Single-Site CrOx Moieties on Silicalite: Highly Active and Stable for Ethane Dehydrogenation with CO2

Article abstract of DOI:10.1007/s10562-018-2348-x

Abstract: Silicalite-1 supported CrO_x is highly active and stable for the ethane dehydrogenation with CO₂, whose ethylene yield reached 37.8%, and only 9.5% of its initial activity was lost after 6?h, which can be attributed to the formation of single-site CrO_x moieties embedded in the framework vacancy defects of zeolite. The strong interaction with silicalite-1 surface as well as the geometry isolation effect may account for the excellent activity and stability of single-site CrO_x moieties. Graphical Abstract: [Figure not available: see fulltext.]

Full text of DOI:10.1007/s10562-018-2348-x

Catalysis Letters
https://doi.org/10.1007/s10562-018-2348-x
Single-Site CrO Moieties on Silicalite: Highly Active and Stable
x
for Ethane Dehydrogenation with CO2
1
1
2
1
1
1
Yanhu Cheng · Tianqi Lei · Changxi Miao · Weiming Hua · Yinghong Yue · Zi Gao
Received: 6 December 2017 / Accepted: 5 March 2018
©
Springer Science+Business Media, LLC, part of Springer Nature 2018
Abstract
Silicalite-1 supported CrO is highly active and stable for the ethane dehydrogenation with CO , whose ethylene yield reached
x
2
3
7.8%, and only 9.5% of its initial activity was lost after 6 h, which can be attributed to the formation of single-site CrO
x
moieties embedded in the framework vacancy defects of zeolite. The strong interaction with silicalite-1 surface as well as
the geometry isolation effect may account for the excellent activity and stability of single-site CrO moieties.
x
Graphical Abstract
40
Cr/silicalite-1
30
20
Cr/SBA-15
10
0
2
4
6
Time on stream (h)
Keywords Ethane dehydrogenation · CrO · Single-site · Silicalite-1 · CO
x
2
1
Introduction
Much effort has been dedicated to the process of catalytic
conversion of light alkanes into their corresponding alkenes
over the past several decades due to the growing demand
for light alkenes. Ethylene is an important raw material for
Electronic supplementary material The online version of this
article (https://doi.org/10.1007/s10562-018-2348-x) contains
supplementary material, which is available to authorized users.
*
Yinghong Yue
Zi Gao
yhyue@fudan.edu.cn
zigao@fudan.edu.cn
Yanhu Cheng
1
Department of Chemistry, Shanghai Key Laboratory
1
4110220008@fudan.edu.cn
of Molecular Catalysis and Innovative Materials, Fudan
University, Shanghai 200433, People’s Republic of China
Tianqi Lei
5110220048@fudan.edu.cn
1
2
Shanghai Research Institute of Petrochemical Technology,
Changxi Miao
Shanghai 201208, People’s Republic of China
miaocx.sshy@sinopec.com
Weiming Hua
wmhua@fudan.edu.cn
Vol.:(0
1
123456789)
3

Y. Cheng et al.
the production of polyethylene, polystyrene, vinyl chloride
and ethylene oxide. Dehydrogenation of ethane (DHE) is
an endothermic reaction that requires relatively high tem-
perature to obtain considerable yield of ethylene, resulting
in a low product selectivity and a fast catalyst deactivation.
Hence, the oxidative dehydrogenation of ethane by oxygen
has been proposed as an alternative to the process. However,
even though great achievements has been made [1, 2], appli-
cation of this technology is still challenged by heat diffusion
and over-oxidation.
supports, and silanol nests are more efficient and helpful
than terminal silanols in the dispersion of Cr species over
the ZSM-5 support.
In the present work, silicalite-1 was chosen as a support
for Cr O . CrO supported on silicalite-1 with abundant
2
3
x
silanol nests was prepared and characterized by XRD, N
2
adsorption, FESEM, IR, TG, H -TPR, UV–Vis and laser-
2
Raman. Its catalytic performance for dehydrogenation of
ethane to ethylene in the presence of CO was investigated
2
and compared with SBA-15 supported one under the similar
Dehydrogenation of ethane with CO to ethylene
CrO dispersion condition. The reason for its high activ-
2
x
(
CO -DHE) is regarded as a promising and advantageous
ity and stability was discussed in term of supported CrO
2
x
strategy for the production of ethylene, since higher equilib-
species.
