Highly reactive Ni-Co/SiO2 bimetallic catalyst via complexation with oleylamine/oleic acid organic pair for dry reforming of methane

Article abstract of DOI:10.1016/j.cattod.2016.07.013

Ni-Co alloy supported on silica is prepared with oleylamine/oleic acid (OAm/OAc) organic pair by incipient wetness impregnation (IWI) method. The alloy formation is confirmed by XRD, TEM, EDX, TPR and XPS and the homogeneous alloy is proved by TEM-EDX (line, mapping and point ID). Compared with monometallic catalysts, in the Ni-Co alloy, the Ni metal size is controlled by the Co addition; also, the Ni-silica interaction is strengthened by the strong interaction between Co and silica; moreover, Ni sintering and oxidation are prevented by the electron transfer from Co. On the other hand, compared with the alloy prepared with single ligand (OAm or OAc), the Ni-Co alloy prepared with the combination of OAm and OAc is homogeneously distributed. This small, highly dispersed, homogeneous Ni-Co alloy strongly interacting with silica support leads to a high conversion of CO₂ and CH₄ which maintains stable over 30 h dry reforming of methane (DRM) reaction without carbon deposition or metal sintering.

Full text of DOI:10.1016/j.cattod.2016.07.013

G Model
Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
CATTOD-10322; No. of Pages9
ARTICLE IN PRESS
Catalysis Today xxx (2016) xxx–xxx
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Highly reactive Ni-Co/SiO2 bimetallic catalyst via complexation with
oleylamine/oleic acid organic pair for dry reforming of methane
Xingyuan Gao, Zhenwei Tan, Kus Hidajat, Sibudjing Kawi∗
Department of Chemical and Biomolecular Engineering, National University of Singapore, Singapore 117576, Singapore
a r t i c l e i n f o
a b s t r a c t
Article history:
Ni-Co alloy supported on silica is prepared with oleylamine/oleic acid (OAm/OAc) organic pair by incipient
wetness impregnation (IWI) method. The alloy formation is confirmed by XRD, TEM, EDX, TPR and XPS and
the homogeneous alloy is proved by TEM-EDX (line, mapping and point ID). Compared with monometallic
catalysts, in the Ni-Co alloy, the Ni metal size is controlled by the Co addition; also, the Ni-silica interaction
is strengthened by the strong interaction between Co and silica; moreover, Ni sintering and oxidation
are prevented by the electron transfer from Co. On the other hand, compared with the alloy prepared
with single ligand (OAm or OAc), the Ni-Co alloy prepared with the combination of OAm and OAc is
homogeneously distributed. This small, highly dispersed, homogeneous Ni-Co alloy strongly interacting
with silica support leads to a high conversion of CO2 and CH4 which maintains stable over 30 h dry
reforming of methane (DRM) reaction without carbon deposition or metal sintering.
Received 27 October 2015
Received in revised form 6 March 2016
Accepted 15 July 2016
Available online xxx
Keywords:
Ni-Co
Alloy
Oleylamine
Oleic acid
Dry reforming of methane (DRM)
© 2016 Elsevier B.V. All rights reserved.
1. Introduction
tion [27,28], we synthesize Ni, Co and Ni-Co supported on silica
catalysts. It is reported that the addition of Co could enhance the
Dry reforming of methane has garnered much attention due
to the discovery of shale gas technology in US, tightened con-
cern about the greenhouse effect and the demand for clean and
renewable energy [1–7]. Ni-based catalysts are more promising
considering the low cost and high reactivity [8–10]. However,
deactivation of catalysts caused by metal sintering and carbon
deposition hinders the commercialization of Ni catalysts [10–13].
Bimetallic catalysts are generally more reactivity and stable than
monometallic catalysts for dry reforming of methane [14–20].
