Acid and redox activity of template-free Al-rich H-BEA* and Fe-BEA* zeolites

Article abstract of DOI:10.1016/j.jcat.2014.06.024

Al-rich template-free BEA^* zeolite (Si/Al 4.6) was synthesized and its structure was analyzed in comparison with the conventional H-BEA^* zeolite of Si/Al 11.3 using XRD, N₂ sorption, SEM, FTIR, ²⁷Al 3Q and ²⁹Si MAS NMR spectroscopy, and DFT calculation of the deprotonation energies. The Al-rich H-BEA^* exhibited a high concentration of Bronsted and Lewis sites, both of high acid strength, although AlSiAl sequences were present in the framework. In Al-rich Fe-BEA^* the exchanged Fe ions, Fe-oxo species, and Fe-oxo oligomers were identified by UV-vis spectroscopy. Cracking of n-decane, alkylation of benzene with benzyl alcohol and hydroamination of styrene with aniline to (anti)Markovnikov phenyl-[2-phenylethyl]amine over H-BEA ^*, and decomposition of N₂O and NH ₃-SCR-NO_x over Fe-BEA^* were investigated in relation to the concentration and nature of acid and Fe-redox sites. The high concentration of Al-related active sites and the highly regular structure of Al-rich beta zeolite are directly manifested in enhanced activity compared with conventional Si-rich beta zeolite.

Full text of DOI:10.1016/j.jcat.2014.06.024

Journal of Catalysis 318 (2014) 22–33
Contents lists available at ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Acid and redox activity of template-free Al-rich H-BEA and Fe-BEA*
*
zeolites
Petr Sazama a, , Blanka Wichterlová , Št eˇ pán Sklenák , Vasile I. Parvulescu , Natalia Candu ,
⇑
a
a
b
b
a
a
a
a
a
Galina Sádovská , Ji rˇ í D eˇ de cˇ ek , Petr Klein , Veronika Pashkova , Petr Štastny
J. Heyrovsk y´ Institute of Physical Chemistry, Academy of Sciences of the Czech Republic, CZ-182 23 Prague 8, Czech Republic
ˇ
´
a
b
University of Bucharest, Department of Organic Chemistry and Catalysis, B-dul Regina Elisabeta 4-12, 030016 Bucharest, Romania
a r t i c l e i n f o
a b s t r a c t
*
Article history:
Al-rich template-free BEA zeolite (Si/Al 4.6) was synthesized and its structure was analyzed in compar-
Received 5 May 2014
Revised 18 June 2014
Accepted 21 June 2014
*
27
29
2
ison with the conventional H-BEA zeolite of Si/Al 11.3 using XRD, N sorption, SEM, FTIR, Al 3Q and Si
MAS NMR spectroscopy, and DFT calculation of the deprotonation energies. The Al-rich H-BEA exhibited
a high concentration of Brønsted and Lewis sites, both of high acid strength, although AlSiAl sequences
*
*
were present in the framework. In Al-rich Fe-BEA the exchanged Fe ions, Fe–oxo species, and Fe–oxo
oligomers were identified by UV–vis spectroscopy. Cracking of n-decane, alkylation of benzene with
benzyl alcohol and hydroamination of styrene with aniline to (anti)Markovnikov phenyl-[2-phenyl-
ethyl]amine over H-BEA , and decomposition of N O and NH -SCR-NO over Fe-BEA were investigated
2 3 x
in relation to the concentration and nature of acid and Fe-redox sites. The high concentration of
Al-related active sites and the highly regular structure of Al-rich beta zeolite are directly manifested in
enhanced activity compared with conventional Si-rich beta zeolite.
Keywords:
Beta (BEA ) zeolite
Acid strength
Cracking of decane
Alkylation of aromatics
Hydroamination of styrene with aniline
Iron zeolites
*
*
*
Ó 2014 Elsevier Inc. All rights reserved.
Fe-beta
NH
3 x
-SCR-NO
Decomposition of N
2
O
1
. Introduction
High-silica zeolites possessing strongly acidic protons, such as
monodimensional 12-ring pores of MOR, and absence of cavities
existing in Y zeolites represent a great advantage for fast diffusion
of bulkier molecules participating in oil processing, dewaxing and
isomerization of paraffins, and benzene alkylation with low olefins,
as well as organic specialty manufacturing. However, a substantial
shortcoming for wider industrial application of the beta zeolite is
the need to use expensive tetraethylammonium hydroxide
(TEAOH) as a template for its hydrothermal synthesis, providing
the beta framework for Si/Al > 11.
*
dealuminated faujasites, mordenites, MFI, and beta (BEA ) zeolites,
represent important zeolite structures employed as catalysts in
fluid catalytic cracking and hydrocracking units converting oil frac-
tions to valuable gasoline and diesel fuels, as well as in production
of petrochemicals [1]. In addition, cationic species in zeolites with
*
low Al concentrations with MFI, BEA , or FER topologies exhibit
extraordinary redox behavior employed in selective oxidation of
Since 2008, the successful synthesis of the template-free and
Al-rich beta zeolite with Si/Al ꢁ 4.5 was reported by Xie et al. [7],
Majano et al. [8], Kamimura et al. [9], and Kubota et al. [10] using
calcined beta seeds added to the aluminosilicate gel. These
procedures yielded a beta zeolite of high crystallinity and well-
hydrocarbons [2], selective conversion of N
lar nitrogen by N O decomposition [3], and SCR of NO/NO
NH [4] or hydrocarbons [5], employed in the end-pipe gases of
2
O/NO/NO
2
to molecu-
2
2
using
3
chemical plants, electric power stations, and exhaust gases of
diesel engines.
developed crystallites with dimensions about 0.3–0.5 lm. As could
*
Zeolite beta (BEA , consisting of three A, B, and C polymorphs)
be expected, the increased concentration of Al in the beta zeolite
with Si/Al ꢁ 4.5 obtained by the template-free route is accompa-
nied by an increased ion-exchange capacity for protons, metal cat-
ions, and metal–oxo complexes.
[
6] is assembled from 4-, 5-, and 6-membered rings forming an
intersecting three-dimensional (3D) channel structure with
1
1
2-membered ring openings of 6.6 ꢀ 6.7 Å (Fig. 1). Its larger 3D
2-ring inner pores compared to 10-ring pores in MFI, only
Very recently, a collaborative effort among several laboratories
[
11,12] yielded a comparison of the structure, framework Al, and
crystal size of the Al-rich beta zeolite (Si/Al 4.6) obtained by
⇑
Corresponding author.
E-mail address: petr.sazama@jh-inst.cas.cz (P. Sazama).
template-free synthesis with those of the standard BEA* (Si/Al
http://dx.doi.org/10.1016/j.jcat.2014.06.024
021-9517/Ó 2014 Elsevier Inc. All rights reserved.
0

P. Sazama et al. / Journal of Catalysis 318 (2014) 22–33
The 15 model: 4 framework Al atoms in one unit cell (Si/Al 15)
23
Al3
Al2
6.6x6.7 Å
Al1
The 5 model: 10 framework Al atoms in one unit cell (Si/Al 5.4)
Al in the Al -O-Si-O-Al chain
3
3
Al3
Al2
Framework of beta zeolite viewed along the [010] axis
Al
O H
Al1
Fig. 1. Illustration of the structure of the beta zeolite and structural models of the zeolite with 4 and 10 Al atoms in one unit cell (molar Si/Al ratio of 15 and 5.4, respectively)
used for the calculation of deprotonation energies of the corresponding Brønsted acid sites.
*
1
2) synthesized using TMAOH and of their activities in several
enhancement of the activity of Al-rich Fe-BEA compared to
*
acid-catalyzed reactions (hydrocarbon alkylation, hydroisomeriza-
tion, and acylation). The authors showed that, in addition to the
higher concentration of Al in the framework, the template-free
beta zeolite exhibited more regular crystalline structure and
conventional Si-rich Fe-BEA in N
SCR-NO to molecular nitrogen.
2
3
O decomposition and NH -
x
slightly larger crystals (0.3–0.5
l
m). While the larger crystals were
2. Experimental
2.1. Preparation of H-BEA and FeH-BEA
connected with faster zeolite deactivation in benzene alkylation
with ethylene, the higher concentration of strong acid sites greatly
increased yields of cracked products in n-decane hydrocracking at
low temperatures. Low activity in the acylation of anisole with
*
*
Al-rich beta zeolite was hydrothermally synthesized by a proce-
dure based on the reports in Refs. [7–9,14] using seeding of cal-
cined beta crystals of Si/Al 11.5 (TZB-212, Tricat) in an amount of
*
acetic anhydride was observed over the Al-rich BEA and its
dealumination to Si/Al 9 was necessary for a high 54% yield of
p-methoxyacetophenone at 99% selectivity due to transport limita-
tions [13]. Kubota et al. [10] showed that steaming of Al-rich BEA
followed by partial dealumination using nitric acid provides a
selective catalyst for high-temperature cracking of hexane to
propene.
0.052 g per 1 g SiO . A starting aluminosilicate mixture with a
2
molar ratio of Al O /SiO /NaOH/H O 1:26.2:17.2:911 was prepared
2
3
2
2
from NaAlO and fumed silica (Cabosil). An amount of 10 g NaAlO2
2
was dissolved in 1000 ml deionized water followed by addition of
42 g NaOH, stirring for 40 min, addition of 96 g silica, stirring for
10 min, and addition of 5 g of beta zeolite seeds, after which mix-
ture homogenization took place for 5 min at room temperature.
The mixture was treated hydrothermally at 120 °C for 125 h in a
2500-ml autoclave with very slow stirring under autogenic pres-
sure. The obtained solid product was separated by filtration,
washed with deionized water, and dried at 80 °C for 6 h. The
zeolite product has a molar Si/Al ratio of 4.6 and was denoted as
BEA-5. The BEA-5 sample was ion-exchanged three times with
1.0 M NH NO , yielding NH -BEA-5 and, after deammonization at
The results obtained to date on the activity of template-free
*
BEA in acid-catalyzed reactions raised new questions on (i) the
*
strength of bridging OH groups in Al-rich BEA , (ii) the structure
and concentration of the metal counter-ions and metal–oxo
species, and (iii) the related activity of H- and metal-Al-rich BEA
zeolite in acid- and redox-catalyzed reactions.
*
In the present study, we compare the synthesized template-free
*
BEA (Si/Al 4.6) and commercial beta zeolite (Si/Al 11.3, synthe-
sized using TMAOH) with respect to (i) the concentration and
strength of the protonic Brønsted sites and the occurrence of Lewis
sites, (ii) the ion-exchange capacity for Fe ions and Fe–oxo com-
plexes, and (iii) the crystal size and texture. These parameters
are related to the activity and selectivity of BEA zeolites with quite
different Al content in their H-form for acid-catalyzed and Fe-BEA
in redox-catalyzed reactions. In this respect, it is demonstrated
that enhancement is attained in cracking of n-decane to olefins
and paraffins as well as transformation of bulkier molecules, the
precursors for synthesis of natural products, pharmaceuticals, cos-
metics, etc. In the alkylation of benzene by benzyl alcohol to diph-
enylmethane and hydroamination of styrene with aniline, the
shape-selective effects, crystal size, and diffusion constraints are
considered in addition to the increased concentrations of active
acid sites. For redox-catalyzed reactions, we demonstrate a marked
increase in the concentration of Fe ions and Fe–oxo species and an
4
3
4
500 °C, obtaining H-BEA-5. The beta zeolite (containing template)
with Si/Al 11.3 was kindly supplied by Zeolyst International (CP
814B -25, Lot. No. 814B-25-1597-77). This type of beta sample is
widely used for preparation of zeolite catalysts and as a standard
for comparing the activities of various zeolites. The sample
denoted as BEA-11 was prepared from this zeolite by calcination
in an ammonia stream at 420 °C for 8 h to remove the template,
according to the procedure first described by Creyghton et al.
[15]. This procedure yields a charge balance of the framework
AlO- by NH ions and thus preserves Al in the framework sites
*
*
+
4
4
[16]. The NH -BEA-11 was deammoniated by calcination in a
4
stream of O at 520 °C for 2 h to H-BEA-11. The chemical composi-
2
*
tions of both H-BEA samples are given in Table 1. Na-forms of beta
+
zeolites for NMR measurements were obtained by Na ion
exchange with NaNO of part of the NH -BEA-5 and H-BEA-11.
3
4

