Journal of Industrial and Engineering Chemistry 43 (2016) 1–12

Contents lists available at ScienceDirect

Journal of Industrial and Engineering Chemistry

journal homepage: www.elsevier.com/locate/jiec

Review

A review on the oil-soluble dispersed catalyst for slurry-phase

hydrocracking of heavy oil
Manh Tung Nguyena,b , Ngoc Thuy Nguyena,b , Joungmo Choa,b , Chulwoo Parka ,
Sunyoung Parka , Jinhwan Junga , Chul Wee Leea,b,*
a
Division of Green Chemistry & Engineering Research, Korea Research Institute of Chemical Technology (KRICT), Daejeon 305-600, Republic of Korea
b
School of Science, University of Science and Technology (UST), Daejeon 305-333, Republic of Korea

A R T I C L E I N F O A B S T R A C T

Article history:
Received 27 May 2016 Nowadays, unconventional-oil becomes a potential candidate for satisfying the world's energy demands
Received in revised form 29 July 2016 due to the scarcity of other energy sources. However, it contains many impurities, such as heavy metal,
Accepted 30 July 2016 sulfur-compounds, and nitrogen-compounds, and leads to quick deactivation of catalyst, high coke
Available online 9 August 2016 formation, and large pressure drop during the operation of a fixed bed or E-bed. Slurry phase
hydrocracking (SHC) with the presence of oil-dispersed catalyst has been proven to be the best solution to
Keywords: overcome those problems. In oil-dispersed, the metal precursor, dispersion ability, and additive strongly
Slurry-phase hydrocracking affect the catalyst performance, and are all reviewed in this paper.
Oil-dispersed catalyst
ã 2016 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights
Heavy oil process
reserved.

Contents

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Oil-soluble dispersed catalyst activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Active species in the catalytic environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Reaction mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Reaction kinetic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Catalysts and catalyst development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Metal precursors in oil-soluble dispersed catalysts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Hydrogen donor for catalyst in slurry-phase hydrocracking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Optimization oil-soluble dispersed catalysts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Catalyst deactivation and regeneration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Conclusions and future prospective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Acknowledgment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

Introduction predicted to increase by 34% between 2014 and 2035 with an

average growth of 1.6% per year [1]. Moreover, fossil fuels
Nowadays, with the sharp increase in world population and the represented 81% of the primary energy fuel in 2010 and are
rapid development of technology, the global energy demand is predicted to stay at 75% in 2035 (Fig. 1) [2]. Fossil fuel can be simply
becoming a critical issue. The primary energy demand has been classified into the following three types: oil, natural gas and coal.
While coal and natural gas can be used as energy source with
minimal preparation after exploitation, oil often requires many
* Corresponding author at: Division of Green Chemistry & Engineering Research,
other treatments before turning to be valuable products. Generally,
Korea Research Institute of Chemical Technology (KRICT), Daejeon 305-600, oil can be classified to four types: conventional oil, heavy oil, extra
Republic of Korea. heavy oil and oil sands bitumen. Currently, conventional oil is used
E-mail addresses: chulwee@krict.re.kr, chulwee2304@gmail.com (C.W. Lee).

http://dx.doi.org/10.1016/j.jiec.2016.07.057
1226-086X/ã 2016 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.
2 M.T. Nguyen et al. / Journal of Industrial and Engineering Chemistry 43 (2016) 1–12

Fig. 1. Primary global energy demand by source: projections for 2035 [2].

as the main source for the petrochemical industry; however up-take of hydrogen; as the results suppress the coke formation,
conventional oil’s capacity is just 30% in total [3] and that source increase total conversion and enhance quality of liquid product.
will probably be exhausted in the near future.
Over the last few decades, interest in unconventional oil has Oil-soluble dispersed catalyst activity
been constantly increased. Processes to convert unconventional oil
especially heavy oil into better quality products such as middle Currently, the research on the mechanism of SHC is still very
distillate products are being widely investigated. Heavy oil has low limited due to the complexity of feedstock. In slurry-phase
American Petroleum Institute gravity (API) and low value; it is a processes, free radicals are created by a thermal cracking reaction
combination of saturate, aromatics, resins, and asphaltenes with of heavy oil components such as asphaltenes and resins, and then
heavy metals, nitrogen, sulfur, oxygen, and other compounds. the hydrogenation occurs on the metal active sites to deactivate
Currently, there are three main processes for the hydro- free radicals [9]. The highly dispersed of oil-soluble dispersed
conversion of heavy oil: fix-bed, ebullated bed and slurry phase. catalyst acts as hydrogen transfer agent and deactivates the
Among them, the slurry phase process (SHC) is the latest intermediate free radical moieties in the liquid phase, thereby
technology with many advantages such as the following: the preventing coke formation and polymerization reactions, and
ability to obtain more than 95 wt.% conversion of vacuum residue reducing the yield of gas products [10–12] (Fig. 2).
and other type of feedstock, including coal. In addition, the low
coke formation benefits in reducing the loss of liquid product in the Active species in the catalytic environment
sediment. Furthermore, the high quality products can be obtained
through a two-stage design that incorporates a fixed-bed hydro- Metal sulfide has been proved to be the active specie in
treating/hydrocracking reactor; flexibility of feedstock; low-cost, hydrocracking processes. It is formed by the reaction between
micron-size deposable catalysts; and a required reactor volume metal precursors in the catalyst and sulfur compounds in the
that is lower than for the E-bed process for a given volume of feedstock or additional sulfur sources [13]. What Cyr et al. [14]
residue feed [4]. have done so far is to conduct SHC experiment with oil-soluble
In the slurry phase hydrocracking process, catalyst can be dispersed catalyst and metal-containing compound additives
divided into two types: heterogeneous solid powder catalysts and (such as iron pentacarbonyl or molybdenum 2-ethyl hexanoate).
homogeneously dispersed catalysts. As the name, heterogeneous They realized that metal compounds might react in situ with sulfur
solid powder catalyst is the powder that contains at least one metal moieties in the feedstock to create the colloids. Then, because of
(cobalt, molybdenum, nickel, iron) in the salt form or sometime, the wetting property, the colloidal particles would stick on the
they are impregnated in a supported material such as alumina, surface of precursor spheroids and inhibit the spheroids from
silica–alumina or carbon nano tube. Heterogeneous solid powder coalescing.
catalyst involves system in which catalyst and reactants are in The activities of metal sulfide have been proposed by Daage and
different physical phases. When the reaction occurs, the reactant Chianelli [15]. Its activity mostly depends on the disordered edge
having high molecular weight deposits on the active sites of planes in the reaction environment. The edge planes often have a
catalyst and blocks the pores; leads to shortening of the catalyst’s high activity, while the basal planes seem to be inert (Fig. 3). In
life [4] and increasing of pressure drop because of the formation of
larger number of solid particle. As the result, operation is
instability and the system is shut-down more frequently.
Unsupported dispersed catalyst has been developed to solve
those problems. It consists of the two catalyst types: water-soluble
dispersed catalyst and oil-soluble dispersed catalyst. Generally, oil-
soluble dispersed catalyst is often more expensive than water-
soluble dispersed catalyst; however, water-soluble dispersed
catalyst has lower catalytic activity due to the fast evaporation
of water and agglomeration of the precursors, finally forming large
particles and causing difficulty to disperse into submicrometer
particle in the feed [5–8]. Meanwhile, oil-soluble dispersed
catalyst can disperse uniformly in the feedstock; favor the rapid
Fig. 2. Catalytically controlled thermal cracking [12].
 M.T. Nguyen et al. / Journal of Industrial and Engineering Chemistry 43 (2016) 1–12 3

For more detail, a series of studies based on the activity of MoS2

were carried out [19–24]. These studies found that the hexagonal
coordination exhibited by unsupported MoS2 is the main factor to
provide hydrogen uptake function and suppress coke formation.
During the reaction, hydrogen can react easily with sulfur ions in
MoS2 that are located in the corner and edge. As a result, the
unsaturated sites and sulfur ion vacancies are formed and then
become active sites for hydrogenation due to the Lewis acid
characteristic. Hydrogen moves to the surface of active specie and
forms Mo–H and S–H moieties. Those connections are unstable
and can react rapidly with the free hydrocarbon radicals that are
generated by thermal cracking and finally decrease the rate of
condensation and aromatization reactions.

