Gated channels in a honeycomb-like zinc-dicarboxylate-bipyridine framework with flexible alkyl ether side chains

Article abstract of DOI:10.1021/ja109317e

Covalent functionalization of 1,4-benzenedicarboxylate (bdc) with methoxyethoxy groups induces conformational freedom in this molecule. Applying these 2,5-disubstituted bdc derivatives in metal-organic framework synthesis together with 4,4′-bipyridine as coligand yields novel honeycomb-like structures. The cylindrical channels of these materials are populated with flexible groups, which act as molecular gates for guest molecules. This allows highly selective sorption of CO₂ over N₂ and CH ₄.

Full text of DOI:10.1021/ja109317e

COMMUNICATION
pubs.acs.org/JACS
Gated Channels in a Honeycomb-like Zinc-Dicarboxylate-Bipyridine
Framework with Flexible Alkyl Ether Side Chains
Sebastian Henke and Roland A. Fischer*
Anorganische Chemie II, Organometallics & Materials, Ruhr-Universit€at Bochum, D-44780 Bochum, Germany
S Supporting Information
b
ABSTRACT: Covalent functionalization of 1,4-benzenedi-
carboxylate (bdc) with methoxyethoxy groups induces
conformational freedom in this molecule. Applying these
2,5-disubstituted bdc derivatives in metal-organic frame-
work synthesis together with 4,40-bipyridine as coligand
yields novel honeycomb-like structures. The cylindrical
channels of these materials are populated with flexible
groups, which act as molecular gates for guest molecules.
This allows highly selective sorption of CO2 over N2 and
CH4.
etal-organic frameworks (MOFs) represent a young,
M
rapidly developing class of porous materials. Due to their
high pore volumes and exceptional specific surface areas, they are
very interesting for a variety of applications, like gas storage,1
catalysis,2 or gas separation.3 MOFs are hybrid materials con-
structed from cationic metal ions or metal-oxo clusters and
anionic linkers. Yaghi and co-workers have demonstrated that the
intrinsic features of the linkers play an important role for the
topology of the network structure.4 In particular, additional
substituents at the carboxylate linkers may introduce control
over alternative topologies.4b Therefore, the design of particular
linkers in combination with the other MOF building blocks repre-
sents a tool for deriving novel MOFs with unusual structures.
Figure 1. Structure of compound 1as as determined by single-crystal
X-ray diffraction. (a) Zinc-centered inorganic building unit, (b) side view
along the c-axis, and (c) unit cell in the [001] direction. The disordered
CH3-O-C2H4 chains of the BME-bdc linker are not shown in panels a
and b; see Figure 2 for a schematic drawing. Zn, O, N, and C atoms are
shown in yellow, red, blue, and gray, respectively. Hydrogen atoms were
omitted for the sake of clarity. In panel c, channel types A and B are
marked in cyan and beige.
Many isoreticular MOFs of the type [M2L2P]n (M = Zn2þ
,
Cu2þ; L = linear dicarboxylate linker; P = neutral pillar, e.g., 1,4-
diazabicylco[2.2.2]octane, pyrazine, or 4,40-bipyridine) are
known.5,6 All these MOFs are constructed from 2D layers of
[M2L2]n, interconnected in the third dimension by the pillar P.
The 2D layers are most often square grids, but polymorphs with
Kagome-type layers were reported recently for some zinc
derivatives.6 A dimeric paddlewheel building unit represents
the characteristic inorganic brick of all known [M2L2P]n frame-
works. Typically, the M2þ ions are five-fold coordinated in a
square planar fashion by four carboxylate oxygen atoms of L and
one axial nitrogen donor atom of P.
Herein we report the discovery of a framework of the
composition [Zn2L2P]n (1) [L = 2,5-bis(2-methoxyethoxy)-1,4-
benzenedicarboxylate (BME-bdc); P = 4,40-bipyridine (bipy)],
which exhibits a honeycomb-like topology that is unknown for
MOFs of the [M2L2P]n family (Figures 1 and 2). The linker
BME-bdc features flexible alkyl ether groups attached to the rigid
phenyl ring, but the particular 2,5-disubstitution increases the
conformational freedom of the carboxylate groups, which are
tilted away from the plane of the phenyl ring. This favors the
formation of a MOF structure with a rod-like structural motif
instead of the otherwise preferred paddlewheel structural motif
of the [M2L2P]n family.
