Water-stabilized three- and four-atom palladium clusters as highly active catalytic species in ligand-free C-C cross-coupling reactions

Article abstract of DOI:10.1002/anie.201303188

Elite cliques: Palladium clusters with three and four atoms were found to be the catalytically active species for ligand-free palladium-catalyzed C-C bond-forming reactions (see picture). These palladium cluster species could be stabilized in water and stored for long periods of time for use on demand with no loss of activity. High yields of products and turnover frequencies (TOFs) of up to 10⁵ h^-1 were observed. Copyright

Full text of DOI:10.1002/anie.201303188

Angewandte
Chemie
DOI: 10.1002/anie.201303188
Palladium Catalysis
Water-Stabilized Three- and Four-Atom Palladium Clusters as Highly
À
Active Catalytic Species in Ligand-Free C C Cross-Coupling
Reactions**
Antonio Leyva-Pꢀrez, Judit Oliver-Meseguer, Paula Rubio-Marquꢀs, and Avelino Corma*
Palladium-catalyzed cross-coupling reactions are recognized
as fundamental transformations in chemistry. The general
mechanism for these reactions under ligand-free conditions
relationship that allows the use of simpler palladium catalysts.
For example, the most common UV-B sunscreen on the
market, 2-ethylhexyl-p-methoxycinnamate, is manufactured
worldwide in multiton quantities by Heck coupling of the
corresponding bromoderivative with palladium nanoparticles
[
1]
[
11]
(
Scheme 1) involves an oxidative addition–reductive elimi-
[2]
nation cycle of a palladium(0) species generated in situ.
Despite considerable effort during last three decades, the
exact nature of the Pd catalytic species still remains unveiled.
on carbon as a catalyst at 1808C and without additional
0
[12]
ligands,
despite the fact that highly active and selective
palladium complexes can operate at lower reaction temper-
atures.
[4,5]
Different palladium compounds, such as salts,
com-
and nanoparti-
are able to catalyze the coupling of bromo com-
[
13,14]
plexes in oxidized or reduced form,
[
15–18]
cles,
pounds. Although palladium complexes catalyze the coupling
at mild temperatures without an induction period, a reaction-
induction time is observed for any palladium catalyst under
heating conditions (> 1308C), as explained by the mechanism
[
2,19]
Scheme 1. General mechanism for palladium-catalyzed cross-coupling
reactions under ligand-free conditions.
in Scheme 1.
Another experimental observation is the
degradation of palladium salts and complexes to form nano-
particles in few minutes under coupling conditions, and
some authors have proposed that the starting Pd active
species are these nanoparticles or some entities leached from
[
6]
0
It is accepted that the oxidative-addition step controls the
overall rate of the coupling, and catalyst optimization has
been a key point of research, since the reactivity of the
organic halide (I > Br@ Cl) is inversely proportional to its
[4,6,20–28]
them.
Herein, we show that regardless of the starting palladium
source, that is, whether it is a salt, a complex, or nanoparticles,
CÀC bond-forming reactions, such as Heck, Sonogashira,
[
3]
availability (Cl > Br@ I) and cost. Although successful
catalytic systems have also been developed with palladium
[
4,5]
salts,
very small amounts of phosphine palladium com-
enable the coupling of iodo-, bromo-, and chloro-
Suzuki, and Stille coupling reactions, of iodo and bromo
derivatives do not proceed at the higher rates until small
[
6–9]
plexes
[
16,28,29]
arenes under mild conditions. The implementation of these
catalysts in advanced organic synthesis has been satisfactory,
and many intermediates en route to pharmaceutical, agro-
chemical, polymer, and materials-science products have been
palladium clusters of three and four atoms are formed
in N-methylpyrrolidone (NMP) under heating: common
experimental reaction conditions for Heck coupling.
After identifying the exact active species, we found that the
three- and four-palladium-atom clusters can be stabilized by
water and stored for at least one month for later use, when
turnover frequencies between 10 and 10 molecules of
product per atom of palladium per hour were observed: the
highest values reported to date for some of the reactions
studied. We also show that not only water but also other
nucleophiles dislodge palladium clusters from nanoparticles,
and how the nature of the nucleophile and the nanoparticle
size both affect cluster formation. Although the palladium
clusters presented herein function at temperatures about
[
2,4–6,12]
[
10]
synthesized according to these protocols. However, indus-
try still relies in many cases on bromo rather than chloro
derivatives, since the former present a good cost/reactivity
5
6
[*] Dr. A. Leyva-Pꢀrez, J. Oliver-Meseguer, P. Rubio-Marquꢀs,
Prof. A. Corma
Instituto de Tecnologꢁa Quꢁmica, Universidad Politꢀcnica de
Valencia-Consejo Superior de Investigaciones Cientꢁficas
Avda. de los Naranjos s/n, 46022 Valencia (Spain)
E-mail: acorma@itq.upv.es
[
**] Financial support by Consolider-Ingenio 2010 (proyecto MULTICAT)
and the “Severo Ochoa” program is acknowledged. A.L.-P. and J.O.-
M. thank the ITQ for a contract. P.R.-M. thanks the MEC for an FPU
contract. The MALDI-TOF analysis was carried out in The
SCSIE_University of Valencia Proteomics Unit, a member of the
ISCIII ProteoRed Proteomics Platform.