2
4
P. Sazama et al. / Journal of Catalysis 318 (2014) 22–33
Table 1
Characteristics of beta zeolites.
Si/Ala
c
a
c
B
b
c
c
S
V
mi
d
S
e
S
EXT
Sample
Al
L
mi
ꢂ1
(m2
ꢂ1
(cm3
ꢂ1
(m2
ꢂ1
(cm3
g )
ꢂ1
(
mmol g
)
g
)
g
)
g
)
H-BEA-11
H-BEA-5
11.3
4.6
1.35
2.97
0.78
1.43
0.25
0.73
617
482
0.169
0.243
333
470
284
12
a
b,c
d
From chemical analysis.
Concentrations of acid Brønsted and Lewis sites, respectively, from FTIR spectra of adsorbed d
Microporous volume.
3
-acetonitrile.
e
Hypothetical surface area in micropores.
ꢂ1
Fe-BEA-5 (11.7 wt.% Fe, Fe/Al 0.73) and Fe-BEA-11 (5.3 wt.%,
resolution of 1 cm by collecting 256 scans for a single spectrum.
*
Fe/Al 0.77) were prepared by ion exchange of the respective BEA
zeolites with aqueous Fe(II)SO solution under nitrogen. Two
grams of the beta zeolite was slurried in 50 ml of deoxygenated
distilled water. Then 7 g of Fe(II)SO * 7H O was added in the form
The extinction coefficients for C„N vibrations associated with
ꢂ1
ꢂ1
4
Brønsted (2298 cm
,
e
l
= 2.05 cm
l
mol
)
and Al-Lewis sites
ꢂ1
ꢂ1
(2325 cm
,
e
= 3.6 cm
mol ) were taken from Ref. [18].
4
2
The UV–vis–NIR reflectance spectra of hydrated Fe-BEA-5 and
-11 were recorded on a Perkin–Elmer Lambda 950 UV–vis–NIR
spectrometer equipped with an integrating sphere for diffuse-
reflectance measurements covered with Spectralon. Spectralon
also served as a reference.
of a pressed pellet. The exchange proceeded for 14 h at 95 °C under
nitrogen. The slurry was then filtered and the Fe-zeolite was
washed with deoxygenated distilled water and dried under
nitrogen followed by heating (2 °C/min) in a stream of nitrogen
at up to 500 °C maintained for 3 h.
The reflectances were recalculated using the Schuster–
Kubelka–Munk function F(R
diffuse reflectance from a semi-infinite layer. F(R
to the absorption coefficient for F(R ) < 0.5. As F(R
substantially exceed this value in the UV region, F(R
corrected using a correction coefficient = 1–1.2 for
F(R ) = 0–10) derived by Klier [19]. The absorption intensities of
the samples were expressed as F(R COR = F(R /8 ), where
)ꢃ(3
3/8 is the conversion coefficient between true absorption and true
scattering coefficient and F(R ).
1
) = (1 ꢂ R
1
)2/2R
1
, where R
) is proportional
) of Fe samples
) were
1
is the
2
.2. Structural analysis
1
1
1
Zeolites were characterized by XRD recorded by a Bruker D8
1
instrument, Bruker AXS, U.S.A., N
an ASAP2020 Micromeritics apparatus, and scanning electron
microscopy (SEM) using a Jeol JSM-03 instrument. The Al MAS
2
sorption at 77 K obtained on
v/g (v/g
1
27
1
)
1
v g
2
7
and Al 3Q MAS NMR spectra were measured on a Brucker Avance
00 MHz (11.7 T) Wide Bore spectrometer using ZrO rotors with a
rotation speed of 12 kHz. Chemical shifts were referenced to the
aqueous solution of Al(NO
. For the 27Al MAS NMR experiment,
high-power decoupling pulse sequences with a /12 (0.7 s)
excitation pulse and 1 s relaxation delay were applied to fully
5
2
1
3 3
)
p
p
l
2
2
.3. Computational details
2
7
hydrated samples. Al 3Q experiments were performed using
the two-pulse z-filtered procedure. A pulse was used for
excitation and a /3 pulse for conversion. Pulses were individually
optimized for each sample. The relaxation delay was 0.5 s. The pre-
sented 2D contour plots are the result of a 2D Fourier transforma-
tion followed by a shearing transformation. Thus, the F1 axis is the
isotropic dimension and F2 gives the MAS spectrum containing the
second-order quadrupolar line shape. Isotropic chemical shifts
were obtained from the first moment analysis of the 2D plot using
the following relationship:
.3.1. Structural models of the beta zeolite
p
Two models of the beta zeolite with P1 symmetry were used to
p
investigate the effect of Si/Al on the deprotonation energies. The
first model (hereafter the 15 model) features 4 Al atoms in one unit
cell (Si/Al 15) and the second model (hereafter the 5 model)
contains 10 Al atoms in one unit cell (Si/Al 5.4) (Fig. 1). All the
ꢂ
AlO
4
entities are compensated for by protons in the AlOHSi
, Al , and Al atoms are substituted for in both the
groups. The Al
1
2
3
models in the same framework positions to allow comparison of
the deprotonation energies. According to our previous experimen-
tal studies on the distribution of Al atoms in beta zeolites with Si/
Al > 11 [16], two different distributions of the framework Al atoms
in two double 6MR units were chosen in the 15 model. One double
diso ¼ ð17dF1 þ 10dF2Þ=27:
29Si MAS NMR single pulse and cross-polarization spectra were
measured at a rotation speed of 5 kHz, with a /6 (1.7 s) excitation
pulse and a relaxation delay of 30 s for single pulse spectra. For the
cross-polarization spectra, pulse sequences with 50% ramp CP pulse,
2
delay were employed. Si high-power decoupling spectra were
decomposed to the Gaussian bands using Microcall Origin 4.1 soft-
ware (Microcall Software Inc., USA). The framework aluminum con-
tent (Si/AlFR) was estimated according to the formula
p
l
6
MR cage features two Al atoms (Al
sequence in one 6MR (Fig. 2A), while the other double-6MR unit
contains two single Al atoms (one of them is Al ) located in two
MRs (Fig. 2B). Two variants of the distribution of Al in the 5 model
were employed. Both feature the same location of the Al SiSiAl
sequence in one 6MR as in the 15 model: however, the Al and
Al atoms also form Al SiAl and Al SiAl sequences in addition to
creating Al SiSiAl as in the 15 model (Fig. 2A). The Al atom forms
an Al SiAl sequence in variant 1 (Fig. 2B), while the Al atom is a
1 2 1 2
and Al ) forming an Al SiSiAl
3
000 ls contact time, high-power decoupling, and 5 s relaxation
6
2
9
1
2
1
2
1
2
1
2
3
3
3
Si=AlFR ¼ I=ð0:25I
1
þ 0:5I
2
Þ;
single Al atom in variant 2. AlSiAl sequences were employed in
the 5 model, since the Si(2Si,2Al) atoms were observed by 29Si
MAS NMR in BEA-5 (see Fig. 3). The starting geometry of both
the models was generated from the experimental orthorhombic
structure of polymorph A of the beta zeolite (see http://www.iza-
structure.org/databases: space group, P4122; cell parameters,
a = 12.632, b = 12.632, c = 26.186 Å). Details of the electronic
structure calculations, geometry optimizations, and molecular
dynamics are given in Supplement I.
where I denotes the intensity of the NMR lines corresponding to
the Si(3Si,1Al), I that of those corresponding to the Si(2Si,2Al)
building unit, and I the total Si intensity; for details see Ref. [17].
1
2
2
9
The concentrations of acidic Brønsted and Al-Lewis sites in
H-BEA zeolites were determined by adsorption of d
3
-acetonitrile
(
13 mbar) on zeolite samples evacuated at 500 °C, carried out at
2
98 K with subsequent evacuation for 15 min at RT. The IR spectra
were recorded on a Nicolet 6700 FTIR spectrometer operating at a

