Peculiar behavior of MWW materials in aldol condensation of furfural and acetone

Article abstract of DOI:10.1039/c4dt00184b

MWW family of different structural types (MCM-22, MCM-49, MCM-56 and MCM-36) was used as catalysts for aldol condensation of furfural and acetone studied in a batch reactor at 100 °C, autogenous pressure and a reaction time of 0-4 h. To establish a relation between physico-chemical and catalytic properties of microporous materials, the samples were characterized by XRD, SEM, N₂ adsorption, FTIR and TGA. It was found that the acidic solids possessed appreciable activity in the reaction and resulted in the formation of products of aldehyde-ketone interaction. Surprisingly, MCM-22 and MCM-49, i.e. three-dimensional materials containing internal supercages, exhibited higher activity than two MCM-36 catalysts with two-dimensional character having larger accessible external surface area due to expansion of the interlayer space by swelling and pillaring treatments. Moreover, all MWW family catalysts gave higher conversion than the large-pore zeolite BEA. Nevertheless, furfural conversion decreased rapidly for all the studied materials due to coke formation. Unexpectedly, the deactivation was found to be more severe for MCM-36 catalysts than for MCM-22 and MCM-49, which was attributed to the reaction taking place also in supercages that are protected by 10-ring channels from severe coking. In contrast the cups located on the external surface were coked rapidly. This journal is the Partner Organisations 2014.

Full text of DOI:10.1039/c4dt00184b

Dalton
Transactions
View Article Online
View Journal | View Issue
PAPER
Peculiar behavior of MWW materials in aldol
condensation of furfural and acetone
Oleg Kikhtyanin,a Pavla Chlubná,b Tereza Jindrováa and David Kubička*a
Cite this: Dalton Trans., 2014, 43,
10628
MWW family of different structural types (MCM-22, MCM-49, MCM-56 and MCM-36) was used as cata-
lysts for aldol condensation of furfural and acetone studied in a batch reactor at 100 °C, autogenous
pressure and a reaction time of 0–4 h. To establish a relation between physico-chemical and catalytic
properties of microporous materials, the samples were characterized by XRD, SEM, N2 adsorption, FTIR
and TGA. It was found that the acidic solids possessed appreciable activity in the reaction and resulted in
the formation of products of aldehyde–ketone interaction. Surprisingly, MCM-22 and MCM-49, i.e. three-
dimensional materials containing internal supercages, exhibited higher activity than two MCM-36 catalysts
with two-dimensional character having larger accessible external surface area due to expansion of the
interlayer space by swelling and pillaring treatments. Moreover, all MWW family catalysts gave higher con-
version than the large-pore zeolite BEA. Nevertheless, furfural conversion decreased rapidly for all the
studied materials due to coke formation. Unexpectedly, the deactivation was found to be more severe for
MCM-36 catalysts than for MCM-22 and MCM-49, which was attributed to the reaction taking place also
in supercages that are protected by 10-ring channels from severe coking. In contrast the cups located on
the external surface were coked rapidly.
Received 19th January 2014,
Accepted 23rd April 2014
DOI: 10.1039/c4dt00184b
www.rsc.org/dalton
Furanic compounds thus can be valorized by aldol conden-
sation with acetone in the presence of suitable heterogeneous
1. Introduction
Production of valuable chemicals as well as fuels from renew- catalysts like MgO–ZrO2,7 alkali and alkaline earth oxides,9–15
able raw materials is a key technological challenge facing alkali-exchanged zeolites (Cs-BEA),16 MCM41,17,18 alumino-
modern society at the beginning of the 21st century. The devel- phosphate oxynitrides,19 ion-exchange resins,20,21 and acti-
opment of efficient technologies, relying on the use of natural vated Mg–Al hydrotalcites.22–26 Alternatively, a biphasic system
renewable resources, is an essential step for ensuring social with NaOH as a catalyst was also proposed.27 The needed
security and stability and mitigating the increasing environ- acetone may be obtained by ketonization of acetic acid28 that
mental threats connected with the extensive use of fossil is an important product of biomass pyrolysis. This route has
fuels.1
attracted significant interest, because hydrogenation/hydro-
Transformation of lignocellulosic and, in particular, waste deoxygenation of these aldol condensation products leads to
biomass into fuels and valuable chemicals is considered to be various hydrocarbons suitable for blending in conventional
one of the most promising directions as it allows avoiding diesel fuel and kerosene.29
competition with food and fodder production.1,2 Cellulose-rich
Aldol condensation of aldehydes and ketones is a classical
feedstocks can be successively treated to form different mono- reaction of organic synthesis, which takes place in the pres-
meric products, generally sugars.3–5 These can be further con- ence of both acidic and basic catalysts. Homogeneous basic
verted by acid hydrolysis to furanic compounds (furfural or catalysts such as sodium hydroxide are most often used for
hydroxymethylfurfural) that can serve as platform chemicals in this purpose.29–31 They possess high activity in conversion of
many applications.6–8 Among them, aldol condensation has reactants and high selectivity towards the desired reaction pro-
great potential as it allows increasing the carbon atom chain ducts. Unfortunately, their environmental friendliness is not
length beyond six carbon atoms, which is the limit for sugars. satisfactory, particularly for large scale applications, due to the
need for neutralization and separation of the homogeneous
catalysts from the liquid products. Consequently, solid basic
catalysts have attracted widespread interest as they could
be helpful in overcoming the deficiencies of the currently
used catalysts. The most studied basic solid catalysts include
metal oxides or double mixed oxides, such as MgO,32
aResearch Institute of Inorganic Chemistry, RENTECH-UniCRE, Chempark Litvínov,
Záluží – Litvínov, 436 70, Czech Republic. E-mail: david.kubicka@vuanch.cz;
Tel: +420-476163735
bJ. Heyrovský Institute of Physical Chemistry, Academy of Sciences of the Czech
Republic, v.v.i., Dolejškova 3, 182 23 Prague, Czech Republic
10628 | Dalton Trans., 2014, 43, 10628–10641
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
MgO–Al2O3,33–35 MgO–ZrO2
and TiO2.37 Recently, also (4.0 Å × 5.5 Å),58 thus forming a three-dimensional crystal
29,36
MOFs have been recognized as highly active catalysts in framework of MCM-22 zeolite with two independent pore
different condensation reactions.38,39 Despite the intensive systems.
research efforts, the heterogeneous basic catalysts suffer from
Formation of MCM-36 occurs through the subsequent swel-
several major drawbacks such as high sensitivity to the ling of the MCM-22(P) precursor by the addition of cetyl-
ambient CO2 and difficulties connected with re-using the trimethylammonium hydroxide followed by pillaring treatment
hydrotalcite-based catalysts in successive catalytic runs,40–42 using tetraethoxysilane (TEOS).53,60 As a consequence, the
which limits their widespread industrial use. Therefore, along space between the “MWW monolayers” is expanded and
with the optimization of the properties of the solid basic cata- internal mesoporous space is generated. Thus, MCM-36 pro-
lysts, attention is focused on alternative solid catalysts for vides free access to all large cups, located on the external
aldol condensation. In this respect, solid acid catalysts have surface of “MWW monolayers”. Nevertheless, it should be
been recognized as promising candidates.
underlined that the pillars consist of amorphous and inactive
Among solid acid catalysts, zeolites are considered versatile silica and thus they dilute the concentration of active sites.
and tunable acidic catalysts exhibiting good activity and Their nature remains unknown.
selectivity in various reactions of organic synthesis ranging
Zeolite MCM-49 possesses nearly the same framework topo-
from petrochemical to fine chemical applications.1,43–48 Zeo- logy as the calcined MCM-22 (only the average unit cell c-par-
lites were also investigated as catalysts for aldol condensation ameter is 0.2 Å longer in MCM-49).61 However, unlike the
in the past.49–51 In our recent work52 it was shown that the latter, MCM-49 is formed by a direct synthesis route, avoiding
acidic zeolites possess rather good activity in aldol conden- the formation of intermediate lamellar MCM-22(P). This is
sation of furfural and acetone and their catalytic properties are achieved by optimizing the composition of the initial reaction
determined by the structural type, as well as by their textural mixture and the conditions of its hydrothermal treatment.
