Journal of Catalysis 369 (2019) 72–85

Contents lists available at ScienceDirect

Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat

Construction and practical application of a novel zeolite catalyst for

hierarchically cracking of heavy oil
Meng Pan a, Jiajun Zheng a,⇑, Yujian Liu b, Weiwei Ning a, Huiping Tian b, Ruifeng Li a,⇑
a
Research Centre of Catalysis & Chemical Technology for Energy, Taiyuan University of Technology, Taiyuan 030024, China
b
Sinopec Research Institute of Petroleum Processing, Beijing 100083, China

a r t i c l e i n f o a b s t r a c t

Article history: The conversion of the heavy fractions of crude oil into transportation fuels and chemical feedstocks is still
Received 14 August 2018 a severe challenge for fluid catalytic cracking (FCC) technology, in which geometrical parameters of the
Revised 23 October 2018 catalyst are of key importance to the accessibility for reactant and product molecules to travel through
Accepted 24 October 2018
the pore space. Here a new core-shell zeolite catalyst with hierarchical pore network from macro- via
Available online 14 November 2018
meso- to micropores is developed by controlling overgrowth of another zeolite shell on the surface of
a commercial Y-zeolite. The zeolite shell is a distinctive nano ZSM-5 crystals. The core-shell zeolite cat-
Keywords:
alyst has a high thermal and hydrothermal stability because of high crystallinity, which can meet the
FCC
Core-shell structure
industrial requirements. More importantly, the heavy oil can be cracked by the designed zeolite catalyst
Y-zeolite step-by-step into the required products. The meso- and macroporous shell facilitates pre-cracking the
Heavy oil large molecules in heavy oil into moderate ones which can directly diffuse into the micropores of the
Gradual cracking Y-zeolite to be cracked selectively into the desired products. Both the pore structure network and acid
Industrialization sites distribution of the core-shell zeolite catalyst in the FCC processing are completely harmonious. As
a practical example, the catalytic cracking of VGO heavy oil is actualized over the new catalyst. The
results showed that the conversion of heavy oil and the yield of gasoline increased by 1.8 wt% and
8.9 wt% respectively, LPG decreased by 6.9 wt%, iso-paraffins and olefins in gasoline fraction increased
by 5.21 wt% and 6.31 wt% respectively, aromatics decreased by 11.55 wt%, because of the graduated
cracking in core-shell zeolite and returned isomerization in nano ZSM-5, compared with commercially
industrial FCC catalyst.
Ó 2018 Elsevier Inc. All rights reserved.

1. Introduction vacuum residues into fractions of motor fuels (mainly gasoline),

undoubtedly, belongs to one of the most important refinery tech-
Zeolites, with their unique pore structure, adjustable acidity nologies [9,10]. Therefore, designing and developing more effective
and high thermal/hydrothermal stability, have caused a revolution FCC catalysts is vital for the industrial production at this stage.
in the fields of heterogeneous catalysis and separation processes Currently, FAU-type Y-zeolite which has a uniform microporous
[1–4]. Especially in the early 1960s, the introduction of zeolites structure with the inlet diameter of 0.74 nm is the primary active
led to groundbreaking developments in the oil refining industry component of the cracking catalyst used in FCC process [11,12].
[5]. Up to now, zeolites have developed into one of the major cat- However, the critical molecular diameters of heavy oil or residues
alytic materials in the petrochemical field [6]. However, the pre- range from 1.2 to 15 nm (boiling point >400 °C, carbon number
sent situation of worldwide reserves of crude oils is challenging: >20) [7]. It is almost impossible that the large molecules in heavy
production of light crude oils is declining and production of heavy oil enter into the micropores of the zeolite to be cracked directly
and extra-heavy oils is increasing. Heavy crude oil brings great dif- [11–15]. Thus, the desired FCC catalysts for processing heavy oil
ficulties for the present refining technologies [7,8]. The fluid cat- need to have improved diffusion abilities and enhanced acid sites
alytic cracking (FCC) unit which converts atmospheric and accessibilities [16–21]. To achieve this goal, many efforts are made
to introduce the meso- and/or macroporosities into zeolite crystals,
namely developing hierarchical zeolitic materials [3,4]. Hierarchi-
⇑ Corresponding authors at: Research Centre of Catalysis & Chemical Technology cal zeolites indeed exhibited significantly enhanced diffusion effi-
for Energy, Taiyuan University of Technology, 79# West Yingze Street, Taiyuan ciency in comparison with purely microporous zeolites. But the
030024, China.
E-mail addresses: [email protected] (J. Zheng), [email protected] (R. Li).

https://doi.org/10.1016/j.jcat.2018.10.032
0021-9517/Ó 2018 Elsevier Inc. All rights reserved.
 M. Pan et al. / Journal of Catalysis 369 (2019) 72–85 73

formation of additional porous systems would more or less cause

the deterioration of the intrinsic crystal structure of zeolite [20,21].
Alternatively, the capability of the catalysts for heavy oil pro-
cessing can be enhanced efficiently by combining the active
matrix, such as mesoporous silicas, aluminosilicates and aluminas
with the zeolite crystal. Presently the commercial FCC catalyst is
essentially a mechanical mixture of Y-zeolite, active matrix (amor-
phous silica-alumina) together with various additives and binders
[2,22]. The role of the active matrix containing meso- and/or
macroporous structure is to pre-crack the bulky molecules into
moderate ones which are able to diffuse into zeolite channels for
further selective cracking [6,7,23]. However, it is reported that
the mechanical mixing of different active components provides
only an unrestricted, open reaction environment where the reac-
tion pathways are random [24,25]. Moreover, the uncertain dis-
tance between two kinds of active sites in the mechanical
mixture also seriously impacts the catalytic activity and selectivity
[26]. Thus, regulating the dispersion and combination of active
matrix and zeolite is a significant direction for designing new
FCC catalysts.
Fig. 1. Developing trend of the meso- and/or macroporous matrix materials in the
Recently, special attentions are paid to the core-shell structured
core-shell FCC catalysts.
catalysts because they can provide the definite reaction pathways
and enhance the synergy of different active sites [24,25,27]. So
many core-shell structured composites with employing zeolite meso- and macropores in the shell are composed of nano ZSM-5
crystal as core and meso- and/or macroporous matrix as shell are crystals. Performance studies show that the crystalline shell gives
developed aiming to process heavy oil more efficiently [12–20]. the zeolite composite a high thermal and hydrothermal stability
For example, Meng et al. [12] synthesized a composite material and the special core-shell structure endows it with an enhanced
Y/ASA by wrapping Y-zeolite in mesoporous silica-alumina (ASA) ability of catalytically cracking bulky molecules. The application
gel which showed higher activity than the mixture-derived cata- of the zeolite composite in the FCC process is also preliminarily
lysts in the catalytic cracking of n-decane and light diesel. Besides, explored in the ACE-Model unit. The results may provide valuable
other core-shell composites with different mesoporous shells, such information for studying chemistry of catalytic cracking reaction
as Al-MCM-41, Al-SBA-15 and TUD-1 have also been reported and give an alternative direction to developing efficient FCC cata-
[13–20]. These core-shell materials all exhibited unique properties lyst for processing heavy oil.
in the reactions involving large molecules because of the close
combination of matrix with zeolite and the excellent connection
between meso- and microporosities. Unfortunately, the poor 2. Experimental
thermal and hydrothermal stability resulted from the amorphous
nature of the mesoporous matrix is a serious problem that needs 2.1. Sample preparation
to be solved. Considering that the transition aluminas (i.e. g-, j-
or c-Al2O3) own higher thermal and hydrothermal stability than Core-shell zeolite composites were prepared by a hydrothermal
the amorphous silicas or aluminosilicates, Jiao et al. [21] prepared method. The detailed process was illustrated in Fig. S1 in the Sup-
a core-shell zeolite Y@c-Al2O3 nanorods composite by employing plementary Material. Typically, 0.8 g of sodium hydroxide (98 wt%,
mesoporous c-Al2O3 as shell. The FCC catalyst derived from this Aladdin), 9.0 g of tetraethylammonium bromide (TEABr, 99 wt%,
core-shell composite showed better contaminant resistant ability Aladdin), 19 mL of silica sol (40 wt%, Qingdao Haiyang Chemical
and gave higher gasoline yield than the reference catalyst in the Co., Ltd.) and 3 mL of ammonia water (25–28 wt%, Aladdin) were
processing of heavy oil. Comparing with the transition aluminas, added to 86 mL of distilled water and stirred at room temperature
zeolite crystals not only possess high thermal and hydrothermal for 0.5 h before 6–20 g of commercial Y-zeolite was added slowly
stability but also have adjustable chemical compositions and thus under vigorous stirring. The mixtures were stirred for another
show more controllable acidic properties which play a key role in 2 h at room temperature and then transferred into an autoclave
the catalytic cracking reactions [6]. Moreover, recent studies have and heated at 180 °C for 0–110 h at autogenous pressure. The as-
shown that the introduction of Brönsted sites in the matrix of synthesized products were recovered by filtrating, washed with
the FCC catalyst can enhance the conversion of heavy oil signifi- distilled water, dried at 90 °C overnight and denoted as Y@NZ
cantly [11,28]. The zeolitization is an effective way of increasing (m)-t (‘‘m” stands for the additive amount of commercial Y-
the Brönsted sites of the matrix. Hence, the zeolitization of zeolite and ‘‘t” stands for the synthesis time). For example, Y@NZ
meso- and/or macroporous matrix should be an important direc- (6)-72 represents the product synthesized by adding 6 g of com-
tion for the development of new FCC catalysts, as shown in mercial Y-zeolite into the above system and at 180 °C for 72 h.
Fig. 1. Considering that the preparation of ordered mesoporous The same synthetic technology and conditions for preparing the
molecular sieves with high crystalline walls is extremely difficult zeolite materials were practiced in the scale-up test in a 10 L
at present, it should be emphasized that the presence of non- autoclave.
ordered mesoporosity is more than sufficient for most applications, Commercial Y-zeolite (named as Y-zeolite) and Commercial
notably catalysis, as reported by White et al [29]. ZSM-5 (named as ZSM-5) used as references were provided by
Herein, a novel core-shell zeolite composite with a non-ordered Sinopec Research Institute of Petroleum Processing. Moreover, a
meso- and macroposous shell and a hardly dissolved commercial nano ZSM-5 zeolite was also synthesized for reference according
Y-zeolite core is prepared by employing the Al and/or Si species to a literature reported by Larsen et al. [30] and denoted as
generated from the dissolution of Y-zeolite as the partial nutrients NanoZSM-5. The Si/Al ratios of these three samples are 2.5, 16.9
for the growth of the shell. More importantly, the walls of the and 15.3 respectively.
74 M. Pan et al. / Journal of Catalysis 369 (2019) 72–85

