MWW-type catalysts for gas phase glycerol dehydration to acrolein

Article abstract of DOI:10.1016/j.jcat.2015.11.010

The materials derived from MWW layered precursor (MCM-22, MCM-36 and ITQ-2) were synthesized with molar ratio SiO₂/Al₂O₃ = 30, characterized by X-ray diffraction (XRD), Fourier transform infrared spectroscopy (FTIR), thermogravimetry (TG), scanning electron microscopy (SEM), energy dispersive X-ray spectrometry (EDX), textural analysis by N₂ physisorption, temperature programmed desorption of ammonia (NH₃-TPD) and diffuse reflectance infrared spectroscopy (DRIFTS) of adsorbed pyridine and evaluated in gas phase glycerol dehydration to acrolein. The delaminated material (ITQ-2) has presented better catalytic performance than MCM-22 zeolite or MCM-36 pillared material, in both glycerol conversion and acrolein selectivity. These results were interpreted based on the textural properties and acidity changes. ITQ-2 excellent performance is due to higher accessibility and improved acidity when compared to pillared MCM-36 or parent MCM-22 zeolite. Long-term stability under either nitrogen or air co-feeding was investigated for ITQ-2.

Full text of DOI:10.1016/j.jcat.2015.11.010

Journal of Catalysis 334 (2016) 34–41
Contents lists available at ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
MWW-type catalysts for gas phase glycerol dehydration to acrolein
a
a
a
a
b
Camila S. Carriço , Fernanda T. Cruz , Maurício B. dos Santos , Daniel S. Oliveira , Heloise O. Pastore ,
a,c
a,c,
⇑
Heloysa M.C. Andrade , Artur J.S. Mascarenhas
a
Laboratório de Catálise e Materiais, Departamento de Química Geral e Inorgânica, Instituto de Química, Universidade Federal da Bahia, R. Barão do Jeremoabo, s/n, Ondina,
0170-280 Salvador, Bahia, Brazil
Grupo de Peneiras Moleculares Micro- e Mesoporosas, Instituto de Química, Universidade Estadual de Campinas, CP 6154, 13083-970 Campinas, São Paulo, Brazil
Instituto Nacional de Ciência e Tecnologia em Energia e Ambiente (INCT-E&A), Universidade Federal da Bahia, R. Barão do Jeremoabo, s/n, Ondina, 40170-280 Salvador,
Bahia, Brazil
4
b
c
a r t i c l e i n f o
a b s t r a c t
Article history:
The materials derived from MWW layered precursor (MCM-22, MCM-36 and ITQ-2) were synthesized
Received 27 May 2015
Revised 10 November 2015
Accepted 18 November 2015
with molar ratio SiO
spectroscopy (FTIR), thermogravimetry (TG), scanning electron microscopy (SEM), energy dispersive X-
ray spectrometry (EDX), textural analysis by N physisorption, temperature programmed desorption of
ammonia (NH -TPD) and diffuse reflectance infrared spectroscopy (DRIFTS) of adsorbed pyridine and
2 2 3
/Al O = 30, characterized by X-ray diffraction (XRD), Fourier transform infrared
2
3
evaluated in gas phase glycerol dehydration to acrolein. The delaminated material (ITQ-2) has presented
better catalytic performance than MCM-22 zeolite or MCM-36 pillared material, in both glycerol conver-
sion and acrolein selectivity. These results were interpreted based on the textural properties and acidity
changes. ITQ-2 excellent performance is due to higher accessibility and improved acidity when compared
to pillared MCM-36 or parent MCM-22 zeolite. Long-term stability under either nitrogen or air co-feeding
was investigated for ITQ-2.
Keywords:
Glycerol dehydration
Acrolein
MCM-22
MCM-36
ITQ-2
Ó 2015 Elsevier Inc. All rights reserved.
1
. Introduction
Biodiesel is now being used worldwide as an alternative fuel to
catalyst; therefore, glycerol dehydration to produce acrolein is a
promising clean and sustainable technology [3,4].
Glycerol dehydration is an acid-catalyzed reaction where the cat-
alyst acidity is a crucial parameter [5]. Among other active catalyst,
such as mixed metal oxides or heteropolyacids, zeolites are promis-
ing catalysts due to their appropriate acidity combined to pore
structure of designable complexity, usually resulting in high glyc-
erol conversion and acrolein selectivity [6–8]. However, during the
reaction the pores of the zeolites can be blocked by coke deposition,
occluding the access to the active sites. Thus, fast catalyst deactiva-
tion is still the main obstacle for industrial application of glycerol
dehydration to acrolein [1], and oxygen co-feeding and new catalyst
compositions are the proposed strategies to solve this problem.
A promising alternative would be the use of materials derived
from layered zeolites, which can produce new materials through
swelling, delamination and pillaring processes, allowing tailoring
the structural, textural and acid properties of these materials [9].
MCM-22 (P) was the first lamellar zeolite known, and it combi-
nes, in the same structure, pores with different diameters, which
allow access to a range of active sites, allowing the study of the
effect of its textural and acid properties in the outcome of several
reactions. We have recently published results on catalytic dehydra-
tion of glycerol to acrolein, proving that H-MCM-22 can be as
replace, at least partially, diesel obtained from oil. This renewable
fuel is mainly produced from transesterification of vegetable oils
using methanol or ethanol, resulting in high amounts of glycerol
obtained as a coproduct. However, the economic viability of bio-
diesel production also depends on the destination of this coprod-
uct. Since its purification for pharmaceutical or cosmetic
applications is very expensive, glycerol could be catalytically con-
verted to many interesting chemicals or intermediates [1].
Acrolein is the glycerol-derived product of largest interest
because it is used for acrylic acid synthesis, which is an important
raw material in the production of polymers used in hygiene
products, such as diapers, detergents and wall paints [1,2]. The
currently used synthesis of acrolein is based on the selective oxida-
x
tion of propene over a complex multicomponent BiMoO -based
⇑
Corresponding author at: Laboratório de Catálise e Materiais, Departamento de
Química Geral e Inorgânica, Instituto de Química, Universidade Federal da Bahia, R.
