Synthesis of mesoporous ZSM-5 zeolites and catalytic cracking of ethanol and oleic acid into light olefins

Article abstract of DOI:10.1016/j.apcata.2019.02.011

Conversion of biomass-derived chemicals into light olefins is a promising method to maintain sustainable development of light olefin industry. In this study, three mesoporous ZSM-5 zeolites (MZSM-5-A, MZSM-5-B and MZSM-5-C) with major pore diameter about 4.8 nm, 16 nm and 22 nm were synthesized using a hydrothermal method by utilizing different templates. The catalytic activity of catalysts was studied by catalytic cracking of ethanol and oleic acid. The influence of reaction temperature on conversion and product selectivity was investigated. The characterization of ZSM-5 samples showed that the orders of the external surface area and mesopore volume were MZSM-5-C > MZSM-5-B > MZSM-5-A > conventional HZSM-5. In ethanol to light olefin reaction, MZSM-5-C achieved the highest light olefin yield (318.3 mL g^?1) and ethylene selectivity (42.3%) at 400 °C. In oleic acid to light olefin reaction, MZSM-5-B achieved a complete conversion of oleic acid at 500 °C, and obtained the highest light olefin selectivity (38.1%) at 550 °C. The difference may be relevant to the size and chemical structure of feedstock molecular as well as the acidity of catalysts. Regardless of ethanol or oleic acid as feedstock, introduction of mesopore in zeolites significantly enhanced the light olefin yield and selectivity.

Full text of DOI:10.1016/j.apcata.2019.02.011

Accepted Manuscript
Title: Synthesis of mesoporous ZSM-5 zeolites and catalytic
cracking of ethanol and oleic acid into light olefins
Authors: Tingting Zhao, Fuwei Li, Hongchang Yu, Shilei
Ding, Zhixia Li, Xinyuan Huang, Xiang Li, Xiaohan Wei,
Zhenlin Wang, Hongfei Lin
PII:
S0926-860X(19)30066-3
DOI:
Reference:
https://doi.org/10.1016/j.apcata.2019.02.011
APCATA 16981
To appear in:
Applied Catalysis A: General
Received date:
Revised date:
Accepted date:
16 October 2018
30 December 2018
8 February 2019
Please cite this article as: Zhao T, Li F, Yu H, Ding S, Li Z, Huang X, Li X, Wei
X, Wang Z, Lin H, Synthesis of mesoporous ZSM-5 zeolites and catalytic cracking
of ethanol and oleic acid into light olefins, Applied Catalysis A, General (2019),
https://doi.org/10.1016/j.apcata.2019.02.011
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

Synthesis of mesoporous ZSM-5 zeolites and catalytic cracking of ethanol and
oleic acid into light olefins
Tingting Zhaoa, Fuwei Lia, Hongchang Yua, Shilei Dinga, Zhixia Lia,*, Xinyuan Huanga, Xiang Lia,
Xiaohan Weia, Zhenlin Wanga, Hongfei Linb
aSchool of Chemistry and Chemical Engineering, Guangxi University, Nanning 530004, People’s
Republic of China
bGuangxi Bossco Environmental Protection Technology Co., Ltd., Nanning 530007, People’s
Republic of China
*Corresponding author. Z. Li; Tel: +86 771 3274209; E-mail address: zhixiali@hotmail.com
Graphical abstract
1

Highlights
 Mesoporous ZSM-5 is synthesized by utilizing template.
 Light olefins are produced from catalytic cracking of ethanol and oleic acid over mesoporous
ZSM-5.
 Mesoporous ZSM-5 shows better catalytic activity than traditional microporous ZSM-5.
 Mesoporous ZSM-5 shows better anti-coking ability than traditional ZSM-5.
Abstract: Conversion of biomass-derived chemicals into light olefins is a promising method to
maintain sustainable development of light olefin industry. In this study, three mesoporous ZSM-5
zeolites (MZSM-5-A, MZSM-5-B and MZSM-5-C) with major pore diameter about 4.8 nm, 16 nm
and 22 nm were synthesized using a hydrothermal method by utilizing different templates. The
catalytic activity of catalysts was studied by catalytic cracking of ethanol and oleic acid. The
influence of reaction temperature on conversion and product selectivity was investigated. The
characterization of ZSM-5 samples showed that the orders of the external surface area and
mesopore volume were MZSM-5-C > MZSM-5-B > MZSM-5-A > conventional HZSM-5. In
ethanol to light olefin reaction, MZSM-5-C achieved the highest light olefin yield (318.3 mL∙g-1)
and ethylene selectivity (42.3%) at 400 °C. In oleic acid to light olefin reaction, MZSM-5-B
achieved a complete conversion of oleic acid at 500 °C, and obtained the highest light olefin
selectivity (38.1%) at 550 °C. The difference may be relevant to the size and chemical structure of
feedstock molecular as well as the acidity of catalysts. Regardless of ethanol or oleic acid as
feedstock, introduction of mesopore in zeolites significantly enhanced the light olefin yield and
selectivity.
Keywords: mesoporous ZSM-5; ethanol; oleic acid; light olefins; catalytic cracking
2

1. Introduction
Light olefins, such as ethylene, propylene and butene, are an important raw material in
petrochemical industry. At present, the main commercial techniques for producing light olefins are
steam cracking (SC) or catalytic cracking of naphtha, light diesel oil and other petroleum products
[1, 2]. With light olefins demand increasing and petroleum resources supply dwindling, exploiting
new non-petroleum raw materials to produce light olefins has attracted significant research
attention. Light olefins have been produced from catalytic cracking of biomass [3, 4],
bio-methanol [5], bio-ethanol [6], fatty acids and triglycerides [7].
Catalytic cracking is an efficient way to produce light olefins from various resources. For
catalytic cracking process, catalyst is a key factor affecting process efficiency and target product
selectivity. Zeolite Y is the main component of typical FCC catalyst, but possesses low activity for
production of light olefins. When zeolite ZSM-5 was mixed with zeolite Y, and used as catalyst for
FCC of vacuum gasoil, more gasoline molecules were converted into light olefins [8]. Cracking of
low value C4-C8 olefinic streams coming from steam cracking with ZSM-5-based catalysts
showed that the propylene/ethylene (P/E) ratio in products was significantly improved [9, 10].
ZSM-5 was used in “Methanol to Olefins” (MTO) reaction, finding that ZSM-5 was stable and
gave propylene as main product [11]. Butenes were produced by dehydro-isomerization of
n-butane on Pt supported ZSM-5 (SiO2/A12O3=480), consequently butenes selectivity around 45%
was obtained even at high conversion (80%) [12]. The outstanding catalytic activity and selectivity
for light olefin production of ZSM-5 was attributed to the proper acidic property and pore structure
dependent-shape selectivity. However, the relatively small size of the pore of ZSM-5 zeolite and
3

