Well-dispersed nickel nanoparticles on the external and internal surfaces of SBA-15 for hydrocracking of pyrolyzed α-cellulose

Article abstract of DOI:10.1039/c8ra09034c

Catalysts comprising nickel supported on SBA-15 were prepared by wet impregnation and co-impregnation methods. Wet impregnation was performed by directly dispersing an Ni(NO₃)₂·6H₂O aqueous solution into SBA-15, whereas in co-impregnation, ethylene glycol (EG) was added to nickel nitrate aqueous solution prior to dispersion into SBA-15. After drying and calcination, NiO/SBA-15w and NiO/SBA-15c were produced. Later, after the reduction process, Ni/SBA-15w and Ni/SBA-15c were obtained. The prepared catalysts were evaluated for the hydrocracking of pyrolyzed α-cellulose. The TEM images revealed that the catalysts prepared by wet impregnation showed inhomogeneous distribution of nickel loading, whereas catalysts prepared by co-impregnation using EG exhibited homogeneous distribution and formed no nickel aggregates. During hydrocracking of pyrolyzed α-cellulose, Ni/SBA-15c with total acidity, nickel loading, particle size, and specific surface area of 7.27 m mol g^?1, 5.20 wt%, 3.17 nm, and 310.0 m² g^?1, respectively, exhibited the best catalytic performance compared to other prepared catalysts with 67.35 wt% conversion of liquid product with maximum selectivity in producing 13.09 wt% of 3-methyl-pentane. Moreover, Ni/SBA-15w with total acidity, nickel loading, particle size, and specific surface area of 10.87 m mol g^?1, 8.15 wt%, 7.01 nm, and 628.0 m² g^?1, respectively, produced 69.89 wt% liquid product without hydrocarbons. Study of selectivity towards the formation of liquid hydrocarbons was carried out via double step hydrocracking using Ni/SBA-15w, and 18.55 wt% of n-hexane was produced in the liquid product.

Full text of DOI:10.1039/c8ra09034c

RSC Advances
View Article Online
View Journal | View Issue
PAPER
Well-dispersed nickel nanoparticles on the external
and internal surfaces of SBA-15 for hydrocracking
of pyrolyzed a-cellulose†
Cite this: RSC Adv., 2019, 9, 1230
*
Wega Trisunaryanti, Endah Suarsih,
Triyono and Iip Izul Falah
Catalysts comprising nickel supported on SBA-15 were prepared by wet impregnation and co-impregnation
methods. Wet impregnation was performed by directly dispersing an Ni(NO3)2$6H2O aqueous solution into
SBA-15, whereas in co-impregnation, ethylene glycol (EG) was added to nickel nitrate aqueous solution
prior to dispersion into SBA-15. After drying and calcination, NiO/SBA-15w and NiO/SBA-15c were
produced. Later, after the reduction process, Ni/SBA-15w and Ni/SBA-15c were obtained. The prepared
catalysts were evaluated for the hydrocracking of pyrolyzed a-cellulose. The TEM images revealed that
the catalysts prepared by wet impregnation showed inhomogeneous distribution of nickel loading,
whereas catalysts prepared by co-impregnation using EG exhibited homogeneous distribution and
formed no nickel aggregates. During hydrocracking of pyrolyzed a-cellulose, Ni/SBA-15c with total
acidity, nickel loading, particle size, and specific surface area of 7.27 m mol gꢀ1, 5.20 wt%, 3.17 nm, and
310.0 m2
g
ꢀ1, respectively, exhibited the best catalytic performance compared to other prepared
catalysts with 67.35 wt% conversion of liquid product with maximum selectivity in producing 13.09 wt%
of 3-methyl-pentane. Moreover, Ni/SBA-15w with total acidity, nickel loading, particle size, and specific
g
surface area of 10.87 m mol gꢀ1, 8.15 wt%, 7.01 nm, and 628.0 m2 ꢀ1, respectively, produced 69.89 wt%
Received 31st October 2018
Accepted 20th December 2018
liquid product without hydrocarbons. Study of selectivity towards the formation of liquid hydrocarbons
was carried out via double step hydrocracking using Ni/SBA-15w, and 18.55 wt% of n-hexane was
produced in the liquid product.
DOI: 10.1039/c8ra09034c
rsc.li/rsc-advances
formation and sintering is an ongoing process. It has been re-
ported that the types of metal and support play an important
role in increasing the efficiency of the conversion process.8
1. Introduction
Green energy production will continue indenitely with
a constant supply of biomass waste. The valorization of biomass
waste has gained much interest because it has become one of
the most promising renewable solutions for the substitution of
fossil resources not only in energy applications but also for the
production of many chemicals and biomaterials.1 One of the
many advantages of biomass-based-fuels is that they reduce
emissions of gases such as CO2, SO2, and NOx and contribute to
global economic sustainability.2–5 Biomass provides a prom-
ising stock for producing renewable fuels and chemicals,
especially for liquid fuels.6,7 However, due to the complexity of
biomass conversion, products are typically obtained with very
low yield and purity. Therefore, upgrading treatments are
required to obtain an excellent conversion process.
