Benchmarking immobilized di- and triarylphosphine palladium catalysts for continuous-flow cross-coupling reactions: Efficiency, durability, and metal leaching studies

Article abstract of DOI:10.1021/cs5020089

Leaching resistance and recyclability, in addition to efficiency, are key parameters which constrain the choice of a catalyst for performing metal-catalyzed cross-coupling reactions in continuous-flow mode. Comparison of commercially available immobilized

Full text of DOI:10.1021/cs5020089

Research Article
pubs.acs.org/acscatalysis
Benchmarking Immobilized Di- and Triarylphosphine Palladium
Catalysts for Continuous-Flow Cross-Coupling Reactions: Efficiency,
Durability, and Metal Leaching Studies
Roberto Greco, Walter Goessler, David Cantillo,* and C. Oliver Kappe*
Institute of Chemistry, University of Graz, NAWI Graz, Heinrichstrasse 28, A-8010 Graz, Austria
S
* Supporting Information
ABSTRACT: Leaching resistance and recyclability, in addition to
efficiency, are key parameters which constrain the choice of a catalyst
for performing metal-catalyzed cross-coupling reactions in continu-
ous-flow mode. Comparison of commercially available immobilized
catalysts is often difficult because literature data are typically obtained
under a wide range of reaction conditions. Here we present a
comparative investigation on some of the most common immobilized
phosphine-based Pd catalysts, namely Pd Tetrakis (polymer bound),
FiberCat 1001, EnCat TPP30, and SiliaCat DPP-Pd. The efficiency,
recyclability, and leaching resistance of each of the catalysts has been
carefully investigated under a standard set of conditions as well as a
selection of literature-based protocols. The data presented herein enable a direct comparison of these catalysts and provide
further insights into the leaching phenomena of these types of ligand-based palladium catalysts.
KEYWORDS: continuous flow, cross-coupling, palladium, immobilized catalysts, leaching
context, several immobilized catalysts have been employed both
in batch procedures and in continuous-flow applications.7 In a
continuous-flow process the catalyst is typically placed in a
packed-bed reactor and the reaction mixture is flown through
the reactor using an appropriate pumping system.8 The
catalytically active material remains located in a specific part
of the reactor during processing of the reaction mixture, and
therefore, the reaction and separation of the catalyst from the
solution (including the product) take place simultaneously. In
this way, catalyst reuse and recycling is simplified, as a catalyst
separation/filtration step can be avoided. Furthermore, as small
amounts of the reaction mixture are in contact simultaneously
with the packed-bed reactor, high local concentrations of
catalyst are achieved, resulting in increased reaction rates in
comparison to that of a batch reaction.8 Thus, high conversions
can be obtained within residence times of a few seconds or
minutes, although on the other hand the same local
concentration effect can also sometimes lead to lower
selectivities.
INTRODUCTION
■
There is little doubt that the development of cross-coupling
chemistry over the past decades has revolutionized modern
synthetic organic chemistry.1 These transition-metal-catalyzed
carbon−carbon and carbon−heteroatom bond forming reac-
tions have had a significant impact on the synthesis of
pharmaceuticals, agrochemicals, and natural products.1,2 As a
recognition of their seminal work, Richard Heck, Ei-ichi
Negishi, and Akira Suzuki were awarded with the Nobel Prize
in Chemistry in 2010.3 The discoveries by these authors of Pd-
catalyzed C−C bond formations changed how many organic
syntheses are conceived and later on inspired other chemists to
develop a wide range of additional cross-coupling reactions,
including carbon−heteroatom coupling.1,2,4 In the case of
carbon−carbon bond forming reactions palladium is the most
commonly used transition-metal catalyst. In a few cases metals
other than palladium are used to perform these cross-couplings
(typically copper or nickel), although the catalytic activity of
these metals is generally significantly lower in comparison to
that of palladium.5
The intense research in the field of cross-coupling chemistry
during the past two decades has led to the development of
“greener” reaction conditions as well as a large number of
highly efficient catalyst/ligand combinations.1−5 Apart from a
plethora of homogeneous catalytic systems, several immobilized
palladium catalysts have also been developed6 and have been
extensively utilized as heterogeneous catalysts, on the premise
that the the solid-supported active metal could be easily
removed from the reaction mixture and reutilized. In this
When cross-coupling chemistry and continuous flow using a
packed-bed heterogeneous catalyst are combined, substantial
amounts of “dissolved” palladium are typically observed in the
reaction mixture collected from the reactor output.9 Ultimately,
deactivation of the catalyst bed and reduced product formation
are inevitable. The leaching phenomenon in cross-coupling
Received: December 14, 2014
Revised: January 13, 2015
Published: January 15, 2015
© 2015 American Chemical Society
1303
DOI: 10.1021/cs5020089
ACS Catal. 2015, 5, 1303−1312

ACS Catalysis
Research Article
reactions is ascribed to the fact that the reaction mechanism for
these kinds of transformations is (quasi)homogeneous and
involves the transformation of Pd0 species into soluble PdII
during the oxidative addition step.10 When an immobilized
catalyst is used in a batch cross-coupling reaction, the leaching
of palladium is not obvious because upon completion the metal
readsorbs onto the solid support (thus acting as a “reservoir” of
soluble, active Pd species). When continuous flow is used, the
situation is different: in continuous-flow mode, the palladium
metal will progressively be “chromatographed” through the
packed-bed catalyst until eventually all Pd will be removed from
the support owing to continuous metal leaching/readsorption.9
This effect, which becomes especially relevant when long-run
experiments are performed, leads to significant amounts of
transition metal being leached out of the reactor, contaminating
the product.
In 2009 our group carried out a detailed investigation on the
mechanism and leaching phenomena for the Pd/C-catalyzed
Mizoroki−Heck reaction of 4-iodobenzonitrile with butyl
acrylate in microwave batch and continuous flow.11 Although
Pd/C is one of the most widely used catalysts for (batch) cross-
coupling chemistry,12 under the applied continuous-flow
conditions a rapid metal leaching and catalyst deactivation
was observed. Notably, novel and more sophisticated supported
Pd catalysts for cross-coupling chemistry are being continuously
reported in the literature.1−5 In particular, incorporation of
ligands to the immobilized catalyst often results in improved
stability and leaching resistance of the solid material.2 In some
instances, continuous-flow applications have been also
performed to test the leaching resistance and hence the
recyclability and durability of the catalysts. However, the flow
experiments have often been performed under very specific
conditions (solvent, temperature) and at very low substrate
concentrations. Low substrate concentrations obviously result
in low amounts of metal being leached from the packed-bed
reactor but compromise the practical use of a procedure that
has no possibility of being scalable. A comparison of the
performance of the different commercially available catalysts is
therefore difficult, as the existing literature data have been
obtained under a wide range of reaction conditions. Herein we
present a benchmarking study of some of the most common
immobilized diarylphosphine- and triarylphosphine-based
palladium catalysts: namely, Pd Tetrakis (polymer-bound),
FiberCat 1001, EnCat TPP30, and SiliaCat DPP-Pd. The
efficiency, durability, and leaching resistance of these supported
catalysts have been studied for continuous-flow Mizoroki−
Heck and Suzuki−Miyaura reactions under analogous as well as
selected literature conditions. The data presented herein enable
a direct comparison of these catalysts and will help to
understand the leaching phenomena of these types of ligand-
based supported Pd catalysts.