rium conversion as well as lower coke formation can be real-
ized by the addition of CO . Meanwhile, over-oxidation of
2
ethane is greatly suppressed [3–5]. Cr-based catalysts show
excellent activity for the reaction, which are usually prepared
by dispersing Cr species on supports with high specific sur-
face areas [5, 6]. Wang et al. once studied Cr O supported
2 Experimental
2.1 Catalyst Preparation
2
3
on various supports for ethane dehydrogenation with CO
Silicalite-1 was synthesized as follow: 25 g tetraethyl-
orthosilicate (TEOS) was added dropwise to 39 g tetrapro-
pylammnoium hydroxide (TPAOH) solution, and the mix-
ture was stirred at 80 °C for 24 h. 26.8 g deionized water
was then added and the resulting mixture was stirred for
another 20 min to get a clear gel. The molar composition of
the clear gel was SiO : 0.4 TPAOH: 30 H O. This gel was
2
and the following order of activities (Cr O /SiO >Cr O /
2
3
2
2
3
ZrO > Cr O /Al O > Cr O /TiO ) was found [5]. Based
2
2
3
2
3
2
3
2
on that, mesoporous SiO materials such as SBA-1, SBA-
2
1
5, MCM-41, and MSU-x are commonly used supports for
6
+
Cr based catalysts and the amount of Cr was found to be
crucial to the catalytic activities [7–12]. However, all the
catalysts exhibited a fast deactivation and the dropping rates
increased with the increases of initial activities.
2
2
transferred to an autoclave and crystallized at 170 °C for
1 day. The obtained product was centrifuged, washed, dried
at 110 °C overnight and then calcined in air at 600 °C for 6 h
to remove the template. SBA-15 was obtained from Nanjing
Pioneer Nanomaterials Science and Technology Co., Ltd. in
Nanjing, China.
Recently ZSM-5 has also been investigated as a support
for metal oxides with regard to ethane or propane dehy-
drogenation in the presence of CO , owing to its unique
2
microporous structure, large specific surface area compared
with oxides, high thermal and hydrothermal stability. Both
textural properties and composition of ZSM-5 exerted an
important influence on the dispersion and the state of Cr
species supported on the catalysts. The amount of Al incor-
porated in the framework of HZSM-5 was reported to have
The supported chromium oxide catalysts were prepared
by impregnating an aqueous solution of Cr(NO ) ⋅9H O on
3
3
2
supports using an incipient wetness method. The impreg-
nated samples were dried at 110 °C overnight and calcined
at 650 °C for 6 h in air. The obtained catalysts are denoted
as Cr/silicalite-1 and Cr/SBA-15, respectively. The loading
amount of chromium species in both catalysts was 0.5 wt%
corresponding to Cr O .
6
+
an impact on the valence state of Cr species. Cr became
6
+
3+
the dominant state and the cycle between Cr and Cr in
the presence of CO operated well when Si/Al ratio was
2
2
3
above 1900 [13]. On the contrary, our previous work showed
Na-form of ZSM-5 favored CrO dispersion in comparison
2.2 Catalyst Characterization
x
to HZSM-5, resulting in better activity in the CO -DHE
2
reaction [14–16]. The deactivation of Cr/NaZSM-5 was
X-ray diffraction (XRD) patterns were recorded on a Per-
also much slower as compared with Cr/HZSM-5 or Cr/
see XD-2 X-ray diffractometer using nickel-filtered Cu K
α
SiO . TS-1 was also found to be a good carrier for Cr based
radiation at 40 kV and 30 mA. The BET surface areas and
2
catalysts [17]. Cr supported on TS-1 with higher Si/Ti ratio
pore volumes of the catalysts were analyzed by N adsorp-
2
exhibited higher activity in the CO -DHE reaction, quite
tion at −196 °C using a Micromeritics ASAP 2000 instru-
ment. Scanning electron microscopy (SEM) was recorded
digitally on a FEI Nova NanoSem 450 microscope oper-
ating at 40 kV. Laser Raman spectra were recorded on a
Horiba JY XPloRA spectrometer using the 532-nm radia-
tion of air-cooled solid state laser as an excitation source.
2
similar to that of Cr/HZSM-5. More recently, surface silanol
rather than surface area of SBA-15 or ZSM-5 support was
found to be crucial for high activity in dehydrogenation reac-
tion [16, 18]. Silanol groups were thought to be responsi-
ble for the interaction between chromium species and the
1
3

Single-Site CrO Moieties on Silicalite: Highly Active and Stable for Ethane…
x
The laser power was 25 mW. The spectra were obtained at
room temperature under ambient conditions with acquire
time of 30 s and accumulation number of 8. The spectral
3 Results and Discussion
3.1 Textual Characterization
−
1
resolution was 2 cm . Diffuse reflectance UV–Vis spectra
(
DRS) were collected on a Shimadzu UV-2450 spectrom-
X-Ray diffraction (XRD) patterns of the prepared catalysts
samples are shown in Fig. 1. Cr/silicalite-1 sample exhib-
its only well crystallized MFI structure, with the typical
characteristic reflections of 2θ = 8.0, 8.9, 23.1, 23.4 and
24.0. No peaks corresponding to chromium oxide are found
eter equipped with an integrating sphere attachment using
BaSO as a standard.