Mesoporous catalysts can improve the development of new
biofuel generation processes because of their important prop-
erties such as high concentration of active sites, high thermal/
hydrothermal stability and enhanced shape selectivity.9 One of
the most widely used mesoporous supports is SBA-15 and it has
a high specic surface area, excellent thermal stability, tunable
porosity, and uniform hexagonal pores and thick pore walls.10
Among the mesoporous materials, SBA-15 with these advan-
tages has attracted much more attention for applications as
catalysts as well as catalyst supports.11,12
Supported nickel catalysts have attracted signicant atten-
tion for their extensive applications in various industrial
processes.13 During biomass conversion, nickel-based catalysts
upgrade the composition of the liquid phase.8 However, nickel-
based catalysts are more easily deactivated because of carbon
deposition and active metal species sintering; during harsh
reactions, the tendency of carbon deposition hinders stable
catalytic performance.14 Designing catalysts with strong metal-
support interactions and well connement of metal particles
is one possible solution for this drawback.10 Small nickel
nanoparticles with strong metal-support interactions reportedly
Cellulosic conversion into fuels is challenging and requires
effective catalytic systems. The search for catalysts with excel-
lent catalytic activity, selectivity, and resistance to coke
Department of Chemistry, Universitas Gadjah Mada, Indonesia. E-mail: wegats@ugm.
ac.id
† Electronic supplementary information (ESI) available. See DOI:
10.1039/c8ra09034c
1230 | RSC Adv., 2019, 9, 1230–1237
This journal is © The Royal Society of Chemistry 2019

View Article Online
Paper
RSC Advances
have limited mobility on the surface and resistance to sinter- using the same reactor at 600 ꢁC for 3 hours and streamed by N2
ing.10,15 A simple wet impregnation process is generally used as gas with 20 ml minꢀ1 ow rate. In the hydrocracking of pyro-
a preparation method. However, this method reportedly causes lyzed a-cellulose, the catalyst and pyrolyzed a-cellulose with the
poor dispersions due to agglomeration during drying and ratio of 1 : 30 were placed into the semi-batch reactor. The
calcination.16 To obtain highly active and dispersed nickel reactor containing the catalyst and pyrolyzed a-cellulose was
catalysts by the impregnation method, several approaches and streamed by H2 gas with 20 ml minꢀ1 ow rate and heated at
techniques have been developed such as using an organic nickel 450 ꢁC for 2 hours. The heart-shaped ask was used to place the
precursor and surface modication.17,18
liquid product. The liquid product was characterized by using
Formerly, it was reported that the polyol-assisted route (co- Gas Chromatography-Mass Spectroscopy (Shimadzu QP2010S).
impregnation) of Ni/SBA-15 catalyst with high dispersion of
small nickel particles results in more active sites for the reac-
2.3 Catalyst characterization
tion, leading to higher catalytic activity.19 In addition, small
The acidity of the catalysts was tested by the gravimetric method
nickel particles with strong metal-support interactions
using NH3 gas as a basic adsorbate. The amount of nickel
prepared via co-impregnation signicantly enhanced the
present in the catalyst was determined by using an atomic
dispersion, resulting in high activity and excellent coke resis-
absorption spectrophotometer (Perkin Elmer PinAAcle 900 T).
The amorphous structure of SBA-15 was analyzed by an X-ray
Diffractometer (Rigaku Miniex 600) using Cu Ka radiation
tance.20,21 A nickel catalyst supported on SBA-15 via co-
impregnation using ethylene glycol was also reportedly
synthesized and it showed excellent stability in the dry
˚
(0.154060 A). The measurement conditions were in the range of
reforming of methane with CO2 to produce syngas and
2q ¼ 20–80ꢁ. The particle sizes of Ni and NiO can be calculated
hydrogen.22 Ni/SBA-15 also showed high activity and achieved
by using the Scherrer equation. The morphology of catalyst
high hydrocarbon yields in the hydrodeoxygenation reaction for
surface was characterized by using Scanning Electron
upgrading lignin-derived bio-oils.23
Microscope-Energy Dispersive Spectroscopy (JEOL JED 2300)
Based on these considerations, nickel catalysts supported on
with an electron beam of 15.0 kV. TEM images were determined
by a JEOL-JEM-1400 microscope with an electron beam of 120
kV. The pore characteristics (Brunauer–Emmett–Teller) of the
catalysts were tested by using a gas adsorption analyzer
SBA-15 were synthesized via wet impregnation and co-
impregnation using ethylene glycol. The catalytic performance
of nickel catalysts in reduced and oxide forms was evaluated for
the hydrocracking of pyrolyzed a-cellulose as a model of
(Quantachrome NovaWin version 11.0). The pore analysis was
biomass. The selectivity of catalysts toward the formation of
liquid hydrocarbons was studied through double step
conducted using N2 gas at 77.3 K.
hydrocracking.
