NiK/yCexZr1-xO2-macroporous Al2O3 catalysts for cracking of vacuum residual oil with steam

Article abstract of DOI:10.1016/j.apcata.2016.07.005

In this study, NiK/yCe_xZr_1-xO₂-macroporous Al₂O₃ catalysts (y?=?0, 10, 20, and 30; x?=?0.73–0.92) were synthesized by dispersing different amounts of Ce_xZr_1-xO₂ phase onto macroporous γ-Al₂O₃ as supports, and then subsequently impregnating Ni and K into the supports, which possessed advantageous properties, such as a high surface area, ordered macropores and high oxygen storage capacity. Moreover, the introduction of Ni and K metals into these supports created more oxygen vacancies in them. We applied these catalysts to the cracking of vacuum residue with steam. It was found that the variation of the liquid yield with the amount of Ce_xZr_1-xO₂ phase showed a volcano pattern. The macropores in the supports played an important role in enhancing the diffusion of large molecules to the active sites, while the high oxygen storage capacity over the Ce_xZr_1-xO₂ phase improved the oxidative cracking rate, thereby increasing the lighter oil fraction from the vacuum residue.

Full text of DOI:10.1016/j.apcata.2016.07.005

Accepted Manuscript
Title: NiK/yCe Zr O -macroporous Al O catalysts for
x
1−x
2
2
3
cracking of vacuum residual oil with steam
Author: Do Thi Lien Chinh Nguyen-Huy Eun Woo Shin
PII:
S0926-860X(16)30360-X
DOI:
Reference:
http://dx.doi.org/doi:10.1016/j.apcata.2016.07.005
APCATA 15942
To appear in:
Applied Catalysis A: General
Received date:
Revised date:
Accepted date:
21-3-2016
13-6-2016
8-7-2016
Please cite this article as: Do Thi Lien, Chinh Nguyen-Huy, Eun
Woo Shin, NiK/yCexZr1-xO2-macroporous Al2O3 catalysts for cracking
of vacuum residual oil with steam, Applied Catalysis A, General
http://dx.doi.org/10.1016/j.apcata.2016.07.005
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

NiK/yCe
steam
x
Zr1-xO
2
-macroporous Al
2
3
O catalysts for cracking of vacuum residual oil with
Do Thi Lien, Chinh Nguyen-Huy, Eun Woo Shin
School of Chemical Engineering, University of Ulsan Daehakro 93, Nam-gu, Ulsan 44610,
South Korea
Corresponding author: Tel.: +82-52-259-2253, Fax: +82-52-259-1689
E-mail address: ewshin@ulsan.ac.kr
1

Graphical abstract
Highlights

Ce
residue
Macroporous structure enhanced the diffusion of large molecules to active sites
High oxygen storage capacity of Ce phase improved the oxidative cracking rate
x
Zr1-x
O
2
were impregnated onto NiK/macroporous Al
2
O
3
for cracking of vacuum



x
Zr1-x
O
2
NiK/20Ce
x
Zr1-xO
2
-MaAl
2
O
3
catalyst showed the best catalytic performance
2

Abstract
In this study, NiK/yCe
.73 - 0.92) were synthesized by dispersing different amounts of Ce
macroporous -Al as supports, and then subsequently impregnating Ni and K into the
x
Zr1-xO
2
-macroporous Al
2
O
3
catalysts (y = 0, 10, 20, and 30; x =
0
x
Zr1-xO
2
phase onto
2
O
3
supports, which possessed advantageous properties, such as a high surface area, ordered
macropores and high oxygen storage capacity. Moreover, the introduction of Ni and K metals
into these supports created more oxygen vacancies in them. We applied these catalysts to the
cracking of vacuum residue with steam. It was found that the variation of the liquid yield with
the amount of Ce
played an important role in enhancing the diffusion of large molecules to the active sites,
while the high oxygen storage capacity over the Ce phase improved the oxidative
cracking rate, thereby increasing the lighter oil fraction from the vacuum residue.
x
Zr1-xO
2
phase showed a volcano pattern. The macropores in the supports
Zr1-xO
x 2
Keywords: Ce
x
Zr1-xO
2
phase; NiK/Al
2
3
O ; Macroporous; Vacuum residue; Catalytic cracking,
Oxidative cracking
1
. Introduction
Heavy oil, extra-heavy oil and bitumen make up about 70% world’s total oil resources of
9
to13 trillion bbl [1]. However, since the demand for fuels increases strongly every year, it is
necessary to convert these heavy oils to light oil which can be used as fuels and feedstock for
the petrochemical industry. Nowadays, many oil processing techniques are being used for
producing fuels from heavy or extra-heavy oil, such as the atmospheric residue and vacuum
3