Among various metals, cobalt is very promising to combine with Ni
as a bimetallic catalyst. It is reported that Ni-Co catalyst with differ-
ent supports show a high coke resistance and reactivity for a wide
range of reactions, such as methane steam reforming, dry reforming
of methane, steam reforming of ethanol, et al. The reason includes
the surface alloy formation, high metal dispersion, solid solution
formation and a smaller size [16,21–26]. However, the synthesis
of Ni-Co alloy supported on inert support (silica) and its appli-
cation in high temperature reactions are rarely reported. Due to
the low cost, high surface area, easy availability, high thermal sta-
bility, controllable pore texture and potential formation of metal
silicate compound which strengthens the metal-support interac-
coke resistance of Ni/Al2O3 catalyst for steam reforming of biomass
due to the synergetic effect of Ni-Co alloy formation and the opti-
mal molar ratio of Co/Ni is 0.25 [29]. Also, excess Ni (more than
80 mol%) will produce more carbon in the Ni-Co bimetallic cata-
lysts applied in DRM [17]. Therefore in this study, the molar ratio
(≈ weight ratio) of Ni to Co of the bimetallic Ni-Co catalyst prepared
is 4:1.
Previously in our group, oleic acid (OAc, C18H34O2) was effective
to increase the reactivity of Ni/SiO2 catalyst for DRM [28]. With
a same aliphatic chain but different functional group, oleylamine
(OAm, C18H35NH2) has been well investigated in the synthesis of
Ni nanoparticles where OAm acts as a surfactant, ligand and/or
reducing agent [30–37]. It was reported that the OAm/OAc organic
pair could control the size of the pure metal nanoparticles [38–42],
however, the combination of OAm-OAc pair with Ni-Co bimetallic
catalysts are rarely reported. Therefore in our work, we study
the effect of Ni-Co alloy formation on silica support and the
OAm/OAc organic pair by comparing five catalysts; (i) 5%Ni/SiO2
prepared with OAm/OAc pair {(OAm + OAc):Ni = (0.1 + 0.1) = 0.2
(molar ratio)}; (ii) 4%Ni + 1%Co/SiO2 prepared with OAm/OAc
pair {(OAm + OAc):Ni = (0.1 + 0.1) = 0.2}; (iii) 5%Co/SiO2 pre-
pared with OAm/OAc pair {(OAm + OAc):Ni = (0.1 + 0.1) = 0.2};
(iv) 4%Ni + 1%Co/SiO2 prepared with OAm (OAm:Ni = 0.2); (v)
4%Ni + 1%Co/SiO2 prepared with OAc (OAc:Ni = 0.2). The five sam-
∗
Corresponding author.
E-mail address: chekawis@nus.edu.sg (S. Kawi).
http://dx.doi.org/10.1016/j.cattod.2016.07.013
0920-5861/© 2016 Elsevier B.V. All rights reserved.
Please cite this article in press as: X. Gao, et al., Highly reactive Ni-Co/SiO2 bimetallic catalyst via complexation with oleylamine/oleic
acid organic pair for dry reforming of methane, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.07.013

G Model
Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
CATTOD-10322; No. of Pages9
ARTICLE IN PRESS
2
X. Gao et al. / Catalysis Today xxx (2016) xxx–xxx
ples are labled as (i) 5Ni-OAmc; (ii) 4Ni1Co-OAmc; (iii) 5Co-OAmc;
(iv) 4Ni1Co-OAm; (v) 4Ni1Co-OAc.
ing rate of 10 ◦C/min from room temperature to 900 ◦C in 80 ml/min
H2/He (5:95) gases.
Thermogravimetric/Differential Thermal Analysis (TG/DTA).
Carbon deposition experiments were conducted by TGA Shimadzu
DTG-50 under atmospheric pressure. The spent catalysts were
loaded in a quartz mesh cell and submitted to heat treatment from
room temperature to 900 ◦C with a ramping rate of 10 ◦C per minute
in 35 ml/min air flow. The% weight loss detected was caused by the
burning of the carbon in the catalyst to form carbon dioxide and lost
to the air. The% weight loss was converted to carbon deposition rate
on the catalyst.
X-ray photoelectron spectroscopy (XPS). Surface analysis of the
catalyst was performed using X-ray photoelectron spectroscopy
(XPS) from a KRATOS AXIS Hsi 165 equipped with Mg-K␣ source
(1253.6 eV). The samples were anchored on the standard sample
stub by double-sided adhesive tapes. The binding energy was ref-
erenced to C 1 s hydrocarbon peak at 284.5 eV.
