Enhancing catalytic activity and stability for CO2 methanation on Ni-Ru/γ-Al2O3via modulating impregnation sequence and controlling surface active species

Article abstract of DOI:10.1039/c3ra47982j

Bimetallic Ni-Ru NPs loaded onto γ-Al₂O₃ were prepared by co-impregnation and sequential impregnation methods and were investigated for CO₂ methanation in the temperature range of 250-500 °C under atmospheric pressure to u

Full text of DOI:10.1039/c3ra47982j

RSC Advances
View Article Online
View Journal | View Issue
PAPER
Enhancing catalytic activity and stability for CO2
methanation on Ni–Ru/g-Al O via modulating
2
3
Cite this: RSC Adv., 2014, 4, 16472
impregnation sequence and controlling surface
active species
Wenlong Zhen, Bo Li, Gongxuan Lu* and Jiantai Maa
ab
ab
b
Bimetallic Ni–Ru NPs loaded onto g-Al
O
2 3
were prepared by co-impregnation and sequential impregnation
ꢀ
methods and were investigated for CO
2
methanation in the temperature range of 250–500 C under
atmospheric pressure to uncover the dependence of activities on surface species. It was found that the
activities of CO methanation were very highly dependent on the preparation sequence. Comparing the
activity results dependence with the characterization results of XRD, H -TPR, BET, TEM-EDX and XPS
2
2
techniques, it was indicated that the dispersion of Ru and Ni was controlled by preparation steps, and
the tendency of Ru segregation on the catalyst surface was identified. This tendency led to the high
activity of the 10Ni–1.0Ru catalyst. Such surface segregation phenomena of Ru in the co-impregnation
process controlled the chemical state of surface Ru species, allowing them to be easily reduced to the
metallic state. In addition, the 10Ni–1.0Ru catalyst performed with high stability and showed almost no
deactivation up to 100 h in long term stability tests. The possible reaction mechanism was proposed, in
Received 26th December 2013
Accepted 25th March 2014
DOI: 10.1039/c3ra47982j
2
which CO was dissociated and was activated on Ru species' surfaces to form carbon species (COads),
www.rsc.org/advances
and then was reacted with activated H on Ni centres to form methane.
Ru, Rh) on oxide supports. Among those catalysts, Ni-based
catalyst was the most extensively studied, for example, Ni/MCM-
Ni/RHA–Al Ni/ZrO –Al catalysts
tion in the atmosphere has resulted in remarkable climate exhibited high activities at low temperature. Small Ni crys-
capture and storage (CCS) is tals in highly dispersed state on the catalyst surface might play
considered as the main large scale method to solve the CO
an important role. One problem of Ni-based catalysts is its
can be converted deactivation due to the interaction of the metal particles with
into many chemicals, such as synthesized urea, salicylic acid, carbon monoxide and formation of mobile nickel sub
Introduction
It is well known that the increase of carbon dioxide concentra- 41, Ni–La/g-Al
O
,
O
,
O
2 3
2
3
2
3
2
15–21
1–4
changes.
At present, CO
2
2
5–8
problem. For its chemical utilization, CO
2
22
methanol, DME and DMC, aliphatic polycarbonate, CH , low carbonyls. Noble metals such as Ru or Rh exhibited good
4
carbon hydrocarbon. One main handicap in CO2 chemical resistant properties to sintering and carbon deposition than
2
3–25
7,10,22,26–30
conversion is its high stability and inertness properties. Ni,
Previous papers indicated that CO can be converted into CH
in fore, Ni–Ru bi-metal catalyst has attracted extensive atten-
the presence of an adequate catalyst.
and showed high activity and stability.
There-
2
4
9–13
31,32
tion.
The Ni–Ru bimetallic clusters on silica caused a strong
improvement in the activity and stability of this bimetallic
0
¼ ꢁ27 kcal molꢁ1 system. It is known that the Ru dispersion and its chemical
33
CO (g) + 4H (g) / CH (g) + 2H O(g), DG
298K
2
2
4
2
state on the support may play a key role in its catalytic perfor-
This reaction is called the Sabatier reaction, operating
34,35
mance,
which means the control of surface species in cata-
ꢀ
14
temperatures at 400 C.
lyst preparation is needed for further understanding.
A number of catalysts were reported to be active for Sabatier
reaction, such as supported VIII metals and noble metals (e.g.,
In this work, two methods, co-impregnation and sequential
impregnation, were used to prepare a series of Ni–Ru/g-Al O3
2
catalysts. By comparison the structure, surface species and
a
State Key Laboratory of Applied Organic Chemistry, College of Chemistry and
catalytic performances of different catalyst, we uncovered the
Chemical Engineering, Lanzhou University, Lanzhou, China. E-mail:
zhenwenlong123@126.com; Tel: +86 18298414700
36
surface species segregation of Ru in Ni–Ru catalysts. This
effect of segregation on catalytic properties was also discussed,
b
State Key Laboratory for Oxo Synthesis and Selective Oxidation, Lanzhou Institute of
and the possible reaction mechanism for CO
proposed.