residue (VR) techniques. Amongst these different methods, the cracking process has attracted
a great deal of attention. In this process, the C-C bond in the hydrocarbon chain is broken
down, forming lighter hydrocarbons. Depending on the factors involved in the cracking of the
C-C bonding, the cracking process can be divided into three main types: thermal cracking,
catalytic cracking, and hydrocracking [2].
Catalytic cracking is a conversion process that can be applied to a variety of feedstocks
ranging from gas oil to heavy oil. The cracking reaction results from the attack of a strong
acid on a hydrocarbon chain to form a carbonium ion [3]. Strong acids come in two
fundamental types, Bronsted and Lewis acids. Bronsted acids are the familiar proton-
containing acids, while Lewis acids are a broader class including inorganic and organic
species formed by positively charged centers. Both kinds have been identified on the surface
of catalysts. The mixture of Bronsted and Lewis acids sometimes depends on the level of
water in the system [4]. In addition to conventional catalytic cracking, the oxidative cracking
process has also been studied. The oxidative cracking of residual oils by methods such as the
atmospheric residue and vacuum residue processes with steam converts the heavy oils into
useful lighter hydrocarbons. The steam provides an alternative resource of oxygen species and
hydrogen for oxidative cracking and hydrogenation [5]. The catalyst used for oxidative
cracking usually has a high oxygen storage capacity (OSC), for example CeO
2
, CeO
2
-ZrO
2
and Zr/FeO [5,6]. Cerium oxide has been extensively used in heterogeneous catalysis because
x
of its ability to store and release oxygen, which is related to its redox properties and thermal
stability [7,8]. The mobility of oxygen species in these materials increases significantly when
4
+
Zr is doped into the CeO
2
lattice [9]. Haneda et al. [10] reported that the use of CeO
2
/ZrO
2
catalysts for soot oxidation resulted in high soot oxidation activity. In another study, the role
4

of Ce
x
Zr1-x
O in propane dehydrogenation to enhance the catalytic stability and decrease the
2
coke combustion temperature was investigated [8].
In this study, NiK/yCe
product yields from the vacuum residue through oxidative cracking. Macroporous Al
MaAl ) is used as a porous support which contains acid sites on the surface [11]. In a
steam environment, these sites promote the catalytic cracking process. In addition, due to the
existence of the macro-mesoporous system in the structure of MaAl , the heavy component
in VR will access the acid sites conveniently, thereby also promoting the catalytic cracking
process. In our previous study [12], the hierarchical macro–mesoporous Al -supported
x
Zr1-xO
2
-MaAl
2
O
3
catalysts are synthesized to produce high light
2
O
3
(
2
O
3
2
O
3
2
O
3
catalyst exhibited better catalytic performance than the mesoporous alumina supported
catalyst in steam cracking, due to the fast diffusion of the reactants to the active sites for
hydrogenation. Along with the macroporosity of the support, the Ce
onto MaAl plays an important role by acting as active sites for the oxidative cracking.
Moreover, the introduction of Ni and K onto yCe -MaAl supports increases the
oxygen storage capacity [13] and promotes the decomposition of steam [14], improving the
oxidative cracking by the Ce phase Accordingly, with NiK/yCe -MaAl
x
Zr1-x
2
O phase dispersed
2
O
3
x
Zr1-xO
2
2
O
3
Zr1-xO
x 2
.
x
Zr1-xO
2
2
O
3
catalysts, not only the accessibility of the large components such as asphaltenes, in the
vacuum residue to the active sites will be increased due to the macro-mesoporous system in
MaAl
2
O
3
, but also additional oxidative cracking will occur over the Ce
x
Zr1-x
O phase,
2
eventually resulting in the enhancement of cracking rates with steam.
2
. Experimental
.1.Catalyst preparation
2
5

To prepare MaAl
2
3
O , 10 g of aluminum-tri-sec-butoxide (ASB, Sigma Aldrich Korea)
was dissolved completely in 10 g of 2-butanol (Sigma Aldrich Korea) by using a magnetic
stirrer. Then, polystyrene (PS) spheres (ca. 320 nm diameter), which were prepared by an
emulsifier-free emulsion polymerization method, were added to the solution under continuous
stirring for 4 hours. The solid product was filtered and washed completely with a mixture of
o
ethanol and water (1Vethanol/2Vwater). The washed solid product was dried in air at 70 C for 24
hours. The PS spheres were removed by calcination in a muffle furnace in air. The
o
o
temperature was ramped at 1 C/min to 600 C and then held at this temperature for 6 hours.
The yCe
excess solution of Ce and Zr precursors. The Ce(NO
ZrO(NO .6H O (Sigma-Aldrich, 99%) salts were each dissolved in 12 ml of distilled
water in separate beakers using sonication and, then, the two solutions were combined in one
beaker. The combined solution was then added to a beaker containing 2 g of MaAl , stirred
x
Zr1-xO
2
-MaAl
2
O
3
supports were synthesized by impregnating MaAl
2
3
O with
3
)
3
.6H O (Sigma-Aldrich, 99%) and
2
3
)
3
2
2
O
3
o
o
at 30 C for 1 hour and left to stand at 30 C for 12 hours. The solution was heated slowly to
o
and kept at 50 C with constant stirring to evaporate the water. After that, the paste samples
o
were dried at 100 C for 12 hours in an oven and then ground with a mortar and pestle. The
o
o
ground powders were calcined in a furnace by heating them from 30 C to 600 C with a ramp
o
rate of 1.5 C/min and then keeping them at this temperature for 6 hours.
Through the incipient wetness impregnation method, NiK/yCe
were synthesized by impregnating an aqueous solution of the precursor salts, Ni(NO
KNO , to reach 10 Wt.% Ni and 5 Wt.% K on the yCe -MaAl supports. The
x
Zr1-xO
2
-MaAl
2
O
3
catalysts
)
3 2
and
3
x
Zr1-xO
2
2
O
3
precursors were dissolved in a predetermined amount of deionized water. One aliquot
containing 200 l of the precursor was evenly dispersed on the calcined supports. The
procedure was repeated until the ratio of liquid to support was 1.0 ml/g. The catalysts were
6

