Selective formation of light olefins from dimethyl ether over MCM-68 modified with phosphate species

Article abstract of DOI:10.1016/j.cattod.2015.08.061

To control the acid sites distribution in MCM-68 zeolite catalyst, phosphate-loaded MCM-68 catalyst was prepared by impregnation with aqueous solution of (NH₄)₂HPO₄ onto MCM-68 and subsequent calcination. Phosphate impregnation was effective for increasing the selectivity to propylene and butylenes. The propylene/ethylene ratio in the products was significantly increased by phosphate modification. The considerable enhancement of the yields of propylene and butylenes could be attributed to diminution of stronger acid sites, especially the acid sites distributed on the external surface of MCM-68 catalyst. Furthermore, phosphate-modified MCM-68 showed high resistance to coke formation, indicating this modification technique is useful to obtain long-lived catalysts in DTO reaction.

Full text of DOI:10.1016/j.cattod.2015.08.061

Catalysis Today 265 (2016) 218–224
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Selective formation of light olefins from dimethyl ether over MCM-68
modified with phosphate species
Sungsik Park, Satoshi Inagaki, Yoshihiro Kubota∗
Division of Materials Science and Chemical Engineering, Yokohama National University, 79-5 Tokiwadai, Hodogaya-ku, Yokohama 240-8501, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 7 June 2015
Received in revised form 12 August 2015
Accepted 26 August 2015
Available online 1 December 2015
To control the acid sites distribution in MCM-68 zeolite catalyst, phosphate-loaded MCM-68 catalyst was
prepared by impregnation with aqueous solution of (NH4)2HPO4 onto MCM-68 and subsequent calcina-
tion. Phosphate impregnation was effective for increasing the selectivity to propylene and butylenes.
The propylene/ethylene ratio in the products was significantly increased by phosphate modification. The
considerable enhancement of the yields of propylene and butylenes could be attributed to diminution
of stronger acid sites, especially the acid sites distributed on the external surface of MCM-68 catalyst.
Furthermore, phosphate-modified MCM-68 showed high resistance to coke formation, indicating this
modification technique is useful to obtain long-lived catalysts in DTO reaction.
Keywords:
MCM-68
Phosphate-modification
DTO
© 2015 Elsevier B.V. All rights reserved.
Propylene
Coking-resistance
1. Introduction
the most feasible because they have huge possibility to improve a
propylene yield by tuning of the catalysts [12,13].
Catalytic conversion of methanol or dimethyl ether into light
olefins, especially propylene, has attracted huge attention because
the world-wide demands of propylene as well as ethylene are
steadily increasing both now and in the near future [1]. As ther-
mal cracking of ethane, supplied from fossil oil, natural gas or shale
gas, gives a high selectivity to ethylene (more than 80 wt%) under
the thermodynamic limitation [2], it satisfies ethylene demand but
not propylene [3]. Therefore, the alternative processes for the pro-
duction of propylene are instantly required. Propylene, which can
be converted into the important and useful chemicals such as pro-
pylene oxide, acrolein, acrylic acid and related polymers, has been
mainly produced through thermal cracking or fluid catalytic crack-
ing (FCC) of naphtha [4]. Recently, on-purpose technologies, such
as methanol-to-olefin (MTO), and dimethyl ether-to-olefin (DTO)
reactions [5–8] as well as olefin interconversion [9], metathesis of
ethylene and butylene [10] and dehydrogenation of propane [11]
have been developed for the production of propylene. Among them,
the DTO and MTO processes, which can apply natural gas and other
non-petroleum feedstocks instead of fossil oils as a resource, are
During the past decades, a variety of solid acid catalysts have
been explored as possible DTO and MTO catalysts and it is acknowl-
edged that the performance of catalysts depends strongly on both
the framework structure and their acidic characteristics [13–18].
ZSM-5 zeolite is the first commercial catalyst used for methanol-
to-gasoline (MTG) process and the most widely studied catalysts
for the DTO and MTO reaction due to its high thermal stabil-
ity and higher resistance to coke deposition [15]. The control of
product selectivity has been the main issue for the MTO process.
Various methods such as incorporation of metallic (Zr [19], Ce
[20], or Ca [21,22]) and/or non-metallic promoters (B [22] or P
[22–30]) into ZSM-5 have been suggested to improve the selec-
tivity to light olefins on ZSM-5 catalysts. Phosphate modification is
known as a feasible method to adjust the acidity of zeolite catalysts
[23–30]. Particularly, modification of zeolite catalysts by phosphate
impregnation has been proven to be effective to change the prod-
uct selectivity in MTG reaction over ZSM-5 [23–30]. Jang et al.
suggested that the modification of ZSM-5 with phosphate species
neutralized its strong acid sites, resulting in a high selectivity to
light olefins by limiting the further reactions from the light olefins
to saturated aliphatic and aromatics [28], while Vedrin et al. sug-
gested that the changes in selectivity were mainly assigned to the
narrowing of pore size due to phosphate compounds bonding to
the zeolite framework, rather than to changes in acid strength [30].
