New Nb and Ta-FAU zeolites - Direct synthesis, characterisation and surface properties

Article abstract of DOI:10.1016/j.cattod.2010.06.018

Niobium and tantalum containing Y zeolites were prepared in a one-pot synthesis and via solid-state ion exchange. The first method led to the incorporation of Nb and Ta into the zeolite framework as documented by XRD, UV-vis, XPS and test reactions. The efficiency of the metal introduction was higher for tantalum than for niobium. The presence of Lewis acid sites was proved by pyridine adsorption and cracking of cumene, whereas basic hydroxyls on the surface were evaluated on the basis of 3-methyl-2-cyclopentenone formation in the cyclisation of acetonyl acetone.

Full text of DOI:10.1016/j.cattod.2010.06.018

Catalysis Today 158 (2010) 170–177
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
New Nb and Ta–FAU zeolites—Direct synthesis, characterisation
and surface properties
M. Trejdaa,∗, A. Wojtaszeka, A. Flocha, R. Wojcieszakb, E.M. Gaigneauxb, M. Zioleka
a Adam Mickiewicz University, Faculty of Chemistry, Grunwaldzka 6, PL-60-780 Poznan, Poland
b Université catholique de Louvain, Unité de Catalyse et Chimie des Matériaux Divisés, Croix du Sud 2/17, 1348 Louvain-la-Neuve, Belgium
a r t i c l e i n f o
a b s t r a c t
Article history:
Available online 3 August 2010
Niobium and tantalum containing Y zeolites were prepared in a one-pot synthesis and via solid-state
ion exchange. The first method led to the incorporation of Nb and Ta into the zeolite framework as
documented by XRD, UV–vis, XPS and test reactions. The efficiency of the metal introduction was higher
for tantalum than for niobium. The presence of Lewis acid sites was proved by pyridine adsorption and
cracking of cumene, whereas basic hydroxyls on the surface were evaluated on the basis of 3-methyl-2-
cyclopentenone formation in the cyclisation of acetonyl acetone.
Keywords:
NbY
TaY one-pot synthesis
Pyridine adsorption
Acetonylacetone cyclisation
Cumene cracking
© 2010 Elsevier B.V. All rights reserved.
1. Introduction
metals from tetrahedral sites towards the extra-framework posi-
tions. This aspect was examined in details for the iron isomorphous
Isomorphous substitution of zeolite lattice atoms, i.e. their
replacements most often by different metal ions, is of great impor-
tance. Thus there is a lot of works on this subject in literature
[1]. It is known that the isomorphous substitution of aluminium
or silicon in zeolite structure can generate unique properties of
the materials obtained. Generally, such modification can be per-
formed by two synthetic routes: (i) incorporation of the desired
metal through crystallisation together with silicon and aluminium
species under hydrothermal conditions or (ii) post-synthesis mod-
ification of zeolites via different treatments [2]. The choice of the
modification route depends on the kind of zeolite structure, ther-
mal/hydrothermal stability of zeolite, available metal source, and
many other parameters.
The classical example of isomorphous substitution is the prepa-
ration of TS-1 material of the MFI structure in which silicon is
substituted with titanium [3]. This microporous molecular sieve
was found to be very interesting in the liquid phase oxidation
of organic compounds with hydrogen peroxide as oxidant. How-
ever, there are a lot of works concerning isomorphous substitution
of other elements into the zeolite lattice, for example B–MFI [4],
Ga–BEA [5], Ge–ITQ-21 [6], Fe–MFI [7], Sn–BEA [8], V–MFI [9],
and Zn–CIT-6 [10]. It should be mentioned that the incorporation
of the so-called heteroatom into zeolite framework via crystalli-
sation under hydrothermal conditions followed by a thermal or
hydrothermal treatment could also lead to the displacement of
substitution into the MFI structure [11].
There are even fewer data concerning metal incorporation into
large pore structures such as for example that of faujasite (FAU). A
possibility of incorporation of boron, gallium, iron and titanium into
the FAU skeleton was described in [2,12,13]. Moreover, the faujasite
structure was modified by non-metal element, i.e. by phosphorous,
to obtain SAPO-37 sample [14]. This material belongs to the alu-
minophosphate family and shows interesting properties induced
by the presence of phosphorous +5 cations in the skeleton. Tak-
ing into account the unique properties of SAPO-37 we focused on
the preparation of FAU type zeolites containing transition met-
als of group five (Nb, Ta—present at +5 oxidation state) instead of
phosphorous. We expected to obtain isolated Lewis acid centres by
incorporation of Nb and Ta cations at the oxidation state +5. The
substitution of these elements into the structure of zeolites ZSM-5
and beta (exhibiting high and medium Si/Al ratios) has been already
studied (e.g. [15–17]) but as far as we know there are no reports on
their incorporation into the zeolite framework of the large pore and
low Si/Al ratio (e.g. FAU structure). Here we have explored this pos-
sibility by performing direct synthesis, without organic structure
directing agent, of Nb and Ta containing Y zeolites.