H2BME-bdc, bipy, and Zn(NO3)2 6H2O were dissolved in a
3
DMF/EtOH mixture and heated to 85 °C. After 48 h, colorless
prismatic crystals of the composition {[Zn2(BME-bdc)2-
(bipy)]n(DMF)2.3(EtOH)0.4} were isolated in good yield
(1as). The amount of solvent guests in the pores of the as
synthesized material was calculated from thermogravimetric
analysis (TGA) and elemental analysis data. The TGA of 1as
Received: October 15, 2010
Published: January 28, 2011
r
2011 American Chemical Society
2064
dx.doi.org/10.1021/ja109317e J. Am. Chem. Soc. 2011, 133, 2064–2067
|

Journal of the American Chemical Society
COMMUNICATION
Figure 3. Depiction of the three different conformations of BME-bdc in
the crystal structure of 1as: 25% of the dicarboxylates in 1as are of type
(a), 50% of type (b) and 25% of type (c). Hydrogen atoms and
disordered CH3-O-C2H4 groups are not shown. Zn, O, and C atoms
are shown in yellow, red, and gray, respectively.
Figure 2. Sketch of the pore aperture of compound 1. The rigid
backbone of the framework is shown in blue, and the flexible alkyl ether
substituents are shown in red.
1; it rather forms an interpenetrated, pillared paddlewheel
structure. Increased conformational freedom may also reduce
the coordinative strain on the Zn2þ ions. In paddlewheel
structures, Zn is forced into the square-pyramidal coordination
geometry, whereas in 1, Zn exhibits its typical quasi-tetrahedral
coordination geometry with less steric crowding. This conforma-
tional freedom of BME-bdc is the key to form a different
structure. In addition, the dipolar interactions of the flexible
linkers within the channels may play a significant role as well.
Furthermore, and very importantly, the steric features, in parti-
cular the lengths of the anionic linker L (BME-bdc) and the
neutral connector P (bipy), must be similar in order to construct
a framework of the topology of 1 (Figure S2). This is nicely
evinced from the crystal structure of [Zn2(BME-bdc)2(dabco)]n
(dabco = 1,4-diazabicyclo[2.2.2]octane). Dabco is another pillar
P, which is very often utilized in [M2L2P]n systems. It is
significantly shorter than bipy and BME-bdc. Therefore, [Zn2-
(BME-bdc)2(dabco)]n cannot form a honeycomb-like structure
analogous to [Zn2(BME-bdc)2(bipy)]n and, in fact, crystallizes
in the conventional pillared paddlewheel structure.8
Interestingly, the use of 2-(2-methoxyethoxy)benzenedicar-
boxylic acid (H2ME-bdc), the monosubstituted analogue of
H2BME-bdc, leads to the (expected) interpenetrated pillared
paddlewheel framework {[Zn2(ME-bdc)2(bipy)]n(DMF)x-
(EtOH)y} (2). Compound 2 has the same network topology
and internal connectivity as [Zn2(bdc)2(bipy)]n, also known as
MOF-508 (Figure S9, Table S2).5b-5d This comparison nicely
shows the importance of 2,5-disubstitution at the bdc linker to
trigger the formation of the unique topology of 1. Accordingly,
we synthesized two related 2,5-substituted H2bdc derivatives,
2,5-bis(3-methoxypropoxy)benzenedicarboxylic acid (H2BMP-
bdc) and 2,5-bis(n-butoxy)benzenedicarboxylic acid (H2BB-
bdc). Analogous syntheses using H2BMP-bdc and H2BB-bdc
shows a weight loss of 16.5% up to 150 °C. This amount fits well
to the reported molecular composition (Figure S8).
Phase purity of 1as is demonstrated by comparison of the
calculated powder X-ray diffraction (PXRD) pattern from the
single-crystal data with the experimental one (Figure S3). Com-
pound 1as crystallizes in the trigonal space group R3c (Table S1).