All the samples were calcined in air at 550 °C for 5 h in order to 2.4. Thermal and hydrothermal stability tests
remove the concealed organic templates in the cages and the pore
channels. The NH+4 form samples were obtained by ion exchange In the thermal treatment, 1 g of H-form sample was calcined at
for three times with 0.5 mol/L of NH4NO3 solution (homemade) 800 °C in dry air for 20 h and then cooled to room temperature. The
at 60 °C for 2 h at each time. The H-form samples were then hydrothermal treatment was performed under the conditions sim-
obtained by calcining the NH+4 form samples at 550 °C for 5 h and ilar to those used in FCC processes according to the literature [31].
marked as HSamples (for example, HY refers to the H-form com- First, 0.2 g of H-form sample was placed in a homemade reactor
mercial Y-zeolite). Note that HY@NZ(6), HY@NZ(10) and HY@NZ and heated to 600 °C in a nitrogen flow with a heating rate of
(20) represent H-form Y@NZ(6)-72, H-form Y@NZ(10)-72 and H- 5 °C/min. Then, the steam with a partial pressure of 0.2 was added
form Y@NZ(20)-72 respectively. into the reactor at the fixed temperature of 600 °C for 5 h. Finally,
the sample was cooled to room temperature in the nitrogen flow.
2.2. Preparation of FCC catalysts
2.5. Catalytic tests
The catalysts used in ACE evaluation were manufactured by
1,3,5-Triisopropylbenzene (TIPB, 95 wt%, Aladdin) cracking
Sinopec Research Institute of Petroleum Processing. First, the zeo-
reactions were conducted at atmospheric pressure using nitrogen
lite samples were ion exchanged in a solution of rare earth chloride
as carrier gas in a continuous flow fixed bed reactor and in a batch
at 90 °C for 1 h. After filtration, washing and drying, the samples
fixed bed micro-activity test unit (MAT), respectively. Prior to the
were calcined at 550 °C for 2 h. The above procedure was repeated
experiments, the H-form samples were made into 20–40 mesh par-
once. Then, the zeolite samples modified by rare earth, binders and
ticles and activated under flowing N2 (50 mL/min) at 550 °C for 2 h.
kaolin were mixed by grinding. The binders were composed of
Then the reactor temperature was decreased to 450 °C. In the con-
pseudoboehmite and amorphous silica-alumina. The catalysts
tinuous flow fixed bed reactor, TIPB was fed into the reactor by a
were obtained by adding distilled water into the above mixture
micro-pump at a rate of 1.0 mL/h and the flow rate of N2 carrier
under stirring and then treated in sequence of spray drying, wash-
was 50 mL/min. The lines were kept over 240 °C using heating
ing and drying. The final solid particles were calcined at 350 °C for
tapes. All of the products were analyzed by an online gas chro-
2 h. The total content of zeolites in the catalyst is about 35 wt%.
matograph (Agilent 1790, FID detector, HP-INNOWax capillary col-
Before the catalytic cracking of heavy oil, the FCC catalysts were
umn). In the batch fixed bed micro-activity test unit (MAT), TIPB
deactivated by a steam treatment in 100% steam at 800 °C for 17 h.
was injected by a syringe pump during 30 s. Liquid products were
condensed and analyzed by Agilent 7890 gas chromatography
2.3. Characterization equipped with a DB-624 capillary column. The conversion, conver-
sion efficiency and selectivity were calculated from the formulas as
XRD patterns were recorded using a Shimadzu XRD-6000 X-ray follows:
diffractometer, which employed Ni-filtered Cu Ka radiation and
conversion ¼ wt =ðwt þ wr Þ  100%;
was operated at 40 kV and 30 mA. The crystallinity of Y-zeolite
phase was determined from the intensities of the peaks located
conversion efficiency ¼ conversion=ðca  mÞ;
at 6.2°, 10.2°, 18.7° and 31.5° using the commercial Y-zeolite as ref-
erence. The crystallinity of ZSM-5 phase was determined from the
selectivity ¼ wi =wt  100%;
intensities of the peaks located at 7.9°, 8.8°, 23.1°, 23.3° and 24.5°
using the commercial ZSM-5 as reference. FTIR spectra of the cal- where wt represents the total weight of the products, wr stands for
cined samples were obtained on a Shimazdu FTIR-8400 spectrom- the weight of the unreacted TIPB, ca represents the acid concentra-
eter in KBr pellets. FTIR spectra of pyridine (Py)-adsorbed samples tion of the catalyst determined by the FTIR spectra of Py-adsorbed
were also recorded on the Shimazdu FTIR-8400 spectrometer. The samples at 150 °C, m represents the mass of the catalyst and wi
samples were activated under a vacuum of 10
2 Pa at 400 °C for stands for the weight of a specific product.
4 h. IR spectra of Py-adsorbed samples were recorded at the tem- The catalytic cracking of heavy oil was carried out in the ACE-
peratures of 150 °C, 250 °C and 350 °C respectively. Crystal size Model unit R+ of KTI Corporation. The feedstock was a specific
and morphology of the as-synthesized samples were investigated VGO delivered by a Chinese refinery and its properties are listed
on a Hitachi S4800 scanning electron microscopy (SEM) coupled in Table S1. Before ACE tests, micro-activity indexes of the deacti-
with energy dispersive spectroscopy (EDS). Crystal faces and inter- vated catalysts were obtained according to ASTM D3907. ACE tests
face structure were studied by TEM (JEOL/JEM-2010) and high res- were conducted at atmospheric pressure with a reaction tempera-
olution TEM (JEM-200CX) coupled with selected area electron ture of 525 °C. The amount of catalyst was 9.0 g and the catalyst/oil
diffraction (SAED). N2 adsorption-desorption measurement was (C/O) ratio was 3.2 g/g. The injection time (TOS) of feed was 32 s.
performed in a NOVA 1200e gas sorption analyzer. Mesopore size After the reaction, the remaining hydrocarbons were stripped from
distribution was calculated using the Barrett-Joyner-Halenda the catalyst bed and the reactor was purged under flowing N2
(BJH) pore size model as applied to the desorption branch of the (120 mL/min) for a period of 10 min. During the cracking and strip-
isotherm. Temperature-programmed desorption of ammonia ping modes, the liquid product was collected in a glass receiver,
(NH3-TPD) was carried out on a TP-5076 apparatus equipped with which was located at the end of the reactor exit and was main-
a thermal conductivity detector (TCD). 100 mg of catalyst was tained at a temperature of
15 °C. The gaseous products were col-
loaded in a quartz tube and supported by silica wool. Then, it lected in a gas receiver by water displacement method. The dry gas
was treated at 550 °C for 2 h in a highly purified He (99.999%) flow. (H2 and C1–C2) and LPG (C3–C4) in the gaseous products were ana-
After being cooled down to 120 °C, the sample was saturated with lyzed using a HP 5880A gas chromatograph. The boiling point dis-
a NH3 (5%)/He gas mixture, followed by being purged with the tribution of the liquid products was determined using a standard
highly purified He at 120 °C for an hour. Then, the temperature simulated distillation according to ASTM D2887. Three liquid frac-
was increased at a rate of 10 °C/min in a highly purified He flow tions were defined as gasoline (C5-221 °C), light cycle oil (LCO,
(30 mL/min), and desorption of NH3 was monitored. In this exper- 221–343 °C), and bottoms (343 °C+). Coke deposited on the cata-
iment, a gas dryer was set up between the sample and the TCD to lyst after cracking was determined by in situ combustion through
avoid the influence of water. the use of a BF QGS-10 instrument.
 M. Pan et al. / Journal of Catalysis 369 (2019) 72–85 75