Figure 1. Structures of the (a) polymer-bound Pd Tetrakis, (b)
FiberCat 1001, (c) EnCat TPP30, and (d) SiliaCat catalysts used in
this work.
durability, as well as reaction selectivity and efficiency in some
cases.2 Two cross-coupling reactions have been selected as
models for our continuous-flow experiments: the Mizoroki−
Heck reaction of 4-iodobenzonitrile (IBN) with n-butyl acrylate
(BA) (in analogy to our previous study on the Pd/C catalyst11)
and the Suzuki−Miyaura reaction of IBN with phenylboronic
acid (PBA) (Scheme 1). The use of aryl bromides in these
Scheme 1. Cross-Coupling Reactions Selected as Models in
a
the Present Study
a
Dehalogenation (D) and homocoupling (H) side products typically
observed in these reactions are also shown.
transformations was troublesome, since their lower reactivity
resulted in only moderate conversions in flow after a single pass
through the packed-bed reactor.11
In all cases the temperature has been kept below 120 °C,
even when low conversions were observed. It has been shown
that at temperatures above 120 °C Pd catalysts are prone to
form Pd black nanoparticles13 which could provoke a faster
deactivation and leaching of the active metal species. Selected
literature-based reaction conditions (temperature, concentra-
tion, base, solvent) have been applied for each of the
immobilized catalysts to independently assess their efficiency
and stability. Furthermore, a set of reaction conditions which
utilizes acetonitrile as solvent and triethylamine (TEA) or
diisopropylethylamine (DIPEA) as base has been concurrently
utilized in all cases. Acetonitrile is a relatively inert, weakly
RESULTS AND DISCUSSION
■
Initial Considerations. Four popular Pd-supported cata-
lysts have been considered in our comparative study (Figure 1).
Polymer-bound Pd Tetrakis, FiberCat 1001, EnCat TPP30, and
SiliaCat DPP-Pd differ in the solid matrix hosting the active
metal catalyst and the way in which the ligand is attached to the
support (covalent bonding, encapsulation) (Figure 1). All of
the catalysts evaluated herein have in common the presence of
a substituted diphenylphosphine (DPP) or triphenylphosphine
(TPP) as ligand. It has been shown that the presence of a
ligand has beneficial effects in terms of catalyst stability and
1304
DOI: 10.1021/cs5020089
ACS Catal. 2015, 5, 1303−1312

ACS Catalysis
Research Article
complexing solvent compatible with all solid-supported
catalysts considered in this study.
M solutions of the substrate in MeCN were treated with 1.5
equiv of the coupling partner (phenylboronic acid or n-butyl
acrylate) and 1.5 equiv of base and mixed with 0.5 mol % of the
supported catalyst. The mixture was heated in a single-mode
microwave reactor for 15 min at temperatures ranging from 70
to 100 °C and the crude product analyzed by GC-FID (see
Table S1 in the Supporting Information). A temperature of 100
°C was not exceeded, as this is the maximum working
temperature recommended by the catalyst vendor. As expected,
the best conversions were obtained at this temperature (56−
57% and 36−40% for the Suzuki and Heck reactions,
respectively) and a relatively good selectivity (90%) was
observed for both transformations. The base employed did not
affect the reaction outcome, and DIPEA and TEA gave
analogous results.
Each of the immobilized catalysts has been tested for both
the model Mizoroki−Heck and Suzuki−Miyaura reactions. In
all cases a set of preliminary microwave (MW) batch
experiments was performed to estimate an appropriate range
of temperatures for the flow experiments.14 Fresh catalyst
cartridges were used for all flow runs, and typically 20 mL of
the reaction mixture or 10 mmol of substrate was processed.
For this purpose a commercial flow instrument (X-Cube,
ThalesNano) was utilized (Figure 2).15 Conversion profiles
We thus decided to move forward and utilize these
conditions (100 °C, TEA or DIPEA as base in MeCN) in
continuous-flow mode (see the Experimental Section for
details). A volume of 20 mL of a 0.5 M reaction mixture was
prepared for both cross-coupling reactions and pumped
through the reactor utilizing fresh catalyst cartridges preheated
at 100 °C. The reactor pump was set to a flow rate of 0.2 mL
min−1, which results in a residence time within the packed-bed
reactor of approximately 2−3 min. Samples were collected from
the reactor output and analyzed by GC-FID to monitor the
reaction conversion and selectivity over time (Figure 3). In the
case of the Heck reaction (Figure 3a) an essentially constant
reaction profile was obtained under steady-state conditions. Full
conversion was obtained within the relatively short contact time
of the reaction mixture with the catalyst bed, although with a
significantly lower selectivity with respect to batch processing
(with formation of ∼25% dehalogenation product and ∼15%
homocoupling). Both effects are typically ascribed to the high
local concentrations of catalyst achieved using a packed-bed
reactor.11 Importantly, after 180 min working time and 10
mmol of starting material processed no drop in the conversion
was noticeable.
For the Suzuki reaction the picture looked completely
different (Figure 3b). In this case the substrate was fully
consumed during the first 40 min of processing (after an initial
period of stabilization) but then dropped dramatically to a
conversion of <10%. Obviously the catalyst was rapidly
degraded during this reaction, and significant amounts of
metal were most likely leached from the support, contaminating
the product.
The irregular contour of the conversion profiles during the
first minutes of the continuous-flow processing is worth noting.
Seemingly, only the desired cross-coupling is initially observed,
followed by a drop in the selectivity until the system stabilizes.
