Towards mesostructured zinc imidazolate frameworks

Article abstract of DOI:10.1002/chem.201101530

The transfer of supramolecular templating to the realm of metal-organic frameworks opens up new avenues to the design of novel hierarchically structured materials. We demonstrate the first synthesis of mesostructured zinc imidazolates in the presence of the cationic surfactant cetyltrimethylammonium bromide (CTAB), which acts as a template giving rise to ordered lamellar hybrid materials. A high degree of order spanning the atomic and mesoscale was ascertained by powder X-ray diffraction, electron diffraction, as well as solid-state NMR and IR spectroscopy. The metrics of the unit cells obtained for the zinc methylimidazolate and imidazolate species are a=(11.43±0.45), b=(9.55±0.35), c=(27.19±0.34)A, and a=(10.98±0.90), b=(8.95±0.95), c=(26.33±0.34)A, respectively, assuming orthorhombic symmetry. The derived structure model is consistent with a mesolamellar structure composed of bromine-terminated zinc (methyl)imidazolate chains interleaved with motionally rigid cationic surfactant molecules in an all-trans conformation. The hybrid materials exhibit unusually high thermal stability up to 300°C, at which point CTAB is lost and evidence for a thermally induced transformation into poorly crystalline mesostructures with larger feature sizes is obtained. Treatment with ethanol effects the extraction of CTAB from the material, followed by facile transformation into pure microporous ZIF-8 nanoparticles within minutes, thus demonstrating a unique transition from a mesostructure into a microporous zinc imidazolate.

Full text of DOI:10.1002/chem.201101530

FULL PAPER
DOI: 10.1002/chem.201101530
Towards Mesostructured Zinc Imidazolate Frameworks
Sebastian C. Junggeburth,[a] Katharina Schwinghammer,[a] Kulpreet S. Virdi,[b]
Christina Scheu,[b] and Bettina V. Lotsch*[a]
Abstract: The transfer of supramolec-
ular templating to the realm of metal–
organic frameworks opens up new ave-
nues to the design of novel hierarchi-
cally structured materials. We demon-
strate the first synthesis of mesostruc-
tured zinc imidazolates in the presence
metrics of the unit cells obtained for
the zinc methylimidazolate and imida-
zolate species are a=(11.43ꢀ0.45), b=
(9.55ꢀ0.35), c=(27.19ꢀ0.34) ꢀ, and
a=(10.98ꢀ0.90), b=(8.95ꢀ0.95), c=
(26.33ꢀ0.34) ꢀ, respectively, assuming
orthorhombic symmetry. The derived
structure model is consistent with a
mesolamellar structure composed of
bromine-terminated zinc (methyl)imid-
ACHTUNGTRENNaUNG zolate chains interleaved with motion-
ally rigid cationic surfactant molecules
in an all-trans conformation. The
hybrid materials exhibit unusually high
thermal stability up to 3008C, at which
point CTAB is lost and evidence for a
thermally induced transformation into
poorly crystalline mesostructures with
larger feature sizes is obtained. Treat-
ment with ethanol effects the extrac-
tion of CTAB from the material, fol-
lowed by facile transformation into
pure microporous ZIF-8 nanoparticles
within minutes, thus demonstrating a
unique transition from a mesostructure
into a microporous zinc imidazolate.
of
cetyltrimethyl
(CTAB), which acts as
the
cationic
ACHTUNGTRENNUNGammonium
surfactant
bromide
template
a
giving rise to ordered lamellar hybrid
materials. A high degree of order span-
ning the atomic and mesoscale was as-
certained by powder X-ray diffraction,
electron diffraction, as well as solid-
state NMR and IR spectroscopy. The
Keywords: mesostructures · metal–
organic frameworks · microporous
materials · organic–inorganic hybrid
composites · zeolites
Introduction
gimes by featuring topologies on the micro-, meso-, and
macroscale, in which microporous zeolites and mesoporous
silica are amongst the most famous representatives of these
ordered porous materials.[8,9] While zeolites comprise 3D
crystalline silicate and alumosilicate frameworks with pore
sizes typically below 2 nm, periodic mesoporous silica
(PMS) are replicas of liquid-crystal-like mesophases. They
combine periodic order on the nanoscale through arrays of
mesopores composed of amorphous silica walls. Since their
discovery by Kresge and co-workers at Mobile in 1992, PMS
materials of the MCM (mobile compound of matter) type
excel by their ease of synthesis, versatility, and chemical as
well as thermal robustness.[8,9] Depending on subtle varia-
tions in the reaction conditions, such as temperature, precur-
sor ratio, and pH, hexagonal (MCM-41), cubic (MCM-48),[8]
and lamellar (MCM-50)[10] mesophases can be obtained, and
myriad silica and non-silica mesophases have been realized
to date.
In recent years, metal–organic frameworks (MOFs) have
garnered attention owing to their exceptional porosity and
hence, application potential, and continue to intrigue due to
their rational synthesis and plethora of intricate framework
topologies.[1] Although currently a major thrust in MOF
chemistry is the extension of pore sizes from the micropore
to the mesopore regime, only a limited number of MOFs
with pore sizes larger than 2 nm have been reported to
date.[2–5] Along similar lines, the design of hierarchical
MOFs with bimodal pore-size distributions featuring micro-
and mesopores would be highly desirable for applications in
which facile diffusion of molecules, coupled with molecular
specificity, is a key requirement, such as in catalysis.[6,7] As
opposed to MOFs, siliceous materials constitute an abun-
dant class of compounds that cover different pore-size re-
Given the successful track record of supramolecular tem-
plating for the design of mesostructured materials, the trans-
fer of mesophase synthesis to the realm of microporous ma-
terials in general, and metal–organic frameworks in particu-
lar, is intriguing. Among the few examples of mesostruc-
tured coordination frameworks reported so far is a Prussian
Blue analogue obtained from specifically designed cationic
templates.[11] Metallomesogens constitute a closely related
class of compounds, featuring metal-containing ionic liquids
with predominantly lamellar structures, in which metallo-
supramolecular metal–ligand arrays are sandwiched between
[a] S. C. Junggeburth, K. Schwinghammer, Prof. Dr. B. V. Lotsch
Department Chemie and Center for Nanoscience (CeNS),
Ludwig-Maximilians-Universitꢁt Mꢂnchen
Butenandtstraße 5-13, 81377 Mꢂnchen (Germany)
Fax : (+49)089-2180-77429
E-mail: bettina.lotsch@cup.uni-muenchen.de
[b] K. S. Virdi, Prof. Dr. C. Scheu
Department Chemie and Center for Nanoscience (CeNS),
Ludwig-Maximilians-Universitꢁt Mꢂnchen
Butenandtstraße 5-13, 81377 Mꢂnchen (Germany)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201101530.
