Efficient hydrogenation of CO2-derived ethylene carbonate to methanol and ethylene glycol over Mo-doped Cu/SiO2 catalyst

Article abstract of DOI:10.1016/j.cattod.2020.07.070

Hydrogenation of ethylene carbonate (EC) to methanol and ethylene glycol is an attractive method for indirect hydrogenation of CO₂ to methanol. Copper-based catalysts have been studied in EC hydrogenation because of their good selective activation ability for carbon-oxygen bonds. Herein, a series of Mo-doped Cu/SiO₂ catalysts were prepared and used for EC hydrogenation. The catalyst with 0.5 wt% Mo dopant exhibited the highest catalytic performance of 80.9 % MeOH yield and 98.6 % EG yield with an excellent stability for at least 140 h. Detailed characterizations revealed that a proper amount of Mo addition could be beneficial to enhancing the copper dispersion and preventing them from aggregation. Moreover, it is demonstrated that the amount of surface Cu⁺ species was increased with some electron-deficient ones generated resulting from the strong interaction between copper and molybdenum species, which may effectively promote the activation of EC. The optimal 0.5 wt% Mo-doped catalyst showed remarkably enhanced efficiency and the apparent activation energy barrier of 115.8 kJ mol^?1 is much lower than that of the Mo-free sample in EC hydrogenation. The insights may bring new possibilities to further design efficient copper-based catalysts for the hydrogenation of carbon-oxygen bonds.

Full text of DOI:10.1016/j.cattod.2020.07.070

Catalysis Today xxx (xxxx) xxx
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Efficient hydrogenation of CO -derived ethylene carbonate to methanol and
2
ethylene glycol over Mo-doped Cu/SiO catalyst
2
a
,
1
a
,
1
a
a
a
a
Youwei Yang , Dawei Yao , Mengjiao Zhang , Antai Li , Yueqi Gao , Busha Assaba Fayisa ,
a
,
b
a,
b
a,
b
,
a,b,
Mei-Yan Wang , Shouying Huang , Yue Wang
*, Xinbin Ma
*
a
Key Laboratory for Green Chemical Technology of Ministry of Education, Collaborative Innovation Center of Chemical Science and Engineering, School of Chemical
Engineering and Technology, Tianjin University, Tianjin 300072, PR China
b
Joint School of National University of Singapore and Tianjin University, International Campus of Tianjin University, Binhai New City, Fuzhou, 350207, PR China
A R T I C L E I N F O
A B S T R A C T
Keywords:
Hydrogenation of ethylene carbonate (EC) to methanol and ethylene glycol is an attractive method for indirect
Copper-based catalyst
Ethylene carbonate
Methanol
hydrogenation of CO
2
to methanol. Copper-based catalysts have been studied in EC hydrogenation because of
catalysts
their good selective activation ability for carbon-oxygen bonds. Herein, a series of Mo-doped Cu/SiO
2
were prepared and used for EC hydrogenation. The catalyst with 0.5 wt% Mo dopant exhibited the highest
catalytic performance of 80.9 % MeOH yield and 98.6 % EG yield with an excellent stability for at least 140 h.
Detailed characterizations revealed that a proper amount of Mo addition could be beneficial to enhancing the
copper dispersion and preventing them from aggregation. Moreover, it is demonstrated that the amount of
Ethylene glycol
Hydrogenation
Carbon dioxide
+
surface Cu species was increased with some electron-deficient ones generated resulting from the strong inter-
action between copper and molybdenum species, which may effectively promote the activation of EC. The
optimal 0.5 wt% Mo-doped catalyst showed remarkably enhanced efficiency and the apparent activation energy
ꢀ 1
barrier of 115.8 kJ mol is much lower than that of the Mo-free sample in EC hydrogenation. The insights may
bring new possibilities to further design efficient copper-based catalysts for the hydrogenation of carbon-oxygen
bonds.
1
. Introduction
conditions, which has attracted extensive attention [7]. In this route,
ethylene carbonate (EC) is used as an intermediate product due to its
Carbon dioxide (CO
2
) is not only a greenhouse gas but also a low-
2
efficient synthesis by cycloaddition of CO into ethylene oxide [8,9].
cost, non-toxic and renewable carbon resource. Its conversion and uti-
lization to valuable chemicals and materials is an attractive approach to
reduce the negative impact on environment as well as to alleviate the
resource shortages [1]. Methanol (MeOH), as a vital chemical feedstock
Sequently, the hydroganation of EC is conducted under mild conditions
to obtain MeOH as well as high value-added ethylene glycol (EG) with
100 % atomic economy [7,10–12]. The EC production via CO
2
reacting
with ethylene oxide has been industrialized in the OMEGA process, in
which followed by hydrolysis of the carbonate to generate EG and
and liquid fuel, has been considered as an important product of CO
conversion [2,3]. However, due to the thermodynamically stable and
inert property of its carbon-oxygen bonds, direct conversion of CO to
MeOH remains a great challenge [4,5]. Harsh reaction conditions are
generally required as well as the single-pass conversion in the direct CO
2
release CO
2
again [1,13]. Thus, combining the OMEGA process with EC
2
hydrogenation step could provide a new way for producing MeOH by
utilization of CO
S1).
2
and obtaining additional product EG as well (Scheme
2
hydrogenation is difficult to improve because of the chemical equilib-
rium [3,6].