o
dried at room temperature for 24 hours and then at 100 C for 12 hours in an oven and
o
subsequently calcined at 600 C for 4 hours in a furnace. The catalysts were reduced in a flow
o
of hydrogen at 600 C for 2 hours. In all of the catalysts, the atomic ratio of Ce to Zr was
equal to 1.0.
2
.2. Catalyst characterization
The textural properties of the NiK/yCe
x
Zr1-xO
2
-MaAl
2
O
3
catalysts were measured by
N
2
adsorption analysis on a Micromeritics ASAP 2020 apparatus (USA). The X-ray
diffraction (XRD) patterns were obtained using a Rigaku RAD-3C diffractometer (Japan)
with Cu Ka radiation ( = 1.5418 Å) at a scan rate of 2° (2θ)/min, operated at 35 kV and 20
mA. The surface morphology and particle size were characterized by field-emission scanning
electron microscopy (FE-SEM, JEOL JSM-600F, Japan). High-resolution transmission
electron microscopy (TEM, JEM-2100F, JEOL) bright-field images were also obtained. The
specimens were prepared by suspending and grinding them in an ethanol solution; two drops
of ethanol solution containing the ground catalyst powder were placed onto a carbon film-
coated copper grid. The Raman spectra were obtained with a DXR Raman Microscope
(Thermo Scientific, USA) using a 532 nm excitation source. The spectral resolution employed
−
1
was 2 cm .
2
.3. Reaction test
o
First, the catalysts were reduced in situ in flowing hydrogen (25 ml/min) at 600 C for 2 h.
Then, the reduced catalyst (0.5 g) was put into the reactor. A mixture of water and nitrogen
was heated to vaporize the water and then entered the reactor at flow rates of 0.3 and 70
ml/min, respectively. The VR feed (0.15 ml/min) was fed into the reactor. The reaction was
o
allowed to proceed at 500 C for 2 hours in a fix-bed reactor (see Fig. S1 in Supplementary
7

Materials). The time factor, W/F
R
, was 0.6 h, where F is the flow rate of VR without toluene
R
and W is the amount of catalyst. In this study, commercial VR was diluted to 10 wt% with
toluene and the diluted liquid was used as a feedstock.
For its analysis, the gas product was sampled through a six-port valve via a sample loop
(0.25 ml) into a chromatograph (YOUNG LIN-Acme 6100 GC, South Korea) using a thermal
conductivity detector (TCD) and HPPLOT Q column. The distribution of the desired liquid
o
o
o
products, such as gasoline (180 C), diesel (180-350 C), and VGO (350-560 C), and the
o
o
5
7
60 C+ fraction (>560 C) was measured by GC-SIMDIS (AC Analytical Controls, Agilent
890). The analysis conditions were based on the ASTM D-2887 (simulated distillation) and
ASTM D-5307 methods. The coke yield was determined by TGA analysis (TG Q50, TA
instrument).
The conversion of the VR and the yield of each product were calculated by balancing the
content of the desired boiling point products in the following equations:
[
퐼푛푝푢푡 푉푅 푤푒ꢀ푔ℎ푡] − [푊푒ꢀ푔ℎ푡 표푓 푎푙푙 푡ℎ푒 푝푟표푑푢푐푡푠]
퐺푎푠 푦ꢀ푒푙푑 (푤푡%) =
× 100
× 100
퐹푒푒푑 푉푅 푤푒ꢀ푔ℎ푡
[푊푒ꢀ푔ℎ푡 표푓 푔푎푠표푙ꢀ푛푒 + 퐷ꢀ푒푠푒푙 + 푉퐺푂 푝푟표푑푢푐푡푠]
퐿ꢀ푞푢ꢀ푑 푦ꢀ푒푙푑 (푤푡%) =
퐹푒푒푑 푉푅 푤푒ꢀ푔ℎ푡
[
푤푒ꢀ푔ℎ푡 표푓 푐표푘푒 푝푟표푑푢푐푒푑 푑푢푟ꢀ푛푔 푟푒푎푐푡ꢀ표푛]
퐶표푘푒 푦ꢀ푒푙푑 (푤푡%) =
× 100
퐹푒푒푑 푉푅 푤푒ꢀ푔ℎ푡
[
푃푟표푑푢푐푡푠 푤푒ꢀ푔ℎ푡 표푓 푑푒푠ꢀ푟푒푑 푏표ꢀ푙ꢀ푛푔 푝표ꢀ푛푡]
퐸푎푐ℎ 푑푒푠ꢀ푟푒푑 푝푟표푑푢푐푡 푦ꢀ푒푙푑 (푤푡%) =
퐹푒푒푑 푉푅 푤푒ꢀ푔ℎ푡
퐶표푛푣푒푟푠ꢀ표푛 (푤푡%) = [퐺푎푠 푦ꢀ푒푙푑] + [푙ꢀ푞푢ꢀ푙푑 푦ꢀ푒푙푑] + [푐표푘푒 푦ꢀ푒푙푑]
3
. Results and discussion
8