Barão do Jeremoabo, s/n, Ondina, 40170-280 Salvador, Bahia, Brazil. Fax: +55 71
3
235 5166.
E-mail address: artur@ufba.br (A.J.S. Mascarenhas).
http://dx.doi.org/10.1016/j.jcat.2015.11.010
021-9517/Ó 2015 Elsevier Inc. All rights reserved.
0

C.S. Carriço et al. / Journal of Catalysis 334 (2016) 34–41
35
active and as selective as ZSM-5, operating at similar conditions
10].
MCM-22 (P) condenses in a three-dimensional structure by
calcination, named MCM-22 zeolite, whose peculiar topology was
designed by International Zeolite Association as MWW (from
Mobil tWenty-tWo). MCM-22 (P) structure consists of layers of
ITQ-2 synthesis proceeded according to the methodology pro-
posed by Concepción et al. [23]. The precursor was mixed with
an aqueous solution of CTABr 29% and an aqueous solution of
tetrapropylammonium hydroxide (TPAOH, 40%). The resulting sus-
pension was refluxed for 16 h at 80 °C. The layers were forced apart
by placing the slurry in an ultrasound bath for 1 h. Separation of
the solids was done by acidification of the medium with concen-
trated hydrochloric acid until the pH < 2, followed by centrifuga-
tion. The resulting material was calcined in the same way as
explained for the other MWW-derived materials.
[
2
.5 nm, which contains a two-dimensional sinusoidal channels,
whose access windows (4.0 ꢀ 5.9 Å) are defined by 10 mem-
bered rings (MR), and surface semicavities (or cups) of
7
.1 ꢀ 7.1 ꢀ 7.0 Å [11].
MCM-36 is obtained by swelling MCM-22 (P) with
The MWW derived materials were converted to their acid form
ꢁ
1
cetyltrimethylammonium bromide (CTABr) and pillaring using
tetraethoxysilane (TEOS) as SiO precursor [12]. Mesopores are
by ion exchange with a solution NH
4
NO
3
1 mol L
(Synth), at
flow
2
ambient temperature for 24 h, followed by calcination in a N
2
ꢁ1
thus created during the post-synthesis modification process,
allowing improved access to protonic sites present in the zeolite
structure, turning this an attractive candidate for catalytic and
sorption applications [13,14]. The pillared derivative is a peculiar
material with double porosity: microporous in crystalline layers
and mesoporous in interlayered region, resulting in a structure
similar to pillared clays.
(50 mL min ) at 550 °C.
2.2. Catalysts characterization
X-ray diffraction (XRD) patterns were collected on Shimadzu
XRD-6000 equipment, operating with a CuK radiation, generated
a
at a voltage of 40 kV, current 30 mA, using a graphite monochro-
+
ꢁ1
Other possibility is to delaminate CTA -swelled MCM-22 (P)
mator, in the region of 1.4–50° 2h at a scan speed of 2° min . Four-
producing an interesting material denominated ITQ-2 [15]. ITQ-2
possesses high external surface area, which allows the reaction
to occur not only inside the 10 MR channels but also in the semi-
cavities (cups) open to the external surface, with 12 MR windows
that confer zeolitic environment to reactants, products and transi-
tion states formed during reaction [16,17].
Improved catalytic performances in acid reactions are reported
in the literature for MWW-derived materials [18,19]. The pillaring
process increases the accessibility to protonic sites, which are
mainly located in mesopores created by silica pillars, resulting in
high catalytic activity of MCM-36 when compared with MCM-22
ier transformed infrared spectra (FTIR) were recorded on a Perkin
ꢁ1
Elmer Spectrum BX spectrometer in the range of 4000–400 cm
,
using 0.1% KBr pellets, by the accumulation of 16 scans with reso-
ꢁ1
lution of 4 cm . Thermogravimetry (TG) was conducted on a Shi-
madzu TGA-50 at temperatures ranging from 25 to 1000 °C, with a
ꢁ1
heating rate b = 10 °C min
,
under nitrogen or air flow
ꢁ1
(50 mL min ). Scanning electron microscopy (SEM) images of var-
ious magnifications were obtained in a Shimadzu SS-550 micro-
scope, with an acceleration voltage of 7–15 kV. The samples were
previously metallized with gold. Elemental analyses were per-
formed by energy dispersive X-ray spectrometry in a Shimadzu
EDX-720. Nitrogen physisorption isotherms were collected at
ꢁ196 °C in Micromeritics ASAP 2020 equipment, using the BET,
BJH and t-plot methods to assess textural properties. The samples
were pretreated at 350 °C for 3 h under vacuum (2 mmHg) to
remove physisorbed species on the sample surface. The acid
amount of catalysts was measured by temperature programmed
[
20]. Similarly, the delamination process increases the acid site
accessibility, by decreasing diffusion limitations, and thus ITQ-2
provides access of larger molecules to the acid sites, usually caus-
ing an increase in activity and selectivity in many acid catalyzed
reactions [21].
Aiming to determine the effects of pillaring and delamination
processes on glycerol conversion and acrolein selectivity, MCM-
desorption of ammonia (NH
2720 equipped with a thermal conductivity detector (TCD) to mea-
sure the NH desorption. The samples were pretreated in He flow
3
-TPD) in a Micromeritics Chemsorb
3
6 and ITQ-2 were synthesized, characterized and evaluated in
gas phase glycerol dehydration, under co-feeding either by
nitrogen or by air. Their catalytic performances were compared
to parent MCM-22 zeolite.
3
ꢁ1
(25 mL min ) at 300 °C for 1 h and cooled to room temperature.
The NH adsorption was carried out at ambient temperature in
3
.9% mol/mol NH /He (25 mL min ) for 1 h. After the chemisorp-
3
ꢁ1
9
2
. Experimental
tion, the system was purged with helium at 150 °C for 1 h to elim-
inate physisorbed ammonia. After cooling to room temperature,
2
.1. Catalysts synthesis
the NH
00 °C (10 °C min ) while monitoring the TCD signal.
