Gold/Palladium Alloy for Carbon-Halogen Bond Activation: An Unprecedented Halide Dependence

Article abstract of DOI:10.1002/asia.201500722

New catalytic activity of gold/palladium alloy nanoclusters (NCs) for carbon-halogen bond activation is demonstrated. In the case of an aryl chloride, the inclusion of gold in a bimetallic catalyst is indispensable to achieve the coupling reactions. Gold has the unique effect of stabilizing palladium, such that Pd²⁺ leached from clusters by means of spillover of chloride during oxidative addition. The thus-formed spillover intermediate further reacts heterogeneously in both Ullmann and Suzuki-type coupling reactions through a new type of mechanism. In the case of an aryl bromide, Ullmann coupling occurs through the spillover of bromide, similar to that of aryl chloride. However, a significant fraction of palladium also leached, which diminished the Ullmann coupling activity of the aryl bromide and, as a result, the order of reactivity was ArCl>ArBr. With regard to the activation of the C-Br bond towards a Suzuki-type reaction, the inclusion of a higher gold content in gold/palladium clusters stabilized palladium to prevent the leaching of Pd²⁺ from the clusters by means of spillover of bromide. The spillover intermediate reacts heterogeneously with PhB(OH)₂, palladium-rich gold/palladium, or pure palladium clusters; the oxidative addition of ArBr favors the extraction of palladium from the clusters, yielding Pd²⁺ intermediates. The extracted intermediates react homogenously (Pd^2+/Pd⁰ catalysis) with PhB(OH)₂, which results in the higher selectivity of the cross-coupling product. An initial step to observe such unprecedented halide dependency, together with the dynamic behavior of palladium on the surface of gold is the oxidative addition of Ar-X. A detailed insight into the first oxidative addition process was also examined by quantum chemical calculations.

Full text of DOI:10.1002/asia.201500722

DOI: 10.1002/asia.201500722
Full Paper
Cluster Compounds
Gold/Palladium Alloy for Carbon–Halogen Bond Activation: An
Unprecedented Halide Dependence
[a]
[b]
[c, d]
[a, d]
Raghu Nath Dhital, Karan Bobuatong, Masahiro Ehara,*
and Hidehiro Sakurai*
Abstract: New catalytic activity of gold/palladium alloy
nanoclusters (NCs) for carbon–halogen bond activation is
demonstrated. In the case of an aryl chloride, the inclusion
of gold in a bimetallic catalyst is indispensable to achieve
the coupling reactions. Gold has the unique effect of stabiliz-
action, the inclusion of a higher gold content in gold/palladi-
um clusters stabilized palladium to prevent the leaching of
2
+
Pd from the clusters by means of spillover of bromide.
The spillover intermediate reacts heterogeneously with
PhB(OH) , palladium-rich gold/palladium, or pure palladium
2
2
+
ing palladium, such that Pd
leached from clusters by
clusters; the oxidative addition of ArBr favors the extraction
2
+
means of spillover of chloride during oxidative addition. The
thus-formed spillover intermediate further reacts heteroge-
neously in both Ullmann and Suzuki-type coupling reactions
through a new type of mechanism. In the case of an aryl
bromide, Ullmann coupling occurs through the spillover of
bromide, similar to that of aryl chloride. However, a signifi-
cant fraction of palladium also leached, which diminished
the Ullmann coupling activity of the aryl bromide and, as
a result, the order of reactivity was ArCl>ArBr. With regard
to the activation of the CÀBr bond towards a Suzuki-type re-
of palladium from the clusters, yielding Pd intermediates.
2
+/
0
The extracted intermediates react homogenously (Pd Pd
catalysis) with PhB(OH) , which results in the higher selectivi-
2
ty of the cross-coupling product. An initial step to observe
such unprecedented halide dependency, together with the
dynamic behavior of palladium on the surface of gold is the
oxidative addition of ArÀX. A detailed insight into the first
oxidative addition process was also examined by quantum
chemical calculations.
Introduction
arises from leached metals after oxidative addition of ArÀX or
through the NC itself. However, a substantial number of re-
ports have appeared that point toward a homogeneous palla-
dium species performing catalysis through the leaching of pal-
Palladium nanoclusters (NCs) are prone to activate an aryl
halide by oxidative addition; the initial step occurs through
cross-coupling with various nucleophiles and Ullmann reduc-
[
2]
ladium species. It has been reported that the oxidative addi-
tion of aryl halides to palladium surface atoms favors their ex-
[
1]
tive homocoupling. A critical issue in palladium NC-catalyzed
activation of ArÀX is the mechanistic nature of these processes.