2. Experimental
2.1. Preparation of zeolites
Preparation of Nb or Ta containing materials of the FAU struc-
ture was based on the modified two-steps procedure reported
originally by Ginter et al. [18]. Firstly, the seed gel was prepared
∗
Corresponding author. Tel.: +48 61 8291305; fax: +48 61 8291505.
E-mail address: tmaciej@amu.edu.pl (M. Trejda).
0920-5861/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.cattod.2010.06.018

M. Trejda et al. / Catalysis Today 158 (2010) 170–177
171
as follows. Sodium hydroxide (Chempur) and sodium aluminate
(Riede-de Haën) were dissolved in H2O. Afterwards, the sodium sil-
icate solution (Aldrich) was added and the mixture was stirred for
about 10 min. The final solution composed of: 10.67 Na2O:Al2O3:10
SiO2:180 H2O was let to age at room temperature for 1 day. In
the second step, sodium hydroxide and sodium aluminate were
dissolved in H2O then sodium silicate was added upon vigorous
stirring. The composition of the obtained feedstock gel was: 4.3
Na2O:Al2O3:10 SiO2:180 H2O. Afterwards, a part of the seed gel
was added to the feedstock gel to get the final composition of: 4.62
Na2O:Al2O3:10 SiO2:180 H2O. The metal source (niobium or tan-
talum pentaethoxide—Aldrich) was added to both, the seed and
feedstock gels. The assumed Si/T ratio (T = Nb or Ta) was 64 or 32.
Then, after vigorous stirring, the gel was put to a polypropylene
bottle and heated in an oven for 5 h at 373 K. The final product was
washed with distilled water and dried at 383 K for 12 h.
For comparison, the modified zeolites were also prepared by
post-synthesis method, i.e. solid-state ion exchange, using com-
mercial sodium Y zeolite (Katalistiks) and group five metal oxides.
Hydrogen form of the zeolite and Nb2O5 (Alfa Aesar) or Ta2O5
(Aldrich) were mechanically mixed (the ratio of protons to Nb or Ta
was assumed as 5) and heated at 973 K in oven for 8 h (temperature
ramp 2 K min−1).
Fig. 1. XRD patterns of Nb and Ta containing zeolites prepared in one-pot synthesis.
2.4. Acetonylacetone cyclisation
The materials were tested in the acetonylacetone cyclisation. A
tubular, down-flow reactor was used in experiments performed at
atmospheric pressure, using nitrogen as a carrier gas. The catalyst
bed (0.05 g) was first activated for 2 h at 623 K under nitrogen flow
(40 cm3 min−1). Afterwards, a 0.5 cm3 of acetonylacetone (Fluka,
GC grade) was passed continuously through the catalyst at 623 K
for 0.5 h. The substrate was delivered with a pump system and
vaporised before it was passed through the catalyst bed in a flow
of nitrogen carrier gas (40 cm3 min−1). The reaction products were
collected downstream of the reactor in a cold trap (mixture of 2-
propanol and liquid nitrogen) and analysed by gas chromatograph
(equipped with a TCD detector and with a silicone SE-30 3 m col-
umn heated at 353 K).
2.2. Catalyst characterisation
X-ray diffraction patterns (XRD) were recorded at room temper-
ature on a Bruker AXS D8 Advance apparatus using Cu K␣ radiation
(ꢀ = 0.154 nm), with a step of 0.05◦.
To establish the Si/Nb or Si/Ta ratio X-ray fluorescence (XRF) was
applied using MiniPal-Philips. The measurements were done using
calibration curves prepared from mixtures of silica and Nb2O5 or
Ta2O5 (Si/Nb or Si/Ta from 5 to 300).
Ultraviolet–visible spectroscopy (UV–vis) was applied. The
spectra were recorded using a Varian-Cary 300 Scan UV-Visible
Spectrophotometer. Catalyst powders, after dehydration at 673 K
for 2 h, were placed into the cell equipped with a quartz window.
The spectra were recorded in the range from 800 to 190 nm. Spec-
tralon was used as a reference material.
SEM images were recorded using Scanning Electron Microscopy
(Philips SEM 515) operating at 15 kV with coating (gold) of the
samples.
Aluminium-27 magic-angle spin solid-state nuclear magnetic
resonance (27Al MAS NMR) was applied. The spectra were recorded
using Bruker Avance DPX300 spectrometer.
X-ray photoelectron spectroscopy (XPS) were performed
with a SSI-X-probe (SSX-100/206) photoelectron spectrometer
equipped with a monochromatic microfocused Al K␣ X-ray source
(1486.6 eV) from Surface Science Instruments.