P. Sazama et al. / Journal of Catalysis 318 (2014) 22–33
25
Fig. 2. (A) Positions of the Al
atoms form Al SiAl and Al SiAl sequences, besides creating Al
atoms in a double-6MR: The 15 model and the 5 model, variant 2 with a single Al
1
and Al
2
atoms forming an Al
1
SiSiAl
2
sequence in a double-6MR: The 15 model (left) and the 5 model (right). In the 5 model, the Al
1
and Al
2
3
1
2
1
SiSiAl
2
. Si atoms are in gray, Al atoms in yellow, O atoms in red, and H atoms in white. (B) Positions of single Al
3
atom (left) and the 5 model, variant 1 with Al
3
SiAl (right). (For interpretation of the
references to color in this figure legend, the reader is referred to the web version of this article.)
A
B
C
BEA-11
1000
8
6
4
2
00
00
00
00
0
BEA-11
BEA-11
BEA-5
BEA-5
BEA-5
0.8
5
10 15 20 25 30 35 40
0.0
0.2
0.4
0.6
1.0
2
(degree)
P/P0
2
Fig. 3. Characteristics of beta zeolites. (A) SEM micrographs, (B) X-ray diffraction patterns, and (C) N -sorption isotherms at 77 K.
2
.3.2. Deprotonation energy
The deprotonation energies of the three protons compensating
the corresponding Brønsted sites [20–22]. Details regarding calcu-
lation of the deprotonation energies employing periodic methods
are discussed elsewhere [21]. The differences between two depro-
tonation energies (i.e., relative deprotonation energies) are used to
compare the acidity of the corresponding two Brønsted sites.
for the three Al atoms that are common to the 15 and 5 models
ꢂ
+
were calculated using the equation Z ꢂ H ? Z + H [20–22]. The
calculated deprotonation energies are a measure of the acidity of

2
6
P. Sazama et al. / Journal of Catalysis 318 (2014) 22–33
2
2
.4. Catalytic reactions
flow rate of the feed (mol of N
W is the weight of the catalyst (kg), in agreement with the reported
reaction rate first-orders with respect to N O [23], NO [24] and
2
O or NO or n-decane per h) and
.4.1. Cracking of n-decane
2
n-Decane cracking was carried out in a fixed-bed flow-through
n-decane [25]. The TOF values were calculated for the respective
reactor at 480–520 °C and atmospheric pressure. The flow rate of
reactions per Al, Brønsted OH, and Fe.
ꢂ6
ꢂ1
n-decane was 1.62 ꢀ 10 mol s at a He/decane molar ratio of
Small crystallites of BEA-5 and BEA-11 (ꢁ0.3–0.5 and ꢁ0.05
lm,
2
9 and with typical catalyst weight of 14 mg, corresponding to
respectively) support their minimum contribution to the overall
ꢂ1
WHSV 120 h . The reaction products were analyzed by an on-line
connected Agilent 6890 gas chromatograph equipped with an
reaction rates for the flow rates (cf. the conditions given in Refs.
[26,27]) employed in NH -SCR-NO , N O, decomposition and dec-
3 x 2
Al
2
O
3
/KCl capillary column and a flame ionization detector.
ane cracking. For the bulkier molecules participating in alkylation
of benzene with benzyl alcohol and hydroamination of styrene
with aniline, carried out in a liquid phase, diffusion constraints cer-
tainly decrease the measured reaction rates. As the BEA-5 zeolite
exhibits small crystals that are, nevertheless, much larger than
BEA-11, the higher conversions obtained over BEA-5 compared to
those over BEA-11 cannot be explained by the diffusion effect.
2.4.2. Alkylation of benzene by benzyl alcohol to diphenylmethane
Benzylation of benzene to diphenylmethane was performed as
follows: Benzene (20 or 50 mmol) was mixed with benzyl alcohol
20 mmol) in a 20 ml glass vial. The dried catalyst (150 mg) was
(
added to this mixture, which was then heated and kept at 80 °C
with stirring (1400 rpm) for 2–24 h. After the reaction, the catalyst
was filtered out and the reaction mixture was analyzed by FID-GC
and MS-GC chromatography (Thermo Electron Corporation
Instrument). Both reagents were from Sigma–Aldrich (P99.0%).
3. Results and discussion
3.1. Synthesis and analysis of the structure of H-BEA
2.4.3. Hydroamination of styrene with aniline to phenyl-[2-
phenylethyl]amine and phenyl-[1-phenylethyl]amine
The beta zeolite with a low Si/Al molar ratio of 4.6 (BEA-5) was
Hydroamination of styrene with aniline was performed in a
stainless steel autoclave loaded with 50 mg of catalyst, by using
prepared by template-free synthesis and the use of seeding crys-
tals. The obtained BEA-5 consisted of well-developed crystals
ꢂ1
1
1
mmol aniline and 2 mmols styrene under autogenic pressure at
50 °C. Toluene (4 ml), a non-polar aprotic solvent, was used. The
0.3–0.5
l
m in size with a high surface area of 482 m g , mainly
2
ꢂ1
in micropores (470 m g ), as estimated from the SEM images
and adsorption isotherms of N at 77 K (Fig. 3 and Table 1). BEA-
11 exhibited a lower micropore surface area (333 m g ) and
reactants and products were analyzed by GC–MS using a Trace
GC 2000 instrument coupled with DSQ MS from Thermo Electron
Corporation. The structure of the resulting products was confirmed
2
2
ꢂ1
3
ꢂ1
higher external surface area (284 m g ) due to its very small
m). The intensities and patterns of the X-ray
diffraction lines for both BEA zeolites are characteristic of the
1
13
by H and C NMR spectroscopy using a Bruker AV 400 spectrom-
eter. Analytical data for the (anti)-Markovnikov adduct are given in
Supplement II.
crystallites (ꢁ0.05
l
*
*
well-developed crystalline structure of BEA topology formed by
intergrowth of polymorphs A and B (Fig. 3; cf. Ref. [28]). The shift
of all the reflections of ca. 0.3 2h to the lower values observed for
BEA-5 corresponds to a higher Al concentration than for BEA-11.
Sharper and more intensive reflections of BEA-5 indicate highly
regular structure and slightly larger crystals.
2
.4.4. Selective catalytic reduction of NO
Selective catalytic reduction was carried out with catalysts
located in a quartz tubular down-flow reactor and using the reac-
tant gases (NO, NH , and O ) and the carrier gas (He). The concen-
x
by ammonia
3
2
trations of reactants and products were read stepwise at steady
state at temperatures from 500 to 200 °C using an UV photometric
In contrast to other zeolites, the structure of beta zeolite is con-
nected with a reversible change in the coordination of the frame-
work Al atoms in the hydrated zeolite, depending on the type of
counterion balancing the framework negative charge. It is gener-
ally accepted that part of the framework Al atoms (with tetrahedral
coordination in Na form) exhibit octahedral coordination in the
analyzer (ABB AO2000-Limas11UV) for NO, NO
infrared absorption photometer (ABB Uras26 EL3020) for N
The experimental conditions for the SCR reaction were as follows:
2
, and NH
3
and an
2
O.
3
ꢂ1
ꢂ1
total gas flow rate 350 cm min and catalyst weight 20.5 mg cor-
*
responding to GHSV 510,000 h ; NO 500 ppm; NH
3
520 ppm; O
2
hydrated protonic form of the BEA zeolite [29–35]. Thus, the coor-
2
%; balance He. NO
NO and NO to N . The concentration of N
all cases was <2 ppm.
x
conversion was defined as the reduction of
dination and local surroundings of the Al sites were analyzed by
2
7
29
2
2
2
O in the products in
the Al and Si MAS NMR spectra of Na-beta zeolites shown in
2
7
Fig. 4 and Supplement III. Al single pulse MAS NMR spectra of
both samples showed the presence of tetrahedral Al atoms
2
7
2.4.5. Decomposition of N
2
O
reflected in the Al resonance with the observed shift around
55 ppm, corresponding to the framework Al atoms. Resonances
Decomposition was performed with 100 mg of the catalyst
2
7
placed in a fixed-bed plug flow-through microreactor in the tem-
perature range 350–425 °C. The catalyst was pretreated in a
helium stream at 450 °C for 2 h and then in an oxygen stream at
with Al observed shifts around 0 ppm reflecting extraframework
octahedral Al atoms, and shifts between 30 and 45 ppm attribut-
able to extraframework penta-coordinated Al atoms or perturbed
tetrahedral Al atoms (framework or extraframework), were not
4
50 °C for 1 h. The feed contained 1000 ppm N
2
O, 500 ppm NO,
2
7
3
% O and 1% H O in He (similar composition to that of the end-
2
2
observed. The Al 3Q MAS NMR experiment provides deeper
insight into the local arrangement of the framework Al atoms in
pipe gases in nitric acid plants) with a total flow rate of
ꢂ1
ꢂ1
27
3
N
00 ml min , corresponding to GHSV 90,000 h . Analysis of
O/NO/NO , N , and O at the reactor outlet was performed using
an Advanced Optima (ABB) IR analyzer, AH (MLU) chemilumines-
cence O/NO analyzer, and Hewlett Packard 5890 II Gas
the zeolite. Two resonances can be recognized in the Al 3Q
2
2
2
2
MAS NMR spectra of both Na-BEA-5 and -11 (Fig. 4A). The predom-
2
7
inant resonance with the
Al isotropic shift at 58.6 ppm
27
N
2
x
(F1 = 59.5 ppm and F2 = 57.0 ppm) and a shoulder with a Al iso-
tropic shift at around 55.8 ppm (F1 = 56.5 ppm and F2 = 54.5 ppm)
were observed. Both resonances are well known to be present in
the Al MAS NMR spectra of BEA zeolites, and the Al 3Q MAS
NMR spectra of Na-BEA-5 and -11 do not differ from those already
reported for Si-rich samples [36–39]. The only remarkable feature
chromatograph.
The rates of N
to N
were calculated from the conversions using the pseudo-first-order
kinetic equation written as r = F/W * (ꢂln(1 ꢂ x)), where F is the
2
2 2 3 x
O decomposition to N and O , and NH -SCR-NO
*
27
*
27
2
over Fe-BEA catalysts and of n-decane cracking over H-BEA

