On the synergistic catalytic properties of bimetallic mesoporous materials containing aluminum and zirconium: The prins cyclisation of citronellal

Article abstract of DOI:10.1002/chem.201002909

Bimetallic three-dimensional amorphous mesoporous materials, Al-Zr-TUD-1 materials, were synthesised by using a surfactant-free, one-pot procedure employing triethanolamine (TEA) as a complexing reagent. The amount of aluminium and zirconium was varied in order to study the effect of these metals on the Bronsted and Lewis acidity, as well as on the resulting catalytic activity of the material. The materials were characterised by various techniques, including elemental analysis, X-ray diffraction, high-resolution TEM, N₂ physisorption, temperature-programmed desorption (TPD) of NH₃, and ²⁷Al MAS NMR, XPS and FT-IR spectroscopy using pyridine and CO as probe molecules. Al-Zr-TUD-1 materials are mesoporous with surface areas ranging from 700-900am² g^-1, an average pore size of around 4anm and a pore volume of around 0.70acm³ g ^-1. The synthesised Al-Zr-TUD-1 materials were tested as catalyst materials in the Lewis acid catalysed Meerwein-Ponndorf-Verley reduction of 4-tert-butylcyclohexanone, the intermolecular Prins synthesis of nopol and in the intramolecular Prins cyclisation of citronellal. Although Al-Zr-TUD-1 catalysts possess a lower amount of acid sites than their monometallic counterparts, according to TPD of NH₃, these materials outperformed those of the monometallic Al-TUD-1 as well as Zr-TUD-1 in the Prins cyclisation of citronellal. This proves the existence of synergistic properties of Al-Zr-TUD-1. Due to the intramolecular nature of the Prins cyclisation of citronellal, the hydrophilic surface of the catalyst as well as the presence of both Bronsted and Lewis acid sites synergy could be obtained with bimetallic Al-Zr-TUD-1. Besides spectroscopic investigation of the active sites of the catalyst material a thorough testing of the catalyst in different types of reactions is crucial in identifying its specific active sites.

Full text of DOI:10.1002/chem.201002909

FULL PAPER
DOI: 10.1002/chem.201002909
On the Synergistic Catalytic Properties of Bimetallic Mesoporous Materials
Containing Aluminum and Zirconium: The Prins Cyclisation of Citronellal
Selvedin Telalovic´,[a] Anand Ramanathan,[b] Jeck Fei Ng,[c] Rajamanickam Maheswari,[d]
Cees Kwakernaak,[e] Fouad Soulimani,[f] Hans C. Brouwer,[e] Gaik Khuan Chuah,[c]
Bert M. Weckhuysen,[f] and Ulf Hanefeld*[a]
Dedicated to Prof. Dr. Hartmut Laatsch, on the occasion of his 65th birthday
Abstract: Bimetallic three-dimensional
amorphous mesoporous materials, Al-
Zr-TUD-1 materials, were synthesised
by using a surfactant-free, one-pot pro-
cedure employing triethanolamine
(TEA) as a complexing reagent. The
amount of aluminium and zirconium
was varied in order to study the effect
of these metals on the Brønsted and
Lewis acidity, as well as on the result-
ing catalytic activity of the material.
The materials were characterised by
various techniques, including elemental
analysis, X-ray diffraction, high-resolu-
tion TEM, N2 physisorption, tempera-
ture-programmed desorption (TPD) of
NH3, and 27Al MAS NMR, XPS and
FT-IR spectroscopy using pyridine and
CO as probe molecules. Al-Zr-TUD-1
materials are mesoporous with surface
areas ranging from 700–900 m2 gÀ1, an
average pore size of around 4 nm and a
pore volume of around 0.70 cm3 gÀ1.
The synthesised Al-Zr-TUD-1 materi-
als were tested as catalyst materials in
the Lewis acid catalysed Meerwein–
Ponndorf–Verley reduction of 4-tert-bu-
tylcyclohexanone, the intermolecular
Prins synthesis of nopol and in the in-
tramolecular Prins cyclisation of citro-
nellal. Although Al-Zr-TUD-1 catalysts
possess a lower amount of acid sites
than their monometallic counterparts,
according to TPD of NH3, these mate-
rials outperformed those of the mono-
metallic Al-TUD-1 as well as Zr-TUD-
1 in the Prins cyclisation of citronellal.
This proves the existence of synergistic
properties of Al-Zr-TUD-1. Due to the
intramolecular nature of the Prins cyc-
lisation of citronellal, the hydrophilic
surface of the catalyst as well as the
presence of both Brønsted and Lewis
acid sites synergy could be obtained
with bimetallic Al-Zr-TUD-1. Besides
spectroscopic investigation of the
active sites of the catalyst material a
thorough testing of the catalyst in dif-
ferent types of reactions is crucial in
identifying its specific active sites.
Keywords: Brønsted acids · hetero-
geneous catalysis · Lewis acids ·
mesoporous materials · synergy
´
[a] S. Telalovic, Dr. U. Hanefeld
[e] C. Kwakernaak, H. C. Brouwer
Gebouw voor Scheikunde, Technische Universiteit Delft
Julianalaan 136, 2628 BL Delft (The Netherlands)
Fax : (+31)15-278-1415
Department of Materials Science and Engineering
Maritime and Materials Engineering
Technische Universiteit Delft
E-mail: U.Hanefeld@tudelft.nl
Mekelweg 2, 2628 CD Delft (The Netherlands)
[b] Dr. A. Ramanathan
[f] F. Soulimani, Prof. Dr. B. M. Weckhuysen
Inorganic Chemistry and Catalysis Group
The Center for Environmentally Beneficial Catalysis
The University of Kansas, 1501 Wakarusa
Bldg. B, Lawrence, KS 66047 (USA)
Debye Institute for Nanomaterials Science
Utrecht University, 3584 CA Utrecht (The Netherlands)
[c] J. F. Ng, Prof. Dr. G. K. Chuah
Department of Chemistry
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201002909.
National University of Singapore, Kent Ridge
Singapore 119260 (Republic Singapore)
Re-use of this article is permitted in accordance with the Terms and
Conditions set out at http://onlinelibrary.wiley.com/journal/10.1002/
TERNN(UG ISSN)1521-3765/homepage/2111_onlineopen.html.
ACHTUNG-
[d] Dr. R. Maheswari
Department of Chemistry, Anna University
Chennai 600025 (India)
Chem. Eur. J. 2011, 17, 2077 – 2088
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2077

Introduction
An alternative approach to studying the synergy between
Brønsted and Lewis acid sites, as seen in the industrially im-
portant Y zeolites, is to synthesise a bimetallic mesoporous
catalyst. This approach necessitates the partial replacement
of aluminium inside an aluminosilicate material possessing
both Brønsted and Lewis acid sites by a transition metal
generating exclusively Lewis acid sites. Framework incorpo-
rated transition metals with different electronic properties
from aluminium would affect the strength of the Brønsted
acid sites already present. Three-dimensional mesoporous
materials lend themselves particularly for such an approach
as diffusion effects are virtually none existent and if synergy
between acid sites does exist, it would improve the catalytic
activity of three-dimensional mesoporous materials to that
of zeolites. Thus diffusion limitations would be reduced
without loss of acidity based activity, relative to zeolites.
This bimetallic approach making use of mesoporous MCM-
41 materials has been explored for Zn–Al and Ce–Al com-
binations and yielded indications for synergy; however, con-
clusive results have not yet been obtained and this necessi-
tates the exploration of other suitable mesoporous matri-
ces.[9,10] However, very recently synthesised Fe–Al bimetallic
catalysts using MCM-41 or SBA-15 as supports proved to
possess a synergistic effect in oxidation of benzyl alco-
hol.[11,12]
TUD-1 is such a porous solid, as it is a well-established
mesoporous material with a three-dimensional structure,
tuneable pore size distribution and surface area reaching
values of 900 m2 gÀ1.[13,14] The main advantage of TUD-1
concerning the study of Brønsted–Lewis acid synergy is that
it is excellent for framework incorporation of different
metals.[15–17] This high degree of framework incorporation is
due to the complexing agent used in the synthesis of TUD-
1, triethanolamine. It forms atrane complexes with different
metals (M), thus ensuring incorporation as isolated metals
rather than metal oxide clusters. By employing this ap-
proach, a wide range of M-TUD-1 with framework incorpo-
rated metals can be prepared and Al-TUD-1 materials pos-
sessing both Brønsted and Lewis acid sites have been ap-
plied in the Lewis acid catalysed Friedel–Crafts alkylation
of phenol.[18] Zr-TUD-1 on the other hand possesses exclu-
sively Lewis acid sites due to a high degree of metal incor-
poration inside the TUD-1 matrix.[17] By combining both
metals inside the TUD-1 matrix it should become possible
to fine-tune the ratio of Lewis and Brønsted acid sites due
to the different types of acidity of the metals.[19] Thus Al-Zr-
TUD-1 should be a showcase material with adjustable
Brønsted and Lewis acidity, ideal to investigate whether syn-
ergy between these two types of acidity really exists and can
be exploited for fine tuning catalytic activity.
Synergy between different types of catalytic sites is a much
sought after objective. Synergy might occur between Lewis
and Brønsted acid sites in monometallic materials or might
be induced by the application of two different active metals.
In this respect, the increased activity of hydrothermally
treated faujasite-type Y zeolites in catalytic cracking reac-
tions has intrigued scientists for decades. Already very early,
the synergy between extra-framework aluminium, Lewis
acid sites formed due to partial release of aluminium from
the framework upon hydrothermal treatment and Brønsted
acid sites was suggested.[1] Enhanced acidity of the catalyst
was explained by direct coordination of Lewis acid sites to
the Brønsted acid site. Very recently, a different type of
mechanism for Brønsted–Lewis acid synergy was proposed
based on NMR experiments and DFT calculations: Lewis
acid sites in the form of extra-framework aluminium coordi-
nate to the oxygen atom nearest to the Brønsted acid site,
though there is no direct interaction between them.[2]
Besides the above-mentioned synergy between Brønsted
and Lewis acid sites in zeolites, there are also other relevant
examples found in mesoporous materials. Immobilisation of
different lanthanide triflates inside the pores of siliceous
SBA-15 led to different Brønsted/Lewis ratios and corre-
sponding catalytic activity in Friedel–Crafts acylation of
naphthalene with p-toluoyl chloride.[3] The synergy found
for these catalytic systems was explained through the con-
finement of large molecules, such as lanthanide triflates
inside the pores of SBA-15, causing significant physical per-
turbation of the surface hydroxyl groups of the host matrix
giving rise to the formation of certain Brønsted type surface
acid sites.
Most common and widespread examples of synergy are
those found in different bimetallic catalysts used in several
important reactions. Synthesis of bimetallic catalysts, such as
amorphous SnO2–ZrO2, so as to achieve synergistic effects
has led to improved catalysts for SO2 reduction produced in
the direct sulfur recovery process (DSRP).[4] Steam or auto-
thermal reforming of hydrocarbons is an important source
of H2 for proton-exchange, membrane fuel cells (H2-
PEMFC). However, reformed gas contains considerable
amounts of CO leading to poisoning of the fuel cell. Pt and
Pd mutually immobilised on perovskite oxide showed higher
activity and stability in the water gas shift reaction in order
to reduce CO compared to their monometallic counter-
parts.[5] The same effect was observed by applying CuO–
CeO2 instead of individual oxides (CuO and CeO2) or their
physical mixtures.[6] The properties of light-metal hydrides,
which are interesting for their high hydrogen storage capaci-
ty, were improved by the addition of Fe and Zr through en-
hanced H2 desorption kinetics.[7] Fe-Ce-ZSM-5 catalysts
showed high NO conversion in the selective catalytic reduc-
tion of NO with NH3 in a wide temperature window and
even in presence of H2O and SO2 relative to Fe-ZSM-5 or
Ce-ZSM-5.[8]
This is the topic of this article: Al-Zr-TUD-1 has been
used to investigate the presence of synergy between Brønst-
ed and Lewis acid sites by performing a systematical varia-
tion of aluminium and zirconium inside the TUD-1 matrix
with constant Si/M ratio instead of increasing the weight
percentage of one metal, while keeping the weight percent-
age of the second constant. Bimetallic mesoporous catalysts
2078
www.chemeurj.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 2077 – 2088

Bimetallic Mesoporous Materials
FULL PAPER
Table 1. Physicochemical and acidic properties of TUD-1-based catalysts.
were compared with their mono-
metallic counterparts Al-TUD-1
and Zr-TUD-1 by using both
spectroscopic methods and cata- Al-TUD-1
[a]
nSi/nM
nSi/nAl nSi/nZr nSi/n(Al+Zr) SBET
[m2 gÀ1
dP,BJH VP,BJH
Total acidity
B/L
G
]
[nm]
A
]
[mmol NH3 gÀ1
]
ratio[b]
25
26.6
31
47
92
–
–
135
69
45
25
26.6
25
28
30
25
956
705
686
877
764
3.7
4.2
4.6
3.3
8.8
0.95
0.70
0.85
0.70
1.23
0.40
0.33
0.32
0.38
0.69
2.41
2.15
2.05
3.21
–
Al-Zr-4.3:1 25
Al-Zr-1.5:1 25
lytic conversions. It will be
shown that the synergistic effect
between different Brønsted and
Lewis acid sites in bimetallic
mesoporous catalysts is very spe-
cific, substrate as well as reac-
tion mechanism dependent.
Al-Zr-1:2
Zr-TUD-1
25
25
[a] Synthesis mixture ratio of silica to the total amount of metal incorporated. [b] Determined by dividing the
area under the Lewis acid region (1460–1440 cmÀ1) and the Brønsted acid region (1557–15378 cmÀ1) found in
spectra recorded at 2008C.
found during the synthesis of zirconium-containing alumino-
Results and Discussion
silicate of BEA structure, Zr-Al-b. The Si/Al ratio of cal-
cined Zr-Al-b was similar to the synthesis gel while the Si/
Zr ratio was higher than the ratio of the synthesis gel.[20]
The 27Al-NMR spectra of the three bimetallic catalysts
are similar and reveal that around 50% of aluminium is tet-
rahedrally incorporated, around 30% is hexa-coordinated
aluminium and is assigned to extra-framework aluminium
responsible for Lewis acidity and finally around 20% of alu-
minium is penta-coordinated (Figure 2). The Al in Al-Zr-
The mesoporous nature of Al-Zr-TUD-1 samples is evi-
denced by the high-resolution transmission electron micro-
graphs (HR-TEM) and representative images are given in
Figure 1. All samples exibit a sponge-like structure typical
for TUD-1. HR-TEM analysis proves the framework incor-
poration of Al and Zr in the TUD-1 matrix. No crystalline
nanoparticles of ZrO2 or Al2O3 were detected, suggesting
that both metals were incorporated into the framework as
expected. This is in line with earlier results for monometallic
Zr-TUD-1 and Al-TUD-1 with Si/M ratio of 25.[17,18] Ac-
cordingly XRD analysis (Figure S1 in the Supporting infor-
mation) further confirmed the mesoporous character of Al-
Zr-TUD-1.
N2-physisorption measurements obtained at 77 K (Fig-
ure S2 and S3 in the Supporting information) and the results
from elemental analysis (ICP-OES) are summarised in
Table 1. The surface area of the three samples varies from
689 to 877 m2 gÀ1 and is not correlated with the amount of
the different metals present. Pore size (dp,BHJ) and pore
volume (VP,BHJ) are also not correlated to the type of metal
present in the Al-Zr-TUD-1 samples. The Si/Zr ratios pres-
ent in calcined samples differ from those in the initial syn-
thesis gel. The more zirconium present in the sample, the
higher is the deviation from the initial synthesis gel Si/AHCTUNG-
Figure 2. 27Al MAS NMR spectrum of Al-Zr-1.5:1 catalyst, representative
for all Al-Zr-TUD-1 materials with Si/M ratio of approximately 25.
T
ACHTUNGTREN(NGNU Al+Zr) ratio of
Figure 1. HR-TEM images of Al-Zr-TUD-1 materials with a Si/M ratio of approximately 25 varying in their Al:Zr ratio: Al-Zr-1.5:1 (left), Al-Zr-1:2
(middle) and Al-Zr-4.3:1 (right). The bars represent 20 nm.
Chem. Eur. J. 2011, 17, 2077 – 2088
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2079

U. Hanefeld et al.
TUD-1 induces therefore both Brønsted and Lewis acid
sites.
cated that Al and Zr are incorporated in the TUD-1 frame-
À
work at the expense of the Si O bonds.
From the XPS measurements, the binding energies for
Si 2p and O 1s photoelectron lines resemble those found in
silicon dioxide (SiO2). The spectra of O 1s, however, can be
deconvoluted in five species for mono- and bimetallic TUD-
1 catalysts (Table 2). The National Institute of Standards
and Technology (NIST) X-ray photoelectron spectroscopy
database give a value of the O 1s binding energy of 532.9Æ
0.9 eV for SiO2 and 530.6Æ0.7 eV for both Al2O3 and
ZrO2.[21]
The binding energy of about 183.7Æ0.1 eV of Zr 3d5/2 is
[21]
higher than the values reported for ZrO2 and corresponds
to the values found in complex zirconium oxides as reported
for ZrSiO4 and zirconium silicate alloy thin films.[23,24] All
zirconium found is incorporated into the SiO2 framework.
Based on NH3 temperature-programmed desorption
(TPD) analysis (Figure 3) of different Al-Zr-TUD-1 cata-
lysts, the acidity is of comparable order in all samples. The
monometallic Zr-TUD-1 possesses a larger amount of weak
acid sites and Al-TUD-1 has a
larger amount of strong acid
Table 2. The XPS binding energy [in eV] of the O 1s, Al 2p, Si 2p and Zr 3d photoelectrons and position of
the Auger lines for Al, Si and Zr.
sites compared to the other cat-
alysts (Table 1). Mutual incor-
poration of Al and Zr into
TUD-1 leads to both a decrease
of weak as well as strong acid
sites and thus to a decrease of
overall acidity.
IR spectra of a KBr pressed
disc of three samples in the
skeletal region (Figure 4) show
typical bands at 1093 cmÀ1 and
Si-TUD-1
Zr-TUD-1
Al-TUD-1
Al-Zr 1:2
Al-Zr 1.5:1
Al-Zr 4.3:1
Al 2p
–
74.1
75.1
74.9
74.4
74.8
Si 2p
103.5
103.5
103.5
103.5
103.5
103.5
Zr 3d5/2
Zr 3d3/2
O 1s (1)
O 1s (2)
O 1s (3)
O 1s (4)
O 1s (5)
–
–
–
183.7
186.2
184.1
186.6
183.7
186.1
183.6
186.0
183.9
186.30
528.33
530.16
531.92
533.01
534.18
528.45
530.24
531.92
533.01
534.18
528.45
530.34
531.92
533.01
534.18
528.56
530.39
531.92
533.01
534.18
528.57
530.34
531.92
533.01
534.18
–
531.92
533.01
534.18
À
The Si-TUD-1 containing Si O bonds only, has three
O 1s peaks with a main peak always present at 533.0 eV.
This main peak is flanked by two peaks with binding ener-
gies of 531.9 and 534.2 eV respectively (Table 2, Figure S4 in
the Supporting Information) The binding energy of O 1s
around 534 eV is due to the presence of air borne species
such as water vapour, carbon dioxide or remaining template.
An alternative explanation of the three peaks suggests that
in Si-TUD-1, an amorphous mesoporous material, three
prominent configurations of Si atoms bonded to an O atom
are present.
Monometallic as well as bimetallic TUD-1 catalysts have
O 1s binding energies that have a lower value than that of
silicon oxide (Table 2). Here, the electronegativity of the Al
(cPauling =1.61) and Zr (cPauling =1.33), as opposed to Si
(cPauling =1.90), renders the oxygen atom to be more ionic.[22]
The mutual repulsions of the electrons in the oxygen ion re-
duces the O 1s1/2 binding energy.
Figure 3. NH3-TPD profile of different TUD-1 catalysts with Si/M ratio
of approximately 25: Zr-TUD-1, Al-TUD-1, Al-Zr-4.3:1, Al-Zr-1.5:1 and
Al-Zr-1:2.
The presence of incorporated metal ions in the mesopo-
rous framework apparently induces two spectral features at
about 528.5 and 530.3 eV respectively. The latter peak can
be ascribed to the “normal” ionic compounds Al2O3 or
ZrO2.[21] It is also very clear from the spectra that it does
not make a difference whether Al or Zr is involved (Fig-
ure S4 in the Supporting Information). They produce nearly
the same spectra. The nature of the Al and Zr species is
only derived from the binding energies of the metallic spe-
cies. The presence of the ionic species does not affect the
fraction of the flanking peaks too much (i.e., O 1s (3) and
O 1s (5) in Table S1 in the Supporting Information), but re-
duces the central peak remarkably (i.e. O 1s (4)). This indi-
a shoulder at 1220 cmÀ1 due to asymmetric stretching vibra-
tions of Si-O-Si bridges. The band at 798 cmÀ1 is caused by
symmetric stretching vibration of Si-O-Si.[25] The signal for
Si-OH or Si-O-M at approximately 970 cmÀ1 is not re-
solved.[23,26,27]
FT-IR of the OH region of the three bimetallic catalysts
and those of Zr-TUD-1 and Al-TUD-1 (Figure 5) show a
band centred around 3745 cmÀ1, asymmetric towards lower
frequencies, commonly assigned to terminal silanol groups.
However, there is a distinct difference between monometal-
lic and bimetallic catalysts. Major bands centred at
3745 cmÀ1 for the bimetallic catalysts have higher values for
2080
www.chemeurj.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 2077 – 2088

Bimetallic Mesoporous Materials
FULL PAPER
bonded hydroxyl groups in the 3700–3400 cmÀ1 region. How-
ever, the presence of Brønsted acid sites of zeolitic strength
should not be excluded. In addition to catalytic experiments
performed with amorphous aluminosilicates proving the ex-
istence of Brønsted acid sites of zeolitic strength, further
spectroscopic evidence has been provided in a recent publi-
cation.[30] However, their density is much lower than that
which can be found in zeolites.
Pyridine FT-IR spectroscopic data: To differentiate between
Brønsted and Lewis acid sites in the Al-Zr-TUD-1 catalysts,
pyridine FT-IR spectroscopy was employed (Figure 6). Pyri-
dine is more reliable probe molecule than ammonia, since
the IR absorption bands do not overlap. Moreover, ammo-
nia decomposes even at rather low temperatures, when ad-
sorbed onto catalysts.[31]
Figure 4. FT-IR skeletal spectra of different TUD-1 catalysts with Si/M
ratio of approximately 25: Zr-TUD-1, Al-TUD-1, Al-Zr-4.3:1, Al-Zr-
1.5:1 and Al-Zr-1:2.
Figure 6. FT-IR spectra after pyridine desorption at 2008C of different
TUD-1 catalysts with Si/M ratio of approximately 25: Al-Zr-4.3:1, Al-Zr-
1.5:1 Al-Zr-1:2, Zr-TUD-1 and Al-TUD-1. B stands for Brønsted acid
site at 1545 cmÀ1 and Lewis acid sites L1 (associated with aluminium)
Figure 5. FT-IR spectra in OH region acquired at 4008C of different
TUD-1 catalysts with Si/M ratio of approx 25: Zr-TUD-1, Al-TUD-1, Al-
Zr-4.3:1, Al-Zr-1.5:1 and Al-Zr-1:2.
and L2 (associated with zirconium) at respectively 1455 and 1448 cmÀ1
.
For the Al-TUD-1 sample, the presence of Lewis acid
sites is indicated by bands at 1455 (L1) and 1623 cmÀ1. How-
ever these bands are not symmetrical and show the presence
of other Lewis acid sites at 1460 and 1620 cmÀ1. The pres-
ence of these bands has already been reported for H-beta,
other zeolites and silica-alumina.[32] The IR band at
1460 cmÀ1 was assigned to iminium ions, formed by attack of
protons on the pyridine complex bound to Lewis acid sites.
Brønsted acidity is clearly present in Al-TUD-1 as indicated
by the band at 1545 cmÀ1 (B).
For Zr-TUD-1 the strength of the Lewis acid site at
1448 cmÀ1 (L2) accompanied by a band at 1612 cmÀ1 seems
to be lower than is the case for Al-TUD-1. While for Al-
TUD-1 temperatures above 4008C are needed to desorb
pyridine, for Zr-TUD-1 temperatures above 2008C are al-
ready sufficient. Based on this we can conclude that Lewis
acidity imparted by partial exchange of aluminium with zir-
conium generates weaker Lewis acid sites. Brønsted acid
sites seem to be absent at outgassing temperatures of 2008C,
which is in line with earlier reports on the synthesis of Zr-
the full width at half maximum compared to monometallic
catalysts. This is an indication of higher degree of heteroge-
neity of different silanol groups inside Al-Zr-TUD-1 cata-
lysts.
In addition to isolated silanol groups all catalysts display a
broad adsorption band in the 3700–3400 cmÀ1 region. Usual-
ly broad absorption in this region is due to the stretching vi-
brations of hydrogen-bonded hydroxyl groups.[28] This under-
lying broad band in this region is more prominently present
in Al-Zr-TUD-1 catalysts than it is in their monometallic
counterparts. For Al-Zr-TUD-1 catalysts a shoulder at
3580 cmÀ1 is clearly distinguishable.
The strong bridged hydroxyl groups (Brønsted acid site)
at around 3613 cmÀ1[29] as seen in zeolites could not be dis-
tinguished in any of the mesoporous TUD-1 samples with
incorporated aluminium; which is typical for mesoporous
materials as they possess lower Brønsted acidity. Further-
more, due to their amorphous nature the presence of
Brønsted acid sites is overshadowed by different hydrogen-
Chem. Eur. J. 2011, 17, 2077 – 2088
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2081