Copper-based catalysts have been widely studied in many ester hy-
drogenation reactions, such as hydrogenation of dimethyl oxalate to EG
[14,15], hydrogenation of methyl acetate to ethanol [16], due to their
Recently, an alternative and indirect method has been brought up for
efficient conversion of CO
2
to MeOH under moderate reaction
excellent ability in carbon-oxygen bond activation and inertness in C C
–
*
Corresponding authors at: Key Laboratory for Green Chemical Technology of Ministry of Education, Collaborative Innovation Center of Chemical Science and
Engineering, School of Chemical Engineering and Technology, Tianjin University, Tianjin 300072, PR China.
E-mail addresses: yuewang@tju.edu.cn (Y. Wang), xbma@tju.edu.cn (X. Ma).
1
These authors contributed equally to this work.
https://doi.org/10.1016/j.cattod.2020.07.070
Received 5 May 2020; Received in revised form 1 July 2020; Accepted 30 July 2020
Available online 6 September 2020
0
920-5861/© 2020 Elsevier B.V. All rights reserved.
Please cite this article as: Youwei Yang, Catalysis Today, https://doi.org/10.1016/j.cattod.2020.07.070

Y. Yang et al.
Catalysis Today xxx (xxxx) xxx
◦
bondcleavage [17,18]. Li and co-workers used copper chromite catalyst
for EC hydrogenation which achieved 60 % and 93 % selectivities to
MeOH and EG respectively [12]. Recently, some mesoporous silica
supported copper-based catalysts, such as Cu/HMS [19], Cu/MCM-41
Finally, the catalyst was calcined at 450 C for 4 h, pressed, crushed, and
sieved to 40 ꢀ 60 mesh.
The Mo doped Cu/SiO
2
catalysts were prepared by incipient wetness
powder with aqueous solu-
impregnation (IWI) method using Cu/SiO
2
[
20], have been employed for continuous EC hydrogenation and ob-
tion of (NH
4
)
6
Mo
7
O
24⋅4H
2
O. Then the catalysts were dried at room
◦
tained excellent catalytic performance benefiting from the high specific
surface areas and suitable pore structures. The role of copper species of
different chemical states as well as the effects of supports and promoters
have also been investigated, which exhibit great influence on the cata-
lytic performance in EC hydrogenation [10,21,22]. It has been reported
temperature for 24 h and calcined at 500 C for 4 h. These samples were
2
noted as xMo-Cu/SiO where x represented the weight percentage of Mo
in the catalyst. For comparison, 1.0Mo/SiO
2
was also prepared by above
method.
+
◦
that balanced Cu and Cu species plays a crucial role in enhancing the
2.2. Catalyst characterization
◦
+
catalytic ability, where the Cu could activate hydrogen and Cu could
function as Lewis acidic sites to polarize the carbon-oxygen bond [23].
Li et al. [24] found introducing suitable amounts of Zr and Mg into
The contents of copper and molybdenum of the samples were
determined by the inductively coupled plasma optical emission spec-
trometer (ICP-OES, Varian Vista-MPX). First, the catalyst was
completely dissolved with HF solution and then a saturated boric acid
solution was added before analysis.
Cu/SiO
2
catalyst could decrease copper particle size and increase the
+
amount of Cu , which were helpful for the adsorption and activation of
EC and resulted in enhanced catalytic activity and stability of EC hy-
drogenation. However, due to EC special structure with two adjacent
alkoxyl groups, the efficient hydrogenation of carbon-oxygen bond is
still challenging [10,25]. Besides copper particle size and valence dis-
tribution, the reaction is found to be closely related to the surface acidity
and basicity [10,26]. It was found that strong and excessive basicity or
acidity of the catalyst could induce the side reactions, such as decar-
bonylation and decarboxylation, and block active sites by strongly
adsorbed reactants, thereby decreasing the EC conversion and product
selectivity. Thus, appropriate pomoter is of great significance to improve
the catalytic performance for EC hydrogenation.
◦
N2 adsorption-desorption isotherms were measured at ꢀ 196 C by
using a Micromeritics Tristar 3000 apparatus. Before testing, the sample
◦
was degassed at 300 C for 4 h. The specific surface area was calculated
by the Brunauer ꢀ Emmett ꢀ Teller (BET) method, and the pore size
distribution was calculated by the Barrett ꢀ Joynerꢀ Halenda (BJH)
method.
X-ray diffraction (XRD) was conducted on Rigaku Model C/max-
2500 diffractometer with Cu-K
α radiation (λ =1.5406 Å). The catalyst
◦
for testing was reduced at 300 C for 4 h in H flow. The pattern was
2
◦
◦
◦
acquired from 10 to 90 with a rate of 8 /min.
In the pioneering work, Poels et al. [27] studied different promoters
Transmission electron microscopy (TEM) image was obtained by
using a JEM-2100 F system electron microscope. The catalyst was first
in Cu/SiO
was
2
catalyst for methyl acetate hydrogenolysis and the activity
arranged in the following order
◦
:
reduced at 300 C for 4 h in H2 flow and then was cooled to room
Mo > Co ≥ Zn ≥ M>Fe > Y>Ni ≥ Mg. Additionally, it was found that
the introduction of Mo promoter could help reduce the particle size and
promote the active metal dispersion [28]. Recent works also reported
that Mo-decorated noble metal catalysts could give enhanced selectivity
temperature. The reduced powder was added to ethanol and dispersed
with sonication. Finally, the solution was dropped onto a copper-grid-
supported carbon membrane, dried and tested immediately.