This effect, particularly pronounced for the Suzuki−Miyaura
reaction, is due to the different affinities of each of the reaction
components to the solid phase of the packed-bed reactor. Thus,
the reactants as well as the formed products are “chromato-
graphed” through the catalyst column, causing an initial
incorrect reaction stoichiometry until steady-state conditions
are reached. As could be observed in Figure 4, the cross-
coupling product is eluted through the packed bed faster than
other reaction products and therefore is the first species
observed in the conversion profile. Butyl acrylate is also
separated and eluted more rapidly than other reagents. This
“chromatographic effect” results in an apparent poor selectivity
at the initial stage of processing, with a greater amount of
homocoupling product being obtained, due to the lower
Figure 2. Schematic representation of the flow setup (a), performed in
a X-Cube reactor from ThalesNano (b). The immobilized catalysts
were loaded in a CatCart stainless steel column reactor (c).
were obtained by GC-FID monitoring of samples collected
from the reactor output. To assess the amount of metal
leaching produced during the flow processing, the Pd content
of the crude reaction mixtures was determined by ICPMS
analysis (see the Experimental Section for details).
Polymer-Bound Pd Tetrakis. This supported catalyst,
developed by Fenger and Le Drian in 1998, consists of a
diphenylphosphinated divinylbenzene (DVB) cross-linked
polystyrene matrix treated with a solution of Pd(PPh3)4 in
benzene.16 In their original work the catalyst was tested for the
batch Suzuki−Miyaura reaction of 4-bromopyridine with
phenylboronic acid. Excellent yields analogous to those
obtained when Pd(PPh3)4 was used as homogeneous catalyst
were obtained, with the advantage that the heterogeneous
catalyst could be easily separated from the reaction mixture and
reused more than five times without any activity decrease.
Recently, it has been shown by Gordon and co-workers that
this catalyst can be applied for continuous-flow Suzuki−
Miyaura reactions involving furylboronic acids,17 although a
detailed investigation on the stability of this particular catalyst
was not carried out.
Our study started with a series of preliminary microwave
batch experiments to evaluate the most appropriate reaction
conditions. As an established literature protocol for cross-
coupling reactions in flow using this catalytic system was not
available, we decided to carry out all our experiments using
acetonitrile as solvent and TEA or DIPEA as base. Thus, ∼0.5
1305
DOI: 10.1021/cs5020089
ACS Catal. 2015, 5, 1303−1312

ACS Catalysis
Research Article
of the corresponding chromatographic effect, see Figure S9 in
the Supporting Information).
ICPMS analyses revealed that significant amounts of Pd had
leached from the packed-bed reactors for both cross-coupling
reactions. Not surprisingly, a higher amount was determined for
the Suzuki−Miyaura (4.61 mg) in comparison to that for the
Mizoroki−Heck crude reaction mixture (2.25 mg) (the amount
of palladium refers to that amount contained in the total crude
reaction mixture), as in the former case the reaction conversion
dramatically dropped after 40 min of processing. However, the
amount of metal measured was small in comparison to the total
amount of palladium contained in each catalyst cartridge. A
close inspection of the packed-bed reactor denoted a substantial
degradation of the material, with copious amounts of Pd black
particles found in the reactor tubing and connections.
FibreCat 1001. A matrix of triphenylphosphine-function-
alized polypropylene fibers serves as support for this palladium
acetate based catalyst (Figure 1b).18 Apart from a number of
batch mode applications,19 this immobilized catalyst has been
utilized in continuous-flow mode for Sonogashira20 and
Suzuki−Miyaura17,21 cross-coupling reactions. In this context,
́
a catalyst reusability study was carried out by Dorman and co-
workers for continuous-flow Sonogashira reactions.20 The
results revealed important substrate-dependent reductions in
conversion after reusing the packed-bed reactors more than five
times.
Our own initial microwave batch experiments showed
excellent conversion (95%) and selectivity (>99%) for the
Mizoroki−Heck reaction on heating to 120 °C for 15 min using
DIPEA as base (see Table S2 in the Supporting Information).
The Suzuki−Miyaura coupling was less efficient with this
catalyst, and at the same temperature only 60% conversion was
achieved. This temperature was thus selected for the long-run
continuous-flow experiments. In this case DIPEA performed
slightly better than TEA as the base for the reaction.
The conversion profiles resulting from the continuous-flow
experiments were somewhat similar to those obtained with the
polystyrene Pd Tetrakis immobilized catalyst (Figure 5). In the
case of the Heck reaction, after an initial period of erratic
selectivity due to the chromatographic effect, a more or less
constant profile with full conversion and ca. 60% selectivity was
achieved (Figure 5a). Evaluation of the retention of the reaction
components in the FiberCat support was analogous to that for
the polystyrene matrix, and similar chromatography curves
were observed (see Figures S10 and S11 in the Supporting
Information). Notably, after 80 min of processing the
conversion started to drop slowly. Similar to the case for
polymer-bound Pd Tetrakis, the Suzuki−Miyaura reaction
provoked a very rapid degradation and deactivation of the
supported catalyst, starting after approximately 20 min of flow
processing (Figure 5b).
Figure 3. Conversion profiles obtained for the continuous-flow
Mizoroki−Heck (a) and Suzuki−Miyaura (b) reactions catalyzed by
polymer-bound Pd Tetrakis. The first ∼10 min, in which no
compound is observed in the GC-FID, corresponds to the dead
volume of the reactor. Conditions: 0.5 M substrate in MeCN, 1.5
equiv of TEA, 1.5 equiv of PhB(OH)2 or n-butyl acrylate, 100 °C, 0.2
mL min−1.
ICPMS analyses of the crude reaction mixtures again
revealed significant amounts of Pd leached from the packed-
bed reactors after processing 20 mL of the reaction mixture
(2.58 and 2.43 mg of Pd for the Mizoroki−Heck and Suzuki−
Miyaura reactions, respectively). However, the quantity of
leached palladium was again relatively moderate in comparison
with the total amount contained in the catalyst cartridge (ca. 15
mg), despite the rapid drop of conversion observed for the
Suzuki−Miyaura reaction. In this particular case the content of
the solid material inside the used catalyst cartridge was also
examined (Figure 6). Apparently, the reaction provoked the
formation of Pd black particles, which gradually moved through
Figure 4. Differences in the retention of the components of the
Mizoroki−Heck reaction within the polymer-bound Pd Tetrakis
catalyst cartridge. Base: TEA.
effective amount of coupling partner until steady-state
concentrations are reached. The effect is even more
pronounced for the Suzuki−Miyaura reaction (for an analysis
1306
DOI: 10.1021/cs5020089
ACS Catal. 2015, 5, 1303−1312

ACS Catalysis
Research Article
Figure 6. Visual representation of the contents of a used Fibrecat 1001
catalyst cartridge after a continuous-flow Suzuki−Miyaura reaction.
The Pd black particles are clearly concentrated in the final part of the
packed-bed reactor after the flow run.
a metal scavenging resin had to be installed into the flow setup.