MCM-68 is a new type of three-dimensional zeolite frame-
work with a 12 × 10 × 10-ring channel system [31]. This framework
∗
Corresponding author at: Division of Materials Science and Chemical Engi-
neering, Yokohama National University, 79-5 Tokiwadai, Hodogaya-ku, Yokohama
240-8501, Japan.
E-mail address: kubota@ynu.ac.jp (Y. Kubota).
http://dx.doi.org/10.1016/j.cattod.2015.08.061
0920-5861/© 2015 Elsevier B.V. All rights reserved.

S. Park et al. / Catalysis Today 265 (2016) 218–224
219
(MSE as the framework type code) [32] has a characteristic struc-
2.2. Characterization
ture in which a straight 12-ring channel intersects with two
inter-connected, tortuous 10-ring channels and possesses an 18-
ring × 12-ring supercage which is accessible only through 10-ring
channels [31,32]. MCM-68 is known to prominent candidates for
designing novel catalysts due to their availability in a wide range of
Si/Al ratios [33]. Such MCM-68 zeolites exhibit unique acid catalytic
properties and are potentially useful as shape-selective catalysts for
the alkylation of aromatics [34–36], as well as for the production
of propylene in hexane-cracking [37,38]. Their use as hydrocarbon
traps has also been reported [39]. In addition, the post-synthetic
isomorphous substitution of Ti for Al in the MSE framework extends
their application as the catalyst for the oxidation of phenol and
olefins with H2O2 as an oxidant [40,41].
As mentioned above, dealumination, one of the post-synthetic
techniques, is practicable to vary the content of Al present in the
framework of MCM-68. Due to the reduced density of acid sites,
such MCM-68 exhibited significantly enhanced selectivity to pro-
pylene with high propylene-to-ethylene (P/E) ratio in DTO reaction
[33,42]. In this work, a phosphate-modified MCM-68 was pre-
pared by (NH4)2HPO4 impregnation. We focused on the changes
in the acidity of MCM-68 and the corresponding changes in cat-
alytic selectivity of the products in the reaction of dimethyl ether
conversion after phosphate modification.
Crystal structures of the obtained solid products were deter-
mined by X-ray powder diffraction (XRD) on an Ultima IV (Rigaku)
using Cu K␣ radiation at 40 kV and 20 mA. The textural proper-
ties of the catalysts were examined by N2 adsorption–desorption
isotherm measurement at −196 ◦C with an Autosorb-iQ analyzer
(Quantachrome Instruments). Before the measurements, all the
samples were evacuated at 400 ◦C for 4 h. Micropore volumes of
the catalysts were calculated from the adsorption isotherm by the t-
plot method. Specific surface areas of the catalysts were calculated
from the adsorption isotherm by the Brunauer–Emmett–Teller
(BET) equation. (For a surface area evaluation, data in the relative
pressure range of 0.05–0.10 are used.) The chemical composi-
tions of zeolites were determined by using inductively coupled
plasma atomic emission spectrometer (ICP-AES, ICPE-9000, Shi-
madzu). The shape and the particle size of the MCM-68 zeolites
were observed by a scanning electron microscope (JSM-7001F,
JEOL).
The properties of acid sites on catalysts were measured by
ammonia temperature-programmed desorption (NH3-TPD) mea-
surement on a BELCAT-B (MicrotracBEL Corp.) equipped with a
thermal conductivity detector (TCD). Before the measurements, the
catalysts (50 mg) charged in a quartz-tube were preheated at 600 ◦C
prior to the measurement under He flow. The TPD data were col-
lected at a ramping rate of 10 ◦C min−1. The number of acid sites
was determined from the area of h-peak [43] in their profiles. For
Fourier transform infrared (FT-IR) spectroscopy analysis (FT/IR-
6100, JASCO), self-supporting zeolite wafer (20–30 mg) was located
between NaCl windows in a cylindrical cell similar to that described
in ref [44 ]. The sample was then heated to 550 ◦C at 5 ◦C min−1
under vacuum, held for 1 h, and cooled to 100 ◦C prior to the adsorp-
tion experiments. Pyridine (ca. 2.7 kPa) was injected into the cell.
The cell was left in vacuum for 10 min to allow physically adsorbed
pyridine to desorb. Spectra were then recorded and averaged over
32 scans between 450 and 4000 cm−1 with 4 cm−1 resolution. The
temperature of the IR cell was progressively increased from 100
to 400 ◦C and the spectrum was recorded at 100, 150, 200, 250,
300, 350 and 400 ◦C. The spectrum was recorded at least two times
for each temperature (5-min interval). The coke contents of the
used catalysts were determined in a thermogravimetirc analyzer
(TG-8120, Rigaku). The temperature was raised from room temper-
ature to 800 ◦C with the rate of 10 ◦C min−1 under air flow (30 cm3
(N.T.P.) min−1).
