Size-Tunable Ni Nanoparticles Supported on Surface-Modified, Cage-Type Mesoporous Silica as Highly Active Catalysts for CO2 Hydrogenation

Article abstract of DOI:10.1021/acscatal.7b02310

In this study, ultrasmall Ni nanoparticles (Ni NPs) were controllably supported in the cage-type mesopores of -COOH-functionalized mesoporous silica SBA-16 (denoted as Ni(x)@S16C, where x is the Ni loading) via wet impregnation under alkaline conditions,

Full text of DOI:10.1021/acscatal.7b02310

Subscriber access provided by READING UNIV
Article
Size-tunable Ni nanoparticles supported on surface-modified, cage-
type mesoporous silica as highly active catalysts for CO2 hydrogenation
Ching-Shiun Chen, Canggih Setya Budi, Hung-Chi Wu, Diganta Saikia, and Hsien-Ming Kao
ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.7b02310 • Publication Date (Web): 02 Nov 2017
Downloaded from http://pubs.acs.org on November 4, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
ACS Catalysis is published by the American Chemical Society. 1155 Sixteenth Street
N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 47
ACS Catalysis
1
2
3
4
5
6
7
8
9
Sizeꢀtunable Ni nanoparticles supported on surfaceꢀ
modified, cageꢀtype mesoporous silica as highly
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
active catalysts for CO hydrogenation
2
a,b*
c
a
c
Ching-Shiun Chen, Canggih Setya Budi, Hung-Chi Wu, Diganta Saikia, and Hsien-Ming
c*
Kao
a
Center for General Education, Chang Gung University, 259, WenꢀHua 1st Rd., Guishan Dist.,
Taoyuan City 33302, Taiwan, Republic of China.
b
Department of Pathology, Chang Gung Memorial Hospital, 5, Fusing St., Guishan Dist.,
Taoyuan City 33302, Taiwan, Republic of China.
c
Department of Chemistry, National Central University, ChungꢀLi, 32001, Taiwan, Republic of
China.
*Corresponding authors
KEYWORDS mesoporous silicas; SBAꢀ16; nickel nanoparticles; CO hydrogenation
2
ACS Paragon Plus Environment
1

ACS Catalysis
Page 2 of 47
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
ABSTRACT
In this study, ultraꢀsmall Ni nanoparticles (Ni NPs) were controllably supported in the cageꢀtype
mesopores of –COOH functionalized mesoporous silica SBAꢀ16 (denoted as Ni(x)@S16C,
where x is the Ni loading) via wet impregnation under alkaline conditions, followed by thermal
reduction. The particle sizes of the Ni NPs ranged from 2.7 to 4.7 nm depend on the Ni loading.
Under the appropriate alkaline conditions, i.e., pH 9, deprotonation of the carboxylic acid groups
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
+
on the cageꢀtype mesopore surfaces endowed the effective incorporation of Ni precursors via
favorable electrostatic interactions, and thus wellꢀdispersed Ni NPs confined in the cageꢀtype
mesopores of SBAꢀ16 were achieved. The combination of the cageꢀtype mesopores and the
surface functionality provided dual beneficial features to confine the immobilized Ni NPs and to
tune their particle sizes. The remarkably enhanced catalytic activities of the Ni(x)@S16C
materials for CO hydrogenation and CH formation were demonstrated. The cageꢀtype SBAꢀ16
2
4
support provided a positive effect for the Ni NPs to enrich the surface sites that can strongly
adsorb CO and CO , thus leading to high catalytic rates for CO and CO hydrogenation. The
2
2
reaction mechanism, catalytic kinetics and active sites were investigated to correlate to the high
reaction rate for CO hydrogenation to form CH .
2
4
Keywords: Carboxylic acid functionalized mesoporous silica SBAꢀ16, cageꢀtype mesopore, CO2
hydrogenation, Ni particles, Carbon monoxide, Methane
ACS Paragon Plus Environment
2

Page 3 of 47
ACS Catalysis
1
2
3
4
5
6
7
8
9
1
. Introduction
The greenhouse effect raised by carbon dioxide (CO ), which is typically generated from the
2
combustion of natural gas, fuels, coal, oil, and wood, has become the worldwide concern. Thus,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
the chemical conversion of CO as a reactant has attracted great interests due to sustainability
2
and environmental issues. Currently, the literature reports indicate that heterogeneous
hydrogenation of CO typically involves the generation of a wide range of organic compounds,
2
1
e.g., carbon monoxide (CO), methanol, formic acid, methane (CH ), and hydrocarbons.
4
However, the CO molecule is difficult to activate due to its thermodynamic stability, leading to
2
a high energy demand in the process of chemical conversion. As a result, the catalytic
hydrogenation of CO into CO and CH is the prime target. The hydrogenation of CO can
2
4
2
0
produce CO via the reverse water gas shift (RWGS) reaction (i.e., H + CO → CO + H O; ꢁH
2
2
2
ꢀ1
=
41.2 kJ mol ). A variety of industrial catalytic technologies, such as methanol synthesis, CH4
reforming with CO , ammonia synthesis, the FischerꢀTropsch reaction, and steam reforming of
2
2
ꢀ9
hydrocarbons, makes the use of the RWGS reaction. On the other hand, the Sabatier reaction,
0
ꢀ1
namely (4H + CO → CH + 2H O; ꢁH = ꢀ292.9 kJ mol ) has recently attracted considerable
2
2
4
2
attention to produce CH as a green fuel to fulfill the requests for renewable carbon sources and
4
1
0ꢀ13
0
sustainable processes.
Although the Sabatier reaction is thermodynamically favorable (ꢁG =
ꢀ1
−
130.8 kJ mol ), the high kinetic barrier associated with this conversion hinders its industrial
applications.
In recent years, metal nanoparticles (NPs) have been extensively applied in catalytic
1
4
applications since these NPs are more catalytically active than their bulk materials. However,
the unique properties of NPs, such as their large surfaceꢀareaꢀtoꢀvolume ratio, leads to easy
aggregation of metal NPs, often resulting in harmful effects on their activity and selectivity. One
ACS Paragon Plus Environment
3