2.5. Cumene cracking
The cumene cracking was performed, using a microcatalytic
pulse reactor inserted between the sample inlet and the column
of a chromatograph. The catalyst bed (0.05 g with a size fraction
of 0.5 mm < Ø < 1 mm) was first activated at 623 K for 2 h under
helium flow (40 cm3 min−1). The cumene (Aldrich) conversion was
studied at 623 K using 5 ␮l pulses of cumene under helium flow
(40 cm3 min−1). The reactant and reaction products: propene and
␣-methyl styrene were analysed using gas chromatograph con-
nected on-line with the microreactor. The reaction mixture was
separated in helium flow (40 cm3 min−1) on a 3 m column filled
with silicone SE-30 and detected by TCD. The column was heated
as follows: 303 K for 2 min then 20 K min−1 to 338 K and kept for
4 min, then 10 K min−1 to 373 K and kept for 18 min.
2.3. Pyridine adsorption
Infrared spectra were recorded with a Bruker Vector 22 FTIR
spectrometer using an in situ cell. Samples were pressed below
0.2 tons cm−2 into thin wafers of ca. 10 mg cm−2 and placed inside
the cell. Catalysts were evacuated at 623 K during 3 h and pyridine
(PY) was then admitted at 423 K. After saturation with PY the sam-
ples were degassed at 423, 473, 523 and 573 K in vacuum for 30 min
at each temperature. Spectra were recorded at room temperature
in the range from 4000 to 400 cm−1. The spectrum without any
sample (“background spectrum”) was subtracted from all recorded
spectra. The IR spectra of the activated samples (after evacuation
at 623 K) were subtracted from those recorded after the adsorption
of PY followed by various treatments. The reported spectra are the
results of this subtraction.
3. Results and discussion
3.1. Synthesis
Niobium and tantalum containing materials prepared via one-
pot synthesis from the final solution of: 4.62 Na2O:Al2O3:10
SiO2:180 H2O:0.15625 (or 0.078125) Nb2O5 or Ta2O5 showed the
faujasite structure as confirmed by XRD patterns shown in Fig. 1.
One can observe the same reflections with identical relative inten-
sities for NaY zeolite and all materials modified by Nb and Ta, i.e.
NbY-64(32) and TaY-64(32) zeolites. The assumed silicon to nio-
bium or tantalum ratios (Si/T where T = Nb or Ta) were 32 or 64
thus the last number in the material symbol stands for Si/T. The

172
M. Trejda et al. / Catalysis Today 158 (2010) 170–177
Table 1
Unit cell parameters (XRD), metal molar concentration (XRF, XPS) and XPS data of prepared samples.
Catalyst
Unit cell parameter
Molar ratios
Metal BE in the sample, eV
Metal BE in bulk oxide, eV
Si/Nb(Ta)—XRF
Si/Nb(Ta)—XPS
Si/Al—XPS
NaY
24.72
24.76
24.77
24.76
24.73
–
>300
72
78
45
–
730
80
15
7
–
–
–
NbY-64
NbY-32
TaY-64
TaY-32
2.2
2.3
2.3
2.3
208.1
207.9
26.7
26.3
207.1 (Nb2O5)
207.1 (Nb2O5)
26.1 (Ta2O5)
26.1 (Ta2O5)
unit cell parameters (a0) of prepared zeolites, estimated from XRD
patterns, are shown in Table 1. In spite of low metal concentration
the materials containing niobium or tantalum show higher value
of a0 parameter than NaY sample. However, the difference is small,
therefore it is not strong evidence for the incorporation of these
metals into zeolite structure. When a metal is incorporated into the
zeolite structure during the one-pot synthesis three possibilities
can be considered: (i) the metal is not able to occupy any positions
in the zeolite structure, (ii) the metal is located in the zeolite skele-
ton and tetrahedrally coordinated with oxygens (like silicon and
aluminium) and (iii) the metal occupies extra framework positions
as cations or metal oxides.
Crystalline niobium(V) oxide and tantalum(V) oxide show main
XRD peaks at 2ꢁ ca. 22.6, 28.3 and 36.6◦. For the zeolites with both
Si/T ratios used in this work no XRD reflections from crystalline
metal oxide phases were observed in XRD patterns. The real amount
of tantalum in the final material is ca. 5 and 8 wt% (calculated for
Ta(V) oxide) for TaY-64 and TaY-32, respectively. As XRD starts to
be sensitive from amount of crystalline phases of ca. 5 wt% one can
expect that if all or significant part of tantalum introduced into the
zeolite existed in the form of extra framework oxide, the XRD pat-
tern should have contained peaks from the metal oxide phase. To
examine more deeply if tantalum is present in the form of oxide
on the material surface the XRD patterns of Ta2O5, NaY and tan-
talum containing zeolites prepared by one-pot synthesis as well
as by post-synthesis method are shown in Fig. 2. The peaks cor-
responding to Ta2O5 are indicated in the figure. One can observe
that the material prepared by post-synthesis technique exhibits all
the characteristic peaks related to tantalum(V) oxide phases. For
TaY-32 only a very broad peak can be seen at 2ꢁ ca. 28.3◦, however
no other peaks which could origin from metal oxide phases are
observed. Therefore, the difference of tantalum state in the final
material depending on the preparation technique is clearly seen.