P. Sazama et al. / Journal of Catalysis 318 (2014) 22–33
27
A
B
Res.(ppm) A (%) Si type
-98.4
3.1
5.9
SiOH
BEA-11
-102.9
-105.1
-110.9
-115.6
Si(3Si,1Al)
Si(3Si,1Al)
Si(4Si)
BEA-11
26.1
52.5
12.4
Si(4Si)
Si/Al(fr) = 12.5
a
b
Res.(ppm) A (%) Si type
BEA-5
-88.9
-93.2
-98.1
2.2
2.0
SiOH
SiOH
17.1
49.0
24.1
5.6
Si(2Si,2Al)
Si(3Si,1Al)
Si(4Si)
-
-
-
104.1
109.6
114.4
Si(4Si)
BEA-5
Si/Al(fr) = 4.8
a
b
-70
-80
-90
-100
-110
-120
-130
29Si chem. shift (ppm)
Fig. 4. MAS NMR spectra of hydrated Na-beta zeolites. (A) 2D plot and F1 and F2 projections of the 27Al 3Q MAS NMR spectra and (B) 29Si MAS NMR spectra with relative areas
of the individual resonances and their attribution to Si(xSiyAl) atoms: (a) experimental data and (b) Gaussian bands of the individual resonances.
that can be connected with the 27Al MAS NMR spectrum of
Na-BEA-5 is the low intensity of the resonance at 55.8 ppm. The
relative intensity of this resonance was reported to increase with
in the BEA zeolites [43]. In addition to the intensities of the indi-
*
vidual IR vibrations, the zeolites differ in the position of the bands
of the noninteracting bridging OH groups at 3614 and 3611 cm
respectively. Both zeolites exhibited substantial intensities in a
broad absorption range from 3620 to 3200 cm , reflecting vibra-
tions of interacting OH groups, a band at 3745 cm of the terminal
SiOH groups, and a low intensity around 3735 cm , indicating
inner SiOH in the defect sites [44]. H-BEA-5 possesses much higher
intensity of both the noninteracting (3614 cm ) and mutually
ꢂ1
,
*
dealumination of the BEA samples, but assignment of both the res-
ꢂ1
ꢂ1
onances to the Al atom in a specific location is not yet clear. Quan-
titative analysis of the 29Si MAS NMR spectra of both the hydrated
Na-beta zeolites (Fig. 4B and Table 1) yielded framework Si/AlFR
ratios of 4.8 and 12.5 for BEA-5 and BEA-11, respectively, compara-
ble with those obtained from the chemical analysis (see Table 1).
Therefore, analysis of both the 27Al and Si MAS NMR spectra of
ꢂ1
ꢂ1
29
ꢂ1
affected (3620–3200 cm ) bridging OH groups. The number of
*
Na-BEA-5 and Na-BEA-11 indicates well-developed regular BEA
structures with a predominance of Al atoms in tetrahedral coordi-
nation in the framework. The 29Si MAS NMR spectrum of BEA-11 is
similar to those reported previously for beta zeolites (Si/Al > 11)
synthesized using an organic template [16], indicating the only
possible Al presence as that of Si(1Al3Si) atoms (Fig. 4) [17,38].
The spectrum of BEA-5 contains an additional resonance at
ꢂ98.1 ppm representing Si(2Si,2Al) atoms in AlSiAl sequences with
two next nearest neighboring Al atoms [31,40]. The spectral contri-
bution of this Si resonance is 17%. In our previous study on a BEA-
3
614
3
611
0.2
SiOHAl
3
630
3580
SiOH
1
1 sample, analysis of the exchange capacity of bare Co(II) ions
showed that the framework Al atoms were present in AlSiSiAl
sequences in 6MRs and as single Al atoms located in different rings
Interacting OH
[
16]. Therefore, the major differences between BEA-5 and BEA-11
are the high concentration of Al in the framework and the occur-
rence of AlSiAl sequences in the framework of BEA-5.
The strength of the bridging SiOHAl protonic sites in H-BEA-11
and also other Si-rich H-zeolites with Si/Al > 12 has been shown
not to be affected by the concentration and distribution of Al atoms
between AlSiSiAl sequences and single Al atoms [41]. However, it is
well known that in Y zeolites with Si/Al ꢁ 3, with AlSiAl sequences
the protonic sites are mutually affected and exhibit much lower
acid strengths of the bridging OH groups [42]. Fig. 5 depicts the
4000
3600
3200
2800
-1
Wavenumber / cm
IR spectra in the region of the OH groups (
m(O–H) vibration mode)
Fig. 5. FTIR spectra of H-BEA-5 (
) and H-BEA-11 (
) zeolites in the region of
for H-BEA-5 and H-BEA-11, which are typical for the OH vibrations
OH groups after evacuation at 450 °C with detail of the bands of Brønsted OH.