ACS Catalysis
Page 4 of 47
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
effective methodology to alleviate this aggregation problem is to utilize an ideal support, either
an inorganic support or an organic capping agent, to stabilize highly dispersed metal NPs. Apart
from organic stabilizing agents, various types of solid supports, including zeolites, TiO2,
mesoporous silicas, hollow silicas, silicon nanotubes, mesoporous carbons, polymers, metalꢀ
organic frameworks (MOFs) and graphene oxide, have been extensively utilized to immobilize
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
15ꢀ31
Ni NPs.
By offering high surface areas, unique pore architectures, and poreꢀsize
confinement, mesoporous silica materials are ideal scaffolds to immobilize metal NPs. However,
one main drawback for the use of pure mesoporous silica is that their mesopores do not provide
sufficiently strong affinity with metal precursors. Therefore, additional surface modifications
with organic functional groups on the mesopores of mesoporous silicas are essential.
Nevertheless, the incorporation of high numbers of highly dispersed, smallꢀsized metal NPs onto
mesoporous silica materials remains a challenge.
Conventionally, transitional metals such as Ni, Ru, Rh, Pd, Pt, Cu and their bimetallic alloys
have been demonstrated to be catalytically active in CO hydrogenation for formation of CO and
2
32ꢀ37
CH4.
Among them, Niꢀbased catalysts have been reported as promising candidates for CO2
hydrogenation due to the low cost, high activity, and earth abundance associated with Ni, which
can be viewed as effective alternatives to noble metal catalysts. Moreover, Ni NPs have been
demonstrated to possess high activities in various hydrogenation processes.
38ꢀ
Extensive studies have focused on the search for silicaꢀbased supports to immobilize Ni NPs.
45
When 2D hexagonal mesoporous silicas, such as MCMꢀ41 and SBAꢀ15, are used as supports,
4
4ꢀ46
however, the aggregation of Ni NPs cannot be avoided.
One way to improve this aggregation
problem is to employ 3D cageꢀlike mesoporous silicas, e.g., SBAꢀ16, to confine the Ni NPs and
thus control their growth. Our previous work demonstrated that cageꢀlike mesoporous SBAꢀ16
ACS Paragon Plus Environment
4

Page 5 of 47
ACS Catalysis
1
2
3
4
5
6
7
8
9
can confine the growth of Ni NPs down to approximately 4 nm, exhibiting exceptionally high
47
reactivity and CH selectivity in CO hydrogenation. However, the methods to enhance the
4
2
reaction rate of CO hydrogenation and the CH selectivity are still ambiguous. Moreover, the
2
4
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2+
effective trapping of Ni ions into the cageꢀlike mesostructure of SBAꢀ16 was believed to
contribute to the high reactivity and selectivity. In continuation of our previous work, a strategy
to combine the carboxylic functionality and cageꢀtype characteristics of mesoporous silica SBAꢀ
1
6 to confine Ni NPs was systematically evaluated in this study. Under the appropriate alkaline
2
+
conditions (at pH 9), a high amount of the Ni precursor was effectively incorporated via
ꢀ
electrostatic interactions induced by the –COO groups. On the other hand, no effective Ni
incorporation was observed under acidic conditions. Scheme 1 illustrates the effect of pH on the
fabrication of Ni NPs on the ‒COOH functionalized SBAꢀ16 support. In view of this pH effect,
the Ni(x)@S16C materials, where x is the Ni loading and S16C stands for the support, i.e., ‒
COOH functionalized SBAꢀ16, were prepared in the solution at pH 9 to achieve a maximum
incorporation of Ni. With the aid of the organic functional groups on the cageꢀlike mesopores,
the present synthesis strategy provided a facile route for the fabrication of wellꢀdispersed Ni NPs
with ultraꢀsmall particle sizes down to 2.7 nm and high Ni loadings up to 21.7 wt%. Most
importantly, the effective size control of the Ni NPs was achieved.
Herein, the Ni(x)@S16C catalysts were used to improve the activity and CH selectivity in
4
CO hydrogenation, displaying considerably superior activity as compared to the cases of the
2
Ni(x)@SiO catalysts where the amorphous SiO was used as a support instead. The catalytic
2
2
kinetics, reaction pathways, mechanisms and active sites for the catalytic hydrogenation of CO2
and CO on the SBAꢀ16ꢀsupported Ni NPs were investigated in details. The correlation between
the enhanced catalytic activity and the particle sizes of Ni NPs was made.
ACS Paragon Plus Environment
5

ACS Catalysis
Page 6 of 47
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 1. Synthesis of Ni(x)@S16C under (a) basic (with formation of Ni NPs) and (b) acidic
(without formation of Ni NPs) conditions.
2. Experimental
2.1. Preparation of SBAꢀ16 functionalized with carboxylic acid groups
Carboxylic acid functionalized cubic mesoporous silica SBAꢀ16 was first synthesized
4
8
following a previously reported procedure. The –COOH functionality was introduced into
SBAꢀ16 via a carboxyethylsilanetriol sodium salt (abbreviated as CES, Gelest, 25% in water), a
silicon precursor bearing the carboxylic acid moiety. Briefly, a triblock copolymer blend of 0.93
g of Pluronic F127 (SigmaꢀAldrich) and 0.19 g of Pluronic P123 (SigmaꢀAldrich), both
functioning as structureꢀdirecting agents, and 4 g of KCl were dissolved in 2 M HCl aqueous
solution (60 mL), followed by the addition of a premixed solution of tetraethyl orthosilicate
(TEOS, SigmaꢀAldrich) and CES with a CES/(TEOS + CES) mole ratio of 20%. Then, the
solution was vigorously stirred at 308 K for 4 h. In this study, the functionalization level was
fixed at 20% based on the amount of silica. The molar ratio of the reaction mixture for the
synthesis of SBAꢀ16 was 0.2/0.8/0.0016/0.0037/2.7/4.4/144 for CES/TEOS/P123/F127/KCl
/
HCl/H O. The reaction mixture was then subjected to hydrothermal treatment at 373 K for 24 h.
2
ACS Paragon Plus Environment
6

