Fuel 237 (2019) 1–9

Contents lists available at ScienceDirect

Fuel
journal homepage: www.elsevier.com/locate/fuel

Full Length Article

Fundamental study of hierarchical millisecond gas-phase catalytic cracking T

process for enhancing the production of light olefins from vacuum residue
⁎ ⁎
Yuanjun Chea, Meng Yuana, Yingyun Qiaoa, Qin Liub, Jinhong Zhanga, , Yuanyu Tiana,b,
a
State Key Laboratory of Heavy Oil Processing, China University of Petroleum, Qingdao 266580, China
b
Key Laboratory of Low Carbon Energy and Chemical Engineering, Shandong University of Science and Technology, Qingdao, Shandong 266590, China

A R T I C LE I N FO A B S T R A C T

Keywords: In order to enhance the production of light olefins from vacuum residue (VR), a novel hierarchical millisecond
Vacuum residue gas-phase catalytic cracking (HM-GCC) process was proposed. The process integrated the pyrolysis of VR and the
Hierarchical catalytic cracking of the pyrolysis oil in the gas phase. Firstly, the effects of reaction temperature on the mil-
Gas-phase catalytic cracking lisecond pyrolysis behavior of VR were investigated, and the results showed that the increase of C2-C4 hydro-
Light olefins
carbons content was not obvious when increasing the temperature from 600 to 750 °C. However, the thermal
cracking as a pretreatment process converted the VR macromolecule into the low molecular pyrolysis oil. After
that, five different catalysts calcium aluminate, FCC catalyst, and ZSM-5 zeolites (Z-40, Z-80, and Z-200) were
investigated to evaluate their activities in catalytic cracking of pyrolysis oil, and among them Z-40 showed the
highest selectivity for C2-C4 hydrocarbons and BTX. For further study, the relative contents of C2-C4 hydro-
carbons were higher for the combination of calcium aluminate and Z-40 (Z + C-cat) than that of the pure Z-40
catalyst. This result was attributed to the fact that Z + C-cat facilitated the cracking of macromolecules and
inhibited hydrogen transfer reactions. Finally, the conversion of VR and pyrolysis oil was studied in the bench
scales with calcium aluminate and Z-40 catalyst. The yield of ethylene, propylene and butenes increased by 4.69,
7.87 and 3.37 wt% respectively in the presence of Z-40.

1. Introduction
With the development of stored energy battery and the strict re-
quirements of the environmental regulations, the electric vehicle has
emerged with the advantages of near-zero pollution and energy saving.
Replacement of fuel vehicles with electric vehicles has become an ir-
reversible trend, which will follow by a sharp drop in transportation
fuel consumption. Thus, petroleum refineries worldwide should be
transformed from the production of transportation fuel (gasoline and
diesel) to light olefins (ethylene, propylene and butenes) and BTX
(benzene, toluene and xylene).
Light olefins and BTX are very important basic chemical raw ma-
terials. Particularly, ethylene and propylene are the most important
light olefins because they are used to produce many petrochemicals
products, such as polypropylene, polyethylene, propylene oxide, ethy-
lene oxide, acrylonitrile, and acrylic acid. At present, about 95% of
ethylene and 60% of propylene in the world are produced using light
raw materials (natural gas, naphtha or light diesel oil) by steam thermal
cracking process [1]. However, due to the depletion of traditional light
petroleum resources [2,3], the very limited flexibility of steam cracking
(in particular, the low propylene to ethylene ratio and the difficulties to
use heavy oil [4]) indicates that it is unappeasable for the market de-
mand. Therefore, to directly produce light olefins through the catalytic
cracking process using heavy oil appears to be a very attractive strategy
to solve this problem [5].
To increase the yield of light olefins, refiners and researchers have
paid more attention to redesigning FCC unit processes in the last few
decades; such as the Indmax process developed by Indian Oil
Corporation [6]; the PetroFCC process licensed by UOP [7]; the deep
catalytic cracking process (DCC) and catalytic pyrolysis process (CPP)
licensed by Sinopec, Research Institute of Petroleum Processing [8,9];
the high severity fluid catalytic cracking process (HS-FCC) developed
by Nippon Oil Corporation [10,11]; and the two-stage riser catalytic
cracking of heavy oil for maximizing propylene yield (TMP) process
proposed by China University of Petroleum [12]. Compared with steam
cracking, these processes have the advantages of wider range of raw
materials, low reaction temperature, and adjustable product distribu-
tion. However, these processes still have some trouble such as the
poison of catalysts, poor atomization, and low conversion etc. when
processing the inferior feedstocks.

⁎
Corresponding authors at: State Key Laboratory of Heavy Oil Processing, China University of Petroleum, Qingdao 266580, China.
E-mail addresses: zhangjh@upc.edu.cn (J. Zhang), tianyy1008@126.com (Y. Tian).

https://doi.org/10.1016/j.fuel.2018.09.088
Received 19 July 2018; Received in revised form 6 September 2018; Accepted 18 September 2018
0016-2361/ © 2018 Published by Elsevier Ltd.
Y. Che et al. Fuel 237 (2019) 1–9