As expected, not all atoms of the flexible substituent chains could
be located in the single-crystal structure refinement due to the
high disorder. Nevertheless, solid-state 13C-MAS NMR and
1
elemental analysis, as well as H and 13C NMR in DCl/D2O/
DMSO-d6, of activated, solvent-free 1dry clearly support the
reported composition (Figures S6 and S7).
The structure of 1as reveals isolated quasi-tetrahedrally co-
ordinated Zn atoms, which are coordinated by one bipy ligand,
one bidentate chelating carboxylate group, and two carboxylate
groups, which bridge in syn-anti fashion to the next zinc center
(Figure 1a). This unit is well known in inorganic coordination
chemistry,7 but, to the best of our knowledge, it has never been
implemented in a robust, three-dimensional MOF structure
before. The persistence of two different types of carboxylate
groups (bridging and chelating) in the case of the activated
(dried) material 1dry is evidenced by solid-state 13C-MAS NMR
and IR spectroscopy (Figures S6 and S14). An infinite chain of
Zn atoms bridged by carboxylate groups runs along the c-axis of
the structure of 1as (Figure 1b). Bipy and BME-bdc cross-link the
zinc-carboxylate chains. This results in a three-dimensional
structure with two different kinds of one-dimensional channels
(Figure 1c). Essentially, the function of bipy in 1as is not that of a
pillar between zinc-carboxylate sheets but rather an additional
connector between infinite zinc-carboxylate rods.
The flexible CH3-O-C2H4 chains of the BME-bdc linkers
point into the channels. In fact, 1 represents a rigid MOF
constructed from a honeycomb-like backbone structure and
has solvent-like properties from the flexible groups, which lead
to a polar, weakly Lewis basic environment within the pores
(Figure 2). One prerequisite for the preference of the obtained
honeycomb topology instead of the trivial tetragonal topology
may be the increased conformational freedom of the carboxylate
groups of the BME-bdc linker as compared to conventional bdc.
Due to the steric repulsion induced by 2,5-disubstitution, copla-
narity of the phenyl ring with the carboxylate groups caused by
π-overlap is no longer preferred.
with Zn(NO3)2 6H2O and bipy in DMF/EtOH yield MOF
3
materials isostructural to 1, as evidenced by PXRD and IR
spectroscopy (Figures S13 and S14). However, using 2,5-sub-
stituted H2bdc for Zn2þ-based MOFs without any coligand
does not lead to new topologies by necessity. We studied the
implementation of BME-bdc and some of the above-mentioned
linkers L into the series of MOF-5 isotypes [Zn4OL3].8
A sample of 1as was activated by soaking in CHCl3 and drying at
140 °Cin vacuo to achieve 1dry. PXRDof1dry shows that the structure
is still intact after guest removal (Figure S3). Moreover, 1dry is
insoluble and stable in common organic solvents (namely EtOH,
THF, MeCN, hexane, acetone, toluene, DMSO), as shown by PXRD
analyses (Figure S4). However, exposing 1dry to water vapor irrever-
sibly transforms the structure to a so-far unknown crystalline phase
(Figure S5). It is supposed that a MOF based on Zn-carboxylates
cannot hold large amounts of moisture without structural decom-
position, if it is not specifically functionalized to be very hydrophobic.9
Three distinct conformations for the BME-bdc linker are
found in 1as (Figure 3). The majority of the linkers exhibit the
conformation shown in Figure 3b, where the carboxylate groups
are twisted about 90° relative to each other. Such a conformation
of the dicarboxylate linker is not possible in bdc, and thus
[Zn2(bdc)2(bipy)]n cannot form the isoreticular framework of
2065
dx.doi.org/10.1021/ja109317e |J. Am. Chem. Soc. 2011, 133, 2064–2067

Journal of the American Chemical Society
COMMUNICATION
Table 1. Adsorbed Volumes of N2, CH4, and CO2 in 1dry at
1 bar
Vads (cm3 g-1)
T (K)
N2
CH4
CO2
77
∼0.1
16
-
60
7
-
156
52
195
273
298
∼0.1
∼0.1
6
33
isotherms at 195, 273, and 298 K show higher uptakes than for
N2, but much less so compared to CO2 at the same temperature.