3. Results and discussion bulk. Moreover, the amount of the nano particles in the as-
synthesized composites is anti-correlated to the amount of the
3.1. Synthesis and characterization of core-shell zeolite composites added Y-zeolite in the initial gel. When adding a small amount of
Y-zeolite (i.e. 6 g per 108 mL gel), the octahedral particles are
Fig. 2A shows the XRD patterns of the commercial Y-zeolite, the almost entirely wrapped by the nano particles in the resultant pro-
commercial ZSM-5 and the as-synthesized composites after 72 h of duct (Fig. 3A and B). The abundant meso- and macropores can be
crystallization. As shown in Fig. 2A, both the characteristic peaks of observed because of the loose accumulation of these nano particles
Y-zeolite and those of ZSM-5 can be simultaneously detected in the (Fig. 3B). The increase of the Y-zeolite in the initial gel leads to the
XRD patterns of the as-synthesized composites. Even though the decrease of the nano particles on the surfaces of the octahedral
amount of the added Y-zeolite reaches up to 20 g per 108 mL gel, particles (Fig. 3C and 3D). When the amount of the added Y-
the peaks at 7.9°, 8.8°, 23.1°, 23.3° and 24.5° assigned to the zeolite reaches up to 20 g per 108 mL gel, the nano particles are
ZSM-5 phase can also be observed in the XRD pattern of the sample too few to wrap the octahedral particles (Fig. 3E and F). Consider-
Y@NZ(20)-72. In addition, no crystalline impure phase can be ing the results of Fig. 2, the octahedral particles in the as-
found. The results indicate that the Y-zeolite and ZSM-5 phases synthesized composites should be Y-zeolite crystals and the nano
coexist in all the synthesized composites. The varying peak inten- particles can be deduced to be ZSM-5 phase.
sities for the different samples also suggest that the relative con- The textural properties of the as-synthesized composites were
tent of those two kinds of zeolite phases in the synthesized characterized using N2 adsorption-desorption measurement and
composites can be adjusted by changing the amount of added Y- compared with those of pure commercial zeolites. The results are
zeolite. Moreover, the zoomed pattern in the range of 22.5–25.0° shown in Fig. 4 and Table 1. The adsorption-desorption isotherms
presents the diffraction peaks of Y@NZ(6)-72 as well as Y@NZ of pure Y-zeolite and ZSM-5 are typical of microporous materials
(10)-72 are broader than those of commercial ZSM-5 zeolite, which (type-I) [34,35]. However, an obvious hysteresis loop can be
indicates that the ZSM-5 crystals in zeolite composites probably observed in the isotherm of Y@NZ(6)-72 or Y@NZ(10)-72 at high
possess a relative small size [32]. Fig. 2B shows FTIR spectra of relative pressure (p/p0 > 0.9) and the slope of the isotherm
the samples, in which the characteristic vibration bands of silica- increases gradually with the relative pressure of N2 (Fig. 4A), indi-
alumina zeolite can be observed, although the bands position and cating the existence of the meso- and macroporous structures [32].
intensity are not completely identical. The bands around 577 cm
1 Especially a sharp increase of the absorption capacity at high rela-
and 547 cm
1 are attributed to the six-membered ring vibration of tive pressure (p/p0 > 0.9) occurs in the isotherm of Y@NZ(6)-72,
Y-zeolite crystal and the five-membered ring vibration of ZSM-5 which is a typical characteristic of the macroporous structures
zeolite, respectively [32,33]. Fig. 2B further confirms the coexis- [36]. The BJH pore size distribution curves derived from desorption
tence of Y-zeolite and ZSM-5 phases in all the as-synthesized branch (Fig. 4B) further confirm the presence of substantial meso-
composites. and macroporous in Y@NZ(6)-72 and Y@NZ(10)-72. As expected,
The commercial Y-zeolite crystals used here are octahedral par- Table 1 makes it clear that the mesopore volume and the external
ticles with smooth surfaces (Fig. S2A and S2B) and the commercial surface area of Y@NZ(6)-72 as well as Y@NZ(10)-72 are much
ZSM-5 zeolite displays a typical coffin-like morphology with the higher than those of pure Y-zeolite or ZSM-5. These meso- and
single crystal size of about 2 lm  1.2 lm  0.5 lm (Fig. S2C and macroporous structures should come from the polycrystalline
S2D). However, the as-synthesized composites display different accumulation of nano ZSM-5, as shown in Fig. 3. The mesopore vol-
morphology, as shown in Fig. 3. First, the microsized coffin-like ume of the as-synthesized composites increases with the content
ZSM-5 crystals don’t appear in the SEM images of the as- of the ZSM-5 phase, which further supports this result. Table 1 also
synthesized composites. Second, although the octahedral particles shows that the as-synthesized composites all have large micropore
of Y-zeolite can still be detected in the SEM images of the as- volumes and micropore areas, larger than those of the commercial
synthesized composites, their surfaces become very rough due to ZSM-5. This result indicates that the channels of the Y-zeolite core
the presence of many spherical particles smaller than 100 nm in are not blocked by the shell. Actually, it is difficult to imagine that
diameter. These nano particles almost exclusively grow on the sur- the two pore systems do not communicate in the case of the
faces of the octahedral particles. It is hard to detect them in the couple between Y-zeolite and ZSM-5 with three-dimensional pore

A (b) B
Intensity (a.u.)

(a) Y
577
(c)
ZSM-5
(b)
23.0 23.5 24.0 24.5
Intensity (a.u.)
Intensity (a.u.)

2 Theta (Degree)

Y
(a)
(c)

(b)
(a) ZSM-5
547
ZSM-5
5 10 15 20 25 30 35 400 600 800 1000 1200 1400
-1
2 Theta (Degree) Wavenumber (cm )

Fig. 2. XRD patterns (A) and FT-IR spectra (B) of the commercial Y-zeolite, the commercial ZSM-5 and the as-synthesized composites: (a) Y@NZ(6)-72; (b) Y@NZ(10)-72; (c)
Y@NZ(20)-72.
76 M. Pan et al. / Journal of Catalysis 369 (2019) 72–85

Fig. 3. SEM images of the as-synthesized composites: (A) and (B) Y@NZ(6)-72; (C) and (D) Y@NZ(10)-72; (E) and (F) Y@NZ(20)-72.

320 0.7
Y Y
A B
Y@NZ(20)-72 0.6 Y@NZ(20)-72
280
Y@NZ(10)-72 Y@NZ(10)-72
Y@NZ(6)-72 Y@NZ(6)-72
Volume Adsorbed (mL/g)

0.5
240 ZSM-5 ZSM-5
dV/dlogD (mL/g)

0.4
200

0.3
160

0.2
120
0.1
80
0.0
0.0 0.2 0.4 0.6 0.8 1.0 10 100
Relative Pressure (P/P0) Pore Diameter (D/nm)

Fig. 4. N2 adsorption-desorption isotherms (A) and BJH pore size distribution curves (B) of the samples.

systems and pores running along the a- and b-axes, respectively [38]. It means that in the catalytic reactions involving bulky mole-
[37]. Furthermore, the abundant meso- and macropores in the cules, the reactants can enter these meso- and macropores first to
shell provide a more efficient path of mass transfer for the be translated into the moderate molecules which can then enter
macromolecules with compared to the micropores in the shell into the micropores of the Y-zeolite core to be further translated.
 M. Pan et al. / Journal of Catalysis 369 (2019) 72–85 77

Table 1
Textural properties of the samples.

Samples SBETa (m2/g) Vmicb (cm3/g) Smicb (m2/g) Sextb (m2/g) Vmesc (cm3/g)
Y 769 0.28 730 39 0.08
Y@NZ(20)-72 656 0.24 622 34 0.08
Y@NZ(10)-72 519 0.17 446 74 0.12
Y@NZ(6)-72 487 0.16 405 82 0.26
ZSM-5 377 0.14 342 35 0.03
a
Value determined by the BET method.
b
Value determined by the t-plot method.
c
Value determined by BJH desorption.