2
methanation was
Chemical Physics, Chinese Academy of Sciences, Lanzhou 730000, P. R. China. E-mail:
gxlu@lzb.ac.cn; Tel: +86 931 4968178
16472 | RSC Adv., 2014, 4, 16472–16479
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
RSC Advances
where X is the conversion, S is the selectivity, P and (P)in are the
reactant/product molar quantities at the exit and at the entrance
of the reactor, respectively.
Experimental
Preparation of the catalysts
A series of Ni–Ru bimetallic catalysts were prepared by co-
impregnation and sequential impregnation methods using
Characterization of the catalysts
Ni(NO
RuCl $xH
precursor. The nickel loading was 10 wt% and ruthenium
loading is 0.5, 1.0, 2.5, and 5.0% on g-Al O (40 mesh, Shanghai
3
)
2
$6H
2
O (AR, Shanghai NO.2 Reagent Co. Ltd) and
3
2
O (AR, Shanghai Shanpu Chem. Corp., Ltd) as a
Temperature-programmed reduction analysis (H
-TPR) was
2
ꢀ
carried out by heating a sample (50 mg) from 25 to 700 C at
ꢀ
ꢁ
1
ꢁ1
2
3
10 C min in a ow of 5 vol% H –Ar mixture (40 mL min ).
2
Meryer Chemical Technology Co., Ltd). Before impregnation,
2
The amount of H consumed was measured by a TCD.
The X-ray diffraction patterns (XRD) of the samples were
recorded on a Rigaku B/Max-RB X-ray diffractometer with a
nickel-ltrated Cu Ka radiation over a 2q range of 15–85 and a
position sensitive detector using a step size of 0.017 and a step
ꢀ
the support g-Al
2
O
3
was stabilized in air at 600 C for 6 h. Aer
impregnation, the catalyst precursor was dried at room
ꢀ
temperature for 24 h and at 110 C for another 24 h. Finally, the
ꢀ
ꢀ
precursor was calcined at 450 C for 5 h. Here, the 10Ni–xRu/g-
Al catalysts were prepared by co-impregnation method,
ꢀ
2 3
O
time of 15 s at 40 mA and 40 kV.
which was denoted as 10Ni–xRu. In sequential impregnation
process, the catalyst that the rst step impregnated Ni and the
X-ray photoelectron spectroscopy (XPS) analysis was per-
formed using a VG Scientic ESCALAB 250Xi-XPS photoelectron
spectrometer with an Al Ka X-ray resource. The binding energies
were calibrated by the C1s binding energy of 284.7 eV.
The specic surface areas (SSA) of the catalysts were deter-

nal step impregnated Ru was denoted as 10Ni–1.0Ru (F),
otherwise, it was denoted as 10Ni (F)–1.0Ru.
Catalytic activity
mined by N adsorption–desorption measurements at 77 K by
2
3
7,38
employing the Brunauer–Emmet–Teller (BET) method (Micro-
meritics apparatus ASAP 2020M).
A reductive pre-treatment was performed before test.
100 mg
of the catalyst was loaded in each run and diluted with equal
Transmission electron microscopy (TEM) and HRTEM
images were taken with a Tecnai-G2-F30 eld emission trans-
mission electron microscope operating at accelerating voltage
of 300 kV.
amount of quartz. The calcined catalysts were pre-reduced in
2 2
situ in a 60% H /N stream for 2 h with a total gas ow of 70 mL
ꢁ
1
ꢀ
ꢀ
ꢁ1
min at 450 C with a heating ramp of 10 C min
.
Catalytic performances tests were conducted at atmospheric
pressure in a xed-bed continuous ow quartz reactor in the
ꢀ
temperature range from 250 to 500 C. A K-type thermocouple
was inserted in the furnace to measure the pre-treatment and Results and discussion
reaction temperature. The reactor was heated in a tube furnace
equipped with a temperature controller. All gases were moni-
tored by calibrated mass ow controllers.