zeotype materials (maximum pore size is typically, 1.2 nm), may significantly limit the
possibilities of upgrading of crude oil and transformations of numerous bulky compounds into fine
chemicals [13, 14]. Recently, the synthesis of zeolites with new structures, in particular those with
large channels where bigger molecules can diffuse into the zeolite crystallite, is an active field of
research.
Early mesoporous materials such as MCM-41 and SBA-15 were synthesized using the
organic-template-assisted approach, in which self-assembled molecular aggregates or
supramolecular assemblies were used as structure-directing agents. These molecular sieves had
well-defined pores up to about 30 nm in diameter [15, 16]. However, the disadvantages of these
materials were low acidity (owing to the low concentration of aluminum) and insufficient thermal
stability (due to the amorphous structure). This is unfavorable for cracking of bulky molecules at
an elevated temperature. Afterwards, a number of researchers started work on preparation of
micro/mesoporous composites to improve surface acidity and hydrothermally stability. In this case,
ZSM-5/MCM-41, zeolite Y/MCM-41, zeolite Beta/MCM-48 and mesoporous ZSM-5 single
crystals were prepared according to either recrystallization of amorphous pore walls in presence of
structure-directing agent, or the use of zeolite seeds as framework-building units, which can
self-organize in mesoporous matrices, or direct assembly of nano-clustered aluminosilicate
precursors–zeolite seeds in crystalline zeolite structures, or utilization of small particle of organic
or inorganic particles (carbon black) as substrate templates [16]. The resulting mesoporous
zeolite-based materials exhibited larger pore volume, higher acidity and hydrothermal stability. In
another research, mesoporous ZSM-5 was directly synthesized from the normal synthesis gel of
ZSM-5 under microwave heating and addition of carbon particles as template [17]. Addition of 40
wt% of carbon led to the formation of mesoporous ZSM-5 (MW-40) with high external surface
4

area (383 m2/g) and mesoporous volume (0.82 cm3/g). The MW-40 exhibited eight times higher
conversion than corresponding standard ZSM-5 without mesopores in the transformation reaction
of 2',4'-dimethoxyacetophenone with 4-methoxybenzaldehyde into vesidryl. Besides, mesopores
can be generated by steaming and acid treatment [18], alkali treatment [19] and heat treatment [20].
However, these methods trend to generate nonuniform mesopores, and the skeleton structure of the
zeolites is frequently destroyed during treatment process.
Most of the synthetic porous zeolites were obtained via solvothermal synthesis by use of
different reaction conditions, reactants, and structure-directing agents. Ahmadpour and Taghizadeh
[21] prepared hierarchical porous ZSM-5 with different pore size by using different templates, e.g.
tetrapropyl ammonium hydroxide (TPAOH), cetyltrimethyl ammonium bromide (CTAB) and
dimethyl octadecyl [3-(trimethoxysilyl)propyl]ammonium chloride (TPOAC), demonstrating that,
in comparison to conventional microporous ZSM-5, the hierarchical porous ZSM-5 with the
secondary porosity centered at ~16 nm achieved a longer catalytic lifetime and a higher propylene
selectivity in MTP reaction. Wen et al. [22] synthesized hierarchical porous ZSM-5 by using
3-triethoxysilypropylmercaptan (CPTES) as a template for catalytic conversion of chloromethane
into light olefins. The obtained mesopore-abundant ZSM-5 improved light olefin selectivity by 24%
and extended the catalytic lifetime by 4.5 times compared to the traditional ZSM-5. The above
results indicate that the introduction of mesopores into ZSM-5 to some extent enhances catalytic
activity and light olefin selectivity, especially when smaller molecules e.g. methanol and
chloromethane are used as feedstocks. However, so far, a few of work has been done to clarify the
effect of pore structure of catalysts on catalytic cracking of a long-chain large molecule reactant. In
this work, three mesoporous ZSM-5 with different pore sizes were synthesized by using different
templates. The obtained ZSM-5 samples were used to catalyze small molecule ethanol and
5

long-chain large molecule oleic acid (both are renewable resources) to produce light olefins. The
effects of pore structure of zeolites and process conditions on conversion and light olefin
selectivity were discussed.
2. Experimental
2.1. Materials
Colloidal silica (Ludox, 40%, Guangzhou Suize Co. Ltd), sodium aluminate (NaAlO2, Al2O3
wt.% = 50%, Sinopharm Chemical Reagent Co. Ltd), sodium hydroxide (NaOH, 96%, Aladdin
Reagents), ammonium nitrate (NH4NO3, analytical grade), tetrapropyl ammonium bromide
(TPABr, 98%, Aladdin), tetrapropyl ammonium hydroxide (TPAOH, 25%, Aladdin), hexadecyl
trimethyl
ammonium
bromide
(CTAB,
99%,
Aladdin),
dimethyl
octadecyl
[3-(trimethoxysilyl)propyl]ammonium chloride (TPOAC, 65%, Macklin Chemical Reagent), and
sulfuric acid (H2SO4, 98%) were used as received. H-type ZSM-5 zeolite (SiO2/Al2O3=227) was
supplied by Nankai University Catalyst Corporation (Tianjin, China) and activated at 110 °C for 6
h before use.
2.2. Preparation of catalysts
Three mesoporous ZSM-5 samples with different pore size were synthesized according to a
hydrothermal method [23-25]. The final molar composition of three synthesis mixtures was
200SiO2–1.0Al2O3–27.8NaOH–10TPABr–3.5TPOAC–14.5H2SO4–3000H2O for MZSM-5-A,
200SiO2–1.0Al2O3–27.8NaOH–5.0TPABr–10CTAB–14.5H2SO4–3000H2O for MZSM-5-B, and
200SiO2–1.0Al2O3–27.8NaOH–10TPAOH–10CTAB–14.5H2SO4–3000H2O for MZSM-5-C. A
typical synthesis process was as follows: first, a certain amounts of NaOH and NaAlO2 were
6

