Nickel Nanoparticles Supported on Nonreducible Mesoporous Materials: Effects of Framework Types on the Catalytic Decomposition of Methane

Full text of DOI:10.1002/bkcs.12168

Communication
DOI: 10.1002/bkcs.12168
BULLETIN OF THE
T. Kim et al.
KOREAN CHEMICAL SOCIETY
Nickel Nanoparticles Supported on Nonreducible Mesoporous
Materials: Effects of Framework Types on the Catalytic Decomposition
of Methane
†,
Taeho Kim,† Bappi Paul,† Dong-il Kwon,† Sung Chan Nam,‡, and Changbum Jo
*
*
†Department of Chemistry and Chemical Engineering, Inha University, 100 Inha-ro, Incheon 22212,
South Korea. *E-mail: jochangbum@inha.ac.kr
‡Greenhouse Gas Laboratory, Korea Institute of Energy Research, 217 Gajeong-ro, Daejeon 34129,
South Korea. *E-mail: scnam@kier.re.kr
Received August 13, 2020, Accepted November 26, 2020
Keywords: Methane decomposition, Hydrogen production, Supported catalyst, Mesoporous materials
Hydrogen is viewed as a key green energy source by the
chemical and petrochemical industries1 because energic use
of hydrogen does not generate greenhouse gas emissions.
As a result, many researches have explored methods of
generating hydrogen in an economical way.2,3 Methane is
the main constituent of biogas and natural gas, and have
the hydrogen/carbon ratio of four, which is higher than that
of any other hydrocarbon. Many processes have been
devised to extract hydrogen from methane, including steam
reforming of methane,4 catalytic decomposition of methane
(CDM),5 and partial oxidation of methane6; however, CDM
is considered to have the greatest potential in terms of its
low CO2 emissions.
The CDM has been widely studied for a practical pur-
pose. During this process, coke is formed as a by-product
of decomposition, and the cokes can block catalyst surfaces
and causes catalyst deactivation.7,8 Many research studies
on active metals,9–17 supports,18 and promotors19 have been
performed to achieve high catalyst activity and stability and
resistance against deactivation during methane decomposi-
tion. These efforts have shown that supported catalysts pro-
vide one of the best ways of improving catalytic activity.
There are various catalyst supports (metal oxides, zeolites,
carbons, silica, and others)5 and all of these supports
increases catalyst activity in one way or another. For exam-
ple, strong metal-supports interact to enhance the dispersity
of active metals and provide resistance to sintering. In turn,
the catalytic performance such as activity and stability was
enhanced.20 Furthermore, the reducibility of support can
affect catalyst activity.21 Sometimes, coke formation occurs
at acid sites on the support material, and thus, the introduc-
tion of basicity to support offers a means of reducing car-
bon formation.22 Recent studies have shown that support
porosity improves catalyst activity.23 Porous materials have
unique structures with internal pores that provide locations
for catalysts. Loading porous material with catalyst
enhances catalyst dispersity, and thus, catalyst activity.24,25
In the present work, we investigated the effects of the
framework types of mesoporous materials on CDM cata-
lytic performance. The supports used in this work were
composed of nonreducible oxides (i.e., SiO2 and Al2O3) to
minimize the effect of support reducibility. Specifically,
mesopore walls were composed of crystalline microporous
MFI zeolite, amorphous silica, and crystalline γ-alumina.
Framework structures were determined by X-ray diffraction
(XRD) (Figure S1). The resulting materials are denoted
“M-Zeo,” “M-SiO2,” and “M-Al2O3,” respectively, where
“M” means mesoporous.
The porosities of individual supports were investigated
using N2 adsorption–desorption isotherms (Figure S2;
Table S1). M-Al2O3 had the lowest surface area and pore
volume. M-zeo exhibited a higher specific surface area
(550 m2/g) than M-SiO2, but its pore volume was lower,
which is attributed to the larger mesopore diameters of M-
SiO2 (see Figure S3 for pore size distributions). Note that
M-SiO2 exhibited mesoporosity only, while M-Zeo
exhibited a hierarchical structure, in which zeolite micro-
pores and intercrystalline mesopores were highly inter-
connected.26 The acidities of M-SiO2 and M-Zeo samples
were characterized by using NH3-TPD (Figure S4). M-Zeo
ꢀ
exhibited unimodal distribution with maxima at ~310 C.
The NH3-desorption peak was shifted to lower temperature
ꢀ
(~190 C) in the case of M-SiO2, indicating that M-SiO2
had a relatively lower acid strength than M-Zeo. For CDM,
nickel was selected as the active metal and was loaded on
these mesoporous supports using a melting method at ~15%
by weight (see Supporting information for experimental
details).