The H2 temperature-programmed reduction (H -TPR) was per-
2
–
of C
–
O bond hydrogenation because of the generated metal-additive
formed on the Micromeritics Autochem II 2920 equipped with a thermal
interaction [29,30]. These studies suggested that Mo could be an
effective promoter to improve the performance of copper-based catalyst
in EC hydrogenation, the role of which is worthy for further study.
conductivity detector (TCD). Firstly, 50 mg catalyst was pretreated in
◦
the quartz tube under He flow at 200 C for 2 h and then cooled to room
◦
temperature. Finally, the catalyst was heated to 800 C in 10 % H /Ar
2
◦
In this work, we fabricated a series of Mo-doped Cu/SiO
2
catalysts
with a heating rate of 10 C/min.
and optimized its contents for EC hydrogenation to produce MeOH and
EG. The conversion of EC and yield of MeOH were remarkedly improved
on the catalyst doped with an appropriate Mo content. The effects of Mo
doping on the structure and physicochemical properties were system-
N O titration was operated on Autochem II 2920 apparatus to
2
determine metallic copper surface area and copper dispersion. In brief,
◦
50 mg of the sample was firstly pretreated in Ar for 1 h at 200 C, then
◦
10 % H /Ar flow was introduced to reduce the catalyst for 2 h at 300 C.
2
◦
atically characterized by N
2
physisorption, XRD, TEM, XPS, in situ FTIR
Next, the sample was cooled to 90 C in Ar flow and then exposed to N O
2
of CO adsorption and so on. It was demonstrated that the addition of Mo
promoter could regulate the chemical state and surface composition of
copper species, consequently affecting the catalytic performance. We
believe these insights of the structure-activity relationship may provide
instructive suggestions for the further design of catalyst for the hydro-
genation reactions.
for 1 h, ensuring the surface metallic copper was entirely oxidized to
Cu O. After purged for 30 min with Ar flow, the sample was reduced
2
◦
again at 300 C by pulse injection of 10 % H /Ar.
2
Raman spectra were collected at room temperature by a Renishaw
invia reflex system. A laser of 532 nm was employed as the spectral
excitation line. The catalyst used for the test was firstly reduced at
◦
3
00 C for 4 h in H
2
flow and then tested immediately.
2
. Materials and methods
X-ray photoelectron spectra (XPS) and Auger electron spectra (AES)
were treated on Perkin-Elmer PHI 1600 ESCA equipped with a mono-
2
.1. Catalyst preparation
chromatic Al K
α
X-ray source (h
ν
=1486.6 eV) to detect surface Cu
◦
species. First, the catalyst was reduced at 300 C for 4 h and then cooled
The Cu/SiO
typical ammonia evaporation method [16], briefly described as follows.
Firstly 26.3 g of Cu(NO O was dissolved in 100 mL of deionized
⋅3H
2
catalyst with 30 wt% Cu loading was prepared by the
2
to room temperature in the H flow. Next, the catalyst was placed into
centrifuge tube and sealed with Ar. Then the sample was transferred to
the holder immediately and outgassed in the chamber. The analysis was
3
)
2
2
ꢀ 8
water, then 90 mL of aqueous ammonia solution (25 wt%) was added.
After the mixture was stirred for 30 min, 44.5 mL of silica sol (30 wt%)
was added dropwise to the above solution. The initial pH value of the
suspension was maintained at 11 ꢀ 12 and stirred for 4 h at room
taken under a vacuum of 1 × 10 Torr. The binding energy was cali-
brated using the C1s peak at 284.6 eV as the reference. The experimental
error was within ±0.2 eV.
The in situ FTIR of CO adsorption was performed on the Nicolet 6700
spectrometer. Briefly, self-supporting wafer was made by about 20 mg
sample and placed in the in situ cell. Next, the sample was reduced at
◦
temperature. Next, the suspension was heated to 80 C for ammonia
evaporation. This process was terminated when the pH value was
◦
◦
decreased to 6 ꢀ 7. The precipitate was separated by filtration, then
300 C for 1 h under 10 % H
flow to obtain the background spectrum. Then CO gas was introduced
2
/Ar and cooled down to 30 C under He
◦
washed with deionized water three times and dried at 110 C overnight.
2

Y. Yang et al.
Catalysis Today xxx (xxxx) xxx
for 30 min. Finally, the spectra were scanned and recorded again under
He flow until they appeared unchanged.
metallic Cu become broader and weaker. Correspondingly, the particle
size calculated by the Scherrer equation (Table S2) decreases from
4
.4 nm to 3.9 nm, suggesting that the addition of Mo species could
reduce the Cu particle size [14]. The copper crystallite size of
.0Mo-Cu/SiO catalyst is 4.3 nm, simillar to the Mo-free Cu/SiO
2
.3. Catalytic activity
1
2
2
catalyst. Combining the fact that the Mo-modified catalysts were
The catalytic activity for the hydrogenation of EC was investigated in
◦
calcined again at a higher temperature (500 C), it is demonstrated that
a stainless-steel fixed-bed reactor with a thermocouple inserted into the
catalyst bed to monitor the reaction temperature. The catalyst pellets
a proper amount of molybdenum addition could alleviate the thermal
migration and aggregation of copper species during the
high-temperature calcination and reduction processes [14].
(
40–60 mesh) were loaded into the reactor and reduced in H
2
flow
◦
(
3 MPa, 100 mL/min) at 300 C for 4 h at a temperature ramping rate of
◦
TEM images of the reduced Cu/SiO
shown in Fig. 1. It is noted that all catalysts have similar morphologies.
The particle size distribution (inset) of Cu/SiO catalyst displays a wide
range with the average diameter at 5.2 nm. After Mo doping, the mean
particle size is reduced to 4.9 nm and 4.6 nm for 0.1Mo–Cu/SiO and
.5Mo–Cu/SiO catalysts respectively with narrowed distributions,
while the average size is 6.1 nm of 1.0Mo–Cu/SiO and some large
particles appeares, in agreement with the XRD results.