Attempts to process large-scale reaction mixtures failed, and the
conversions dropped dramatically after a few hours of
processing.25c
Figure 5. Conversion profiles obtained for the continuous flow
Mizoroki−Heck (a) and Suzuki−Miyaura (b) reactions catalyzed by
Fibercat 1001. The first ∼10 min, in which no compound is observed
in the GC-FID, corresponds to the dead volume of the reactor.
Conditions: 0.5 M substrate in MeCN, 1.5 equiv of DIPEA, 1.5 equiv
of PhB(OH)2 or n-butyl acrylate, 120 °C, 0.2 mL min−1.
Our study again started with a series of preliminary
microwave batch experiments to establish adequate reaction
conditions for this catalyst (see Table S3 in the Supporting
Information). Apart from an evaluation of the general
acetonitrile/DIPEA conditions, other base and solvent systems
utilized in the literature25 were tested. In MeCN and DIPEA or
TEA as base, low conversions (<25%) were observed in all
cases after 15 min at 120 °C, in contrast to the high conversions
achieved with Pd Tetrakis and FibreCat 1001 under analogous
conditions. When EtOH was used as solvent instead and
tetrabutylammonium acetate (TBAOAc) was used as base,
higher conversions (60% and 75% for Suzuki−Miyaura and
Mizoroki−Heck reactions, respectively) and excellent selectiv-
ities were achieved (Table S3, entries 4 and 13). Ultimately
both solvent/base systems were translated into flow for
comparative purposes.
The flow experiments carried out using the two systems and
the Pd Encat catalyst highlight the importance of the solvent/
base choice on the stability of the packed-bed reactor. When
the Mizoroki−Heck coupling was performed in acetonitrile
with TEA as base, a constant amount of product was observed
(after the initial irregular selectivity due to the chromatographic
effect) on processing 20 mL of the reaction mixture (10 mmol)
(Figure 7a). Even when the amount of reaction mixture was
doubled (40 mL), no drop in conversion could be seen (see
Figure S1 in the Supporting Information). In stark contrast,
when EtOH was used as the solvent in combination with
TBAOAc as base (conditions based on the work of the Ley
group25) a constant decrease in the product formation was
observed during flow processing (Figure 7b). Again, Suzuki−
Miyaura reactions resulted in much faster degradation of the
the column with the flow of the reaction mixture. Thus, at the
end of the flow run most of the metal was deposited in the final
part of the packed-bed reactor. ICPMS analyses of the different
portions of the solid material confirmed higher amounts of Pd
in the final part of the packed-bed reactor after the Suzuki−
Miyaura reaction (Figure 6). This catalyst degradation with
migration of Pd through the supported catalyst would indeed
result in a much shorter contact time of Pd-containing material
with the reaction mixture. This effect, together with the metal
leaching, can explain the significant reduction in conversion
observed.
PdEncat TPP30. Polyurea encapsulated Pd is a popular,
commercially available class of heterogeneous catalysts
introduced by the Ley group in 2002.22 The polyurea matrix
provides the material with a good ability to ligate metal species
and has been used in a wide range of applications, including
cross-coupling reactions.23 Although considered a hetereoge-
neous catalyst, it was shown by Broadwater and McQuade24
that most likely soluble, catalytically active Pd species leached
from the support are responsible for catalysis in cross-coupling
reactions, with the encapsulated palladium species serving as a
reservoir for soluble palladium. Suzuki−Miyaura cross-coupling
reactions using PdEncat in a packed-bed reactor in continuous-
flow mode have also been described.25 Importantly, metal
contamination in the products was observed in some cases and
1307
DOI: 10.1021/cs5020089
ACS Catal. 2015, 5, 1303−1312

ACS Catalysis
Research Article
experiment (also 8 h) utilizing this catalyst for a continuous-
flow Suzuki−Miyaura reaction in THF/H2O/EtOH as solvent
clearly exhibited a significant decrease in conversion.28a
A wide range of reaction conditions and solvent/base
combinations were evaluated for the SiliaCat DPP-Pd catalyst
in our preliminary batch experiments. These incorporated
THF/H2O and THF/H2O/EtOH mixtures as solvents as well
as K2CO3, KOH, and KOAc as bases, in addition to our
standard MeCN/DIPEA (or TEA) conditions concurrently
used for all catalysts (see Table S4 in the Supporting
Information). Remarkably, this catalyst had a relatively low
performance in acetonitrile in batch processing, for both the
Suzuki−Miyaura and Mizoroki−Heck reactions. Conversions
considerably increased when mixtures including protic solvents
and inorganic bases were employed. Thus, full conversion and
complete selectivity were observed in H2O/THF and H2O/
THF/EtOH mixtures and K2CO3 and KOH as bases, even at
temperatures as low as 60−70 °C (see entries 5 and 6 in Table
S4).
Given the good results obtained for a variety of solvent/base
combinations, a large number of flow experiments were carried
out using the SiliaCat catalyst. Although the conversion for the
Mizoroki−Heck reaction in acetonitrile was not very high
(Figure 8a), excellent selectivity was achieved. Interestingly, the
Figure 7. Conversion profiles obtained for the continuous-flow
Mizoroki−Heck reactions catalyzed by Pd Encat TPP30 using the (a)
acetonitrile/TEA and (b) ethanol/TBAOAc solvent/base systems.
The first ∼10 min, in which no compound is observed in the GC-FID,
corresponds to the dead volume of the reactor. Conditions: (a) 0.5 M
or (b) 0.25 M substrate, 1.5 equiv of base, 1.5 equiv of n-butyl acrylate,
120 °C, 0.2 mL min−1.
immobilized catalyst (Figures S2 and S3 in the Supporting
Information).
Despite the constant production of the desired Mizoroki−
Heck cross-coupling product using acetonitrile as solvent,
significant amounts of Pd (6.25 mg) were detected in the crude
reaction mixture by ICPMS after processing 20 mmol of the
substrate. Interestingly, lower amounts of Pd were found in the
ethanolic crude mixture (1.03 mg), probably reflecting the fact
that the catalyst is degraded or poisoned but that the Pd does
not pass the packed-bed reactor filter. As for the previous
catalyst studies, higher amounts of Pd were found in the
Suzuki−Miyaura reaction mixtures (detailed information on the
ICPMS Pd determination for all reactions is collected in Table
S5 in the Supporting Information).
SiliaCat DPP-Pd. SiliaCat DPP-Pd is a commercially
available sol−gel entrapped catalyst made out of an amorphous
organosilica matrix functionalized with a diphenylphosphine
ligand bound to Pd(II).26 This immobilized catalyst has
previously been evaluated for continuous-flow applications by
Figure 8. Conversion profiles obtained for the continuous-flow
Mizoroki−Heck (a) and Suzuki−Miyaura (b) reactions catalyzed by
SiliaCat DPP-Pd. The first ∼10 min, in which no compound is
observed in the GC-FID, corresponds to the dead volume of the
reactor. Conditions: 0.5 M substrate in MeCN, 1.5 equiv of TEA, 1.5
equiv of PhB(OH)2 or n-butyl acrylate, 120 °C, 0.2 mL min−1.