4
Thermogravimetric (TG) analysis of the support was
conducted on a thermogravimetric analyzer TGA800 with
o
a ramp rate of 5 C/min in the presence of purified N from
after CrO supported, indicating that chromium oxides are
2
x
2
5 to 950 °C. Before measurements, the samples were
well dispersed on silicalite-1. A good dispersion of chro-
dehydrated for 20 h and then rehydrated at ambient tem-
perature in a closed vessel under an 80% relative humidity
for 2 days to set the hydration degree. TG analysis of the
catalysts after reaction was performed in an air flow on the
same apparatus with a ramp rate of 10 °C/min to determine
the amount of coke deposited on the catalyst during the
reaction process.
mium species can also be obtained on the SBA-15 support,
as evidenced by the absence of CrO reflections (Fig. 1b).
x
SEM images of the two supports are presented in Fig. 2.
Silicalite-1 is well crystallized without any presence of
amorphous materials and exhibits a uniform crystallite size
distribution with the average crystallite size about 200 nm.
SBA-15 displays a chain-like shape. The length of each sec-
tion of the chain is around 800 nm with a diameter around
220 nm.
DRIFTS data were recorded on an FTIR (Nicolet 6700)
equipped with an MCT detector cooled by liquid N . The
2
DRIFTS cell was fitted with CaF windows and a heating
Textural properties of the samples are characterized by N
2
2
cartridge. Samples were activated in the He atmosphere at
adsorption and the results are listed in Table 1. The nitrogen
adsorption isotherm of SBA-15 (shown in supporting infor-
mation) displayed a clear H1-type hysteresis loop at high
relative pressure value, typical of mesostructured materials,
which was nearly the same as that after Cr loading (Cr/SBA-
15). Specific surface areas and pore volume of silicalite-1
4
50 °C for 4 h, and cooled to 300 °C under flowing He
(
50 ml/min), and treated in flowing N (15 ml/min) for
2
2
0 min before the scanning. Spectra were collected at 300 °C
in order to exclude the impact of water on the samples under
−
1
flowing He, with a resolution of 4 cm and accumulation
of 32 scans.
changed little after CrO supporting, while that of SBA-15
x
Temperature-programmed reduction (TPR) profiles were
obtained on a Micromeritics AutoChem II apparatus loaded
with 100 mg of catalyst. The experiments were carried out in
decreased a little, due to the final higher temperature calci-
nation procedure (650 °C). The above results indicated that
both of the support channels were not blocked by the CrO
x
1
1
0 vol% H /Ar flowing at 30 ml/min, with a ramping rate of
species, which is in accordance with XRD results. It should
be noted that the specific surface area as well as the pore
volume of silicalite-1 is much lower in comparison to that
of the mesoporous SBA-15 support.
2
0 °C/min. H consumption was monitored using a thermal
2
conductivity detector.
2.3 Catalytic Activity Tests
Catalytic tests for ethane dehydrogenation in the presence of
CO were performed at 650 °C in a fixed-bed flow microre-
2
actor at atmospheric pressure. The catalyst load was 200 mg,
and each sample was pretreated at 650 °C for 2 h under a
nitrogen flow prior to the reaction. The gas reactant con-
tained 3 vol% ethane, 15 vol% carbon dioxide and with the
balance consisting of nitrogen at a total flow rate of 30 ml/
min. The hydrocarbon reaction products were analyzed using
an on-line gas chromatograph (GC) equipped with a 6 m
Porapak Q packed column and a flame ionization detector
b
23.1
7
.9
8
.9
23.4
2
4.0
a
(
FID). The gaseous products were analyzed on-line using a
second GC equipped with a thermal conductivity detector
10
20
30
40
50
o
(
TCD) and a carbon molecular sieve 601 column. The reac-
2 Theta ( )
tion data in this work were reproducible, with a variation of
less than 5%.
Fig. 1 XRD patterns of Cr/silicalite-1 (a) and Cr/SBA-15 (b)
1
3

Y. Cheng et al.
Fig. 2 SEM images of Cr/sili-
calite-1 (a) and Cr/SBA-15 (b)
Table 1 The textural and
surface properties of the
supports
Samples
Surface area
Pore volume
Silanol amount
2
2
m /ga
3
cm /ga
3
mmol/ga
m /g
cm /g
mmol/g
SBA-15
Silicalite-1
542
330
475
325
0.86
0.27
0.75
0.26
0.25
0.21
0.17
0.11
aData referred to the support after chromium oxide supported
3.2 Surface Characterization
3738 3728
Since surface silanol groups of the supports are crucial for
the high activity of Cr based catalysts in ethane dehydro-
genation reaction [16, 18], the amount of silanols on the
support surface was determined by TG analyses. Based on
previous results that the weight loss below 100 °C attributed
to the removal of physical adsorbed water, the loss between
3510
b'
b
a'
1
00 and 200 °C assigned to the elimination of chemical
a
adsorbed water and the weight loss above 200 °C related to
the dehydroxylation by the condensation of silanols [19–21].