Diffuse reflectance infrared spectra (DRIFTS) of adsorbed pyri-
3
-TPD was performed by heating the sample from 30 °C to
ꢁ1
8
2 2 3
MCM-22 (P) (SiO /Al O = 30) was prepared as described previ-
ously [10]. To produce MCM-22, the layered precursor material
dine were obtained by using a Perkin Elmer Spectrum 400 spectrom-
eter, operating with a reaction chamber from Harrick and a praying
mantis accessory. The samples were pretreated at 350 °C for 1 h
prior to the pyridine adsorption, in order to remove any adsorbed
water, and then a droplet of liquid pyridine was added on the sample
surface. The chamber was heated at 150 °C under flowing argon for
at least 30 min, or until no changes were detected in the spectrum.
ꢁ1
was calcined by heating under dry nitrogen (1 °C min ) up to
50 °C and keeping at this temperature for 6 h under dry air
5
ꢁ
1
(
50 mL min ).
MCM-36 was obtained according to the procedure reported by
Maheshwari et al. [22], mixing the MCM-22 (P) slurry with an
aqueous solution of 29% CTABr and an aqueous solution of 40%
tetrapropylammonium hydroxide (TPAOH). The pH of the resulting
mixture was 13.8. The mixture was allowed to stir for 16 h at room
temperature, after which, the particles were recovered by repeated
cycles of centrifugation and washing. The swollen MCM-22 (P)
powder was mixed with TEOS (tetraethoxysilane), stirred for
ꢁ1
Spectra were collected using 50 scans at 4 cm of resolution.
2.3. Catalytic tests
MWW-precursor derived catalysts were evaluated in gas phase
glycerol dehydration to acrolein, at atmospheric pressure, in a con-
tinuous flow vertical reactor on borosilicate glass at 320 °C. An
aqueous 36.6% glycerol solution was added to the reactor packed
with 50 mg of the catalyst, with the aid of a peristaltic pump oper-
2
5 h at 80 °C under nitrogen atmosphere, then filtered and dried
at room temperature. The resulting solid was hydrolyzed in water
pH = 8, controlled with NaOH) for 6 h at 40 °C, then filtered, dried
(
at room temperature and calcined using the same method applied
to MCM-22 zeolite.
ꢁ1
ꢁ1
ating at 0.33 mL min and a nitrogen or air flow (30 mL min ).

3
6
C.S. Carriço et al. / Journal of Catalysis 334 (2016) 34–41
During 10 h of reaction, the condensable products were collected
hourly in a cold trap containing a 0.1% solution of hydroquinone
and analyzed by a Perkin Elmer Clarus 500 gas chromatograph
operating with a flame ionization detector (GC-FID) and OV-315
column (30 m ꢀ 0.32 mm id).
2
of the SiO pillars maintained the basal distance in essentially
the same value as the swollen material, and this is consistent with
the production of MCM-36 pillared material [13,14]. Finally,
sonication of the swollen material causes partial amorphization,
due to disordered sheet-like structure expected for ITQ-2
[15,16]. No peaks were observed at low angle region for ITQ-2 sam-
ple, suggesting an efficient exfoliation during the synthesis
procedure.
The glycerol conversion and selectivity to acrolein were quanti-
fied according to the following equations:
Glycerol conversion ð%Þ ¼ 100 ꢀ ðnglycerol;in ꢁ nglycerol;outÞ=nglycerol;in
ð1Þ
FTIR spectra (Fig. 2) show that the MWW bands are preserved
+
upon pillaring and delamination processes. In CTA -MCM-22 sam-
ꢁ1
ple bands between 3000 and 2800 cm are observed and were
assigned to C–H stretching vibrations in –CH3 and –CH2 groups,
corresponding to the presence of CTABr, which was used in the
Acrolein selectivity ð%Þ ¼ 100 ꢀ nacrolein;formed=nglycerol;consumed
ð2Þ
where nglycerol,in is the initial amount of matter of glycerol (mol);
glycerol,out is the final amount of matter of glycerol (mol); nacrolein,-
swelling step. Wider bands were observed in MCM-36 material
n
ꢁ1
due to the presence of SiO
2
pillars. A band at ꢂ960 cm in the
formed is the amount of matter of acrolein (mol) in the effluents;
and nglycerol,consumed is the consumed amount of matter of glycerol
spectrum of ITQ-2 indicates the presence of Si(OH) terminal
groups, that increase upon delamination [24].
and corresponds to the difference between nglycerol,in and nglycerol,out
.
TG experiments were performed in the calcined samples of
MCM-22, ITQ-2 and MCM-36 and observed a flat line in this tem-
perature range, ruling out any possibility of residues due to precur-
sors used in the preparation.
Elemental analyses by EDX (Table 1) are in good agreement
with nominal values, indicating that no significant changes in
The formaldehyde, acetaldehyde, propionaldehyde and
acetone selectivities were confirmed by high performance
liquid chromatography (HPLC) after derivatization with
2
,4-dinitrophenylhydrazine. The HPLC analyses were performed
on an Agilent Series 1100 equipped with a Bruker C18 column
250 mm ꢀ 2.1 mm) using as mobile phase a gradient mixture of
water: acetonitrile.
(
the SiO
ITQ-2 from MCM-22. However, for MCM-36, the experimental
SiO /Al molar ratio (51) is larger than the nominal value (30),
2 2 3
/Al O molar ratio occurred during the preparation of
2
2 3
O
3
. Results and discussion
due to the addition of SiO during the pillaring step to form the pillars.
2
The morphologic analyses by SEM of materials derived from
3.1. Catalysts preparation and characterization
MWW-type precursors are shown in Fig. 3.
MCM-22 is formed by platelet crystallites aggregated in toroidal
particles [10]. MCM-36 and ITQ-2 are composed of platelet
agglomerates, resulting from the disaggregation of the toroidal
particles of MCM-22, which suggest that the material also delam-
inate during the pillaring process [25,26].
The XRD of the materials derived from MWW precursor is pre-
sented in Fig. 1.
MCM-22 (P), containing hexamethyleneimine (HMI) in the
structure, shows the typical pattern found in the literature [10]
for this material, with a low angle diffraction at 2h = 3.22°
The isotherms of the materials derived from layered precursor
are presented in Fig. 4.