Chem. Eur. J. 2012, 18, 2143 – 2152
ꢃ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2143

amphiphile layers. These organic–inorganic hybrid architec-
tures, which have been prepared, for example, from alkyl-
ACHTUNGTRENNUNGimidazolium surfactants and various metal salts, combine
liquid-crystal templating with coordination chemistry, yet
often at the expense of stable, highly condensed struc-
tures.[12,13]
emulsifiers.[22] Heating the microemulsion to reflux at 1008C
for 2 h furnished a white precipitate, which was isolated and
washed to remove excess surfactant. The ligand to metal
ratio significantly affects the product composition in the
system Zn/MeIM/CTAB, leading to ZIF-8 for a 2:1 ratio,
while a new phase, denoted mesostructured imidazolate
framework (MIF-1) in the following, starts to nucleate at
ligand to metal ratios of 1:1 and smaller. When using water
as a reaction medium instead of heptane/1-hexanol, quanti-
tative formation of ZnO is observed. If CTAC is employed
instead of CTAB, a product similar to MIF-1, denoted MIF-
3, is obtained. Reaction of zinc acetate and imidazole (IM)
at equimolar amounts under otherwise similar conditions
gives rise to a zinc imidazolate species denoted as MIF-2 in
the following.
Recently, Ryoo et al. have developed an interesting class
of hierarchically micro- and mesoporous MFI zeolite nano-
sheets stacked into pillared arrays mediated by gemini-type
surfactants.[14,15] Along similar lines, hierarchically porous sil-
icates were obtained from layered zeolites pillared by arylic
silsesquioxanes, thus generating multifunctional hybrid ma-
terials combining acidic and basic properties in the inorganic
and organic compartments of the structure, respectively.[16]
Similar biphase architectures may be expected for a sub-
class of MOFs that combines coordinative bonds with zeo-
lite-like topologies, christened zeolitic imidazolate frame-
works (ZIFs).[17–19] Owing to the formal replacement of the
bridging oxygen in zeolites by imidazolate (IM) units, the
cages and hence, pores in ZIFs are augmented versions of
the secondary building units in zeolites. As the local geome-
try, in particular the bonding angle of the Zn-IM-Zn units,
in ZIFs largely corresponds to the Si-O-Si linkage in zeo-
lites, the design of zinc imidazolate (Zn-IM) analogues of
MCM type or even hierarchically porous mesophases in the
presence of a structure-directing agent appears highly plau-
sible. The chemical diversity of imidazolate-derived ligands
containing functional groups renders zinc imidazolates in-
herently functional, as opposed to pure mesoporous silica
architectures. Whereas the latter typically have to be con-
verted into their surface-modified analogues by post-func-
tionalization,[20,21] the corresponding mesostructured zinc
imidazolates bear functional groups already homogeneously
attached to the framework.
X-ray powder diffraction: The products MIF-1 (Zn/MeIM/
CTAB), MIF-2 (Zn/IM/CTAB) and MIF-3 (Zn/MeIM/
CTAC) obtained with a ligand to metal ratio of 1:1 were
characterized by X-ray powder diffraction (XRD) (Figure 1
Considering the well-established structural analogy be-
tween zeolites and ZIFs on the one hand, and the large
family of periodic mesoporous silicates and mesolamellar
zeolites on the other hand, we would like to address the elu-
sive class of mesostructured imidazolate frameworks,
dubbed MIFs. Although these hybrid materials present bi-
modal organization rather than hierarchical porosity, they
may nevertheless lay the grounds for the future synthesis of
micro- and mesoporous metal–organic frameworks.
Figure 1. XRD powder patterns of different lamellar MIFs: a) CTAB/
MeIM (MIF-1), b) CTAC/MeIM (MIF-3), and c) CTAB/IM (MIF-2). The
first peaks correspond to the basal spacing of the lamellar mesostruc-
tures.
and Figures S1 and S2 in the Supporting Information). All
the XRD patterns show no resemblance with any of the
starting materials, nor do they correspond to the respective
ZIFs (ZIF-8 and ZIF-4 for MeIM and IM, respectively),[17]
indicating the formation of new phases and essentially total
conversion of the starting materials. Notably, all XRD data
exhibit a set of four sharp, equidistant peaks starting at low
angles, which are assigned to the (00l) series of reflections
of a lamellar mesophase. The small peak widths and the ap-
pearance of reflections up to the fourth order indicate a
rather high degree of long-range order along the stacking di-
rection and hence, layer registry.
The presumed 001 peaks (2q=4.28 for MIF-1 and 2q=
4.48 for MIF-2) correspond to lamellar structures with a
basal spacing of d001 =27.19 ꢀ for MIF-1 and d001 =26.33 ꢀ
for MIF-2, respectively. A similar scenario holds for MIF-3
(d001 =27.01 ꢀ), thus giving rise to an increasing basal spac-
Here we report on the synthesis and characterization of
the first lamellar MIF species and document their close rela-
tionship to microporous zeolitic imidazolate frameworks.
Results
Synthesis and characterization: In an attempt to study the
nucleation and growth of ZIFs using micellar nanoscale re-
actors, a zinc precursor and imidazole ligand were combined
under
cetyltrimethyl
ammonium chloride (CTAC), or other trimethylalkylammo-
nium bromide surfactants with C10 to C18 alkyl chains as
reverse
microemulsion
conditions
using
ACHTUNGTRENNUNG
ACHTUNGTRENNUNG
2144
www.chemeurj.org
ꢃ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 2143 – 2152

Microporous Materials
FULL PAPER
ing with increasing ligand and halogenide size in the order
CTAB/IM<CTAC/MeIM<CTAB/MeIM (Figure 1). This
preliminary finding suggests the incorporation of ligand and
surfactants, including counterions, into the as-obtained ma-
terials.