2. Experimental
2.1. Catalyst preparation
For the synthesis of MCM-68 with Si/Al molar ratio of 10, col-
loidal silica (Ludox HS-40, DuPont, 40 wt% SiO2), de-ionized water
and Al(OH)3 (Pfaltz & Bauer) were mixed for 10 min. Aqueous KOH
solution (5.93 mmol g−1) was added to the solution, and stirred
for further 30 min. Then, N,N,Nꢀ,Nꢀ-tetraethylbicyclo[2.2.2]oct-7-
ene-2,3:5,6-dipyrrolidinium diiodide (TEBOP2+ (I–)2) was added
as a structure-directing agent (SDA), and the mixture was stirred
for another 4 h. The resulting mixture with a molar composi-
tion 1.0SiO2–0.1TEBOP2+(I–)2–0.375KOH–0.1Al(OH)3–30H2O was
taken into a Teflon-lined autoclave, and kept statically at 160 ◦C
for 16 d in a convection oven. After quenching the autoclave in an
ice bath for 30 min, the solid part was separated by centrifugation,
washed several times with de-ionized water until the pH value
of the decanted water reached around 7, and dried overnight at
100 ◦C. The as-synthesized MCM-68 zeolite was calcined at 650 ◦C
for 10 h to eliminate SDA. After the calcination, the potassium
+
form of MCM-68 was converted into the NH4 form by exchanges
2.3. Catalytic reactions
in 0.5 mol L−1 aqueous solution of NH4NO3 at 80 ◦C for 12 h, fol-
lowed by filtration and washing. This procedure was repeated four
Dimethyl ether (partial pressure: 5.0 kPa) was introduced into
the top of the reactor (a down-flow quartz-tube microreactor with
a 9-mm inner-diameter) with He (40 cm3 (N.T.P.) min−1). Each zeo-
lite catalyst was pelletized without any binder, roughly crushed and
then sieved to obtain catalyst pellets with 500–600 ␮m in size. Prior
to running the reaction, 100 mg of catalyst pellets were placed in
the fixed bed of the reactor. The temperature of electric furnace was
raised to the pretreatment temperature with the rate of 10 ◦C min−1
under air flow (40 cm3 (N.T.P.) min−1). After pretreatment at 550 ◦C
for 1 h, the temperature was adjusted to reaction temperature
under He flow (40 cm3 (N.T.P.) min−1). The reactants and products
were analyzed on DB-5 capillary column (i.d. 0.53 mm; length 60 m;
thickness of the stationary phase 5.00 ␮m; Agilent Technology) and
an HP-PLOT/Q capillary column (i.d. 0.53 mm; length 30 m; thick-
ness of the stationary phase 40.0 ␮m; Agilent Technology) using
a GC-2014 (Shimadzu) with a flame ionization detector (FID). The
conversion of dimethyl ether, the selectivity of the products, the
yield of the products and material balance were calculated on the
carbon-basis of the inlet amount of dimethyl ether.
times and NH4 form of MCM-68 was dried at 100 ◦C for 12 h
+
and calcined at 550 ◦C for 4 h to obtain the proton form of MCM-
68.
MCM-68 zeolite with Si/Al molar ratio of 60 was prepared by
treating of potassium form of MCM-68(10) with aqueous HNO3
solution (2 mol L−1, 100 mL of solution for 1 g of zeolite) at 80 ◦C for
2 h. After cooling to room temperature, the solid part was separated
by centrifugation, washed several times with de-ionized water, and
dried overnight at 100 ◦C.
The phosphorus loaded MCM-68 zeolites were prepared from
proton forms of MCM-68 with Si/Al molar ratios of 10 and 60. For
the phosphorus impregnation, 0.5 g of zeolites were impregnated
in aqueous solution of (NH4)2HPO4 (0.21 mol L−1, 1.0 mL) at room
temperature for 30 min. Then, phosphate-modified MCM-68 was
dried at 100 ◦C for 12 h and calcined at 550 ◦C for 4 h to obtain
the proton form of MCM-68. In this paper, the phosphate-modified
MCM-68(60) is designated as P/MCM-68(60). The number in paren-
theses is the Si/Al molar ratio.

220
S. Park et al. / Catalysis Today 265 (2016) 218–224
and phosphate impregnation treatments (Fig. 1). As shown in SEM
images in Fig. 2, the particle size of ion-exchanged MCM-68 zeo-
lites was about 100 nm. The SEM images of dealuminated (Fig. 1b)
and further phosphate-modified MCM-68 (Fig. 1c) indicate that
the particle size and the morphology did not change with acid
treatment and further phosphate treatment. The specific surface
area and micropore volume are listed in Table 1. The shapes of
N2 adsorption–desorption isotherms of MCM-68(60) and P/MCM-
68(60) were very similar (typical type-I isotherms), indicating that
phosphate impregnation did not significantly change the micro-
pore feature of MCM-68 (see Fig. S3). On the other hand, phosphate
impregnation reduced the both micropore volume and specific sur-
face area. Temperature programmed desorption (TPD) profiles of
NH3 of MCM-68 indicate two peaks at around 150–200 ◦C (l-peak)
and 350–450 ◦C (h-peak) as shown in Fig. 3. The area of the h-peak
corresponding to the amount of acid sites did not change signifi-
cantly during phosphate impregnation (Table 1).