2
2
and xMo–Cu/SiO catalysts are
2
C/min from room temperature. After reduction, the catalyst was
cooled down to the reaction temperature. Then, the EC solution (10 wt%
in 1,4-dioxane) was injected continuously from the top of the reactor
2
through a high-pressure pump and was vaporized in the stream of H
gas.
2
2
0
2
The liquid products were condensed and analyzed off-line with an
Agilent Micro GC 6820 instrument with an HP-INNOWAX capillary
2
column (Hewlett-Packard, 30 m ×0.32 mm ×0.50
μ
m). To ensure the
In order to identify the existence states of molybdenum species,
Raman spectra of the reduced catalysts were collected [31,32], and the
accuracy and repeatability of the results, 2–3 samples were measured
under each reaction condition. For the catalyst stability test, the reaction
was performed continuously and the sample was collected every 2 or
ꢀ 1
results are illustrated in Fig. 2. An obvious and broad peak at 893 cm
is observed in the 1.0 wt% Mo-doped catalyst, which could be ascribed
do-
3
h.
to the vibrational of Mo-O/Mo = O bond in the oligomeric MoO
x
mains [32,33]. Notably, as the Mo loading decreases, the peak intensity
is dramatically declined. It implies that Mo species is highly dispersed on
the catalyst surface when the Mo dopant is less than 1.0 wt%, consistent
with the results of no observable Mo species in XRD and TEM. The
3
. Results and discussion
3
2 2
.1. Textural properties of Cu/SiO and xMo-Cu/SiO catalysts
2
distinct peak at 1.0 Mo-Cu/SiO suggests that molybdenum species has
The metal contents of the as-prepared catalysts measured by ICP-OES
are listed in Table 1. The copper loadings of all catalysts are about 30 wt
, while molybdenum loadings vary in the range of 0.1–1.0 wt% as
designed. The N adsorption-desorption isotherms and pore size distri-
accumulated to some extent, which maybe not able to effectively sepa-
rate the copper species and result in the aggregation [14,34].
%
0
We further measured the metallic copper surface area (SCu) and
2
2
copper dispersion (DCu) by N O titration. As shown in Table 1 and
butions are displayed in Fig. S1. All the samples possess Langmuir type
IV isotherms with hysteresis loops, corresponding to the typical meso-
porous structure [14]. As shown in Table S1, texture parameters for all
0
Table S2, with the increasing Mo loading, SCu and DCu are first increased
and then decreased and the highest DCu (14.6 %) is obtained on
0
.5Mo–Cu/SiO
2
catalyst, in line with the XRD and TEM results. Notably,
the SCu of 1.0Mo–Cu/SiO is found smaller than that of the Mo-free Cu/
SiO catalyst, indicating that some surface copper particles were
aggregated or covered by excessive Mo species [14].
catalysts are close to each other. The specific surface areas are over
0
2
2
4
40 m /g and average pore sizes are around 6.7 nm, indicating that the
2
introduction of molybdenum species has no significant effect on the
porosity of the catalyst.
XRD patterns of the reduced catalysts are presented in Fig. S2. The
◦
3.2. Chemical states of cooper species of Cu/SiO
catalysts
2 2
and xMo-Cu/SiO
broad peak that appears at 2θ of 22 in all samples is attributed to
◦
amorphous silica. The characteristic peak at 36.4 belongs to Cu
2
O
◦
◦
phase (JCPDS 65-3288), and the peaks cencered at 43.3 and 50.5
could be ascribed to metallic Cu phase (JCPDS 04-0836). No obvious
diffraction peaks of Mo species could be observed, indicating the pres-
ence of highly dispersed Mo species on the catalysts. When the Mo
addition is increased to 0.5 wt%, the characteristic diffraction peaks of
H
2
-TPR profiles of the as-prepared catalysts were performed to
investigate the reducibility of copper species. As shown in Fig. 3, it is
observed that the reduction peak of Cu/SiO
2
catalyst is mainly
◦
concentrated at 216 C, which is attributed to the overlapped reduction
◦
peaks of the well-dispersed CuO to Cu species and copper phyllosilicate
+
to Cu species [35]. With the increase of Mo content, the reduction
Table 1
temperature is gradually decreased and then increased, consistent with
the copper dispersion [36]. Moreover, the peaks of the Mo-doped cata-
lysts are more symmetrical and sharper compared with the Mo-free
Physicochemical properties of Cu/SiO
2
and xMo-Cu/SiO
2
catalysts.
a
a
b
0
Cu
c
+ d
+ e
Catalyst
M
Cu
M
Mo
d
Cu
S
X
Cu
S
Cu
2
ꢀ 1
cat
2
ꢀ 1
g
cat
(
wt%)
(wt%)
(nm)
(m
g
)
(%)
(m
)
Cu/SiO
2
catalyst, suggesting narrower size distribution and more uni-
Cu/SiO
2
31.2
30.6
/
5.2
4.9
26.1
25.2
27.5
35.7
9.9
◦
form dispersion of copper particles. Additionally, a weak peak at 290 C
0
0
1
.1Mo-
Cu/
0.10
14.0
19.2
10.6
is observed on Cu/SiO
2
catalyst, corresponding to the reduction of large
◦
CuO particles [35]. The broad shoulder peak around 273 C on
SiO
2
.5Mo-
Cu/
30.2
31.7
0.49
1.05
4.6
6.1
28.0
22.3
40.7
32.3
2
1.0Mo-Cu/SiO could be ascribed to the overlapped reduction peaks of
large CuO particles, and some of which may interact with molybdenum
SiO
2
◦
species [37]. The weak peak of 1.0Mo-Cu/SiO
2
above 600 C may be
.0Mo-
Cu/
attributed to the reduction of molybdenum oxides. For comparison, the
SiO
2
1.0Mo/SiO
2
sample was prepared, which showed a wide range of
◦
a
b
c
reduction temperature at 400ꢀ 650 C. It has been reported that highly
Cu and Mo contents obtained by ICP-OES analysis.
dispersed molybdenum oxides could be reduced at a relatively low
Cu particle size measured from the TEM image.