dissolved in H2O to form solution A. A certain amount of Ludox and templates were dispersed in
H2O to form solution B. Then, solution A was added into solution B drop by drop with agitation to
form gel mixture. A certain amount of moderate concentrated H2SO4 was added to the mixture to
adjust gel pH to 7, and continuously stirred for 4 h. Finally, the resultant gel was transferred into a
stainless-steel Teflon-lined autoclave and crystallized at 180 °C for 48 h. The solid product was
filtered, washed with distilled water until the pH of the filtrate was 7, and then dried at 105 °C
overnight. The product was afterwards calcined at 550 °C for 6 h to remove the organic templates.
The obtained sample was defined as Na-ZSM-5 zeolite. HZSM-5 zeolite was obtained by
ion-exchanging Na-ZSM-5 zeolite three times with 1 M NH4NO3 solution (liquid/solid = 10/1
(g/g)) at 90 °C for 2 h. The solid sample was filtered, dried at 105 °C overnight, and calcined at
550 °C for 6 h. The HZSM-5 was tableted, granulated and sieved to obtain 20-30 mesh catalyst.
2.3. Characterizations of catalysts
The crystalline structure and pore structure of the synthesized ZSM-5 were determined by
X-ray diffraction (XRD) and N2 adsorption-desorption method, respectively. The acidic properties
were analyzed by the temperature-programmed desorption of ammonia (NH3-TPD) and infrared
adsorption of pyridine (Py-IR). The detail about determination was revealed in previous study [26].
The analysis of Py-IR is as follows. Catalysts samples were activated at 500 °C under vacuum
(10-3 Pa) for 1.5 h and cooled to room temperature in an IR cell prior to the adsorption of pyridine.
Pyridine adsorption proceeded at ambient temperature for 30 min and followed by desorption at
150 °C or 300 °C for 20 min. After cooling to room temperature, IR spectra were collected using
an FT-IR spectrometer (Thermo Nicolet380). The concentrations of Brønsted acid and Lewis acid
sites were calculated from the integral intensities of individual adsorption bands at 1540 cm-1
7

(Brønsted acid sites) and 1446 cm-1 (Lewis acid sites) [27]. The actual SiO2/Al2O3 molar ratio was
determined by inductively coupled plasma optical emission spectroscopy (ICP−OES, Agilent
7700e).
2.4. Catalytic performance tests
Catalytic performance of catalysts was evaluated by catalytic cracking of ethanol and oleic
acid in a fixed-bed quartz tube reactor (inner diameter: 20 mm; length: 380 mm). As a typical run,
the catalyst (0.5 g) was loaded into reactor, and heated to a desired temperature in N2 stream (the
flow rate: 40 mL/min). The feedstock (0.87 g) was then injected into the reactor by a syringe pump.
After flowing out from the reactor, the reaction products were cooled. Liquid products were
weighed and analyzed by GC-MS (QP5000, Shimadzu, Japan) equipped with a DB-WAX fused
silica capillary column (30 m 0.25 mm  0.25 μm). The gaseous product was collected with a gas
collecting bag, and analyzed by GC referring to the method as described in our previous study
[26].
The amount of coke deposition on the catalysts was calculated from weight loss during
calcination of the used catalysts at 550 °C for 3 h. 0.5 g of the used catalysts were dispersed in 2
mL dichloromethane, and the dichloromethane-dissolved compounds (soluble coke) were analyzed
by GC-MS. Temperature programmed oxidation (TPO) of carbon deposition was performed to
determine the properties of coke [26].
The yields of total gas production (YTGP) and light olefins (YLO), the selectivity to light olefins
(SLO), feedstock conversion and coke yield (Ycoke) were calculated by the following equations,
respectively:
YTGP (mL/g) = VTGP/M0
(1)
8

SLO (%) = x(C2H4 ) + x(C3H6 ) + x(C4H8)
YLO (mL/g) = YTGP·SLO
(2)
(3)
(4)
(5)
Conversion (%) = [1- (Mu/M0)] × 100%
Ycoke (%) = (Mc0– Mc1)/M0 × 100%
where VTGP is the total volume of gas product excluding N2 (mL), M0 is the mass of the feedstock,
x is the molar percentage content of light olefins in the gas product (%), Mu is the mass of the
unreacted feedstock, Mc0 is the mass of the used catalysts before calcination (g), and Mc1 is the
mass of the used catalyst after calcination (g). The data of YTGP, YLO, SLO, Conversion and Ycoke are
the average value of three parallel experiments.
2.5. Stability experiments of catalysts
The stability of the catalysts was determined by ten continuous cracking-regeneration tests.
The used catalyst was regenerated by calcination at 550 °C for 3 h in air atmosphere to remove the
deposited coke.
3. Results and discussion
3.1. Characterization of the catalysts
The low-angle and the wide-angle XRD patterns of the catalyst samples are indicated in Fig.
1. The main peaks observed in the XRD patterns (2θ = 7.8, 8.8, 23.2 and 23.8°) of all samples are
assigned to the ZSM-5 orthorhombic structure [23]. The lattice parameters of the three ZSM-5
samples were calculated and depicted in Table 1, which were compared with that of conventional
HZSM-5. The relative crystallinity (was calculated based on the sum of peak areas between 2θ =
22.5−25° from XRD pattern of ZSM-5 samples compared to that of conventional HZSM-5) is 85.8%
9

