A self-supported SS-fiber@meso-HZSM-5 core-shell catalyst: Via caramel-assistant synthesis toward prolonged lifetime for the methanol-to-propylene reaction

Article abstract of DOI:10.1039/c6ra08944e

A self-supported SS-fiber@meso-HZSM-5 core-shell catalyst was essentially designed and engineered from micro- to macro-scale by caramel-assistant hydrothermal synthesis. The significant role of caramel during the crystallization process was revealed in detail. Caramel not only created the mesoporosity in the ZSM-5 crystals, but also released acid under hydrothermal synthesis conditions which lowered the zeolite crystallinity. By taking advantage of the mesopore development in a hierarchical micro-meso-macropore structure with favourably-tuned acidic properties, such a catalyst provided a dramatically prolonged lifetime of 845 h (>90% conv.) with high propylene selectivity (e.g., 48%) in the MTP reaction. The hierarchical pore structure development mainly increased the accommodation capacity of the zeolite shell for receiving formed coke thereby leading to a dramatically prolonged lifetime in the MTP reaction.

Full text of DOI:10.1039/c6ra08944e

RSC Advances
View Article Online
View Journal | View Issue
PAPER
A self-supported SS-fiber@meso-HZSM-5 core–
shell catalyst via caramel-assistant synthesis
toward prolonged lifetime for the methanol-to-
propylene reaction†
Cite this: RSC Adv., 2016, 6, 48387
Jia Ding, Zhiqiang Zhang, Lupeng Han, Chunzheng Wang, Pengjing Chen,
*
*
Guofeng Zhao, Ye Liu and Yong Lu
A self-supported SS-fiber@meso-HZSM-5 core–shell catalyst was essentially designed and engineered
from micro- to macro-scale by caramel-assistant hydrothermal synthesis. The significant role of caramel
during the crystallization process was revealed in detail. Caramel not only created the mesoporosity in
the ZSM-5 crystals, but also released acid under hydrothermal synthesis conditions which lowered the
zeolite crystallinity. By taking advantage of the mesopore development in a hierarchical micro–meso–
macropore structure with favourably-tuned acidic properties, such a catalyst provided a dramatically
prolonged lifetime of 845 h (>90% conv.) with high propylene selectivity (e.g., 48%) in the MTP reaction.
The hierarchical pore structure development mainly increased the accommodation capacity of the
zeolite shell for receiving formed coke thereby leading to a dramatically prolonged lifetime in the MTP
reaction.
Received 7th April 2016
Accepted 10th May 2016
DOI: 10.1039/c6ra08944e
www.rsc.org/advances
effective since the secondary mesopore network can shorten
the diffusion path length.12 One widely used top-down method
1. Introduction
to create intracrystalline mesopores is desilication, involving
controllably leaching silicon in an alkaline medium (e.g., NaOH,
quaternary ammonium hydroxides, Na2CO3, or mixtures of
them). Recently, Milina et al. reported a series of desilicated
ZSM-5 catalysts with hierarchical pores, and specially correlated
the methanol-to-hydrocarbons lifetime of the zeolites with the
global pore connectivity.13 Dual-templating with surfactants,
block copolymers, organosilane agents or hard templates is
another bottom-up method to obtain secondary porosity.14–19
Zhang et al. synthesized a mesopore-containing ZSM-5 catalyst
by using spent catalyst in which the deposited coke could act as
a hard template, and the as-synthesized catalyst exhibited
a 3-fold increase in lifetime for the MTP reaction.20 Besides the
modication of ZSM-5, microstructured engineering is also an
efficient strategy to improve the catalyst performance.21–26 For
example, the monolithic metal-ber@HZSM-5 core–shell cata-
lyst possessed enhanced heat/mass transfer and delivered
4-fold longer lifetime than the common particulate HZSM-5 in
the MTP process.21–23 However, further prolonged lifetime of the
metal-ber@HZSM-5 catalyst is particularly desirable to meet
the requirement of commercial run (500–600 h, Lurgi's MTP
process).27 To accomplish this goal, the hierarchical porous
structure design and tailoring of such microbrous-engineered
ZSM-5 core–shell catalyst is a promising strategy.
With the ever-increasing demand for propylene and the deple-
tion of petroleum resources,1 the methanol-to-propylene (MTP)
reaction catalyzed by ZSM-5 zeolite catalysts has attracted
extensive attention, since methanol can be readily generated
from coal, natural gas, and renewable biomass.2 As typical
crystalline microporous sieves, ZSM-5 possesses a large surface
area which provides enough active sites for acid catalysis.3
However, the regular pores in molecular dimensions also seri-
ously affect the stability of zeolite catalysts. In the case of MTP,
coke (bulky molecules such as polycyclic aromatics, high
molecular weight polymeric carbon, amorphous carbon, etc.)
will accumulate inside and even block the microporous chan-
nels, severely weakening the diffusion capacity, (i.e., the rate of
reactants accessing or products leaving from the active sites),
and nally, a rapid deactivation is caused.4
Toward the prolonged lifetime by signicantly retarding
coke deposition, numerous efforts have been devoted to modify
the catalyst, including acidity adjustment, size- or morphology-
controllable synthesis and hierarchical-tailoring of the pore
structure.5–11 Particularly, the hierarchical-tailoring is highly-
Shanghai Key Laboratory of Green Chemistry and Chemical Processes, School of
Chemistry and Molecular Engineering, East China Normal University, Shanghai
200062, China. E-mail: gfzhao@chem.ecnu.edu.cn; ylu@chem.ecnu.edu.cn
† Electronic supplementary information (ESI) available: Experimental details;
characterizations of the prepared catalysts and test results over the catalysts.
