Palladium(0) nanoparticles on glass-polymer composite materials as recyclable catalysts: A comparison study on their use in batch and continuous flow processes

Article abstract of DOI:10.1002/adsc.200700510

Palladium particles were generated by reduction of palladate anions bound to an ion exchange resin inside microreactors. The size and distribution of the palladium particles differed substantially depending on the degree of cross-linking and the density of ion exchange sites on the polymer/glass composites, the latter parameter having a larger influence than the former. The polymer phase of the composite materials was used for the loading with clusters composed of palladium particles which are 1 to 10 nm in diameter. The reactivity and stability of six different palladium-doped polymer/glass composite samples for transfer hydrogenations was investigated both under conventional and microwave heating in the batch mode as well as under continuous flow conditions using the cyclohexene-promoted transfer hydrogenation of ethyl cinnamate as a model reaction. Regarding the heating method it was found that catalysts that are composed of larger metal particles perform better under microwave irradiating conditions whereas samples with smaller particle sizes perform better under conventional heating. Comparing batch experiments with flow-through experiments the latter technique gives better conversion. Reusability was better in microwave heated experiments than in traditional heating.

Full text of DOI:10.1002/adsc.200700510

FULL PAPERS
DOI: 10.1002/adsc.200700510
Palladium(0) Nanoparticles on Glass-Polymer Composite
Materials as Recyclable Catalysts: A Comparison Study on their
Use in Batch and Continuous Flow Processes
a
b
c
d
d
Klaas Mennecke, Raul Cecilia, Toma N. Glasnov, Susanne Gruhl, Carla Vogt,
e
f
c,
b,
Armin Feldhoff, M. A. Larrubia Vargas, C. Oliver Kappe, * Ulrich Kunz, *
and Andreas Kirschning *
a,
a
Center ofBiomolecular Drug Research (BMWZ) and Institute ofOrganic Chemistry, Leibniz Universität Hannover,
Schneiderberg 1B, 30167 Hannover, Germany
Fax : (+49)-511-762-3011; e-mail: andreas.kirschning@oci.uni-hannover.de
Institute ofChemical Process Engineering (ICVT), Clausthal University ofTechnology, Leibnizstr. 17, 38678
b
Clausthal-Zellerfeld, Germany
E-mail: kunz@icvt.tu-clausthal.de
Christian Doppler Laboratory for Microwave Chemistry and Institute of Chemistry, Karl Franzens University,
c
Heinrichstrabe 28, 8010 Graz, Austria
E-mail: oliver.kappe@uni-graz.at
Institute ofInorganic Chemistry, Leibniz Universität Hannover, Callinstraße 9, 30167 Hannover, Germany
Institute ofPhysical Chemistry and Electrochemistry, Leibniz Universität Hannover, Callinstraße 3-3A, 30167 Hannover,
Germany
Chemical Engineering Department, University ofMalaga, Campus de Teatinos, 29071 Malaga, Spain
d
e
f
Received: October 23, 2007; Published online: March 17, 2008
Supporting information for this article is available on the WWW under http://asc.wiley-vch.de/home/.
Abstract: Palladium particles were generated by re- flow conditions using the cyclohexene-promoted
duction ofpalladate anions bound to an ion ex-
transfer hydrogenation of ethyl cinnamate as a
change resin inside microreactors. The size and dis- model reaction. Regarding the heating method it was
tribution of the palladium particles differed substan- found that catalysts that are composed of larger
tially depending on the degree ofcross-linking and
the density ofion exchange sites on the polymer/
metal particles perform better under microwave irra-
diating conditions whereas samples with smaller par-
glass composites, the latter parameter having a larger ticle sizes perform better under conventional heat-
influence than the former. The polymer phase of the ing. Comparing batch experiments with flow-through
composite materials was used for the loading with experiments the latter technique gives better conver-
clusters composed ofpalladium particles which are 1
sion. Reusability was better in microwave heated ex-
to 10 nm in diameter. The reactivity and stability of periments than in traditional heating.
six different palladium-doped polymer/glass compo-
site samples for transfer hydrogenations was investi- Keywords: enabling technologies; hydrogenation;
gated both under conventional and microwave heat- immobilization; microreactors; microwave heating;
ing in the batch mode as well as under continuous palladium
Introduction
cause oftheir large sur af ce-to-volume ratio. Usually,
catalytically active nanoparticles are prepared from a
metal salt, a reducing agent, and a stabilizer and are
The role ofmetal nanoparticles, particularly Pd(0)
clusters as active catalytic species has increased dra- supported on an oxide, charcoal, a zeolite or a
matically in recent years. This field, sometimes polymer. Physical methods for preparation, such as
termed “semi-heterogeneous catalysis”, is located at the electrochemical route developed by Reetz, are
[
2]
[3]
[4]
[
5]
[6]
[
7,8]
the frontier between homogeneous and heterogene- also numerous.
Currently, several diverse protocols
[1]
ous catalysis. Nanoparticles often offer higher cata- for preparing nanoparticles are being pursued such as
[
9]
[9,10]
lytic efficiency per gram than larger size materials, be- impregnation,
co-precipitation,
deposition/pre-
Adv. Synth. Catal. 2008, 350, 717 – 730
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
717

Klaas Mennecke et al.
FULL PAPERS
Figure 1. PASSflow reactor system as employed in the present work: a) Precipitation polymerization inside a megaporous
glass carrier; b) monolithic porous glass polymer composite Raschig rings inserted into a c) PASSflow reactor (rings are
aligned on a perforated tube; gaskets between the rings prevent bypass); d) housing is made of PEEK polymer. The flow
streams from an annular gap through the ring walls and irregular microchannels inside the glass/polymer composite and
leaves through the perforated inner tube (9 mm ring diameter).
[
11]
[9,12]
cipitation,
sol-gel,
gas-phase organometallic ofvolumetric lf ow rates thereby reducing mass trans-
[14] [15] [32–34]
[13]
deposition,
sonochemical,
micro-emulsion,
fer phenomena.
[16]
[17]
laser ablation, electrochemical, and cross-linking
modalities.
The catalytic flow system showed excellent proper-
ties in transfer hydrogenations of alkenes, alkynes,
[18]
Polymers provide stabilization for Pd(0) nanoparti- benzyl ethers and aromatic nitro compounds. Addi-
cles through the framework or by binding weakly to tionally, we demonstrated that the Pd species per-
the nanoparticle surface through a heteroatom that forms CÀC cross coupling reactions such as the
plays the role ofa ligand. In this context, poly( N- Suzuki–Miyaura, the Heck, and the Sonogashira reac-
[
31]
vinyl-2-pyrrolidone) has been employed for nanopar- tions under continuous flow conditions.
ticle stabilization and catalysis most frequently.
This pre-
liminary work had focussed on the scope of applica-
Other polymers have recently been used as efficient bility ofPd(0) species inside these continuous fl ow re-
[
18,19]
[
20]
supports for nanoparticle catalysis such as polyurea,
actors. At that point we speculated that the procedure
polyacrylonitrile and/or polyacrylic acid, polysilane had generated Pd(0) nanoparticles because ofthe
[
21]
[
22]
[23]
micelles with cross-linked shells,
polysiloxanes,
high reactivity ofthe catalytic system and its long li fe -
poly(N,N-dialkylcarbodi- time. In view ofthe importance ofligand- fr ee, Pd-cat-
[
24]
poly(4-vinylpyridine),
[25]
[26]
[27]
A
C
H
T
R
E
U
N
G
imide),
perbranched aromatic polyamides.
By taking advantage ofionic stabilization ofpalla-
polyethylene glycol,
c2h8]itosan
and hy- alyzed transformations in the industrial context, it is
[
of fu ndamental importance to relate the nature ofthe
polymeric support, its composition, its swelling prop-
dium(0) particles we demonstrated that these parti- erties and its texture with the performance, stability
cles can be generated after reduction of palladate and reusability ofmetal nanoparticles in catalytic
anions loaded on anion exchange resins [obtained by transformations for which hardly any information has
[29]
[
24b,35]
precipitation polymerization inside the void volume been accumulated so far.
In addition, it is un-
using chlormethylvinylbenzene known how the mode ofheating (thermal heating or
[
30]
ofmegaporous glass
VBC) and divinylbenzene (DVB) as cross-linker].
[
31]
(
microwave irradiation) as well as the type ofprocess
Beneficially, the deposition of palladium was carried (batch or continuous flow) chosen, effects nanoparti-
out on the surface of this monolithic glass/polymer cle-mediated catalysis. For investigating the influence
[
36]
composite material inside a PASSflow reactor, a con- ofmicrowave irradiation
on nanoparticle structure
tinuous flow system which was developed in our labo- and the catalytic process in general, the PASSflow re-
ratories (Figure 1). This reactor concept has the actor shown in Figure 1c was fitted into a commer-
unique feature that it can be operated at a wide range cially available single mode microwave reactor de-
718
asc.wiley-vch.de
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 717 – 730