for MZSM-5-A, 71.1% for MZSM-5-B and 54.7% for MZSM-5-C. The decrease in crystallinity of
the synthesized ZSM-5 is probably attributed to the weaker interaction between templates
(TPOAC and CTAB) and aluminosilicate species, which leads to a more difficult deposition of
aluminosilicate species on templates. Some amorphous species were likely formed during the
hydrothermal crystallization process [28, 29]. The crystal sizes of zeolites were calculated with
Scherrer Equation using XRD data. The crystal sizes of the synthesized ZSM-5 are 37~69 nm
(shown in Table 1), which are larger than MFI nanozeolite as reported in Valtchev and Mintovas’
research [30], but smaller than the referential conventional ZSM-5 (89 nm). The smaller crystal
sizes are thought to be more favorable for shortening diffusion path and increasing diffusion rate
of the reactants and products molecules.
The N2 adsorption-desorption isotherms and the pore size distribution for each sample are
illustrated in Fig. 2. As can be seen, the conventional HZSM-5 presents a typical I-type isotherm
without any hysteresis loop even at a high relative pressure, indicating the existence of
microporous structure [13, 24]. For the three synthesized ZSM-5, N2 adsorption amounts increase
significantly and an IV-type isotherm with a hysteresis loop at a high relative pressure
(P/P0=0.7-0.99) are observed. In particular, MZSM-5-B and MZSM-5-C show a larger hysteresis
loop than MZSM-5-A, indicating that there are more mesoporous structure in MZSM-5-B and
MZSM-5-C. Fig. 2(b) shows the pore size distributions of the four ZSM-5 obtained by using
Barrett-Joyner-Halenda (BJH) method [31]. As can be seen, the pore size distribution broadens in
this order: HZSM-5 < MZSM-5-A < MZSM-5-B < MZSM-5-C. The major pore diameter is 2.5
nm for HZSM-5, 4.8 nm for MZSM-5-A, 16 nm for MZSM-5-B, and 22 nm for MZSM-5-C. Table
2 summarizes the textural properties of all ZSM-5 samples. All ZSM-5 samples show a high
specific surface area (> 400 m2∙g-1). Compared with HZSM-5, external surface areas (Sext), total
10

pore volumes (Vtotal), mesopore volumes (Vmeso), and average pore diameter (Daver) of the
synthesized ZSM-5 significantly increase, indicating that TPOAC and CTAB as templates
substantially improve the mesopore formation in catalysts. We deduce that mesostructure could be
generated by aggregation of the surfactant molecules, whereas the crystallization of microporous
ZSM-5 is directed by quaternary ammonium groups within the mesoporous structure [29].
Mesoporous domain in fact plays an important role in scattering microporous zeolites and active
sites. Although the SBET and Sext of the synthesized ZSM-5 in our work are lower than those of
two-dimensional MCM-36 and ITQ-2, but are much higher than those of MCM-22P and MCM-56
MWW zeolites [32]. Furthermore, Sext occupies >57% of total surface area of the fabricated
ZSM-5 zeolites, thus, mesoporous structure in present study are expected to play an important role
in contributing on catalytic activity and selectivity.
The NH3-TPD profiles of the ZSM-5 samples are illustrated in Fig. 3 and Table 3. As can be
seen, two obvious NH3 desorption peaks appeared at 100−200 °C and 400 °C are observed for all
ZSM-5 samples, which are ascribed to the desorption of NH3 adsorbed on the weak acid sites and
strong acid sites, respectively [33, 34]. All ZSM-5 samples exhibit a similar total acidity, indicating
that the kinds of template and pore structure have little effect on the total acidity, and the acidity is
mainly determined by the SiO2/Al2O3 ratio.
Fig. 4 shows the Py-IR spectra of the ZSM-5 samples at different desorption temperature. As
can be seen in Fig. 4(a), three characteristic bands at 1446 cm-1, 1490 cm-1and 1540 cm-1 of all
samples are observed. The IR band at 1446 cm-1 is attributed to Lewis (L) acid sites, mainly
derives from Lewis aluminum, being an electron pair acceptor (binding pyridine in a coordination
fashion). The IR band at 1540 cm-1 can be ascribed to Brønsted (B) acid sites, mainly deriving
from PyrH+ created by protonation. Infrared band appearing at 1490 cm-1 is related to the vibration
11

of the pyridine ring on both B and L acid sites [24]. The concentrations of B and L acid sites
determined based on the intensity of these bands are shown in Table 4. As can seen, the
concentration of B acid sites is much lower than that of L acid sites for all samples, and with
increasing the desorption temperature to 300 °C, the concentration of all acid sites substantially
decrease (Fig. 4b). This indicates that the ZSM-5 samples prepared in present study are L
acid-dominant materials, and among them, MZSM-5-B possesses more B acid sites.
3.2. Catalyst Performance
3.2.1 Conversion of ethanol into light olefins
In ethanol to light olefins (ETO) reaction, ethanol was catalytically cracked at 200−500 °C in
N2 stream over four ZSM-5 samples. Conversion and YLO are exhibited in Fig. 5. As can be seen in
Fig. 5, at the lowest temperature, 200 °C, the conversions for four ZSM-5 samples are lower than
60%, but the YLO are less than 20 mL∙g-1. At the highest temperature, 500 °C, the conversions from
the synthesized ZSM-5 are higher than 95%, but the conversion from conventional HZSM-5
reaches only 89%. Meanwhile, the conversion increases with the rise of the pore size of catalysts,
and meets the maximum value over MZSM-5-C (99.4%) at 500 °C. As shown in Fig. 5(b), the YLO
from different catalysts increase with raising reaction temperature until 300 °C, and maintain a
constant value at a higher temperature. Sample MZSM-5-C gives the highest YLO (> 310 mL∙g-1)
whereas the conventional HZSM-5 obtains the lowest value (about 262 mL∙g-1) at 300−500 °C.
These results demonstrate that introduction of mesopore into zeolites significantly enhanced
ethanol conversion and light olefin selectivity.
Fig. 6 shows the gaseous product selectivity from cracking of ethanol at different temperature
over the synthesized ZSM-5 samples. As shown in Fig. 6(a), the selectivity to ethylene increases
12