See DOI: 10.1039/c6ra08944e
As previously noted, biomass-based templates were utilized
to create mesopores for cost-effective hierarchical zeolite
This journal is © The Royal Society of Chemistry 2016
RSC Adv., 2016, 6, 48387–48395 | 48387

View Article Online
RSC Advances
Paper
syanthesis.5,28–32 Most notably, as a partially carbonized carbo-
Ammonia solution (25 wt%) of 2 g was dissolved in 20 g
hydrate, caramel obtained by simple hydrothermal pre- water, followed by adding 20 g glucose under stirring. The
treatment of glucose, can be regarded as a mesopore template mixture was then transferred into a stainless-steel Teon-lined
in the transitional state between so mesopore template and autoclave and heated at 150 ꢀC for periods ranging from 0 to 4 h.
hard mesopore template. This inspires us to try this affordable, Aer diluted by water, the brown caramel solution (10 wt%) was
eco-friendly and easily-obtained caramel as
a mesopore obtained (Fig. S1†).
template for the synthesis of ZSM-5 zeolite with developed
intracrystalline mesopores. Herein, we attempt to render self-
supported meso-HZSM-5 catalysts essentially designed and
2.3. Synthesis of SS-ber@meso-HZSM-5
engineered from micro- to macro-scale by direct growth of Circular chips (dia. 16.1 mm) of microbrous substrate were
meso-ZSM-5 zeolite onto a thin-sheet sinter-locked 20 mm seeded by silicalite-1 using dip-coating method. To remove the
stainless steel bers (SS-ber) matrix (15 vol% SS-ber, 85 vol% contamination, the microbrous substrate was sonicated in
void volume) with the aid of caramel-assistant hydrothermal acetone for 30 min, thoroughly washed using water, and then
synthesis. Indeed, caramel not only hierarchically-tailored the dried at 100 ꢀC in air. The dip-coating process used 2 wt%
ZSM-5 shell with micro–meso–macropore system, but also silicalite-1 (50–100 nm) suspension and its pH was adjusted to
spontaneously tuned its acidic properties. Interestingly, as- 10 using 25 wt% NH3$H2O prior to dip coating. Seeding of
prepared SS-ber@meso-HZSM-5 catalyst demonstrated silicalite-1 was performed by dipping the microbrous
dramatically prolonged lifetime in the MTP reaction process, substrate into the above suspension for 15 min with ultrasonic
ꢀ
mainly due to both the shortened micropore diffusion path processing, followed by drying at 100 C for 10 h.
length as a result of a secondary mesopore network and the
favorably-tuned acidic properties.
A certain amount of NaAlO2, TPAOH and NaOH were dis-
solved in deionized water under stirring at room temperature,
followed by TEOS adding whilst stirring for 1 h to obtain a clear
sol. Subsequently, the caramel solution was dropwise added
into the above sol under vigorous stirring. The molar ratio of
TEOS/TPAOH/NaOH/NaAlO2/H2O in such sol was 1/0.25/0.18/
0.01/250, and caramel/SiO2 mass ratio was tuned to be 0.6,
1.2, 1.8 or 2.4, respectively. Finally, 70 mL sol and 10 circular
chips of microbrous substrate (silicalite-1 seeded) were sealed
into the Teon-lined stainless steel autoclave, and statically
crystallized at 180 ꢀC for 48 h. The as-synthesized microbrous-
structured SS-ber@meso-ZSM-5 composites were rinsed thor-
oughly with water, dried in oven at 100 ꢀC overnight, and
calcined in air at 550 ꢀC for 5 h with a heating rate of 2 ꢀC minꢁ1
to remove the organic templates. The as-calcined SS-
ber@meso-ZSM-5 samples were immersed into an aqueous
solution of 1 M NH4Cl at 80 ꢀC for 1 h, dried at 100 ꢀC and
calcined at 550 ꢀC for 5 h to obtain the SS-ber@meso-HZSM-5
product catalysts denoted as SS-ber@meso-HZSM-5(x h, y),
where “x” and “y” represent caramelization time (h) and
caramel/SiO2 mass ratio, respectively.
2. Experimental
2.1. Materials
A sinter-locked microbrous structure consisting of 15 vol%
316-L stainless steel bers (SS-ber, 20 mm in dia.) and 85 vol%
voidage (Western Metal Material Co., Ltd.) was employed as
monolithic substrate for directly growing ZSM-5 crystals on it.
The following raw materials including tetraethylorthosilicate
(TEOS, A.R., Sinopharm Chemical Reagent Co., Ltd.), sodium
aluminate (NaAlO2, C.P., Sinopharm Chemical Reagent Co.,
Ltd.), tetrapropylammonium hydroxide (TPAOH, 25 wt%,
Sinopec Co., Ltd.) and sodium hydroxide (NaOH, A.R., Sino-
pharm Chemical Reagent Co., Ltd.) were used as silica,
aluminum, template and hydroxide sources for synthesizing of
silicalite-1 seeds and ZSM-5. Ammonium chloride (NH4Cl, A.R.,
Sinopharm Chemical Reagent Co., Ltd.) was used for the
aggregation of silicalite-1 gel and ion exchange of ZSM-5.
Ammonium hydroxide (NH3$H2O, A.R., Sinopharm Chemical
Reagent Co., Ltd.) was used to adjust the pH of seeding
suspension. Caramel solution was prepared with D(+)-glucose
(C6H12O6, A.R., Sinopharm Chemical Reagent Co., Ltd.) and
ammonium hydroxide.