Fig. 3. Rim–edge theory [15]. Reaction mechanism

their model; the rim site shows the hydrogenation function and To determine mechanism of reactions in SHC, Du et al. [13]
the rupture of the C

S bond while the edge sites are active for only experimented on Karamay vacuum gas oil (KLVGO), Karamay
the rupture of the C

S bond. vacuum residue (KLVR) and Venezuelan atmospheric residue
During the SHC reaction, the unsupported MoS2 exhibits a (V-AR) with molybdenum naphthenate acting as an oil-soluble
hexagonal coordination, and then the corner and edge ions in MoS2 dispersed catalyst precursor. The product was analyzed to
are removed quickly. As a result, the coordinatively unsaturated determine the value of two effect parameters RG (the ratio of
sites (CUS) and sulfur ion vacancies are formed. Due to the Lewis i-C4H10 and n-C4H10) and isoparaffin/n-paraffin ratio. The small
acid characteristic of sulfur ion vacancies, CUS can easily adsorb value of these effect parameters indicates that the reaction creates
molecules with unpaired electrons. fewer isomerization products; the reaction is followed by a free
In a catalytic environment, it is obvious that fewer layers could radical mechanism. Otherwise, the reaction is followed by
enable more benefit in total catalyst activity. However, due to the carbonium ion mechanism. The gas product data can be seen in
anisotropic layered stacked structure and the interaction of the the Table 1.
support with active phases, investigations on catalyst dispersion It is obvious that the gas product from thermal cracking and
have been plagued with many difficulties. Liang et al. [16] catalytic cracking have the same composition, and paraffin yields
established a procedure for quantitatively determining the size were above 96% volume including a 73–78% volume yield of
of MoS2 slabs from the analysis of the diffraction pattern of poorly methane and ethane. Therefore, the mechanism of slurry phase
crystalline MoS2. As a result, they found that by increasing the hydrocracking is the same as that of thermal hydrocracking,
pressure, the catalyst could completely destack and the activity of a namely the free radical mechanism.
catalyst that consists of a single layer of MoS2 decreased only Normally, the catalyst is activated before hydrocracking
slightly. The well dispersion and single layer of MoS2 can be seen in reaction. That procedure is often named the sulfidation; there
Fig. 4 [17]. are two kind of sulfidation: ex-situ and in-situ sulfidation. Oil-
On the other hand, Strausz [7,18] used molybdenum naph- dispersed catalyst precursors are mixed with carbonaceous
thenate and nickel di-2-ethylhexanoate as metal precursors for compound; under suitable thermal conditions, it is decomposed
SHC. Those precursors were chemically reacted with the micellar to solid, nano-sized catalyst particles. The sulfur sources are often
dispersed asphaltenes in the bitumen and formed the metal in the carbonaceous compound, but sometime, hydrogen sulfide is
complexes. During the hydrocracking reaction, micelle–metal used, especially in the case of ex-situ sulfiding. Temperature of
complex breaks down and form metal oxide/sulfide active species. sulfidation is ranged from 300 to 400  C [25], and hydrogen
However, not all of the complexes are broken down; the micelles pressure is about 1000–2500 psig.
that have strong affinity between metal and asphaltenes are very As what has been mentioned before, after sulfidation, the
stable and cannot be decomposed under the reaction condition. As coordinatively unsaturated sites, sulfur ion vacancies are both
the results, it will be kept inside the coke. formed and distributed uniformly. Then, the mixture of catalyst

Fig. 4. High-resolution transmission electron micrographs (HRTEM) of the THF insoluble fraction of the recycle stream of the EST process, showing the dispersion of the
catalyst (left) and a single MoS2 layer with a superimposed atomic model (right) [17].
4 M.T. Nguyen et al. / Journal of Industrial and Engineering Chemistry 43 (2016) 1–12

Table 1
Gaseous product distribution of feedstock oils at thermal hydrocracking and catalytic hydrocracking [13].

Product yield of KLVGO (vol%) Product yield of KLVR (vol%) Product yield of V-AR (vol%)
a b
TH CH TH CH TH CH
CH4 57.10 56.04 58.13 58.19 59.17 58.62
C2H6 18.04 17.78 18.90 16.66 17.99 17.90
C2H4 0.29 0.27 0.26 0.26 0.19 0.19
C3H8 13.37 13.51 13.51 13.35 13.45 13.61
C3H6 0.63 0.98 0.69 0.67 0.82 0.94
i-C4H10 1.86 1.82 2.32 2.25 1.92 1.85
n-C4H10 2.55 2.71 2.75 2.94 2.56 2.77
trans-2-C4H8 0.18 0.23 0.10 0.10 0.14 0.12
1-C4H8 0.20 0.27 0.12 0.12 0.13 0.13
cis-2-C4H8 0.15 0.10 0.15 0.14 0.13 0.11
i-C5H12 2.15 2.49 1.24 1.33 1.29 1.35
n-C5H12 1.82 2.21 0.78 0.85 0.87 0.91
cis-2-C5H10 0.03 0.04 0.20 0.16 0.14 0.18
CO 1.14 1.13 0.56 0.62 0.76 0.91
CO2 0.45 0.37 0.33 0.32 0.41 0.45
C6+ 0.04 0.05 0.06 0.04 0.03 0.06
a
Thermal hydrocracking.
b
Catalytic hydrocracking.

will be transferred to mix again with the feedstock in the case of

ex-situ sulfidation; or reaction will carry out directly by continuing
increase temperature. At higher temperature, the thermal cracking
happens. The poly-aromatic compounds are being broken down in
the weak positions such as C

C, C

S, C

O and etc. and finally
become the smaller free radicals. Meanwhile, H2 molecule moves
to the active sites and forms the intermediate connections Mo–H or
S–H; those connections are not stable and easily to give hydrogen
free radical to other compounds. Consequently, deactivation of free
radical along with the saturation of unsaturated compound is
occurred constantly. Thermal hydrocracking and catalytic hydro-
cracking have been reported to occur in cascade [26]. It means the
vacuum residue is hydrocracked gradually to form the lighter
products following the order: vacuum gas oil, middle distillate,
Fig. 5. Proposed kinetic mode for the hydrocracking of heavy oils [28].
naphtha and so forth. The efficiency of a hydrocracking reaction is
made by the well combination of thermal cracking and hydro-
genation to obtain the desired products. Vaccum gas oil ðVGOÞ : rVGO ¼ k1 y2R
ðk5 þ k6 þ k7 ÞyVGO ð2Þ
In summary, with the presence of oil-dispersed catalyst
precursors, coke and gas yields decrease significantly while the
quality of the liquid product improves. Metal sulfide has been Middle distillate : rMD ¼ k2 y2R þ k5 yVGO
ðk8 þ k9 ÞyMD ð3Þ
formed by the ex-situ or in-situ reaction; and then provides
hydrogenation function. Those active sites saturate aromatics in
the feed, saturate olefins, deactivate the free radicals formed in the Naphtha : rN ¼ k3 y2R þ k6 yVGO þ k8 yMD
k10 yN ð4Þ
cracking, inhibit condensation and polymerization and finally
protect themselves from poisoning by coke [27].
Gases : rG ¼ k4 y2R þ k7 yVGO þ k9 yMD þ k10 yN ð5Þ
Reaction kinetic
The kinetic model was incorporated into an isothermal plug-
Most of kinetic models have been reported in the literature flow reactor model to ignore all the limitations in the mass
propose that the hydrocracking reactions of heavy oil are followed transfers. Then they were solved using Runge–Kutta method.
by the first-order reaction [28]. However, Sanchez and Ancheyta dyi
[29] has presented another point of view when they considered the 1