The decrease in the micropore volume and the specific surface
area after phosphate impregnation is probably due to the partial
pore mouth blockage. Although NH3-TPD profiles cannot distin-
guish between Brønsted and Lewis acidity, the decrease in the area
of h-peak after phosphate impregnation indicates that phosphate
compounds are incorporated into MCM-68 by interacting with the
acid sites.
Fig. 1. XRD patterns of (a) MCM-68(10), (b) MCM-68(60), (c) P/MCM-68(60) and
(d) MCM-68(84). The phosphate-loaded samples are designated as P/MCM-68. The
number in parentheses is Si/Al molar ratio.
The cracking of cumene or 1,3,5-triisopropylbenzene (TIPB) was
performed at 300 ◦C under atmospheric pressure in a pulse-type
quartz-tube (i.d. 4 mm) microreactor (see Fig. S1) under a stream of
helium (30 cm3 (N.T.P.) min−1). The catalyst amount, dose amounts
of cumene and TIPB are 20 mg, 0.8 ␮L, and 0.6 ␮L, respectively.
The products were analyzed by TCD on GC equipped with stain-
less steel column (i.d. 3.0 mm; length 6.0 m) packed with Silicone
OV-1 (60–80 mesh).
3.2. FT-IR of pyridine adsorbed on MCM-68 zeolites
The FT-IR spectra of the MCM-68 zeolites activated at 500 ◦C for
1 h under vacuum and cooled down at 100 ◦C are shown in Fig. S4.
The ranges of 3500–4000 cm−1 in the FT-IR spectra were focused
on for detecting surface hydroxyl groups. FT-IR peaks at ca. 3789,
3747, 3734, 3667 and 3626 were observed for the MCM-68(10) in
agreement with the findings of B. Gil et al. [45]. As can be observed
in Fig. S4, the differences between the FT-IR spectra of MCM-68(10)
and MCM-68(60) were obviously seen, while the FT-IR spectra of
3. Results and discussion
3.1. Effect of phosphorus on the structure and acidity of MCM-68
The XRD patterns of dealuminated MCM-68(60) showed that
each sample maintained the MSE structure during dealumination
Fig. 2. Typical SEM images of (a) MCM-68(10), (b) MCM-68(60) and (c) P/MCM-68(60).

S. Park et al. / Catalysis Today 265 (2016) 218–224
221
Table 1
Physicochemical properties of catalysts used in this work.
Catalyst (Si/Al)a
P/Ala
Specific surface
Micropore volume
Number of acid sites
area (m2 g−1
)
(cm3 g−1
)
(mmol g−1
)
b
c
d
MCM-68(60)
P/MCM-68(60)
MCM-68(84)
nil
1.9
nil
498
416
501
0.182
0.147
0.186
0.208
0.187
0.142
a
Bulk Si/Al and P/Al molar ratios were determined by ICP-AES analysis.
b
c
Specific surface areas of the catalysts were calculated using the Brunauer–Emmett–Teller (BET) equation on the N2 adsorption isotherms.
Micropore volumes of the catalysts were calculated by the t-plot method based on the adsorption isotherms.
The number of acid sites was estimated by NH3-TPD using TCD detector.
d
MCM-68(60) and P/MCM-68(60) in the region of 3500–4000 cm−1
did not change. The peak at 3747 cm−1 is attributed to the O
H
vibrations in the Si(OH) groups on the external surface and that of
at 3734 cm−1 is attributed to those in the Si(OH) inside a site defect
[45 ]. The removal of Al atoms in the framework of MCM-68(10) by
dealumination distinctly reduced the intensity of the peak which is
due to O H vibrations in Si(OH)Al (3626 cm−1), while the intensity
of the peak which is due to O H vibrations in Si(OH) (3728 cm−1
increased.
)
Fig. 4 shows FT-IR spectra in the region of 1400–1600 cm−1
after adsorption of pyridine on MCM-68(60) and P/MCM-68(60).
FT-IR peaks at 1543 and 1454 cm−1, which are attributed to pyri-
dinium ions (pyridine chemisorbed on Brønsted acid sites) and
pyridine interacting with Lewis acid sites, were observed. These
results indicate that both Brønsted and Lewis acid sites exist on
MCM-68(60) and P/MCM-68(60). With the increase in tempera-
ture, the intensity of the peaks at 1543 and 1454 cm−1 decreased.
In the case of P/MCM-68(60), the intensity of the peak at 1454 cm−1
significantly decreased with the increase in temperature and dis-
appeared at the temperature of 400 ◦C, while the corresponding
peaks were still detected over MCM-68(60). Although the obser-
vation of temperature dependence of pyridine adsorbed on the
Brønsted and Lewis acid sites could give us the information related
to the relative strength of Brønsted and Lewis acid sites, the quan-
titative determination of the number of Brønsted and Lewis acid
sites on MCM-68 seems inapplicable because the molar extinc-
tion coefficients over MCM-68 have not been investigated, to our
best knowledge. Despite we have not been quantifying the num-
ber of strong or weak Brønsted and Lewis acid sites on MCM-68,
the results of IR experiments indicate that the strength of Lewis
acid sites on P/MCM-68(60) can be expected to be much weaker
than that on the MCM-68(60) since the peaks at 1454 cm−1 due to
the pyridine interacted with Lewis acid sites almost disappeared
or decreased readily with increasing temperatures. These results
Fig. 3. NH3-TPD profiles of (a) P/MCM-68(60), (b) MCM-68(60) and (c) MCM-68(84).