Palladium(0) Nanoparticles on Glass-Polymer Composite Materials
FULL PAPERS
[
37]
signed for continuous flow processing.
Several exchange resin (see Figs. S1 and S2 ofSupporting In-
recent publications have demonstrated the feasibility formation).
and benefits of carrying out metal-mediated flow
chemistry under microwave conditions, although the
influence of microwave irradiation on the heterogene- Governing Factors for Pd Nanoparticle Size
ous metal catalyst itselfhas not been studied in
[38]
detail. One ofthe advantages ofmicrowave dielec-
As the Pd particle size will have a strong effect on the
tric heating over conventional heating in the work de- catalytic performance under classical thermal as well
scribed here is the fact that strongly absorbing materi- as microwave irradiating conditions, we first studied
als such as metal particles can be heated selectively in the influence of the flow rate during borohydride re-
[
36,39]
a reaction mixture.
It can therefore be imagined duction inside the composite material on the palladi-
that the intrinsic temperature localized around those um particle size. Loading resins at a low flow rate re-
particles is significantly higher than that of the bulk sulted in a shell-like distribution ofpalladate, whereas
[36,39]
solution.
In the development ofour reactor con-
high flow rate leads to a uniform palladium distribu-
cept (Figure 1) we incorporated the opportunity to se- tion. The palladium distribution can be seen from the
lectively heat (“activate”) the palladium nanoparticles black colour in broken composite rings (Supporting
by microwave irradiation since both the chosen reac- Information; Figure S1). Taking advantage of the
tor material (PEEK) as well as the porous glass poly- flow-through conditions with the composite Raschig
mer composite Raschig rings can be considered to be rings inside the new reactor concept (Figure 1) reduc-
[
40]
microwave transparent (see below).
Our detailed tion was performed under different flow conditions:
studies in this field are disclosed in this account which a) pure diffusive operation (batch with shaking, no
focuses on the catalytic performance of Pd nanoparti- forced flow through the pore volume of the samples),
À1
cles under different conditions and experimental set- b) an intermediate regime flow rate (10 mLmin )
ups.
where external diffusion controls the process and c)
pure kinetic conditions under flow operation
À1
(
110 mLmin ).
Results and Discussion
Reduction ofthe ion-exchanged palladate under
flow conditions yielded smaller Pd clusters with a nar-
rower particle size distribution and a better dispersion
throughout the whole polymeric phase compared to
Ion Exchange
Generating Pd(0) species inside the PASSflow reactor reduction under diffusion control as judged by trans-
is achieved by pumping an aqueous solution of mission electron microscopy (TEM) (Figure 2; Sup-
sodium tetrachloropalladate (Na PdCl ) through the porting Information; Figure S3) The fact that the
2
4
PASSflow reactor loaded with anion exchange resin 1 average particle size did not decrease after increasing
À1
(
chloride form). This process leads to instantaneous the flow conditions from 10 to 110 mLmin clearly
ion exchange (Scheme 1). However, the stoichiometry reveals a smaller diffusion limitation for the borohy-
ofthis process is not known as three possible species
dride than for the palladate anion. Consequently, a
À1
2a–c can be considered. Based on EPMA (electron flow rate of 10 mLmin was chosen as standard pro-
probe microanalysis) and EDXS (energy dispersive tocol for all reduction procedures. Also the nature of
X-ray spectroscopy) scanning profiles, which allowed the polymer composition has a pronounced influence
[
41]
us to determine both the palladium as well as chlorine on the Pd particle size. Thus, we prepared a set of
profiles inside the polymer/glass composite material, polymers which differed in the degree of cross-linking
we propose that ionic species 2a is present on the ion (5.3, 10, 20 and 30%wt DVB). These were functional-
Scheme 1. Formation ofPd(0) nanoparticles inside the PASS fl ow reactor.
Adv. Synth. Catal. 2008, 350, 717 – 730
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
719

Klaas Mennecke et al.
FULL PAPERS
[
42]
With reference to a mathematical model, it can
be assumed that during the precipitation polymeri-
zation process (Figure 1a) the polymeric particles
grow from a starting nucleus to which polymeric
layers are being added, leading to an onion-like shell
structure with a decreasing gradient ofcross-linking
from the core to the outer shell. The observed lack of
sensitivity to the content ofcross-linker DVB be-
tween 5.3 and 10% wt, the small change in the 20%
and finally, the pronounced shift in the 30% wt
sample could in principle be ascribed to the particle
morphology. Thus, a polymeric phase containing 30%
DVB also has rigid spaces between polymer chains in
the outer shell, preventing agglomeration and as a
consequence leading to reduced Pd cluster size.
An additional four polymeric samples were pre-
pared that differed in the density of cationic sites. We
expected that site isolation will reduce the process of
agglomeration ofPd particles upon of rmation as well
during catalytic transformations. These samples were
prepared by adding styrene to the mixture ofmono-
Figure 2. Pd cluster size distribution collected from TEM
micrographs (see also Supporting Information; Figure S3)
(
polymer: VBC/DVB (5.3%); loaded samples by diffusion mers thereby gradually exchanging VBC by the inert
based on 70 Pd particles; and loaded samples under flow spacer styrene. The molar ratio ofstyrene and VBC
conditions based on 50 Pd particles.
was 1:1, 10:1 and 30:1 (for detailed polymer composi-
tion see Supporting Information, Table S2). However,
the properties ofthe polymeric samples with a high
degree ofcross-linking (30% DVB) were only uf rther
ized with Pd particles under flow conditions. Again, studied for the sample styrene/VBC 1:1 as TEM anal-
the cluster size distribution was analyzed by TEM yses for all other highly cross-linked samples turned
(
Figure 3). The particle sizes did not differ for sam- out to be inconclusive, which hampered precise judge-
ples containing 5.3 and 10% DVB, a small shift to ments.
smaller clusters was encountered for the 20% DVB
Spacing the ion exchange sites in samples contain-
sample and a large change can be noted when the ing 5.3% DVB (styrene/VBC molar ratio from 0:1 to
degree ofcross-linking was raised to 30% DVB.
1:1) results in dilution ofthe large Pd agglomerates
Figure 3. Left: TEM bright-field images of Pd clusters of different polymeric phases (5.3, 10, 20 and 30% DVB crosslinker)
prepared under intermediate regime flow conditions. Scale bars indicate 100 nm. Right: Pd cluster size distribution based on
50 Pd clusters.
720
asc.wiley-vch.de
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 717 – 730