with the rise of reaction temperature, and meets the maximum value (42.3%) over MZSM-5-C at
400 °C. It is worth noting that the propylene/ethylene mass ratios (P/E values) in products for all
catalysts are less than 1.0, and dramatically decrease with increasing reaction temperature. This
indicates that ethylene is the main light olefin product from cracking of ethanol at the tested
temperature range. Meanwhile, as can be seen in Fig. 6(b), a little amount of C1-C4 alkanes are
formed, and the selectivity slightly increases with raising temperature. CO and CO2 were not
detected in this reaction likely due to the extremely low content. Some other gas components such
as n-pentane, isobutylene and butadiene, were also probably produced but could not be identified
with the present standard gas.
It is thought that ethanol conversion is controlled by a single-site surface reaction, and
strongly depends on the diffusion of reagents and products in the catalyst particles [35].
MZSM-5-C contains larger mesopore structure (major pore diameter is 22 nm) than other
synthesized ZSM-5. This provides an accessible tunnel for diffusion of feedstock and products
molecular. The high Sext of MZSM-5-C allows exposing more active sites and helping reactants
better access to active sites [36]. Coke formation is restrained owing to the medium acidity of
catalyst as well as the diminishing secondary reaction (due to the low diffusion resistance of
product molecules). The suitable porous structure and acidic properties would account for the high
catalytic activity of the synthesized MZSM-5-C in ETO reaction.
3.2.2. Conversion of oleic acid into light olefins
In oleic acid to light olefins (OATO) reaction, oleic acid was catalytically cracked at
380−600 °C in N2 stream over the four ZSM-5 samples. Conversion and YLO are shown in Fig. 7.
The rapid conversion of oleic acid occurs in two temperature ranges: 380  400 °C and 450 
13

550 °C. At 380  400 °C, oleic acid rapidly evaporates, which is accompanied by the preliminary
cracking, decarbonylation and decarboxylation reaction. At 450  550 °C, cracking and
decarbonylation reactions of oleic acid amplify, evidenced by the release of a large of light olefins
and CO (Fig. 8). At a high temperature e.g. > 500 °C, the conversions on different catalysts are all
higher than 90%, indicating that a temperature higher than 500 °C is needed for effective cracking
of oleic acid. MZSM-5-B achieves a complete conversion at a temperature from 500 to 600 °C,
indicating the high activity of MZSM-5-B. Contrarily, the conversion over HZSM-5 reaches the
maximum value (98%) at 550 °C, and then drops at a higher temperature. This may be attributed to
the fact that the relative small pore size trends to cause fast deactivation of catalyst at a high
temperature [37]. Among the four tested ZSM-5 samples, MZSM-5-B achieved the highest
conversion and YLO (Fig. 7b), indicating that the porosity with main pore size about 16 nm would
be the most suitable for mass transfer of oleic acid molecular and providing reaction sites.
The gaseous product distribution from cracking of oleic acid at different temperature over
different ZSM-5 samples is presented in Fig. 8. For all ZSM-5 samples, SLO increase significantly
with raising reaction temperature, indicating that reaction temperature plays an important role to
determine product yield and selectivity. At 550 °C, three mesoporous ZSM-5 samples give the
maximum SLO, and the SLO for four ZSM-5 samples are in the order: MZSM-5-B (38.1%) >
MZSM-5-A (36.7%) > MZSM-5-C (36.2%) > HZSM-5 (31.3%). Conversion, YLO and SLO
considered, MZSM-5-B is the best catalyst for cracking of oleic acid into light olefins. Besides, it
is found that the P/E values are higher than 1.0, indicating that propylene is the main light olefin
product, which is different from that of ETO reaction. The P/E values decrease dramatically with
increasing reaction temperature, indicating that secondary decomposition of C3H6 could occur,
consequently leading to the formation of more C2H4 [38, 39].
14

3.2.3. Coke analysis
After cracking reaction, the used catalysts were collected, and calcined at 550 °C. The amount
of coke deposited on the catalysts was calculated from weight loss of the catalysts during
calcination. The results are shown in Table 5. As can be seen, for both ETO and OATO reactions,
the coke yields on the three synthesized ZSM-5 are lower than conventional HZSM-5, indicating
that the anti-coking ability of catalysts was markedly improved by introduction of mesopores into
zeolites [21, 40]. The TPO profiles of the used HZSM-5, MZSM-5-C and MZSM-5-B, collected
from cracking ethanol and oleic acid are shown in Fig. 9. As seen in Fig. 9(a), in comparison to
HZSM-5, MZSM-5-C releases less CO and CO2 (as evidenced by the weaker peaks), indicating
less coke deposited on MZSM-5-C. In the case of cracking of oleic acid (in Fig. 9b), the total peak
area of CO and CO2 from MZSM-5-B is much less than that from HZSM-5, indicating less coke
deposited on MZSM-5-B. The temperature corresponding to peak height (Tp) on TPO curve of
HZSM-5 is 545 °C, which is little lower than Tp on MZSM-5-B (560 °C). This indicates that more
thermo-oxidative-sensitive light compounds were deposited on HZSM-5. It is worth noting that
the CO peak on the TPO curve of HZSM-5 is much stronger than that of CO2, suggesting that the
small pore in traditional ZSM-5 could affect the diffusion of O2, consequently leading to
incomplete oxidation of coke. In addition, the catalysts derived from cracking of oleic acid
produce a much stronger oxidation peaks than that from ethanol, indicating that more coke was
formed on catalysts when large molecule oleic acid was used as feedstock. The results are also
coincident with the analysis on Ycoke as shown in Table 5.
The dissoluble components of coke from cracking of ethanol and oleic acid at 500 °C are
presented in Table 6. As can be seen, ethanol-derived coke mainly consists of hydrocarbons, while
the composition of oleic acid-derived coke is complicated, including hydrocarbons, aromatics,
15

esters and ketones. It is also found that some heavier coke compounds are easily formed when
oleic acid is used as feedstock compared to ethanol.
3.2.4. Liquid product analysis and cracking reaction mechanism
The liquid products from cracking of ethanol at 500 °C over MZSM-5-C mainly contain
o-xylene (28.5%), 2-methylpropyl-cyclopentane (19%) and 3, 5, 5-trimethyl-cyclohexene (16.2%).
Combined with the results of previous research, the reaction path for ethanol is concluded, as
shown in Scheme 1. Ethanol is converted to C2H4 directly via an intramolecular dehydration or
undergoes an intermolecular dehydration to form CH3CH2OCH2CH3, and CH3CH2OCH2CH3 is
then converted to C2H4. Ethanol can be converted to C3H6 via oligomerization reaction between
ethylene and carbene species, and converted to C4H8 via ethylene dimerization [41]. C3H6 and
C4H8 subsequently produce cycloparaffins and aromatics via cyclization, dehydrogenation and
aromatization reactions. The presences of o-xylene, 2-methylpropyl-cyclopentane and 3, 5,
5-trimethyl-cyclohexene in liquid product, as well as chain- and cyclo-paraffins in coke confirm
the above speculation.
Liquid products from cracking of oleic acid (Table 7) at 500 °C over MZSM-5-B contain
about 97.1% aromatic hydrocarbons and a small amount of esters. Combined with the current
reports, the reaction pathway of oleic acid is illustrated in Scheme 2. Breaking of CC bond in
oleic acid could occur at β position of C=C bond and α position of –COOH [7], which leads to the
formation of light olefins and short-chain paraffins with a simultaneous release of CO2 and CO.
Subsequently, these short-chain paraffins could be converted to light olefins via further cracking
reaction. Besides, those light olefins and paraffins could also produce aromatic hydrocarbons
through series further reactions in the pores of the ZSM-5 zeolites, such as cyclization,
16