Page 7 of 47
ACS Catalysis
1
2
3
4
5
6
7
8
9
After filtration and washed with deionized water, the precipitate was collected and air dried at
343 K. The templates (Pluronic F127 and P123) were removed by acid extraction. Typically, the
asꢀsynthesized sample (0.35 g) was dispersed in 125 mL of a H SO solution (48 wt%) with
2
4
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
vigorous stirring at 368 K for 24 h. The final mesoporous silica denoted as S16C was used as the
support to fabricate the Ni NPs.
2.2. Preparation of Ni(x)@S16C and Ni(x)@SiO2
Fabrication of the Ni NPs inside the mesopores of –COOH functionalized SBAꢀ16 (i.e., S16C)
was performed by wet impregnation of 0.1 g of S16C with 10 mL of Ni(NO ) ⋅6H O (Sigmaꢀ
3
2
2
Aldrich) with different concentrations (0.01, 0.025, and 0.05 M), followed by a thermal
reduction process under a flow of Ar/H (95%/5%). The pH of the mixture was subsequently
2
controlled by the dropwise addition of a 0.1 M NaOH aqueous solution. A pale green precipitate
only formed at a solution pH of 9. The pale green precipitate was then filtered, washed with
deionized water, and dried at room temperature. In general, both –COOH and –SiOH functional
ꢀ
ꢀ
groups on the surface of S16C can be deprotonated to become –COO and –SiO functional
2
+
groups to attract Ni ions under appropriate conditions. Figure S1a shows the titration curve of
the S16C support obtained by using a 0.01 M NaOH(aq) solution. The support S16C (0.04 g) was
initially dispersed in 10 mL of pure water (pH 7.0) could generate a weakly acid solution with a
pH value of 3.8. According to the acidꢀbase titration experiments, the equivalent point of the
support S16C titrated by 0.1 M NaCl aqueous solution was around pH 7.3. As reported in the
ꢀ
51
literature, the –SiOH group can become –SiO species when the pH value is higher than ~7.5.
In addition, as revealed by the zeta potential measurements, as shown in Figure S2b, the surface
of the support S16C became negatively charged when the solution pH was increased from 3 to 9
due to deprotonation of both carboxylic acid and silanol groups on the surface of S16C under
ACS Paragon Plus Environment
7

ACS Catalysis
Page 8 of 47
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
alkaline conditions. The preparation of Ni(x)@S16C by wet impregnation at different pHs (i.e.,
3, 5, 7 and 9) was performed. It was found that a pale green precipitate was not well formed
unless by keeping the solution pH to be 9, as shown in Figure S3c. On the other hand, only few
green solid products were formed at low pHs, e.g., at pHs 3 and 6. The intensity of XRD pattern
of each reduced powder obtained at different pHs also gradually increased as the pH was
increased (Figure S3d). In view of this pH effect, the Ni(x)@S16C materials were prepared at
solution of pH 9 in order to achieve the maximum incorporation levels of Ni precursors. The
alkaline condition allows the –COOH groups to be deprotonated, confirmed by the zeta potential
measurements shown in Figure S2b, and therefore have favorable electrostatic interactions with
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
+
the Ni precursors. At pH 9, the optimum adsorption of Ni ions can be achieved with the Ni
loading, measured by ICPꢀAES, which is consistent with the theoretical Ni amount. The detailed
characterization methods for these Ni(x)@S16C samples are described in Supporting
Information (SI).
All the samples were ground and thermally treated at 873 K for 6 h under an atmosphere of
ꢀ1
flowing Ar/H (95%/5%) with a heating rate of 1 K min . The resulting dark gray powder was
2
then kept under vacuum to dry. The Ni amount (wt%) was determined by inductively coupled
plasmaꢀatomic emission spectroscopy (ICPꢀAES). For comparison purposes, the Ni(x)@SiO2
catalysts were prepared with similar Ni loadings as the cases of the Ni(x)@S16C catalysts.
2
ꢀ1
Typically, 1 g of SiO (SigmaꢀAldrich, surface area = 300 m g ) was impregnated with 20 mL
2
of a Ni(NO ) aqueous solution with 0.044 M for 5.9 wt%, 0.094 M for 12 wt% and 0.16 M for
3
2
2
0 wt%, respectively. Afterward, the Ni(x)@SiO catalysts were calcined in air and reduced
2
under H at 773 K for 5 h before use.
2
2
.3. Catalytic CO hydrogenation tests
2
ACS Paragon Plus Environment
8

Page 9 of 47
ACS Catalysis
1
2
3
4
5
6
7
8
9
The Ni(x)@S16C and Ni(x)SiO catalysts were reduced at 673 K for 5 h before reactions. A
2
fixedꢀbed reactor (0.95 cm outer diameter) was equipped to perform all catalytic tests for CO2
and CO hydrogenation over the Ni(x)@S16C and Ni(x)@SiO catalysts at atmospheric pressure.
2
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The CO and CO hydrogenation reactions were conducted by feeding streams of H /CO (1:1)
2
2
2
ꢀ1
and H /CO (3:1), respectively, at a rate of 100 mL min into the reaction vessel. The gas for
2
reaction after 30 min was sampled for analyzing the product distribution as reaction rate. All the
products were analyzed via gas chromatography (GC) quipped with a thermal conductivity
detector (TCD) through a 12 ft PorapakꢀQ column. The turnover frequencies (TOFs) were
ꢀ1
23
calculated based on the formula: TOF = [conversion × mL s for CO or CO × 6.02 × 10
2
ꢀ
1
ꢀ1
(
molecules mol )]/[24400 (mL mol ) × Ni sites]. The peak areas corresponding to CH and CO
4
were used to obtain the carbon balance data, which reached approximately 95ꢀ100%.
.4. Measurements for Fourier transform infrared (FTIR) spectroscopy
The inꢀsitu diffuse reflectance infrared Fourier transform (DRIFT) spectroscopy were
2
ꢀ1
conducted with a 1 cm resolution for 256 scans by using a Nicolet 5700 FTIR spectrometer,
which was equipped with a mercury cadmium telluride (MCT) detector. A heating cartridge was
attached on the DRIFT cell (Harrick) with ZnSe windows to heat the sample to 773 K. A stream
of CO was passed through the Niꢀbased catalysts for 30 min at room temperature, and then the
IR spectra of the Niꢀbased catalysts with adsorbed CO were obtained. Finally, the residual
gaseous CO was purged for 60 min using a He stream. In a similar way, the IR spectra of the coꢀ
adsorption of H and CO were obtained after passing CO through the Niꢀbased catalysts for 30
2
2
2
min at room temperature. The residual gaseous CO was ensured to be purged out for 60 min
2
using a He stream, followed by reaction with a H stream. At the end, the IR spectra were
2
recorded in a H stream at elevated temperatures.
2
ACS Paragon Plus Environment
9

