Lanthanum oxide-modified Cu/SiO2 as a high-performance catalyst for chemoselective hydrogenation of dimethyl oxalate to ethylene glycol

Article abstract of DOI:10.1021/cs400574v

Several lanthanum oxide-modified Cu/SiO₂ (La-Cu/SiO₂) catalysts synthesized by urea-assisted gelation and postimpregnation were used for the vapor-phase chemoselective hydrogenation of dimethyl oxalate (DMO) into ethylene glycol (EG). The 1.0La-Cu/SiO₂-u catalyst with 1.0 wt % La loading was found to have the highest activity, whereas the catalyst with higher La loading at 3.0 wt % adversely affected the deterioration of catalytic activity. H₂-TPR results revealed that strong interactions between La promoters and Cu species substantially changed catalyst reducibility and made some Cu²⁺ species on catalyst precursors difficult to be reduced. Several positive variations induced by the introduction of La were confirmed by in situ XRD, N₂O chemisorption, H₂-TPD, X-ray Auger electron spectroscopy, and in situ FT-IR of chemisorbed CO. These variations included increased Cu metallic dispersion, improved ability for H₂ activation, elevated surface concentration of Cu⁺ species, and enhanced stability of catalyst nanostructure. The formation of unique Cu-O-La bonds between LaO_x and cupreous species located at interfacial sites was presumably responsible for the improved catalytic performance and stability of La-Cu/SiO₂-u catalyst.

Full text of DOI:10.1021/cs400574v

Research Article
pubs.acs.org/acscatalysis
Lanthanum Oxide-Modified Cu/SiO as a High-Performance Catalyst
2
for Chemoselective Hydrogenation of Dimethyl Oxalate to Ethylene
Glycol
Xinlei Zheng, Haiqiang Lin,* Jianwei Zheng, Xinping Duan, and Youzhu Yuan*
State Key Laboratory of Physical Chemistry of Solid Surfaces and National Engineering Laboratory for Green Chemical Production of
Alcohols-Ethers-Esters, College of Chemistry and Chemical Engineering, Xiamen University, Xiamen 361005, China
*
S Supporting Information
ABSTRACT: Several lanthanum oxide-modified Cu/SiO (La-
2
Cu/SiO ) catalysts synthesized by urea-assisted gelation and
2
postimpregnation were used for the vapor-phase chemo-
selective hydrogenation of dimethyl oxalate (DMO) into
ethylene glycol (EG). The 1.0La-Cu/SiO -u catalyst with 1.0
2
wt % La loading was found to have the highest activity, whereas
the catalyst with higher La loading at 3.0 wt % adversely
affected the deterioration of catalytic activity. H -TPR results
2
revealed that strong interactions between La promoters and Cu
species substantially changed catalyst reducibility and made
2+
some Cu species on catalyst precursors difficult to be reduced.
Several positive variations induced by the introduction of La were confirmed by in situ XRD, N O chemisorption, H -TPD, X-ray
2
2
Auger electron spectroscopy, and in situ FT-IR of chemisorbed CO. These variations included increased Cu metallic dispersion,
+
improved ability for H activation, elevated surface concentration of Cu species, and enhanced stability of catalyst nanostructure.
2
The formation of unique Cu−O−La bonds between LaO and cupreous species located at interfacial sites was presumably
x
responsible for the improved catalytic performance and stability of La-Cu/SiO -u catalyst.
2
KEYWORDS: copper, lanthanum oxide, hydrogenation, dimethyl oxalate, ethylene glycol
1
. INTRODUCTION
generally believed to be pivotal in determining catalytic activity
and stability. Various synthesis methods including ammonia
Ethylene glycol (EG) is an important chemical feedstock used
in many industrial processes and has versatile commercial
applications. Commercial EG is mainly produced from
petroleum-derived ethylene through the hydration of ethylene
oxide. Considering the depletion of petroleum resources, an
efficient production of EG from syngas through a two-step
process consisting of the coupling of CO with nitrite esters to
dimethyl oxalate (DMO) and heterogeneous catalytic hydro-
2
22−25
13
evaporation, urea-assisted precipitation,
impregnation,
1
5
1
and ion-exchange have been successfully developed to
produce highly dispersed Cu nanoparticles (NPs) on silica
support with high Cu loading. A synergistic effect exists
2
0
+
between Cu and Cu species, suggesting the dissociative
0
+
activation of H molecules on Cu sites as Cu species polarize
2
and activate ester groups in DMO molecules. This phenom-
enon emphasizes the importance of generating and retaining
1
−5
genation of DMO to EG is attracting considerable interest.
0
+
the appropriate surface distribution of Cu and Cu species
during the activation of catalyst precursor and the catalytic
However, several problems remain unsolved despite consid-
erable efforts in the study of the DMO hydrogenation process.
To date, EG selectivity and catalytic stability still need
improvement, the interaction with active sites remains
controversial, and the process of catalyst deactivation requires
further elucidation. A deeper understanding of the generic
concepts is necessary for more rational catalyst design.
2
,19,22,23,25−29
operation of DMO hydrogenation.
To improve
the poor lifespan of Cu catalyst on silica support mainly
ascribed to the agglomeration of Cu NPs and deterioration of
metal dispersion during long-term catalytic operation, enhanc-
ing the interaction between Cu and its silica support or
introducing a promoter to stabilize Cu NPs is considered
beneficial. As a weakly acidic carrier, silica poorly interacts with
metallic Cu, thereby facilitating the sintering of Cu NPs and the
loss of the active cupreous surface as evidenced by the color
Inexpensive Cu-based heterogeneous catalysts are highly
6
−8
active for the chemoselective hydrogenation of DMO to EG.
Significant advancements in fabricating efficient and stable Cu-
based catalysts, as well as the preliminary understanding of the
catalytic essence of DMO hydrogenation, have been achieved in
the past decade.
catalyst supported by silica carriers, Cu dispersion and
interaction between Cu species and support or promoter are
2
,9−21
Regarding the extensively studied Cu
Received: July 18, 2013
Revised: October 5, 2013
Published: October 14, 2013
©
2013 American Chemical Society
2738
dx.doi.org/10.1021/cs400574v | ACS Catal. 2013, 3, 2738−2749