(
d = 2.74 nm) which is an indicative of its layered nature. The
calcination produces deep changes in the XRD patterns due to
the condensation of the lamellae into the 3D structure of
MCM-22 zeolite. By swelling MCM-22 (P) with CTA cations, the
low angle peak is shifted to lower values (2h = 1.74°), indicating
the increase of basal spacing from 2.74 to 5.08 nm. Pillaring with
TEOS followed by calcination makes small modifications into the
XRD pattern, a peak at nearly the same position (2h = 1.56°) of
the swollen precursor is observed, indicating that the formation
MCM-22 isotherm is of type Ib [27], characteristic of microp-
orous materials, but is also observed the presence of secondary
mesopores created by crystallite pilling. MCM-36 preserves this
isotherm profile, except for the shape of the hysteresis loop that
changes dramatically. ITQ-2 isotherm is of type IV, characteristic
of mesoporous materials, due to the presence of secondary meso-
pores created by the exfoliation step. The hysteresis loop, which
is associated with the pores shape and structure, shows that
+
MCM-36
ITQ-2
ITQ-2
MCM-36
+
CTA -MCM-22
+
CTA - MCM-22
MCM-22
MCM-22 (P)
MCM-22
MCM-22 (P)
1
0
20
30
40
50
4000 3600 3200 2800 1600 1200 800
400
-1
2
θ (degree)
Wavenumber (cm )
Fig. 1. XRD profiles of the materials derived from MWW precursor obtained with
SiO /Al = 30.
Fig. 2. Infrared profiles of the MWW precursor derived materials obtained with
SiO /Al = 30.
2
2
O
3
2
2 3
O

C.S. Carriço et al. / Journal of Catalysis 334 (2016) 34–41
37
Table 1
Elemental and textural analysis of the MWW precursor derived materials.
BET (m2
g
ꢁ1
)
S
micro (m2
g
ꢁ1
)a
S
external (m2
g
ꢁ1
)a
V
micro (cm3
g
ꢁ1
)a
V
meso (cm3
g
ꢁ1
)b
d
meso (nm)b
Sample
SiO
2
2
/Al O
3
S
MCM-22
MCM-36
ITQ-2
28.4
51.1
37.2
565
603
852
435
175
169
130
428
684
0.191
0.074
0.072
0.327
0.319
0.564
10.2
10.4
19.8
a
Determined by BET and t-plot methods.
Determined by BJH method.
b
5
4
3
2
1
00
00
00
00
00
0
MCM-36
MCM-22
ITQ-2
0.0
0.2
0.4
0.6
0.8
1.0
Relative Pressure (P/P0)
Fig. 4. Adsorption isotherms of the MWW precursor derived materials. Filled and
hollow symbols indicate adsorption and desorption branches, respectively.
MCM-22 and ITQ-2 have plate-shaped particles aggregated
(
hysteresis loop of H3 type), and MCM-36 presents slit-shaped
pores (hysteresis loop of H4 type) [28].
The textural properties calculated from the isotherms (Table 1)
show a decrease in the micropore volume and the creation of a rel-
atively significant mesoporosity, as indicated by the increase in the
mesopore volume observed for the pillared and delaminated sam-
ples. The remarkable decrease in micropore volume was caused by
the disappearance of MWW supercage during the layers expan-
sion. The increase in external surface area observed for MCM-36
and ITQ-2 is coherent with the mesopores formation and delami-
nation observed for each material. Based on Table 1 results and
the analysis of literature data which study MWW type zeolites in
catalytic reactions [29–32], it can be concluded that textural prop-
erties depend on the quality of pillaring and delamination steps.
Results from textural analyses show that materials derived
from MWW-type precursor present improved access to the active
sites, allowing studying the influence of pores shape, size and
structure in the glycerol dehydration reaction.
3
The acid sites densities were investigated by NH -TPD (Fig. 5)
and quantification is presented in Table 2.
MCM-22 profile shows moderate and strong acid sites, des-
orbing at 253 and 421 °C, respectively. The pillaring process causes
a significant decrease in total density of acid sites for MCM-36 [25],
probably due to the silica pillars, which increase the total mass of
non-acidic material [33]. Desorption peaks are shifted to lower
temperatures (243 and 416 °C), indicating lower acid site strength
than MCM-22. The delaminated material (ITQ-2) presents a similar
profile to the MCM-22, also shifted to lower temperatures (207 and
3
78 °C), with a higher total density of acid sites, probably due to
the exposition of the sites located in the surface cups, but with
Fig. 3. SEM images of the MWW precursor derived materials: (a) MCM-22; (b)
MCM-36 and (c) ITQ-2.
lower acid site strength than MCM-22. The acid site strength

3
8
C.S. Carriço et al. / Journal of Catalysis 334 (2016) 34–41
MCM-22
MCM-36
B + L
L
B
ITQ-2
ITQ-2
MCM-22
1
550
1500
1450
1400
-1
MCM-36
Wavenumber (cm )
Fig. 6. FTIR spectra of adsorbed pyridine collected at 150 °C for MWW precursor
derived materials.
pH adjustment after exfoliation was already reported in the litera-
ture [36,37].
1
00 200 300 400 500 600 700 800
Temperature (°C)
3.2. Catalytic test
3
Fig. 5. NH -TPD profiles of the MWW precursor derived materials.
The glycerol conversion using materials derived from MWW-
type precursor is shown in Fig. 7. Catalytic results show that all
materials are highly active in gas phase glycerol dehydration. In
the first hours of reaction, MCM-22 has presented higher conver-
sion than ITQ-2 or MCM-36, but deactivates by coke formation
and after 3 h, glycerol conversion is inferior to that observed for
MCM-36 catalyst. Based on the results in Fig. 7 and Table 3,
MCM-36 presents the best glycerol conversion, remaining with
conversions over 80% during 6 h of reaction. On the other hand,
ITQ-2 showed a gradual decrease since the first hours, which is
related to the high available acidity of this material.