However, basing on XRD patterns, the presence of some tantalum
oxide in TaY-32 sample should not be excluded.
The lack of peaks related to metal oxide phases in XRD pat-
terns of NbY and TaY zeolites prepared in one-pot synthesis does
not exclude the possibility of the presence of such phase in the
material, because small metal clusters and/or small quantities of
oxides could not be detected by the XRD technique. To examine
in detail the possibility of the presence of extra framework metal
oxides in the samples prepared, the UV–vis spectra of zeolites were
recorded. Fig. 3 shows the spectra of tantalum containing Y zeolites
prepared by both, one-pot and post-synthesis methods as well as
of bulk tantalum oxide (as a reference sample). The spectra of tan-
talum containing zeolites prepared in the one-pot hydrothermal
synthesis are characterised by a band at 221 nm, which is typical of
tetrahedrally coordinated tantalum species. The zeolite containing
tantalum incorporated via post-synthesis method shows a differ-
ent spectrum with a broad band and an adsorption maximum at ca.
270 nm. This band is typical of octahedrally coordinated tantalum,
e.g. in tantalum(V) oxide [19]. It is seen as the main band for bulk
tantalum(V) oxide (Fig. 3). The UV–vis spectra of niobium contain-
ing Y zeolites prepared by synthesis and post-synthesis methods
are shown in Fig. 4. Moreover, the spectrum of niobium(V) oxide
is also included in the figure. The octahedral coordinated niobium
species gives a characteristic adsorption band at ca. 330 nm. The
band at similar position is observed for both, niobium(V) oxide
and Y zeolite modified by post-synthesis technique (Nb2O5/Y). In
the case of NbY-32 the band corresponded to octahedral coordi-
nated niobium species is not observed. Therefore, one can exclude
the presence of niobium oxide in this sample and confirm the
tetrahedral coordination of niobium species. Thus UV–vis spectra
confirmed the XRD results.
Fig. 2. XRD patterns of Ta containing zeolites prepared in one-pot synthesis and
post-synthesis methods.
Fig. 3. UV–vis spectra of tantalum containing Y zeolites and Ta2O5.

M. Trejda et al. / Catalysis Today 158 (2010) 170–177
173
Fig. 4. UV–vis spectra of niobium containing Y zeolites and Nb2O5.
Fig. 5. MAS 27Al NMR spectra.
The incorporation of niobium or tantalum into the faujasite
structure could disturb the location of at least part of aluminium
present in the synthesis mixture in the framework positions or
even displace aluminium located in the framework positions to
extra-framework ones. To check the location of aluminium in the
zeolites prepared in one-pot synthesis 27Al MAS NMR spectroscopy
was applied and the spectra are shown in Fig. 5. The zeolites pre-
pared in hydrothermal synthesis show a characteristic peak at
ca. 54 ppm, which is attributed to tetrahedrally coordinated alu-
minium. Moreover, the peak typical of octahedrally coordinated
aluminium (at ca. 0 ppm) is not detected for all samples, which
proved that all aluminium species were incorporated into the zeo-
lite lattice. The location of aluminium in the framework positions
confirms the well-developed crystallinity of the zeolite structure.
Moreover, SEM images of tantalum and niobium-containing sam-
ples (Fig. 6) show that the particles of zeolites are similar in size and
shape. One cannot observe the presence of some separated phases
which could inform about the formation of metal oxide in the extra
framework positions for samples with Si/Nb(Ta) ratio = 64. How-
ever, the presence of such phase should not be excluded for lower
Si/Nb(Ta) ratio, especially in the case of tantalum containing sam-
ple. The average particle size of zeolites estimated from SEM images
is between 0.5 and 1 ␮m. The same particle sizes for all materials
were also inferred from XRD patterns, according to Scherrer’s equa-
tion. It points out that both the nature of the heteroatom (Nb or Ta)
and the silicon to metal (Nb or Ta) ratio did not impact on the crystal
morphology.
The efficiency of Nb or Ta incorporation into the faujasite
structure in the one-pot synthesis was examined by XRF using cali-
bration curves. The real Si/T ratios in the final zeolites are shown in
Table 1. It is well marked that the assumed and real Si/T ratios are
a little bit higher for TaY-64(78) and for TaY-32(45). This demon-
Fig. 6. SEM images of: (A) NbY-64; (B) NbY-32; (C) TaY-64; (D) TaY-32.