ACS Catalysis
Page 10 of 47
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2
.5. Temperatureꢀprogrammed desorption (TPD) and temperatureꢀprogrammed hydrogenation
(
TPH)
All TPD and TPH experiments were performed under pure He and H /He (20/80) streams at a
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ꢀ1
flow rate of 100 mL min in a fixedꢀbed flow system, respectively. The CO and CO streams
2
ꢀ1
(100 mL min ) were fed over the catalysts for 30 min at room temperature to achieve saturation
for the TPD and TPH experiments. A Kꢀtype thermocouple inserted into the catalyst bed was
ꢀ
1
used to measure the temperature, which was increased from 300 to 650 K at a rate of 10 K min
for obtaining TPD and TPH profiles. All signals were measured with a VG Smart IQ+ 200D
mass spectrometer and the desorbed products were carried to the vacuum chamber through a leak
−
7
valve. The operating pressure in the chamber was approximately 3 × 10 mbar, and the base
ꢀ10
pressure in the chamber was approximately 5 × 10 mbar.
2
.6 Measurements of Ni surface area
The H ꢀTPD technique was employed to measure the number of surface Ni sites per unit mass
2
of catalyst, which was carried out on a TPD system equipped with a thermal conductivity
detector (TCD). The sample (0.2 g) was used and reduced in H stream at 673 K for 5 h, and then
2
was cooled to room temperature from 673 K under the H environment for saturated adsorption.
2
An Ar stream was used to remove the physically addsorbed H at room temperature for 2 h. The
2
ꢀ1
Ni sample with H chemisorption was heated to 773 K ramping at 10 K min in the Ar flow (30
2
ꢀ1
mL min ). The desorbed H was detected continuously with increasing temperature by a TCD.
2
The area of H detected by TCD was quantified by sampling 1 mL of 10% H /Ar to calculate the
2
2
amount of H desorption. The number of Ni sites was determined by means of peak area
2
assuming the adsorption stoichiometry of H/Ni = 1:1. The average surface density of the Ni
19
ꢀ2
metal was 1.54 × 10 Ni atoms m .
ACS Paragon Plus Environment
10

Page 11 of 47
ACS Catalysis
1
2
3
4
5
6
7
8
9
2
.7 Xꢀray photoelectron spectroscopy (XPS)
The oxidation state of the catalyst surface was examined by XPS, using a Thermo VGꢀScientific
Sigma Probe spectrometer at the Precision Instrument Center of the College of Engineering at
the National Central University, Taiwan. The spectrometer was equipped with anAl Kα Xꢀray
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ꢀ
19
source (1486.6 eV; 1 eV = 1.602 × 10 J) operated at 108 W and a hemispherical analyzer
operated at a pass energy of 50 eV. The instrument was typically operated with an analysis
ꢀ9
chamber pressure of approximately 1 × 10 Torr.
3
3
. Results and Discussion
.1. Structural characterization of the Ni(x)@S16C samples
The wideꢀangle Xꢀray diffraction (XRD) patterns of the Ni NPs deposited on ‒COOH
functionalized SBAꢀ16, as shown in Figure 1A, exhibited three major peaks at 2θ values of
4
4.5°, 51.8°, and 76.4°, which can be indexed to the (111), (200), and (220) characteristic facets
of crystalline Ni, respectively. The average size of the Ni NPs was estimated from the Scherrer
equation using the fullꢀwidth at halfꢀmaximum (FWHM) value of the (111) peak and listed in
Table1. The Ni particle size was strongly dependent on the initial concentration of the Ni
precursor used in the preparation of the Ni catalyst. The particle size of the Ni NPs immobilized
on S16C slightly grew with the increase in the Ni loading from 5.9 to 21.7 wt%. For comparison
purposes, the XRD patterns of Ni(x)@SiO with similar Ni loadings (5.6ꢀ20.0 wt%) are shown in
2
Figure S2 (SI). Large Ni particles with sizes between 7.4 and 15.8 nm were formed on SiO2
using a wide range of Ni loadings from 5.6 to 20.0 wt%. Based on the above observations, it can
be inferred that the SBAꢀ16 support functionalized with –COOH groups may provide strong
2
+
ꢀ
interactions between Ni and the –COO groups, leading to a confinement effect on the
formation of Ni NPs due to the cageꢀlike mesopores of SBAꢀ16. Thus, the cageꢀtype mesopores
ACS Paragon Plus Environment
11

ACS Catalysis
Page 12 of 47
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
of ‒COOH functionalized SBAꢀ16 can effectively confine the Ni nanoparticles down to small
sizes ranging from 2.7ꢀ4.7 nm, even if the Ni amount was increased to 21.7 wt%. The Ni surface
area and dispersion for all the Ni(x)@SiO and Ni(x)@S16C catalysts are also compared in
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Table S1, revealing that Ni NPs confined in the cageꢀtype mesopores of S16C apparently
exhibited higher values than Ni deposited on SiO2.
The smallꢀangle XRD patterns (Figure 1B) of Ni(x)@S16C and pristine SBAꢀ16 show one
major diffraction peak at 2θ = 0.8° for the (110) plane, corresponding to cubic SBAꢀ16 with
cageꢀtype mesopores (Im3m symmetry). As the Ni loading was increased, the intensity of the
diffraction peak gradually decreased. It suggests that incorporation of Ni NPs in ‒COOH
functionalized SBAꢀ16 causes some degradation on the mesostructure. The other possible reason
is that the formed Ni nanoparticles partially fill the pores, resulting in reduction of the electronic
contrast between the mesopores and the silica framework. As a result, a less intense (110)
diffraction peak was observed for Ni(x)@S16C.
All the nitrogen adsorptionꢀdesorption isotherms, as shown in Figure 1C, of the Ni(x)@S16C
samples belong to type IV isotherms with H2 hysteresis loops according to IUPAC
classifications. The characteristic shape of hysteresis indicated the presence of mesopores. The
width of the hysteresis loops was affected by the amounts of Ni NPs, and became narrower when
more Ni NPs were loaded. The pore size distribution curves are provided in Figure 1D. The
surface area, pore volume, and pore size of Ni(x)@S16C are summarized in Table 1. As seen in
2
ꢀ1
Table 1, when the amount of Ni NPs was increased, the surface area decreased from 664 m g
2
ꢀ1
for pristine SBAꢀ16 to 231 m g for Ni(21.7)@S16C and the pore volume also decreased from
3
ꢀ1
0.68 to 0.24 cm g , accordingly. The cage pore size has the similar trend due to the blockage of
cages as the Ni loading was increased.
ACS Paragon Plus Environment
12