Catalytic performance seems to be a combination of improved
porosity and acid site amount and distribution for each type of cat-
alyst. Although ITQ-2 has a higher mesoporosity than MCM-36 (see
Table 1), it is important to understand that the structure of these
materials is quite different. ITQ-2 is constituted by exfoliated lay-
ers and its area is primarily external area, while MCM-36 is a pil-
lared material, in which MWW layers are separated by SiO2
pillars. These differences should be accounted in considering coke
formation in these catalysts. However, it would be expected that if
the accessibility is restricted, deactivation will be faster.
decrease may be attributed to a partial dealumination during the
swelling or pillaring processes [32]. Based on the analysis of
NH -TPD results and literature data with materials derived from
3
MWW-type precursors [29–32], it can be concluded that acid prop-
erties are deeply dependent on the efficiency of pillaring and
delamination processes, since these steps influence in the distribu-
tion/exposition of the active sites.
The nature of acid sites was investigated by DRIFTS of adsorbed
pyridine and spectra are shown in Fig. 6. It was possible to detect
ꢁ1
the bands associated with Brönsted acid sites at 1540 cm and
ꢁ1
Lewis acid sites in the region of 1450–1440 cm . In addition, a
ꢁ1
band at nearly 1490 cm was observed in the spectra of Fig. 6,
which can be attributed to pyridine adsorbed on both Brönsted
and Lewis acid sites, as well as H-bonded pyridine. Since the bands
ꢁ1
at 1540 and 1440–1450 cm have approximately equal extinction
coefficients [34], the ratio of their integral intensities can be used
as a direct ratio of Brönsted and Lewis acid site densities (B/L).
From the ratio B/L and total density of acid sites, it was possible
to calculate the density of Brönsted acid sites (BAS) and Lewis acid
sites (LAS) shown in Table 2. The pillaring process resulted in a
decrease mainly in the Lewis acid sites for MCM-36, while
Brönsted acid sites are not affected. On the other hand, exfoliation
promoted only a slight decrease in the Brönsted acid sites density,
but increased the Lewis acid sites density. This could be a conse-
quence of extra-framework aluminum (EFAL) formation inside
ITQ-2 pores during its preparation, contributing for improved
acidity [35]. The precipitation of extra-framework material during
Recently, Park et al. [38] studied the coke formation mechanism
for H-ZSM-5, showing that the condensation of coke precursors is
promoted within its narrow pore structure, which results in the
rapid deactivation of the catalyst. For aluminosilicophosphate
nanosphere, on the other hand, condensed carbonaceous com-
pounds are observed and the deactivation is delayed due to the rel-
atively uninterrupted diffusion of materials (reactants, products, or
potential coke precursors).
Table 2
3
Acid properties of the MWW precursor derived materials by NH -TPD and DRIFTS-pyridine.
Sample
NH
3
-TPD
DRIFTS-pyridine
Density of acid sites (mmol g 1
ꢁ
)
BAS (mmol g
ꢁ1
)
LAS (mmol g )
ꢁ1
B/L
Tm,1 (°C)
Tm,2 (°C)
MCM-22
MCM-36
ITQ-2
253
243
207
421
416
378
1.19
0.78
1.27
0.35
0.37
0.31
0.84
0.41
0.96
0.42
0.90
0.32

C.S. Carriço et al. / Journal of Catalysis 334 (2016) 34–41
39
It is important to mention that MCM-36 has a lower total acid
site density (Table 2), but stronger acid sites than ITQ-2. Probably,
the stronger acid sites present in the MCM-36 are responsible for
the better stability. The surface coverage by carbonaceous deposits
would be a consequence of a delicate balance between the
oligomerization that occurs preferentially on weak acid sites and
leads to coke formation, and cracking of the coke precursors on
strong acid sites, leading to different products, including light com-
pounds. This explanation has been previously used in the study of
alkylation of isobutane, comparing catalysts of different acid
strength and its impact on coke formation [39].
The almost parallel conversion profiles observed in Fig. 7a for
MCM-22 and ITQ-2 suggest that the pore system where the reac-
tion takes place is very similar in these two solids, that is, in these
cases, the zeolitic acid sites (in surface cups and/or inside the inter-
layer channel) control the behavior of the conversion along time,
while the values of conversion are related to the concentration of
available sites in each material.
1
00
9
8
7
6
5
4
3
2
1
0
0
0
0
0
0
0
0
0
0
MCM-22 (30)
MCM-36 (30)
ITQ-2 (30)
0
1
2
3
4
5
6
7
8
9
10
Time (h)
Unlikely, MCM-36 is the least selective solid, because its acid
and structural properties probably favor the formation of non-
condensable by-products, such as CO, CO and CH , that were not
2 4
(
a)
1
00
monitored in this work. The less complete molar balance observed
for this catalyst is an indication that there is an increase in the for-
mation of gases [40]. It is necessary to consider that, at least in
MCM-22 (30)
MCM-36 (30)
ITQ-2 (30)
9
8
7
6
5
4
3
2
1
0
0
0
0
0
0
0
0
0
0
2
part, SiO pillars contain silanol groups that can catalyze parallel
reactions. On the other hand, the initial conversion remains
approximately constant due to the fact that the mesoporosity
keeps the large spaces available for reaction until coke build up
attains a certain point – around ca. 6 h of reaction – where it irre-
versibly blocks the active sites and one sees a sudden drop in con-
version values.
Some other products were observed in the condensate (Table 3),
such as acetol (1-hydroxy-acetone), acetaldehyde, formaldehyde,
acetone, and propionaldehyde. The by-products production is in
accordance with the reaction mechanism proposed by Martinuzzi
et al. [41].
Since ITQ-2 and MCM-22 acid sites densities are similar, the
results of coke formation have been correlated with the mesopore
volume. The difference is that the mesopore volume is almost dou-
ble in ITQ-2 than in MCM-22. After 10 h of catalytic run, the pores
are filled with coke, as recently suggested by Decolatti et al. [42].
Because of this, the amount of coke on ITQ-2 is almost double than
in the MCM-22.
According to the literature [43,44], the ratio B/L influences the
product distribution. The greater the ratio of Brönsted to Lewis acid
sites, the higher the selectivity to acrolein. On the other hand, the
greater is the ratio of Lewis to Brönsted acid sites, the higher the
selectivity to hydroxyacetone (acetol). However, in the herein
reported results, it was not possible to see a clear relationship
between product distribution and the ratio B/L.