and acidic characteristics. The highest conversion of furfural When hexamethyleneimine is used as a structure directing
was observed when using wide pore zeolites with a three- agent, MCM-49 generally has a lower Si/Al molar ratio (<13)
dimensional crystalline framework, the best result being than MCM-22P does (>13). A factor that influences the for-
obtained over zeolite BEA. It was also shown that during the mation of either MCM-22P or MCM-49 is the molar ratio of the
reaction, the activity of all investigated zeolites decreased due organic cation Ro (SDA) to the inorganic cations Ri (e.g. Na, K,
to the formation of carbonaceous deposits inside their micro- or Rb). When the ratio Ro/Ri is typically less than 2.0, the for-
pores. Therefore, to compete with the solid basic materials as mation of MCM-49 is favoured. By analogy with MCM-22, the
catalysts for aldol condensation it is necessary to optimize the crystalline framework of MCM-49 consists of two independent
properties of zeolites with the aim to increase their activity in pore systems, which are both accessible through 10-ring aper-
the reaction and to improve their stability, i.e. resistance to tures with diameters of 4.0 × 5.9 Å for the sinusoidal channels
coking. Layered two-dimensional materials exhibit a better and 4.0 × 5.4 Å for the entrances to the supercages.62
accessibility of their active sites than the classical three-dimen-
Apart from MCM-22, zeolite MCM-49 crystallizes through
sional zeolites and could thus make the activity enhancement its delaminated intermediate designated as MCM-56. MCM-56
possible. About fifteen zeolite frameworks have been found to is synthesized from the same synthetic gel as MCM-49 but
form two-dimensional zeolite forms and among them MWW is with a more careful control for crystallization time and temp-
the first and most famous example.53–56
erature.63 The structure of this zeolite consists of disordered
In this respect, materials with MWW framework topology agglomeration of “MWW monolayers” without in-registry
occupy a special place among a large variety of different micro- alignment in the c direction and no interlayer links. Due to
and mesoporous materials as this family consists of more than this, MCM-56 has a higher concentration of large cups
ten members, e.g. MCM-22, MCM-36, MCM-49, MCM-56 and exposed to the crystal surface than the MCM-49.
others, which have the same MWW topology but a different
Owing to their properties, MWW materials have been used
arrangement of the layers (different interlayer bonding and in a number of different applications, and it has been found
distances).57 The unifying feature of all MWW materials is the that they can be considered promising catalysts for several
presence a layered crystalline unit in their structure, so-called organic reactions, for example for isobutene–2-butene alky-
“MWW monolayer”, with a thickness of 25 Å, which contains a lation,64 as well as alkylation of benzene with propylene.65 The
two-dimensional 10-MR sinusoidal channel pore system and uniqueness of these catalysts is that they combine the pro-
large cups (7.1 Å Ø, 7.0 Å depth) on the [0 0 1] crystal perties of both conventional zeolite systems and layered
surface.58,59 As a result of agglomeration of such layered materials. The expansion of the areas of application of MWW
species during hydrothermal treatment of aluminosilicate gel materials necessitates a deeper understanding of the impact of
in the presence of hexamethyleneimine (HMI) an intermediate structural features on their activity in conversion of organic
product is formed. It is designated as MCM-22 (P), the layered molecules.
precursor. Upon calcination of this reaction product, the con-
This paper presents a comparative study focusing on
densation of the monolayers occurs leading to creation of a (i) characterization of MCM-22, MCM-36, MCM-49 and MCM-56
new independent pore system, constituted by large supercages materials with a number of physico-chemical methods, and
(7.1 Å Ø, 18.2 Å height) accessible through 10-ring openings (ii) investigation of their catalytic properties in aldol condensation
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 10628–10641 | 10629

View Article Online
Paper
Dalton Transactions
of furfural and acetone. In addition, wide-pore zeolite Beta
that has shown the best performance among the zeolite cata-
lysts investigated previously in aldol condensation of furfural
and acetone was included in the study for comparison.
Table 1 Molar chemical compositions of reaction mixtures for the
synthesis of MWW zeolites
Sample
Al2O3
SiO2
Na2O
HMI
H2O
MCM-22
1
1
1
1
1
30
46
36
25
25
2.52
3
3
2.45
2.45
10
20
20
8.15
8.15
815
1960
1875
440
MCM-36(1)
MCM-36(2)
MCM-49
2. Experimental
2.1. Synthesis of MWW materials
MCM-56
440
2.1.1. MCM-22. 7.31 g of 50% water solution of NaOH was
added to 400 g of water. 7.44 g of sodium aluminate (40–45%
Na2O, 50–56% Al2O3, Aldrich) was added and stirred until it
dissolved. Then, 232.29 g of Ludox AS-30 was added. The
mixture became thick and was stirred until a homogeneous
gel was formed. Then, 38.34 g of HMI was added and the final
mixture was stirred for 2 hours. The reaction mixture was
charged into a Teflon-lined steel autoclave. Crystallization pro-
ceeded at 143 °C with inside stirring and autogenous pressure
for 96 hours. The solid product was collected by filtration,
washed with distilled water and dried in an oven at 60 °C
overnight.
2.1.3. MCM-36(2). The preparation method of this sample
was substantially the same as that for MCM-36(1), with the
only exception that only 62 g of Ludox AS-30 was taken.
2.1.4. MCM-49. 9 g of 50% NaOH water solution was
added to 396 g of water. 9.75 g of sodium aluminate (40–45%
Na2O, 50–56% Al2O3, Aldrich) was added and stirred until dis-
solving. Then, 77.1 g of silicon dioxide (Aldrich, nanopowder,
particle size 10–20 nm) was added. The mixture was stirred for
10 minutes and then 40.8 g of HMI was added and the gel was
stirred for 15 minutes. Crystallization proceeded at 143 °C
under agitation and autogenous pressure for 93 hours. The
solid product was collected by filtration, washed with distilled
water and dried in the oven at 60 °C overnight.
MCM-22P was heated in nitrogen at 500 °C for 3 hours and
a temperature ramp of 2 °C min−1, cooled to about 130 °C,
and then calcined in air at 540 °C for 6 hours with the ramp of
2 °C min−1. The calcined MCM-22 product was ion-exchanged
MCM-49 was heated in nitrogen at 482 °C for 3 hours and a
temperature ramp of 3 °C min−1, cooled to about 130 °C, and
then calcined in air at 540 °C for 6 hours with a ramp of 2 °C
min−1. The calcined product MCM-49 was ion-exchanged into
+
into NH4 form by treating four times with 1.0 M NH4NO3
solution for 4 hours at room temperature.
+
NH4 form by treating it four times with 1.0 M NH4NO3 solu-
2.1.2. MCM-36(1). 1.16 g of NaOH was dissolved in 251.2 g
of water and mixed with 1.68 g of sodium aluminate (40–45%
Na2O, 50–56% Al2O3, Aldrich). Then, 80 g of Ludox AS-30 was
added. The mixture became thick and was stirred until a
homogeneous gel was formed. Then, 38.34 g of HMI was
added and the final mixture was stirred for 2 hours. The reac-
tion mixture was charged into a Teflon-lined steel autoclave.
Crystallization proceeded at 150 °C under agitation and auto-
genous pressure for 7 days. The solid product was collected
by filtration, washed with distilled water and dried in the oven
at 60 °C overnight.
tion for 4 hours at room temperature.
2.1.5. MCM-56. The preparation method of this sample
was substantially the same as that for MCM-49, with the only
exception that the time of hydrothermal synthesis was 33 hours.
Molar compositions of reaction mixtures used for the syn-
thesis of different MWW samples are presented in Table 1.
The relationship between different structures of the MWW
family is depicted in Scheme 1.
2.2. Characterization methods
10 g of uncalcined MCM-22(P) was added to 200 ml of 25% The crystallographic structures of MWW zeolites were deter-
solution of C16TMA-OH prepared by ion-exchange from chlor- mined by X-ray powder diffraction using a Bruker AXS D8
ide form. The slurry was stirred overnight at ambient tempera- Advance diffractometer, working with the Cu Kα line (λ =
ture. The product MCM-22SW was separated by centrifugation, 0.154 nm) in the 2θ range of 4–40°.
properly washed with water and dried at 60 °C.
