Communications
DOI: 10.1002/anie.200901241
Catalytically Active MOFs
Heterogeneous Catalytic Oxidation by MFU-1:
A Cobalt(II)-Containing Metal–Organic Framework
Markus Tonigold, Ying Lu, Bjꢀrn Bredenkꢀtter, Bernhard Rieger, Stefan Bahnmꢁller,
Julia Hitzbleck, Gerhard Langstein, and Dirk Volkmer*
Porous metal–organic frameworks (MOFs) are a rapidly
emerging class of multifunctional hybrid materials that might
be useful for diverse technical applications, such as gas or
liquid adsorption and separation, molecular recognition, or
catalysis.[1] Combining polycarboxylate ligands and (transi-
tion) metal ions, moderately robust MOFs can be prepared;
1,4-benzenedicarboxylate (bdc, terephthalic acid) and 4,4’-
biphenyldicarboxylate (bpdc) are often used as linkers.
Highly porous non-interpenetrated frameworks, such as the
well-known MOF-5 ([Zn4O(bdc)3])[2] or IRMOF-9 ([Zn4O-
(bpdc)3])[3] then form. These microporous MOFs generally
show good thermal stabilities (decomposition occurs at T>
3508C). A fundamental disadvantage, however, is their low
hydrolytic stability: Decomposition of the framework occurs
rapidly when the gas or liquid phase contains a small amount
of water,[4] which imposes severe limitations on their usage in
catalytic oxygenation reactions, in which water is a major
reaction product. Preliminary attempts to use MOF-5 as a
photocatalyst have been reported recently;[5] however, the
fact that these frameworks contain Lewis acidic zinc(II) ions
only imposes severe limitations on their use in redox catalytic
applications in general.
oxidations using a MOF catalyst.[7] They used an enantiomer-
ically pure manganese complex of a modified salen ligand as a
building block to construct a three-dimensional porous
framework. A distinct approach towards heterogeneous
asymmetric catalysis based on a homochiral metal–organic
framework was recently proposed by Lin et al.[8] However,
industrial oxidation or oxygenation reactions typically require
very high turn-over numbers (TONs) and frequencies
(TOFs), which have not been achieved to date by current
MOF catalysts.
To produce thermally and hydrolytically stable redox-
active MOFs, our initial efforts focused on the isostructural
replacement of a single zinc ion by an open-shell transition
metal ion M within the tetranuclear {Zn4O} coordination unit
of MOF-5. However, all attempts in this direction led to
heteronuclear MOFs containing trinuclear coordination units
(for example, [MZn2(bpdc)3(dmf)2], M = CoII, NiII, CdII),
which are structurally different from MOF-5.[9] A search of
the CSD database, however, led to the tetranuclear complex
[CoII O(3,5-dmpz)6] (3,5-dmpz = 3,5-dimethylpyrazolate),[10]
4
which is a structural analogue of basic zinc acetate, [Zn4O-
(OAc)6], the prototypic secondary building unit of MOF-5
(Figure 1). Thus, reacting the ligand 1,4-bis[(3,5-dimethyl)-
pyrazol-4-yl]benzene (H2bdpb)[11] with suitable cobalt(II)
salts under solvothermal conditions led to formation of a
series of novel cobalt(II)-based MOF compounds,[12] one of
Conceptually different approaches have been reported to
circumvent the intrinsic disadvantages of MOF-5-type frame-
works. Fischer et al. reported the gas-phase deposition of
volatile organometallic complexes in the open cavities of
MOF-5. Subsequent photolytic or reductive cleavage of the
precursors led to catalytically active metal clusters (Cu, Pd,
Au) that are finely dispersed in the MOF-5 framework.[6]
Nguyen, Hupp et al. were among the first to present
which, [CoII O(bdpb)3] (MFU-1),[13] is described herein.
4
Single crystals of MFU-1 were grown under solvothermal
conditions from solutions of the ligand H2bdpb and cobalt(II)
chloride in DMF at 1208C (see the Supporting Information).
For a more efficient bulk synthesis of the phase-pure
compound, we employed a microwave system, which reduced
the reaction time from several days to a few minutes. X-ray
powder diffraction studies (Supporting Information, Fig-
ure S2) showed that microcrystals obtained from the micro-
wave-assisted synthesis are structurally identical to those
obtained from solvothermal synthesis, but they are signifi-
cantly smaller and more uniform in size (Supporting Infor-
mation, Figures S3, S4).[16]
[*] M. Tonigold,[+] Dr. Y. Lu,[+] Dr. B. Bredenkꢀtter, Prof. D. Volkmer
Institut fꢁr Anorganische Chemie II—Materialien und Katalyse
Universitꢂt Ulm, Albert-Einstein-Allee 11, 89081 Ulm (Germany)
Fax : (+49)731-50-23039
E-mail: dirk.volkmer@uni-ulm.de
Prof. B. Rieger
WACKER-Lehrstuhl fꢁr Makromolekulare Chemie
Technische Universitꢂt Mꢁnchen
¯
MFU-1 crystallizes in space group P43m with a =
Lichtenbergstrasse 4, 85747 Garching (Germany)
15.963 ꢀ.[17] The bdpb ligands and {Co4O} units are linked
into a non-interpenetrated network (Figure 2) of low density
(1calcd = 0.43 gcmꢀ3). The structure of MFU-1 is similar to
MOF-5, which has a CaB6-type framework topology. The
MFU-1 network encloses octahedral {Co4O(dmpz)6} nodes
that are reminiscent of the {Zn4O(CO2)6} secondary building
units of MOF-5. Phenylene rings constituting the edges of the
cubic CaB6 network. The framework has three-dimensional
intersecting channels that encompass almost spherical voids
Dr. S. Bahnmꢁller, Dr. J. Hitzbleck, Dr. G. Langstein
Bayer MaterialScience AG, BMS-CD-NB-NT
Gebꢂude B211, 51368 Leverkusen (Germany)
[+] These authors contributed equally to this work.
Supporting information (17 pages) for this article (details on the
synthesis of MFU-1, experimental methods and equipment used for
investigations, and procedures for investigations of catalytic reac-
7546
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 7546 –7550