ZSM-5 zeolite has been widely used in the catalytic cracking process Thermal Cracking Gas Phase Catalytic Cracking
for enhancing the yield of light olefins due to its shape-selectivity
[13,14]. However, ZSM-5 zeolite mainly contains micropores, which
are not large enough to contain larger molecules to allow the fast dif-
fusion of substances. In addition, the heavy oil macromolecules accu-
Vacuum Light
mulated on the surface of the zeolite are easy to form coke, thereby Gasoil
Residue Olefins
blocking the catalyst pores.
Furthermore, because the distillation range of VR is mainly above
500 °C, hydrocarbons mainly undergo liquid phase reaction in the ex-
isting process. The free radicals which split from hydrocarbon can
disperse quickly in the gas phase whereas they are surrounded by Fig. 2. Schematic representation of the hierarchical conversion of VR.
surrounding molecules like a “cage” in the liquid phase. Thus, to dis-
perse the formed free radicals, it is necessary to overcome the extra
bench-scale fluidized bed-fixed bed dual-reactor. The effects of tem-
energy barriers spread out from the “cage” in the liquid phase. This is
perature on the thermal behavior of VR millisecond pyrolysis were in-
the so-called “radical cage effect” [15,16]. Therefore, how to improve
vestigated firstly. Then five catalysts including calcium aluminate cat-
the gas-phase reaction ratio and suppress the liquid-phase reaction will
alyst, FCC catalyst, and ZSM-5 zeolites with the silica-alumina ratios of
become the key problem for the next generation of heavy oil processing
40, 80 and 200 were used to study the product distribution (especially
technology.
the yield of C2-C4 hydrocarbons and BTX) of the gas-phase catalytic
To solve the problems mentioned above, the author proposes the
cracking of pyrolysis oil. The combination of ZSM-5 and calcium alu-
hierarchical millisecond gas-phase catalytic cracking (HM-GCC) pro-
minate catalysts was tested for optimizing catalyst in terms of high C2-
cess. As shown in Fig. 1, the heavy oil is sprayed into the downer-type
C4 hydrocarbons selectivity. Finally, the analysis of the yields and
reactor through a high-efficiency nozzle and quickly contacts with high
compositions of the gas and liquid products obtained by a larger scale
temperature regenerated alkaline catalyst. After millisecond pyrolysis
experiment provided an insight into the potential application of VR to
reaction, the produced oil-gas and coked catalyst are separated by a
chemicals.
horizontal cyclone separator at the bottom of the dower reactor. The
high temperature gas oil contacts and mixes with the regenerated cat-
alyst in another downer-type reactor directly, and takes place milli-
2. Experimental section
second catalytic cracking without condensate separation. The coked
catalyst enters a riser-type gasifier for combustion and regeneration;
2.1. Feedstock and catalyst
and then returns to the downer reactor for recycling. The products are
piped to a fractionation tower and cut into dry gas, light olefins, BTX,
The heavy feedstock was the vacuum residue (VR) provided by
and fuel oil. The major advantages of this process are as follows: fully
Shida Changsheng Energy Technology Co., Ltd. (Shandong Province,
use the heat of pyrolysis gas-oil; overcome the “cage effect” of the liquid
China) and its properties were shown in Table 1. The density, viscosity,
phase reaction, and maximize the light olefins yield.
and Conradson carbon residue (CCR) of the VR were 1020 kg/m3,
Via this HM-GCC process, the VR is mainly converted into light
3985 mm2/s, and 22.38 wt%, respectively.
olefins and BTX. The process is essentially a hierarchical conversion of
The FCC catalysts, ZSM-5 zeolites with different silica-alumina ra-
the inferior heavy oil. Fig. 2 clarifies that the VR macromolecule can be
tios of 40, 80, and 200 (Z-40, Z-80, and Z-200), and calcium aluminate
pyrolyzed to produce gas-oil which has lower molecules; the gasoil is
catalysts were used in the experiments and their properties were shown
further catalytically cracked into light olefins. Besides, to improve
in Section 3.2.1. The Z-40 and calcium aluminate were physically
catalytic performance in the gas-phase cracking gasoil, catalyst with
mixed to prepare the Z + C-cat catalyst. The catalysts were formed by
micro-mesoporous hierarchical structure is needed [17–19]. Thus, the
tablet machine after fully grinding and blending. And the particle sizes
catalyst should have the hierarchical porosity of the pore and the ability
of the catalysts distributed between 0.075 and 0.090 mm.
to easier transport and access to the active sites [3].
The purpose of this work is to investigate the fundamental char-
acteristics of the HM-GCC process using an analytical Py-GC/MS and a

Regenerated
Heat-carriers
Heavy oil Heavy oil
Regenerated
Gasoil catalyst

Downer Dry Gas

Pyrolysis Downer
Reactor Riser-type
Catalysis Light Olefins
Catalyst
Cracking LPG
Regenerator
Reactor
Fractionator
Coked
Catalyst BTX
Aromatics
Separator Separator
Steam and Air Fuel Oil

Fig. 1. Schematic diagram of the HM-GCC process.

2
Y. Che et al. Fuel 237 (2019) 1–9

Table 1
Properties of Shida Changsheng vacuum residue.
Density (20 °C), kg/m3 Viscosity (100 °C), mm2·s−1 CCRa, wt% C, wt% H, wt% S, wt% N, wt% O, wt% H/C

1020 3985 22.38 85.75 9.81 2.57 0.64 1.23 1.36

a
CCR is the Conradson carbon residue.

