Silicon nitride as a new support for copper catalyst to produce acrolein via selective oxidation of propene with very low CO2 release

Article abstract of DOI:10.1016/j.jcat.2019.09.019

To obtain thermally stable copper-based catalysts with excellent catalytic performance in the selective propene (C₃H₆) oxidation reaction, amorphous silicon nitride (Si₃N₄) was selected to anchor active copper species using a deposition–precipitation approach followed by air calcination at different temperatures. We found that the thermal treatment temperature remarkably modified the reactivity of the copper catalyst, and the 800 °C-calcined 10 wt% Cu/Si₃N₄ showed superior catalytic activity with 24.0% propene conversion and 86.2% acrolein selectivity at 325 °C, featuring a turnover frequency as high as 80.6 h^?1. With the help of transmission electron microscopy, X-ray absorption fine structure, and in situ X-ray diffraction, we have identified that the larger Cu₂O species account for the highly efficient formation of acrolein at 325 °C. From the CO temperature-programmed reduction results, we have confirmed that the presence of surface copper hydroxyls (Cu–OH) was closely related to the acrolein selectivity, since they favored CO₂ generation at the beginning of the reaction. Furthermore, surface copper hydroxyls can be effectively tuned by optimizing the air calcination temperature and thus improve the catalytic activity.

Full text of DOI:10.1016/j.jcat.2019.09.019

Journal of Catalysis xxx (xxxx) xxx
Contents lists available at ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Silicon nitride as a new support for copper catalyst to produce acrolein
via selective oxidation of propene with very low CO2 release
Ling-Ling Guo a,b, Jing Yu c, Miao Shu a,b, Lu Shen d, Rui Si a,e,
⇑
a Shanghai Institute of Applied Physics, Chinese Academy of Sciences, Shanghai 201800, China
b University of Chinese Academy of Science, Beijing 100049, China
c Shanghai Institute of Measurement and Testing Technology, Shanghai 200233, China
d Division of China, TILON Group Technology Limited, Shanghai 200090, China
e Shanghai Synchrotron Radiation Facility, Zhangjiang Laboratory, Shanghai 201204, China
a r t i c l e i n f o
a b s t r a c t
Article history:
To obtain thermally stable copper-based catalysts with excellent catalytic performance in the selective
propene (C3H6) oxidation reaction, amorphous silicon nitride (Si3N4) was selected to anchor active copper
species using a deposition–precipitation approach followed by air calcination at different temperatures.
We found that the thermal treatment temperature remarkably modified the reactivity of the copper cat-
alyst, and the 800 °C-calcined 10 wt% Cu/Si3N4 showed superior catalytic activity with 24.0% propene
Received 10 June 2019
Revised 11 September 2019
Accepted 15 September 2019
Available online xxxx
conversion and 86.2% acrolein selectivity at 325 °C, featuring a turnover frequency as high as 80.6 hꢀ1
.
Keywords:
Silicon nitride
Copper-based catalyst
Selective oxidation of propene
Surface hydroxyls
With the help of transmission electron microscopy, X-ray absorption fine structure, and in situ X-ray
diffraction, we have identified that the larger Cu2O species account for the highly efficient formation of
acrolein at 325 °C. From the CO temperature-programmed reduction results, we have confirmed that
the presence of surface copper hydroxyls (Cu–OH) was closely related to the acrolein selectivity, since
they favored CO2 generation at the beginning of the reaction. Furthermore, surface copper hydroxyls
can be effectively tuned by optimizing the air calcination temperature and thus improve the catalytic
activity.
Structure–activity relationship
Ó 2019 Elsevier Inc. All rights reserved.
1. Introduction
and many other reactions [46–48]. Among them, oxidized copper
materials have also been discovered to be active for selective oxi-
Acrolein is an important chemical intermediate and has been
widely used in resin production and organic synthesis [1–4]. The
selective oxidation of propene by molecular oxygen to provide
high-valued acrolein (CH2@CHACH3 + O2 ? CH2@CHACHO + H2O)
has been extensively explored, but is accompanied by several by-
products such as acetone, propanal, propylene oxide, and carbon
dioxide [5–10]. Since the 1960s, two major catalyst systems for
this reaction, bismuth molybdates [11–22] and copper-based cata-
lysts [10,23–35], have been widely investigated. However, how to
remarkably improve the acrolein formation and simultaneously
minimize the generation of by-products, especially CO2 (2CH2-
dation of C3H6 to acrolein [49,50].
In industry, selective oxidation of propene is currently carried
out under near-atmospheric pressure at 330 °C or above [32,51].
Thus, much effort has been made in fundamental research to
improve both propene conversion and acrolein selectivity for
copper-based catalysts [27,29,31]. For instance, Adams and Jen-
nings were the first to report the use of cuprous oxide as a catalyst
for selective propene oxidation, obtaining ~10% propene conver-
sion and 60–80% selectivity for acrolein [23,24]. Then Schüth and
co-workers applied copper phthalocyanine as a copper precursor
that was extremely diluted (below 0.1%) in a silica matrix, which
showed moderate activity (conversion 40.1%; selectivity 25.5%) at
475 °C [29]. In 2014, Yang et al. reported vacuum–thermal treated
Cu/SBA-15 catalysts resulting in better catalytic stability, while
propene tends to be overoxidized to form CO2 with increasing reac-
tion temperature (30% propene conversion and 41% acrolein selec-
tivity at 300 °C) [31]. As a result of these findings, it seemed that
there was no good balance between propene conversion at high
@CHACH3 + 9O2 ? 6CO2 + 6H2O), is still
a
major challenge
[10,35-38]. On the other hand, copper-based catalysts have been
broadly used in selective hydrogenation [39–41], CO oxidation
[42,43], the water-gas shift [44,45], methanol dehydrogenation,
⇑
Corresponding author at: Shanghai Institute of Applied Physics, Chinese
Academy of Sciences, Shanghai 201800, China.
E-mail address: sirui@sinap.ac.cn (R. Si).
https://doi.org/10.1016/j.jcat.2019.09.019
0021-9517/Ó 2019 Elsevier Inc. All rights reserved.
Please cite this article as: L.-L. Guo, J. Yu, M. Shu et al., Silicon nitride as a new support for copper catalyst to produce acrolein via selective oxidation of
propene with very low CO2 release, Journal of Catalysis, https://doi.org/10.1016/j.jcat.2019.09.019

2
L.-L. Guo et al. / Journal of Catalysis xxx (xxxx) xxx
temperature and acrolein selectivity at low temperature under
realistic conditions [52,53].
30), where y is the copper content as a mass percentage (y = [Cu/
{Cu + Si3N4}] ꢂ 100%). For the synthesis of 10Cu/SBA-15–800,
detailed procedures were the same as for 10Cu/Si3N4-800, except
that the support was different. The Cu2O particles (10 wt%) were
physically mixed with Si3N4 powders by manual grinding for
20 min and denoted as b-Cu2O/Si3N4. For the leaching of 10Cu/
Si3N4-800, 0.8 g of the powders was washed for 4 h with 100 mL
hydrochloric acid aqueous solution (HCl, 5% aq.) at 30 °C under
stirring. After being centrifuged and washed in water (pH 7), the
sample was dried at 80 °C for 12 h and then air-calcined at
300 °C for 4 h.
Other researchers tried to introduce noble metal (Au) into this
system to improve the acrolein selectivity to 80–90% at 300 °C or
above, but the corresponding conversion of C3H6 decreased to less
than 1% [54,55]. Therefore, effectively enhancing the activity of
copper-based catalysts, as well as improving acrolein selectivity
under the propene oxidation process conditions (>300 °C, 1 atm),
still remains a great challenge. This is probably due to difficulties
in rational synthesis of thermally stable catalysts and control of
the interactions between copper and support [34,56,57]. Here,
the selection of an appropriate support plays a crucial role in
achieving high propene conversion and acrolein selectivity for
the selective oxidation of C3H6.
2.2. Catalyst characterization
In this paper, we report the successful synthesis of a series of
copper-based catalysts anchored on a new type of support, silicon
nitride (Si3N4), which takes multiple advantages [58–63] of (i) its
outstanding mechanical and thermal properties at elevated tem-
peratures, which have never been used in the selective propene
oxidation reaction; (ii) better thermally and chemically resistant
properties than silica, which is usually used as an oxide support
in Cu-based catalysts for this reaction. As a result, in this work,
we have found that the temperature of thermal treatment on the
as-dried samples remarkably tuned their catalytic activity for
selective oxidation of propene to produce acrolein. It is very inter-
esting that the acrolein selectivity was significantly promoted over
the 800 °C-calcined 10% Cu/Si3N4 catalyst, achieving 24.0% propene
conversion and 86.2% acrolein selectivity (yield > 20%) at 325 °C.
The enhanced reactivity can be attributed to the interaction
between the copper species and the thermally stable support of
the catalyst, as well as the partial elimination of surface copper
hydroxyls (Cu–OH) after high-temperature calcination, which
favors adsorbing C3H6 to form CO2 in this reaction.
2.2.1. Inductively coupled plasma atomic emission spectroscopy
The actual copper concentrations of the catalysts were analyzed
by inductively coupled plasma atomic emission spectroscopy (ICP-
AES; Optima 5300DV, PerkinElmer). The air-calcined samples
(fresh catalysts) were used directly for characterization. First,
0.1 g catalyst (accurate to 0.0001 g) was added to 2 mL hydrofluo-
ric acid under continuous stirring until the powder was dissolved
adequately. Second, the as-formed SiF4 was removed via evapora-
tion. Then, almost 3 mL of nitric acid was introduced and the solu-
tion was kept slightly boiling for 2 h. Finally, the solution was
cooled to nearly 25 °C and diluted for the ICP-AES test.
2.2.2. Nitrogen adsorption–desorption
Nitrogen adsorption–desorption measurements were carried
out at 77 K on an ASAP2020-HD88 analyzer (Micromeritics Co.).
The air-calcined samples (fresh catalysts) were used directly for
characterization. The powder samples were degassed at 250 °C
under vacuum (<100
lmHg) for 4 h before being analyzed. The
Brunauer–Emmett–Teller (BET) specific surface areas (SBET) were
calculated based on the data in the relative pressure (P/P0) range
between 0.05 and 0.20. The pore size distribution (dp) of the tested
sample was calculated from the desorption branch of the iso-
therms according to the Barrett–Joyner–Halenda (BJH) method.
2. Experimental
2.1. Catalyst preparation
2.2.3. Transmission electron microscopy
2.1.1. Materials
Transmission electron microscopy (TEM) experiments were
performed and scanning transmission electron microscopy–energy
dispersive spectrometry (STEM-EDS) elemental mapping results
were obtained on a FEI Tecnai G2 F20 microscope operating at
200 kV. The fresh (as calcined in air) and used (after 325 °C reac-
tion and exposure to air) catalysts were used for characterization.
The tested sample was prepared by suspension in ethanol, and
then a drop of this very dilute suspension was cast onto a
carbon-film-coated Mo grid. The as-formed sample grid was dried
sufficiently before being loaded into the TEM sample holder.
Amorphous silicon nitride (Si3N4, 20 nm, SBET = 67 m2/g, 99.9%),
a-silicon nitride (800 nm, 99.9%), and b-silicon nitride (1 lm,
99.9%) were obtained from Shanghai Yao Tian Nano Material Co.
SBA-15 (SBET = 600–800 m2/g, >99.5%) was purchased from Nanjing
JCNANO Technology Co. Copper nitrate (Cu(NO3)2ꢁ3H2O, 98.0–
102.0%), sodium carbonate (Na2CO3, 99.8%), hydrochloric acid
(HCl, 36.0–38.0%), and cuprous oxide (Cu2O, > 90.0%) were pur-
chased from Sinopharm Chemical Reagent Co. and used without
purification.
2.1.2. Catalyst preparation
2.2.4. X-ray absorption fine structure
The copper–silicon nitride catalysts were synthesized via a
deposition–precipitation method followed by thermal treatment
at different temperatures. Typically, 0.38 g Cu(NO3)2ꢁ3H2O
The X-ray absorption fine structure (XAFS) spectra at the Cu K-
edge (E0 = 8979 eV) were collected at the BL14W1 beamline of the
Shanghai Synchrotron Radiation Facility (SSRF) at a typical energy
of the storage ring of 3.5 GeV under the ‘‘top-up” mode with a con-
stant current of 250 mA. Fresh (as calcined in air) and used (after
325 °C reaction) catalysts were used for characterization. To obtain
accurate valence and coordination information on the used cata-
lysts, these samples were transferred into sample tubes and sealed
with Ar, after being cooled to room temperature under the reaction
gas. Then ca. 30 mg powder was pressed into a solid pellet sealed
with Kapton tape before the XAFS test. The XAFS data were
recorded in the transmission mode with a Si(1 1 1) monochroma-
tor and Oxford ion chambers. The energy was calibrated with the
absorption edge of pure Cu foil (K-edge, 8979 eV). Athena and Arte-
mis codes were applied to extract the data and fit the profiles. For
(1.6 mmol) was dissolved in 100 mL Millipore H2O (>18 M
X
ꢁcm)
under vigorous stirring and mixed with 0.9 g Si3N4 powders. The
pH of the Cu precursor and Si3N4 slurry was adjusted to ~ 9 by
0.5 M Na2CO3 aqueous solution and then aged for 4 h at 25 °C.
The mixture was centrifuged and washed three times with Milli-
pore water. The as-obtained solid was dried at 75 °C under vacuum
for 12 h and then calcined in still air at different temperatures for
4 h (ramp rate: 2 °C/min). The as-prepared 10Cu/Si3N4 (copper
loading amount: 10 wt%) catalyst calcined at x°C is denoted as
10Cu/Si3N4-x (x = 400, 600, 800, or 1000). For comparison, Cu/
Si3N4 was also synthesized with different copper loading amounts
calcined at 800 °C and labeled as yCu/Si3N4-800 (y = 5, 10, 20, and
Please cite this article as: L.-L. Guo, J. Yu, M. Shu et al., Silicon nitride as a new support for copper catalyst to produce acrolein via selective oxidation of
propene with very low CO2 release, Journal of Catalysis, https://doi.org/10.1016/j.jcat.2019.09.019