¼ ri ð6Þ
d LHSV
second-order for the hydrocracking of vacuum residue due to the
first and quick disappear of the most reactive compound in the Finally, all the kinetic parameters were determined using the
residue. optimization of Marquardt [31] and have been shown in the Table 2
The five-lump kinetic model has been used to investigate together with the corresponding activation energies.
kinetic of hydrocracking reaction (Fig. 5) [28]. From those results, some important comments can be pointed
With the assumptions that hydrocracking of vacuum residue out. First, naphtha cracking is extremely difficult to occur and can
follows second order whiles other reactions follow the first order be observe in the investigated conditions. Second, the kinetic
[29,30]. The reaction rates for each reactant have been estimated parameters of residue hydrocracking exhibited the following
by the following equations [29]: order: k1 > k2 > k4 > k3, indicated high selectivity toward VGO,
followed by middle distillates and gases, with naphtha having
Residue : rR ¼
ðk1 þ k2 þ k3 þ k4 Þy2R ð1Þ the lowest selectivity.
 M.T. Nguyen et al. / Journal of Industrial and Engineering Chemistry 43 (2016) 1–12 5

Table 2
Kinetic parameters of the proposed model [29].

Pressure = 6.9 MPa Pressure = 8.3 MPa Pressure = 9.8 MPa

a a
Rate constant value Activation energy, EA Rate constant value Activation energy, EA Rate constant valuea Activation energy, EA
(kcal/mol) (kcal/mol) (kcal/mol)
        
380 C 400 C 420 C 380 C 400 C 420 C 380 C 400 C 420 C
Residue
k1 0.0012 0.0035 0.0129 53.3 0.0022 0.0047 0.0197 49.0 0.0025 0.0060 0.0206 47.7
k2 0.0012 0.0029 0.0128 53.1 0.0011 0.0057 0.0215 66.3 0.0011 0.0039 0.0171 62.4
k3 0.0002 0.0008 0.0028 59.0 0.0003 0.0010 0.0029 51.2 0.0004 0.0009 0.0031 45.9
k4 0.0013 0.0019 0.0058 33.4 0.0008 0.0023 0.0054 43.6 0.0005 0.0021 0.0040 31.2

VGO
k5 0.0025 0.0060 0.0135 37.9 0.0048 0.0055 0.0244 36.5 0.0070 0.0150 0.0353 36.3
k6 0.0011 0.0030 0.0081 45.9 0.0016 0.0043 0.0086 37.4 0.0022 0.0055 0.0090 31.8
k7 0.0008 0.0017 0.0036 35.1 0.0009 0.0019 0.0038 33.0 0.0010 0.0021 0.0040 31.2

Middle distillate
k8 0.0027 0.0060 0.0127 34.8 0.0028 0.0065 0.0146 37.5 0.0028 0.0070 0.0166 39.8
k9 0 0 0 0 0 0 0 0 0

Naphtha
k10 0 0 0 0 0 0 0 0 0

1
a
The rate constants for the residue are given in units of wt.% h , whereas the rate constants for the VGO, middle distillates, and naphtha are given in units of h
1.

1

Fig. 6. Comparison between experimental and estimated product compositions at

9.8 MPa: () 380  C, (*) 400  C, and (&) 420  C [28]. Fig. 7. Configurations of SPR: (a) slurry-phase reactor; (b) bubble column reactor;
(c) stirred tank reactor [32].

The comparison between experimental and calculated product Generally, slurry phase hydrocracking with oil-soluble
composition obtained by solving Eqs. (1)–(5) is shown in Fig. 6. dispersed catalyst are often operated between 420–450  C. Metal
The deviations between both the experimental and estimated removal can be increased when temperature is increased, however,
values were lower than 5% on average. It indicates that the five- it is has side effect when coke formation is also increased.
lump kinetic model is found to represent adequately to the Hydrogen pressure is often kept from 1000 to 2500 psig. Increasing
hydrocracking of heavy oil. hydrogen pressure causes the decrease in coke yield [33,34] but it
seems to be not economical. Catalyst concentration is another
Catalysts and catalyst development important issue, depends on the activity of metal, catalyst
concentration might be varied between 300–2500 ppm. Resident
Nowadays, when all of the convenient oil sources are being time ranges from 20 to 180 min [35]. Some of the typical
exhausted, the inconvenient oil sources are getting dirtier, and advantages and disadvantage of SPR are listed as following Table 3
require a high process technique to satisfy the world energy’s [10,32].
demands. Currently, slurry phase hydrocracking is the latest In the slurry phase hydrocracking technique, catalysts always
upgrade heavy oil technology. It has been investigated and attract huge interest from the scientists. Thus far, oil-soluble
perfected for decades and gradually affirmed its important dispersed catalyst has proved to be the most excellent catalyst with
position in the future of petroleum industry. almost full conversion, the ability to work in super high impurity
Slurry phase reactors (SPR) are worked based on the fact that environments and low coke formation. However, the synthesis cost
catalyst is suspended in a liquid in batch mode or it might be move is the most disadvantage of oil-soluble dispersed catalysts. In this
concurrently or counter-currently with the gas flow. Nowadays, vein, many papers and patents have been published; as the
three main kinds of reactors are used in laboratory and industrial scientists have looked for cheaper and better metal precursors, or
scale; those are: slurry phase reactors, bubble column reactor and tried to improve catalyst activity by combining different metals, or
stirred tank reactor (Fig. 7a) [32]. by adding some additives.
6 M.T. Nguyen et al. / Journal of Industrial and Engineering Chemistry 43 (2016) 1–12

Table 3
Typical advantages and disadvantages of slurry phase reactors [10,32].