3. Results and discussion
2. Experimental
3.1 Catalyst characterization
2.1 Catalyst preparation
During impregnation, nickel salt solution was assumed to be
adsorbed easily by mesoporous SBA-15 due to capillary pres-
sure.24 In fact, the SBA-15 support is not completely hydrophilic,
which can cause incomplete wetting and inhomogeneous nickel
dispersion on pores during impregnation.25 The addition of
ethylene glycol into the nickel nitrate aqueous solution has
been reported to decrease the surface tension of impregnation
solutions because it can act as a surfactant, which is preferred
for better wetting of the support and increasing the rate of
diffusion and infusion into deeper mesoporous channels of the
catalyst support.22 During the drying process, the solvent was
gradually removed. At this stage, the concentrated solvent with
lower surface tension could diffuse deeply at rst and was
transported toward the internal surface of the mesoporous
channels via capillary forces.22 At the end of drying, the catalyst
exhibited strong interaction with the internal surface of the
support and inhibited the redistribution of the impregnation
solution; also, uniform distribution of the active component in
the mesopore channels could be achieved. In contrast, the
SBA-15 as a support material was supplied by Green Stone Swiss
Co. Ltd. and Ni(NO3)2$6H2O as a salt precursor was supplied by
Merck. Nickel supported on SBA-15 catalysts were set as
10.0 wt% and were prepared by the wet and co-impregnation
methods.20 The only difference between the wet and co-
impregnation methods was the addition of a dened amount
of ethylene glycol (Ni/EG ratio ¼ 1) during the dissolution of the
salt precursor. The mixtures were stirred and impregnated in
the SBA-15 support, followed by standing overnight; they were
evaporated and dried at 100 ꢁC. Aerwards, the dried solids
were calcined at 500 ꢁC for 5 hours in the air. At this stage, the
solids were denoted as NiO/SBA-15c and NiO/SBA-15w. A further
step was the reduction process, where the solids were streamed
by 20 ml minꢀ1 H2 gas at 400 ꢁC for 3 hours to prepare zero
valences of nickel (denoted as Ni/SBA-15c and Ni/SBA-15w).
2.2 Catalyst test
The hydrocracking of pyrolyzed a-cellulose was carried out by precursor exhibited weak interaction with the support without
using a set of cracking reactor. A semi-batch reactor system EG; the concentrated solvent adsorbed by the mesoporous pores
using stainless steel tube with an inner diameter of 2.80 cm, could be back diffused by higher surface tension and it
outer diameter of 3.10 cm, and length of 21.00 cm was used. migrated toward the external surface of the mesoporous chan-
Prior to the hydrocracking process, a-cellulose was pyrolyzed nels, resulting in a concentration build up on the external rim of
This journal is © The Royal Society of Chemistry 2019
RSC Adv., 2019, 9, 1230–1237 | 1231

View Article Online
RSC Advances
Paper
Table 1 Physical and chemical properties of catalysts
Nickel
Acidityb
Pore volumec
Average pore
Ni species
loadinga (wt%)
(m mol gꢀ1
)
SBET (m2 gꢀ1
)
(cm3 gꢀ1
)
diameterc (nm)
particle sized (nm)
c
Samples
SBA-15
—
3.99
4.85
5.06
10.87
7.27
623.80
419.80
422.20
628.00
310.00
1.05
0.80
0.86
0.88
0.64
6.75
7.66
8.19
5.60
8.24
—
NiO/SBA-15w
NiO/SBA-15c
Ni/SBA-15w
Ni/SBA-15c
5.07
5.25
8.15
5.20
9.05
2.91
7.01
3.17
a
Nickel loadings were determined by an atomic absorption spectrophotometer. b Acidity of the catalysts was determined gravimetrically using NH3
gas as a basic adsorbate. c SBET and pore volumes and diameters of the catalysts were calculated using BET and BJH theory, respectively. d Particle
size was calculated using Scherrer's equation.
the support. As a consequence, inhomogeneous distributions of splitting of pores in the Ni/SBA-15w catalyst since it has the
the precursor were observed.22
lowest pore diameter. As seen in Table 1, Ni/SBA-15w has the
The loading of Ni on SBA-15 affected the acidity of the highest loading of nickel, which might be affected by the
catalyst. As presented in Table 1, the presence of nickel impregnation process; thus, the splitting of pores cannot be
increased the acidity of the catalysts and as the concentration of avoided. As a consequence, it can increase the specic surface
nickel increased, the acidity increased. This was due to the area due to the decrease in average pore diameter.
Lewis acid (electron acceptor) sites provided by nickel.