Thus, opinions differ as to whether palladium NC catalysis
2
+
2+
traction from the NCs to yield a Pd intermediate (ArPd X)
that reacts homogeneously with the nucleophile (Scheme 1,
[
3]
cycle A). Circumstantial evidence has been found to support
the hypothesis that oxidative addition promotes palladium
[
a] Dr. R. N. Dhital, Prof. Dr. H. Sakurai
Division of Applied Chemistry, School of Engineering
Osaka University, 2-1 Yamada-oka, Suita
Osaka 565-0871 (Japan)
2
+
leaching and the formation of Pd intermediates in cross-cou-
pling reactions and palladium(0) species are generated to close
[
4]
the catalytic cycle. The resulting palladium(0) species can fur-
ther participate in the reaction or redeposit on the surface of
the cluster. Alternatively, in the absence of a stabilizing ligand,
palladium(0) species could deactivate through the formation
of inert palladium black. Recently, we introduced a new strat-
egy to capture the leached palladium species by using gold
NCs as a stabilizing ligand; this is the case in pure palladium
Fax: (+81)6-68794593
E-mail: hsakurai@chem.eng.osaka-u.ac.jp
[
b] Dr. K. Bobuatong
Department of Chemistry, Faculty of Science and Technology
Rajamangala University of Technology Thanyaburi
Thanyaburi, Pathum Thani, 12110 (Thailand)
[
c] Prof. Dr. M. Ehara
Institute for Molecular Science
Research Center for Computational Science
Myodaiji, Okazaki 444-8585 (Japan)
E-mail: ehara@ims.ac.jp
[
4d]
NC-catalyzed hydrodechlorination (HDC). Consequently, gold
2
+
prevents Pd to leach out from the cluster by means of the
spillover of chloride and the reaction further proceeds through
the heterogeneous pathway (Scheme 1, cycle B).
[
d] Prof. Dr. M. Ehara, Prof. Dr. H. Sakurai
Japan Science and Technology Agency
With regard to bimetallic systems, for palladium alloyed with
gold, the catalytic activity and product selectivity is often ob-
served to be better than that with the monometallic counter-
ACT-C 4-1-8 Honcho, Kawaguchi, Saitama 332-0012 (Japan)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/asia.201500722.
Chem. Asian J. 2015, 10, 2669 – 2676
2669
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
spillover intermediate C is sup-
ported by quantum chemical
calculations (see below).
Moreover, the bimetallic NC
had a higher catalytic activity to-
wards chloroarenes than the
bromo equivalents (Scheme 2c).
For example, a comparison of
the observed reaction rates for
the decay of chlorotoluene (k
)
)
CÀCl
and
bromotoluene
(k
CÀBrl
showed that the value of k
CÀCl
À2 À1
(
5.6”10 h ) was nearly three
times that of
k
(2.0”
CÀBrl
À2 À1 [8a,10,9a]
1
0
h ).
The significant
Scheme 1. Metal NC-catalyzed CÀX (X=Cl, Br, I) bond transformations.
[5,6]
parts, probably because of the alloy effect.
In addition, the
homopalladium catalyst can become more stabilized with the
incorporation of gold owing to its unique electron-withdraw-
ing effect to neighboring primary metal atoms and through
[6]
the formation of a stable alloy. Herein, an unprecedented
halide dependence in the mechanistic pathway (homo- versus
heterogeneous), together with the peculiar effect of gold in
gold/palladium alloy NC-catalyzed carbon–halogen bond acti-
vation are demonstrated.
Results and Discussion
CÀCl and CÀBr Bond Activation in Ullmann Coupling
Ullmann coupling is a copper-mediated reductive homocou-
pling of aromatic halides, for which, however, stoichiometric
amounts of copper salts are required together with high reac-
[7]
tion temperatures (ꢀ2008C). We have recently introduced
a new strategy to overcome these limitations by using a gold/
palladium alloy stabilized by poly(N-vinyl-2-pyrrolidone) (PVP)
[
8a]
as a catalyst. The use of such catalyst enables us to perform
the Ullmann coupling of aryl chlorides at low temperature (27–
3
58C), which provides a high selectivity and excellent yield to
the desired products. The reactions were performed in the
N,N-dimethylformamide DMF/H O (1:2) solvent system in the
2
presence of 150 mol% of KOH at 35–458C under an argon at-
[8a,10]
mosphere (Scheme 2a).
Neither pure gold nor pure palla-
dium clusters could catalyze the reaction, which indicated that
alloying of the metals was crucial for coupling to occur. An es-
sential step, which determines the characteristic differences in
the catalytic activities of gold, gold/palladium, and palladium
NCs is the oxidative addition of chloroarene followed by the
spillover of chloride (intermediate C in cycle C, Scheme 2). Both
gold and palladium participate in breaking the CÀCl bond syn-
ergistically and this process might prevent the formation of
Scheme 2. Bimetallic gold/palladium NC-catalyzed Ullmann coupling of aryl
halide.
differences in the rates of activation of CÀCl and CÀBr showed
an unprecedented halide dependence on the catalytic activity
of bimetallic gold/palladium catalyst, otherwise the order of re-
activity should be ArBr>ArCl. In the course of our study on
the elucidation of such inverse halide dependence, we hap-
pened to observe a significant amount of palladium leaching
2
+
Pd intermediates. Intermediate C further reacts heterogene-
ously and ultimately delivers the coupling product. DMF acts
as a reducing agent and the hydrogen-transfer process from
DMF was found to be the rate-limiting step. The formation of
Chem. Asian J. 2015, 10, 2669 – 2676
www.chemasianj.org
2670
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
from the bimetallic clusters in the reaction with bromotoluene,
as confirmed by inductively coupled plasma atomic emission
spectroscopy (ICP-AES), UV/Vis spectroscopy, TEM, and XRD
pure gold clusters, which indicated that no oxidant (O ) was
2
available in the reaction because to occur it would require an
[
12d]
external homocoupling oxidant, such as oxygen.