Effect of temperature and preparation method on catalytic
activity
In order to investigate the inuence of active species and
effective active sites on catalytic performance, the 10Ni–1.0Ru,
2 2 2 2
The H and CO reactants were mixed at a H : CO ratio of
4
: 1 and N was added as a carrier gas and internal standard for
2
10Ni (F)–1.0Ru, 10Ni–1.0Ru (F) and 10Ni–1.0Ru (used, aer
gas analysis. They were introduced into the reactor at a molar
ratio of H : CO : N
¼ 4 : 1 : 4 and the total ow rate were set to
0 mL min . The gas hourly space velocity (GHSV) was varied
long-term reaction) catalysts were selected as the representa-
tives and results were exhibited in Fig. 1. Fig. 1a showed that the
2
2
2
ꢁ1
9
ꢁ
1
CO
2
conversion increased with the reaction temperature
between 3000 and 15000 h . GHSV was varied by changing the
catalyst mass while keeping the reactant ows constant. The gas
phase products were analyzed on two on-line chromatographs
equipped with thermal-conductivity detectors (TCD). Hydro-
carbons as well as oxygenated products were separated with a
GC112A (SDPTOP, Shanghai in China; column: HP-5, 30 m ꢂ
ꢀ
increase (from 250 to 400 C), and followed by the gradual
decrease (above 400 C) over all the samples. Fig. 1b showed the
ꢀ
4
selectivity of CH and by-product (CO) at the different reaction
temperatures over all the samples. It was found that CH
4
ꢀ
selectivity was 100% below 400 C. CO was produced only above
ꢀ
4
00 C. Among 10Ni–1.0Ru, 10Ni–1.0Ru (F) and 10Ni (F)–1.0Ru,
0
.53 mm ꢂ 0.5 mm) and analyzed by means of a ame ionization
ꢁ2
the 10Ni–1.0Ru catalytic activity was higher (CO
2
conversion
detector (FID). The detection limit for the products was 1 ꢂ 10
ꢀ
was 82.7%, the CH selectivity was 100%) at 400 C, suggesting
4
2 4
vol% at the given conditions. CO conversion, CH yield, CH
4
that more effective Ni and Ru species could located on the
surface of 10Ni–1.0Ru catalyst. The surface free energy of Ru
and CO selectivity were dened as follows:
ꢀ
ꢁ
ꢁ2
ꢁ2
was 3.05 J m , Ni was 2.45 J m , therefore, part of Ru atoms
CO
2
ꢂ ðN
2
Þ
Þ
in
in
X
CO2 ð%Þ ¼ 1 ꢁ N
ꢂ 100
39
had been moved onto surface layer during co-impregnation,
2
ꢂ ðCO
CH
2
which led to surface segregation. As a result, more effective Ni
and Ru species could be located on the surface of 10Ni–1.0Ru
catalyst. For 10Ni–1.0Ru (used), the catalytic activity was lower
than other samples. This was possibly due to the change of
active species on the surface of 10Ni–1.0Ru (used) during the
long-term reaction.
4
S
CH4 ð%Þ ¼ CO
ꢂ 100
ꢂ 100
2
þ CH
4
þ CO þ 2C
2 6
H
CO
S
COð%Þ ¼ CO
2
þ CH
4
þ CO þ 2C
2 6
H
This journal is © The Royal Society of Chemistry 2014
RSC Adv., 2014, 4, 16472–16479 | 16473

View Article Online
RSC Advances
Paper
ꢀ
structure (JCPDS# 87-0712). The peaks at 39.2, 42.3, and 44.0
could be assigned to the diffraction peaks of Ru with a hexag-
onal close-packed structure (JCPDS# 70-0274), while the peaks
ꢀ
at 37 and 67 were assigned to the diffraction of g-Al O phase.
2
3
ꢀ
ꢀ
The other weak diffraction peaks at 32 and 61 were also
detected, which were the characteristic peaks of NiAl
2 4
O .
In Fig. 2, the similar intensities of peaks assigned to (100),
200) and (220) facets of Ni implied that the preparation
(
methods had no signicant inuence on the distribution of Ni
ꢀ
species. However, the intensity of the peak at 44.0 assigned to
the (101) facet of 10Ni–1.0Ru (F) sample was weaker compared
with 10Ni–1.0Ru. The difference of surface Ru species among
1
0Ni–1.0Ru, 10Ni–1.0Ru (F) and 10Ni (F)–1.0Ru was resulted
from the different impregnation sequences. In addition, the
peaks of spinel (NiAl ) phase were observed, which was
originated from the calcination of the nickel species located in
2 4
O
40
the cationic decient sites of alumina. For 10Ni–1.0Ru cata-
lyst, there was no signicant difference before and aer reac-
tion, suggesting that the 10Ni–1.0Ru catalyst was stable during
the reaction process.
TEM analysis
In order to better understand the role of Ni and Ru in the
catalyst, the TEM and HRTEM images of Ni and Ru impreg-
nated g-Al O were taken (Fig. 3a–f). The particle sizes distri-
2
3
bution was showed in the Fig. 3a–c. It could be noticed that
nickel and ruthenium particles were small and uniformly
Fig. 1 Effect of temperature on the CO
and CO selectivity (b) at the methanation of CO
2
conversion (a) and the CH
over the 10Ni–1.0Ru,
4
2
1
0Ni–1.0Ru (F), 10Ni (F)–1.0Ru and 10Ni–1.0Ru (used) catalysts; dispersed in the 10Ni–1.0Ru and 10Ni–1.0Ru (F) samples.
reaction conditions: 100 mg catalysts, H : CO
2
2
¼ 4 : 1, GHSV ¼ 9000
However, the agglomerative phenomenon occurred on 10Ni (F)–
.0Ru catalyst. This reected the difference of preparation and
calcination process. The average size of Ni and Ru particles on
0Ni–1.0Ru was about 7–9 nm (Fig. 3a). The d-spacing of adja-
cent fringe for Ni nanocrystals was 0.204 nm, that could be
ꢁ1
h
, 1 atm.