Figure S5 shows XRD patterns for Ni/M-Zeo, Ni/M-
SiO2, and Ni/M-Al2O3. These results confirmed the suc-
cessful loading of metallic nickel after reduction on all
supports, as indicated by characteristic nickel peaks at
44.36ꢀ and 51.72ꢀ.27 In the case of Ni/M-Al2O3, two addi-
tional XRD peaks (19.4ꢀ and 32.0ꢀ) were observed, which
were assigned to nickel-aluminate with a spinel structure.
This suggested a considerable portion of the Ni reacted
Bull. Korean Chem. Soc. 2020
© 2020 Korean Chemical Society, Seoul & Wiley-VCH GmbH
Wiley Online Library
1

Communication
ISSN (Print) 0253-2964 | (Online) 1229-5949
BULLETIN OF THE
KOREAN CHEMICAL SOCIETY
Figure 2. (a) Conversion of methane as a function of reaction time
for Ni/M-SiO2 (blue), Ni/M-Zeo (black), and Ni/M-Al2O3 (red).
(b) Carbon yield of individual supported nickel catalyst. Reaction
temperature: 550ꢀC, CH4/N2 = 0.5 WHSV: 2.95 h−1
.
best initial activity (~26.5%), which decreased in the fol-
lowing order: Ni/M-SiO2 > Ni/M-Zeo > Ni/M-Al2O3. It
would appear that the high dispersity of supported Ni
nanoparticles <10 nm is related to catalyst activity. Interest-
ingly, despite well-dispersed nickel in Ni/M-Al2O3, a con-
siderable amount of this nickel was converted to nickel
aluminate, which might reduce the initial activity.29
As CDM proceeded, CH4 conversion was decreased as a
function of reaction time for all Ni-supported catalysts,
although their deactivation rates differed. In this study, cat-
alytic stability was defined as the time required to achieve
5% CH4 conversion. The Ni/M-Al2O3 catalyst was most
rapidly deactivated to reach the 5% CH4 conversion within
3 h. Catalyst stability increased in the order of Ni/M-
Zeo < Ni/M-SiO2. In the case of Ni/M-SiO2 catalyst, the
initial activity was retained for 5 h and then slowly
deactivated. Ni/M-SiO2 reached 5% CH4 conversion after
20 h of reaction, which was the best result observed.
It is well known that methane decomposes to carbon and
hydrogen, and thus, carbon yield also provides a measure
of catalyst performance (i.e., hydrogen productivity). Car-
bon yield can be calculated using Eq. (1), as previously
described.24,30
Figure 1. STEM images and their corresponding EDS elemental
mapping images in STEM mode. (a) Ni/M-Al2O3, (b) Ni/M-Zeo,
and (c) Ni/M-SiO2.
with alumina to form a nickel-alumina compound during
the reduction process.28
Figure 1 shows representative scanning transmission
electron microscopy (STEM) images for Ni/M-Al2O3, Ni/
M-Zeo, and Ni/M-SiO2. Particle size distributions for
supported Ni nanoparticles were derived from several TEM
images (Figures S6–S9). We considered that 10–20 nm
nickel particles appeared to be loaded inside mesopores,
whereas larger particles were located on support surfaces in
case of all. Ni/M-SiO2 had the highest loading rate in
mesopores. In terms of uniformity, Ni/M-Zeo sample
showed a broad distribution of particle diameters for
supported Ni. In addition, N2 isotherms for Ni-supported
samples showed a notable decrease in mesopore volume,
indicating that a large portion of Ni nanoparticles was
located inside mesopores (Figure S10).
Weight of formed coke
Carbon yield =
ð1Þ
Weight of Ni in catalyst
The weight of coke formed can be calculated using ther-
mogravimetric analysis (TGA) results. Figure S11 shows
TGA results for reacted-Ni/M-SiO2, reacted-Ni/M-Zeo, and
reacted-Ni/M-Al2O3. All samples experienced an abrupt
mass loss from 400ꢀC to 600ꢀC, which was attributed to
the decomposition of the cokes produced in air. Therefore,
mass losses in the range of 400–600ꢀC are associated with
the weight of coke formed during CDM reaction. The car-
bon yield was determined using Eq. (1); and results are
summarized in Figure 2(b). As was expected, Ni/M-SiO2
provided the highest carbon yield, which suggests high ini-
tial activity and catalyst stability.