2.4. Characterizations
Zeolite coatings grown on the microbrous substrate were
characterized by inductively coupled plasma-atomic emission
spectroscopy (ICP-AES; ICP Thermo IRIS Intrepid II XSP; USA),
N2 adsorption–desorption (BET-MAX; Japan), X-ray diffraction
2.2. Preparation of silicalite-1 seeds and caramel solution
Silicalite-1 (MFI-type) seeds were prepared according to the re- (XRD; Rigaku Ultima IV, Cu K; Japan), scanning electron
ported method.33 TPAOH was mixed with water, and then TEOS microscopy (SEM; Hitachi S-4800; Japan), transmission electron
was dropped into the above solution under vigorously stirring microscopy (TEM; Tecnai G2 F30; USA), temperature-
until obtaining a clear precursor sol with a molar composition programmed desorption of ammonia (NH3-TPD; Xianquan TP-
of SiO2/0.35TPAOH/19.6H2O. Aer homogenizing at room 5080; China), Fourier transform infrared spectroscopy as well
temperature for 24 h, the mixture was hydrothermally treated at as infrared spectroscopy of pyridine adsorption (FT-IR and Py-
90 ꢀC for 96 h. Aer cooling, NH4Cl was added into the obtained IR; Nicolet Nexus 670; USA), and thermogravimetric analysis
silicalite-1 gel under stirring. Silicalite-1 seeds were aggregated (TGA; NETZSCH STA 449 F3 Jupiter; Germany). The mass of
and separated by ltration. Before use, silicalite-1 seeds were zeolite was estimated by the weight change of microbrous
dried at 100 ꢀC for 24 h and ground into powder.
substrates before and aer hydrothermal synthesis.
48388 | RSC Adv., 2016, 6, 48387–48395
This journal is © The Royal Society of Chemistry 2016

View Article Online
Paper
RSC Advances
2.5. Catalytic tests in the MTP reaction
As-obtained SS-ber@meso-HZSM-5 core–shell catalysts were
tested in the MTP reaction as described in detail elsewhere.22
The test was performed using a xed-bed quartz tube reactor
ꢀ
(i.d., 16 mm; reactor length of 300 mm) at 450 C and weight
hourly space velocity (WHSV) of 10 (or 1) hꢁ1 under 0.1 MPa.
Circular chips (16.1 mm diameter) of the SS-ber@meso-HZSM-
5 catalysts were packed layer-up-layer into the tube reactor with
HZSM-5 mass of 0.4 g. Note that the diameter of 0.1 mm larger
than the i.d. of the tubular reactor was retained deliberately to
avoid the appearance of gap between the reactor wall and the
edges of catalyst chips thereby preventing the gas bypassing.
Methanol (A.R., 99.5%, Sinopharm Chemical Reagent Co., Ltd.;
directly used without treatment) was fed continuously at
desired rate using a high performance liquid chromatography
(HPLC) pump, followed by vaporization at 200 ꢀC in highly
puried N2 ow. The gas effluent from evaporator was 30 vol%
gaseous methanol in N2. The effluent gas at outlet of the reactor
was quantitatively analyzed by an on-line Shimadzu 2014 gas
Fig. 1 (A) SEM images of the sinter-locked SS-fiber before (A) and after
chromatography-ame ionization detector (GC-FID) with a 30 m (B) the seeding of silicalite-1, showing an uniform seeded substrate
surface with irregular crystallite 50–100 nm in size; (C) photograph of
the macroscopic SS-fiber@meso-HZSM-5; (D and E) SEM images of
SS-fiber@meso-HZSM-5 catalyst, showing uniform and dense growth
of zeolite crystals along with fibers to form a core–shell structure.
HP-Plot Q capillary column (Agilent). Methanol conversion was
calculated on the basis of reactor inlet and outlet concentra-
tions. Selectivity to hydrocarbons was dened as the molar ratio
of carbon in the hydrocarbon formed versus the methanol
converted.
loading of 18.7 wt% was achieved aer only two synthesis cycles
with a ZSM-5 shell thickness of 4–5 mm (Fig. 1E); in the absence
of caramel, however, more than three cycles were needed to load
3. Results and discussion
3.1. Microstructure and features of hierarchical SS-
ber@meso-HZSM-5
comparative zeolite content on SS-ber, indicating that adding
a certain amount of caramel facilitated the ZSM-5 shell growth.
Moreover, aer ultrasonic treatment in petroleum ether for 1 h,
the weight loss of SS-ber@meso-HZSM-5 catalysts was negli-
gible, indicating good mechanical strength of the core–shell
structure. Clearly, SS-ber@meso-HZSM-5(3 h, 1.2) contained
typical MFI structure, being evidenced by XRD pattern with
major diffraction peaks located at 2q of 7.9ꢀ and 8.9ꢀ and the
characteristic quintet at 23–25ꢀ (Fig. 2A; JCPDS card no. 00-044-
0002). The SS-ber@meso-HZSM-5(3 h, 1.2) presented a specic
surface area (SSA) of 73 m2 gꢁ1 and a total pore volume of
SEM images of the SS-ber before and aer seeding with
silicalite-1 are shown in Fig. 1. Clearly, the SS-ber matrix
possessed 3D sinter-locked microbrous network with entirely
open but irregular submillimeter-scaled macropores, which was
favorable for the ZSM-5 growth due to the big and highly
accessible void space (Fig. 1A). Aer simply sonicating in
silicalite-1 suspension, seeds (MFI-type, 50–100 nm) were
homogeneously and rmly anchored on the surface of the one-
dimension rodlike SS-ber (Fig. 1B), which assured the identity
of crystallization condition for the perfect core–shell structure
formation. Moreover, such ultrasonic-assisted seeding method
could be easily implemented in large scale. Subsequently, self-
supported SS-ber@meso-HZSM-5 catalysts, with hierarchical
micro–meso–macropore structure, were prepared via an in situ
hydrothermal synthesis using the above seeded SS-ber
substrates, and specically, caramel was added into the
synthesis sol as mesopores template. Fig. 1C–E show the
geometry, morphology, and structural/textural properties of our
representative catalyst prepared using a synthesis sol with
caramel/SiO2 mass ratio of 1.2 (caramelized for 3 h), and the
catalyst is denoted as SS-ber@meso-HZSM-5(3 h, 1.2). In
comparison with the pristine SS-microbrous structure
(Fig. 1A), it was clear that a uniform and dense ZSM-5 zeolite
shell grew continuously along with the SS-ber to form hierar-
chical 3D porous ber@zeolite core–shell structure (Fig. 1D and
Fig. 2 Phase structure and texture of SS-fiber@meso-HZSM-5 cata-
lyst. (A) XRD pattern, showing the formation of MFI-type zeolite shell;
(B) N2 adsorption–desorption isotherms and BJH mesopore size
distribution (inset), indicating the hierarchical porous structure of the
E). Notably, with the presence of caramel, a high ZSM-5 zeolite HZSM-5 zeolite shell.