The amount of silanols of the two supports was calculated
and listed in Table 1. Silicalite-1 has less silanols than SBA-
4
000
3500
3000
-1
Wavenumber (cm )
1
5, due to its relatively lower surface area and crystalline
structure.
A deep understanding of hydroxyl groups can be obtained
Fig. 3 IR spectra of silicalite-1 (a), Cr/silicalite-1 (a′), SBA-15 (b)
and Cr/SBA-15 (b′)
by infrared spectroscopy. The results are displayed in Fig. 3.
−
1
A distinct peak at 3740 cm was presented in either of the
supports, which was assigned to the terminal silanol groups
well with the literature results that both types of the silanol
groups were responsible for the dispersion of chromium spe-
cies [16, 29, 30].
[
22–25]. The intensity of this peak of silicalite-1 is weaker
as compared with that of SBA-15, which is in consistent
with the TG measurements. Another broad peak at around
3.3 State of Cr Species
−
1
3
500 cm was only observed on silicalite-1, which was in
relation to silanol nests [25–28]. This means that besides the
silanol amount, silanol distribution is also different between
the two supports. Both silanol nests and terminal silanol
appeared on silicalite-1 while only terminal silanol groups
were observed on SBA-15. The intensities of both types of
silanol groups on the supports decreased after the introduc-
tion of chromium oxide, as shown in Fig. 3. This agreed
The state of Cr species can be obtained from UV–Vis dif-
fuse reflectance measurements, and the results are shown
in Fig. 4. Two peaks centered at 272 and 370 nm are
observed for both of the catalysts, which can be assigned
2
−
6+
to O →Cr charge transfer transition of monochromate
species in tetrahedral coordination [8–11, 31–33]. The peak
6+
at around 468 nm is also associated with Cr , which is more
1
3

Single-Site CrO Moieties on Silicalite: Highly Active and Stable for Ethane…
x
b'
980
6
00
807
b
550
890
1010
a'
2
72
a
3
70
a
b'
b
5
00 600 700 800
4
68
6
37
b'
b
6
25
a'
a
a'
a
600
800
1000
1200
-
1
Raman Shift (cm )
2
00
400
600
800
Wavelength (nm)
Fig. 5 Laser-Raman spectra of Cr/silicalite-1 (a), spent Cr/silicalite-1
a′), Cr/SBA-15 (b) and spent Cr/SBA-15 (b′)
(
Fig. 4 UV–Vis spectra of Cr/silicalite-1 (a), spent Cr/silicalite-1 (a′),
Cr/SBA-15 (b) and spent Cr/SBA-15 (b′)
3
74
382
3
22
obvious on the spectrum for Cr/SBA-15 than that of Cr/
silicalite-1 [8–11, 31–33]. The intensity of 468 nm peak is
believed to increase with the degree of the oligomerization
in the chromium species [8, 34]. The peak at about 625 nm
resulting from the typical octahedral symmetry transition
of tiny crystalline chromium oxides is not found on either
of the two fresh catalysts [12, 32]. A tiny peak at around
b
a
6
37 nm was found on fresh Cr/silicalite-1 and disappeared
after the reaction. It was not related to the crystalline chro-
mium oxides, which can be confirmed by the laser-Raman
tests discussed later. Considering silicalite-1 with silanol
2
00
400
600
800
Temperature (°C)
nest groups has a higher affinity for H O [35], this peak may
2
Fig. 6 H -TPR spectra of Cr/silicalite-1 (a) and Cr/SBA-15 (b)
2
result from the adsorption of two water molecules on the tet-
rahedral CrO species and transformed it into octahedral Cr
x
species, which is similar to that of Cr/BEA [36]. The above
Table 2 H -TPR results and coke deposition of the two chromium
2
oxide catalysts
results show that CrO species are well dispersed on both
x
of silicalite-1 and SBA-15 supports due to their abundant
Catalysts
H consumption
Coke amount Carbon
2
surface silanols and low CrO loadings. However, the degree
(mmol/g)
(%)
balance
%)
x
(
of oligomerization of the chromium species is different, that
is, a little higher oligomerization degree over Cr/SBA-15.
Laser-Raman spectroscopy was employed to elucidate the
molecular nature of the Cr species dispersed on the supports,
Cr/SBA-15
Cr/silicalite-1
0.093
0.094
1.3
0.6
95.0
94.3
−
1
and the results were depicted in Fig. 5. Band at 550 cm ,
assigned to crystalline Cr O , was not observed on either
its oligomerization degree of the chromium species is lower
2
3
−
1
of the spectra for the fresh catalysts, confirming a high dis-
than that of Cr/SBA-15. Another broad band at ∼890 cm
persion of Cr species on both supports. Two intense bands
was also present in the spectra of Cr/silicalite-1 catalyst.
This broad Raman feature can be assigned to the Cr mono-
oxo species embedded in the framework vacancy defect
sites, which is confirmed by DFT calculations [30]. It seems
that the structure of Cr species of silicalite-1 is also different
from that of SBA-15.