In order to further reveal the structure and composition of the proves this result, which clearly exhibits the polycrystal-like
as-synthesized composites, the sample Y@NZ(10)-72 was charac- diffraction pattern of MFI framework. The above mentioned results
terized by TEM, EDS and SAED. The results are shown in Fig. 5 make clear that the as-synthesized composite has a core-shell
and Fig. S3. Interestingly, a unique core-shell structure can be structure with wrapping Y-zeolite in a meso- and macroporous
observed in the TEM image (Fig. 5A). The core displays a typical shell. More importantly, the walls of the meso- and macropores
octahedral morphology and the shell is made up of many nano par- are composed of the nano ZSM-5 crystals. To the best of our knowl-
ticles. A large number of meso- and macropores exist in the shell, edge, this is the first report on the synthesis of such a core-shell
which is consistent with the results of Fig. 3 and Fig. 4. The core- zeolite composite.
shell structure is also detected by the EDS analysis which shows
the Si/Al ratios of the core and the shell are different (Fig. 5B and 3.2. Formation mechanism of core-shell zeolite composites
Fig. S3). The Si/Al ratio of the octahedral core is 3.5, which is close
to the 3.4 of the commercial Y-zeolite (obtained by the EDS In order to explore the formation mechanism of the core-shell
analysis, as shown in Fig. S4). However, the Si/Al ratio of the nano zeolite composites, two experiments were conducted. Without
particles in the shell is 17.7 differs vastly from the 3.4 of the adding Y-zeolite in the initial gel, an amorphous phase was
Y-zeolite. In general, the chemical composition of a given zeolite obtained after crystallization for 72 h (Fig. S5A). When employing
is an important characteristic which defines its properties [34]. sodium aluminate instead of Y-zeolite in the initial gel, the resul-
The different Si/Al ratios indicate that the core and the shell are tant product after 72 h of crystallization was pure microsized
composed of different zeolite phases. The core-shell structure ZSM-5 with the crystal size of about 10 lm  8 lm  5 lm
and the meso- and macropores in the shell are also visible in the (Fig. S5B). However, these microsized ZSM-5 crystals can’t be
high-resolution TEM (HRTEM) image (Fig. 5C). It’s notable that detected in the sample Y@NZ(10)-72 (Fig. S5C). The result indicates
many meso- and macropores connect with the core directly that the Y-zeolite plays a crucial role in the formation of the core-
(marked by red circles in Fig. 5C). More interestingly, the HRTEM shell zeolite composites.
image in Fig. 5D presents very clear lattice fringes with different To gain deeper insight into the formation process of the core-
orientations, indicating that the nano particles in the shell are shell zeolite composites, the products collected at different synthe-
highly crystalline. The SAED image (the inset in Fig. 5D) further sis times according to the formula of the sample Y@NZ(10)-72

A B

C D

Fig. 5. Characterization results of the sample Y@NZ(10)-72: (A) TEM image; (B) EDS analysis; (C) and (D) HRTEM images.
78 M. Pan et al. / Journal of Catalysis 369 (2019) 72–85

were studied by XRD, SEM and TEM. The results are presented in Y-zeolite core from excessive dissolution and the crystallization
Fig. 6 and Fig. S6. XRD patterns (Fig. 6A) show that only the sharp of nano ZSM-5 in the shell is completed within 72 h.
characteristic peaks of Y-zeolite can be detected in the initial gel. A possible formation mechanism is proposed based on the
By increasing the synthesis time to 20 h, the intensity of the char- above studies and illustrated in Fig. 7. Initially, the alkaline
acteristic peaks of Y-zeolite decreased but the characteristic peaks hydrothermal system led to a partial dissolution of the Y-zeolite
of ZSM-5 phase did not appear. The result indicates that the growth crystals (Fig. 7a to b). As a result, the Al and/or Si species were
of the nano ZSM-5 crystals should lag behind the dissolution of the extracted from the parent Y-zeolite [41]. The extracted Al and/or
Y-zeolite [39]. When extending the synthesis time to 40 h, a partial Si species together with the supplemental silicon species and tem-
crystalline ZSM-5 phase appeared. It developed into a high crys- plates created an appropriate environment for the formation of the
talline phase after 72 h of synthesis. A further extension of the syn- ZSM-5 precursors. In order to minimize the potential energy [42],
thesis time to 110 h had little effect on the crystallinity of ZSM-5 the formed precursors then aggregated on the surface of
phase, suggesting that the crystallization of ZSM-5 phase in the Y-zeolite crystal (Fig. 7c). The following and continual crystalliza-
zeolite composites may be completed within 72 h. These observa- tion resulted in a complete transformation of the precursors into
tions were further associated with the SEM images. In the initial the nano ZSM-5 crystals. The formation of the nano crystals rather
gel, Y-zeolite displayed a defined octahedral morphology than microsized crystals may be associated with the supply of the
(Fig. 6C). After 20 h of synthesis, a slight dissolution occurred on Al species because the Al species needed for the growth of ZSM-5
the surface of the Y-zeolite crystal (Fig. 6D). Meanwhile, a few crystals can only be obtained from the dissolution of the
worm-like particles appeared on the surface (the inset in Y-zeolite which is a slow and gradual process, especially after
Fig. 6D). These worm-like particles can also be observed in the 40 h of synthesis (see Fig. 6A). Moreover, the entirely different
TEM image (Fig. S6). They seem to be aggregating towards the structures between Y-zeolite and ZSM-5 prevented the oriented
Y-zeolite crystals. Ren et al. [40] reported that the worm-like par- attachment of the nano crystals and then avoid the formation of
ticles formed at the early stage and then aggregate to form the con- a closely intergrown layer [43]. Finally, the loosely accumulational
densed aggregates during the crystallization of ZSM-5 zeolite from nano ZSM-5 crystals constituted a meso- and macroposous shell on
a diluted heterogeneous system. From this we may certainly infer the Y-zeolite core (Fig. 7d). During the synthesis, when increasing
that the worm-like particles in the sample Y@NZ(10)-20 are the the addition amount of Y-zeolite in the initial gel, the residual
precursors of the ZSM-5 zeolite. By increasing the synthesis time Y-zeolite in the resultant products should increase because the dis-
to 40 h, plenty of worm-like particles appeared on the surface of solution of Y-zeolite was slight. Moreover, we can assume that the
Y-zeolite and almost wrap the crystal. It is worth noting that the absolute quantity of nano ZSM-5 crystals in the resultant products
intensity of the characteristic peaks of Y-zeolite in the XRD pat- was constant because most of the conditions for the growth of
terns reduced very slowly after 40 h of synthesis, as shown in them, including the supplemental silicon species, alkalinity and
Fig. 6A. Bouizi et al. [37] reported that the formation of the shell templates, were fixed. Thus, an increase of Y-zeolite in the initial
can protect the core crystals from dissolution during the prepara- gel can lead to a decrease in the relative content of the nano
tion of core-shell zeolite composites. This consequence is in agree- ZSM-5 in the resultant product. That is, we can adjust the content
ment with our observations. The worm-like particles on the of shell in the prepared zeolite composites by changing the amount
surface of Y-zeolite gradually translated into spherical nano ZSM- of added Y-zeolite. It was previously reported that the core-shell
5 crystals within the time range of 72 h (see Fig. 3D and Fig. 5D). zeolite composites composed of Y-zeolite and ZSM-5 were hard
When extending the synthesis time to 110 h, the size and shape to be prepared [44]. In general, there are two common approaches
of the nano ZSM-5 crystals showed little change (Fig. 6B). More- to synthesize the core-shell zeolite composites. One involves an
over, a relatively complete Y-zeolite core can still be observed. epitaxial or heteroepitaxial growth of one zeolite on the other, such
The results further indicate that the formed shell prevents the as the preparation of chabazite/sodalite, X-A and ZSM-5/Silicalite-1

A B
(e)

(d)
Intensity (a.u.)

(c)

(b)

(a)

5 10 15 20 25 30 35
2 Theta (Degree)

C D E

Fig. 6. (A) XRD patterns of the samples with different synthesis times: (a) Y@NZ(10)-0; (b) Y@NZ(10)-20; (c) Y@NZ(10)-40; (d) Y@NZ(10)-72; (e) Y@NZ(10)-110. (B) SEM
image of Y@NZ(10)-110 after being crushed by 20 MPa. (C) SEM image of Y@NZ(10)-0. (D) SEM image of Y@NZ(10)-20. (E) SEM image of Y@NZ(10)-40.
 M. Pan et al. / Journal of Catalysis 369 (2019) 72–85 79

Fig. 7. Schematic representation of the formation process of the core-shell zeolite composites.