ACS Catalysis
Research Article
change from “gray or black” as a fresh catalyst to “copper red”
light blue precipitate was separated by high-speed centrifuga-
tion, washed three times with deionized water, dried in a
vacuum overnight at 393 K, and calcined at 623 K for 2 h in air.
The yielded azure powder of catalyst precursor was denoted as
7
,22
as a deactivated catalyst.
From a catalyst precursor with a
cupric phyllosilicate phase in which copper atoms are entrapped
into a silica network at the atomic level, enhanced metal−
support interaction and improved Cu metallic dispersion can be
achieved by appropriate thermal reduction activation.
xLa-Cu/SiO -u (x = 0.1, 0.5, 1.0, and 3.0), where x is the preset
2
weight percentage of La and u represents a synthesis method,
Previous studies on the mechanism of catalytic cycles and
catalyst deactivation in Cu-catalyzed DMO hydrogenation
indicate that one of the most convenient and feasible means
i.e., urea-assisted gelation. Cu/SiO -u catalyst was prepared
2
following the same synthesis procedure for xLa-Cu/SiO -u
2
except that no La(NO ) ·6H O additive was added. Cu/La O -
3
3
2
2
3
to solve the problems of existing Cu/SiO catalysts in terms of
2
u catalyst with La O as a support was also synthesized by urea-
2 3
efficiency, stability, and physical property can be achieved by
finding appropriate catalyst promoters. Promoters are required
to interact appropriately with the Cu component and silica
support to generate a fitting and stable catalytic setup. In the
assisted gelation, replacing Ludox AS-40 colloidal silica with an
appropriate concentration of aqueous La(NO ) ·6H O.
3
3
2
For comparison, two La O -mixed Cu/SiO -u catalysts were
2
3
2
also prepared: one by a synthesis method of postimpregnation,
and the other by mechanical mixing. In a typical procedure, a
certain quantity of aqueous La(NO ) ·6H O solution was
case of boric oxide-doped Cu/SiO catalyst, a remarkably
2
prolonged lifespan for DMO hydrogenation into EG or ethanol
has been observed, and the strong interaction between boric
oxide and surface cupreous species is assumed to be responsible
3
3
2
added dropwise onto a 100 mL beaker containing Cu/SiO -u
2
precursor. The obtained slurry was stirred for 12 h at room
temperature (RT), dried overnight at 393 K, and calcined for 2
h at 623 K in air, yielding the catalyst precursor denoted as
2
2,30
for the stabilizing effect.
In metallurgy, the rare-earth
lanthanum oxide (LaO ) is sometimes used as an additive to
x
improve the resistance of metal or alloy against sintering,
1
.0La-Cu/SiO -i for the postimpregnation method. Another
31
2
oxidation, corrosion, and wearing. La O is also often utilized
2
3
catalyst denoted as 1.0La-Cu/SiO -m for mechanical mixing
2
as a promoter for heterogeneous metal catalysts in various
32
was prepared by mechanically blending La(NO
granules and Cu/SiO -u precursor using a planetary ball mill
with agate container and spheres.
.2. Catalyst Characterization. X-ray diffraction (XRD)
3 3 2
) ·6H O
catalytic reactions, such as water−gas shift, syngas methana-
3
3
34
35
2
tion, lean-burn NO removal, ammonia decomposition,
x
36
and methane decomposition. The promotional effects of
2
La O reportedly originate from several aspects, such as the
2
3
patterns of catalysts were collected on a PANalytical X’pert Pro
Super X-ray diffractometer using Cu Kα radiation (λ = 0.15418
nm) with a scanning angle (2θ) range of 10° to 90°. The tube
voltage and current were 40 kV and 30 mA, respectively. For in
situ XRD measurement, the catalyst precursor was placed in a
stainless steel holder and covered with a 0.1 mm-thick
textural change of catalyst to form special active sites or to
33,34
increase metal dispersion,
electronic transfer to decrease
the electron density of active metal and stabilize active sites in
32
their appropriate valence state by electron trapping force, and
synergistic effect between La-involved sites, as well as other
3
5,36
active sites.
However, the exact nature of the promotional
beryllium plate. A 5% H −95% Ar mixture was introduced at
2
effect of La remains unclear. The introduction of LaO into a
x
−1
a flow rate of 50 mL min , and the temperature was ramped
Cu/SiO catalyst may elicit promotional effects on Cu metallic
2
from RT to 473, 523, 573, 623, 673, 723, 773, 873, 973, 1073,
dispersion to form special surface cupreous species with a
thermally stable catalyst structure, among others. Although
La O -supported Cu catalysts are occasionally used for
−1
and 1173 K at a rate of 2 K min . The Cu crystallite size was
calculated with the Scherrer equation using the full width at
half-maximum (fwhm) of the Cu(111) diffraction peak at 2θ =
2
3
hydrogenation, a detailed study on LaO -modified Cu/SiO2
x
4
3.2°.
catalyst systems for DMO hydrogenation has not yet been
reported.
N adsorption−desorption isotherms were measured by the
2
static volumetric method at 77 K using a Micromeritics TriStar
In this study, a series of LaO -modified Cu/SiO catalysts
x
2
II 3020 surface area and pore analyzer. Before N physisorption
was designed and fabricated following urea-assisted gelation or
impregnation, and their catalytic performances were compre-
hensively evaluated in the vapor-phase hydrogenation of DMO
into EG. Significant improvements in catalytic activity and
stability can be achieved by properly controlling the atomic
ratio of La/Cu and conducting appropriate activation of the
catalyst precursor. To gain further insight on the promotional
effect of LaO on Cu/SiO , a series of catalyst characterizations
2
measurement, all samples were degassed at 393 K for 1 h and
evacuated at 573 K for 3 h to remove physically adsorbed
impurities. The specific surface area (SBET) was calculated by
the Brunauer−Emmett−Teller (BET) method, adopting
isotherm data within the relative pressure (P/P ) range of
0
0.05−0.2. The total pore volume (V ) was derived from the
p
adsorbed N volume at a relative pressure of approximately
x
2
2
was performed.
0.995, and mesopore size distributions were derived using the
desorption branch of isotherm using the Barrett−Joyner−
Halenda method.
2. EXPERIMENTAL SECTION
The H2 temperature-programmed reduction (H -TPR)
2
.1. Catalyst Preparation. All catalysts contained a
2
profiles of the catalysts were measured on a Micromeritics
Autochem II 2920 instrument connected with a Hiden Qic-20
mass spectrometry (MS) system. In a typical procedure, 100
constant, preset Cu loading of 15 wt %. LaO -modified Cu/
x
SiO -u catalysts with various La loadings were prepared by
2
22−25
urea-assisted gelation as previously described.
Cu(NO ) ·
3 2
3
H O weighing 4.6 g and a certain amount of La(NO ) ·6H O
mg of catalyst was pretreated at a flow of 20% O
2
−80% Ar (50
mL min ) in a quartz U-tube reactor at 523 K for 1 h. After
cooling down to RT in high-purity Ar, a flow of 5% H
2
3 3
2
−1
were dissolved in 100 mL of ammonia aqueous solution in a
round-bottom flask. Subsequently, 3.0 g of urea was added
before slowly dropping 12.0 g of 40 wt % Ludox AS-40 colloidal
silica under mechanical stirring. The mixture was vigorously
stirred and aged at 343 K in an oil bath for 4 h. The obtained
−95% Ar
2
−1
(50 mL min ) was passed through the catalyst bed. A
temperature ramping program from RT to 1073 K at a rate of 5
K min− was used, and H consumption was monitored with
1
2
2
739
dx.doi.org/10.1021/cs400574v | ACS Catal. 2013, 3, 2738−2749

ACS Catalysis
Research Article
MS signals at m/e = 2 in the multiple ion detection (MID)
mode.
completely remove chemisorbed surface species. After cooling
down to RT, the catalyst wafer was exposed to high-purity CO
(760 Torr) for 30 min. IR spectra were collected at different
evacuation times and referenced to the background spectrum of
623 K reduced catalyst, which was collected before CO-soaking
in a high vacuum at the same temperature.
Characterizations of X-ray photoelectron spectroscopy
(XPS) and Auger electron spectroscopy (XAES) were
conducted on a Quantum 2000 Scanning ESCA Microprobe
instrument (Physical Electronics) equipped with an Al Kα X-
ray radiation source (hν = 1486.6 eV). The activated catalyst
was carefully collected and sealed under the protection of Ar
H temperature-programmed desorption (H -TPD) profiles
2
2
were collected on the same equipment for H -TPR measure-
2
ment. In a typical procedure, 100 mg of catalyst in a quartz U
tube was first reduced in 5% H −95% Ar at 623 K for 4 h. After
2
the activated catalyst had cooled down to RT, a flow of high-
−
1
purity H (50 mL min ) was introduced for 1 h, followed by a
2
−1
purge of high-purity Ar (50 mL min ) to completely remove
unabsorbed H so that the MS signal baseline reached the
2
stable state. Finally, the H -TPD profile was collected from RT
2
−1
to 1073 K at a ramping rate of 5 K min in a flow of high-
purity Ar (50 mL min ). Desorbed H was monitored with
−1
atmosphere after reduction in a flow of 5% H −95% Ar at 623
2
2
MS signals at m/e = 2 in MID mode.
K for 4 h. The collected sample was compressed into a thin disk
in a glovebox and transferred to the XPS apparatus analysis
chamber. Binding energies were calibrated using the Si 2p peak
at 103.7 eV as a reference. Experimental errors were within
±0.2 eV.
Transmission electron microscopy (TEM) images were
obtained on a TECNAI F-30 transmission electron microscope
operating at an acceleration voltage of 300 kV. Elemental
distribution in catalyst was determined by energy-dispersive X-
ray spectroscopy (EDX) mapping technique in scanning TEM
2.3. Measurement of Catalytic Activity and Durability.
Catalytic evaluation of vapor-phase DMO hydrogenation was
conducted in a continuous flow mode using a stainless-steel
fixed-bed tubular reactor equipped with a computer-controlled
autosampling system. In a typical run, 100 mg of catalyst
precursor (40 to 60 meshes) was placed in the center of the
reactor (7 mm internal diameter), and the top side of the
catalyst bed was packed with adequate quartz powder (40 to 60
(STEM) mode. To prepare an appropriate sample for TEM
observation, catalyst powder was ultrasonically dispersed in
ethanol at RT for 30 min and transferred onto carbon-coated
copper or molybdenum grid by dipping.
The actual metal contents of catalyst were analyzed by
inductively coupled plasma−optical emission spectrometry
(
ICP-OES) on a Thermo Electron IRIS Intrepid II XSP. The
copper surface areas and dispersions of the catalysts were
determined on a Micromeritics Autochem II 2920 apparatus
meshes). The catalyst was activated in a flow of 5% H
2
−95% Ar
−1
atmosphere (50 mL min ) at 623 K for 4 h with a ramping
−1
with a thermal conductivity detector (TCD) by N O
rate of 2 K min . The reactor was cooled to the target reaction
was fed into the reactor
2
chemisorption and H pulse reduction methods based on the
temperature (443−473 K), and H
2
2
stoichiometry of [2Cu(s) + N O → Cu O(s) + N ], where
using a mass flow controller. The system pressure was precisely
2
2
2
37
Cu(s) represents surface copper atom. Initially, 100 mg of
controlled at 3 MPa with a back-pressure regulator. DMO
−1
methanol solution (0.02 g mL ) was pumped into the reactor
using a Series III digital HPLC pump (Scientific Systems, Inc.).
The outlet stream was sampled using an automatic Valco 6-port
valve system and analyzed using an online gas chromatograph
catalyst precursor was reduced in 5% H −95% Ar (50 mL
2
−1
min ) at 623 K for 4 h and cooled down to 333 K. Pure N O
20 mL min ) was passed through the catalyst bed for 1 h to
2
−1
(
completely oxidize surface copper atoms into Cu O. H pulse
2
2
(
Agilent 7890A) equipped with a flame ionization detector and
reduction was conducted in a flow of high-purity Ar (50 mL
−1
RTX-Wax capillary column (30 m × 0.25 mm ×0.25 μm) at an
interval of 1 h.
min ) at a temperature of 573 K to guarantee that the
chemisorbed oxygen can immediately and thoroughly react
with the high-purity H2 supplied from a 0.4 mL loop.
3
. RESULTS AND DISCUSSION
Calibrated TCD was used to quantify the reserved H in the
2
outlet gas in which moisture was removed with a 13 X zeolite
dehydration trap. Pulse dosing was repeated at an interval of 3
min until the detected TCD peaks for the latest two or more
pulses were constant in the area. The volume of the consumed
3.1. Catalyst Crystalline Phase and Morphology. The
XRD patterns of Cu/SiO -u, Cu/La O -u, and xLa-Cu/SiO -u
2
2
3
2
catalyst precursors thermally treated in air at 623 K but not
subjected to reductive activation with H are shown in Figure 1.
2
H was calculated by subtracting the sum of the residual
No diffraction peak except quite faint ones ascribable to cupric
2
volumes of all unsaturated H pulses from the total H injection
phyllosilicate phases is observed on Cu/SiO -u and La-Cu/
2
2
2
volume. Copper dispersion was calculated by dividing the
amount of surface copper atom by the total number of
supported copper atoms per gram of catalyst. The metallic
copper surface area was calculated based on an atomic copper
SiO -u catalysts with <1 wt % La loading. Meanwhile, La
2
loading in La-Cu/SiO -u catalyst increases to a high level at 3.0
2
wt %, a series of XRD peaks at 2θ = 17.8°, 24.5°, 30.5°, 43.0°,
43.9°, and 46.9°, which is attributed to lanthanum carbonate
1
9
−2
38
surface density of 1.46 × 10 Cu atoms m .
hydroxide (JCPDS 00-049-0981), emerges and indicates that
The Fourier-transform infrared (FT-IR) spectra of catalyst
precursors and chemisorbed CO species on activated catalysts
were collected on a Nicolet 6700 spectrometer at a special
urea-assisted gelation enables most species from the La(NO3)3
precursor to precipitate and mix with Cu/SiO2 catalyst
precursors mainly in the form of lanthanum carbonate
hydroxide. For cupric phyllosilicate with low crystallinity at
2θ = 31.0°, 34.8°, 57.2°, 63.3°, and 71.2° (JCPDS 00-003-
0219), a series of weak diffraction peaks is present in the XRD
−1
resolution of 4 cm . For CO-adsorption experiments, a
stainless steel cell with CaF windows and a high-vacuum turbo
2
pump system that allowed in situ thermal treatment up to 873
K under different atmospheres or high vacuum were used. In a
typical procedure, 30 mg of catalyst was compressed into a self-
supporting wafer and carefully loaded into the in situ cell. The
patterns of xLa-Cu/SiO -u catalyst precursors, indicating that
2
the introduction of La(NO ) into catalyst synthesis systems
3
3
does not significantly disturb the formation of cupric
catalyst wafer was reduced for 4 h in a flow of 5% H −95% Ar
phyllosilicate, which is believed to be the origin of the high
2
−1
(
50 mL min ) at 623 K and evacuated for 30 min to
Cu dispersion of Cu/SiO -u catalyst in urea-assisted gelation.
2
2
740
dx.doi.org/10.1021/cs400574v | ACS Catal. 2013, 3, 2738−2749