−
1
at approximately 980 and 1010 cm were present in the
spectrum of Cr/SBA-15 catalysts, which can be ascribed
to monochromate and polychromate species, respectively
−
1
[
7, 10, 32, 37–41]. The band at ∼ 600 and 807 cm was
associated with SBA-15 support [38, 39]. However, only
−
1
a weak band of 980 cm was observed for Cr/silicalite-1.
H -TPR is carried out to characterize the redox ability of
2
−
1
The absence of the band of 1010 cm indicates a lack of
polymeric chromium species on silicalite-1 support, showing
Cr species of the catalysts, the results are illustrated in Fig. 6
and Table 2. There is only one reduction peak observed on
1
3

Y. Cheng et al.
the Cr/silicalite-1 spectrum, which is attributed to the reduc-
tion of Cr(VI) species present on the catalyst. A peak at
Cr(III) ions formed from the reduction of Cr(VI) are con-
sidered to be more active than the dispersed Cr(III) ions on
the fresh catalysts [7, 12], which is usually regarded as the
main active species for the dehydrogenation reaction. This
indicates that the Cr(III) ions formed on silicalite-1 are prob-
ably more effective for dehydrogenation as compared with
those formed on SBA-15.
3
1
74 °C with a shoulder at 322 °C was found on Cr/SBA-
5, associated with the reduction of monochromates and
polychromates species, respectively [7]. The reduction
temperature for Cr/silicalite-1 is slightly higher than that
for Cr/SBA-15, indicating a higher interaction of Cr species
with silicalite-1 support. Another very weak peak at around
Turnover frequency was calculated taking the amount of
reducible Cr(VI) species as a measure of surface active sites.
TOF based on the analysis at 10 min is given in Table 3. As
6
70 °C was observed in the spectrum for Cr/silicalite-1,
which might be related to the reduction of Cr(III) to Cr(II)
[
42, 43]. H consumption of the two catalysts is almost the
expected, TOF of CrO species over Cr/silicalite-1 is higher
2
x
6
+
equal, showing the same amount of reducible surface Cr ,
which may come from the same Cr loading as well as the
good dispersion of Cr species on both catalysts.
than that of Cr/SBA-15 catalyst, confirming that CrO spe-
x
cies on silicalite-1 are more active in dehydrogenation. From
laser Raman spectra, both monochromate and polychromate
species appeared over the fresh Cr/SBA-15 catalyst, formed
by reaction with terminal silanol groups on SBA-15 surface;
whereas Cr mono-oxo species embedded in zeolitic defect
sites formed over Cr/silicalite-1 besides the monochromate
species. Considering that monochromate and polychromate
species showed no obvious different activities in the pro-
3.4 Catalytic Performance
Dehydrogenation of ethane over Cr/silicalite-1 in the pres-
ence of CO was studied, and the results are given in Table 3
2
and Fig. 7. Mass transfer and heat transfer limitations can be
excluded according to the Weisz-Prater analysis and Mears
analysis (shown in the supporting information) [44, 45]. The
major product formed in the reaction is ethylene, and the
minor product is methane. Data of carbon balance calcu-
lated (shown in Table 2) suggested that considerable use was
made of the raw materials on both catalysts. High activity
as well as high selectivity was obtained over Cr/silicalite-1
pane dehydrogenation with the presence of CO [10], Cr
2
species embedded in zeolitic defect sites seem to exhibit
higher dehydrogenation activity than those anchored on the
external surface. The difference in activities may come from
different surroundings of CrO species [30]. Silanol nests
x
were believed to possess a higher local density of hydrox-
yls than terminal silanols [26]. Thus, a stronger interaction
catalyst despite of its low CrO loading (0.5 wt%). This
between CrO species and supports would be obtained,
x
x
activity is much higher than that over Cr/SBA-15 catalyst
which may lead to higher dehydrogenation ability of CrO
x
with the same CrO loading. The result is quite surprising
species. Similar results are obtained by Dzwigaj et al. when
comparing the catalytic behaviors of two NbSiBEA zeolites
with different Nb(V) species [46]. Catalyst containing only
x
since the amount of reducible Cr(VI) is almost the same on
the two catalysts, as shown in H -TPR measurements. The
2
Table 3 Evaluation results
of the two chromium oxide
catalysts at 650 °C
Conv. (%)a
Selectivity (%)a
Yield (%)a
−1 b
Samples
TOF (h
)
CH4
C H
C H
4
2
4
2
Cr/SBA-15
Cr/Silicalite-1
24.9 (11.8)
40.2 (36.4)
7.2 (4.8)
6.2 (6.0)
92.8 (95.2)
93.8 (94.0)
23.1 (11.3)
37.8 (34.2)
30.0
47.9
a
The values outside and inside the brackets are obtained at 10 min and 6 h, respectively
b
6+
TOF was defined for the (reacted C H molecules)/(reducible Cr atoms×time)
2
6
Fig. 7 Regenerations of Cr/
silicalite-1 (a) and Cr/SBA-15
A
B
(
b) at 550 °C
3
2
1
9
6
3
2
1
1
1.2
5.9
0.6
0
5
10
15
20
25
0
5
10
15
20
25
Time on stream (h)
Time on stream (h)
1
3

Single-Site CrO Moieties on Silicalite: Highly Active and Stable for Ethane…
x
isolated framework mononuclear Nb(V) was more active
than the one containing a mixture of framework mononu-
clear and extra-framework polynuclear Nb(V) in the con-
version of ethanol/acetaldehyde mixture into 1,3-butadiene,
MPV reduction of crotonaldehyde with ethanol and etherifi-
cation of 4-methoxybenzyl alcohol with 2-butanol. Isolated
framework mononuclear Nb(V) was obtained by dispersing
the Nb precusors in vacant T-atom sites of BEA created
by concentrated acid dealumination. This two-step post-
synthesis method was applied in preparing a lot of transi-
tion metal-containing BEA zeolites, such as NiSiBEA [47],
CuSiBEA [48], FeSiBEA [49] and CrSiBEA [50], analogous
conclusion about their activities had been drawn.