Nitrogen adsorption/desorption isotherms were measured
Pillaring was carried out with 12.44 g of MCM-22SW in on a Micromeritics GEMINI II 2370 volumetric Surface Area
373 ml of TEOS (Aldrich). The mixture was stirred and heated Analyzer at −196 °C to determine the surface area, pore
at 85 °C under reflux overnight. The solid was isolated by cen- volume and pore size distribution. Before the sorption
trifugation and dried at ambient temperature. Then, about measurements, all samples were degassed on a Micromeritics
1500 ml of water was added to 14.37 g of dried powder and FlowPrep 060 instrument under helium at 300 °C (heating rate
stirred overnight (hydrolysis). The product was centrifuged 10 °C min−1) for 4 h. The specific surface area was evaluated
again and dried at 60 °C. Final calcinations were carried out by the BET method66 using adsorption data in the range of
under nitrogen at 482 °C for 3 hours and a temperature ramp relative pressure from p/p0 = 0.05 to p/p0 = 0.25. The t-plot
of 3 °C min−1, cooled to about 130 °C and then calcined in air method67 was applied to determine the volume of micropores
at 540 °C for 6 hours with a ramp of 2 °C min−1. The calcined (Vmic). The adsorbed amount at relative pressure p/p0 = 0.98
+
product MCM-36(1) was ion-exchanged into NH4 form by reflects the total adsorption capacity (Vtot).
treating four times with 1.0 M NH4NO3 solution for 4 hours at
room temperature.
The concentration of Lewis (L) and Brønsted (B) acid sites
was determined after adsorption of d3-acetonitrile (Ac) by FTIR
10630 | Dalton Trans., 2014, 43, 10628–10641
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
Scheme 1 MWW family presented by MCM-22P, MCM-49 and MCM-56 prepared by direct synthesis and MCM-36 prepared by post-synthetic
modification of MCM-22P. The red dot-line presents the assumed transformation of delaminated form MCM-56 into regular 3D MCM-49 if the syn-
thesis time is prolonged.
spectroscopy on a Nicolet Protégé 460 Magna with a trans- from room temperature to 900 °C under the flow of nitrogen
mission DTGS and MTC/A detector. Zeolites were pressed into (20 mL min−1, Linde 3.0). Approximately15 mg of sample was
self-supporting wafers with a density of 8.0–12.0 mg cm−2 and heated in an open alumina crucible.
activated in situ at 450 °C overnight. Before adsorption d3-
2.3. Reaction studies
acetonitrile was degassed by freezing and thawing cycles. d3-
Acetonitrile was adsorbed at ambient temperature for 30 min
at a partial pressure of 900 Pa, followed by desorption for
20 min. All spectra were recorded with a resolution of 4 cm−1
by collecting 128 scans for a single spectrum at room tempera-
ture. The experimental data were recalculated to a normalized
wafer (density of 10 mg cm−2). The concentrations of Lewis
and Brønsted acid sites were evaluated from the integral inten-
sities of bands at 2323 cm−1 (Lewis acid sites) and at
2294 cm−1 (Brønsted acid sites) using extinction coefficients
The properties of zeolite samples were investigated in aldol
condensation of furfural and acetone. The catalytic experi-
ments were carried out in a 200 ml stirred batch reactor (Parr
autoclave) at 100 °C. Before the start of the catalytic runs, 0.5 g
of catalyst was mixed together with a mixture of 19.75 g of
acetone and 3.25 g of furfural (acetone/furfural molar ratio 10/1)
and loaded into the autoclave. After initiation of the heating
the desired temperature was achieved in ∼60 min, and the
autoclave was kept at T = 100 °C for an additional 0, 2 or 4 h.
Analysis of the reaction products was performed on an
Agilent 7890A GC unit equipped with a flame ionization detec-
tor using a HP-5 capillary column (30 m/0.32 ID/0.25 μm). The
obtained products were identified based on the standard refer-
ence compounds. Catalytic activity is expressed in terms of
conversion, which is defined as the fraction of furfural which
has reacted. Selectivity was calculated by dividing the percen-
tage of furfural moles presented in the products by total moles
of converted furfural.
ε (L) = 3.6 cm μmol−1 and ε (B) = 2.05 cm μmol−1 68
.
The presented IR spectra both for d3-acetonitrile and for
2,6-di-tert-butyl-pyridine (DTBP) are not correlated to the stan-
dard mass. A large probe molecule DTBP was used to deter-
mine the concentration of acid sites located on the external
surface of zeolites.69 The adsorption of DTBP was performed
at 150 °C at equilibrium probe vapor pressure with the zeolite
wafer for 15 min. Desorption proceeded at the same tempera-
ture for 1 h followed by collection of spectra at room
temperature.
The size and shape of zeolite crystals were examined by
scanning electron microscopy (SEM, JEOL, JSM-5500LV). For
the measurement, crystals were coated with a thin platinum
layer by sputtering in the vacuum chamber of a BAL-TEC
SCD-050.
Thermogravimetric analysis of the zeolite catalysts was
performed using a TA Instruments TGA Discovery series
equipment and operating at a heating ramp of 10 °C min−1
3. Results and discussion
3.1. Catalyst characterization
3.1.1. XRD. XRD patterns of the MWW materials are
depicted in Fig. 1. Although all the prepared materials possess
the same MWW topology of the layers, the differences in the
layer arrangement can be recognized on the basis of XRD pat-
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 10628–10641 | 10631

View Article Online
Paper
Dalton Transactions
under relatively severe conditions. In the case of MCM-36 the
length of crystals is about 1 μm. It was pointed out that the
large crystals of MCM-36 might have been previously hidden
in the aggregates and became visible only upon swelling and
pillaring treatments.70
3.1.3. N2 adsorption. The textural parameters of the pre-
pared MWW materials were evaluated based on nitrogen iso-
therms (Fig. 3(A)). MCM-22 and MCM-49 zeolites have very
similar isotherms typical of microporous materials according
to the IUPAC classification.71 It was mentioned above that
MCM-22 and MCM-49 crystallize as aggregates of thin sheets.
As a result we observe a narrow hysteresis loop at higher rela-
tive pressure in both zeolites corresponding to nitrogen con-
densation in interlayer voids of the aggregates. The BET
surface areas for MCM-22 and MCM-49 are similar, 505
and 455 m2 g−1, as well as the micropore volumes, 0.151
and 0.165 cm3 g−1, respectively (for more details see Table 2).
Delaminated MCM-56 has the nitrogen isotherm with a dis-
tinctive hysteresis loop corresponding to the adsorption in the
interlayer space of delaminated MWW layers. As MCM-56 is
the intermediate to MCM-49 and cannot be fully crystalline
(discussion vide supra), also the BET surface area is lower in
comparison to MCM-49, 401 vs. 455 m2 g−1. Nevertheless, both
materials differ noticeably in the micropore and total volume
when the micropore volume for MCM-56 and MCM-49 is 0.113
and 0.165 cm3 g−1 and the total pore volume 0.701 and
0.459 cm3 g−1, respectively (see Table 2). The higher micropore
volume in MCM-49 is caused by the presence of the regular
three-dimensional framework with a 10-ring two-dimensional
channel system and the presence of supercages. On the other
hand, in MCM-56 the supercages were not formed as the layers
are not condensed and irregularly aggregated. The layers are
expected to be connected only by hydrogen bonds. The pillared
materials MCM-36 possess isotherms with a gradual increase
in the adsorbed amount in the relative pressure range p/p0
0.01–0.3. It is typical of mesopore materials with the mesopore
size below 4 nm.72 Also the high surface areas for both
MCM-36 materials, 606–762 m2 g−1, indicate the formation of
a new mesopore structure. The micropore volumes decreased
in comparison to MCM-22, 0.062–0.097 cm3 g−1. This is
related to the presence of amorphous silica pillars, which form
Fig. 1 XRD patterns of MWW samples. (1) MCM-22, (2) MCM-36(1),
(3) MCM-36(2), (4) MCM-49, (5) MCM-56.
terns. The XRD patterns of MCM-22 and MCM-49, as the
three-dimensional zeolites, differ from the others and show
three well-defined peaks corresponding to (100) (2θ = 7.1–7.2°),
(101) (2θ ≈ 8°) and (102) (2θ ≈ 10°) reflections, which is in
good agreement with previous reports70 and proves good crys-
tallinity and phase purity of these materials. In contrast, the
two-dimensional MWM forms (MCM-56 and MCM-36) have a
broad band between 8° and 10° indicating the disappearance
of atomic order in the third dimension, along the c-axis. In
addition, both MCM-36 samples are distinctly characterized by
a high intensity low angle peak around ≈2 °C, corresponding
to a d-spacing of about 5 nm. It corresponds well to an idea of
MWW layers of ca. 2.5 nm thick separated by around 2.5 nm
interlayer space filled partially with amorphous silica pillars.53
Similarly, in MCM-56 the broad band between 8° and 10° indi-
cates that the material is composed of thin layers that extend
in the a and b directions with a lack of order in the c-axis direc-
tion.63 As the intermediate to MCM-49, delaminated MCM-56
is not fully crystalline and an amorphous part is most probably
present evaluated from lower intensities of diffraction lines in
comparison to fully crystalline MCM-49. Based on the analysis
of XRD patterns it can be concluded that a series of materials
with MWW topology differing in layer arrangement was pre-
pared in the pure phase (with the exception of MCM-56, which
has an amorphous part present).
a
considerable part of the material. Alike MCM-56, in
MCM-36 materials supercages were not formed and MWW
layers have only large cups on the surface and 10-sinusoidal
channel system inside the layers. The mesopore size distri-
bution of pillared MCM-36 (1) and (2) materials (displayed in
Fig. 3(B)) is centered at 2.3 and 2.5 nm, nicely corresponding
to the data determined from XRD patterns (see section 3.1.1).