0
1
2
3
4
5
6
7
8
9
10
Time (h)
(
b)
Fig. 7. Glycerol conversion and acrolein selectivity over MWW catalysts. Feed
composition: T = 320 °C; 36.6% glycerol; glycerol flow = 0.33 mL min , catalyst
weight = 0.10 g, N flow = 30 mL min .
2
ꢁ
1
ꢁ
1
Table 3
Product selectivity during gas phase glycerol dehydration after 2 and 10 h of time on
stream.
3.3. Characterization of spent catalysts
Catalyst
MCM-22
MCM-36
ITQ-2
Conversion (%)
Acrolein yield (%)
99.8 (20.2)
49.9 (4.3)
89.0 (10,5)
6.7 (0.1)
58.1 (17.6)
44.4 (7.2)
The spent catalysts were characterized by TG under air after
0 h of time on stream. From TG data in Fig. 8, three events were
1
Molar selectivity (%)
Acrolein
Acetol
Formaldehyde
Acetaldehyde
Acetone
Allylic alcohol
Propionic acid
Propionaldehyde
Others
detected in the region of 300–800 °C: (i) around 450 °C, corre-
sponding to the burning of thermally unstable coke, probably
formed on the external surface of crystallites; (ii) around 560 °C,
referring to the burning of coke deposited on the MWW supercages
or surface cups and (iii) around 670 °C, carbonaceous deposits in
the 10 MR channel [10].
These results suggested that the coke formed is similar in all
three materials and that it could be removed by thermal treatment
under oxidative atmosphere.
50.1 (21.3)
2.0 (0)
4.6 (13.6)
5.5 (15.3)
0.2 (0.4)
0.3 (0)
4.2 (0)
0.1 (0.7)
33.0 (48.7)
7.5 (1.3)
1.6 (0.2)
0.7 (5.6)
0.8 (5.6)
0.1 (0.1)
0 (0)
1.1 (0.2)
0.1 (0.4)
88.1 (86.5)
76.5 (40.1)
7.3 (1.4)
4.0 (17.3)
4.8 (20.7)
0.3 (1.2)
0.4 (0.2)
5.3 (0.3)
0.3 (2.8)
1.2 (16.1)
Coke (%)a
24.6
20.4
37.0
For evaluating the efficiency of a thermal treatment at 500 °C
for 1 h, under air flow (30 mL min ), to remove coke, an ITQ-2
a
Coke content after 10 h of run determined by TG under air flowing.
ꢁ1

4
0
C.S. Carriço et al. / Journal of Catalysis 334 (2016) 34–41
1
00
90
1
00
MCM-22
MCM-36
ITQ-2
Cycle 1
Cycle 2
Cycle 3
8
7
6
5
4
3
2
1
0
0
0
0
0
0
0
0
0
9
8
7
6
0
0
0
0
Air
N2
0
200
400
600
800
1000
0
2
4
6
8 10 12 14 16 18 20 22 24 26 28 30 32
Temperature (°C)
Time (h)
(
a)
(
a)
1
00
Cycle 1
Cycle 2
Cycle 3
9
8
7
6
5
4
3
2
0
0
0
0
0
0
0
0
MCM-22
MCM-36
ITQ-2
Air
10
0
N2
2
00
400
600
800
1000
0
2
4
6
8 10 12 14 16 18 20 22 24 26 28 30 32
Temperature (°C)
Tempo (h)
(
b)
(
b)
Fig. 8. (a) Thermogravimetry and (b) derivative thermogravimetry of MWW
derived materials after 10 h of catalytic test.
Fig. 10. Glycerol conversion (a) and acrolein selectivity (b) during glycerol
dehydration over ITQ-2 catalyst during 3 cycles of 10 h each. Feed composition:
T = 320 °C; 36.6% glycerol; glycerol flow = 0.33 mL min , catalyst weight = 0.10 g,
ꢁ
1
ꢁ
1
gas flow = 30 mL min
.
1
9
8
7
6
5
4
3
2
1
000
00
00
00
00
00
00
00
00
00
1
00
9
9
8
8
7
7
5
0
5
0
5
0
the mass loss was 10.3%, corresponding to 80% of the total carbon
deposits. Above 500 °C, only highly stable coke is still present in
the catalyst, corresponding to 5.8% of mass loss and only 19.8% of
total carbon deposited in the ITQ-2 catalyst.
3.4. Longtime stability and oxygen co-feeding
Catalytic results obtained during three cycles of 10 h, interca-
lated by a thermal treatment at 500 °C for 1 h, are shown in
ꢁ1
0
Fig. 10, for both nitrogen or air under 30 mL min flow.
0
20 40 60 80 100 120 140 160
In terms of glycerol conversion, ITQ-2 performance was fully
re-established after 3 reaction cycles in nitrogen flow. The small
differences in the glycerol values can be accounted for by experi-
mental errors. On the other hand, the initial acrolein selectivity is
not the same between the first and the second cycle, confirming
that the thermal treatment under air flowing did not remove com-
pletely the coke deposits. From the TG/DTG data (Figs. 8 and 9), it is
possible to see that only thermally unstable coke was removed
(ca. 80%), leaving more stable species unchanged (ca. 20%). Still
comparing the acrolein selectivity under nitrogen flow, one could
Time (min)
Fig. 9. Evaluation of coke removal by burning at 500 °C, for 1 h, under air flow
ꢁ
1
(
30 mL min ) by thermogravimetry.
coked sample was submitted to thermogravimetry experiments
and results are shown in Fig. 9.
From room temperature up to 500 °C, one could observe a mass
loss of 13.3%, due to thermally unstable coke. After 1 h at 500 °C,

C.S. Carriço et al. / Journal of Catalysis 334 (2016) 34–41
41
observe that the selectivity losses with time on stream are less pro-
nounced in the third than in the second reaction cycle.
[14] Y.J. He, G.S. Nivarthy, F. Eder, K. Seshan, J.A. Lercher, Synthesis,
characterization and catalytic activity of the pillared molecular sieve MCM-
36, Microporous Mesoporous Mater. 25 (1998) 207–224.