L.-L. Guo et al. / Journal of Catalysis xxx (xxxx) xxx
3
X-ray absorption near-edge spectroscopy (XANES), the experimen-
tal absorption coefficients as functions of energies, (E), were
2.2.8. Temperature-programmed reduction by carbon monoxide
l
Carbon monoxide (CO) TPR experiments using CO as reductant
were performed on a mass spectrometer (Tilon GRP Technology
Limited LC-D200M). The fresh (as calcined in air) catalysts were
used for characterization. The sample (100 mg, 20–40 mesh) was
pretreated for 30 min in a flow of 5% H2/Ar (30 mLꢁminꢀ1) at
300 °C, and then cooled down to 30 °C and switched to He
(50 mLꢁminꢀ1) for 10 min to purge the reaction system. Finally,
the sample was heated from room temperature to 500 °C with a
step of 10 °Cꢁminꢀ1 under a mixture gas flow of 5% CO/He
(20 mLꢁminꢀ1). During the measurement, the signals of He (m/
z = 4), CO2 (m/z = 44), H2O (m/z = 18), CO (m/z = 28), and H2 (m/
z = 2) were detected. The reported CO2 and H2 data were normal-
ized by dividing by the corresponding He standard signal (m/z = 4).
obtained by background subtraction and normalization procedures
and reported as ‘‘normalized absorption.” On the basis of the nor-
malized XANES profiles, the valences of copper can be acquired via
a linear combination fit by means of bulk references (Cu foil for
Cu0, Cu2O for Cu+, and CuO for Cu2+). For the extended X-ray
absorption fine structure (EXAFS), the Fourier-transformed (FT)
data in R space were analyzed using CuO and Cu2O models for
Cu–O and Cu–Cu contributions. The parameters for the description
of electronic properties (e.g., correction to the photoelectron
energy origin, E0) and local structure environment including CN,
bond distance (R), and Debye–Waller factor around the absorbing
atoms were allowed to vary in the fitting process. The fitted ranges
for k and R spaces were chosen to be k = 3–12 Åꢀ1 with R = 0.8–
3.8 Å (k3 weighted).
2.2.9. Temperature-programmed surface reaction
Temperature-programmed surface reaction (TPSR) measure-
ment of the sample was carried out on a mass spectrometer (Tilon
GRP Technology Limited LC-D200M). Fresh (as calcined in air) cat-
alysts were used for characterization. The sample (100 mg, 20–40
mesh) was pretreated for 30 min in 5% H2/Ar (30 mLꢁminꢀ1) or
20% O2/Ar (30 mLꢁminꢀ1) at 300 °C before measurement. After
being cooled to room temperature, the sample was ramped to
400 °C with a heating rate of 10 °Cꢁminꢀ1 in a continuous C3H6/
O2/He (2.5/2.5/45, 50 mLꢁminꢀ1) mixture gas flow. The signals of
He (m/z = 4), C3H5 (m/z = 41), O2 (m/z = 32), acrolein (m/z = 56),
and CO2 (m/z = 44) were tracked. The reported data of C3H6, O2,
CO2, and acrolein were normalized by dividing by the correspond-
ing He standard signal (m/z = 4).
2.2.5. X-ray diffraction
The ex situ X-ray diffraction (XRD) patterns were determined
on a Bruker D8 Advance diffractometer (40 kV, 40 mA) using CuKa1
radiation (k = 1.5406 Å) and a scanning rate of 4°/min. The scan-
ning angle was collected from 10° to 90° using a step of 0.02°. Fresh
(as calcined in air) and used (after 325 °C reaction and exposure to
air) catalysts were used for characterization. About a 50-mg sam-
ple was spread flat on a quartz sample holder for each test. A
lm-scale alumina disc was used to calibrate the 2h angles, and a
quartz sample holder was applied to hold the powder sample.
The crystallite size of Cu/Cu2O/CuO was calculated by the Scherrer
equation and the instrumental broadening was corrected by esti-
mating the (1 1 1) peak width of
lm-scale alumina. With the
2.2.10. N2O chemisorption
least-squares refinement software ‘‘LAPOD,” cell dimensions from
powder data were calculated according to Cohen’s method
[64,65] with the calibration of 2h angle via the (1 1 1) peak of
N2O chemisorption in a Builder PCSA-1000 instrument was
used to determine the Cu dispersion. The fresh (as-calcined in
air) catalysts were used for characterization. First of all, the cata-
lyst went through H2 TPR procedure as described above, and the
H2 consumption amount (X) was corresponded to the total reduc-
tion amount of CuO to Cu. Second, the sample was cooled to 35 °C
and flushed with helium (30 mLꢁminꢀ1) for 0.5 h, and then flushed
with 20% N2O/N2 gas mixture (30 mLꢁminꢀ1) for 1 h at 35 °C.
Finally, the sample was purged with He again for 0.5 h and run
the H2 TPR procedure again. The H2 consumption amount (Y)
was corresponded to the amount of surface Cu2O reduced to Cu.
The Cu dispersion (D) was derived as D = 2Y/X ꢂ 100%.
lm-scale alumina.
2.2.6. In situ X-ray diffraction
In situ X-ray diffraction (XRD) was carried out on a PANalytical
X’pert3 powder diffractometer (40 kV, 40 mA) equipped with an
Anton Paar XRK-900 reaction chamber in the 2h range 10°–90°
(scan rate 4°/min). The fresh (as calcined in air) catalysts were used
for characterization. Almost a 15-mg powder sample was loaded
into a ceramic sample holder with a diameter of 10 mm and a
depth of 1 mm. The tested powders were heated from room tem-
perature to each preset temperature (100, 200, and 300 °C) at a
constant rate of 10 °C/min in 5% H2/Ar flow (50 cm3ꢁminꢀ1) and
the corresponding XRD pattern was recorded after each tempera-
ture was maintained for 10 min. Then the sample was cooled to
room temperature under the H2/Ar flow and the XRD patterns of
different atmospheres at 30 °C were collected (5% H2/Ar ? He ?
1% C3H6/Ar + 1%O2/N2). Finally, the sample was ramped up to
300 °C for 320 min in a 1% C3H6/Ar (20 cm3ꢁminꢀ1) and 1% O2/N2
(20 cm3ꢁminꢀ1) mixture with XRD pattern collection every 40 min.
2.2.11. In situ diffuse reflectance infrared Fourier transform
spectroscopy
In situ diffuse reflectance infrared Fourier transform spec-
troscopy (DRIFTS) spectra were collected on a Bruker Vertex 70
FTIR spectrometer equipped with a mercury cadmium telluride
(MCT) detector to detect the hydroxyl part in the reduced catalysts
with an acquisition time of 40 s. The fresh (as calcined in air) cat-
alysts were used for characterization. Before the test, almost 15 mg
of the sample was pretreated in situ in the reaction cell in a5% H2/
Ar (30 mLꢁminꢀ1) gas mixture kept at 300 °C for 0.5 h and then
cooled to 30 °C in the same atmosphere, followed by pure N2 purg-
ing for 20 min.
2.2.7. Temperature-programmed reduction by hydrogen
Hydrogen temperature-programmed reduction (H2 TPR) was
applied to determine the pretreatment temperature under hydro-
gen. The measurements were carried out on a Micromeritics Auto-
chem II 2920 instrument. Fresh (as calcined in air) catalysts were
used for characterization. Prior to the measurement, the catalyst
(20–40 mesh, ~100 mg) was pretreated for 30 min in a flow of 5%
O2/He (30 mLꢁminꢀ1) at 300 °C (10 °Cꢁminꢀ1). The test was carried
out from room temperature to 600 °C (10 °Cꢁminꢀ1) at a ramp of
10 °Cꢁminꢀ1 under 5% H2/Ar (30 mLꢁminꢀ1). A thermal conductivity
detector (TCD) was used to detect the changes of hydrogen
concentration.
2.3. Catalytic tests
The catalytic performance of the catalysts (100 mg, 40–60
mesh) was determined in a fixed-bed reactor. The gas feed was
composed of C3H6 (2.5 cm3ꢁminꢀ1), O2 (2.5 cm3ꢁminꢀ1), and 10%
N2/Ar (45 cm3ꢁminꢀ1) from Jinan Deyang Corporation with a total
flow rate of 50 cm3ꢁminꢀ1. The corresponding space velocity was
30,000 cm3ꢁhꢀ1ꢁg . The catalysts were pretreated for 30 min with
ꢀ1
cat
a 75% H2/Ar (30 cm3ꢁminꢀ1) gas mixture at 300 °C. Then, once the
Please cite this article as: L.-L. Guo, J. Yu, M. Shu et al., Silicon nitride as a new support for copper catalyst to produce acrolein via selective oxidation of
propene with very low CO2 release, Journal of Catalysis, https://doi.org/10.1016/j.jcat.2019.09.019

