CAU-3: A new family of porous MOFs with a novel Al-based brick: [Al 2(OCH3)4(O2C-X-CO2)] (X = aryl)

Article abstract of DOI:10.1039/c2dt12005d

A new family of Al-based MOFs denoted as CAU-3 (CAU = Christian-Albrechts- Universitaet) was discovered in the solvothermal system Al ³⁺/aryldicarboxylic acid/NaOH/methanol by applying high-throughput-methods. The three compounds reported in th

Full text of DOI:10.1039/c2dt12005d

View Article Online / Journal Homepage / Table of Contents for this issue
This article is published as part of the Dalton Transactions themed issue entitled:
Coordination chemistry in the solid state
Guest Editor Russell E. Morris
Published in Issue 14, Volume 41 of Dalton Transactions
Articles in this issue include:
Communications
Highly oriented surface-growth and covalent dye labeling of mesoporous metal–organic
frameworks
Florian M. Hinterholzinger, Stefan Wuttke, Pascal Roy, Thomas Preuße, Andreas Schaate,
Peter Behrens, Adelheid Godt and Thomas Bein
Papers
Supramolecular isomers of metal–organic frameworks: the role of a new mixed donor
imidazolate-carboxylate tetradentate ligand
Victoria J. Richards, Stephen P. Argent, Adam Kewley, Alexander J. Blake, William Lewis and
Neil R. Champness
Hydrogen adsorption in the metal–organic frameworks Fe2(dobdc) and Fe2(O2)(dobdc)
Wendy L. Queen, Eric D. Bloch, Craig M. Brown, Matthew R. Hudson, Jarad A. Mason, Leslie
J. Murray, Anibal Javier Ramirez-Cuesta, Vanessa K. Peterson and Jeffrey R. Long
Visit the Dalton Transactions website for the latest cutting inorganic chemistry
www.rsc.org/publishing/journals/dt/

Dynamic Article Links
Dalton
Transactions
Cite this: Dalton Trans., 2012, 41, 4164
www.rsc.org/dalton
PAPER
CAU-3: A new family of porous MOFs with a novel Al-based brick:
[Al2(OCH3)4(O2C-X-CO2)] (X = aryl)†
Helge Reinsch, Mark Feyand, Tim Ahnfeldt and Norbert Stock*
Received 21st October 2011, Accepted 7th December 2011
DOI: 10.1039/c2dt12005d
A new family of Al-based MOFs denoted as CAU-3 (CAU = Christian-Albrechts-Universität) was
discovered in the solvothermal system Al3+/aryldicarboxylic acid/NaOH/methanol by applying high-
throughput-methods. The three compounds reported in this article [Al2(OCH3)4BDC],
[Al2(OCH3)4BDC-NH2] and[Al2(OCH3)4NDC] (BDC = 1,4-benzenedicarboxylate;
NDC = 2,6-naphtalenedicarboxylate) are all based on the same unprecedented inorganic building unit
[Al12(OCH3)24]12+, which is a dodecameric cyclic aluminium-methanolate-cluster. The material
CAU-3-NDC was found to exhibit the highest surface area as well as the highest micropore volume of all
Al-based MOFs reported until now.
on the presence and the nature of guest molecules inside the
Introduction
pores.9 Another example was reported for the Zr4+-based MOFs
called UiO-66, -67, and 68.10,11 In this case, Zr6O4(OH)4
-
12+
During the past few years, the research on highly porous MOFs
with tailored properties has become a main objective for scien-
tists in the field of porous materials.1–5 This relatively new class
of materials is built up from inorganic vertices, most often metal
ions or cationic metal-oxo-clusters, which are connected to each
other using polytopic organic linker-molecules. These organic
parts of the network usually consist of organic anions often
bearing carboxylate- or phosphonate-groups. To create pores by
separating the inorganic vertices from each another, the linker
molecule is often based on a rigid aromatic unit like benzene or
naphthalene. Once the reaction conditions for the formation of
such a hybrid compound are known, similar reaction parameters
should in principle allow the incorporation of further functiona-
lized and also larger organic molecules into the structure, while
keeping the inorganic brick unchanged. This approach towards
the synthesis of new compounds is commonly referred to as iso-
reticular synthesis.6 There are only few examples for the syn-
thesis of isoreticular families of MOFs, consisting of more than
two materials based on the same inorganic building unit.7–12 An
example is the MIL-88-series, which is based on dicarboxylate
ions but contains trimeric M(III)3-μ3O7+-clusters, where M(III)
stands for a Cr3+ or Fe3+. These compounds exhibit enormous
changes in their lattice parameters, depending not only upon the
metal ion (Fe3+, Cr3+) and the dicarboxylate-molecule, but also
clusters are twelvefold connected by linear dicarboxylate ions to
form a fcu-net, which corresponds to a fcc-packing of inorganic
building blocks. Due to their outstanding chemical stability, the
functionalized analogues of the UiO-66 (based on 1,4-benzene-
dicarboxylate) have shown to be ideal candidates for further
post-synthetic modification reactions.12
One reason for this limited number of examples is due to
the complexity of solvothermal reactions. Small changes of the
organic linker molecule have often a strong influence on the
solubility as well as the acid–base and the coordination proper-
ties. Thus, the reaction conditions have to be established and
optimized for each organic linker molecule separately. In this
context, high-throughput-methods have proven to be a highly
valuable tool for the intentional synthesis of MOFs.13,14 Apply-
ing this methodology, we were recently able to synthesize the
amino-functionalized Al-MIL-53,15 a MOF with lozenge-shaped
channels bearing NH2-groups which can be further chemically
modified, and Cr-MIL-101-NDC,16 which exhibits giant pores
with a diameter of 4.6 nm. The miniaturization of the reaction
vessels, the parallelization of the synthesis and the automated
characterization open the opportunity to screen even complex
reaction systems with large parameter-spaces. Therefore, chemi-
cal trends can be easily identified, and synthesis parameters can
be rapidly optimized with a rather low consumption of starting
materials.