Fig. 4. IR spectra of pyridine adsorbed on (a) MCM-68(60) and (b) P/MCM-68(60).
The upper curves are related to a low temperature and the bottom curves are related
to a high temperature during IR measurements. Pretreatment conditions: temper-
ature, 550 ◦C; period, 1 h under vacuum.
Fig. 5. Catalytic activity in the cracking of 1,3,5-triisopropylbenzene (TIPB) or
cumene over (᭹) MCM-68(60) and (ᮀ) P/MCM-68(60) at 300 ◦C. Reaction condi-
tions: catalyst, 20 mg; temperature, 300 ◦C; Pretreatment conditions: temperature,
550 ◦C; period, 1 h; air flow rate, 40 cm3 (N.T.P.) min−1
.

222
S. Park et al. / Catalysis Today 265 (2016) 218–224
indicate that the phosphate-modification on MCM-68 decreased
the number of strong Lewis acid sites by covering them.
thus evaluates the acid sites on both the external and internal sur-
faces of MCM-68 catalysts [37]. The initial TIPB conversion over
MCM-68(60) was higher than that of P/MCM-68(60) and its cat-
alytic activity maintained above 40% after 8 pulses of TIPB. In the
case of P/MCM-68(60), the catalytic activity for the conversion of
TIPB was very low from the first reaction pulse and then dropped
rapidly to less than 1% at the third pulses of TIPB. In contrast, in
terms of the cumene conversion, the initial activity of both parent
and phosphate-modified MCM-68(60) was high (approximately
93%) and the catalytic activity for cumene conversion over MCM-
68(60) and P/MCM-68(60) was maintained almost constant.
The higher catalytic activity of parent MCM-68(60) in the
TIPB conversion could be attributed to the large amount of
acid sites on the external surface. The low catalytic activity of
3.3. Catalytic cracking of 1,3,5-triisopropylbenzene and cumene
on MCM-68 zeolites
Fig. 5 shows the results of the catalytic cracking of 1,3,5-
triisopropylbenzene (TIPB) and cumene on parent and P/MCM-
68(60) at 300 ◦C. Catalytic cracking of TIPB occurs only at acid sites
on the external surface of MCM-68 because TIPB molecules are
much larger than the pore diameters of the 12R and 10R micro-
pores within the MSE framework. In contrast, cumene molecules,
which are much smaller than TIPB molecules, can penetrate the 10R
micropores in the MSE framework. Catalytic cracking of cumene
Fig. 6. Dimethyl ether conversion and product distributions in the DTO reactions over (a, d) MCM-68(60), (b, e) P/MCM-68(60) and (c, f) MCM-68(84). Reaction conditions:
catalyst, (a–c) 100 mg and (d–f) 50 mg; temperature, 400 ◦C; He flow rate, 40 cm3 (N.T.P.) min−1. Pretreatment conditions: temperature, 550 ◦C; period, 1 h; air flow rate,
40 cm3 (N.T.P.) min−1
.

S. Park et al. / Catalysis Today 265 (2016) 218–224
223
Table 2
DTO reactions over MCM-68 and phosphate-loaded MCM-68 (time on stream = 5 min).a
Catalyst [Si/Al]b
Conv. [%]c
Product distribution [C-%]d
M.B. [C-%]e
P/Ef
Content of coke
g
[mg-coke (g-cat)−1
]
MeOH
C1 + C2 + C3
C2=
C3=
C4
C4=
C5 + C5=
≥C6
Aromatics
MCM-68(60)
P/MCM-68(60)
MCM-68(84)
96.2
96.7
93.1
4.0
6.7
4.7
1.3
0.3
0.8
2.2
1.0
2.6
37.2
43.4
40.8
5.9
1.1
3.4
21.4
25.4
20.9
11.4
10.5
12.7
12.2
9.5
10.0
4.3
2.1
4.2
83.7
95.7
85.8
16
42
16
56
23
39
a
Reaction conditions: catalyst weight, 50 mg; W/F = 10.0 g-cat h mol−1; pellet size, 500–600 ␮m; He gas flow rate, 40 cm3 (N.T.P.) min−1; partial pressure of DME = 4.9 kPa;
reaction temperature, 400 ◦C. Pretreatment conditions: 550 ◦C, 1 h, air flow rate, 40 cm3 (N.T.P.) min−1
.
b
Values in parentheses are Si/Al ratios determined by ICP-AES analysis.
c
DME conversion = {1 − (C-atoms of DMEoutput)/(C-atoms of DMEinput)} × 100.
d
Product yield = {(C-atoms of the product)/(C-atoms of DMEinput − C-atoms of DMEoutput)} × {(DME conv. (C-%)}.