This journal is © The Royal Society of Chemistry 2016
RSC Adv., 2016, 6, 48387–48395 | 48389

View Article Online
RSC Advances
Paper
Fig.
3 TEM images of SS-fiber@meso-HZSM-5 catalyst, directly
revealing the significant mesoporosity development.
Fig. 4 (A) XRD patterns of SS-fiber@meso-HZSM-5 synthesized by
using caramel with different pre-caramelization time lengthes; (B) XRD
patterns of SS-fiber@meso-HZSM-5 synthesized with different
caramel/SiO2 mass ratio; (a) SS-fiber@meso-HZSM-5(0 h, 0.6), (b) SS-
fiber@meso-HZSM-5(1 h, 0.6), (c) SS-fiber@meso-HZSM-5(2 h, 0.6),
(d) SS-fiber@meso-HZSM-5(3 h, 0.6), (e) SS-fiber@meso-HZSM-5(3 h,
1.2), (f) SS-fiber@meso-HZSM-5(3 h, 1.8), (g) SS-fiber@meso-HZSM-
5(3 h, 2.4).
0.048 cm3 gꢁ1 (Fig. 2B; including the SS-ber mass), mainly
attributed to a micropore SSA of 358 m2 gꢁ1 with a micopore
volume of 0.17 cm3 gꢁ1 in zeolite shell, which was comparable
with those (350 m2 gꢁ1 and 0.16 cm3 gꢁ1) of pure HZSM-5.
Fig. S2† shows the TGA proles of zeolite shell in the as-
synthesized catalysts without calcination. The weight loss of
SS-ber@HZSM-5 (synthesized without caramel) was ꢂ13.1%,
assigned to the decomposition of TPA+, which was ꢂ11.6%
lower than sample SS-ber@meso-HZSM-5(3 h, 1.8) due to the
decomposition of caramel, indicating that caramel was retained
in the zeolite shell aer synthesis. In consequence, ZSM-5 shell
showed a mesoporous feature with external SSA of 32 m2 gꢁ1
and volume of 0.09 cm3 gꢁ1 as well as a Barrett–Joyner–Halenda
(BJH) pore size distribution at ꢂ25 nm aer calcination in air;
more interestingly, some macropores were also created with
a size distribution centered at ꢂ94 nm likely due to the
nonuniformity of caramel molecules (Fig. 2B). The TEM images
(Fig. 3) of ZSM-5 crystals directly revealed the signicant mes-
oporosity development while small grains with clear lattice
fringes were clearly seen. These results indicated the hierar-
chical porous structure of the SS-ber@meso-HZSM-5 catalyst
obtainable using caramel as mesopore template.
Effect of caramel/SiO2 mass ratio on the zeolite crystallinity
of SS-ber@meso-HZSM-5 was investigated in the range of 0.6
to 2.4, with the results as shown in Fig. 4B and S4.† Note that the
caramel template aer pre-caramelizing for 3 h was utilized and
zeolite content in those microbrous-structured catalysts was
controlled to ꢂ18% to avoid the effect of shell thickness. Fig. 4B
shows the XRD patterns of SS-ber@meso-HZSM-5 synthesized
with different caramel/SiO2 mass ratios. Specially, the
composite samples were crashed to destroy the robust core–
shell structure to obtain the powdered ZSM-5 to avoid any effect
on calculation of the relative crystallinity, such as zeolite
content and slight nonuniformity of shell. Clearly, the crystal-
linity was trivially affected by the caramel/SiO2 mass ratio varied
from 0.6 to 1.8 but clearly weakened at 2.4 (Fig. S4†), which was
mainly due to the pH value reduction of the synthesis system
under the crystallization conditions. By combining above
information with the fact that the pH values of the original
synthesis sols kept almost constant (Fig. S3B†), it was suggested
that caramel could further caramelize under crystallization
conditions to release acid thereby lowering the synthesis system
pH value. Not surprisingly, when the caramel/SiO2 mass ratio
was increased up to 2.4, the micropore volume of zeolite shell
decreased to 0.10 cm3 gꢁ1 with a low micropore SSA of only 207
m2 gꢁ1, indicating the poorer shell crystallinity (Fig. 5A and
Table 1). This observation was consistent with the diffraction
intensity decline in XRD patterns (Fig. 4B) and the crystal
morphology evolution from crystallized to amorphous (Fig. 6).
It should be noticed that the caramel template also provided
ability to tune the acidic properties of HZSM-5 shell. Fig. 7
shows the NH3-TPD proles of SS-ber@meso-HZSM-5 core–
shell catalysts with various caramel/SiO2 mass ratios. All
samples offered dual-peak NH3-TPD proles, indicating the
existence of two types of acid sites. The low-temperature
3.2. Effects of pre-caramelization time and caramel/SiO2
ratio
Specically, the effect of pre-caramelization time on the struc-
tural and textural properties of the SS-ber@meso-HZSM-5
catalysts was investigated. As shown in the XRD patterns in
Fig. 4A, all samples were characterized of typical MFI structure.