Christian-Albrechts-Universität zu Kiel, Institut für Anorganische
Chemie, Max-Eyth-Str. 2, 24118 Kiel, Germany. E-mail: stock@
ac.uni-kiel.de; Fax: +49 431 8801775; Tel: +49 431 8801675
†Electronic supplementary information (ESI) available: Crystallographic
data, details of the Rietveld refinements and further XRPD data, selected
bond lengths, TG curves. CCDC reference numbers 799242–799244 for
CAU-3-NDC, CAU-3-BDC, and CAU-3-BDC-NH2, respectively. For
ESI and crystallographic data in CIF or other electronic format see DOI:
10.1039/c2dt12005d
Recently we have started a systematic high-throughput investi-
gation on the role of the solvent in the synthesis of Al-based
MOFs. In contrast to MOFs based on divalent cations, only few
Al-based MOFs have been reported in the literature since mostly
very small μm-sized crystals have been obtained, which compli-
cates the structure determination. The frameworks of these
known porous Al-MOF structures contain only six different
4164 | Dalton Trans., 2012, 41, 4164–4171
This journal is © The Royal Society of Chemistry 2012

Al–O bricks.17–23 This rather limited number of structures is in
sharp contrast to the diversity of polynuclear species that are
known from the solvolysis reactions of aluminium salts.24,25 For
this reason, the field of Al3+-based MOFs still is a promising
chemical system for the discovery of new materials. Due to the
high thermal and chemical stability of the known Al-based
MOFs, as well as to the cost-effective availability of the usually
nontoxic starting materials, these yet unknown compounds could
be of high interest for commercial applications.
current effects. The molecular modelling software used was
Materials Studio 5.0.33 Sorption experiments were performed
using a BEL JAPAN INC. Belsorpmax
.
Synthesis and high-throughput-investigations
Discovery and synthesis optimisation of CAU-3-BDC,
[Al2(OCH3)4BDC]. The compound CAU-3-BDC (1) was dis-
covered in a high throughput-experiment using AlCl3·6H2O, ter-
ephthalic acid (H2BDC), methanol and 2 M methanolic NaOH as
starting materials. After the reaction at 125 °C for 5 h, the ten-
dencies in the product formation shown in Fig. 1 could be
observed. Exact amounts of starting materials can be found in
Table S5.† The molar ratio Al3+:H2BDC was kept constant at 4,
the absolute amount of starting material was increased from row
to row and from column to column, the amount of base was
raised.
The compound was obtained from a highly diluted basic sol-
ution. Under more acidic and more concentrated conditions the
well known compounds CAU-1 and MIL-53 are formed. Fur-
thermore an unknown product of very low crystallinity was
observed. Although the material obtained under these conditions
already exhibited a remarkable porosity (apparent specific
surface area of ABET ∼1200 m2 g−1), the crystallinity of the
sample was rather low. A detailed high-throughput-investigation
(∼200 reactions) led to an improved synthesis procedure for
CAU-3-BDC and the crystallinity was improved substantially.
Therefore, the Al3+-source was varied (nitrate, chloride and per-
chlorate), as well as the molar ratios and the absolute amounts of
starting materials. We also investigated the influence of H2O on
the product formation. While the use of Al(ClO4)3·9H2O or the
addition of small amounts of water led to the formation of X-ray
amorphous products or Al2O3, we observed highest crystallinity
only for very small concentrations of H2BDC and large excess
of Al(NO3)3·9H2O and NaOH. To achieve an optimum of crys-
tallinity we also varied the heating program. The progress during
this synthesis optimization is visualized in Fig. 2.
Especially linear dicarboxylic acids have shown to be useful
for the synthesis of isoreticular compounds, not only but also
due to the commercial availability of a variety of dicarboxylic
acids with different size and functionalization. In combination
with one-dimensional chains of corner-sharing AlO6-octahedra,
several porous compounds based on the MIL-53 structure are
known, showing interesting sorption behaviour as well as high
thermal stabilities.26–30 Another example for a dicarboxylate-
based Al-MOF is CAU-1, which contains twelvefold connected
octanuclear Al-oxo-clusters.23,31 Herein we report the synthesis
and characterization of a new family of porous MOFs based on
linear aryldicarboxylate ions, which contain the novel dodeca-
meric Al-based brick [Al12(OCH3)24]12+
.
Experimental
Materials and methods
Chemicals. AlCl3·6H2O
(Riedel-de
Haen,
≥99%),
Al(NO3)3·9H2O (Merck, ≥99%), H2BDC (Aldrich, ≥98%),
H2BDC-NH2 (Fluka, ≥98%), NaOH (Baker, ≥97%), methanol
(BASF, purum), and N,N-dimethyl-formamide (BASF, tech.)
were used as purchased. 2,6-Naphthalenedicarboxylic acid was
synthesized by hydrolysis of dimethyl 2,6-naphthalenedicarbox-
ylate (Aldrich, 98%).