1
1
XRD analysis
The crystal structures of the fresh and used catalysts were indexed to the (111) crystalline plane of face-centered cubic Ni
characterized by XRD method, the results were given in Fig. 2. lattice. The d-spacings of 0.209 nm and 0.234 nm could be
The diffraction patterns of the 10Ni–1.0Ru, 10Ni (F)–1.0Ru, indexed to as the (101) and (100) crystalline plane of hexagonal
1
0Ni–1.0Ru (F) and 10Ni–1.0Ru (used) catalysts showed similar close-packed Ru lattice, respectively. For 10Ni–1.0Ru (F) and
typical peaks. The peaks at 44.5, 51.7, and 76.3 were the 10Ni (F)–1.0Ru, the average particle sizes were 5–8 and 6–10 nm,
ꢀ
respectively (Fig. 3b–c). The d-spacings of Ni(111) and Ru(101)
were also given in Fig. 3e and f, which suggested that crystalline
features of Ni and Ru particles on 10Ni–1.0Ru (F) and 10Ni (F)–
characteristic peaks of metallic Ni with a face-centered cubic
1.0Ru were similar to that on 10Ni–1.0Ru. This result was in
accordance with those results of XRD characterization.
2
H -TPR analysis
H -TPR measurements were carried out to investigate the
2
reducibility of the 10Ni–1.0Ru, 10Ni (F)–1.0Ru and 10Ni–1.0Ru
(F) catalysts and the interaction between impregnated metal
particles and supports. Fig. 4 showed TPR proles of 10Ni–
ꢀ
1
5
.0Ru, 10Ni (F)–1.0Ru and 10Ni–1.0Ru (F) calcined at 450 C for
h. All the catalysts showed two main reduction regions at 225
and 400 C. The reduction peaks of surface Ni species and Ru
species were at 400 and 225 C, respectively. The shapes of Ru
ꢀ
ꢀ
species reduction peak of 10Ni–1.0Ru and 10Ni–1.0Ru (F)
catalysts were similar, suggesting that those Ru species were
same. The results were in good agreement with the results of
Fig. 2 XRD patterns of the 10Ni–1.0Ru, 10Ni–1.0Ru (F), 10Ni (F)–1.0Ru
and 10Ni–1.0Ru (used) catalysts.
16474 | RSC Adv., 2014, 4, 16472–16479
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
RSC Advances
Fig. 3 TEM and HRTEM images of catalysts with different preparation
methods: (a) and (d) 10Ni–1.0Ru; (b) and (e) 10Ni–1.0Ru (F); (c) and (f)
1
0Ni (F)–1.0Ru; (g) EDS spectrum of 10Ni–1.0Ru.
Fig. 5 Isotherms (a) and pore size distributions (b).
hysteresis loop. The corresponding pore size distributions were
very narrow at 10.0 nm (Fig. 5b), indicating the presence of
mesopore structure in all the catalysts.
The Brunauer–Emmett–Teller specic surface areas (SBET
were determined by N isotherms at 77 K (as shown in Table 1).
The SBET of 10Ni–1.0Ru, 10Ni–1.0Ru (used), 10Ni–1.0Ru (F),
)
2
1
1
S
2 3
0Ni (F)–1.0Ru and g-Al O were 149.5, 149.0, 148.5, 144.3 and
69.6, respectively. 10Ni–1.0Ru and 10Ni–1.0Ru (F) had larger
BET than that of 10Ni (F)–1.0Ru. Meanwhile, 10Ni–1.0Ru had
the biggest pore volume and average pore size among these
three samples, which beneted gas molecules diffusing into the
holes.
Fig. 4
H
2
-TPR profiles of the 10Ni–1.0Ru, 10Ni–1.0Ru (F), 10Ni (F)–
ꢀ
1
.0Ru and 10Ni–1.0Ru (used) catalysts calcined at 450 C 5 h.
XRD characterization. The Ru species of 10Ni (F)–1.0Ru was
different from those of 10Ni–1.0Ru and 10Ni–1.0Ru (F). The
difference in the Ru species was one of the main inuencing
factors on the different catalytic activities. The shapes or
intensities of Ni species reduction peak were very similar for the
three catalysts mentioned above, implying that the surface Ni
species of all the catalysts was almost same. Compared 10Ni–
Table 1 Structural parameters obtained from N
analysis
2
adsorption isotherms
a
b
S
BET
Pore volume
Average pore
size (nm)
2
ꢁ1
3
ꢁ1
c
Samples
g-Al
(m g
)
(cm g
)
2
O
3
169.6
149.0
149.5
144.3
148.5
0.4876
0.4049
0.4736
0.3759
0.3699
11.50
11.45
12.67
9.941
10.27
1
1
0Ni–1.0Ru (used)
0Ni–1.0Ru
1.0Ru (used) with 10Ni–1.0Ru, it was found that the Ru species
changed during the long-time reaction.