2.2. Experimental apparatus and product analysis
2.2.1. Py-GC/MS
The consists of the Py-GC/MS experiments were a Pyroprobe 5200
Series pyrolyser (CDS Analytical Inc.) and a gas chromatography/mass
spectrometry (GC/MS, Techcomp). For a typical experiment,
0.40 ± 0.01 mg of oil sample was put into a quartz tube. The pyrolysis
temperature was set to 600–750 °C at a heating rate of 10 °C/ms and
held for 30 s. For simulation of the hierarchical gas-phase catalytic
cracking experiments, the oil sample and catalyst were conducted in a
layer method [20]; 0.4 mg of oil sample and 0.6 mg of catalyst were
separated by a quartz wool layer. The oil sample was first pyrolyzed at
600 °C and then the pyrolysis gas-oil passed through the catalyst layers,
occurring the catalytic cracking at 600 °C. The GC separation of product
vapors was done by a DB-5MS column (60 m × 0.25 mm, 0.25 μm film
thickness) with Helium (99.999%) carrier gas (0.8 mL/min). The GC
inlet was 300 °C and a split ratio of 50:1 was used. The oven was pro-
grammed to start at 35 °C (hold for 15 min) and then ramp to 200 °C
with 3 °C/min, and then raised to 300 °C at the heating rate of 5 °C/min
and hold for 10 min. The mass spectrometer operates at 70 eV EI mode
with a mass spectrum from m/z 35 to 500. The chromatographic peaks
were determined based on the NIST mass spectral library. Because the
chromatographic peak area of a compound is linear with its quantity,
the peak area % value of each product can be compared to reveal the
change in its relative content [21].
2.2.2. Fluidized bed-fixed bed reactor experimental apparatus
The schematic diagram of fluidized bed-fixed bed experimental
apparatus was shown in Fig. 3. VR and steam were preheated to 300 °C
and mixed in a furnace and then continuously fed into the reactor by
pumps. The oil-steam mixture was injected into the fluidized bed re-
actor and suffered pyrolysis reaction. The formed oil gas separated by a
steel filter was transferred to the fixed bed reactor and occurred gas-
phase catalytic cracking reaction. A two-step condensation process was
used to condense the cracking gas. The liquid products and non-
condensable gas were measured by electronic balance and gas flow
meter, respectively.
2.2.3. Analysis and characterization
The compositions of cracking gas and liquid products were analyzed
by a Scion gas chromatograph (GC) and a Varian GC respectively, ac-
cording to the standard test method. A coke analyzer was used to de-
termine the coke yield deposited on the catalyst. Nitrogen adsorption/
desorption isotherms were measured using a static volumetric appa-
ratus (Micromeritics) at 77 K to determine the specific surface area and
pore volume. The concentration of Lewis and Brønsted acid sites was
determined after adsorption of pyridine (Py) by FTIR spectroscopy
(Thermofisher). Temperature programmed desorption of ammonia
(NH3-TPD) technique was used to measure the acidic properties. The
crystal size and shape of ZSM-5 were examined by scanning electron
microscopy (SEM, JEOL JSM-7800F microscope).
3. Results and discussion
3.1. Millisecond pyrolysis of vacuum residue
Vacuum residue (VR) has higher contents of polynuclear aromatics,
heteroatoms, contaminant mental, and higher CCR than light oil.
Considering the inferior property of VR, pyrolysis is always used as a
non-catalytic pre-treatment process to produce light gas-oils and reduce
the content of trace metal. In this work, Py-GC/MS was used to study
the effect of temperature on VR millisecond pyrolysis process.
Fig. 4 shows the total ion chromatograms from millisecond pyrolysis
of VR at 600 °C. The representative pyrolysis products identified by GC/
MS are light olefins (such as ethene, propene, and butenes), C5+ al-
kenes (1-alkenes, iso-alkenes, and cycloalkenes), alkanes (n-alkanes,
iso-alkanes, and cycloalkanes), alkadienes, aromatics, and other spe-
cies. The relative contents of identified product were listed in Table 2.
The relative contents of C2, C3, and C4 hydrocarbons in the tem-
perature of 600–700 °C showed a monotonic increase (Table 2), which

Oil feeding
system
Heating belt
Furnace
Oil Pump Condenser

Fixed bed Cooling

reactor water
Water Pump Steam Condenser
generator
Ethanol Gas flow
meter
Gas flow
O2 N2 meter
Second oil collector
First oil collector
Fluidized bed
Gas supplying reactor
Product collecting
system Reaction system
system
Fig. 3. Schematic diagram of the fluidized bed—fixed bed reactor experimental apparatus.

3
Y. Che et al. Fuel 237 (2019) 1–9

8 o
6x10 600 C
8
5x10
8
4x10 1-alkenes n-alkanes

Abundance
8
3x10
C11 C13 C15 C19 C22
C17
8 C25
2x10 C9

1x10
8 C27

0 10 20 30 40 50 60 70 80 90 100
R.T. (min)
Fig. 4. Total ion chromatogram of the pyrolysis products of VR pyrolyzed at 600 °C.

Table 2 rate of addition reaction in the liquid phase was much faster than that
Percentage Hydrocarbon Distribution for Pyrolysis Products. in the gas phase, which would lead to the reduction of olefins. On the
Products 600 °C 650 °C 700 °C 750 °C
other hand, the higher temperature was beneficial to the dehy-
drogenation reaction, resulting in the decrease of alkanes. The data in
C2 (%) 1.16 1.53 1.89 1.98 Table 2 showed that the relative content of alkadienes and aromatics
C3 (%) 3.87 4.53 5.56 5.93 increased from 2.65% to 7.45% and 4.67% to 11.18%, respectively. The
C4 (%) 3.80 4.52 5.99 6.07
C5+ alkenes (%) 45.72 50.84 51.96 52.29
increase of alkadienes with the temperature is mainly due to the de-
1-alkenes 33.17 36.05 36.41 37.49 hydrogenation of olefins and the further cracking of CeC bonds in lager
iso-alkenes 9.58 10.23 9.68 6.16 olefins [27]. The enhancement of aromatics is attributed to the increase
Cycloalkenes 2.96 4.56 5.87 8.64 of aromatization reactions of alkenes and alkadienes and the dehy-
alkanes (%) 35.06 25.72 19.50 13.59
drogenation of cycloalkanes with the increase of temperature.
n-alkanes 28.41 19.97 14.61 10.45
iso-alkanes 5.02 4.28 4.05 2.45
cycloalkanes 1.62 1.47 0.84 0.70
3.2. Gas-phase catalytic cracking of the pyrolysis gas-oil
alkadienes (%) 2.65 3.54 4.79 7.45
aromatics (%) 4.67 5.50 7.51 11.18
others (%) 3.06 3.82 2.80 1.50 3.2.1. Catalyst characterization
The surface area and pore structure of catalysts, presented in
Table 3, were evaluated by N2 adsorption isotherms. The FCC catalyst
could be explained that higher temperature was beneficial to the rup- mainly possesses mesopores with higher average diameter whereas the
ture of CeH and CeC bonds; thus the formation of small molecule pore volume of ZSM-5 zeolite was mainly micropore. ZSM-5 zeolites
products was increased. In addition, Vafi and Bu deemed that the with different Si/Al2 ratios have the same value of micropore volume
higher the temperature the greater the fraction of the feed would react while Z-200 has the highest mesopore volume (0.05 cm3/g). The sur-
in the gas phase [22,23]. Under gas phase conditions, a larger amount face areas of Z-40, Z-80 and Z-200 were 342, 339 and 398 m2/g, re-
of light olefins will be provided because of the fact that the β-scission spectively. Calcium aluminate has the lowest surface area of only
reaction is more favorable than either addition or hydrogen transfer 10.6 m2/g compared with the FCC and ZSM-5 catalysts.
reaction [24]. However, as the temperature further increased to 750 °C, Fig. 5(a) shows the NH3-TPD profiles of the ZSM-5 zeolites. Typi-
the increase becomes slower. This may be explained by the elimination cally, characteristic of the NH3-TPD curves of the zeolites exhibits two
of free radicals. The relative content of alkenes increased with the in- desorption peaks. The low temperature region of 180–400 °C and the
crease of temperature, and olefins are effective inhibitors of free radical high temperature region of 400–650 °C, as shown in Fig. 5(a), corre-
decomposition; thus the increase of temperature will result in a de- spond to the weak and strong acid sites, respectively.[28] It can be
crease in the decomposition rate. observed that the strong acidity of the zeolite decreased with increasing
It can be found that 1-alkenes is the predominant type in the mil- Si/Al molar ratio and the weak acidity of Z-40 was the highest com-
lisecond pyrolysis products. Previous researchers [25] found that the pared with Z-80 and Z-200. The pyridine-FTIR spectra were obtained to
formation of 1-alkenes was attributed to the rupture of alkyl free ra- determine the Brønsted and Lewis acidities of the catalysts, as shown in
dicals. The formation of iso-alkenes was mainly due to the cracking of Fig. 5(b). The band at 1545 cm−1 corresponds to Brønsted acid sites
isomeric alkyl substituents attaching to aromatic rings and naphthenic
rings. The Diels-Alder reaction was the main way for the formation of Table 3
cycloalkenes [26]. Meanwhile, the aliphatic free radicals and naph- Surface area and pore volume of catalysts.
thenic free radicals produced by the cracking reaction could capture H
Items FCC catalyst Z-40 Z-80 Z-200 Calcium
radicals and then produce alkanes, which is the cause of formation of n- aluminate
alkanes and iso-alkanes and cycloalkanes. It can be seen from Table 2
that the relative content of alkanes significantly decreases from 35.06% BET surface area 78.69 341.83 339.86 398.32 10.6
(m2/g)
to 13.59% with the increase of reaction temperature whereas the C5+
Total pore volume 0.14 0.17 0.18 0.19 0.06
alkenes increases from 45.72% to 52.29%. The ratios alkenes/alkanes (cm3/g)
monotonically increased from 1.30 to 3.85 with the increase of tem- Micro pore volume 0.02 0.14 0.14 0.14 0
peratures, which indicated that higher temperatures favored production (cm3/g)
of alkenes over alkanes. As mentioned above, the higher temperature Meso pore volume 0.12 0.03 0.04 0.05 0.06
(cm3/g)
was available to enhance the gas-phase cracking. On the one hand, the