In contrast,
[
10]
analysis. Thus, to explain the reactivity order of CÀCl>CÀBr,
we proposed that, unlike the aryl chloride (Scheme 2d,
path D), the organic bromide induced the leaching process
bimetallic clusters did exhibit coupling activities (Figure 1). For
(
Scheme 2d, path F), and therefore, the leached species disfa-
vored promotion of the Ullmann coupling reaction owing to
a lack of stabilizing ligand. The catalytic cycle, therefore, would
occur through a heterogeneous pathway (Scheme 2d, path D
or E), or otherwise the activity for the aryl bromide should be
high.
CÀCl Bond Activation in Suzuki-Type Coupling
The Suzuki–Miyaura coupling reaction, which leads to the for-
mation of new CÀC bonds, are key in organometallic chemis-
[1]
try. Aryl chlorides, which are more economical and readily
available substrates, have rarely been used owing to the high
activation barriers associated with difficult oxidative insertion
of palladium(0) species into the CÀCl bond. Our protocol for
the activation of ArÀCl by using the gold/palladium alloy was
also investigated for Suzuki–Miyaura coupling. The reaction be-
tween equimolar amounts of 4-bromobenzoic acid and
PhB(OH)2 was carried out in the presence of 300 mol% of
6 4 2 2
Figure 1. The coupling reaction between 4-Cl-C H CO H and PhB(OH) , as
catalyzed by Au, Pd, and Au/Pd NCs.
[
11]
K CO at 278C under aerobic conditions in aqueous media.
2
3
Because gold was unable to activate the CÀCl bond by under-
example, when Au Pd was used, 40% of the cross-coupling
going oxidative addition, the cross-coupling product, biphenyl-
0.8
0.2
product (1) was observed. To our surprise, compound 2 was
4
-carboxylic acid (1), was not observed, but biphenyl (2) and
still obtained in 38% yield (yield based on PhB(OH) ; i.e., 2:1
oxygenation side product, phenol (3), were observed in 74 and
2
molar ratio), which was competitive with 1, despite none of
oxygenated product 3 being produced. The activities for 1 and
2
6% yield, respectively, through an oxidative coupling path-
[12]
20À
way (Scheme 3, cycle G).
The depicted Au
cluster-cata-
2
continued to increase upon increasing the palladium content
up to 50%, and 50% of 1 and 50% of 2 was obtained with
Au Pd . The catalytic activity decreased as the amount of
lyzed oxidative coupling was recently reproduced by quantum
[12c]
chemical calculations.
Notably, boron peroxide, formed by
oxygen reduction during the homocoupling reaction, pro-
0.5
0.5
12b,c
gold decreased; however, the ratios of 1 and 2 were not al-
tered. Notably, under similar reaction conditions, gold-rich bi-
metallic clusters exhibited better activities than those of palla-
motes the conversion of PhB(OH) into 3 as side product.
2
To avoid oxidative homocoupling of PhB(OH)2 (Scheme 3,
cycle G), reactions were performed under degassing (freeze–
pump–thaw cycle) conditions. No coupling was observed with
dium-rich bimetallic clusters. For example, Au Pd catalyst af-
0.8
0.2
forded 40% of 1 and 38% of 2, instead only 8% of each 1 and
was observed with the Au Pd catalyst. This result indicat-
2
0.2
0.8
ed that, although pure gold had no activity, the inclusion of
higher contents of gold as the nearest heteroatoms had a sig-
nificant role in promoting the coupling reaction. Because of
the difficulty in the oxidative insertion of palladium(0) species
into the high-strength ArÀCl bond, no reaction was observed
with pure palladium. Notably, in the previous reports, Pd/PVP-
catalyzed coupling of aryl iodides or bromides as coupling
[
13]
partners were depicted, but not with aryl chlorides.
The kinetics data showed the formation of constant ratios of
[
11]
1
and 2 over each time interval. This suggests that both
products are formed in a single catalytic cycle and the reaction
2
+
0
is unlikely to follow the conventional Pd /Pd mechanism. In
addition, tetrahydroxydiboron was detected as a side product
1
1
from PhB(OH) by B NMR spectroscopy during the course of
2
[
11]
reaction, which indicated that the homocoupling product (2)
Scheme 3. Gold NC-catalyzed oxidative homocoupling of PhB(OH)
2
.
ND=not determined, T.M.=transmetallation, R.E.=reductive elimination.
probably formed through a formal metathesis-type reaction. To
Chem. Asian J. 2015, 10, 2669 – 2676
www.chemasianj.org
2671
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
determine that the catalytic process was truly heterogeneous,
the metal content in the filtrate after the reaction was analyzed
by ICP-AES. The leached metals (Au and Pd) were below the
limits of detection (<50 ppb). Although we aware that reclus-
terization processes may occasionally lead to inconclusive re-
sults, it is believed that this experiment is likely to support the
idea that the process is heterogeneous. Although the details of
the reaction mechanisms are still unclear, the mechanism in-
volving a single metal (Au or Pd) catalytic site is excluded and
a new model of oxidative addition of the CÀCl bond on the
surface of gold/palladium (intermediate C, Scheme 4) is pro-
6 4 2 2
Figure 2. The coupling reaction between 4-Br-C H CO H and PhB(OH) cata-
lyzed by Au, Pd, and Au/Pd NCs.
in the case of 4-bromobenzoic acid, a higher yield of 2 main-
tains competition with the formation of 1 in the presence of
gold-rich gold/palladium clusters, and 2 decreased with de-
creasing gold content in bimetallic catalysts. Such a unique
trend reveals that the reaction mechanisms are unlikely to be
similar in CÀCl or CÀBr bond activation and probably different
in the oxidative addition step.