Advantages Disadvantage
 Operate almost isothermally  Uncertain scale up
 Control temperature easily  Catalyst sedimentation
 Interphase contact well  Catalyst recovery and reuse
 Large liquid hold up
 High conversion rate
 Operational flexibility
 Low pressure drop
 Low construction and operational costs
 Low catalyst consumption compares with other processes.
 Easy addition of catalyst

Metal precursors in oil-soluble dispersed catalysts maltenes. Under hydrotreatment conditions, MoS2 is formed. Mo-
DTC has higher sulfiding efficiency than Mo-DTP and is expected to
Oil-soluble dispersed catalyst precursors are created by the be more active. He also found that MoS2 or any type of solid Mo
combination of an oxide, a sulfide or a salt of metal from group IV derived from those catalysts are very fine particles with low
through VIII including transition metal-based catalysts derived crystallinity and are easily dispersed in feedstock.
from the organic acid salt or metal–organic compounds of Mo, V, Beside the catalysts based on Mo, many scientists have tried
W, Cr, Fe etc. [36]. with different metal precursors. Zhang et al. [49] performed slurry
Table 4 below shows some typical oil-soluble catalysts phase residue hydrocracking with nickel oil-soluble dispersed
precursors that have been used so far. catalyst and a feedstock of LiaoHe atmospheric residue. Fig. 8
Among many metal precursors, molybdenum seems to have shows the results of thermal hydrocracking and catalytic
attracted the most interest due to its high hydrogenation activity in hydrocracking with 300 mg nickel/g feedstock.
sulfide form compared with others. Hooman et al. [47] conducted a It is obvious that the presence of nickel oil-soluble dispersed
series of SHC experiments with Cold Lake vacuum residue using catalyst significantly improved the product quality and decreased
Mo-octoate and Mo-micelle precursors. The reactions have been the coke yield. By experiment, the author also concluded that the
carried out in 1 h at 415–445  C and 13.8 MPa of hydrogen. conversion of feedstock to distillate is thermally controlled and
Consequently, coke suppression, residue conversion and hydrogen practically independent of both the catalyst concentration and
uptake were observed. In addition, the optimization catalyst hydrogen pressure. In addition, the active catalytic species of
concentration was also found at 600 ppm with 84 wt.% of residue dispersed catalyst are the different nickel sulfide crystals:
conversion and 2.9 wt.% of coke compared with 22 wt.% of coke in hexagonal NiS, millerite NiS and cubic Ni3S4.
thermal hydrocracking under the same conditions. A series of different catalysts based on different metals was
Watanabe et al. [48] investigated the activities of used in the research of Bearden [50]. Those catalyst precursors
Mo-dithiocarbamate (Mo-DTC) and Mo-dithiophosphate (Mo- were nickel octoate, cobalt resinate, iron naphthenate, molybde-
DTP) in vacuum residue. In his research, the Mo complexes and num resinate, vanadium resinate and chromium resinate, and they
derived MoS2 were found to be distributed in both asphaltenes and were mixed separately with heavy oil (Jobo and Cold Lake Crude).

Table 4
Oil-soluble dispersed catalysts precursors for slurry phase hydrocracking.

Licenser Catalyst components Feed Catalyst concentration Main results Ref

no.
Headwaters Technology Ni, Mo etc. oil-soluble PFO 4000 ppm BTX yield 17.0%; gasoline and diesel yield [37]
Innovation, Llc catalysts 72.3%
University of British MoS2 catalyst Vacuum residue 100–1800 ppm Liquid yield achieved 88% [38]
Columbia, Vancouver,
Canada

Exxon Research and Molybdenum alicyclic or Heavy oil with 50–200 ppm 50 wt.% reduction of CCR [39]
Engineering Co. naphthenate CCR > 5% Solid, non-colloidal catalyst
Fe2O3 and Mo naphthenate Cold Lake crude oil 50–200 ppm in-situ 50 wt.% reduction of CCR, coke yield 1% [40]
Iron molybdenum Cold Lake crude oil 0.5–2.0 wt.% Conversion > 50% [41]
Solid particle with low SA and
PV
CrO3tert-butyl alcohol Heavy oil with CCR 5– 0.1–2.0 wt.% Conversion 80–85 wt.% [42]
50 wt.% Solid chromium-containing
catalyst
Alberta Oil Sands Technology Iron pentacarbonyl or Athabasca bitumen 0.1–0.5 wt.% well-dispersed 90% Conversion 0.3 wt.% coke yield [43]
and Research Authority molybdenum 2-ethyl with 50 wt.% diluent colloidal particles
hexanoate
Mo, Ni acetylacetonates or 2- Athabasca bitumen 50–300 ppm mixture of Low coke yield [44]
ethyl hexanoate asphaltene and metal doped
coke
Universal Oil Products Co. Non-stochiometric vanadium Wyoming sour crude Well-dispersed colloidal High Ni, V removal [45]
sulfide oil particles
Headwaters Technology Bimetallic catalyst (CoMo, Heavy oil 100 ppm metal Increase asphaltene conversion (up to [46]
Innovation NiMo, FeMo . . . ) 86%) and reduce coke (down to 1.18%)
 M.T. Nguyen et al. / Journal of Industrial and Engineering Chemistry 43 (2016) 1–12 7

As has been mentioned before, one of the most disadvantages of

oil-soluble dispersed catalyst is high production cost. In order to
overcome that problem, the mixtures of oil-soluble dispersed
catalyst and other cheaper catalysts have been considered. The
mixture of oil-soluble dispersed catalyst and heterogeneous
catalyst has been tested [52]. In that research, Mo oil-soluble
dispersed was used as the homogeneous part, and the heteroge-
neous part was Co–Mo catalyst on g-Al2O3 support. Surprisingly, a
rapid decrease in the conversion of feedstock was observed as the
amount of homogeneous catalyst in the reactor increased. It
appears as though the supported catalyst ties up some of the
dispersed Mo, thereby deactivating the supported catalyst; as a
result the conversions are similar to runs with just-dispersed Mo,
but at a lower concentration.
The ideal of using bimetallic catalyst is attracting considerable
interests. Wang et al. [53] investigated a hydrocracking reaction
Fig. 8. Effect of nickel oil-soluble dispersed catalyst on slurry-phase hydrocracking with the combination of molybdenum based oil-soluble dispersed
[49]. and iron-based water-soluble dispersed catalyst precursors. He
found out that water-soluble dispersed catalyst not only inhibits
coke formation but sacrifices the sites of coke deposition.
The active species were formed in-situ. After one hour reaction, Consequently, the active Mo site could be saved from the
generally, all of the metals were effective in not only suppression of deposition of coke and finally improve the quality of the product.
coke and gas formation but also showing the high levels of The mixture of two oil-soluble dispersed catalysts has been
desulfurization, demetallization and Conradson carbon conver- tested for SHC of LiaoHe Vacuum Residue [54]. A series of
sion. However, the molybdenum gave the best results in terms of experiments used Co-naph or Ni-naph or their mixture as the
liquid yield (95.8–97.2 wt.%) and coke (0.5–0.8 wt.%). Oil-soluble catalyst. In the experiments, nickel showed better performance
dispersed catalyst based on iron has the greatest advantage in than cobalt, and the presence of Ni–Co combination just produced
terms of cost but seems to be not very effective, with the highest a simple mathematical addition effect on the coke inhibition and
coke formation (5.2 wt.%) and gas yield (2.8 wt.%) (Fig. 9). conversion of feedstock. However, the combinations of Co or Ni
Panariti et al. [51] has also worked with various types of oil- precursor with Mo precursor have been proven to greatly benefit
soluble dispersed catalysts. The result is shown in the Table 5. The both the inhibition of coke and the product quality [46]. Cobalt
catalytic performances of the tested metals follow the order: precursor was chosen as the second transition metal and was
Mo > Ni  Ru > Co > V > Fe. The reason for the difference in catalyst mixed with molybdenum precursor with a mole ratio varied from
activity might due to the dissimilar in the active form of 1:98 to 1:10. The inventors found that the active species of metal in
molybdenum and the others. Molybdenite disperses into feedstock the vacuum residue are CoS and CoMoS; and that their combina-
in the form of monolayer or crystallites with a limit of two or three tion benefited not only economic but also enhanced asphaltene
layers; the average dimensions of the layer are only about 20–40 Å, conversion and reduced sediment. The reason for that phenome-
while the other metals often form a three-dimensional pyrrhotite- non has been explained by the synergetic effect of Mo and Co. Due
like crystalline structure with a crystal size of 200–300 Å. This to the electronic configurations, [Kr] 4d55s1 (cobalt) has a stronger
means there is a significantly different surface area, which results attraction to sulfur than [Ar] 3d74s2 (molybdenum). Finally, when
in much lower in catalyst activity. the molecular complex of those metals is formed, cobalt becomes a

Fig. 9. Performances of different monometal oil-soluble dispersed catalysts [50].