2
8
P. Sazama et al. / Journal of Catalysis 318 (2014) 22–33
ꢂ1
terminal SiOH groups (3745 cm ) was very small in BEA-5, com-
pared with their high intensity for small crystallites and the high
1 2 1 2
that both the Al and Al atoms also form Al SiAl and Al SiAl
sequences because of a higher number of Al atoms in the 5 model
(Table 2). Two variants of the distribution of the framework Al
*
external surface area of BEA-11. Neither of the BEA zeolites exhib-
ited extraframework Al–OH groups of vibrations at approx.
atoms were employed in the 5 model, one with an Al
an Al SiAl chain (variant 1) and the other one with a single Al
(variant 2). Al is a single Al atom in the 15 model and the corre-
sponding Brønsted OH group yields a hydrogen bond with an oxy-
gen atom of a SiOSi chain in the 6MR, whereas there is no such
hydrogen bond in either of the variants of the 5 model. Despite
the absence of a hydrogen bond in both variants, the deprotonation
3
atom forming
ꢂ1
3
650 cm , supporting the presence of Al atoms in the framework
3
3
atom
sites.
Quantitative analysis of the IR intensities of the characteristic
C„N vibrations ( (C„N mode) of adsorbed d -acetonitrile (spectra
3
m
3
not shown) distinguishing between Brønsted and Lewis sites indi-
cated that H-BEA-5 contained approximately twice the concentra-
tion of Brønsted sites and a three times higher concentration of
Lewis sites than H-BEA-11 (Table 1). While Na-BEA-5 with Al
energy is slightly higher by 0.3 (single Al
kcal mol than that of the 15 model (Fig. 2B, Table 2).
3
) and 1.1 (Al
3
in Al
3
SiAl)
ꢂ1
+
atoms charge-balanced by Na ions contained Al atoms exclusively
The increase in the deprotonation energies for the 5 model rel-
ꢂ1
in T
d
coordination (absence of elevated intensity at 30–45 ppm and
ative to the 15 model (up to 6.6 kcal mol ) is smaller than the dif-
0
ppm), the presence of Lewis sites in H-beta zeolites is not surpris-
ference between the lowest and highest deprotonation energies
ing, as the occurrence of unsaturated Al atoms (with electron
acceptor properties) preserving their location in the framework
calculated for the individual Brønsted sites of AlOHSi groups for
ꢂ1
the 15 model (i.e.,
D
= 8.7 kcal mol because EDP = 81.6 (Al
1
) and
*
ꢂ1
as well as those in extraframework positions is typical of BEA
2
90.3 (Al ) kcal mol ). Thus calculations show that the strengths
*
topology [16,45]. If the Al atoms in H-BEA remain in the frame-
of the protonic acid sites in the 5 and 15 models does not signifi-
cantly differ, being in the range of differences in the deprotonation
energy values among the protons associated with AlOHSi groups at
individual T sites of the framework. This finding is also consistent
with the small shift in the OH stretching frequency in the FTIR
spectra.
work, the concentration of total framework Al (obtained from
2
9
Si MAS NMR, Table 1) should equal the sum of the concentration
of Brønsted sites and twice the concentration of Lewis sites:
+ 2c . The obtained concentrations of Brønsted and Lewis
cAl = c
B
L
sites for both BEA-5 and BEA-11 zeolites (Table 1, differences <8%
from total Al concentrations) indicate that a highly predominant
number of the Al atoms exhibiting a Lewis character preserve their
location in the framework positions. The total Al content in BEA-5
increased ca. 2.2 times compared to BEA-11, while the concentra-
tions of Brønsted and Lewis sites increased by 1.8 and 2.9 times,
respectively. We can speculate that AlSiAl sequences present only
in BEA-5 might be a particular reason for the increased concentra-
tion of Lewis sites.
It can be concluded that the synthesized BEA-5 zeolite with Si/
Al 4.6 with high concentration of Al located in T coordination in
d
the framework is formed from small well-developed crystallites
*
of BEA topology, without mesopores and of negligible external
surface (ꢁ3%), unusually low compared to the conventional beta
zeolites. To accommodate a high concentration of Al in the beta
zeolite, in addition to AlSiSiAl sequences occurring in one ring
and single Al atoms located in different rings in BEA-11, the AlSiAl
sequences are formed in the framework of BEA-5. These sequences
could lead to an increased fraction of Al atoms possessing electron-
acceptor Lewis character in H-BEA-5, but still preserving the Al
location in the framework. The acid strength of protons of the
bridging OH groups as a total is high not significantly decreased,
as supported by the very small shift in the structural OH vibrations
and less-than-significant differences in the deprotonation energies
for OH groups related to the individual Si–Al sequences.
Although the relationship between the shift in the frequency of
the stretching vibration mode of bridging OH groups and the acid
strength is not completely straightforward, the increase in the
ꢂ1
OH stretching frequency from 3611 to 3614 cm for noninteract-
ing SiOHAl groups of BEA-5 might reflect the lower strength of
some Brønsted sites. For example, a shift in the OH vibrations of
ꢂ1
1
0 and 20 cm was reported by Chu and Chang [46] for evidently
less acidic GaOHSi and FeOHSi groups, respectively, compared with
the Al-substituted ZSM-5 framework. In this light, the shift of
ꢂ1
3
cm in the OH vibration does not indicate a significant change
3.2. Acid-catalyzed reactions
in the acid strength of the zeolite OH groups. It implies that,
although the BEA-5 zeolite contains a substantially higher concen-
tration of Al than BEA-11, its bridging OH groups preserved their
high acid strength. Nevertheless, lower strength of some of the
OH groups adjacent to AlSiAl sequences cannot be completely
excluded.
3.2.1. Cracking of n-decane
A linear relationship between the concentration of the bridging
OH groups and the cracking rate was reported for beta and ZSM-5
zeolites [47,48]. Corma et al. [49] showed that the beta zeolite with
7
Si/Al 12 provides slightly higher activity in cracking C –C14 alkanes
The deprotonation energies of the Brønsted sites Al
OHSi, and Al OHSi were calculated for both the 15 and 5 mod-
els, corresponding to Si/Al 15 and 5.4, respectively (see Figs. 1 and
), to estimate how the acid strength of the individual OH groups
depends on Si/Al and the distribution of Al atoms in the frame-
work. Our calculations for the 15 model for Al and Al forming Al1-
SiSiAl (Fig. 2A, Table 2) show that the OH group belonging to Al is
free, while the other OH group of Al OHSi forms a hydrogen bond
with the oxygen atom of a SiOSi chain in the 6MR. Our results
1
OHSi,
than the traditionally employed USY. They assumed that the
absence of large cavities in the beta zeolite might be responsible
for higher selectivity for olefins and lower rate of coke formation
than with USY [49]. However, the necessity of using an expensive
organic template for synthesis of the beta zeolite limits its use in
large-scale catalytic cracking. Fig. 6 depicts the rate of n-decane
cracking over H-BEA-5 compared with that of H-BEA-11 and a
commercially employed H-USY-6. The reaction rate over H-BEA-5
is approximately twice as high as that over H-BEA-11 and
H-USY-6. This is also reflected in the higher apparent activation
Al
2
3
2
1
2
2
1
2
reveal higher deprotonation energies for Al
the 5 model of BEA-5 (
1
OHSi and Al
DP 6.6 and 3.7 kcal mol , respectively)
2
OHSi in
ꢂ1
D
E
energies for H-BEA-11. The TOF values calculated per Brønsted site
ꢂ1
than in the same acidic sites in the 15 model of BEA-11 (Table 2),
since both Al OHSi and Al OHSi yield hydrogen bonds with the
oxygen atom of the SiOSi chains in the 5MR and 6MR, respectively.
The higher deprotonation energy of the OH group belonging to Al
at 500 °C are similar for H-BEA-5 (840 h
)
and H-BEA-11
ꢂ1
1
2
(833 h ), while those calculated per Al differ, being lower for
H-BEA-5, as it contains a higher relative concentration of Lewis
sites, which are assumed not to be active in decane cracking
(Table 3). The selectivity of n-decane cracking to individual
2
agrees with the fact that the corresponding hydrogen bond is
stronger in the 5 model (the OH. . .O distance is 1.71 Å) than in
the 15 model (the OH. . .O distance is 1.79 Å). It should be noted
*
products was comparable for both H-BEA zeolites, as given in Sup-
2 4
plement IV, with selectivity 41.1 and 42.0% for C –C olefins, 30.6

P. Sazama et al. / Journal of Catalysis 318 (2014) 22–33
29
Table 2
Deprotonation energies (EDP) of the corresponding Brønsted acid sites for the calculated 15 (4 Al atoms in one unit cell, Si/Al 15) and 5 models (10 Al atoms in one unit cell, Si/Al
.4).
5
ꢂ1
AlOHSi group
The 15 model
Sequence
The 5 model
Sequence
D
E
DP (kcal mol
)
ꢂ1
ꢂ1
EDP (kcal mol
)
E
DP (kcal mol
)
Al
Al
Al
1
OHSi
2
OHSi
3
OHSi
Al
Al
1
SiSiAl
SiSiAl
2
2
in 6MR
in 6MR
in 6MR
81.6
90.3
86.8
Al
Al
Al
1
1
3
SiSiAl
SiSiAl
SiAl ,
2
in 6MR and A
in 6MR and Al
l
SiAl
SiAl
88.2
94.0
87.9
87.1
6.6
3.7
1.1
0.3
1
2
a c
2
Single Al
3
b c
3
single Al ,
a
b
c
Variant 1 of the 5 model with the Al
Variant 2 of the 5 model with the single Al
Both variants 1 and 2 feature the same location of Al
3
3
atom forming a Al SiAl sequence.
3
atom.
1 2
SiSiAl .
of larger crystallites of H-BEA-5 with predominant microporous
surface (97% of the total surface area) in contrast to that in H-
BEA-11 exhibiting high external surface area (see Table 1). When
an excess of benzene (benzene/BzOH 2.5) was used (Fig. 7A), a fur-
ther increase in the conversion of BzOH up to 85% and high selec-
tivity (56.6%) for DPM were achieved. After 24 h, even in an excess
of benzene, the conversion increased to 99% and the selectivity for
DPM remained almost unchanged. In addition, no significant
change in the activity was observed in three successive regenera-
tion cycles (Supplement V).
7
7
6
6
.5
.0
.5
.0
lnA = 9.0
Ea = 12.1 kJ.mol
-1
lnA = 9.4
Ea = 18.8 kJ.mol
-
1
H-BEA-5
H-BEA-11
3
.2.3. Hydroamination of styrene with aniline
Scheme 2 illustrates amination of styrene by aniline to phenyl-
0
.00120
0.00125
0.00130
0.00135
1
/T (K-1
)
[2-phenylethyl]amine and phenyl-[1-phenylethyl]amine. Fig. 7B
shows high conversion of aniline in styrene hydroamination
Fig. 6. Cracking of n-decane as a function of temperature over H-BEA-5 and H-BEA-
1 shown as an Arrhenius plot with the pre-exponential factors lnA and activation
(
87%) with high selectivity for the Markovnikov isomer (68%) over
H-BEA-5, markedly exceeding the aniline conversion over H-BEA-
1 (16%) with Markovnikov selectivity of 58%. To the best of our
knowledge, the H-BEA-5 zeolite provides the highest yields in phe-
nyl-[2-phenylethyl]amine and phenyl-[1-phenylethyl]amine
reported for this reaction in the literature to date (cf., e.g., Ref.
50], where the maximum conversion was 75%, with selectivity
1
energies.
1
1 4 5
and 29.5% for C –C hydrocar-
paraffins, and 28.2 and 28.5% for C+
bons at comparable decane conversions of 76.3 and 80.6% for
H-BEA-5 and H-BEA-11, respectively.
[
3
.2.2. Alkylation of benzene with benzyl alcohol to diphenyl methane
A threefold increase in the conversion of benzyl alcohol (BzOH)
for the Markovnikov isomer of 35% under the same conditions).
In Ref. [50], analyzing the activity and regioselectivity in beta zeo-
lites, the Brønsted acid sites were assumed to orient the hydroam-
ination toward the anti-Markovnikov adducts, while the Lewis acid
sites oriented it towards Markovnikov regioisomers. While more
than a fourfold increase in aniline conversion can be attributed
to an increase in the concentration of acid sites, the higher regiose-
lectivity for the Markovnikov adduct is not correlated with an
increase in the relative concentration of Lewis sites in H-BEA-5
compared to H-BEA-11. It can be speculated that for the amination
was obtained with H-BEA-5 compared to H-BEA-11 (data for a
reaction time of 2 h are shown in Fig. 7A, accompanied by a sub-
stantial increase in the selectivity for diphenylmethane (DPM),
from 16.6% to 39.6% after 2 h, and half selectivity for dibenzyl ether
(
DBE), but twice the increase in selectivity for dibenzylated ben-
zene (DBB); see Scheme 1 and Fig. 7A). Data for reaction times of
h and 24 h are given in Supplement IV. The increase in conver-
sion of BzOH over H-BEA-5 and in selectivity for DPM and DBB
and the decrease in the selectivity for DBE) can be explained by
4
*
(
reaction occurring in the pores of BEA zeolites, other parameters
an increase in the concentration of Brønsted and Lewis sites (cf.
Table 1). Some of the increase in the yield of the desired DPM
might also be explained by the reaction taking place inside pores
might also contribute to the regioselectivity. Structure-directed
shape selectivity with respect to regioisomers (cf. Ref. [51]) might
play some role inside zeolite pores.
Table 3
Cracking of n-decane over H-BEA-5 compared to H-BEA-11 and H-USY-6 (Si/Al 6, Zeolyst Int. CBV712, Lot 712014001708).
ra (molC10H22 kgcat
ꢂ1 ꢂ1
h
)
TOFb (molC10H22 molOH
ꢂ1 ꢂ1
h
)
ꢂ1 ꢂ1
TOFc (molC10H22 molAl
h )
Sample
T (°C)
Conversion of C10
H
22 (%)
H-BEA-11
480
53.8
54.1
59.2
645
650
748
–
833
–
478
481
554
500
20
5
H-BEA-5
H-USY-6
480
72.5
76.3
78.6
1077
1201
1286
–
840
–
363
404
433
5
5
00
20
480
44.9
43.8
55.5
498
481
676
–
–
–
209
202
284
5
5
00
20
a
b
c
Reaction rate calculated from the first-order kinetics.
TOF per Brønsted OH group.
TOF per Al.