Page 13 of 47
ACS Catalysis
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1. (A) Wide and (B) small angle XRD patterns of Ni(x)@S16C as a function of Ni
loading, (C) N adsorptionꢀdesorption isotherms and (D) cage pore size distribution curves of
2
Ni(x)@S16C with different Ni loading amounts.
The successful surface functionalization of the SBAꢀ16 support with carboxylic acid
1
3
functional groups was confirmed by C cross polarizationꢀmagic angle spinning (CPMAS)
NMR, as depicted in Figure S3A (SI). As seen in Figure S3A, the peaks at 7 ppm and 27 ppm are
attributed to the C and C carbons of ethyl groups from CES, while the peak at 178 ppm is
1
2
contributed from the carboxylic acid moiety from CES, demonstrating the presence of –COOH
functional groups in the S16C support. Furthermore, the absence of the characteristic NMR
signals of the copolymer Pluronic templates revealed that the surfactants were completely
13
removed from the S16C silica framework. The characteristic C NMR signals of –CH –CH –
2
2
13
13
COOH detected in solid state C CPMAS NMR spectrum of S16C matched well to the C
ACS Paragon Plus Environment
13

ACS Catalysis
Page 14 of 47
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
solution NMR spectrum of the liquid CES itself. The triplet peaks at 76ꢀ78 ppm are due to the
CDCl solvent, as shown in Figure S3(B).
3
There are no peaks related to the T species detected for the support SBAꢀ16, as shown in the
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
29
29
Si MAS NMR spectra (Figure S3(C)). The Si MAS NMR spectra of S16C (Figure S3(D))
n
4
consists of two different peak regions. One region is for the Q sites consisting of Q (–112
3
2
45
ppm), Q (–102 ppm) and Q (ꢀ92 ppm), which are attributed from the TEOS condensation.
m
3
2
The other region corresponds to different T sites, namely T [(CSi(OSi) )], T [CSi(OH)(OSi) ],
3
2
1
and T [CSi(OH) (OSi)] species centered at –67, –60 and –52 ppm, respectively, which are
2
m
29
generated from CES that bears the –COOH moiety. The appearance of the T sites in the Si
MAS NMR spectrum gives direct support of the successful incorporation of carboxylic acid
functional group into the mesopore surfaces of cageꢀtype SBAꢀ16. The quantitative assessment
of the degree of functionalization of the organic moiety can be estimated from the relative
n
m
integrated intensities of Q and T NMR signals (ca. 20.2 %), which is in good agreement with
that expected according to the composition of initial mixture (20% CES). T groups are clearly
observed in the S16C sample, which confirm the presence of the CES moiety. Therefore, both
1
3
29
C and Si solidꢀstate NMR results confirm the presence of the CES moiety in the support
ꢀ
S16C. The carboxylic acid groups on the mesopores of SBAꢀ16 were converted to –COO groups
2
+
at pH 9 and, therefore, could have effective electrostatic interactions with the Ni ions. The Ni
ions could bind to the negatively charged surfaces of the mesopores and then confine in the cageꢀ
type mesopores at nanoscale dimensions even with a high loading of 21.7 wt%.
ACS Paragon Plus Environment
14

Page 15 of 47
ACS Catalysis
1
2
3
4
5
6
7
8
9
Table 1. Textural properties of Ni(x)@S16C
Ni loading ABET
Vp
Cage size Entrance size Ni particle size Ni particle size
a
2
ꢀ1 b
3
ꢀ1 c
d
e
f
g
(
wt %)
(m g )
664
(cm g ) (nm)
(nm)
3.5
(nm)
—
(nm)
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
0
5
1
0.68
0.46
0.42
0.24
7.7
6.5
5.0
4.9
—
.9
414
3.4
2.7
3.6
2.6±0.9
3.7±0.9
5.1±0.9
4.2
1.7
366
3.2
2
231
3.2
4.7
a
b
c
d
Ni loading determined by ICPꢀAES; A
is the surface area; V is the pore volume; Cage
BET
p
e
pore sizes were obtained from the nitrogen adsorption isotherms; Entrance pore sizes were
f
obtained from the nitrogen desorption isotherms; Average particle size of Ni NPs estimated
g
from XRD patterns; Average particle size of Ni NPs estimated from TEM images.
The successful incorporation of the Ni NPs as well as their particleꢀsize distribution as a function
of the Ni loading was further explored by transmission electron microscopy (TEM)
measurements. A wellꢀordered unfilled pore structure was clearly observed for the pristine SBAꢀ
16 sample in the TEM image shown in Figure 2(a). For all the Ni(x)@S16C samples (Figure 2(bꢀ
d)), ultraꢀsmall Ni NPs were well confined within the cage pores of the S16C support. The Ni NP
size was tuned down to approximately 3 nm for Ni(5.9)@S16C by lowering the initial molar
concentration of the Ni precursor, and tuned to approximately 5 nm for Ni(21.7)@S16C by a
high initial molar concentration of the Ni precursor. The average particle sizes of Ni NPs with
standard deviations, estimated from the TEM images, are also listed in Table 1.
ACS Paragon Plus Environment
15

ACS Catalysis
Page 16 of 47
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 2. TEM images of Ni(x)@S16C, where x = (a) 0, (b) 5.9, (c) 14.2, and (d) 21.7 wt%.
The mesostructure of Ni(14.2)@S16C was further analyzed by highꢀresolution TEM
(HRTEM), as shown in Figure 3. The magnified HRTEM image and SAED pattern in Figure
3(bꢀc) clearly demonstrate that the Ni NPs in the S16C support are highly crystalline, indicating
distinct lattice fringes with a spacing of 0.203 nm and consistent with the (111) lattice plane of
the metallic Ni. The energy dispersive Xꢀray spectrum (EDS) of Ni(14.2)@S16C clearly
ACS Paragon Plus Environment
16

Page 17 of 47
ACS Catalysis
1
2
3
4
5
6
7
8
9
indicated that the Ni(14.2)@S16C sample mainly contains Si, O and Ni elements (Figure 3dꢀg)
and confirmed well dispersion of Ni NPs in the S16C matrix.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. TEM images of (a) the Ni NPs in Ni(14.2)@S16C, (b) displaying the dꢀspacing of the
Ni NPs; (c) selectedꢀarea electron diffraction (SAED) pattern of Ni(14.2)@S16C; (dꢀg) energy
dispersive spectroscopy (EDS) selected area and mapping of silicon, oxygen and nickel.
Notably, the Ni NPs were immobilized in the mesopores of SBAꢀ16, which suggested that the
presence of the –COOH functionality facilitated the formation of sizeꢀcontrollable Ni NPs
without aggregation. The role of the carboxylic acid groups during the formation of the Ni NPs
ꢀ1
was probed by FTIR spectroscopy (Figure S4, SI). The vibrational band at 1725 cm ,
corresponding to the –COOH groups in SBAꢀ16, disappeared after fabrication of the Ni NPs. We
2+
anticipate that the –COOH groups deprotonated at pH 9 and attracted the cationic Ni precursor
to enter the mesopores via electrostatic interactions during the initial experimental procedure.
Most of the –COOH groups decomposed when a high reduction temperature was employed to
ensure the formation of the Ni NPs.
ACS Paragon Plus Environment
17