7
08C higher than those used with the best palladium
complexes, and although they are not active with less
activated chloro derivatives, they are nevertheless useful for
the synthesis of a large number of compounds that are
accessible by palladium catalysis under these ligand-free
conditions, including industrially important products, such as
2-ethylhexyl-p-methoxycinnamate.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201303188.
Angew. Chem. Int. Ed. 2013, 52, 1 – 7
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

.
Angewandte
Communications
revealed that the palladium salts and complexes decomposed
to form 2–10 nm palladium nanoparticles in a few minutes
when the reaction was carried out under anhydrous con-
ditions (see Figure S3), in agreement with previous
[
7,36]
reports.
However, the reaction-induction period continues
for approximately two additional hours after the formation in
situ of the palladium nanoparticles; a similar induction period
was observed when palladium was incorporated as preformed
nanoparticles. The results clearly indicate that the palladium
nanoparticles generated in situ are not the catalytically active
species.
To find out which were the catalytically active palladium
species, we analyzed reaction samples before and after the
induction time by electrospray ionization mass spectrometry
(
ESIMS) and matrix-assisted laser desorption/ionization
coupled to time-of-flight (MALDI-TOF) mass spectrometry
see Figure S3). The results indicated that the reaction started
(
only when palladium clusters with a mass below 500 Da
appear, and not before. The evolution of the catalytic system
was also studied by UV/Vis spectroscopy. According to the
[
37]
jellium model, metal particles in the subnanometer regime
present a typical wavelength absorption in the visible region
as a function of their size, and the corresponding fluorescence
signal when irradiated at that wavelength. In contrast,
palladium nanoparticles present only a weak, broad, non-
[38]
fluorescence signal in the UV/Vis region.
When the cyclic vinyl bromide 1 in the presence of
palladium nanoparticles (3.9 nm) was submitted to anhydrous
coupling conditions and the reaction was followed by gas
chromatography–mass spectrometry (GC–MS) and absorp-
tion and emission UV/Vis spectroscopy, only the nonfluor-
escence plasmon band corresponding to the palladium nano-
particles at approximately 420 nm was observed during the
induction time. However, as soon as a new band at about 300–
Figure 1. Top graph: Conversion–time plot for the Heck coupling
between 1 and 2a with different palladium catalysts in anhydrous
media and in the presence of water: A) Pd(OAc) , no water; B) Pd
2
nanoparticles, no water; C) Pd nanoparticles with added water. Bottom
graph: Corresponding UV/Vis fluorescence emission spectra (irradia-
tion wavelength: 370 nm) for the anhydrous samples (curves A and B)
at different time intervals. Aqueous samples (curve C) gave a similar
emission spectrum to that at the maximum rate.
3
4
50 nm appeared, and clear fluorescence bands appeared at
50 nm upon irradiation, the reaction started, and conversion
was observed. Interestingly, the same absorption and emission
bands (450 nm) appeared immediately after irradiation at
approximately 350 nm appeared immediately when water
(1% v/v) was introduced into the reaction medium at reaction
time zero, and the induction period disappeared (see curve C
in Figure 1). The induction period also disappeared when
Figure 1 shows the kinetic results for the cross-coupling
reaction between the vinyl bromide 1 and the butyl acrylate
a under the catalysis of palladium acetate or monodisperse
palladium nanoparticles of approximately 3.9 nm in diameter
in solution (see Figure S1 in the Supporting Information). The
two palladium catalysts gave a typical sigmoidal curve under
anhydrous conditions. Other palladium precursors, including
[30]
2
[
31]
[35]
water was introduced and [Pd dba ], complexes 4 and 5,
2
3
and Pd/C (1 wt%) were used as catalysts. The addition of the
free oxime of complex 4 to the water-treated nanoparticles
did not produce any change in the reaction kinetics; this result
discards the possibility of any role of the ligand on the active
species (see Figure S4). The addition of more water and
acidification or basification did not change the kinetic profile
observed, which discards the possibility that the effect of
water is derived from the better dissolution of potassium
acetate (see Figure S4). Therefore, we can assume, in a first
approximation, that water promotes the formation of palla-
dium active sites from palladium nanoparticles in solution. In
control experiments without added palladium, no significant
conversion occurred.