The type of obtained isotherms reflects the structural differ-
ences between the samples. Thus, isotherms of MCM-22 and
MCM-49 with the same structural characteristics are very close
to those typical of microporous solids,73 showing a fast
increase in the adsorbed amount occurring at very low p/p0,
followed by long almost horizontal plateau extending up to the
high p/p0 values. A very narrow hysteresis loop observed for
these samples at high p/p0 values indicates the presence of
3.1.2. SEM. Fig.
2 shows SEM images of the MWW
materials. The morphology of the individual crystals of all
materials shows a thin plate-like shape that is a characteristic
of MWW. The samples differ somewhat with regard to the size
of individual crystals and their aggregation. The crystal size
can be influenced by the synthesis time as is demonstrated on
the row MCM-22 (96 hours) ≥ MCM-49 (93 hours) > MCM-56
(33 hours, the intermediate to MCM-49). The individual crys-
tals of MCM-22, MCM-49 and MCM-56 have a length of
0.2–0.5 μm and a thickness of about 0.1 μm. In the case of
MCM-36 materials the aggregation of individual crystals is
visible as they underwent swelling and pillaring treatments
10632 | Dalton Trans., 2014, 43, 10628–10641
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
Fig. 2 SEM images of zeolites samples. (1) MCM-22, (2) MCM-36(1), (3) MCM-36(2), (4) MCM-49, (5) MCM-56.
some slit-shaped, wide mesopores (or narrow macropores) or proves a slight decrease in the amount of micropores in this
platelet particles,73 most probably originating from the aggre- sample due to the absence of bonding between
gation of the lamellar particles.
“MCM-22 monolayers”. In higher p/p0 region of the N2 iso-
A change in the shape of the isotherms for two MCM-36 therm the adsorption amount increases very fast due to a well-
samples reflects the structural changes in these materials caused developed interlamellar space formed by MCM-22 sheets
by the swelling followed by pillaring. The marked increase in the arrayed in a disordered manner forming the so-called partially
adsorption capacity of MCM-36 up to p/p0 = 0.3–0.4 evidences delaminated structure.63,76
capillary condensation and indicates the presence of mesopores
The values of the calculated specific surface areas and total
created by pillaring. At higher p/p0 there is a clear hysteresis loop pore volumes are summarized in Table 2.
of type H4 characteristic of layered materials.74,75
All the prepared samples possess high BET area. The
In comparison with MCM-22, the isotherm of MCM-56 observed differences in BET area between the samples are in
shows a slightly lower N2 adsorption amount at p/p0 = 0–0.6. It full agreement with their structural features. The smallest BET
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 10628–10641 | 10633

View Article Online
Paper
Dalton Transactions
Fig. 3 (A) Nitrogen isotherms of the investigated MWW materials. Open circles – adsorption branch, closed circles – desorption branch. (B) The
mesopore size distribution for pillared MCM-36 materials.
comparison with MCM-36(2) (Table 2). Taking into account
the difference in Si/Al ratio in the reaction mixture for the
preparation of these samples (Table 1), it may be assumed that
the mesopore volume of the final MCM-36 materials is deter-
mined not only by the pillaring procedure, but also by the pro-
perties of the intermediate MCM-22P material.
Table 2 also shows that the BET area of BEA used as a refer-
ence sample differs only slightly from the values determined
for MWW samples. However, the BEA sample has a signifi-
cantly higher Vtot due to the well-developed mesoporosity of
this material.
3.1.4. Acidity. IR spectra of MCM-22, MCM-36, MCM-49
and MCM-56 in the region of OH stretching vibration are
shown in Fig. 4. The well expressed absorption band at
3745 cm−1 is assigned to the terminal silanol groups exposed
at the external surface. This band is the most intensive for two
MCM-36 samples compared with MCM-22 and MCM-49 which
can be explained by the presence of silicate pillars in the struc-
ture and the increased access to the MWW sheets, i.e.
increased external specific surface area as indicated also by
the nitrogen physisorption results (Table 2). Similarly, the rela-
Table 2 Textural properties of MWW materials determined from nitro-
gen adsorption measured at 77 K
BET
Sext.
Vmic
Vtot
Sample
(m2 g−1
)
(m2 g−1
)
(cm3 g−1
)
(cm3 g−1
)
H-MCM-22
H-MCM-36(1)
H-MCM-36(2)
H-MCM-49
H-MCM-56
H-BEA(25)
505
762
606
455
401
608
176
656
430
134
181
222
0.151
0.062
0.097
0.165
0.113
0.174
0.449
0.553
0.398
0.459
0.701
1.017
BET – specific surface area (from the BET method). Vmic – micropore
volume (from the t-plot method). Vtot – total pore volume at p/p0 = 0.97.
area (401 m2 g−1) is observed for delaminated MCM-56 sample
representing a disordered agglomeration of MCM-22 monolayers
without in-registry alignment in the c direction with no inter-
layer links. For MCM-22 and MCM-49 samples the BET area
increases to 505 and 455 m2 g−1, respectively, due to the order-
ing of “MCM-22 monolayers” in the vertical direction resulting
in the formation of a three-dimensional crystalline framework
and creation of 12-ring supercages between the layers. The
further growth of the c parameter of the crystalline framework
due to the introduction of the silica pillars not only influences
the BET area, but also results in a substantial increase of the
external surface area for both MCM-36 samples (Table 2).
The largest volume of micropores is observed for MCM-22
and MCM-49 samples as their pore system consists of “MWW
monolayers” and large cylindrical supercages between them. It
is obvious that delaminated materials not having supercages
and channels connecting them have lower micropore volume
regardless of whether the layers remain disordered (MCM-56)
or connected by siliceous pillars (MCM-36). At the same time,
the disordered arrangement of “MWW monolayers” in
MCM-56 generates significant inner mesoporosity in this
sample. It was shown76,77 that pillaring may also result in the
formation of mesopores. However, according to ref. 77, the
volume of the formed mesopores depends on the pillaring pro-
cedure. Hence, it is not surprising that the two MCM-36
samples studied here have different mesopore volumes;
MCM-36(1) exhibits considerably larger mesopore volume in
Fig. 4 Region of OH stretching vibration of the studied zeolites.
(1) MCM-22, (2) MCM-36(1), (3) MCM-36(2), (4) MCM-49, (5) MCM-56.
10634 | Dalton Trans., 2014, 43, 10628–10641
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
tively low intensive band for Si–OH groups in
MCM-56 material correlates well with the low external surface
area and also the fact that MCM-56 is not a fully crystalline
material (discussion vide supra) (Table 2).
Table 3 Acidic properties of the samples determined by FTIR using
d-acetonitrile and DTBP as probe molecules
CB (DTBP)
CB (Ac)
CL (Ac)
Si/Al
(Ac)b
Sample
(mmol g−1
)
(mmol g−1
)
(mmol g−1
)
The band at 3620 cm−1 is attributed to the acidic bridging
Si(OH)Al hydroxyls associated with the framework Al. The
intensity of this band is the highest for MCM-22 and MCM-49
due to a high structural ordering of these zeolites. It is also
seen that both delamination (MCM-56) and pillaring
(MCM-36) result in a partial degradation of the Brønsted acid
centers. This is in accordance with previous results.64,78 The
difference between the two MCM-36 samples should be also
noted. The intensity of the band at 3620 cm−1 is slightly lower
for MCM-36(1) in comparison with MCM-36(2). Fig. 4 also
shows that spectra of OH-groups of MCM-22, MCM-49 and
H-MCM-22
H-MCM-36(1)
H-MCM-36(2)
H-MCM-49
H-MCM-56
H-BEAa
0.044
0.067
0.053
0.036
0.050
0.125
0.335
0.196
0.249
0.317
0.210
0.353
0.247
0.177
0.245
0.537
0.380
0.220
19.1
29
21.6
11
16
—
a Concentration of acid sites in BEA zeolite was determined by
pyridine-FTIR. b Calculated from the Ac adsorption data.
of the band at 3620 cm−1. The lower concentrations of BAS in
MCM-36 can be explained by the presence of silica pillars,
which generally dilute the concentrations of active sites. It was
already mentioned above that MCM-56 is not fully crystalline
and it is also reflected in its acidity. Since the MCM-36 materials
were prepared from MCM-22P precursors differing in Si/Al
ratio, it is reasonable that the final pillared MCM-36 also differ
in the acid site concentrations. It should be additionally noted
that MCM-49 and MCM-56 possess an increased concentration
of Lewis acid sites (LAS), which is most probably a consequence
of their higher Al content in comparison with MCM-22 and
MCM-36 materials.