Figure 2(a) shows CH4 conversions as a function of reac-
tion time over Ni/M-Zeo, Ni/M-SiO2, and Ni/M-Al2O3.
The initial activity was defined as the CH4 conversion
obtained after reaction for 1 h. Ni/M-SiO2 exhibited the
TEM and Raman spectroscopy were used to determine
the characteristics of the coke formed on the catalyst. For
Bull. Korean Chem. Soc. 2020
© 2020 Korean Chemical Society, Seoul & Wiley-VCH GmbH
www.bkcs.wiley-vch.de
2

BULLETIN OF THE
Communication
KOREAN CHEMICAL SOCIETY
three catalysts, the coke was formed as carbon nanotube
(CNT). TEM shows most of the formed CNTs in Ni/M-
SiO2 and Ni/M-Zeo had around 10–20 nm, though excep-
tionally some of them were ~50 nm in Ni/M-SiO2 (Figures
S12–S14). Figure S15 shows g-band (~1580 cm−1) and d-
band (~1350 cm−1) peaks were detected in Ni/M-SiO2 and
Ni/M-Zeo, whereas a low intensity g-band and d-band peak
was observed in Ni/M-Al2O3 due to the small amount of
coke produced.31 We also confirmed the presence of CNTs
by XRD, as it produces a characteristic peak at ~26ꢀ
(Figure S16).32
previous paper.26 Nickel nanoparticles were supported on
the support using the melting method. CDM reaction was
performed in a fixed-bed reactor system using quartz reac-
tor (see detailed description in Supporting Information).
Acknowledgment. This work was supported by Inha Uni-
versity Research Grant (INHA-63143).
Supporting Information. Additional supporting informa-
tion may be found online in the Supporting Information
section at the end of the article.
There are two main causes of supported Ni catalyst deacti-
vation during CDM reactions, that is, metal sintering and
coke formation. To determine the cause of deactivation, we
first investigated changes in Ni particle size distribution after
CDM. Figure S17 shows nickel particle sizes were not chan-
ged, which indicates deactivation was not caused by sintering
during the reaction. In addition, when the N2 adsorption
behaviors of the spent catalysts were almost restored after air
regeneration, indicating that textural properties might not be
changed during the CDM reactions. Therefore, we believe
the deactivation was primarily due to coke formation, as has
been previously reported.8,23 The catalytic activity of the
spent Ni/M-SiO2 catalyst can be restored by coke removal
through air calcination (Figure S18).
As regards catalyst stability, we observed deactivation
was slower for higher pore volumes. According to XRD
and TEM measurements, the cokes formed may be multi-
walled CNTs.30 Thus, it seems that higher mesopore vol-
umes might increase the amount of CNTs formed, which
might enhance catalyst stability of supported Ni catalysts
during CDM. To support the effect mesopore volume,
Ni supported on bulk MFI was prepared and compared
the CDM performance with Ni/M-Zeo (Figure S19). Deac-
tivation rate is faster for Ni/bulk-MFI, as compared to Ni-
M-Zeo.
We also found that total carbon production when non-
reducible mesoporous supports were used was strongly
related to pore volume. Higher pore volumes might provide
more space for the formation of multi-layered CNTs, which
means coke capacity is closely connected to catalytic abil-
ity. Accordingly, for a high quantity of H2 generation,
higher pore volume should be considered a driver of cata-
lyst activity and stability during CDM.
In conclusion, the catalytic performance of Ni supported
materials in CDM reaction is highly related to the amount
of mesopore volume of supports. This might be attributed
to that higher pore volumes might provide more space for
the multi-layered CNTs (i.e., cokes), which means coke
capacity is closely connected to catalytic ability.
References
1. D. Parra, L. Valverde, F. J. Pino, M. K. Patel, Renew. Sust.
Energ. Rev. 2019, 101, 279.
2. A. M. Abdalla, S. Hossain, O. B. Nisfindy, A. T. Azad, M.
Dawood, A. K. Azad, Energy Convers. Manag. 2018,
165, 602.
3. J. D. Holladay, J. Hu, D. L. King, Y. Wang, Catal. Today
2009, 139, 244.
4. T. L. LeValley, A. R. Richard, M. Fan, Int. J. Hydrog.
Energy 2014, 39, 16983.
5. W. J. Jang, J. O. Shim, H. M. Kim, S. Y. Yoo, H. S. Roh,
Catal. Today 2019, 324, 15.