3
0
P. Sazama et al. / Journal of Catalysis 318 (2014) 22–33
A
B
100
100
xBzOH
sDPM
sDBE
sDBB
xaniline Markovnikov
80
60
40
20
0
80
Markovnikov
anti-
Markovnikov
60
40
anti-Markovnikov
20
xaniline
0
H-BEA-11
H-BEA-5
H-BEA-11
H-BEA-5
*H-BEA-5
Fig. 7. Activity of H-BEA-5 in acid-catalyzed reactions compared to H-BEA-11. (A) Conversion of benzyl alcohol (xBzOH) in alkylation of benzene and selectivity to diphenyl
*
methane (DPM) and by-products dibenzyl ether (DBE) and dibenzylated benzene (DBB) at benzene/BzOH 1 after 2 h. ( Excess of benzene at benzene/BzOH 2.5.) (B)
Conversion of aniline (xaniline) in hydroamination of styrene and selectivity for Markovnikov and anti-Markovnikov products.
in the resulting Fe-BEA-5 compared to Fe-BEA-11 exhibited signif-
icantly higher intensities of the O ? Fe(III) ligand to metal charge
ꢂ1
transfer (LMCT) at 36,000–48,000 and 28,000 cm connected with
Diphenylmethane(DPM)
isolated Fe(III)-oxo species with T
bridging or polynuclear Fe(III)–oxo complexes [27,54], respectively
Fig. 8). The possible presence of Fe(II) species [26,27] was not
d h
and O coordination [53] and
(
O
detected in the vis range because of the low intensities of their
d–d transitions [26] and their overlap in the MLCT region. Although
various Fe(III) species could not be analyzed quantitatively because
of lack of knowledge of the extinction coefficients of the individual
LMCT transitions, the spectral intensities indicate a significantly
higher population of counterion Fe species in Fe-BEA-5 compared
to Fe-BEA-11. This corresponds to differences in Fe content (11.7
and 5.3 wt.%, respectively) and in the content of framework Al
atoms in the beta zeolites (Table 1). The increased ion-exchange
capacity of Fe-BEA-5 is also supported by the lower intensity of
catalyst
-H
2
O
Dibenzylether(DBE)
OH
Dibenzylated benzene(DBB)
ꢂ1
the absorption, around 18,000 cm in Fe-BEA-5 compared to Fe-
Scheme 1. Alkylation of benzene with benzyl alcohol.
BEA-11, reflecting the lower concentration of small Fe oxide-like
neutral species not charge-balanced by the negative framework
in Fe-BEA-5.
As Fe-catalysts undergo severe hydrothermal treatment, partic-
H
N
ularly in the SCR-NO
tems, the UV–vis–NIR spectra were monitored also on Fe-BEA
pretreated in 10% H O at 600 °C for 12 h. This resulted in a lower
x
process in active diesel engine exhaust sys-
NH
2
*
N
H
2
catalyst
toluene
+
+
423 K, 24h
Fe-BEA-11
Fe-BEA-5
Fe-BEA-11HT
Fe-BEA-5HT
Phenyl-[1-
phenylethyl]amine
Phenyl-[2-
phenylethyl]amine
Markovnikov
anti-Markovnikov
Scheme 2. Hydroamination of styrene with aniline.
8
*
3.3. Structure analysis and redox-catalyzed reactions over Fe-BEA
Fe counterion species charge-balanced by the framework
ꢂ
AlO
NH
4
in Si-rich zeolites are known to be highly active sites in
[4] and decomposition of N O to molecular N and
[52]. Increased Fe(II) ion-exchange capacity of BEA-5 for
3
-SCR-NO
x
2
2
O
2
1
0000
20000
30000
40000
50000
counterion Fe species similar to those in traditional Si-rich Fe-
-1
*
Wavenumber (cm )
BEA was assumed. The Fe(II) ion exchange occurred under a non-
*
oxidizing atmosphere, but exposure of Fe-BEA in a hydrated state
Fig. 8. Diffuse reflectance UV–vis–NIR spectra of hydrated Fe-BEA-5 and Fe-BEA-11
catalysts and catalysts treated in a stream of 10% water vapor at 600 °C for 12 h (Fe-
BEA-5HT and Fe-BEA-11HT).
to air oxidized a predominant part of Fe(II) to Fe(III) species. UV–
vis–NIR spectra monitored for estimation of the state of Fe species

P. Sazama et al. / Journal of Catalysis 318 (2014) 22–33
31
degree of clustering of counterion Fe species into Fe-oligomers and
formation of inactive Fe oxides in H-BEA-5 compared to Fe-BEA-11,
The effect of a high concentration of Al in the framework and a
corresponding high concentration of counterion Fe sites is also
ꢂ1
reflected in the lower relative band intensities at ꢁ18,000 cm
manifested in the decomposition of N
this is in the presence of NO), depicted for Fe-BEA-5 and Fe-BEA-
11 in Fig. 10. Pseudo-first-order kinetic plots with respect to N
[23] and zero-order plots with respect to N and O [57]
2 2 2
O to N and O (note that
and preservation of a large portion of the counterion Fe species
ꢂ1
manifested by the absorption in the region 28,000–48,000 cm
.
2
O
These findings indicate better stabilization of a high concentration
of counterion Fe species and framework Al in Fe-BEA-5 under the
severe conditions of hydrothermal calcination.
2
2
(ꢂln(1 ꢂ x) = rN2O(W/F)) were used to obtain the reaction rate for
temperatures of 350–425 °C (Fig. 10A). An approximately one
2
and one-half times higher rate of N O decomposition over Fe-
*
3
.3.1. NH
3
-SCR-NO
x
and decomposition of N
2
O over Fe-BEA
BEA-5 than over Fe-BEA-11 was obtained (Fig. 10B). The apparent
ꢂ1
Fig. 9 depicts the temperature dependence of NO conversion to
, pseudo-first-order reaction rates, and Arrhenius plots for the
activation energies of 125 and 132 kJ mol for Fe-BEA-5 and Fe-
ꢂ1
N
2
BEA-11, respectively, are slightly lower than 140 kJ mol reported
NH
lysts pretreated in a stream of 10% H
high concentration of Fe counterion species in Al-rich Fe-BEA-5
caused greatly improved conversion of NO over the low tempera-
ture range from 200 to 300 °C, important for catalyst application in
diesel exhaust gases. The rates for conversion of NO to N over Fe-
BEA-5 and steamed Fe-BEA-5HT catalysts (Fig. 9B) are more than
twice as high as those of Fe-BEA-11 and Fe-BEA-11HT, respectively.
3 x
A similar activity increase in NH -SCR-NO over the Al-rich Fe-BEA
3
-SCR-NO
x
reaction over Fe-BEA-5 and Fe-BEA-11 and the cata-
for N
lites [58]. The E
2
O decomposition in the presence of NO over Fe-ZSM-5 zeo-
value is slightly higher for the more active Fe-
2
O in He at 600 °C for 12 h. The
a
BEA-5, but it is compensated for by the approximately two-
order-higher pre-exponential factor A (lnA = 16 vs. 14). Not so for
NH -SCR-NO , where identical TOFSCR values per Fe were found
3 x
in both Fe-BEA catalysts. The TOFdeN2O per Fe over Fe-BEA-5
reached roughly only 70% of that over Fe-BEA-11 (Table 4). This
implies that the concentration of the Fe sites most active in N
decomposition did not follow in linear proportionality the increase
in the Al concentration of BEA-5. The most active sites for N
decomposition over Fe-MFI, Fe-FER, and Fe-BEA have been identi-
fied as bare Fe(II) ions, whose divalent state is stabilized by an Al–
Si–Si–Al sequence in 6MRs, even in the oxidizing atmosphere pres-
ent during the decomposition reaction [26,27,59–62]. The
x
*
x
2
2
O
is reported in the recent patent assigned to BASF [55]. The apparent
2
O
*
activation energies obtained from the Arrhenius plots (Fig. 9C) for
ꢂ1
all the Fe catalysts investigated here range from 40 to 47 kJ mol
,
in agreement with those reported in the literature for Fe species in
ꢂ1
zeolites (E
a
= 36–51 kJ mol in the temperature range 200–300 °C
[
24,26,56]). However, E
a
over Fe-BEA-5 are compensated for by a
2
increased activity in N O decomposition with Fe-BEA-5 reported
pre-exponential factor A (lnA from ꢂ1 up to 2.6) several orders
of magnitude higher for the most active Fe-BEA-5 catalyst. This
can be seen on similar types of active counterion Fe sites, and
the higher activity of Fe-BEA-5 is caused by their higher concentra-
tion. This assumption is clearly evident from equal TOFSCR values
here can be attributed to the increased concentration of these Fe
ions, but it does not strictly correspond to the increase in the con-
centration of framework aluminum.
3 x 2
For both NH -SCR-NO and N O decomposition, the use of Fe-
BEA-5 containing an increased concentration of counterion Fe spe-
cies is clearly beneficial. Nevertheless, while the sum of concentra-
tions of all types of counter Fe ions and Fe–oxo species active in
(
per Fe) obtained over Fe-BEA-5 and Fe-BEA-11 in the whole
temperature range studied (Table 4).
A
B
C
100
-7
100
80
60
40
20
0
-
8
8
6
4
0
0
0
lnA = 2.6
Ea = 45 kJ.mol
-
1
-9
lnA = 2.3
Ea = 47 kJ.mol
-
1
Fe-BEA-5
lnA = -0.3
Ea = 40 kJ.mol
-
10
-1
Fe-BEA-11
Fe-BEA-5HT
Fe-BEA-11HT
20
0
lnA = -1
Ea = 40 kJ.mol
-1
-11
200
250
300
350
400
450
500
210
250
290
330
0.0016
0.0018
0.0020
0.0022
Temperature (°C)
-1
Temperature (°C)
1/T (K )
Fig. 9. NH
3
-SCR-NO over Fe-BEA-5 and Fe-BEA-11 and catalysts treated in a stream of 10% H
2
O in He at 600 °C for 12 h (Fe-BEA-5HT and Fe-BEA-11HT). Reaction conditions:
ꢂ1
4
90 ppm NO, 520 ppm NH
3
, 2% O
2
in He, GHSV 510,000 h . (A) Conversion of NO
x
as a function of temperature. (B) First-order reaction rate rSCR as a function of temperature.
(
C) Arrhenius plot with the pre-exponential factors lnA and apparent activation energies.
Table 4
Composition of Fe-BEA* and TOF and reaction rate r for NH
3
-SCR-NO
x
2
and decomposition of N O.
Sample
c
Fe
Fe/
Al
NH
10 °C
ra
3
-SCR-NO
2
Decomposition of N O
(wt.%)
2
260 °C
ra
400 °C
425 °C
b
TOFb
ra
c
ra
TOFc
TOF
(h
TOF
(h
ꢂ1 ꢂ1
ꢂ1
ꢂ1 ꢂ1
ꢂ1
ꢂ1 ꢂ1
ꢂ1
ꢂ1 ꢂ1
ꢂ1
(
molNO kg
h
)
)
(molNO kg
h
)
(h
)
(molN2O kg
h
)
)
(molN2O kg
h
)
(h
)
Fe-BEA-5
Fe-BEA-
11.7
5.3
0.73
0.77
31.4
13.2
15.0
13.9
92.1
40.6
43.9
42.8
1.60
1.05
0.76
1.10
3.80
2.43
1.81
2.54
1
1
a
b
c
Reaction rate calculated from the first-order kinetics.
ꢂ1 ꢂ1
TOF (molNO molFe
h ).
ꢂ1 ꢂ1
TOF (molN2O molFe
h ).