Fig. S3A† shows that pH values of caramel solution decreased
with the increase of pre-caramelization time, indicating that
a longer time of 3 h would release more acid beforehand, and
thus the caramel/SiO2 mass ratio could be increased to 1.8
(mixture without severe solidication during crystallization)
while the ratio was only 0.6 if the pre-caramelization time
reduced to 2 h or less. Moreover, increasing pre-caramelization
time to 4 h or more was unsuitable for SS-ber@meso-HZSM-5
synthesis since highly carbonized caramel was detrimental to
form homogeneous precursor. Notably, the mesopore SSA and
volume increased along with the pre-caramelization time (Table
S1†), indicating the appropriate pre-caramelization time was
crucial to coordinate caramelization of caramel template with
zeolite crystallization.
ꢀ
desorption peak centered at 190–220 C was attributed to the
interaction of NH3 with the weak acid sites in the HZSM-5 shell,
and the high-temperature desorption peaking at 400–450 ꢀC to
48390 | RSC Adv., 2016, 6, 48387–48395
This journal is © The Royal Society of Chemistry 2016

View Article Online
Paper
RSC Advances
Fig. 5 (A) N2 adsorption–desorption isotherms and (B) BJH mesopore
size distributions of (a) SS-fiber@meso-HZSM-5(3 h, 0.6), (b) SS-
fiber@meso-HZSM-5(3 h, 1.2), (c) SS-fiber@meso-HZSM-5(3 h, 1.8)
and (d) SS-fiber@meso-HZSM-5(3 h, 2.4) catalysts with various
caramel/SiO2 mass ratios.
strong acid sites. As shown in Table 1, when caramel/SiO2 mass
ratio was increased from 0.6 to 1.2, strong acid amount slightly
decreased from 0.10 to 0.09 mmol gꢁ1. Further increasing the
caramel/SiO2 mass ratio to 1.8, strong acid amount
pronouncedly decreased to 0.05 mmol gꢁ1, accompanied by
a severe loss of total acid amount from 0.23 to 0.13 mmol gꢁ1
.
The nature of acidity were further characterized by using IR
pyridine adsorption experiments (Fig. 7B and Table 1), in which
the band at 1455 cmꢁ1 was assigned to Lewis acid sites, 1490
cmꢁ1 was associated with Lewis and Brønsted acid sites and
1545 cmꢁ1 was representative of Brønsted acid sites due to the
protonation of pyridine.34 The Brønsted acid amount of the
zeolite shell in SS-ber@meso-HZSM-5 catalysts decreased with
increasing caramel amount, indicating that the active sites for
catalysis reaction can be tuned by our caramel-assistant
synthesis. Meanwhile, the Brønsted/Lewis acid ratio also
decreased with increasing caramel amount. Essentially, the
change of acid amounts was closely associated with the ever-
increased SiO2/Al2O3 molar ratio in HZSM-5 shell, which was
determined by ICP-AES (Table 1). FT-IR spectra showed that the
band intensity at 3610 cmꢁ1, corresponding to the Brønsted
acid sites,35,36 decreased with adding caramel (Fig. 8), further
conrming the reduction of active sites. The above results
clearly indicated that the caramel impeded Al entering into the
zeolite framework while such impediment effect became
intense with the caramel adding amount.
Most notably, use of caramel template indeed brought the
increase of mesopore volume and external SSA of the zeolite
shell in the SS-ber@meso-HZSM-5. Adding a small amount of
caramel with a caramel/SiO2 mass ratio of 0.6 led to a little
increase in the mesopore volume (0.04 cm3 gꢁ1 vs. 0.03 cm3 gꢁ1
for the SS-ber@HZSM-5;21 Fig. 5A and Table 1). Increasing the
ratio to 1.2 tripled the mesopore volume (0.09 cm3 gꢁ1; Table 1),
while further increasing it up to 2.4 enhanced the mesopore
volume by a factor of 3 (0.12 cm3 gꢁ1; Table 1). Moreover, the
pore size distributions showed a decentralized trend along with
the increase of the caramel/SiO2 mass ratio (Fig. 5B). As a whole,
the addition of caramel originated the ever-increased meso-
porosity of the zeolite shell but clearly decreased its crystallinity.
This journal is © The Royal Society of Chemistry 2016
RSC Adv., 2016, 6, 48387–48395 | 48391

View Article Online
RSC Advances
Paper
Fig. 8 FT-IR spectra of (a) SS-fiber@meso-HZSM-5(3 h, 0.6), (b) SS-
fiber@meso-HZSM-5(3 h, 1.2), (c) SS-fiber@meso-HZSM-5(3 h, 1.8), (d)
SS-fiber@meso-HZSM-5(3 h, 2.4).
ratio of 3/7 at 450 ꢀC (Table 2 and Fig. S6†). Clearly, the SS-
ber@meso-HZSM-5(3 h, 1.8) was the best catalyst, delivering
the longest lifetime of 31 h, as well as higher propylene selec-
tivity of 47.9% and C3H6/C2H4 molar ratio of 4.8. It should be
noticed that the total acid amount of SS-ber@meso-HZSM-5(3
h, 1.8) was only 0.13 mmol gꢁ1, and according to our previous
work, such low acid amount was adverse to the stability of ZSM-
5 for MTP reaction.21–24 However, it was also reported that MFI
Fig. 6 SEM images of (A1 and A2) SS-fiber@meso-HZSM-5(3 h, 0.6), zeolite nanosheets catalyst with lower acidity (ꢂ0.05 mmol gꢁ1
)
was required to obtain the best MTP performance due to its
well-developed mesoporosity.37 In combining this information
with the fact that our SS-ber@meso-HZSM-5(3 h, 1.8) with
much lower acid amount than the SS-ber@meso-HZSM-5(3 h,
0.6) showed dramatically prolonged lifetime, we believed that
the developed mesopore system in the HZSM-5 shell was para-
mount to the prolonged lifetime of our structured HZSM-5
catalysts in the MTP reaction. Indeed, as clearly shown in
Fig. 9, the lifetime of the SS-ber@meso-HZSM-5 catalysts was
positively related with the ratio of Vmeso/Vtotal when the caramel/
SiO2 mass ratio was below 1.8, while the SS-ber@meso-HZSM-
5(3 h, 2.4) was inconsonant due to its lower crystallinity (Fig. 4B
and S4†) and ultra-low Brønsted acid amount (Fig. 7B and Table
1). Not surprisingly, the SS-ber@meso-HZSM-5(3 h, 0.6)
showed the shortest lifetime of only 5 h due to its minimum
mesopore volume. Such results proved the essential benet of
the mesopore development toward the stability improvement of
the SS-ber@meso-HZSM-5 in the MTP reaction process.