174
M. Trejda et al. / Catalysis Today 158 (2010) 170–177
incorporation obtained from XRF (which are related to the bulk)
are different from those estimated by XPS (Table 1). Whereas for
niobium containing samples XRF and XPS data are quite similar,
for tantalum zeolites the amount of Ta estimated by XPS is higher.
It suggests the higher concentration of tantalum species on the
materials surface. However, as evidenced by UV–vis spectroscopy
described above tantalum does not show the octahedral coordina-
tion, therefore the presence of tantalum oxide could be suggested
only basing on XRD and SEM results.
3.2. Metal state and location
Information concerning the metal location in zeolite frame-
work or extra-framework positions can be deduced from its
coordination. Tetrahedrally coordinated species should be found
for the framework located metals, whereas octahedrally coordi-
nated ones should be related to the extra-framework species.
As shown in the 27Al MAS NMR spectra described in the pre-
vious section, all aluminium species are present in tetrahedral
sites. Thus they are located in the skeleton of zeolites. More-
over, the Si/Al ratio estimated from XPS results (Table 1) is
almost the same for all the samples. The sodium amount was
also estimated in the final material and expressed as Al/Na ratio.
The obtained value was ca. 1, however the difference of sodium
content in all samples was small (in the range of experimen-
tal error) and therefore is out of discussion. For niobium and
tantalum incorporated via the one-pot synthesis the UV–vis spec-
tra pointed out the same metal coordination as for aluminium,
i.e. tetrahedral. Moreover, no octahedral niobium or tantalum
species were found, contrary to the samples prepared by the
post-synthesis method. Therefore, the tetrahedral coordination of
Nb and Ta species in zeolites obtained in one-pot hydrothermal
synthesis suggests the location of both metals in the zeolite frame-
work.
Fig. 7. Spectra of pyridine adsorbed at 423 K followed by evacuation at 473 K for
0.5 h: (a) TaY-32; (b) TaY-64; (c) NbY-32; (d) NbY-64.
strates that tantalum is rather easily incorporated into the zeolite. It
is not the case for niobium containing samples. For NbY-32 sample
the efficiency of niobium incorporation is ca. 50% and even much
lower for NbY-64. However, much more difficult incorporation of
niobium into the zeolites does not exclude the possible location
of Nb in the zeolite framework. The results of efficiency of metal
For further investigation of the metal state and location in the
niobium and tantalum-containing Y zeolites XPS spectroscopy was
applied. The results obtained for these samples as well as for bulk
niobium(V) and tantalum(V) oxides are collected in Table 1. Only
Fig. 8. Spectra of: (a) pyridine adsorbed at 423 K; and after evacuation at: (b) 423 K; (c) 473 K; (d) 523 K; (e) 573 K.

M. Trejda et al. / Catalysis Today 158 (2010) 170–177
175
Fig. 9. Spectra of: (a) pyridine adsorbed at 423 K; and after evacuation at: (b) 423 K; (c) 473 K; (d) 523 K; (e) 573 K.
one metal oxidation state was found for both niobium and tantalum
and it was estimated as +5. The incorporation of metals at the oxi-
dation state +5 into the zeolite framework changes the local lattice
charge by the formation of positively charged tetrahedral species.
This could imply the formation of Lewis acids sites in niobium and
tantalum-containing zeolites. Indeed, the presence of Lewis centres
was concluded on the basis of pyridine adsorption followed by FTIR
measurements (described below). It should be mentioned that the
origin of such centres cannot be assigned to the aluminium species
since they are present in framework positions.
the decrease of metal content estimated on the materials sur-
face.
3.3. Pyridine adsorption
Aiming the identification of the type of acidic centres present
on the zeolite surfaces, infrared spectroscopy combined with the
adsorption of pyridine was applied. After the interaction of pyridine
with Lewis acid sites (LAS) the characteristic bands at ∼1450 and
∼1610 cm−1 appear [20]. The intensity of the first band is related
to the number of LAS, whereas the position of the second band
characterises the strength of LAS [20]. Pyridine also interacts with
Brønsted acid sites (BAS), which is manifested by the infrared band
at ∼1550 cm−1 and two other bands in the 1620–1640 cm−1 range.