Methods. Most reactions were carried out using our 24-high-
throughput reactor system.32 The upscaled synthesis was per-
formed in custom made Teflon inserts in steel autoclaves with a
volume of 30 mL. The high-throughput X-ray analyses were per-
formed in transmission geometry using a STOE HT powder dif-
fractometer equipped with a xy-stage and an image plate
detector (IPDS) system (Cu-Kα1 radiation). Temperature depen-
dent X-ray powder diffraction (XRPD) data was measured on a
STOE Stadi-P diffractometer in transmission geometry equipped
with an image plate detector (IPDS) using Cu-Kα1 radiation.
High-precision X-ray powder diffraction data for the structure
solution was collected on a STOE Stadi-P powder diffractometer
equipped with a linear position sensitive detector (PSD) system
(monochromated Cu-Kα1 radiation) in transmission geometry.
XRPD data for the structure refinement was recorded on a Pana-
lytical X-pert Highscore diffractometer in reflection-geometry.
MIR spectra were recorded on an ATI Matheson Genesis spec-
trometer in the spectral range of 4000–400 cm−1 using the KBr
disk method. FT-Raman spectra were recorded on a Bruker IFS
66 FRA 106 in the range of 0–3300 cm−1 using a Nd/YAG-
Laser (1064 nm). The thermogravimetric analyses were recorded
using an NETZSCH STA 409 CD analyzer. The samples were
heated in Al2O3 crucibles at a rate of 4 K min−1 under a flow of
air (25 ml min−1). The TG data were corrected for buoyancy and
The optimized synthesis procedure of 1 in the 24 reactor
system is as follows: a mixture of Al(NO3)3·9H2O (45.2 mg,
0.120 mmol), terephthalic acid (H2BDC; 2.5 mg, 0.015 mmol)
and a solution of NaOH in methanol (2 M, 60 μL, 0.120 mmol)
was suspended in methanol (1.340 mL). The reactor was heated
up to 125 °C in 12 h. The temperature was held for 5 h and the
reactor was allowed to cool down to room temperature in 4 h.
Fig. 1 Results of the high-throughput investigation for the discovery of
CAU-3-BDC.
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 4164–4171 | 4165

For the up-scaling of the reaction we used again our 30 mL
custom-made autoclave with a Teflon insert. For this synthesis,
Al(NO3)3·9H2O (540 mg, 1.44 mmol), H2BDC-NH2 (40 mg,
0.220 mmol) and a solution of NaOH in methanol (2 M, 720 μL,
1.44 mmol) were suspended in methanol (16.0 mL). The reactor
was heated up to 125 °C in 12 h. The temperature was held for
3 h and the reactor was allowed to cool down to room tempera-
ture in 1 h.
After the filtration, a yellow microcrystalline product was
obtained. Thermogravimetric measurements revealed the pres-
ence of an X-ray amorphous byproduct, which was removed by
applying the same procedure as for CAU-3-BDC. The crystalli-
nity of CAU-3-BDC-NH2 decreased slightly during this treat-
ment (Fig. S2†). Therefore, XRPD data of the as-synthesized
product was used for the structure refinement while all other
analytical data was measured for the activated material. Elemen-
tal analysis on an activated sample (160 °C/12 h/0.1 mbar)
stored under ambient conditions: found: C: 31.02%, H: 3.81%,
N: 1.96%; These values differ from the assumed formula [Al2
(OCH3)4(O2C-C6H3NH2-CO2)]·3.6H2O which are calculated to
be: C: 34.1%, H: 5.7%, N: 3.3%. After dissolution in
D2O/NaOD we observed in the NMR-spectrum, that CAU-3-
NH2 not only contains aminoterephthalate ions, but also tereph-
thalate anions and N-methylated aminoterephthalate ions. We
attribute the discrepancy between the experimental values and
the ideal formula to this in situ conversion of the linker and
to small amounts of X-ray amorphous Al-species (see also
TG-measurement). Further investigations of this phenomenon
are in progress.
Fig. 2 Improvement of the crystallinity of the products during the HT-
assisted synthesis optimization.
Scale-up of the reaction was performed in a 30 mL custom-
made autoclave with Teflon insert. For this synthesis, the reaction
parameters were stepwise adjusted to the larger reactor.
Al(NO3)3·9H2O (540 mg, 1.44 mmol), H2BDC (40 mg,
0.240 mmol) and a solution of NaOH in methanol (2 M, 720 μL,
1.44 mmol) were suspended in methanol (16.0 mL). The reactor
was heated up to 125 °C in 12 h. The temperature was held for
3 h and the reactor was allowed to cool down to room tempera-
ture in 1 h.
After the filtration, a white microcrystalline powder was
obtained. Thermogravimetric analysis and TEM-images
revealed, that a large amount of an X-ray amorphous byproduct
was formed. To remove this byproduct, 100 mg of the reaction
product were treated with 10 mL of DMF in a microwave oven
(Biotage Initiator) at 150 °C for 1 h under stirring. To remove
the DMF, the filtrated solid was treated with 10 mL methanol
and heated up in the microwave oven to 100 °C for 1 h.