́ ́
the groups of Alcaz ar and
ar27 and Pagliaro.28,29 Notably, Alcaz
co-workers were able to perform a continuous Suzuki−Miyaura
reaction in H2O/THF for more than 8 h (0.125 M, 0.2 mL
min−1) without observing a decrease in the catalytic activity.27a
In contrast, in the work by Pagliaro and co-workers a long-run
1308
DOI: 10.1021/cs5020089
ACS Catal. 2015, 5, 1303−1312

ACS Catalysis
Research Article
conversion profile was not stable until ca. 40 min of processing
time. During the first stage of the processing a higher
conversion (∼35%) was obtained, which then dropped and
stabilized at ca. 20% after 40 min. This behavior is possibly due
to a modification of the metal species upon contact with the
reaction mixture that partially decreases its initial activity. A
similar pattern was observed for the Suzuki−Miyaura reaction
(Figure 8b), with higher initial conversion rates which slowly
stabilized during flow processing.
As was experienced in batch processing, combinations with a
protic solvent and an inorganic base resulted in very good
conversions and selectivities in continuous flow. The
conversion profiles depicted in Figure 10 were obtained using
We decided to increase the residence time for the Mizoroki−
Heck reaction by using a larger packed-bed reactor containing 2
g of supported catalyst. Under these conditions full conversion
was obtained during the complete processing time but with
decreased selectivity (∼80%) (see Figure S4 in the Supporting
Information). An Ommifit column reactor was utilized for this
experiment. In contrast to the stainless steel CatCart catalyst
cartridges, the Omnifit columns are made out of transparent
borosilicate glass and we therefore could observe in situ
changes in the appearance of the solid material during the
processing (Figure 9). The SiliaCat DPP-Pd catalyst has an
Figure 10. Conversion profiles obtained for the continuous-flow
Mizoroki−Heck (a) and Suzuki−Miyaura (b) reactions catalyzed by
SiliaCat DPP-Pd. The first ∼10 min, in which no compound is
observed in the GC-FID, corresponds to the dead volume of the
reactor. Conditions: 0.25 M substrate in THF/EtOH/H2O, 1.5 equiv
of K2CO3, 1.25 equiv of PhB(OH)2 or n-butyl acrylate, 80 °C, 0.2 mL
min−1.
the THF/EtOH/H2O and K2CO3 system described by Pagliaro
and co-workers.28 The reaction outcome rapidly stabilized, and
a constant profile was achieved. It is possible that in this case
the Pd catalyst is “preactivated” in contact with the hot protic
solvent, with can be easily oxidized by Pd(II) to form Pd(0),30
thus avoiding the initial stabilization time necessary when
MeCN was used as solvent. Interestingly, under these
conditions SiliaCat DPP-Pd performed much better for the
Suzuki−Miyaura reaction (Figure 10b) than for the Mizoroki−
Heck reaction (Figure 10a). Almost full conversion and
excellent selectivity was obtained during all of the processing
for the coupling with PhB(OH)2. For the Mizoroki−Heck
reaction a constant 65% conversion was obtained. This solvent/
base system gave the best results in terms of conversion and
catalyst stability. Several other solvent/base systems were
tested, including THF/H2O/KOH, THF/H2O/K2CO3, and
MeOH/H2O/K2CO3 (the corresponding conversion profiles
are collected in the Supporting Information).
Figure 9. Visual appearance of SiliaCat DPP-Pd packed in a column
reactor during the processing of a Mizoroki−Heck reaction using
acetonitrile as solvent and TEA as base (a) in contact with hot solvent,
(b) after 30 min, (c) after 60 min, and (d) after 120 min. The reaction
stream flows from the bottom to the top of the packed-bed reactor.
intense orange color which is maintained in contact with
acetonitrile even when the temperature is increased (Figure
9a). However, as soon as the reaction mixture was introduced
in the packed-bed reactor (Figure 9b) the catalyst material
gradually turned black, indicating the formation of Pd(0) in the
supported catalyst. Ultimately all of the supported catalyst
acquired a black coloration (Figure 9c,d). This catalyst
modification during the reaction is most probably responsible
for the decrease of catalytic activity observed for the smaller
scale experiment (Figure 8a).
1309
DOI: 10.1021/cs5020089
ACS Catal. 2015, 5, 1303−1312

ACS Catalysis
Research Article
ICPMS analyses of all crude reaction mixtures obtained from
the reactor output indicated a significantly lower amount of Pd
being leached from the solid support with respect to that for
the three previously studied supported catalysts. The amounts
of metal encountered in the mixtures obtained with the
MeCN/TEA system (Figure 8) were 0.81 and 1.25 mg for the
Mizoroki−Heck and Suzuki−Miyaura reactions, respectively.
For the THF/EtOH/H2O/K2CO3 system the amount was
even lower (332 and 39 μg for the Mizoroki−Heck and
Suzuki−Miyaura reactions). The solid material from the
catalyst cartridge utilized for this Suzuki−Miyaura reaction
(Figure 10b) was placed out of the steel cylinder for a closer
inspection. Similar to the reaction carried out in acetonitrile,
the catalyst had acquired a black color as a result of the
formation of Pd(0) black. More importantly, clear catalyst
degradation could be visually observed at the first part of the
material (Figure 11). Apparently, the Pd had been gradually
leaching and catalyst degradation were still observable. An
obvious question arises from the possibility that the silica matrix
itself can retain Pd species by a simple physical affinity with the
polar media and whether the presence of the phosphine ligand
covalently bound to the support is essential for the low
leaching. In order to address this question, a simple packed-bed
reactor was built in a column containing unmodified silica gel.
Then, a solution of Pd(PhPh3)4 in acetonitrile was pumped
through the column reactor at room temperature. The solid
material rapidly acquired a dark coloration while, after a second
pass of the Pd solution, the solvent became completely
colorless. The column was then utilized for a Suzuki−Miyaura
reaction using the MeCN/TEA system. Notably, when only
solvent was passed through the reactor, no leaching was
appreciable even on heating to 120 °C. However, as soon as the
reaction mixture was introduced, a rapid leaching was observed.
The conversion rapidly dropped, and it could be clearly
observed how the Pd Tetrakis was “washed out” from the silica
(for a detailed description, see Figures S12 and S13 in the
Supporting Information). This simple experiment demonstra-
ted that the polarity of the silica matrix itself has no crucial
effect on the recyclability of the SiliaCat DPP-Pd and the
immobilized ligand plays an essential role in the Pd retention.