The concentration of external Brønsted acid sites deter-
mined by DTBP-FTIR reflects differences in structural features
of the investigated MWW materials. The largest concentration
of the external BAS is observed for MCM-36 and MCM-56,
which correlates with their larger external surface and the
increased concentration of large cups exposed to the outside
and hence the improved accessibility of acid sites. Similarly,
the lower concentration of external BAS in MCM-22 and
MCM-49 is due to their lower external surface area and thus
less cups per gram of catalyst.
MCM-56 contain
a very weak band between 3660 and
3670 cm−1, which is characteristic of OH groups located on
extra-framework aluminum (EFAL) centers.79 It was suggested
that some of these aluminum-containing species may have
rather strong Brønsted-type acidity.80
It is generally accepted that acid sites in the MWW family
materials can have in principle three different locations
depending on packing arrangement of “MWW monolayers”.62
For delaminated MCM-56 these centers can be present in the
sinusoidal 10-ring channels and in cups located on the exter-
nal surface of the crystals. In addition to these two possible
locations, MCM-22 and MCM-49 also contain acid sites in
large supercages formed by condensation of “MWW mono-
layers” that are accessible by another network of 10-ring pores.
Similarly to MCM-56, MCM-36 material does not contain large
supercages. However, it can be expected that the pillaring
procedure can also result in a decrease of accessibility to acid
sites located in the cups due to their blocking by the silicate
pillars.
Small basic molecules, such as pyridine or d3-acetonitrile,
can easily enter the 10-ring channels of zeolites and can be
used as probes for acid sites located inside supercages and the
2D sinusoidal pore system of MWW family materials. Conver-
sely, large basic molecules, such as 2,6-di-tert-butyl-pyridine
(DTBP) cannot enter the 10-ring, so they are adsorbed only on
the Brønsted acid sites located on the external surface of such
zeolitic materials.69 Thus, DTBP was used to determine the
concentration of the external acid sites of the prepared MWW
materials, while the total concentration of Brønsted and Lewis
acid sites, respectively, was evaluated by FTIR of the adsorbed
d3-acetonitrile (Ac).
Table 3 also shows that the reference BEA sample possesses
a relatively high concentration of BAS, both total and those
located on the external surface, 0.353 and 0.125 mmol g−1
,
respectively, which is a consequence of high aluminum
content and its good textural characteristics (Table 2). Besides,
the high concentration of acid sites in BEA determined by
DTBP-FTIR can be partially attributed to the penetration of
DTBP into the pores of BEA and its interaction with both exter-
nal and internal Brønsted acid sites.69
3.2. Catalyst performance
The concentrations of acid sites in the studied zeolites
determined with both probing molecules, Ac and DTBP, is
3.2.1. Conversion and selectivity. It was shown in ref. 52
presented in Table 3. It can be seen that the highest concen- that consecutive interaction of furfural and acetone over acidic
tration of Brønsted acid sites (BAS), 0.335 mmol g−1 and and basic catalysts results in the formation of the following
0.317 mmol g−1, is observed for three-dimensional MWW zeo- main products: FAc (produced by dehydration of an intermedi-
lites with the largest micropores volume, MCM-22 and ate alcohol) and F2Ac (Scheme 2). A distinctive feature of this
MCM-49, respectively. It is in good agreement with the high reaction over zeolites is the formation of an additional reaction
intensity of the band of bridged OH-groups at 3620 cm−1 for product, (FAc)2, which is formed by dimerization of FAc, i.e. it
these samples (Fig. 4). The concentration of BAS in MCM-36 is not a direct aldol condensation product. Additionally, self-
and MCM-56 is lower, which is in line with the lower intensity condensation of acetone also takes place to a small extent
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 10628–10641 | 10635

View Article Online
Paper
Dalton Transactions
Scheme 2 Aldol condensation between furfural and acetone over acidic catalysts.52
leading to diacetone alcohol (DAA) that is subsequently de- BET area and acidity of the reference BEA sample, this might
hydrated and affords mesityl oxide (Scheme 2). be surprising and indicates that other factors, such as the
Furfural conversion over MWW materials in aldol conden- specific structural features of MWW materials, play an impor-
sation of furfural and acetone is depicted in Fig. 5 and 6. tant role. Among the MWW samples, MCM-22 was the most
Fig. 5(A) shows furfural conversion obtained over different active in aldol condensation with furfural conversion of 60%
catalysts after 2 hours of the reaction. It is seen that the activity after 2 hours of the reaction. The conversion of furfural over
of all MWW samples is higher than the activity of BEA zeolite MCM-49 and MCM-56 was only slightly lower being 55 and
which was used as the reference sample. It can be concluded 52%, respectively, after 2 hours. Finally, the lowest conversion
that MWW materials exhibit better activity than the other zeo- of 30 and 35% was obtained over MCM-36(2) and MCM-36(1),
lites investigated so far for the aldol condensation of furfural respectively, i.e. the pillared materials. It can be concluded
and acetone as the following activity sequence was reported that introduction of silicate pillars between the “MWW mono-
previously: BEA > FAU > MOR > MFI.52 Considering the high layers” resulted in deterioration of the catalytic properties of
Fig. 5 (A) Furfural conversion obtained on different catalysts after 2 hours of the reaction at 100 °C and autogenous pressure. (B) Dependence of
furfural conversion on the duration of reaction at 100 °C and autogenous pressure.
10636 | Dalton Trans., 2014, 43, 10628–10641
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
Fig. 6 (A) Selectivity towards the main reaction products in aldol condensation of furfural and acetone over different catalysts after 2 hours at
100 °C and autogenous pressure. (B) Dependence of the yield of reaction products on furfural conversion for all studied MWW samples at 100 °C
and autogenous pressure and reaction time from 0 to 4 hours.
these materials in aldol condensation. This is quite surprising, confirm that aldol condensation of furfural and acetone over
taking into account the larger external surface and better acidic catalysts takes place as one and the same sequence of
accessibility of acid sites which can be deduced from the successive transformations, and furfural conversion is kineti-
DTBP-FTIR results according to which the concentration of the cally governed and no molecular sieving effect is observed.
external BAS is 20–50% larger in MCM-36 than in MCM-22. This could mean that at least the secondary reactions proceed
Moreover, increased activity of MCM-36 as compared with on the external surface. However, when comparing the concen-
MCM-22 was reported for a number of organic reactions.64,78,81 tration of the external BAS with the observed conversion
Additional insight can be obtained when looking at the time (Table 3, Fig. 5), they do not correlate. In fact, the conversion
dependence of furfural conversion. Fig. 5(B) shows that at after 2 hours appears to decrease with increasing concen-
treac. = 0 (i.e. when reaching T = 100 °C inside the batch stirred tration of external Brønsted acid sites, i.e. those determined by
reactor, which takes about 1 hour) the furfural conversion over DTBP-FTIR (Table 3). This suggests that some of the reactions
MCM-36(1) is slightly higher than over MCM-22 catalyst, 20 vs. might proceed at the internal surface that is significantly
18%. When analyzing the shape of the conversion as a func- larger in the case of MCM-22 and MCM-49 (Table 2) that also
tion of time curve it can be evidenced that the reaction rate exhibit higher conversion than both MCM-36 catalysts.
over MCM-22 exceeds that observed for MCM-36(1) and BEA.