Material balance = (Total C-atoms of products and DMEoutput)/(C-atoms of DMEinput) × 100.
e
f
Propylene/ethylene molar ratio in the products.
g
Catalysts used in DTO reaction for 305 min were used in thermogravimetric analysis. The weight loss observed from 300 to 700 ◦C in a thermogravimetirc analysis was
regarded as coke contents.
phosphate-modified MCM-68(60) in the TIPB conversion reaction
indicates that the number of acid sites on the external surface
of P/MCM-68(60) is lower than that of MCM-68(60). Despite the
low catalytic activity of P/MCM-68(60) in TIPB cracking reaction, it
showed approximately similar catalytic activity in cumene crack-
ing reaction to that of parent MCM-68(60). These results indicate
selective covering of external acid sites in MCM-68(60) during
phosphate modification.
P/MCM-68(60), the increase in catalyst loading increased the yield
of propylene from ca. 40 to ca. 50 C-atom %, while that of aro-
matics and ethylene did not drastically change (Fig. 6b and e). In
order to clarify the effect of treatment with phosphate on acid-
ity of MCM-68, MCM-68 with the Si/Al molar ratio of 84, which
possesses the lesser amount of acid sites than MCM-68(60), was
applied in DTO reaction at the same reaction conditions (shown in
Fig. 6c and f). When catalyzed by MCM-68(84), the increase in the
yield of propylene and butylenes and the subsequent decrease in
the yield of aromatics and paraffins were observed. This is in good
agreement with our previous research reporting a catalytic perfor-
mance of MCM-68 with various Si/Al molar ratios [33,42]. On the
other hand, there is slight difference in product distributions, that
is, the reaction over phosphate-modified P/MCM-68(60) resulted
in higher yields of propylene and butylene compared to that over
unmodified MCM-68. Furthermore, the propylene/ethylene (P/E)
ratio in products significantly increased with phosphate modifica-
tion (from 16 to 42). These results indicate that a certain active sites,
which may be responsible for the formation of aromatics (especially
pentamethylbenezene and hexamethylbenzene) and ethylene,
decreased by phosphate impregnation on MCM-68(60). One of the
most undesirable side reactions in the DTO reaction is olefin aroma-
tization. The P/MCM-68(60) showed decreased aromatic selectivity
compared with MCM-68(60) and MCM-68(84), probably due to a
decrease in the number of strong Lewis acid sites, which is known
to catalyze aromatization and hydrogen transfer reactions.
The spent catalysts were characterized by TG/DTA to deter-
mine their coke contents. A distinct weight loss, accompanied by
an exothermic DTA peak, was observed at the temperature range
of around 300–700 ◦C. This is mainly related to combustion of the
carbonaceous material deposited on the catalysts during DTO reac-
tions. According to the results from TG/DTA, the amount of coke
deposited on P/MCM-68(60) was lower than that of MCM-68(60)
(Table 2 and Fig. S5). These results indicate that phosphate impreg-
nation on MCM-68 catalyst could improve the tolerance for coke
deposition during DTO reaction. Coke, which should be considered
as a mixture of hydrogen deficient residues, originates mainly from
aromatics [46], thus aromatization and hydrogen transfer reactions
are highly important contributors to coke deposition. The enhanced
durability to coke deposition during DTO reaction over P/MCM-
68(60) is probably due to the decrease in the number of stronger
Lewis acid sites, which take part in the formation of aromatics
from the conversion of dimethyl ether, by phosphate impregnation
(Fig. 4).
3.4. Catalytic performance of MCM-68 in DTO reactions
In our previous study, we investigated the effect of Si/Al molar
ratio of MCM-68 on the catalytic performance for DTO reaction.
The large number of acid sites on parent MCM-68(10) produced
more C2, C3 and C4 paraffins, aromatics and hydrocarbons that have
carbon numbers greater than 6. On the other hand, the optimization
of the number of acid sites by acid treatment gives MCM-68 with
sufficient acid sites for the propylene production in DTO reactions
[33,42]. This is because the undesirable consecutive reactions from
the light olefins were suppressed by the optimization of the number
of the acid sites on MCM-68.
In this work, we have compared the catalytic performance
of dealuminated MCM-68(60) with phosphate-modified MCM-
68(60). Comparison of the activity and selectivity over MCM-68(60)
and P/MCM-68(60) in DTO reactions are presented in Fig. 6 (a and
b, respectively). When the contact-time, W/F, (g h mol−1) was 20,
complete conversion of dimethyl ether to hydrocarbons over MCM-
68 and P/MCM-68(60) occurred. In order to compare the product
selectivity, the conversion level was tuned at around 95% by
decreasing the catalyst loading from 100 mg to 50 mg (see Table 2).
The product selectivity and DME conversion in DTO reactions over
MCM-68(60) and P/MCM-68(60) are also shown in Fig. 6d and
e, respectively). Phosphate modification on MCM-68(60) slightly
reduced the conversion since the more unreacted methanol and
dimethyl ether were observed in the DTO reaction over P/MCM-
68(60) in comparison with that over MCM-68(60). The selectivity to
light olefins, except ethylene, greatly increased by phosphate mod-
ification, whereas the selectivity of aromatics (benzene, toluene,
xylenes, 1,2,4-trimethylbenzene, 1,2,3-trimethylbenzene, 1,2,3,5-
tetramethylbenzene, 1,2,4,5-tetramethylbenzene, pentamethyl-
benzene and hexamethylbenezene) and light paraffins (ethane,
propane, butane and isobutane) decreased by phosphate modifi-
cation.