ACS Catalysis
Page 18 of 47
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Although we have observed from TEM images (Figures 2 and 3) that most of the Ni
nanoparticles are uniformly distributed in the pores of S16C support, still it is possible that some
nanoparticles may be deposited outside the pores. A close examination of the TEM images
indicated very few aggregated Ni nanoparticles outside the pores. The nitrogen adsorptionꢀ
desorption data listed in Table 1 also confirm the formation of Ni nanoparticles inside the pores
as the surface area and pore volumes decrease with the increase in the loading of Ni
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ꢀ
ꢀ
nanoparticles. The –COO and –SiO functional groups on S16C can effectively attract the
2
+
positive Ni ions into the cage structure. For example, the Ni(21.7)@S16C provided larger Ni
size with 5.1±0.9 nm than a low Ni loading catalysts, but was apparently smaller than the cage
size of pure S16C (7.7 nm). Therefore, our results suggested that most of the Ni nanoparticles
were well distributed within the S16C matrix.
The oxidation state associated with the different size and loading amount of Ni NPs
immobilized in the mesoporous silica frameworks can be further explored by XPS
measurements. The Ni XPS spectra of Ni(x)@S16C were compared and depicted in Figure S5
(SI). As observed in Figure S5, the binding energy of Ni 2p3/2 signal exhibited at 854.5 eV for
5.9 wt%, at 854.5 eV for 14.2 wt% and slightly shifted to 854.9 eV for 21.7 wt%, respectively.
The surface of reduced Ni(x)@S16C might be partially oxidized under atmosphere environment
during the sample transfer process before XPS measurements. Thus, the binding energy of Ni
5
0ꢀ52
2p3/2 signals in Figure S5 was likely toward the formation of NiO.
However, the relative shift
for Ni 2p3/2 signal to higher energy could also provide the clue for Ni NPs interacting with the
S16C support, when the Ni loading was increased to 21.7 wt%. The Ni(21.7)@S16C sample
with higher loading and larger particle size might lead to higher likelihood that the Ni NPs were
situated outside the cage pores of S16C than other samples with low Ni loading. In this case, it is
ACS Paragon Plus Environment
18

Page 19 of 47
ACS Catalysis
1
2
3
4
5
6
7
8
9
possible that most of –COOH groups were sacrificially decomposed after reduction treatment,
therefore the interaction of Ni NPs with oxygen bonded SiꢀOH group might influence the
oxidation state of Ni NPs. For low Ni loadings of 5.9 wt% and 14.2 wt%, the majority of Ni NPs
should confine inside the cage pore and weakly interact with the small amount of SiꢀOH groups
in pore structure. Increasing the Ni loading up to 21.7 wt% could cause the additional supportꢀ
metal interaction for Ni NPs located on the surface of the mesoporous silica, As a result, the XPS
peak slightly shifted to a higher energy for Ni(21.7)@S16C.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3.2. Catalytic hydrogenation of CO on the Ni(x)@S16C and Ni(x)@SiO catalysts
2
2
The hydrogenation of CO was employed to examine the catalytic efficiencies of the
2
Ni(x)@S16C and Ni(x)@SiO catalysts. The TOFs were estimated for all the Ni catalysts based
2
on the number of Ni surface sites, as measured by H chemisorption at room temperature.
2
The overall TOFs for CO hydrogenation on the Ni(x)@S16C and Ni(x)@SiO catalysts are
2
2
achieved at 598ꢀ773 K, as revealed in Figure 4A. The major products detected during CO2
hydrogenation were CO and CH . The Ni(x)@S16C catalysts exhibited higher TOFs for CO2
4
hydrogenation than the Ni(x)@SiO catalysts at all the investigated temperatures. The catalytic
2
activity of Ni(x)@S16C increased with a decrease of the Ni loading and particle size. The
temperatureꢀdependent TOFs for the formation of CO and CH analyzed from Figure 4A are
4
shown in Figures 4B and 4C. The TOFs for CO on the Ni(x)@S16C catalysts followed the order
of 5.9 wt% (2.7 nm) > 14.2 wt% (3.6 nm) > 21.7 wt% (4.7 nm), but the rate for CH formation
4
was enhanced upon increasing the Ni particle size. Figure S6 compares the CO conversion,
2
selectivity of CO and CH as a function of temperature, based on the results in Figure 4. As seen
4
in Figure S6, a larger Ni loading might lead to higher reaction conversion since the overall
surface sites increased with the Ni loading. Although both Ni(x)@S16C and Ni(x)@SiO with
2
ACS Paragon Plus Environment
19

ACS Catalysis
Page 20 of 47
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
similar Ni loadings exhibited similar CH selectivity at all temperatures, the Ni size of the former
4
was significantly smaller than that of the latter.
The WeiszꢀPrater criterion (NWꢀP) was frequently used for negligible diffusion
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
53,54
limitation.
The number of NWꢀP smaller than 0.3 should have negligible mass transfer
53,54
limitations.
The values of NWꢀP for both Ni(21.7)@S16C and Ni(20) @SiO catalysts for
2
CO hydrogenation at 673 and 773 K were calculated to obtain extremely small numbers as
2
listed in Table S2, indicating that the turnover rates in this study were not affected by diffusion
limitations of reactants.
Figure 4. Comparison of the reaction rates for CO hydrogenation on the Ni(x)@S16C and
2
Ni(x)@SiO catalysts as a function of temperature: (A) overall CO hydrogenation; (B) CO
2
2
formation and (C) CH formation. A H /CO stream with a 1:1 ratio at a total flow rate of 100
4
2
2
ꢀ1
mL min was passed over 5 mg of the catalysts.
The CO conversion and selectivity of both products as a function of time at 773 K is revealed
2
in Figure 5. All Ni(x)@S16C catalysts appeared high stability within 48 h. The XRD spectra of
ACS Paragon Plus Environment
20