2. Experimental section
2.1. Materials and preparation
5%Ni/SiO2 and 5%Co/SiO2 catalyst with OAm/OAc were prepared
by incipient wetness impregnation (IWI) method. Firstly, 1.3 g
Nickel(II) nitrate hexahydrate or Cobalt(II) nitrate hexahydrate
99.999% trace metals basis (from Sigma-Aldrich) was dissolved
in D.I. water; then oleylamine and oleic acid (molar ratio to Ni
is 0.1 for each) were added followed by sonication; 5 g spherical
silica (from Kanto Chemical Co.: particle size = 40–60 ␮m, specific
surface area = 753 m2/g, mean pore size = 7.5 nm) was introduced.
4%Ni + 1%Co/SiO2 catalysts with OAm, OAc and OAm/OAc were pre-
pared as above. Briefly, 1.04 g Ni nitrate salt and 0.26 g Co nitrate
salt were added into DI water; then OAm (molar ratio to Ni + Co is
0.2), OAc (molar ratio to Ni + Co is 0.2) and OAm/OAc (molar ratio
to Ni is 0.1 each) were mixed with the above solution under son-
ication, followed by the impregnation on 5 g silica gel. After that,
the samples were impregnated at room temperature for several
hours and dried overnight at 60 ◦C. All catalysts were calcined in
box furnace at 700 ◦C for 4 h.
2.3. Reactivity test
The dry reforming of methane (DRM) reactions was conducted
in a quartz reactor. The catalyst was loaded in the middle of
the quartz-tube packed with quartz wool on both sides. Under
atmospheric pressure, the catalyst was heated from room tem-
perature to a reaction temperature of 700 ◦C with a ramping rate
of 20.0 ◦C/min. Reactions were performed with Gas-Hourly-Space-
Velocity (GHSV) of 72,000 ml h−1 g−1(cat.). The molar ratio of feed
gas is CH4/CO2/N2 = 1/1/1. Prior to the catalytic reaction, the cata-
lysts were reduced in-situ in purified H2 (purity = 99.99%) at 700 ◦C
for 1 h. 0.05 g of the catalyst was tested. The effluent gas stream
was analyzed by an Agilent HP-GC equipped with two columns of
5A and Porapak Q.
2.2. Catalyst characterisation
X-ray Diffraction (XRD). The X-ray diffraction patterns of
reduced catalysts were collected in
a step-scan mode at
2␪ = 40◦–50◦ on a Bruker D8 Advance powder diffractometer, where
Cu target K␣-ray (operating at 40.0 kV and 30.0 mA) was used as
the X-ray source. The reduced sample was prepared by reducing
the fresh catalyst in purified H2 under 700 ◦C for one hour.
Transmission Electron Microscopy (TEM). Reduced samples
were first reduced in 10 ml/min of pure H2 gas at 700 ◦C for 1 h
for TEM characterization. The images of the reduced and spent cat-
alysts were obtained using a JEOL JEM-2010 electron microscope.
Reduced samples were further characterized by an OXFORD INCA
EDS analyzer which was operated in the range of 100–400 keV.
N2O pulse Chemisorption. N2O chemisorption experiments
were conducted using Thermo Scientific TPDRO 1100 series sys-
tem equipped with a thermal conductivity detector (TCD). 0.05 g of
fresh catalyst was reduced at 700 ◦C for one hour with a flow of 5%
H2/N2 of 30 ml/min with a ramping rate of 20 ◦C/min. After cooling
down, pulse chemisorption was performed at 90 ◦C with nitrogen
as the carrier gas and N2O as the pulse gas. After chemisorption of
N2O, the sample was treated by a temperature programmed reduc-
tion until 700 ◦C under 30 ml/min of 5% H2/N2 with a ramping rate
of 10 ◦C/min. The metal dispersion, specific surface area and particle
size were provided automatically.
3. Results and discussion
3.1. Effect of addition of cobalt into Ni/SiO2 with OAm/OAc
organic pair
In order to show the effect of cobalt in Ni/SiO2 catalyst, we
prepare three catalysts: 5Ni-OAmc, 4Ni1Co-OAmc and 5Co-OAmc.
These three samples are all prepared with OAm/OAc pair so it is fair
to compare the mono- and bimetallic samples.