4
L.-L. Guo et al. / Journal of Catalysis xxx (xxxx) xxx
reactor was cooled to 30 °C in the reductive atmosphere, 10% N2/Ar
was switched and purged for 10 min. Finally, the reaction gases
were fed into the reactor, which was heated to the designed tem-
peratures (275, 300, 325, and 350 °C) and the catalytic data were
collected at time on stream for 7 h. The outlet gases were detected
and analyzed online using a Shimadzu gas chromatograph (GC-
9160) every 40 min. A Porapak Q and a 5A molecular sieve-
packed column with a thermal conductivity detector (TCD) were
used to analyze CO2, C3H6, and N2. C3H6, acrolein, propene oxide,
aldehyde, propanal, and acetone were analyzed by an AE-FFAP cap-
illary column with a flame ionization detector (FID). The carbon
balance for each test was calculated to be 100 3%.
to 600–1000 °C, the experimental values gradually decreased,
which was ascribed to the partial oxidation of Si3N4 to SiO2 at
the elevated air-calcination temperatures. On the other hand, we
applied H2 TPR experiments to investigate the reducibility of these
catalysts. As shown in Fig. S1 in the Supplemental Information, a
single broad peak was detected in the region of 150–350 °C for
each sample. With the increase of air-calcination temperature,
the center of the reduction peak shifts to a higher temperature,
indicating a larger Cu(II) component formed on the surface of the
silicon nitride support or a strengthened interaction between the
metal (Cu) and support (Si3N4). Furthermore, the H2 consumption
rates in this temperature range are equal to the theoretical surface
oxygen numbers in this work, implying the complete reduction of
Cu(II) to Cu(0); see Table 1 [66].
All the calculations, including propene conversion, selectivity
(Sx) and yield of acrolein and other by-products, and acrolein for-
mation rate (racrolein), are defined for the reaction aC3H6 ? bCx as
follows,
As shown in the XRD patterns (Fig. 1a), a broad weak peak was
detected at about 23° for 10Cu/Si3N4 calcined at 800 °C, corre-
sponding to the generation of amorphous SiO2, and became stron-
ger with increased calcination temperature. However, the pattern
is complicated for 10Cu/Si3N4-1000, with other phases appearing,
nin ꢀ nout
C3H6
nin
C3H6
conversionð%Þ ¼
ꢂ 100;
C3H6
including
a-SiO2 (PDF#01-083-2466), which was caused by the
ba nx
severe decomposition of Si3N4 support. Furthermore, we analyzed
the oxidation degree of Si3N4 (oxidized to SiO2) for these samples
on the basis of the ICP-AES data (Fig. S2). Remarkably, the oxida-
tion degree of Si3N4 exceeded 60% at 1000 °C, confirming that the
structure of 10Cu/Si3N4-1000 was dramatically changed. However,
10Cu/Si3N4-800 maintained its intrinsic structure with a relatively
lower oxidation degree of 29% in experiments. For the 800 °C-
calcined 10Cu/SBA-15, the original structure of the SBA-15 support
P
Sxð%Þ ¼
ꢂ 100;
ab nx
Yieldð%Þ ¼ conversion ꢂ selectivity ꢂ 100;
conversion ꢂ Sacrolein ꢂ nCin
3H6
racrolein
¼
¼
;
mCu
was completely converted into a-SiO2 (Fig. S3), indicating its rela-
conversion ꢂ Sacrolein ꢂ nCin
3H6
racrolein
;
tively poorer thermal stability than that of Si3N4. On the other
hand, no Cu/Cu2O/CuO characteristic peaks were observed for
10Cu/Si3N4-400 in Fig. 1a, indicating the formation of amorphous
copper species. The diffraction peaks assigned to CuO (PDF#00-
045-0937) were clearly determined and become stronger, with a
calculated average grain size from 27 to 33 nm (see Table 1), when
the calcination temperature increased from 600 to 800 °C. Further-
more, as shown in Fig. 1b, the peak intensities of CuO in different
samples are almost identical when Cubulk reaches 10 wt% under
the same testing conditions (power of X-ray source, thickness of
sample loading, etc.), revealing their similar crystallinity. In addi-
tion, we applied N2O titration to calculate the Cu dispersion (D)
of the catalyst, which was directly associated with the surface-
active sites during the reaction (Fig. S4 and Table 1). The Cu disper-
sion of 10Cu/Si3N4-400 was 45.2%, much higher than that of 10Cu/
Si3N4-800 (10.8%), 20Cu/Si3N4-800 (11.9%), or 10Cu/SBA-15-800
(12.3%), further confirming the smaller size of copper species dis-
tributed on the surface of the former, while larger copper species
existed in the latter [35]. Besides the amorphous Si3N4, the
mcat
conversion ꢂ Sacrolein ꢂ nin
C3H6
Turnover frequecyðTOFÞ ¼
;
D ꢂ nCu
where nx is the partial moles of the analyzed products in this reac-
tion (aldehyde, propane, acetone, propene oxide, acrolein, and CO2).
3. Results and discussion
3.1. Structural characterization of fresh copper–silicon nitride
catalysts
Table 1 summarizes the textural properties of copper–silicon
nitride catalysts after they are calcined at different temperatures.
The bulk concentrations of copper (Cubulk) were determined by
ICP-AES and the value of 10Cu/Si3N4 was close to the designated
number when the sample was calcined at a low temperature of
400 °C. However, when the calcination temperature was increased
Table 1
Physical properties of copper–silicon nitride samples.
Sample
Cubulk (wt.%)a H2 consumption (mmolꢁgꢀ1
)
SBET (m2gꢀ1
)
dp (nm) Vp (cm3gꢀ1
)
DXRD (nm)d DTEM (nm)e Cu dispersion (%)f DXRD (nm)g
Experimentalb Theoreticalc
Si3N4
–
–
–
67
76
94
60
4
16
30
23
26
60
0.17
0.49
0.49
0.31
0.02
–
–
27
33
32
–
–
45.2
–
10.8
–
–
10Cu/Si3N4-400
10Cu/Si3N4-600
10Cu/Si3N4-800
10.0
9.7
9.3
1.56
1. 53
1.46
1.35
1.57
1.53
1.46
1.35
3.6 1.0
–
38 37
–
4.9
26
35
36
10Cu/Si3N4-1000 8.6
a
b
c
d
e
f
Determined by ICP-AES.
Integrated H2 TPR peak area from room temperature to 500 °C.
Calculated by assuming the complete reduction of CuO to Cu.
Particle size of as-calcined samples calculated from the XRD data by the Scherrer equation.
Statistic from over 100 particles in TEM images.
Determined by N2O titration.
Particle size of 325 °C used samples calculated from the XRD data by the Scherrer equation.
g
Please cite this article as: L.-L. Guo, J. Yu, M. Shu et al., Silicon nitride as a new support for copper catalyst to produce acrolein via selective oxidation of
propene with very low CO2 release, Journal of Catalysis, https://doi.org/10.1016/j.jcat.2019.09.019

L.-L. Guo et al. / Journal of Catalysis xxx (xxxx) xxx
5
Fig. 1. XRD patterns of fresh copper–silicon nitride catalysts: (a) 10 wt% Cu with different calcination temperatures; (b) calcined at 800 °C with different copper
concentrations.
crystallized Si3N4, including
gated for the present copper catalysts (10Cu/
10Cu/b-Si3N4-800). The diffraction peaks of CuO, along with the
support ( -Si3N4 or b-Si3N4), were clearly observed, giving a hint
a
-Si3N4 and b-Si3N4, were also investi-
copper species were observed (Fig. 3a–d). Fig. 3c verifies the cop-
per dispersion on the surface of Si3N4 with the help of STEM-EDS
mappings. But for 10Cu/Si3N4-800, Fig. 3e–g show unevenly large
(20 nm and above) copper species, revealing the growth of copper
during high-temperature (800 °C) air calcination. Based on the
TEM data, we can obtain the average particle size for copper
component as 38 nm with a rather wide distribution (Fig. 3h), in
accordance with the XRD results.
To investigate the electronic structure (oxidation state) and
local coordination structure (distance and coordination number)
of the copper species supported on Si3N4, we conducted XAFS mea-
surements. The XANES profiles of the as-calcined 10Cu/Si3N4-400,
10Cu/Si3N4-600, 10Cu/Si3N4-800, and 10Cu/Si3N4-1000 in Fig. 4a
show edge jump energy similar to that of the CuO reference in
the region 8980–8990 eV. This implies that the present copper spe-
cies were in the oxidized Cu2+ state for these samples. Even 10Cu/
Si3N4-400 exhibits a different edge shape. As shown in Fig. 4b, the
related EXAFS spectrum for 10Cu/Si3N4-400 displays a dominant
peak at approximately 1.95 Å derived from the first shell of CuAO,
plus a secondary CuACu shell at a distance of ca. 2.94 Å, corre-
sponding to the CuAOACu structure in the CuO model (see Table 2).
Compared with the bulk CuO reference, the lower intensity of the
CuACu shell in 10Cu/Si3N4-400 indicates its lower crystallinity. For
other samples such as 10Cu/Si3N4-600, 10Cu/Si3N4-800, and 10Cu/
Si3N4-1000, two additional CuACu characteristic peaks are present
at distances of 3.10 and 3.85 Å (Fig. 4b and Table 2), which can be
assigned to the highly crystallized CuO structure in these
a-Si3N4-800 and
a
of the excellent thermal stability of these Si3N4 materials (Fig. S5).
Nitrogen absorption measurements were carried out to detect
the textural properties of as-calcined catalysts. The adsorption/
desorption isotherms and the corresponding pore-size distribution
on typical samples (Si3N4, 10Cu/Si3N4-400, 10Cu/Si3N4-600, 10Cu/
Si3N4-800, and 10Cu/Si3N4-1000) are shown in Fig. 2a and 2b,
respectively. These curves display a typical type IV isotherm with
an H3 hysteresis loop in the P/P0 range 0.8–1.0, indicating textural
porosity. Particularly, 10Cu/Si3N4-1000 showed a very low value of
BET surface area (4 m2 gꢀ1) compared with all the other samples,
which was caused by the collapse of its structure at the elevated
calcination temperature. Table 1 shows that the average pore sizes
(dp) of 23–60 nm for copper-silicon nitride samples are obviously
larger than that of the pure Si3N4 support (16 nm). Furthermore,
the BJH pore volumes (Vp) gradually decrease with air-calcination
temperatures above 600 °C (see Table 1). These results demon-
strate that while the average pore sizes were largely maintained,
the surface areas and pore volumes of 10Cu/Si3N4 dropped contin-
uously after high-temperature air calcination.
The morphology of as-calcined copper–silicon nitride samples
was investigated by TEM. For 10Cu/Si3N4-400, the smaller (<5
nm) nanoparticles are uniformly distributed on the surface of
Si3N4 with an average size of 3.6 1.0 nm, and no aggregated
Fig. 2. Nitrogen adsorption–desorption curves of fresh copper–silicon nitride catalysts: (a) adsorption–desorption isotherm; (b) BJH pore-size distribution.
Please cite this article as: L.-L. Guo, J. Yu, M. Shu et al., Silicon nitride as a new support for copper catalyst to produce acrolein via selective oxidation of
propene with very low CO2 release, Journal of Catalysis, https://doi.org/10.1016/j.jcat.2019.09.019