[45–47]. However, the approach is limited by the structural simi- have a visible change. Fig. 8A shows that the crystallinities of the Y-
larities (at least the same secondary building units) of the two zeo- zeolite phase and the ZSM-5 phase in HY@NZ(10) declined by
litic materials which are nonexistent in the cases of Y-zeolite and about 45.1 percent and 43.7 percent respectively after the
ZSM-5. Alternatively, the core-shell zeolite composites can be hydrothermal treatment. It is worth noting that the crystallinities
achieved through the preliminary adsorption of nanoseeds on the of the pure H-form commercial Y-zeolite and the pure H-form
core surface followed by their secondary growth to form a contigu- commercial ZSM-5 decreased by about 43.2 percent and 39.9 per-
ous shell layer [37,44,48]. The approach provides an advanced level cent respectively after the hydrothermal treatment under the same
of freedom in the combination of different zeolite structures. Nev- conditions, as shown in Fig. S7. The results indicate that the zeolite
ertheless, in order to successfully prepare a core-shell zeolite com- composite has a high hydrothermal stability, very close to the com-
posite by this method, keeping the chemical equilibrium at the mercial zeolites. Moreover, the crystalline meso- and macroposous
solution-core crystal interface is essential. In other words, it shell should preserve completely after the thermal or hydrother-
requires the partial overlapping of the framework compositions mal treatment. This inference is further proved by the SEM images,
and crystallization conditions between core and shell zeolites as shown in Fig. 8B to D. After the thermal or hydrothermal treat-
[37,44]. But such an overlapping is completely absent between ment, the meso- and macroposous shell composed by the nano
FAU-type Y-zeolite and MFI-type ZSM-5 [44]. The situation leads ZSM-5 crystals can also be observed on the surface of the Y-
to the rapid dissolution of the core zeolite during the secondary zeolite. And the morphology and size of the nano ZSM-5 crystals
growth. Furthermore, it causes the detachment of the seeds and show little change. The above results indicate that the prepared
the change of the chemical equilibrium at the solution-core crystal core-shell zeolite composites have a high thermal and hydrother-
interface. In consequence, it prevents from the forming of a core- mal stability. And the zeolite composites have high mechanical
shell zeolite composite composed by Y-zeolite and ZSM-5. Consid- strength bearing a pressure of 30 MPa. Moreover, the nano ZSM-
ering that the dissolution of core zeolite during the secondary 5 particles embedded/accreted in the surface of Y-zeolite do not
growth is inevitable, we employed the extracted Al and/or Si spe- escape or remove before and after thermal and hydrothermal treat-
cies resulting from the dissolution of the Y-zeolite core as the par- ments, indicating a solid connection between these two zeolites.
tial nutrients for the growth of the shell here. Through controlling Probably a chemical bond formed in the interface between core
the speed of the dissolution of the core as well as the growth of the and shell in the zeolite composites [27,46].
shell, a new dynamic equilibrium was established. Thus, a core- In order to highlight the catalytic performance of the core-shell
shell zeolite composite with Y-zeolite as the core and loosely accu- zeolite composites, two reference catalysts were prepared. One
mulational nano ZSM-5 crystals as the meso- and macroposous was a mechanical mixture of the H-form commercial Y-zeolite
shell can be obtained. This unique method has been successfully and the H-form commercial ZSM-5 with the mass ratio of 1/1
used in preparing many other zeolite composites, such as BFZ, Y- and denoted as HY+Z. The other was a mechanical mixture of the
b and ZSM-5@SAPO-34 in our previous work [33–35,39]. H-form commercial Y-zeolite and the H-form NanoZSM-5 with
the mass ratio of 1/1 and marked as HY+NZ. The characterizations
3.3. Performances of core-shell zeolite composites of the NanoZSM-5 are shown in Figs. S8–S10 and Table S2. Because
of the small particle size, the NanoZSM-5 exhibits the high meso-
Considering that the thermal and hydrothermal stability is vital pore volume and external surface area. Fig. 9 shows the SEM
for the practical application of a zeolite catalyst, this property was images of these two reference catalysts. The SEM image of HY+Z
tested on the H-form core-shell zeolite composite. Fig. 8A shows (Fig. 9A) shows a physical mixing of Y-zeolite and microsized
the XRD patterns of the parent HY@NZ(10) and the sample after ZSM-5. However, the sample HY+NZ exhibits a physical mixing of
thermal or hydrothermal treatment. Obviously, little change hap- Y-zeolite and the aggregate of nano ZSM-5 crystals, as shown in
pened in the intensities of the diffraction peaks after the thermal Fig. 9B. The average diameter of these nano ZSM-5 crystals is about
treatment at 800 °C for 20 h, suggesting that the zeolite composite 100 nm. Because of the aggregation, the abundant intercrystalline
has an extremely high thermal stability. After the hydrothermal meso- and macropores can be observed. It should be mentioned
treatment at 600 °C for 5 h, the intensities of the diffraction peaks that the ratio of Y-zeolite phase to ZSM-5 phase in the composite
80 M. Pan et al. / Journal of Catalysis 369 (2019) 72–85

A B
(c)

Intensity (a.u.)

(b)

(a)

5 10 15 20 25 30 35
2 Theta (Degree)

Fig. 8. (A) XRD patterns of the samples: (a) HY@NZ(10); (b) HY@NZ(10) after thermal treatment; (c) HY@NZ(10) after hydrothermal treatment. (B) SEM image of HY@NZ(10).
(C) SEM image of HY@NZ(10) after thermal treatment. (D) SEM image of HY@NZ(10) after hydrothermal treatment.

Fig. 9. SEM images of HY+Z (A) and HY+NZ (B).

HY@NZ(10) is very close to that in HY+Z or HY+NZ. Thus, these ture. Table 2 further shows that the sample HY+NZ possesses the
three catalysts are selected to make comparative analysis in the highest concentration of the Brönsted acid sites while the compos-
following catalytic reactions. The Si/Al ratios of HY+Z, HY+NZ and ite HY@NZ(10) has a lowest value. The concentrations of the total
HY@NZ(10) obtained by the EDS analysis are 4.7, 4.5 and 5.0 acid sites also follow the same trend, which is consistent with the
respectively. results of the NH3-TPD measurement. Moreover, Table 2 also indi-
The acidic properties of the composite HY@NZ(10) and the two cates that the concentrations of the acid sites decrease when
reference catalysts were characterized by NH3-TPD and pyridine increasing the temperature. This trend is more pronounced among
(Py)-adsorbed FTIR spectra measurements. The results are pre- the Lewis acid sites, so the values of B/L increase as the tempera-
sented in Fig. 10 and Table 2. NH3-TPD curves (Fig. 10A) indicate ture rises. When comparing the three catalysts, it is found that
that the acid amount of the sample HY+NZ is slightly more than the temperature has the biggest impact on the acid concentration
that of HY+Z. Comparing with these two reference catalysts, the of the composite HY@NZ(10), especially the concentration of the
composite HY@NZ(10) possesses a lower acid amount distinctly. Brönsted acid sites. That is, the strength of the Brönsted acid sites
In the FTIR spectra of Py-adsorbed samples (Fig. 10B to D), the in the composite HY@NZ(10) is lower than that in the reference
bands at 1450 cm
1 and 1545 cm
1 are assigned to the Lewis catalysts. NH3-TPD results (Fig. 10A and Table 2) also shows that
and Brönsted acid sites, respectively [49]. In view of the dominant the composite HY@NZ(10) mainly has the weak and medium acid-
role of the Brönsted acid sites in the catalytic cracking reactions ity. By contrast, the strong acid sites are few.
[50], the bands locating at 1545 cm
1 are carefully compared. The catalytic cracking of 1,3,5-triisopropylbenzene (TIPB) in a
The FTIR spectra show that the intensities of these bands decrease continuous flow fixed bed reactor was used as a probe reaction
with the order of HY+NZ > HY+Z > HY@NZ(10) at each tempera- to evaluate the catalytic performance of the composite HY@NZ
 M. Pan et al. / Journal of Catalysis 369 (2019) 72–85 81

HY+Z HY+Z
A B
HY+NZ HY+NZ
HY@NZ(10) HY@NZ(10)
TCD Signal (a.u.)

Intensity (a.u.)
200 300 400 500 600 1400 1450 1500 1550 1600
o -1
Temperature ( C) Wavenumber (cm )

C HY+Z D HY+Z
HY+NZ HY+NZ
HY@NZ(10) HY@NZ(10)

Intensity (a.u.)
Intensity (a.u.)

1400 1450 1500 1550 1600 1400 1450 1500 1550 1600
-1 -1
Wavenumber (cm ) Wavenumber (cm )

Fig. 10. Characterizations of the acidic properties: (A) NH3-TPD curves; (B) FTIR spectra of Py-adsorbed samples at 150 °C; (C) FTIR spectra of Py-adsorbed samples at 250 °C;
(D) FTIR spectra of Py-adsorbed samples at 350 °C.

Table 2
Concentration of the acid sites derived from FTIR spectra of Py-adsorbed samples and NH3-TPD curves.

Samples 150°Ca (lmol/g) B/L 250°Ca (lmol/g) B/L 350°Ca (lmol/g) B/L Aciditiesb (lmol/g)
Lewis Brönsted Lewis Brönsted Lewis Brönsted Weak Medium Strong Total
HY+Z 195 295 1.5 142 288 2.0 59 240 4.1 274 151 152 577
HY+NZ 198 348 1.8 138 359 2.6 57 307 5.4 302 190 142 634
HY@NZ(10) 103 249 2.4 49 246 5.0 22 163 7.4 209 149 90 448
a
Value derived from FTIR spectra of Py-adsorbed samples.
b
Value derived from NH3-TPD curves.