3
2
P. Sazama et al. / Journal of Catalysis 318 (2014) 22–33
r
deN2O
-1 -1
A
B
C
(
mol.kg .h )
2
1
0
3
.80
4
3
2
1
0
T
Fe-BEA-5 Fe-BEA-11
1.0
0.8
0.6
0.4
0.2
0.0
425 °C
400 °C
375 °C
350 °C
Fe-BEA-5
Fe-BEA-5
2
.43
Fe-BEA-11
lnA = 24.1
-1
1
.60
.05
Ea = 132 kJ.mol
1
Fe-BEA-11
-
1
0.70
lnA = 22.5
Ea = 126 kJ.mol
0
.44
.24
-1
0
0.18
-2
0.0014
0.00
0.05
0.10
0.15
0.20
0.25
350
375
400
425
0.0015
1/T (K )
0.0016
-1
-1
W/F (kg.h.mol )
Temperature (°C)
ꢂ1
Fig. 10. Decomposition of N
2
O over Fe-BEA-5 and Fe-BEA-11. Reaction conditions: 1000 ppm N
2
O, 500 ppm NO, 3% O
2
2
, and 1% H O in He, GHSV 90,000 h . (A) Pseudo-first-
order kinetic plot ꢂln(1 ꢂ x) = rdeN2O(W/F). (B) First-order reaction rates rdeN2O as a function of temperature. (C) Arrhenius plot with the pre-exponential factors lnA and
apparent activation energies.
NH
3
-SCR-NO
x
is proportional to the increased concentration of
O decomposition,
selectivity towards formation of diphenyl methane at the expense
of diphenyl ether corresponds to the increased concentration of
Brønsted sites.
framework Al, for the most active sites in N
2
i.e., bare Fe(II) ions balanced by AlSiSiAl sequences in 6MRs, this
relationship does not quantitatively follow the concentration of
Al in the framework. This observation implies differences in the
distribution of framework Al atoms in Fe-BEA-5 and Fe-BEA-11
balancing individual Fe counterion species. That is also indicated
by the presence of AlSiAl sequences found in BEA-5 (cf. Si(2Al2Si)
atoms in Fig. 4) in contrast to BEA-11 containing only AlSiSiAl
sequences in 6MRs and single Al atoms, as reported in Ref. [16].
It might be speculated that the presence of AlSiAl sequences
matches the occurrence of AlSiSiAl sequences in 6MRs charge bal-
An advantage of a high concentration of framework Al atoms in
*
Al-rich BEA also lies in an increased concentration of counter Fe
ion species with exceptional redox behavior reflected in the
increased rates of N
ular nitrogen. While the concentration of Fe–oxo and oligomeric
Fe–oxo counter species active in NH -SCR-NO followed the
increased concentration of framework Al atoms, the N O decompo-
2 3 x
O decomposition and NH -SCR-NO to molec-
3
x
2
sition occurring over bare Fe(II) ions is not strictly proportional to
*
the increased Al concentration in Al-rich BEA .
ancing bare Fe(II) ions active in N
take part in charge balancing various counter Fe–oxo up to oligo-
meric Fe–oxo species active in NH -SCR-NO However, the
2
O decomposition, but that they
Beside the advantage of Al-rich beta zeolite with high ion-
exchange capacity for protons and/or metal, metal–oxo ion species,
it brings about the advantage of the less costly demanded synthesis
without employing an organic template and typical highly regular
crystals with high surface area of inner pores and negligible exter-
nal surface area. That presumes enhanced exploiting of the shape
selectivity of the three-dimensional 12MRs pores, particularly in
reactions of bulkier organic molecules.
3
x
.
detailed description of the distribution of Al atoms in the frame-
*
work of Al-rich BEA and its consequence for divalent ion-exchange
capacity will be matters for further study.
4
. Conclusions
Acknowledgments
The high concentration of Al in the framework of the template-
free synthesized beta zeolite (Si/Al 4.6) compared with conven-
tional Si-rich beta zeolite (Si/Al 11.3) provides charge balance for
the corresponding high concentration of protons and/or counter
divalent or trivalent oxo metal ion species. In addition to the high
content of framework Al, the synthesized crystals of Al-rich beta
This study was supported by the Czech Science Foundation
(
Projects # P106/11/0624 and # GA14-10251S), by the Technology
Agency of the Czech Republic (Project # TA01021377) and by RVO
61388955.
#
zeolite (0.3–0.5 lm) exhibit highly regular pore structure with
negligible external surface area, implying high shape selectivity
effects of inner pores compared to very small crystallites (nm
range) of Si-rich beta zeolites that commonly exhibit large external
surfaces.
Appendix A. Supplementary material
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.jcat.2014.06.024.
Although a much higher concentration of protonic sites is pres-
ꢂ
ent in the Al-rich beta zeolite associated with AlO
4
, these Brønsted
References
sites represent strongly acidic sites of strength comparable to that
ꢂ
[
[
[
1] A. Corma, Chem. Rev. 95 (1995) 559–614.
of the protons of Si-rich beta zeolites. Nevertheless, some AlO
4
2] D.E. De Vos, M. Dams, B.F. Sels, P.A. Jacobs, Chem. Rev. 102 (2002) 3615–3640.
3] F. Kapteijn, J. Rodriguez-Mirasol, J.A. Moulijn, Appl. Catal., B: Environ. 9 (1996)
25–64.
exhibit an increased tendency toward perturbation of the T–O
framework bonds with formation of Al-Lewis electron-acceptor
sites, but preserving their location in the framework. These find-
ings are manifested in the increased conversion in n-decane crack-
ing, alkylation of benzene with benzyl alcohol to diphenylmethane,
and hydroamination of styrene with aniline to phenyl-[2-phenyl-
ethyl]amine and phenyl-[1-phenylethyl]amine, correspondingly
to the increased concentration of Brønsted acid sites. The differ-
[
[
4] S. Brandenberger, O. Krocher, A. Tissler, R. Althoff, Catal. Rev. Sci. Eng. 50
(
2008) 492–531.
5] Y. Traa, B. Burger, J. Weitkamp, Microporous Mesoporous Mater. 30 (1999) 3–
41.
[6] Ch. Baerlocher, L.B. McCusker, D.H. Olson, Atlas of Zeolite Framework Types,
th revised edition, 2007.
[
[
6
7] B. Xie, J. Song, L. Ren, Y. Ji, J. Li, F.S. Xiao, Chem. Mater. 20 (2008) 4533–4535.
8] G. Majano, L. Delmotte, V. Valtchev, S. Mintova, Chem. Mater. 21 (2009) 4184–
4191.
*
ences in Al distribution between conventional H-BEA (containing
*
[9] Y. Kamimura, S. Tanahashi, K. Itabashi, A. Sugawara, T. Wakihara, A.
Shimojima, T. Okubo, J. Phys. Chem. C 115 (2011) 744–750.
10] Y. Kubota, K. Itabashi, S. Inagaki, Y. Nishita, R. Komatsu, Y. Tsuboi, S. Shinoda, T.
Okubo, Chem. Mater. 26 (2014) 1250–1259.
only AlSiSiAl pairs and single Al atoms) and Al-rich H-BEA (pos-
sessing also AlSiAl sequences) are not significantly reflected in
the acid strength and activity per Brønsted site. The increased
[