[
Pd dba ] (dba = dibenzylideneacetone), an oxime pallada-
2 3
[
31–33]
cycle complex 4,
and a phosphine palladium complex
[
34,35]
5,
and Pd/C (1 wt%) showed a similar sigmoidal curve
under anhydrous conditions (see Figure S2). The turnover
frequencies (TOFs), measured after the induction period,
differed slightly depending on the starting palladium source.
High-resolution transmission electron microscopy (HRTEM)
together with dynamic light scattering (DLS) measurements
From the results obtained up to this point, we could
conclude that: a) palladium nanoparticles are not the active
species in the Heck reaction, but rather small palladium
2
www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 1 – 7
These are not the final page numbers!

Angewandte
Chemie
clusters, which are responsible for the UV/Vis signal at
50 nm; b) contact of the palladium nanoparticles with water
instantaneously generates palladium species that absorb at
50 nm. Thus, if water is introduced into the reaction media,
clusters not only catalyze the Heck coupling of vinyl bromide
1 but also that of the vinyl iodide 6 (see Figure S6). However,
4
4
the palladium clusters described above are formed instanta-
neously, and the induction period disappears.
To estimate the size of the palladium species whose
formation was promoted by water, we irradiated palladium
nanoparticles in aqueous solution at different wavelengths
from 296 to 548 nm, which correspond to palladium clusters of
the conversion of 6 into 7 was not higher than 50% even after
+
À
2
–13 atoms according to the jellium model. Significant
emission signals (450 nm) were only observed in the range
36–370 nm, which corresponds to small clusters with three
prolonged reaction times. The addition of BuN I (20 mol%)
inhibited the reaction, which suggests that the catalyst is
poisoned after the release of iodide. Similar results were
obtained with Pd(OAc) , [Pd dba ], and complexes 4 and 5.
3
and four metal atoms. We isolated the mixture of products 3
and confirmed that they did not absorb or emit in the blue
region and therefore did not interfere with the UV/Vis
measurements. The possibility that the adsorption of synthetic
ligands, solvent, or reactants, or changes in particle shape or
metal-surface roughness could produce the fluorescence
emission was discarded on the basis of control experiments
and a literature search. Indeed, we cannot talk of surface area
in molecular entities such as subnanometric metal clusters.
Treatment of the palladium nanoparticles with any other
reagent of the coupling reaction, including the vinyl bromide,
did not produce the UV/Vis-active species at room temper-
ature. However, if the vinyl bromide or the alkene was
introduced into a reactor that already contained the nano-
particles and water, the reaction started without an induction
period. These results strongly suggest that water promotes the
formation of subnanometric palladium clusters with three and
four atoms from nanoparticles and stabilizes these clusters,
and that these species catalyze the Heck coupling reaction
between 1 and 2a at the highest rate, either in the presence of
water or under anhydrous conditions. We cannot distinguish
whether the minimal catalytic activity found during the
induction time, before the boost in the reaction rate, is due
to palladium nanoparticles or minor amounts of three- and
four-atom palladium clusters that have already formed.
If we accept that water is able to promote the formation of
and stabilize the active sub-nanoclusters, then it should be
possible to prepare and store separately the three- and four-
atom-cluster catalytic species, to be used later on demand. To
check this possibility, we added Pd(OAc)2 or palladium
nanoparticles (3.9 nm) to an aqueous solution of NMP (water/
Pd molar ratio: 3000) and kept the mixture a flask for
2
2
3
To estimate the relative rate of conversion of aryl iodides and
bromides without the interference of poisoning, we tested
p-bromoiodobenzene (8c) as a substrate for the coupling
reaction. The results clearly indicate that iodides are activated
much faster than bromides (see Figure S7), in accordance
with the more facile oxidative addition of the former.