Scheme 4. Possible heterogeneous pathway in the gold/palladium-catalyzed
coupling reaction
.
posed. The thus-formed spillover intermediate C reacts hetero-
geneously to give two equivalent of cross-coupling product (1)
and one equivalent of 2. Because this is completely different
from conventional Suzuki coupling, we therefore refer to the
present coupling reaction hereafter as Suzuki-type coupling,
which only occurs on the surface of the gold/palladium alloy.
Two exclusively different catalytic cycles (hetero- and homo-
geneous) are depicted in Scheme 5. Similar to aryl chloride,
spillover of bromide (Scheme 5, intermediate I, cycle I) was pro-
posed in the gold-rich gold/palladium-catalyzed reaction. The
thus-formed intermediate I is proposed to react heterogene-
ously with PhB(OH) and affords products 1 and 2 in a single
2
catalytic cycle. Two factors are suspected for the reactions pro-
CÀBr Bond Activation in Suzuki-Type Coupling
Consequently, coupling reactions between 4-bromobenzoic
acid and PhB(OH) were also carried out under anaerobic con-
2
ditions (Figure 2). As expected, when Au Pd was used, 34%
0
.8
0.2
of 1 was observed. Compound 2 was still obtained in 34%
yield, which was competitive with 1. Unlike the reaction with
4
-Cl-C H CO H, the yield of 1 continued to increase and that of
6 4 2
2
decreased upon increasing the palladium content in gold/
palladium clusters. For example, 48% of 1 formed with a palla-
dium content 60% in the bimetallic catalyst and the selectivity
of 1 increased with increasing amount of palladium in gold/
palladium, accordingly. Product
2 was maximized with
Au Pd and decreased with decreasing gold content in gold/
0.6
0.4
palladium. Finally, quantitative (>99%) yield of cross-coupling
product 1 and no homocoupling product (2) was observed
with pure palladium clusters.
In summary, the trend for the formation of 1 and 2 in the
coupling between 4-halobenzoic acids and PhB(OH) is depen-
2
dent on both the nature of halides as a coupling partner and
compositions of metals in bimetallic clusters. With 4-chloro-
benzoic acid, products 1 and 2 were observed in competition
in all compositions of gold and palladium in alloy. In contrast,
Scheme 5. Proposed reaction mechanisms for gold- (cycle I) and palladium-
rich (cycle J) gold/palladium-catalyzed coupling reactions.
Chem. Asian J. 2015, 10, 2669 – 2676
www.chemasianj.org
2672
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
gressing heterogeneously. First, the gold-rich alloy catalyst has
a higher coverage of gold atoms on the surface, and therefore,
bromide might go onto the surface of gold through the spill-
over mechanism. Thus, neighboring gold atoms probably pre-
(Au Pd and Au Pd ) before and after the reaction (Table 1
0.8
0.2
0.6
0.4
and Figure S3 in the Supporting Information). These results are
consistent with our hypothesis that gold could stabilize palla-
dium in gold-rich clusters and couplings might occur heteroge-
neously. In contrast, significant differences were observed in
the TEM images and XRD patterns of palladium-rich gold/palla-
dium clusters (Au Pd , Au Pd and pure Pd), which was
2
+
vent both the formation of Pd and the extraction of palladi-
um from the clusters. Second, if palladium leaches into the so-
2
+
0
lution, it should follow the conventional mechanism (Pd /Pd )
to give 1 exclusively and 2 should not compete with 1. On the
other hand, in palladium-rich gold/palladium alloys, a signifi-
cant amount of palladium is located on the surface of the clus-
ter. These surface palladium atoms have a homopalladium-like
0.4
0.6
0.2
0.8
probably due to changes in the morphologies of the catalysts
after use (Figure S2 in the Supporting Information). For exam-
ple, TEM images showed that after the reaction the sizes of
palladium-rich clusters increased by nearly 1 nm. XRD studies
of broad (111) peaks associated with a small face-centered
cubic crystal of gold/palladium catalysts before use were shift-
ed to a higher 2q value (towards Pd region) as a result of the
reactions (Table 1 and Figure S3 in the Supporting Informa-
tion). Figure 3 shows typical examples of XRD patterns of gold-
(
PdÀPd) character rather than that of heteronuclear gold–palla-
dium. Therefore, a large fraction of 4-bromobenzoic acids un-
dergo oxidative addition locally on palladium atoms to form
2
+
oxidized Pd , which leads to its extraction from the NCs
Scheme 5, cycle J, intermediate J). Intermediate J reacts homo-
(
geneously with PhB(OH) and ultimately delivers the cross-cou-
2
pling product as the final product. It is important to note that
the proposed intermediates I and J are supported by DFT cal-
culations, as discussed below. Amatore and co-workers proved
that gold@palladium (Au core and Pd shell) NC-catalyzed
[
4a]
Suzuki coupling occurred through the leaching of palladium.
Their report partly supports the current proposed hypothesis
that bimetallic gold/palladium catalysts with higher palladium
contents could catalyze the coupling reaction through a homo-
genous pathway.