4
Y. Che et al. Fuel 237 (2019) 1–9

-75 (a) Z-200 (b)

Z-40 Z-80
Z-80 Z-40
-60 Z-200

Absorbance ( a.u.)
Intensity (a.u.)
-45

-30

-15

100 200 300 400 500 600 700 1600 1550 1500 1450 1400
o -1
Desorption temperature ( C ) Wavemumber (cm )

Fig. 5. Characterizations of the ZSM-5 zeolites with (a) NH3-TPD and (b) py-FTIR.

Fig. 6. SEM images of ZSM-5: (a) Z-40; (b) Z-80; (c) Z-200.

(BASs) and the band at 1445 cm−1 is related to the Lewis acid sites
(LASs) [1,29]. The concentration of BASs and LASs decreased in the
following order: Z-40 > Z-80 > Z-200. The SEM images were showed
in Fig. 6. It could be seen from Fig. 6(a-1) and (a-2) that the crystals of
ZSM-5 zeolite have typical hexagonal shape forming larger agglomer-
ates. Comparing the Fig. 6(a-1), (b) and (c), it could be found that the
crystal size of ZSM-5 with different silica-alumina ratios was in the
order Z-40 > Z-200 > Z-80.
3.2.2. Effect of catalysts on enhancing the production of light olefins and
BTX.
In Section 3.1, we found that the relative content of light olefins did
not enhance tremendously even increased the temperature to 750 °C. As
mentioned above, ZSM-5 zeolites are commonly used to produce more
light olefins because of the shape selectivity. To understand the effect of
different catalysts on the gas-phase catalytic cracking of pyrolysis oil,
the FCC catalyst and solid base catalyst (calcium aluminate) were tested
as the reference experiments.
Enhancing the formation of light olefins is the main objectives of
gas-phase catalytic cracking of pyrolysis gas-oil. The relative content of
C2, C3 and C4 hydrocarbons obtained over different catalysts are pre-
sented in Fig. 7. The experimental data showed that the relative content
of C2-C4 hydrocarbons increased from 9.11% without catalyst to about
11.99% over calcium aluminate and improvements were also reported
at lower increasing extent for FCC catalyst (11.26%). However, the
relative contents of C2-C4 hydrocarbons were 38.78%, 34.78%, and
31.66% for Z-40, Z-80, and Z-200, respectively. This was due to the
high shape-selectively and high surface area of ZSM-5 zeolites. The
proposed mechanism suggests that the cracking reaction of hydro-
carbons on calcium aluminate catalysts is carried out by free radical
mechanisms [30]. The formation of small molecule hydrocarbons needs
higher temperature to break the bond energy. Since the reaction tem-
perature of the gas-phase cracking is the same as the pyrolysis tem-
perature, the increase in light olefins is not much. It is well known that

5
Y. Che et al. Fuel 237 (2019) 1–9

C2 Calcium Aluminate
20
C3 10 FCC
Z-40
16 C4
Z-80
Relative Content (%)

Relative Content (%)

Z-200
12
6

8
4

4
2

0
Without Calcium
0
FCC Z-40 Z-80 Z-200
6 7 8 9 10 11 12 13 14 15 16 17
Catalyst Aluminate
Carbon number
Fig. 7. Effect of catalysts on the production of C2, C3 and C4 hydrocarbons.
Fig. 8. Effect of catalysts on the production of aromatics.