P. Sazama et al. / Journal of Catalysis 318 (2014) 22–33
33
[
11] B. Yilmaz, U. Muller, M. Feyen, S. Maurer, H. Zhang, X. Meng, F.S. Xiao, X. Bao,
W. Zhang, H. Imai, T. Yokoi, T. Tatsumi, H. Gies, T. De Baerdemaeker, D. De Vos,
Catal. Sci. Technol. 3 (2013) 2580–2586.
12] T. De Baerdemaeker, B. Yilmaz, U. Muller, M. Feyen, F.S. Xiao, W. Zhang, T.
Tatsumi, H. Gies, X. Bao, D. De Vos, J. Catal. 308 (2013) 73–81.
13] P. Sazama, Z. Sobalik, J. Dedecek, I. Jakubec, V. Parvulescu, Z. Bastl, J. Rathousky,
H. Jirglova, Angew. Chem., Int. Ed. 52 (2013) 2038–2041.
14] H. Zhang, B. Xie, X. Meng, U. Muller, B. Yilmaz, M. Feyen, S. Maurer, H. Gies, T.
Tatsumi, X. Bao, W. Zhang, D. De Vos, F.S. Xiao, Microporous Mesoporous
Mater. 180 (2013) 123–129.
[36] J.A. Van Bokhoven, D.C. Koningsberger, P. Kunkeler, H. Van Bekkum, A.P.M.
Kentgens, J. Am. Chem. Soc. 122 (2000) 12842–12847.
[37] D.M. Roberge, H. Hausmann, W.F. Holderich, Phys. Chem. Chem. Phys. 4 (2002)
3128–3135.
[38] L. Capek, J. Dedecek, B. Wichterlova, J. Catal. 227 (2004) 352–366.
[39] J. Dedecek, S. Sklenak, C. Li, F. Gao, J. Brus, Q. Zhu, T. Tatsumi, J. Phys. Chem. C
113 (2009) 14454–14466.
[40] J. Perez-Pariente, J. Sanz, V. Fornes, A. Corma, J. Catal. 124 (1990) 217–223.
[41] P. Sazama, J. Dedecek, V. Gabova, B. Wichterlova, G. Spoto, S. Bordiga, J. Catal.
254 (2008) 180–189.
[
[
[
[
15] E.J. Creyghton, S.D. Ganeshie, R.S. Downing, H. Van Bekkum, J. Mol. Catal. A:
Chem. 115 (1997) 457–472.
16] L. Capek, J. Dedecek, P. Sazama, B. Wichterlova, J. Catal. 272 (2010) 44–54.
17] C.A. Fyfe, Y. Feng, H. Grondey, G.T. Kokotailo, H. Gies, Chem. Rev. 91 (1991)
[42] N. Wang, M. Zhang, Y. Yu, Microporous Mesoporous Mater. 169 (2013) 47–53.
[43] C. Paze, S. Bordiga, C. Lamberti, M. Salvalaggio, A. Zecchina, G. Bellussi, J. Phys.
Chem. B 101 (1997) 4740–4751.
[44] P. Sazama, B. Wichterlova, J. Dedecek, Z. Tvaruzkova, Z. Musilova, L. Palumbo,
S. Sklenak, O. Gonsiorova, Microporous Mesoporous Mater. 143 (2011) 87–96.
[45] I. Kiricsi, C. Flego, G. Pazzuconi, W.O. Parker, R. Millini, C. Perego, G. Bellussi, J.
Phys. Chem. 98 (1994) 4627–4634.
[
[
1
525–1543.
[
18] B. Wichterlova, Z. Tvaruzkova, Z. Sobalik, P. Sarv, Microporous Mesoporous
Mater. 24 (1998) 223–233.
[
[
[
[
[
19] K. Klier, J. Opt. Soc. Am. 62 (1972) 882.
[46] C.T.W. Chu, C.D. Chang, J. Phys. Chem. 89 (1985) 1569–1571.
[47] S. Kotrel, M.P. Rosynek, J.H. Lunsford, J. Catal. 182 (1999) 278–281.
[48] A.F.H. Wielers, M. Vaarkamp, M.F.M. Post, J. Catal. 127 (1991) 51–66.
[49] A. Corma, P.J. Miguel, A.V. Orchilles, Appl. Catal., A: Gen. 138 (1996) 57–73.
[50] M. Ciobanu, A. Tirsoaga, P. Amoros, D. Beltran, S.M. Coman, V.I. Parvulescu,
Appl. Catal., A: Gen. 474 (2014) 230–235.
[51] B. Wichterlova, J. Cejka, J. Catal. 146 (1994) 523–529.
[52] G.D. Pirngruber, M. Luechinger, P.K. Roy, A. Cecchetto, P. Smirniotis, J. Catal.
224 (2004) 429–440.
[53] S. Bordiga, R. Buzzoni, F. Geobaldo, C. Lamberti, E. Giamello, A. Zecchina, G.
Leofanti, G. Petrini, G. Tozzola, G. Vlaic, J. Catal. 158 (1996) 486–501.
[54] G.D. Pirngruber, P.K. Roy, R. Prins, Phys. Chem. Chem. Phys. 8 (2006) 3939–
3950.
20] M. Brandle, J. Sauer, J. Am. Chem. Soc. 120 (1998) 1556–1570.
21] U. Eichler, M. Brandle, J. Sauer, J. Phys. Chem. B 101 (1997) 10035–10050.
22] J. Sauer, J. Mol. Catal. 54 (1989) 312–323.
23] F. Kapteijn, G. Marban, J. Rodriguez-Mirasol, J.A. Moulijn, J. Catal. 167 (1997)
2
56–265.
24] S. Brandenberger, O. Krocher, A. Tissler, R. Althoff, Appl. Catal., B: Environ. 95
2010) 348–357.
[
(
[
[
25] L. Riekerti, J.Q. Zhou, J. Catal. 137 (1992) 437–452.
26] P. Sazama, B. Wichterlova, E. Tabor, P. Stastny, N.K. Sathu, Z. Sobalik, J.
Dedecek, S. Sklenak, P. Klein, A. Vondrova, J. Catal. 312 (2014) 123–138.
27] P. Sazama, N.K. Sathu, E. Tabor, B. Wichterlova, S. Sklenak, Z. Sobalik, J. Catal.
[
2
99 (2013) 188–203.
[
[
28] M.M.J. Treacy, J.M. Newsam, Nature 332 (1988) 249–251.
29] E. Bourgeat-Lami, P. Massiani, F. Di Renzo, P. Espiau, F. Fajula, T. Des Courieres,
Appl. Catal. 72 (1991) 139–152.
30] L.W. Beck, J.F. Haw, J. Phys. Chem. 99 (1995) 1076–1079.
31] R.B. Borade, A. Clearfield, Microporous Mater. 5 (1996) 289–297.
32] L.C. De Menorval, W. Buckermann, F. Figueras, F. Fajula, J. Phys. Chem. 100
[55] S. Maurer, M. Wycisk, J. Petry, S. Deuerlein, W. Zhang, Ch. Shi, H. Takashi,
WO2013118063.
[56] M. Iwasaki, K. Yamazaki, H. Shinjoh, Appl. Catal., A: Gen. 366 (2009) 84–92.
[57] K. Sun, H. Xia, E. Hensen, R. Van Santen, C. Li, J. Catal. 238 (2006) 186–195.
[58] H. Xia, K. Sun, Z. Liu, Z. Feng, P. Ying, C. Li, J. Catal. 270 (2010) 103–109.
[59] J. Dedecek, L. Capek, P. Sazama, Z. Sobalik, B. Wichterlova, Appl. Catal., A: Gen.
391 (2011) 244–253.
[
[
[
(
1996) 465–467.
[
[
[
33] P.J. Kunkeler, B.J. Zuurdeeg, J.C. Van Der Waal, J.A. Van Bokhoven, D.C.
Koningsberger, H. Van Bekkum, J. Catal. 180 (1998) 234–244.
34] G.H. Kuehl, H.K.C. Timken, Microporous Mesoporous Mater. 35–36 (2000)
[60] Z. Sobalik, J. Novakova, J. Dedecek, N.K. Sathu, E. Tabor, P. Sazama, P. Stastny, B.
Wichterlova, Microporous Mesoporous Mater. 146 (2011) 172–183.
[61] Z. Sobalik, P. Sazama, J. Dedecek, B. Wichterlova, Appl. Catal., A: Gen. 474
(2014) 178–185.
[62] E. Tabor, K. Zaveta, N.K. Sathu, A. Vondrova, P. Sazama, Z. Sobalik, Catal. Today
175 (2011) 238–244.
5
21–532.
35] J. Penzien, A. Abraham, J.A. Van Bokhoven, A. Jentys, T.E. Muller, C. Sievers, J.A.
Lercher, J. Phys. Chem. B 108 (2004) 4116–4126.