The product was dried at room temperature in air, and
further activated under vacuum for the sorption experiments
(160 °C/12 h/0.1 mbar). Since the crystallinity slightly decreases
during this process (Fig. S1), the structure solution and refine-
ment were performed using the XRPD measurements of the “as
synthesized” compound. All other analytical data was measured
for the activated product stored under ambient conditions prior
to the measurements.
Discovery and synthesis optimisation of CAU-3-NDC,
[Al2(OCH3)4NDC]. The synthesis of CAU-3-NDC (3) needed
much more effort, since it is highly sensitive to every single para-
meter during the synthesis, and every small change leads to the
formation of crystalline byproducts, whose structures and com-
positions are a subject of current research. Without the use of
high-throughput-methods, optimization of the synthesis con-
ditions (∼800 reactions) would have been hardly possible. Due
to the large number of reactions that were performed, a detailed
description of the HT-investigations is not given. The varied par-
ameters comprise the Al3+-source (nitrate, chloride and perchlor-
ate) and the molar ratios and absolute amounts of starting
materials. Due to the sensitivity of the reaction towards the
thermal process, the chemical composition of the starting
mixture had to be screened using different heating programs.
This finally led to the optimized synthesis of 3 in the 24 reactor
system, which is as follows: a mixture of AlCl3·6H2O (29.3 mg,
0.12 mmol), 2,6-naphtalenedicarboxylic acid (H2NDC; 7.5 mg,
0.03 mmol) and a 2 M solution of NaOH (52 μL, 0.1 mmol) in
methanol was suspended in methanol (448 μL) and heated to
130 °C in 1 h. The temperature was kept for 4 h and the reactor
was allowed to cool down in 1 h. The white, as-synthesized
product contains traces of sodium chloride (see also structure
refinement) and X-ray amorphous byproducts like residual linker
molecules in the pores. EDX measurements showed molar ratios
of Al: Cl ranging from 2.5 to 3. IR-spectroscopy proved the pres-
ence of residual naphthalene dicarboxylic acid. Elemental analy-
sis on a thermally activated sample (160 °C/12 h/0.1 mbar)
stored under ambient conditions prior to the measurement:
Elemental analysis on an activated sample: found: C: 35.43%,
H: 4.07%. Calculated values, based on the deduced formula
[Al2(OCH3)4(O2CC6H4CO2)]·3.6H2O: C: 35.4%, H: 5.7%.
Discovery and synthesis optimisation of CAU-3-BDC-NH2,
[Al2(OCH3)4BDC-NH2]. Starting from the optimized reaction
conditions for 1, the synthesis of the amino-functionalized ana-
logue CAU-3-BDC-NH2 (2) was attempted. Surprisingly, the
reaction conditions are very similar. For the optimized synthesis
of 2 in the multiclave, a mixture of Al(NO3)3·9H2O (52 mg,
0.139 mmol), 2-aminoterephthalic acid (H2BDC-NH2, 2.5 mg,
0.014 mmol) and a 2 M solution of NaOH in methanol (69 μL,
0.138 mmol) were suspended in methanol (1.330 mL). The
reactor was heated up to 125 °C in 8 h. The temperature was
held for 6 h and the reactor was allowed to cool down to room
temperature in 1 h.
4166 | Dalton Trans., 2012, 41, 4164–4171
This journal is © The Royal Society of Chemistry 2012

found: C: 47.97%, H: 3.81%. Calculated values, based on the
ideal formula [Al2(OCH3)4(O2CC10H6CO2)]: C: 48.9%, H:
4.6%. Although these values are in quite good agreement, the
described material still contains the mentioned byproducts.
All measurements were performed with the as-synthesized
microcrystalline product. The only activation step for the sorp-
tion measurements was heating under vacuum, which hardly
influences the crystallinity (Fig. S3†). Several attempts were
made, to perform a scale-up of the synthesis and to further acti-
vate the as-synthesized compound. The products of the synthesis
in larger reactors exhibit a much lower specific surface area,
although, based on the XRPD measurements, no differences
were observed. Further activation steps were attempted in several
solvents, but the raw material decomposes in water, DMF and
even ethanol. The solvent treatment in methanol led only to a
reduction of the amount of chloride ions.
The absolute amounts synthesized are in all three cases quite
low. In the case of CAU-3-BDC and CAU-3-BDC-NH2, we
usually obtain ∼25 mg of fully activated sample from one up-
scaled reaction. One reason is the low overall concentration of
the linker molecule that is necessary to obtain the title com-
pounds. The other reason is the elaborate activation procedure to
remove the X-ray amorphous byproducts. In the case of CAU-3-
NDC, around 8 mg are obtained from one HT-reaction.
by 1 and 0.5%. The results of the final Rietveld refinement are
shown in Fig. S5.† The C–C distances, C–O distances of the
methoxy groups as well as the C–N distances were restrained
and an overall temperature factor was used. The refinement led
to good structure indicators of RBragg = 0.59%, Rwp = 4.17% and
GoF = 1.71. Selected bond lengths are given in Table S2.†
The indexing and lattice parameter refinement of the exper-
imental powder pattern of 3 was carried out with the Stoe
WinxPow software package.36 The hexagonal cell parameters a
= 23.1533(5) Å and c = 40.4120(9) Å were obtained. For the
construction of a model of CAU-3-NDC, we started from the
structure of CAU-3-BDC. The space group was converted to P1
and the terephthalate ions were replaced by naphthalene dicar-
boxylate ions after adjusting the cell parameters to the ones
obtained from the indexing process, using Materials Studio 5.0.