This effect is in agreement with previous observations which
indicated that the addition of external ligands often induces
leaching of the metals.31
CONCLUSION
■
A comparative study of the efficiency, durability, and metal
leaching resistance of some of the most common immobilized
diarylphosphine- and triarylphosphine-based palladium catalysts
toward continuous flow cross-coupling reactions has been
performed. Pd Tetrakis (polymer-bound), FiberCat 1001,
EnCat TPP30, and SiliaCat DPP-Pd have been independently
evaluated using two model reactions: the Mizoroki−Heck
reaction of 4-iodobenzonitrile with butyl acrylate and the
Suzuki−Miyaura reaction of the same halide with phenyl-
boronic acid. A set of reaction conditions combining
acetonitrile as solvent and triethylamine or diisopropylethyl-
amine as base has been concurrently used in all cases to directly
compare the performance of the immobilized catalyst.
Furthermore, selected literature protocols have been also
reevaluated in terms of catalyst efficiency and durability.
Except for SiliaCat DPP-Pd, the Suzuki−Miyaura cross-
coupling produced a much higher amount of metal leaching
and catalyst degradation than the Mizoroki−Heck reaction.
The choice of an appropriate solvent/base system is crucial for
the durability of the supported catalyst. Thus, in the case of Pd
EnCat TPP30 a constant drop in conversion was observed
when EtOH as solvent and TBAOAc as base were utilized for
the Mizoroki−Heck reaction, while in MeCN a much more
constant conversion profile was achieved.
Variable amounts of Pd leached from the packed-bed
reactors were always present in the crude reaction mixtures
collected from the reactor output (for a comparison of Pd
leaching data for all experiments performed, see Table S5 in the
Supporting Information). In this context, SiliaCat DPP-Pd
combined with THF/EtOH/H2O as solvent and K2CO3 as
base gave the best results regarding catalyst efficiency and
leaching resistance. However, significant degradation of the
immobilized catalyst was also observable in these cases.
Negligible amounts of metal leaching are only achieved when
Figure 11. Visual appearance of the content of a used SiliaCat DPP-Pd
catalyst cartridge after a continuous-flow Suzuki−Miyaura reaction
with the THF/EtOH/H2O/K2CO3 system. The Pd content of
different fragments of the solid material, as determined by ICPMS
analyses, is also incorporated. Degradation of the solid material at the
beginning of the reactor is visually observable (gray coloration).
moved through the packed-bed reactor during the reaction
processing, becoming a gray material. This hypothesis was
confirmed by ICPMS analyses of the used solid material
(Figure 11), which detected a higher concentration of Pd in the
final part of the catalyst cartridge. Although no drop of
conversion was observed during our flow experiment and the
amount of metal contaminating the reaction mixture was not
very high, this analysis revealed that most likely when larger
reaction mixtures are processed all of the catalyst material will
be degraded, ultimately losing its activity. A negligible amount
of Pd leaching could only be obtained when very dilute reaction
mixtures were processed. Thus, following a procedure
described by the group of Pagliaro,29 solutions for the
Mizoroki−Heck and Suzuki−Miyaura reactions containing
0.05 M substrate concentration were processed using fresh
catalyst cartridges. Although, as expected, the metal leached
from the reactor was very low (15 μg and 5 μg of Pd), such
dilute conditions arguably are not very useful for practical
preparative purposes.
The results described above show SiliaCat DPP-Pd to be a
superior catalyst with respect to those in the other case studies,
especially in terms of stability and durability, even though metal
1310
DOI: 10.1021/cs5020089
ACS Catal. 2015, 5, 1303−1312

ACS Catalysis
Research Article
rather diluted reaction mixtures (e.g., 0.05 M) are processed;
this is not practical for preparative purposes.
established from 1.000 g of Pd/L standard (CPI International).
Indium served as the internal standard.
General Procedure for Microwave Batch Experiments.
In a 2 mL Pyrex vial were placed the Pd catalyst (0.5 mol %), 4-
iodobenzonitrile (0.5 mmol), phenylboronic acid (Suzuki−
Miyaura) (1.5 equiv) or n-butyl acrylate (Mizoroki−Heck) (1.5
equiv), the corresponding base (see Tables S1−S4 in the
Supporting Information), and 1 mL of solvent. The vial was
capped and heated in a single-mode microwave reactor
(Biotage Initiator) for 15 min at the desired temperature.
The reaction mixture was subsequently cooled to room
temperature with compressed air, filtered, and analyzed by
GC FID.
Continuous-Flow Procedures. Procedure 1: Flow Re-
actions in MeCN as Solvent and TEA or DIPEA as Base
(Figures 3, 5, 7, and 8). A stock solution containing 4-
iodobenzonitrile (0.5 M), phenylboronic acid (Suzuki−
Miyaura) or n-butyl acrylate (Mizoroki−Heck) (1.5 equiv),
and the corresponding base (1.5 equiv) in MeCN was
prepared. A fresh catalyst cartridge was placed in the X-Cube
flow reactor, and the system was set at the desired temperature
and 20 bar of back pressure while pure acetonitrile was pumped
at a rate of 0.2 mL min−1. When the temperature and pressure
were stable, the reactor inlet was switched to the reaction
mixture, and 20 mL of the stock solution was pumped through
the reactor before switching the inlet back to solvent. The
crude reaction mixture was collected from the output, and the
conversion was monitored by GC-FID.
The role of the covalent bonding of the ligand with the silica
matrix for the stability of the SilicaCat DPP-Pd catalyst could
be assessed by using a packed-bed reactor of unmodified silica
loaded with Pd(PPh3)4. In this case, a rapid metal leaching and
drop of conversion was observed during a continuous Suzuki−
Miyaura reaction.
Given the problem of metal leaching and catalyst stability
observed in most continuous-flow metal-catalyzed cross-
coupling reactions,7,9 it appears that in many cases it would
be more appropriate to use a suitable homogeneous metal
catalyst/ligand system. This would lead to more predictable and
reproducible transformations, as the catalyst loading could be
accurately adjusted. If necessary, recycling of the homogeneous
metal catalyst could also be implemented and likely future
research in this area will become increasingly important.32
Notably, the SiliaCat DPP-Pd catalyst system described herein
has shown superior leaching resistance and stability with
respect to the other three case studies. Further development of
this or other silica-based immobilized catalysts might lead to
more leaching-resistant supports in the future, at least on a
laboratory scale. Other strategies to circumvent the problem of
metal leaching in continuous-flow cross-couplings using
immobilized catalysts, such as “capture and release” systems33
and recirculating reactors,34 can also be useful on a small scale
when telescoped reactions are not required.