6
L.-L. Guo et al. / Journal of Catalysis xxx (xxxx) xxx
Fig. 3. TEM images (a, b, e–g), the related STEM-EDS mapping results (c), and particle-size histograms for copper species (d, h) in fresh copper–silicon nitride catalysts: (a–d)
10Cu/Si3N4-400; (e–h) 10Cu/Si3N4-800.
Fig. 4. Cu K-edge XANES profiles (a) and EXAFS fitting results in R space (b) of fresh copper–silicon nitride samples.
Table 2
Cu K-edge EXAFS fitting results (R: distance; CN: coordination number;
r
2: Debye–Waller factor;a 4E0: inner potential correctionb) of as-calcined copper–silicon nitride samples.
Sample
CuAO
R (Å)
CuACu
R (Å)
CuACu
R (Å)
CuACu
R (Å)
CN
CN
CN
CN
CuO
1.948
4
2.884
4
3.115
6
3.727
2
10Cu/Si3N4-400
10Cu/Si3N6-600
10Cu/Si3N4-800
10Cu/Si3N4-1000
1.95 0.01
1.95 0.01
1.95 0.01
1.95 0.01
3.6 0.1
4.2 0.2
4.1 0.2
3.6 0.2
2.94 0.01
2.91 0.01
2.90 0.01
2.91 0.01
1.3 0.2
3.9 0.5
4.2 0.4
3.5 0.4
–
–
–
–
3.10 0.01
3.10 0.01
3.10 0.01
4.2 0.6
4.3 0.5
3.3 0.5
3.85 0.03
3.85 0.03
3.84 0.03
1.2 0.5
1.3 0.4
1.0 0.5
a
r
D
2 = 0.0045 Å2 (CuAO) or 0.007 Å2 (CuACu) for all analyzed samples.
E0 = 11.8 eV for all analyzed samples.
b
high-temperature air-calcined catalysts. These findings are in good
agreement with the XRD (Fig. 1a) and TEM (Fig. 3) results.
As discussed above, we found that the copper species can be
anchored successfully onto the silicon nitride support via a
deposition–precipitation method, and the particle size of obtained
copper catalysts can be tuned effectively by selecting the appropri-
ate air-calcination temperature. With the help of TEM, XRD,
XANES, and EXAFS analyses, we have confirmed that the amor-
phous copper oxide species were highly dispersed on the Si3N4
support for 10Cu/Si3N4-400, while the CuO crystals with larger size
Please cite this article as: L.-L. Guo, J. Yu, M. Shu et al., Silicon nitride as a new support for copper catalyst to produce acrolein via selective oxidation of
propene with very low CO2 release, Journal of Catalysis, https://doi.org/10.1016/j.jcat.2019.09.019

L.-L. Guo et al. / Journal of Catalysis xxx (xxxx) xxx
7
appeared for these copper–silicon nitride catalysts calcined at ele-
vated temperature.
experimental points from 160 to 320 min. As shown in Fig. 5,
acrolein and CO2 are the dominant products (>98%), and the other
by-products can be neglected. Fig. 5a presents the propene conver-
sion and selectivity of each product on the Cu/Si3N4 catalysts with
different Cu loading amounts (5, 10, 20, and 30 wt%) at 325 °C
(see Fig. S6). Although the acrolein selectivity is located in a very
narrow region of 82.7–87.6%, the propene conversion of
10Cu/Si3N4-800 shows a significantly higher value (24.0%) than
for all the other samples calcined at 800 °C (9.0–11.1%). On the
other hand, reaction temperature also plays an important role in
the catalytic performance of these copper–silicon nitride catalysts.
Fig. 5b and Fig. S7 show that the acrolein selectivity increased
3.2. Catalytic performance
To explore the catalytic performance of copper–silicon nitride
catalysts, the selective oxidation of propene to acrolein at 275–
350 °C has been carefully evaluated (see Table S1). All the tested
samples were pretreated at 300 °C in a reducing atmosphere
(75% H2/N2) for 30 min before the catalytic measurement. The
reported data on conversion of propene and selectivity for acrolein
or other by-products were averaged from the relatively stable
Fig. 5. Catalytic performance of copper–silicon nitride catalysts: (a) different Cu loading amounts at 325 °C, (b) different reaction temperatures for 10Cu/Si3N4-800, and (c)
different air-calcination temꢀp1eratures for 10Cu/Si3N4 at 325 °C. (d) Comparison of acrolein selectivities at similar propene conversions for different space velocities (10Cu/
Si3N4-400, 50,000 cm3ꢁhꢀ1ꢁgcat; 10Cu/Si3N4-600, 30,000 cm3ꢁhꢀ1ꢁg ; 10Cu/Si3N4-800, 35,700 cm3ꢁhꢀ1ꢁg ) at 325 °C. (e) TOF values for acrolein formation at 325 °C. (f)
ꢀ1
ꢀ1
cat
cat
Apparent activation energy (Ea). Reaction conditions: 100 mg catalyst, C3H6:O2:N2/Ar = 5:5:90, 50 mLꢁminꢀ1
.
Please cite this article as: L.-L. Guo, J. Yu, M. Shu et al., Silicon nitride as a new support for copper catalyst to produce acrolein via selective oxidation of
propene with very low CO2 release, Journal of Catalysis, https://doi.org/10.1016/j.jcat.2019.09.019

8
L.-L. Guo et al. / Journal of Catalysis xxx (xxxx) xxx
ꢀ1
rapidly from 65.2% (275 °C) to 90.6% (300 °C), and then slowly
decreased to 78.1% (350 °C). However, the propene conversion
was proportionally raised in the region 275–325 °C (3.7%?
13.0%?24.0%) and then dropped slightly to 20.6% at 350 °C. Gener-
ally speaking, the catalytic activity on the current copper–silicon
nitride catalyst was greatly modified when the Cu loadings and
reaction temperatures were changed. Among the catalysts, 10Cu/
Si3N4-800 exhibited superior catalytic activity for the reaction of
selective propene oxidation to acrolein at 325 °C.
between
5 wt%
(131.0 mmolꢁhꢀ1ꢁg
)
and
10 wt%
Cu
(136.9 mmolꢁhꢀ1ꢁg ), and then decreases severely for 20 wt%
ꢀ1
Cu
(24.0 mmolꢁhꢀ1ꢁg ) and 30 wt% (21.5 mmolꢁhꢀ1ꢁg ), confirming
ꢀ1
ꢀ1
Cu
Cu
that 10 wt% Cu is optimized in our study. To gain further insight
into the reactivity of the catalyst, the turnover frequency (TOF) of
acrolein was estimated on the basis of the copper dispersion using
N2O titration. In Fig. 5e and Table 3, it is clear that 10Cu/Si3N4-800
attained the highest TOF value of 80.6 hꢀ1, significantly higher than
those of 10Cu/Si3N4-400 (22.1 hꢀ1), 20Cu/Si3N4-800 (12.8 hꢀ1), and
10Cu/SBA-15–800 (2.4 hꢀ1). Meanwhile, the acrolein formation
rate and the TOF value increased sharply with the reaction
temperature from 275 to 325 °C, indicating that a higher reaction
temperature is more beneficial for copper–silicon nitride catalysts
air-calcined at 800 °C. When normalized by catalyst weight, the
acrolein formation rates of 10Cu/Si3N4 in Table 3 shows a severe
Furthermore, we investigated the effect of air-calcination tem-
perature (400, 600, 800, and 1000 °C) of copper–silicon nitride
samples on the catalytic activity for selective oxidation of propene
(see Fig. S8). In Fig. 5c, kept at 325 °C, the acrolein selectivity was
increased monotonically from 73.8% (10Cu/Si3N4-400) to 86.2%
(10Cu/Si3N4-800), while the related propene conversion was
decreased from 34.7% (10Cu/Si3N4-400) to 19.8% (10Cu/Si3N4-
600) and then partially recovered to 24.0% (10Cu/Si3N4-800). To
rigorously compare the acrolein selectivity on these samples, the
catalytic tests were performed at similar propene conversion
(~20%) by regulating the space velocity. As shown in Fig. 5d, the
selectivity of acrolein could be ordered as 10Cu/Si3N4-
400 < 10Cu/Si3N4-600 < 10Cu/Si3N4-800 when reacted at 325 °C,
especially for a long time period, indicating that the acrolein selec-
tivity can be optimized effectively by tuning the calcined temper-
atures of copper–silicon nitride catalysts. In particular, the 10Cu/
Si3N4-800 exhibits promoted catalytic activity for selective oxida-
tion of propene to acrolein with a superhigh selectivity of 86.2%
for acrolein (<13% for CO2) at 325 °C, which is higher than those
of all the previously reported copper-based catalysts [10,23–35].
In addition, the acrolein formation rate of 10Cu/Si3N4-800 reaches
decline (15.7 mmolꢁhꢀ1ꢁgcꢀa1t ? 1.1 mmolꢁhꢀ1ꢁg ) when the pre-
ꢀ1
cat
treated temperature increased from 400 to 1000 °C, except that
the 800 °C air-calcined catalyst exhibited an enhanced acrolein for-
mation rate of 12.7 mmolꢁhꢀ1ꢁg , which is much higher than not
ꢀ1
cat
only the other loading catalysts calcined at the same temperature,
but also the reported values in the literature [27,29]. Particularly,
for 10Cu/Si3N4-ꢀ11000, either the acrolein formation rate of
12.8 mmolꢁhꢀ1ꢁgCu or 1.1 mmolꢁhꢀ1ꢁg or the TOF of 7.5 hꢀ1 was
ꢀ1
cat
nearly an order of magnitude lower than that of 10Cu/Si3N4-800,
which was probably caused by the significant structural changes
of the Si3N4 support after air calcination at 1000 °C (Fig. 1a and
S2). In Fig. 5f, the apparent activation energies (Ea) calculated from
the Arrhenius plots at low propene conversions (<15%) were
88.8 1.3, 99.7 2.7, and 108.9 6.6 kJꢁmolꢀ1 for 10Cu/Si3N4-400,
10Cu/Si3N4-800, and 10Cu/Si3N4-1000, respectively, revealing that
the larger copper species supported on silicon nitride (10Cu/Si3N4-
800) require a slightly higher barrier energy to activate the reac-
tion than the smaller 10Cu/Si3N4-400 catalyst. The 10Cu/Si3N4-
800 and 10Cu/Si3N4-1000, which possess similar average sizes of
copper species (~33 nm), exhibited a slightly different apparent
activation energy, implying that the superior catalytic performance
lies in the maintaining of such an amorphous, partially oxidized
Si3N4 surface, which probably affected the activation of propene
or O2 in this reaction. Moreover, in order to explore the effect of
crystallinity of the support on catalytic activity, copper catalysts
136.9 mmolꢁhꢀ1ꢁg at 325 °C, which is higher than those from the
ꢀ1
Cu
very active 1 wt% Cu/SiO2 (57.1 mmolꢁhꢀ1ꢁg , 325 °C) [27] and
ꢀ1
Cu
5 wt% Cu/SiO2 (24.9 mmolꢁhꢀ1ꢁg , 325 °C) [27] at similar reaction
ꢀ1
Cu
temperatures, and even higher than the very high-temperature-
reacted CuPc/SBA-15 catalyst (90.5 mmolꢁhꢀ1ꢁg at 475 °C) [29]
ꢀ1
Cu
(Fig. S9), revealing the remarkable catalytic performance of our
10Cu/Si3N4-800 in the selective oxidation of propene to acrolein
reaction.
Table 3 summarizes the results of propene conversion and acro-
lein selectivity, as well as the formation rate of acrolein, over these
copper–silicon nitride catalysts. For comparison, pure Si3N4 was
applied to this reaction at 325 °C and did not convert any propene
(Fig. S10). The acrolein formation rates normalized by copper
weight in Table 3 show that Cu efficiency is nearly identical
on
a-Si3N4 and b-Si3N4 air-calcined at 800 °C were evaluated at
325 °C. As shown in Fig. S11, compared with the amorphous
Si3N4-supported sample (Fig. S6b), the acrolein selectivity of these
catalysts (10Cu/Si3N4-800 with amorphous Si3N4,
a-Si3N4, and
Table 3
Catalytic performance for selective oxidation of propene to acrolein over copper–silicon nitride catalysts.
Rate (mmolꢁhꢀ1ꢁg
)
Rate (mmolꢁhꢀ1ꢁg
)
TOF (hꢀ1
)
ꢀ1
ꢀ1
cat
Sample
T (°C)
Conversion (%)
Sacrolein (%)
Cu
Si3N4
325
325
325
275
300
325
350
325
325
325
325
325
325
325
475
325
325
0
0
0
0
0
10Cu/Si3N4-400
10Cu/Si3N4-600
10Cu/Si3N4-800
10Cu/Si3N4-800
10Cu/Si3N4-800
10Cu/Si3N4-800
10Cu/Si3N4-1000
5Cu/Si3N4-800
20Cu/Si3N4-800
30Cu/Si3N4-800
34.7
19.8
3.7
13.0
24.0
20.6
2.0
10.6
9.0
11.1
13.8
6.0
73.8
80.6
65.2
90.6
86.2
78.1
86.0
87.6
82.7
84.4
83.8
85.6
38.3
59.0
40.5
27.1
157.5
100.6
16.2
77.6
136.9
106.7
12.8
131.0
24.0
21.5
70.9
31.5
4.5
15.7
9.8
1.5
7.2
12.7
9.9
1.1
5.7
4.6
5.7
7.1
3.2
0.4
0.3
0.6
1.2
22.1
59.1a
9.5
45.6
80.6
62.7
7.5a
–
12.8
12.6a
41.7a
18.5a
2.4
10Cu/a-Si3N4-800
10Cu/b-Si3N4-800
10Cu/SBA-15-800
CuPc/SBA-15(Ref.29)
1 wt% Cu/SiO2(Ref.27)
5 wt% Cu/SiO2(Ref.27)
1.9
14.9
2.3
90.5
57.1
24.9
–
–
–
7.5
Note: Reaction conditions: 100 mg catalyst, C3H6:O2:N2/Ar = 5:5:90, 50 mLꢁminꢀ1
.
a
The values were estimated based on the copper dispersion of 10Cu/Si3N4-800, which exhibited similar average size, according to XRD results.
Please cite this article as: L.-L. Guo, J. Yu, M. Shu et al., Silicon nitride as a new support for copper catalyst to produce acrolein via selective oxidation of
propene with very low CO2 release, Journal of Catalysis, https://doi.org/10.1016/j.jcat.2019.09.019