The high-angle XRD patterns of the catalysts are shown in
Comparing the oxide form and zero valences of the catalysts, it Fig. 1. In all high-angle diffractograms, it is shown that the
was shown that the reduced form of nickel (Ni) supported on characteristic peak of SBA-15 corresponded to 2q of approxi-
SBA-15 exhibited higher acidity than the oxide form of nickel mately 20–30ꢁ, resulting in a broad peak. Aer impregnation of
(NiO). This was because the reduced form of nickel has more nickel into SBA-15, it was clearly shown that the characteristic
orbitals to accept electrons, whereas the oxide form of nickel peaks of SBA-15 are relatively similar. This result indicates that
has less orbitals because of the presence of oxygen.
aer the impregnation of nickel into SBA-15, the structure of
The physical properties of the catalysts, as presented in SBA-15 was not damaged.
Table 1, show that the loading of nickel on SBA-15 decreases the
The impregnation of nickel into SBA-15 apparently results in
pore volume due to shrinkage of silica walls by the heat treat- peaks that correspond to the presence of NiO and Ni in the
ment or partial blockage of mesoporous channels by nickel catalysts. Catalysts prepared via the wet impregnation method
nanoparticles. The impregnation of nickel on SBA-15 also showed three strong and sharp diffraction peaks at 2q of around
decreased the specic surface area of the catalysts except for Ni/ 37ꢁ, 43ꢁ and 63ꢁ, indicating that nickel species were present in
SBA-15w. This may have been caused by the splitting of pores by the form of cubic NiO (JCPDS no. 01-1239); the peak at 2q of
the aggregates of impregnated Ni during the impregnation around 42.7ꢁ indicated that nickel species were present in the
process. The average pore diameter may also conrm the form of cubic Ni (JCPDS no. 02-0850). Furthermore, catalysts
Fig. 1 Wide angle XRD patterns of catalysts (a) SBA-15, (b) Ni/SBA-15w, (c) Ni/SBA-15c, (d) NiO/SBA-15w, and (e) NiO/SBA-15c.
1232 | RSC Adv., 2019, 9, 1230–1237
This journal is © The Royal Society of Chemistry 2019

View Article Online
Paper
RSC Advances
prepared via co-impregnation showed weak and diffuse
diffraction peaks of NiO and Ni. This conrmed that NiO and Ni
particles were very small and highly dispersed on the SBA-15
support. The average particle size of NiO was evaluated by the
Scherrer equation. The average particle sizes for NiO/SBA-15w,
NiO/SBA-15c, Ni/SBA-15w, and Ni/SBA-15c were 9.05 nm,
2.91 nm, 7.01 nm, and 3.17 nm, respectively. By this evaluation,
it was conrmed that NiO particles prepared via co-
impregnation have smaller particle sizes and possibly supe-
rior catalytic performance. In addition, the characteristic peaks
of NiO that were also observed for the reduced form of catalysts
(Ni/SBA-15w and Ni/SBA-15c) were caused by the incomplete
reduction process.
Brunauer–Emmet–Teller (BET) theory explains the physical
adsorption of gas molecules on a solid surface. It applies to the
system of multilayer adsorption and usually utilizes probing
gases that do not chemically react with material surfaces as
adsorbates to quantify the specic surface area. Nitrogen is the
most commonly employed gaseous adsorbate used for surface
probing by BET methods. For this reason, the standard BET
analysis is mostly conducted at the boiling temperature of N2
(77 K). As presented in Fig. 2, the nitrogen adsorption/
desorption isotherms are typical of a reversible type IV
isotherm with a hysteresis loop based on IUPAC classication
scheme for mesoporous materials with narrow pore size
distribution of the cylindrical pores.
The pore size distributions are presented in Fig. 3, and they
were analyzed by using BJH desorption data. By processing the
data, we inferred that the best pore distribution was observed
for SBA-15, which was indicated by the narrowest peak formed
due to its homogeneous pore size. Aer the impregnation
process, there were several changes in terms of pore distribu-
tion. As seen in Fig. 3, the pore distribution of catalysts aer
impregnation of nickel shied to the right, indicating that pore
diameters have changed. These changes might be caused by the
shrinkage of SBA-15 wall due to the heating treatment or the
formation of new pores as an effect of the impregnation
process. Therefore, the diameter changes of the pores will vary
Fig. 3 Pore distributions of catalysts.
the diameter size and the pores will become more
heterogeneous.
The SEM images of the catalysts are shown in Fig. 4. These
images reveal that the SBA-15 morphology consisted of short
rod-like aggregates. Aer impregnation, it can be seen in Fig. 4
that the SEM image looks brighter. This was due to the presence
of nickel. In SEM, when an electron is ejected to the specimen
that consists of a metal, the secondary electron will be captured,
making the image look brighter due to the physical properties
of the metal. Comparing the SEM images of the nickel catalysts
prepared by wet impregnation and co-impregnation (Fig. 4), it
was clearly shown that there were more nickel agglomerations
during wet impregnation on the external surface of SBA-15, as
indicated by the rough surface. The Ni/SBA-15 catalyst with
most loaded nickel tended to form aggregates rapidly during
calcination and reduction processes, leading to the formation
of severe agglomeration (Fig. 4D).