The model was submitted to a full energy minimization without
an optimization of the unit cell constants with the universal force
field (UFF) implemented in the software. For this simulation, the
aluminum atoms were replaced by iron due to their very similar
ionic radii (0.67 vs. 0.69) and the missing parameters for octa-
hedrally coordinated aluminum in the parameter set. Van der
Waals interactions were represented by a classical 12-6 Lennard
Jones potential. The convergence criteria were set to 1.0 × 10−4
kcal mol−1 and 0.005 kcal mol−1 Å−1 and 5.0 × 10−5 Å (displa-
cement) respectively. The obtained structural model possesses
lower symmetry than the parent-structure CAU-3-BDC and exhi-
Structure determination and refinement
ˉ
bits the space group R3, due to the break of the symmetry
The experimental XRPD pattern of CAU-3-BDC was success-
fully indexed with Topas Academics34 as a hexagonal crystal
system with the lattice parameters a = 21.0480(4) Å and c =
34.8305(7) Å with a goodness of fit of 23. Based on the extinc-
caused by the naphthalene dicarboxylate ions (atomic coordi-
nates and a simulated powder pattern of this model can be found
in Table S3 and Fig. S6†). This structural model was used to
carry out the Rietveld refinement. Due to the lower crystallinity
and the two-fold number of atomic parameters caused by the
lower symmetry, the atomic positions of the NDC2− ions were
restrained and an overall temperature factor was used. The final
ˉ
tion conditions, the rhombohedral space group R3m was
suggested by the program. The structure solution was carried out
successfully with the Expo200435 software package using Direct
ˉ
Methods. Starting from the space group R3m, the positions of
refinement leads to satisfying structure indicators (RBragg =
the aluminum based brick were determined. By recycling these
fragments in a new intensity extraction, in addition parts of the
BDC2− ion were localized. This starting model was completed
by force field calculations with the software package Forcite
implemented in Materials Studio 5.0.33 For the calculations, the
universal force field was used without an optimization of the cell
parameters. The completed model was refined with Rietveld
techniques using Topas Academics.34 The final Rietveld refine-
ment involved 15 background parameters, 15 atomic parameters,
4 temperature factors, 1 scale factor and 2 cell parameters. The
peak shape was modeled using a pearson VII function and aniso-
tropic peak broadening effects were taken into account using a
spherical harmonics series. The C–C distances and the C–O dis-
tances of the methoxy groups were restrained. The final Rietveld
plot shown in Fig. 3 led to satisfying structural model indicators
(RBragg = 0.4%, Rwp = 4.76% and GoF = 1.64). Selected bond
lengths and the asymmetric unit can be found in the supporting
informations in Fig. S4 and Table S1,† respectively.
4.47%, Rwp = 9.41% and GoF = 4.80). The final Rietveld plot is
shown in Figure S7† and selected bond lengths are given in
Table S4.† Two reflections at 32.62° and 45.42° are observed
which are due to NaCl as an impurity. The final parameters of all
three refinements are summarized in Table 1.
The structural model of CAU-3-BDC was used for the Riet-
veld refinement of CAU-3-BDC-NH2. The difference Fourier
calculations showed electron densities at a distance of 1.5 Å
from the lateral carbon atoms of the phenyl ring. Assigning these
electron densities to a nitrogen atom with a site occupation of
0.25 led to a significant lowering of the R-values Rbragg and Rwp
Fig. 3 Final Rietveld-plot for 1. Measured intensities are in black, cal-
culated intensities are in red, the difference plot is blue. The vertical bars
mark the Bragg-positions.
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 4164–4171 | 4167

Table 1 Final parameters obtained from the Rietveld-refinements
CAU-3-
BDC-NH2
CAU-3-BDC
CAU-3-NDC
Empirical formula Al2O8C12H16
Al2O8C12N1H17 Al2O8C16H18
M g mol−1
Crystal system
Space group
a/pm
342.21
357.23
368.25
Rhombohedral
ˉ
R3
Rhombohedral Rhombohedral
ˉ
ˉ
R3m
R3m
2110.51(5)
3488.8(1)
13458.1(8)
18
2093.5(1)
3481.3(2)
13214(1)
18
2320.56(2)
4063.51(4)
18950.(3)
18
c/pm
V/106 pm3
Z
Rwp/%
RBragg/%
GoF
4.76
0.4
1.64
4.17
0.59
1.71
9.41
4.47
4.8
weighted Durbin- 0.594
Watson statistic
0.312
0.120
Fig. 5 Schematic representation of a part of the distorted fcu-net. The
white bonds represent the connectivity in the xy-plane of the inorganic
brick, the black ones the connectivity along the z-axis.
Fig. 6 The two different types of cavities in the fcu-framework. Differ-
ent colours emphasize the ABC stacking of the of the Al-based bricks
12+
[Al12(OCH3)24
]
.
Fig. 4 Dodecameric building unit and its connectivity mode in the fra-
mework of CAU-3-BDC. Methyl-groups are omitted for clarity.
The twelvefold connectivity by dicarboxylate units leads to
the formation of a fcu-net (Fig. 5).