◦
0
temperature, while bulk ones could be reduced at above 600 C [31].
Thus, it is suggested that excessive Mo species in 1.0Mo-Cu/SiO would
Metallic Cu surface area (SCu) obtained from N
2
O titration.
d
e
+
+
+
0
Cu /(Cu + Cu ) calculated from Cu LMM AES spectra.
2
+
0
+
0
Cu surface area (SCu) estimated on the basis of SCu and XCu assuming the Cu
agglomerate during the high-temperature treatment, consistent with the
Raman spectra.
+
and Cu have the same surface area and atomic sensitivity factor.
3

Y. Yang et al.
Catalysis Today xxx (xxxx) xxx
2 2
Fig. 1. TEM images of the reduced Cu/SiO and xMo-Cu/SiO catalysts (inset:particle size distribution).
2 2 2
Fig. 3. H -TPR profiles of the calcined Cu/SiO and xMo-Cu/SiO catalysts.
2 2
Fig. 2. Raman spectra of the reduced Cu/SiO and xMo-Cu/SiO catalysts.
species [34].
XPS was carried out to investigate surface chemical state of copper
species of the reduced catalysts. As shown in Fig. 4(a), there are two
peaks at binding energies of 932.4 eV and 952.5 eV, typically assigned
to Cu 2p3/2 and Cu 2p1/2 peaks respectively [14]. The absence of Cu 2p
◦
+
Since Cu and Cu species have rather similar binding energies, Cu
LMM AES spectra were detected to determine the different copper spe-
cies [36]. As shown in Fig. 4(b), the broad and asymmetric peak pre-
2
sented on Cu/SiO catalyst is deconvoluted into two symmetrical peaks
2
+
satellite peak at 942ꢀ 944 eV indicates that the Cu species has been
◦
+
centered at 569.7 eV and 572.7 eV, which are attributed to Cu and Cu ,
respectively [35]. Notably, as the Mo content increases, the binding
+
0
completely reduced to Cu and Cu under the reduction condition [16],
as shown by XRD. Further, a slight shift of Cu2p3/2 peak to higher
binding energy is observed after doping Mo species, suggesting the
strong interaction and charge transfer between copper and molybdenum
+
energy of Cu gradually shifts to higher value and arrives at 573.1 eV,
further indicating the charge transfer from copper to molybdenum
species [14]. It is demonstrated that the addition of Mo species could
4

Y. Yang et al.
Catalysis Today xxx (xxxx) xxx
2 2
Fig. 4. Cu2p XPS spectra (a) and Cu LMM AES spectra (b) of the reduced Cu/SiO and xMo-Cu/SiO catalysts.
+
+
◦
decrease the electron density of Cu species resulting from the strong
interaction. The similar electronic interaction was reported on Mo pro-
moted Ir-based catalyst [30].
and the latter is Cu ꢀ CO adsorption domain. Further increasing the
-1
temperature to 100 C, the band around 2108 cm is sharply reduced,
while that around 2124 cm- comes out (Fig. S3), indicating the stronger
1
+
+
-1
The molar proportion of surface Cu (XCu) calculated from the AES
CO adsorption at 2124 cm [14].
+
deconvolution result is listed in Table 1. It can be seen that the XCu of Mo
Based on above discussion, the asymmetric peaks in Fig. 5 could be
ꢀ 1
doped catalysts is increased compared with the Mo-free Cu/SiO
2
catalyst
deconvoluted into three symmetrical peaks centered at 2080 cm ,
ꢀ 1
ꢀ 1
and reaches the maximum value of 40.7 % at 0.5 wt% Mo dopants. By
2
2108 cm and 2124 cm , respectively. The 0.5Mo-Cu/SiO catalyst is
◦
+
◦
ꢀ 1
assuming the Cu and Cu species have the same surface area and
given the largest Cu ꢀ CO adsorption at 2080 cm , indicating that it
has the highest metallic copper dispersion, in line with the TEM and TPR
results. The FTIR of CO adsorption could be used to verify the amount of
+
+
atomic sensitivity factor, the surface area of Cu (SCu) could be esti-
◦
+
+
mated on the basis of SCu and XCu [16]. Correspondingly, the SCu is
+
+
+
varied with the Mo loading as well. The SCu is first increased and ob-
surface Cu species as well [39]. The integral area of Cu ꢀ CO band
2
+
+
tained the maximum value of 19.2 m /g when the Mo content is
(ACu, Table S2) and the SCu of these reduced catalysts as the function of
Mo loading (Fig. S4) exhibit good consistency, and both of them reach
the highest value at the 0.5 wt% Mo-doped sample, which confirms the
accuracy and reliability on the quantitative measurement of the copper
species [16]. In addition, compared to the Mo-free sample, the vibration
increased to 0.5 wt%, and then sharply decreased with further increase
of Mo species (Table 1).