(10) in comparison with the reference catalysts. Fig. 11A to C meso- and macropores coming from the nano ZSM-5 crystals.
shows the changes of TIPB conversion and product selectivities Fig. 11D further shows that the conversion efficiency over
with time-on-stream (TOS). The composite HY@NZ(10) as well as HY@NZ(10) was much higher than that over HY+Z or HY+NZ at
the sample HY+NZ exhibited a higher initial activity than the sam- the initial stage of the reaction. Although the conversion efficiency
ple HY+Z although the conversion decreases gradually with the over HY@NZ(10) reduced to a level similar to that of HY+NZ after
reaction time. Considering that the acid amounts of the three cat- 280 min of reaction, it should be mentioned that the catalyst con-
alysts follow the order of HY+NZ > HY+Z > HY@NZ(10) as shown in tacts with the reactants for only a few seconds in each cycle of the
Fig. 10 and Table 2, the improved activity should not be related to practical FCC process. That is, the initial conversion efficiency has a
the acid amount. It’s known that the critical diameter of the TIPB higher reference value for the practical application. The above
molecule (0.95 nm) is larger than the pore diameters of both the results indicate that the special structure of the composite
Y-zeolite (0.74 nm) and ZSM-5 (0.54 nm) [51]. Thus, the catalytic HY@NZ(10) is beneficial to the conversion of bulky molecules.
cracking of TIPB over these catalysts must be diffusion controlled Comparing with the conversion, the product distribution is of
[52]. The increased conversion over HY+NZ or HY@NZ(10) should greater concern because it provides the information concerning
be attributed to the large external surface and the abundant the extent of cracking degree [53]. The main products of TIPB
82 M. Pan et al. / Journal of Catalysis 369 (2019) 72–85

TIPB DIPBs IPB Bz Others TIPB DIPBs IPB Bz Others

100 100 100 100
A B
80 80 80 80

Product Sel. (wt %)

Product Sel. (wt %)

TIPB Conv. (wt %)

TIPB Conv. (wt %)

60 60 60 60

40 40 40 40

20 20 20 20

0 0 0 0
40 80 120 160 200 240 280 40 80 120 160 200 240 280
Time on stream (min) Time on stream (min)

TIPB DIPBs IPB Bz Others 0.8

100 100 HY+Z
D
C HY+NZ

Conversion Efficiency (wt %/µmol)

0.7 HY@NZ(10)
80 80
Product Sel. (wt %)

0.6
TIPB Conv. (wt %)

60 60
0.5

40 40
0.4

20 20 0.3

0.2
0 0 40 80 120 160 200 240 280
40 80 120 160 200 240 280
Time on stream (min) Time on stream (min)

Fig. 11. Results of the catalytic cracking TIPB over (A) HY+Z, (B) HY+NZ and (C) HY@NZ(10) in the continuous flow fixed bed reactor and comparison of conversion efficiency
over different catalysts (D). Reaction conditions: catalyst, 0.2 g; feed rate, 1.0 mL/h; N2, 50 mL/min; temperature, 450 °C.

cracking are diisopropylbenzene isomers (DIPBs), isopropylben- (i.e. 40 wt%) over the different catalysts were obtained according to
zene (IPB), benzene (Bz) and others. Quite a lot of DIPBs were pro- the fitted curves between the conversion and the selectivity and
ducted over HY+Z as well as HY+NZ and the deep cracking products shown in Fig. 12B. The results indicate that the meso- and macro-
(IPB and Bz) were very little. Interestingly, the selectivity of IPB porous shell benefited the generation of DIPBs. This is to be
over HY@NZ(10) was much higher than that of HY+Z or HY+NZ expected because the abundant meso- and macropores in the shell
when compared at the same TOS. For example, the initial selectiv- contribute to the primary cracking of TIPB and the produced DIPBs
ity of IPB over HY@NZ(10) was up to 27.4 wt% which was five times molecules are too large to enter the micropores of nano ZSM-5
higher than that over HY+Z. After 280 min of reaction, this value of crystals to be further cracked [54]. However, the DIPBs molecules
HY@NZ(10) was still about three times higher than that of HY+Z. can enter the channels of the Y-zeolite to be cracked into smaller
The results indicate that the special core-shell structure gives the molecules. Under the current conditions, the main product of
composite HY@NZ(10) a better cracking ability than the reference DIPBs cracking over the Y-zeolite is IPB. This conclusion is further
catalysts HY+Z and HY+NZ. proved by the catalytic cracking of 1,3-diisopropylbenzene (DIPB)
To clarify the relationship between the structure and the cat- over the pure Y-zeolite under similar conditions as the TIPB crack-
alytic performance, the catalytic cracking of TIPB was conducted ing, as shown in Fig. S12. Based on the aforementioned studies, a
on the pure Y-zeolite and the zeolite composites with different possible reaction pathway of TIPB cracking over the zeolite com-
shell thickness in the continuous flow fixed bed reactor. The con- posites in the continuous flow fixed bed reactor is proposed and
tent of the Y-zeolite core was kept consistent in each reaction by illustrated in Fig. 13. The TIPB was cracked in the meso- and
varying the weight of the catalyst. For reference, the TIPB cracking macroporous shell first to produce the main products of DIPBs.
was also conducted on the pure NanoZSM-5. The changes of TIPB The special core-shell structure allowed the DIPBs molecules to
conversion and product selectivities versus time are shown in enter the micropores of the Y-zeolite core directly to be further
Fig. S11. Fig. 12A shows the effect of the meso- and macroporous cracked within the confined spaces with the strong acid centers.
shell on the TIPB conversion. It can be seen that the increase of Therefore, the core-shell composite catalyst had a better cracking
the shell promoted the conversion of TIPB effectively. This effect ability than the mixture-derived catalysts. Regarding the low selec-
became more obvious after 280 min of reaction. That is, the deac- tivity of benzene, it could be explained by the fact that the short
tivation rate decreased gradually with increasing the shell. The contact between the catalysts and the reactants under the fast flow
selectivities of IPB and DIPBs corresponding to the same conversion cannot ensure the complete cracking of the reactants.
 M. Pan et al. / Journal of Catalysis 369 (2019) 72–85 83

70 35 100
40min A IPB B
60 280min 30 DIPBs
90
50 25

DIPBs Sel. (wt %)

TIPB Conv. (wt %)

IPB Sel. (wt %)

80
40 20

30 15
70

20 10
60
10 5

0 0 50
HY HY@NZ(20) HY@NZ(10) HY@NZ(6) HY HY@NZ(20) HY@NZ(10) HY@NZ(6) HNanoZSM-5

Fig. 12. Effect of the meso- and macroporous shell on TIPB conversion (A) and product selectivities (B).

Fig. 13. Possible reaction pathway of TIPB cracking over the zeolite composites in the continuous flow fixed bed reactor.

The TIPB cracking was then conducted in a batch fixed bed more iso-paraffins and olefins while less aromatics were produced
micro-activity test unit (MAT). In this reactor, the reactants had over HY@NZ-C, compared to HY+Z-C. The results indicate that the
enough time to contact with the catalysts and thus large amounts cracking reaction and the isomerization were promoted simultane-
of Bz was produced. The results are shown in Fig. 14. HY+NZ as ously on HY@NZ-C while the aromatization was inhibited. The
well as HY@NZ(10) also exhibited a higher activity than HY+Z. improved cracking performance of HY@NZ-C is attributed to its
Moreover, the highest selectivity of Bz over HY@NZ(10) further special core-shell structure, which has been demonstrated by the
proved that the core-shell composite catalyst had a better cracking above probe reaction. More specifically, the designed core-shell
ability than the mixture-derived catalysts. To understand the real zeolite composite provides a meso- and macroporous shell to
catalytic performance of the core-shell zeolite composite in the improve the pre-cracking of the bulky molecules in heavy oil.
fluid catalytic cracking of heavy oil, a cracking catalyst derived The pre-cracked products can directly enter the micropores of
from HY@NZ(10) (denoted as HY@NZ-C) was evaluated in the the Y-zeolite core to be selectively cracked into the desired prod-
ACE-Model unit R+ provided by KTI Corporation and compared ucts with molecular precision and uniformity. The isomerisation
with a commercial catalyst prepared by using HY+Z as the main performance of HY@NZ-C should be promoted by the weaker
active component (denoted as HY+Z-C). The physicochemical prop- Brönsted acid of HY@NZ(10) compared with HY+Z as shown in
erties of the cracking catalysts are shown in Table S3. Compared Fig. 10 and Table 2. It is reported that the weak-medium acid sites
with HY+Z-C, although HY@NZ-C possesses lower textural proper- play a key role in the isomerisation [56]. Because the isomerisation
ties, it has a higher micro-activity index after deactivation. Table 3 and cracking occured simultaneously, more iso-paraffins were
and Table 4 summarize the results of the ACE evaluation. Table 3 obtained over HY@NZ-C. Moreover, based on the mechanism of
shows that the conversion of heavy oil over HY@NZ-C was higher aromatization reaction [57], the low aromatization ability of
than that over HY+Z-C. Furthermore, it should be highlighted that HY@NZ-C should be related to the small amount of Lewis acid sites
HY@NZ-C exhibited an increased gasoline yield by 8.9 wt% com- (see Fig. 10 and Table 2) which inhibited the hydrogen transfer
pared with that of HY+Z-C. It is known that the FCC gasoline is reaction. At last, it should be mentioned that the core-shell com-
the important blending component of gasoline products [55]. The posites were prepared only using the conventional raw materials,
component analysis of the gasoline products (Table 4) shows that without using any expensive secondary templates, and the yield
84 M. Pan et al. / Journal of Catalysis 369 (2019) 72–85

40
Bz IPB DIPBs Others
A 50 B

30
40
TIPB Conv. (wt %)

Product Sel. (wt %)

30
20

20

10
10

0 0
HY HY+Z HY+NZ HY@NZ(10) HY HY+Z HY+NZ HY@NZ(10)

Fig. 14. Results of the catalytic cracking TIPB in the batch fixed bed micro-activity test unit: (A) TIPB conversion; (B) distribution of the liquid products. Reaction conditions:
catalyst/reactant, 0.25 of HY and 0.5 of Others (g/g); temperature, 450 °C.