Palladium(0) Nanoparticles on Glass-Polymer Composite Materials
FULL PAPERS
(
clusters) in a less dense environment ofindividual Pd
At this point a general trend can be drawn. By in-
nanoparticles (see Figure 4). The TEM micrographs ceasing the styrene content smaller clusters and small-
at higher magnification clearly reveal that the large er nanoparticles are generated.
Pd agglomerates are in fact a dense spherical packing
ofsmaller Pd nanoparticles. This is particularly well
documented for the “diluted” samples. For samples Selective Heating Using Microwave Irradiation
with 5.3% DVB, the increase ofstyrene content leads
to larger distances between the ion exchange sites The ability ofa speci fc material to convert micro-
thereby resulting not only in the dilution ofthe small
agglomerates but in an overall decrease ofthe Pd
nanoparticle size as is shown in Figure 5.
wave energy into heat at a given frequency and tem-
perature is determined by the so-called loss tangent
(tan d), expressed as the quotient, tan d=e’’/e’, where
e’’ is the dielectric loss, indicative of the efficiency
with which electromagnetic radiation is converted
into heat, and e’ is the dielectric constant, describing
the ability ofmolecules to be polarized by the electric
[
36]
field. A reaction medium with a high tan d at the
standard operating frequency of a microwave reactor
(
2.45 GHz) is required for good absorption and, con-
sequently, for efficient heating. As indicated above
our reactor design should in principle allow for the se-
[36]
lective heating ofthe strongly microwave absorbing
palladium nanoparticles by microwave energy, since
both the chosen reactor material (PEEK) as well as
the porous glass polymer composite Raschig rings
should be microwave transparent. In order to verify
this hypothesis we have conducted an extensive set of
microwave irradiation experiments with all the mate-
rials used in the reactor construction. The difference
in the microwave absorbance ofthe various compo-
site materials was readily demonstrated by comparing
heating profiles when suspended in a microwave
Figure 4. TEM micrographs ofsamples without styrene ( left)
and with a 1:1 molar ratio ofstyrene and VBC ( right).
[36]
transparent solvent
such as carbon tetrachloride.
For these experiments carried out in a batch micro-
wave reactor, a fully microwave transparent quartz re-
action vessel and an accurate internal fiber-optic tem-
[
43]
perature probe was utilized. Exposure ofa stirred
sample ofcarbon tetrachloride to 100 W constant
magnetron output power for 5 min led to virtually no
detectable heating ofthe solvent (Figure 6, curve a).
A small heating effect raising the bulk temperature of
the mixture was seen under identical conditions for a
sample ofPEEK suspended in the microwave trans-
parent solvent (curve b), and for one of the standard
glass Raschig rings (without polymer and Pd) dis-
played in Figure 1b (curve c), confirming the more or
less microwave transparent nature ofthe base reactor
materials. In contrast, a suspension ofone monolithic
glass/polymer composite Raschig ring possessing
anion exchange functionality [that is, 1, Figure S1 in
Supporting Information (VBC, 5.3 wt% DVB)] dem-
onstrated significant absorbance of microwave energy
and conversion into heat (curve d), probably as a
result ofthe ionic character ofthe polymer matrix, in-
[36]
voking an ionic conduction heating mechanism. Im-
portantly, however, under the same conditions the
Figure 5. Pd cluster size distribution (based on 40 Pd parti-
cles) ofsamples with de if rf ent styrene content collected
from TEM pictures (see Supporting Information; Fig- palladated Raschig rings 3 proved to be strongly mi-
ure S6).
crowave absorbing raising the bulk temperature of
Adv. Synth. Catal. 2008, 350, 717 – 730
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
721

Klaas Mennecke et al.
FULL PAPERS
Figure 6. Microwave heating profiles for materials suspended in carbon tetrachloride. Experiments were carried out using a
single-mode microwave instrument in a 10 mL sealed quartz vessel with magnetic stirring (fiber-optic temperature measure-
ment) using 100 W constant magnetron power output for 5 min. Profile a: pure carbon tetrachloride. Profile b: PEEK
sample suspended in carbon tetrachloride. Profile c: one glass Raschig ring (cf. Figure 1b) suspended in carbon tetrachloride.
Profile d: glass/polymer composite 1 (5.3% DVB, 1:1 styrene/VBC) suspended in carbon tetrachloride. Profile e: palladated
glass/polymer composite 3 (5.3% DVB, 1:1 styrene/VBC) suspended in carbon tetrachloride.
[31]
the microwave transparent solvent from room temper- lent activity in CÀC cross-coupling reactions,
ature to 708C within about one minute ofirradiation tried to keep the set-up simple thus avoiding the use
curve e). This dramatic heating effect is clearly the oftwo organic coupling partners. There of re, in the
we
(
consequence ofthe selective absorbance ofmicro-
present study, the transfer hydrogenation of ethyl cin-
wave irradiation by the metal particles on the poly- namate 4 to dihydrocinnamate 5 served as benchmark
meric support. It should be stressed that the measured reaction (Table 1). In principal, four modes of con-
temperature of ca. 708C merely reflects the bulk tem- ducting this transformation were studied in detail
perature ofthe solvent and not the intrinsic tempera-
ture localized around the palladium particles which thermal heating as well as b) MW irradiation in a
may be significantly higher. batch wise operation and continuous flow hydrogena-
which are transfer hydrogenations at 708C with a)
We have additionally recorded the heating profiles tions again with c) thermal as well as d) MW irradia-
for two other solvents under constant 100 W micro- tion. In this context, it is ofimportance to understand
wave irradiation conditions, namely for the weak mi- how the nature and local environment ofthe Pd(0)
crowave absorber toluene (tan d=0.040) and the particle on the polymer support is correlated with its
strong absorber ethanol (tan d=0.941). As expected, reactivity and thus affects the efficiency of the hydro-
toluene proved to be significantly less microwave ab- genation process under the four differing process con-
sorbing than the palladated Raschig ring/carbon tetra- ditions. As a result a correlation between the metal
chloride suspensions, whereas ethanol was significant- particle size (before and after several runs) and the
ly stronger absorbing (data not shown).
efficiency of the hydrogenation can be drawn. Fur-
thermore, the undoubtedly important role ofthe sol-
vent under MW irradiation was investigated.
Catalytic Transfer Hydrogenations
In the initial experiment the transfer hydrogenation
was carried out under batch as well as continuous
Having established the factors that govern the distri- flow conditions [0.072M 4 in EtOH/cyclohexene
À1
bution on the polymeric phase and the particle diame- (1:1), 708C; flow rate=2 mLmin for flow through
ter ofpalladium particles, as well as having studied
the suitability ofthe PASS lf ow reactor under micro-
mode] using Raschig rings with a polymeric composi-
tion of5.3% DVB as cross-linker and only VBC as
wave irradiating conditions, we focussed on the cata- major second monomer. The flow-through reaction
lytic performance of these nanoparticles. Although we was carried out with glass/polymer composite rings
have shown that these Pd nanoparticles show excel- which were incorporated into a type ofPASS lf ow re-
722
asc.wiley-vch.de
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 717 – 730

Palladium(0) Nanoparticles on Glass-Polymer Composite Materials
FULL PAPERS
Table 1. Transfer hydrogenation of ethyl cinnamate under batch and continuous flow conditions at 708C bulk temperature
5.3% DVB, VBC; palladate loading and reduction were conducted under two different conditions: sole diffusion and kinet-
(
ic regime).
Batch/flow Solvent
mode
Reaction time (h) until full transformation
Pd loading at a flow rate of
Pd loading by pure diffusion,
conventional heating
Pd loading by pure
diffusion, MW
À1
10 mLmin , conventional heating
Batch
Flow
Batch
Flow
EtOH/cy-
clohexene
EtOH/cy-
clohexene
Tol/cyclo-
hexene
14 h
14 h
2 h
14 h
3 h
3.25 h
3.75 h
3.75 h
4.5 h
2 h
Tol/cyclo-
hexene
24 h
48 h
actor shown in Figure 1c. In Figure 7a (conventional bulk temperature) perform better under continuous
heating) ten experiments are depicted, which were flow conditions than in the batch mode (Figure 7c
carried out under identical conditions except that five and Figure 7d). Now, polymeric backbones that ac-
different composite materials were employed. The re- commodate larger Pd nanoparticles show better cata-
sults are normalized with respect to the Pd content of lytic reactivity compared to those that are loaded
the samples (see Supporting Information; Figure S7). with smaller ones. However, the trends noted for the
They clearly demonstrate that the catalyst activity experiments conducted with conventional heating pre-
changes with the nature ofthe polymeric phase which
itselfdetermines the Pd cluster and particle size (see
also Figure 5) In essence, dilution ofthe Pd clusters
vail in the same manner, namely that continuous flow
conditions give better results for transfer hydrogena-
tions than the batch mode and that ethanol is the sol-
with respect to the individual nanoparticles by adding vent ofchoice in comparison to toluene. Importantly,
styrene as a spacer monomer enhances the reactivity Figure 7a-d clearly reveal that microwave irradiation
independent from the flow method (e.g., compare leads to an improved performance irrespective of the
double column 1 with double column 2 in Figure 7a). Pd-doped polymeric sample employed and irrespec-
Furthermore, the smaller the Pd particle size the tive ofwhether the hydrogenation is pe rof rmed in
higher the reactivity ofthe catalyst in these trans ef r
hydrogenations regardless ofwhether batch or contin-
batch or flow mode. Again, we would like to stress
that here conventional and microwave heating results
uous flow conditions are employed (e.g., compare are difficult to compare as far as the genuine “reac-
double column 2 with 3 in Figure 7a). As a general tion temperature” is concerned due to the selective
trend in heterogeneous catalysis this fact is expected heating ofthe metallic nanoparticles by micro-
[
39,40]
in terms ofthe great increase in the speci if c catalytic
active surface area of Pd. Additionally, continuous
waves.
The inversion in the reactivity pattern with respect
flow commonly results in better performance com- to particle size and heating method already described
pared to batch mode hydrogenations due to the in- could be associated with the special characteristics of
crease in the mass transfer processes of the reactants. a microwave irradiated process. Employing micro-
The relationship between Pd particle size and reactiv- wave dielectric heating, the power converted to heat
ity is similar when transfer hydrogenations were con- in an irradiated particle is proportional to the volume
[
44]
ducted in toluene although transfer hydrogenation ofthe particle
whereas the heat transferred from a
proceeds more efficiently in ethanol (Figure 7b). particle to the environment is proportional to the
When switching from conventional to microwave heat exchange surface area. Hence, bigger particles
heating transfer hydrogenations at 708C (measured have a bigger volume to surface area ratio than small-
Adv. Synth. Catal. 2008, 350, 717 – 730
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
723