the distinct pore structure and the accessibility of the acidic sites of
important role in dehydrogenation reactions to form BTX. The LASs
zeolites influence the catalytic activity. It has been reported that the
promoted the dehydrogenation, resulting in the high yield of BTX.
small pore size of the zeolite could inhibit the hydrogen transfer reac-
However, the Z-80 and Z-200 zeolites have the lower amount of Lewis
tions due to the fact that smaller pore entrances do not accommodate
acid sites in (as shown in Fig. 5(b)) to complete the hydrogen transfer
bulky bimolecular reaction [31,32]. ZSM-5 zeolite exhibited a smaller
reaction and thus lead to a lower production of BTX compared with Z-
pore structures than FCC catalyst, as shown in Table 3. Hence, ZSM-5
40. Furthermore, Yang and co-workers [37] deemed that the formed
zeolites produced more C3 and C4 hydrocarbons whereas FCC catalyst
alkenes entered the weak acid site and underwent the Diels-Alder cy-
favored aromatization reactions resulting in a low enhancement in C2-
clization reactions and then formed the cycloalkanes; the cycloalkanes
C4 hydrocarbons yield.
underwent dehydrogenation-aromatization reaction on the strong acid
The highest C3 hydrocarbons relative content of 16.92% was ob-
sites and produced the aromatics. On the other hand, the crystal size of
tained for Z-40, compared with 11.88% and 12.82% obtained for Z-80
Z-40 was larger than that of Z-80 and Z-200, as showed in Fig. 6. The
and Z-200, respectively. However, the highest C4 hydrocarbons relative
larger crystallites could provide longer pores and thus increased the
content was achieved for Z-80 (20.95%). The higher relative content of
bimolecular reaction, resulting in a higher BTX relative content [38].
C2-C4 hydrocarbons over Z-40 compared with Z-80 and Z-200 might be
The relative content of C9+ aromatics for FCC catalyst is higher than
ascribed to the fact that the higher amount of BASs increased the
that for Z-40, Z-80 and Z-200 catalyst. This is due to the fact that the
cracking reaction. Additionally, this result was because the larger
pores of the FCC catalyst are larger than that of ZSM-5; thus the larger
zeolite crystallite (Z-40) was more favored to the secondary reaction,
molecule components are easier enter into the pores of FCC catalyst and
such as protolytic cracking of carbenium ions and ring-opening reaction
suffer aromatization reaction.
[33]. The C2 hydrocarbons formation could be explained by the
As shown in Fig. 9, 1-Alkenes, accounting for 33.17% of the pyr-
cracking of higher olefin precursors by the classical β-scission, which is
olysis oil (without catalyst), decreased to 8.46%, 4.82% and 5.05% for
closely associated with reaction temperature. Thus, the increase of C2
FCC catalysts, Z-80 zeolites, and Z-200 zeolites, respectively. Especially,
hydrocarbons relative content was inapparent. The light olefins dis-
for Z-40 zeolites, the 1-alkenes were completely cracked. This may be
tribution was much favored to propylene, in HM-GCC process, resulting
due to the high reactivity of 1-alkenes. The formation of carbenium ions
in a high propylene to ethylene ratio (P/E was 8.6 over Z-40). The P/E
from olefins can easily proceed through the cracking, isomerization,
ratio of thermal cracking in this study was 3.3, which was much lower
oligomerization, dehydrogenation, and H-transfer reaction to form new
than that of gas-phase catalytic cracking. This indicated that the HM-
GCC process was a feasible option for requiring the market demand.
Aromatics as the widely used commodities, especially monocyclic 90
aromatics such as benzene, toluene, and xylene (BTX) form an im- cycloalkenes cycloalkanes
portant segment of petrochemical industry. It is meaningful to produce 80 iso-alkenes iso-alkanes
more BTX while maximizing the production of light olefins by gas- 70 1-alkenes n-alkanes
phase catalytic cracking of pyrolysis oil. The distribution of aromatics
Relative Content (%)

for the five catalysts is illustrated in Fig. 8. BTX relative content for 60
catalyst Z-40 was obvious higher than other catalysts. The alkanes 50
formed through thermal cracking could be directly dehydrogenated on
the LASs to generate the corresponding alkenes. And the alkenes 40
(generated from alkanes and thermal cracking) moved to the strong 30
acid sites and generated alkylcarbenium ion, and then rearranged and
cracked via β-scission. Research found that the carbenium ions having 20
relative longer lifetime at strong acid catalysts [34,35]. Thus, the in- 10
trinsic strong acid of Z-40 zeolite (as shown in Fig. 5(a)) can facilitate
the cyclization reactions and produce more cycloolefins. Then, the cy- 0
Without Calcium FCC Z-40 Z-80 Z-200
cloolefin intermediates can suffer hydrogen transfer which occurs on
Catalyst Aluminate
the larger amount of LASs of the Z-40 and thus form more BTX products
[29]. Mochizuki and co-workers [36] found that LASs played an Fig. 9. Effect of catalysts on the production of alkenes and alkanes.