Catalysts prepared by co-impregnation using ethylene glycol
showed fewer nickel aggregates, as conrmed by Fig. 4C and E.
These SEM images conrm that the addition of ethylene glycol
may resist the formation of aggregates in the catalyst, which can
be a drawback. These SEM images correspond to Scherrer's
evaluation of the average size of nickel particles present in the
catalysts based on X-ray diffraction data. The aggregates from
nickel loading caused increase in peak intensities due to large
nickel particles.
Nickel particles with smaller sizes and stronger metal-
support interactions are favored due to their better resistance
on sintering and coke formation. Nanoparticles with high
surface areas tend to minimize the aggregation energy. Large
nickel particles may be obtained from the sintering of nickel
nanoparticles that have loose contact with the support due to
weak metal-support interaction.
The above results are conrmed by TEM images of the
catalysts, as shown in Fig. 5. The typical structure of SBA-15
observed was an ordered mesoporous structure. The TEM
images of catalysts prepared via wet and co-impregnation show
a very clear distinction. For the catalysts prepared via wet
Fig. 2 Nitrogen adsorption/desorption isotherms of N2 gas of (a) SBA-
15, (b) Ni/SBA-15c, (c) Ni/SBA-15w, (d) NiO/SBA-15c, (e) NiO/SBA-15w
catalysts.
This journal is © The Royal Society of Chemistry 2019
RSC Adv., 2019, 9, 1230–1237 | 1233

View Article Online
RSC Advances
Paper
Fig. 4 SEM images of (A) SBA-15, (B) NiO/SBA-15w, (C) NiO/SBA-15c, (D) Ni/SBA-15w, and (E) Ni/SBA-15c catalysts.
impregnation, NiO is usually present in a form of bulk oxide in place. This connement effect resulted in the homogeneous
and is difficult to deliver to the channels, which may cause and well dispersion suppressing the growth of metal particles,
aggregation on the external surface of the support.10 Fig. 5B and which could improve resistance to coking and sintering.
D show the large aggregates of nickel particles on the external
surface of SBA-15 due to their inability to enter the channels. In
3.2 Catalytic activity in the hydrocracking of pyrolyzed a-
this case, small nickel particles only lled a small part of the
cellulose
channels because of the inhomogeneous distributions of nickel
particles that were formed due to inability of water as a solvent In the investigation of catalytic activity, the prepared catalysts
to suppress the redistribution and agglomeration during were used in the hydrocracking of pyrolyzed a-cellulose. Prior to
drying, calcination, and reduction processes; this result corre- the hydrocracking process, a-cellulose should be pyrolyzed to
sponded to previous results.22 Meanwhile, on catalysts prepared obtain the liquid form of a-cellulose. Pyrolysis was carried out
via co-impregnation, as seen in Fig. 5C and E, nickel particles by heating a-cellulose at 600 ꢁC while being streamed by N2 gas
were anchored inside the channels of SBA-15 due to the pres- (20 cm3 minꢀ1). The pyrolysis products included a carbon-rich
ence of EG, which could conne the small nickel nanoparticles solid called char, an incondensable gas fraction and
Fig. 5 TEM images of (A) SBA-15, (B) NiO/SBA-15w, (C) NiO/SBA-15c, (D) Ni/SBA-15w, (E) Ni/SBA-15c catalysts.
1234 | RSC Adv., 2019, 9, 1230–1237
This journal is © The Royal Society of Chemistry 2019

View Article Online
Paper
RSC Advances
Table 2 Major products of pyrolyzed a-cellulosea via pyrolysis based
on GC-MS data
a hydrocracking reactor. This hydrogenation process can
simultaneously be called as hydrodeoxygenation because at this
stage, hydrogenation as well as deoxygenation reactions can
occur. The term hydrocracking was used because during the
heating process, the components of bio-oil tend to be poly-
merized, which was then cracked by H2 gas to obtain the desired
product. This catalytic activity test at this stage was done by
using the prepared catalysts. The percentages of product
distribution in the hydrocracking process are shown in Table 3.
The investigation of the catalyst performance on hydro-
cracking of pyrolyzed a-cellulose is shown in Table 3. The
results showed that the hydrocracking process resulted in three
phases of the product, which were a liquid fraction, gas fraction,
and coke. As expected, the hydrocracking process with a poor
conversion value of the liquid product was thermal hydro-
cracking without a catalyst. The catalytic cracking over catalysts
resulted in more liquid products, and the thermal cracking
resulted in more gas products. This result could be caused by
the differences in the reaction mechanisms.27,28 In thermal
cracking, the reaction occurs through the formation of radical
ions triggered by a high temperature without a catalyst to
produce short carbon chains, whereas in catalytic hydro-
cracking, the reaction takes place via the formation of carbo-
cations to produce longer carbon chains.