Klaas Mennecke et al.
FULL PAPERS
Figure 7. Transfer hydrogenations of ethyl cinnamate (4) in batch as well as flow mode (refer to Table 1; abscissa: for exam-
ple, 5.3 (1:1) re fe rs to a polymeric sample composed of5.3% DVB cross-linker and a 1:1 ratio ofVBC and sytrene). Reactiv-
ity profiles (mol product/ming Pd) of different Pd(0)-doped polymeric phases in transfer hydrogenations of cinnamate 4. Fig-
ure 7a and Figure 7b refer to conventional heating (708C) in batch as well as flow mode in ethanol and toluene, respectively.
Figure 7c and Figure 7d) refer to the corresponding set of experiments under MW irradiation conditions (bulk temperature
708C).
er particles, thereby increasing the heat accumulated usability ofthe catalytic samples with a high palladi-
in the particles and as a result providing a hypotheti- um content. However, samples with low palladium
cal higher hot spot effect. Such an effect, which has loading are deactivated rather rapidly (curves high-
not been possible until now to be theoretically mod- lighted with arrows). The same trend is observed with
[39]
elled, is under current investigation in our laborato- toluene as solvent (Figure 8b) even though deactiva-
ries.
tion only occurs after more runs (curves highlighted
with arrows). A different behaviour was observed
under microwave irradiating conditions. All samples
displayed a stable performance over ten runs in etha-
nol as well as in toluene. The behaviour ofthe non-di-
luted sample 5.3 (0:1) is out ofthis general trend,
since a clear deactivation is observed after the sixth
ReusabilityStudies and Aging
a) ReusabilityStudy
Having established the performance of individual Pd- run. This might be explained by the higher energy
doped polymer samples under different conditions, conversion stress suffered by the Pd particles, since in
we then investigated the reusability ofthese catalyst
a non-microwave absorbing solvent all the heat to
samples, starting with batch operations. Conditions reach the desired bulk temperature comes from the
were maintained as described in the section “Catalytic conversion ofthe whole power input solely absorbed
Transfer Hydrogenations” and the results for up to by the Pd particles.
ten runs are summarized in Figure 8a to d. Using eth-
Deactivation ofnoble metal-supported heterogene-
anol as solvent (Figure 8a) results in an improved re- ous catalysts can occur by a) sintering ofmetal crys-
724
asc.wiley-vch.de
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 717 – 730

Palladium(0) Nanoparticles on Glass-Polymer Composite Materials
FULL PAPERS
Figure 8. Performance of Pd-doped polymers after repeated transfer hydrogenations of 4 (batch mode).
tallites, b) formation of side products which poison crowave irradiating conditions. The fact that no cata-
the metal surface, c) chemical changes involving the lytic activity loss was found for the same sample indi-
metal (e.g., oxidation, leaching) and d) chemical and cates that no sintering ofmetal crystallites had oc-
physicochemical changes involving the support. curred.
Kralik et al. reported a possible deactivation caused
by the combination ofthe hydrogenolysis ofthe poly-
[
45]
mer backbones and a Pd leaching process.
In the b) Aging of Nanoparticles
present case, degradation ofthe polymeric phase can
be neglected. In addition, formation of poisoning side Commonly, fresh Raschig rings contain Pd particles
products has never been reported for this kind of which slightly differ in size, depending how loading
transfer hydrogenations. The amount of leaching of was performed (diffusive 4–5 nm; convective 6 nm).
Pd was analyzed for all batch experiments (thermal When transfer hydrogenations were performed under
and microwave heating) and all flow-through experi- batch conditions the particles did not change in size,
th
ments (thermal conditions). We commonly collected even not after the 10 run. Only the agglomerates
th
samples after the 5 run for determining the degree slightly increase during the catalytic process. MW irra-
ofPd leaching by ICP-MS. We of und that leaching
was commonly very low (between 0.002 and 0.1% Pd) cle size after the 10 run (diffusive 6 nm; convective
with reference to the amount of Pd initially employed 6–10 nm). In contrast to Pd particles that have been
diation leads to a slight increase ofaverage Pd parti-
th
(
see Supporting Information; Table S4). loaded under convective conditions, the diffusively
Therefore, we also conducted analytical studies on loaded samples show substantial increases in particle
determining Pd particle size changes after catalysis as well as cluster size. This can be ascribed to the fact
utilizing STEM analysis. In Figure 9 a comparison be- that the Pd particles are almost exclusively located in
tween fresh samples (5.3% DVB; 0:1) of composite the outer sphere of the Raschig rings when diffusion
material and the aged samples after the tenth run controlled loading had been conducted (see Figure S1
(
both heating methods in ethanol) are depicted. A in Supporting Information) which leads to higher
slight increase in the cluster size can be observed for local concentrations. In contrast, loading under con-
the sample which had been heated under convention- vective flow conditions leads to more even distribu-
al conditions, while almost no agglomeration was tion throughout the rings. Changes in particle as well
found in the sample which had been heated under mi- as cluster size during transfer hydrogenation are mar-
Adv. Synth. Catal. 2008, 350, 717 – 730
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
725

Klaas Mennecke et al.
FULL PAPERS
Figure 9. Dark-field STEM pictures of the non-diluted composite before hydrogenation and after 10 runs.
ginal irrespective ofwhether toluene or ethanol
serves as solvent.
fects. At low Pd content the amount of microwave
energy that can be absorbed and dissipated by the
The current data do not give a conclusive and ho- catalyst drops. However, an inversion ofthe reactivity
mogeneous picture. However, the following trends with the particle diameter changing from traditional
can be summarized: a) Higher degrees ofcross-link-
heating to microwave heating was observed. That
ing lead to Pd species with lower catalytic activity. means that smaller particles are more active in tradi-
This, however, results in glass-polymer composite ma- tional heating whereas bigger particles perform better
terials that show prolonged activity. b) In contrast, in microwave heating. This can be explained by hot
polymers with smaller degrees ofcross-linking or
higher dilution ofPd particles on their sur fa ces (using
spot effects which depend on the particle diameter.
The bigger the size, the higher the possible tempera-
styrene as third monomeric component) show higher ture gradient between particle and bulk liquid be-
[
46]
catalytic activity which more rapidly erode with each comes.
A different behaviour from the aspect of
run.
catalyst reusability was observed for different heating
methods. Low to intermediate active site dilution is
beneficial if the catalyst is conventionally heated.
Under microwave heated samples no loss ofcatalytic
activity was found.
Conclusions
In summary, it was demonstrated that the PASSflow
Flow chemistry in conjunction with microwave
[
47]
technique can be successfully utilized in a catalytic heating is still in its infancy. Because ofthe limited
process using transition metal catalysts. This study re- penetration depth ofmicrowave irradiation into ab-
veals that this technique can be extended to micro- sorbing media, the scale-up ofmicrowave-assisted re-
wave heated operations. The influence of catalyst dis- actions is inherently problematic using conventional
tribution inside the composites was investigated for batch processing. Using microwaves, the solution to
the transfer hydrogenation of ethyl cinnamate as a the scale-up problem therefore has to involve a flow
model reaction. It was observed that the reactivity of process and we are currently exploring different strat-
the catalysts prepared is different in traditional com- egies realizing this process.
pared to microwave heating. The reason for the dif-
ferent performance is related to the presence of two
effects. On one hand a high active site dilution produ-
ces a decrease in the nanoparticle size leading to a
Experimental Section
higher surface area and coupled with this to a higher
catalytic activity. On the other hand the same dilution
General Methods
produces a decrease in the Pd content ofthe catalysts
TEM measurements were conducted on a Philips CM 200
leading to a small rate enhancement by hot spot ef- microscope at 200 kV. STEM micrographs were recorded on
726
asc.wiley-vch.de
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 717 – 730