ACS Catalysis
Research Article
particle sizes from a synthesis system containing urea,
La(NO ) , and Cu(NO ) . Generally, for urea-assisted gelation
3
3
3 2
synthesis in an alkaline environment, silicate anions arising
from silica dissolution can react with cuprammonia complexes
39
to form insoluble copper phyllosilicate whiskers. However,
under the same environment, no compound like La CuO or
2
4
CuLa O can be formed from a reaction between copper and
2
4+δ
lanthanum species. Instead, lanthanum carbonate hydroxide
and basic cupric carbonate are preferentially precipitated to
form Cu/La O -u catalyst precursor.
2
3
For the 1.0La-Cu/SiO -i catalyst precursor produced by
2
postimpregnation method, the characteristic diffraction peaks of
cupreous phyllosilicate seem to be analogous to that with the
Cu/SiO -u catalyst precursor (Supporting Information Figure
2
S1-c), indicating that the introduction of a small quantity of La
Figure 1. XRD patterns of 623 K-calcined catalyst precursors. (a) Cu/
SiO -u, (b) 0.1La-Cu/SiO -u, (c) 0.5La-Cu/SiO -u, (d) 1.0La-Cu/
onto the surface of Cu/SiO -u and the subsequent thermal
2
2
2
2
treatment at 623 K induces unobvious changes in the crystalline
SiO -u, (e) 3.0La-Cu/SiO -u, and (f) Cu/La O -u.
2
2
2
3
phase of Cu/SiO -u catalyst. The intensities of the XRD peaks
2
of the cupreous phyllosilicate of 1.0La-Cu/SiO -m produced by
2
FT-IR characterization, which is a useful technique for
distinguishing copper phyllosilicate phase, was also adopted.
The FT-IR spectra shown in Figure 2 confirm the presence of
mechanical mixing are weaker than those of Cu/SiO -u, 1.0La-
39
2
Cu/SiO -u, and 1.0La-Cu/SiO -i catalyst precursors (Support-
2
2
ing Information Figure S1-d) mainly because of the crystallinity
deterioration of cupreous phyllosilicate caused by long-time
high-energy milling.
In general, metallic Cu NPs highly dispersed in catalyst
2
+
0
support tend to be oxidized to cupric oxide (E
= 0.3419
Cu /Cu
+
0
V) or cuprous oxide (E
= 0.521 V) when exposed to air,
Cu /Cu
even at RT. Thus, in situ XRD was adopted to monitor the
phase evolution of Cu/SiO -u, 1.0La-Cu/SiO -u, 3.0La-Cu/
2
2
SiO -u, and Cu/La O -u catalysts upon reductive activation in
2
2
3
5
% H −95% Ar atmosphere at different temperatures. For Cu/
2
SiO -u and 1.0La-Cu/SiO -u catalysts, the characteristic
2
2
diffraction peaks for copper phyllosilicate disappear with
increased reduction temperature to 523 K (Figures 3A,B).
Figure 2. FT-IR spectra of as-calcined catalyst precursors (a) Cu/
SiO -u, (b) 1.0La-Cu/SiO -u, (c) 1.0La-Cu/SiO -i, and (d) 1.0La-Cu/
2
2
2
SiO -m. The I670/I800 denotes the intensity ratio of two IR bands
2
−1
located at 670 and 800 cm .
copper phyllosilicate in xLa-Cu/SiO -u catalyst precursors as
2
−1
evidenced by the δ band at 670 cm and the νSiO shoulder
OH
−1
peak at 1042 cm . Because the asymmetric ν band of silica
SiO
−1
at 1118 cm and the νSiO band of copper phyllosilicate at 1042
cm overlap with each other, the relative contents of copper
−1
phyllosilicate in different catalysts are compared using the ratio
of integrated intensity of δ band of copper phyllosilicate at
OH
6
70 cm− and symmetric ν
1
band of silica at 800 cm
−1
SiO
39
denoted as I670/I800
. The value of I670/I800 for 1.0La-Cu/SiO2-
u precursor is 0.153, which is clearly higher than that for Cu/
SiO -u (0.099), 1.0La-Cu/SiO -i (0.097), and 1.0La-Cu/SiO -
m (0.03). This finding indicates that introducing La species into
Figure 3. In situ XRD patterns of as-calcined catalysts as a function of
2
2
2
the reduction temperature under a 5% H −95% Ar atmosphere. (A)
2
Cu/SiO -u and (B) 1.0La-Cu/SiO -u.
2
2
Cu/SiO -u by urea-assisted gelation substantially improves the
2
formation of cupric phyllosilicate.
The absence of evident XRD peaks for CuO or other
cupreous substances further confirms that Cu atoms are finely
dispersed into silica supports by urea-assisted gelation. By
contrast, Cu/La O -u catalyst affords two strong diffraction
Moreover, diffraction peaks are observed at 2θ = 43.1°, 50.2°,
0
and 73.9°, which are characteristic of Cu (JCPDS 04-0836).
Thus, most Cu2+ species entrapped in silica support can be
2
+
reduced by H at ≤523 K. However, the Cu species on the
2
3
2
peaks at 2θ = 35.6° and 38.2°, which are characteristic of CuO
JCPDS 01-089-2531). This result signifies that cupreous
species are inclined to form copper oxides with relatively large
3.0La-Cu/SiO -u and Cu/La O -u catalysts are difficultly
reduced by H at 523 K. The characteristic Cu diffractions
2
2
3
0
(
2
of these two catalysts emerge at a much higher reduction
2
741
dx.doi.org/10.1021/cs400574v | ACS Catal. 2013, 3, 2738−2749

ACS Catalysis
Research Article
Figure 4. TEM images of 1.0La-Cu/SiO -u: (A) as-calcined catalyst precursor; (B) as-reduced catalyst; (C) particle size distribution; (D) HRTEM
2
of as-reduced catalyst; (E) HAADF-STEM image of as-reduced catalyst. Elemental EDX maps of as-reduced catalyst corresponded to the orange
square region in E. Four maps represent the measured Si K intensity, Cu K intensity, La M intensity, and overlap intensity, respectively.
temperature of 573 and 723 K, respectively (Supporting
Information Figure S2). This result strongly indicates that a
strong interaction exists between Cu and La species, and this
temperature of 1073 K, indicating the high thermal stability of
lanthanum carbonate hydroxide. Meanwhile, highly dispersed
lanthanum carbonate hydroxide on 1.0La-Cu/SiO -u and
2
2
+
interaction tends to hinder the reduction of Cu species to
3.0La-Cu/SiO -u catalysts are partly decomposed or reduced
2
0
Cu . With increased reduction temperature from 523 to 873 K,
to facilitate the high dispersion of lanthanum oxide NPs at a
relatively low temperature of 623 K, as evidenced by the
weakening of characteristic diffraction peaks. In summary, La
0
the diffraction peaks of Cu nanocrystallite on 1.0La-Cu/SiO -u
2
catalyst does not evidently change, reflecting the relatively high
thermal stability of the catalyst. Furthermore, the particle sizes
of copper NPs on Cu/SiO -u, 1.0La-Cu/SiO -u, and 3.0La-Cu/
species appropriately introduced into Cu/SiO -u can strongly
2
interact with Cu species on the catalyst precursor to hinder the
2
2
SiO -u catalysts were calculated using the Scherrer equation
reduction of Cu2+ species to some extent, as well as to retard
2
0
based on the fwhm of the Cu(111) diffraction peak at 2θ =
surface migration and accumulation of nascent Cu NPs into
43.1°, as shown in Figures 3 and Supporting Information
large particles.
Figure S2. Interestingly, the size of copper NPs follows the
The TEM images of the Cu/SiO -u, 1.0La-Cu/SiO -u, and
2
2
order Cu/SiO -u > 1.0La-Cu/SiO -u > 3.0La-Cu/SiO -u at
3.0La-Cu/SiO -u catalyst precursors (Figures 4A and Support-
2
2
2
2
almost all treatment temperatures, indicating that the
introduction of La species is beneficial for stabilizing small-
sized Cu NPs during reductive activation. The diffraction peaks
of lanthanum carbonate hydroxide at 2θ = 17.8°, 24.5°, and
ing Information Figure S3) clearly reveal that numerous
whisker-shaped copper phyllosilicate particles are present
before reductive activation. After reduction at 623 K, only
traces of the whisker-shaped copper phyllosilicates remain, and
3
0.5° observed on the catalyst precursor Cu/La O -u do not
a large number of cross-linked SiO spheres, whose surface is
2
3
2
disappear until the temperature increases to 723 K, and the
complete transformation to La O phase is achieved at a high
densely covered with black metallic Cu NPs <5 nm in size, are
formed. The big rod-shaped particles observed in the TEM
2
3
2
742
dx.doi.org/10.1021/cs400574v | ACS Catal. 2013, 3, 2738−2749