3
.1. Catalyst characterizations
Fig. 1 shows the SEM images of the polystyrene, MaAl
2
O
3
and NiK/yCe
x
Zr1-x
O -
2
MaAl catalysts. The macroporous structure of MaAl is built by making use of hard
2
O
3
2
O
3
template-polystyrene spheres (average diameter 320 nm, Fig. 1A). After removing the PS
o
spheres by calcination at 600 C for 6 h, the ordered macro-mesoporous system is found in the
structure (Fig. 1B). Because of the shrinking of the hard templates during the calcination
process, the diameter of the macropores is smaller than that of the polystyrene spheres [15].
The wall thickness and diameter of the macropores are about 75 nm and 215 nm, respectively.
The SEM images of the NiK/yCe
According to these images, the macropores of MaAl
Ce phase during the addition of Ce and Zr. The consistent result shows that an
increase in the Ce and Zr loading amounts onto MaAl results in a decrease in the surface
x
Zr1-xO
2
-MaAl
2
O
3
catalysts are shown in Figs. 1C-F.
2
O
3
are significantly covered with the
x
Zr1-xO
2
2
O
3
2
2
3
3
areas (from 211.6 m /g to 80.9 m /g) and pore volumes (from 0.48 cm /g to 0.19 cm /g) of the
catalysts (Table 1). Moreover, adding Ce and Zr to MaAl reduces the interaction between
Ni and MaAl , which promotes the aggregation of the Ni particles on the surface of the
2
O
3
2
O
3
support [16]. This causes an increase in the nickel particle sizes with increasing amounts of
Ce and Zr (Table 1).
Fig. S2 (see in Supplementary Materials) shows the nitrogen adsorption-desorption
isotherms and pore size distributions of the NiK/yCe
x
Zr1-xO
2
-MaAl
2
3
O catalysts. The nitrogen
adsorption-desorption isotherms of the catalysts (Fig. S2-A) represent a typical type-IV
isotherm with H3-shaped hysteresis loop which doesn’t exhibit any limiting adsorption at
high P/P . This corresponds to capillary condensation in the mesopores and the existence of
o
narrow slit-shaped pores in the materials [17]. The occurrence of the hysteresis loop at higher
relative pressures of around 0.8–0.98 and the continuous nitrogen uptake near the saturation
9

pressure indicate the presence of macropores in the catalysts [18]. In the pore size distribution
curves of the NiK/yCe -MaAl catalysts (Fig. S2-B), one peak around 7 nm is
x
Zr1-xO
2
2
O
3
obviously shown. This peak becomes broader and shifts to a higher pore diameter as the
amounts of Ce and Zr increase.
Fig. 2 shows the XRD patterns of the NiK/yCe
calcination at 500 °C and reduction in H at 600 °C for 2 h, where all of the characteristic
XRD peaks for the crystalline structure of all of the catalysts are to be found. The peaks at
x
Zr1-xO
2
/MaAl
2
O
3
catalysts after
2
2
θ=28.64o (NiK/10Ce
x
Zr1-x
Zr1-x
, Ce0.83Zr0.0.17
O
2
-MaAl
2
O
3
), 2θ=28.76o (NiK/20Ce
) are assigned to the reflection peaks of the solid
, respectively. These peaks have
a broad shape which is indicative of the good dispersion of the ceria–zirconia crystallites over
MaAl . With increasing amounts of Ce and Zr, the position of the peak for the Ce
phase shifts to a higher two theta value (Table 2), suggesting a decrease in the lattice constant
x
Zr1-x
O
2
-MaAl
2
O ) and
3
o
2
θ=28.91 (NiK/30Ce
x
O
2
-MaAl
2
O
3
solutions of Ce0.92Zr0.08
O
2
O
2
and Ce0.73Zr0.27
O
2
2
O
3
x
Zr1-xO
2
4
+
because of the incorporation of the smaller Zr ions into the CeO
2
lattice [8]. The chemical
compositions of the solid solution Ce in all of the catalysts are estimated from the
Zr1-xO
x 2
calibration curve generated by plotting the two theta value of the (111) peak as a function of
the atomic concentration of cerium using the JCPDS database. The estimated values of x are
in the range of 0.73 - 0.92 (Table 2). In addition, strong sharp peaks for Ni metal phase are
o
o
o
detected at 2 = 44.43 ; 51.91 ; 76.51 (JCPDS Card No. 04-0850), confirming the reduction
of NiO to Ni under hydrogen media. The strong interaction between the Ni species and
MaAl
peaks of which are detected at 2=36.4 ; 45.5 and 65.3 (JCPDS Card No.10-0339). The
amounts of Ni species changed to NiAl phase are simply estimated on the basis of XRD
patterns [20]. The estimated values for relative ratio of NiAl phase to total Ni species are
2
O
3
induces the formation of surface nickel aluminate (NiAl
2
O
4
) [19], the characteristic
o
o
o
2
O
4
2
O
4
1
0

listed in Table 4. Without Ce
x
Zr1-x
O
2
phase, there is a strong interaction between Ni and
/Ni species (0.39). In the presence of Ce
alumina, resulting in a high value of NiAl
2
O
4
x
Zr1-xO
2
phase, the values are relatively low in the range of 0.07-0.09, indicating that Ce
x
Zr1-x
2
O phase
inhibits the interaction between Ni and alumina.
In the Raman spectra of the NiK/yCe
x
Zr1-xO
2
/MaAl
2
O
3
catalysts after reduction in H at
2
6
4
00 °C for 2 h (Fig. 3A), the Ce-O and Ce- ( : a vacant site) bands can be clearly seen at
-1
-1
67 cm and 610 cm , respectively. The intensity ratio ICe- /ICe-O shows the relation between
the Ce-O bonding and Ce- bonding. An increase of the ICe- /ICe-O value indicates a relative
increase in the quantity of oxygen vacancies. The formation of oxygen vacancies in a catalyst
depends on many factors, such as the calcination temperature, reduction or oxidation media
and doping of metal into the catalyst [13,21,22]. Fig. 4 shows the Raman spectra of 30Ce Zr1-
x
o
O
x 2
-MaAl
2
O
3
after calcination at 600 C for 6 hours, 30Ce
x
Zr1-xO
2
-MaAl
2
3
O after reduction in
H
2
x
, and NiK/30Ce Zr1-x
O
2
-MaAl
2
O
3
after reduction in H
2
, respectively. The peak of the Ce-O
-1
-1
band at 467 cm is clearly present, while there is no peak for Ce- at around 615 cm in the
case of 30Ce
MaAl after reduction, the peak of Ce- at 615 cm appears (Fig. 4b), indicating that
oxygen vacancies form over the Ce phase after reduction in H and that the
x
Zr1-xO
2
-MaAl
2
O
3
after calcination (Fig. 4a). However, for 30Ce
x
Zr1-x
2
O -
-1
2
O
3
x
Zr1-xO
2
2
4
+
3+
3+
4+
transformation of Ce to Ce occurs. The substitution of two Ce for one Ce in the lattice
generates one oxygen vacancy in the lattice of CeO [23]. Moreover, the doping of Ni species
2
onto the supports increases the number of the oxygen vacancies due to the formation of
charge-compensating oxygen vacancies neighboring the dopant, which is evidenced by the
Raman spectra of 30Ce
Zr1-xO
x 2
-MaAl
2
O
3
after reduction and NiK/30Ce
x
Zr1-xO
2
-MaAl
2
O
3
after reduction (Fig. 4 (spectra b-c)).
3
.2. Cracking VR over NiK/Ce
x
Zr1-xO
2
/MaAl
2
O catalysts
3
1
1