IR spectroscopy: The presence of surfactant in the products
is also unambiguously evidenced by IR spectroscopy
(Figure 3 and Figure S3 in the Supporting Information). The
Interestingly, the observed d00l values closely correspond
to that of crystalline CTAB[23] (P21/c, a=5.638(1), b=
7.260(2), c=52.072(7) ꢀ; b=93.78, hence d002 ꢁ26.03 ꢀ, 2q
ꢁ4.48), which is known to crystallize with an antiparallel bi-
layer arrangement of interdigitated hydrophobic chains sep-
arated by a polar headgroup layer. For comparison, proto-
typic MCM-50 silica templated by CTAB exhibits a basal-
plane spacing of around 31 ꢀ, whereas other lamellar meso-
phases obtained with CTAB or CTAC as structure-directing
agents, such as zinc phosphates, aluminum phosphates,
GeX2/CTAB superstructures (X=S, Se), metallotropic
liquid crystals, or zinc oxide, exhibit basal spacings ranging
from 17 up to 48 ꢀ.[24–29] The observed variations in basal
spacing are generally attributed either to different composi-
tions of the inorganic layer sandwiched between the surfac-
tants, or the formation of phases with different surfactant
layer packing (single vs. bilayer), or different degrees of in-
terdigitation and tilt angles of the hydrophobic surfactant
chains.[28,30] Notably, for all compounds the intensity of the
002 peaks is lower than that of all higher order 00l reflec-
tions (Figure 1), which may be indicative of “accidental ex-
tinction” resulting from a particular constellation of scatter-
ing factors for the organic and inorganic layers.[14,31]
Figure 3. IR spectra of a) MIF-1 (CTAB/MeIM), b) CTAB, c) MeIM and
d) ZIF-8.
To verify the incorporation of surfactant and formation of
layered phases templated by the respective surfactant, dif-
ferent mesophases were synthesized in the system MeIM/
CTAB using alkylammonium surfactants with different
chain lengths (C13H30NBr to C21H46NBr). As expected, the
(00l) reflections for surfactants shorter than CTAB are shift-
ed to higher angles, while those pertaining to surfactants
with longer chains are shifted to lower angles (Figure 2),
which adds evidence to the incorporation of the surfactants
into the products and similar geometric arrangements of the
different surfactants.
strong CH-stretching vibrations of CTAB (CTAC) at
2850 cmꢂ1 and 2915 cmꢂ1 (n
ACTHNUTRGENUN(G CH2)) are the dominating fea-
ture of the IR spectra of MIFs, whereas the broad n(NH)-
stretching region characteristic of solid imidazole and 2-
methylimidazole extending from 3200 down to 1800 cmꢂ1 is
essentially absent, as analogously found in ZIF-8 and ZIF-4
(Figure 3 and Figure S3 in the Supporting Information).[32]
This clearly suggests the presence of deprotonated imidazo-
late moieties in the products, although different local envi-
ronments compared to the respective ZIFs are likely as seen
by slight shifts in the band positions.
Solid-state NMR spectroscopy: The 13C MAS solid-state
NMR spectra of MIF-1, MIF-2, MeIM/IM and CTAB, as
well as that of ZIF-8/ZIF-4 are shown in Figure 4 and Fig-
ure S4 in the Supporting Information. In the following, the
NMR spectra of MIF-1 will be discussed exemplarily, while
the major findings equally hold for MIF-2. The presence of
both CTAB and MeIM is again confirmed, and it is evident
that the spectrum of both MIFs is not a mere overlay of the
spectra of their constituents. Two of the three low-field sig-
nals pertaining to MeIM (d=150.1, 148.0, 124.8 ppm for C3-
C1) in MIF-1 are deshielded with respect to the free,
proton-bearing base MeIM (d=144.1, 125.0, 115.6 ppm,
Table S2 in the Supporting Information) and show a differ-
ent splitting pattern. The strongest MeIM signals in MIF-1
correspond closely to those found in ZIF-8 (d=150.0,
Figure 2. XRD powder patterns of MIFs based on MeIM ligands ob-
tained in the presence of different surfactants with varying chain lengths
(C13H30NBr to C21H46NBr), suggesting the incorporation of the surfac-
tants and their role as templates.
Chem. Eur. J. 2012, 18, 2143 – 2152
ꢃ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2145

B. V. Lotsch et al.
its only two signals, one pertaining to the two nitrogen
atoms in MeIM (d=ꢂ169.9 ppm) and the other one to
CTAB (d=ꢂ331.3 ppm). This clearly points to the deproto-
nation of the MeIM moiety, rendering the chemical shifts of
the two nitrogen atoms almost identical. However, the
width of the MeIM nitrogen is significantly broader
(109 Hz) than that of the CTAB nitrogen (29 Hz) and also
than the MeIM signal in ZIF-8 at d=ꢂ179 ppm (30 Hz).
This finding corroborates the assumption that although the
MeIM unit does not bear a proton, the nitrogen atoms are
not equivalent. Furthermore, the MeIM signal in MIF-1 is
shifted downfield with respect to that of MeIM in ZIF-8 by
d=ꢂ9.2 ppm. It has been noted earlier that the chemical
shift of the basic nitrogen in imidazoles is very sensitive to
the local environment and subtle changes in the hydrogen
bonding strength. According to these studies, the observed
deshielding in MIF-1 would be consistent with coordination
of the imidazole moiety to Zn2+ ions, which, owing to differ-
ences in their coordination sphere, are slightly less strong
Lewis acids than those in ZIF-8.[33–35]
It has been pointed out previously that the chemical shifts
and peak widths of the different carbon atoms of the surfac-
tant are a sensitive probe of the conformation of the tem-
plate confined in inorganic hosts and host–guest interfacial
interactions.[36,37] The 13C signals of all MIFs show a down-
field shift of the inner-chain methylene and terminal carbon
of the surfactant chain (C4–C15 and C16/17, for labeling see
the scheme in Figure 4) by about 1–2 ppm with respect to
that typically found for CTAB molecules occluded in hexag-
onal mesostructures, and are consistent with those found for
lamellar mesostructures. In particular, a chemical shift of
the methylene carbons around d=31 ppm has been assigned
to a more mobile gauche-conformation found predominantly
in hexagonal mesostructures, whereas a peak at about d=
33 ppm corresponds to a more rigid all-trans conformation,
which is found in well-ordered lamellar morphologies.[36,38]
Free CTAB and CTAC with an established all-trans-confor-
mation exhibit peaks at d=32.0 and 33.6 ppm, respectively.
The chemical shifts of the respective methylene carbons in
MIFs-1/2 (Figure 4 and Figure S4 in the Supporting Informa-
tion) and MIF-3 (Figure S4 in the Supporting Information)
are found around d=32.8 ppm and 33.5 ppm, respectively,
which clearly indicates a predominant all-trans conformation
at room temperature. Furthermore, the chemical shift of the
methylene carbon C2 is significantly shifted downfield in all
MIFs compared to CTAB/CTAC (Table S2 in the Support-
ing Information). This can be taken as evidence of an inter-
action between the surfactant head groups and the inorganic
slabs sandwiched between them.[37] Restricted motion due to
an ordered, close-packed surfactant arrangement as well as
strong interfacial interactions is further confirmed by the
high cross-polarization efficiency for C2 and C1 in MIF-1
compared to the free surfactant,[37,39] as well as by the ob-
served line broadening in MIF-1 compared to CTAB
(115 Hz vs. 60 Hz).