ACS Catalysis
Research Article
Table 1. Physicochemical Characterization of Several Supported La-Cu/SiO Catalysts
2
a
a
b
2
−1
c
3
−1
d
e
f
2
−1
catalyst
Cu loading, wt %
La loading, wt %
SBET
,
m g
V , cm g
D , nm
DCu, %
SACu, m g
p
p
SiO -u
−
−
−
0.84
108.2
106.2
321.6
347.9
342.5
338.2
268.8
334.2
22.4
0.34
0.59
0.70
0.84
0.73
0.86
0.52
0.79
0.10
0.70
0.31
9.9
21.1
7.5
−
−
−
−
2
1.0La/SiO -u
2
Cu/SiO -u
15.8
15.6
15.3
15.1
15.1
15.8
17.9
15.2
0
28.2
30.9
32.6
34.6
32.1
29.4
28.9
31.3
32.4
33.9
31.5
30.1
2
0
0
1
1
3
.1La-Cu/SiO -u
0.07
0.38
0.77
0.77
2.50
70.0
0.92
7.8
2
.5La-Cu/SiO -u
7.1
2
.0La-Cu/SiO -u
8.4
2
g
.0La-Cu/SiO -u
7.4
2
.0La-Cu/SiO -u
7.6
2
h
h
Cu/La O -u
20.7
7.8
6.2
7.3
2
3
1
.0La-Cu/SiO2-i
.0La-Cu/SiO -m
307.0
305.6
26.4
26.0
27.5
1
15.5
0.86
5.0
27.3
2
a
b
c
d
e
Determined by ICP-OES analysis. BET specific surface area. Pore volume that obtained from P/P = 0.99. Pore size. Dispersion of metallic Cu
0
f
g
determined by N O chemisorption and H pulse reduction. SA : Cu metallic surface area per gram of catalyst. The catalyst was calcined and
reduced at 723 K. Determined by TEM images.
2
2
Cu
h
image of the Cu/La O -u catalyst precursor are assumed to be
the La contents of these catalysts are slightly lower than the
2
3
38
lanthanum carbonate hydroxide crystals. Moreover, a number
of large, black metallic Cu particles are found after high-
temperature reduction at 823 K (Supporting Information
Figure S3-c). The mean particle sizes of these Cu particles are
approximately consistent with the in situ XRD results for the
Cu/SiO -u and La-Cu/SiO -u catalysts. In addition, surface
preset values. The N physisorption measurement revealed that
2
the Cu/SiO -u and xLa-Cu/SiO -u catalysts prepared by urea-
2
2
assisted gelation all exhibit type IV isotherms (Supporting
4
3
Information Figure S5), which are usually given by
mesoporous adsorbents. The shape of hysteresis loop
associated with capillary condensation taking place in
mesopores is changed to some extent as a result of introducing
La species into synthetic system of urea-assisted gelation for
Cu/SiO -u catalysts. The BET surface areas of xLa-Cu/SiO -u
2
2
characterization by nanoscale elemental STEM-EDX mapping
was also used to determine the surface distributions of Cu and
La species on the La-Cu/SiO -u catalysts. The distribution
2
2
2
2
−1
2
−1
domains of the Cu and La species clearly overlap each other on
both nonactivated and reduced catalysts (Figures 4 and
Supporting Information Figure S4), demonstrating that via
urea-assisted gelation, the La and Cu species are highly
dispersed into the silica texture and are in close contact with
each other. Moreover, the appropriate reduction activation of
catalysts ranging from 334.2 m g to 347.9 m g are slightly
higher than that of the Cu/SiO -u catalyst (321.6 m g ).
2
−1
2
Moreover, the pore volume of unmodified Cu/SiO -u catalyst is
2
3
−1
0.70 cm g and the introduction of La species increases the
pore volumes of xLa-Cu/SiO -u catalysts to 0.73−0.86 cm g .
The different copper phyllosilicate contents in the xLa-Cu/
3
−1
2
La-Cu/SiO -u catalyst precursors can retain the high dispersion
SiO -u catalyst precursors, as determined by FTIR character-
2
2
of the La and Cu species. During reductive activation by H2,
some Cu species trapped in the copper phyllosilicate texture
undergo gradual reduction into metallic Cu along with electron
donation from the activated H-species and tend to aggregate
into metallic Cu NPs on the silica support surface. The nascent
ization (Figure 2), may be responsible for these changes. By
contrast, the introduction of La by impregnation and
mechanical mixing has a negative effect on the specific surface
area of catalyst. The BET surface areas of the 1.0La-Cu/SiO -i
2
2
−1
and 1.0La-Cu/SiO -m catalysts are 307.0 and 305.6 m g ,
2
metallic Cu species act as active sites for H activation to
respectively, which are all lower than that of the Cu/SiO -u
catalyst (321.6 m g ). A control experiment on 1.0La-Cu/
2
2
2
+
2
−1
promote the reduction of Cu further. Moreover, the partial
3+
reduction of La into LaO , with La in a valence lower than +3,
SiO -u catalyst using 723 K as calcination/reduction temper-
x
2
cannot be excluded when neighboring Cu active sites for H
ature shows decreases in values of SBET, Cu dispersion (DCu),
2
40
activation are abundant. The strong interaction resulting from
the close contact between La and Cu species in La−O−Cu
bonds can hinder the transformation of the positive-valence Cu
and Cu metal surface area (SA ) (Table 1).
Cu
High DCu and large SACu are generally believed to be pivotal
in achieving high catalytic activity and stability in chemo-
selective hydrogenation catalyzed by Cu-based heterogeneous
catalysts. The DCu and SACu results as determined by classic
N O chemisorption are shown in Table 1. La-Cu/SiO -u
3
2,41
species into metallic Cu to some extent.
Furthermore, as an
alkaline metallic oxide, La oxides react with silica to form a La−
O−Si silicate phase, which leads to surface modification of the
2
2
4
2
silica support. Thus, the La species introduced into the Cu/
catalysts with La loadings below 1 wt % all exhibit higher DCu
SiO -u catalyst through urea gelation can act as chemical
and SACu values than those of the Cu/SiO -u catalyst. However,
2
2
bridges between Cu NPs and the silica surface via La−O−Cu
and La−O−Si bonds to enhance metal−support interaction.
This enhancement can improve the stability of Cu NPs on the
catalyst surface during catalyst activation and catalytic run.
a La loading of 3 wt % results in decreased DCu and SACu.
Analogous to the interpretation for the increased BET surface
area induced by the introduction of La species via urea-assisted
gelation, the larger amount of copper phyllosilicate compounds
3
.2. Physicochemical Properties. The chemical compo-
formed in the xLa-Cu/SiO -u catalyst precursors at appropriate
La loadings may be the main cause of the improved Cu metal
2
sitions and textural properties of all catalysts are summarized in
Table 1. The copper loadings for all catalysts from urea-assisted
gelation are close to the preset value of 15 wt %, as determined
by ICP-OES. This result is mainly due to the precipitation of
most of the Cu species from the Cu(NO ) reagents into
dispersion. The DCu and SACu of 1.0La-Cu/SiO -i and 1.0La-
2
Cu/SiO -m are slightly lower than those of the Cu/SiO -u
2
2
catalyst, possibly indicating that the direct contact between
La(NO ) and copper phyllosilicates in the Cu/SiO -u catalyst
3
2
3 3
2
copper phyllosilicate during urea-assisted gelation. However,
precursor disrupts the mesoporous structure of the Cu/SiO -u
2
2
743
dx.doi.org/10.1021/cs400574v | ACS Catal. 2013, 3, 2738−2749