6
Y. Che et al.
compounds [39,40]. In contrast to the decrease of the 1-alkenes, the
relative contents of iso-alkenes increased obviously for FCC, Z-80, and
Z-200 catalysts. The reason was probably that the isomerization reac-
tion of 1-alkenes occurred and produced a large amount of iso-alkenes.
Besides, the iso-alkenes contained bulky isomers that were too large to
enter the pores of the zeolites to be converted. The relative contents of
n-alkanes reduced from 28.41% in pyrolysis oil to 15.03%, 9.05%,
12.87% and 13.72% for FCC, Z-40, Z-80 and Z-200 catalysts, respec-
tively. The relative contents of iso-alkanes also decreased to some ex-
tent for all catalysts. Compared to olefins, n-alkanes and iso-alkanes had
a lower reactivity towards cracking attributed to a more difficult for-
mation of carbenium ions. However, due to the presence of the LASs,
the alkanes could form a carbenium ion and caused the bimolecular
reaction. The largest reduction in n-alkanes was Z-40 zeolites because it
contained the most BASs and LASs.
3.2.3. The combination of ZSM-5 and calcium aluminate
As mentioned above, the small pore structure and strong acid
strength of ZSM-5 limited the use of them for the pyrolysis of heavy oil.
Recently, considerable attention has been focused on the preparation of
meso-microporous composite materials, which can improve the cata-
lytic activity and product selectivity in the cracking reaction. In this
study, the enhancement of light olefins yield from pyrolysis oil over the
Z-40 and calcium aluminate combination catalyst (Z + C-cat) was in-
vestigated.
Table 4 shows the relative contents of product distributions pro-
duced in the absence of the catalyst and presence of calcium aluminate
and FCC catalyst. The relative content for gasoline (C5-C12) and diesel
(C13-C22) fractions increased whereas the C23+ fraction was completely
converted over calcium aluminate catalyst. For FCC catalyst, the ga-
soline fraction increased obviously while the diesel and C23+ fractions
decreased. Besides, it can be seen from Figs. 8 and 9 that fewer n-al-
kanes and more 1-alkenes were formed over the calcium aluminate
compared with no catalyst, and the production of aromatics for calcium
aluminate catalyst was the lowest compared with other catalysts. This
may be due to the fact that the calcium aluminate catalyst was a base
catalyst which followed by radical reaction mechanism without hy-
drogen transfer. A large amount of 1-alkenes with lower molecular
would be favored in the increase of propylene and butene by catalytic
cracking. However, the gasoline fraction for FCC catalyst mainly con-
tains aromatics and iso-alkenes, which suffered from diffusional lim-
itations in catalytic cracking by ZSM-5 zeolites. To sum up, we chose
the calcium aluminate catalyst instead of FCC catalyst to combine with
ZSM-5 zeolites.
Fig. 10 shows the effect of the Z-40/(Z-40 + calcium aluminate)
ratio on the pyrolysis oil cracking at 600 °C. With the increase of the
blend ratio, the C2 hydrocarbon content slightly increased with in-
creasing Z-40 content, whereas the relative content of C3 and C4 hy-
drocarbons increased remarkably and reached a maximum value. Also,
excess amount of ZSM-5 had the adverse effect for the production of C3
and C4 hydrocarbons. Thus, the blend ratio at 0.7 could be the optimum
to enhance the production of C2-C4 hydrocarbons. The addition of
calcium aluminate was benefit to generate light olefins. The excess
amount of Z-40 led to the consumption of light olefins, which was due
to the fact that light olefins participated in the bimolecular reaction and
then formed aromatics and paraffins.
Table 4
Product distributions pyrolysis VR and catalytic cracking of pyrolysis oil using
calcium aluminate and FCC catalyst.
C5-C12 (%) C13-C22 (%) C23+ (%)
No Catalyst 43.85 35.93 11.11
Calcium Aluminate 51.95 36.06 0
FCC catalyst 63.21 23.3 2.24
7
Fuel 237 (2019) 1–9
10
21 C2
C3
20
C4 8
Relative Content (%)
19
6
18
17
4
16
15
2
14
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
ZSM-5/(ZSM-5+Calcium aluminate)
Fig. 10. Effect of the ratio of Z-40 and calcium aluminate on pyrolysis oil
catalystic cracking.
The HM-GCC process contained the fast pyrolysis of VR and the
catalytic cracking of pyrolysis gas-oil. The small-molecule pyrolysis gas-
oil could enter the pores of Z-40, whereas the large-molecule pyrolysis
oil would suffer thermal cracking or form coke on the surface of Z-40.
Calcium aluminate, which mainly contained marco/meso-pores, was
beneficial to the cracking of large-molecule pyrolysis oil. Besides, the
addition of calcium aluminate inhibited the hydrogen transfer reaction
by the reduction of acid sites per unit weight. Thus, replacing single Z-
40 with Z + C-cat catalysts resulted in the following two most im-
portant advantages: (1) It reduced the acid strength and concentration
of the pure Z-40 and thereby decreased the hydrogen transfer and ar-
omatization reactions; (2) It facilitated the cracking of bulkier mole-
cules and thus promoted the mass transport in zeolites [41,42].
3.3. Pyrolysis and gas-phase catalytic cracking in a dual-reactor
For further study the pyrolysis of VR and the gas-phase catalytic
cracking of pyrolysis gas-oil, the combination of a fluidized bed reactor
and a fixed bed reactor was used to investigate the yields and compo-
sitions of the gas and liquid products. In the fluidized bed reactor,
calcium aluminate was used as the catalyst and the reaction tempera-
ture was 600 °C; the catalyst/oil and steam/oil weight ratio were about
8 and 0.6, respectively. After the pyrolysis of the VR, the pyrolysis gas-
oil was transferred to the fixed bed reactor which was heated to 600 °C
and loaded ZSM-5 zeolites. The run in the fluidized bed reactor was the
pretreatment of the VR and used as the reference. The yields and
compositions of products were listed in Table 5.
As seen in Table 5, the product yields in the presence of ZSM-5
zeolites were significantly different than those in the absence of the
ZSM-5. The increase of gas yield was 19.47 wt% over the ZSM-5 when
calculated on the basis of the gas yield from the reference gas without
ZSM-5 zeolites. The yield of ethylene, propylene and butenes increased
by 4.69 wt%, 7.87 wt% and 3.37 wt% respectively with the Z-40 when
compared with the reference light olefins. The yield of light alkanes
(including ethane, propane and butane) was also increased due to the
hydrogen transfer reaction. These changes in light olefins were coin-
cident the Py-GC/MS results. In addition, the yields of liquid products
in the presence of ZSM-5 (53.24 wt%) were lower than those without
ZSM-5 (73.71 wt%).
The thermal cracking and catalytic cracking reaction occurred si-
multaneously in the fixed bed reactor. The pyrolysis oil suffered the
secondary cracking and produced lower molecules which were more
accessible to the ZSM-5 zeolites. The cracking of pyrolysis oil in the gas
phase could overcome the restriction of the “radical cage effect” and
reduce mass transfer limitations [15,22]. In addition, Bastiani and co-
workers found that the reduction of hydrogen transfer reactions in the
Y. Che et al. Fuel 237 (2019) 1–9