It is possible to see in Fig. 10a that initial deactivation occurs for
both nitrogen and air, but in a lesser extension in the last case. ITQ-2
reaches higher values of glycerol conversion under oxidative condi-
tions. Acrolein is still the major product, with high selectivity
[
15] A. Corma, V. Fornés, S.B. Pergher, T.L.M. Maesen, J.G. Burglass, Delaminated
zeolite precursors as selective acidic catalysts, Nature 396 (1998) 353–356.
16] A. Corma, V. Fornés, J. Martínez-Triguero, S.B. Pergher, Delaminated zeolites:
combining the benefits of zeolites and mesoporous materials for catalytic uses,
J. Catal. 186 (1999) 57–63.
[
(
Fig. 10b), but trace amounts of acetic, acrylic and propionic acids
[17] P. Frontera, F. Testa, R. Aiello, S. Candamano, J.B. Nagy, Transformation of
MCM-22(P) into ITQ-2: the role of framework aluminium, Microporous
Mesoporous Mater. 106 (2007) 107–114.
were detected in the condensate. Acrolein selectivity increased by
using air instead of nitrogen as carrier gas during catalytic tests,
mainly because oxygen favors a continuous removal of coke precur-
sors that regenerates acid sites at reaction temperature. The
decrease of acrolein selectivity in each reaction cycle is a conse-
quence of coke formation that changes accessibility and acidity of
the catalysts. Even though acrolein selectivity decreases during a
[
18] H.-K. Min, M.B. Park, S.B. Hong, Methanol to olefin conversion over H-MCM-22
and H-ITQ-2 zeolites, J. Catal. 271 (2010) 186–194.
[
19] Z. Zhang, W. Zhu, S. Zai, M. Jia, W. Zhang, Z. Wang, Synthesis, characterization
and catalytic properties of MCM-36 pillared via the MCM-56 precursor, J.
Porous Mater. 20 (2013) 531–538.
[20] J. Wang, X. Tu, W. Hua, Y. Yue, Z. Gao, Role of the acidity and porosity of
MWW-type zeolites in liquid-phase reaction, Microporous Mesoporous Mater.
142 (2011) 82–90.
1
0 h run, it has been almost fully regenerated in the following cycles.
[
[
[
21] H.J. Jung, S.S. Park, C.H. Shin, Y.K. Park, S.B. Hong, Comparative catalytic studies
on the conversion of 1-butene and n-butane to isobutene over MCM-22 and
ITQ-2 zeolites, J. Catal. 245 (2007) 65–74.
22] S. Maheshwari, C. Martinez, M.T. Portilla, F.J. Llopis, A. Corma, M. Tsapatsis,
Influence of layer structure preservation on the catalytic properties of the
pillared zeolite MCM-36, J. Catal. 272 (2010) 298–308.
23] P. Concepción, C. Lopez, A. Martinez, V.E. Puntes, Characterization and catalytic
properties of cobalt supported on delaminated ITQ-6 and ITQ-2 zeolites for the
Fischer–Tropsch synthesis reaction, J. Catal. 228 (2004) 321–332.
4
. Conclusions
Catalysts derived from MWW-type precursors are active and
selective in gas phase dehydration of glycerol to acrolein. The
delaminated material ITQ-2 showed higher selectivity to acrolein
because of the improved textural properties when compared to
MCM-22. The pillared material MCM-36 presented higher glycerol
conversion, but lower selectivity to acrolein, as a combined result
[24] A. Corma, U. Diaz, V. Fornés, J.M. Guil, J. Martínez-Triguero, E.J. Creyghton,
Characterization and catalytic activity of MCM-22 and MCM-56 compared
with ITQ-2, J. Catal. 191 (2000) 218–224.
[
25] A. Lacarriere, F. Luck, D. Swierczynski, F. Fajula, V. Hulea, Methanol to
hydrocarbons over zeolites with MWW topology: effect of zeolite texture and
acidity, Appl. Catal. A 402 (2011) 208–217.
2
of acid sites density decreases and obstruction of sites by SiO pil-
lars. The main deactivation mechanism is coke formation, but the
ITQ-2 catalyst can be regenerated by burning under oxidative con-
ditions. The ITQ-2 is the best candidate for a stable catalyst for the
dehydration of glycerol to acrolein under oxidative conditions
among the MWW-derived materials here studied.
[
26] M.M. Antunes, S. Lima, A. Fernandes, M. Pillinger, M.F. Ribeiro, A.A. Valente,
Aqueous-phase dehydration of xylose to furfural in the presence of MCM-22
and ITQ-2 solid acid catalysts, Appl. Catal. A 417–418 (2012) 243–252.
[27] S.J. Gregg, K.S. Singh, Adsorption Surface Area and Porosity, second ed.,
Academic Press, London, 1982.
[
28] J. Rouquerol, F. Rouquerol, K.S.W. Singh, Adsorption by Powders and Porous
Solids: Principles, Methodology and Applications, Academic Press, London,
1998.
Acknowledgment
[
29] J. Wang, S. Jaenicke, G.K. Chuah, W.M. Hua, Y.H. Yue, Z. Gao, Acidity and
porosity modulation of MWW type zeolites for Nopol production by Prins
condensation, Catal. Commun. 12 (2011) 1131–1135.
The authors acknowledge the financial support from the
Conselho Nacional de Desenvolvimento Científico e Tecnológico
(
2
[30] S.T. Yang, J.Y. Kim, J. Kim, W.S. Ahn, CO capture over amine-functionalized
MCM-22, MCM-36 and ITQ-2, Fuel 97 (2012) 435–442.
CNPq – Process n°. 476657/2010-5).
[
31] E. Dumitriu, D. Meloni, R. Monaci, V. Solinas, Liquid-phase alkylation of phenol
with t-butanol over various catalysts derived from MWW-type precursors, C.
R. Chim. 8 (2005) 441–456.
References
[
32] B. Liu, H. Huo, Q. Meng, S. Gao, Synthesis of ITQ-2 zeolite under static
conditions and its properties, Sci. China B 49 (2006) 148–154.