Moreover, the prolonged lifetime along with the development of
the mesopore structure was accompanied by a visible increase
(B1 and B2) SS-fiber@meso-HZSM-5(3 h, 1.2), (C1 and C2) SS-fiber@-
meso-HZSM-5(3 h, 1.8) and (D1 and D2) SS-fiber@meso-HZSM-5(3 h,
2.4), showing the morphology evolution with different caramel
amounts.
Fig. 7 (A) NH3-TPD profiles and (B) Py-IR spectra of the ZSM-5 shell of
(a) SS-fiber@meso-HZSM-5(3 h, 0.6), (b) SS-fiber@meso-HZSM-5(3 h,
1.2), (c) SS-fiber@meso-HZSM-5(3 h, 1.8), (d) SS-fiber@meso-HZSM-
5(3 h, 2.4).
+
of the C5 heavy products formation, which was also likely
attributed to the secondary porosity in zeolitic crystals that
improved the intracrystalline diffusion and nally reduced the
steric limitations for converting large molecules.38
As shown in Table 2, the C4 hydrogen transfer index (C4-
HTI)39,40 of the SS-ber@meso-HZSM-5(3 h, 1.8) was higher than
the SS-ber@meso-HZSM-5(3 h, 0.6), probably due to the higher
framework defects which were evidenced by the increased
intensity of inner silanol groups inside the zeolite crystals at
3727 cmꢁ1 in FT-IR spectra35,36 (Fig. 8). It is well accepted that
high C4-HTI generally is accompanied with high coking rate.39
This observation portended that an appropriate caramel/SiO2
mass ratio (to obtain more mesopores but with ensured crys-
tallinity) in the as-studied range was promising to improve the
catalytic performance (Fig. S5†).
3.3. Catalyst performance for MTP reaction
The SS-ber@meso-HZSM-5 core–shell catalysts prepared with
different caramel/SiO2 mass ratio were initially tested in MTP
reaction with a high WHSV of 10 hꢁ1 and methanol to N2 molar
48392 | RSC Adv., 2016, 6, 48387–48395
This journal is © The Royal Society of Chemistry 2016

View Article Online
Paper
RSC Advances
Table 2 Results for the MTP reaction, catalyzed by SS-fiber@meso-HZSM-5 catalystsa
Selectb (%)
Conv.
P/Ec Lifetimed C4-HTIe
+
¼
¼
Sample
(%)
CH4 C2H4 C2H6 C3H6 C3H8 C4H8 C4H10 C5
C2 –C4
SS-ber@meso-HZSM-5(3 h, 0.6) 100
SS-ber@meso-HZSM-5(3 h, 1.2) 100
SS-ber@meso-HZSM-5(3 h, 1.8) 100
2.2
2.0
1.1
2.0
16.0
11.8
6.7
0.2
0.1
0.1
0.1
38.4
43.6
47.9
48.1
3.3
1.3
0.5
0.4
14.8
15.8
14.2
12.9
9.0
9.1
10.2
9.9
16.1 69.2
16.3 71.2
19.3 68.8
20.1 67.5
1.6
2.5
4.8
5.0
5
22
31
10
0.38
0.37
0.42
0.43
SS-ber@meso-HZSM-5(3 h, 2.4)
99.4
6.5
a
Reaction conditions: WHSV ¼ 10 hꢁ1, 450 ꢀC, 0.1 MPa, zeolite 0.4 g. b Data were collected in steady-state associated with the highest propylene
formation; calculation was on the basis of carbon. c C3H6/C2H4 molar ratio. d Life time till the reaction was quitted at ꢂ90% conversion. e C4-HTI ¼
C4H10/(C4H10 + C4H8).
However, only 3.4 wt% coke was formed in our SS-ber@meso-
HZSM-5(3 h, 1.8) shell aer 31 h on stream, corresponding to an
average coking rate of 1.0 mg gzeoliteꢁ1 hꢁ1. In contrast, for the
SS-ber@meso-HZSM-5(3 h, 0.6), even though with much lower
C4-HTI, 4.6 wt% coke was formed in its zeolite shell aer only
5 h on stream, showing more than ve times higher coking rate
than the SS-ber@meso-HZSM-5(3 h, 1.8). This observation was
mainly due to its smaller mesopore volume as well as external
SSA (Table 1) and weaker intensity of isolated silanol groups on
external surface (Fig. 8).
3.4. Stability
Fig. 10 Methanol conversion and product selectivity for MTP vs. time
on stream using SS-fiber@meso-HZSM-5(3 h, 1.8). Reaction condi-
tions: 450 ꢀC, WHSV ¼ 1 hꢁ1, 0.1 MPa, 30 vol% methanol in N2, zeolite
0.4 g. Testing was stopped at ꢂ90% conversion.
The hierarchically designed and engineered SS-ber@meso-
HZSM-5(3 h, 1.8) was also tested at 450 ꢀC, under the indus-
trial relevant conditions with WHSV of 1 hꢁ1. As expected, this
catalyst showed a dramatic stability improvement, delivering
a single-run lifetime of 845 h with the terminated methanol
conversion of 90%. In contrast, the SS-ber@HZSM-5 synthe-
sized without caramel only delivered a lifetime of 210 h under
the identical conditions.21,22 The methanol conversion was
maintained at >99% over the rst 540 h and then experienced
a very slow decline to 90% within another 300 h (Fig. 10). The
propylene selectivity increased to ꢂ40% aer only 50 h running,
and then climaxed and remained at ꢂ45%. Notably, although
higher propylene selectivity was obtained, C2–C4 olen selec-
tivity decreased by ꢂ10% within the 845 h testing due to the
reduction of ethylene and butene selectivities (Fig. 10). In this
+
course, the selectivity to C5 heavy products increased from
ꢂ18% at the initial to ꢂ30% at the end.