ACS Catalysis
Research Article
catalyst and results in further blockage of some copper surface
approximately 433 K. For the Cu/La O -u catalyst, the
2
3
sites. The low Cu dispersion of the Cu/La O -u catalyst is
observed reduction peak at 954 K may be ascribed to the
partial reduction of the La species, which comes in close
2
3
mainly ascribed to the low porosity and inadequate surface area
2
+
of the La O3 support, which do not favor the efficient
contact to the metallic Cu formed by the reduction of Cu at
low temperatures. The introduction of La species into the Cu/
2
dispersion of the Cu species, particularly for a Cu loading as
high as 15 wt %.
SiO -u catalyst precursor via the impregnation method also
2
3
.3. H -TPR and H -TPD. The effect of the La species on
slightly increases the reduction temperature (ca. 531 K) for the
2
2
the reducibility of the La-Cu/SiO catalyst precursors was
1.0La-Cu/SiO -i catalyst (Figure 5).
2
2
determined using H -TPR characterization. The H -TPR
H surface adsorption behaviors on the La-Cu/SiO catalysts
2
2
2
2
profiles (Figure 5) show only one H consumption peak over
were analyzed using the H -TPD technique. The H -TPD
2
2
2
profiles for the Cu/SiO -u and xLa-Cu/SiO -u catalysts, which
2
2
were activated by H at 623 K for 4 h, all show two desorption
2
peaks located in the 300 to 350 K and 700 to 900 K
temperature ranges, respectively (Figure 6). To determine the
Figure 5. H -TPR profiles of as-calcined catalyst precursors. (a) Cu/
2
SiO -u, (b) 0.1La-Cu/SiO -u, (c) 0.5La-Cu/SiO -u, (d) 1.0La-Cu/
2
2
2
SiO -u, (e) 3.0La-Cu/SiO -u, (f) Cu/La O -u, (g) 1.0La-Cu/SiO -i,
2
2
2
3
2
and (h) 1.0La-Cu/SiO -m.
2
Figure 6. H -TPD profiles of as-reduced catalysts. (a) SiO , (b) 1.0La/
2
2
SiO -u, (c) Cu/SiO -u, (d) 0.1La-Cu/SiO -u, (e) 0.5La-Cu/SiO -u, (f)
2
2
2
2
the Cu/SiO -u and xLa-Cu/SiO -u catalyst precursors. This
H -TPR peak corresponds to the reduction of the Cu species
2
2
2
2+
1.0La-Cu/SiO
2
-u, (g) 3.0La-Cu/SiO
2
-u, and (h) Cu/La O -u.
2 3
trapped in the silica texture in the form of copper phyllosilicate
into metallic Cu. The H -TPR peak for the Cu/SiO -u catalyst
origins of the two H-adsorption species, a H -TPD test was also
2
2
2
precursor is broad and centered at 527 K. Minute amounts (0.1
conducted on a pure silica support. The presence of a weak H2
or 0.5 wt %) of La species introduced into the xLa-Cu/SiO -u
desorption peak in the 670 to 900 K range on the H -TPD
2
2
catalyst precursors appear to have no effect on the reduction
behavior of the Cu species. As the La loading is increased to
profile for SiO indicates that the observed H desorption at
high temperatures corresponds to a number of strongly
2
2
2
+
1
5
.0 wt %, the center of the reduction peak slightly shifts from
27 to 535 K. Further increasing the La loading to 3.0 wt %
chemisorbed H species on the silica support surface (Figure
2
44
6). Similarly, an investigation by Bettahar et al. on H2
chemisorption over silica also showed the existence of surface
H-species, which were proposed to be formed during the
thermal treatment of silica with H . Generally, H desorption
increases the reduction temperature to 543 K. In the in situ
XRD characterization under 5% H −95% Ar atmosphere, XRD
peaks for Cu species emerge at 523 K on Cu/SiO -u and
2
0
2
2
2
1
.0La-Cu/SiO -u catalysts but are only observable at higher
peaks at low temperature ranges correspond to chemisorbed H
2
2
45−49
temperature (573 K) on 3.0La-Cu/SiO -u catalyst (Figures 3
on the metallic Cu surface.
This ascription is supported by
2
and Supporting Information Figure S2-B), also confirming that
appearance of the 323 K H desorption peak only in silica-
based catalysts containing Cu. The intensities of the two H2
2
the introduction of the La species into the Cu/SiO -u catalyst
2
2
+
significantly retards the reduction of the Cu species. Two
reduction peaks centered at 745 and 954 K, which are
significantly higher than the reduction temperature for the Cu/
SiO -u catalyst precursor, are observed on the Cu/La O -u
desorption peaks of the La-Cu/SiO -u catalysts are all clearly
2
affected by La loading. As revealed by N O chemisorption, the
2
DCu and SACu for the La-Cu/SiO -u and Cu/SiO -u catalysts
2
2
significantly differ even though the actual Cu loadings for these
catalysts are nearly identical at ca. 15 wt %. For the 1.0La-Cu/
SiO -u catalyst, which possesses the highest D and SACu
2
2
3
catalyst precursor. This result is consistent with the in situ XRD
characterization of the phase evolution of the Cu/La O -u
2
3
2
Cu
catalyst precursor in a reductive atmosphere (Supporting
Information Figure S2−B). In that analysis, the XRD peaks
for CuO only disappear after the temperature is increased to
among all catalysts, the strongest H desorption peaks are
located at 323 and 805 K, implying that surface Cu atoms are
highly correlated with the formation of the two chemisorbed H-
2
7
23 K. La O reduction by H usually occurs at extremely high
species. However, compared with the unmodified Cu/SiO -u
2
3
2
2
temperatures (above 1173 K). However, with the use of metals
capable of activating H and supplying active H-species, La O
catalyst, the 3.0La-Cu/SiO -u catalyst with slightly lower D
2
Cu
and SACu exhibits the 323 and 805 K H desorption peaks with
2
2
3
2
may be reduced at relatively low temperatures. For example,
slightly higher intensities, suggesting that the La species are also
4
0
Bell and Rieck have observed that lanthanum oxides in the
vicinity of Pd NPs can be partially reduced by H2 at
involved in H chemisorption and storage on the La-Cu/SiO -u
2
2
catalysts. Therefore, the introduction of La species into the Cu/
2
744
dx.doi.org/10.1021/cs400574v | ACS Catal. 2013, 3, 2738−2749

ACS Catalysis
Research Article
SiO -u catalyst significantly enhances H activation and
2
2
promotes the surface concentration of the active H-species.
These promotional effects may originate from (1) the presence
of appropriate La species, which improves the dispersion of Cu
atoms trapped in the silica texture and stabilizes the metallic Cu
formed by reductive activation in small size, and in turn, the
amount of surface metallic Cu atom that can dissociatively
activate H molecules and bond with active H-species increases;
2
(
2) the appropriate modification of the silica surface by La
species, which can increase the number of special adsorption
sites capable of strongly chemisorbing H-species; and (3) H2
spillover or transfer from metallic Cu to the silica support,
which is possibly enhanced by a special Cu-LaO -SiO interface.
x
2
3.4. Chemical States of Surface Species. Prior to XPS
measurements, all catalysts were reductively activated by H at
2
623 K for 4 h. The XPS spectra of Cu 2p (Figure 7) show that,
Figure 8. Cu LMM XAES spectra of as-reduced catalysts. (a) Cu/
SiO -u, (b) 0.1La-Cu/SiO -u, (c) 0.5La-Cu/SiO -u, (d) 1.0La-Cu/
2
2
2
SiO -u, and (e) 3.0La-Cu/SiO -u.
2
2
Table 2. Deconvolution Results of XPS and Cu LMM XAES
of La-Cu/SiO Catalysts
2
a
b
KE, eV
AP, eV
Cu 2p3/2
BE, eV
Cu+
Cu0
Cu+
Cu0
XCu , %
c
+
catalyst
Cu/SiO -u
914.0
914.0
918.0
918.2
1846.9
1846.9
1850.9
1851.1
932.9
932.9
46.9
49.7
2
0
0
1
3
.1La-Cu
SiO -u
/
2
.5La-Cu
SiO -u
914.0
914.0
918.0
918.0
918.0
1846.9
1846.9
1850.9
1850.9
1850.9
932.9
932.9
932.9
51.6
54.9
/
2
.0La-Cu
SiO -u
/
2
.0La-Cu
SiO -u
914.0
1846.9
56.0
/
2
Figure 7. Cu 2p XPS spectra of as-reduced catalysts. (a) Cu/SiO -u,
2
a
b
c
+
Kinetic energy. Auger parameter. Intensity ratio between Cu and
(
b) 0.1La-Cu/SiO -u, (c) 0.5La-Cu/SiO -u, (d) 1.0La-Cu/SiO -u, (e)
2
2
2
+
0
(Cu + Cu ) by deconvolution of Cu LMM XAES spectra.
3
.0La-Cu/SiO -u, and (f) Cu/La O -u.
2
2
3
over the activated Cu/SiO -u and xLa-Cu/SiO -u catalysts, the
interaction between Cu and La species, possibly from the
formation of a distinct chemical bond (Cu−O−La). Thi+s
strong interaction may not only increase the amount of Cu
species but also prevent Cu NPs on the catalyst surface from
thermally migrating and accumulating.
2
2
Cu 2p3/2 and Cu 2p1/2 peaks appear at binding energies of
32.8 and 952.7 eV, respectively. This result suggests that most
9
2+
of the Cu species in both the Cu/SiO -u and the xLa-Cu/
SiO -u catalyst precursors can be reduced to Cu and/or Cu in
2
0
+
2
the low valence state during reductive activation by H at 623 K
For the Cu/La O -u catalyst with high La content, the La
3d5/2 XPS peak at 835.1 eV is attributed to the La species
(Supporting Information Figure S6). No obvious La 3d5/2 XPS
2
2
3
3
+
because of the absence of a shakeup satellite peak located
50
approximately 10 eV higher than the Cu 2p3/2 binding energy.
On the Cu/La O -u catalyst, a shakeup satellite peak at 942.8
peak is observed on the xLa-Cu/SiO -u (x = 0.1, 0.5, 1.0, and
2
3
2
2
+
eV, characteristic of the Cu species, is observed, confirming
that a temperature of 623 K is insufficient to reduce the Cu/
La O -u catalyst because of the strong interaction between La
3.0) catalysts, possibly because of the relatively low La loading.
Therefore, an accurate determination of the chemical states and
surface distribution of the La species is not accomplished.
However, the possibility of the existence of partially reduced
LaO species on the H reduction-activated La-Cu/SiO -u
2
3
and Cu species. This observation is consistent with the H -TPR
2
and in situ XRD results. Given that the Cu 2p and Cu 2p1/2
3
/2
x
2
2
+
0
binding energies for Cu and Cu are nearly the same, Cu
LMM XAES is usually adopted to discriminate these peaks. A
broad, asymmetric Auger kinetic energy peak ranging from 904
to 926 eV is observed on the XAES data for the activated Cu/
SiO -u and xLa-Cu/SiO -u catalysts (Figure 8), strongly
catalysts could not be excluded.
3.5. In Situ FT-IR Study of CO Chemisorption. FT-IR
investigation of CO chemisorption is also useful in distinguish-
ing various surface cupreous species on the supported Cu
catalysts. All catalysts were activated by H at 623 K for 4 h
2
2
2
0
+
indicating the simultaneous occurrence of Cu and Cu species
on these catalysts. Deconvolution of the Auger kinetic energy
peak into two symmetric peaks at ca. 914 and 918 eV, which
correspond to Cu and Cu , respectively, can help determine
the surface distribution of these two species. Deconvolution
results (Table 2) show that the proportion of surface Cu+
species gradually increases with increasing La loading. This
result further proves our hypothesis on the existence of a strong
prior to CO adsorption at RT for 0.5 h. The FT-IR spectra of
the strongly absorbed CO species, which are the reserved
species after evacuation for 0.5 h, are shown in Figure 9. The
+
0
activated La-Cu/SiO -u catalysts containing La display a
2
−1
significantly stronger IR band at ca. 2120 cm compared
with that of the unmodified Cu/SiO -u catalyst. According to
2
51
+
literature, the CO species chemisorbed on Cu sites exhibit
significantly higher stability than those on Cu or Cu sites.
2+
0
2
745
dx.doi.org/10.1021/cs400574v | ACS Catal. 2013, 3, 2738−2749