6. P. Corbo, F. Migliardini, Int. J. Hydrog. Energy 2007, 32, 55.
7. A. S. A. Al-Fatish, A. A. Ibrahim, A. H. Fakeeha, M. A.
Soliman, M. R. H. Siddiqui, A. E. Abasaeed, Appl. Catal. A:
Gen. 2009, 364, 150.
8. H. F. Abbas, W. M. A. Wan Daud, Int. J. Hydrog. Energy
2010, 35, 1160.
9. Z. Bian, S. Das, M. H. Wai, P. Hongmanorom,
ChemPhysChem 2017, 18, 3117.
10. S. A. Abbas, A. Ma, D. Seo, Y. J. Lim, K.-D. Jung, K. M.
Nam, Bull. Kor. Chem. Soc. 2020, 41, 1080.
11. Shin. H, Oh S, Jun. H, Oh M, Bull. Kor. Chem. Soc.. 2020,
https://doi.org/10.1002/bkcs.12141
12. S. Manivannan, D.-K. Kang, K. Kim, Bull. Kor. Chem. Soc.
2019, 40, 297.
13. J. Lee, J. W. Ha, Bull. Kor. Chem. Soc. 2019, 40, 461.
14. S. Ahn, S. Manivannan, Y. Seo, Y. Kim, Bull. Kor. Chem.
Soc. 2019, 40, 522.
15. Y. S. Jeong, Y. J. Jang, S.-J. Park, Y. J. Lee, Bull. Kor.
Chem. Soc. 2019, 40, 642.
16. M. Arumugam, M. Y. Choi, Bull. Kor. Chem. Soc. 2020,
41, 304.
17. Hong, J. W, Bull. Kor. Chem. Soc. 2020, https://doi.org/10.
1002/bkcs.12125
18. U. P. M. Ashik, W. M. A. Wan Daud, H. F. Abbas, Renew.
Sust. Energ. Rev. 2015, 44, 221.
19. M. Pudukudy, Z. Yaakob, Z. S. Akmal, Appl. Surf. Sci. 2015,
353, 127.
20. A. K. Singh, X. Q, ChemCatChem 2013, 5, 652.
21. L. Azancot, L. F. Bobadilla, J. L. Santos, J. M. Córdoba,
M. A. Centeno, J. A. Odriozola, Int. J. Hydrog. Energy 2019,
44, 19827.
Experimental
22. H. Özdemir, M. A. Faruk Öksüzömer, M. Ali Gürkaynak, Int.
J. Hydrog. Energy 2010, 35, 12147.
23. T. Sakae, O. Hitoshi, Y. Ichiro, O. Kiyoshi, Appl. Catal. A
Gen. 2001, 217, 101.
Mesoporous silica and γ-alumina were purchased from
Sigma-Aldrich (St. Louis, MO, USA). Mesoporous MFI
zeolites are synthesized in the laboratory according to the
Bull. Korean Chem. Soc. 2020
© 2020 Korean Chemical Society, Seoul & Wiley-VCH GmbH
www.bkcs.wiley-vch.de
3

BULLETIN OF THE
Communication
KOREAN CHEMICAL SOCIETY
24. R. Couttenye, M. Devila, S. Suib, J. Catal. 2005, 233, 317.
25. G. Naresh, V. Vijay Kumar, C. Anjaneyulu, J. Tardio, S. K.
Bhargava, J. Patel, A. Venugopal, Int. J. Hydrog. Energy
2016, 41, 19855.
26. C. Jo, K. Cho, J. Kim, R. Ryoo, Chem. Commun. 2014, 50, 4175.
27. D. Pribat, C. Deniz, N. Karatepe, Y. H. Lee, M. Razeghi, S.
Baik, Carbon Nanotubes, Graphene, and Associated Devices
VI, SPIE Press, Bellingham, WA 2013.
28. C. Ragupathi, J. J. Vijaya, L. J. Kennedy, J. Saudi Chem.
Soc. 2017, 21, S231.
29. B. S. Liu, A. C. T, Appl. Catal. A Gen. 2003, 244, 181.
30. U. P. M. Ashik, W. M. A. Wan Daud, J.-i. Hayashi, Renew.
Sust. Energ. Rev. 2017, 76, 743.
31. J. Barbier, Appl. Catal. 1986, 23, 225.
32. P. Lingyu, L. Yongdan, C. Jiuling, C. Liu, Y. S. L. Jerry,
Catal. Today 2002, 74, 145.
Bull. Korean Chem. Soc. 2020
© 2020 Korean Chemical Society, Seoul & Wiley-VCH GmbH
www.bkcs.wiley-vch.de
4