8 M.T. Nguyen et al. / Journal of Industrial and Engineering Chemistry 43 (2016) 1–12

Table 5
Effect of active metals on product yield and quality [51].

Catalyst precusor No catalyst Mo naph Fe naph Ni naph CoRe VRe RuAA

Feedstock conversion (wt.%) 89.6 84.9 87.6 83.2 83.7 85.2 84.1

Product distribution (wt.%)

C1-C4 10.9 7.3 10.0 12.6 11.7 13.2 11.7
C5- 200  C 33.5 38.5 35.3 32.9 35.4 32.0 30.7
200–350  C 16.6 23.8 16.2 21.2 19.2 19.6 21.8
350–500  C 6.7 12.0 6.9 11.5 9.4 8.2 11.8
> 500  C 10.4 15.1 12.4 16.8 16.3 14.8 18.6
Coke 21.9 3.3 19.2 5.0 8.0 12.2 5.4
C5-asphaltenes 8.9 14.3 10.5 13.5 14.1 13.9 15.3

Product quality
HDM (%) >99 82 98 87 89 93 78
S in 200–350  C 3.2 1.9 3.2 2.3 3.0 2.9 2.3
S in 350–500  C 4.9 2.6 4.8 3.3 4.5 4.4 3.3

promoter, attracts the sulfur and then transfers it to the

molybdenum molecular nearby.
Jeon et al. [55], successfully chemically synthesized bimetallic
CoMo oil-soluble dispersed catalyst precursor while in the
previous studies, that kind of precursor was simply prepared by
physical mixing two different monometallic catalysts. Compared
with monometallic and physical mixtures of catalyst, the new
catalyst precursor showed a better performance, enhanced of
liquid yield and reduced of both gas and coke (Table 6).
On the other hand, the multi-metallic liquid catalyst is another
choice for the slurry-phase hydrocracking of heavy oil. Multime-
tallic liquid catalyst can also be classified into two types: water-
dispersed and oil-dispersed. The water-dispersed type often
consists of two metals; one from group VIB and another from
group VIII [56]. During the preparation process, the oxide of a
group VIB metal such as molybdenum or tungsten is dissolved in
ammonia aqueous to form an absolute solution. Then, ex-situ Fig. 10. Product distribution for SHC with multi-metallic liquid catalyst at different
sulfidation steps are carried out to form slurry then the slurry is temperature [57].
promoted with a group VIII metal, for example: nickel, cobalt; and
finally that mixture is introduced with hydrocarbon oil. In the case Consequently, the multimetallic liquid catalyst seems to be a
of oil-dispersed catalyst, the major parts of the metal components promising candidate in slurry-phase hydrocracking due to the
often include nickel, iron, molybdenum, manganese, cobalt and the cheap price and quite high conversion with low coke formation; in
like [57]. Because those metals are obtained from the recovery addition, recycling the catalyst has been proven to provide even
process of industrial waste materials, the cost would greatly better results compared with using fresh catalyst.
reduce. A series of SHC experiment with multi-metallic has been
carried out [57]. 50–200 ppm based on metal of UPC-21 has been Hydrogen donor for catalyst in slurry-phase hydrocracking
used as a catalyst; reaction temperatures were investigated from
430  C to 460  C and initial pressure of hydrogen was about In upgrading heavy oil, hydrogenation is the most important
8–12 MPa. UPC-21 is kind of metal-organic framework that has function of the catalyst. Therefore, theoretically, the presence of a
general formation: [Cu3(L)1.5(H2O)3]3DEF20H2O, where L is the hydrogen donor would be very helpful for the enhancement of
ligand [58]. As a result, conversion rate is quite high at 80–90% with hydrogen uptake, resulting in increased conversion and the
a little coke formation (<1%). Fig. 10 has shown the results of reduction of coke.
experiments that were conducted at different temperatures. Generally, the hydrogen donors in the petroleum industry could
Furthermore, a part of the bottom oil was brought back to the be classified to three types based on the different types of hydrogen
reactor to investigate recycle reactions. Table 7 shows experiment that are provided: anionic hydrogen, radical hydrogen, molecular
conditions and the main results of those experiments. hydrogen. Fisher et al. [59] conducted a series of experiments with
tetralin acting as a hydrogen donor. The results were quite
Table 6 impressive: even with a heterogeneous catalyst of CoMo based on
Performance of monometallic, physical bimetallic and chemical bimetallic oil- alumina, the conversion was about 60%, the desulphurization and
soluble dispersed catalysts [51]. demetallization reached a high level, especially for nickel and
Catalyst No catalyst Co naph Mo octo Co + Mo CoMo
vanadium, and coke formation was just around 1%.
Shi and Que [54] have tested performance of various oil-soluble
Gas yield (wt.%) 21.16 14.77 13.78 13.22 12.38
Liquid yield (wt.%) 76.27 83.77 85.41 86.00 87.08
dispersed catalysts in slurry-phase hydrocracking of LiaoHe
<200  C 19.40 14.33 16.34 16.59 16.65 Vacuum Residue (LHVR) with and without the presence of tetralin
200–350  C 22.42 23.64 22.28 23.44 24.57 (THN) which acted as a hydrogen radical donor. As a result, the
350–500  C 21.37 28.05 26.04 27.53 27.76 performance of the catalyst mixture between an oil-soluble
>500  C 13.08 19.39 19.11 18.44 18.10
dispersed catalyst precursor and hydrogen donor would effectively
Coke yield (wt.%) 2.56 1.46 0.81 0.78 0.55
inhibit coke formation and yield a good upgrade of LHVR. In
 M.T. Nguyen et al. / Journal of Industrial and Engineering Chemistry 43 (2016) 1–12 9

Table 7
Recycling experiments with multimetallic liquid catalyst. [57].

Reaction temperature,  C 440 440 445 445

Hydrogen partial pressure, Mpa 10 10 10 10
Hydrogen–oil ratio, Mm3/m3 757/1 800/1 737/1 800/1
Recycling ratio (fresh raw material/bottom oil) 100 66/34 100 70/30
Total volume hourly space velocity, 1/h 1.10 1.14 1.13 1.14
Volume hourly space velocity of fresh raw material, h
1 1.10 0.75 1.13 0.80