Page 21 of 47
ACS Catalysis
1
2
3
4
5
6
7
8
9
Ni(x)@S16C catalysts after the timeꢀonꢀstream tests were shown in Figure S7, revealing the Ni
particle size of 2.7 nm for 5.9 wt%, 4.1 nm for 14.2 wt% and 5.5 nm for 21.7 wt%. The particle
size of Ni did not seem to be influenced by the longꢀterm catalytic reaction at 773 K.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 5 Timeꢀdependent CO hydrogenation for (A) overall CO conversion; (B) CO selectivity
2
2
ꢀ1
and (C) CH selectivity. A H /CO stream with a 1:1 ratio at a total flow rate of 30 mL min was
4
2
2
passed over 50 mg of the catalysts at 773 K.
The Arrhenius plots for CO hydrogenation, which provided the apparent activation energies,
2
on the Ni(x)@S16C and Ni(x)@SiO catalysts were further investigated, as shown in Figure 6.
2
All the experiments were performed in the temperature range of 598ꢀ673 K. The apparent
activation energies were estimated from the slopes, obtaining values of 48.4 kJ/mol for
Ni(5.9)@S16C, 52.6 kJ/mol for Ni(14.2)@S16C and 56.8 kJ/mol for Ni(21.7)@S16C, as shown
in Figure 6. For comparison, the apparent activation energies for the Ni(x)@SiO catalysts with
2
ꢀ1
5
.6ꢀ20.0 wt% Ni loadings were ~82.9–94.0 kJ mol , which were significantly higher than the
cases of the Ni(x)@S16C catalysts.
ACS Paragon Plus Environment
21

ACS Catalysis
Page 22 of 47
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 6. Arrhenius plots for the TOFs of CO hydrogenation on the Ni(x)@S16C andNi@SiO2
2
catalysts.
Figure S8 (SI) shows the CH selectivity versus the CO conversion on the reduced
4
2
Ni(x)@S16C catalysts. The CO conversion was changed using different flow rate of the
2
reactants and the weight of catalyst. It can obtain that the selectivity of CH decreased with a
4
decreasing CO conversion and approached approximately zero at very low conversion on all the
2
Ni(x)@S16C catalysts. It was indicative of a typical stepwise reaction in which CO is an
5
5
intermediate in the process of CH formation. Therefore, CO hydrogenation on the
4
2
Ni(x)@S16C catalysts is proposed to follow a consecutive pathway, where CO is the key
intermediate in CO methanation. According to our previous investigations, the consecutive
2
pathway of CO → CO → CH was assumed to be the essential steps during CO hydrogenation
2
4
2
56
on the Ni(x)@SiO catalysts, regardless of the Ni particle size. Thus, the enhancement of the
2
CH formation on Ni(x)@S16C might depend on the reaction pathway for CO hydrogenation.
4
ACS Paragon Plus Environment
22

Page 23 of 47
ACS Catalysis
1
2
3
4
5
6
7
8
9
Figure 7A compares the catalytic activities of CO hydrogenation to CH on the Ni(x)@S16C and
4
Ni(x)@SiO catalysts at different temperatures. The CO methanation reaction was performed in
2
the temperature range of 573ꢀ673 K. As expected, all the Ni(x)@S16C catalysts possessed higher
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
rates of CO hydrogenation than the Ni(x)@SiO catalysts, where increasing the Ni loading and
2
particle size led to considerable improvements to the TOF. The Arrhenius plots for CO
hydrogenation on the Ni(x)@S16C and Ni(x)@SiO catalysts are shown in Figures 7B and 7C.
2
On the Ni@SiO catalysts, the apparent activation energies displayed similar values regardless of
2
the Ni loading, but the apparent activation energies on the Ni(x)@S16C catalysts decreased from
78.3 kJ/mol to 62.1 kJ/mol when the Ni loading was increased from 5.6 to 21.7 wt%.
Figure 7. (A) Comparison of the reaction rates for CO hydrogenation for the Ni(x)@S16C and
Ni(x)@SiO catalysts as a function of temperature; (B) and (C) Arrhenius plots for the TOFs of
2
CO hydrogenation on the Ni(x)@S16C and Ni(x)@SiO catalysts. A H /CO stream with a 3:1
2
2
ꢀ1
ratio at a total flow rate of 100 mL min was passed over 5 mg of the catalysts.
3
.3. CO ꢀ and COꢀTPD on the Ni(x)@S16C catalysts
2
ACS Paragon Plus Environment
23

ACS Catalysis
Page 24 of 47
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
In this study, small Ni NPs with sizes of 2.7ꢀ4.7 nm confined in S16C that could induce high
catalytic activities for CO hydrogenation and effectively enhance CH formation compared to
2
4
the Ni(x)@SiO catalysts. In our previous studies, small Ni NPs typically decreased CH4
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
56
formation due to low H coverage on the small Ni NPs. This special phenomena on the
2
Ni(x)@S16C catalysts might result from the relationship between the active sites and CO and
2
CO adsorption. We used TPD to study the desorption behavior of CO and CO from the active
2
sites on Ni(x)@S16C catalysts. Figure 8 compares the TPD profiles of CO and CO desorbed
2
from the Ni(x)@S16C catalysts under a He stream. As shown in Figure 8A, CO desorbed at
2
~
355 K from Ni(5.6)@S16C, and a new peak associated with CO desorption appeared at 460 K
2
upon increasing the Ni loading.
Figure 8. TPD plots of (A) CO and (B) CO desorbed from the Ni(x)@S16C catalysts in a He
2
ꢀ1
stream with a 10 K min heating rate. All CO and CO adsorptions were performed in a pure 100
2
ꢀ1
mL min stream at atmospheric pressure and 298 K for 40 min.
ACS Paragon Plus Environment
24

Page 25 of 47
ACS Catalysis
1
2
3
4
5
6
7
8
9
The TPD results of CO desorption from the Ni(x)@SiO catalysts are also shown in Figure
2
2
S9A (SI), revealing similar desorption behavior for all the Ni loadings. The desorption peak at
360 K retained similar intensities for all the Ni(x)@SiO catalysts, but the intensity of the
~
2
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
second peak at high temperatures was markedly enhanced as well as the desorption temperature
shifted from 520 K to 476 K as Ni loading increased from 5.6 wt% to 20 wt%. In general, CO2
can be considered as an acidic gas, and thus the desorption temperature represents the strength of
the basic sites to bind CO . Literature reports have indicated that the CO ꢀTPD profiles on Ni
2
2
surfaces show the desorption temperatures between 373ꢀ423 K for weakly basic sites, 423ꢀ573 K
57,58
for moderately basic sites and above 773 K for strongly basic sites.
As shown in Figures 8A
and S9A, the CO desorbed from Ni(x)@S16C and Ni(x)@SiO might indicate the existence of
2
2
weakly and moderately basic sites on these surfaces. Notably, the relative intensities for CO2
desorption from the Ni(x)@S16C catalysts were larger than those from the Ni(x)@SiO2
catalysts, implying that the Ni(x)@S16C catalysts might induce higher CO coverage.
2
The formation of CH during CO hydrogenation likely depended on the efficiency for CO
4
2
hydrogenation due to the reaction following the consecutive pathway. Thus, the COꢀTPD results
might shed light on the active sites for CO adsorption. Figure 8B reveals the TPD profiles for
saturated CO adsorption on the Ni(x)@S16C samples. Small amounts of CO desorption were
observed from Ni(5.9)@S16C at ~365 and 478 K, and then three intense desorption bands
gradually formed at ~365, 405 and 478 K upon increasing the Ni concentration to 14.2 wt% and
2
1.7 wt%. Figure S9B (SI) compares the COꢀTPD on the Ni(x)@SiO catalysts, revealing a
2
single and weak peak at 365 K. These results indicate that Ni(x)@S16C offers a larger number of
sites and greater CO adsorption ability than Ni(x)@SiO Assignments of the CO desorption
2
.
temperature from the Ni surface to specific adsorption sites have been reported in the literature.
ACS Paragon Plus Environment
25