To study the formation of the palladium clusters under
anhydrous conditions, we prepared different palladium-on-
carbon solids with particle sizes ranging from 2.5 to 5.7 nm
and tested them in the Heck reaction of p-bromoanisole (8l)
under anhydrous or aqueous conditions (see Figure S8). The
results show that the amount of palladium that is leached into
solution without water is significantly lower, as measured by
inductively coupled plasma atomic emission spectroscopy
(ICP-AES) analysis of the reaction at different reaction times,
and that the reaction starts and proceeds only after 2 h. When
8l was absent, no palladium leached into solution. These
results indicate that the mechanism of palladium leaching in
the absence of water involves the oxidative addition of the
bromoarene to the outer shell of palladium atoms on the
nanoparticle. If this hypothesis is correct, the induction time
should correspond to the amount of palladium exposed on the
solid, as indeed is the case (see Figure S8). HRTEM and DLS
measurements have shown that palladium salts and com-
plexes decompose rapidly to nanoparticles with different sizes
under reaction conditions. The formation of nanoparticles
with different sizes explains the different TOF values found
for the homogeneous catalysts under anhydrous conditions
after the induction time. When water is present, the dislodg-
ing of palladium atoms from the nanoparticles is far more
efficient and also proportional to the amount of palladium
atoms exposed. These results together suggest that the
limiting step in the initiation of the reaction is the efficient
removal of palladium atoms from the nanoparticle surface,
either by water or the aryl bromide, and the subsequent
formation of the atomic clusters. They nicely fit the mecha-
1
month. After this period of time, part of the solution was
added as a catalyst to a mixture of vinyl bromide 1 and butyl
acrylate 2a. No induction period was observed, and the initial
reaction rate per palladium atom was the same as when the
clusters were freshly generated in situ by performing the
reaction in the presence of water (see Figure S5). Therefore,
we can conclude that a) clusters with three and four palladium
atoms are active species for the Heck CÀC bond-forming
[
22]
[4]
nisms proposed by de Vries
and Reetz, Amatore and
[
21]
[28]
[2,6]
Jutand, Ananikov and Beletskaya, and others. We also
found that the addition of water once the reaction had
stopped conferred new catalytic activity to the deactivated
palladium catalyst. It is accepted that deactivation of palla-
dium mainly occurs after agglomeration. If this hypothesis is
correct, the addition of water to commercially available
palladium black should improve the reaction. Indeed, we
observed just such an effect (see Figure S9).
reaction studied; b) water generates these active species from
palladium nanoparticles formed in situ from palladium
complexes, salts, or nanoparticles; c) it is possible to form
and stabilize ex situ these palladium aggregates with three
and four atoms in aqueous media and to store them for a long
time for used on demand. These subnanometric palladium
Angew. Chem. Int. Ed. 2013, 52, 1 – 7
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

.
Angewandte
Communications
Although the chemical basis for the formation of these
subnanometric palladium clusters is not clear at this stage,
a likely explanation is a disruption of the nanoparticle Stern
shell upon the addition of water. The relatively unhindered
nucleophilic oxygen atom would remove and somewhat
stabilize the subnanometric palladium clusters. The enhance-
ment of the reaction rate by minor amounts of water during
palladium-catalyzed reactions under the reaction conditions
used in this study is a widely observed effect that has not been
UV-B sunscreen 2-ethylhexyl-p-methoxycinnamate was syn-
thesized in high yield with a quantity of palladium of only
3 ppm. Remarkably, a variety of iodo and bromo derivatives
underwent Sonogashira coupling with an array of alkynes
5
with TOF values of up to 10 molecules of product per atom of
palladium per hour. As far as we know, this catalytic system is
[
43–45]
the most active reported for such a reaction.
ICP-AES
analysis of the reaction mixture showed that copper was
introduced by the reagents into the reaction mixture in
a quantity of less than 0.2 ppm, and no change in the catalytic
activity was found when CuI (300 ppm) was added. These
results confirm that the reaction is truly copper-free. When
vinyl bromide 1 was used as a substrate and kinetic experi-
ments were combined with UV/Vis spectroscopic measure-
ments of the reaction solution, we again found that palladium
clusters of three and four atoms were formed under anhy-
drous conditions at the time at which the reaction started and
products were formed (see Figure S12). As observed for the
Heck reaction, the clusters were formed rapidly, and the
induction time was suppressed if water was added. The
stability of the palladium clusters in solution over time was
also evaluated for the Sonogashira reaction (see Figure S13).
The results were similar to those obtained for the Heck
reaction and indicated unchanged catalytic activity for at least
1 month.
[39–42]
explained satisfactorily to date.
Thus, the stabilization of
the three- and four-atom active palladium clusters by water
provides a plausible explanation for the positive effect of
water on the rate of these reactions. If this hypothesis is
correct, other relatively unhindered nucleophiles, such as
amines, might give the same effect. Indeed, cyclohexylamine
promoted the formation of the palladium clusters more
efficiently than water, as assessed by fluorescence measure-
ments; consequently, the Heck coupling of bromobenzene
(8h) with 2a proceeded faster (see Figure S10). The reaction
mixture with water was darker than that with cyclohexyl-
amine, and the reaction mixture under anhydrous conditions
without cyclohexylamine was darker still (see Figure S10).
The catalytic activity decreased with the degree of substitu-
tion at the nitrogen atom of the amine, and for the same
degree of substitution, the more nucleophilic the amine was,
the higher the activity was (see also Figure S10). In other
words, both steric and electronic factors control the formation
of palladium clusters, and an optimum combination is found
in the relatively unhindered, hard nucleophile cyclohexyl-
amine. In accordance, extremely basic nucleophiles, such as
first-generation phosphazenes, promoted the coupling effi-
ciently. The use of secondary aliphatic amines also led to an
improvement in the reaction rate as compared to that with
water, but tertiary amines and cyclic amines performed less
well. Furthermore, the use of different alcohols only led to
a decrease in the reaction rate and the final conversion with
respect to the results with water, since their steric hindrance is
higher.
