEM images of the as-calcined CuO/SiO sample and the schematic illustration of the
core-shell structure of the rod-like section.
20

c
1 nm
d
Cu
6
nm
Cu-SiO2
interface
SiO2
Figure 4. TEM/STEM images of the hydrogen-reduced Cu/SiO
2
catalyst and the schematic illustration
of the core-shell structure of the highly dispersed copper particles.
21

5
21 K
a
fresh
spent
5
01 K
c
b
d
3
50
400
450
500
550
573
Temp.(K)
Figure 5. H
first reduction profile of the as-calcined fresh CuO/SiO
as-calcined fresh Cu/SiO sample after N O surface oxidation, (c) the first reduction profile of spent
2
-TPR profiles of the as-calcined CuO/SiO
2
sample and the spent Cu/SiO
2
catalyst. (a) the
2
sample, (b) the second reduction profile of the
2
2
sample, (d) the second reduction profile of the spent sample after N
2
O surface oxidation.
0
Cu LMM
Cu
Cu 2p
+
9
32.9
Cu
568.2
9
52.9
p1/2
2
p3/2
572.2
2
31.6%
E
E
D
D
38.9%
C satellite
satellite
B
28.1%
B
9
34.7
satellite
A
9
54.7
satellite
940
5
75
570
Binding Energy/eV
565
9
70
960
950
930
Binding Energy/eV
Figure 6. Cu 2p XPS and Cu LMM XAES spectra of (A) the as-calcined CuO/SiO
hydrogen-reduced Cu/SiO catalyst, and (C) the spent Cu/SiO
samples after hydrogen-reduction at 483 K (D) and 573 K (E).
2
sample, (B) the
2
2
catalyst after passivation and its derived
22

60
50
40
30
20
10
EtOH
CH3OH
Et
h
yl
Acetate
0.3
0.2
0.1
CH3CHO
Alkane
40
0
20
60
80
100
MA conversion (%)
Figure 7. Effect of MA conversion on the product selectivity during MA hydrogenation over the
Cu/SiO catalyst. Reaction conditions: 498 K, 1.0 MPa, MA/H =2.5/97.5 vol%.
2
2
23

1
0
8
6
4
2
0
60
A
CH3OH
EtOH
4
0
0
2
Ethyl acetate
Acetyldehyde
1
0
Alkane
400
600
800
1000
H2 pressure/kPa
1
1
1
5.0
2.5
0.0
6
0
Et
B
O
H
0
.92-1.0 MPa
CH3OH
40
2
0
te
ta
ce
a
yl
h
Et
Al
ka
n
e
1
0
7
.5
Acetyldehyde
5
.0
0.02 0.04
0.06
0.08
MA/H2
Figure 8. Effects of H
2
(A) and MA (B) partial pressures on the reaction rate and the product selectivity
during MA hydrogenation on the Cu/SiO
2
catalyst. Reaction conditions：A: 513 K，PMA = 35 kPa,
GHSV = 6000 mL/(gcat.h); B: 523 K, PH2 = 920-1000 kPa, GHSV = 6000 mL/(gcat.h).
24

100
MA Conversion 95%
80
60
40
EtOH
MeOH
4
2
0
Ethyl Acetate
Acetyldehyde
Alkane
80
0
20
40
60
100
Time-on-stream (h)
Figure 9. Stability test of the Cu/SiO
2
catalyst for MA hydrogenation. Reaction conditions: 523 K, 1.0
MPa, MA/H
2
= 3/97 vol%, GHSV = 6000 mL/(gcat.h).
H2O
CO
2
3
00
400
500
600
T(K)
700
800
900
73
8
2
Figure 10. Temperature-programmed desorption profile of the spent Cu/SiO catalyst under He flow.
25

2
Figure 11. TEM/STEM images of the spent Cu/SiO catalyst.
26

5
28 K
Cu(0)
3
61 K
Cu(+)
4
00
500
600
700
800
Temp.(K)
Figure 12. MA-TPD profiles over the referencing Cu/SiO
2
catalysts containing different Cu species.
The desorption at 361 K was originated from traces of water in the dosing gas.
27

2
Scheme 1. Possible surface reaction routes for methyl acetate hydrogenation on the Cu/SiO catalyst.
2
Table 1. Effect of reaction temperature on the catalytic performance of the Cu/SiO catalyst
MA
Product Selectivity (%)
Acetaldehyde EtOAc MeOH
Temp. (K)
Conv.% Alkanes*
EtOH
30.8
4
5
5
93
18
28
26.5
81.1
96.2
0.2 (CH
4
)
---
32.7
12.2
2.5
36.4
34.2
33.6
0.4 (50% C
1.7 (75% C
2
2
H
H
6
6
)
)
0.2
0.5
52.9
61.7
Reaction conditions: MA/H
= 3.5/96.5 vol%, 6000 mL/(gcat.h), 1.0 MPa.
2
*Data in the brackets indicate the selectivity of C H in total alkanes (CH +C H ).
2 6 4 2 6
2
Table 2. MA conversion over the Cu/SiO catalyst
Catalyst Loading MA(l)
H
2
W/F
Contact
time MA
(
g)
(μl/min) (mL/min) (g.s/mL) (g.h/molMA) Conv.%
0
0
0
0
0
.3
2
4
6
4
6
17
34
51
34
51
1.06
0.53
0.35
0.26
0.18
274.6
127.1
81.9
86.6
49.8
27.8
16.7
11.4
.3
.3
.15
.15
68.3
43.2
Reaction conditions: 498 K, 1.0 MPa, MA concentration: 2.5% (by GC analysis using a 2% MA/He
calibration gas). For a specified loading of catalyst (0.15 or 0.3 g), MA conversion linearly decreased
as the feeding rate increased, making the amount of MA converted nearly constant. This suggests that
effect of external diffusion is excluded.
28

2
Table 3. Results of control experiments on the Cu/SiO catalyst
Temp.
K）
493
18
493
18
Conv.
(%)
Sel. (%)
Alkane
Reaction
MA+H a
（
Acetaldehyde EtOAc Methanol Ethanol
2
26.5
81.1
77.3
91.8
0.6
0.19
0.40
0.22
0.93
0.07
0.24
0.38
0.85
65
32.6
12.2
0
36.4
34.2
0
30.8
52.9
99.4
98.2
5
b
EA+H
2
5
0
0
EtOH+H c
498
498
35(C
0.5(C
2
2 6
H )
d
EtOH+N
2
24
2
H
4
)
99.5
a
b
c
d
=3.2/96.8 vol%, GHSV = 6000 mL/(gcat.h), 1.0 MPa;
=3.0/97.0 vol%, GHSV = 6600 mL/(gcat.h), 1.0 MPa;
MA/H
EA/H
EtOH/H
EtOH/N
2
2
=2.5/97.5 vol%, GHSV = 6600 mL/(gcat.h), 1.0 MPa;
=2.5/97.5 vol%, GHSV = 6000 mL/(gcat.h), 0.5 MPa.
2
2
29