to crystalline Cr O appeared in the spectrum [51–55]. The
2 3
−1
−1
intensity ratio of 980 cm peak to 1010 cm peak decreased
at the same time, confirming the aggregation of the dispersed
CrO species on Cr/SBA-15 during the reaction process. High
x
catalytic activity comes from good dispersion of CrO species,
x
the above aggregation would lower the dispersion of CrO spe-
x
cies, resulting in a reduced ethane conversion. In the contrast,
no such aggregation occurred for Cr/silicalite-1 catalyst since
neither the UV–Vis spectrum nor the laser-Raman spectrum
changed after the 6 h reaction. These findings are consistent
with the regeneration tests, since the air treatment cannot re-
disperse the aggregated species on the catalysts.
The changes of chromium species on the two catalysts
during the reaction process are quite different and this may
3.5 Catalyst Deactivation
result from the different structure of CrO species formed
x
on the two supports. Chromium species were believed to
migrate on the supports in the calcination or reaction pro-
cess and the type of the species would then change [30,
It can also be seen from Fig. 7 that ethane conversion
dropped gradually with reaction time over both of the two
catalysts. However, the decreasing rate was very slow for Cr/
silicalite-1, only 9.5% of its initial activity was lost during
the first 6 h. While for Cr/SBA-15, the dropping data was
much larger, which was up to 58% during the same period.
The conversion decrease is commonly thought to result from
coke deposition on the surface of the catalyst, as confirmed
by the TG analysis. Only 0.6 wt% of coke is deposited on
the Cr/silicalite-1 after 6 h reaction, while the deposits on
Cr/SBA-15 is much higher (1.3 wt%) after the same period
despite of its lower ethane conversion. This means that coke
is more difficult to form on Cr species embedded in defect
zeolite sites.
32, 36]. In our present work, although CrO species got a
x
good dispersion on the fresh Cr/SBA-15 catalyst owing to
the abundant surface silanol groups of SBA-15, they suf-
fered frequent redox cycles during the reaction. As a result,
chromium species were easy to mobile and encounter each
other and combined into Cr O nanoparticles, especially
2
3
under high reaction temperature (650 °C). Furthermore,
since Cr O is more acidic than chromates, carbonaceous
2
3
deposits are easier to be produced after the formation of
Cr O . Owing to a higher local hydroxyl density of silanol
2
3
nests [26], a stronger interaction between CrO species and
x
surface was obtained over Cr/silicalite-1 catalyst. The strong
interaction as well as the geometry isolation would show
a confinement effect on chromium species, inhibiting their
Attempt was also made to regenerate the above catalysts
after 6 h on stream by re-calcination of the spent catalysts
in flowing air at 550 °C for 2 h followed by purge with N
mobility. In this way, CrO species embedded in defect zeo-
2
x
for another 0.5 h. As shown in Fig. 7, the original activity
of Cr/silicalite-1 was fully restored, without any noticeable
deactivation being detected even after the second regenera-
tion. The full recovery of the activity after the air treatment
is owing to the removal of carbonaceous species deposited
on the catalyst. In contrast, the original activity cannot be
fully recovered for Cr/SBA-15. Small deactivation was
detected after each regeneration. 4.4 and 12.9% of the initial
activity was lost after the first and the second regeneration,
respectively.
lites were more stable and more difficult to contact with each
other, thus avoiding aggregating into tiny crystalline Cr O .