Procedure 2: Flow Reactions in THF/EtOH/H2O and K2CO3
as Base (Figure 10). Two separate stock solutions (A and B)
were prepared. Solution A contained 4-iodobenzonitrile (0.83
M) in THF. Solution B contained phenylboronic acid (Suzuki−
Miyaura) or n-butyl acrylate (Mizoroki−Heck) (0.45 M) and
K2CO3 (0.55 M) in 1/1 EtOH/H2O. A modified flow setup
incorporating two feeds was employed. Feed A was set to 0.055
mL min−1 using THF as solvent. Feed B containing 1/1 EtOH/
H2O was set to 0.155 mL min−1. At these flow rates the
substrate/coupling partner proportion is 1/1.25 and the
substrate/ base proportion is 1/1.5. A fresh catalyst cartridge
was placed in the X-Cube flow reactor, and the system was set
at the desired temperature and 20 bar of back pressure while
pure THF and EtOH/H2O were pumped through the system.
When the temperature and pressure were stable, the reactor
inlets were switched to the corresponding stock solutions A and
B. The two reaction streams were mixed immediately before
the heated column reactor in a T-mixer. The system was run
until 10 mmol of the substrate had been processed (8.3 mL of
solution A) before switching the inlets back to solvent. The
crude reaction mixture was collected from the output, and the
conversion was monitored by GC-FID.
EXPERIMENTAL SECTION
■
General Considerations. GC-FID analysis was performed
on a standard GC instrument with a flame ionization detector,
using an HP5 column (30 m, 0.250 mm, 0.025 mm). After 1
min at 50 °C the temperature was increased in 25 °C min−1
steps up to 300 °C and kept at 300 °C for 4 min. The detector
gas for the flame ionization was H2 and compressed air (5.0
quality). Curves drawn in the conversion profiles correspond to
average plots of the experimental data. Solvents and chemicals
were obtained from standard commercial vendors and were
used without any further purification. The following catalyst
cartridges were obtained from ThalesNano: Pd Tetrakis,
polymer bound (0.5 mmol of Pd/g, THS X1178, ca. 565 mg
of catalyst/cartridge); Pd Fibrecat 1001 (0.5 mmol of Pd/g,
THS X1187, ca. 300 mg of catalyst/cartridge); Pd EnCat
TPP30 (0.4 mmol of Pd/g, THS X1125, ca. 530 mg of catalyst/
cartridge); SiliaCat DPP-Pd (0.2 mmol of Pd/g, ca. 220 mg of
catalyst/cartridge). A fresh cartridge was used for every
continuous-flow experiment described herein.
Leaching Studies. The amount of Pd leached from the
immobilized catalyst for each of the flow runs was determined
by ICPMS analysis of the crude reaction mixture. Thus, the
solution collected from the reactor was evaporated under
reduced pressure to remove all volatile compounds. The
resulting residue was dissolved in acetonitrile/concentrated
nitric acid to give a homogeneous solution, and the solution
was diluted with nitric acid to 40 mL and placed in a vial for
microwave digestion. Microwave-assisted acid digestion was
carried out in an MLS UltraClave IV instrument. The
temperature was ramped up in 30 min to 250 °C and kept at
this temperature for a further 30 min. After appropriate dilution
Pd was quantitatively determined at m/z 105 with an Agilent
7500ce inductively coupled plasma mass spectrometer. A
calibration was performed with an external calibration curve
ASSOCIATED CONTENT
■
S
* Supporting Information
Supplementary Tables and Figures. This material is available
free of charge via the Internet at http://pubs.acs.org.The
following file is available free of charge on the ACS Publications
website at DOI: 10.1021/cs5020089.
Preliminary MW-batch experiments, ICPMS data,
continuous-flow conversion profiles, separation of the
reaction components within the catalysts, and continu-
ous-flow experiment with packed unmodified silica
(PDF)
1311
DOI: 10.1021/cs5020089
ACS Catal. 2015, 5, 1303−1312

ACS Catalysis
Research Article
(25) (a) Lee, C. K. Y.; Holmes, A. B.; Ley, S. V.; McConvey, I. F.; Al-
Duri, B.; Leecke, C. A.; Santos, R. C. D.; Seville, J. P. K. Chem.
Commun. 2005, 2175−2177. (b) Leeke, G. A.; Santos, R. C. D.; Al-
Duri, B.; Seville, J. P. K.; Smith, C. J.; Lee, C. K. Y.; Holmes, A. B.;
McConvey, I. F. Org. Process Res. Dev. 2007, 11, 144−148.
(c) Baxendale, I. R.; Griffiths-Jones, C. M.; Ley, S. V.; Tranmer, C.
K. Chem. Eur. J. 2006, 12, 4407−4416.
AUTHOR INFORMATION
Corresponding Authors
*E-mail for D.C.: david.cantillo@uni-graz.at, .
*E-mail for C.O.K.: oliver.kappe@uni-graz.at.
Notes
■
The authors declare no competing financial interest.
(26) (a) Pandarus, V.; Gingras, G.; Bel
Pagliaro, M. Org. Process Res. Dev. 2012, 16, 117−122. (b) Lemay, M.;
Pandarus, V.; Simard, M.; Marion, O.; Tremblay, L.; Beland, F. Top.
Catal. 2010, 53, 1059−1062.
(27) (a) Munoz, J. M.; Alcazar, J.; de la Hoz, A.; Dıaz-Ortiz, A. Adv.
́
and, F.; Ciriminna, R.;
ACKNOWLEDGMENTS
■
́
R.G. gratefully acknowledges the University of Ferrara (grant:
Fondo Giovani-2012) for a fellowship. We thank ThalesNano
for providing the catalyst cartridges and technical support.
́
́
̃
Synth. Catal. 2012, 354, 3456−3460. (b) Egle, B.; Munoz, J. M.;
̃
́
Alonso, N.; De Borggraeve, W. M.; de la Hoz, A.; Dıaz-Ortiz, A.;
Alcazar, J. J. Flow. Chem. 2014, 4, 22−25.
(28) (a) Pandarus, V.; Gingras, G.; Bel
Pagliaro, M. Org. Process Res. Dev. 2014, 18, 1550−1555. (b) Pandarus,
V.; Gingras, G.; Beland, F.; Ciriminna, R.; Pagliaro, M. Org. Process Res.
Dev. 2014, 18, 1556−1559.
(29) Pandarus, V.; Ciriminna, R.; Gingras, G.; Bel
́
REFERENCES
■
́
and, F.; Ciriminna, R.;
(1) Johansson Seechurn, C. C. C.; Kitching, M. O.; Colacot, T. J.;
Snieckus, V. Angew. Chem., Int. Ed. 2012, 51, 5062−5085.