Table 4 shows the product distributions after VR cracking with steam over the
o
NiK/yCe
time factor W/F
produces a lower liquid yield (50.12 %) and conversion (57.26%). In the presence of
NiK/MaAl , however, the main reactions were suggested to be catalytic cracking and steam
reforming [24], which increases the liquid yields from 50.12% to 64.32%. The gas yield in the
presence of NiK/MaAl is higher than that without the catalyst (Table 4), indicating that
x
Zr1-xO
2
-MaAl
2
O
3
catalysts for 2 hours. The reaction temperature is 500 C and the
R
is 0.6 h. Without the catalyst, the main reaction is thermal cracking, which
2
O
3
2
O
3
this catalyst has high acidity and that the existence of the macropores accelerates the diffusion
rate of large molecules to the acidic sites and Ni surface for the catalytic cracking and steam
reforming reactions, respectively. Along with the catalytic cracking over the acid sites, the
presence of CO gas in the gaseous products (Table 5) suggests that the steam reforming
2
occurs over the Ni/support [25–27]; this is a secondary reaction derived from the light
fractions, especially alkyl benzene. The alkyl chains of alkylbenzenes react with steam,
transforming the aromatics and CO . Therefore, this process has a high selectivity toward
2
aromatics such as benzene and toluene [25,28], causing an increase in the gasoline and diesel
yields in the product distribution. However, the steam reforming of high molecular weight
hydrocarbons is a complex process, including several consecutive and parallel reactions, some
of which may result in the formation of coke.
With increasing Ce and Zr loading amounts, the surface areas of the catalysts decrease
and the nickel particle sizes increase significantly (Table 1), implying a decrease in the
catalytic activity of the acidic sites over MaAl
catalytic cracking over MaAl becomes less active due to the coverage of MaAl
Ce phase. Instead, the oxidative cracking over Ce phase increases. Oxidative
cracking is a process mainly involving the breakage of the C-C bonds between the aromatic
2
3
O , as well as that of the Ni metal sites. The
2
O
3
2
O
3
by
x
Zr1-xO
2
x
Zr1-xO
2
1
2

groups in both maltene and asphaltene in the residue and coke [28] to form lighter oils and
release CO gas, indicating that this process enhances the VGO yields in the product. It is
observed that the VGO yield increases as the Ce and Zr loading amounts increase (Table 4)
Moreover, the amount of CO also increases in the gas products (Table 5), proving that CO
gas is generated by the oxidative cracking over the Ce phase, as well as by the steam
reforming on the Ni particles. The level of CO gas formation indicates the oxygen storage
capacity (OSC) in the catalysts [21]. With increasing amounts of Ce and Zr, the relative
portion of CO in the gas composition gradually rises (Table 5). In contrast, the catalytic
cracking over the acidic sites over MaAl lessens, due to the surface coverage of MaAl
with Ce phase, which is evidenced by a rapid decrease in the gas yield (Table 4) and a
gradual decrease of C2 in the gas composition (Table 5) with increasing amounts of Ce and Zr
12,30]. The catalytic performance in this study is strongly related to the competition or
2
2
2
Zr1-xO
x 2
2
2
2
O
3
2
O
3
x
Zr1-x
O
2
[
combination of the catalytic cracking, steam reforming and oxidative cracking over each
active site. The highest value of the liquid yield obtained in this study is 73.85% for the
NiK/20Ce
x
Zr1-xO
2
-MaAl
2
3
O catalyst (Table 4). As y increases up to y=20, the liquid yield
increases gradually and then decreases again at y=30. We can observe clearly the competition
of these reactions in Table 4; the reaction conversion is almost unchanged as the Ce and Zr
loading amounts increase from 0% to 20%. Then, as the Ce and Zr loading amounts further
increase to 30%, the reaction conversion decreases.
Fig. 5 shows the proposed scheme for the reaction pathways of VR cracking with steam
over the NiK/Ce
x
Zr1-x
O
2
-MaAl
2
O
3
catalysts. In the process of catalytic cracking, the C-C
to forms light oils.
bonds in the heavy oils are broken down by the acidic sites on MaAl
2
O
3
The catalytic cracking reaction mechanism is identified as the protolytic and -scission
mechanism [31]. If the protolytic mechanism is followed, methane, ethane and even ethylene
1
3