2
3
Coke deposition and transformation of active sites are
commonly responsible for the deactivation of dehydrogena-
tion catalysts [7, 9]. The former is a temporary deactivation
that can be recovered in the regeneration process, while the
latter is usually irreversible. Thus, a steady state of active
sites is needed in order to obtain high stability of supported
catalysts. In this work, the confinement effect of silanol nests
on zeolites is discovered, which is not only helpful in sup-
plying more active chromium species but also benefits in
maintaining their states. This will be meaningful in catalysts
designing for high-temperature reactions.
The different stability and reusability may come from
different deactivation mechanism. Figure 4 illustrated the
UV–Vis spectra of the two spent catalysts after 6 h. A new
weak peak at about 625 nm assigned to tiny crystalline chro-
mium oxides appeared in the spectrum for Cr/SBA-15 [12,
3
2], indicating that dispersed CrO species were gradually
4 Conclusion
x
transformed from monochromates and polychromates into
crystalline Cr O on SBA-15 during the reaction process. The
Cr-based catalysts were prepared using silicalite-1 and SBA-
2
3
result is more evident in the laser-Raman spectrum of spent
15 as the supports. Good dispersion of CrO species were
x
Cr/SBA-15 shown in Fig. 5. New band at 550 cm− assigned
1
observed on both of the supports because of abundant silanol
1
3

Y. Cheng et al.
groups on the surface. However, Cr/silicalite-1 exhibits
much higher activity and stability in the ethane dehydroge-
19. Hartmeyer G, Marichal C, Lebeau B, Rigolet S, Caullet P, Her-
nandez J (2007) J Phys Chem C 111:9066
2
0. Benamor T, Michelin L, Lebeau B, Marichal C (2012) Micropo-
rous Mesoporous Mater 147:334
nation in the presence of CO . 37.8% ethylene yield can be
2
obtained, which is 1.6 times as that over Cr/SBA-15. After
reaction for 6 h, 9.5% of the initial activity of Cr/silicalite-1
was lost, while for Cr/SBA-15, the data is 58%. The differ-
ence in catalytic behaviors can be attributed to the different
21. Wang LF, Yang RT (2011) J Phys Chem C 115:21264
22. Armaroli T, Simon LJ, Digne M, Montanari T, Bevilacqua M,
Valtchev V, Patarin J, Busca G (2006) Appl Catal A 306:78
2
2
3. Qin GL, Zheng L, Xie YM, Wu CC (1985) J Catal 95:609
4. Habib S, Salamé P, Launay F, Herledan VS, Marie O, Zhao W,
Tušar NN, Gédéon A (2007) J Mol Catal A 271:117
CrO species formed on the surface. Both monochromate
x
and polychromate species appeared over the Cr/SBA-15
catalyst by esterifying with terminal silanol groups on the
surface, while Cr mono-oxo species embedded in zeolite
defects formed over Cr/silicalite-1 by reaction with silanol
nests of silicalite-1. Cr mono-oxo species show higher dehy-
drogenation activity and stability against aggregation due to
25. Dessau RM, Schmitt KD, Kerr GT, Woolery GL, Alemany LB
(
1987) J Catal 104:484
2
6. Zecchina A, Bordiga S, Spoto G, Marchese L, Petrini G, Leofanti
G, Padovan M (1992) J Phys Chem 96:4985
27. Zecchina A, Bordiga S, Spoto G, Marcbese L, Petrid G, Leofanti
G, Padovan M (1992) J Phys Chem 96:4991
2
2
8. Heitmann GP, Dahlhoff G, HÖlderich WF (1999) J Catal 186:12
9. Spoto G, Bordiga S, Garrone E, Ghiotti G, Zecchina A (1992) J
Mol Catal 74:175
the stronger interaction between CrO species and silanol
x
nests. Since silanol nests can only be found in zeolite, zeolite
with abundant of silanol nests would be a perfect support for
such supported catalysts.
30. Gao J, Zheng YT, Tang YD, Jehng JM, Grybos R, Handzlik J,
Wachs IE, Podkolzin SG (2015) ACS Catal 5:3078
3
1. Ayari F, Mhamdi M, Álvarez-Rodríguez J, Guerrero Ruiz AR,
Delahay G, Ghorbel A (2013) Appl Catal B 134–135:367
3
2. Weckhuysen BM, Wachs IE, Schoonheydt RA (1996) Chem Rev
Acknowledgements This work was supported by the National Natu-
ral Science Foundation of China (91645201), the National Key R&D
Program of China (2017YFB0602204) and the Science & Technology
Commission of Shanghai Municipality (13DZ2275200).
9
6:3327
3
3
3. Zhao XH, Wang XL (2010) Catal Lett 135:233
4. Janas J, Gurgul J, Socha RP, Kowalska J, Nowinska K, Shishido
T, Che M, Dzwigaj S (2009) J Phys Chem C 113:13273
5. Zhang K, Lively RP, Noel JD, Dose ME, McCool BA, Chance
RR, Koros WJ (2012) Langmuir 28:8664
3
3
3
3
3
6. Dzwigaj S, Shishido T (2008) J Phys Chem C 112:5803
7. Hardcastle FD, Wachs IE (1988) J Mol Catal 46:173
8. Kim D, Tatibouet JM, Wachs IE (1992) J Catal 136:209
9. Kumar MS, Hammer N, Rønning M, Holmen A, Chen D, Walms-
ley JC, Øye G (2009) J Catal 261:116
References
1
2
.