[
1] A. Talebian-Kiakalaieh, N.A.S. Amin, H. Hezaveh, Glycerol for renewable
acrolein production by catalytic dehydration, Renew. Sustain. Energy Rev. 40
[
33] J.O. Barth, A. Jentys, J. Kornatowski, J.A. Lercher, Control of acid–base
properties of new nanocomposite derivatives of MCM-36 by mixed oxide
pillaring, Chem. Mater. 16 (2004) 724–730.
(
2014) 28–59.
[
2] B. Katryniok, S. Paul, F. Dumeignil, Recent developments in the field of catalytic
dehydration of glycerol to acrolein, ACS Catal. 3 (2013) 1819–1834.
3] K. Omata, S. Izumi, T. Murayama, W. Ueda, Hydrothermal synthesis of W–Nb
complex metal oxides and their application to catalytic dehydration of glycerol
to acrolein, Catal. Today 201 (2013) 7–11.
[
34] H. Knözinger, in: G. Ertl, H. Knözinger, F. Schüth, J. Weitkamp (Eds.),
Handbook of Heterogeneous Catalysis, vol. 2, Wiley-VCH, 2008, p. 1154.
35] Z. Wang, L. Wang, Y. Jiang, M. Hunger, J. Huang, Cooperativity of Brønsted and
Lewis acid sites on zeolite for glycerol dehydration, ACS Catal. 4 (2014) 1144–
[
[
1147.
[
[
4] A. Ulgen, W.F. Hoelderich, Conversion of glycerol to acrolein in the presence of
WO /TiO catalysts, Appl. Catal. A 400 (2011) 34–38.
3 2
5] B. Katryniok, S. Paul, V. Belliere-Baca, P. Rey, F. Dumeignil, Glycerol
dehydration to acrolein in the context of new uses of glycerol, Green Chem.
[
[
[
36] M.K. de Pietre, F.A. Bonk, C. Rettori, F.A. Garcia, H.O. Pastore, Delaminated
vanadoaluminosilicate with [V, Al]-ITQ-18 structure, Microporous
Mesoporous Mater. 156 (2012) 244–256.
37] M.K. de Pietre, F.A. Bonk, C. Rettori, F.A. Garcia, H.O. Pastore, [V, Al]-ITQ-6:
novel porous material and the effect of delamination conditions on V sites and
their distribution, Microporous Mesoporous Mater. 145 (2011) 108–117.
38] H. Park, Y.S. Yun, T.Y. Kim, K.R. Lee, J. Baek, J. Yi, Kinetics of the dehydration of
glycerol over acid catalysts with an investigation of deactivation mechanism
by coke, Appl. Catal. B 176–177 (2015) 1–10.
1
2 (2010) 2079–2098.
[
[
[
6] Y.T. Kim, K.D. Jung, E.D. Park, A comparative study for gas-phase dehydration of
glycerol over H-zeolites, Appl. Catal. A 393 (2011) 275–287.
7] Y.T. Kim, K.D. Jung, E.D. Park, Gas-phase dehydration of glycerol over ZSM-5
catalysts, Microporous Mesoporous Mater. 131 (2010) 28–36.
8] C.J. Jia, Y. Liu, W. Schmidt, A.H. Lu, F. Schuth, Small-sized HZSM-5 zeolite as
highly active catalyst for gas phase dehydration of glycerol to acrolein, J. Catal.
[
39] B.O. Dalla Costa, C.A. Querini, Isobutane alkylation with butenes in gas phase,
Chem. Eng. J. 162 (2010) 829–835.
2
69 (2010) 71–79.
[
40] M. Käldström, N. Kumar, T. Heikkilä, D.Yu. Murzin, Pillared H-MCM-36
mesoporous and H-MCM-22 microporous materials for conversion of
levoglucosan: influence of varying acidity, Appl. Catal. A 397 (2011) 13–21.
41] I. Martinuzzi, Y. Azizi, J.-F. Devaux, S. Trejak, O. Zahraa, J.-P. Leclerc, Reaction
mechanism for glycerol dehydration in the gas phase over a solid acid catalyst
determined with on-line gas chromatography, Chem. Eng. Sci. 116 (2014)
[
9] U. Díaz, Layered materials with catalytic applications: pillared and
delaminated zeolites from MWW precursors, ISRN Chem. Eng. 20 (2012),
http://dx.doi.org/10.5402/2012/537164. ID537164.
[
[
[
10] C.S. Carriço, F.T. Cruz, M.B. Santos, H.O. Pastore, H.M.C. Andrade, A.J.S.
Mascarenhas, Efficiency of zeolite MCM-22 with different SiO /Al molar
2
2 3
O
ratios in gas phase glycerol dehydration to acrolein, Microporous Mesoporous
Mater. 181 (2013) 74–82.
118–127.
42] H.P. Decolatti, B.O. Dalla Costa, C.A. Querini, Dehydration of glycerol to acrolein
using H-ZSM5 zeolite modified by alkali treatment with NaOH, Microporous
Mesoporous Mater. 204 (2015) 180–189.
[
[
[
11] S. Lawton, M.E. Leonowicz, R. Partridge, P. Chu, M.K. Rubin, Microporous
Mesoporous Mater. 23 (1998) 109–117.
12] C.T. Kresge, W.J. Roth, K.G. Simmons, J.C. Vartuli, Crystalline Oxide Material, US
Patent 5,229,341, 1993, to Mobil Oil Corporation.
13] W.J. Roth, C.T. Kresge, J.C. Vartuli, M.E. Leonowicz, A.S. Fung, S.B. McCullen,
MCM-36: the first pillared molecular sieve with zeolite properties, Stud. Surf.
Sci. Catal. 94 (1995) 301–308.
[
43] Y.T. Kim, K.-D. Jung, E.D. Park, Gas-phase dehydration of glycerol over silica–
alumina catalysts, Appl. Catal. B 107 (2011) 177–187.
[
44] G.S. Foo, D. Wei, D.S. Sholl, C. Sievers, Role of Lewis and Brönsted acid sites in
the dehydration of glycerol over niobia, ACS Catal. 4 (2014) 3180–3192.