10Ni (F)–1.0Ru
10Ni–1.0Ru (F)
a
b
BET
Obtained from BET method. Total pore volume taken from the
) of 0.99.
Average pore diameter determined from the adsorption data of the
nitrogen adsorption volume at a relative pressure (P/P
0
c
As shown in Fig. 5a, the nitrogen adsorption–desorption
isotherms of all samples exhibited the IV type with H3 type isotherms using BJH method.
This journal is © The Royal Society of Chemistry 2014
RSC Adv., 2014, 4, 16472–16479 | 16475

View Article Online
RSC Advances
Paper
XPS analysis
assigned to Ni2p3/2 (Ni), Ni2p3/2 (NiAl
2
O
4
), Ni2p1/2 (Ni), and
Ni2p1/2 (NiAl O ), respectively. It implied that Ni species on
2
4
0
Fig. 6a showed typical XPS survey (wide-scan) spectra of the
2+
surface were Ni (Ni) and Ni (NiAl O ), which was strongly
2
4
10Ni–1.0Ru (fresh and used), 10Ni–1.0Ru (F), and 10Ni (F)–
veried by the results of XRD.
1.0Ru. Obviously, C, O, Al, Ni, and Ru elements with compa-
The Ru3d spectra of all the catalysts were shown in Fig. 6b–e.
The peaks at 280.1 (Ru3d5/2) and 284.2 eV (Ru3d3/2) were
observed for 10Ni–1.0Ru (fresh), 10Ni–1.0Ru (F), and 10Ni (F)–
rable peak intensities were presented on the surfaces of all the
samples, which were veried by EDS characterization (Fig. 3g).
Here, the XPS spectra of the Ni2p and Ru3d of the four catalysts
were presented in Fig. 6b–e. For the Ni2p, it was observed that
the binding energies at 852.5, 857.9, 869.9, and 876.7 eV were
4
1
1
.0Ru, suggesting that metallic Ru was present on the surface.
The peak at 281.0 eV were only present for 10Ni–1.0Ru (F) and
0Ni (F)–1.0Ru, and the intensity was weak, indicating that a
small part of Ru species was RuO for 10Ni–1.0Ru (F) and 10Ni
1
2
42
(
F)–1.0Ru. The RuO2 originated from the incompletely
reduced catalysts aer calcined at 450 C for 5 h. For the 10Ni–
ꢀ
2
1.0Ru (used) catalyst (Fig. 6e), the peaks assigned to RuO were
also observed, which might originate from the partial oxidation
of the ruthenium surface. Simultaneously, the C1s peak at 284.7
eV indicated there was contaminated carbon on the surface of
43,44
all samples.
These results veried that the active species (Ru)
had changed on the surface of 10Ni–1.0Ru (used) aer reaction.
Combined with the results of catalytic activity experiments, the
higher catalytic activity of 10Ni–1.0Ru might result from the Ru
0
species (Ru ).
The catalyst prepared by the co-impregnation method
exhibited the good catalytic performance. Therefore, The details
of Ru loading, H : CO molar ratios, GHSV, and stability were
2 2
discussed in the subsequent sections.
Effect of Ru loading on catalytic activity
The catalytic activities of various 10Ni–xRu (x ¼ 0.5, 1.0, 2.5, and
5.0) catalysts prepared by co-impregnation method, with Ru
loadings varying from 0.5 to 5.0 wt% were studied at tempera-
ꢀ
ꢁ1
tures ranging from 250 to 500 C at a GHSV of 9000 h , a
H : CO molar ratio of 4 : 1. The CO2 conversion increased
2
2

rstly with the reaction temperature increase from 250 to 400
ꢀ
ꢀ
C, then decreased (above 400 C) (Fig. 7a). This might be due to
the deactivation of the catalyst at higher temperature (above 400
ꢀ
4
C). Fig. 7b showed the selectivity of CH and CO. It was noticed
that CO was the only by-product, which was produced on the
ꢀ
surface of 10Ni–0.5Ru and 10Ni–5.0Ru catalysts above 350 C.
ꢀ
The CH selectivity was always 100% from 250 to 350 C for all
4
samples, and was less than 100% for the 10Ni–1.0Ru and 10Ni–
ꢀ
2
.5Ru catalysts above 400 C. Obviously, the conversion
ꢀ
maximum was 82.71% for the 10Ni–1.0Ru at 400 C. It was
suggesting that 10Ni–1.0Ru might provide more active sites.
Effects of H
Different H
2
: CO
: CO
2
molar ratio
2
2
molar ratios were studied at a temperature
ꢀ
ꢁ1
ranging from 250 to 500 C, GHSV was 9000 h over the 10Ni–
1
1
.0Ru catalyst. CO
.0Ru at different H
2
conversion and CH
4
selectivity over 10Ni–
2
: CO molar ratios and temperatures were
2
shown in Fig. 8.