dehydrogenation and aromatization [4]. H2 could be released from the above dehydrogenation and
aromatization reaction, which would react with CO to produce methanol. The resultant methanol
would subsequently react with oleic acid on the surface of catalysts (H+ site) to form esters.
Therefore, a lot of aromatic hydrocarbons are present in liquid products while a considerable
amount of ester compounds are present in oleic acid-derived coke.
3.3. Stability of the catalysts
The stability of the catalysts was determined by repeating the cracking and regeneration
experiments 10 times using ethanol and oleic acid over the optimized catalyst. The results are
shown in Fig. 10 and Fig. 11. As seen in Fig. 10, YLO and SLO from ethanol decrease by less than
10% after ten cycles. Meanwhile, C2H4 selectivity decreases from 42% to 36% after ten cycles.
For MZSM-5-B, as shown in Fig. 11(a), the YTGP after several cycles are higher than that of the
1
fresh catalyst, while YLO keeps at a constant value (200 mL·g ). Besides, the SLO, C3H6 and C2H4
selectivity decrease slightly with increasing the cycle number (Fig. 11(b)). These results show that
the synthesized catalyst MZSM-5-B and MZSM-5-C have good stability and reusability.
Based on the above study, it is clear that regardless of ETO and OATO reactions, the
synthesized ZSM-5 with mesopore structure showed a higher catalytic activity than conventional
ZSM-5. However, there are some differences in the selected optimum catalyst for cracking of
ethanol and oleic acid. MZSM-5-C and MZSM-B exhibited the best catalytic performance in ETO
and OATO reaction, respectively. The difference may be relevant to the size and chemical
structure of feedstock molecular as well as the acidity of catalysts. For cracking of large oleic acid
molecular, a catalyst with both large pore channel and moderate acid sites (B acid sites) is required.
The large pore channel is helpful for quick mass and heat diffusion, while the B acid sites are the
17

active sites for cracking and deoxygenation reaction. Therefore, MZSM-B is the most suitable for
OATO reaction.
Table 8 lists the comparison between our obtained results and those reported in current
literatures. In Wang’s research [23], the mesoporous ZSM-5 with pore size centered around 3.0 nm
showed the higher light olefin selectivity (40.9%) to that of micro-ZSM-5 (31.0%) in MTO
reaction. Compared to Wang’s research, MZSM-5-C (major pore diameter about 22 nm) in this
work achieved higher light olefin selectivity (47.2%), including C2H4 (42.3%). In Bi’s study [6],
ethanol as feedstock obtained a higher selectivity to C2H4. This is essentially in agreement with
our results. Furthermore, the light olefin yields obtained in present study are 0.32−0.34 kg of
olefins/kg of ethanol and oleic acid, namely 32−34 wt%, which are much higher than those from
cracking of oleic acid [26], bio-oil [38] and cellulose [39]. Besides, C3H6 selectivity of OATO
reaction in this study is 5.3% higher than Li’s study [26]. The high Si/Al2 ratio (> 200) led to the
formation of weak acid-dominant ZSM-5 zeolite in this study. The weak acidity and mesoporous
structure maybe help to restrain coking and thus prolong the lifetime of catalysts [23].
4. Conclusion
Three kinds of ZSM-5 zeolites with different pore structure (MZSM-5-A, MZSM-5-B and
MZSM-5-C) were successfully synthesized using a hydrothermal method by utilizing different
templates. The resultant ZSM-5 zeolites were used for production of light olefins from catalytic
cracking of ethanol and oleic acid. The synthesized ZSM-5 zeolites contained abundant mesopore,
and had higher external surface area and mesopore volume than conventional ZSM-5. In ETO
reaction, MZSM-5-C (major pore diameter about 22 nm) obtained the highest light olefin yield
(318.3 mL∙g-1) and C2H4 selectivity (42.3%) at 400 °C. In OATO reaction, MZSM-5-B (with major
18

pore diameter about 16 nm) achieved a complete conversion at 500 °C, and obtained the maximal
light olefin selectivity (38.1%) at 550 °C. The difference may be relevant to the size and chemical
structure of feedstock molecular as well as the acidity of catalysts.
Notes
The authors declare no competing financial interest.
Acknowledgements
The authors are grateful for supports from the National Natural Science Foundation of China
(21566004), the Guangxi Natural Science Foundation (2015GXNSFAA139036), the Scientific
Research Foundation of Guangxi University (XGZ120081, XTZ140787, XJPZ160713), and
2018–2019 “Guangxi University college students’ innovative and entrepreneurial training plan”
(No. 201810593185).
19