Palladium(0) Nanoparticles on Glass-Polymer Composite Materials
FULL PAPERS
a
JEOL JEM-2100F-UHR microscope at 200 kV in
more. Reduction was achieved by addition ofa solution of
HAADF contrast. Samples were collected after crushing
composite Raschig rings into fine powder, dispersion of the
material in a volatile liquid (ethanol) and finally subjecting
it to a fine mesh carbon net which is electron transparent.
EPMA-EDXS scanning was carried out with a Cameca SX
0.2M NaBH (6 g, 158.6 mmol) in water until gas evolution
4
stopped. The rings were washed with water (500 mL) and
methanol (500 mL) before gently shaking them in a solution
of2N HCl (510 mL). A tf er washing the rings with water
(until pH 7) and methanol the rings are dried in high
vaccum.
100 Microprobe analyser which had a Cambridge CS44
SEM camera incorporated in combination with an EDAX
spectrometer. Ion exchange measurements ofpalladate were
achieved by UV-VIS spectroscopy ofthe palladate solution
using a Hach DR 4000V Spectrophotometer (430 m wave-
Flowprocedure: Loading under flow conditions was typi-
cally carried out by incorporation o fi vf e Raschig rings
inside a stainless steel PASSflow reactor (reactor set-up de-
picted in Figure 1c). The palladate ion exchange was opti-
mized at different flow rates (10–110 mL). The graphic that
describes the initial rate ofion exchange vs. flow rate re-
length). NMR spectra were recorded with a Bruker DPX-
1
4
(
00 spectrometer at 400 MHz ( H NMR) and at 125 MHz
1
3
À1
C NMR) in CDCl using tetramethylsilane as the internal
veals that at 110 mLmin the ion exchange rate is no
3
standard. Mass spectra (EI) were obtained at 70 eV with a
type VG Autospec apparatus (Micromass). GC analyses
were conducted using a Hewlett–Packard HP 6890 Series
GC System equipped with a SE-54 capillar column (25 m,
Macherey–Nagel) and a FID detector 19231 D/E. Flash
column chromatography was performed on Merck silica gel
longer dependent on the flow rate (Supporting Informa-
tion). As a result under these flow conditions external diffu-
sion can be neglected. The loading procedures for preparing
the various catalyst samples were conducted stepwise. For
the polymeric samples that are composed of5.3%, 10%,
20% and 30% DVB (styrene/VBC=0:1) and 5.3% DVB
(styrene/VBC=1:1) ion exchange was achieved within 2 h
by circulating an aqueous solution (0.028M; 1:1 ratio ion ex-
change capacity and palladate) ofNa Cl Pd·3H O (refer
60 (230–400 mesh). Commercially available reagents and
dry solvents were used as received. All reactions were car-
ried out in a continuous flow apparatus from Chelona
GmbH, Potsdam.
2
4
2
also to Table 1 ofthe Supporting In of rmation) through the
reactor. Except for the kinetic studies, the flow rate was
À1
kept at 10 mLmin . In the following, the reactor was flush-
Reactor Concept
ed for one hour with distilled water at a flow rate of
À1
Rings are aligned on a perforated tube inside the reactor.
The spaces between the rings were sealed with KALREZꢁ
or VITONꢁ polymer. These gaskets prevent bypass. The
whole structure forces the fluid to flow from an annular gap
through the ring walls and leaving the reactor through the
5 mLmin before a solution of borohydride in water (1.5 g,
39.65 mmol; 200 mL) was pumped through the PASSflow re-
À1
actor at a flow rate of 10 mLmin (except for the kinetic
studies). The loading was terminated by three washing
À1
steps: (a) distilled water, 1 h, 5 mLmin ; (b) 100 mL 2M
[33]
À1
À1
perforated inner tube. For a detailed description see ref.
HCl, 9 mLmin and (c) distilled water, 9 mLmin until the
pH at the outlet was determined to be 7. For the polymeric
samples that are composed of5.3% DVB (styrene/VBC =
Composite Materials
1
0:1) and 30% (1:1) the loading procedure was as follows:
All samples that were prepared without the use ofstyrene
as monomer were obtained by copolymerization ofvinyl-
benzyl chloride (VBC, 90% purity), in the presence ofdi-
vinylbenzene (DVB, 65% purity) as cross-linking monomer
À1
(
a) ion exchange within 2 h by circulating (10 mLmin ) a
solution ofTHF/water (0.028M; 1:1 ratio ion exchange ca-
pacity and palladate) ofNa Cl Pd·3H O; b) THF/water, 1 h,
2
4
2
À1
5
2
5
9
mLmin ; c) borohydride solution in THF/water (0.75 g,
0 mmol; 100 mL), 10 mLmin ; (d) THF/water, 1 h,
mLmin ; (e) 11 mL 2m HCl in 100 mL THF/water,
mLmin and finally (f) rings washed in batch with a mix-
(
5.3 to 30% wt) using precipitation polymerization inside a
À1
fractal porous Raschig ring. In contrast, all catalysts that
also contain styrene as monomer were prepared by altering
the above-mentioned procedure by additionally adding sty-
rene (99% purity) to the polymerization mixture. The
amount in mass percent and mol percent used for the prepa-
ration of the different polymerization mixtures are listed in
Table 1 ofthe Supporting In of rmation. All polymer samples
obtained by precipitation polymerization were functional-
ized in the following by amination with triethylamine in tol-
uene which afforded the corresponding polymers functional-
ized with quaternary ammonium chloride sites.
À1
À1
ture THF/water for 12 h. As solvent mixtures THF/water
3:1) for the samples 5.3% DVB (styrene/VBC=10:1) and
THF/water (1:3) for the sample 30% DVB (styrene/BVC=
:1) turned out best suited.
(
1
Transfer Hydrogenation in Batch and Continuous
Flow Modes under Conventional Heating
Batch Mode: In a typical experiment a 10-mL reaction
vessel was filled with 5 mL of the appropriate solvent mix-
ture EtOH:cyclohexene=1:1 or toluene:cyclohexene=1:1,
two palladated Raschig rings 3 (Table 1; Supporting Infor-
mation) and 60 mL (0.36 mmol) ofethyl cinnamate under ni-
trogen. The vial was heated to 708C and shaken. After the
time listed in Table 1, the rings were removed by filtration,
rinsed with ethyl acetate and the combined organic phases
were removed under reduced pressure. Typically, a colour-
less oil (yield: 64 mg; 100%) was collected. The rings were
Palladium Loading Procedures
Batch procedure: Raschig rings (100 g; 5.3% DVB, load-
ing=0.17 mmol/ring) were washed with 2N HCl (510 mL)
for 0.5 h and water until the eluent showed pH 7. Then the
Raschig rings were shaken in a 0.03M solution ofsodium
tetrachloropalladate (5 g, 15.3 mmol) in water under N (O
exclusion) atmosphere until the colour ofthe solution had
turned from red to yellow. The rings were washed with
water (1.7 L) until no palladate salts leave the rings any-
2
2
Adv. Synth. Catal. 2008, 350, 717 – 730
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
727