4 2 2
CO conversion and CH selectivity increased when H : CO
2
molar ratios were increased from 2 : 1 to 6 : 1 at the same
reaction temperature, but corresponding CO selectivity
Fig. 6 All the samples XPS survey spectra (a); Ni2p and Ru3d XPS
spectra for (b) 10Ni (F)–1.0Ru, (c) 10Ni–1.0Ru (F), (d) 10Ni–1.0Ru, and
2
decreased. Fig. 8 showed that CO conversion was 94.75% and
ꢀ
(
e) 10Ni–1.0Ru (used).
CH selectivity was 100% (at 400 C, H : CO molar ratio was
4
2
2
16476 | RSC Adv., 2014, 4, 16472–16479
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
RSC Advances
Fig. 8 The CO
2
conversion (a) and the selectivity (b) over 10Ni–1.0Ru
Fig. 7 Effect of temperature and Ru loading on the CO
2
2 2
conversion (a) catalyst at different H : CO molar ratio (ranging from 2 : 1 to 6 : 1);
ꢁ1
and the selectivity (b) over 10Ni–xRu (x ¼ 0.5, 1.0, 2.5, 5.0) catalysts; reaction conditions: 100 mg catalyst, GHSV ¼ 9000 h , 1 atm.
reaction conditions: 100 mg catalysts, H : CO
2
2
¼ 4 : 1, GHSV ¼ 9000
ꢁ1
h
, 1 atm.
ꢁ
1
curves, the optimum GHSV was 9000 h , meaning the entire
catalyst works efficiently.
6
: 1). According to the results, it could be found that increasing
H : CO molar ratio was benecial to CO methanation.
2
2
2
Long-term stability tests
Fig. 10 showed the catalytic activity during 100 h on CO
2
Effects of GHSV
ꢀ
methanation over 10Ni–1.0Ru at 400 C and H
2
: CO
2
molar
ꢁ
1
To study the effect of GHSV on the activity, the GHSV was varied ratio was 4 : 1, at 9000 h . Obviously, CO conversion remained
2
ꢁ1
between 3000 and 15 000 h . Fig. 9 showed that the catalytic above 82% and CH selectivity was almost 100% during long-
4
activity of 10Ni–1.0Ru was tested at different GHSV, H
2
: CO
2
term over 10Ni–1.0Ru. The by-product (CO) was only produced
molar ratio was 4 : 1 and temperatures ranging from 250 to on the 10Ni–1.0Ru catalyst aer 80 hours. However, CO selec-
ꢀ
5
00 C.
tivity was less than 0.5%.
It was found that the temperature curves could be divided
into two domains. First, CO
between 3000 and 6000 h . As the GHSV increased, the velocity reductive in carbon dioxide methanation, the metallic nickel
for the reactants through the catalyst bed became larger; (Ni) of catalyst surface could be gradually re-oxidized to nickel
therefore, the contact time on the surface of the catalyst bed oxide (NiO) by CO2 or by water produced, deactivating the
4
There were two reasons for the decrease of CH selectivity
conversion decreased at GHSV aer 80 hours. First of all, although the atmosphere was rather
2
ꢁ1
0
became shorter. Thus, the thermodynamic equilibrium catalyst during the long-term reaction. The nickel species (Ni
conversions for the reactants were not achieved due to insuffi- or Ni ) located at the cationic decient sites of g-Al O phase
2 3
2
+
cient contact time. In addition, the number of the reactants strongly interacted with the carrier, resulting in the formation
through the catalyst bed became larger with the increase of of solid solution of nickel oxide (NiO) and nickel aluminate
ꢀ
39,45,46
GHSV; consequently, the catalyst was not able to provide suffi- spinel phase (NiAl
2
O
4
) at 400 C.
The second reason was
. Noble metals
cient active sites for redundant reactants. In addition, CO
2
that the noble metal Ru was also oxidized to RuO
conversion almost remained constant at 400, 450 and 500 C (Ru, Pt, Rh, etc.) had better resistance to carbon deposition than
2
ꢀ
ꢁ
1
23ꢁ25
when GHSV varied from 6000 to 15 000 h , and the other three Ni,
and enhanced the high activity and stability of the
conversions with low carbon
The decrease of CH selectivity was due to
ꢀ
temperature curves (250, 300 and 350 C) showed a slow upward catalysts and achieved high CO
2
7,10,26–28
trend. In addition, regarding the mentioned-temperature formation.
4
This journal is © The Royal Society of Chemistry 2014
RSC Adv., 2014, 4, 16472–16479 | 16477

View Article Online
RSC Advances
Paper
Scheme 1 The proposed a possible reaction mechanism of CO
methanation over 10Ni–1.0Ru catalyst.
2
2
Fig. 9 The CO conversion over 10Ni–1.0Ru catalyst at different
ꢁ
1
GHSV (ranging from 3000 to 15 000 h ); reaction conditions: 100 mg
catalyst, H : CO
¼ 4 : 1, 1 atm.