The highest conversion value of the liquid product was ob-
tained by hydrocracking using Ni/SBA-15w as a catalyst. This
was due to the highest loading of nickel present in the catalyst.
By looking at the percentage of liquid product obtained and the
nickel content in the catalyst comparatively, it was inferred that
Ni/SBA-15w gives the highest yield of the liquid product
compared to the other prepared catalysts with the nickel
content of 8.15%. Ni/SBA-15c obtained by co-impregnation gave
67.35% yield of liquid product with only 5.20% of nickel present
in the catalyst. These results show that Ni/SBA-15 prepared by
co-impregnation can be compared to the one prepared by wet
impregnation in terms of liquid product yields with less amount
of nickel present in the catalyst. However, in the view of liquid
fraction per Ni loading also presented in Table 3, NiO/SBA-15w
exhibited the maximum value of 13.04 with no selectivity
towards the formation of hydrocarbons. The value of liquid
fraction per Ni loading was obtained from the division of the
yield of the liquid fraction by Ni loading.
Content (wt%)
Name
Mol. formula
7.20
9.44
6.82
10.27
2.51
2.78
4.49
2.12
Formic acid
2,3-Butanedione
Acetic acid
1-Hydroxy-2-propanone
Oxirane (butoxymethyl)
1-Hydroxy-2-butanone
2-Furancarboxaldehyde
2-Furanmethanol
CH2O2
C4H6O2
C2H4O2
C3H6O2
C7H14O2
C4H8O2
C5H4O2
C5H6O2
a
Yield of pyrolyzed a-cellulose (liquid) is 45.42 wt%.
a condensable vapor fraction composed of H2O and a complex
mixture of organic compounds (Table 2).
As expected, the result of the pyrolysis process was the
conversion of a-cellulose into oxygenated compounds. The
resulting compounds were analyzed by using gas
chromatography-mass spectrometry; they predominantly con-
tained carboxylic acid compounds (formic acid; acetic acid),
ketones (2,3-butanedione; 1-hydroxy-2-propanone; 1-hydroxy-2-
butanone), and furan compounds (2-furancarboxaldehyde; 2-
furanmethanol). This result corresponds to that of a former
investigation.26 The mixture of these compounds was called as
bio-oil. Bio-oils are generally presented as dark brown organic
liquids, which have a special smoky odor. The obtained bio-oil
was a yellow clear liquid with a very strong odor. With time, this
yellow clear liquid turned dark brown and the presence of
particles could be observed. This occurred because of the
polymerization process during the aging of oil and also due to
the oxidation process, which might occur because of the contact
with oxygen. Based on these results, we inferred that bio-oil was
composed of unstable compounds.
The upgrading of bio-oil was conducted by the hydro-
cracking process using the prepared catalysts. In the catalytic
hydrocracking of bio-oil from pyrolyzed a-cellulose, we ob-
tained a more stable liquid, which was called as biofuel. This
hydrocracking process is considered to be the bio-renery of
biomass; in this case, a-cellulose was converted into bio-oil by
pyrolysis and nally into biofuel by the hydrocracking process.
The hydrocracking process was performed by a hydrogenation
reaction at 450 ꢁC in the presence of H2 (20 cm3 minꢀ1) in
Table 3 Product distribution of hydrocracking process
Hydrocracking product
Nickel
loading (wt%)
Liquid
fraction (wt%)
Gas fraction
(wt%)
Liquid fraction
per Ni loading
Catalyst
Coke (wt%)
Thermal
SBA-15
NiO/SBA-15w
NiO/SBA-15c
Ni/SBA-15w
Ni/SBA-15c
—
—
5.07
5.25
8.15
5.20
48.61
85.96
66.13
60.28
69.89
67.35
51.39
13.16
31.26
38.86
28.82
31.83
0.00
0.88
2.61
0.86
1.29
0.82
—
—
13.04
11.48
8.58
12.95
This journal is © The Royal Society of Chemistry 2019
RSC Adv., 2019, 9, 1230–1237 | 1235

View Article Online
RSC Advances
Paper
SBA-15c (Fig. 6) before and aer hydrocracking of pyrolyzed a-
cellulose, it is conrmed that the Ni/SBA-15c catalyst (Fig. 6C
and D) has excellent stability towards hydrocracking of pyro-
lyzed a-cellulose as there is only slight blocking of the pores of
the catalyst due to slight formation of coke. On the other hand,
the TEM image of Ni/SBA-15w (Fig. 6A and B) aer hydro-
cracking of pyrolyzed a-cellulose is quite darker, indicating that
the formation of coke in the reaction was substantial and may
cause catalyst deactivation.
The selectivity of liquid product to the hydrocarbon forma-
tion is shown in Table 4. In Table 4, the resulting compounds
are grouped into carboxylic acid, hydroxyl ketone, furan
compound, and hydrocarbons (alkene; alkane). Based on the
table showing four major products resulting from hydro-
cracking, it is found that thermal hydrocracking produced
a long carboxylic acid. During thermal hydrocracking, a hydro-
carbon (dodecane) was also obtained but only in a small
amount of 4.21 wt%. For the SBA-15 catalyst without impreg-
nation by nickel, it was found that the catalyst was not selective
to furan compound. All other catalysts impregnated by nickel
were selective to the production of carboxylic acid, hydroxyl
ketone, furan, and hydrocarbons.