The sorption of gas mixtures in porous materials can be
reliably estimated from single-component sorption isotherms.10
Therefore, we determined the initial slopes in the Henry region
of the adsorption isotherms of 1dry (Figures S15-S17). The
ratios of the slopes were used to estimate the sorption selec-
tivities for CO2 over N2 and for CO2 over CH4. From these data,
the CO2/N2 selectivity is 84:1 at 273 K (611:1 at 195 K and 40:1
at 298 K). To the best of our knowledge, this selectivity is slightly
higher than the highest selectivity reported previously at this
temperature.11 The CO2/CH4 selectivity is 9:1 at 273 K (77:1 at
195 K and 9:1 at 298 K). The sorption selectivity of 1dry is not
due to a simple size-sieving effect. CH4 is bigger than N2, but it
has a higher polarizability and may interact more strongly with
the polar pore surface of 1dry. Therefore, increased adsorption is
expected. These data indicate that 1dry can compete with the
benchmark MOFs for highly selective CO2 capture.11,12 This
substantiates the concept of using alkyl ether side chains. The
flexibility of these substituents will have a major effect on the
performance of 1dry in the adsorption of multicomponent gas
mixtures. Therefore, breakthrough measurements as well as
multicomponent adsorption measurements are a matter of on-
going work in our laboratories. However, the reactivity of 1dry
toward water suggests that the material will not have real-life
application in CO2 capture. Using other metal centers instead of
Zn2þ, e.g., Mg2þ and Co2þ, exhibiting a suitable coordination
chemistry, isoreticular congeners of 1 may be accessible, which
are likely to be much more water tolerant and may therefore be of
potential interest for CO2 capture under industrial conditions.
In conclusion, we have demonstrated a novel concept of using
tailored linker functionalization to trigger otherwise inaccessible
MOF network topologies for an important MOF family of highly
variable [M2L2P]n compositions. The particular alkyl ether sub-
stituents do not interfere with the MOF synthesis (e.g., avoiding
open-framework formation) but rather have a major impact on the
conformational freedom of the carboxylate groups of the linker,
which is the key for obtaining a previously unknown, functionalized
honeycomb-like structure. The obtained MOF shows ultrahigh
selective sorption of CO2 over N2 and CH4, suggesting possible
applications in capturing CO2 from flue gases or upgrading natural
gas by CO2/CH4 separations, if water-tolerant isoreticular deriva-
tives can be made. The concept is extendable to a series of
isoreticular frameworks by employing other 2,5-disubstituted bdc
derivatives. In principle, a wide variety of functional groups can be
integrated to design the coordination space of the material for
specific applications in gas adsorption and separation. Future work
will examine how the sorption properties of the materials depend
on the nature and the combinations of the substituents with
particular respect to the concept of multivariate MOFs introduced
by Yaghi and co-workers.13
Figure 4. Sorption isotherms recorded on 1dry. Adsorption and desorp-
tion branches are shown with closed and open symbols, respectively.
The material 1dry is expected to show selective sorption
properties toward small polar molecules due to its polar, sol-
vent-like CH3-O-C2H4 chains.8 In order to analyze the sorp-
tion behavior of 1dry, N2, CO2, and CH4 sorption isotherms were
recorded at various temperatures (Figure 4). Interestingly, 1dry
does not adsorb any nitrogen at 77 K. Raising the temperature to
195 K increases the sorption capacity only slightly. At 77 K, the
methoxyethoxy chains have quite low thermal energy and lock
the pore aperture windows, keeping out nitrogen molecules like a
molecular gate. Raising the temperature to 195 K increases the
thermal motion of the flexible chains, which makes it easier for N2
to penetrate into the channels of 1dry. Increasing the temperature
further to 273 or 298 K decreases the N2 sorption capacity again
toward almost zero, as expected for thermodynamic reasons.