Table 5 Acknowledgments
Comparison of yields and compositions of products in a dual-reactor.
calcium aluminate (reference) calcium aluminate + Z-40 The authors acknowledge the financial support provided by the
National Natural Science Foundation of China (21576293, 21706287
gas yield (wt%) 17.03 36.50 and 21576294), the Natural Science Foundation of Shandong Province
gas composition (wt%)
(ZR2018QB008), and the Fundamental Research Funds for the Central
ethylene 5.09 9.78
propylene 3.18 11.05
Universities (18CX02117A).
butenes 0.69 4.06
ethane 2.39 4.13 References
propane 0.71 1.25
butanes 0.41 0.36
[1] Hussain AI, Aitani AM, Kubů M, Čejka J, Al-Khattaf S. Catalytic cracking of Arabian
others 4.56 5.87
Light VGO over novel zeolites as FCC catalyst additives for maximizing propylene
Liquid yield* (wt%) 73.71 53.24 yield. Fuel 2016;167:226–39.
Liquid composition (wt%) [2] Joshi JB, Pandit AB, Kataria KL, Kulkarni RP, Sawarkar AN, Tandon D. Petroleum
Gasoline 18.52 18.17 residue upgradation via visbreaking: a review. Ind Eng Chem Res 2008;47:8960–88.
Diesel 27.23 18.52 [3] Bari Siddiqui MA, Aitani AM, Saeed MR, Al-Khattaf S. Enhancing the production of
light olefins by catalytic cracking of FCC naphtha over mesoporous ZSM-5 catalyst.
VGO 19.13 11.92
Top Catal 2010;53:1387–93.
Heavy oil 8.83 4.64
[4] Corma A, Corresa E, Mathieu Y, Sauvanaud L, Al-Bogami S, Al-Ghrami MS. Crude oil
Coke yield (wt%) 9.26 10.26
to chemicals: light olefins from crude oil. Catal Sci Technol 2017;7:12–46.
[5] Alotaibi FM, González-Cortés S, Alotibi MF, Xiao T, Al-Megren H, Yang G.
* The total yield of gasoline, diesel, VGO and heavy oil. Enhancing the production of light olefins from heavy crude oils: turning challenges
into opportunities. Catal Today 2018.
[6] Das AK. Process for catalytic cracking of petroleum based feed stocks. United States
gas phase was more evident than that in the liquid phase [42]. Thus, it
Patent 5846402, Dec.08; 1998.
was clear that gas-phase cracking in the fixed bed reactor promoted the [7] Rusty MP, Lawrence LU. FCC process with improved yield of light olefins. United
conversion of the heavy molecules (diesel, VGO and heavy oil) to light States Patent 7312370, Dec.25; 2007.
hydrocarbons and olefin. However, the yield of gasoline changed [8] Meyers RA. Handbook of petroleum refining processes. McGrawHill Professional;
2003.
slightly. This may be due to the production of a large amount of BTX [9] Dharia D, Letzsch WS, Kim HN, Mccue D, Chapin L. Increase light olefins produc-
and the further cracking of diesel and VGO. Furthermore, the genera- tion. Hydrocarbon Process 2004;83:61–4.
tion of coke in the presence of ZSM-5 was 10.26 wt%, which was little [10] Parthasarathi RS, Alabduljabbar SS. HS-FCC High-severity fluidized catalytic
cracking: a newcomer to the FCC family. Appl Petrochem Res 2014;4:441–4.
higher than this without ZSM-5. This result indicated that gas-phase [11] Fujiyama, Yuichiro. Process for fluid catalytic cracking of oil. United States Patent
catalytic of pyrolysis oil formed a small amount of coke. It has been 5904837, May.18; 1999.
reported that the formation of coke could be less severe in gas phase [12] Chaohe Y, Xiaobo C, Jinhong Z, Chunyi L, Honghong S. Advances of two-stage riser
catalytic cracking of heavy oil for maximizing propylene yield (TMP) process. Appl
than that in the liquid phase, because much lower agglomeration re- Petrochem Res 2014;4:435–9.
actions occurred in the former status [23,24]. Thus, the reduction of [13] Rahimi N, Karimzadeh R. Catalytic cracking of hydrocarbons over modified ZSM-5
coke production was beneficial to the extension of catalyst life. zeolites to produce light olefins: a review. Appl Catal A 2011;398:1–17.
[14] AlKhattaf S. The influence of alumina on the performance of FCC catalysts during
hydrotreated VGO catalytic cracking. Energy Fuels 2003;17:62–8.
[15] Braden DA, Parrack EE, Tyler DR. Solvent cage effects. I. Effect of radical mass and
4. Conclusion size on radical cage pair recombination efficiency. II. Is geminate recombination of
polar radicals sensitive to solvent polarity? Coord Chem Rev 2001;211:279–94.
[16] Liang WJ. Heavy Oil Chemistry. Dongying, Shandong, China: China University of
A novel process hierarchical millisecond gas-phase catalytic Petroleum Press; 2000.
cracking (HM-GCC) process was proposed in this study. The process [17] García-Martínez J, Johnson M, Valla J, Li K, Ying JY. Mesostructured zeolite
could enhance the production of light olefins through millisecond Y—high hydrothermal stability and superior FCC catalytic performance. Catal Sci
Technol 2012;2:987–94.
pyrolysis of VR and gas-phase catalytic cracking of pyrolysis oil using [18] Serrano DP, Escola JM, Pizarro P. Synthesis strategies in the search for hierarchical
Z + C-cat. The pyrolysis products identified by GC/MS were light ole- zeolites. Chem Soc Rev 2013;42:4004–35.
fins, C5+ alkenes, alkanes, and aromatics. The production of alkenes [19] Na K, Choi M, Ryoo R. Recent advances in the synthesis of hierarchically nano-
porous zeolites. Microporous Mesoporous Mater 2013;166:3–19.
was enhanced with the increase of reaction temperature. The gas-phase
[20] Zhang M, Resende FLP, Moutsoglou A. Catalytic fast pyrolysis of aspen lignin via
catalytic cracking of pyrolysis oil was tested using the calcium alumi- Py-GC/MS. Fuel 2014;116:358–69.
nate, FCC, Z-40, Z-80, and Z-200 catalysts. The highest relative contents [21] Lu Q, Yang X, Dong C, Zhang Z, Zhang X, Zhu X. Influence of pyrolysis temperature
and time on the cellulose fast pyrolysis products: analytical Py-GC/MS study. J Anal
of C3 and C4 hydrocarbons were 16.92% obtained from Z-40 and
Appl Pyrol 2011;92:430–8.
20.95% achieved from Z-80 respectively. Due to the large pore struc- [22] Vafi K, McCaffrey WC, Gray MR. Minimization of coke in thermal cracking of
tures, FCC catalyst showed higher selectivity for the C9+ aromatics, Athabasca vacuum residue in a high-temperature short-residence time continuous
whereas Z-40 obtained larger amount of BTX. The calcium aluminate flow aerosol reactor. Energy Fuels 2012;26:6292–9.
[23] Bu W, Gray MR. Kinetics of vapor-phase cracking of bitumen-derived heavy gas oil.
was suitable for combination with Z-40 for further enhancing the pro- Energy Fuels 2013;27:2999–3005.
duction of C2-C4 hydrocarbons. With the increase of Z-40/(Z- [24] Wu G, Katsumura Y, Chihiro Matsuura A, Ishigure K, Kubo J. Comparison of liquid-
40 + calcium aluminate) ratio, the relative content of C3 and C4 hy- phase and gas-phase pure thermal cracking of n-hexadecane. Ind Eng Chem Res
1996;35:4747–54.
drocarbons increased remarkably and reached to the maximum at the [25] Zeng M, Yuan W, Wang Y, Zhou W, Zhang L, Qi F. Experimental and kinetic
ratio of 0.7. While the excess amount of Z-40 had the adverse effect for modeling study of pyrolysis and oxidation of n-decane. Combust Flame
the production of C3 and C4 hydrocarbons. A bench scale fluidized bed- 2014;161:1701–15.
[26] Liang WJ, Que GH, Liu CG, Yang QS. Petroleum chemistry. 2nd ed. Dongying,
fixed bed reactor was used to further study the HM-GCC process. The Shandong, China: China University of Petroleum Press; 2009.
yield of ethylene, propylene and butenes increased by 4.69, 7.87 and [27] Zhang J, Che Y, Wang Z, Qiao Y, Tian Y. Coupling process of heavy oil millisecond
3.37 wt% respectively with the presence of Z-40. These results in- pyrolysis and coke gasification: a fundamental study. Energy Fuels
2016;30:6698–708.
dicated that the HM-GCC process was a feasibility technology for en-
[28] Al-Khattaf SS, Ali SA. Catalytic cracking of Arab super light crude oil to light olefins:
hancing the production of light olefins. an experimental and kinetic study. Energy Fuels. 2018;32:2234–44.
[29] Su X, Zan W, Bai X, Wang G, Wu W. Synthesis of microscale and nanoscale ZSM-5
zeolites: effect of particle size and acidity of Zn modified ZSM-5 zeolites on ar-
omatization performance. Catal Sci Technol 2017;7:1943–52.
5. Notes [30] Lemonidou AA, Vasalos IA. Preparation and evaluation of catalysts for the pro-
duction of ethylene via steam cracking: effect of operating conditions on the per-
The authors declare no competing financial interest. formance of 12CaO-7Al 2O3 catalyst. Appl Catal 1989;54:119–38.