Table 3
Results of the ACE evaluation.

Catalysts Conversion (wt %) Product distribution (wt %)

Dry Gas LPG Gasoline LCO Bottoms Coke
HY+Z-C Base Base Base Base Base Base Base
HY@NZ-C +1.8
0.3
6.9 +8.9
1.0
0.8 +0.1

Table 4
Compositions of the gasoline products in the ACE evaluation (wt %).

Catalysts n-Paraffins iso-Paraffins Olefins Naphthenes Aromatics

HY+Z-C Base Base Base Base Base
HY@NZ-C
0.64 +5.21 +6.31 +0.66
11.55

was very high (e.g. 93.9% of Y@NZ(20)-72). Therefore, the produc- of VGO. The achieved materials with the special structure together
tion cost of the composite catalyst HY@NZ-C should be very close with the innovative synthesis strategy may open a door to develop
to that of the present commercial FCC catalysts. Considering the an excellent FCC catalyst for processing heavy oil.
outstanding performances, we think that the prepared core-shell
zeolite composites have great potential in practical application. Acknowledgements

4. Conclusions This work is supported by the Joint Funds of the National Nat-
ural Science Foundation of China-China Petroleum and Chemical
In this work, the core-shell zeolite composites with employing Corporation (the State Key Program Grant No. U1463209); the
the hardly dissolved commercial Y-zeolites as cores and the poly- National Natural Science Foundation of China (Grant Nos.
crystalline nano ZSM-5 aggregates as meso- and macroposous 21371129; 21376157) and Sinopec (No. 116050).
shells were successfully synthesized for the first time. The difficul-
ties of preparing such composites associated with the existing Notes
methods were overcome by using the Al and/or Si species gener-
ated from the dissolution of the Y-zeolite as the partial nutrients The authors declare no competing financial interest.
for the growth of the shell to establish a new chemical equilibrium
at the solution-core crystal interface. The relative content of the
meso- and macroposous shell in the zeolite composites can be Appendix A. Supplementary material
adjusted to meet different demands of industry by changing the
added amounts of Y-zeolite. Because of the high crystallinity, the Supplementary data to this article can be found online at
prepared core-shell zeolite composites presented a good thermal https://doi.org/10.1016/j.jcat.2018.10.032.
and hydrothermal stability. More importantly, the special core-
shell structure endowed the prepared composites with an References
enhanced ability of catalytically cracking bulky molecules. Specifi-
[1] C. Martínez, A. Corma, Inorganic molecular sieves: preparation, modification
cally, the macro- and mesoporous shell improved the pre-cracking
and industrial application in catalytic processes, Coordin. Chem. Rev. 255
of bulky molecules and the pre-cracked products could directly (2011) 1558–1580.
enter the micropores of the Y-zeolite core to be selectively cracked [2] W. Vermeiren, J.P. Gilson, Impact of zeolites on the petroleum and
into the desired products with molecular precision and uniformity. petrochemical industry, Top. Catal. 52 (2009) 1131–1161.
[3] V. Valtchev, G. Majano, S. Mintova, J. Pérez-Ramírez, Tailored crystalline
Thus, the FCC catalyst developed from the core-shell zeolite com- microporous materials by post-synthesis modification, Chem. Soc. Rev. 42
posite exhibited excellent catalytic properties in catalytic cracking (2013) 263–290.
 M. Pan et al. / Journal of Catalysis 369 (2019) 72–85 85