In situ FTIR spectra of CO adsorption could be also used to detect the
surface Cu species [14,38]. As shown in Fig. 5, the asymmetric CO ab-
sorption peak with a weak shoulder one is observed in the all reduced
+
bands of Cu ꢀ CO adsorption are gradually blue-shifted with the Mo
+
Cu/SiO
2
and Mo-modified catalysts. Since no CO adsorption is found on
, the peaks are attributed to the CO adsorption at copper
addition, suggesting the decreasing electronic density of surface Cu
1
.0Mo/SiO
2
species [14,30]. Correspondingly, the integral area ratio under the peaks
2
+
◦
-1
-1
sites. Generally, the CO adsorption on Cu and Cu sites is considered to
disappear after purge at room temperature because of the weak inter-
action [16]. But it has been observed that some detectable CO adsorp-
of 2124 cm and 2108 cm displays a volcano-trend change as the Mo
loading increases, and maximizes the value of 1.64 at 0.5 wt% Mo
+
content (Table S2), indicating the change of electronic density of Cu
◦
◦
tion could occur at 27 C in small Cu crystallites, which would
disappear when further increasing the temperature [38]. Then in order
to identify the peaks in FTIR spectra of CO adsorption, we tested one
sample purged under different temperatures (Fig. S3). It is obvious that
species, agreed with the AES analysis.
Thus, it could be concluded that an appropriate amount of Mo
dopant could not only enhance the dispersion of copper active species,
+
but also facilitate the generation of more electron-deficient Cu sites
ꢀ 1
the band at 2080 cm disappears when the temperature is increased to
because of the strong interaction between copper and molybdenum
species. However, when the Mo content is increased to 1.0 wt%, the
promoting effects drops sharply due to the agglomeration of Mo species
and its coverage on active sites.
◦
5
0 C, whereas the asymmetric band is remained and accompanied with
◦
a shoulder peak, suggesting the former is the peak of Cu ꢀ CO adsorption
3
.3. Catalytic performance
2
The catalytic performances were tested over Cu/SiO and xMo-Cu/
2
SiO catalysts to investigate the effect of Mo content on vapor-phase EC
hydrogenation to MeOH and EG. As shown in Fig. 6, the EC conversion
and products yields could be significantly improved by introducing
suitable amount of Mo dopants. For comparison, 1.0Mo/SiO
2
was also
tested but there was barely EC converted (not given). Under the WHSV
ꢀ 1
(
weight hourly space velocity) of EC at 0.75 h , all catalysts exhibit 100
EC conversion and >97.5 % EG yield, while the MeOH yield is
remarkably changed with the different Mo loadings. It maybe because
the generation of CO or CO via the decarbonylation and decarboxyl-
%
2
ation reactions [12,26], which would reduce the yield of MeOH. The
possible reactions are shown in the supporting information (Eq.S1-S6).
The high yield of EG could be ascribed to few by-products of C3ꢀ 4 al-
cohols, such as 1,2-propanediol and 1,2-butanediol, formed by the
Fig. 5. In situ FTIR spectra of CO absorption of the reduced Cu/SiO
Cu/SiO catalysts.
2
and xMo-
Guerbet reactions. The few surface basic sites of SiO
2
support, as well as
◦
2
the relative low reaction temperature of 180 C could both effectively
5

Y. Yang et al.
Catalysis Today xxx (xxxx) xxx
◦
Fig. 6. The catalytic performance of Cu/SiO
2
and xMo-Cu/SiO
2
catalysts: conversion of EC (a), yields of MeOH (b) and EG(c) (Reaction conditions: 180 C, 3 MPa,
H
2
/EC = 180 mol/mol), (d) stability test of the 0.5Mo-Cu/SiO
2
catalyst (WHSV =1.0 hꢀ 1).
reduce the Guerbet reactions. Similar results were reported in the lit-
eratures using Cu/SiO for EC hydrogenation [40,41]. When the WHSV
increases, the EC conversion of Cu/SiO decreases rapidly, whereas that
of 0.5Mo-Cu/SiO could maintain 93.3 % at 2.0 h . The optimal per-
formance of 80.9 % yield of MeOH and 98.6 % yield of EG is obtained at
2
2
ꢀ 1
2
.0 h- WHSV on the 0.5 wt% Mo-doped sample. When Mo content is
1
1
increased to 1.0 wt%, the catalytic activity becomes even worse than the
blank one due to the partial coverage and aggregation of active species.
2
Moreover, we selected the best-performing 0.5Mo-Cu/SiO catalyst to
further evaluate the long-term stability in EC hydrogenation. The result
in Fig. 6(d) exhibits that the 0.5 wt% Mo doped catalyst possesses
excellent lifespans. The conversion of EC maintains constant at about
1
00 % and stable yields of MeOH and EG are also observed during 140 h
of time on stream.
3
.4. The role of Mo dopant
Fig. 7. Arrhenius-type plots for EC hydrogenation of Cu/SiO
2
and 0.5Mo-Cu/
2
In order to better understand the role of Mo in the Cu/SiO catalyst
SiO catalysts.
2
for EC hydrogenation, we further employed Arrhenius plots to analyze
the apparent activation energy barriers (Ea) of Cu/SiO and 0.5Mo-Cu/
2
catalyzing the ester hydrogenation reactions. In order to get a further
SiO
2
catalysts. As shown in Fig. 7, the Ea of Cu/SiO
2
catalyst is 153.7 kJ
+
insight, catalytic activity and SCu as a function of Mo loading was
ꢀ
1
ꢀ 1
mol , while that for the 0.5Mo-Cu/SiO
2
catalyst is 115.8 kJ mol
,
investigated. As shown in Fig. 8, along with the increase of Mo loading,
the EC conversion and MeOH yield get improved, which are similar to
suggesting Mo dopant could greatly reduce the Ea and enhance the
catalytic efficiency. In fact, some studies have reported that Mo-doped
Ru-based catalyst is beneficial for activating carbonyl-containing com-
+
+
the change of SCu. It is implied that the enhancement of Cu sites could
make great contribution to the catalytic performance. Some researchers
–
pounds and increasing selectivity of C
found that Mo-decorated Ir-based catalyst could form Ir-MoO
via the interaction between Ir and MoO species, which could promote
–
O hydrogenation [29]. It is also
+
found that the high proportion of Cu species could facilitate the con-
x
interface
version of intermediates during the process of EC hydrogenation [40],
x
+
because Cu can act as electrophilic or Lewis acidic sites to adsorb and
adsorption and activation of aldehyde group [30]. The carbon-oxygen
bonds of EC is quite stable and difficult to be activated because of the
resonance stabilization effect of two adjacent alkoxy groups [10,25].