Product distribution, wt.%

C1–C4 (gas) 4.76 5.50 4.96 7.40
C5-180  C 9.27 9.60 10.28 13.80
180–350  C 24.08 27.30 27.41 29.60
350–524  C 37.50 53.10 37.62 45.40
<524  C 75.61 96.30 80.27 96.20
>524  C 25.39 4.60 20.90 5.00
Hydrogen loss 1.13 0.92 1.18 1.18
Total yield 101.0 100.92 101.18 101.18

addition, the conversion per coke unit was significantly reduced source does not affect the sulfidation of Mo precursor; in other
with the presence of THN (Table 8). words, the sulfur compounds existing in the feedstock provide a
The only disadvantage of using a hydrogen donor relates to the sufficient amount of sulfur for the activation of the catalyst
energy point of view. Huge amounts of hydrogen donor were used, precursor. That point of view was also confirmed in Del Bianco
from 10% [54] to 100% [58], in mass compared with feedstock; and et al’s research [35]. The addition of a phosphorus source increased
this led to the consumption a great deal of energy during the the vanadium removal in the upgrading of heavy stock due to the
reaction and in the next distillation stage. formation of a P–V oil-insoluble compound [51,60–62]. However, it
does not have much effect on hydrogen consumption, coke
Optimization oil-soluble dispersed catalysts formation or the quality of liquid product [51]. Therefore,
generally, the performance of the catalyst seems to be independent
Catalyst activity is the most important issue in any type of on the organic group bonded to the active metal.
catalytic reaction. It affects not only the total conversion yield and The ratio of hydrogen sulfide to hydrogen could greatly affect
the desired products but also on production costs. In the previous the number of active sites CUS [24]. As what has been mentioned
section, active metals were demonstrated to be a critical factor that before, the SH groups attached to the sulfided Mo and CoMo
affects the activity of catalysts and using molybdenum oil-soluble moieties seems respond to the acid property of the catalyst. Petit
dispersed catalyst with the presence of promoters such as cobalt or et al. [63] recently presented evidence for the presence of Brønsted
nickel could be more beneficial than other cases. However, there acid sites not only on the support but also on sulfided Mo, CoMo,
are still many other factors that may affect the performance of oil- and NiMo phases, using dimethylpyridine as the probemolecule.
soluble dispersed catalysts such as precursor solubility, degree of Topsøe et al. [64] and Rana et al. [65] conducted research on the
dispersion or the pretreatment of the catalyst. process to form an SH group from anion vacancies with the
Panariti et al. [51] has experimented with different oil-soluble presence of hydrogen and hydrogen sulfide (Fig. 11).
dispersed catalysts based on Mo precursor: molybdophosphonic During the operation, metal precursors are converted to sulfide
compound, dithiocarbamate complex Mo3S7-(DTC)4, MoNaph, forms; therefore it is expected to save more resident time and limit
MoAA, PMA and the two Mo-based ultrafine materials: Mo2N the deactivation if the catalyst is presulfided before contacting
and Mo2C. As a result, the hydrogen consumptions in the case of with the feed. Presulfiding could be performed either in-situ or ex-
Mo2N and Mo2C were found to be lower and the coke formations situ. Generally, ex-situ is more beneficial than in-situ for two main
were higher than other cases. This suggests that with lower reasons. First of all, in-situ sulfidation requires a complex operation
dispersion, catalyst activity would be lower. condition due to the time required to complete the sulfidation
On the other hand, the performance of the dithiocarbamate reaction [66–68] (Fig. 12).
complex has not shown a significant difference with the other oil- In addition, according to research [69] on CoMo catalyst,
soluble dispersed catalysts. This means the addition of a sulfur increasing the sulfiding temperature from 300  C to 400  C could

Table 8
Slurry-phase hydrocracking of oil-soluble dispersed catalyst with and without a hydrogen donor [54].

Catalyst, mg/g Tetralin wt.% Coke wt.% Conversion wt.% Conversion per unit coke

Co naphthalene Ni naphthalene
0 0 0 9.79 65.20 6.66
10 7.02 63.92 9.11
200 0 0 2.39 56.40 23.60
10 1.39 54.98 39.55
160 40 0 2.51 58.20 23.19
10 1.59 56.00 35.22
100 100 0 2.61 59.00 22.61
10 1.73 57.42 33.19
40 160 0 2.99 59.50 19.90
10 1.93 57.91 30.00
0 200 0 3.34 61.21 18.33
10 2.29 59.35 25.92
10 M.T. Nguyen et al. / Journal of Industrial and Engineering Chemistry 43 (2016) 1–12

Fig. 11. Formation of an SH group in a hydrocracking reaction [66].

supporting itself and, therefore, will not fall down to the bottom of
reactor. It has diameter of around 1.83 m when the reactor is open.
Shot coke is a combination of individual particles that often have
spherical or slightly ellipsoidal shape and the diameter is about
1–4 mm. The advantage of that coke is that it is easy to be removed
from the reactor; however it has lower value than sponge coke
(Fig. 13). Needle coke is a premium-grade coke that is made from
special petroleum feedstock. It has a crystalline broken needle
structure and is chemically produced through the cross linking of
condensed aromatic hydrocarbon during the coke formation.
Purge coke is produced by flexi-coking and is used as a fuel in coke-
burning boilers. Catalyst coke is the mixture of coke and catalyst. It
is formed when carbon is deposited on catalysts, is used in various
refining processes, and is burned off and used as a fuel in the
refining process.
Generally, the catalyst will be deactivated when the metal
active site is covered by carbonaceous or heavy metal or other
types of impurities [75]. In general, the formation and deposition
Fig. 12. Comparison of in-situ and ex-situ (actiCAT) presulfiding [67].
of carbonaceous or coke are the main reasons for catalyst
deactivation; however, until now, not so many researches on that
produces a major increase in the degree of sulfiding of molybde- phenomena has been published. For the fixed-bed catalyst, the
num species and results in increased activity of the catalyst. deactivation of the catalyst occurs quickly happened when the
However, on an industrial scale, in-situ presulfiding does not allow coke blocks catalyst pores [5,76], while with oil-soluble dispersed
for the optimal temperature control and optimal heat release; catalyst, it is likely different.
therefore the temperature is always kept lower 320  C to prevent In fact, the recycling of catalysts, especially oil-soluble
local overheating. dispersed catalysts in slurry phase hydrocracking often leads to
many difficulties. The major problem of recycling homogeneous
Catalyst deactivation and regeneration catalysts is that, in the end of reaction, the catalyst is mixed with
products at the nanometer scale so it is nearly impossible or very
The most disadvantage of oil-soluble dispersed catalyst costly to recover.
comparing with other types of catalyst is its price. Therefore, In the precipitate, the metal precursor changed into the sulfide
along with the research on developing new catalysts, it is necessary form. In the case of the supported catalyst, before recycling, the
to do research on catalyst deactivation and the reusability of those solid mixture in the bottom of the reactor will be treated by
catalysts. In slurry phase hydrocracking, the reaction time is often burning several times to remove the coke from the catalyst and
very short and the catalyst deactivation is not really significant. then the solid mixture will be directly remixed with the feedstock
However, on an industrial scale, the long catalyst lifetime has [77]. However, with oil-soluble dispersed catalysts, the metal
numerous benefits not only in production cost, but also operation sulfides created after reactions are very fine crystals. These crystals
and maintenance expenses. are often stuck inside the coke; therefore, when recycling to the
Basically, coke can be classified into four types: sponge coke, feedstock, the amount of solid loading will be increased.
shot coke, needle coke, purge coke and catalyst coke [70–73]. Rezaei et al. [38,78,79] have done experiments with various
Sponge coke has a spongelike appearance with pores and bubbles concentrations of Mo octoate precursor and then coke was recycled
stuck inside the solid coke. That type of coke has the ability of back again and again to the feed to investigate the reusability of the

Fig. 13. Sponge coke (a) and shot coke (b) [74].
 M.T. Nguyen et al. / Journal of Industrial and Engineering Chemistry 43 (2016) 1–12 11

Significant advantages have been obtained by gaining an

understanding of the role oil-soluble dispersed catalyst in
slurry-phase hydrocracking. Improvement of catalyst performance
is on the top priority to reduce the production cost as well as
undesired product that can harm for environment. Currently,
catalyst design attracts many attentions, the inventions of
bimetallic and multimetallic liquid catalysts are both the impres-
sive steps ahead; however, it is still very potential in the future. In
the other hand, using additive in hydrocracking is not a new ideal,
but it is very efficiency. Hydrogen donor is a great example; it
enhances the hydrogenation and results in better performance of
catalyst during reaction. Coke formation is another critical issue,
until now, the knowledge that relates to it is still very limited.
Investigation of coke and pre-treatment method before recycling
to the reactor is believed to be a crucial part of slurry phase
hydrocracking prospective.