3.5. Insight into the prolonged lifetime
Coking is well known to be the main cause for the HZSM-5
deactivation in the MTP process12 and therefore it was partic-
ularly desirable to study their coking behavior for gaining
insight into the dramatic stability improvement by the hierar-
chical pores design. Fig. 11 shows the comparison of SS-
ber@meso-HZSM-5(3 h, 1.8) catalysts before and aer 845 h
on stream for MTP reaction under WHSV of 1 hꢁ1. Clearly, there
were large amounts of spherical particles formed on the surface
of the used catalyst aer undergoing long-term MTP reaction
(Fig. 11B). In order to further identify such species, used catalyst
ꢀ
was calcined at 550 C in air and such spherical particles dis-
appeared (Fig. 11C), which indicated that they were coke
deposit formed in the long-term reaction. As shown in the XRD
patterns in Fig. 11D, the used catalyst still possessed MFI
structure and higher crystallinity. In comparison with the fresh
SS-ber@meso-HZSM-5(3 h, 1.8) catalyst, the strong acid
amount (mainly attributed to the Brønsted acid amount) of
zeolite shell aer reaction was ultra-low (Fig. 11E). Over the
used SS-ber@meso-HZSM-5(3 h, 1.8) catalyst, coke amount in
the zeolite shell was determined by thermal gravity analysis
(TGA) to be ꢂ16.2 wt% aer 845 h running, corresponding to an
Fig. 9 Lifetime (asterisks) and Brønsted acid amount (circles) as
a function of Vmeso/Vtotal ratio.
average coking rate of 1.9 ꢃ 10ꢁ4 g gzeolite hꢁ1, which was
ꢁ1
This journal is © The Royal Society of Chemistry 2016
RSC Adv., 2016, 6, 48387–48395 | 48393

View Article Online
RSC Advances
Paper
Fig. 11 All-in-one for the SS-fiber@meso-HZSM-5(3 h, 1.8) catalysts before and after 845 h on stream for the MTP reaction. (A) Photograph and
(B) SEM images of the used catalyst, (C) SEM images of the used catalyst after calcination at 550 ꢀC in air, (D) XRD patterns and (E) NH3-TPD
profiles of the zeolite shell, (F) TGA profile of the zeolite shell in the used catalyst, (G) N2 adsorption–desorption isotherms as well as (H) Barrett–
Joyner–Halenda (BJH) mesopore size distribution of the monolithic catalyst with a zeolite content of ꢂ17.7% and (I) the changes of the zeolite
shells texture properties.
slightly higher than the coking rate of the shell in the hierarchical micro–meso–macropore system coupled with
ꢁ1
previously-reported SS-ber@HZSM-5 (1.2 ꢃ 10ꢁ4 g gzeolite
facilely-tuned acidic properties, has been developed by caramel-
hꢁ1).22 Fig. 11G–I show that the micropore volume of the cata- assistant hydrothermal synthesis. Such catalysts, with high
lyst decreased by 25% aer 845 h on stream while the micropore selectivity to propylene, show dramatic stability improvement in
SSA decreased by 28%. In contrast, both mesopore volume and the MTP process, because of the mesopore development in the
external SSA showed a dramatic decrease, strongly indicating hierarchical micro–meso–macropore structure with favourably-
that the mesopores were essentially responsible for accommo- tuned acidic properties.
dating the formed coke during the MTP reaction. Notably,
although the zeolite shell of the used catalyst still retained 75%
Acknowledgements
micropore volume compared to the fresh catalyst, the used SS-
ber@meso-HZSM-5(3 h, 1.8) showed a big decline of the acid
sites with almost complete loss of the strong acid sites
(Fig. 11E); this was the main cause for its activity loss. In these
This work was nancially supported by the NSF of China
(21473057, 21322307, U1462129, 21273075), and the “973
program” (2011CB201403) from the MOST of China.
microporous channels, coke favorably deposited on the
Brønsted acid sites associated with a partial blocking the
micropore mouth (ꢂ25%; Fig. 11I). The above results revealed
Notes and references
that under the industrial relevant conditions with WHSV of
1 hꢁ1, our hierarchical design mainly increased the accommo-
dation capacity of zeolite shell for receiving formed coke
thereby avoiding fast micropore mouth blocking. As a result,
the microbrous-structured SS-ber@meso-HZSM-5 catalyst
1 H. Torres Galvis and K. de Jong, ACS Catal., 2013, 3, 2130.
2 P. Tian, Y. Wei, M. Ye and Z. Liu, ACS Catal., 2015, 5, 1922.
3 A. Corma, Chem. Rev., 1995, 95, 559.
4 G. F. Froment, J. D. Meyer and E. G. Derouane, J. Catal., 1990,
provided a dramatically prolonged lifetime in the MTP reaction.
124, 391.
5 C. Mei, P. Wen, Z. Liu, Y. Wang, W. Yang, Z. Xie, W. Hua and
Z. Gao, J. Catal., 2008, 258, 243.
6 F. Xu, M. Dong, W. Gou, J. Li, Z. Qin, J. Wang and W. Fan,
4. Conclusions
In summary, a promising self-supported SS-ber@meso-HZSM-
5 core–shell catalysts to be used in the MTP process, with
enhanced mesopore structure in the micro-engineered
Microporous Mesoporous Mater., 2012, 163, 192.
7 A. Z. Ruiz, H. Li and G. Galzaferri, Angew. Chem., Int. Ed.,
2006, 45, 5282.