Moreover, pyridine can also interact with hydroxyls on the zeo-
lite surface forming weak hydrogen bounds and giving adsorption
bands at 1596 and 1445 cm−1. The spectra after pyridine adsorp-
tion at 423 K followed by evacuation at 473 K for 0.5 h are shown in
Fig. 7. For all zeolites prepared in the one-pot synthesis, the bands
at 1598 cm−1 and 1444 cm−1 dominate. The first band at 1598 cm−1
is assigned to the pyridine adsorbed via weak hydrogen bound on
hydroxyl groups which do not exhibit strong acidity or are not
acidic (e.g. silanol groups and/or Nb–OH, Ta–OH species). The sec-
ond band at 1444 cm−1 covers both species, i.e. pyridine adsorbed
on LAS and hydrogen-bounded pyridine. The overlapping of both
bands does not allow to calculate the number of LAS. However, the
existence of LAS centres can be validated by the weak band typi-
cal of pyridine adsorbed on LAS at ca. 1610 cm−1. Moreover, for all
samples the band at 1490 cm−1 is observed. This band is assigned to
pyridine adsorbed on both Lewis and Brønsted acid sites. When the
BAS are not present or exist at a low concentration, the intensive
band at 1490 cm−1 can testify of the presence of LAS. This is the case
for all materials prepared. Only the spectrum of NbY-64 (Fig. 7d)
shows a distinct band at 1550 cm−1, which is typical of pyridine
adsorbed on BAS centres. Such centres were also observed on NaY
zeolite prepared according to the procedure applied in this work.
This band is hardly seen for NbY-32 and TaY-64 zeolites (Fig. 7b
The information concerning the metal location in the zeolite can
also be understood from its binding energy. The data in Table 1
show the XPS binding energies obtained for materials prepared
in hydrothermal synthesis and for the bulk niobium(V) and tan-
talum(V) oxides. It can be found that the energies recorded for
niobium and tantalum-containing zeolites were higher than those
characteristic of bulk metal oxides. The increase in the binding
energy can be attributed to different surroundings of niobium or
tantalum species in zeolites than in bulk niobium(V) or tanta-
lum(V) oxides. Different neighbourhoods of metal species can be
obtained when Si–O–Nb or Si–O–Ta link is created. Such links
are formed when the transition metal is incorporated into the
zeolite framework, thus the XPS results support the conclusion
about the location of metals (Nb and Ta) in the zeolite frame-
work. The lower binding energies for Si/Nb(Ta) = 32 than those for
Si/Nb(Ta) = 64 can be explained by the higher concentration of tran-
sition metal when the Si/Nb(Ta) ratio is lower. Such an increase
in the concentration can induce the interaction between two Nb
or Ta species leading to decreasing binding energy. This effect is
more pronounced for tantalum zeolites. In spite of low number of
Ta atoms calculated per unit cell, such an interaction is possible.
It can occur mainly on the materials surface, where a higher con-
centration of tantalum was estimated from XPS (Table 1). These
results can also give some important information concerning the
isolation of metal species in the zeolites. As one can observe the
difference of binding energies between Nb2O5 or Ta2O5 and nio-
bium or tantalum included to zeolite, respectively, increases with

176
M. Trejda et al. / Catalysis Today 158 (2010) 170–177
Table 2
Table 4
The estimation of LAS strength of niobium and tantalum containing zeolites.
Average conversion and selectivity in cracking of cumene at 623 K.
Catalyst
Lewis acid sites occupied by pyridine after evacuation at
Catalyst
Conversion of cumene, %
Selectivity, %
Propene
423 K
473 K
523 K
573 K
␣-Methyl styrene
NbY-64
NbY-32
TaY-64
TaY-32
100%
100%
100%
100%
51%
55%
64%
48%
34%
20%
20%
11%
7%
3%
2%
2%
NbY-64
Nb Y-32
TaY-64
TaY-32
12
10
8
61
48
42
0
39
52
58
2
100
Catalyst—0.05 g; cumene pulse = 5 ␮l; total He flow—40 cm3 min−1
.
and c) and invisible for TaY-32 sample (Fig. 7d). Therefore, one can
postulate that LAS centres and OH groups are the dominant ones
on the surface of NbY and TaY zeolites.
the reaction is strongly supported by the much higher selectivity
to DMF (33% for NbY-64 and 7% for TaY-64). Among all the cat-
alysts studied, NbY-64 reveals several times higher yield of DMF.
For niobium and tantalum zeolites with the Si/T ratio of 32, the
conversion observed can be related to the real metal (Nb or Ta) con-
centration in the zeolite (Table 1). The TaY-32, which shows real
Si/Ta = 45, is more active than NbY-32 (Si/Nb = 72). However, not
only the amount of transition metal influences the activity but also
the nature of metal. If one compares TaY-64 and NbY-32 samples
having similar numbers of tantalum and niobium species, differ-
ent conversions of acetonylacetone are observed, i.e. 19% and 39%,
respectively. The higher conversion on NbY-32 zeolite should be
related to the privileged role of Brønsted acid centres present in
NbY-32 material, as confirmed by the results of cumene cracking
described below.
Analysis of the products distribution clearly indicates that with
increasing number of Nb or Ta atoms in the zeolite framework,
the yield of MCP increases. It means that the number of basic OH
groups active in the production of MCP increases with increasing
number of metal (Nb or Ta) atoms. The indicator of basicity, i.e.