Klaas Mennecke et al.
FULL PAPERS
dried under reduced pressure and stored under nitrogen at-
mosphere.
stant magnetron output power for 5 min. The fibre-optic
probe was introduced into the reaction vessel protected by a
sapphire immersion well. When obtaining the temperature
Continuous Flow: The PASSflow reactor (Figure 1c) was
charged with three palladated Raschig rings 3 (Table 1), and
connected to the HPLC pump. A freshly prepared mixture
ofEtOH:cyclohexene or toluene:cyclohexene (50 mL, 1:1)
was used as a stock solution. Initially, the pump was primed
by injecting this reaction solvent mixture while maintaining
profiles of the solid materials only CCl was employed as
4
solvent media. The experiments were performed as de-
scribed above by adding one piece of: PEEK, Raschig rings,
polymer composite 1 and polymer composite 3 (5.3% DVB,
1:1 styrene/VBC).
À1
a flow rate of 2 mLmin . During the period of lf ow stabili-
zation, the PASSflow reactor was checked for leakage and
kept outside ofthe oven. The reactor was subsequently in-
troduced into the oven while the pumping ofsolvent lf ow is
stopped. Thereafter, the inlet was switched to the actual re-
action mixture ofEtOH:cyclohexene or toluene:cyclohex-
ene (14 mL, 1:1), containing 90 mL (0.53 mmol) ofethyl cin-
Transfer Hydrogenation in Batch and Continuous-
Flow under Microwave Irradiation
Batch Mode Microwave Processing: In a typical experiment
a 10-mL Pyrex reaction vessel was filled with 5 mL of the
appropriate solvent mixture EtOH:cyclohexene=1:1 or tol-
uene:cyclohexene=1:1, two palladated Raschig rings 3
(Table 1; Supporting Information) and 60 mL ofethyl cinna-
mate. The vial was subsequently sealed and irradiated for
30 min at a set temperature of70 8C (IR control) and con-
stant stirring ofthe mixture. A ft er 30 min, an aliquot sample
was taken and the degree oftrans ef r hydrogenation deter-
mined by HPLC-analysis at 215/254 nm.
À1
namate, while maintaining a flow rate of 2 mLmin (reac-
tion times are listed in Table 1). An aliquot sample was
taken and the degree oftrans ef r-hydrogenation was deter-
mined by GC-analysis. The reactor was washed with the sol-
vent mixture (20 mL) and the product was collected after re-
moval ofthe combined solvents under reduced pressure.
Prior to the next reaction the PASSflow reactor was washed
by switching the inlet to a reservoir with EtOAc for 10 min
and then to the reservoir with the reaction solvent mixture
without ethyl cinnamate for another 10 min.
Continuous FlowMicro wa ve Processing:
The PASSflow
reactor (Figure 1c) was charged with two palladated Ra-
schig rings 3 (Table 1), tightened and connected to the CEM
Voyager module and HPLC pump. A freshly prepared mix-
ture ofEtOH:cyclohexene or toluene:cyclohexene (50 mL,
Spectroscopic Data for Ethyl Dihydrocinnamate 5
1
:1) was used as a stock solution. Initially, the pump is
1
H NMR (200 MHz, CDCl , CDCl =7.26 ppm): d=7.28 (m,
primed by injecting this reaction solvent mixture while
3
3
À1
5
8
3
7
H, 2-H, 6-H), 4.16 (q, J=7.1 Hz, 2H, 10-H), 2.99 (t, J=
maintaining a flow rate of 2 mLmin . During the period of
.1 Hz, 2H, 7-H), 2.66 (t, J=8.1 Hz, 8-H), 1.27 (t, J=7.1 Hz,
flow and back pressure stabilization, the PASSflow reactor
is checked for leakage and kept outside of the microwave
cavity. The reactor is subsequently introduced into the
cavity while the pumping ofsolvent lf ow is stopped. There-
after, the reaction parameters (608C, 30–50 W magnetron
power, 2 min ramp time and 180 min hold time) are loaded
and the inlet is switched to the actual reaction mixture of
EtOH:cyclohexene or toluene:cyclohexene (13 mL, 1:1),
containing 90 mL ofethyl cinnamate, while maintaining a
13
H, 11-H);
C NMR (100 MHz, CDCl3, CDCl =
3
7.36 ppm): d=173.2 (q, C-9), 140.8 (q, C-1), 128.7 (t, C-2,
C-6), 128.6 (t, C-3, C-5), 126.5 (t, C-4), 60.7 (s, C-10), 36.2 (s,
C-8), 31.3 (s, C-7), 14.5 (p, C-11); GC-MS(EI): m/z (%)=
+
1
78 (20) [M ].
Microwave Experiments
À1
All microwave experiments were performed with a single-
mode CEM Discover/Voyager microwave instrument at
flow rate of 2 mLmin . The flow is maintained in a closed
loop (the outlet is introduced into the reaction mixture re-
servoir) for 180 min. An aliquot sample was taken and the
degree oftrans ef r-hydrogenation was determined by HPLC
analysis at 215/254 nm. Prior to the next reaction the PASS-
flow reactor was washed by switching the inlet to a reservoir
with EtOAc for 10 min and then to the reservoir with the re-
action solvent mixture without ethyl cinnamate for another
2450 MHz controlled irradiation, using standard Pyrex vials,
custom-made high purity quartz vessels or the appropriate
PASSflow reactor (Figure 1c). The temperature profiles of
the solvents were monitored by using a fibre-optic probe.
When performing PASSflow experiments the fiber-optic
probe was mounted on the outside wall ofthe PASS fl ow re-
TM
actor with a thin Teflon tape. To maintain stable power
1
0 min. For a picture ofthe reactor set-up re ef r to the Sup-
and temperature during the reaction, the microwave cavity
and outside surface of the PASSflow reactor was actively
cooled by compressed air to remove latent heat. It should
be noted that it was only possible to monitor the tempera-
ture ofthe outside reactor wall corresponding roughly to
the temperature ofthe bulk reaction mixture and not the
actual temperature near the metal particles inside the reac-
tor.
porting Information.
Stability/Reactivity Studies
Microwave conditions for batch and continuous flow experi-
ments were used as described above. A matrix often runs in
sequence was performed with all six types of palladated
Rashig rings 3 under batch as well as under continuous flow
conditions. The degree oftrans ef r-hydrogenation was deter-
mined by HPLC-analysis at 215/254 nm.
Heating Profiles of Materials/Solvents (Figure 6)
In a typical experiment a 10-mL quartz reaction vessel con-
taining a magnetic stir bar was filled with 5 mL of the ap-
Kinetic Studies with Palladated Raschig Rings
propriate solvent (CCl , EtOH, toluene) or solvent/material
mixture. The vial was closed and irradiated with 100 W con-
Thermal conditions for batch and continuous flow experi-
ments were used as described above. An aliquot sample was
4
728
asc.wiley-vch.de
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 717 – 730