48394 | RSC Adv., 2016, 6, 48387–48395
This journal is © The Royal Society of Chemistry 2016

View Article Online
Paper
RSC Advances
8 S. Ivanova, C. Lebrun, E. Vanhaecke, C. Pham-Huu and 25 S. Ivanova, B. Louis, B. Madani, J. P. Tessonnier,
B. Louis, J. Catal., 2009, 265, 1.
9 Q. Wang, S. Xu, J. Chen, Y. Wei, J. Li, D. Fan, Z. Yu, Y. Qi,
M.-J. Ledoux and C. Pham-Huu, J. Phys. Chem. C, 2007,
111, 4368.
Y. He, S. Xu, C. Yuan, Y. Zhou, J. Wang, M. Zhang, B. Su 26 Y. L. Jiao, X. D. Yang, C. H. Jiang, C. Tian, Z. M. Yang and
and Z. Liu, RSC Adv., 2014, 4, 21479.
J. S. Zhang, J. Catal., 2015, 332, 70.
10 X. Meng and F. Xiao, Chem. Rev., 2014, 114, 1521.
11 L. Chen, X. Li, J. Rooke, Y. Zhang, X. Yang, Y. Tang, F. Xiao
and B. Su, J. Mater. Chem., 2012, 22, 1738.
27 H. Koempel and W. Liebner, in Studies in Surface Science and
Catalysis, ed. F. B. Noronha, M. Schmal and E. F. Sousa-
Aguiar, Elsevier, 2007, p. 261.
´
12 M. Milina, S. Mitchell, P. Crivelli, D. Cooke and J. Perez- 28 D. Nandan, S. K. Saxenab and N. Viswanadham, J. Mater.
´
Ramırez, Nat. Commun., 2014, 3922.
Chem. A, 2014, 2, 1054.
´
13 M. Milina, S. Mitchell, D. Cooke, P. Crivelli and J. Perez- 29 I. Ali, M. Thabet, K. El-Nasser, A. Hassan and T. Salama,
´
Ramırez, Angew. Chem., Int. Ed., 2015, 54, 1591.
Microporous Mesoporous Mater., 2012, 160, 97.
14 J. J. Zhao, Z. L. Hua, Z. C. Liu, Y. S. Li, L. M. Guo, W. B. Wu, 30 M. Kustova, K. Egeblad, K. K. Zhu and C. H. Christensen,
X. Z. Cui, M. L. Ruan, H. R. Chen and J. L. Shi, Chem.
Commun., 2009, 7578.
15 J. Zhou, Z. L. Hua, Z. C. Liu, W. Wu, Y. Zhu and J. L. Shi, ACS
Catal., 2011, 1, 287.
16 M. Choi, H. S. Cho, R. Srivastava, C. Venkatesan, D. H. Choi
and R. Ryoo, Nat. Mater., 2006, 5, 718.
17 L. J. Gu, D. Ma, S. D. Yao, C. L. Wang, W. J. Shen and
X. H. Bao, Chem. Commun., 2003, 958.
Chem. Mater., 2007, 19, 2915.
31 R. White, A. Fischer, C. Goebel and A. Thomas, J. Am. Chem.
Soc., 2014, 136, 2715.
32 W. H. Wang, G. Li, W. G. Li and L. P. Liu, Chem. Commun.,
2011, 47, 3529.
33 A. E. Persson, B. J. Schoeman, J. Sterte and J. E. Otterstedt,
Zeolites, 1994, 14, 557.
34 C. A. Emeis, J. Catal., 1993, 141, 347.
18 K. Shen, W. Qian, N. Wang, C. Su and F. Wei, J. Am. Chem. 35 P. Sazama, B. Wichterlova, J. Dedecek, Z. Tvaruzkova,
Soc., 2013, 135, 15322.
19 N. Wang, W. Qian, K. Shen, C. Su and F. Wei, Chem.
Commun., 2016, 52, 2011.
20 Q. Zhang, S. Hu, L. L. Zhang, Z. J. Wu, Y. J. Gong and T. Dou,
Green Chem., 2014, 16, 77.
21 X. Y. Wang, M. Wen, C. Z. Wang, J. Ding, Y. Sun, Y. Liu and
Y. Lu, Chem. Commun., 2014, 50, 6343.
22 M. Wen, X. Y. Wang, L. P. Han, J. Ding, Y. Sun, Y. Liu and
Y. Lu, Microporous Mesoporous Mater., 2015, 206, 8.
23 M. Wen, J. Ding, C. Z. Wang, Y. K. Li, G. F. Zhao, Y. Liu and
Y. Lu, Microporous Mesoporous Mater., 2016, 221, 187.
24 J. Ding, L. P. Han, M. Wen, G. F. Zhao, Y. Liu and Y. Lu,
Catal. Commun., 2015, 72, 156.
Z. Musilova, L. Palumbo, S. Sklenak and O. Gonsiorova,
Microporous Mesoporous Mater., 2011, 143, 87.
36 K. Barbera, F. Bonino, S. Bordiga, T. Janssens and P. Beato, J.
Catal., 2011, 280, 196.
37 Y. Kim, J.-C. Kim, C. Jo, T.-W. Kim, C.-U. Kim, S.-Y. Jeong and
H.-J. Chae, Microporous Mesoporous Mater., 2016, 222, 1.
38 C. H. Christensen, K. Johannsen, I. Schmidt and
C. H. Christensen, J. Am. Chem. Soc., 2003, 125, 13370.
39 L. Huang, P. Wang, J. Li, J. Wang and W. Fan, Microporous
Mesoporous Mater., 2016, 223, 230.
40 C. M. Tsang, P.-S. E. Dai, F. P. Mertens and R. H. Petty, Prepr.
Symp.–Am. Chem. Soc., Div. Fuel Chem., 1994, 39, 367.
This journal is © The Royal Society of Chemistry 2016
RSC Adv., 2016, 6, 48387–48395 | 48395