[4] K. Möller, T. Bein, Mesoporosity-a new dimension for zeolites, Chem. Soc. Rev. zeolite under hydrothermal treatment conditions, Ind. Eng. Chem. Res. 43
42 (2013) 3689–3707. (2004) 4066–4072.
[5] E.T.C. Vogt, B.M. Weckhuysen, Fluid catalytic cracking: recent developments [32] H. Zhang, Y. Ma, K. Song, Y. Zhang, Y. Tang, Nano-crystallite oriented self-
on the grand old lady of zeolite catalysis, Chem. Soc. Rev. 44 (2015) 7342– assembled ZSM-5 zeolite and its LDPE cracking properties: effects of
7370. accessibility and strength of acid sites, J. Catal. 302 (2013) 115–125.
[6] A. Primo, H. Garcia, Zeolites as catalysts in oil refining, Chem. Soc. Rev. 43 [33] Q. Zhao, B. Qin, J. Zheng, Y. Du, W. Sun, F. Ling, X. Zhang, R. Li, Core-shell
(2014) 7548–7561. structured zeolite-zeolite composites comprising Y zeolite cores and nano-b
[7] W. Chen, D. Han, X. Sun, C. Li, Studies on the preliminary cracking of heavy oils: zeolite shells: synthesis and application in hydrocracking of VGO oil, Chem.
contributions of various factors, Fuel 106 (2013) 498–504. Eur. J. 257 (2014) 262–272.
[8] J. Zhang, H. Shan, X. Chen, C. Li, C. Yang, Multifunctional two-stage riser [34] J. Zheng, Q. Zeng, Y. Zhang, Y. Wang, J. Ma, X. Zhang, W. Sun, R. Li, Hierarchical
catalytic cracking of heavy oil, Ind. Eng. Chem. Res. 52 (2013) 658–668. porous zeolite composite with a core-shell structure fabricated using b-zeolite
[9] A. Corma, C. Martínez, L. Sauvanaud, New materials as FCC active matrix crystals as nutrients as well as cores, Chem. Mater. 22 (2010) 6065–6074.
components for maximizing diesel (light cycle oil, LCO) and minimizing its [35] J. Zheng, G. Wang, M. Pan, D. Guo, Q. Zhao, B. Li, R. Li, Hierarchical core-shell
aromatic content, Catal. Today 127 (2007) 3–16. zeolite composite ZSM-5@SAPO-34 fabricated by using ZSM-5 as the nutrients
[10] U. Navarro, C.A. Trujillo, A. Oviedo, R. Lobo, Impact of deactivation conditions for the growth of SAPO-34, Microporous Mesoporous Mater. 206 (2015) 114–
on the acidity of Y zeolites used in the formulation of FCC catalysts, studied by 120.
FTIR of adsorbed CO, J. Catal. 211 (2002) 64–74. [36] M. Thommes, K. Kaneko, A.V. Neimark, J.P. Olivier, F. Rodriguez-Reinoso, J.
[11] B. Wang, C. Han, Q. Zhang, C. Li, C. Yang, H. Shan, Studies on the preliminary Rouquerol, K.S.W. Sing, Physisorption of gases, with special reference to the
cracking of heavy oils: the effect of matrix acidity and a proposal of a new evaluation of surface area and pore size distribution (IUPAC Technical Report),
reaction route, Energy Fuels 29 (2015) 5701–5713. Pure Appl. Chem. 87 (2015) 1051–1069.
[12] Q. Meng, B. Liu, J. Piao, Q. Liu, Synthesis of the composite material Y/ASA and [37] Y. Bouizi, I. Diaz, L. Rouleau, V.P. Valtchev, Core-shell zeolite microcomposites,
its catalytic performance for the cracking of n-decane, J. Catal. 290 (2012) 55– Adv. Funct. Mater. 15 (2005) 1955–1960.
64. [38] J.C. Groen, W. Zhu, S. Brouwer, S.J. Huynink, F. Kapteijn, J.A. Moulijn, J. Pérez-
[13] A.C.F. Coriolano, C.G.C. Silva, M.J.F. Costa, S.B.C. Pergher, V.P.S. Caldeira, A.S. Ramírez, Direct demonstration of enhanced diffusion in mesoporous ZSM-5
Araujo, Development of HZSM-5/AlMCM-41 hybrid micro-mesoporous zeolite obtained via controlled desilication, J. Am. Chem. Soc. 129 (2007) 355–
material and application for pyrolysis of vacuum gasoil, Microporous 360.
Mesoporous Mater. 172 (2013) 206–212. [39] J. Zheng, Y. Yi, W. Wang, K. Guo, J. Ma, R. Li, Synthesis of bi-phases composite
[14] L. Jia, X. Sun, X. Ye, C. Zou, H. Gu, Y. Huang, G. Niu, D. Zhao, Core-shell zeolites MFZ and its hierarchical effects in isopropylbenzene catalytic
composites of USY@Mesosilica: synthesis and application in cracking heavy cracking, Microporous Mesoporous Mater. 171 (2013) 44–52.
molecules with high liquid yield, Microporous Mesoporous Mater. 176 (2013) [40] N. Ren, B. Subotić, J. Bronić, Y. Tang, M.D. Sikirić, T. Mišić, V. Svetličić, S. Bosnar,
16–24. T.A. Jelić, Unusual pathway of crystallization of zeolite ZSM-5 in a
[15] Y. Lv, X. Qian, B. Tu, D. Zhao, Generalized synthesis of core-shell structured heterogeneous system: phenomenology and starting considerations, Chem.
nano-zeolite@ordered mesoporous silica composites, Catal. Today 204 (2013) Mater. 24 (2012) 1726–1737.
2–7. [41] D. Zhai, L. Zhao, Y. Liu, J. Xu, B. Shen, J. Gao, Dissolution and absorption: a
[16] Y. Wang, D. Cui, Q. Li, Synthesis, characterization and influence parameters on molecular mechanism of mesopore formation in alkaline treatment of zeolite,
the overgrowth of micro/mesoporous Y-zeolite-MCM-41 composite material Chem. Mater. 27 (2014) 67–74.
under acidic conditions, Microporous Mesoporous Mater. 142 (2011) 503–510. [42] G.O. Brunner, A proposal for a mechanism of nucleation in zeolite synthesis,
[17] T. Jiang, L. Qi, M. Ji, H. Ding, Y. Li, Z. Tao, Q. Zhao, Characterization of Y/MCM-41 Zeolites 12 (1992) 428–430.
composite molecular sieve with high stability from Kaolin and its catalytic [43] R.L. Penn, J.F. Banfield, Imperfect oriented attachment: dislocation generation
property, Appl. Clay Sci. 62–63 (2012) 32–40. in defect-free nanocrystals, Science 281 (1998) 969–971.
[18] M.S. Hamdy, G. Mul, TUD-1-encapsulated HY zeolite: a new hierarchical [44] Y. Bouizi, L. Rouleau, V.P. Valtchev, Factors controlling the formation of core-
microporous/mesoporous composite with extraordinary performance in shell zeolite-zeolite composites, Chem. Mater. 18 (2006) 4959–4966.
benzylation reactions, ChemCatChem 5 (2013) 3156–3163. [45] T. Wakihara, S. Yamakita, K. Iezumi, T. Okubo, Heteroepitaxial growth of a
[19] P. Waller, Z. Shan, L. Marchese, G. Tartaglione, W. Zhou, J.C. Jansen, T. zeolite film with a patterned surface-texture, J. Am. Chem. Soc. 125 (2003)
Maschmeyer, Zeolite nanocrystals inside mesoporous TUD-1: a high- 12388–12389.
performance catalytic composite, Chem. Eur. J. 10 (2004) 4970–4976. [46] Z. Wang, W. Wan, J. Sun, W. Carrillo-Cabrera, D. Grüner, X. Yin, S. Qiu, G. Zhu,
[20] A. Ishihara, K. Kimura, A. Owaki, K. Inui, T. Hashimoto, H. Nasu, Catalytic X. Zou, Epitaxial growth of core-shell zeolite X-A composites, CrystEngComm
cracking of VGO by hierarchical ZSM-5 zeolite containing mesoporous silica- 14 (2012) 2204–2212.
aluminas using a curie point pyrolyzer, Catal. Commun. 28 (2012) 163–167. [47] A. Ghorbanpour, A. Gumidyala, L.C. Grabow, S.P. Crossley, J.D. Rimer, Epitaxial
[21] W. Jiao, X. Wu, G. Li, T. Xue, Y. Wang, Y. Tang, Core-shell zeolite Y@c-Al2O3 growth of ZSM-5@Silicalite-1: a core-shell zeolite designed with passivated
nanorod composites: optimized fluid catalytic cracking catalyst assembly for surface acidity, ACS Nano 9 (2015) 4006–4016.
processing heavy oil, ChemCatChem 9 (2017) 2574–2583. [48] Y. Bouizi, L. Rouleau, V.P. Valtchev, Bi-phase MOR/MFI-type zeolite core-shell
[22] A. Corma, J. Martı́nez-Triguero, C. Martı́nez, The use of ITQ-7 as a FCC zeolitic composite, Microporous Mesoporous Mater. 91 (2006) 70–77.
additive, J. Catal. 197 (2001) 151–159. [49] Y. Wang, J. Ma, F. Ren, J. Du, R. Li, Hierarchical architectures of ZSM-5
[23] M. Falco, J. Retuert, A. Hidrobo, C. Covarrubias, P. Araya, U. Sedran, Catalytic nanocrystalline aggregates with particular catalysis for lager molecule
performance of silica-aluminas synthesised with the help of chitosan reaction, Microporous Mesoporous Mater. 240 (2017) 22–30.
biopolymer, Appl. Catal. A: Gen. 366 (2009) 269–274. [50] P. Bai, M. Xie, U.J. Etim, W. Xing, P. Wu, Y. Zhang, B. Liu, Y. Wang, K. Qiao, Z.
[24] K. Pinkaew, G. Yang, T. Vitidsant, Y. Jin, C. Zeng, Y. Yoneyama, N. Tsubaki, A Yan, Zeolite Y mother liquor modified c-Al2O3 with enhanced Brönsted acidity
new core-shell-like capsule catalyst with SAPO-46 zeolite shell encapsulated as active matrix to improve the performance of fluid catalytic cracking
Cr/ZnO for the controlled tandem synthesis of dimethyl ether from syngas, catalyst, Ind. Eng. Chem. Res. 57 (2018) 1389–1398.
Fuel 111 (2013) 727–732. [51] Z. Qin, B. Shen, X. Gao, F. Lin, B. Wang, C. Xu, Mesoporous Y zeolite with
[25] G. Yang, C. Xing, W. Hirohama, Y. Jin, C. Zeng, Y. Suehiro, T. Wang, Y. homogeneous aluminum distribution obtained by sequential desilication-
Yoneyama, N. Tsubaki, Tandem catalytic synthesis of light isoparaffin from dealumination and its performance in the catalytic cracking of cumene and
syngas via Fischer-Tropsch synthesis by newly developed core-shell-like 1,3,5-triisopropylbenzene, J. Catal. 278 (2011) 266–275.
zeolite capsule catalysts, Catal. Today 215 (2013) 29–35. [52] J. Ding, M. Wang, L. Peng, N. Xue, Y. Wang, M.Y. He, Combined desilication and
[26] J. Zečević, G. Vanbutsele, K.P. de Jong, J.A. Martens, Nanoscale intimacy in phosphorus modification for high-silica ZSM-5 zeolite with related study of
bifunctional catalysts for selective conversion of hydrocarbons, Nature 528 hydrocarbon cracking performance, Appl. Catal. A: Gen. 503 (2015) 147–155.
(2015) 245–248. [53] L. Xu, X. Ji, S. Li, Z. Zhou, X. Du, J. Sun, F. Deng, S. Che, P. Wu, Self-assembly of
[27] X. Zhang, Q. Guo, B. Qin, Z. Zhang, F. Ling, W. Sun, R. Li, Structural features of cetyltrimethylammonium bromide and lamellar zeolite precursor for the
binary microporous zeolite composite Y-Beta and its hydrocracking preparation of hierarchical MWW zeolite, Chem. Mater. 28 (2016) 4512–4521.
performance, Catal. Today 149 (2010) 212–217. [54] K.A. Mahgoub, S. Al-Khattaf, Catalytic cracking of hydrocarbons in a riser
[28] R. Feng, S. Liu, P. Bai, K. Qiao, Y. Wang, H.A. Al-Megren, M.J. Rood, Z. Yan, simulator: the effect of catalyst accessibility and acidity, Energy Fuels 19
Preparation and characterization of c-Al2O3 with rich Brönsted acid sites and (2005) 329–338.
its application in the fluid catalytic cracking process, J. Phys. Chem. C 118 [55] D. Gao, A. Duan, X. Zhang, Z. Zhao, H.E.Y. Qin, C. Xu, Synthesis of CoMo
(2014) 6226–6234. catalysts supported on EMT/FAU intergrowth zeolites with different
[29] R.J. White, A. Fischer, C. Goebel, A. Thomas, A sustainable template for morphologies and their hydro-upgrading performances for FCC gasoline,
mesoporous zeolite synthesis, J. Am. Chem. Soc. 136 (2014) 2715–2718. Chem. Eur. J. 270 (2015) 176–186.
[30] A. Petushkov, S. Yoon, S.C. Larsen, Synthesis of hierarchical nanocrystalline [56] Y. Ono, A survey of the mechanism in catalytic isomerization of alkanes, Catal.
ZSM-5 with controlled particle size and mesoporosity, Microporous Today 81 (2003) 3–16.
Mesoporous Mater. 137 (2011) 92–100. [57] X. Su, W. Zan, X. Bai, G. Wang, W. Wu, Synthesis of microscale and nanoscale
[31] S. Sombatchaisak, P. Praserthdam, C. Chaisuk, J. Panpranot, An alternative ZSM-5 zeolites: effect of particle size and acidity of Zn modified ZSM-5 zeolites
correlation equation between particle size and structure stability of H-Y on aromatization performance, Catal. Sci. Technol. 7 (2017) 1943–1952.