The product distributions were definitely different between the
cases of MCM-68(60) and P/MCM-68(60), when using 100 mg of
catalysts (see Fig. 6). When catalyzed by MCM-68(60), the increase
in catalyst loading from 50 to 100 mg significantly increased the
yield of ethylene and aromatics, particularly for pentamethyl-
benzene and hexamethylbenzene (Fig. 6a and d). In the case of
Conversely, the low ethylene selectivity observed over P/MCM-
68(60) is probably due to the changes in the number of strong
Brønsted acid sites because the formation of ethylene is proceed
through carbenium cation resulted by the cracking reactions of
intermediates [47]. Bleken et al. reported catalytic results of MTO

224
S. Park et al. / Catalysis Today 265 (2016) 218–224
reaction over SSZ-13 and SAPO-34 catalysts [ 48]. The high P/E ratios
in products were obtained when catalyzed by SSZ-13, which has the
same topology and the number of acid sites, but is slightly more
acidic due to the framework composition. They concluded that the
differences in acid strength might affect the product distribution
in MTO reactions. By considering the energy level of the primary
carbenium ions, which is higher than that of the secondary and
tertiary carbenium ion [49], the formation of ethylene, which has
to go through primary carbenium ions, proceed readily with the
existence of the stronger acid sites on catalysts. Although we could
not obtain a reliable result showing the decrease in the number of
strong Brønsted acid sites on MCM-68 by phosphate modification,
we believe that the high P/E ratio in products obtained P/MCM-68
is probably due to the decrease in the number of strong Brønsted
acid sites.
[4] A. Corma, F.V. Melo, L. Sauvanaud, F.J. Ortega, Appl. Catal. A 265 (2004) 195.
[5] A. Sardesai, S. Lee, Energy Sources 27 (2005) 489.
[6] S.A. Tabak, S. Yurchak, Catal. Today 6 (1990) 307.
[7] C.D. Chang, A.J. Silvestri, J. Catal. 47 (1977) 249.
[8] C.D. Chang, Stud. Surf. Sci. Catal. 36 (1988) 127.
[9] G. Zhao, J. Teng, Z. Xie, W. Jin, W. Yang, Q. Chen, Y. Tang, J. Catal. 248 (2007)
29.
[10] J.C. Mol, J. Mol. Catal. A: Chem. 213 (2004) 39.
[11] M.A. Chaar, D. Patel, H.H. Kung, J. Catal. 109 (1988) 463.
[12] G.A. Olah, Catal. Lett. 93 (2004) 1.
[13] K. Asami, Q. Zhang, X. Li, S. Asaoka, K. Fujimoto, Catal. Today 106 (2005) 247.
[14] M. Bjørgen, F. Joensen, K. Lillerud, U. Olsbye, S. Svelle, Catal. Today 142 (2009)
90.
[15] M. Stöcker, Micropor. Mesopor. Mater. 29 (1999) 3.
[16] L. Wu, E.J.M. Hensen, Catal. Today 235 (2014) 160.
[17] U. Olsbye, S. Svelle, M. Bjørgen, P. Beato, T.V.W. Janssens, F. Joensen, S.
Bordiga, K.P. Lillerud, Angew. Chem. Int. Ed. 51 (2012) 5810.
[18] H. Yamazaki, H. Shima, H. Imai, T. Yokoi, T. Tatsumi, J.N. Kondo, Angew. Chem.
Int. Ed. 50 (2011) 1853.
[19] T. Zhao, T. Takemoto, N. Tsubaki, Catal. Commun. 7 (2006) 647.
[20] A. Iida, R. Nakamura, K. Komura, Y. Sugi, Chem. Lett. 37 (2008) 494.
[21] H. Ito, K. Ooyama, S. Yamada, M. Kume, N. Chikamatsu, U.S. Pat. (2007)
0032379.
4. Conclusions
[22] K. Omata, Y. Yamazaki, Y. Watanabe, K. Kodama, M. Yamada, Ind. Eng. Chem.
Res. 48 (2009) 6256.
[23] J.C. Védrine, A. Auroux, P. Dejaifve, V. Ducarme, H. Hoser, S. Zhou, J. Catal. 73
(1982) 147.
[24] A. Rahman, G. Lemay, A. Adnot, S. Kaliaguine, J. Catal. 112 (1988) 453.
[25] M. Kaarsholm, F. Joensen, J. Nerlov, R. Cenni, J. Chaouki, G.S. Patience, Chem.
Eng. Sci. 62 (2007) 5527.
[26] J. Liu, C. Zhang, Z. Shen, W. Hua, Y. Tang, W. Shen, Y. Yue, H. Xu, Catal.
Commun. 10 (2009) 1506.
[27] D.V. Vu, Y. Hirota, N. Nishiyama, Y. Egashira, K. Ueyama, J. Jpn. Petrol. Inst. 53
(2010) 232.
