Angewandte
Chemie
DOI: 10.1002/anie.200906166
Homogeneous Catalysis
A Recyclable Nanoparticle-Supported Palladium Catalyst for the
Hydroxycarbonylation of Aryl Halides in Water**
Sebastian Wittmann, Alexander Schꢀtz, Robert N. Grass, Wendelin J. Stark, and Oliver Reiser*
Separation and recycling of homogeneous catalysts is consid-
ered a main objective in the area of green chemistry.[1] One
approach to such an environmentally benign process is based
on the development of solid-supported catalysts exhibiting
high activity even in aqueous media.[2] Palladium complexes
have received tremendous interest as powerful catalysts for
coupling reactions of aryl halides[3] with great industrial
potential. Hence, efforts have been made to immobilize them
on diverse soluble and insoluble supports, such as inorganic
solids,[4] polymers,[5] dendrimers,[6] perfluorinated tags,[7] and
nanoparticles.[8]
Heterogeneous supports allow efficient recycling by
filtration, albeit a substantial decrease in the activity of the
immobilized catalysts is frequently observed. Soluble sup-
ports usually require a second solvent for the selective
precipitation of the supported catalyst out of the reaction
mixture (e.g. JandaJel, ROMP-gel, MeOPEG) or for extrac-
tion into the orthogonal liquid phase (e.g. into a fluorinated
solvent).
Nanoparticles are considered a semi-heterogeneous sup-
port since they are readily dispersed and exhibit an intrinsi-
cally high surface area, which is combined with excellent
accessibility of the surface-bound catalytic sites. Some
particles are even amenable to magnetic separation, thus,
avoiding the need for catalyst separation by filtration.
However, examples for nanoparticle-grafted catalysts show-
ing equal or even superior activity when compared to their
homogeneous counterparts are still scarce.[9]
Ultimately, a catch–release system for a homogeneous
catalyst would be a remedy for rate limitations caused by solid
supports provided that a low contaminant level of the catalyst
in solution is maintained. Herein we describe a palladium
complex, noncovalently grafted on Co/C nanoparticles that
dissociates from the nanoparticles for the course of the
hydroxycarbonylation of aryl halides in water (1008C) and is
then recaptured on the graphene-layer of the magnetic
nanoparticles (room temperature).[10] The high magnetic
remanence of the nanoparticle powder permits its efficient
separation and recycling.
Grass et al. have reported the preparation of cobalt
nanoparticles coated with multiple graphene layers that are
produced by a scalable one-step flame spray synthesis
process.[11] We could demonstrate that the carbon shell of
these particles can be suitably functionalized to allow the
covalent attachment of catalysts,[12] however, multiple syn-
thetic steps are required. Pyrene moieties interact with
graphene layers through p–p stacking,[13] and some examples
exploiting this were recently published showing the non-
covalent modification of carbon nanotubes, for example, with
biomolecules[13c] or olefin metathesis initiators and cata-
lysts.[14] We envisaged that carbon-encapsulated Co/C nano-
magnets are equally suited for such a reversible functional-
ization with pyrene units, which would greatly facilitate the
immobilization of catalysts. In addition, the remarkable
stability of the Co/C particles along with the exceptionally
high saturation magnetization (158 emugÀ1) fulfills all the
requirements for catalysis in aqueous media under basic
conditions,[12] in sharp contrast to most nanoparticles fre-
quently used as catalyst support, for example, magnetite@sil-
ica.[8]
To test the feasibility of our catch–release concept, we
synthesized nitrophenyl pyrene derivative 1,[15] which allows a
facile quantification of the adsorption of 1 on the Co/C
particles at room temperature and the level of desorption
from the surface at reaction temperature (1008C). Briefly, a
dispersion of Co/C nanoparticles in water was sonicated in the
presence of an excess of 1 (Scheme 1). The loading typically
obtained by the noncovalent grafting was assessed by means
of UV/Vis spectroscopy.[13] To this end, the nitrophenyl-
functionalized particles 2 were subjected to basic hydrolysis
(1m NaOH(aq), 12 h) upon which the concentration of the
generated nitrophenolate was determined against a likewise
hydrolyzed standard solution of 1 (Figure 1, top). Thus, it was
found that a loading of 0.2 mmolgÀ1 was established, a value
being almost twice as high as that obtained by covalent
modification.[13]
[*] Dipl.-Chem. S. Wittmann, Dr. A. Schꢀtz, Prof. Dr. O. Reiser
Institut fꢁr Organische Chemie, Universitꢀt Regensburg
Universitꢀtsstrasse 31, 93053 Regensburg (Germany)
Fax: (+49)941-943-4121
E-mail: oliver.reiser@chemie.uni-regensburg.de
Dr. R. N. Grass, Prof. Dr. W. J. Stark
To estimate the degree of desorption of 1 from the carbon
surface at elevated temperatures, the nitrophenyl-tagged
nanoparticles 2 in boiling water were filtered through a
sinter funnel, as a result 49% of the nitrophenyl moieties
dissociated from the surface of the nanobeads (Figure 1,
bottom). In an alternative procedure, repeated (four or eight
times) magnetic decantation of nanomagnets 2 from the hot
supernatant solution caused desorption of 60% and 76% of 1,
respectively. Importantly, under the same treatment with
Institut fꢁr Chemie- und Bioingenieurwissenschaften, Department
Chemie und Angewandte Biowissenschaften, ETH Zꢁrich HCI E 107
Wolfgang-Pauli-Strasse 10, 8093 Zꢁrich (Switzerland)
Fax: (+41)44-633-1083
[**] This work was supported by the Deutsche Forschungsgemeinschaft
(Re 948/8-1, “GLOBUCAT”) and the International Doktoranden-
kolleg NANOCAT (Elitenetzwerk Bayern).
Supporting information for this article is available on the WWW
Angew. Chem. Int. Ed. 2010, 49, 1867 –1870
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1867