Therefore, the decrease of Ea may result from the promoted adsorption
and activation of carbon-oxygen bonds of EC due to the introduction of
Mo dopant.
polarize carbo-oxygen groups [14,16]. It is observed that the Mo dopant
+
could increase XCu as well as produce the electron transfer from copper
to molybdenum due to their strong interaction, which maybe in favor of
+
stabilizing the surface Cu species [42]. Moreover, the generated
+
electron-deficient Cu sites are more conducive to the adsorption and
activation of the carbon-oxygen group, thus promoting the conversion
+
◦
It is generally accepted that the synergy of Cu and Cu is the key in
6

Y. Yang et al.
Catalysis Today xxx (xxxx) xxx
City (18JCQNJC06100) and National Postdoctoral Program for Inno-
vative Talents of China (BX20180211).
Appendix A. Supplementary data
Supplementary material related to this article can be found, in the
online version, at doi:https://doi.org/10.1016/j.cattod.2020.07.070.
References
[
[
[
[
1] W.T. Lee, A.P. van Muyden, F.D. Bobbink, Z. Huang, P.J. Dyson, Angew. Chem. Int.
Ed. 58 (2019) 557–560.
2] K.W. Ting, T. Toyao, S.M.A. Hakim Siddiki, K.-i. Shimizu, ACS Catal. 9 (2019)
3
685–3693.
3] C. Huang, J. Wen, Y. Sun, M. Zhang, Y. Bao, Y. Zhang, L. Liang, M. Fu, J. Wu, D. Ye,
L. Chen, Chem. Eng. J. 374 (2019) 221–230.
4] Z.-Q. Wang, Z. Xu, S.-Y. Peng, M. Zhang, G. Lu, Q.-S. Chen, Y. Chen, G.-C. Guo, ACS
Catal. 5 (2015) 4255–4259.
+
[5] K. Chen, H. Fang, S. Wu, X. Liu, J. Zheng, S. Zhou, X. Duan, Y. Zhuang, S.C.
Fig. 8. The variations of EC conversion, MeOH yield and Cu surface area
E. Tsang, Y. Yuan, Appl. Catal. B 251 (2019) 119–129.
against the Mo loading.
[
[
6] A. Bansode, A. Urakawa, J. Catal. 309 (2014) 66–70.
7] Z. Han, L. Rong, J. Wu, L. Zhang, Z. Wang, K. Ding, Angew. Chem. Int. Ed. 51
(
2012) 13041–13045.
efficiency as confirmed by the significant decrease of Ea. However,
excessive Mo addition will deteriorate the catalytic performance. As can
[
[
8] J. Sun, S. Zhang, W. Cheng, J. Ren, Tetrahedron Lett. 49 (2008) 3588–3591.
9] J. Sun, W. Cheng, W. Fan, Y. Wang, Z. Meng, S. Zhang, Catal. Today 148 (2009)
361–367.
be seen, the 1.0Mo-Cu/SiO
2
catalyst and Cu/SiO
2
catalyst have the
+
[
[
10] H. Liu, Z. Huang, Z. Han, K. Ding, H. Liu, C. Xia, J. Chen, Green Chem. 17 (2015)
4281–4290.
similar SCu, while the former one displayed lower conversion of EC and
◦
yield of MeOH (Fig. 8), which could be ascribed to its smaller SCu
11] X. Wu, L. Ji, Y. Ji, E.H.M. Elageed, G. Gao, Catal. Commun. 85 (2016) 57–60.
causing insufficient hydrogen dissociation [16]. It has been reported
that when the amount of surface metallic copper is sufficient, the cata-
lytic activity of ester hydrogenation is primarily related to the area of the
[12] C. Lian, F. Ren, Y. Liu, G. Zhao, Y. Ji, H. Rong, W. Jia, L. Ma, H. Lu, D. Wang, Y. Li,
Chem Commun. 51 (2015) 1252–1254.
[
13] J. Fu, N. Jiang, D. Ren, Z. Song, L. Li, Z. Huo, Int. J. Hydrogen Energy 41 (2016)
9
118–9122.
+
surface Cu sites [16]. It is illustrated that the synergy catalysis with a
[
14] Z. He, H. Lin, P. He, Y. Yuan, J. Catal. 277 (2011) 54–63.
◦
+
balanced Cu and Cu species is also of great significance in EC hydro-
genation, consistent with the reported literatures [19,24].
[15] P. Ai, M. Tan, N. Yamane, G. Liu, R. Fan, G. Yang, Y. Yoneyama, R. Yang,
N. Tsubaki, Chemistry 23 (2017) 8252–8261.
[
[
16] Y. Wang, Y. Shen, Y. Zhao, J. Lv, S. Wang, X. Ma, ACS Catal. 5 (2015) 6200–6208.
17] P. Ai, M. Tan, P. Reubroycharoen, Y. Wang, X. Feng, G. Liu, G. Yang, N. Tsubaki,
Catal. Sci. Technol. 8 (2018) 6441–6451.