This high connectivity is remarkable, since only few examples
of such MOFs have been reported. Besides the well known
UiO-66,10 for example the frameworks of CAU-123 and its Ti-
analogue MIL-12540 exhibit this connectivity mode. While the
latter two MOFs show a distorted pseudo-bcc packing of clus-
ters, the assembly in CAU-3 leads to a fcc-packing like in
UiO-66. Due to the anisotropic shape of the cluster a strongly
distorted packing is observed (Fig. S7†).
Results and discussion
Structural description
The structure of CAU-3-BDC is based on dodecameric
[Al12(OCH3)24]12+ cations, to which twelve carboxylate moieties
are coordinated, each one bridging two Al-ions (Fig. 4).
Thus, the inorganic units are composed of twelve edge-
sharing AlO6-polyhedra. The edge-sharing oxygen atoms are
part of methanolate ions, while the other oxygen-atoms result
from the coordination of the bridging carboxylate-groups.
Similar dodecameric clusters have been only observed twice in
molecular complexes containing the transition metals Fe3+ and
Mn3+/Cr3+ ions.37,38 To the best of our knowledge this is the first
time, that this cluster has been observed in Al chemistry as well
as in the chemistry of MOFs. Recently, the Mn3+/Cr3+ building
block was proposed as a possible brick for the formation of new
metal–organic frameworks.39 Besides the octanuclear cyclic
cluster incorporated in the framework of CAU-1,23 this building
unit is the second example of an wheel-shaped aluminium car-
boxylate, whose occurrence is again strictly limited to the incor-
poration into a metal–organic framework.
Accordingly, the [Al12(OCH3)24]12+ clusters are connected
sixfold in their xy-plane and sixfold alternating in both directions
along the z-axis of the cyclic cluster. The resulting network con-
tains tetrahedral and octahedral cavities which are strongly dis-
torted due to the anisotropic shape of the Al-containing brick
(Fig. 6).
Assuming a spherical shape, the diameters of the tetrahedral
and octahedral cavities are approximately 10 and 11 Å (calcu-
lated based on van-der-Waals radii), respectively. In reality larger
molecules could be accommodated and based on the estimated
available free space, the incorporation of rod-shaped guests with
maximum length of 27 Å should be possible. The triangular
apertures of these cavities differ only slightly in size and should
be accessible for molecules up to a diameter of 7 Å.
4168 | Dalton Trans., 2012, 41, 4164–4171
This journal is © The Royal Society of Chemistry 2012

The dimensionality and shape of the pores differ only slightly
for compound 2, since the amino-group is statistically distributed
over the four possible positions of the aromatic ring. The larger
linker molecule in 3 leads to an extended, non-interpenetrating
framework containing distorted tetrahedral and octahedral cav-
ities of ∼14 Å and ∼15 Å in diameter, assuming a spherical
guest. Based on the structure, rod-shaped guests with maximum
length of 38 Å should fit into the octahedral cavities.
min−1 (Fig. S9, S10 and S11†). For 1, the first weight loss of
−16.5% corresponds to the removal of incorporated solvent mol-
ecules. At higher temperatures (∼200 °C), the decomposition of
the frameworks and thus the structural collapse proceeds in two
steps (calc.: 58.7%, obs.: 60.9%). The product formed in the end
is weakly crystalline Al2O3. In the case of CAU-3-BDC, we
were also able to prove the structural rigidity of the framework
during the activation process. Temperature-dependent XRPD
data (Fig. 8) demonstrates, that no uncommon cell parameter
shifts can be observed.
While the chemical decomposition of 1 shows a stepwise
mechanism in the TG curve, the structure of the framework col-
lapses directly during the second weight loss. The increased
stability during the TDXRPD experiment compared to the TG-
data is attributed to the different experimental set ups.
The decomposition of CAU-3-BDC-NH2 (2) is similar to that
of 1. Like for CAU-3-BDC, the decomposition starts after the
removal of adsorbed solvent (10.3%) at a temperature of
∼180 °C. The two last weight losses correspond quite well to the
decomposition of 2 (calc.: 66.4%, obs.: 62.4%), although the
difference could be attributed to a small amount of X-ray-amor-
phous Al-oxo-species.
The decomposition of 3 during the TG-experiment shows a
similar stability (∼180 °C), but due to the observed byproducts,
we did not attribute the weight losses to defined steps of
decomposition.
Spectroscopic and thermal properties
The vibrational spectra (Fig. 7) of the three title compounds are
very similar. The characteristic bands for the carboxylate
vibrations around 1580 cm−1 and 1420 cm−1 clearly show the
presence of the dicarboxylate ions coordinating to the Al3+-ions.
The aliphatic C–H vibrations at 2950 cm−1 and 2840 cm−1 are
due to the bridging methanolate ions in the Al-based brick. In
the case of as synthesized CAU-3-NDC, the absorption band
around 1700 cm−1 is attributed to residual naphthalene dicar-
boxylic acid molecules occluded in the pores.
The band at 1250 cm−1 (C–N-vibration) in the spectrum of
CAU-3-BDC-NH2 proves the incorporation of aminoterephtalic
acid, although the characteristic NH2-bands around 3450 cm−1
are not observed, probably due to the presence of hydrogen-
bonded water inside the pores. In the Raman-spectra, especially
the aromatic C–C-vibrations between 1640 cm−1 and 1380 cm−1
are well resolved.