The Stille coupling was also evaluated under the reaction
conditions with the stabilized three- and four-atom palladium
clusters. A TOF value of 200000 molecules of product per
0
atom of palladium per hour was observed for the coupling
between bromoderivative 8g and tetrabutyltin (12a). This
level of catalytic activity is, to our knowledge, the highest ever
[
46]
reported for the Stille reaction. Again, catalyst activity was
only observed when palladium aggregates of three and four
metal atoms were formed (see Figure S14). Control experi-
ments without added palladium did not show any significant
conversion. The Suzuki coupling of phenylboronic acid with
different aryl halides proceeded under typical conditions and
with less palladium by one order of magnitude than the
amount required for Heck, Sonogashira, and Stille coupling
The extension of our study to other substrates for
palladium-catalyzed CÀC cross-coupling reactions was then
[
47]
reactions (Table 1).
As observed for the other reaction
envisaged. If the oxidative addition to the palladium clusters
is the rate-limiting step of the coupling reaction and the
formation of the clusters only depends on the presence of
water (or other nucleophiles) under the general reaction
conditions described above, other CÀC coupling reactions
types, when a palladium salt (Pd(OAc) ), a palladium com-
2
plex ([Pd dba ], palladacyles, phosphine complexes), or
2
3
palladium nanoparticles were used as catalysts, an induction
period was observed that disappeared when water was added.
Palladium nanoparticles, whether they had been treated
with water or not, were inactive as catalysts for the intra-
molecular amination of ortho-alkynyl anilines and for the
“click” reaction, two well-known reactions catalyzed by small
amounts of Lewis acids, including palladium salts (see Fig-
ure S15). These results indicate that the palladium clusters are
not positively charged, which is in agreement with the
expectation that catalysts with high electron density on the
palladium site should show superior reactivity in the oxidative
addition step. Measurement of the z potential of the reaction
mixture when the atomic palladium clusters were generated
with cyclohexylamine showed negative values, which infer
that the palladium clusters are anionic in nature. In any case,
the three- and four-atom palladium clusters seem to present
an electronic density suitable for the highly efficient oxidative
should in principle also be feasible with excellent catalytic
turnovers. Different iodo- and bromoarenes were coupled
with acrylates (Heck), alkynes (Sonogashira), organotin
compounds (Stille), and phenylboronic acids (Suzuki) in the
presence of the palladium clusters described above with
excellent initial catalytic activities and in good final yields
(
Table 1; see Figure S11 for a complete list of results with the
relevant structures). Note that the catalytic activity is
calculated with respect to the total amount of palladium
and not with respect to the amount of catalytically active
species, which according to the experiments with Pd/C may
account for less than 10% of the total palladium present.
The Heck reaction proceeded with different iodo- and
bromoarenes and acrylates in high yields, and the multiton
4
www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 1 – 7
These are not the final page numbers!

Angewandte
Chemie
Table 1: Results of the Heck, Sonogashira, Stille, and Suzuki cross-coupling reactions of different iodo-
and bromoarenes under palladium-cluster catalysis in the presence of water.
different coupling reactions. Fur-
thermore, chloro derivatives did
not react at all under these reaction
conditions, regardless of the cou-
pling partner and the amount of
catalyst used. This result indicates
that the palladium clusters are not
able to promote the oxidative
addition to start the catalytic cycle.
In conclusion, three- and four-
atom palladium clusters are
formed from palladium nanoparti-
cles and are responsible for the
catalytic activity of palladium salts,
complexes, and nanoparticles in
NMP under heating conditions.