Palladium(0) Nanoparticles on Glass-Polymer Composite Materials
FULL PAPERS
taken at periods of30 min reaction time until the trans ef r
hydrogenation was complete or no further improvement of
the conversion could be monitored. The degree oftrans ef r
hydrogenation was determined by GC analysis.
[6] a) M. T. Reetz, W. Helbig, J. Am. Chem. Soc. 1994, 116,
7401–7402; b) M. T. Reetz, S. A. Quaiser, Angew.
Chem. Int. Ed. Eng. 1995, 34, 2240–2241; Angew.
Chem. 1995, 107, 2461–2463; c) M. T. Reetz, R. Brein-
bauer, K. Wanninger, Tetrahedron Lett. 1996, 37, 4499–
4
502; d) M. T. Reetz, G. Lohmer, Chem. Commun.
1
996, 1921–1923.
[
7] a) Nanoparticles, (Ed.: G. Schmid), Wiley-VCH, Wein-
heim, 2004; b) R. G. Finke, in: Metal Nanoparticles.
Synthesis Characterization and Applications, (Eds.:
D. L. Feldheim, C. A. Foss Jr.), Marcel Dekker, New
York, 2002, chap. 2, p 17; c) Nanoparticles and Nano-
structured Films. Preparation, Characterization and Ap-
plications, (Ed.: J. H. Fendler), Wiley-VCH, Weinheim,
Acknowledgements
Financial support for this work was given by the DFG (grant
Ki 397/6–1 and KU 853/3–1), the Austrian Science Fund
(
FWF, grant I18-N07), the Fonds der Chemischen Industrie
and in part by the Max-Buchner-Stiftung for which we are
grateful.
1
998; d) G. Schmid, in: Nanoscale Materials in Chemis-
try, (Ed.: K. J. Klabunde), Wiley-Interscience, New
York, 2001, p 15.
[
8] Recent reviews on catalytic applications: a) M. A. El-
Sayed, Acc. Chem. Res. 2001, 34, 257–264; b) R. M.
Crooks, M. Zhao, L. Sun, V. Chechik, L. K. Yeung,
Acc. Chem. Res. 2001, 34, 181–190; c) T. Yonezawa, N.
Toshima, Polymer- Stabilized Metal Nanoparticles:
Preparation, Characterization and Applications: in: Ad-
vanced Functional Molecules and Polymers, (Ed.: H. S.
Nalwa), Accelerated Development, 2001, chapter 3,
p 65; d) A. B. R. Mayer, Polym. Adv. Technol. 2001, 12,
References
[1] a) D. Astruc, F. Lu, J. R. Aranzaes, Angew. Chem.
2
7
1
005, 117, 8062–8083; Angew. Chem. Int. Ed. 2005, 44,
852–7872; b) V. Farina; Adv. Synth. Catal. 2004, 346,
553–1582; c) N. T. S. Phan, M. van der Sluys, C. W.
Jones, Adv. Synth. Catal. 2006, 348, 609–679.
[
2] a) L. Djakovitch, M. Wagner, C. G. Hartung, M. Beller,
K. Kçhler, J. Mol. Catal. A: Chem. 2004, 219, 121–130;
b) A. Biffis, M. Zecca, M. Basato, J. Mol. Catal. A:
Chem. 2001, 173, 249–274; c) A. F. Schmidt, L. V. Ma-
metova, Kinet. Catal. 1996, 37, 406–408; d) A. Wali,
S. M. Pillai, V. K. Kaushik, S. Satish, Appl. Catal. A:
General 1996, 135, 83–93; e) F. Zhao, M. Arai, React.
Kinet. Catal. Lett. 2004, 81, 281–289; f) A. Biffis, M.
Zecca, M. Basato, Eur. J. Inorg. Chem. 2001, 1131–
9
6–106; e) H. Bçnnemann, R. Richards; Synth. Meth-
ods Organomet. Inorg. Chem. 2002, 10, 209–224; f) A.
Roucoux, J. Schulz, H. Patin, Chem. Rev. 2002, 102,
3
757–3778; g) M. Moreno-Manas, R. Pleixats, Acc.
Chem. Res. 2003, 36, 638–643; h) B. F. G. Johnson, Top.
Catal. 2003, 24, 147–159.
[
9] X.-D. Mu, D. G. Evans, Y. Kou, Catal. Lett. 2004, 97,
1
51–154.
1
133.
[
[
10] P. Claus, A. Brꢂckner, C. Mçhr, H. Hofmeister, J. Am.
[
[
3] a) J. P. M. Niederer, A. B. J. Arnold, W. F. Hçlderich, B.
Spliethoff, B. Tesche, M. T. Reetz, H. Bçnnemann, Top.
Catal. 2002, 18, 265–269; b) N. Toshima, Y. Shiraishi, T.
Teranishi, M. Miyake, T. Tominaga, H. Watanabe, W.
Brijoux, H. Bçnnemann, G. Schmid, Appl. Organomet.
Chem. 2001, 15, 178–196; c) M. T. Reetz, H. Schulen-
burg, M. Lopez, B. Spliethoff, B. Tesche, Chimia 2004,
Chem. Soc. 2000, 122, 11430–11439.
11] A.-I. Kozlov, A. P. Kozlova, K. Asakura, Y. Matsui, T.
Kogure, T. Shido, Y. Iwasawa, J. Catal. 2000, 196, 56–
6
5.
[
[
12] A. Martino, S. A. Yamanaka, J. S. Kawola, D. A. Loy,
Chem. Mater. 1997, 9, 423–429.
13] U.-A. Paulus, U. Endruschat, G.-J. Feldmeyer, T.-J.
5
8, 896–899.
Schmidt, H. Bçnnemann, R.-J. Behm, J. Catal. 2000,
4] a) C. P. Mehnert, D. W. Weaver, J. Y. Ying, J. Am.
Chem. Soc. 1998, 120, 12289–12296; b) C. P. Mehnert,
Chem. Commun. 1997, 2215–2216; c) L. Djakovitch, K.
Kçhler, J. Am. Chem. Soc. 2001, 123, 5990–5999;
d) R. G. Heidenreich, J. G. E. Krauter, J. Pietsch, K.
Kçhler, J. Mol. Catal. A: Chem. 2002, 182–183, 499–
1
95, 383–393.
[
[
[
14] Y. Mizukoshi, R. Oshima, Y. Maeda, Y. Nagata, Lang-
muir 1999, 15, 2733–2737.
15] S. Papp, I. Dekany, Colloid Polym. Sci. 2001, 279, 449–
4
58.
16] C. B. Hwang, Y.-S. Fu, Y.-L. Lu, S.-W. Jang, P.-T. Chou,
5
09; e) K. Kçhler, R. G. Heidenreich, J. G. E. Krauter,
J. Pietsch, Chem. Eur. J. 2002, 8, 622–631; f) K. Oku-
mura, K. Nota, K. Yoshida, M. Niwa, J. Catal. 2005,
C.-R. C. Wang, S.-J. Yu, J. Catal. 2000, 195, 336–341.
[17] K.-T. Wu, Y.-D. Yao, C.-R. C. Wang, P. F. Chen, E.-T.
2
31, 245–253; g) L. Artok, H. Bulut, Tetrahedron Lett.
Yeh, J. Appl. Phys. 1999, 85, 5959–5961.
2
004, 45, 3881–3884; h) S. Mandal, D. Roy, R. V.
[18] R. P. Andres, J.-D. Bielefeld, J.-I. Henderson, D.-B.
Janes, V.-R. Kolagunta, C.-P. Rubiak, W. J. Mahoney,
R.-G. Osifchin, Science 1996, 273, 1690–1693.
Chaudhari, M. Sastry, Chem. Mater. 2004, 16, 3714–
3
724; i) K. Kçhler, W. Kleist, S. S. Prçckl, Inorg. Chem.
2
007, 46, 1877–1883.
[19] N. Toshima, in: Fine Particles Sciences and Technology.
From Micro- to NewParticles , (Ed.: E. Pellizzetti),
Kluwer, Dordrecht, 1996, p 371.
[
5] a) M.-C. Daniel, D. Astruc, Chem. Rev. 2004, 104, 293–
46; b) M. Brust, M. Walker, D. Bethell, D. J. Schiffrin,
R. Whyman, J. Chem. Soc., Chem. Commun. 1994,
3
[20] S. V. Ley, C. Mitchell, D. Pears, C. Ramarao, J.-Q. Yu,
8
2
3
01–802; c) D. I. Gittins, F. Caruso, Angew. Chem.
001, 113, 3089–3092; Angew. Chem. Int. Ed. 2001, 40,
001–3004.
W.-Z. Zhou, Org. Lett. 2003, 5, 4665–4668.
[21] M. M. Demir, M. A. Gulgun, Y. Z. Menceloglu, B.
Erman, S. S. Abramchuk, E. E. Makhaeva, A. R.
Adv. Synth. Catal. 2008, 350, 717 – 730
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
729