It was shown in ref. 52 that the conversion over large-pore
This is a clear indication of catalyst deactivation that was three-dimensional zeolites in aldol condensation of furfural
reported also previously for this reaction system.52 The aspect and acetone is considerably higher compared with medium-
of coke formation in these zeolites and its impact on catalyst pore three-dimensional or wide-pore mono-dimensional
performance will be discussed in the present article.
materials. This has been attributed to the constrained space
Apart from furfural conversion, selectivity towards the main within MFI and MOR pores for the formation of aldol conden-
products is an important parameter of the overall catalyst sation products. Consequently, the observed, though very low
evaluation. The selectivity data are summarized for all investi- conversion, over MFI and MOR was attributed to the activity of
gated catalysts after 2 hours of the reaction (Fig. 6(A)). Gener- the external acid sites.52 From this point of view, it is hard to
ally, the composition of the reaction products obtained over believe that the sinusoidal 10-ring channels of MWW materials
MWW samples corresponds to that observed for zeolites with could be responsible for the high activity of these materials in
different structural types.52 For all the catalysts in the present aldol condensation, which is even higher than that of BEA
study FAc is the main reaction product formed with selectivity zeolite, while MFI was virtually inactive although its 10-ring
≥80%. In accordance with the generally accepted reaction pores are slightly larger than those of MWW materials (MFI,
network, this compound is formed on acidic sites by dehy- 5.3 × 5.6 and 5.1 × 5.5 Å vs. MWW, 5.1 × 4.1 and 5.5 × 4.0 Å).
dration of the corresponding alcohol, FAc–OH.52 Other reac- This points to the conclusion that apart from the external
tion products include F2Ac formed by condensation of a surface acid sites, the acid sites located in the supercages of
furfural molecule with FAc as well as (FAc)2. It was shown in MWW are of great importance for catalytic transformations.
ref. 52 that the latter compound is a result of dimerization of Considering the dimensions of acetone and furfural, 4.1 × 2.4
two (olefinic) FAc molecules. Apparently, the formation of this × 2.7 Å and 5.2 × 3.5 × 1.7 Å, these molecules can reach the
dimer is a distinctive feature of aldol condensation of furfural supercages of MWW through the 10-ring pores, where they
and acetone carried out in the presence of acidic catalysts as could form the first aldol condensation product, FAc (8.9 × 3.6
this product was not observed when hydrotalcites, i.e. basic × 2.2 Å), while the formation of the second aldol condensation
catalysts, were used.33 The yields of FAc, F2Ac and (FAc)2, product, F2Ac (13.3 × 5.3 × 1.9 Å), in the supercages is highly
respectively, increased linearly with the growth of furfural con- unlikely. Since FAc is a rather flat molecule it might be able to
version (Fig. 6(B)) with no obvious dependence on the struc- diffuse out from the supercages through the 10-ring channels.
tural features of the studied catalysts. The obtained results In contrast, the intersections of the 10-ring pores in MFI may
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 10628–10641 | 10637

View Article Online
Paper
Dalton Transactions
not accommodate even FAc, which would explain the very low
activity of MFI reported earlier.52 More importantly, it can help
explain the significantly higher conversion over MCM-22 and
MCM-49 in comparison with MCM-36 catalysts that possess
only cups at the external surface, but no supercages. This con-
clusion is supported by the catalytic performance of the two
MCM-36 catalysts, where MCM-36(1) shows higher activity
than MCM-36(2) despite having lower total concentration of
acid sites (Table 3) but higher concentration of surface acid
sites. It proves that the interaction of furfural and acetone over
these samples generally takes place in the large cups of the
external surface of crystallites.
In lamellar MCM-56, the concentration of acidic centers
determined by DTBP-FTIR is also slightly lower than that in
MCM-36(2) (0.50 vs. 0.53 mmol g−1), but its activity in aldol
condensation is close to that of MCM-22 and MCM-49 (Fig. 5).
MCM-56 is an intermediate crystalline product in the MCM-49
synthesis: the increase of crystallization time results in the
condensation of the lamellar monolayers of MCM-56 into the
three-dimensional MCM-49 framework. It could be assumed
that the MCM-56 sample may contain a small admixture of
condensed MCM-49, which is confirmed by the splitting of the
XRD reflection at 2θ = 8–10°. In that case there should be avail-
able supercages and the observed MCM-56 activity would be
higher than what would be expected solely on the basis of the
concentration of acid sites determined by DTBP-FTIR.
Taking into account the structural features of MWW
materials, an assumption could be made that their catalytic
activity may be considerably influenced by the location and
concentration of Brønsted acid sites. Nevertheless, with
respect to Lewis acid sites the situation is not as clear. Com-
paring the acidic properties of the samples (Table 3) with their
activity in furfural conversion (Fig. 5) it can be concluded that
the catalytic properties of MWW materials show poor corre-
lation with the concentration of Lewis acid sites. This obser-
vation is in total agreement with our results obtained for
different zeolites used as catalysts in aldol condensation.52
However, there is not yet enough information available to
make a definitive conclusion about the role of Lewis acid sites
both in the transformation of the feed molecules and in coke
formation. Apparently, a comprehensive understanding of the
impact of various acid sites is possible only after an additional
study of a series of molecular sieve samples having the same
structure but different Brønsted/Lewis acid sites ratios.
Fig. 7 DTG profiles of spent MWW materials obtained in oxidative
atmosphere (air). (1) MCM-22, (2) MCM-36(1), (3) MCM-36(2), (4)
MCM-49, (5) MCM-56.
region consists of two peaks, at 70–80 °C and 140–150 °C
corresponding to the residual organic substances (acetone,
ethanol, furfural, mesityl oxide) and water removal. The weight
loss at this stage is equal to 4.4–6.4% for all the samples. The
position and intensity of the DTG peaks located in the HT
region show rather poor dependence on structural features of
MWW. Nevertheless, several conclusions can be made. Clearly
distinguishable peaks with maxima at 320–340 °C and
507–563 °C are observed on the DTG curves of all samples.
Since MWW materials have a dual pore system, it could be
assumed that the two peaks on the DTG curve correspond to
the removal of coke from large cups on the external surface of
the zeolite crystals (T = 320–340 °C) and from the 10-ring sinu-
soidal channels (T = 507–563 °C). The determined weight loss
associated with each of these peaks is in line with this assump-
tion. The largest weight loss in the range of T = 320–340 °C is
observed for MCM-36 and MCM-56 samples (Table 4) with
Table 4 Results of TGA for MWW zeolite samples carried out under
oxidative (oxygen) atmospheres
Maximum
value on DTG
curve (°C)
Temperature
region (°C)
Weight
loss (%)
Sample
3.2.2. Deactivation of catalysts. Based on the conversion
time dependence, it has been concluded that the activity of
MWW materials in aldol condensation continuously decreases
with increasing length of the experiment (Fig. 5B). All the cata-
lysts after reaction are black in color suggesting that they have
accumulated a significant amount of carbonaceous deposits
(coke). Hence, after thorough washing with acetone, the
samples were studied by TGA/DTG to evaluate the content and
possibly also the distribution of carbonaceous deposits on the
spent catalysts.
MCM-22
20–200
200–750
72; 145
331
563
72; 144
341
526
72; 146
334
521
72; 170
332
521
5.0
6.0
17.4
4.6
7.3
12.5
4.4
7.6
12.1
6.4
5.3
11.8
6.2
MCM-36(1)
MCM-36(2)
MCM-49
20–200
200–750
20–200
200–750
20–200
200–750
MCM-56
20–200
200–750
72; 151
320
507
Two regions of weight loss, low-temperature (LT) and high-
temperature (HT), are observed on DTG curves (Fig. 7). The LT
7.8
10.8
10638 | Dalton Trans., 2014, 43, 10628–10641
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
easily accessible external large cups due to the delamination aldol condensation activity of these materials in comparison
and pillaring. On the other hand, MCM-22 with a well-develo- with conventional microporous crystalline materials, zeolites,
ped microporosity (Table 2) has the highest weight loss at T ≈ as it is apparent from the comparison with the reference
560 °C (17.4%). However, the MCM-49 sample with crystal sample, a wide-pore three-dimensional BEA sample.
structure and value of micropore volume similar to those of
MCM-22 appears to be an exception from these considerations.
Taking into account the high concentration of LAS in MCM-49
(Table 3), it may be assumed that these active sites could be
4. Conclusions
inactive or less active in coke formation on the MCM-49
A number of MWW materials (MCM-22, MCM-49, MCM-56
crystals.
and MCM-36) were used as catalysts for the aldol condensation
3.2.3. Conversion and selectivity over coked samples. To
of furfural and acetone. Their performance was compared to
evaluate the influence of carbonaceous deposits on the cataly-
that of large-pore zeolite BEA in a batch reactor at 100 °C, auto-
tic properties of MWW materials, the catalyst samples after
genous pressure and a reaction time of 0–4 h. To establish the
reaction were separated from the organic phase, washed with
relationship between the physico-chemical and catalytic pro-
excess of ethanol, dried at 120 °C for 12 hours and re-used in
perties of these micro/mesoporous materials, the samples
the aldol condensation under the same reaction conditions.
were characterized by XRD, SEM, N2 adsorption, FTIR and
The results are presented in Table 5.