ACS Catalysis
Research Article
Table 3. Performance of Vapor-Phase Chemoselective
a
Hydrogenation of DMO to EG over La-Cu/SiO Catalysts
2
selec., %
c
STY
,
EG
b
−1 −1
catalyst
conv., %
EG
MG others
g g‑cat
h
1
.0La/SiO -u
0.3
91.3
97.4
97.9
98.5
96.9
79.5
3.8
5.5
52.5
60.1
76.3
80.3
67.8
23.2
4.8
94.5
47.3
39.5
22.4
17.6
31.2
76.7
95.2
33.8
69.5
0.0
0.3
0.5
1.3
2.2
1.0
0.1
0.0
0.7
0.0
0.00
2
Cu/SiO -u
0.38
0.46
0.59
0.62
0.52
0.14
0.00
0.50
0.05
2
0
0
1
1
3
.1La-Cu/SiO -u
2
.5La-Cu/SiO -u
2
.0La-Cu/SiO -u
2
d
.0La-Cu/SiO -u
2
.0La-Cu/SiO -u
2
Cu/La O -u
2
3
1
.0La-Cu/SiO2-i
.0La-Cu/SiO -m
96.6
25.9
65.5
30.3
Figure 9. In situ FT-IR spectra of chemisorbed CO on as-reduced
catalysts. (a) 1.0La/SiO -u, (b) Cu/SiO -u, (c) 0.1La-Cu/SiO -u, (d)
1
2
2
2
2
a
Reaction conditions: T = 453 K, P(H ) = 3.0 MPa, H /DMO = 80,
2
2
0
.5La-Cu/SiO -u, (e) 1.0La-Cu/SiO -u, (f) 3.0La-Cu/SiO -u, and (g)
2 2 2
−1 −1
b
WLHSV = 1.5 g g−cat h . Others include EtOH, 1,2-BDO, and
1
DMO
Cu/La O -u.
2
3
c
,2-PDO. STY represents the space time yield of EG for the fresh
EG
−1
−1
catalysts, gram of EG per gram of catalyst per hour (g g−
h ).
cat
+
Moreover, the IR band for linear CO-Cu usually appears in the
d
The catalyst was calcined and reduced at 723 K.
1
−1
2
110 cm− to 2160 cm range. Therefore, this IR band i+s
reasonably ascribed to the linearly bonded CO on the Cu
species. Given that the Cu/La O -u catalyst cannot be reduced
2
3
2
+
at 623 K and most of the cupreous species are Cu , no strongly
chemisorbed CO is observed (Figure 9g). La species seem to
be incapable of chemisorbing CO because no observable IR
band for chemisorbed CO species can be found on the 1.0La/
SiO -u catalyst (Figure 9a). The La incorporation-induced
2
+
increase in surface Cu species, as determined by FT-IR
characterization of CO chemisorption, is consistent with the
aforementioned XAES and XPS results and further confirms
our supposition that a strong interaction exists between Cu and
La species. In addition, the slight red-shifting of the IR band for
+
the linearly bonded CO on the Cu species as the La loading
increases may be due to the effect of the amount of neighboring
Figure 10. DMO hydrogenation over (A) Cu/SiO -u and (B) 1.0La-
2
Cu/SiO -u catalysts as a function of WLHSVDMO. Reaction conditions:
alkaline LaO , which exhibits high electron affinity, on the
2
x
+
T = 453 K, P(H ) = 3.0 MPa, H /DMO = 80.
electron state of the Cu species. Similarly, a previous study by
Pestryakov and Davydov showed that the shift in the CO
stretching vibration frequency of the reduced CuO/La O /
2
2
41
3.0La-Cu/SiO -u (Table 3). Apparently, the introduction of La
2
3
2
+
Al O3 catalyst to the low-frequency range of Cu -CO
species into the Cu/SiO -u catalyst within the appropriate
2
2
complexes may be due to electronic donor−acceptor
content range significantly improves in the catalytic perform-
interactions of Cu ions with the oxygen ions of La O .
ance of the Cu/SiO -u catalyst for DMO hydrogenation. The
2
3
2
3
.6. Catalytic Activity and Stability. In vapor-phase
1.0La-Cu/SiO -i catalyst from the impregnation method also
shows superior DMO conversion (96.6%) and EG selectivity
2
DMO hydrogenation, a series of successive hydrogenation
reactions, including DMO hydrogenation to methyl glycolate
(65.5%) compared with those of the Cu/SiO -u catalyst, further
2
(
MG), MG hydrogenation to EG, and deep hydrogenation of
confirming that the La species promote the DMO hydro-
EG to ethanol, simultaneously occur. A number of reaction
factors, including pressure, space velocity, temperature, and
relative ratio of substrates, significantly affect catalytic perform-
ance. In the present study, an evaluation of the catalytic
performance of all catalysts in vapor-phase DMO hydro-
genation was conducted under identical conditions: 453 K, 3.0
MPa, DMO weight liquid hourly space velocity (WLHSVDMO)
genation activity of the Cu/SiO catalyst. The low catalytic
2
performance of the 1.0La-Cu/SiO -m catalyst from the
2
mechanical mixing method may be due to the serious damage
on the Cu/SiO -u catalyst precursor structure induced by the
2
long, high-energy milling treatment. This damage was revealed
by XRD and static N -physisorption characterizations.
2
In general, high metal dispersion is assumed to promote the
0.5 g g‑cat−
1
h
−1
to 3.5 g g
−1 −1
h , and H /DMO = 80. At a
hydrogenation activity of Cu/SiO catalysts,
16,52
which can be
=
‑cat
2
2
−
1
−1
low WLHSVDMO of 1.5 g g‑cat h , Cu/SiO -u and most of the
affected by several parameters like promoters/additives and
2
La-Cu/SiO -u catalysts display high hydrogenation activity,
pretreatment conditions. The control experiments on 1.0La-
2
with a DMO conversion above 90% (Table 3). Increasing the
Cu/SiO -u catalyst using 723 K (presumably converting
2
−1
−1
WLHSVDMO to 3.5 g g‑cat
h
results in a significant decrease
LaCO OH to La O ) as calcination/reduction temperature
3
2
3
in DMO conversion. However, a DMO conversion of 69.2%
show drops in SBET, SA , and lower activity (Tables 1, 3, and
cu
can still be obtained over the optimum 1.0La-Cu/SiO -u
4), probably indicating that the La species formed at 623 K may
be better and the sintering of Cu NPs may occur under higher
treatment temperatures. The measured DCu for the different
catalysts listed in Table 1 appear to correlate well with their
2
catalyst (Figure 10). The La loading clearly affects the catalytic
activity and EG selectivity, which follows the order 1.0La-Cu/
SiO -u > 0.5La-Cu/SiO -u > 0.1La-Cu/SiO -u > Cu/SiO -u >
2
2
2
2
2
746
dx.doi.org/10.1021/cs400574v | ACS Catal. 2013, 3, 2738−2749