Fig. 6 TEM images of Ni/SBA-15w (A) before and (B) after hydro-
cracking of pyrolyzed a-cellulose and those of Ni/SBA-15c (C) before
and (D) after hydrocracking of pyrolyzed a-cellulose.
Comparing the selectivities of nickel catalyst impregnated
into SBA-15 in the production of biofuel, we observed that the
catalysts prepared by co-impregnation, i.e., both forms of NiO
The production of coke aer hydrocracking process is shown and Ni impregnated on SBA-15 showed selectivity toward the
in Table 3. Based on the table, it is shown that the formation of formation of hydrocarbons due to the formation of a branched
coke in NiO and Ni impregnated into SBA-15 via co-impregnation alkane (3-methyl-pentane). On the other hand, the prepared
was very much lower compared to that observed for the catalysts catalysts via wet impregnation showed negligible selectivity
prepared via wet impregnation; also, it was even smaller towards hydrocarbon formation. This was because no hydro-
compared to that observed for SBA-15. This fact supports the carbons were obtained in the one step of hydrocracking. The
previous observation reported by many researchers: co- investigation of two-step hydrotreatment of bio-oils with very
impregnation by using ethylene glycol can reduce the forma- contrasting compositions has been reported to give low oxygen
tion of coke, which can affect catalyst deactivation.
contents in the product.29 The selectivity towards hydrocarbon
When the reaction occurs and coke is formed, it might cause was shown aer the second hydrocracking was conducted using
decrease in catalytic activity due to deactivation of the catalyst Ni/SBA-15w and n-hexane was produced. When the resulting
active sites. Comparing the TEM images of Ni/SBA-15w and Ni/ hydrocarbons were compared, 3-methyl-pentane obtained from
Table 4 Selectivity of liquid product on hydrocracking of pyrolyzed a-cellulose
Product percentage (wt%)
Ni/SBA-15w
Compound
Thermal
SBA-15
NiO/SBA-15w
NiO/SBA-15c
4.59
1 step
2 step
Ni/SBA-15c
Dodecanoic acid
Decanoic acid
Tetradecanoic acid
Acetic acid
11.84
2.92
3.15
18.92
12.30
17.12
11.01
15.45
22.51
12.11
7.21
Propanoic acid
Ethanedioic
12.31
17.99
1-Hydroxy-2-propanone
1-Hydroxy-2-butanone
2-Furancarboxaldehyde
1-Methyl-4(pent-1-en-4-on-2-yl) cyclohexane
n-Hexane
17.48
9.86
11.66
11.12
7.70
9.27
8.17
14.57
9.84
18.55
13.43
13.09
3-Methyl-pentane
Dodecane
10.58
4.21
1236 | RSC Adv., 2019, 9, 1230–1237
This journal is © The Royal Society of Chemistry 2019

View Article Online
Paper
RSC Advances
the hydrocracking process using catalysts via impregnation was
more preferable because it has a higher octane number
compared to n-hexane.
6 N. R. Singh, W. N. Delgass, F. H. Ribeiro and R. Agrawal,
Environ. Sci. Technol., 2010, 44, 5298.
7 M. Patel and A. Kumar, Renewable Sustainable Energy Rev.,
2016, 58, 1293.
8 J. Grams, N. Potrzebowska, J. Goscianska, B. Michalkiewicz
and A. M. Ruppert, Int. J. Hydrogen Energy, 2016, 41, 8656.
9 C. Perego and A. Bosetti, Microporous Mesoporous Mater.,
2011, 144, 28.
10 Q. L. Zhang, M. Z. Wang, T. F. Zhang, Y. R. Wang, X. S. Tang
and P. Ning, RSC Adv., 2015, 5, 94016.
11 S. Lin, L. Shi, M. M. L. Ribeiro Carrott, P. J. M. Carrott,
J. Rocha, M. R. Li and X. D. Zou, Microporous Mesoporous
Mater., 2011, 142, 526.
12 M. Tao, Z. Sin, X. Meng, Z. Bian and Y. Lv, Fuel, 2017, 188,
267.
13 S. Qiu, X. Zhang, Q. Liu, T. Wang, Q. Zhang and L. Ma, Catal.
Commun., 2013, 42, 73.
14 J. Lu, B. Fu, M. C. Kung, G. Xiao, J. W. Elam, H. H. Kung and
P. C. Stair, Science, 2012, 335, 1205.
15 Q. Zhang, T. Wu, P. Zhang, R. Qi, R. Huang, X. Songa and
L. Gao, RSC Adv., 2014, 4, 51184.
16 B. W. Lu and K. Kawamoto, RSC Adv., 2012, 2, 6800.
17 F. Li, X. D. Yi and W. P. Fang, Catal. Lett., 2009, 130, 335.
18 J. F. Chen, Y. R. Zhang, L. Tan and Y. Zhang, Ind. Eng. Chem.
Res., 2011, 50, 4212.