TGA. It was found that the acidic solids possessed appreciable
It is seen that furfural conversion over both re-used
activity in the reaction and their use resulted in the formation
MCM-36 samples decreased drastically to about 5–7% from
of products of aldehyde–ketone interaction. Surprisingly,
about 30–35% obtained over the fresh catalysts (Table 5). At
MCM-22 and MCM-49, i.e. condensed materials containing
the same time, the conversion of furfural over re-used MCM-22
supercages, exhibited higher activity than two MCM-36 cata-
and MCM-49 samples decreased only to 33–38% from the origi-
lysts having larger accessible external surface area than the
nal 55–60% conversion of furfural. The results suggest that
condensed materials due to delamination and pillaring. This
rapid formation of carbonaceous deposits which are poorly
was explained by the participation of the active sites located in
soluble in ethanol occurs in the cups located both on the exter-
the supercages in the reaction in addition to the active sites
nal surface of MWW monolayers and in the mesoporous pil-
present in the cups located on the external surface of MWW
lared space between these layers. On the other hand, in
materials. Moreover, all MWW family catalysts gave higher
MCM-22 and MCM-49 possessing large supercages, the coking
conversion than the large-pore zeolite BEA. Nevertheless, fur-
rate and growth of the carbonaceous deposits are limited,
fural conversion decreased rapidly for all catalysts due to coke
which makes it possible to remove such carbonaceous deposits
formation. Unexpectedly, the deactivation was found to be
from supercages by a simple washing with ethanol. Taking
more severe for MCM-36 catalysts than for MCM-22 and
into account the structural differences between condensed and
MCM-49, which was attributed to the reaction taking place
laminated/pillared MWW materials it can be assumed that the
also in supercages that are protected by 10-ring channels from
main reason for the different catalytic properties of the
severe coking. In contrast, the cups located on the external
studied samples MWW is concerned with the presence of
surface were coked rapidly which resulted in low conversion
supercages located between “MWW monolayers” which are
over MCM-36 catalysts as well as in a shift in product selectiv-
accessible through 10-ring pores.
ities. The selectivity to F2Ac and (FAc)2, i.e. products that can
It follows that the presence of cups on the surface of
be formed due to their dimensions only on the external
“MWW monolayers” or supercages (double cups) in the struc-
surface, was lower over coked catalyst than over fresh ones as a
ture of MWW materials introduces a positive impact on the
result of decreased concentration of the active sites on the
external surface.
Table 5 Comparison of catalytic properties of fresh and used MWW
materials. T = 100 °C, t = 2 hours, autogenous pressure
Acknowledgements
Selectivitya (%)
This publication is a result of project no. P106/12/G015 sup-
Sample
Furfural conversion (%)
FAc
F2Ac
(FAc)2
ported by the Czech Science Foundation which is being
carried out in the UniCRE centre (CZ.1.05/2.1.00/03.0071)
whose infrastructure was supported by the European Regional
Development Fund and the state budget of the Czech Repub-
lic. The authors are very grateful to Jiří Čejka and Dana Vitvar-
ová, both from the J. Heyrovský Institute of Physical
Chemistry, Academy of Sciences of the Czech Republic, for
stimulating discussions and FTIR characterization of the
catalysts, respectively.
MCM-22
MCM-22-coked
MCM-36(1)
MCM-36(1)-coked
MCM-36(2)
MCM-36(1)-coked
MCM-49
60.4
33.0
34.6
5.6
29.7
7
82.1
91.7
83.7
6.7
3.5
3.2
0
2.8
0
11.5
4.8
13.1
0
12.4
5.1
9.2
4.5
100
84.8
94.9
85.4
92.5
55.0
37.6
5.4
2.9
MCM-49-coked
a FAc–OH was not observed among the reaction products.
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 10628–10641 | 10639

View Article Online
Paper
Dalton Transactions
27 R. M. West, Z. Y. Liu, M. Peter, C. A. Gärtner
and J. A. Dumesic, J. Mol. Catal. A: Chem., 2008, 296,
18–27.
28 T. N. Pham, D. Shi and D. E. Resasco, Appl. Catal., B, 2014,
145, 10–23.
References
1 D. Kubička, I. Kubičková and J. Čejka, Catal. Rev. Sci. Eng.,
2013, 55, 1–78.
2 C. Somerville, H. Youngs, C. Taylor, S. C. Davis and
S. P. Long, Science, 2010, 329, 790–792.
3 G. W. Huber and A. Corma, Angew. Chem., Int. Ed., 2007,
46, 7184–7201.
4 D. L. Klass, Biomass for Renewable Energy, Fuels and Chemi-
cals, Academic Press, San Diego, 1998.
29 C. J. Barrett, J. N. Chheda, G. W. Huber and J. A. Dumesic,
Appl. Catal., B, 2006, 66, 111–118.
30 R. M. West, Z. Y. Liu, M. Peter, C. A. Gärtner and
J. A. Dumesic, J. Mol. Catal. A: Chem., 2008, 296, 18–27.
31 R. Xing, A. V. Subrahmayam, H. Olcay, W. Qi, P. van
Walsum, G. Pendse and G. W. Huber, Green Chem., 2010,
12, 1933–1946.
5 Y. Sun and J. Cheng, Bioresour. Technol., 2002, 83, 1–11.
6 X. Hu, L. Wu, Y. Wang, Y. Song, D. Mourant, R. Gunawan,
M. Gholizadeh and C.-Z. Li, Bioresour. Technol., 2013, 133, 32 A. S. Ndou, N. Plint and N. J. Coville, Appl. Catal., A, 2003,
469–474. 251, 337–345.
7 J. N. Chheda, G. W. Huber and J. A. Dumesic, Angew. 33 L. Hora, V. Kelbichová, O. Kikhtyanin, O. Bortnovskiy and
Chem., Int. Ed., 2007, 46, 7164–7183. D. Kubička, Catal. Today, 2014, 223, 154–162.
8 D. R. Dodds and R. A. Gross, Science, 2007, 318, 1250–1251. 34 D. Tichit, M. N. Bennani, F. Figueras, R. Tessier and
9 C. Gutsche, D. Redmore, R. Buriks, K. Nowotny, J. Kervennal, Appl. Clay Sci., 1998, 13, 401–415.
H. Grassner and C. Armbruster, J. Am. Chem. Soc., 1967, 89, 35 B. F. Sels, D. E. de Vos and P. A. Jacobs, Catal. Rev. Sci.
1235. Eng., 2001, 43, 443–488.
10 Y. Shigemasa, K. Yokoyama, H. Sashiwa and H. Saimoto, 36 W. Shen, G. A. Tompsett, K. D. Hammond, R. Xing,
Tetrahedron Lett., 1994, 35, 1263.
11 G.-W. Wang, Z. Zhang and Y.-W. Dong, Org. Process Res.
Dev., 2004, 8, 18.
12 Z. Zhang, Y.-W. Dong and G.-W. Wang, Chem. Lett., 2003,
32, 966.
13 V. K. Díez, C. R. Apesteguía and J. I. Di Cosimo, J. Catal.,
2006, 240, 235.
14 C. Noda, G. P. Alt, R. M. Werneck, C. A. Henriques and
J. L. F. Monteiro, Braz. J. Chem. Eng., 1998, 15, 120.
15 I. Di Cosimo, V. K. Diez and C. R. Apesteguia, Appl. Catal.,
A, 1996, 137, 149.
16 M. J. Climent, A. Corma, S. Iborra and J. Primo, J. Catal.,
1995, 151, 60.
F. Dogan, C. P. Grey, W. C. Conner Jr., S. M. Auerbach and
G. W. Huber, Appl. Catal., A, 2011, 392, 57–68.
37 P. A. Zapata, J. Faria, M. P. Ruiz and D. E. Resasco, Top.
Catal., 2012, 55, 38–52.
38 A. Dhakshinamoorthy, M. Opanasenko, J. Čejka and
H. Garcia, Adv. Synth. Catal., 2013, 355, 247–268.
39 A. Dhakshinamoorthy, M. Opanasenko, J. Čejka and
H. Garcia, Catal. Sci. Technol., 2013, 3, 5209–5240.
40 D. P. Debecker, E. M. Gaigneaux and G. Busta, Chem. –
Eur. J., 2009, 15, 3920–3393.
41 C. Xu, Y. Gao, X. Liu, R. Xin and Z. Wang, RSC Adv., 2013,
3, 793–801.