ACS Catalysis
Research Article
Table 4. TOF of La-Cu/SiO Catalysts for the Vapor-Phase
Chemoselective Hydrogenation of DMO to EG
addtion is increased to 1.0 wt %, there is only a gradual increase
2
a
+
in the amount of Cu site associated with almost no changes in
0
that of the Cu one. Further increase of La amount to 3.0 wt %
selec., %
+
0
will cause drops in the amounts of both Cu and Cu sites,
probably due to the blockage of La species. Therefore, the
increased catalytic activity for DMO hydrogenation may be
mainly ascribed to the surface concentrations of active sites for
c
WLHSV
,
TOF,
DMO
−1
−
1
b
−1
catalyst
g g‑cat
h
conv., %
EG
MG others
h
Cu/SiO -u
4.0
4.0
13.3
22.7
7.4
92.6
88.7
0.0
0.0
6.4
2
0
0
1
1
3
.1La-Cu
SiO -u
11.3
10.1
both H and DMO molecule activation, which are assumed to
2
/
2
0
+
be Cu and Cu sites, respectively. The H -TPD results clearly
2
.5La-Cu
SiO -u
5.0
6.0
4.5
4.0
1.5
4.0
1.5
23.6
28.5
24.9
3.6
15.2
22.7
16.1
2.3
84.8
77.2
83.9
97.7
95.2
87.0
69.5
0.0
0.1
0.0
0.0
0.0
0.0
0.0
12.7
17.6
12.4
2.1
prove that the concentrations of two types of chemisorbed H-
/
2
species all reach the maximum on the most active 1.0La-Cu/
.0La-Cu
SiO -u
+
/
SiO -u catalyst. With respect to the Cu catalytic sites for DMO
2
2
.0La-Cu
activation, the cupreous species located at the interface between
Cu NPs and the catalyst support are likely stabilized in the +1
valence state as a result of the electron donor−acceptor transfer
between interfacial Cu atoms and the adjacent oxygen atoms of
d
/
SiO -u
2
.0La-Cu
SiO -u
/
2
Cu
/
3.8
4.8
2.8
3
2,41
La O -u
the oxide support.
On the basis of this hypothesis, the
2
3
+
1
.0La-Cu
26.5
25.9
13.0
30.3
14.2
4.9
chemical state of the Cu species on the Cu/SiO -u catalyst
2
/
SiO2-i
may be quite different from that on the La-Cu/SiO -u catalyst
2
1
.0La-Cu
SiO -m
because of the existence of a strong interaction between
/
2
adjacent Cu and La species, as confirmed by H -TPR and in
2
a
Reaction conditions: T = 453 K, P(H ) = 3.0 MPa, H /DMO = 80.
2
2
situ XRD characterizations. Therefore, the difference between
b
c
Others include EtOH, 1,2-BDO and 1,2-PDO. TOF values were
+
the chemical states of the Cu species on the Cu/SiO -u and
d
2
calculated according to the Cu dispersion in Table 1. The catalyst was
La-Cu/SiO -u catalysts may be partially responsible for their
2
calcined and reduced at 723 K.
significantly dissimilar TOFs for DMO catalytic hydrogenation.
To date, the high deactivation and inadequate lifespan of Cu-
based hydrogenation catalysts remain significant problems that
hinder the practical industrialization of vapor-phase hydro-
genation of DMO to EG. The long-term catalytic performances
hydrogenation activities. However, the catalytic activity is not
simply determined by the amount of surface metallic Cu atoms
because the TOF values for the Cu/SiO -u and xLa-Cu/SiO -u
2
2
of the optimal 1.0La-Cu/SiO -u and unmodified Cu/SiO -u
catalysts significantly change with the La loading (Table 4).
2
2
catalysts were compared under conventional reaction con-
ditions, namely, 473 K, 3.0 MPa, H /DMO = 80, and
The optimal 1.0La-Cu/SiO -u catalyst with the highest DMO
2
−1
hydrogenation performance possesses a TOF of 17.6 h , which
2
−1
−1
WLHSVDMO = 1.5 g g‑cat
h
(Figure 11). The 1.0La-Cu/
is significantly higher than that of unmodified Cu/SiO -u (6.4
2
−1
h ). This result strongly suggests that the catalytic cycles of
vapor-phase DMO hydrogenation are not related exclusively to
the surface Cu sites. For the vapor-phase DMO hydrogenation
0
catalyzed with heterogeneous Cu-based catalysts, surface
adsorption and activation on the catalytic sites of substrates
and of reaction intermediates, including H , DMO, and MG
2
molecules, as well as the surface reactions of the activated
species, are all vital catalytic steps. To date, although the
catalytic mechanism of DMO hydrogenation remains unclear, a
number of assumptions have been preliminarily proposed to
interpret the complicated hydrogenation behaviors of Cu-based
2
catalysts. For example, Chen et al. suggested that H2
0
dissociative activation on Cu surfaces and the CO bond in
+
DMO molecules are polarized and activated on Cu sites. As
can be seen from Table 1, the measurement of N O
2
chemisorption reveals that the values of DCu and SACu tend
to increase when an appropriate La amount is introduced into
Figure 11. DMO hydrogenation over Cu/SiO -u and 1.0La-Cu/SiO -
2
2
u catalysts as a function of time on stream. Reaction conditions: T =
1
the Cu/SiO -u catalysts by urea-assisted gelation. From the
473 K, P(H ) = 3.0 MPa, H /DMO = 80, WLHSV
h .
= 1.5 g g‑cat−
2
2
2
DMO
−1
deconvolution results of Cu LMM XAES listed in Table 2, the
+
+
0
+
+
XCu ratio (XCu = Cu /(Cu + Cu )) also increases with the
increase of La addition. Combining the results of N O
SiO -u catalyst retains its high catalytic activity, with no
significant changes, during DMO hydrogenation for 200 h. On
2
2
chemisorption and XPS, we can make an estimation on the
0
+
amounts of surface Cu and Cu sites. The results show that the
the other hand, the unmodified Cu/SiO -u catalyst is clearly
2
+
0
amounts of Cu and Cu sites on the La-free Cu/SiO -u
deactivated within 80 h. These results indicate the superior
catalytic stability of the modified catalyst for long-term
hydrogenation operations. In general, Cu sintering induced
by thermal surface mass transfers, structural collapse, or erosion
of the catalyst support is believed to be the main cause of
2
2
−1
2
−1
catalyst are about 13.6 m g and 15.3 m g , respectively.
When 0.1 wt % La is added, the amounts of Cu and Cu sites
on the 0.1La-Cu/SiO -u catalyst increase to 15.6 m g and
5.7 m g , respecively. We consider that the increases are
+
0
2
−1
2
2
−1
1
5
3,54
mainly resulted from increases of DCu and SACu and the strong
catalyst deactivation.
The Cu/SiO -u catalyst seems to be
2
interactions between the La and Cu species. When the La
quite stable within the first 45 h on stream but suddenly lost its
2
747
dx.doi.org/10.1021/cs400574v | ACS Catal. 2013, 3, 2738−2749