Acknowledgment
Fig. 14. The XRD diffractogram of coke samples recovered after different
hydroconversion experiments [78]. This work was supported by the National Research Council of
Science & Technology (NST) grant by the Korea government (MSIP)
(No. CRC-14-1-KRICT).
catalyst. They announced some important conclusions. Firstly,
metal precursors were transferred into sulfide form and 90% of
References
them were well dispersed and were recovered in the solid product
(coke). Even if the solid loading is proportional to the recycling [1] British Petroleum Company, BP Energy Outlook 2016, British Petroleum Co.,
time, when it is lower than 7.5%, there is no significant effect to London, 2016. p. 13, Retrieved from: https://www.bp.com/content/dam/bp/
catalyst activity. Rezaei et al. also showed that the recycling pdf/energy-economics/energy-outlook-2016/bp-energy-outlook-2016.pdf.
[2] REN21—Renewable Energy Policy Network for the 21st Century Renewables
number depends on the initial concentration of the catalyst and 2012—Global Status Report, (2012) . Retrieved from: http://www.ren21.net/
type of catalyst, but then when the initial concentration of a Portals/0/documents/Resources/GSR2012_low%20res_FINAL.pdf.
catalyst is higher than a certain number, the recycling time no [3] J.G. Speight, Catal. Today 98 (2004) 55.
[4] R. Paul Robinson, Petroleum and its products, in: A. James Kent (Ed.),
longer depends on that factor anymore. In addition, the coke yield Handbook of Industrial Chemistry and Biotechnology, 12th ed, Springer, 2012
increased rapidly when the concentration of Mo in the recycled p. 699.
coke was lower than 1.2%, indicating that the catalyst has been [5] Y. Li, J. Wang, L. Jiang, Z. Zhang, J. Liu, S. Ren, B. Zhao, Y. Jia, US 6004454, 1999.
[6] G. Que, C. Men, C. Meng, A. Ma, J. Zhou, W. Deng, Z. Wang, B. Mu, C. Liu, D. Liu, S.
deactivated quickly. Therefore, both the Mo concentration in the Liang, B. Shi, US 6660157, 2003.
recycled coke and the age of the coke-catalyst mixture are the [7] Hui Luo, Wenan Deng, Jingjie Gao, Weiyu Fan, Guohe Que, Energy Fuels 25
determining factors to the catalyst deactivation and reusability. (2011) 1161.
[8] E. Furimsky, Catalyst for Upgrading Heavy Petroleum Feed, Elsevier,
Fig. 14 shows an XRD diffractogram of coke samples in a series
Amsterdam, 2007 p. 404.
of recycling experiment. The peaks at 33 and 59 confirm the [9] Julius Scherzer, Adrian J. Gruia, Other applications of hydrocracking, in: Julius
exists of MoS2 crystal planes 11 0 and 1 0 0, respectively. By looking Scherzer, A.J. Gruia (Eds.), Hydrocracking Science and Technology, CRC Press,
at the peak intensity ratio between MoS2 crystal planes and 1996 p. 253.
[10] Yanet Villasana, Miguel A. Luis-luis, Franklin J. Mendez, Henry Labrador,
graphite, it is clear that those ratios are decreased following the Joaquin L. Brito, in: Umesh Chandra Sharma, Sri Sivakumar, Ram Prasad (Eds.),
recycle times. That trend is due to the increase of coke fraction in Upgrading and Hydrotreating of Heavy Oils and Residua, Energy Science and
the recycled coke. The broad of peak areas of MoS2 in all Technology Vol. 3: Oil and Natural Gas, Studium Press LLC, 2015 p. 319.
[11] Christine W. Curtis, Kan Joe Tsai, James A. Guin, Ind. Eng. Chem. Res. 26 (1987)
experiments indicates that there is no significant change or 12.
agglomeration of MoS2 during recycling. [12] Russell R. Chianelli, Mohammad H. Siadati, Myriam Perez De la Rosa, Gilles
Aldridge and Bearden [40] announced a process in which both Berhault, Jess P. Wilcoxon, Roby Bearden Jr., Billie L. Abrams, Catal. Rev.: Sci.
Eng. (2006) 1.
oil-soluble dispersed Mo naphthenate and a solid iron compound [13] Hui Du, Ming Li, Yuyang Ren, Yajing Duan, Appl. Petrochem. Res. 5 (2015) 89.
were used. After the reaction, a solid comprised of molybdenum [14] T. Cyr, L. Lewkowicz, B. Ozum, US 5578197, 1996.
and iron compound was recovered and used as the catalyst for the [15] M. Daage, R.R. Chianelli, J. Catal. 149 (1994) 414.
[16] K.S. Liang, R.R. Chianelli, F.Z. Chien, S.C. Moss, J. Non-Cryst. Solids 79 (1986)
next hydrocracking reaction. As a result, a catalyst comprised of 251.
molybdenum and the iron-containing coal ash showed good [17] Giuseppe Bellussi, Giacomo Rispoli, Daniele Molinari, Alberto Landoni, Paolo
activity retention in recycle runs not only for coke and light gas Pollesel, Nicoletta Panariti, Roberto Millini, Erica Montanari, Catal. Sci.
Technol. 3 (2013) 176.
suppression but also for metal removal and Conradson carbon
[18] Strasz, Fuel Energy Abstr. 37 (1996) 176.
conversion. [19] M. Breysse, E. Furimsky, S. Kasztelan, M. Lacroix, G. Perot, Catal. Rev.: Sci. Eng.
44 (2002) 651.
Conclusions and future prospective [20] S. Kasztelan, G.B. McGarvey, J. Catal. 147 (1994) 476.
[21] G.B. McGarvey, S. Kasztelan, J. Catal. 148 (1994) 149.
[22] M. Breysse, S. Kasztelan, Actual. Chim. 5 (2002) 612.
For the past few decades, renewable, nuclear and bio energy [23] M. Sun, J. Adjaye, A.E. Nelson, Appl. Catal. A 263 (2004) 131.
have been developing rapidly, however, these could not satisfy the [24] Jorge Ancheyta, Mohan S. Rana, Edward Furimsky, Catal. Today 109 (2005) 3.
[25] R. Prada Silvy, P. Grange, F. Delannay, B. Delmon, Appl. Catal. 46 (1989) 113.
ever-increasing energy demands, especially for transportation and [26] C. Yang, H. Zheng, F. Du, C. Xu, Reaction Characteristics and Mechanism of
production. That demand requires a huge supply of distillate fuels Residuum Hydrocracking, University of Petroleum, State Key Laboratory of
that are often produced from the refinery of conventional oil. Heavy Oil Processing, 2011. http://www.anl.gov/PCS/acsfuel/preprint%20ar-
chive/Files/43_3_BOSTON_08-98_0668.pdf.
However, the capacity of conventional oil has been reducing [27] G. James Speight, Hydrocracking, in: G. James Speight (Ed.), Heavy and Extra-
quickly since the 1950s. In contrast, unconventional oil is so heavy Oil Upgrading Technologies, Elsevier Inc., 2013 p. 96.
abundant and is gaining increasing attention. [28] S. Sánchez, M.A. Rodríguez, J. Ancheyta, Ind. Eng. Chem. Res. 44 (2005) 9409.
12 M.T. Nguyen et al. / Journal of Industrial and Engineering Chemistry 43 (2016) 1–12