8
Y. Che et al.
[31] Corma A, González-Alfaro V, Orchillés AV. The role of pore topology on the be-
haviour of FCC zeolite additives. Appl Catal A 1999;187:245–54.
[32] Zhu X, Liu S, Song Y, Xu L. Catalytic cracking of C4 alkenes to propene and ethene:
influences of zeolites pore structures and Si/Al2 ratios. Appl Catal A
2005;288:134–42.
[33] Al-Sabawi M, Lasa HD. Influence of zeolite crystallite size on methyl-cyclohexane
catalytic conversion products. Fuel 2012;96:511–23.
[34] Guisnet M, Gnep NS, Aittaleb D. Doyemet YJ. Conversion of light alkanes into
aromatic hydrocarbons: VI. Aromatization of C2–C4 alkanes on H-ZSM-5 —reaction
mechanisms. Appl Catal A 1992;23:255–75.
[35] Long H, Wang X, Sun W, Xiong G, Wang K. Effect of acidity on n-octene reaction
over potassium modified nanoscale HZSM-5. Fuel 2008;87:3660–3.
[36] Mochizuki H, Yokoi T, Imai H, Namba S, Kondo JN, Tatsumi T. Effect of desilication
of H-ZSM-5 by alkali treatment on catalytic performance in hexane cracking. Appl
Catal A 2012;449:188–97.
[37] Yang X, Liu J, Fan K, Rong L. Hydrocracking of Jatropha oil over non-sulfided PTA-
Fuel 237 (2019) 1–9
NiMo/ZSM-5 catalyst. Sci Rep 2017;7:41654.
[38] Konno H, Tago T, Nakasaka Y, Ohnaka R, Nishimura J, Masuda T. Effectiveness of
nano-scale ZSM-5 zeolite and its deactivation mechanism on catalytic cracking of
representative hydrocarbons of naphtha. Microporous Mesoporous Mater
2013;175:25–33.
[39] Vogt ET, Weckhuysen BM. Fluid catalytic cracking: recent developments on the
grand old lady of zeolite catalysis. Chem Soc Rev 2015;44:7342–70.
[40] Hollander MAD, Wissink M, Makkee M, Moulijn JA. Gasoline conversion: reactivity
towards cracking with equilibrated FCC and ZSM-5 catalysts. Appl Catal A
2002;223:85–102.
[41] Siddiqui MAB, Aitani AM, Saeed MR, Al-Yassir N, Al-Khattaf S. Enhancing propy-
lene production from catalytic cracking of Arabian Light VGO over novel zeolites as
FCC catalyst additives. Fuel 2011;90:459–66.
[42] Bastiani R, Lam YL, Henriques CA, Teixeira da Silva V. Application of ferrierite
zeolite in high-olefin catalytic cracking. Fuel 2013;107:680–7.