To verify our hypothesis, we used several spectroscopic tech-
niques to examine structural and morphological changes in
the clusters through the reaction (Table 1). Elemental analysis
was performed by means of ICP-AES after the reactions. In the
reactions catalyzed by gold-rich gold/palladium clusters
Figure 3. XRD profiles of Au-rich Au0.8Pd0.2 (a) and Pd-rich Au0.2Pd0.8 (b) NC
before and after use.
rich Au Pd and palladium-rich Au Pd catalysts before and
0.8
0.2
0.2
0.8
(
Au Pd and Au Pd ), the amounts of gold and palladium
after the reaction. Notably, the broad peak (amorphous nature)
associated with the 111 plane of pure palladium before use
became sharp (crystalline nature) after the reaction.
These results suggest that palladium leaches out
during the course of reaction and probably redepo-
0.8
0.2
0.6
0.4
ions were below the limits of detection (<50 ppb of Au and
Table 1. Spectroscopic analysis of catalysts before and after use.
sits on the clusters, otherwise the morphologies of
[
a]
[b]
Catalyst
Size of catalysts [nm]
Before use
Leaching of Pd
(ICP-AES) [%]
Shift of 111
the clusters would not change. These results lead us
to conclude that a homogenous process occurs in
coupling catalyzed by palladium-rich gold/palladium
or pure palladium clusters.
After use
peak position
Au0.8Pd0.2
Au0.6Pd0.4
Au0.4Pd0.6
Au0.2Pd0.8
Pd1.0
3.1Æ1.1
3.8Æ1.3
4.0Æ1.1
3.9Æ1.2
4.0Æ1.1
3.2Æ1.0
3.7Æ1.2
4.9Æ1.4
4.9Æ1.2
5.0Æ1.0
<0.2
<0.1
0.40
0.77
1.3
no shift
no shift
[b]
[b]
[b]
0.20 (2q)
0.40 (2q)
0.42 (2q)
CÀCl Activation on Au/Pd Alloy
[
a] As determined by TEM. [b] Shift towards the higher 2q, as determined by XRD.
An important step that determines the characteristic
differences in the catalytic activity of monometallic
<
44 ppb of Pd), which indicated that the amounts of leached
and bimetallic gold/palladium NCs in the Ullmann coupling of
ArÀCl is oxidative addition, which results in CÀCl bond dissoci-
ation followed by the spillover of chloride (Scheme 2, interme-
palladium ions should be less than 0.2 and 0.1%, respectively.
Instead, 0.4, 0.77, and 1.3% leached palladium was observed
[
8a]
(
Table 1) in the reactions catalyzed by Au Pd , Au Pd , and
diate C).
The high catalytic activity of the gold/palladium
0
.4
0.6
0.2
0.8
pure Pd, respectively. In all cases, leached gold was below the
limits of detection. These results additionally support our hy-
pothesis that the reactions catalyzed by gold-rich clusters
might occur heterogeneously and the reactions catalyzed by
NCs is due to the substrate being adsorbed effectively onto
the alloy surface and subsequent activation barriers of oxida-
tive addition being significantly low, which is unlikely to occur
[
8a]
in the gold NC. The spillover of chloride prevents the forma-
[14]
2+
palladium-rich clusters might occur homogeneously.
tion of Pd species and the thus-formed intermediate C reacts
Next, we did not observe clear differences in the TEM
images and XRD patterns of gold-rich gold/palladium clusters
heterogeneously in both Ullmann and Suzuki-type coupling re-
actions. The spillover of chloride is likely to be a crucial factor
Chem. Asian J. 2015, 10, 2669 – 2676
www.chemasianj.org
2673
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
in these reactions only proceeding in the presence of a bimetal-
lic catalyst.
This leaching process results in the experimentally observed in-
verse halogen dependence, as shown below. The free energy
changes (DDG298) of the leaching processes in both ArÀBr and
ArÀCl were evaluated by using a solvation model density
CÀBr Activation on Au/Pd Alloy
(
SMD) continuum model developed by Truhlar and co-work-
[
20]
An unprecedented halide dependence on the reactivity (Cl>
Br) was observed in the Ullmann coupling of aryl halides. Let
us turn to CÀBr bond activation on the gold/palladium NC.
The coupling reaction occurs on gold/palladium NCs with an
alloy ratio of Au Pd –Au Pd , as seen in Figure 2. The rele-
ers, for various gold/palladium NCs with various ratios of pal-
ladium, namely, Au2 Pd NCs (n=0, 1, 4, 10, 20), which repre-
0Àn
n
sent, for example, gold, Au Pd , Au Pd , and palladium
0.8
0.2
0.5
0.5
NCs. Also, the leached species, C H Pd(H O) X, were considered
6
5
2
2
with the coordination of two water molecules in SMD (details
are given in the Supporting Information). The energetics of the
reaction shown in Equation (1) depend on the defect site and
gold/palladium ratio, and are summarized in Tables S2 in the
Supporting Information. The free energy changes in each gold/
palladium NC are plotted in Figure 5 for ArÀCl and ArÀBr. In
0.8
0.2
0.5
0.5
vant reaction center or site can be well represented by the
Au Pd NC model. Thus, we performed calculations for the
10
10
oxidative addition of ArÀBr on some structures of Au Pd
1
0
10
NCs, which were thermally accessible. Detailed results for CÀBr
bond activation at various sites on Au Pd NCs predicted by
1
0
10
[15,16]
the genetic algorithm (GA),
followed by DFT calculations,
Equation (1), the energy of Au2 Pd
is dependent on
nÀ1
0Àn
are given in Table S1 in the Supporting Information. DFT calcu-
a defect site of the Au2 Pd
NC owing to the leaching of
0Àn
nÀ1
[17]
lations with the M06 functional were performed by using
palladium complex, and therefore, the calculated free energy
changes have some range, as shown by the error bars in
Figure 5.