[28] H.-G. Jang, H.-K. Min, S.B. Hong, G. Seo, J. Catal. 299 (2013) 240.
[29] M. Dyballa, E. Klemm, J. Weitkamp, M. Hunger, Chem. Ing. Tech. 85 (2013)
1719.
MCM-68(60) and phosphate-loaded MCM-68(60) zeolites were
synthesized by the post-treatment of hydrothermally synthesized
MCM-68(10). The results obtained from the DTO reactions indi-
cate that the P/MCM-68(60) is an effective catalyst for the selective
production of light olefins such as propylene and butylene. The
impregnation of MCM-68 with phosphate efficiently changed the
product distributions in DTO reaction. Almost 70 C-atom % selec-
tivity to propylene and butylenes with ca. 50 C-atom % selectivity
to propylene was achieved for the reaction over P/MCM-68(60).
Moreover, the phosphate modification improved the resistance to
coke deposition and decreased the selectivity toward aromatics
and ethylene. This is probably due to the selective decrease in the
number of acid sites on external surface of MCM-68 as confirmed
by TIPB and cumene cracking reactions. Furthermore, the phos-
phate modification decreased the number of strong Lewis acid sites
by covering them, providing the resultant catalyst P/MCM-68(60)
with high resistance to coke deposition during DTO reaction. The
high P/E ratio over P/MCM-68(60) is probably due to the removal
of strong Brønsted acid sites with phosphate modification.
[30] J.C. Vedrine, A. Auroux, P. Dejaifve, V. Ducarme, H. Hoser, S. Zhou, J. Catal. 73
(1982) 147.
[31] D.L. Dorset, S.C. Weston, S.S. Dhingra, J. Phys. Chem. B 110 (2006) 2045.
[32] C. Baerlocher, L.B. McCusker, D.H. Olson, Atlas of Zeolite Framework Types,
6th ed., Elsevier, Amsterdam, 2007 http://www.iza-structure.org/databases/.
[33] Y. Kubota, S. Inagaki, Top. Catal. 58 (2015) 480.
[37] Y. Kubota, S. Inagaki, K. Takechi, Catal. Today 226 (2014) 109.
[34] T. Shibata, S. Suzuki, H. Kawagoe, K. Komura, Y. Kubota, Y. Sugi, J.-H. Kim, G.
Seo, Micropor. Mesopor. Mater. 116 (2008) 216.
[35] T. Shibata, H. Kawagoe, H. Naiki, K. Komura, Y. Kubota, Y. Sugi, J. Mol. Catal. A:
Chem. 297 (2009) 80.
[36] S. Ernst, S.P. Elangovan, M. Gerstner, M. Hartmann, S. Sauerbeck, Abstr. 14th
Int. Zeol. Conf. (2004) 982.
[38] S. Inagaki, K. Takeshi, Y. Kubota, Chem. Commun. 46 (2010) 2662.
[39] S.P. Elangovan, M. Ogura, S. Ernst, M. Hartmann, S. Tontisirin, M.E. Davis, T.
Okubo, Micropor. Mesopor. Mater. 26 (2006) 210.
[40] M. Sasaki, Y. Sato, Y. Tsuboi, S. Inagaki, Y. Kubota, ACS Catal. 4 (2014)
2653.
Acknowledgements
This work was supported in part by CREST, JST.
Appendix A. Supplementary data
[41] Y. Kubota, Y. Koyama, T. Yamada, S. Inagaki, T. Tatsumi, Chem. Commun. 46
(2008) 6224.
[42] S. Park, Y. Watanabe, Y. Nishita, T. Fukuoka, S. Inagaki, Y. Kubota, J. Catal. 319
(2014) 265.
[43] M. Niwa, K. Katada, Catal. Surv. Jpn. 1 (1997) 215.
[44] H. Yamazaki, H. Shima, H. Imai, T. Yokoi, T. Tatsumi, J.N. Kondo, J. Phys. Chem.
C 116 (2012) 24091.
Supplementary material related to this article can be found,
in the online version, at http://dx.doi.org/10.1016/j.cattod.2015.08.
061.
[45] B. Gil, G. Kosˇová, J. Cejka, Micropor. Mesopor. Mater. 129 (2010) 256.
[46] W.G. Appleby, J.W. Gorbson, G.W. Good, Ind. Eng. Chem. Process Des. Dev. 1
(1962) 102.
References
[47] G.A. Olah, J. Org. Chem. 66 (2001) 5943.
[48] F. Bleken, M. Bjørgen, L. Palumbo, S. Bordiga, S. Svelle, K.-P. Lillerud, U. Olsbye,
Top. Catal. 52 (2009) 218.
[1] J. Verstraete, V. Coupard, C. Thomazeau, P. Etienne, Catal. Today 106 (2005) 62.
[2] E.H. Edwin, J.G. Balchen, Chem. Eng. Sci. 56 (2001) 989.
[3] A. Farshi, F. Shaiyegh, S.H. Burogerdi, A. Dehgan, Pet. Sci. Technol. 29 (2011)
875.
[49] P. Vogel, Study Organic Chemistry, 4th. Ed., 1985, pp. 167.