References
[1] Y. Wei, Z. Liu, G. Wang, Y. Qi, L. Xu, P. Xie, Y. He, Stud. Surf. Sci. Catal. 158 (2005)
12231230.
[2] Y. K. Park, C.W. Lee, N.Y. Kang, W.C. Choi, S. Choi, S.H. Oh, D.S. Park, Catal. Surv. Asia
14 (2010) 7584.
[3] G.W. Huber, S. Iborra, A. Corma, Chem. Rev. 106 (2006) 40444098.
[4] X. Dong, Z. Chen, S. Xue, J. Zhang, J. Zhou, Y. Liu, Y.P. Xun, Z.M. Liu, RSC Adv. 3 (2013)
2578025787.222
[5] M. Rostamizadeh, A. Taeb, J. Ind. Eng. Chem. 27 (2015) 297–306.
[6] J.D. Bi, M. Liu, C.S. Song, X.S. Wang, X.W. Guo, Appl. Catal. B: Environ. 107 (2011)
68–76.
[7] B. Periyasamy, Fuel 158 (2015) 479487.
[8] G. Bellussi, P. Pollesel, Stud. Surf. Sci. Catal. 158 (2005) 12011211.
[9] A.H. Tullo, Chem. Eng. News 82 (2004) 2024.
[10] H. Boelt, Proc. European Petrochemical Industry Forum, Amsterdam (2000) 13–14.
[11] R. Wei, C. Li, C. Yang, H. Shan, J. Energ. Chem. 20 (2011) 261–265.
[12] G.D. Pimgruber, K. Seshan, J.A. Lercher, J. Catal. 186 (1999) 188–200.
[13] M. Bjørgen, S. Svelle, F. Joensen, J. Catal. 249 (2007) 195–207.
[14] H.V. Xuan, S. Nguyen, T.T. Dang, B.M.Q. Phan, D.A. Nguyen, U. Armbruster, Catalysts 5
(2015) 1692–1703.
[15] C.T. Kresge, M.E. Leonowicz, W.J. Roth, J.S. Beck, Nature 359 (1992) 710–712.
[16] J. Čejka, S. Mintova, Catal. Rev. 49 (2007) 457–509.
20

[17] J. B. Koo, N. Jiang, S. Saravanamurugan, M. Bejblová, Z. Musilová, J. Čejka, S.E. Park, J.
Catal. 276 (2010) 327–334.
[18] A.E.W. Beers, J.A.V. Bokhoven, K.M. de Lathouder, F. Kapteijn, J.A. Moulijn, J. Catal. 218
(2003) 239–248.
[19] C.S. Mei, P.Y. Wen, Z.C. Liu, H.X. Liu, Y.D. Wang, W.M. Yang, Z.K. Xie, W.M. Hua, J. Catal.
258 (2008) 243–249.
[20] C.M. Zhang, Q. Liu, Z. Xu, K.S. Wang, Micropor. Mesopor. Mat. 62 (2003) 157–163.
[21] J. Ahmadpour, M. Taghizadeh, J. Catal. 23 (2015) 184–194.
[22] D.F. Wen, Q. Liu, Z.Y. Fei, Y.R. Yang, Z.X. Zhang, X. Chen, J.H. Tang, M.F. Cui, X. Qiao,
Ind. Eng. Chem. Res. 57 (2018) 446–455.
[23] Q.Y. Wang, S.T. Xu, J.R. Chen, Y.X. Wei, J.Z. Li, D. Fan, Z. Yu, RSC Adv. 4 (2014)
21497–21491.
[24] F. Yaripour, Z. Shariatinia, S. Sahebdelfar, A. Irandoukht, J. Nat. Gas Sci. Eng. 22 (2015)
260–269.
[25] Q. Zhang, S. Hu, L.L. Zhang, Z.J. Wu, Y.J. Gong, T. Dou, Green Chem. 16 (2013) 77–81.
[26] F.Li, S. Ding, Z. Wang, Z. Li, L. Li, C. Gao, Z. Zhong, H. Lin, C. Chen, Energy fuels 32
(2018) 5910–5922.
[27] D. Zhang, X. Li, S. Liu, X. Zhu, W. Xin, S. Xie, P. Zeng, L. Xu, Chin. J. Catal. 32 (2011)
1747–1754.
[28] T. Xue, H.P. Liu, Y. Zhang, H.H. Wu, P. Wu, M.Y. He, Micropor. Mesopor. Mat. 242 (2017)
190–199.
[29] K. Na, C. Jo, J. Kim, K. Cho, J. Jung, Y. Seo, R.J. Messinger, B.F. Chmelka, R. Ryoo, Science
333 (2011) 328–332.
21

[30] V. Valtchev, S. Mintova, Micropor. Mesopor. Mat. 43 (2001) 41–49.
[31] Q. Liu, D.F. Wen, Y.R. Yang, Z.Y. Fei, Z.X. Zhang, X. Chen, J.H. Tang, M.F. Cui, X. Qiao,
Appl. Surf. Sci. 435 (2017) 945–952.
[32] J. Shi, G. Lu, G. Cao, Sino-Global Energy 15 (2010) 66–74, in Chinese.
[33] F.J. Huang, D. Mao, X. Zhai, Q. Guo, Appl. Catal. A: Gen. 520 (2016) 99–104.
[34] C.S. Triantafillidis, A.G. Vlessidis, L. Nalbandian, N.P. Evmiridis, Micropor. Mesopor. Mat.
47 (2001) 369–388.
[35] A. Takahashi, X. Wei, Q. Wu, T. Furukawa, I. Nakamura, H. Shimada, Appl. Catal. A: Gen.
467 (2013) 380–385.
[36] W. J. Roth, P. Nachtigall, R. E. Morris, J. Čejka, Chem. Rev. 114(2014) 4807−4837.
[37] J.W. Jun, Z. Hasan, C.U. Kim, S.Y. Jeong, S.H. Jhung, J. Nanosci. Nanotechnol. 16 (2016)
4377–4385.
[38] F.Y. Gong, Z. Yang, C.G. Hong, W.W. Huang, S. Ning, Z.X. Zhang, Y. Xu, Q.X. Li,
Bioresour. Technol. 102 (2011) 92479254.
[39] S. Zhang, M. Yang, J. Shao, H. Yang, K. Zeng, Y. Chen, J. Luo, F.A. Aqblevor, H. Chen, Sci.
Total Environ. 628 (2018) 350–357.
[40] Y. Tao, H. Kanoh, L. Abrams, K. Kaneko, Cheminform 37 (2010) 896–910.
[41] A. Morschbacker, Polym. Rev. 49 (2009) 79–84.
22