These clusters can be stored in
aqueous solution to be used on
demand and catalyze the Heck,
Sonogashira, Stille, and Suzuki
coupling reactions of different
iodo and bromo derivatives under
ligand-free, industrially viable con-
ditions in high yields and with
unprecedented turnover frequen-
cies in some cases. The activation
of chloro derivatives is not possible
with this catalytic system. The
results presented herein connect
with those found for gold catalysts
and point to the potential of met-
allic clusters as highly active cata-
1
2
3
4
5
À1 [a]
[b]
X
R
R /R /R /R
Product Pd [ppm] TOF [h
]
Yield [%]
0
2
5
4
3
3
3
3
3
3
4
5
3
3
4
4
3
I
I
I
I
I
I
I
I
I
p-COMe (8a)
H (8b)
H (8b)
H (8b)
H (8b)
p-Br (8c)
p-Br (8c)
p-OMe (8d)
p-OMe (8d)
R =nBu (2a)
9a
9b
3
30
2.3ꢂ10
1.2ꢂ10
2.0ꢂ10
7.2ꢂ10
3.2ꢂ10
3.6ꢂ10
3.1ꢂ10
3.2ꢂ10
8.8ꢂ10
2.0ꢂ10
8.4ꢂ10
8.2ꢂ10
3.1ꢂ10
1.8ꢂ10
2.1ꢂ10
–
99
99
73
83
68
99
99
99
99
99
99
91
99
99
99
<5
2
R =nBu (2a)
3
R =nOct (10a)
11a
11b
15a
9c
11c
9d
300
300
30
300
300
300
3
3
300
30
300
300
300
0
3
0
3
R =Ph (10b)
5
R =Ph (14a)
2
R =nBu (2a)
3
R =nOct (10a)
2
R =nBu (2a)
2
R =2-ethylhexyl (2b)
9e
3
Br p-COH (8e)
Br p-COH (8e)
Br o-COH (8 f)
Br p-COMe (8g)
Br p-COMe (8g)
Br p-COMe (8g)
Br p-COMe (8g)
Br p-COMe (8g)
Br p-COMe (8g)
Br p-COMe (8g)
Br p-COMe (8g)
Br H (8h)
Br H (8h)
Br H (8h)
Br H (8h)
Br H (8h)
Br H (8h)
Br H (8h)
Br H (8h)
Br H (8h)
R =nOct (10a)
11d
11e
13a
9a
11 f
11g
11g
13b
13b
13c
15b
9b
3
R =Ph (10b)
4
R
=(nBu) (12a)
4
(
4)
2
R =nBu (2a)
3
R =Ph (10b)
3
R =nOct (10a)
3
R =nOct (10a)
4
5
[c]
R
R
R
=(nBu) (12a)
=(nBu) (12a)
=Ph, Me (12b)
2.0ꢂ10
–
70
(
(
(
4)
4)
4)
4
4
4
5
<5
4
4
5
2
3
3
3
3
3
2
3
3
7.9ꢂ10
1.0ꢂ10
8.8ꢂ10
5.2ꢂ10
6.6ꢂ10
1.2ꢂ10
2.4ꢂ10
2.0ꢂ10
8.0ꢂ10
7.6ꢂ10
–
2.0ꢂ10
4.9ꢂ10
1.1ꢂ10
4.7ꢂ10
–
44
61
3
[c]
R =Ph (14a)
0.3
300
300
300
300
300
300
300
30
2
R =nBu (2a)
64
99
99
96
99
94
97
87
<5
83
99
3
R =nOct (10a)
11a
11h
11i
[48,49]
3
lysts for organic reactions.
Our
R =m-tolyl (10c)
3
results do not contradict those
found for poly(vinylpyrrolidone)
R =o-tolyl (10d)
3
R =p-MeOC H (10e) 11j
6
4
3
R =o-MeOC H (10 f) 11k
(PVP)
solids,
palladium-supported
in which case it is
6
4
3
R =n-dodecyl (10g)
11l
15a
15a
11m
11n
9 f
[50,51]
5
R =Ph (14a)
thought that unsaturated palla-
dium atoms, such as those present
at the edges and corners of crystals,
are the active species. Indeed, the
smaller the palladium crystallites,
the larger the number of unsatu-
rated atoms will be. In our case, the
fact that three- and four-palladium-
atom clusters are present makes
the number of such unsaturated
sites very large, and the corre-
5
R =Ph (14a)
0
3
3
3
4
2
Br p-Me (8i)
Br o-OMe,p-naphthyl (8j) R =nOct (10a)
Br o,o,p-iPr (8k)
Br p-OMe (8l)
Br p-OMe (8l)
Br p-OMe (8l)
R =nOct (10a)
300
300
300
300
0
3
2
[c]
R =2-ethylhexyl (2b)
35
3
2
R =nBu (2a)
9d
9d
40
<5
50
2
R =nBu (2a)
3
2
R =Ph (10b)
11j
300
2.8ꢂ10
[
[
a] The initial turnover frequency (TOF ) was calculated on the basis of the total amount of palladium.
b] Product yields after 24 h. [c] Reaction temperature: 1608C.