L.-L. Guo et al. / Journal of Catalysis xxx (xxxx) xxx
9
b-Si3N4) was almost the same, while the propene conversion
decreased in the order 10Cu/Si3N4-800 (24.0%) > 10Cu/ -Si3N4-
3.3. Structural characterization of used copper–silicon nitride catalysts
a
800 (13.8%) > 10Cu/b-Si3N4-800 (6.0%), confirming that the amor-
phous Si3N4 support was a promising candidate in this reaction.
To highlight the superiority of the amorphous silicon nitride
support to this higher-temperature selective oxidation reaction, a
conventional SBA-15 was selected as a support for copper (10Cu/
SBA-15-800) and was prepared via a similar procedure. The stabil-
ity tests of 10Cu/SBA-15-800, 10Cu/Si3N4-800, and 10Cu/Si3N4-400
were conducted at 325 °C (see Fig. 6 and S12), and the average pro-
pene conversion for 10Cu/SBA-15-800 was just 1.6% at 325 °C in
the initial 400 min, which was much less than that for 10Cu/
Si3N4-800 (24%) and 10Cu/Si3N4-400 (35%). This can be attributed
to the structural collapse of SBA-15 after air calcination at 800 °C
(Fig. S3), while the structure of the Si3N4 support was maintained
(Fig. 1a). Moreover, 10Cu/SBA-15-800 featured a very long induc-
tion period: the selectivity for acrolein was gradually increased
from 22 to 40% in almost 400 min, while the values for 10Cu/
Si3N4-400 and 10Cu/Si3N4-800 were almost twice as high, and dis-
played a slowly rising tendency in approximately 1200 min. Nota-
bly, the propene conversion of 10Cu/Si3N4-400 was higher than
that of 10Cu/Si3N4-800 at the beginning of the reaction, but
declined linearly throughout the whole reaction process, while
the propene conversion of 10Cu/Si3N4-800 nearly leveled off after
400 min, revealing the excellent stability of the larger copper–sili-
con nitride catalyst after air calcination at 800 °C for acrolein for-
mation via selective oxidation of propene.
The XRD patterns of the used copper–silicon nitride catalysts
(Fig. 7 and S13) show obvious diffraction peaks at 36.52°, 42.42°,
61.55°, and 73.74°, which can be assigned to the (1 1 1), (2 0 0),
(2 2 0), and (3 1 1) planes of Cu2O (PDF#00-005-0667), respec-
tively. This indicates that the CuO species in fresh catalysts have
been transformed to the Cu2O component during the selective oxi-
dation of propene. Similar structural transformations were
detected in 10Cu/a-Si3N4-800, 10Cu/b-Si3N4-800, and 10Cu/SBA-
15-800 (Figs. S14 and S15). Furthermore, we have calculated the
average grain sizes of Cu2O in these samples by using the Scherrer
equation (see Table 1). Clearly, the diffraction peaks became shar-
per with increased air-calcination temperature or reaction temper-
ature, revealing the formation of Cu2O (Fig. 7 and Table 1). Unlike
the 325 °C-used 10Cu/Si3N4 samples,
a fraction of CuO was
observed in both 350 °C- and 275 °C-used 10Cu/Si3N4-800
(Fig. 7b) and 20Cu/Si3N4-800 or 30Cu/Si3N4-800 (Fig. S13), imply-
ing that part of the Cu2O species could be overoxidized to the
CuO structure under these reaction conditions. Thus, the existence
of this CuO phase could lead to a decreased propene conversion or
acrolein formation rate (see Table 3, Figs. S6 and S7).
Fig. 8 shows TEM images of used copper–silicon nitride sam-
ples. For 10Cu/Si3N4-400, the uniform copper species with an aver-
age size less than 5 nm were still highly dispersed on the surface of
the Si3N4 support (Fig. 8a–8d), while nonuniform larger copper
species were observed for 10Cu/Si3N4-800 (Fig. 8e, 8f).
Fig. 6. Stability test of (a) 10Cu/SBA-15-800 and (b) 10Cu/Si3N4-800 at 325 °C. Reaction conditions: 100 mg catalyst, C3H6:O2:N2/Ar = 5:5:90, 50 mLꢁminꢀ1
.
Fig. 7. XRD patterns of used copper–silicon nitride catalysts: (a) different 10Cu/Si3N4 samples after reaction at 325 °C; (b) 10Cu/Si3N4-800 after reaction at different
temperatures.
Please cite this article as: L.-L. Guo, J. Yu, M. Shu et al., Silicon nitride as a new support for copper catalyst to produce acrolein via selective oxidation of
propene with very low CO2 release, Journal of Catalysis, https://doi.org/10.1016/j.jcat.2019.09.019