Figure 4. 13C MAS solid-state NMR spectra of a) MIF-1 (CTAB/MeIM),
b) MeIM, c) CTAB, d) ZIF-8. Signals marked with an asterisk are spin-
ning sidebands.
125.2 ppm), which suggests complete deprotonation of the
MeIM ligand and a similar coordination environment. The
peaks at d=150.1/148.0 ppm are both tentatively assigned to
C3, yet owing to the observed splitting we infer at least two
crystallographically nonequivalent MeIM sites in MIF-1.
The above scenario is confirmed by the 15N solid-state
NMR spectrum recorded for MIF-1 (Figure 5), which exhib-
Figure 5. 15N MAS solid-state NMR spectra of a) MIF-1 (CTAB/MeIM),
b) ZIF-8, and c) CTAB. The signal marked with an asterisk is a spinning
sideband.
2146
www.chemeurj.org
ꢃ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 2143 – 2152

Microporous Materials
FULL PAPER
Elemental analysis: Whereas ZIFs based on bivalent metal
ions feature metal to ligand ratios of M2+/IMꢂ =1:2, the ra-
tional compositions of MIFs 1–3 were consistently and re-
producibly determined as Zn2+/IMꢂ/Brꢂ (or Clꢂ)/CTA+ =
1:1:2:1 (CTA+ =C19H42N+; Tables S4–S9 in the Supporting
Information). The Zn/Br ratio of 1:2 was further corroborat-
ed by energy dispersive X-ray spectroscopy (Figure S6 in
the Supporting Information). Hence, the tentative formulae
of MIF-1 and MIF-2 are [ZnACHTUNGTRNE(UNG MeIM)2/2Br2]CTA and
[Zn(IM)2/2Br2]CTA, respectively. Thus, the incorporation of
excess halogenide ions (e.g., CTAB+Brꢂ) at the expense of
ligand anions with respect to ZIF-8, the assumption of a tet-
rahedral zinc coordination environment, and electroneutrali-
ty arguments require the formation of zinc imidazolate
chains Zn(IM)2/2 rather than zinc imidazolate networks
Zn(IM)4/2 as observed in ZIFs. We assume anionic corner-
ꢂ
sharing Zn(IM)2/2Br2 chains electrostatically bonded to pos-
itively charged CTA+ ions. Similar zinc imidazolate chains,
albeit without surfactant templates, have been reported in
the literature (see below).[40,41]
Transmission electron microscopy (TEM): The information
extractable from the XRD pattern, measured in reflection
mode, was limited by the preferred orientation of the lamel-
lae perpendicular to the [001] direction. Therefore, selected
area diffraction (SAD) experiments were conducted by
TEM to gather in-plane information about the crystal struc-
ture. As has been reported for ZIFs,[42] MIFs are extremely
unstable under the electron beam and hence, imaging by
TEM was strongly hampered. Most of the areas imaged im-
plied the region to be amorphous, which can be attributed
to electron-beam induced damage of the structure. Howev-
er, in some areas the initial crystalline structure and its dis-
integration could be depicted with SAD by a rapid change
in the diffraction pattern from rectangular arranged spots to
diffused rings within seconds.
Figure 6. SAD patterns taken parallel to the c axes for MIF-1 (top) and
MIF-2 (bottom). Due to beam sensitivity only short acquisition times
were possible resulting in highly noisy patterns.
To extract information on crystallinity, phase purity, and
possible long-range order of the sample, SAD measure-
ments were therefore performed on MeIM and IM struc-
tures for very short acquisition times. These SAD patterns
unambiguously ascertain the existence of a highly crystalline
majority phase in the MIF samples studied (Figure 6). The
relatively high noise in the patterns can be attributed to the
very short acquisition times for capturing the images and
the acquisition to be undertaken as soon as the region was
exposed to the electron beam. The d values d1a and d1b for
MIF-1 and d2a and d2b for MIF-2 were determined to be
(11.43ꢀ0.45), (9.55ꢀ0.35) and (10.98ꢀ0.90), (8.95ꢀ
0.95) ꢀ, respectively (Figure 6). The tolerance values for
these measurements are considerably higher than in conven-
tional SAD experiments; however, beam sensitivity hin-
dered more accurate measurements. A value of 10.98 ꢀ has
previously been reported as the translational parameter of a
zigzag zinc imidazolate chain.[41] As such, we assume that
the found values represent the in-plane lattice parameters of
the respective unit cells (see below). If we assume the struc-
ture to be orthorhombic (see next section), the alternating
intensity in the spots in Figure 6 (top, MIF-1), which is not
observed for MIF-2, probably stems from the 200/400 reflec-
tions giving rise to high intensity and the 100/300 reflections
resulting in less intensity. This may point to systematic ex-
tinctions, which due to the thickness of the sample and dy-
namic scattering become visible in the electron diffraction
patterns with intensities considerably less than that of the
(h00): h=2n planes. The EDX spectrum from a crystalline
domain in MIF-1 is shown in Figure S6 in the Supporting In-
formation, which clearly indicates the presence of Zn and
Br.
Structure discussion: On the basis of the above spectroscop-
ic evidence and diffraction results from TEM, we put for-
ward a tentative model for the structure of MIF-2 based on
simulations using the software package “Materials Studio”
(Figure 7).[43] The structures of other MIFs may be con-
structed analogously. According to elemental analysis, we
assume that the inorganic part is made up of Zn-IM chains
rather than Zn(IM)2 nets. The metrics of the unit cell can be
extracted from XRD and TEM, respectively, yielding a=
Chem. Eur. J. 2012, 18, 2143 – 2152
ꢃ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2147

B. V. Lotsch et al.
condition 00l: l=2n. In principle, we cannot distinguish be-
tween an AAA-type stacking (00l: l=n) with a basal spac-
ing of 26.3 ꢀ and an ABAB-type stacking, with extinction
of the 001 reflection owing to a glide plane and a doubled
basal spacing of 52.6 ꢀ. An AAA-type model structure con-
structed based on the above considerations is shown in
Figure 7. Although details of the structure, such as confor-
mation and tilting of the Zn-IM chain, exact CTA+ tilt
angle, and, in particular, the relative arrangement of the
CTA+ and [ZnBr2IM]ꢂ units, cannot be predicted unambig-
uously based on the data available, the proposed arrange-
ment represents a reasonable model consistent with the ex-
perimental data, including powder XRD (Figure S7 in the
Supporting Information).