Therefore, the adsorption at 273 and 298 K can be attributed to
adsorption on the outer surface of the MOF powder and not in
the porous structure.
The CO2 sorption measurements at 195, 273, and 298 K show
typical type I isotherms with high uptake (Table 1). At 195 K, a
steep increase of the CO2 uptake takes place in the low-pressure
region (0.00-0.05 bar), and a small hysteresis is visible in the
desorption branch. The hysteresis can be attributed to slow
kinetics of adsorption. The uptake of CO2 at 273 K and 1 bar is
∼500 times higher than the uptake of N2 (∼10 times higher at
195 K; ∼300 times higher at 298 K) (Table 1). CH4 sorption
2066
dx.doi.org/10.1021/ja109317e |J. Am. Chem. Soc. 2011, 133, 2064–2067

Journal of the American Chemical Society
COMMUNICATION
’ ASSOCIATED CONTENT
(10) (a) Myers, A. L.; Prausnitz, J. M. AIChE J. 1965, 11, 121–127.
(b) Bae, Y.-S.; Farha, O. K.; Spokoyny, A. M.; Mirkin, C. A.; Hupp, J. T.;
Snurr, R. Q. Chem. Commun. 2008, 4135–4137. (c) Bae, Y.-S.; Farha,
O. K.; Hupp, J. T.; Snurr, R. Q. J. Mater. Chem. 2009, 19, 2131–2134.
(11) An, J.; Geib, S. J.; Rosi, N. L. J. Am. Chem. Soc. 2010, 132, 38–39.
(12) (a) Banerjee, R.; Furukawa, H.; Britt, D.; Knobler, C.; O’Keeffe,
M.; Yaghi, O. M. J. Am. Chem. Soc. 2009, 131, 3875–3877. (b)
Demessence, A.; D’Alessandro, D. M.; Lin Foo, M.; Long, J. R. J. Am.
Chem. Soc. 2009, 131, 8784–8786. (c) Wang, B.; C^ote, A. P.; Furukawa,
H.; O’Keeffe, M.; Yaghi, O. M. Nature 2008, 453, 207–212. (d)
Vaidhyanathan, R.; Iremonger, S. S.; Dawson, K. W.; Shimizu,
G. K. H. Chem. Commun. 2009, 5230–5232. (e) Choi, H.-S.; Suh,
M. P. Angew. Chem., Int. Ed. 2009, 48, 6865–6869.
S
Supporting Information. Detailed synthesis procedures,
b
CIF files of 1as and 2, and further analytical data (Figures S1-S17).
This material is available free of charge via the Internet at http://
pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
roland.fischer@ruhr-uni-bochum.de
(13) Deng, H.; Doonan, J. Ch.; Furukawa, H.; Ferreira, R. B.;
Towne, J.; Knobler, C. B.; Wang, B.; Yaghi, O. M. Science 2010, 327,
846–850.
’ ACKNOWLEDGMENT
The authors gratefully thank Prof. C. W. Lehmann and A.
Dreier from the Max-Planck Institute f€ur Kohlenforschung
(M€uhlheim an der Ruhr, Germany) for collecting single-crystal
X-ray diffraction data of 1as. The Priority Program 1362 of the
German Research Foundation on “Metal-Organic Frameworks”
is acknowledged for funding. The Research School of the Ruhr-
Universit€at Bochum and the Fonds der chemischen Industrie are
acknowledged for fellowships (S.H.).
’ REFERENCES
(1) (a) Eddaoudi, M.; Kim, J.; Rosi, N.; Vodak, D.; Wachter, J.;
O’Keeffe, M.; Yaghi, O. M. Science 2002, 295, 469–472. (b) Millward,
A. R.; Yaghi, O. M. J. Am. Chem. Soc. 2005, 127, 17998–17999. (c) Dinca,
M.; Long, J. Angew. Chem., Int. Ed. 2008, 47, 6766–6779.
(2) (a) Lee, J. Y.; Farha, O. K.; Roberts, J.; Scheidt, K. A.; Nguyen,
S. T.; Hupp, J. T. Chem. Soc. Rev. 2009, 38, 1450–1459. (b) Farrusseng,
D.; Aguado, S.; Pinel, C. Angew. Chem., Int. Ed. 2009, 48, 7502–7513.