ACS Catalysis
Page 26 of 47
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
The peaks observed at approximately 365 and 405 K were likely ascribed to linear CO
chemisorption, and the desorption peak at 478 K might be derived from a bridgingꢀtype of CO
5
9
adsorption.
.4. CO ꢀ and COꢀTPH on the Ni(x)@S16C catalysts
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
2
To identify the relationship between the CO desorption state and CO/CH formation, TPH
2
4
was further investigated on the CO ꢀ and COꢀadsorbed Ni(x)@S16C catalysts under a H /He
2
2
(
20/80) stream. Figure 9 shows the CO ꢀTPH results, where CO , H O, CO and CH molecules
2 2 2 4
are detected at elevated temperature. On Ni(5.9)@S16C, adsorbed CO was hydrogenated to
2
desorbed CO at 360 K. Upon increasing the Ni content to 14.2 wt% and 21.7 wt%, a weak CO2
signal was observed at ~360 K, while the desorption temperatures of CO derived from CO2
hydrogenation appeared at ~360, 423 and 500 K. The temperatures for CO desorption were
similar to the results of the COꢀTPD profiles in Figure 8B, implying that the step of CO2
hydrogenated to CO might occur below 360 K, before desorbing CO at increased temperatures.
However, the relative intensity of desorbed CO at 500 K, corresponding to CO bound to the
bridging sites, was small compared to that in the COꢀTPD results in Figure 8B. The obvious CH4
desorption peak at ~ 520 K in the TPH of CO for all the Ni(x)@S16C likely corresponded to the
2
hydrogenation of a bridging CO intermediate.
ACS Paragon Plus Environment
26

Page 27 of 47
ACS Catalysis
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ꢀ1
Figure 9. TPH plots of CO desorbed from the Ni(x)@S16C catalysts in a 100 mL min H /He
2
2
ꢀ1
(
20/80) stream with a 10 K min heating rate. CO adsorption was performed in a pure 100 mL
2
ꢀ1
min CO stream at atmospheric pressure and 298 K for 40 min.
2
The CH formation correlated to adsorbed CO species on Ni(x)@S16C was further identified
4
using the TPH experiments for CO adsorption, as shown in Figure 10. As expected, the marked
CH desorption was also present at 520 K via the hydrogenation of adsorbed CO, but the peaks
4
for CO desorption at 365 K and 405 K persisted, and the peak at 478 K associated with the
adsorption of bridging CO completely disappeared. This observation suggests that the adsorption
of bridging CO on Ni(x)@S16C may be the important species for CH formation. Thus, CH4
4
likely formed from the hydrogenation of adsorbed bridging CO. The CO ꢀTPH profiles of
2
Ni(20)@SiO are shown in Figure S10 (SI) to compare with the reactions on Ni(x)@S16C,
2
revealing that the weak CO and CH peaks occurred at ~500 K. The results were attributed to the
4
ACS Paragon Plus Environment
27

ACS Catalysis
Page 28 of 47
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
simultaneous formation of CO and CH at ~500 K resulting from intermediate dissociation in the
4
process of CO hydrogenation.
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ꢀ1
Figure 10. TPH plots of CO desorbed from the Ni(x)@S16C catalysts in a 100 mL min H /He
2
ꢀ1
(20/80) stream with a 10 K min heating rate. CO adsorption was performed in a pure 100 mL
2
ꢀ1
min CO stream at atmospheric pressure and 298 K for 40 min.
3.5. IR spectra of CO adsorbed on Ni(x)@S16C
Figure 11A shows the IR spectra of CO adsorbed on Ni(x)@S16C to investigate the active
sites associated with CO hydrogenation. CO initially adsorbed on all the Ni(x)@S16C catalysts
2
ꢀ1
at 298 K, exhibiting wide IR bands with peaks centered at 2042, 2013, 1900 and 1830 cm . The
relative intensities of the CO peaks increased upon increasing the Ni loading. The vibrational
stretching bands of CO adsorbed on the Ni surfaces have been reported in several literature
6
0ꢀ62
ꢀ1
reports.
The peaks at 2042 and 2013 cm are attributed to linear CO adsorption on the Ni
ꢀ1
surface. The 1900 and 1830 cm peaks may correspond to bridging Ni CO species and 3ꢀfold or
2
ACS Paragon Plus Environment
28

Page 29 of 47
ACS Catalysis
1
2
3
4
5
6
7
8
9
60ꢀ62
4
ꢀfold CO adsorption, respectively.
Temperatureꢀdependent IR spectra of CO adsorbed on
ꢀ1
Ni(14.2)@S16C are shown in Figure 11B. The bands for linear CO at ~ 2042 and 2013 cm and
ꢀ1
bridging CO at ~1900 cm greatly diminished as the temperature increased to 573 K, but the
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ꢀ1
peak at 1830 cm , assigned to multiꢀbonded Ni CO or Ni CO species, remained intense even at
4
3
573 K due to the strong adsorption for the multiꢀbonded CO species. The IR results are in
reasonable accordance with the COꢀTPD measurements on Ni(x)@S16C shown in Figure 8B,
indicating that the desorbed peaks at 365 and 405 K might be associated with linear CO and that
the peak at 478 K corresponds to bridging CO. However, multiꢀbonded CO might strongly bind
to the Ni surface, causing difficultly for observation in the TPD profiles.
Figure 11. (A) IR spectra of CO adsorbed on the Ni(x)@S16C catalysts and (b) temperatureꢀ
dependent IR spectra of CO on the Ni(14.2)@S16C catalyst.
ACS Paragon Plus Environment
29