0
addition of aryl iodides and bromides, and if the oxidative
addition is the rate-limiting step of the reaction, the rate of the
coupling depends only on the aryl halide and is independent
sponding TOF values very high.
of the second reaction partner. This mechanistic hypothesis is Experimental Section
corroborated by a comparison of the initial reaction rates of
General procedure: Potassium acetate (58 mg, 0.6 mmol) was placed
in a 1.5 mL vial equipped with a magnetic stir bar, and anhydrous
NMP (1 mL) was added. The desired volume of the palladium catalyst
solution, water (0.5 mmol), the corresponding halide (0.5 mmol), and
the coupling counterpart (acrylate or boronic acid: 0.6 mmol,
different aryl halides. For example, with iodobenzene (8b),
À1
a TOF value of around 5000 h was observed for the Heck,
0
Sonogashira, and Suzuki coupling, around half the value
À1
(
2000 h ) was observed for bromobenzene (8h) under the
1.2 equiv; alkyne or stannane: 1 mmol) were then added, and the
same reaction conditions, and with p-bromoacetophenone
vial was sealed and placed in a preheated oil bath at 1358C. The
reaction mixture was stirred, and aliquots of 0.05 mL were taken
À1
(
8g) a value of around 20000 h was observed for the four
Angew. Chem. Int. Ed. 2013, 52, 1 – 7
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!

.
Angewandte
Communications
periodically and diluted in diethyl ether (1.5 mL) for GC analysis with
nitrobenzene as an external standard. The same aliquots without
nitrobenzene were used for UV/Vis spectroscopic measurements.
[26] A. K. Diallo, C. Ornelas, L. Salmon, J. R. Aranzaes, D. Astruc,
Angew. Chem. 2007, 119, 8798; Angew. Chem. Int. Ed. 2007, 46,
8644.
[
27] K. Kçhler, W. Kleist, S. S. Prçckl, Inorg. Chem. 2007, 46, 1876.
[28] V. P. Ananikov, I. P. Beletskaya, Organometallics 2012, 31, 1595.
29] A. F. Schmidt, V. V. Smirnov, Top. Catal. 2005, 32, 71.
[30] L. Brandsma, H. D. Verkruijsee, Synthesis 1978, 290.
Received: April 16, 2013
Revised: August 13, 2013
Published online: && &&, &&&&
[
[
31] L. Botella, C. Nꢀjera, Angew. Chem. 2002, 114, 187; Angew.
Chem. Int. Ed. 2002, 41, 179.
Keywords: clusters · cross-coupling reactions ·
.
[32] D. A. Alonso, C. Nꢀjera, M. C. Pacheco, Adv. Synth. Catal. 2002,
44, 172.
33] D. A. Alonso, C. Nꢀjera, M. C. Pacheco, J. Org. Chem. 2002, 67,
588.
homogeneous catalysis · ligand-free conditions · palladium
3
[
[
5
[
[
[
1] X.-F. Wu, P. Anbarasan, H. Neumann, M. Beller, Angew. Chem.
010, 122, 9231; Angew. Chem. Int. Ed. 2010, 49, 9047.
2] N. T. S. Phan, M. Van Der Sluys, C. W. Jones, Adv. Synth. Catal.
006, 348, 609.
3] A. F. Littke, G. C. Fu, Angew. Chem. 2002, 114, 4350; Angew.
Chem. Int. Ed. 2002, 41, 4176.
34] S. D. Walker, T. E. Barder, J. R. Martinelli, S. L. Buchwald,
2
Angew. Chem. 2004, 116, 1907; Angew. Chem. Int. Ed. 2004, 43,
1871.
2
[
[
[
[
[
[
35] T. E. Barder, S. D. Walker, J. R. Martinelli, S. L. Buchwald, J.
Am. Chem. Soc. 2005, 127, 4685.
36] R. W. Y. Man, A. R. C. Brown, M. O. Wolf, Angew. Chem. 2012,
[
[
4] M. T. Reetz, J. G. de Vries, Chem. Commun. 2004, 1559.
5] A. H. M. de Vries, J. M. C. A. Mulders, J. H. M. Mommers,
H. J. W. Henderickx, J. G. de Vries, Org. Lett. 2003, 5, 3285.
6] D. A. Alonso, C. Nꢀjera, Chem. Soc. Rev. 2010, 39, 2891.
7] N. Marion, S. P. Nolan, Acc. Chem. Res. 2008, 41, 1440.
8] R. Martin, S. L. Buchwald, Acc. Chem. Res. 2008, 41, 1461.
9] D. S. Surry, S. L. Buchwald, Angew. Chem. 2008, 120, 6438;
Angew. Chem. Int. Ed. 2008, 47, 6338.
1
24, 11512; Angew. Chem. Int. Ed. 2012, 51, 11350.
37] J. Zheng, C. Zhang, R. M. Dickson, Phys. Rev. Lett. 2004, 93,
77402.
0
[
[
[
[
38] M. Hyotanishi, Y. Isomura, H. Yamamoto, H. Kawasaki, Y.
Obora, Chem. Commun. 2011, 47, 5750.
39] T. Rosner, J. Le Bars, A. Pfaltz, D. G. Blackmond, J. Am. Chem.
Soc. 2001, 123, 1848.
40] P. Cotugno, A. Monopoli, F. Ciminale, N. Cioffi, A. Nacci, Org.
Biomol. Chem. 2012, 10, 808.