will be found in the gas products, while C3 and C4 gases will be formed by the -scission
mechanism. In our previous study [12], C2 gases were generated from the catalytic cracking
over the acidic sites of the supports, not from thermal cracking. In this study, with increasing
amounts of Ce and Zr, the C2 gas composition gradually decreases, due to the loss of the
acidic sites on MaAl
2
3
O (Table 5). Along with the catalytic cracking, deposition and
polymerization reactions occur, forming to coke. That is evidenced by dectection of a graphite
O
peak at 2=25.2 in the XRD patterns of the spent catalysts, as shown in Fig. 6.
In this study, a parallel reaction with the catalytic cracking is oxidative cracking. In
previous studies, it was proposed that the reaction mechanism for oxidation depends on the
redox properties of CeO
2
[32–34]. In general, the oxygen can adsorb onto the reduced surface
of the Ce phase in the vicinity of the hydrocarbon and, subsequently, electron transfer
x
Zr1-xO
2
3
+
3+
4+
from the surface donor (Ce ) to the adsorbed oxygen occurs as follows: Ce - + O  Ce -
2
-
(
O-O) [6]. The formation of active oxygen species in this study can be explained by steam
decomposition. The active oxygen species produced from steam decomposition over ZrO [14]
2
can oxidize the carbonaceous residue or induce the oxidative cracking of heavy oil [5]. The
active oxygen species on the surface of a catalyst can be divided into two groups: electrophilic
and nucleophilic. Electrophilic oxygen comprises electron deficient adsorbed species such as
−, O
2−
and O which are generated from the activation of O , whereas nucleophilic oxygen
−
O
2
2
2
includes saturated species such as the terminal oxygen group (M O) and bridging oxygen
2
−
(
M–O–M), where the active oxygen species exist in a nominal O state [32]. In this case, the
carbon in heavy oil adsorbs on the nucleophilic oxygen in the lattice of CeO and the active
oxygen species on the surface of the cerium oxide become unstable. As a result, cracking of
2
4+
3+
the heavy oil takes place via oxidation and the Ce is reduced to Ce [29]. In this process, the
surface mobility of the active oxygen species in the CeO lattices is a key step. In this study,
2
1
4

Ce
which enhances the oxidative cracking of heavy oil [9]. In fact, the oxidation of hydrocarbons
results in the formation of CO as one of the main gases. The detection of CO gas
demonstrates that the active oxygen species are employed to decompose the heavy
components into lighter fractions and CO . The Ce phase detected by XRD, as shown in
x
Zr1-xO
2
solid solution plays an important role by acting as an oxygen species conductor
2
2
2
2
O
3
Fig. 6, and the oxygen vacancies detected by Raman spectroscopy for the spent catalysts are
decreased (Table 3), implying that the formation rate of the active oxygen species generated
4
+
from the decomposition of steam is faster than the reduction rate of Ce . Besides the oxygen
species react with heavy oils, a large number of the active oxygen species react with Ni metal,
leading to the formation of NiO phase in the spent catalysts (Fig. 6).
4
. Conclusion
NiK/yCe
x
Zr1-xO
2
-macroporous Al
2
O
3
catalysts were prepared for the process of
o
cracking VR with steam, which was conducted in a fixed-bed reactor at 500 C to improve the
liquid yield and the reaction conversion. The macro-mesoporous structure plays an important
role in reducing the diffusion time of large molecules to the active sites, thereby promoting
the catalytic cracking over MaAl
steam reforming on the Ni particles. However, since the Ce
oxygen storage capacity, the oxidative cracking was enhanced when it was impregnated into
MaAl , whereas the catalytic cracking was inhibited, due to the loss of the acidic sites. The
2
O
3
, the oxidative cracking over the Ce
x
Zr1-x
O phase and the
2
Zr1-xO
x 2
phase possesses a high
2
O
3
catalyst with y=20 showed the best catalytic performance (the highest liquid yield) resulting
from the competitive process of catalytic cracking, steam reforming and oxidative cracking.
Acknowledgement
This research was financially supported by the research fund of the University of Ulsan.
1
5

References
[
1] H. Alboudwarej, J. (John) Felix, S. Taylor, R. Badry, C. Bremner, B. Brough, Oilf. Rev.
2006) 34–53.
(
[
[
[
2] J.H. Gary, G.E. Handwerk, Petroleum Refining, Taylor & Francis, 2001.
3] B.S. Greensfelder, H.H. Voge, G.M. Good, Ind. Eng. Chem. 41 (1949) 2573–2584.
4] J.G. Speight, The Chemistry and Technology of Petroleum, in: CRC Press, 1999: p.
9
34.
[
[
5] E. Fumoto, T. Tago, T. Masuda, Energy & Fuels. 20 (2006) 1–6.
6] M. Machida, Y. Murata, K. Kishikawa, D. Zhang, K. Ikeue, Chem. Mater. 20 (2008)
4
489–4494.
[
[
7] J. Kašpar, P. Fornasiero, M. Graziani, Catal. Today. 50 (1999) 285–298.
8] B.K. Vu, E.W. Shin, J.-M. Ha, S.K. Kim, D.J. Suh, W.-I. Kim, H. L. Koh, Y. G. Choi,
Appl. Catal. A Gen. 443-444 (2012) 59–66.
[
9] T. Murota, T. Hasegawa, S. Aozasa, H. Matsui, M. Motoyama, J. Alloys Compd. 193
(1993) 298–299.
[
[
[
[
10] M. Haneda, R. Taguchi, M. Hattori, J. Ceram. Soc. Japan. 123 (2015) 414–418.
11] M. Trueba, S.P. Trasatti, Eur. J. Inorg. Chem. 2005 (2005) 3393–3403.
12] C. Nguyen-Huy, E.W. Shin, Fuel. 169 (2016) 1–6.
13] A.B. Kehoe, D.O. Scanlon, G.W. Watson, Chem. Mater. 23 (2011) 4464–4468.
1
6