The thermal stability of all three MOFs was investigated in air
atmosphere up to at least 800 °C with a heating rate of 4 K
Sorption properties
For the sorption measurements, the samples were activated in
vacuum (10−2 mbar) at 160 °C for 12 h. The XRPD patterns of
the samples after the sorption experiments can be found in the
supporting information (Fig. S1–S3). The nitrogen isotherms
(Fig. 9) were measured at 77 K. The BET-method was applied to
calculate the apparent surface area and the micropore volumes
were calculated from the amount adsorbed at p/p0 = 0.5.
The sorption experiments confirm the tendencies that we
expected. The incorporation of the amino-group leads to a
decrease of micropore volume as well as apparent surface area,
Fig. 7 IR- (top) and Raman-spectra (bottom) of the different CAU-3
Fig. 8 Temperature-dependent XRPD patterns of CAU-3-BDC (1),
MOFs.
measured under air.
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 4164–4171 | 4169

Fig. 10 Water vapour isotherms for 1 and 2 measured at 25 °C. Blue
triangles for the NH2-functionalized compound 2, red squares for 1.
Empty symbols represent the desorption-, filled symbols represent the
adsorption-branch.
Fig. 9 N2-Sorption isotherms for 1, 2 and 3 measured at 77 K.
Table 2 Apparent specific surface areas and Vmic values of the
CAU-3-MOFs
Acknowledgements
Compound
ABET m2 g−1
ALangmuir m2 g−1
Vmic cm3 g−1
This work has been financially supported by the DFG (SPP
1362). The research leading to these results has received funding
from the European Community’s Seventh Framework Pro-
gramme (FP7/2007–20013) under grant agreement n° 228862′.
CAU-3-BDC
CAU-3-BDC-NH2
CAU-3-NDC
1550
1250
2320
1920
1520
2750
0.64
0.53
0.95
Notes and references
while the enlargement of the linker molecule results in a drasti-
cally increased uptake. The BET-surfaces as well as the micro-
pore volumes are summarized in Table 2.
1 G. Ferey, Chem. Soc. Rev., 2008, 37, 191.
2 Z. Wang and S. M. Cohen, Chem. Soc. Rev., 2009, 38, 1315.
3 S. Kitagawa, R. Kitaura and S. Noro, Angew. Chem., Int. Ed., 2004, 43,
2334.
The apparent specific surface areas and the micropore volume
measured for CAU-3-NDC 3 are to the best of our knowledge
the highest reported so far for any literature known Al-MOF.
Changing the adsorbate from N2 to H2O, the influence of the
functional group is clearly visible (Fig. 10). The use of H2O
vapour leads to strong hysteresis. The absolute amount of
adsorbed water is similar for both CAU-3-BDC and CAU-3-
BDC-NH2, but due to the presence of the polar amino group, the
adsorbed amount of water vapour at lower partial pressures is
higher in CAU-3-NH2.
4 C. Janiak, Dalton Trans., 2003, 2781.
5 R. Long and O. M. Yaghi, Chem. Soc. Rev., 2009, 38, 1213.
6 O. M. Yaghi, M. O’Keeffe, N. W. Ockwig, H. K. Chae, M. Eddaoudi and
J. Kim, Nature, 2003, 423, 705.
7 T. Devic, P. Horcajada, C. Serre, F. Salles, G. Maurin, B. Moulin,
D. Heurtaux, G. Clet, A. Vimont, J.-M. Greneche, B. Le Ouay,
F. Moreau, E. Magnier, Y. Filinchuk, J. Marrot, J.-C. Lavalley, M. Daturi
and G. Ferey, J. Am. Chem. Soc., 2010, 132, 1127.
8 M. Eddaoudi, J. Kim, N. Rosi, D. Vodak, J. Wachter, M. O’Keeffe and
O. M. Yaghi, Science, 2002, 295, 469.
9 C. Serre, C. Mellot-Draznieks, S. Surblé, N. Audebrand, Y. Filinchuk and
G. Férey, Science, 2007, 315, 1828.
Although this is a purely qualitative sorption study, the
influence of the amino group with its ability for hydrogen
bonding is obvious.
10 J. H. Cavka, S. Jakobsen, U. Olsbye, N. Guillou, C. Lamberti, S. Bordiga
and K. P. Lillerud, J. Am. Chem. Soc., 2008, 130, 13850.
11 A. Schaate, P. Roy, A. Godt, J. Lippke, F. Waltz, M. Wiebcke and
P. Behrens, Chem.–Eur. J., 2011, 17, 6643.
12 M. Kandiah, S. Usseglio, S. Svelle, U. Olsbye, K. P. Lillerud and
M. Tilset, J. Mater. Chem., 2010, 20, 9848.
13 S. Bauer, C. Serre, T. Devic, P. Horcajada, J. Marrot, G. Férey and
N. Stock, Inorg. Chem., 2008, 47, 7568.
Conclusion
Summarizing our results, we have synthesized three new Al-
containing MOFs by applying our high-throughput-methods and
characterized them in detail. This new family of MOFs contains
an unprecedented brick, [Al12(OCH3)24]12+, which is twelvefold
connected by dicarboxylate molecules to form a fcu-net. This
inorganic unit is the second example for the occurrence of a
wheel-shaped aluminium cluster in a metal–organic framework.
All three compounds of the CAU-3-family are highly porous and
exhibit BET-surface areas larger than 1200 m2 g−1 and micro-
pore volumes higher than 0.5 cm3 g−1. They are thermally stable
up to at least 180 °C in air. The sorption properties are altered
upon incorporation of a functional group or a larger linker mol-
ecule, respectively.
14 P. Maniam and N. Stock, Inorg. Chem., 2011, 50, 5085.
15 T. Ahnfeldt, D. Gunzelmann, T. Loiseau, D. Hirsemann, G. Férey,
J. Senker and N. Stock, Inorg. Chem., 2009, 48, 3057.
16 A. Sonnauer, F. Hoffmann, M. Fröba, L. Kienle, V. Duppel,
M. Thommes, C. Serre, G. Férey and N. Stock, Angew. Chem., Int. Ed.,
2009, 48, 3791.
17 C. Volkringer, D. Popov, T. Loiseau, G. Férey, M. Burghammer,
C. Riekel, M. Haouas and F. Taulelle, Chem. Mater., 2009, 21, 5695.
18 C. Volkringer, D. Popov, T. Loiseau, N. Guillou, G. Ferey, M. Hauuas,
F. Taulelle, C. Mellot-Draznieks, M. Burghammer and C. Riekel, Nat.
Mater., 2007, 6, 760.
19 C. Volkringer, T. Loiseau, N. Guillou, G. Ferey, M. Haouas, F. Taulelle,
N. Audebrand, I. Margiolaki, D. Popov, M. Burghammer and C. Riekel,
Cryst. Growth Des., 2009, 9, 2927.
20 C. Volkringer, T. Loiseau, M. Haouas, F. Taulelle, D. Popov,
M. Burghammer, C. Riekel, C. Zlotea, F. Cuevas, M. Latroche,
4170 | Dalton Trans., 2012, 41, 4164–4171
This journal is © The Royal Society of Chemistry 2012

D. Phanon, C. Knöfelv, P. L. Llewellyn and G. Ferey, Chem. Mater.,
2009, 21, 5783.
30 S. Couck, J. F. Denayer, G. V. Baron, T. Remy, J. Gascon and F. Kapteijn,
J. Am. Chem. Soc., 2009, 131, 6326.
21 C. Volkringer, T. Loiseau, N. Guillou, G. Férey, M Haouas, F. Taulelle,
E. Elkaim and N. Stock, Inorg. Chem., 2010, 49, 9852.
22 C. Volkringer, T. Loiseau, N. Guillou, G. Ferey and E. Elkaım, Solid
State Sci., 2009, 11, 1507.
31 T. Ahnfeldt, J. Moellmer, V. Guillerm, R. Staudt, C. Serre and N. Stock,
Chem.–Eur. J., 2011, 17, 6462.
32 E. Biemmi, S. Christian, N. Stock and T. Bein, Microporous Mesoporous
Mater., 2009, 117, 111.
23 T. Ahnfeldt, N. Guillou, D. Gunzelmann, I. Margiolaki, T. Loiseau,
G. Férey, J. Senker and N. Stock, Angew. Chem., Int. Ed., 2009, 48,
5163.
24 W. Schmitt, E. Baissa, A. Mandel, C. E. Anson and A. K. Powell, Angew.
Chem., Int. Ed., 2001, 40, 3577.
25 E. A. Mainicheva, O. A. Gerasko, L. A. Sheludyakova, D. Y. Naumov,
M. I. Naumova and V. P. Fedin, Russ. Chem. Bull., 2006, 55, 267.
26 T. Loiseau, C. Serre, C. Huguenard, G. Fink, F. Taulelle, M. Henry,
T. Bataille and G. Ferey, Chem.–Eur. J., 2004, 10, 1373.
27 I. Senkovska, F. Hoffmann, M. Fröba, J. Getzschmann, W. Böhlmann and
S. Kaskel, Microporous Mesoporous Mater., 2009, 122, 93.
28 A. Comotti, S. Bracco, P. Sozzani, S. Horike, R. Matsuda, J. Chen,
M. Takata, Y. Kubota and S. Kitagawa, J. Am. Chem. Soc., 2008, 130,
13664.
33 Materials Studio Version 5.0, Accelrys Inc., San Diego, CA, 2009.
34 Topas Academics 4.2, Coelho Software, 2007.
35 A. Altomare, M. Cavalli, R. Calandro, C. Cuocci, C. Giacovazzo,
A. Gagliardi, A. G. G. Moliterni and R. Rizzi, EXPO2004. Program for
Solving Crystal Structures from Powder Data by direct Methods, 2004.
36 STOE WinXPOW version 2.11, Stoe & Cie GmbH, 2005, Darmstadt,
Germany.
37 T. C. Stamatatos, A. G. Christou, C. M. Jones, B. J. O’Callaghan,
K. A. Abboud, T. A. O’Brien and G. Christou, J. Am. Chem. Soc., 2007,
129, 9840.
38 M. Helliwell, A. A. Smith, S. J. Teat and R. E. P. Winpenny, Inorg. Chim.
Acta, 2003, 354, 49.
39 D. J. Tranchemontagne, J. L. Mendoza-Cortes, M. O’Keeffe and
O. M. Yaghi, Chem. Soc. Rev., 2009, 38, 1257.
29 J. Gascon, U. Aktay, M. D. Hernandez-Alonso, G. P. M. Van Klink and
F. Kapteijn, J. Catal., 2009, 261, 75.
40 M. Dan-Hardi, C. Serre, T. Frot, L. Rozes, G. Maurin, C. Sanchez and
G. Ferey, J. Am. Chem. Soc., 2009, 131, 10857.
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 4164–4171 | 4171