(3) (a) Li, J.-R.; Kuppler, R. J.; Zhou, H.-C. Chem. Soc. Rev. 2009, 38,
1477–1504. (b) Chen, B.; Xiang, S.; Qian, G. Acc. Chem. Res. 2010, 43,
1115–1124. (c) Zhang, J.-P.; Chen, X.-M. J. Am. Chem. Soc. 2008, 130,
6010–6017.
(4) (a) Furukawa, H.; Kim, J.; Ockwig, N. W.; O’Keeffe, M.; Yaghi,
O. M. J. Am. Chem. Soc. 2008, 130, 11650–11661. (b) Eddaoudi, M.;
Kim, J.; O’Keeffe, M.; Yaghi, O. M. J. Am. Chem. Soc. 2002, 123, 376–377.
(c) Eddaoudi, M.; Kim, J.; Vodak, D.; Sudik, A.; Wachter, J.; O’Keeffe, M.;
Yaghi, O. M. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 4900–4904.
(5) See for example: (a) Dybtsev, D.; Chun, H.; Kim, K. Angew.
Chem., Int. Ed. 2004, 43, 5033–5036. (b) Ma, B.-Q.; Mulfort, K. L.;
Hupp, J. T. Inorg. Chem. 2005, 44, 4912–4914. (c) Chen, B.; Liang, C.;
Yang, J.; Contreras, D. S.; Clancy, Y. L.; Lobkovsky, E. B.; Yaghi, O. M.;
Dai, S. Angew. Chem., Int. Ed. 2006, 45, 1390–1393. (d) Chun, H.;
Dybtsev, D. N.; Kim, H.; Kim, K. Chem. Eur. J. 2005, 11, 3521–3529. (e)
Horike, S.; Matsuda, R.; Tanaka, D.; Matsubara, S.; Mizuno, M.; Endo,
K.; Kitagawa, S. Angew. Chem., Int. Ed. 2006, 45, 7226–7230. (f) Chen,
B.; Ma, S.; Hurtado, E. J.; Lobkovsky, E. B.; Zhou, H. C. Inorg. Chem.
2007, 46, 8490–8492.
(6) (a) Chun, H.; Moon, J. Inorg. Chem. 2007, 46, 4371–4373. (b)
Friꢀsiꢁc, T.; Reid, D. G.; Halasz, I.; Stein, R. S.; Dinnebier, R. E.; Duer,
M. L. Angew. Chem., Int. Ed. 2010, 49, 712–715. (c) Kondo, M.;
Takashima, Y.; Seo, J.; Kitagawa, S.; Furukawa, S. Cryst. Eng. Commun.
2010, 12, 2350–2353.
(7) (a) Singh, B.; Long, J. R.; Fabrizi de Biani, F.; Gatteschi, D.;
Stavropoulos, P. J. Am. Chem. Soc. 1997, 119, 7030–7047. (b) Kumar, U.;
Thomas, J.; Thirupathi, N. Inorg. Chem. 2010, 49, 62–72.
(8) Henke, S.; Schmid, R.; Grunwaldt, J.-D.; Fischer, R. A. Chem.
Eur. J. 2010, 48, 14296–14306.
(9) (a) Greathouse, J. A.; Allendorf, M. D. J. Am. Chem. Soc. 2006,
128, 10678–10679. (b) Schr€ock, K.; Schr€oder, F.; Heyden, M.; Fischer,
R. A.; Havenith, M. Phys. Chem. Chem. Phys. 2008, 10, 4732–4739. (c)
Low, J. J.; Benin, A. I.; Jakubczak, P.; Abrahamian, J. F.; Faheem, S. A.;
Willis, R. R. J. Am. Chem. Soc. 2009, 131, 15834–15842. (d) Nguyen, J.
L.; Cohen, S. M. J. Am. Chem. Soc. 2010, 132, 4560–4561.
2067
dx.doi.org/10.1021/ja109317e |J. Am. Chem. Soc. 2011, 133, 2064–2067