[
[
14] K. Sayama, H. Arakawa, J. Phys. Chem. 97 (1993) 531–533.
15] C. Nguyen-Huy, V.H. Pham, D.K. Kim, D.-W. Kim, S.H. Oh, J.S. Chung, E. W. Shin,
Appl. Catal. A Gen. 468 (2013) 305–312.
[
[
[
16] P. Turlier, H. Praliaud, P. Moral, G.A. Martin, J.A. Dalmon, Appl. Catal. 19 (1985)
87–300.
2
17] K.S.W. Sing, D.H. Everett, R. a. W. Haul, L. Moscou, R. a. Pierotti, J. Rouquérol, T.
Siemieniewska, Pure Appl. Chem. 57 (1985) 603–619.
18] D. Han, X. Li, L. Zhang, Y. Wang, Z. Yan, S. Liu, Microporous Mesoporous Mater.
1
58 (2012) 1–6.
[
[
19] E. Heracleous, A. Lee, K. Wilson, A. Lemonidou, J. Catal. 231 (2005) 159–171.
20] J.D. Martin, Using XPowder: A software package for Powder X-Ray diffraction
analysis, Spain, 2004. www.xpowder.com.
[
21] S. Bedrane, C. Descorme, D. Duprez, Catal. Today. 75 (2002) 401–405.
[
22] L. Liotta, A. Macaluso, G.. Arena, M. Livi, G. Centi, G. Deganello, Catal. Today. 75
(2002) 439–449.
[
[
23] J.R. McBride, K.C. Hass, B.D. Poindexter, W.H. Weber, J. Appl. Phys. 76 (1994) 2435.
24] J. Juan-Juan, M.. Román-Martınez, M.. Illán-Gómez, Appl. Catal. A Gen. 264 (2004)
́
1
69–174.
[
[
25] D. Duprez, Appl. Catal. A Gen. 82 (1992) 111–157.
26] Y. Matsumura, T. Nakamori, Appl. Catal. A Gen. 258 (2004) 107–114.
1
7

[
[
[
27] J.L. Pinilla, P. Arcelus-Arrillaga, H. Puron, M. Millan, Fuel. 109 (2013) 303–308.
28] D. Duprez, A. Miloudi, G. Gelahay, R. Maurel, J. Catal. 90 (1984) 292–304.
29] M. Dejhosseini, T. Aida, M. Watanabe, S. Takami, D. Hojo, N. Aoki, T. Arita, A.
Kishita, T. Adschiri, Energy & Fuels. 27 (2013) 4624–4631.
[
30] E.F. Iliopoulou, E.V. Antonakou, S.A. Karakoulia, I.A. Vasalos, A.A. Lappas, K.S.
Triantafyllidis, Chem. Eng. J. 134 (2007) 51–57.
[
[
31] A. Corma, A.V. Orchillés, Microporous Mesoporous Mater. 35-36 (2000) 21–30.
32] Z. Guo, B. Liu, Q. Zhang, W. Deng, Y. Wang, Y. Yang, Chem. Soc. Rev. 43 (2014)
3
480–524.
[
[
33] H. Huang, Y. Xu, Q. Feng, D.Y.C. Leung, Catal. Sci. Technol. 5 (2015) 2649–2669.
34] P. Pal, S.K. Pahari, A. Sinhamahapatra, M. Jayachandran, G.V.M. Kiruthika, H.C.
Bajaj, RSC Adv. 3 (2013) 10837.
1
8

List of Figures Captions
Figure 1. SEM images of (A) Polystyrene, (B) MaAl
2
O
3
support, (C) NiK/MaAl
2
3
O , (D)
NiK/10Ce
x
Zr1-xO
2
-MaAl
2
O
3
, (E) NiK/20Ce
Zr1-xO
x 2
-MaAl , (F) NiK/30Ce
2
O
3
x
Zr1-x
O -
2
MaAl catalysts.
2
O
3
Figure 2. XRD patterns of fresh NiK/yCe
x
Zr1-x
O
2
-MaAl
2
O
3
catalysts. Symbols represent the
,  Ni,  - Al
XRD characteristic peaks of each phase: Ce
x
Zr1-x
O
2
,  NiAl
2
O
4
2
3
O .
Figure 3. Raman spectra of catalysts: (A) fresh NiK/yCe
reduction) and (B) spent NiK/yCe -MaAl catalysts.
x
Zr1-xO
2
/MaAl
2
O
3
catalysts
(
x
Zr1-xO
2
2
O
3
Figure 4. Raman spectra of (a) 30Ce
x
Zr1-x
O
2
-MaAl
2
O
3
after calcination, (b) 30Ce
O after reduction.
x
Zr1-x
O -
2
MaAl after reduction, and (c) NiK/30Ce
2
O
3
x
Zr1-x
O
2
-MaAl
2
Figure 5. Schematic illustration of proposed pathway for cracking of VR with steam over
NiK/yCe -MaAl catalysts.
x
Zr1-xO
2
2
O
3
Figure 6. (A) XRD patterns of spent NiK/yCe
x
Zr1-x
O
2
-MaAl
2
O
3
catalysts (Symbols represent
,  NiAl , 
) and comparison of Ce and Ce peaks in XRD patterns of (B) fresh NiK/yCe
the XRD characteristic peaks of each phase:  Graphite,  NiO,  Ce
Zr1-xO
x 2
2
O
4
3
+
4+
- Al
2
O
3
x
Zr1-
O
x 2
-MaAl
2
O
3
catalysts and (C) spent NiK/yCe
x
Zr1-xO
2
-MaAl
2
3
O catalysts.
1
9