.
Valente JS, Herrera HA, Solorzano RQ, Ángel P, Nava N, Masso
A, Nieto JML (2014) ACS Catal 4:1292
Zhu HB, Rosenfeld DC, Anjum DH, Sangaru SS, Saih Y, Chikh
SO, Basset JM (2015) J Catal 329:291
4
4
0. Hamdy MS, Mul G (2015) Appl Catal B 174–175:413
1. Weckhuysen BM, Bensalem A, Schoonheydt RA (1998) J Chem
Soc Faraday Trans 94:2011
3
4
.
.
Wang SB, Zhu ZH (2004) Energy Fuels 18:1126
Shishido T, Shimamura K, Teramura K, Tanaka T (2012) Catal
Today 185:151
4
4
2. Zhu ZD, Hartmann M, Maes EM, Czernuszewicz RS, Kevan L
(
2000) J Phys Chem B 104:4690
5
6
.
.
Wang SB, Murata K, Hayakawa T, Hamakawa S, Suzuki K (2000)
Appl Catal A 196:1
3. Davydov L, Reddy EP, France P, Smirniotis PG (2001) J Catal
2
03:157
Nakagawa K, Okamura M, Ikenaga N, Suzuki T, Kobayashi T
4
4
4. Mears DE (1971) Ind Eng Chem Process Des Dev 10:541
(
1998) Chem Commun 9:1025
5. Oyama ST, Zhang XM, Lu JQ, Gu YF, Fujitani T (2008) J Catal
7
8
.
.
Baek J, Yun HJ, Yun D, Choi Y, Yi J (2012) ACS Catal 2:1893
Michorczyk P, Ogonowski J, Zenczak K (2011) J Mol Catal A
2
57:1
4
4
4
4
5
6. Kyriienkoa PI, Larina OV, Popovych NO, Soloviev SO, Millot Y,
Dzwigaj S (2016) J Mol Catal A 424:27
3
49:1
9
.
Liu LC, Li HQ, Zhang Y (2006) Catal Today 115:235
7. Srebowata A, Baran R, Łomot D, Lisovytskiy D, Onfroy T,
Dzwigaj S (2014) Appl Catal B 147:208
1
0. Wang Y, Ohishi Y, Shishido T, Zhang QH, Yang W, Guo Q, Wan
HL, Takehira K (2003) J Catal 220:347
8. Baran R, Grzybek T. Onfroy T, Dzwigaj S (2016) Microporous
Mesoporous Mater 226:104
1
1. Takehira K, Ohishi Y, Shishido T, Kawabata T, Takaki K, Zhang
QH, Wang Y (2004) J Catal 224:404
9. Kocemba I, Rynkowski J, Gurgul J, Socha RP, Łatka K, Krafft JM,
Dzwigajd S (2016) Appl Catal A 519:16
1
1
2. Liu LC, Li HQ, Zhang Y (2006) J Phys Chem B 110:15478
3. Mimura N, Okamoto M, Yamashita H, Oyama ST, Murata K
0. Perez-Ramirez J, Gallardo-Llamas A (2005) Appl Catal A
(
2006) J Phys Chem B 110:21764
2
79:117
1
1
1
1
1
4. Cheng YH, Zhang F, Zhang Y, Miao CX, Hua WM, Yue YH, Gao
Z (2015) Chin J Catal 36:1242
5
5
1. Chakrabarti A, Wachs IE (2015) Catal Lett 145:985
2. Marth MS, Aun AW, Curry-Hyde HE, Baiker A (1992) J Catal
5. Zhang F, Wu RX, Yue YH, Yang WM, Gu SY, Miao CX, Hua
WM, Gao Z (2011) Microporous Mesoporous Mater 145:194
6. Cheng YH, Miao CX, Hua WM, Yue YH, Gao Z (2017) Appl
Catal A 532:111
1
33:415
5
5
5
3. Dines TJ, Inglis S (2003) Phys Chem Chem Phys 5:1320–1328
4. Lee EL, Wachs IE (2007) J Phys Chem C 111:14410
5. Moisii C, Deguns EW, Lita A, Callahan SD, van de Burgt LJ,
Magana D, Stiegman AE (2006) Chem Mater 18:3965
7. Kumar R, Mukherjee P, Pandey RK, Rajmohanan P, Bhaumik A
(
1998) Microporous Mesoporous Mater 22:23
8. Cheng YH, Zhou LB, Xu JX, Miao CX, Hua WM, Yue YH, Gao
Z (2016) Microporous Mesoporous Mater 234:370
1
3