[18]
the Gaussian 09 suite of programs.
0À
The energetics of the oxidative addition of ArÀBr on Au
2
and Au Pd are nearly the same as those of ArÀCl (Fig-
10
10
[
8a,19]
ure 4).
The calculated activation energy barrier is too high
C H X þ 2 H O þ Au Pd ! C H PdðH OÞ X þ Au Pd
6
5
2
20Àn
n
6
5
2
2
20Àn
nÀ1
À1
À
(
123.9 kJmol ) in the case of Au , whereas it is significantly
20
ðX ¼ Cl, BrÞ
reduced on the Au Pd NC. Therefore, CÀBr bond activation
10
10
ð1Þ
proceeds favorably on the Au Pd NC for the ArÀCl case. The
1
0
10
reaction is initiated by the adsorption of ArÀBr to form a p
complex between ArÀBr and metal atoms on the facet site of
The leaching of surface palladium atoms from the gold/pal-
ladium NC could occur in the catalytic cycle of the ArÀBr case,
unless surrounding PVP domain prevented the desorption of
À
the Au20 NC and on the PdÀPd bridge site of the Au Pd NC.
10
10
It appears that the direct coordination effect plays an impor-
tant role in this step.
Two pathways are possible for
the dissociative adsorption of
ArÀBr. One is bond activation at
a single palladium atom to yield
Int_A2 followed by a spillover of
aryl and bromide to generate
Int_B2. The calculated energy
barriers of these steps are 42.0
À1
and 3.2 kJmol , respectively.
The direct dissociation of ArÀBr
is also possible through the acti-
À1
vation barrier of 29.5 kJmol .
The trend in the calculated acti-
vation barrier of CÀX dissocia-
tion on the gold/palladium NC is
consistent with the conventional
trend of the reactivity of aromat-
ic halides (Br>Cl).
Another relevant factor to
consider herein is the leaching
of surface palladium atoms from
the gold/palladium NC, which
might also be possible after ArÀ
X adsorption, for example, in the
case of Au2 Pd NC, the leach-
0Àn
n
ing of palladium(II) complex
Figure 4. Energy diagram for the oxidative addition of ArÀBr on Au20À and Au10Pd10 NCs. TS=transition state,
occurs, as shown in Equation (1). Int=intermediate.
Chem. Asian J. 2015, 10, 2669 – 2676
www.chemasianj.org
2674
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
in palladium-rich gold/palladium or pure palladium clusters.
2
+
The thus-formed Pd
intermediate reacted homogenously
2
+
0
(
Pd /Pd catalysis) with PhB(OH) and delivered the cross-cou-
2
pling product as the only product. This unique catalysis
bridges the gap between homo- and heterogeneous mecha-
nisms in organic transformations and may open up a new con-
cept for developing other bimetallic systems for CÀX bond ac-
tivation.
Experimental Section
Typical Procedure for Ullmann Coupling of Chloroarene or
Bromoarene
Figure 5. Calculated free energy changes (DDG298) of the palladium complex
Ullmann coupling reactions of chloro- or bromoarenes were carried
out by using an organic synthesizer (EYELA, PPS-2510). Chloro- or
bromoarene (0.1 mmol) and the required amount of base (300 or
leaching process for Au20ÀnPd
ArÀCl.
n
(n=0, 1, 4, 10, 20) in the case of ArÀBr and
1
50 mol%) were placed in a test tube (f=15 mm). A hydrosol of
Au–PVP, Pd–PVP, or Au/Pd–PVP (5 mm, 0.4 mL, 2 at%) was added
palladium complex; this process is unlikely in the case of ArÀ
Cl, except for the palladium-rich gold/palladium NC. Therefore,
catalysis in the activation of the CÀBr bond towards Ullmann
coupling might occur on the surface of the gold/palladium NC
through the formation of spillover intermediate Int_B2 (hetero-
geneous, path E in Scheme 2), similar to that of ArÀCl. The for-
mation of such a bromide spillover intermediate was supposed
to be a crucial factor for the formation of 2 through the heter-
ogeneous pathway (cycle I, intermediate I in Scheme 5) in the
gold-rich gold/palladium-catalyzed Suzuki-type reaction. How-
to DMF (1 mL) and Milli-Q water (1.6 mL; DMF/H O, 1:2 ratio, 3 mL
2
final volume) and argon was supplied from a balloon. The reaction
was stirred vigorously (1300 rpm) at the required temperature for
the desired time. The reaction was quenched with 0.5m HCl and
extracted with ethyl acetate (4”10 mL). GC was used to determine
the yields of products by using hexadecane as an internal stan-
dard. To determine the yield, the extracted organic layer was dried
over excess Na SO and evaporated in vacuo. The crude product
2
4
was isolated by preparative TLC (ethyl acetate in hexane). The for-
1
mation of product was confirmed by H NMR spectroscopic analy-
sis by comparison with an authentic sample.