3.1.1. XRD results
Fig. 1 shows a peak at 44.43◦ in 5Ni-OAmc which refers to metal-
lic Ni reflection and a peak at 44.26◦ in 5Co-OAmc which refers
to metallic Co reflection [17,21,24,29]. In 4Ni1Co-OAmc, one peak
between 44.43◦ and 44.26◦ is shown, which suggests the NiCo alloy
formation. Similar results are also found in Ni-Co with different
The surface areas of the catalysts were determined by the
Micromeritics ASAP 2010 system. The surface area was calculated
using Brunauer–Emmett–Teller (BET) method using the adsorption
data. Nitrogen gas was utilized as the probe gas to estimate the pore
size of meso-pores under the liquid nitrogen temperature of 77 K.
Before analysis, 0.05 g of nano-catalysts was degased at 300 ◦C for
12 h to remove the moisture. The catalyst weight was quantified by
measuring before and after degasing of the sample.
H2-Temperature Programmed Reduction (H2-TPR). Hydrogen-
temperature programmed reduction (H2-TPR) experiments were
conducted using Quantachrome ChemBET 3000 TPD/TPR system
equipped with a thermal conductivity detector (TCD). Fresh cat-
alysts (30 mg) were loaded into a glass U-tube secured by quartz
wool and placed inside a furnace. They were heated with the ramp-
Fig. 1. XRD patterns of reduced catalysts: (a) 5Ni-OAmc01; (b) 4Ni1Co-OAmc01; (c)
5Co-OAmc01.
Please cite this article in press as: X. Gao, et al., Highly reactive Ni-Co/SiO2 bimetallic catalyst via complexation with oleylamine/oleic
acid organic pair for dry reforming of methane, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.07.013

G Model
Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
CATTOD-10322; No. of Pages9
ARTICLE IN PRESS
X. Gao et al. / Catalysis Today xxx (2016) xxx–xxx
3
Fig. 2. TEM images of reduced catalysts: (a) 5Ni-OAmc01; (b) 4Ni1Co-OAmc01; (c) 5Co-OAmc01.
supports [17,21,24,29]. Saw [43] also found the similar trend in
3.1.4. N2O-chemisorption results and BET data
NiCu bimetallic catalyst, which is attributed to the NiCu alloy for-
mation.
It is clearly shown that compared with the monometallic
catalysts, Ni-Co alloy has the smallest particle size, which is incon-
sistence with the XRD and TEM results. Also, Ni-Co alloy has the
highest metal dispersion and specific surface area according to both
chemisorption and BET data, which is beneficial in promoting the
reactivity. As for the much lower metal dispersion and surface area,
it is because at 700 ◦C only a few Co is reduced so that the active
sites are limited.
3.1.2. TEM results (reduced)
Fig. 2 shows that 4Ni1Co-OAmc has a smaller metal particle
size after reduction than monometallic catalysts. The average of
the Ni-Co alloy is 3.1 nm, smaller than Ni (3.5 nm) and Co (3.9 nm).
Formation of an alloy can decrease the particle size. Since Co is dif-
ficult to be reduced from the solid solution, the alloy particle size is
effectively controlled by the Co [12]. A smaller metal size produces
a higher surface area and more active sites for the reaction [44].
3.1.5. XPS resulsts
It is shown in Fig. 4(a) and Table 2(a) that when Ni-Co alloy
is formed, the binding energy of Ni 2p3/2 shifts from 856.4 eV to
856.1 eV, which means Ni withdraws electrons and the electron
density is higher; Fig. 4(b) and Table 2(b) show that compared with
5Co-OAmc, the binding energy of Co 2p shifts from 782.0 eV to
782.4 eV in the Ni-Co alloy, which means Co donates electrons. The
binding energy change of Ni and Co indicates the intimate inter-
action between Ni and Co in the Ni-Co alloy [16,21,22,26]. Also
it is reported that due to the surface segregation, Co can modify
the Ni surface electronically [19]. Due to the electron donation of
neighbor Co, a higher electron density of Ni is formed, which can
protect Ni from being oxidized and sintering during the reaction
[16,17,45–49].