42 S. Abelló, D. Vijaya-Shankar and J. Pérez-Ramírez, Appl.
Catal., A, 2008, 342, 119–125.
17 B. M. Choudary, M. L. Kantam, P. Sreekanth,
T. Bandopadhyay, F. Figueras and A. Tuel, J. Mol. Catal. A: 43 E. Taarning, C. Osmundsen, X. Yang, B. Voss, S. Andersen
Chem., 1999, 142, 361.
18 J. Yu, S. Y. Shiau and A. Ko, Catal. Lett., 2001, 77, 165.
and C. Christensen, Energy Environ. Sci., 2011, 4(3),
793–804.
19 M. J. Climent, A. Corma, H. García, R. Guil-López, S. Iborra 44 C. Perego and D. Bianchi, Chem. Eng. J., 2010, 161(3),
and V. Fornés, J. Catal., 2001, 193, 385. 314–322.
20 H. Naka, Y. Kaneda and T. Kurata, J. Oleo Sci., 2001, 50, 45 H. van Bekkum and H. W. Kouwenhoven, Stud. Surf. Sci.
813. Catal., 2007, 168, 947–998.
21 V. Subramoni, V. K. Rajeevan and B. R. Prasad, Ion Exch. 46 A. Corma, S. Iborra and A. Velty, Chem. Rev., 2007, 107(6),
Technol., 1984, 367. 2411–2502.
22 J. C. A. A. Roelofs, A. J. van Dillen and K. P. de Jong, Catal. 47 P. Mäki-Arvela, B. Holmbom, T. Salmi and D. Murzin,
Today, 2000, 60, 297. Catal. Rev. Sci. Eng., 2007, 49(3), 197–340.
23 M. J. Climent, A. Corma, S. Iborra and A. Velty, Catal. Lett., 48 M. Bejblová, D. Procházková and J. Čejka, ChemSusChem,
2002, 79, 157. 2009, 2, 486.
24 C. Noda Pérez, C. A. Pérez, C. A. Henriques and 49 E. Dumitriu, V. Hulea, I. Fechete, A. Auroux, J.-F. Lacaze
J. L. F. Monteiro, Appl. Catal., A, 2004, 272, 229.
25 S. Abelló, F. Medina, D. Tichit, J. Pérez-Ramírez,
and C. Guimon, Microporous Mesoporous Mater., 2001, 43,
341–359.
J. C. Groen, J. E. Sueiras, P. Salagre and Y. Cesteros, 50 A. Ungureanu, S. Royer, T. V. Hoang, D. Trong On,
Chem. – Eur. J., 2005, 11, 728.
26 F. Winter, X. Xia, B. P. C. Hereijgers, J. H. Bitter, A. J. van
E. Dumitriu and S. Kaliaguine, Microporous Mesoporous
Mater., 2005, 84, 283–296.
Dillen, M. Muhler and K. P. de Jong, J. Phys. Chem. B, 2006, 51 T. Komatsu, M. Mitsuhashi and T. Yashima, Stud. Surf. Sci.
110, 9211.
Catal., 2002, 142, 667–674.
10640 | Dalton Trans., 2014, 43, 10628–10641
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
52 O. Kikhtyanin, V. Kelbichová, D. Vitvarová, M. Kubů and 67 B. C. Lippens and J. H. de Boer, J. Catal., 1965, 4, 319.
D. Kubička, Catal. Today, 2014, 227, 154–162.
53 W. J. Roth, C. T. Kresge, J. C. Vartuli, M. E. Leonowicz,
68 B. Gil, S. I. Zones, S.-J. Hwang, M. Bejblová and J. Čejka,
J. Phys. Chem. C, 2008, 112, 2997.
A. S. Fung and S. B. McCullen, Stud. Surf. Sci. Catal., 1995, 69 A. Corma, V. Fornés, L. Forni, F. Márquez, J. M. Triguero
94, 301. and D. Moscotti, J. Catal., 1998, 179, 451–458.
54 F. S. O. Ramos, M. K. De Pietre and H. O. Pastore, RSC Adv., 70 P. Chlubná, W. J. Roth, A. Zukal, M. Kubů and J. Pavlatová,
2013, 3, 2084. Catal. Today, 2012, 179, 35–42.
55 W. J. Roth, P. Nachtigall, R. E. Morris, P. S. Wheatley, 71 K. S. W. Sing, D. H. Everett, F. A. W. Haul, L. Mouscou,
V. R. Seymour, S. E. Ashbrook, P. Chlubná, L. Grajciar,
M. Položij, A. Zukal, O. Shvets and J. Čejka, Nat. Chem.,
2013, 5, 628.
R. A. Pierotti, J. Rouquerol and T. Siemieniewska, Pure
Appl. Chem., 1985, 57, 603.
72 M. Kruk, M. Jaroniec and A. Sayari, J. Phys. Chem. B, 1997,
101, 583.
73 F. Rouquerol, J. Rouquerol and K. Sing, Adsorption by
Powders and Porous Solids, Principles Methodology and Appli-
cations, Academic Press, London, 1999.
74 J. S. Beck, J. C. Vartuli, W. J. Roth, M. E. Leonowicz,
C. T. Kresge, K. D. Schmitt, C. T.-W. Chu, D. H. Olson,
E. W. Sheppard, S. B. McCullen, J. B. Higgins and
J. L. Schlenker, J. Am. Chem. Soc., 1992, 114, 10834.
56 W. J. Roth, P. Nachtigall, R. E. Morris and J. Čejka, Chem.
Rev., 2014, DOI: 10.1021/cr400600f, article ASAP.
57 W. J. Roth, in Introduction to Zeolite Science and Practice, ed.
J. Čejka, H. V. Bekkum, A. Corma, F. Schuth, Elsevier
Science, 2007, vol. 168.
58 M. E. Leonowicz, S. L. Lawton, R. D. Partridge, P. Chu and
M. K. Rubin, Science, 1994, 264, 1910.
59 S. L. Lawton, M. E. Leonowicz, R. D. Partridge, P. Chu and
M. K. Rubin, Microporous Mesoporous Mater., 1998, 23, 75 O. Franke, J. Rathousky, G. Schulz-Ekloff, J. Stárek and
109–117.
A. Zukal, in Zeolites and Related Microporous Materials: State
of the Art 1994, Studies in Surface Science and Catalysis,
ed. J. Weitkamp, H. G. Karge, H. Pfeifer and W. Hölderich,
vol. 84, Part B, Elsevier, Amsterdam, 1994, p. 77.
60 S. Maheshwari, C. Martínez, M. T. Portilla, F. J. Llopis,
A. Corma and M. Tsapatsis, J. Catal., 2010, 272, 298–308.
61 S. Lawton, A. Fung, G. Kennedy, L. Alemany, C. Chang,
G. Hatzikos, D. Lissy, M. Rubin, H. Timken and 76 A. Corma, U. Diaz, V. Fornés, J. M. Guil, J. Martínez-Tri-
S. Steuernagel, J. Phys. Chem., 1996, 100, 3788. guero and E. J. Creyghton, J. Catal., 2000, 191, 218–224.
62 K. Liu, S. Xie, S. Liu, G. Xu, N. Gao and L. Xu, J. Catal., 77 S. Laforge, P. Ayrault, D. Martin and M. Guisnet, Appl.
2011, 283, 68–74. Catal., A, 2005, 279, 79–88.
63 G. G. Juttu and R. F. Lobo, Microporous Mesoporous Mater., 78 A. Corma, V. Fornes, J. Martinez-Triguero and S. B. Pergher,
2000, 40, 9–23. J. Catal., 1999, 186, 57–63.
64 Y. J. He, G. S. Nivarthy, F. Eder, K. Seshan and J. A. Lercher, 79 A. Lercher, C. Grundling and G. Eder-Mirth, Catal. Today,
Microporous Mesoporous Mater., 1998, 25, 207–224. 1996, 27, 353.
65 Y. Zhang, H. Xing, P. Yang, P. Wu, M. Jia, J. Sun and T. Wu, 80 B. Gil, B. Marszałek, A. Micek-Ilnicka and Z. Olejniczak,
React. Kinet. Catal. Lett., 2007, 90, 45–52. Top. Catal., 2010, 53, 1340.
66 S. Brunauer, P. H. Emmet and E. Teller, J. Am. Chem. Soc., 81 E. J. A. Schweitzer and P. F. Van den Oosterkamp, Micro-
1938, 62, 309.
porous Mesoporous Mater., 1998, 20, 397.
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 10628–10641 | 10641