10
L.-L. Guo et al. / Journal of Catalysis xxx (xxxx) xxx
Fig. 8. TEM images (a, b, e, f) and the related particle-size histogram for copper species (c) and STEM-EDS mapping results (d) of used copper–silicon nitride catalysts: (a–d)
10Cu/Si3N4-400; (e, f) 10Cu/Si3N4-800.
Fig. 9a displays the XANES profiles of used catalysts. With the
help of various standards (Cu foil for Cu0, Cu2O for Cu+, and CuO
for Cu2+), we have determined their oxidation states as Cu(I) after
the selective oxidation of propene. The related EXAFS fitting results
(see Fig. 9b and Table 4) show that all the investigated samples
after the 325 °C reaction contained the CuAO (R = 1.87 Å) first shell
and the CuACu (R = 3.00 Å) second shell of Cu2O.
accompanied by a very weak Cu2O (2 0 0) diffraction peak with
slowly increased diffraction intensity during the following reaction
process (320 min). This is ꢀp1robably owing to the higher space
velocity (160,000 cm3ꢁhꢀ1ꢁgcat) in the in situ XRD experiment than
that of the ex situ sample used in the catalytic reaction
(30,000 cm3ꢁhꢀ1ꢁg ). The combination of in situ (Fig. 10a) and ex
ꢀ1
cat
situ XRD data (Fig. 7a) suggested that the small copper species
Therefore, on the basis of the TEM, XRD, and XANES/EXAFS
results, we found that the Cu2O-like species were generated during
the hydrogen pretreatment followed by the selective propene oxi-
dation process over the studied copper–silicon nitride catalysts.
underwent an evolution of Cu(II) ? Cu(0) ? Cu(I), and the cuprous
oxide species were not overoxidized to copper oxide species in the
reaction atmosphere at 325 °C.
The structural transformation was more obvious on 10Cu/Si3N4-
800 than on 10Cu/Si3N4-400 under the same testing conditions. As
shown in Fig. 10b, a very sharp characteristic peak of Cu(1 1 1) was
detected at 43.0°, together with two weak diffraction peaks of Cu
(0 0 2) and Cu(0 2 2) at 50.1° and 73.6°. All the peaks of CuO van-
ished after the sample was reduced at 300 °C, indicating that the
highly crystallized copper oxide species can be reduced to metallic
copper at 300 °C. Furthermore, once the sample reacted at 325 °C
in the mixture of C3H6 and O2, the diffraction peaks of Cu2O
(1 1 1), (2 0 0), (2 2 0), and (3 1 1) appeared. At the same time,
the Cu(1 1 1) peak was weakened and soon disappeared. No other
CuO diffraction peaks were found even after 320 min of reaction,
which was consistent with the ex situ XRD results (Fig. 7a). There-
fore, the larger Cu2O species in 10Cu/Si3N4-800 is of crucial impor-
tance for the acrolein formation.
It was reported that the high-temperature treatment of Cu spe-
cies on nitrogen-containing material might result in an atomically
dispersed copper species, which has exhibited significant activity
for the oxygen reduction reaction [67]. To verify the possibility in
our system, 10Cu/Si3N4-800 was leached with 5% aq. HCl for 4 h
to remove the weakly bound CuO particles and to compare the
reactivity before and after leaching. It can be clearly seen that
3.4. Identification of active sites
To investigate the structural evolution of copper species under
the activation and reaction conditions and further determine the
active species in acrolein formation via selective oxidation of pro-
pene, we performed in situ XRD measurements on different cop-
per–silicon nitride catalysts. For 10Cu/Si3N4-400 (see Fig. 10a),
the tested sample underwent a temperature-rising reduction pro-
cess (5% H2/Ar) followed by a 320-min reaction under a propene
and oxygen mixture (molar ratio 1:1). No diffraction peaks were
observed until the reduction temperature rose to 300 °C, where
the characteristic peak of Cu(1 1 1) at 43.0° appeared. This implies
that the copper oxide species can be reduced to metallic copper at
300 °C, which was in accord with the H2 TPR results (Fig. S1 and
Table 1). Moreover, the Cu(1 1 1) peak remained unchanged when
cooled to 30 °C or under other atmospheres (He or C3H6 + O2). The
selective propene oxidation reaction was conducted after the gas
was switched to a mixture of C3H6 and O2 and warmed to 325 °C,
where the Cu(1 1 1) peak disappeared and a broader diffraction
peak at ca. 36.4° of Cu2O (1 1 1) was generated immediately,
Please cite this article as: L.-L. Guo, J. Yu, M. Shu et al., Silicon nitride as a new support for copper catalyst to produce acrolein via selective oxidation of
propene with very low CO2 release, Journal of Catalysis, https://doi.org/10.1016/j.jcat.2019.09.019

L.-L. Guo et al. / Journal of Catalysis xxx (xxxx) xxx
11
Fig. 9. Cu K-edge XANES profiles (a) and EXAFS fitting results in R space (b) of used copper–silicon nitride catalysts.
Table 4
Cu K-edge EXAFS fitting results (R: distance; CN: coordination number;
r
2: Debye–Waller factora; 4E0: inner potential correctionb) of 325 °C used catalysts.
Sample
CuAO
R (Å)
CuACu
R (Å)
CuACu
R (Å)
CN
CN
CN
Cu2O
1.849
2
3.019
12
10Cu/Si3N4-400
10Cu/Si3N6-600
10Cu/Si3N4-800
10Cu/Si3N4-1000
1.90 0.01
1.87 0.01
1.87 0.01
1.85 0.01
2.6 0.2
2.4 0.2
2.2 0.2
1.7 0.2
3.00 0.02
3.01 0.01
3.01 0.01
2.99 0.02
6.4 1.2
8.6 1.3
8.9 1.3
8.0 2.0
–
–
–
–
–
–
2.59 0.04
2.3 1.0
a
r
D
2 = 0.0046 Å2 (CuAO) or 0.0203 Å2 (CuACu) for all analyzed samples.
E0 = 10.8 eV for all analyzed samples.
b
Fig. 10. In situ XRD patterns of (a) 10Cu/Si3N4-400 and (b) 10Cu/Si3N4-800 under H2 reduction at different temperatures, which were then reacted at 325 °C for 320 min.
the large CuO species was effectively removed (Fig. S16), and the
rest of the highly dispersed copper species exhibited extremely
poor catalytic activity with ~1% propene conversion over this
reaction (Fig. S17), demonstrating that higher reactivity on
10Cu/Si3N4-800 indeed originated from larger Cu2O species.
Furthermore, commercial bulky Cu2O particles (~35 nm, 10 wt%)
were physically mixed with a silicon nitride material with a copper
dispersion of 14.6%, namely b-Cu2O/Si3N4 (Fig. S18 and S19). When
reacted at 325 °C, the sample exhibited an acrolein selectivity of
65.4% and a very low propene conversion of only 4.8% (Fig. S20),
resulting in a TOF value of 8.4 hꢀ1, almost one order of magnitude
lower than the 10Cu/Si3N4-800 catalyst reacted under the same
conditions, indicating that the silicon nitride is of crucial impor-
tance for fabricating highly active catalysts and the interaction
between the metal and support in 10Cu/Si3N4-800 greatly affected
the reactivity of the catalyst. Combining this with the in situ XRD
results (Fig. 10), we can conclude that 10Cu/Si3N4-800 with larger
Cu2O supported on the thermally stable silicon nitride is more
active than 10Cu/Si3N4-400 with smaller Cu2O for the reaction of
selective propene oxidation to acrolein at a relatively higher reac-
tion temperature of 325 °C.
Besides the active sites, the instantaneous surface reaction
remarkably influenced their catalytic behaviors. Hence, we per-
formed the temperature-programmed surface reaction (TPSR) over
copper–silicon nitride catalysts. CO2 and acrolein as two main
products of this reaction were purposely monitored. As shown in
Fig. 11a, CO2 started to react at nearly 175 °C over 10Cu/Si3N4-
400, but above 200 °C over 10Cu/Si3N4-800 (Fig. 11b), indicating
that a higher reaction temperature was required for the larger cop-
per species, which was in accordance with the apparent activation
Please cite this article as: L.-L. Guo, J. Yu, M. Shu et al., Silicon nitride as a new support for copper catalyst to produce acrolein via selective oxidation of
propene with very low CO2 release, Journal of Catalysis, https://doi.org/10.1016/j.jcat.2019.09.019

12
L.-L. Guo et al. / Journal of Catalysis xxx (xxxx) xxx
Fig. 11. Temperature-programmed surface reaction (TPSR) results for (a) 10Cu/Si3N4-400 and (b) 10Cu/Si3N4-800, which were detected by C3H5 (m/z = 41), O2(m/z = 32),
acrolein (m/z = 56), and CO2 (m/z = 44). Reaction conditions: 100 mg catalyst, C3H6:O2:N2/Ar = 5:5:90, 50 mLꢁminꢀ1
.
energy results (Fig. 5f). Furthermore, the CO2 amounts of these two
samples increased sharply with the reaction temperature until
another product (acrolein) was generated. Also, 10Cu/Si3N4-400
produced a significantly larger amount of CO2 than 10Cu/Si3N4-
800, associated with the surface structure of this catalyst. The main
products, including CO2 and acrolein, were clearly observed, and
the acrolein intensity of 10Cu/Si3N4-800 at 325 °C was greater than
that of 10Cu/Si3N4-400, which may result in completely different
selectivity of acrolein.
Furthermore, we applied CO TPR tests to detect the properties
of surface oxygen or hydroxyls over copper–silicon nitride cata-
lysts [68,69]. To mimic the structural evolution during reaction
pretreatment, hydrogen reduction at 300 °C was used to activate
the catalyst. For Si3N4 in Fig. S21a, no CO2 or H2O signals were
detected, confirming that no reaction between the hydroxyls on
the pure Si3N4 support and the CO gas happened. So the generation
of H2 and CO2 in Fig. 12 and Fig. S21b during CO TPR could be
attributed to surface hydroxyls activated by metal (Cu) to form
CuAOH species after the reduction pretreatment step via the reac-
tion of CO + 2OHꢀ ? CO2" + H2" + O2ꢀ with the stoichiometric ratio
of ~1:1 for 10Cu/Si3N4-800 and 10Cu/Si3N4-400 [68]. The amounts
of CO2 and H2 decreased with the calcination temperature and
show the order 10Cu/Si3N4-400 > 10Cu/Si3N4-600 > 10Cu/Si3N4-
800, implying that the number of CuAOH species can be greatly
tuned by selecting the appropriate calcination temperature. Mean-
while, we applied the in situ DRIFTS measurements to detect the
hydroxyls on the copper–silicon nitride catalysts after reduction
at 300 °C for 30 min. As shown in Fig. S22, the vibration peaks
corresponding to terminal OAH species and SiAO asymmetric
stretching vibrations can be observed at 3735, 1250, and
1125 cmꢀ1 for 10Cu/Si3N4-400 and 10Cu/Si3N4-800. This result
verified the formation of terminal SiAOH species, which was
favored in photocatalytic reactions [70-72]. In addition, the red
shift of hydroxyls in DRIFTS was observed for the copper–silicon
nitride catalysts, if they were compared with the pure Si3N4. This
indicates that a strong interaction between copper and silicon
nitride formed during the deposition and the following calcination
process, leading to enhanced catalytic activity for selective pro-
pene oxidation to produce acrolein [73]. On the other hand, the
CuAOH band at nearly 3538 cmꢀ1 was not observed in DRIFTS,
probably because of the limited amount of CuAOH on the surface
of 10Cu/Si3N4-400 and 10Cu/Si3N4-800 [74].
More importantly, the rates of CO2 production determined by
TPSR over both 10Cu/Si3N4-400 and 10Cu/Si3N4-800 were closely
proportional to the amounts of surface hydroxyls (Fig. S23). For
the catalytic performance of 10Cu/Si3N4-400 in Fig. S8a, it is inter-
esting that the selectivity for acrolein was increased linearly from
39 to 74% in the initial 200 min and then became stable, while the
selectivity for CO2 decreased rapidly from 58 to 23% at the same
time. This indicates that the surface CuAOH can induce the reac-
tion between C3H6 and O2 to form CO2 preferentially at the begin-
ning of the oxidation reaction, and the acrolein product increases
as CO2 decreases with the consumption of CuAOH during the
sequential period. For 10Cu/Si3N4-600, the induction time was
shortened to 150 min (Fig. S8c), while for 10Cu/Si3N4-800 with less
CuAOH in Fig. S6b, the selectivity for acrolein and CO2 was almost
Fig. 12. CO TPR profiles of (a) 10Cu/Si3N4-400 and (b) 10Cu/Si3N4-800 detected by H2 (m/z = 2) and CO2 (m/z = 44).
Please cite this article as: L.-L. Guo, J. Yu, M. Shu et al., Silicon nitride as a new support for copper catalyst to produce acrolein via selective oxidation of
propene with very low CO2 release, Journal of Catalysis, https://doi.org/10.1016/j.jcat.2019.09.019