2
+
ever, the process of leaching Pd
intermediates from
Au2 Pd to afford Au Pd and C H Pd(H O) Br is also en-
Typical Procedure for Suzuki–Miyaura Coupling between 4-
Halobenzoic Acid and Phenylboronic Acid
0Àn
n
20Àn
nÀ1
6
5
2
2
ergetically favorable (homogeneous pathway). Exothermicity in
the case of ArÀBr increases as the ratio of palladium in the
gold/palladium NC increases, as clearly seen in Figure 5. This is
consistent with experimental observations that the leaching of
palladium occurs in conventional Suzuki–Miyaura coupling
with a palladium-rich gold/palladium catalyst.
The coupling reactions were carried out by using an organic syn-
thesizer (EYELA, PPS-2510). Phenylboronic acid (18.3 mg,
0.15 mmol), 4-bromobenzoic acid, or 4-chlorobenzoic acid
(0.15 mmol) and K CO (63.5 mg, 0.45 mmol, 300 mol%) were
2 3
placed in a test tube (f=20 mm). A hydrosol of Au–PVP, Pd–PVP,
or Au/Pd–PVP (3 mL, 1 mm, 2 at%) was added and the mixture was
stirred vigorously (1300 rpm) at 278C for 24 h in the open air or
under degassed conditions. The reaction was quenched with 0.5m
HCl and the product was extracted with ethyl acetate (4”10 mL).
The extracted organic layer was dried (Na SO ) and concentrated in
Conclusion
New catalytic activities of bimetallic gold/palladium alloy for
CÀCl and CÀBr bond activations, together with unique halogen
dependency in the coupling reactions, were demonstrated.
The inclusion of gold had a remarkable effect on stabilizing
2
4
vacuo. The crude product was dissolved in [D ]acetone and ana-
6
lyzed by NMR spectroscopy. Dioxane was used as an internal NMR
standard to calculate the yields.
2
+
palladium and led to Pd
leaching from the clusters by
means of spillover of chloride. This spillover intermediate fur-
ther reacted heterogeneously in both Ullmann and Suzuki-type
coupling reactions. A significant fraction of palladium leached
out, which diminished the Ullmann coupling activity and, as
a result, the order of reactivity was ArÀCl>ArÀBr. In the case
of CÀBr activation, the gold-rich gold/palladium cluster had
a significant effect on stabilizing palladium and prevented
Acknowledgements
We would like to thank Setsiri Haesuwannakij, Department of
Functional Molecular Science, School of Physical Sciences,
Graduate University for Advanced Studies (SOKENDAI), and
Professor Yasuhiro Uozumi, Institute for Molecular Science, for
the TEM measurements. M.E. thanks Dr. B. Boekfa in Kasetsart
University for valuable discussions.
2
+
Pd from leaching. The spillover intermediate further reacted
heterogeneously with PhB(OH) and provided the cross-cou-
2
pling and metathesis-type homocoupling (derived from
Keywords: alloys · cluster compounds · CÀX activation · gold ·
PhB(OH) ) products in a single catalytic cycle. In contrast, the
2
palladium
oxidative addition of 4-bromobenzoic acid favored the extrac-
2
+
tion of palladium from the clusters to yield Pd intermediates
Chem. Asian J. 2015, 10, 2669 – 2676
www.chemasianj.org
2675
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[
[
1] a) N. Miyaura, A. Suzuki, Chem. Rev. 1995, 95, 2457–2483; b) A. F. Littke,
G. C. Fu, Angew. Chem. Int. Ed. 2002, 41, 4176–4211; Angew. Chem.
Chem. 2014, 3, 668–684; c) S. Karanjit, M. Ehara, H. Sakurai, Chem. Asian
J. 2015, DOI: 10.1002/asia.201500535; d) M. Moreno-MaÇas, M. PØrez, R.
Pleixats, J. Org. Chem. 1996, 61, 2346–2351; e) R. N. Dhital, A. Muruga-
doss, H. Sakurai, Chem. Asian J. 2012, 7, 55–59.
2
002, 114, 4350–4386; c) L. Yin, J. Liebscher, Chem. Rev. 2007, 107,
33–173; d) A. Dhakshinamoorthy, A. M. Asiri, H. Garcia, Chem. Soc. Rev.
015, 44, 1922–1947.
2] a) M. B. Thathagar, J. E. ten Elshof, G. Rothenberg, Angew. Chem. Int. Ed.
006, 45, 2886–2890; Angew. Chem. 2006, 118, 2952–2956; b) A. V.
1
2
[13] a) Y. Li, X. M. Hong, D. M. Collard, M. A. El-Sayed, Org. Lett. 2000, 2,
2385–2388; b) Y. Li, M. A. El-Sayed, J. Phys. Chem. B 2001, 105, 8938–
8943; c) Y. Li, E. Boone, M. A. El-Sayed, Langmuir 2002, 18, 4921–4925;
d) P. J. Ellis, I. J. S. Fairlamb, S. F. J. Hackett, K. Wilson, A. F. Lee, Angew.
Chem. Int. Ed. 2010, 49, 1820–1824; Angew. Chem. 2010, 122, 1864–
1868.