2
2
Conclusion
A series of bimetal Ni–Ru catalysts had been successfully
prepared by co-impregnation and sequential impregnation
methods. The obtained catalysts performed high catalytic
activities, selectivities, and excellent stabilities for CO metha-
2
nation. Besides, it was also found that the segregation
phenomenon of Ru occurred on the catalyst surface in the co-
impregnation preparation process, by which more active Ni and
Ru species (metallic Ru) could be provided on the surface of
1
0Ni–1.0Ru catalyst. The high catalytic activity was obtained at
ꢁ
1
2 2
Ru loading of 1.0%, GHSV 9000 h and higher H : CO molar
ratio. The 10Ni–1.0Ru showed more stable and highly active
properties during the long-term reaction. Based on the char-
Fig. 10 Long-term (100 h) stability tests over 10Ni–1.0Ru catalyst; acterizations, it was proposed that CO2 was dissociated and
reaction conditions: 100 mg catalyst, H
2
: CO
2
¼ 4 : 1, GHSV ¼ 9000
activated to form carbon species (COads) on Ru species surface,
then reacted with activated H on Ni species to form methane.
ꢁ1
ꢀ
h
, 1 atm, 400 C.
RuO located on the surface of the catalyst. Therefore, all the
phenomena mentioned above indicated that 10Ni–1.0Ru cata-
2
Acknowledgements
lyst performed the excellent catalytic activity and stability.
This work is supported by the 973 and 863 Programs of the
Department of Sciences and Technology of China
(
2013CB632404, 2012AA051501) and the NSF of China (grant
Possible reaction mechanism of CO
.0Ru
Scheme 1 gave a possible reaction mechanism of CO
nation on 10Ni–1.0Ru. The methanation of CO involved the
conversion of CO to CO prior to methanation, and the subse-
quent reaction followed the same mechanism as CO
2
methanation on 10Ni–
no. 21173242), respectively.
1
2
metha-
2
Notes and references
2
1 D. B. Dell, Amico, F. Calderazzo, L. Labella, F. Marchetti and
G. Pampaloni, Chem. Rev., 2003, 103, 3857.
2 M. Mikkelsen, M. Jorgensen and F. C. Krebs, Energy Environ.
Sci., 2010, 3, 43.
3 W. Wang and J. L. Gong, Front. Chem. Sci. Eng., 2011, 5, 2.
4 Z. Y. Wang, H. C. Chou, J. Wu, D. P. Tsai and G. Mul, Appl.
Catal., A, 2010, 380, 172.
5 X. D. Xu and J. A. Moulijn, Energy Fuels, 1996, 10, 305.
6 H. Yang, Z. Xu, M. Fan, R. Gupta, R. B. Slimane, A. E. Bland
and I. Wright, J. Environ. Sci., 2008, 20, 14.
47,48
methanation.
The rst step of the methanation reaction could be the
chemisorption of CO on the catalyst and the dissociation of
CO into carbon species (COads) and oxygen species (Oads) on
the surface. At the same time, H
C. Swalus et al. studied H adsorption on metallic Ni. Their work
had emerged a picture of relatively facile H dissociation on,
and strong binding of atomic H to, most transition metals.
2
2
49
2
is dissociated into H on Ni.
2
2
49,50
The second step of carbon species (COads) was dissociated into
C and O on the surface of the catalysts. C could react with H to
7 A. J. Hunt, E. H. K. Sin, R. Marriott and J. H. Clark,
ChemSusChem, 2010, 3, 306.
51
4 2
produce CH on metallic Ru, and H and O atoms formed H O.
16478 | RSC Adv., 2014, 4, 16472–16479
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
RSC Advances
8
9
G. Ferey, C. Serre, T. Devic, G. Maurin, H. Jobic, 30 M. Ku ´s mierz, Catal. Today, 2008, 137, 429.
P. L. Llewellyn, G. D. Weireld, A. Vimont, M. Daturi and 31 S. Hwang, J. Lee, U. G. Hong, J. H. Baik, D. J. Koh, H. Lim and
J. S. Chang, Chem. Soc. Rev., 2011, 40, 550.
I. K. Song, J. Ind. Eng. Chem., 2013, 19, 698.
A. Beulsa, C. Swalusa, M. Jacquemina, G. Heyenb, 32 D. J. Elliott and J. H. Lunsford, J. Catal., 1979, 57, 11.
A. Karelovica and P. Ruiza, Appl. Catal., B, 2012, 113–114, 2. 33 C. Crisafulli, S. Scir `e , S. Minic `o and L. Solarino, Appl. Catal.,
1
1
1
0 P. Sabatier and J. B. Senderens, C. R. Hebd. Seances Acad. Sci.,
A, 2002, 225, 1.
1902, 134, 514.
34 Z. Kowalczyk, K. Stołecki, W. Rar ´o g-Pilecka, E. Mi ´s kiewicz,
1 A. D. Tomsett, T. Hagiwara, A. Miyamoto and T. Inui, Appl.
Catal., 1986, 26, 391.