́
(2) (a) Science of Synthesis: Cross Coupling and Heck Type Reactions;
Molander, G. A., Wolf, J. P., Larhed, M., Eds.; Georg Thieme Verlag:
Stuttgart, Germany, 2013; Vols. 1−3. (b) Metal Catalyzed Cross-
́
and, F.; Drobod,
M.; Jina, O.; Pagliaro, M. Tetrahedron Lett. 2013, 54, 1129−1132.
(30) Brazier, J. B.; Nguyen, B. N.; Adrio, L. A.; Barreiro, E. M.;
Leong, W. P.; Newton, M. A.; Figueroa, S. J. A.; Hellgardt, K.; Hii, K.
K. M. Catal. Today 2014, 229, 95−103.
(31) Kobayashi, S.; Miyamura, H. Aldrichim. Acta 2013, 46, 3−19.
(32) (a) Li, P.; Moore, J. S.; Jensen, K. F. ChemCatChem 2013, 5,
1729−1733. (b) Ormerod, D.; Bongers, B.; Porto-Carrero, W.; Ciegas,
S.; Vijt, G.; Lefevre, N.; Lauwers, D.; Brusten, W.; Buekenhoudt, A.
RSC Adv. 2013, 3, 21501−21510. (c) Peeva, L.; Arbour, J.; Livingston,
A. Org. Process Res. Dev. 2013, 17, 967−975. (c) Peeva, L.; da Silva
Burgal, J.; Vartak, S.; Livingston, A. G. J. Catal. 2013, 306, 190−201.
(33) Suzuki, Y.; Laurino, P.; McQuade, D. T.; Seeberger, P. H. Helv.
Chim. Acta 2012, 95, 2578−2588.
Coupling Reactions and More; de Meijere, A., Brase, S., Oestreich, M.,
̈
Eds.; Wiley-VCH: Weinheim, Germany, 2014.
(3) (a) Negishi, E.-i. Angew. Chem., Int. Ed. 2011, 50, 6738−6764.
(b) Suzuki, A. Angew. Chem., Int. Ed. 2011, 50, 6722−6737.
(4) (a) Bariwal, J.; van der Eycken, E. Chem. Soc. Rev. 2013, 42,
9283−9303. (b) Monnier, F.; Taillefer, M. Angew. Chem., Int. Ed.
2009, 48, 6954−6971. (c) Ley, S. V.; Thomas, A. W. Angew. Chem., Int.
Ed. 2003, 42, 5400−5449.
(5) (a) Takahashi, T.; Kanno, K.-i. In Modern Organonickel Chemistry;
Tamaru, Y., Ed.; Wiley-VCH: Weinheim, Germany, 2005; pp 41−53.
(b) Copper-Mediated Cross-Coupling Reactions; Evano, G., Blanchard,
N., Eds.; Wiley: Hoboken, NJ, 2014.
́
(6) (a) Molnar, A. Chem. Rev. 2011, 111, 2251−2320. (b) Anderson,
(34) (a) Pavia, C.; Ballerini, E.; Bivona, L. A.; Giacalone, F.; Aprile,
C.; Vaccaro, L.; Gruttadauria, M. Adv. Synth. Catal. 2013, 355, 2007−
2018. (b) Petrucci, C.; Strappaveccia, G.; Giacalone, F.; Gruttadauria,
M.; Pizzo, F.; Vaccaro, L. ACS Sustainable Chem. Eng. 2014, 2, 2813−
2819.
E. B.; Buchmeiser, M. R. ChemCatChem 2012, 4, 30−33.
(7) Noel, T.; Buchwald, S. L. Chem. Soc. Rev. 2011, 40, 5010−5029.
̈
(8) Frost, C. G.; Mutton, L. Green Chem. 2010, 12, 1687−1703.
(9) Cantillo, D.; Kappe, C. O. ChemCatChem 2014, 6, 3286−3305.
(10) (a) Pagliaro, M.; Pandarus, V.; Ciriminna, R.; Bland, F.; Demma
̀
Cara, P. ChemCatChem 2012, 4, 432−445. (b) Ananikov, V. P.;
Beletskaya, I. P. Organometallics 2012, 31, 1595−1604. (b) Kashin, A.
S.; Ananikov, V. P. J. Org. Chem. 2013, 78, 11117−11125.
(11) Glasnov, T. N.; Findenig, S.; Kappe, C. O. Chem. Eur. J. 2009,
15, 1001−1010.
(12) Yin, L.; Liebscher, J. Chem. Rev. 2007, 107, 133−173.
(13) de Vries, J. G. Dalton Trans. 2006, 421−429.
(14) Glasnov, T. N.; Kappe, C. O. Chem. Eur. J. 2011, 17, 11956−
11968.
(15) The ThalesNano X-Cube system is no longer commercially
available and has been replaced by the H-Cube Pro, which enables
working with identical cartridges, but under slightly decreased
maximum conditions (150 °C, 100 bar) in “no hydrogen” mode.
(16) Fenger, I.; Le Drian, C. Tetrahedron Lett. 1998, 39, 4287−4290.
(17) Trinh, T. N.; Hizartidis, L.; Lin, A. J. A.; Harman, D. G.;
McCluskey, A.; Gordon, C. P. Org. Biomol. Chem. 2014, 12, 9562−
9571.
(18) Carole, W.; Colacot, T. J. Chim. Oggi 2010, 28, XXIII−XXV.
(19) (a) Colacot, T. J.; Carole, W. A.; Neide, B. A.; Harad, A.
Organometallics 2008, 27, 5605−5611. (b) Colacot, T. J. Top. Catal.
2008, 48, 91−98.
̈
́
(20) Borcsek, B.; Bene, G.; Szirbik, G.; Dorman, G.; Jones, R.; Urge,
L.; Darvas, F. ARKIVOC 2012, 5, 186−195.
(21) Estrada, G. O. D.; Flores, M. C.; Silva, J. F. M.; de Souza, R. O.
M. A.; e Miranda, L. S. M. Tetrahedron Lett. 2012, 53, 4166−4168.
(22) Ramarao, C.; Ley, S. V.; Smith, S. C.; Shirley, I. M.; DeAlmeida,
N. Chem. Commun. 2002, 10, 1132−1133.
(23) (a) Kuang, Y. Y.; Chen, F. R. Helv. Chim. Acta 2009, 92, 897−
903. (b) Carpita, A.; Ribecai, A. Tetrahedron Lett. 2009, 50, 204−207.
(24) Broadwater, S. J.; McQuade, D. T. J. Org. Chem. 2006, 71,
2131−2134.
1312
DOI: 10.1021/cs5020089
ACS Catal. 2015, 5, 1303−1312