U. Hanefeld et al.
TUD-1 leading to framework incorporation of tetravalent
zirconium.[17]
Al-Zr-TUD-1 samples show the presence of all the bands
present in Al-TUD-1 and Zr-TUD-1. Varying the Al/Zr
ratio, but keeping the Si/M ratio constant, leads to variation
of the ratio of different Lewis acid sites L1/L2. However,
the relation between Lewis acid sites L1 and L2 in Al-Zr-
TUD-1 samples does not follow the ratio between the differ-
ent metals incorporated into the TUD-1 matrix.
The Lewis/Brønsted ratio in all three samples varies
(Table 1). For the bimetallic samples containing large
amounts of aluminium the Lewis/Brønsted ratio is smaller
than that found in Al-TUD-1. However, the sample with the
highest amount of zirconium incorporated, Al-Zr-1:2, has a
larger Lewis/Brønsted ratio than Al-TUD-1, 3.21 compared
to 2.41. Overall, five catalysts with varying Lewis/Brønsted
ratios are thus available.
CO FT-IR spectroscopic data: CO is a weak base that ad-
sorbs end-on through the carbon on polarised sites.[29] In
contrast to more basic ammonia and pyridine, CO is capable
of differentiating between sites of very similar acid strength.
Its small size, weak basicity, low reactivity (at low tempera-
tures) and sensitivity make it ideal for investigation of sam-
ples with both Brønsted and Lewis acid sites.[33]
Due to the similarities of the shifts of nOH and nCO modes
of all Al-Zr-TUD-1 materials and monometallic TUD-1 cat-
alysts only the Al-Zr-1.5:1 FT-IR spectra are discussed here
(Figure 7, nOH and nCO modes of other catalysts, see Support-
ing information Figure S5 and S6).
Figure 7. FT-IR difference spectra following CO adsorption obtained at
77 K of Al-Zr-1.5:1 sample: Top: n(OH) region [the spectra are presented
as difference plots: from the measured spectra after adsorption of CO, a
spectrum of the corresponding pre-treated catalyst has been subtracted—
a positive contribution represents peaks than are growing as a result of
CO adsorption whereas negative contributions represent peaks that are
reduced in intensity upon CO adsorption]. Bottom: the n(CO) region in
which Brønsted acid sites are marked by capital letter B, while Lewis
acid sites corresponding to different metals are marked by capital letters
L1 for aluminium and L2 for zirconium.
In the difference spectra of nOH mode of Al-Zr-1.5:1 at
very low pressures of CO
a broad band appears at
3506 cmÀ1, without development of a clear negative band.
At such low pressures, CO will first interact with stronger
acid sites as Brønsted acid sites and hydroxyl groups stron-
ger than isolated silanol groups. The broadness of this band
suggests heterogeneity of the sample. Even though no clear
negative band has developed the red shift of this band
(DnOH) relative to isolated silanol groups is equal to
241 cmÀ1. This value lies between the DnOH for silanol
groups (90 cmÀ1) and DnOH values for strong Brønsted acid
sites found in zeolites (300 cmÀ1), implying the presence of
medium strong Brønsted acid sites.[34,35] Isolated silanol
3468 cmÀ1, owing to the solvent effect.[36] The very sharp
peak developed at 3650 cmÀ1 at high CO pressure is related
to the measuring equipment.
groups at these low pressures are unperturbed and appear
The accompanied nCO mode at low pressure of CO shows
two bands appearing at 2229 and 2193 cmÀ1, the latter band
À1
À
as a sharp peak at 3750 cm . Isolated Si OH groups, usual-
ly encountered at 3746 cmÀ1, display a blueshift upon cool-
having much higher intensity. The blueshift of CO (DnCO)
ing to 3750 cmÀ1, a known temperature effect.[29]
relative to the free molecule (nCO liquid is 2138 cmÀ1) is
91 cmÀ1 for the band at 2229 cmÀ1, which is usually associat-
ed with strong Lewis acid sites related to highly coordina-
With gradual increase of CO pressure the sharp band at
3750 cmÀ1 associated with weak silanol groups gradually de-
creases and is accompanied by development of a positive
band at 3660 cmÀ1 with a shoulder at 3598 cmÀ1, once all iso-
lated silanol groups are saturated. At the same time nega-
tive bands at 3746 and 3741 cmÀ1 develope, with a shoulder
at 3716 cmÀ1. The redshift of isolated silanol groups DnOH is
equal to 87 cmÀ1, which is common for silanol groups.[33,34]
The broad band at 3506 cmÀ1 associated with Brønsted acid
sites and strong hydroxyl groups undergo a redshift to
tively unsaturated extra framework Al3+ [37,38] The low inten-
.
sity of this band indicates that the amount of extra frame-
work Al species is quite low. The band at 2193 cmÀ1, which
has a much higher intensity, is associated with CO adsorbed
on Zr4+ Lewis acid sites, based on CO FT-IR measurements
on Zr-TUD-1 and literature values.[39,40] Lewis acid sites re-
lated to Zr4+ with a lower frequency shift (DnCO =58 cmÀ1)
are weaker than Lewis acid sites associated with extra
2082
www.chemeurj.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 2077 – 2088