L.-L. Guo et al. / Journal of Catalysis xxx (xxxx) xxx
13
ˇ
ˇ
´
ˇ
[10] J. Terzan, P. Djinovic, J. Zavašnik, I. Arcon, G. Zerjav, M. Spreitzer, A. Pintar, Alkali
and earth alkali modified CuOx/SiO2 catalysts for propylene partial oxidation:
what determines the selectivity, Appl. Catal. B Environ. 237 (2018) 214–227.
[11] Y.-H. Han, W. Ueda, Y. Moro-Oka, Lattice oxide ion-transfer effect
demonstrated in the selective oxidation of propene over silica-supported
bismuth molybdate catalysts, Appl. Catal. A Gen. 176 (1999) 11–16.
[12] L.J. Deiner, J.G. Serafin, C.M. Friend, S.G. Weller, J.A. Levinson, R.E. Palmer,
Insight into the partial oxidation of propene: the reactions of 2-propen-1-ol on
clean and O-covered Mo(110), J. Am. Chem. Soc. 125 (2003) 13252–13257.
[13] S.R.G. Carrazán, C. Martín, R. Mateos, V. Rives, Influence of the active phase
structure Bi-Mo-Ti-O in the selective oxidation of propene, Catal. Today 112
(2006) 121–125.
unchanged in the whole reaction process. These catalysts can be
ranked in terms of acrolein selectivity: 10Cu/Si3N4-400 < 10Cu/
Si3N4-600 < 10Cu/Si3N4-800. Therefore, we can obtain the conclu-
sion that the silicon nitride material was vital in constructing the
highly active copper-based catalyst, and the higher calcination
temperature in preparation can eliminate the surface Cu-OH spe-
cies, and thus lower the selectivity for CO2 and simultaneously
improve the acrolein selectivity.
[14] S. Pudar, J. Oxgaard, K. Chenoweth, A.C.T. van Duin, W.A. Goddard, Mechanism
of selective oxidation of propene to acrolein on bismuth molybdates from
quantum mechanical calculations, J. Phys. Chem. C 111 (2007) 16405–16415.
[15] A.B. Getsoian, V. Shapovalov, A.T. Bell, DFT+U investigation of propene
oxidation over bismuth molybdate: active sites, reaction intermediates, and
the role of bismuth, J. Phys. Chem. C 117 (2013) 7123–7137.
[16] A. Nell, A.B. Getsoian, S. Werner, L. Kiwi-Minsker, A.T. Bell, Preparation and
characterization of high-surface-area Bi(1–x)/3V1–xMoxO4 catalysts, Langmuir 30
(2014) 873–880.
[17] A.B. Getsoian, Z. Zhai, A.T. Bell, Band-gap energy as a descriptor of catalytic
activity for propene oxidation over mixed metal oxide catalysts, J. Am. Chem.
Soc. 136 (2014) 13684–13697.
[18] B. Farin, A.H.A.M. Videla, S. Specchia, E.M. Gaigneaux, Gaigneaux, Bismuth
molybdates prepared by solution combustion synthesis for the partial
oxidation of propene, Catal. Today 257 (2015) 11–17.
[19] L. Bui, R. Chakrabarti, A. Bhan, Mechanistic origins of unselective oxidation
products in the conversion of propylene to acrolein on Bi2Mo3O12, ACS Catal. 6
(2016) 6567–6580.
[20] Z. Zhai, M. Wütschert, R.B. Licht, A.T. Bell, Effects of catalyst crystal structure
on the oxidation of propene to acrolein, Catal. Today 261 (2016) 146–153.
[21] G.I. Panov, E.V. Starokon, M.V. Parfenov, B. Wei, V.I. Sobolev, L.V. Pirutko,
Quasi-catalytic identification of intermediates in the oxidation of propene to
acrolein over a multicomponent Bi-Mo catalyst, ACS Catal. 8 (2018) 1173–
1177.
[22] G.I. Panov, M.V. Parfenov, E.V. Starokon, A.S. Kharitonov, Investigation of
propene oxidation to acrolein by the method of ultralow conversion: a new
mechanism of the reaction, J. Catal. 368 (2018) 315–323.
[23] C.R. Adams, T.J. Jennings, Investigation of the mechanism of catalytic oxidation
of propylene to acrolein and acrylonitrile, J. Catal. 2 (1963) 63–68.
[24] C.R. Adams, T.J. Jennings, Mechanism studies of the catalytic oxidation of
propylene, J. Catal. 3 (1964) 549–558.
4. Conclusions
In summary, an amorphous silicon nitride was selected to
immobilize a series of copper oxides through a deposition–precipi
tation approach. The fresh catalysts calcined at different tempera-
tures exhibited completely different catalytic performance for
selective propene oxidation to acrolein. The as-synthesized 10Cu/
Si3N4-800 with good structural stability showed superior catalytic
activity with 24.0% propene conversion and 86.2% acrolein selec-
tivity at 325 °C, ꢀa1ttaining an acroleiꢀn1 formation rate value of
136.9 mmolꢁhꢀ1gCu or 12.7 mmolꢁhꢀ1gcat and a turnover frequency
(TOF) of 80.6 hꢀ1. With the aid of multiple characterization means,
we have identified that the larger Cu2O species and the interaction
between the active species and the silicon nitride with a partially
oxidized surface account for the origin of a highly effective reaction
between C3H6 and O2 to form acrolein at 325 °C. Furthermore, we
have verified that the surface CuAOH species favor the formation
of CO2 during the initial reaction period, displaying an induction
time. By optimizing the air-calcination temperature, the activity
of copper–silicon nitride catalysts can be improved through the
elimination of surface CuAOH species. The unique thermal stability
of silicon nitride is of crucial importance for the fabrication of
highly active catalysts reacted at higher temperatures.
[25] K.H. Schulz, D.F. Cox, Surface reactions of acrolein and propionaldehyde on
cuprous oxide(100): nonselective oxidation and enolate-mediated side
reactions to C3 products, J. Phys. Chem. 97 (1993) 3555–3564.
[26] J.B. Reitz, E.I. Solomon, Propylene oxidation on copper oxide surfaces:
electronic and geometric contributions to reactivity and selectivity, J. Am.
Chem. Soc. 120 (1998) 11467–11478.
[27] O.P.H. Vaughan, G. Kyriakou, N. Macleod, M. Tikhov, R.M. Lambert, Copper as a
selective catalyst for the epoxidation of propene, J. Catal. 236 (2005) 401–404.
[28] W. Su, S. Wang, P. Ying, Z. Feng, C. Li, A molecular insight into propylene
epoxidation on Cu/SiO2 catalysts using O2 as oxidant, J. Catal. 268 (2009) 165–
174.
[29] H. Tüysüz, J.L. Galilea, Highly diluted copper in a silica matrix as active catalyst
for propylene oxidation to acrolein, Catal. Lett. 131 (2009) 49–53.
[30] K.L. Bøyesen, K. Mathisen, Exposing the synergistic effect between copper and
vanadium in AlPO-5 during the selective oxidation of propene, Catal. Today
229 (2014) 14–22.
Acknowledgements
Financial support from the National Science Foundation of
China (Grant 21773288) and the National Key Basic Research Pro-
gram of China (2017YFA0403402) is acknowledged.
Appendix A. Supplementary material
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.jcat.2019.09.019.
[31] C.-H. Liu, N.-C. Lai, J.-F. Lee, C.-S. Chen, C.-M. Yang, SBA-15-supported highly
dispersed copper catalysts: vacuum-thermal preparation and catalytic studies
in propylene partial oxidation to acrolein, J. Catal. 316 (2014) 231–239.
[32] W. Song, D.M.P. Ferrandez, L. van Haandel, P. Liu, T.A. Nijhuis, E.J.M. Hensen,
Selective propylene oxidation to acrolein by gold dispersed on MgCuCr2O4
spinel, ACS Catal. 5 (2015) 1100–1111.
[33] D. Düzenli, D.O. Atmaca, M.G. Gezer, I. Onal, A density functional theory study
of partial oxidation of propylene on Cu2O(001) and CuO(001) surfaces, Appl.
Surf. Sci. 355 (2015) 660–666.
[34] K.L. Bøyesen, T. Kristiansen, K. Mathisen, Dynamic redox properties of
vanadium and copper in microporous supports during the selective
oxidation of propene, Catal. Today 254 (2015) 21–28.
[35] N.-C. Lai, M.-C. Tsai, C.-H. Liu, C.-S. Chen, C.-M. Yang, Efficient selective
oxidation of propylene by dioxygen on mesoporous-silica-nanoparticle-
supported nanosized copper, J. Catal. 365 (2018) 411–419.
[36] X. Deng, B.K. Min, X. Liu, C.M. Friend, Partial oxidation of propene on oxygen-
covered Au(111), J. Phys. Chem. B 110 (2006) 15982–15987.
[37] Y. Cui, Y. Xia, J. Zhao, L. Li, T. Fu, N. Xue, L. Peng, X. Guo, W. Ding, Super high
selectivity of acrolein in oxidation of propene on molybdenum promoted
hierarchical assembly of bismuth tungstate nanoflakes, Appl. Catal. A Gen. 482
(2014) 179–188.
[38] D.M.P. Ferrandez, I.H. Fernandez, M.P.G. Teley, M.H.J.M. de Croon, J.C.
Schouten, T.A. Nijhuis, Kinetic study of the selective oxidation of propene
with O2 over Au-Ti catalysts in the presence of water, J. Catal. 330 (2015) 396–
405.
[39] F. Liao, Y. Huang, J. Ge, W. Zheng, K. Tedsree, P. Collier, X. Hong, S.C. Tsang,
Morphology-dependent interactions of ZnO with Cu nanoparticles at the
References
[1] J.L. Callahan, R.K. Grasselli, E.C. Milberger, H.A. Strecker, Oxidation and
ammoxidation of propylene over bismuth molybdate catalyst, Ind. Eng.
Chem. Prod. Res. Dev. 9 (1970) 134–142.
[2] A.-R. Leeds, Xylidine-acrolein, J. Am. Chem. Soc. 5 (1883) 1–2.
[3] T. Kano, T. Hashimoto, K. Maruoka, Asymmetric 1,3-dipolar cycloaddition
reaction of nitrones and acrolein with a bis-titanium catalyst as chiral lewis
acid, J. Am. Chem. Soc. 127 (2005) 11926–11927.
[4] K.-H. Dostert, C.P. O’Brien, F. Ivars-Barcelo´, S. Schauermann, H.-J. Freund,
Spectators control selectivity in surface chemistry: acrolein partial
hydrogenation over Pd, J. Am. Chem. Soc. 137 (2015) 13496–13502.
[5] Z. Xi, N. Zhou, Y. Sun, K. Li, Reaction-controlled phase-transfer catalysis for
propylene epoxidation to propylene oxide, Science 292 (2001) 1139–1141.
[6] B.-K. Min, C.-M. Friend, Heterogeneous gold-based catalysis for green
chemistry: low-temperature CO oxidation and propene oxidation, Chem.
Rev. 107 (2007) 2709–2724.
[7] Y. Lei, F. Mehmood, S. Lee, J. Greeley, B. Lee, S. Seifert, R.-E. Winans, J.-W. Elam,
R.-J. Meyer, P.-C. Redfern, D. Teschner, R. Schlögl, M.-J. Pellin, L.-A. Curtiss, S.
Vajda, Increased silver activity for direct propylene epoxidation via
subnanometer size effects, Science 328 (2010) 224–228.
[8] S.J. Khatib, S.T. Oyama, Direct oxidation of propylene to propylene oxide with
molecular oxygen: a review, Catal. Rev. 57 (2015) 306–344.
[9] D.M. Driscoll, W. Tang, S.P. Burrows, D.A. Panayotov, M. Neurock, M. McEntee,
J.R. Morris, Binding sites, geometry, and energetics of propene at
nanoparticulate Au/TiO2, J. Phys. Chem. C 121 (2017) 1683–1689.
Please cite this article as: L.-L. Guo, J. Yu, M. Shu et al., Silicon nitride as a new support for copper catalyst to produce acrolein via selective oxidation of
propene with very low CO2 release, Journal of Catalysis, https://doi.org/10.1016/j.jcat.2019.09.019