[10] C. Torborg, M. Beller, Adv. Synth. Catal. 2009, 351, 3027.
[11] I. P. Beletskaya, A. V. Cheprakov, Chem. Rev. 2000, 100, 3009.
[12] IL 1996 – 118817 in Israeli, Bromine Compounds Ltd., Israel,
[
[
41] F. Zhao, M. Shirai, M. Arai, J. Mol. Catal. A 2000, 154, 39.
42] A. Sud, R. M. Deshpande, R. V. Chaudhari, Catal. Commun.
1999.
2007, 8, 183.
[
[
[
[
13] R. B. Bedford, Chem. Commun. 2003, 1787.
[
43] J.-C. Hierso, A. Fihri, R. Amardeil, P. Meunier, H. Doucet, M.
Santelli, V. V. Ivanov, Org. Lett. 2004, 6, 3473.
14] J. Dupont, C. S. Consorti, J. Spencer, Chem. Rev. 2005, 105, 2527.
15] D. Astruc, Tetrahedron: Asymmetry 2010, 21, 1041.
16] K. Okamoto, R. Akiyama, H. Yoshida, T. Toshida, S. Kobayashi,
J. Am. Chem. Soc. 2005, 127, 2125.
[
[
44] A. Kçllhofer, H. Plenio, Adv. Synth. Catal. 2005, 347, 1295.
45] G. Borja, A. Monge-Marcet, R. Pleixats, T. Parella, X. Cattoꢁn,
M. Wong Chi Man, Eur. J. Org. Chem. 2012, 3625.
[
17] C. P. Mehnert, D. W. Weaver, J. Y. Ying, J. Am. Chem. Soc. 1998,
[
[
[
46] D. A. Albisson, R. B. Bedford, P. Noelle Scully, S. E. Lawrence,
Chem. Commun. 1998, 2095.
47] R. K. Arvela, N. E. Leadbeater, M. S. Sangi, V. A. Williams, P.
Granados, R. D. Singer, J. Org. Chem. 2005, 70, 161.
48] J. Oliver-Meseguer, J. R. Cabrero-Antonino, I. Domꢂnguez, A.
Leyva-Pꢃrez, A. Corma, Science 2012, 338, 1452.
120, 12289.
[
18] B. M. Choudary, S. Madhi, N. S. Chowdari, M. L. Kantam, B.
Sreedhar, J. Am. Chem. Soc. 2002, 124, 14127.
[
[
19] M. Weck, C. W. Jones, Inorg. Chem. 2007, 46, 1865.
20] M. T. Reetz, E. Westermann, Angew. Chem. 2000, 112, 170;
Angew. Chem. Int. Ed. 2000, 39, 165.
[
49] A. S. K. Hashmi, Science 2012, 338, 1434.
[
[
[
21] C. Amatore, A. Jutand, Acc. Chem. Res. 2000, 33, 314.
22] J. G. de Vries, Dalton Trans. 2006, 421.
23] A. V. Gaikwad, G. Rothenberg, Phys. Chem. Chem. Phys. 2006,
[
50] J. Le Bars, U. Specht, J. S. Bradley, D. G. Blackmond, Langmuir
1999, 15, 7621.
[
51] P. J. Ellis, I. J. S. Fairlamb, S. F. J. Hackett, K. Wilson, A. F. Lee,
8
, 3669.
24] M. B. Thathagar, J. E. ten Elshof, G. Rothenberg, Angew. Chem.
006, 118, 2952; Angew. Chem. Int. Ed. 2006, 45, 2886.
Angew. Chem. 2010, 122, 1864; Angew. Chem. Int. Ed. 2010, 49,
[
[
1820.
2
25] A. V. Gaikwad, A. Holuigue, M. B. Thathagar, J. E. ten Elshof,
G. Rothenberg, Chem. Eur. J. 2007, 13, 6908.
6
www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 1 – 7
These are not the final page numbers!

Angewandte
Chemie
Communications
Palladium Catalysis
A. Leyva-Pꢁrez, J. Oliver-Meseguer,
P. Rubio-Marquꢁs,
A. Corma*
&&&& — &&&&
Water-Stabilized Three- and Four-Atom
Palladium Clusters as Highly Active
Catalytic Species in Ligand-Free CÀC
Elite cliques: Palladium clusters with
three and four atoms were found to be the
catalytically active species for ligand-free
palladium-catalyzed CÀC bond-forming
water and stored for long periods of time
Cross-Coupling Reactions
for use on demand with no loss of activity.
High yields of products and turnover
5
À1
frequencies (TOFs) of up to 10 h were
observed.
reactions (see picture). These palladium
cluster species could be stabilized in
Angew. Chem. Int. Ed. 2013, 52, 1 – 7
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7
These are not the final page numbers!