Thermal behavior: The thermal behavior was investigated in
situ by combined DTA-TG measurements (Figure 8 and Fig-
ure S8 in the Supporting Information) as well as ex situ ex-
periments in sealed glass ampoules. A slight mass loss (6%)
associated with an endothermic event is seen around 2108C
in MIF-1, followed by further decomposition in two steps at
300–3508C (mass loss 44%) and above 4208C. Mass loss in
MIF-2 is less well resolved and proceeds in three broad
steps between 210 and 4508C. The mass loss slightly above
2008C may be due to the commencing decomposition of
Figure 7. Structural model of MIF-2 (CTAB/IM) viewed along [010], con-
structed with the program Materials Studio.[43]
10.98, b=8.95, and c=26.33 ꢀ. For simplicity, we use a
primitive orthorhombic unit cell as working model, although
we cannot exclude monoclinic metrics as found in CTAB.
Since the observed basal spacing is reminiscent of that in
crystalline CTAB, an interdigitated monolayer arrangement
featuring an antiparallel stacking of surfactant molecules
rather than a tail-to-tail bilayer arrangement likely consti-
tutes the surfactant slab.[23] As stated above, the metrics of
the ab plane is consistent with the metrics of a IM-Zn-IM
zigzag-chain running along the a axis. The b parameter is
then determined by the spacing of adjacent chains, which is
influenced by the packing of the surfactant molecules ex-
tending along the stacking direction.[41] According to this
model, two [ZnBr2IM]ꢂ units may be accommodated in a
single unit cell, which requires two CTA+ molecules accord-
ing to elemental analysis and for charge compensation. For
ꢂ
the Zn Br bond length within the tetrahedral Zn environ-
ment a value of 2.3 ꢀ was adopted according to common
ꢂ
Zn Br bond lengths, and Br···H contacts in our model
(3.1 ꢀ) slightly exceed the sum of van der Waals radii for Br
(1.83 ꢀ) and H (1.10 ꢀ).[44,45] The hydrogen atoms were
added by the software and H···H distances between different
alkyl chains were set to about 2.5 ꢀ (CTAB: 2.5 ꢀ).
Assuming a basal spacing of 26.3 ꢀ obtained from XRD
and an all-trans conformation of the CTA+ ions ascertained
by solid-state NMR spectroscopy, the CTAB molecules have
to be inclined with respect to the c axis, forming a tilted
monolayer arrangement. The tilt angle with the ab plane
can be determined by adding up all space increments given
by bonds and van der Waals contacts along the c axis and
taking into account the projected length of the CTAB mole-
cule along the same direction.[46,47] Using this scheme, a tilt
angle of roughly 628 for the CTA+ molecules can be ex-
tracted (CTAB: 658). The SAD results imply a perfect regis-
try between the organic and inorganic slabs, which is rarely
observed in lamellar mesophases.
Figure 8. DTA/TG of a) MIF-1 (CTAB/MeIM), b) CTAB, c) MeIM
d) MIF-2 (CTAB/IM), e) CTAB and f) IM.
CTAB is characterized by an ABAB-type stacking and
hence, the observed 00l reflections are indexed based on the
2148
www.chemeurj.org
ꢃ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 2143 – 2152

Microporous Materials
FULL PAPER
CTAB into cetylbromide and trimethylamine and subse-
quent gradual loss of the decomposition product (boiling
point of cetylbromide: 1908C) as observed in pure CTAB
(Figure 8B).
Notably, the thermal stability of all MIFs exceeds that of
the individual components (MeIM/IM and CTAB/CTAC, re-
spectively) alone, indicating an exceptional thermal robust-
ness of the hybrid material compared to other lamellar
mesoACHTUNGTRENNUNG
phases.[26,48] Small endothermic events around 1008C
(MIF-1) and between 70 and 1008C (MIF-2) may correlate
with a reversible trans$gauche conformational phase transi-
tion due to enhanced mobility of the CTA+ chains familiar
from other lamellar mesophases. This is corroborated by ex
situ XRD studies recorded of MIF-1 heated in evacuated
glass ampules in steps of 508C up to 3208C. No changes
were detectable in the XRD pattern of as-obtained MIF-1
up to a temperature of 2908C, at which point decomposition
of the samples into a brownish resin commenced. A white
sublimate was identified by NMR spectroscopy as CTAB.
The amorphous product obtained after heating for 30 min at
3208C still contains residual CTAB and MeIM as confirmed
by IR and solid-state NMR spectroscopy (Figures S9 and
S10 in the Supporting Information). Based on the TG re-
sults and consistent with elemental analysis of the calcined
product, about 70% of CTAB is lost upon calcination up to
3208C, indicating retention of about 30% within the materi-
al (Table S10 in the Supporting Information). The XRD pat-
terns of the residue exhibit a small shoulder to the primary
beam decay at very small angles (0.6–0.88, d values of 14.6–
11.0 nm), which suggests the formation of a disordered
Figure 9. IR spectra obtained after refluxing MIF-1 in EtOH at 1008C
for different times: a) untreated, b) 5 min, c) 10 min, d) 1 h, e) 2 h, and
f) ZIF-8.
Throughout the CTAB extraction process, a white powder
was observed at all times, suggesting a heterogeneous trans-
formation of MIF-1 into ZIF-8. Figures 9 and 10 outline the
structural evolution of ZIF-8 as a function of ethanol treat-
ment time. In the sequence of IR spectra a gradual disap-
ꢂ
pearance of the CTAB C H stretching bands (2850 and
2915 cmꢂ1) and an increase of the ZIF-8 signature in the fin-
gerprint region is observed.
The formation of ZIF-8 is evidenced in the XRD spectra
by the gradual increase of the most prominent ZIF-8 peak
(011) as time progressed (Figure 10). At the same time, the
intensity of the 00l peaks of MIF-1 diminish. The onset of
ZIF-8 formation is visible even after 5 min of ethanol treat-
ment at 1008C. The fact that sharp ZIF-8 reflections are ob-
served even at short reaction times suggests that the trans-
formation proceeds by direct conversion of MIF-1 rather
than dissolution of the MIF, followed by nucleation and
growth of the new phase. Accordingly, the intensity of the
011 reflection is consistently higher than that of the 112 and
222 peaks at all times, which is different from a homogene-
mesoACHTUNGTRENNUNGstructure with a large distribution of pore sizes upon
thermal treatment (Figure S11 in the Supporting Informa-
tion). Porosity may be generated by a rearrangement of the
zinc imidazolate chains into 2D or 3D mesophases concomi-
tant with partial CTAB release. However, the observed
mesoACHTUNGTRENNUNGstructure may also be due to textural mesoporosity
originating from gradual framework destruction as CTAB
and its volatile decomposition products are released.
Conversion into ZIF-8: In order to explore the framework
stability upon removing CTAB, MIF-1 was treated with eth-
anol at 1008C. Remarkably, the obtained product turned out
to be pure ZIF-8. Hence, the reaction occurring in ethanol
can be described by Equation (1).
1
1
½ZnðMeIMÞ2=2Br2ꢃCTA ! = ZIF-8 þ = ZnBr2 þ CTAB
2
2
ð1Þ
MIF-1 was completely converted to ZIF-8 within 1 h,
while when using an ultrasonic bath at 808C the reaction
was completed within 5 min. The conversion yield is about
80–90%. However, upon treating the lamellar network with
strong acids like HBr and HCl in EtOH, it was completely
destroyed and a clear solution was obtained.[49] Treatment
with mild solvents such as CHCl3 left the structure intact.
Figure 10. XRD patterns illustrating the structural evolution of ZIF-8
after refluxing MIF-1 in EtOH at 1008C for different times: a) untreated,
b) 5 min, c) 10 min, d) 1 h, e) 2 h, and f) simulated powder pattern of
ZIF-8.
Chem. Eur. J. 2012, 18, 2143 – 2152
ꢃ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2149

B. V. Lotsch et al.
Figure 11. SEM images of MIF-1 treated with EtOH at 1008C for different times. Untreated (left), 1 h (middle), 2 h (right). Note the different length
scales.
ous nucleation and growth scenario described by Venna
et al. for the synthesis of ZIF-8 from methanol.[50] Analysis
of the 001 (MIF-1) and 011 (ZIF-8) peak areas as a function
of time reveals an exponential decay or increase, respective-
ly, corresponding to a first-order reaction kinetics (Fig-
ure S12 in the Supporting Information). The MIF-1!ZIF-8
transformation was also followed by SEM imaging after dif-
ferent ethanol treatment times (Figure 11). Unreacted MIF-
1 exhibits a characteristic platelet-like morphology consis-
tent with its lamellar structure. After extraction for 1 h, the
coexistence of material with a layer-like morphology and
particles is visible, whereas after 2 h only agglomerates of
particulate material are observed. From these images, we
infer the gradual formation of rather monodisperse, spheri-
cal ZIF-8 nanoparticles with dimensions around 100 nm
from an intermediate stage of ZIF-8 particles embedded in
a matrix of unreacted lamellar MIF-1.
tion between the Zn-IM chains and cationic surfactants,
which is consistent with anionic [Zn(IM)2/2Br2]ꢂ chains sand-
wiched between cationic CTA molecules, “glueing” them to-
gether. This could also explain the fact that even at temper-
atures as high as 3208C, about 30% of CTAB is retained in
the structure. The observation of crystalline domains in the
MIF samples by SAD is remarkable, as it implies a periodic
mesophase that is not only ordered on the mesoscale like
MCM-type materials, but in addition it exhibits translational
order on the atomic scale. In general, this type of ordered
arrangement on two different length scales is rarely found in
mesophase materials.[31] Also, the observed crystallinity im-
plies excellent layer registry between the inorganic and or-
ganic slabs, which is consistent with strong interactions at
the inorganic–surfactant interface as suggested by our
model.
Conclusion
Discussion
With the above first example of a zinc imidazolate meso-
phase, we have transferred the concept of mesostructured
silica materials to zinc imidazolates, which is rationalized by
the similar Si-O-Si and Zn-IM-Zn bonding geometries.
Thus, the structural analogy between microporous ZIFs and
zeolites has been demonstrated to also hold for mesostruc-
tured silica materials on the one hand, and mesostructured
zinc imidazolate coordination networks on the other. Com-
prehensive characterization of the obtained MIF materials
gives evidence of a lamellar hybrid structure composed of
bromine-terminated zinc imidazolate chains, which are sand-
wiched by cationic surfactant layers. The observed facile in
situ transformation into ZIF-8 nanocrystals suggests a close
topological relationship between MIFs and ZIFs, thus
boding well for a new generation of mesostructured imida-
zolate frameworks derived from different types of imidazo-
late linkers. What is more, MIFs can adopt a range of func-
tional groups attached to the imidazole linkers, demonstrat-
ed here by the use of methylimidazole, which are introduced
into the framework in a one-step reaction and homogene-
ously distributed, rendering post-modification and grafting
reactions as familiar from functionalized silica materials un-
The facile transformation of MIF-1 into ZIF-8 without prior
dissolution of the starting material suggests a close topologi-
cal relationship between the zinc imidazolate network in
MIF-1 and ZIF-8. Hence, MIFs may be considered a meta-
stable precursor for ZIFs, the latter being generated by ex-
traction of CTAB from the interlayer space. In addition, the
transformation may be considered an alternative pathway
for the synthesis of ZIF nanocrystals, which is feasible in dif-
ferent solvents and under a range of reaction conditions.
Nevertheless, the stoichiometry of MIFs found by elemental
analysis and EDX does not allow for Zn(IM)2-type net-
works with connectivities in three dimensions. 1D zinc imi-
dazolate chains terminated by bromine ions on either side
account for both the observed elemental composition and
geometric requirements. The inorganic slabs sandwiched be-
tween the surfactants show layer thicknesses around 5.4 ꢀ,
consistent with an interdigitated CTAB monolayer, which is
likewise found in CTAB and several other metallomesogens
based on, for example, the [Zn(CnH2nꢂ1IM)2Cl2] moiety.[40]
The observed exceptional thermal stability of the lamellar
MIFs is further testament to the strong electrostatic interac-
2150
www.chemeurj.org
ꢃ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 2143 – 2152

Microporous Materials
FULL PAPER
necessary. Although evidence for the transformation of
mesoACHTUNGTRENNUNGporous into microporous materials is still pending, the
Acknowledgements
We thank Dr. Markus Dçblinger for TEM measurements and Dr. Steffen
Schmidt for technical support and helpful discussions, as well as Christian
Minke for EDX, SEM, and solid-state NMR measurements. We are in-
debted to Geraldine Rꢂhl and Danica Radmanovac for practical support.
K.S.V. would like to thank Elitenetzwerk Bayern for financial support.
Financial support by the Nanosystems Initiative Munich (NIM), Center
for Nanoscience (CeNS) and Fonds der Chemischen Industrie (FCI) is
gratefully acknowledged.
grounds have been laid for further studies in this system,
which will include the synthesis of 2D and 3D zinc imidazo-
late mesophases and surfactant removal in order to liberate
mesoporous imidazolate frameworks.
Experimental Section
Cetyltrimethylammonium bromide (CTAB, C19H42NBr, Fluka 98%), ce-
tyltrimethylammonium chloride (CTAC, C19H42NCl, Molekula), decyltri-
methylammonium bromide (DTAB, C13H30NBr), tetradecyltrimethylam-
monium bromide (TTAB, C17H38NBr), dodecyltrimethylammonium bro-
mide (DTAB, C15H34NBr), trimethyloctadecylammonium bromide
(TMODA, C21H46NBr, all Aldrich, 98%), imidazole, 2-methylimidazole
(both Aldrich, 99%), 1-hexanol (Aldrich, 98%), n-heptane (99%),
chloroform (Merck, p. a.), and zinc acetate dihydrate (ZnACTHNUGTRNEUNG(OAc)2·2H2O,
Merck, 99.5%) were purchased from commercial sources and used with-
out further purification.
[1] O. M. Yaghi, G. Li, H. Li, Nature 1995, 378, 703.
[2] H.-L. Jiang, Y. Tatsu, Z.-H. Lu, Q. Xu, J. Am. Chem. Soc. 2010, 132,
5586.
[3] Y. K. Park, S. B. Choi, H. J. Nam, D.-Y. Jung, H. C. Ahn, K. Choi,
H. Furukawa, J. Kim, Chem. Commun. 2010, 46, 3086.
[4] N. Klein, I. Senkovska, K. Gedrich, U. Stoeck, A. Henschel, U. Mu-
eller, S. Kaskel, Angew. Chem. 2009, 121, 10139; Angew. Chem. Int.
Ed. 2009, 48, 9954.
[5] B. Mu, P. M. Schoenecker, K. S. Walton, J. Phys. Chem. C 2010, 114,
6464.
Syntheses: Two separate microemulsions were formed by adding 0.10m
aqueous IM/MeIM (675 mL) and 0.10m aqueous zinc acetate dihydrate
(675 mL) to two separate mixtures containing 0.05m CTAB (0.993 g),
0.50m 1-hexanol, and 0.50m n-heptane (50 mL each). The separated mi-
croemulsions were stirred vigorously for 10 min at room temperature,
subsequently combined and refluxed at 1008C for 2 h. When using MeIM
a white precipitate can be observed after 5 min of stirring, while a white
precipitate is observed when using IM after cooling to room temperature.
The white powder was isolated by centrifugation at 13 000 rpm for
10 min, and washed with cold ethanol or chloroform. The yield of the air-
dried powder was 35 mg (90%). For surfactants other than CTAB the
same concentrations were used. ZIF-8 was obtained by reaction of MeIM
and zinc nitrate in methanol according to literature protocols.[41] ZIF-4
was produced according to the procedure described by Park et al.[17]
CTAB extraction experiments with ethanol were performed by refluxing
(1008C oil-bath temperature) MIF (400 mg) in ethanol (25 mL) for up to
24 h in a round bottom flask.
[6] M. Hartmann, Angew. Chem. 2004, 116, 6004; Angew. Chem. Int.
Ed. 2004, 43, 5880.
[7] F.-S. Xiao, L. Wang, C. Yin, K. Lin, Y. Di, J. Li, R. Xu, D. S. Su, R.
Schlçgl, T. Yokoi, T. Tatsumi, Angew. Chem. 2006, 118, 3162;
Angew. Chem. Int. Ed. 2006, 45, 3090.
[8] C. T. Kresge, M. E. Leonowicz, W. J. Roth, J. C. Vartuli, J. S. Beck,
Nature 1992, 359, 710.
[9] J. S. Beck, J. C. Vartuli, W. J. Roth, M. E. Leonowicz, C. T. Kresge,
K. D. Schmitt, C. T. W. Chu, D. H. Olson, E. W. Sheppard, J. Am.
Chem. Soc. 1992, 114, 10834.
[10] G. van den Bossche, R. Sobry, F. Fontaine, J.-M. Clacens, Z. Gabeli-
ca, J. Appl. Crystallogr. 1997, 30, 1065.
[11] X. Roy, M. J. MacLachlan, Chem. Eur. J. 2009, 15, 6552.
[12] J. Y. Z. Chiou, J. N. Chen, J. S. Lei, I. J. B. Lin, J. Mater. Chem. 2006,
16, 2972.
[13] G. Schwarz, Y. Bodenthin, Z. Tomkowicz, W. Haase, T. Geue, J.
Kohlbrecher, U. Pietsch, D. G. Kurth, J. Am. Chem. Soc. 2011, 133,
547.
Physical measurements: Combined differential thermal and thermogravi-
metric analyses (DTA/TG) were carried out on a TG-DTA thermo bal-
ance (type 92–2400, Setaram). All samples were heated at a rate of
38C minꢂ1 under a helium atmosphere. Powder X-ray diffraction (XRD)
data were collected on a D8 diffractometer (Bruker) and StadiP powder
diffractometer (Stoe), both using CuKa1 radiation. Infrared (IR) spectra
in the range 650–4000 cmꢂ1 were recorded in reflection geometry using a
Spectrum BXII FT-IR spectrometer (Perkin–Elmer) equipped with a Du-
raSamplIR II diamond ATR device (Smith Detection). Determination of
the elements H, C, N, Cl, and Br was carried out by oxidative pulping
using a commercial elemental analyzer system with helium as carrier gas
(VARIO EL, Elementar Analysensysteme GmbH). The zinc content was
determined by inductive coupled plasma atomic emission spectrscopy
(ICP-AES) on a VISTA simultan spectrometer (Varian). A JSM-6500F
scanning electron microscope (SEM, Jeol) with integrated EDX detector
(Oxford type 7418) was used to obtain particle size and morphology in-
formation. The software package Materials Studio 5.5 (Accelrys) was
used for setting up a suitable structural model. 13C and 15N solid-state
cross polarization magic angle spinning nuclear magnetic resonance (CP-
MAS NMR) spectra were recorded at room temperature with a DSX
Avance 500 FT spectrometer (Bruker) equipped with a commercial
4 mm MAS NMR double-resonance probe at a magnetic field of 11.7 T,
the chemical shifts being referenced with respect to TMS (13C) and nitro-
methane (15N) as external standards. The different NMR parameters for
the single measurements are summarized in Table S1 (Supporting Infor-
mation). For transmission electron microscopy, samples were prepared by
drop coating the solutions on Cu grids with lacey carbon backing (Plano
GmbH). TEM experiments were performed on a JEOL 2011 microscope
equipped with a LaB6 thermal source, operated at 200 kV.
[14] M. Choi, K. Na, J. Kim, Y. Sakamoto, O. Terasaki, R. Ryoo, Nature
2009, 461, 246.
[15] K. Na, M. Choi, W. Park, Y. Sakamoto, O. Terasaki, R. Ryoo, J.
Am. Chem. Soc. 2010, 132, 4169.
[16] A. Corma, U. Diꢄaz, T. Garciꢄa, G. n. Sastre, A. Velty, J. Am. Chem.
Soc. 2010, 132, 15011.
[17] K. S. Park, Z. Ni, A. P. Cꢅte, J. Y. Choi, R. Huang, F. J. Uribe-
Romo, H. K. Chae, M. O’Keeffe, O. M. Yaghi, Proc. Natl. Acad. Sci.
USA 2006, 103, 10186.
[18] R. Banerjee, A. Phan, B. Wang, C. Knobler, H. Furukawa, M.
OꢄKeeffe, O. M. Yaghi, Science 2008, 319, 939.
[19] M. Sturm, F. Brandl, D. Engel, W. Hoppe, Acta Crystallogr. Sect. B
1975, 31, 2369.
[20] K. K. Tanabe, C. A. Allen, S. M. Cohen, Angew. Chem. 2010, 122,
9924; Angew. Chem. Int. Ed. 2010, 49, 9730.
[21] M. Kim, S. J. Garibay, S. M. Cohen, Inorg. Chem. 2011, 50, 729.
[22] W. J. Rieter, K. M. L. Taylor, H. An, W. Lin, W. Lin, J. Am. Chem.
Soc. 2006, 128, 9024.
[23] A. R. Campanelli, L. Scaramuzza, Acta Crystallogr. Sect. C 1986, 42,
1380.
[24] Q. Huo, D. I. Margolese, U. Ciesla, D. G. Demuth, P. Feng, T. E.
Gier, P. Sieger, A. Firouzi, B. F. Chmelka, F. Schꢂth, G. D. Stucky,
Chem. Mater. 1994, 6, 1176.
[25] Y. Wang, X. Sun, H. Li, Mater. Sci. Eng. B 2010, 167, 177.
[26] Y. Z. Khimyak, J. Klinowski, J. Chem. Soc. Faraday Trans. 1998, 94,
2241.
[27] L. Chen, P. J. Klar, W. Heimbrodt, N. Oberender, D. Kempe, M.
Frçba, Appl. Phys. Lett. 2000, 77, 3965.
Chem. Eur. J. 2012, 18, 2143 – 2152
ꢃ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2151

B. V. Lotsch et al.
[28] J. D. Martin, C. L. Keary, T. A. Thornton, M. P. Novotnak, J. W.
Knutson, J. C. W. Folmer, Nat. Mater. 2006, 5, 271.
[29] Y. Sun, P. Afanasiev, M. Vrinat, G. Coudurier, J. Mater. Chem. 2000,
10, 2320.
[41] G. A. V. Martins, P. J. Byrne, P. Allan, S. J. Teat, A. M. Z. Slawin, Y.
Li, R. E. Morris, Dalton Trans. 2010, 39, 1758.
[42] J. Cravillon, S. Mꢂnzer, S.-J. Lohmeier, A. Feldhoff, K. Huber, M.
Wiebcke, Chem. Mater. 2009, 21, 1410.
[30] Y. Tan, E. M. P. Steinmiller, K.-S. Choi, Langmuir 2005, 21, 9618.
[31] K. Lund, N. Muroyama, O. Terasaki, Microporous Mesoporous
Mater. 2010, 128, 71.
[43] Materials Studio 5.5, Accelrys Inc, San Diego.
[44] M. Mantina, A. C. Chamberlin, R. Valero, C. J. Cramer, D. G. Truh-
lar, J. Phys. Chem. A 2009, 113, 5806.
[32] E. J. Baran, E. G. Ferrer, M. C. Apella, Monatsh. Chem. 1991, 122,
21.
[45] M. Jochum, H.-G. Unruh, H. Baernighausen, J. Phys. Condens.
Matter 1994, 6, 5751.
[33] K. Bouchmella, S. G. Dutremez, B. Alonso, F. Mauri, C. Gervais,
Cryst. Growth Des. 2008, 8, 3941.
[46] C. Liang, Y. Shimizu, M. Masuda, T. Sasaki, N. Koshizaki, Chem.
Mater. 2004, 16, 963.
[34] S. Kuroki, S. Ando, I. Ando, A. Shoji, T. Ozaki, G. A. Webb, J. Mol.
Struct. 1990, 240, 19.
[35] J. A. Webb, J. E. Klijn, P. A. Hill, J. L. Bennett, N. S. Goroff, J. Org.
Chem. 2004, 69, 660.
[36] Y. Z. Khimyak, J. Klinowski, Phys. Chem. Chem. Phys. 2001, 3, 616.
[37] L. Q. Wang, G. J. Exarhos, J. Liu, Adv. Mater. 1999, 11, 1331.
[38] K. Schmidt-Rohr, H. W. Spiess, Multidimensional Solid State NMR
and Polymers, Academic Press, London, 1994.
[39] L.-Q. Wang, J. Liu, G. J. Exarhos, B. C. Bunker, Langmuir 1996, 12,
2663.
[47] M. Meyn, K. Beneke, G. Lagaly, Inorg. Chem. 1993, 32, 1209.
[48] S. Oliver, A. Kuperman, A. Lough, G. A. Ozin, Chem. Commun.
1996, 1761.
[49] M. Benjelloun, P. Van Der Voort, P. Cool, O. Collart, E. F. Vansant,
Phys. Chem. Chem. Phys. 2001, 3, 127.
[50] S. R. Venna, J. B. Jasinski, M. A. Carreon, J. Am. Chem. Soc. 2010,
132, 18030.
[51] D. Kumar, S. Varma, G. K. Dey, N. M. Gupta, Microporous Mesopo-
rous Mater. 2004, 73, 181.
Received: May 18, 2011
[40] C. K. Lee, M. J. Ling, I. J. B. Lin, Dalton Trans. 2003, 4731.
Published online: January 9, 2012
2152
www.chemeurj.org
ꢃ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 2143 – 2152