MCP/DMF ratio changes significantly depending on the number and
nature of transition metal (Table 2). This phenomenon supports
the location of Nb and Ta in the zeolite framework and their +5
oxidation state compensated by four oxygens from the tetrahedra
and one basic OH group. The formation of Nb(V)␦+–OH␦− in the
case of the isomorphous substitution of niobium into mesoporous
molecular sieves was reported before and supported by different
experimental techniques [23].
The strength of Lewis acid centres can be estimated on the basis
of the drop of intensity of IR bands after evacuation of adsorbed
pyridine at different temperatures. Figs. 8 and 9 reveal the results
of such evacuation for TaY and NbY, respectively. Due to the low
concentration of BAS in the samples, the intensity of the band at
1490 cm−1, after evacuation at different temperatures, can give
an information concerning the LAS strength. The decrease of LAS
calculated from the areas of 1490 cm−1 band are also shown in
Table 2. The increase in the tantalum content (from Si/Ta = 78 to
Si/Ta = 45 for TaY-64 and TaY-32, respectively) leads to a weaker
chemisorption of pyridine demonstrated by the lower intensity of
the 1490 cm−1 band after evacuation at a desired temperature. The
same is observed for niobium containing Y zeolites. The strength of
LAS lowers when Ta or Nb forming Lewis acid centres can interact
with each other, as can happen at higher concentration of metal
atoms.
3.4. Acetonylacetone cyclisation
For the evaluation of the acid/base properties in the reac-
tion conditions, the acetonylacetone transformation was carried
out. In the presence of acid centres 2,5-dimethylfuran (DMF)
is formed, whereas the presence of basic centres leads to the
formation of 3-methyl-2-cyclopentenone (MCP) [21,22]. If acid
centres dominate on the catalyst surface the selectivity ratio
MCP/DMF is <1. The ratio of MCP/DMF > 1 indicates that basicity
is dominant. When MCP/DMF ∼ 1, the catalyst has both acid/base
character.
3.5. Cumene cracking
The results of acetonylacetone cyclisation carried out on nio-
bium and tantalum containing Y zeolites prepared by the one-pot
synthesis are shown in Table 3. It can be pointed out that the higher
metal concentration in the sample the higher the conversion of
acetonylacetone is observed for the desired nature of metal (Nb
or Ta). For the materials with the assumed Si/T ratio of 64, the
niobium containing sample shows a little bit higher activity (23%
of acetonylacetone conversion) than the tantalum-containing one
(19% of acetonylacetone conversion) in spite of the fact that the real
molar concentration of Nb is much lower than Ta (Table 1). There-
fore, the higher conversion for NbY-64 sample can be explained
by the presence of Brønsted acid sites observed for this material
(Fig. 7d). The participation of BAS present in NbY-64 sample in
Cracking of cumene is the test reaction used mainly for char-
acterisation of strong acidic centres [24,25]. When strong Brønsted
acid centres are involved in the catalytic process, the main products
formed are benzene and propene. The transformation of cumene
can also take place in the presence of electron-transfer centres.
In the later case, it leads to the formation of different products,
primarily styrene and ␣-methyl styrene.
The results of cumene cracking reaction carried out on Nb and
Ta containing zeolites are shown in Table 4. The lowest activity was
observed for TaY-32 sample (2% of cumene conversion). The same
catalyst showed 100% selectivity to ␣-methyl styrene. It is impor-
tant to stress that for this sample, not even traces of Brønsted acid
sites were detected by pyridine adsorption. The presence of only
Lewis acid sites that can act as electron-transfer centres leads to
the high selectivity to ␣-methyl styrene. The other samples which
contain at least traces of BAS are more active in cracking of cumene.
Moreover, they show also a significant selectivity to propene. This
product dominates for the sample showing the highest number of
BAS, i.e. NbY-64.
Table 3
The conversion and selectivity in the acetonylacetone cyclisation at 623 K.
Catalyst
Conversion, %
Selectivity, %
Yield, %
DMF
MCP/DMF
DMF
MCP
MCP
NbY-64
NbY-32
TaY-64
TaY-32
23
39
19
56
33
3
7
67
97
93
99
7.6
1.2
1.3
0.6
15.4
37.8
17.7
55.4
2
32
13
99
According to the real number of Nb and Ta atoms in the zeolite
framework, the samples can be ordered as follows: NbY-64 ꢀ NbY-
32 ≤ TaY-64 < TaY-32. With increasing number of these atoms, the
selectivity to ␣-methyl styrene increases because it is related to
the number of Lewis acid sites. In the same order the strength of
1
Catalyst—0.05 g; acetonylacetone volume—0.5 cm3; reaction time—0.5 h. DMF—2,5-
dimethylfuran; MCP—3-methyl-2-cyclopentenone.

M. Trejda et al. / Catalysis Today 158 (2010) 170–177
177
acid centres decreases, as shown in the study of pyridine adsorp-
tion/desorption, and it reflects the decrease in cumene conversion.
[2] S. Han, K.D. Schmitt, S.E. Schramm, P.T. Reischman, D.S. Shihabi, C.D. Chang, J.
Phys. Chem. 98 (1994) 4118.
[3] M. Taramasso, G. Perego, B. Notari, US Patent 4,410,501 (1983).
[4] W.Q. Xu, S.L. Suib, C.L. Oyoung, J. Catal. 144 (1993) 285.
[5] S.G. Hegde, R.A. Abdullah, R.N. Bhat, P. Ratnasamy, Zeolites 12 (1992) 951.
[6] T. Blasco, A. Corma, M.J. Diaz-Cabanas, F. Rey, J. Rius, G. Sastre, J.A. Vidal-Moya,
J. Am. Chem. Soc. 126 (2004) 13414.
4. Conclusions
To the best of our knowledge the Nb and Ta containing FAU
zeolites have been successfully synthesised for the first time. Tan-
talum is easier included into the faujasite structure than niobium.
For assumed Si/Nb(Ta) ratio = 64, niobium and tantalum were found
to be present exclusively in the tetrahedrally coordinated positions
indicating the introduction of Nb and Ta (both in an oxidation state
of +5) into the zeolite skeleton. XPS binding energies confirmed
that these metals are located in environments other than in bulk
oxides, i.e. in the zeolite skeleton. The excess of local positive charge
in the zeolite framework in which Nb or Ta is located can be com-
pensated by basic OH groups. These are active in the formation of
MCP in acetonylacetone cyclisation. The Lewis acid centres were
present in all zeolites prepared in this work and their presence was
confirmed by pyridine adsorption and by the results of the cumene
cracking reaction.
[7] G. Calis, P. Frenken, E. de Boer, A. Swolfs, M.A. Hefni, Zeolites 7 (1987) 319.
[8] A. Corma, L.T. Nemeth, M. Renz, S. Valencia, Nature 412 (2001) 423.
[9] J. Kornatowski, B. Wichterlova, J. Jirkovsky, E. Löffer, W. Pitz, J. Chem. Soc.,
Faraday Trans. 92 (1996) 1067.
[10] T. Takewaki, L.W. Beck, M.E. Davis, Topics Catal. 9 (1999) 35.
[11] G. Spoto, A. Zecchina, G. Berlier, S. Bordiga, M.G. Clerici, L. Basini, J. Mol. Catal.
A: Chem. 158 (2000) 107.
[12] Y.-N. Ju, Z.-H. Shen, J. Zhao, J.-Q. Zhao, X.-L. Wang, Acta Phys. Chim. Sinica 22
(2006) 28.
[13] M.L. Occelli, G. Schwering, C. Fild, H. Eckert, A. Auroux, P.S. Iyer, Microporous
Mesoporous Mater. 34 (2000) 15.
[14] B.M. Lok, et al., J. Am. Chem. Soc. 106 (1984) 6092.
[15] L. Kevan, A.M. Prakash, J. Am. Chem. Soc. 120 (1998) 13148.
[16] Y.S. Ko, W.S. Ahn, Microporous Mesoporous Mater. 30 (1999) 283.
[17] A. Corma, F.X. Llabres, I. Xamena, C. Prestipino, M. Renz, S. Valencia, J. Phys.
Chem. C 113 (2009) 11306.
[18] D.M. Ginter, A.T. Bell, C.J. Radke, in: M.L. Occelli, H.E. Robson (Eds.), Synthesis of
Microporous Materials, Vol. 1. Molecular Sieves, Van Nostrand Reinhold, New
York, 1992, p. 6.
[19] N.K. Mal, M. Fujiwara, A. Bhaumik, J. Non-Cryst. Solids 353 (2007) 4116.
[20] S. Khabtou, T. Chevreau, J.C. Lavalley, Microporous Mater. 3 (1994) 133.
[21] R.M. Dessau, Zeolites 10 (1990) 205–206.
Acknowledgements
[22] J.J. Alcaraz, B.J. Arena, R.D. Gillespie, J.S. Holmgren, Catal. Today 43 (1998)
89.
[23] M. Ziolek, I. Sobczak, A. Lewandowska, I. Nowak, P. Decyk, M. Renn, B.
Jankowska, Catal. Today 70 (2001) 169.
[24] D. Best, B.W. Wojciechowski, J. Catal. 47 (1977) 11.
[25] M. Marczewski, B.W. Wojciechowski, Can. J. Chem. Eng. 60 (1982) 622.
Polish Ministry of Science and Higher Education (grants
118/COS/2007/03 and N N204 032536) and COST D36/0006/06 are
acknowledged for a financial support.
References
[1] B.J. Nagy, R. Aiello, G. Giordano, A. Katovic, F. Testa, Z. Kónya, I. Kiricsi, Molecular
Sieves-Science and Technology, vol. 5, 2007, p. 365.