Table 1 Lattice parameters of four different ZSM-5 zeolites.
Crystal sizea
(nm)
Lattice constants (nm)
Relative
Samples
Vcell (nm3)
crystallinity (%)
a
b
c
HZSM-5
100
85.8
71.1
54.7
2.003
2.004
2.004
2.004
1.995
1.983
1.984
1.982
1.328
1.331
1.336
1.328
5.306
5.291
5.310
5.276
89
56
69
37
MZSM-5-A
MZSM-5-B
MZSM-5-C
aThe crystal size was calculated by using Scherrer equation: D=Kλ/(Bcosθ), where K is a constant, λ is X ray
wavelength, B is half width and θ is diffraction angle (2θ=23.2 º).
23

Table 2 Textural properties of four different ZSM-5 zeolites.
e
a
b
b
c
b
d
Daver
SBET
Smicro
Sext
Vtotal
Vmicro
Vmeso
Samples
(m2∙g-1)
(m2∙g-1)
258
188
132
0
(m2∙g-1)
(cm3∙g-1)
(cm3∙g-1)
0.13
0.09
0.06
0
(cm3∙g-1)
(nm)
3.1
4.0
4.5
7.5
HZSM-5
459
201
0.35
0.22
MZSM-5-A
MZSM-5-B
MZSM-5-C
439
251
0.40
0.31
469
337
0.52
0.46
405
405
0.76
0.76
aSBET (BET surface area) was obtained from the adsorption isotherm.
bSmicro (micropore surface areas) and Sext (external surface areas) and Vmicro (micropore volumes) were
calculated using t-plot method.
cVtotal (total pore volumes) was obtained at the relative pressure (p/p0) = 0.99.
dVmeso = Vtotal - Vmicro.
eDaver (average pore diameter) was obtained from average pore size.
24

Table 3 Acidity properties from NH3-TPD of four catalysts.
n(SiO2)/
n(Al2O3)a
Total acid
(mmol/g)
Percentage of acid sites in total acidity (%)
Samples
Strong/weak
Weak acid
69.2
Strong acid
30.8
HZSM-5
227
222
212
226
0.52
0.53
0.56
0.51
0.44
0.77
0.47
0.28
MZSM-5-A
MZSM-5-B
MZSM-5-C
56.6
43.5
67.7
32.1
78.4
21.6
aSiO2/(Al2O3) molar ratio of the four ZSM-5 zeolites was determined by ICP-OES.
25

Table 4 Quantitative analysis of acidic sites types of ZSM-5 samples by Py-IR spectra.
150 °C desorption
300 °C desorption
Samples
B(mmol/g) L (mmol/g)
B/L
0.05
0.06
0.11
0.07
B(mmol/g) L (mmol/g)
B/L
0
HZSM-5
0.01
0.01
0.02
0.01
0.21
0.18
0.18
0.14
0
0
0.03
0.04
0.04
0.03
MZSM-5-A
MZSM-5-B
MZSM-5-C
0
0.01
0
0.25
0
B–Brönsted acid amount; L–Lewis acid amount
26

Table 5 Coke yield (Ycoke) from catalytic cracking of ethanol and oleic acid over the
synthesized ZSM-5 at 500 °C, WHSV=4.5h-1.a
Ycoke (%)
Samples
Ethanol
Oleic acid
1.17 ± 0.03
0.66 ± 0.08
0.90 ± 0.02
0.82 ± 0.03
HZSM-5
0.73 ± 0.04
0.28 ± 0.06
0.16 ± 0.03
0.70 ± 0.04
MZSM-5-A
MZSM-5-B
MZSM-5-C
aThe data is reported as mean values ± standard deviation, n = 3.
27

Table 6 Components of soluble coke formed from catalytic cracking of ethanol and oleic acid at
500 °C.
Area %
Name of compound
Molecular formula Structural formula
HZSM-5 MZSM-5-C
Catalytic cracking of ethanol
Octane, 3,5-dimethyl-
Nonane, 3,7-dimethyl-
Cyclododecane
25.23
8.33
̶
̶
C10H22
C11H24
8.70
34.77
13.04
̶
C12H24
Decane, 2,3,5-trimethyl-
1-Pentadecene
16.67
8.14
41.64
C13H28
C15H30
Others
No hit compound
43.48
HZSM-5 MZSM-5-B
Catalytic cracking of oleic acid
Methyl octyl ketone
C10H20O
21.62
̶
O
Octane, 3,4,5,6-tetramethyl-
Benzene, undecyl-
C12H26
C17H28
C18H36O
̶
10.72
27.03
̶
Methyl n-hexadecyl ketone
10.81
5.41
6.93
O
O
Heptadecanoic acid, methyl ester C18H36O2
̶
O
O
Octadecanoic acid, methyl ester
Others
C19H38O2
̶
47.06
35.29
O
No hit compound
35.13
28

Table 7 Compounds in liquid products from catalytic cracking of oleic acid over MZSM-5-B.
Sample (area%)
Molecular
formula
NO.
RT (min)
4.980
Name of compound
Oleic acid
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
benzene
C6H6
3.5
8.1
6.4
21.5
6.3
3.4
5.0
3.9
4.6
2.4
1.1
3.3
4.4
5.8
2.2
3.9
5.0
1.4
1.9
1.8
1.4
2.7
5.939
8.291
8.665
9.128
Toluene
C7H8
Ethylbenzene
p-Xylene
o-Xylene
Benzene, 1,2,3-trimethyl-
Benzene, 1-ethyl-2-methyl-
Indane
C8H10
C8H10
C8H10
C9H12
C9H12
C9H10
C9H10
C10H12
C10H12
C10H10
C10H10
C10H8
C11H14
C10H12
C11H10
C12H12
C12H12
C16H24
C18H28
C19H34O2
10.732
11.487
12.347
12.550
13.253
14.361
14.550
14.695
15.282
15.941
16.534
17.120
18.769
18.939
24.978
25.588
26.677
Indene
(E)-1-Phenyl-1-butene
1H-Indene, 2,3-dihydro-5-methyl-
1H-Indene, 1-methyl-
2-Methylindene
Naphthalene
Benzene, (1,2-dimethyl-1-propenyl)-
Naphthalene, 1,2,3,4-tetrahydro-
Naphthalene, 1-methyl-
Naphthalene, 2,6-dimethyl-
Naphthalene, 2,3-dimethyl-
Indane, 1-nonyl-
Naphthalene, 1-hepty-1,2,3,3-tetrahydro
9,12-Octadecadiennoic acid, methyl ester
29