4
. Conclusion
[
[
[
[
[
[
18] H. Yue, Y. Zhao, S. Zhao, B. Wang, X. Ma, J. Gong, Nat. Commun. 4 (2013)
2
339–2346.
In summary, we found that molybdenum could be an effective
19] X. Chen, Y. Cui, C. Wen, B. Wang, W.-L. Dai, Chem Commun. 51 (2015)
13776–13778.
dopant to improve the catalytic performance of Cu/SiO
hydrogenation of CO -derived EC to produce MeOH and EG. The
optimal 0.5 wt% Mo-doped Cu/SiO catalyst exhibits remarkable per-
2
catalyst for the
20] F. Deng, N. Li, S. Tang, C. Liu, H. Yue, B. Liang, Chem. Eng. J. 334 (2018)
2
1
943–1953.
2
21] Y. Ding, J. Tian, W. Chen, Y. Guan, H. Xu, X. Li, H. Wu, P. Wu, Green Chem. 21
(2019) 5414–5426.
formance of 100 % EC conversion, 98.6 % EG and 80.9 % MeOH yields
and shows a excellent stability during the 140 h long-term reaction. But
when excessive amount of Mo is introduced, the catalytic performance
would be deteriorated due to the surface coverage and aggregation of
the active sites. Furthermore, it is demonstrated that with the addition of
appropraite Mo species, the aggregation of copper particle is suppressed
and the Cu dispersion is promoted. Meanwhile, according to the quan-
22] C. Zhang, L. Wang, J. Liu, Y. Yang, P. He, Y. Cao, J. Chen, H. Li, ChemCatChem 10
(
2018) 4617–4628.
23] F. Li, L. Wang, X. Han, Y. Cao, P. He, H. Li, Int. J. Hydrogen Energy 42 (2017)
2144–2156.
[24] J. Tian, W. Chen, P. Wu, Z. Zhu, X. Li, Catal. Sci. Technol. 8 (2018) 2624–2635.
[
[
[
25] P.A. Dub, T. Ikariya, ACS Catal. 2 (2012) 1718–1741.
26] J. Kim, N. Pfander, G. Prieto, ChemSusChem 13 (2020) 2043–2052.
27] D.S. Brands, E.K. Poels, A. Bliek, Appl. Catal. A 184 (1999) 279–289.
+
tative characterizations of copper species, the surface area of Cu sites is
[28] X. Chen, H. Li, H. Luo, M. Qiao, Appl. Catal. A 233 (2002) 13–20.
[
[
29] L. Chen, Y. Zhu, H. Zheng, C. Zhang, Y. Li, Appl. Catal. A 411-412 (2012) 95–104.
increased as well as more electron-deficient ones formed benifiting from
the strong interaction between copper and molybdenum species, which
may effectively decrease the apparent activation energy barrier of the
hydrogenation of EC. The insights could enlighten us to further design
efficient copper-based catalyst for the hydrogenation of carbon-oxygen
bonds.
30] Y.-M. Xu, W.-B. Zheng, Y.-M. Hu, C. Tang, A.-P. Jia, J.-Q. Lu, J. Catal. 381 (2020)
2
22–233.
[31] M. Shetty, K. Murugappan, T. Prasomsri, W.H. Green, Y. Rom ´a n-Leshkov, J. Catal.
3
31 (2015) 86–97.
[
32] E. Kordouli, B. Pawelec, K. Bourikas, C. Kordulis, J.L.G. Fierro, A. Lycourghiotis,
Appl. Catal. B 229 (2018) 139–154.
[33] J.V. Silveira, J.A. Batista, G.D. Saraiva, J. Mendes Filho, A.G. Souza Filho, S. Hu,
X. Wang, Vib. Spectrosc. 54 (2010) 179–183.
[
[
34] A. Yin, C. Wen, X. Guo, W.L. Dai, K. Fan, J. Catal. 280 (2011) 77–88.
Declaration of Competing Interest
35] Y. Zhao, S. Zhao, Y. Geng, Y. Shen, H. Yue, J. Lv, S. Wang, X. Ma, Catal. Today 276
(
2016) 28–35.
[
[
36] G. Cui, X. Meng, X. Zhang, W. Wang, S. Xu, Y. Ye, K. Tang, W. Wang, J. Zhu,
M. Wei, D.G. Evans, X. Duan, Appl. Catal. B 248 (2019) 394–404.
37] R.G. Kukushkin, O.A. Bulavchenko, V.V. Kaichev, V.A. Yakovlev, Appl. Catal. B
163 (2015) 531–538.
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
[
[
[
[
38] A. Dandekar, M.A. Vannice, J. Catal. 178 (1998) 621–639.
39] D. Yao, Y. Wang, Y. Li, Y. Zhao, J. Lv, X. Ma, ACS Catal. 8 (2018) 1218–1226.
40] W. Chen, T. Song, J. Tian, P. Wu, X. Li, Catal. Sci. Technol. 9 (2019) 6749–6759.
41] M. Zhou, Y. Shi, K. Ma, S. Tang, C. Liu, H. Yue, B. Liang, Ind. Eng. Chem. Res. 57
(2018) 1924–1934.
Acknowledgements
We are grateful for the supports from the National Key R&D Program
of China (2018YFB0605801), the National Natural Science Foundation
of China (21706184, 21938008), Natural Science Foundation of Tianjin
[
42] Y. Zhao, H. Zhang, Y. Xu, S. Wang, Y. Xu, S. Wang, X. Ma, J. Energy Chem. 49
(
2020) 248–256.
7