E. Wilczkowska and Z. Karpi n´ ski, Appl. Catal., A, 2008, 342,
35.
2 P. J. Lunde and F. L. Kester, Ind. Eng. Chem. Process Des. Dev., 35 F. Solymosi, A. Erdohelyi and T. Bansagi, J. Catal., 1981, 68,
974, 13, 27. 371.
3 K. R. Thampi, J. Kiwi and M. Graetzel, Nature, 1987, 327, 506. 36 Y. Ni, W. Peng, A. Sun, W. Mo, J. Hu, T. Li and G. Li, J. Ind.
1
1
1
4 K. P. Brooks, J. Hu, H. Zhu and R. J. Kee, Chem. Eng. Sci.,
007, 62, 1161.
5 G. D. Weatherbee and C. H. Bartholomew, J. Catal., 1981, 68,
7.
6 D. E. Peebles, J. M. Goodman and J. M. White, J. Phys. Chem.,
983, 87, 4378.
Eng. Chem., 2010, 16, 503.
37 F. Ocampo, B. Louis and A. C. Roger, Appl. Catal., A, 2009,
369, 90.
38 F. Ocampoa, B. Louisa, L. Kiwi-Minskerb and A. C. Roger,
Appl. Catal., A, 2011, 392, 36.
39 H. L. Skriver and N. M. Rosengaard, Phys. Rev. B: Condens.
Matter Mater. Phys., 1992, 46, 7157.
40 A. Corma, V. Forn ´e s, R. M. Mart ´ı n-Aranda and F. Rey, J.
Catal., 1992, 134, 58.
41 J. Okal, M. Zawadzki and W. Tylus, Appl. Catal., B, 2011, 101,
548.
42 D. D. Sarma and C. N. R. Rao, J. Electron Spectrosc. Relat.
Phenom., 1980, 20, 25.
43 Y. C. Lin, C. Y. Lin and P. W. Chiu, Appl. Phys. Lett., 2010, 96,
133110.
2
1
1
1
1
1
2
2
6
1
7 M. Yamasaki, H. Habazaki, K. Asami, K. Izumiya and
K. Hashimoto, Catal. Commun., 2006, 7, 24.
8 G. Du, S. Lim, Y. Yang, C. Wang, L. Pfefferle and G. L. Haller,
J. Catal., 2007, 249, 370.
9 F. Guo, W. Chu, H. Xu and T. Zhang, Chin. J. Catal., 2007, 28,
4
29.
0 M. Cai, J. Wen, W. Chu, X. Cheng and Z. Li, J. Nat. Gas Chem.,
011, 20, 318.
2
1 F. W. Chang, M. S. Kuo, M. T. Tsay and M. C. Hsieh, Appl.
Catal., A, 2003, 247, 309.
2 M. Agnelli, M. Kolb and C. Mirodatos, J. Catal., 1994, 148, 9.
44 X. L. Cai, X. F. Dong and W. M. Lin, J. Nat. Gas Chem., 2008,
17, 98.
2
2
3 A. T. Ashcro, A. K. Cheetham, M. L. H. Green and 45 E. D. Dimotakis and T. J. Pinnavaia, Inorg. Chem., 1990, 29,
P. D. F. Vernon, Nature, 1991, 352, 225. 2393.
4 P. D. F. Vernon, M. L. H. Green, A. K. Cheetham and 46 J. A. Wang, X. Bokhimi, A. Morales and O. Novaro, J. Phys.
A. T. Ashcro, Catal. Today, 1992, 13, 417. Chem. B, 1999, 103, 299.
5 J. R. Rostrup-Nielsen and J. H. B. Hansen, J. Catal., 1993, 144, 47 J. L. Falconer and A. E. Za ˘g li, J. Catal., 1980, 62, 280.
2
2
3
8.
48 M. Marwood, R. Doepper and A. Renken, Appl. Catal., A,
1997, 151, 223.
2
2
6 J. R. Rostrup-Nielsen, Catal. Rev.: Sci. Eng., 2004, 46, 247.
7 E. Akpana, Y. Sun, P. Kumar, H. Ibrahim, A. Aboudheir and 49 C. Swalus, M. Jacquemin, C. Poleunis, P. Bertrand and
R. Idem, Chem. Eng. Sci., 2007, 62, 4012.
8 D. L. Trimm, Catal. Today, 1999, 49, 3.
9 A. L. Kustov, A. M. Frey, K. E. Larsen, T. Johannessen,
P. Ruiz, Appl. Catal., B, 2012, 125, 41.
50 K. D. Rendulic, A. Winkler and H. P. Steinr u¨ ck, Surf. Sci.,
1987, 185, 469.
2
2
J. K. Nørskov and C. H. Christensen, Appl. Catal., A, 2007, 51 P. D. Reed, C. M. Comrie and R. M. Lambert, Surf. Sci., 1976,
20, 98. 59, 33.
3
This journal is © The Royal Society of Chemistry 2014
RSC Adv., 2014, 4, 16472–16479 | 16479