Klaas Mennecke et al.
FULL PAPERS
Khokhlov, V. G. Matveeva, M. G. Sulman, Macromole-
cules 2004, 37, 1787–1792.
22] S. Kidambi, J. H. Dai, J. Lin, M. L. Bruening, J. Am.
Chem. Soc. 2004, 126, 2658–2659.
23] B. P. S. Chauhan, J. S. Rathore, T. Bandoo, J. Am.
Chem. Soc. 2004, 126, 8493–8500.
24] a) W. Solodenko, K. Mennecke, C. Vogt, S. Gruhl, A.
Kirschning, Synthesis 2006, 1873–1882; b) A. Drelin-
kiewicza, A. Waksmundzka, W. Makowski, J. W. Sobc-
zak, A. Krol, A. Zieba, Catal. Lett. 2004, 94, 143–156.
25] Y.-B. Liu, C. Khemtong, J. Hu, Chem. Commun. 2004,
applications ofthis reactor, see: a) M. C. Bagley, R. L.
Jenkins, M. C. Lubinu, C. Mason, R. Wood, J. Org.
Chem. 2005, 70, 7003–7006; b) T. N. Glasnov, D. J.
Vugts, M. M. Koningstein, B. Desai, W. M. F. Fabian,
R. V. A. Orru, C. O. Kappe, QSAR Comb. Sci. 2006,
25, 509–518; c) J. S. Smith, F. J. Iglesias-Sigueenza, I. R.
Baxendale, S. V. Ley, Org. Biomol. Chem. 2007, 5,
2758–2761.
[
[
[
[
38] a) I. R. Baxendale, C. M. Griffiths-Jones, S. V. Ley,
G. K. Tranmer, Chem. Eur. J. 2006, 12, 4407–4416;
b) G. Shore, S. Morin, M. G. Organ, Angew. Chem.
[
[
[
[
3
98–399.
26] U. R. Pillai, E. Sahle-Demessie, J. Mol. Catal. A 2004,
22, 153–158.
2
006, 118, 2827–2832; Angew. Chem. Int. Ed. 2006, 45,
2
761–2766, and references cited therein; c) P. Watts,
2
S. J. Haswell, Chem. Eng. Technol. 2005, 28, 290–301.
39] a) J. R. Thomas, Catal. Lett. 1997, 49, 137–141; b) W. L.
Perry, D. W. Cooke, J. D. Katz, A. K. Datye, Catal.
Lett. 1997, 47, 1–4.
40] The selective heating ofmetallic nanoparticles by mi-
crowaves has received considerable attention lately,
and applications have ranged from improving catalytic
performance and enhancing nanoparticle growth, to
various biomedical applications. For example, see:
a) Review: M. Tsuji, M. Hashimoto, Y. Nishizawa, M.
Kubokawa, T. Tsuji, Chem. Eur. J. 2005, 11, 440–452;
b) J. A. Gerbec; D. Magana, A. Washington, G. F.
Strouse, J. Am. Chem. Soc. 2005, 127, 15791–15800;
c) J. C. Yu, X. Hu, Q. Li, L. Zhang, Chem. Commun.
27] M. Adlim, M. Abu Bakar, K. Y. Liew, J. Ismail, J. Mol.
Catal. A 2004, 212, 141–149.
28] D. Tabuani, O. Monticelli, A. Chincarini, C. Bianchini,
[
[
F. Vizza, S. Moneti, S. Russo, Macromolecules 2003, 36,
4
294–4301.
29] M. Kralik, A. Biffis, J. Mol. Catal. A: Chem. 2001, 177,
13–138.
30] K. Sundmacher, H. Kꢂnne, U. Kunz, Chem. Ing. Tech.
998, 70, 267–271.
[
[
[
1
1
31] W. Solodenko, H. Wen, S. Leue, F. Stuhlmann, G. Sour-
kouni-Argirusi, G. Jas, H. Schçnfeld, U. Kunz, A.
Kirschning, Eur. J. Org. Chem. 2004, 3601–3610.
[
[
32] Reviews: a) G. Jas, A. Kirschning, Chem. Eur. J. 2003,
9
, 5708–5723; b) A. Kirschning, W. Solodenko, K.
2
005, 2704–2706; d) M. J. Kogan, N. G. Bastus, R.
Mennecke, Chem. Eur. J. 2006, 12, 5972–5990.
33] a) U. Kunz, A. Kirschning, H. Schçnfeld, W. Soloden-
ko, J. Chromatogr. A 2003, 1006, 241–249; b) U. Kunz,
A. Kirschning, H.-L. Wen, W. Solodenko, R. Cecilia,
C. O. Kappe, T. Turek, Catal. Today 2005, 105, 318–
Amigo, D. Grillo-Bosch, E. Araya, A. Turiel, A. Labar-
ta, E. Giralt, V. F. Puntes, Nano Lett. 2006, 6, 110–115;
e) G. S. Vanier; Synlett 2007, 131–135; f) B. C. Ranu,
A. Saha, R. Jana, Adv. Synth. Catal. 2007, 349, 2690–
2
696.
3
24; c) U. Kunz, H. Schçnfeld, W. Solodenko, G. Jas, A.
[
[
41] B. Corain, K. Jerabek, P. Centomo, P. Canton, Angew.
Chem. 2004, 116, 977-980; Angew. Chem. Int. Ed. 2004,
Kirschning, Ind. Eng. Chem. Res. 2005, 44, 8458–8467.
34] Applications: a) A. Kirschning, C. Altwicker, G.
Dräger, J. Harders, N. Hoffmann, U. Hoffmann, H.
Schçnfeld, W. Solodenko, U. Kunz, Angew. Chem.
[
[
[
4
3, 959–962;.
42] This assumption was verified by using a mathematical
model which also demonstrated that larger amounts of
added cross-linker lead to considerable degree ofcross-
linking in the outer shell. U. Kunz, C. Altwicker, U.
Limbeck, U. Hoffmann, J. Mol. Catal. A, Chem. 2001,
2
3
001, 113, 4118–4120; Angew. Chem. Int. Ed. 2001, 40,
995–3998; Angew. Chem. 2001, 113, 4118–4120; b) W.
Solodenko, U. Kunz, G. Jas, A. Kirschning, Bioorg.
Med. Chem. Lett. 2002, 12, 1833–1835.
1
77, 21–32.
43] a) J. M. Kremsner, C. O. Kappe, J. Org. Chem. 2006, 71,
651–4658; b) M. Hosseini, N. Stiasni, V. Barbieri,
35] a) M. Zecca, M. Kralik, M. Boaro, G. Palma, S. Lora,
M. Zancato, B. Corain, J. Mol. Catal. A 1998, 129, 27–
[
[
[
4
3
4; b) A. Biffis, B. Corain, Z. Cvengrosova, M. Hronec,
K. Jerabek, M. Kralik, Appl. Catal. A: Gen. 1995, 124,
55–365; c) J. K. Cho, R. Najman, T. W. Dean, O. Ichi-
hara, C. Muller, M. Bradley, J. Am. Chem. Soc. 2006,
28, 6276–6277.
C. O. Kappe, J. Org. Chem. 2007, 72, 1417–1424.
44] A. C. Metaxas, R. J. Meredith, Industrial Microwave
Heating, in: IEE Power Engineering Series, London,
3
1
983, Vol. 4.
45] J. Kurfꢂrstova, M. Hajek, Res. Chem. Intermed. 2004,
0, 673–681.
1
36] Recent reviews on microwave chemistry: a) C. O.
Kappe, Angew. Chem. 2004, 116, 6408–6443; Angew.
Chem. Int. Ed. 2004, 43, 6250–6284; b) B. L. Hayes, Al-
drichimica Acta 2004, 37, 66–77; c) B. A. Roberts,
C. R. Strauss, Acc. Chem. Res. 2005, 38, 653–661;
d) A. de La Hoz, A. Diaz-Ortiz, A. Moreno, Chem.
Soc. Rev. 2005, 34, 164–178; e) L. Perreux, A. Loupy,
Tetrahedron 2001, 57, 9199–9223.
3
[46] The effect of hot spot effects as described here are
highly dependent ofthe l of w rate. Compared to
À1
2 mLmin
flow rate employed here, the hot spot
should be totally suppressed at flow rates of
À1
180 mLmin : R. Cecilia, U. Kunz, T. Turek, Chem.
Eng. Proc. 2007, 47, 870–881.
[37] CEM Voyager continuous flow single mode microwave
[47] For a recent review, see: T. N. Glasnov, C. O. Kappe,
system, see Experimental Section for details. For recent
Macromol. Rapid Commun. 2007, 28, 395–410.
730
asc.wiley-vch.de
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 717 – 730