14
L.-L. Guo et al. / Journal of Catalysis xxx (xxxx) xxx
materials’ interface in selective hydrogenation of CO2 to CH3OH, Angew. Chem.
Int. Ed. 50 (2011) 2162–2165.
[56] J. Lu, M. Luo, H. Lei, X. Bao, C. Li, Epoxidation of propylene on NaCl-modified
VCe1ꢀx Cux oxide catalysts with direct molecular oxygen as the oxidant, J.
Catal. 211 (2002) 552–555.
[57] X. Yang, S. Kattel, K. Xiong, K. Mudiyanselage, S. Rykov, S.D. Senanayake, J.A.
Rodriguez, P. Liu, D.J. Stacchiola, J.G. Chen, Direct epoxidation of propylene
over stabilized Cu+ surface sites on titanium-modified Cu2O, Angew. Chem. Int.
Ed. 54 (2015) 11946–11951.
[58] R. Xia, Y. Zhang, Q. Zhu, J. Qian, Q. Dong, F. Li, Surface modification of nano-
sized silicon nitride with BA-MAA-AN tercopolymer, J. Appl. Polym. Sci. 107
(2008) 562–570.
[40] F. Liao, Z. Zeng, C. Eley, Q. Lu, X. Hong, S.C.E. Tsang, Electronic modulation of a
copper/zinc oxide catalyst by a heterojunction for selective hydrogenation of
carbon dioxide to methanol, Angew. Chem. Int. Ed. 51 (2012) 5832–5836.
[41] C. Xu, G. Chen, Y. Zhao, P. Liu, X. Duan, L. Gu, G. Fu, Y. Yuan, N. Zheng,
Interfacing with silica boosts the catalysis of copper, Nat. Commun. 9 (2018)
1–10.
[42] A.M. Harzandi, J.N. Tiwari, H.S. Lee, H. Jeon, W.J. Cho, G. Lee, J. Baik, J.H. Kwak,
K.S. Kim, Efficient CO oxidation by 50-facet Cu2O nanocrystals coated with CuO
nanoparticles, ACS Appl. Mater. Interfaces 9 (2017) 2495–2499.
[43] B. Huang, H. Kobayashi, T. Yamamoto, S. Matsumura, Y. Nishida, K. Sato, K.
Nagaoka, S. Kawaguchi, Y. Kubota, Solid-solution alloying of immiscible Ru and
Cu with enhanced CO oxidation activity, J. Am. Chem. Soc. 139 (2017) 4643–
4646.
ˇ
[59] S. Lojanová, P. Tatarko, Z. Chlup, M. Hnatko, J. Dusza, Z. Lencéš, P. Šajgalík,
Rare-earth element doped SiN/SiC micro/nano-composites-RT and HT
mechanical properties, J. Eur. Ceram. Soc. 30 (2010) 1931–1944.
[60] J. Hu, L. Fang, P.-W. Zhong, A.-Q. Tang, B. Yin, Y. Li, Preparation and properties
of Ni-Co-P/nano-sized Si3N4 electroless composite coatings, Surf. Interface
Anal. 44 (2012) 450–455.
[61] J. Liu, Y. Gu, F. Li, F. Li, L. Lu, H. Zhang, S. Zhang, Catalytic nitridation
preparation of high-performance Si3N4(w)-SiC composite using Fe2O3 nano-
particle catalyst: experimental and DFT studies, J. Eur. Ceram. Soc. 37 (2017)
4467–4474.
[62] Z. Hu, W. Huo, T. Zhu, J. Liu, Z. Xie, Preparation of -Si3N4 nanorods assembled
nanobelts by crystallizing amorphous Si3N4 powders, Physica E Low-Dimens.
Syst. Nanostruct. 112 (2019) 137–141.
[63] M. Biesuz, P. Bettotti, S. Signorini, M. Bortolotti, R. Campostrini, B. Bahri, O.
Ersen, G. Speranza, A. Lale, S. Bernard, G.D. Sorarù, First synthesis of silicon
[44] J.A. Rodriguez, J. Graciani, J. Evans, J.B. Park, F. Yang, D. Stacchiola, S.D.
Senanayake, S. Ma, M. Pérez, P. Liu, J.F. Sanz, J. Hrbek, Water-gas shift reaction
on
a highly active inverse CeOx/Cu(111) catalyst: unique role of ceria
nanoparticles, Angew. Chem. Int. Ed. 48 (2009) 8047–8050.
[45] Z. Zhang, S.-S. Wang, R. Song, T. Cao, L. Luo, X. Chen, Y. Gao, J. Lu, W.-X. Li, W.
Huang, The most active Cu facet for low-temperature water gas shift reaction,
Nat. Commun. 8 (2017) 1–10.
[46] M.D. Marcinkowski, C.J. Murphy, M.L. Liriano, N.A. Wasio, F.R. Lucci, E.C.H.
Sykes, Microscopic view of the active sites for selective dehydrogenation of
formic acid on Cu(111), ACS Catal. 5 (2015) 7371–7378.
[47] M. Tamura, T. Kitanaka, Y. Nakagawa, K. Tomishige, Cu sub-nanoparticles on
Cu/CeO2 as an effective catalyst for methanol synthesis from organic carbonate
by hydrogenation, ACS Catal. 6 (2016) 376–380.
[48] H. Yang, Y. Chen, X. Cui, G. Wang, Y. Cen, T. Deng, W. Yan, J. Gao, S. Zhu, U.
Olsbye, J. Wang, W. Fan, A highly stable copper-based catalyst for clarifying the
catalytic roles of Cu0 and Cu+ species in methanol dehydrogenation, Angew.
Chem. Int. Ed. 57 (2018) 1836–1840.
[49] J. Yu, L. Kevan, Effects of reoxidation and water vapor on selective partial
oxidation of propylene to acrolein in copper(II)-exchanged X and Y zeolites, J.
Phys. Chem. 95 (1991) 6648–6653.
[50] J. He, Q. Zhai, Q. Zhang, W. Deng, Y. Wang, Active site and reaction mechanism
for the epoxidation of propylene by oxygen over CuOx/SiO2 catalysts with and
without Cs+ modification, J. Catal. 299 (2013) 53–66.
[51] M. Tonelli, M. Aouine, L. Massin, V.B. Bacab, J.M.M. Millet, Selective oxidation
of propene to acrolein on FeMoTeO catalysts: determination of active phase
and enhancement of catalytic activity and stability, Catal. Sci. Technol. 7
(2017) 4629–4639.
[52] L. Liu, X.P. Ye, J.J. Bozell, A comparative review of petroleum-based and bio-
based acrolein production, ChemSusChem 5 (2012) 1162–1180.
[53] H.-C. Wu, C.-S. Chen, C.-M. Yang, M.-C. Tsai, J.-F. Lee, Decomposition of large Cu
crystals into ultrasmall particles using chemical vapor deposition and their
application in selective propylene oxidation, ACS Appl. Mater. Interfaces 10
(2018) 38547–38557.
nanocrystals in amorphous silicon nitride from
Nanotechnology 30 (2019) 255601.
[64] J.I. Langford, Powder pattern programs, J. Appl. Crystallogr. 4 (1971) 259–260.
[65] J.I. Langford, The accuracy of cell dimensions determined by Cohen’s method
of least squares and the systematic indexing of powder data, J. Appl.
Crystallogr. 6 (1973) 190–196.
[66] Z.-Q. Wang, Z.-N. Xu, S.-Y. Peng, M.-J. Zhang, G. Lu, Q.-S. Chen, Y. Chen, G.-C.
Guo, High-performance and long-lived Cu/SiO2 nanocatalyst for CO2
hydrogenation, ACS Catal. 5 (2015) 4255–4259.
[67] F. Li, G.-F. Han, H.-J. Noh, S.-J. Kim, Y. Lu, H.Y. Jeong, Z. Fu, J.-B. Baek, Boosting
oxygen reduction catalysis with abundant copper single atom active sites,
Energy Environ. Sci. 11 (2018) 2263–2269.
[68] Y. Wang, Q. Yuan, Q. Zhang, W. Deng, Characterizations of unsupported and
supported rhodium-iron phosphate catalysts effective for oxidative
carbonylation of methane, J. Phys. Chem. C 111 (2007) 2044–2053.
[69] Q. Fu, H. Saltsburg, M. Flytzani-Stephanopoulos, Active nonmetallic Au and Pt
species on ceria-based water-gas shift catalysts, Science 301 (2003) 935–938.
[70] H. Yoshida, C. Murata, T. Hattori, Screening study of silica-supported catalysts
for photoepoxidation of propene by molecular oxygen, J. Catal. 194 (2000)
364–372.
a preceramic polymer,
[71] L. Yuliati, M. Tsubota, A. Satsuma, H. Itoh, H. Yoshida, Photoactive sites on pure
silica materials for nonoxidative direct methane coupling, J. Catal. 238 (2006)
214–220.
[54] C.L. Bracey, A.F. Carley, J.K. Edwards, P.R. Ellis, G.J. Hutchings, Understanding
the effect of thermal treatments on the structure of CuAu/SiO2 catalysts and
their performance in propene oxidation, Catal. Sci. Technol. 1 (2011) 76–85.
[55] S. Belin, C.L. Bracey, V. Briois, P.R. Ellis, G.J. Hutchings, T.I. Hyde, G. Sankar,
CuAu/SiO2 catalysts for the selective oxidation of propene to acrolein: the
impact of catalyst preparation variables on material structure and catalytic
performance, Catal. Sci. Technol. 3 (2013) 2944–2957.
[72] W.H. Saputera, H.A. Tahini, E.C. Lovell, T.H. Tan, A. Rawal, K.-F. Aguey-Zinsou,
D. Friedmann, S.C. Smith, R. Amal, J. Scott, Cooperative defect-enriched SiO2 for
oxygen activation and organic dehydrogenation, J. Catal. 376 (2019) 168–179.
[73] J. Joseph, E.D. Jemmis, Red-, blue-, or no-shift in hydrogen bonds: a unified
explanation, J. Am. Chem. Soc. 129 (2007) 4620–4632.
[74] M.D. Driessen, V.H. Grassian, Oxidation of methyl to methoxy on oxidized Cu/
SiO2, J. Phys. Chem. 99 (1995) 16519–16522.
Please cite this article as: L.-L. Guo, J. Yu, M. Shu et al., Silicon nitride as a new support for copper catalyst to produce acrolein via selective oxidation of
propene with very low CO2 release, Journal of Catalysis, https://doi.org/10.1016/j.jcat.2019.09.019