[29] S. Sánchez, J. Ancheyta, Energy Fuels 21 (2007) 653. [56] Kaidong Chen, Bruce E. Reynolds, Darush Farshid, US 7678732, 2010.
[30] Z. Yingxian, L. Da, China Petrol. Process. Petrochem. Technol. 13 (2011) 24. [57] Guohe Que, Cungui Men, Chunxu Meng, An Ma, Jiashun Zhou, Wenan Deng,
[31] D.W. Marquard, J. Soc. Ind. Appl. Math. 2 (1963) 431. Zongxian Wang, Baoquan Mu, Chenguang Liu, Dong Liu, Shichang Liang, Bin
[32] Cristian J. Calderon, Jorge Ancheyta, Energy Fuels 30 (2016) 2525. Shi, US 6660157 B2, 2003.
[33] C.H. Batholomew, Catalyst deactivation in hydrotreating of residua: a review, [58] Minghui Zhang, Liangliang Zhang, Zhenyu Xiao, Qinhui Zhang, Rongming
in: C. Michael Oballa, S. Stuart Shih (Eds.), Catalytic Hydroprocessing of Wang, Fangna Dai, Daofeng Sun, Sci. Rep. 6 (2016) 20672.
Petroleum and Distillates, CRC Press, 1994 p. 1. [59] Ian P. Fisher, Oakville, Nicolas G. Samman, Mississauga, US 4485004, 1984.
[34] J. Ancheyta, M.S. Rana, E. Furimski, Catal. Today 109 (2005) 3. [60] R. Ranganathan, J.-M.D. Denis, B.B. Pruden, US 4214977, 1980.
[35] Alberto Del Bianco, Nicoletta Panariti, Sabatino Di Carlo, Jeanne Elmouchnino, [61] R. Bearden, C.L. Aldridge, US 4244839, 1981.
Bernard Fixari, Pierre Le Perchec, Appl. Catal. A: Gen. 94 (1993) 1. [62] R. Bearden, C.L. Aldridge, US 4295995, 1981.
[36] Shuyi Zhang, Dong Liu, Wenan Deng, Guohe Que, Energy Fuels 21 (2007) 3057. [63] C. Petit, F. Maug’e, J.-C. Lavalley, Stud. Surf. Sci. Catal. 106 (1997) 156.
[37] Zhihua Wu, Zhenhua Zhou, Bing Zhou, US 8142645 B2, 2012. [64] N.-Y. Topsøe, H. Topsøe, F.E. Massoth, J. Catal. 119 (1989) 52.
[38] Hooman Rezaei, Shahrzad Jooya Ardakani, Kevin J. Smith, Energy Fuels 26 [65] M.S. Rana, B.N. Srinivas, S.K. Maity, G. Murali Dhar, T.S.R. Prasada Rao, J. Catal.
(2012) 6540. 195 (2000) 31.
[39] R.J. Bearden, C.L. Aldridge, US 4226742, 1980. [66] P. Zeuthen, P. Blom, B. Muegge, F.E. Massoth, Appl. Catal. A 68 (1991) 117.
[40] C.L. Aldridge, R.J. Bearden, US 4066530, 1978. [67] P. Dufresne, N. Brahma, F. Labruyere, M. Lacroix, M. Breysse, Catal. Today 29
[41] R.J. Bearden, C.L. Aldridge, US 4295995, 1981. (1996) 251.
[42] R.J. Bearden, C.L. Aldridge, US 4579838, 1986. [68] E. Furimsky, Appl. Catal. A 171 (1998) 177.
[43] T. Cyr, L. Lewkowicz, B. Ozum, US 5578197, 1996. [69] R. Prada Silvy, P. Grange, F. Delannay, B. Delmon, Appl. Catal. 46 (1989) 113.
[44] O.P. Strausz, US 6068758, 2000. [70] M. Siskin, S.R. Keleme, M.L. Gorbaty, D.T. Ferrughelli, L.D. Brown, C.P. Eppig, R.J.
[45] J.G. Gatsis, US 4194967, 1980. Kenned, Energy Fuels 20 (2006) 2117.
[46] Bing Zhou, Zhenhua Zhou, Zhihua Wu, US 7842635 B2, 2010. [71] S.R. Kelemen, M. Siskin, M.L. Gorbat, D.T. Ferrughelli, P.J. Kwiatek, L.D. Brown,
[47] Hooman Rezaei, Shahrzad Jooya Ardakani, Kevin J. Smith, Energy Fuels 26 C.P. Eppig, R.J. Kenned, Energy Fuels 21 (2007) 927.
(2012) 2768. [72] S. Birghila, I. Carazeanu Popovici, A. Dumitru, Rom. J. Phys. 56 (2011) 976.
[48] I. Watanabea, M. Otakeb, M. Yoshimotob, K. Sakanishic, Y. Koraia, I. Mochida, [73] Anthony Andrews, Richard K. Lattanzio, Petroleum Coke: Industry and
Fuel 81 (2002) 1515. Environmental Issues, Congressional Research Service, (2013) . Retrieved
[49] Zhang Shuyi, Deng Wenan, Luo Hui, Liu Dong, Que Guohe, Energy Fuels 22 from: http://www.nam.org/CRSreport/.
(2008) 3583. [74] D. John Elliott, Managing Shot Coke: Design & Operation, January, (2009) . p. 1,
[50] Ruby Bearden Jr., US 4134825, 1979. XP55035572, Retrieved from: http://www.fwc.com/publications/tech_papers/
[51] N. Panariti, A. Del Biancoa, G. Del Pieroa, M. Marchionnab, Appl. Catal. A: Gen. files/ManagingShotCoke-Design&OperationA-4Rev1.pdf.
204 (2000) 203. [75] J.W. Gosselink, W.H.J. Stork, Ind. Eng. Chem. Res. 36 (1997) 3354.
[52] Dietmar Kennepohl, Emerson Sanford, Energy Fuels 10 (1996) 229. [76] H.S. Wengt, G. Eigenberger, J.B. Butt, Chem. Eng. Sci. 30 (1975) 1341.
[53] Z. Wang, H. Zhang, A. Guo, G. Que, Prepr Pap.—Am. Chem. Soc. Div. Fuel Chem [77] Roby Bearden Jr., Zhiguo Hou, Martin Leo Gorbaty, David Thomas Ferrughelli,
43 (1998) 486. https://web.anl.gov/PCS/acsfuel/preprint%20archive/Files/ Ronald Damian Myers, US 6511937B1, 2003.
43_3_BOSTON_08-98_0486. pdf. [78] Hooman Rezaei, Shahrzad Jooya Ardakani, Kevin J. Smith, Energy Fuels 26
[54] Bin Shi, Guohe Que, Prepr. Pap.—Am. Chem. Soc. Div. Fuel Chem. 48 (2003) 722. (2012) 6540.
https://web.anl.gov/PCS/acsfuel/preprint%20archive/Files/48_2_New% [79] Hooman Rezaei, Xuebin Liu, Shahrzad Jooya Ardakani, Kevin J. Smith, Maureen
20York_10-03_0684.pdf. Bricker, Catal. Today 150 (2010) 44.
[55] Sang Goo Jeon, Jeong-Geol Na, Chang Hyun Ko, Kwang Bok Yi, Nam Sun Rho,
Seung Bin Park, Energy Fuels 25 (2011) 4256.