ACS Catalysis
Research Article
catalytic activity in relatively short time (from 50 to 80 h). We
have repeated the experiment several times, and the sharp
deactivation is always observed. Among possible causes for
catalyst deactivation, catalyst poisoning and blocking of active
sites by carbon deposit could be excluded since the
regeneration of deactivated catalyst by calcination and
rereduction did not restore the catalytic activity. It is found
AUTHOR INFORMATION
■
Corresponding Authors
(H.L.) Phone: +86-592-2182287. Fax: +86 592 2182287. E-
mail: hqlin@xmu.edu.cn.
(Y.Y.) Phone: +86-592-2181659. Fax: +86 592 2183047. E-
*
*
mail: yzyuan@xmu.edu.cn.
Notes
that the Cu particle size of Cu/SiO -u increased obviously after
2
The authors declare no competing financial interest.
the serious catalyst deactivation. Therefore we would like to
ascribe the main cause for catalyst deactivation to the
deterioration of Cu dispersion. However, the growth of Cu
particles during catalytic operation of DMO hydrogenation
should be a continuous and gradual process, which is unlikely
to induce the sharp decay of catalytic activity. As we know, the
ACKNOWLEDGMENTS
■
We gratefully acknowledge the financial support from the
National Basic Research Program of China (2011CBA00508),
the Natural Science Foundation of China (21173175 and
20923004), the Research Fund for the Doctoral Program of
Higher Education (20110121130002), and the Program for
Changjiang Scholars and Innovative Research Team in
University (IRT1036).
synergistic effect between Cu and Cu+ species exists in
0
0
catalytic hydrogenation of DMO. The Cu sites activate H
2
molecules and Cu+ species polarize and activate the ester
groups of DMO molecules. Thus, we speculate that the size of
Cu particles and the strength of interaction between metal and0
support can greatly influence the surface distribution of Cu
REFERENCES
■
+
+
and Cu species. For the Cu species, which possibly locate in
(
1) Yue, H. R.; Zhao, Y. J.; Ma, X. B.; Gong, J. L. Chem. Soc. Rev.
the interface between Cu particles and SiO support, the
2012, 41, 4218−4244.
2
amount may significantly decrease when the growth of Cu
particle exceeds a special extent due to the obvious change of
interfacial state. This and other results likewise demonstrate
that the hydrogenation of DMO is a structure-sensitive
(2) Chen, L. F.; Guo, P. J.; Qiao, M. H.; Yan, S. R.; Li, H. X.; Shen,
W.; Xu, H. L.; Fan, K. N. J. Catal. 2008, 257, 172−180.
(
(
3) Kerr, R. A.; Service, R. F. Science 2005, 309, 101.
4) Xu, G. H.; Li, Y. C.; Li, Z. H.; Wang, H. J. Ind. Eng. Chem. Res.
2
2,25
1995, 34, 2371−2378.
5) Xu, Z. N.; Sun, J.; Lin, C. S.; Jiang, X. M.; Chen, Q. S.; Peng, S.
Y.; Wang, M. S.; Guo, G. C. ACS Catal. 2013, 3, 118−122.
6) Brands, D. S.; Poels, E. K.; Bliek, A. Appl. Catal., A 1999, 184,
reaction.
Therefore, the decrease of surface cupreous
(
species and the serious destruction of synergetic catalysis
0
+
between Cu and Cu active sites can lead to the sharp decay in
catalytic activity. The considerably enhanced catalytic stability
(
2
79−289.
of the 1.0La-Cu/SiO -u catalyst may be ascribed to the
(7) Wall, R. G. U.S. Patent 4,149,021, 1979.
(8) Bartley, W. J. U.S. Patent 4,628,128, 1986.
2
stabilizing effect of the La species on the surface cupreous
species via strong interaction or bonding between them.
(
9) Lin, L.; Pan, P. B.; Zhou, Z. F.; Li, Z. J.; Yang, J. X.; Sun, M. L.;
Yao, Y. G. Chin. J. Catal. 2011, 32, 957−969.
10) Wang, S. R.; Li, X. B.; Yin, Q. Q.; Zhu, L. J.; Luo, Z. Y. Catal.
Commun. 2011, 12, 1246−1250.
11) Yin, A. Y.; Guo, X. Y.; Dai, W. L.; Fan, K. N. Acta Chim. Sin.
009, 67, 1731−1736.
12) Yin, A. Y.; Guo, X. Y.; Dai, W. L.; Fan, K. N. J. Phys. Chem. C
010, 114, 8523−8532.
13) Yin, A. Y.; Guo, X. Y.; Dai, W. L.; Li, H. X.; Fan, K. N. Appl.
Catal., A 2008, 349, 91−99.
14) Li, Z. X.; Qian, Z. G.; Zhao, X. G.; Xiao, W. D. Chem. React. Eng.
(
4. CONCLUSIONS
(
2
(
2
The catalytic behaviors of vapor-phase, chemoselective hydro-
genation of DMO over a series of La-modified Cu/SiO2
catalysts were investigated. Marked improvements in both
hydrogenation activity and catalytic stability were observed after
(
introduction of a La promoter. The 1.0La-Cu/SiO -u catalyst
2
prepared by urea-assisted gelation and which has a low La
loading of 1 wt % displayed the highest catalytic performance.
The existence of strong interactions between the La promoters
(
Technol. 2004, 20, 121−128.
(15) Yin, A. Y.; Guo, X. Y.; Fan, K. N.; Dai, W. L. ChemCatChem
2010, 2, 206−213.
and the Cu species was deduced from H -TPR and in situ XRD
2
(
16) Zhu, Y. Y.; Wang, S. R.; Zhu, L. J.; Ge, X. L.; Li, X. B.; Luo, Z. Y.
results. N O chemisorption and H -TPD characterizations
2
2
Catal. Lett. 2010, 135, 275−281.
17) Zhao, L.; Zhao, Y. J.; Wang, S. P.; Yue, H. R.; Wang, B.; Lv, J.;
Ma, X. B. Ind. Eng. Chem. Res. 2012, 51, 13935−13943.
18) Yue, H. R.; Zhao, Y. J.; Zhao, L.; Lv, J.; Wang, S. P.; Gong, J. L.;
Ma, X. B. AIChE J. 2012, 58, 2798−2809.
19) Gong, J. L.; Yue, H. R.; Zhao, Y. J.; Zhao, S.; Zhao, L.; Lv, J.;
further showed that the metallic dispersion of Cu and surface
concentration of chemisorbed H-species were all enhanced by
(
the introduction of La promoters into the Cu/SiO catalyst. In
2
(
+
addition, the amount of surface Cu species, which are assumed
to be involved in the activation of DMO molecules, also
increased, as confirmed by X-ray Auger electron spectroscopy
and in situ FT-IR of chemisorbed CO. The formation of
(
Wang, S. P.; Ma, X. B. J. Am. Chem. Soc. 2012, 134, 13922−13925.
(20) Yin, A. Y.; Guo, X. Y.; Dai, W. L.; Fan, K. N. J. Phys. Chem. C
2
(
009, 113, 11003−11013.
21) Wen, C.; Cui, Y. Y.; Yin, A. Y.; Fan, K. N.; Dai, W. L.
ChemCatChem 2013, 5, 138−141.
distinct Cu−O−La bonds between LaO and the cupreous
x
species located at interfacial sites may be responsible for the
significantly improved catalytic performance and stability.
(
22) He, Z.; Lin, H. Q.; He, P.; Yuan, Y. Z. J. Catal. 2011, 277, 54−
6
3.
ASSOCIATED CONTENT
(23) Lin, H. Q.; Zheng, X. L.; He, Z.; Zheng, J. W.; Duan, X. P.;
Yuan, Y. Z. Appl. Catal., A 2012, 445−446, 287−296.
24) Lin, H. Q.; Duan, X. P.; Zheng, J. W.; Zheng, X. L.; He, P.;
Yuan, Y. Z.; Yang, Y. H. RSC Adv. 2013, 3, 11782−11789.
(25) Huang, Y.; Ariga, H.; Zheng, X. L.; Duan, X. P.; Takakusagi, S.;
Asakura, K.; Yuan, Y. Z. J. Catal. 2013, 307, 74−83.
■
*
S
Supporting Information
(
XRD, in situ XRD, HAADF-STEM images, elemental EDX
maps, and La 3d XPS spectra of some catalysts. This material is
available free of charge via the Internet at http://pubs.acs.org.
2
748
dx.doi.org/10.1021/cs400574v | ACS Catal. 2013, 3, 2738−2749

ACS Catalysis
Research Article
(
26) Monnier, J. R.; Hanrahan, M. J.; Apai, G. J. Catal. 1985, 92,
19−126.
27) Rao, R.; Dandekar, A.; Baker, R. T. K.; Vannice, M. A. J. Catal.
997, 171, 406−419.
28) Sitthisa, S.; Sooknoi, T.; Ma, Y. G.; Balbuena, P. B.; Resasco, D.
E. J. Catal. 2011, 277, 1−13.
29) Wang, Y. N.; Duan, X. P.; Zheng, J. W.; Lin, H. Q.; Yuan, Y. Z.;
Ariga, H.; Takakusagi, S.; Asakura, K. Catal. Sci. Technol. 2012, 2,
637−1639.
30) Zhao, S.; Yue, H. R.; Zhao, Y. J.; Wang, B.; Geng, Y. C.; Lv, J.;
Wang, S. P.; Gong, J. L.; Ma, X. B. J. Catal. 2013, 297, 142−150.
31) Zhou, Y. F.; Yang, Y. L.; Qi, X. W.; Jiang, Y. W.; Yang, J.; Ren, X.
J.; Yang, Q. X. J. Rare Earths 2012, 30, 1069−1074.
32) Kam, R.; Selomulya, C.; Amal, R.; Scott, J. J. Catal. 2010, 273,
3−81.
33) Kok, E.; Scott, J.; Cant, N.; Trimm, D. Catal. Today 2011, 164,
97−301.
34) Ozawa, M.; Toda, H.; Suzuki, S. Appl. Catal., B 1996, 8, 141−
55.
35) Liu, H. C.; Wang, H.; Shen, J. H.; Sun, Y.; Liu, Z. M. Catal.
Today 2008, 131, 444−449.
36) Cunha, A. F.; Mahata, N.; Orfao
1
(
1
(
(
1
(
(
(
7
(
2
(
1
(
́
(
̃
, J. J. M.; Figueiredo, J. L. Energy
Fuel 2009, 23, 4047−4050.
(
(
37) Dandekar, A.; Vannice, M. A. J. Catal. 1998, 178, 621−639.
38) Li, G. G.; Shang, M. M.; Geng, D. L.; Yang, D. M.; Peng, C.;
Cheng, Z. Y.; Lin, J. CrystEngComm 2012, 14, 2100−2111.
(
39) Toupance, T.; Kermarec, M.; Lambert, J. F.; Louis, C. J. Phys.
Chem. B 2002, 106, 2277−2286.
(40) Rieck, J. S.; Bell, A. T. J. Catal. 1985, 96, 88−105.
(41) Pestryakov, A. N.; Davydov, A. A. Appl. Surf. Sci. 1996, 103,
4
(
79−483.
42) Vidal, H.; Bernal, S.; Baker, R. T.; Finol, D.; Omil, J. A. P.;
Pintado, J. M.; Rodríguez-Izquierdo, J. M. J. Catal. 1999, 183, 53−62.
43) Sing, K. S. W.; Everett, D. H.; Haul, R. A. W.; Moscou, L.;
Pierotti, R. A.; Rouquerol, J.; Siemieniewska, T. Pure Appl. Chem.
985, 57, 603−619.
44) Boudjahem, A. G.; Monteverdi, S.; Mercy, M.; Ghanbaja, D.;
Bettahar, M. Catal. Lett. 2002, 84, 115−122.
45) Wilmer, H.; Genger, T.; Hinrichsen, O. J. Catal. 2003, 215,
(
́
1
(
(
1
(
(
88−198.
46) Waugh, K. C. Solid State Ionics 2004, 168, 327−342.
47) Ahlers, J. A.; Grasser, J. A.; Loveless, B. T.; Muggli, D. S. Catal.
Lett. 2007, 114, 185−191.
48) Genger, T.; Hinrichsen, O.; Muhler, M. Catal. Lett. 1999, 59,
37−141.
49) Dong, X.; Zhang, H. B.; Lin, G. D.; Yuan, Y. Z.; Tsai, K. R.
(
1
(
Catal. Lett. 2003, 85, 237−246.
(50) Kim, K. S. J. Electron Spectrosc. 1974, 3, 217−226.
(51) Hadjiivanov, K.; Knozinger, H. Phys. Chem. Chem. Phys. 2001, 3,
1
(
132−1137.
52) Lambert, S.; Cellier, C.; Grange, P.; Pirard, J. P.; Heinrichs, B. J.
Catal. 2004, 221, 335−346.
53) Lin, J. D.; Zhao, X. Q.; Cui, Y. H.; Zhang, H. B.; Liao, D. W.
Chem. Commun. 2012, 48, 1177−1179.
54) Wen, C.; Cui, Y. Y.; Dai, W. L.; Xie, S. H.; Fan, K. N. Chem.
Commun. 2013, 49, 5195−5197.
(
(
2
749
dx.doi.org/10.1021/cs400574v | ACS Catal. 2013, 3, 2738−2749