Bimetallic Mesoporous Materials
FULL PAPER
framework Al3+. This reconfirms the results of the pyridine
FT-IR.
With the increase of CO pressure additional bands
become evident. Between the high intensity band at
2159 cmÀ1 (generally associated to CO adsorbing to silanol
groups) and 2191 cmÀ1 band (shifted due to solvent effect) a
new band appears at 2177 cmÀ1 that can be assigned to
Brønsted acid sites according to CO FT-IR measurements
performed on Al-TUD-1 and literature values.[41,42] With fur-
ther increase of CO pressure, bands associated with liquid
like CO at 2138 cmÀ1 develop. A very weak band at
2110 cmÀ1 is present as well and has been explained by CO
interacting with pairs of ions through oxygen ends.[37]
Substantial shifts in the position of Brønsted acid sites in
different Al-Zr-TUD-1 catalysts have not been observed as
would be expected if synergistic interaction between Lewis
and Brønsted acid sites were present (Table 3).
Scheme 1. Meerwein–Ponndorf–Verley reduction of 4-tert-butylcyclohex-
anone with isopropanol.
Table 3. Summary of the FT-IR data on the TUD-1 materials: nOH and
nCO modes upon CO adsorption at 77 K for Zr-TUD-1, Al-Zr-4.3:1, Al-
Zr-1.5:1, Al-Zr-1:2 and Al-TUD-1.
nOH
[cmÀ1
DnOH
[cmÀ1
nCO
[cmÀ1
nCO
[cmÀ1
nCO
[cmÀ1
]
Si-OH-Al
Figure 8. Reduction of 4-tert-butylcyclohexanone (2 mmol) in isopropanol
(4 mL) at 808C in the presence of 1,3,5-triisopropylbenzene (0.1 mL) as
an internal standard, using different bimetallic catalysts (50 mg) with Si/
M ratios of approximately 25: Al-Zr-4.3:1, Al-Zr-1.5:1 and Al-Zr-1:2.
[a]
[a]
U
]
A
]
R
]
G
]
N
Al3+
Zr4+
Zr-TUD-1
Al-Zr-4.3:1
Al-Zr-1.5:1
Al-Zr-1:2
3514
3524
3506
3525
3506
233
223
241
222
241
–
2200
2190
2193
2193
–
–
2227
2229
2232
2230
2177
2177
–
Using aluminium-rich Al-Zr-4.3:1, 30% conversion was ob-
tained after 24 h compared to 64% obtained with the Zr-
rich catalyst, Al-Zr-1:2. The more zirconium present the
more active the catalyst in this Lewis acid catalysed reac-
tion. Comparison of Al-Zr-1.5:1 with monometallic Al-
TUD-1 and Zr-TUD-1 and their physical mixtures
(Figure 9) demonstrates absence of synergy between Lewis
and Brønsted acid sites in the Lewis acid catalysed MPV re-
duction of 4-tert-butylcyclohexanone; the physical mixture
Al-TUD-1
2177
[a] nOH mode of a broad band appearing at low CO pressure in the differ-
ence spectra.
Overall extensive spectroscopic investigation of the sur-
face of different Al-Zr-TUD-1 catalysts based on pyridine
FT-IR led to the conclusion that partial exchange of alumi-
nium by zirconium in Al-TUD-1 leads to different propor-
tions of Lewis and Brønsted acid sites, but not necessarily to
the increase of their strength according to CO FT-IR.
Performance of Al-Zr-TUD-1 as a catalyst: The best way to
probe whether synergy between Lewis and Brønsted acid
sites in Al-Zr-TUD-1 catalysts exists is to employ them in
different Lewis and/or Brønsted acid catalysed reactions. In
that way the presence of different kinds of acid sites or in-
crease in their strength or amount due to mutual incorpora-
tion of aluminium and zirconium in the TUD-1 matrix
should reflect itself in increased catalytic activity: synergy.
For that reason the catalysts were tested in the Lewis acid
catalysed Meerwein–Ponndorf–Verley (MPV) reduction and
À
the C C bond formation reactions with the Prins reaction.
This reaction is catalysed both by Lewis and Brønsted acids
and it was performed intermolecularly (nopol synthesis) and
intramolecularly (isopulegol synthesis).
From the study of different Al-Zr-TUD-1 catalysts in the
MPV reduction of 4-tert-butylcyclohexanone (Scheme 1), it
is clear that the increase in amount of zirconium present in
the catalysts leads to an increase of conversion (Figure 8).
Figure 9. Comparison of bimetallic Al-Zr-1.5:1 catalyst with Al-TUD-1
and Zr-TUD-1 and their physical mixture in reduction of 4-tert-butylcy-
clohexanone (2 mmol) in isopropanol (4 mL) at 808C in the presence of
1,3,5-triisopropylbenzene (0.1 mL) as an internal standard, using different
amounts of activated catalysts. In all reactions the trans:cis ratio of 4-tert-
butylcyclohexanol was always around 87:13.
Chem. Eur. J. 2011, 17, 2077 – 2088
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2083

U. Hanefeld et al.
showed the same catalytic behaviour as the bimetallic cata-
lyst. The MPV reduction is catalysed by coordination of an
alcohol and ketone or an aldehyde to the Lewis acid form-
ing a six-membered transition state enabling a carbon-to-
carbon hydride transfer (Scheme 2).[17,43–45] The more Lewis
acid sites a catalysts possesses the more active is the cata-
lyst.
Scheme 2. Mechanism of the Meerwein–Ponndorf–Verley reduction.
In the nopol synthesis (Scheme 3), a Brønsted and Lewis
acid catalysed Prins reaction, no synergy was observed,
either. While all of the catalysts were active in the conver-
Scheme 3. Intermolecular Prins nopol synthesis from b-pinene and para-
ACHTUNGTRENNUNGformaldehyde in toluene at temperature of 808C.
Figure 10. Nopol synthesis performed in toluene (4 mL) at 808C using
paraformaldehyde (2 mmol), 1,3,5-triisopropylbenzene (0.5 mmol) as in-
ternal standard, b-pinene (1 mmol) and activated catalyst (50 mg). A)
Results obtained with bimetallic catalysts Al-Zr-1:2, Al-Zr-1.5:1 and Al-
Zr-4.3:1. B) Results obtained with monometallic catalysts Zr-TUD-1 and
Al-TUD-1.
sion of b-pinene, the yield of nopol was low (Figure 10). Al-
TUD-1 and Al-Zr-4.3:1-TUD-1 had the highest activity (full
conversion of b-pinene within 1 h). The highest selectivity
towards nopol was around 30% obtained after 1 h reaction
time with Zr-TUD-1 (Figure 10). An increase of reaction
time led to product degradation. The major side products in
nopol synthesis were, according to GC-MS analysis, isomeri-
sation products such as limonene, camphene and terpino-
lene.
However, excellent results were obtained in Prins cyclisa-
tion of (Æ)-citronellal (Scheme 4). All three samples of Al-
Zr-TUD-1 catalysts outperformed their monometallic coun-
terparts Al-TUD-1 and Zr-TUD-1 (Figure 11). This, despite
of the fact that both Al-TUD-1 and Zr-TUD-1 possess
larger amounts of acid sites according to NH3 TPD. Alumi-
nium-rich Al-Zr-TUD-1 catalysts led to high initial rates,
but full conversions are not reached as also in the case of
the catalyst containing the smallest amount of aluminium,
Al-Zr-1:2. Selectivity in all cases was above 95% and major
isomers were (Æ)-isopulegol and (Æ)-neo-isopulegol.
Both types of Prins reaction, intramolecular citronellal
cyclisation and intermolecular Nopol synthesis, can be cata-
lysed by Brønsted and/or Lewis acid sites. Exclusively
Brønsted acid catalysts catalyse the Prins cyclisation via the
formation of carbenium ion intermediates (Scheme 5).[46,47]
In the Prins cyclisation of citronellal by means of an intra-
molecular reaction, a more stable tertiary carbocation is
formed. Clearly intramolecular reaction will more readily
Scheme 4. Intramolecular Prins cyclisation of citronellal in toluene at
808C.
occur than intermolecular reaction as is the case in Nopol
synthesis were the availability of the reacting alkene group
at the rather hydrophilic surface (based on the FTIR results
in Figure 5) is lower. This difference is even more enhanced
for catalysts containing both the Brønsted and Lewis acid
sites. Due to activation of both reacting groups, the alkene
2084
www.chemeurj.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 2077 – 2088

Bimetallic Mesoporous Materials
FULL PAPER
Based on these results we can conclude that synergy in
Al-Zr-TUD-1 obtained in the Prins cyclisation of citronellal
is much more complex and more subtle than can be ex-
plained by increase of Brønsted acidity due to incorporation
of a different Lewis acid. All three samples of Al-Zr-TUD-1
show synergistic properties in the Prins cyclisation of citro-
nellal, even though the proportions of Lewis acid sites to
Brønsted acid sites are different according to pyridine FT-
IR. The sample Al-Zr-1:2, with predominately Lewis acidic
character has a lower initial rate than more aluminium-rich
Al-Zr-TUD-1 catalysts, but eventually reaches full conver-
sion first. Demonstration of synergy could not be repeated
À
for other C C bond formation reactions. A possible explan-
ation could be that in many of the heterogeneously cata-
lysed reactions, the match between the strength of acid sites
and substrates plays an important role. In another study,
even though not explicitly mentioned as synergy, a similar
observation was made for the same intramolecular Prins re-
action. By varying the F/OH ratio during synthesis of nano-
scopic magnesium fluorides, unexpected catalytic properties
were obtained in the Prins cyclisation of citronellal.[51,52] In
that study it was also spectroscopically proven that the dif-
ferent combinations and variable strength of Lewis and
Brønsted acid sites are responsible for increased activity.
Another very important difference between Prins cyclisa-
tion and nopol synthesis is the difference between intramo-
lecular and intermolecular reaction. Intramolecular reac-
tions generally proceed much more rapidly and under much
milder reaction conditions than their intermolecular coun-
terparts, since the two reacting groups are already in close
proximity to one another. Indeed, this is one of the underly-
ing reasons for the great catalytic power of enzymes.[53,54]
Figure 11. Intramolecular Prins cyclisation of (Æ)-citronellal (4 mmol) in
toluene (5 g) at 808C using different TUD-1 catalysts (50 mg) with Si/M
ratio of approximately 25: Al-Zr-4.3:1, Al-Zr-1.5:1 Al-Zr-1:2, Al-TUD-1
and Zr-TUD-1.
Scheme 5. The Brønsted acid catalysed Prins cyclisation of citronellal.
and the carbonyl group, by Lewis acid sites and the presence
of Brønsted acid sites in close proximity, the rate of the re-
action is further enhanced as is the case with bimetallic Al-
Zr-TUD-1 materials in the Prins cyclisation of citronellal. It
was already proposed earlier that the desired heterogeneous
catalysts for the cyclisation of citronellal should have strong
Lewis and weak Bronsted acid sites.[48] It is believed that cit-
ronellal coordinates in an orientation favourable for ring
closure were both the oxygen of the aldehyde group, as the
electron-rich double bonds are attached to Lewis acid sites,
in this case zirconium. In the transition state, protonation of
the oxygen atom occurs through neighbouring Brønsted hy-
droxyl groups from the surface of the support.[48–50] Subse-
quently, hydrogen is abstracted from the isopropyl group
and the ring is closed to form isopulegol (Scheme 6).
Conclusion
Al-Zr-TUD-1, a three-dimensional mesoporous material
containing varying amounts of aluminium and zirconium,
was synthesised by using triethanolamine as complexing
agent. Framework incorporation of Al and Zr is evidenced
by absence of Al2O3 or ZrO2 phases, as proven by X-ray dif-
fraction, HR-TEM and XPS studies. Extensive FT-IR analy-
sis with pyridine and CO as probe molecules did not allow
us to identify synergistic properties due to incorporation of
zirconium as a Lewis acid in addition to aluminium generat-
ing both Lewis as well as Brønsted acidity.
The search for synergistic properties was therefore per-
formed with catalytic experiments. Synergy was not encoun-
tered if Al-Zr-TUD-1 catalysts were applied in Lewis acid
catalysed Meerwein–Ponndorf–Verley reduction of 4-tert-bu-
tylcyclohexanone or the intermolecular Prins nopol synthe-
sis. However, synergistic properties of Al-Zr-TUD-1 cata-
lysts were clearly evidenced in the Prins cyclisation of citro-
nellal, owing to proximity effects of reacting groups in this
intramolecular reaction. Along with the excellent properties
of TUD-1 type materials allowing framework incorporation
of different metals as well as a systematic variation of two
Scheme 6. The Brønsted/Lewis acid catalysed Prins cyclisation of citro-
nellal, a new carbon–carbon bond is formed between an alkene and an
aldehyde.
Chem. Eur. J. 2011, 17, 2077 – 2088
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2085

U. Hanefeld et al.
chemical shifts are reported with respect to AlACTHNUTGRNEG(UN NO3)3 as external stan-
dard at d=0 ppm.
metals rather than merely increasing the weight percentage
of one metal, while keeping the weight percentage of the
second constant. While we do not have yet the right spectro-
scopic technique in combination with probe molecules to
discriminate between the sites so to identify synergistic ef-
fects, catalytic reactions can pinpoint synergistic properties.
However, it is only possible for specific reactions, namely
the Prins cyclisation of citronellal. In another type of Prins
reaction (nopol synthesis), a synergistic effect could not be
observed, owing to the intermolecular nature of the reaction
and hydrophilic surface of the catalyst. In other words, syn-
ergistic effects seem to be reaction mechanism dependent.
The XPS measurements were performed with a PHI 5400 ESCA provid-
ed with a dual Mg/Al anode X-ray source, a hemispherical capacitor ana-
lyser and a 5 keV ion-gun. Powdered catalyst samples were pressed into
clean indium foil (Alfa Products, purity 99,9975%) with a thickness of
0.5 mm and subsequently placed on a flat specimen holder. The input
lens optical axis to the analyser was at a take off angle of 158 with respect
to the sample surface normal. The input lens aperture used was 3.5ꢂ
1.0 mm. All spectra were recorded with non-monochromated magnesium
radiation. The X-ray source was operated at an acceleration voltage of
13 keV and a power of 200 W. A survey spectrum was recorded between
0 and 1000 eV binding energy using pass energy of 71.95 eV and step size
of 0.25 eV. The spectra of the separate photoelectron and Si-Auger elec-
tron lines were recorded with pass energy of 35.75 eV and step size of
0.2 eV. The Zr-Auger electron line was recorded with pass energy of
89.45 eV and step size of 0.5 eV. The conditions for the spectra are sum-
marised in Table S2 of Supporting Information. The spectra were evaluat-
ed with Multipak 8.0 software (Physical electronics). Firstly, the satellite
photoelectron lines were substracted from the spectrum. Next the energy
scale was aligned adopting a value of 103.5Æ0.2 eV for the binding
energy of the Si 2p photoelectron line present in the Si-TUD-1 carrier
implying a binding energy of 532.9Æ0.2 eV for the O 1s line.[21] Then, the
background intensity was subtracted from the spectra using a Shirley
method.[55] Afterwards, the spectra were fitted with (symmetrical) mixed
Gauss–Lorentz functions by using the linear least-square method to re-
solve the chemical states of the constituting components. The peaks de-
scribing sample Si-TUD-1 were kept fixed during the deconvolution of
the Al and Zr loaded catalysts.
Experimental Section
Catalyst synthesis: Monometallic Al-TUD-1 and Zr-TUD-1 were syn-
thesised according to previous reports and were described earlier in full
detail.[17,18]
Al-Zr-TUD-1: Al-Zr-TUD-1 materials were synthesised by using trietha-
nolamine (TEA, ꢀ99.0%, Fluka) as a complexing agent in a one pot sur-
factant-free procedure based on the sol–gel technique. Al-Zr-TUD-1 ma-
terials with constant Si/
ratios were synthesised by adjusting the molar ratio of SiO2/xAl2O3/
yZrO2/tetraethylammonium hydroxide (TEAOH)/(0.5–1) TEA/(10–20)
H2O. In a typical synthesis (Al/Zr=1:3), aluminum(III) isopropoxide
ACHTUNGTRENNU(G Al+Zr) molar ratio of 25 with varying Al/Zr
A
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
Temperature-programmed desorption (TPD) of ammonia was carried out
on a Micromeritics TPR/TPD 2900 equipped with a thermal conductivity
detector (TCD). The sample (30 mg) was pre-treated at 823 K to remove
volatile components. Prior to the TPD measurements the samples were
saturated with ammonia gas at 393 K. This procedure was repeated three
times. The measurements were only started when as much as possible
physisorbed NH3 was removed. Desorption of NH3 was monitored in the
(0.51 g, 98+ %, Aldrich) and zirconium(IV) propoxide solution (0.32 g,
70 wt% in 1-propanol, Aldrich), dissolved in a 1:1 mixture of isopropanol
(HPLC grade, Fisher Chemicals, 0.013% H2O) and absolute ethanol (J.
T. Baker, 0.2% H2O), was added to tetraethyl orthosilicate (17.3 g; 98%,
Aldrich). After stirring for a few minutes, a mixture of TEA (12.5 g;
ꢀ99.0%, Fluka) and water (9.4 g) was added, followed by addition of
TEAOH (10.2 g; 35 wt% in H2O, Aldrich) under vigorous stirring. The
clear gel obtained after these steps was then aged at room temperature
for 12–24 h and dried at 988C for 12–24 h, followed by hydrothermal
treatment in a Teflon-lined autoclave at 1808C for 4–24 h and final calci-
nation in the presence of air up to 6008C with a temperature ramp of
range between 393 and 823 K at a ramp rate of 10 KminÀ1
.
Skeletal FTIR spectra were measured in the 1500–600 cmÀ1 region. FTIR
spectra of KBr diluted wafers of samples were recorded using a Perkin–
Elmer Spectrum One instrument. In total 19 scans were taken with a res-
olution of 4 cmÀ1
.
18CminÀ1. Al-Zr-TUD-1 samples with constant Si/
ACTHNUTRGNEUNG(Al+Zr) ratio of ap-
proximately 25 and varying Al/Zr ratio of 3, 1 and 0.33 were prepared
and are denoted as Al-Zr-4.3:1, Al-Zr-1.5:1 and Al-Zr-1:2, respectively,
based on ICP results.
FT-IR spectra of the OH region were measured in 3900–3000 cmÀ1
region. FTIR spectra of self-supported wafers were recorded with a
Thermo Nicolet FT-IR Nexus instrument. Self-supported wafers were
pre-treated at 5008C in three-window cells (CaF2) under a flow of He. In
Catalysts characterisation: Chemical analysis of Si, Al and Zr were per-
formed in duplet by dissolving the samples in 1% HF (48% in H2O,
99.99 + % based on metal basis, Aldrich) and 1.25% H2SO4 (99.999%,
Aldrich) solution and measuring them with inductively coupled plasma—
optical emission spectroscopy (ICP-OES) on a Perkin–Elmer Optima
3000DV instrument. The textural properties of the materials were charac-
terised by volumetric N2 physisorption at 77 K using Micromeritics
ASAP 2010 equipment. Prior to the physisorption experiment, the sam-
ples were dried overnight at 573 K (pꢁ10–2 Pa). From the nitrogen sorp-
tion isotherms, the specific surface area SBET, the pore diameter dP, BJH and
the pore volume VP,BJH were calculated.
total 128 scans were taken with resolution of 4 cmÀ1
.
A Perkin–Elmer 2000 FT-IR instrument was used to record FT-IR spec-
tra after pyridine desorption at various temperatures. Self supported cat-
alyst wafers (18–25 mg/16 mm) were pressed at 3 bar pressure applied for
10 s. The wafer was placed inside a glass cell with KBr windows and sub-
sequently evacuated to 10À6 bar followed by drying at 3008C (38C minÀ1
)
for 1 h. The cell was cooled down to room temperature and the IR spec-
trum was collected. Then the temperature of the cell was raised to 508C
and the sample was brought into contact with pyridine vapours
(3.1 mbar) for 10 min. Afterwards by applying vacuum for 30 min physi-
sorbed and loosely bound pyridine was removed. FT-IR spectra were re-
corded under vacuum under various conditions by increasing the temper-
ature (38C minÀ1) from 50 to 4508C. For each spectrum 25 scans were re-
High-resolution transmission electron microscopy (HR-TEM) was per-
formed on a Philips CM30UT electron microscope with a LaB6 filament
as the source of electrons operated at 300 kV. Samples were mounted on
Quantifoilꢁ carbon polymer supported on a copper grid by placing a few
droplets of a suspension of ground sample in ethanol on the grid, fol-
lowed by drying at ambient conditions.
corded with resolution of 4 cmÀ1
.
CO adsorption experiments were performed with Perkin–Elmer 2000
FTIR instrument. Self supporting wafers were prepared by applying
3 bar pressure for 10 s. The wafers were placed in a stainless steel IR
transmission cell (12–17 mg per 13 mm) equipped with CaF2 windows.
The cell was evacuated at 10À8 bar followed by drying at 3008C (ramp
rate 38C minÀ1) for 1 h. Subsequently the wafers were cooled down to
77 K with liquid N2. A background spectrum was taken prior to CO ex-
posure. CO was introduced as 10% CO in He (Linde gas) (0.1 mbar to
Powder X-ray diffraction (XRD) patterns were obtained on a Philips PW
1840 diffractometer equipped with a graphite monochromator using CuKa
radiation.
27Al MAS NMR experiments were performed at 9.4 T on a Varian VXR-
400 S spectrometer operating at 104.2 MHz with a pulse width of 1 ms.
4 mm Zirconia rotors with a spinning speed set to 6 kHz were used. The
2086
www.chemeurj.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 2077 – 2088

Bimetallic Mesoporous Materials
FULL PAPER
30 mbar) at 77 K. For each spectrum 25 scans were recorded with a reso-
[10] P. Kalita, N. M. Gupta, R. Kumar, J. Catal. 2007, 245, 338–347.
[11] A. M. Balu, A. Pineda, K. Yoshida, J. M. Campelo, P. L. Gai, R.
Luque, A. A. Romero, Chem. Commun. 2010, 46, 7825–7827.
[12] A. Gervasini, C. Messi, P. Cartini, A. Ponti, N. Ravasio, F. Zacche-
ria, J. Catal. 2009, 262, 224–234.
lution of 4 cmÀ1
.
Catalysts testing: The catalytic experiments were performed in dried
glassware using Schlenk techniques. The anhydrous solvents were used as
received.
[13] J. C. Jansen, Z. Shan, L. Marchese, W. Zhou, N. v. d. Puil, T. Masch-
meyer, Chem. Commun. 2001, 713–714.
For the Meerwein-Ponndorf-Verley reductions, 4-tert-butylcyclohexanone
(2 mmol; 99%, Aldrich), isopropanol (4 mL; 99.5%, Aldrich) and 1,3,5-
triisopropylbenzene (0.1 mL; 96%, Aldrich, internal standard) were
loaded in the Schlenk flask containing the activated Al-Zr-TUD-1 cata-
lyst (50 mg; 6008C, 10 h, 18CminÀ1). The reaction flask was immersed
into a preheated oil bath at 808C. Periodically samples were withdrawn
(20 mL) and analysed on a GC (Shimadzu GC-17 A gas chromatograph)
equipped with a 25 mꢂ0.32 mmꢂ0.25 mm chiral column ChrompackTM
Chirasil-Dex CB and a FID detector. The reactants and products were
identified by comparison with the retention times of authentic samples
and additionally by NMR spectroscopy as described earlier.[17] Employing
an isotherm (1208C) following retention times were recorded: 1,3,5-tri-
´
[14] S. Telalovic, A. Ramanathan, G. Mul, U. Hanefeld, J. Mater. Chem.
2010, 20, 642–658.
[15] Z. Shan, E. Gianotti, J. C. Jansen, J. A. Peters, L. Marchese, T.
Maschmeyer, Chem. Eur. J. 2001, 7, 1437–1443.
[16] R. Anand, M. S. Hamdy, U. Hanefeld, T. Maschmeyer, Catal. Lett.
2004, 95, 113–117.
´
[17] A. Ramanathan, M. C. C. Villalobos, C. Kwakernaak, S. Telalovic,
U. Hanefeld, Chem. Eur. J. 2008, 14, 961–972.
[18] R. Anand, R. Maheswari, U. Hanefeld, J. Catal. 2006, 242, 82–91.
´
[19] S. Telalovic, J. F. Ng, R. Maheswari, A. Ramanathan, G. K. Chuah,
U. Hanefeld, Chem. Commun. 2008, 38, 4631–4633.
[20] B. Rakshe, V. Ramaswamy, A. V. Ramaswamy, J. Catal. 1999, 188,
252–260.
ACHTUNGTRENNUNGisopropylbenzene (3.86 min), 4-tert-butylcyclohexanone (4.86 min), cis-4-
tert-butylcyclohexanol (5.4 min) and trans-4-tert-butylcyclohexanol
(5.8 min).
[21] NIST X-ray Photoelectron Spectroscopy Database, Version 3.5
(Web Version), C. D. Wagner, A. V. Naumkin, A. Kraut-Vass, J. W.
Allison, C. J. Powell, J. R. Rumble, Jr., 2007, http://srdata.nist.gov/
xps/.
[22] a) L. C. Allen, J. Am. Chem. Soc. 1989, 111, 9003–9014; b) A. L.
Allred, J. Inorg. Nucl. Chem. 1961, 17, 215–221.
[23] A. Infantes-Molina, J. Mꢅrida-Robles, P. Mairales-Torres, E. Finoc-
chio, G. Busca, E. Rodrꢆguez-Castellꢇn, J. L. G. Fierro, A. Jimꢅnez-
Lꢇpez, Microporous Mesoporous Mater. 2004, 75, 23–32.
[24] G. B. Rayner, D. Kang, Y. Zhang, G. Lucovsky, J. Vac. Sci. Technol.
B 2002, 20, 1748–1758.
[25] M. S. Morey, G. D. Stucky, S. Schwarz, M. Frçba, J. Phys. Chem. B
1999, 103, 2037–2041.
[26] M. A. Camblor, A. Corma, J. Perez-Pariente, J. Chem. Soc. Chem.
Commun. 1993, 557–559.
[27] J. El Haskouri, S. Cabrera, C. Guillem, J. Latorre, A. Beltran, D.
Beltran, M. Dolores Marcos, P. Amoros, Chem. Mater. 2002, 14,
5015–5022.
[28] B. Bonelli, M. F. Ribeiro, A. P. Antunes, S. Valange, Z. Gabelica, E.
Garrone, Microporous Mesoporous Mater. 2002, 54, 305–317.
[29] M. S. Holm, S. Svelle, F. Joensen, P. Beato, C. H. Christensen, S.
Bordiga, M. Bjørgen, Appl. Catal. A 2009, 356, 23–30.
[30] D. G. Poduval, J. A. R. van Veen, M. S. Rigutto, E. J. M. Hensen,
Chem. Commun. 2010, 46, 3466–3468.
In the intermolecular nopol synthesis catalysts (50 mg) were dried at
1208C under vacuum for 1 h. To the dried catalyst materials paraformal-
dehyde (2 mmol, 0.07 g; 95%, Merck) was added followed by dry toluene
(4 mL; 99.8%, Aldrich) and 1,3,5-triisopropylbenzene (0.5 mmol, 0.10 g;
96%, Aldrich) as internal standard. Finally, b-pinene (1 mmol, 0.14 g;
99%, Aldrich) was added. The reaction was started by submerging the
reaction mixture into a preheated oil bath at 808C. Samples (20 mL) were
analysed by GC (Shimadzu GC-2014 gas chromatograph) equipped with
a 50 mꢂ0.53 mmꢂ1.0 mm CP-Sil 5CB column and using a FID detector.
Employing an isotherm (1108C) following retention times were recorded:
b-pinene (2.01 min), nopol (7.24 min) and 1,3,5-triisopropylbenzene
(10.40 min).
The intramolecular Prins reaction was performed as reported earlier.[50]
The catalysts were dried at 1008C overnight. (Æ)-Citronellal (4 mmol;
ꢀ95.0%, Aldrich) and solvent (5 g), toluene (99.8%, Aldrich) or tert-bu-
tanol (ꢀ99.5%, Aldrich), were added to the catalyst (50 mg). The reac-
tions were performed at 808C or at room temperature. Samples were
withdrawn periodically and analysed by GC (Agilentꢃs HP5 column). The
1
different isomers were identified by H NMR spectroscopy.
Acknowledgements
[31] T. Barzetti, E. Selli, D. Moscotti, L. Forni, J. Chem. Soc. Faraday
Trans. 1996, 92, 1401–1407.
[32] M. Guisnet, P. Ayrault, C. Coutanceau, M. F. Alvarez, J. Datka, J.
Chem. Soc. Faraday Trans. 1997, 93, 1661–1665.
[33] W. Daniell, U. Schubert, R. Glçckler, A. Meyer, K. Noweck, H.
Knçzinger, Appl. Catal. A 2000, 196, 247–260.
S.T. thanks NWO (Nederlandse Organisatie voor Wetenschappelijk On-
derzoek) for a Mozaꢄek grant. For the cover illustration we are grateful
to James E. Busse, The University of Kansas, Center for Environmentally
Beneficial Catalysis.
[34] I. Mirsojew, S. Ernst, J. Weitkamp, H. Knçzinger, Catal. Lett. 1994,
24, 235–248.
[35] B. Bonelli, I. Bottero, N. Ballarini, S. Passeri, F. Cavani, E. Garrone,
J. Catal. 2009, 264, 15–30.
[36] H. Knçzinger, S. Huber, J. Chem. Soc. Faraday Trans. 1998, 94,
2047–2059.
[1] C. Mirodatos, D. Barthomeuf, J. Chem. Soc. Chem. Commun. 1981,
39–40.
[2] S. Li, A. Zheng, Y. Su, H. Zhang, L. Chen, J. Yang, C. Ye, F. Deng,
J. Am. Chem. Soc. 2007, 129, 11161–11171.
[3] S. Selvakumar, N. M. Gupta, A. P. Singh, Appl. Catal. A 2010, 372,
130–137.
[37] A. Zecchina, S. Bordiga, C. Lamberti, G. Spoto, L. Carnelli, C.
Otero Arean, J. Phys. Chem. 1994, 98, 9577–9582.
[38] B. Onida, L. Borello, B. Bonelli, F. Geobaldo, E. Garrone, J. Catal.
2003, 214, 191–199.
[39] A. V. Kucherov, A. N. Shigapov, A. V. Ivanov, T. N. Kucherova,
L. M. Kustov, Catal. Today 2005, 110, 330–338.
[4] G. B. Han, N.-K. Park, S. H. Yoon, T. J. Lee, K. L. Yoon, Appl.
Catal. A 2008, 337, 29–38.
[5] Y. Sekine, H. Takamatsu, S. Aramaki, K. Ichishima, M. Takada, M.
Matsukata, E. Kikuchi, Appl. Catal. A 2009, 352, 214–222.
[6] F. MariÇo, G. Baronetti, M. Laborde, N. Bion, A. Le Valant, F.
Epron, D. Duprez, Int. J. Hydrogen Energy 2008, 33, 1345–1353.
[7] A. Kale, N. Bazzanella, R. Checchetto, A. Miotello, Appl. Phys.
Lett. 2009, 94, 204103-1–3.
[40] T. N. Vu, J. van Gestel, J. P. Gilson, C. Collet, J. P. Dath, J. C.
Duchet, J. Catal. 2005, 231, 453–467.
[41] F. Wakabayashi, J. N. Kondo, K. Domen, C. Hirose, J. Phys. Chem.
1995, 99, 10573–10580.
[8] G. Carja, G. Delahay, C. Signorile, B. Coq, Chem. Commun. 2004,
1404–1405.
[42] C. Pazꢅ S. Bordiga, C. Lamberti, M. Salvalaggio, A. Zecchina, G.
Bellussi, J. Phys. Chem. B 1997, 101, 4740–4751.
[9] M. Selvaraj, A. Pandurangan, K. S. Seshadri, P. K. Sinha, K. B. Lal,
Appl. Catal. A 2003, 242, 347–364.
Chem. Eur. J. 2011, 17, 2077 – 2088
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2087