3.1.3. TEM-EDX results
We can see from Line-EDX that the intensity of Ni and Co
changes simultaneously; in other words, Ni and Co co-exist in one
particle which clearly indicates the formation of Ni-Co alloy. This
is also proved by the mapping results that the signal of Ni and Co
well overlap each other [21]. The intensity difference is due to the
different loadings of both metals. This result is in accordance with
the XRD results that Ni-Co alloy is well formed (Fig. 3).
Fig. 3. TEM-EDX line and mapping results of reduced 4Ni1Co-OAmc sample.
Please cite this article in press as: X. Gao, et al., Highly reactive Ni-Co/SiO2 bimetallic catalyst via complexation with oleylamine/oleic
acid organic pair for dry reforming of methane, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.07.013

G Model
Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
CATTOD-10322; No. of Pages9
ARTICLE IN PRESS
4
X. Gao et al. / Catalysis Today xxx (2016) xxx–xxx
Fig. 5. TPR profile of fresh catalysts.
Fig. 4. XPS profile of fresh catalysts: (a) Ni 2p; (b) Co 2p.
3.1.6. TPR results
We can see two major peaks in 5Ni-OAmc. One peak is at around
500 ◦C, referring to the NiO weakly interacting with the support; the
other one is at around 700 ◦C, referring to the reduction of Ni from
silicate framework [50–52]. In 5Co-OAmc, the reduction tempera-
ture is beyond 800 ◦C, which suggests the very strong interaction
between Co ions and silica matrices [53]. The strong interaction
between Co and silica is also proved by XPS that the binding energy
of Co 2p in both 4Ni1Co-OAmc and 5Co-OAmc is equal to or higher
than 782.0 eV, which is reported to be the evidence of the strong
interaction of Co ion with silica support even the incorporation of
Co ion in the silica framework [53–57]. With the addition of Co into
Ni (4Ni1Co-OAmc), the low temperature peak becomes not obvi-
ous while the higher temperature peak becomes relatively larger;
moreover, both peaks shift to higher temperatures compared with
Ni alone (5Ni-OAmc) but still lower than the pure Co (5Co-OAmc).
In 4Ni1Co-OAmc, the reduction temperature between pure Ni and
pure Co again proves the interaction between Ni and Co [20]. Fur-
thermore, the enhancement of the reduction temperature in Ni-Co
alloy compared with Ni alone is attributed to the very strong inter-
action between Co and silica. This similar promotional effect of
Co on enhancing the Ni-Co interaction with the support is also
reported in other works [22–24]. This strong MSI is beneficial to the
reactivity. If a strong metal-support interaction exists, the partially
reduced support will migrate on the surface of metal and create
highly active sites on the boundary between metal and support
[17,18,56,58,59]; in the meanwhile, the support will destroy the
large metal particles and improve the metal dispersion [17], leading
to a high and stable catalytic performance (Fig. 5).
Fig. 6. Catalytic performances for dry reforming of methane reaction. Reaction con-
ditions: 700 ◦C, GHSV = 72,000 ml h−1 g-1(cat.), Wcat = 0.05 g, CO2:CH4:N2 = 1:1:1.
alloy is attributed to the smaller size, higher metal dispersion and
surface area, stronger MSI and the electron-rich environment for Ni.
As for the very low conversion in 5Co-OAmc, it is because the metal
is not fully reduced at 700 ◦C so that the active sites are very lim-
ited, which is in consistence with the N2O-chemisorption results
(Fig. 6).
3.1.8. TGA results
From Table 3 we can see that Ni alone produces carbon while
Ni-Co alloy has no carbon over 30 h reaction. The small size and
strong metal-support interaction cause this excellent coke resis-
tance, which maintains the high stability and reactivity. As for
5Co-OAmc, it is reasonable that no carbon is formed because the
reactivity is very low so that the methane decomposition is very
limited.
3.1.7. Catalytic performances
The Ni-Co alloy performs the best that over 30 h reaction time,
CO2 and CH4 conversion maintains at around 82% and 79% respec-
tively. While for Ni alone, CO2 drops from 81% to 77% and CH4 drops
from 78% to 74% over 30 h. The better catalytic stability of Ni-Co
Please cite this article in press as: X. Gao, et al., Highly reactive Ni-Co/SiO2 bimetallic catalyst via complexation with oleylamine/oleic
acid organic pair for dry reforming of methane, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.07.013