4. Conclusions
In this study, the preparation of catalysts via the co-
impregnation method was successfully applied for NiO/SBA-
15c and Ni/SBA-15c catalysts, which showed excellent activity
and resistance towards coke formation in the hydrocracking of
pyrolyzed a-cellulose. The TEM images reveal that catalysts
prepared by wet impregnation show inhomogeneous distribu-
tion of nickel loading, whereas catalysts prepared by co-
impregnation using EG exhibit homogeneous distribution and
form no nickel aggregates. During hydrocracking of pyrolyzed a-
cellulose, Ni/SBA-15c with total acidity, nickel loading, particle
size, and specic surface area of 7.27 m mol gꢀ1, 5.20 wt%,
3.17 nm, and 310.0 m2 gꢀ1, respectively, exhibited best catalytic
performance among other prepared catalysts with 67.35 wt%
conversion of liquid product with maximum selectivity in
producing 13.09 wt% of 3-methyl-pentane. Moreover, Ni/SBA-
15w with total acidity, nickel loading, particle size, and
specic surface area of 10.87 m mol gꢀ1, 8.15 wt%, 7.01 nm, and
628.0 m2 gꢀ1, respectively, produced 69.89 wt% liquid product
without hydrocarbons. Study of selectivity towards the forma-
tion of liquid hydrocarbon was carried out through double step
hydrocracking using Ni/SBA-15w, and 18.55 wt% of n-hexane
was produced in the liquid product.
19 T. Xie, L. Y. Shi, J. P. Zhang and D. S. Zhang, Chem. Commun.,
2014, 50, 7250.
20 X. Y. Lv, J. F. Chen, Y. S. Tan and Y. Chang, Catal. Commun.,
2012, 20, 6.
21 D. P. Liu, X. Y. Quek, W. N. E. Cheo, R. Lau, A. Borgna and
Y. H. Yang, J. Catal., 2009, 266, 380.
22 S. Qiu, Q. Zhang, W. Lv, T. Wang, Q. Zhang and L. Ma, RSC
Adv., 2017, 7, 24551.
Conflicts of interest
There are no conicts to declare.
23 S. Jin, Z. Xiao, C. Li, X. Chen, L. Wang, J. Xing, W. Li and
C. Liang, Catal. Today, 2014, 234, 125.
24 P. Munnik, P. E. de Jongh and K. P. de Jong, Chem. Rev.,
2015, 115, 6687.
25 J. R. A. Sietsma, J. D. Meeldijk, M. Versluijs-Helder,
A. Broersma, J. A. van Dillen, P. E. de Jongh and K. P. de
Jong, Chem. Mater., 2008, 20, 2921.
Acknowledgements
The authors thank Indonesian Ministry of Research Technology
and Higher Education and Universitas Gadjah Mada for the
nancial support under the scheme of PTUPT research grant
2018 (Contract Number: 1923/UN1/DITLIT/DIT-LIT/LT/2018).
References
26 R. M. Santos, A. O. Santos, E. M. Sussuchi, J. S. Nascimento,
A. S. Lima and L. S. Freitas, Bioresour. Technol., 2015, 196, 43.
27 M. Pongsendana, W. Trisunaryanti, F. W. Artanti, I. I. Falah
and S. Sutarno, Korean J. Chem. Eng., 2017, 34, 2591.
28 W. Trisunaryanti, S. Purwono and A. Putranto, Indones. J.
Chem., 2008, 8, 342.
29 S. Kadarwati, S. Oudenhoven, M. Schagen, X. Hu,
M. G. Perez, S. R. A. Kersten and R. J. M. Westerhof, J.
Anal. Appl. Pyrolysis, 2016, 118, 136.
1 A. V. Bridgwater, Chem. Eng. J., 2003, 91(2–3), 87.
2 S. V. Vassilev, D. Baxter, L. K. Andersen and C. G. Vassileva,
Fuel, 2013, 105, 19.
3 A. Pandey, T. Bhaskar, M. Stocker, and R. Sukumaran, Recent
Advances in Thermochemical Conversion of Biomass, Elsevier,
Boston, 2015, pp. 3–30.
¨
4 Z. Tan, K. Chen and P. Liu, Renewable Sustainable Energy
Rev., 2015, 41, 368–378.
5 T. Nomiyama, N. Aihara, A. Chitose, M. Yamada, and S. Tojo,
Biomass aslocal resource: Research approaches to sustainable
biomass systems, Academic Press, Boston, 2014, pp. 7–17.
This journal is © The Royal Society of Chemistry 2019
RSC Adv., 2019, 9, 1230–1237 | 1237