Figure 1
2
0

Fig 2
2
1

ure 2
Figure 3
2
2

Figure 4
2
3

Figure 5
2
4

Figure 6
2
5

List of Tables
Table 1. Textural properties of NiK/Ce
x
Zr1-x
O
2
-MaAl
2
O
3
catalysts
Ni particle NiAl
2
O
4
/Ni
Surface
Pore
size
(nm)
Pore
volume
(cm /g)
a
size
(nm)
species b
Sample
area
2
3
(m /g)
(Mole ratio)
MaAl
2
O
3
211.6
134.5
7.4
7.9
0.48
0.27
-
-
7
7
.1
.3
0.39
0.08
0.07
0.09
NiK/MaAl
2
O
3
NiK/10Ce
NiK/20Ce
NiK/30Ce
x
Zr1-x
Zr1-x
Zr1-x
O
2
O
2
O
2
-MaAl
-MaAl
-MaAl
2
2
2
O
3
O
3
O
3
109.3
100.8
80.9
8.4
7.6
8.0
0.27
0.25
0.19
1
1.0
x
x
1
4.5
a: the values were obtained from TEM images (Fig. S3 in Supplementary Materials).
b: the values were simply calculated on the basis of the XRD patterns.
2
6

Table 2. Compositions and lattice constants of Ce
x
Zr1-xO
2
estimated from the standard XRD patterns
Experimental data
Sample
Standard data
(111) peak position
Estimated composition
nd
Standard JCPDS
CeO2
(111) peak position
NiK/MaAl
2
O
3
nd
28.54
28.87
29.25
29.89
NiK/10Ce
NiK/20Ce
NiK/30Ce
x
Zr1-x
Zr1-x
Zr1-x
O
2
O
2
O
2
-MaAl
-MaAl
-MaAl
2
2
2
O
3
O
3
O
3
28.64
28.76
28.91
Ce0.92Zr0.08O2
Ce0.83Zr0.17O2
Ce0.73Zr0.27O2
Ce0.75Zr0.25O2
Ce0.50Zr0.50O2
Ce0.36Zr0.54O2
x
x
nd:
not
dectected
Table 3. Estimated ICe3+/Ce4+ and ICe- /Ce-O values in fresh and spent catalysts
Sample
ICe3+/Ce4+ (XRD)
ICe- /Ce-O (Raman)
Fresh catalysts
Spent catalysts
Fresh catalysts
Spent catalysts
NiK/MaAl
2
O
3
-
-
-
-
NiK/10Ce
NiK/20Ce
NiK/30Ce
x
Zr1-x
Zr1-x
Zr1-x
O
2
O
2
O
2
-MaAl
-MaAl
-MaAl
2
2
2
O
3
O
3
O
3
0.64
0.35
0.29
0.52
0.29
0.24
0.54
0.29
0.36
0.29
0.23
0.21
x
x
2
7

Table 4. Product distributions and conversions obtained from cracking of VR with steam over NiK/yCe
x
Zr1-xO
2
-MaAl
2
3
O catalysts in
o
fixed-bed reactor (Reaction temperature = 500 C, Reaction time = 2 hours, time factor, W/F
R
, = 0.6 h)
Products (%)
Samples
Liquid
Coke (%)
yield Conversion
Gas
Gasoline Diesel
VGO
21.11
39.49
41.77
47.27
50.17
42.05
Residue
76.03
42.74
20.63
21.34
21.72
33.78
(%)
Vacuum residue (VR)
No catalyst (Blank test)
0
0
2.86
0
-
-
5.5
1.49
6.23
6.28
6.36
7.06
9.14
1.64
1.61
1.78
2.96
1.58
50.12
64.32
70.59
73.85
62.56
57.26
79.37
78.66
78.28
66.22
NiK/MaAl
2
O
3
13.44
6.29
1.47
2.08
16.32
17.04
17.32
13.45
NiK/10Ce
NiK/20Ce
NiK/30Ce
x
Zr1-x
Zr1-x
Zr1-x
O
2
O
2
O
2
-MaAl
-MaAl
-MaAl
2
2
2
O
3
O
3
O
3
x
x
2
8

Table 5. Gas products distribution obtained from cracking of VR with steam in fixed-bed
reactor at t=1 hour in presence of NiK/yCe
x
Zr1-xO
2
-MaAl
2
O
3
catalysts (reaction
o
temperature = 500 C, time factor, W/F
R
, = 0.6 h)
Gas products (%)
C1 CO
Samples
2
C2
C3
C4
NiK/MaAl
2
O
3
18.10 16.81 26.39
20.21 24.54 18.16
32.13 32.27 9.17
25.70 36.38 17.19
22.32
23.36
23.09
16.43
16.38
13.73
3.34
4.30
NiK/10Ce
NiK/20Ce
NiK/30Ce
x
Zr1-x
Zr1-x
Zr1-x
O
O
O
2
2
2
-MaAl
-MaAl
-MaAl
2
2
2
O
3
O
3
O
3
x
x
2
9