U. Hanefeld et al.
[43] D. Klomp, T. Maschmeyer, U. Hanefeld, J. A. Peters, Chem. Eur. J.
2004, 10, 2088–2093.
[50] Z. Yongzhong, N. Yuntong, S. Jaenicke, G.-K. Chuah, J. Catal. 2005,
229, 404–413.
[44] C. R. Graves, H. Zhou, C. L. Stern, S. T. Nguyen, J. Org. Chem.
2007, 72, 9121–9133.
[51] S. M. Coman, P. Patil, S. Wuttke, E. Kemnitz, Chem. Commun.
2009, 460–462.
[45] R. Mello, J. Martꢆnez-Ferrer, G. Asensio, M. A. Gonzꢈlez-NfflÇez, J.
Org. Chem. 2007, 72, 9376–9378.
[52] M. D. Gracia, A. M. Balu, J. M. Campelo, R. Luque, J. M. Marinas,
A. A. Romero, Appl. Catal. A 2009, 371, 85–91.
[46] P. Mäki-Arvela, N. Kumar, V. Nieminen, R. Sjçholm, T. Salmi, D. Y.
Murzin, J. Catal. 2004, 225, 155–169.
[47] A. M. Balu, J. M. Campelo, R. Luque. A. A. Romero, Org. Biomol.
Chem. 2010, 8, 2845–2849.
[53] T. D. H. Bugg, Introduction to Enzyme and Coenzyme Chemistry,
2nd ed., Blackwell Publishing, Oxford, 2004, pp. 33–35.
[54] S. Chandrasekhar, Res. Chem. Intermed. 2003, 29, 107–123.
[55] D. A. Shirley, Phys. Rev. B 1972, 5, 4709–4714.
[48] G. K. Chuah, S. H. Liu, S. Jaenicke, L. J. Harrison, J. Catal. 2001,
200, 352–359.
[49] F. Neatu, S. Coman, V. I. Pârvulescu, G. Poncelet, D. De Vos, P.
Jacobs, Top. Catal. 2009, 52, 1292–1300.
Received: October 8, 2010
Published online: January 21, 2011
2088
www.chemeurj.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 2077 – 2088