[14] In the presence of PVP matrix, the reclusterization process might be
faster and we are only able to detect partial amounts of leached palla-
dium by ICP-AES.
2
Gaikwad, A. Holuigue, M. B. Thathagar, J. E. ten Elshof, G. Rothenberg,
Chem. Eur. J. 2007, 13, 6908–6913.
3] a) S. MacQuarrie, J. H. Horton, J. Barnes, K. McEleney, H.-P. Loock, C. M.
Crudden, Angew. Chem. Int. Ed. 2008, 47, 3279–3282; Angew. Chem.
[
[
2
008, 120, 3324–3328.
4] a) P.-P. Fang, A. Jutand, Z.-Q. Tian, C. Amatore, Angew. Chem. Int. Ed.
011, 50, 12184–12188; Angew. Chem. 2011, 123, 12392–12396; b) Z.
2
[15] R. L. Johnston, Dalton Trans. 2003, 4193–4207.
Niu, Q. Peng, Z. Zhuang, W. He, Y. Li, Chem. Eur. J. 2012, 18, 9813–9817;
c) N. T. S. Phan, M. Van Der Sluys, C. W. Jones, Adv. Synth. Catal. 2006,
[16] F. Pittaway, L. O. Paz-Borbón, R. L. Johnston, H. Arslan, R. Ferrando, C.
Mottet, G. Barcaro, A. Fortunelli, J. Phys. Chem. C 2009, 113, 9141–9152.
[17] Y. Zhao, D. Truhlar, Theor. Chem. Acc. 2008, 120, 215–241.
[18] Gaussian 09, Revision B.01, M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E.
Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Men-
nucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian,
A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara,
K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O.
Kitao, H. Nakai, T. Vreven, J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro,
M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, T. Keith,
R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S.
Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox,
J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Strat-
mann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L.
Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dan-
nenberg, S. Dapprich, A. D. Daniels, O. Farkas, J. B. Foresman, J. V. Ortiz,
J. Cioslowski, D. J. Fox, Gaussian, Inc., Wallingford CT, 2010.
[19] K. Bobuatong, S. Karanjit, R. Fukuda, M. Ehara, H. Sakurai, Phys. Chem.
Chem. Phys. 2012, 14, 3103–3111.
3
48, 609–679; d) S. Karanjit, A. Jinasan, E. Somsook, R. N. Dhital, Y. Sato,
K. Motomiya, K. Tohji, H. Sakurai, Chem. Commun. 2015, 51, 12724–
2727.
1
[
5] a) F. Gao, D. W. Goodman, Chem. Soc. Rev. 2012, 41, 8009–8020; b) R.
Ferrando, J. Jellinek, R. L. Johnston, Chem. Rev. 2008, 108, 845–910;
c) G. J. Hutchings, Chem. Commun. 2008, 1148–1164; d) L. Zhang, A.
Wang, J. T. Miller, X. Liu, X. Yang, W. Wang, L. Li, Y. Huang, C.-Y. Mou, T.
Zhang, ACS Catal. 2014, 4, 1546–1553.
[
6] a) C. Rossy, J. Majimel, E. Fouquet, C. Delacôte, M. Boujtita, C. Labrug›re,
M. TrØguer-Delapierre, F.-X. Felpin, Chem. Eur. J. 2013, 19, 14024–
1
2
4029; b) J. B. Xu, T. S. Zhao, S. Y. Shen, Y. S. Li, Int. J. Hydrogen Energy
010, 35, 6490–6500.
[
[
7] a) F. Ullmann, J. Bielecki, Chem. Ber. 1901, 34, 2174–2185; b) F. Ullmann,
Justus Liebigs Ann. Chem. 1904, 332, 38–81.
8] a) R. N. Dhital, C. Kamonsatikul, E. Somsook, K. Bobuatong, M. Ehara, S.
Karanjit, H. Sakurai, J. Am. Chem. Soc. 2012, 134, 20250–20253; b) B.
Boekfa, E. Pahl, N. Gaston, H. Sakurai, J. Limtraku, M. Ehara, J. Phys.
Chem. C 2014, 118, 22188–22196.
9] a) R. N. Dhital, C. Kamonsatikul, E. Somsook, Y. Sato, H. Sakurai, Chem.
Commun. 2013, 49, 2542–2544; b) I. Blakey, Z. Merican, L. Rintoul, Y.-M.
Chuang, K. S. Jack, A. S. Micallef, Phys. Chem. Chem. Phys. 2012, 14,
[20] A. V. Marenich, C. J. Cramer, D. G. Truhlar, J. Phys. Chem. B 2009, 113,
6378–6396.
[
3
604–3611.
10] R. N. Dhital, C. Kamonsatikul, E. Somsook, H. Sakurai, Catal. Sci. Technol.
013, 3, 3030–3035.
11] R. N. Dhital, H. Sakurai, Chem. Lett. 2012, 41, 630–632.
12] a) C. Adamo, C. Amatore, I. Ciofini, A. Jutand, H. Lakmini, J. Am. Chem.
Soc. 2006, 128, 6829–6836; b) R. N. Dhital, H. Sakurai, Asian J. Org.
[
Manuscript received: July 24, 2015
2
Accepted Article published: September 10, 2015
Final Article published: September 30, 2015
[
[
Chem. Asian J. 2015, 10, 2669 – 2676
www.chemasianj.org
2676
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim