One-Pot Synthesis of Zeolitic Strong Solid Bases: A Family of Alkaline-Earth Metal-Containing Silicalite-1

Article abstract of DOI:10.1002/chem.201501894

Fabricating stable strong basic sites in well-preserved crystallized zeolitic frameworks still remains a difficult issue. Here, we reported a family of MFI-type metallosilicate zeolites, AeS-1 (Ae: alkaline-earth metal ions of Mg, Ca, Sr or Ba; S-1: silic

Full text of DOI:10.1002/chem.201501894

DOI: 10.1002/chem.201501894
Full Paper
&
Zeolites
One-Pot Synthesis of Zeolitic Strong Solid Bases:
A Family of Alkaline-Earth Metal-Containing Silicalite-1
Yu Zhou,[a] Yanhua Jin,[a] Meng Wang,[b] Wei Zhang,[a] Jingyan Xie,[a] Jing Gu,[a]
Haimeng Wen,[a] Jun Wang,*[a] and Luming Peng*[b]
Abstract: Fabricating stable strong basic sites in well-pre-
served crystallized zeolitic frameworks still remains a difficult
issue. Here, we reported a family of MFI-type metallosilicate
zeolites, AeS-1 (Ae: alkaline-earth metal ions of Mg, Ca, Sr or
Ba; S-1: silicalite-1) through a direct one-pot hydrothermal
method involving the acidic co-hydrolysis/condensation of
the silica precursor with the Ae salts. Step-by-step full char-
acterizations were designed and conducted for in-depth dis-
cussion of the Ae status in AeS-1. Strong basicity (H_ꢀ22.5–
26.5) was detected in AeS-1. The basicity was further con-
firmed by CO2 sorption measurements, 13C NMR spectra of
chloroform-adsorbed samples, and 1H! C and 1H! Si
cross-polarization magic-angle spinning NMR spectra of
ethyl cyanoacetate-adsorbed samples. The results of Knoeve-
nagel condensations demonstrated the excellent solid base
catalysis of AeS-1, which showed high activity, reusability,
and shape-selectivity, all of which are explained by Ae-de-
rived zeolitic intracrystalline strong basic sites.
13
29
AlO4À, in which the former is neutral whereas the latter usually
exhibits acidic properties. Traditionally, basic zeolites can be
prepared by ion exchange,[13,22,23] impregnation,[24,25] and nitri-
dation.[12,26] Adding other metal salts during the one-step hy-
drothermal synthesis of zeolites delivers highly isolated metal
ion-related active sites[27–32] and is also an efficient way to
adjust the acid/base properties.[15,19,33–35] Such synthesis allows
the advantages of preserving the zeolitic crystal framework
and avoidance of further post-treatment; therefore, it is an
energy- and time-saving process. However, previously reported
heteroatomic zeolites presented only weak basicity.[13,20,21] Alka-
line-earth (Ae) metals have a low electronegativity of approxi-
mately 0.9–1.2, benefiting strong basic sites. For example,
magnesia is the most common solid base used in industry.[11,36]
Indeed, Ae ions have been incorporated into aluminophos-
phates for adjusting acidity,[37,38] and tetrahedrally coordinated
framework magnesium species have been observed in the mi-
croporous Mg–Si–O and Al–Si–O materials derived from metal
silsesquioxanes.[39] However, introduction of Ae ions into alumi-
nosilicate or pure silica zeolites through a direct one-pot syn-
thesis has never been reported, except one patent,[40] which
was short of characterization proofs.
Introduction
Zeolites are a class of well-known environmentally benign cat-
alytic materials with large surface areas and versatile but very
stable aluminosilicate crystalline frameworks.[1–5] In the chemi-
cal industry, zeolites have extensive applications as solid acidic
catalysts with unique shape-selective properties because of
their tunable acidities plus uniform pores.[6–10] In the past sever-
al decades, much research has been devoted to the applica-
tions of zeolites in basic catalysis because of their special sub-
nanometer micropores, which provide highly dispersed basic
sites and increased selectivity in various reactions.[8,11–16] The
basicity, pore structure, and stability all affect the catalytic per-
formance of basic zeolites;[8,16–19] as a result, it is a challenge to
create highly stable and active strong basic sites in zeolite
frameworks while preserving the well-defined intracrystalline
channel structures.
Zeolite basicity originates from the negative charge of the
framework oxygen atoms.[13,20,21] The common aluminosilicate
zeolite has two primary tetrahedral structural units of SiO4 and
[a] Dr. Y. Zhou,+ Y. Jin,+ W. Zhang,+ J. Xie, Dr. J. Gu, Dr. H. Wen, Prof. J. Wang
State Key Laboratory of Materials-Oriented Chemical Engineering
College of Chemistry and Chemical Engineering
Here, we report a family of novel zeolite series, AeS-1 (Ae: al-
kaline-earth metals of Mg, Ca, Ba, or Sr), in which the alkaline-
earth metal ions are incorporated into the widely used MFI-
structured pure silica zeolite, silicalite-1 (abbreviated as S-1).[41]
Another reason to choose S-1 (rather than an aluminosilicate
zeolite) as the candidate is that this work aims to fabricate
strong zeolitic base catalysts and the Al-derived acid sites in
aluminosilicate may compromise the created Ae-related basic
sites. The synthesis is achieved in a direct one-pot hydrother-
mal system involving an unconventional acidic pretreatment
procedure for co-hydrolysis/condensation of the silica source
Nanjing Tech University (former Nanjing University of Technology)
Nanjing 210009 (P. R. China)
E-mail: junwang@njtech.edu.cn
[b] M. Wang,+ Dr. L. Peng
Key Laboratory of Mesoscopic Chemistry of MOE
College of Chemistry and Chemical Engineering
Nanjing University, Nanjing 210093 (P. R. China)
E-mail: luming@nju.edu.cn
[+] These authors contributed equally to this work
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201501894.
Chem. Eur. J. 2015, 21, 15412 – 15420
15412
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
with Ae salts. Full characteriza-
tions are performed on these
AeS-1 zeolites, particularly on
the typical series MgS-1, includ-
ing solid-state 29Si and 25Mg
NMR spectroscopy. The obtained
zeolites present superior strong
basicity (H_ꢀ22.5–26.5), and
their catalytic performances are
tested in the typical base-cata-
lyzed Knoevenagel condensa-
tion, demonstrating their high
and stable activity as well as the
special shape-selectivity.
Table 1. Textural properties and catalytic performance in Knoevenagel condensations.
Sample
Si/Mg[a]
[molmolÀ1
Unit cell parameters[b]
Surface area
[m2 gÀ1
]
Basic strength
[H_]
]
a [Š]
b [Š]
c [Š]
V [Š3]
S-1
1
136
125
71
34
25
19.7208
19.8554
19.8118
19.9894
20.0358
20.0360
20.0342
19.4624
20.0332
20.0524
20.0646
20.0971
20.1176
20.1063
20.1100
19.9619
13.0746
13.1969
13.2404
13.2772
13.3976
13.3597
13.3156
12.9194
5165.39
5254.32
5263.27
5333.84
5400.20
5381.95
5364.69
5019.27
387
389
382
380
378
371
367
368
<15.0
22.5–26.5
MgS-1(200)
MgS-1(100)
MgS-1(50)
MgS-1(30)
MgS-1(20)
MgS-1(10)
MgO/S-1
22.5–26.5
22.5–26.5
22.5–26.5
22.5–26.5
22.5–26.5
15.0–18.4
14
35
[a] Si/Mg molar ratios for the final solid products analyzed by ICP. [b] Unit cell volume was determined by
using the MDI Jade software (Jade 7 XRD Pattern Processing Software[42]).
Results and Discussion
MgS-1 zeolites: primary characterizations
Scheme 1 illustrates the synthesis of AeS-1(n) (n=Si/Ae molar
ratio in the initial synthesis gel), where the silica source tetrae-
thylorthosilicate (TEOS) and corresponding Ae (Mg, Ca, Sr, or
Ba) ions are co-hydrolyzed and condensed in a moderately
acidic environment, followed by basic aging and hydrothermal
crystallization. By varying the Mg content in the initial synthe-
sis solution, a series MgS-1(n) zeolites are synthesized (Table 1).
No crystal structure is obtained when the Mg2+ content is too
high (nꢁ10). The XRD patterns for calcined MgS-1 display
a set of featured peaks indexed to the MFI structure (Figure 1).
By contrast, only an amorphous phase is produced under the
corresponding basic co-hydrolysis conditions (Figure 1a), indi-
cating the important role of the unconventional acidic hydroly-
sis for this synthesis.
Generally, the acidic conditions for hydrolysis and condensa-
tion of silica precursors offers a lower reaction rate than
a basic one,[43] allowing more opportunities for isolation and
dispersion of metal ions and, thus, promoting the bond forma-
tion between metal ions and silanol groups.[29,33] In this way, it
Figure 1. a) XRD patterns for calcined MgS-1. b) Variation of unit-cell volume
with Si/Ae molar ratio for all AeS-1 samples. c,d) 25Mg NMR spectra for the
solution at different stages during the co-hydrolysis process of the silica and
Mg precursors at: c) pH 1, and d) pH 10. Asterisks denote artifacts.
is more likely to generate the desirable Si–O–Ae–O–Si linkage
(known as the primary building blocks for further growth of
the heteroatomic zeolite crystal) in the gel mixture, ultimately
promoting the incorporation of Ae ions into the S-1 framework.
To confirm the above speculation, the 25Mg NMR spectrum of
the initial synthetic solution (Figure 1c and d) were measured.
The sharp peak at 0 ppm ascribed to the free Mg2+
(Mg(H2O)62+) disappears after hydrolysis at pH 1 for 20 h (Fig-
ure 1c), indicating that the Mg2+ is bonded to the silica pre-
cursors. Conversely, when hydrolysis occurs at pH 10, signifi-
cant amounts of free Mg2+ can still be detected after 20 h (Fig-
ure 1d), implying that under the conventional basic conditions,
Scheme 1. Schematic diagram of the synthesis procedure for alkaline-earth
metal framework-substituted MFI zeolites.
Chem. Eur. J. 2015, 21, 15412 – 15420
www.chemeurj.org
15413
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
the Mg2+ species are unable to link with silanols efficiently to
create the primary zeolite-forming units. Moreover, the hydro-
lyzates of Mg2+ under basic conditions can further hamper the
crystallization process, generating only an amorphous structure
(Figure 1a). Therefore, the acidic hydrolysis is the key step for
successfully obtaining AeS-1 in the present strategy.
The Brunauer–Emmett–Teller (BET) surface areas of these
MgS-1 samples are approximately 380 m2 gÀ1 (Table 1) and sim-
ilar SEM images are observed compared with the parent
sample, S-1 (Figure S1 in the Supporting Information), which
features the MFI phase. More notably, the unit-cell volumes
expand for all MgS-1 samples, achieving the maximum at n=
30 (Table 1 and Figure 1b), consistent with the larger radius of
Mg2+ (0.072 nm) over Si4+ (0.040 nm). In contrast, no incre-
ment of the unit-cell volume is detected for the supported
counterpart MgO/S-1. This comparison is indicative of the in-
corporation of Mg ions into the MFI framework.[29,33]
The FTIR spectra for MgS-1 (Figure 2a) all exhibit a new
shoulder absorbance band in the wavenumber region of ap-
proximately 970–1000 cmÀ1, which is absent for S-1 and MgO/
S-1. Furthermore, all the MgS-1 products show a broad absorp-
tion band at approximately 245 nm in their UV/Vis spectra (Fig-
ure 2b); whereas, this peak is also absent in the control sam-
ples of S-1, MgO/S-1, and MgO. In addition, this UV/Vis peak is
not seen in the initial gelation stage, however, it emerges and
becomes stronger during the crystallization process (Figure S2
in the Supporting Information). This comparison implies the
gradual formation of Mg-containing S-1 during the one-pot
synthesis, which is different from the simply MgO impregnated
counterpart.
Figure 3. a,b) TEM images of MgS-1(30), and c) the simultaneously recorded
small-angle electron diffraction (SAED) pattern from the particle. d) SEM, and
e,f) elemental mapping images of Mg and Si on the Mg-S-1(30) sample.
proving the intrinsic MFI crystal structure as demonstrated by
the XRD patterns (Figure 1a). The inductively coupled plasma
mass spectrometry (ICP-MS) analysis confirms the existence of
Mg atoms in the solid products (Table 1), and elemental map-
ping for MgS-1(30) shows that Mg ions distribute uniformly
throughout the whole crystal and even in well-crystallized
pieces (Figure 3d–f).
The TEM image of MgS-1(30) displays the well-defined uni-
formly ordered microporous channels and clear electron dif-
fraction pattern of the selected area (Figure 3a–c), further
MgS-1 zeolites: further NMR analysis
29Si NMR spectroscopy has been used extensively to illustrate
the local structure of zeolites.[44,45] Figure 4 shows the 29Si
magic-angle spinning (MAS) NMR spectra for S-1, the MgS-1
1
29
series, and MgO/S-1, as well as the H! Si cross-polarization
(CP) MAS NMR spectra for MgS-1(30) and MgS-1(10). Fifteen
relatively sharp peaks between À108 and À118 ppm (right
panel of Figure 4) attributable to Q4 [(Si(OSi)4] sites appears on
S-1, reflecting good crystallinity of the highly silicious parent
material. The spectral assignment of these peaks to crystallo-
graphically inequivalent Si atoms can be found in previous
studies.[46,47] Two additional peaks emerge at À98 and
À102 ppm for all the 29Si MAS NMR spectra of the MgS-1
series. With the increase of Mg content, the intensity of the
former peaks continuously increases, whereas the latter peaks
increase at first, before reaching a maximum for the MgS-1(30)
sample, and then decrease (Figure 4). By contrast, the 29Si NMR
spectrum for MgO/S-1 is similar to that of S-1, that is, no peak
at À98 or À102 ppm is observed. Such comparison suggests
that the two signals at À98 and À102 ppm are not assigned
to some susceptibility effect caused by extra-framework-doped
Mg species and should be attributed to the variation of the in-
ternal Si skeletal environment in the presence of Mg salts
during the hydrothermal process. Generally, two possible varia-
Figure 2. a) FTIR, and b) UV/Vis spectra for MgS-1.
Chem. Eur. J. 2015, 21, 15412 – 15420
www.chemeurj.org
15414
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
tions will cause the emergence of more positive Si signals. One
is the formation of defectively crystallized zeolitic structures
such as terminal Q3 [Si(OSi)3OH] sites. Secondly, the 29Si chemi-
cal shift becomes more positive by introducing other metal
ions, for example, trivalent Al ions, into the traditional zeolite
framework because of decreased shielding.[44] To make a clearer
1
identification of the signals at À98 and À102 ppm, the H–29Si
CP MAS NMR spectra were measured for all the MgS-1 series
as well as S-1 and MgO/S-1 (Figure S3 in the Supporting Infor-
mation). Selected spectra for the MgS-1(30) and MgS-1(10)
samples are presented in Figure 4 to give a comparison with
the 29Si MAS NMR spectra. According to the principle of cross
1
polarization, the signal in the H–29Si CP MAS NMR spectrum is
attributed to the transfer of magnetization from 1H to 29Si
mediated by the dipolar interaction, which is a through-space
rather than through-bond interaction. Here, the CP MAS ex-
periment is utilized to detect the selectively produced signal
Figure 4. 29Si MAS NMR spectra for S-1, the MgS-1(n) series, and MgO/S-1 in
1
only from 29Si in close proximity to H nuclei, in other words,
29
comparison to 1H! Si CP MAS NMR spectra for the MgS-1(n) series. The
1
29Si MAS NMR spectra in the frequency of À107 to À119 ppm are zoomed in
no signal can be observed if there is no H existing near the
29Si atoms. As shown in Figure S3 (in the Supporting Informa-
on the right.
1
tion), no signal is observed in the H–29Si CP MAS NMR spectra
of S-1 and MgO/S-1, in accordance with the well-crystallized
structure of these two samples (all the silica exists in the form
of Q4 [(Si(OSi)4] as demonstrated by the 29Si MAS NMR spectra),
1
which provides no possible H atoms near the 29Si to produce
the CP signal. A single peak at À98 ppm is observed in each
MgS-1 sample and this can be attributed to terminal Q3 [Si(O-
Si)3OH] groups (Figures 4 and Figure S3 in the Supporting In-
formation). The intensity of this peak increases with more Mg
ions, consistent with the presence of more cracks on the crys-
1
tal surface. No resonance at À102 ppm is detected in the H–
29Si CP MAS NMR spectra of any of the MgS-1 samples, sug-
gesting that the signals at À102 ppm in the 29Si MAS NMR
spectra of the MgS-1 series do not arise from the defectively
crystallized zeolitic structure. Therefore, this signal is tentatively
assigned to the contribution of the framework Mg species.[43]
Moreover, with the increase of the Mg content, two variations
can be observed in the 29Si MAS NMR spectra of the MgS-1
series: 1) the peak intensity at À102 ppm increases at low Mg
content, reaching the highest value for MgS-1(30), and then
decreases; 2) the line widths of the Q4 resonances (right panel
of Figure 4) first become significantly broader with increasing
Mg content, then slightly narrower. The above variations are in
accordance with the enlargement of the unit-cell volumes (Fig-
ure 1b), and are also in agreement with the existence of frame-
work Mg ions at the lower Mg contents and the emergence of
extra-framework magnesia at higher Mg loadings.[45]
25Mg MAS NMR spectroscopy was employed for more in-
sight into the state of the Mg species in MgS-1(30). 25Mg iso-
topically labeled precursor was used to prepare a 25Mg isotope
enriched 25MgS-1(30) sample to enhance the NMR signal. Nev-
ertheless, no signal was observable, implying the large quadru-
polar coupling constants of Mg ions in the Mg-incorporated
distorted framework environment. Therefore, we designed
a successive process, in which the 25MgS-1(30) sample is cal-
cined, followed with water adsorption, and then water removal
through calcination, to monitor the variation of 25Mg signal in
Figure 5. 25Mg MAS NMR spectra for 25MgO and 25MgS-1(30) treated succes-
sively by in situ calcination, water adsorption, and finally water removal
through calcination, as well as Mg(OH)2 for reference. Line shape simulations
for Mg(OH)2 (iso=13.5, CQ=3.15 MHz, Q=0.0) and 25MgS-1(30) adsorbed
with water (iso=35, CQ=3.15 MHz, Q=0.0) are shown for comparison.
each step. The designing of above experiment was inspired by
the early 27Al NMR studies of aluminosilicate zeolites,[48,49]
which reported that the framework Al species changed from
tetrahedral to octahedral by adsorbing water and reversed
back to tetrahedral by desorbing the water, while the extra-
framework Al species gave no such variation. Therefore, the
purpose of our experiment is to detect the possible variation
of the Mg species in MgS-1(30) sample by water adsorption. As
shown in Figure 5, although no signal appears for the calcined
25MgS-1(30) sample, a wide signal ranging from approximately
30 to À230 ppm is detected after it is exposed to moisture.
The simulated spectrum verifies that this peak is associated
with a similar quadrupolar coupling constant (CQ) and asym-
Chem. Eur. J. 2015, 21, 15412 – 15420
www.chemeurj.org
15415
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
metry parameter (Q) to the octahedrally coordinated Mg spe-
cies in Mg(OH)2. However, the different isotropic (iso) chemical
shift excludes the possible formation of nano-sized Mg(OH)2
during the water adsorption process. Therefore, the signal of
the water-adsorbed 25MgS-1(30) can be interpreted as follows:
water adsorption has relaxed the much-distorted framework
Mg-containing structure (with mostly tetrahedral Mg species)
into octahedrally coordinated Mg species through the binding
of two H2O molecules, leading to a much smaller quadrupolar
coupling constant and the appearance of a 25Mg signal. The
peak disappears again after water removal, indicating that the
Mg species may return back to the distorted framework struc-
ture. The phenomenon also implies that the formed framework
Mg species are stable in contact with water.
work Mg2+-derived strong base to be potentially very impor-
tant for catalytic applications.[30,51]
CaS-1, SrS-1, and BaS-1 zeolites
Parallel to MgS-1 and following the same acidic co-hydrolysis
synthetic route shown in Scheme 1, we also synthesized and
characterized three other Ae containing S-1 species (CaS-1,
SrS-1, and BaS-1). Highly crystallized Ca-, Sr-, and Ba-containing
S-1 can be prepared and show similar structure variation to
MgS-1; for example, the expanding of the unit-cell volume and
the new peaks in the IR and UV/Vis spectra (Table S1, Figur-
es S4 and S5 in the Supporting Information). In addition, all of
these samples show the strongly basic strength of H_ꢀ22.5–
26.5. However, compared with the radius of Mg2+ (0.072 nm),
the much larger radii of Ca2+ (0.099 nm), Sr2+ (0.112 nm), and
Ba2+ (0.135 nm) make them more difficult to be incorporated
into the zeolite framework; as a result, their synthetic windows
are narrower than that of MgS-1. Actually, Ca (Sr, Ba)S-1 crystals
can only be obtained with lower Ae contents (nꢁ20 for CaS-1,
and nꢁ50 for SrS-1 of BaS-1) with the most expanded unit-
cell volume being detected at n=100 for Ca, Sr, and Ba-con-
taining samples (Figure S3b–d in the Supporting Information).
Base properties
The basicity of the obtained MgS-1 was investigated by the
Hammett indicator method, carbon dioxide sorption, and 13C
MAS NMR analysis of 13CHCl3 adsorbed over zeolites. The Ham-
mett indicator method shows that the MgS-1 series all present
strong basicity with H_ in the range of approximately 22.5–
26.5, much higher than S-1 itself (H_<15.0) or the impregnat-
ed counterpart MgO/S-1 (H_ꢀ15.0–18.4; Table 1), indicating
that the strong basicity has been generated through our one-
pot hydrothermal synthesis. The 13C MAS NMR spectra of
13CHCl3 adsorbed over MgO/S-1, S-1, and MgS-1(30) (Figure 6a)
indicate that MgO/S-1 shows the same chemical shift as S-1
whereas a signal centred at a higher frequency is observed for
MgS-1(30), further indicating the enhanced basic strength of
MgS-1.[50] The CO2 adsorption of MgS-1, S-1, and MgO/S-1 were
assessed at 1008C. No CO2 uptake was detected in the adsorp-
tion–desorption isotherms of the S-1 and MgO/S-1 samples,
suggesting that CO2 molecules are hardly adsorbed on these
two weakly basic samples under these conditions.
Catalytic studies
The catalytic activities of the AeS-1 series were tested in the
Knoevenagel condensation of benzaldehyde and ethyl cyanoa-
cetate (Scheme S1 in the Supporting Information), which is
a probe reaction typically used for evaluating basic catalysts.[52]
The target product, ethyl-2-cyano-3-phenyl acrylate, is detect-
ed exclusively with the tested catalysts. As shown in Figure 7,
the conversion of ethyl cyanoacetate over pure S-1 (30.6%) is
only slightly higher than the non-catalytic system (24.7%),
whereas AeS-1 catalysts exhibit dramatically enhanced activi-
ties (Figure 7 and Table S1 in the Supporting Information).
Over MgS-1(30), CaS-1(100), SrS-1(100), and BaS-1(100), the
conversions reach their respective maximum values of 95.2,
96.2, 91.9, and 91.6%, all of which are about twice those over
the supported counterparts, AeO/S-1. The conversion by MgS-
1(30) even exceeds the traditionally strong solid base of pure
MgO (93%, Figure 7). Furthermore, in term of TOF (turnover
frequency), MgS-1(30) (40 hÀ1, Figure 7) is much more active
than MgO (0.74 hÀ1, Figure 7); this is probably due to the non-
porosity of the latter. A five-run catalytic recycling test was ap-
plied to investigate the reusability of selected AeS-1 samples
and good reusability is observed with MgS-1(30) and CaS-
1(100) (Table S2 in the Supporting Information).
By contrast, MgS-1(30) exhibits apparent CO2 uptake of
about 21 mggÀ1 (0.477 mmolgÀ1; Figure 6b), indicating that
incorporating Mg2+ into the S-1 framework can significant im-
prove the basicity of zeolite, allowing the capture of CO2 at
high temperature. It is rational to expect this zeolitic frame-
For further comparison, pure ZSM-5 and Fe-containing het-
eroatomic ZSM-5 zeolite, FeZSM-5(100), were prepared accord-
ing to ref. [33] (the corresponding textural properties are in
Table S3 in the Supporting Information) and their catalytic per-
formances were assessed under the same conditions. The clas-
sical ZSM-5, with aluminum metal ions and is often used as
a solid acid catalyst, shows an inferior conversion of 34.9%
(Figure 7). In previous work, FeZSM-5(100) showed good activi-
ty in catalyzing the Knoevenagel condensation of benzalde-
Figure 6. a) 13C MAS NMR spectra of 13CHCl3 adsorbed over S-1, MgO/S-1,
and MgS-1(30); b) CO2 sorption isotherm of MgS-1(30) at 373 K.
Chem. Eur. J. 2015, 21, 15412 – 15420
www.chemeurj.org
15416
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
of the framework Mg species
but also excluding the possibility
formation of MgO in the pores.
As shown in Scheme 2, the neg-
atively charged oxygen atoms
bonded to the framework Ae
species contribute to the basicity
of the AeS-1 samples.
Framework Ae ions of AeS-1
zeolites possess low electronega-
tivity and are covalently bonded
to the oxygen atoms in the MFI
skeleton, which enables electron
transfer from the metal ions to
the oxygen atoms; thus, gener-
ating more negative oxygen
Figure 7. Knoevenagel condensation of benzaldehyde with ethyl cyanoacetate over various catalysts. Reaction
conditions: catalyst 0.1 g, benzaldehyde 8 mmol, ethyl cyanoacetate 8 mmol, ethanol 5 mL, 708C, 4 h. *: reaction
time: 12 h, reaction temperature: 908C.
hyde with malononitrile;[33] however, if more inert ethyl cya-
noacetate[53,54] instead of malononitrile is used as the substrate,
the conversion drops to 41.1%. It still gives a low conversion
of 45.9% even when the reaction temperature is raised from
70 to 908C and at the much elongated time of 12 h (Figure 7),
indicating that Fe ions in the MFI framework cannot provide
suitable basicity for efficiently catalyzing the Knoevenagel con-
densation of benzaldehyde and ethyl cyanoacetate. All the
comparison results strongly demonstrate the much better ac-
tivity of the AeS-1 zeolites, which is most probably associated
with the well-isolated framework Ae2+-derived strongly basic
sites.
Scheme 2. Potential basic site in MgS-1 samples.
atoms that can be the origin of the strongly basic sites. Con-
versely, alkaline-earth metal oxide (AeO) impregnated ana-
logues present weaker basicity with much inferior catalytic ac-
tivity than the AeS-1 series, suggesting that the extra-frame-
work Ae ions are unable to introduce the strongly basic sites.
It is also interesting to note that these AeS-1 catalysts can
be similarly active after full exposure to air without any in situ
activation, that is, the strongly solid basic sites are capable of
resisting possible contamination by water and CO2 in air
(Table S2 in the Supporting Information). Moreover, strong and
stable basic sites within the well-defined zeolitic intracrystalline
micropores allow us to expect a possible shape-selective effect
in the base-catalyzed reaction. Indeed, by comparing catalytic
performances of MgS-1 and pure MgO in Knoevenagel con-
densations with varied substrates (Table S4 in the Supporting
Information), it can be found that the activity of the former is
sensitive to the molecular size of the substrates, whereas the
latter is not, revealing the significant influence of the uniform
micropores and implying a reactant-controlled shape-selectivi-
ty effect on the catalytic process.
To confirm the above speculation, 1H! C and 1H! Si CP
MAS NMR studies of ethyl cyanoacetate (the substrate of the
Knoevenagel condensation) adsorbed on the zeolites were car-
13
29
ried out. For S-1 and MgS-1(30), similar 1H! C signals, but
13
1
29
very different H! Si CP signals were observed in the MAS
NMR spectra of ethyl cyanoacetate (Figure 8).
Three resonances at À98, À102, and À114 ppm can be ob-
1
29
served for MgS-1(30) in the H! Si CP MAS NMR spectrum.
The intensities of the former two peaks are comparable to
each other but stronger than the latter one. It is not surprising
to observe a strong signal for the peak at À98 ppm, which
originates from Si close to H (Si–O–H groups). The resonance
at À102 ppm, arising from Si–O–Mg, is enhanced significantly
in CP with ethyl cyanoacetate, which results from the strong
dipolar coupling effect between H in the adsorbed ethyl cya-
noacetate and Si atoms in the Si–O–Mg environment. The re-
sults indicate that the framework Mg2+ site is indeed such
a basic site that the acidic substrate tend to be adsorbed
nearby. Otherwise, inert resonance should have been observed
at À102 ppm. In contrast, the spectrum for ethyl cyanoacetate
adsorbed over S-1 exhibits very small signals, implying very
weak interactions because of the weak basicity of S-1. The
above observations prove that the framework Mg species are
responsible for the strongly basic sites.
Based on the above results, we tentatively propose basic
sites in AeS-1 samples (Scheme 2). Taking MgS-1(30) as an ex-
ample, ion exchange tests show that the obtained AeS-1 sam-
ples are capable of ion exchange, and the counter cations are
Na+ and Mg2+ for MgS-1 samples (Table S5 in the Supporting
Information). No signal attributable to the nano-sized MgO (a
wide peak centered at about 26 ppm)[55] is detected in the
25Mg MAS NMR spectra for 25MgS-1(30) (Figure 5), suggesting
that rare MgO forms in the pores after calcination. In addition,
expansion rather than shrinkage of the unit-cell parameters is
observed after calcination of the MgS-1(30) sample (Figure S6
in the Supporting Information), not only revealing the stability
Chem. Eur. J. 2015, 21, 15412 – 15420
www.chemeurj.org
15417
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
species, which are responsible for the better catalytic per-
formance. This work provides the first family of strong solid
base materials with well-defined microporous zeolitic structure,
which could be potentially very important for heterogeneous
base-catalyzed shape-selective reactions.
Experimental Section
Synthesis
All chemicals were of analytical grade and used as received. The
typical synthesis procedure of MgS-1 was as follows: First, an aque-
ous solution of hydrochloric acid (12m) was added into the aque-
ous solution of TEOS (10.0 g TEOS, 32.5 deionized water) under in-
tense stirring at pH 1.0; this was followed by addition of a precalcu-
lated amount of magnesium nitrate hexahydrate. The resultant
mixture was stirred at 208C for 20 h to obtain complete co-hydrol-
ysis of TEOS with the magnesium salt. TPABr (tetrapropylammoni-
um bromide, 1.2 g) was then introduced into the mixture with stir-
ring at room temperature. After that, aqueous NaOH solution
(12.5m) was added to induce the gelation at pH 5.0–6.0 and the
mixture was stirred for 0.5 h. Finally, the pH value of the mixture
was regulated to 10.0 by the careful addition of NaOH solution
(12.5m). After aging at room temperature for 10 h, the slurry was
transferred to a Teflon-lined stainless steel autoclave and left to
crystallize statically at 1808C for a preset time. Solid product was
separated by centrifugation, washed with deionized water, and air-
dried. The obtained as-synthesized samples were calcined at 5508C
for 5 h in an air stream to obtain the final as-calcined solids. The
products are designated MgS-1(n), where n stands for the Si/Mg
molar ratio in the initial gel. Accordingly, CaS-1(n), SrS-1(n), and
BaS-1(n) were synthesized by employing calcium nitrate tetrahy-
drate, strontium nitrate, and barium nitrate, respectively. Silicalite-1
was synthesized by similar procedures as above without adding al-
kaline-earth salts.
13
29
Figure 8. 1H! C and 1H! Si CP MAS NMR spectra of ethyl cyanoacetate
adsorbed over S-1 and MgS-1(30).
Conclusion
Alkaline-earth metal-containing MFI-structured Al-free silicalite-1
zeolites, named as AeS-1(Ae: alkaline-earth metal ions of Mg,
Ca, Sr, and Ba; S-1: slilicalite-1), were obtained through the
one-pot hydrothermal approach involving the acidic co-hydrol-
ysis/condensation of the silica source with Ae salts. Typical
samples of MgS-1 exhibit apparent enlargement of unit-cell
volume, additional IR vibrations at approximately 970–
1000 cmÀ1, and new UV/Vis absorbance at approximately
245 nm, which are all absent for the parent pure S-1 and the
impregnated analogue MgO/S-1, suggesting that variation of
the pure silica MFI framework must have occurred through the
introduction of Mg ions. Moreover, the unique 29Si MAS NMR
signal at À102 ppm for MgS-1, which is missing for S-1 and
MgO/S-1, can be assigned to the framework Mg species by
For comparison, the conventional method for the synthesis of het-
eroatomic zeolites was employed to synthesize the typical control
samples for MgS-1(100), CaS-1(100), SrS-1(100), and BaS-1(100).
TEOS and the corresponding alkaline-earth salts were co-hydro-
lyzed in basic media (using the aqueous solution of sodium hy-
droxide instead of hydrochloric acid), with the other procedures
being the same as the above acidic co-hydrolysis route. The other
control samples, AeO/S-1 (Ae=Mg, Ca, Sr, or Ba) were obtained by
incipient wetness impregnation of alkaline-earth salts onto silica-
lite-1. A calculated amount Mg(NO3)2·6H2O (0.0804 g) was dis-
solved in water (1 mL) and then mixed with S-1 powder (1 g). After
stirring for 2 h, the mixture was evaporated at 808C to remove the
majority of the water and then dried at 808C for 24 h. After that,
the solid product was calcined at 5508C for 5 h in an air stream,
giving the final MgO/S-1 sample with the MgO loading amount of
1.87 wt%. In other words, the Si/Mg ratio of MgO/S-1 is same as
MgS-1(30). Other AeO/S-1 (Ae=Ca, Sr, or Ba), with the Ae content
the same as AeS-1(100), were prepared similarly. The 25Mg isotopi-
cally labeled MgS-1(30) and MgO/S-1 were synthesized by using
the 25Mg isotopically labeled precursor 25Mg(NO3)2. Pure ZSM-5 in
sodium form and ferric-containing heteroatomic ZSM-5 (named
FeZSM-5(100)) were prepared according to ref. [33]. The commer-
cial powder MgO (content ꢂ98%) was purchased from West Long
Chemical Co., China. Highest impurity content: ignition loss (2.0%),
carbonate (1.5%), hydrotrope (0.5%), sulfate (0.2%), sodium
(0.05%), calcium (0.02%), chloride (0.01%). The nitrogen sorption
results and XRD pattern of MgO are shown in Figures S7 and S8 in
the Supporting Information.
careful analysis of the 29Si and H–29Si CP MAS NMR spectra.
1
The proposal is further supported by the 25Mg NMR analysis
for the 25Mg isotope enriched 25MgS-1(30) sample that was
subjected to calcination, water adsorption, and re-calcination
pretreatments. Therefore, though still short of direct evidence
of the detailed exact structure of the Mg species in MgS-1 zeo-
lites, the step-by-step comparative and complementary charac-
terizations of this work allow us to propose the formation of
framework Mg2+ species, alternatively, isolated Mg2+ species
covalently bonded to the framework oxygen. This does not ex-
clude the existence of extra-framework MgO species; indeed,
these are detectable when the Mg loading is high in the
MgS-1 series. Catalytic assessment indicates stable and higher
activity over MgS-1 for the Knoevenagel condensation reaction
than over the MgO-impregnated analogue and other control
catalysts. Systematic base analyses demonstrate that MgS-1 ex-
hibits strong basicity that stems from the framework Mg2+
Chem. Eur. J. 2015, 21, 15412 – 15420
www.chemeurj.org
15418
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
tact time of 4 ms and a recycle delay of 1 s. The chemical shifts of
1H, 13C, and 29Si were externally referenced to tetramethylsilane,
and the chemical shift of 25Mg was referenced to 0.1m aqueous
MgSO4 solution, both of which were set at 0.0 ppm.
Characterization
X-ray diffraction (XRD) patterns were collected on a SmartLab dif-
fractometer (Rigaku Corporation) equipped with a rotating anode
Cu source (9 kW) at 45 kV and 200 mA, from 5 to 508 with a scan
rate of 28minÀ1. The morphology of the samples was tested with
scanning electron microscopy (HITACHI S-4800). Energy-dispersive
X-ray (EDX) analysis and elemental mapping were also obtained on
this instrument with an acceleration voltage of 20 kV. Chemical
compositions of samples were analyzed by using a Thermo Fisher
iCAP 6300 inductively coupling plasma (ICP) spectrometer. BET sur-
face areas were measured at the temperature of liquid nitrogen by
using a Micromeritics ASAP2020 analyzer. Thermogravimetric (TG)
analysis was carried out with a NETZSCH 449 F3 instrument in dry
air at a heating rate of 108CminÀ1. The FTIR spectra were obtained
with a Nicolet 8700 FTIR spectrometer. Diffuse reflectance UV/Vis
spectra were measured with the Shimadzu UV2600 spectrometer.
CO2 sorption experiments were conducted on a Micromeritics
ASAP 2020m. The sample was pretreated at 573 K for 6 h in a He
atmosphere, and maintained at 373 K during the sorption process.
Catalytic assessment
The catalytic activity was assessed through the Knoevenagel con-
densation reaction by using different substrates, such as benzalde-
hyde and ethyl cyanoacetate. In a typical test, benzaldehyde
(8 mmol), ethyl cyanoacetate (8 mmol), and ethanol (5 mL) were
mixed under a nitrogen atmosphere and heated to 708C, followed
by the addition of catalyst (0.1 g). The reaction slurry was stirred
for 4 h under reflux conditions. After reaction, n-dodecane (0.85 g)
was added as the internal standard. The reaction mixture was cen-
trifuged and the liquid was analyzed by gas chromatography (Agi-
lent GC 7890B equipped with a flame-ionization detector and a ca-
pillary column, HP-5; 30 m”0.32 mm”0.25 mm). In addition, a five-
run recycling test was carried out. The catalyst was recovered by
filtration, washed with ethanol, calcined at 5508C, and then reused
for the next run.
The base strengths of the samples were measured by using
a series of Hammett indicators, including 2,4-dinitroaniline (pKa =
15.0), 4-nitroaniline (pKa =18.4), benzidine (pKa =22.5), 4-chloroani-
line (pKa =26.5), and aniline (pKa =27.0).[56] These indicators are
listed in Table S6 in the Supporting Information, together with
their colors. They were commercial reagents purified by recrystalli-
zation. Each indicator was dissolved in benzene for the preparation
of a 0.5 wt% solution. The solid sample (100 mg) was pretreated at
6008C for 2 h, and then transferred into benzene (2 mL) under a ni-
trogen atmosphere, followed with the addition of 2–3 drops of in-
dicator solution. The color of the indicator adsorbed on the solid
was observed visually. Solids with a base intensity (H_) that ex-
ceeds the pKa of the indicator will exhibit a color change.
Acknowledgements
The authors greatly thank the National Natural Science Foun-
dation of China (Nos. 21136005, 21303084, 21222302, and
21476109), the National Basic Research Program of China (No.
2013CB934800), the Jiangsu Province Science Foundation for
Youths (No. BK20130921), the Specialized Research Fund for
Doctoral Program of Higher Education (No. 20133221120002),
and the Project of Priority Academic Program Development of
Jiangsu Higher Education Institutions (PAPD).
Solid-state NMR experiments were performed on a Bruker Avance
III spectrometer at an external magnetic field of 9.4 T with a 7 mm
double-tuned MAS probe at a spinning rate of 6 kHz. 29Si MAS
1
Keywords: alkaline-earth metals · heterogeneous catalysis ·
hydrothermal synthesis · solid bases · zeolites
NMR spectra were recorded with high-power H decoupling using
a short excitation pulse of p/8 (1.1 s) with a recycle delay of 10 s
and with 5000 scans. 1H! Si cross polarization (CP) MAS NMR
29
spectra were acquired with a contact time of 1 ms, recycle delay of
0.5 s, and 20000 scans. 29Si shifts were referenced to tetramethylsi-
lane (TMS) at 0 ppm. 25Mg MAS NMR spectra for zeolites were re-
corded using a rotor-synchronized Hahn-echo pulse sequence for
spin-5/2 quadrupolar nuclei (p/6–t–p/3–t–acq.). A p/6 pulse length
of 4.2 ms and a recycle delay of 0.1 s were used. 25Mg solution NMR
was used to monitor the environment of the Mg species during
different stages of the co-hydrolysis step of MgS-1(30) as well as
direct hydrolysis at pH 10.
[1] C. S. Cundy, P. A. Cox, Chem. Rev. 2003, 103, 663–701.
[2] M. E. Davis, Nature 2002, 417, 813–821.
[3] A. Primo, H. Garcia, Chem. Soc. Rev. 2014, 43, 7548–7561.
[4] K. Na, C. Jo, J. Kim, K. Cho, J. Jung, Y. Seo, R. J. Messinger, B. F. Chmelka,
R. Ryoo, Science 2011, 333, 328–332.
[5] W. R. Gunther, Y. Wang, Y. Ji, V. K. Michaelis, S. T. Hunt, R. G. Griffin, Y.
Romµn-Leshkov, Nat. Commun. 2012, 3, 1109.
[6] A. Corma, Chem. Rev. 1995, 95, 559–614.
[7] U. Olsbye, S. Svelle, M. Bjørgen, P. Beato, T. V. W. Janssens, F. Joensen, S.
Bordiga, K. P. Lillerud, Angew. Chem. Int. Ed. 2012, 51, 5810–5831;
Angew. Chem. 2012, 124, 5910–5933.
[8] R. J. Davis, J. Catal. 2003, 216, 396–405.
[9] G. J. Kramer, R. A. Santen, C. A. Emeis, A. K. Nowak, Nature 1993, 363,
529–531.
[10] T. L. M. Maesen, R. Krishna, J. M. Baten, B. Smit, S. Calero, J. M. C. San-
chez, J. Catal. 2008, 256, 95–107.
[11] H. Hattori, Chem. Rev. 1995, 95, 537–558.
[12] F. Dogan, K. D. Hammond, G. A. Tompsett, H. Huo, W. C., Jr. Conner,
S. M. Auerbach, C. P. Grey, J. Am. Chem. Soc. 2009, 131, 11062–11079.
[13] T. C. Keller, S. Isabettini, D. Verboekend, E. G. Rodrigues, J. PØrez-Ramír-
ez, Chem. Sci. 2014, 5, 677–684.
[14] X. F. Zhang, E. S. M. Lai, R. Martin-Aranda, K. L. Yeung, Appl. Catal. A
2004, 261, 109–118.
NMR analysis of 13C-enriched CHCl3 adsorbed on zeolites was per-
formed as follows:
(10 mmolgÀ1 of CHCl3 for each sample) was introduced to the de-
A
certain amount of 13C-enriched CHCl3
hydrated zeolite sample by a vacuum line before the sample was
packed into a 7.0 mm rotor in a N2 glovebox. 1H! C CP MAS
13
NMR experiments were performed with a contact time of 3 ms and
a recycle delay of 4 s.
NMR analysis of ethyl cyanoacetate adsorbed on zeolites was car-
ried out as follows: Ethyl cyanoacetate (0.1 mmol) and ethanol
(5 mL) were mixed under a nitrogen atmosphere and heated to
708C; this was followed by addition of the catalyst (0.3 g) and stir-
ring for 4 h. Then, the adsorbate-loaded catalyst was heated to
353 K under vacuum to remove ethanol. The sample was finally
[15] Y. Ono, J. Catal. 2003, 216, 406–415.
[16] J. Weitkamp, M. Hunger, U. Rymsa, Microporous Mesoporous Mater.
2001, 48, 255–270.
1
13
packed into a 7.0 mm rotor in a N2 glovebox. H! C CP MAS and
29
1H! Si CP MAS NMR measurements were performed with a con-
[17] R. J. Davis, E. J. Doskocil, S. Bordawekar, Catal. Today 2000, 62, 241–247.
Chem. Eur. J. 2015, 21, 15412 – 15420
www.chemeurj.org
15419
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[18] M. J. Climent, A. Corma, S. Iborra, A. Velty, J. Mol. Catal. A 2002, 182–
183, 327–342.
[19] G. Busca, Chem. Rev. 2010, 110, 2217–2249.
[39] N. Maxim, P. C. M. M. Magusin, P. J. Kooyman, J. H. M. C. van Wolput,
R. A. van Santen, H. C. L. Abbenhuis, Chem. Mater. 2001, 13, 2958–2964.
[40] J. M. Garces, US Patent 4,732,747, 1988.
[41] E. M. Flanigen, J. M. Bennett, R. W. Grose, J. P. Cohen, R. L. Patton, R. M.
Kirschner, J. V. Smith, Nature 1978, 271, 512–516.
[42] N. Hedin, G. J. DeMartin, W. J. Roth, K. G. Strohmaier, S. C. Reyes, Micro-
porous Mesoporous Mater. 2008, 109, 327–334.
[43] H. P. Lin, C. Y. Mou, Acc. Chem. Res. 2002, 35, 927–935.
[44] K. J. D. MacKenzie, M. E. Smith, Multinuclear Solid-State Nuclear Magnetic
Resonance, Ch. 4, Elsevier, Oxford, 2002.
[45] C. A. Fyfe, Y. Feng, H. Grondey, G. T. Kokotailo, H. Gies, Chem. Rev. 1991,
91, 1525–1543.
[20] D. Barthomeuf, Catal. Rev. Sci. Eng. 1996, 38, 521–612.
[21] R. A. Schoonheydt, P. Geerlings, E. A. Pidko, R. A. Santen, J. Mater. Chem.
2012, 22, 18705–18717.
[22] U. D. Joshia, P. N. Joshib, S. S. Tamhankarb, V. V. Joshic, C. V. Rodeb, V. P.
Shiralkarb, Appl. Catal. A 2003, 239, 209–220.
[23] L. Martins, W. Hçlderich, D. Cardoso, J. Catal. 2008, 258, 14–24.
[24] P. E. Hathaway, M. E. Davis, J. Catal. 1989, 116, 263–279.
[25] M. D. Romero, G. Ovejero, A. Rodríguez, J. M. Gómez, I. gueda, Ind.
Eng. Chem. Res. 2004, 43, 8194–8199.
[46] C. A. Fyfe, H. Grondey, Y. Feng, G. T. Kokotailo, J. Am. Chem. Soc. 1990,
112, 8812–8820.
[47] H. Morell, K. Angermund, A. R. Lewis, D. H. Brouwer, C. A. Fyfe, H. Gies,
Chem. Mater. 2002, 14, 2192–2198.
[48] J. A. van Bokhoven, D. C. Koningsberger, P. Kunkeler, H. Bekkum, A. P. M.
Kentgens, J. Am. Chem. Soc. 2000, 122, 12842–12847.
[49] J. Jiao, S. Altwasser, W. Wang, J. Weitkamp, M. Hunger, J. Phys. Chem. B
2004, 108, 14305–14310.
[50] M. Sµnchez-Sµnchez, T. Blasco, F. Rey, Phys. Chem. Chem. Phys. 1999, 1,
4529–4535.
[51] X. B. Wang, X. F. Zhang, Y. Wang, H. O. Liu, J. S. Qiu, J. Q. Wang, W. Han,
K. L. Yeung, ACS Catal. 2011, 1, 437–445.
[52] E. Angelescu, O. D. Pavel, R. Birjega, R. Zavoianu, G. Costentin, M. Che,
Appl. Catal. A 2006, 308, 13–18.
[53] M. Boronat, M. J. Climent, A. Corma, S. Iborra, R. Montón, M. Sabater,
Chem. Eur. J. 2010, 16, 1221–1231.
[54] K. Sugahara, T. Kimura, K. Kamata, K. Yamaguchi, N. Mizuno, Chem.
Commun. 2012, 48, 8422–8424.
[26] H. K. Min, S. H. Cha, S. B. Hong, Chem. Commun. 2013, 49, 1115–1117.
[27] M. Hartmann, L. Kevan, Chem. Rev. 1999, 99, 635–663.
[28] Y. T. Meng, H. C. Genuino, C. H. Kuo, H. Huang, S. Y. Chen, L. C. Zhang, A.
Rossi, S. L. Suib, J. Am. Chem. Soc. 2013, 135, 8594–8605.
[29] Y. J. Wu, J. Wang, P. Liu, W. Zhang, J. Gu, X. J. Wang, J. Am. Chem. Soc.
2010, 132, 17989–17991.
[30] W. B. Fan, R. G. Duan, T. Yokoi, P. Wu, Y. Kubota, T. Tatsumi, J. Am. Chem.
Soc. 2008, 130, 10150–10164.
[31] Y. K. Hwang, J. S. Chang, S. E. Park, D. S. Kim, Y. U. Kwon, S. H. Jhung,
J. S. Hwang, M. S. Park, Angew. Chem. Int. Ed. 2005, 44, 556–560;
Angew. Chem. 2005, 117, 562–566.
[32] A. Corma, L. T. Nemeth, M. Renz, S. Valencia, Nature 2001, 412, 423–
425.
[33] J. Gu, Y. H. Jin, Y. Zhou, M. J. Zhang, Y. J. Wu, J. Wang, J. Mater. Chem. A
2013, 1, 2453–2460.
[34] A. Corma, R. M. Martín-Aranda, F. Sµnchez, J. Catal. 1990, 126, 192–198.
[35] S. H. Park, H. Liu, M. Kleinsorge, C. P. Grey, B. H. Toby, J. B. Parise, Chem.
Mater. 2004, 16, 2605–2614.
[36] B. Q. Xu, J. M. Wei, H. Y. Wang, K. Q. Sun, Q. M. Zhu, Catal. Today 2001,
68, 217–225.
[55] A. V. Chadwick, I. J. F. Poplett, D. T. S. Maitland, M. E. Smith, Chem. Mater.
1998, 10, 864–870.
[56] J. I. Take, N. Kikuchi, Y. Yoneda, J. Catal. 1971, 21, 164–170.
[37] S. K. Saha, S. B. Waghmode, H. Maekawa, K. Komura, Y. Kubota, Y. Sugi,
Y. Oumi, T. Sano, Microporous Mesoporous Mater. 2005, 81, 289–303.
[38] L. Wang, Y. P. Xu, B. C. Wang, S. J. Wang, J. Y. Yu, Z. J. Tian, L. W. Lin,
Chem. Eur. J. 2008, 14, 10551–10555.
Received: May 14, 2015
Published online on September 2, 2015
Chem. Eur. J. 2015, 21, 15412 – 15420
www.chemeurj.org
15420
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim