A theoretical DFT-based and experimental study of the transmetalation step in Au/Pd-mediated cross-coupling reactions

Article abstract of DOI:10.1002/chem.201301840

In this work a combined theoretical and experimental investigation of the cross-coupling reaction involving two metallic reaction centers, namely gold and palladium, is described. One metal center (Au) hereby is rather inert towards change in its oxidation state, whereas Pd undergoes oxidative insertion and reductive elimination steps. Detailed mechanistic and energetic studies of each individual step, with the focus on the key transmetalation step are presented and compared for different substrates and ligands on the catalytic Pd center. Different aryl halides (Cl, Br, I) and aryl triflates were investigated. Hereby the nature of the counteranion X turned out to be crucial. In the case of X=Cl and L=PMe₃ the oxidative addition is rate-determining, whereas in the case of X=I the transmetalation step becomes rate-determining in the Au/Pd-cross-coupling mechanism. A variety of Au-Pd transmetalation reaction scenarios are discussed in detail, favoring a transition state with short intermetallic Au-Pd contacts. Furthermore, without a halide counteranion the transmetalation from gold(I) to palladium(II) is highly endothermic, which confirms our experimental findings that the coupling does not occur with aryl triflates and similar weakly coordinating counteranions - a conclusion that is essential in designing new Au-Pd catalytic cycles. In combination with experimental work, this corrects a previous report in the literature claiming a successful coupling potentially catalytic in both metals with weakly coordinating counteranions. A helping hand: For the transmetalation from gold(I) to palladium(II), halide ions play a crucial role (see scheme). Without the formation of a strong gold(I)-halogen bond (e.g., with sulfonate), the transmetalation does not proceed.

Full text of DOI:10.1002/chem.201301840

FULL PAPER
DOI: 10.1002/chem.201301840
A Theoretical DFT-Based and Experimental Study of the Transmetalation
Step in Au/Pd-Mediated Cross-Coupling Reactions
Max M. Hansmann,[a, b] Markus Pernpointner,*[b] Renꢀ Dçpp,[a] and
A. Stephen K. Hashmi*[a, c]
Abstract: In this work a combined the-
oretical and experimental investigation
of the cross-coupling reaction involving
two metallic reaction centers, namely
gold and palladium, is described. One
metal center (Au) hereby is rather
inert towards change in its oxidation
state, whereas Pd undergoes oxidative
insertion and reductive elimination
steps. Detailed mechanistic and ener-
getic studies of each individual step,
with the focus on the key transmetala-
tion step are presented and compared
for different substrates and ligands on
the catalytic Pd center. Different aryl
halides (Cl, Br, I) and aryl triflates
were investigated. Hereby the nature
of the counteranion X turned out to be
crucial. In the case of X=Cl and L=
PMe3 the oxidative addition is rate-de-
termining, whereas in the case of X=I
the transmetalation step becomes rate-
determining in the Au/Pd-cross-cou-
pling mechanism. A variety of Au–Pd
transmetalation reaction scenarios are
discussed in detail, favoring a transition
state with short intermetallic Au–Pd
contacts. Furthermore, without a halide
counteranion the transmetalation from
gold(I) to palladium(II) is highly endo-
thermic, which confirms our experi-
mental findings that the coupling does
not occur with aryl triflates and similar
weakly coordinating counteranions—
a conclusion that is essential in design-
ing new Au–Pd catalytic cycles. In com-
bination with experimental work, this
corrects a previous report in the litera-
ture claiming a successful coupling po-
tentially catalytic in both metals with
weakly coordinating counteranions.
Keywords: density functional calcu-
lations · gold · halides · palladium ·
transmetalation
Introduction
Cross-coupling reactions have become one of the key pro-
[1]
À
cesses for C C bond-formation in organic chemistry.
Whereas one transition metal, for example, palladium, read-
ily undergoes oxidative addition and reductive elimination
processes, the organic coupling partner is introduced by
transmetalation from a redox stable main-group or
ACHTUNGTNERgNUNG ano-
ACHTUNGTRENNUNGmetallic reagent most commonly derived from B, Sn, Li,
Mg, Si, and so on (Figure 1). In the case of a transmetalation
reagent that is based on a d-block metal, such as seen in the
Sonogashira reaction (Cu), it is possible to classify these
transformations as bimetallic catalytic systems.[2] These cou-
[a] M. M. Hansmann, Dr. R. Dçpp, Prof. Dr. A. S. K. Hashmi
Organisch-Chemisches Institut
Ruprecht-Karls-Universitꢀt Heidelberg, Im Neuenheimer Feld 270
69120 Heidelberg (Germany)
E-mail: hashmi@hashmi.de
[b] M. M. Hansmann, Dr. M. Pernpointner
Institut fꢁr Physikalische Chemie, Theoretische Chemie
Ruprecht-Karls-Universitꢀt Heidelberg, Im Neuenheimer Feld 229
69120 Heidelberg (Germany)
Figure 1. Key steps involved in the general palladium-catalyzed cross-
coupling catalysis cycle.
E-mail: Markus.Pernpointner@pci.uni-heidelberg.de
[c] Prof. Dr. A. S. K. Hashmi
pling reactions combine the advantages of two metal centers
in a catalytic cycle, whereas at the same time they dramati-
cally increase the complexity of the system; both metals
having their own coordination sphere, ligand sets, and
Chemistry Department, Faculty of Science
King Abdulaziz University, Jeddah 21589 (Saudi Arabia)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201301840.
Chem. Eur. J. 2013, 00, 0 – 0
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
1
&
ÞÞ
These are not the final page numbers!

redox-potentials. The various coupling mechanisms known
in the literature are mainly discerned by the type of metals
carrying the species to be coupled (Figure 1).
a counteranion that is weakly coordinating for the Au catal-
ysis and, at the same time, is efficient enough in generating
a driving force in the transmetalation step, an issue that is
a major focus of this article. The counteranion problem was
nicely solved for a very special system by Blum et al.[5a] uti-
lizing a palladium-catalyzed cleavage of an allylic ester by
formation of the palladium-p-allyl species circumventing the
use of strongly coordinating halogen atoms, even if in this
specific case it is questionable if palladium on its own can
catalyze this transformation.[3b,15]
We were highly interested in understanding the mecha-
nism and the thermodynamics of the gold–palladium trans-
metalation step from a theoretical and experimental point
of view with a focus on the counteranion involved in this
process to develop new dual-metal-catalyzed reactions. Very
recently, Espinet et al.[16] analyzed the transmetalation step
and prompted us to publish our results on this dual-catalytic
system. The crucial effect of the counteranion in the trans-
metalation process has not been studied so far but seems in
our opinion essential to understand and develop dual-cata-
lyzed reactions.
In the recent past the role of gold as a cross-coupling cat-
alyst has been discussed, but seemingly homogenous gold
catalysts are not able to efficiently undergo oxidative addi-
1
[3]
tion with R X. This creates an “orthogonality”[3b] in catal-
ysis, which might lead to synthetically interesting selectivi-
ties, the assumed disadvantage might be fruitfully exploited
by a combination of two metal-catalyzed cycles exhibiting
orthogonal reactivity as shown in the seminal work by
Hashmi et al.[4] and Blum et al.[5] Although the transmetala-
tion of organogold-compounds to several redox-active tran-
sition metals such as Pd,[3b,4–6] Rh,[7] Ni,[8] Zr,[9] Ru, and Fe[10]
has been studied by employing a stoichiometric amount of
organogold species,[11] it is a challenging task to develop
a double catalytic cycle in which the transmetalation reagent
À
À
À
(R Au L) is generated in situ in a separate catalysis cycle
(Figure 2).
As the gold-catalysis cycle in general is dependent on
+
À
a highly cationic gold species (L Au ; usually generated by
À
À
silver-abstraction of the corresponding L Au Cl com-
As the transmetalation from the more noble (more elec-
tronegative) gold to the less noble (less electronegative) pal-
ladium seems thermodynamically counterintuitive, a detailed
theoretical investigation supporting the experimental find-
ings was of prominent interest. Here we report a computa-
tional study of the whole catalytic cycle from Figure 1 with
two different ligand systems (monodentate and bidentate)
pound)[12] and the transmetalation is driven by the formation
À
of the Au X bond, a seemingly conflictive scenario is gener-
ated. One solution to this “counteranion problem” is the use
of terminal alkynes, the corresponding Au-acetylides can be
À
À
generated from the L Au Cl complexes without activation
by silver additives, leading to an Au-catalyzed Sonogashira-
type reaction.[3a,13] Besides the use of terminal alkynes, the
concept of dual-catalytic Pd/Au cycles is very scarce.[5a,14]
One major drawback recently pointed out by Gagnꢃ and
Weber is the fact that Pd/Au redox-incompatibilities can be
crucial.[6e] In our opinion, the lack of further reports on Pd/
Au dual catalysis is additionally affected by the need for
on palladium, namely [Pd
ACHTUNGTRENNUNG
phosphino)ethane) and [PdAHCTNUGTRENNNUG
1
2
À
ments R =phenyl and R =vinyl in the form of Me3PAu
vinyl. Whereas vinyl gold intermediates involved in the AuI
catalytic cycle are experimentally[17] and theoretical-
ly[18].well-studied, relatively little is known about their reac-
Figure 2. Dual-metal catalysis: combining palladium and gold catalysis.
&
2
&
www.chemeurj.org
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Au/Pd-Mediated Cross-Coupling Reactions
FULL PAPER
tivity as transmetalation agents.[6] Two major aspects deserve
special attention: First, it was experimentally observed that
Table 1. Total electronic energies of the global reaction (Figure 4) for
different methods.
X
EDFT/B3LYP
EMP2/RECP
À
different halogen atoms X in the Ph X species lead to very
[kcalmolÀ1
]
[kcalmolÀ1
]
different yields in the overall reaction.[4,6c] All steps were
therefore calculated for X=Cl (lowest reactivity) and X=I
(highest reactivity) starting from the oxidative addition step.
Secondly, a non-chelating (PMe3) and a chelating (dmpe)
ligand on the Pd substrate were considered to clarify the
mechanistic details in the crucial transmetalation step
(Figure 3).
Cl (a)
Br (b)
I (c)
OTf (d)
OTs (e)
À43.728
À48.100
À52.723
À38.245
À30.541
À38.906
À43.933
À47.585
–
–
ing solvent effects for these neutral species.[26] For all ligands
and methods the cross-coupling reaction between an orga-
nogold(I) species and an aryl halide or aryl triflate is ther-
modynamically clearly exothermic and exergonic (with only
a small entropy contribution), whereas the reactivity within
the halides increases from the chloride to the iodide. This is
À
in accord with experimentally determined Ar X bond disso-
ciation energies, which amount to 96 (X=Cl), 81 (X=Br),
and 65 kcalmolÀ1 (X=I).[27] The same trend is consistent
Figure 3. The key step: transmetalation from gold to palladium.
À
with the bond-dissociation energies for the Au X fragment
obtained in high-level CCSD(T) calculations and FT-ICR-
MS experiments corroborating the DFT trends.[28] The ex-
periments show that aryl iodides lead to high yields in corre-
sponding cross-coupling reactions, whereas only few conver-
sions are known for aryl chlorides.[4,6c] A possible reason for
the significant decrease of the overall thermodynamics for
Computational Details
All calculations were performed by employing the Gaussian 03 program
package.[19] The theoretical approach was the framework of density func-
tional theory (DFT)[20] in combination with relativistic pseudo-potentials
for both Au[21] and Pd[22] and the hybrid B3LYP functional.[23] The
pseudo-potential basis sets and the all-electron basis sets for the light ele-
ments were of cc-pVDZ quality for the combined two-center systems and
of cc-pVTZ quality for the smaller subunits. Due to the size of the trans-
metalation complex, augmented or triple zeta bases could not be em-
ployed in this case. All transition states were uniquely characterized by
occurrence of one imaginary frequency and sought by the quadratic syn-
chronous transit (QST3) algorithm.[24] As soon as the starting material,
transition state, and product structures were available, the resulting reac-
tion profile was verified by following the intrinsic reaction coordinate
(IRC) as described in the literature.[25] Free reaction enthalpies were cal-
culated for standard conditions (1 atm, 298 K) and are not corrected.
the TfOÀ and TsOÀ counteranions is the much weaker Au
À
À
À
OTf and Au OTs bond (compared with the Au I bond).
On average the MP2 values are higher by about 5 kcal
molÀ1 but the relative trend in the overall thermodynamics
is comparable for both methods. The thermodynamics of the
isolated transmetalation step (Figure 5) are shown in
Table 2.
Hereby the substrate and product structures with varying
counteranion X were energetically optimized at the DFT/
Computational Results
Global thermodynamic analysis of the cross-coupling reac-
tion and the transmetalation step: As shown in Figure 4 and
Table 1, initially for the overall cross-coupling reaction the
differences in the electronic energy of the reactants and
products were calculated at the DFT/B3LYP and MP2/
RECP level of theory for X=Cl, Br, I, trifluoromethanesul-
fonate (OTf), and 4-toluenesulfonyl (OTs; 5a–e) disregard-
Figure 5. Thermodynamics of the transmetalation reaction.
Table 2. Thermodynamics of the transmetalation step (Figure 5) for vari-
ous basis sets, methods, and tightness in the optimization.
X
E
E
E
E
[a]
[b]
[c]
[d]
[kcalmolÀ1
]
[kcalmolÀ1
]
[kcalmolÀ1
]
[kcalmolÀ1
]
Cl (a)
Br (b)
I (c)
OTf (d) 14.174
OTs (e) 13.991
2.316
2.129
1.857
2.291
2.664
1.856
–
2.172
2.093
1.738
–
0.414
1.971
4.246
–
–
–
–
[a] DFT/B3LYP, cc-pVDZ, normal convergence. [b] DFT/B3LYP, cc-
pVDZ, tight convergence. [c] DFT/B3LYP, cc-pVTZ, normal conver-
gence. [d] MP2/RECP, cc-pVDZ, normal convergence.
Figure 4. Global reaction for the palladium-catalyzed cross-coupling of
organogold species.
Chem. Eur. J. 2013, 00, 0 – 0
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
3
&
ÞÞ
These are not the final page numbers!

M. Pernpointner, A. S. K. Hashmi et al.
B3LYP and MP2/RECP level of theory together with the
cc-pVDZ, cc-pVTZ, and LANL2DZ basis sets. The tight-
ness of structure optimization was analyzed but proved to
have only little influence.
Au to Pd via a 4-centered transition-state 9a,c followed by
cis–trans-isomerization leading to the same transmetalation
products (Figure 6). Both key transition states have in
common a trigonal-pyramidal coordination sphere at the
AuI center, whereas the transition state 9a,c would be
pseudo-trigonal-pyramidal and 8a,c square-planar in the co-
ordination sphere of PdII.
Several attempts to locate a transition state 9a,c accord-
ing to mechanism II with various different methods and
basis-sets failed. Based on our results, species 9a,c is a transi-
The isolated transmetalation step is slightly endergonic at
all levels of theory with comparable results for the individu-
al halides (X=Cl, Br, I) at the DFT level (Table 2). Increas-
ing the tightness of structure optimization, namely the
threshold of the energy and norm of gradient vector, had
only little influence on the thermodynamics but uses consid-
erably more computational resources. The transition from
a VDZ to a VTZ basis did not influence the DFT results sig-
nificantly, justifying the use of the double zeta basis for all
further calculations, which is important because a triple zeta
basis cannot be employed for the large two-center com-
plexes due to numerical bottlenecks.[29] The oxygen-bridged
ligands are significantly raised in energy (X=OTf, OTs),
which indicates a smaller tendency for the gold–oxygen
bond formation as observed for the overall thermodynamics.
No accurate experimental bond dissociation energies for the
ent species not exhibiting the required properties at
able to a transition state. Interestingly, we were able to
locate a transition state according to 8a,c, which connects
ACHTUNGTRENNUNGtribACHTUNGTRENNUNGut-
AHCTUNGTRENNUNG
À
À
the Au C bond-breaking and Au X bond-formation event
in the coordination sphere of the palladium complex
(Figure 7).
À
Remarkably, in these transition states the Au Pd distance
(2.722 for X=Cl and 2.804 ꢄ for X=I) is much shorter
than the van der Waals radii of gold and palladium com-
bined. This indicates a significant metal–metal interaction
lowering the activation energy for the R-group transfer.
This is in good agreement with the very recent publications
of Espinet et al. reporting on a structurally equal transition
À
PMe3Au X systems are available, which would be of great
value to gain further insight into the transmetalation
step.[28,30] To extend the systematic basis set test, we also
tested the smaller LANL2DZ Pd basis set and an aug-cc-
pVDZ Pd basis. Both variations led to very similar results in
the transmetalation step, from which we deduced that an
augmentation of the double zeta basis is not relevant and
would lead to serious computational bottlenecks.
À
state with a nearly identical Pd Au contact of 2.8 ꢄ in the
case of the corresponding chloro-arsan complex.[16] Further-
more, in this report we analyze the influence of the counter-
anion and give a complete theoretical analysis of the catalyt-
ic cycle, necessary to understand the rate-determining step
in the total transformation.
Analysis of the Au/Pd transmetalation mechanism employ-
ing monodentate phosphine ligands: We initially started our
analysis of the transmetalation mechanism with the trans-
The complete reaction path for the transmetalation
through the trans complex is shown for both the chlorine
and iodine counteranions in Figure 8. As expected, the trans
complexes trans-1a,c are, by about 11 kcalmolÀ1, more
stable than the corresponding cis compounds (cis-1a,c),
which are generated initially in the oxidative addition event
and proved to be a sink on the energy potential surface.[32]
Substitution of the PMe3 ligand (associative or dissociative)
followed by coordination leads to the Pd-h2-vinyl complexes
(7a,c), which requires approximately 12 kcalmolÀ1 in the ab-
sence of substantial entropic effects. Based on the work by
Farina et al.[33a] and Amatore and Jutand et al.[33b] on the
[PdXPhACHTUNGTRENNUNG(PMe3)2] complex (trans-1a,c; X=Cl (a) or I (c)),
formed after cis–trans-isomerization of the corresponding cis
complex (cis-1a,c) after oxidative addition, and analyzed the
two main transmetalation mechanisms:[31] Mechanism I) a
dissociative or associative ligand substitution of the phos-
phine ligand with the vinyl gold compound leading to the in-
termediate 7a,c followed via a 4-centered intramolecular
transition state 8a,c to the final transmetalation products, or
mechanism II) a direct intermolecular vinyl-transfer from
Figure 6. Model for a potential transmetalation mechanism with monodentate phosphine ligands.
&
4
&
www.chemeurj.org
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Au/Pd-Mediated Cross-Coupling Reactions
FULL PAPER
(10a,c) different scenarios are imaginable: The direct disso-
À
À
ciation of 10a,c into the X Au PMe3 molecule (4a,c) and
the T-shaped structure (compound 11) seems electronically
non-favored, but reasonable based on the large entropy con-
tribution. Alternatively, an associative pathway with a sol-
vent molecule (e.g., MeCN) or a phosphane ligand also
seems reasonable to generate the final transmetalation com-
À
plex 3 (Figure 8). It is noteworthy to highlight that the Au
Pd contacts, not only in the transmetalation transition states
(8a,c), but also in the intermediates before and after the
transmetalation step (7a,c and 10a,c) are very short (see the
Supporting Information) clearly demonstrating the impor-
tance of metal–metal interactions.
The overall thermodynamic data for the transmetalation
step compared to the cis-complex [PdXPhL2] (cis-1a,c) is
DGtrans = +1.4 kcalmolÀ1 (X=Cl) and DGtrans =À0.9 kcal
molÀ1 (X=I), respectively, is in accord with the thermody-
namic results for the transmetalation reaction employing the
chelate ligand on Pd (Table 2). Interestingly, the transmeta-
lation thermodynamics are highly endergonic DGtrans = +
13.5 kcalmolÀ1 (X=Cl); DGtrans = +10.2 kcalmolÀ1 (X=I) if
compared with the trans-[PdXPhL2] complex (trans-1a,c),
which is in good agreement with the experimental and theo-
retical results by Espinet et al.[16a] studying the favored (ex-
ergonic) transmetalation from Pd to Au starting with
[PdR2L2]. The here studied corresponding reverse and ther-
modynamically upwards process can only be realized by
coupling this process to a reductive elimination event push-
ing the equilibrium towards the desired cross-coupling prod-
ucts (see section below).
Analysis of the Au/Pd transmetalation mechanism employ-
ing a chelate ligand: In the second section of this discussion
(see above) we could demonstrate a reasonable transmetala-
tion mechanism based on the exchange of a monodentate
phosphine ligand with the vinyl gold fragment followed by
an intramolecular R-group transfer via a transition state
Figure 7. Transition state structures (8a,c) for the transmetalation step
with X=Cl and X=I.
À
showing close Pd Au contacts. It is interesting to point out
Stille reaction, in which a vinyl-transfer reagent is first
bound in the coordination sphere of palladium prior to the
transmetalation step as well, it seems very likely that the
formation of the corresponding mixed Pd-vinyl-Au species
7a,c is generated through a dissociative mechanism from
trans-1a,c.
that this mechanism is closely related to the proposed mech-
anism of the deeply studied Stille coupling.[33] Interestingly,
these transmetalation studies always employed monodentate
ligands on the palladium center, easily opening a free coor-
dination site by dissociation of a ligand on palladium. Our
experiments, however, demonstrated that chelating ligands
such as 1,1’-bis(diphenylphosphino)ferrocene (dppf) were
very efficient ligands for the Pd-catalyzed cross-coupling of
organogold species,[4,6c] raising the question how this trans-
metalation process can proceed if the ligand seemingly
blocks the free coordination-site on the metal center that is
essential for organogold pre-coordination prior to transme-
talation. In Figure 9 our initial ideas are summarized on
how a bidentate ligand can influence the change in the reac-
tion pathway for the transmetalation step.
The theoretical results exhibit only a small difference on
the activation energies for the transmetalation for the differ-
ent counteranions (DG°(I) =26.0 kcalmolÀ1
;
DG°
=
(Cl)
23.4 kcalmolÀ1) and in the energies of the product com-
plexes 10a,c, which is seemingly in contrast to the experi-
mental findings that only for strongly activating ligands will
chlorine compounds undergo the cross-coupling reaction.[4,6c]
This can be explained by a change in the rate-determining
step from transmetalation to oxidative addition depending
on the counteranion, which will be approved in the section
below and in the Experimental Section.
Pathway I is in analogy to Figure 6, a concerted mecha-
nism passing through a four-membered transition state with
local trigonal-bipyramidal coordination at palladium (9cha–
e). Analogous to the case of the monodentate phosphine
To generate the cis-[PdR2L2] precursor 3 from the trans-
metalation products still bound in the coordination sphere
Chem. Eur. J. 2013, 00, 0 – 0
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
5
&
ÞÞ
These are not the final page numbers!

M. Pernpointner, A. S. K. Hashmi et al.
Figure 8. Reaction profile for the transmetalation pathway with monodentate ligands.
ligand system we were unable, even with very high efforts,
To classify the feasibility of transmetalation pathway II
À
to locate such a transition state, which is in accord with the
search for a similar transition state in the Stille cross-cou-
pling.[33c] From a computational point of view it may be pos-
sible that an even larger basis together with a more sophisti-
cated method such as coupled cluster theory would lead to
a localizable state, but as far as our tests revealed, such
a transition state does not seem to be accessible. As a conse-
quence, we suggest a multistep mechanism according to the
dissociative pathways II or III: In II the Pd/X bond is heter-
olytically broken (13ch) and the vacant coordination site oc-
cupied by the vinyl carrier (14ch), whereas in III a chelate
ring opening is discussed.
we calculated the bond-dissociation energy for the Pd X
bond (13ch) in Figure 9 and obtained 128.98 (X=Cl), 123.66
(X=Br), and 113.79 kcalmolÀ1 (X=I), which indicates an
energy barrier that is far too high for this pathway to occur
in the gas phase under standard conditions. Certainly, sol-
vents can considerably decrease this barrier by interaction
with the resulting ions. For acetonitrile we simulated this
solvation effect by adding one molecule of acetonitrile to
the complex 13ch and obtained a noticeable energetic stabili-
zation of 29.2 kcalmolÀ1.
In contrast to the theoretical prediction, a considerable in-
fluence of the solvent could not be observed experimental-
ly,[4] and therefore does not support an ionic dissociative
&
6
&
www.chemeurj.org
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Au/Pd-Mediated Cross-Coupling Reactions
FULL PAPER
number. A fivefold coordina-
tion with the substrate occupy-
ing the apex of a quadratic pyr-
amid immediately led to a re-
verse process forming the
educts as we showed computa-
tionally. The employment of
a larger basis set did not change
these findings, having their
origin in a high-lying pz LUMO
on Pd making the metal center
unfit for a fast attack by the
substrate. In Figure 10 the com-
plete reaction path for the
transmetalation under a chelate
ring opening is presented for all
three halides (X=Cl, Br, and
I).
In a first step, the vinyl gold
species (2) adds to the vacant
coordination site on Pd being
available after the opening of
the chelate ring (15cha–c). The
required energy for this step
amounts to 7–9 kcalmolÀ1 (elec-
tronic) and 19–21 kcalmolÀ1
(free enthalpy; because this
process is entropically dis-
AHCTUNGTERGaNNUN dvantageous), being quite in-
Figure 9. Proposed mechanisms for the transmetalation step utilizing a chelating ligand.
dependent of the X substituent
employed. These energy differ-
mechanism. Mechanisms of this type are only feasible if the
counteranions are very weakly bound to the metal center,
which is the case for, for example, triflate anions, but not for
halides (e.g., see the discussion of the cationic Heck reac-
tion[34]). For the current system, a triflate ligand (X=OTf)
does not undergo any transmetalation reaction with Au and
Pd as metal centers (see the section below). As a conse-
quence, a purely dissociative mechanism including a cleavage
ences lie within an accessible range for a thermal reaction
and would allow for a ring opening. In the resulting inter-
ch
À
mediates 15 a–c (Figure 10), the Au Pd distance amounts
to about 3.32 ꢄ and is an indication for a possible interac-
tion between the Au and Pd metal centers as seen with the
monodentate phosphine ligands as well. Via a cyclic transi-
tion state, analogous to the ones in Figure 7, a concerted
À
À
Au C bond-breaking and an Au X bond-forming takes
À
À
À
of the Pd X bond is very unlikely. A variation of this path-
place, followed by liberation of the X Au PMe3 species
from the complexes 15cha–c, yielding the T-shaped inter-
mediate 17. This amount of energy required (ꢀ30 kcal
molÀ1) is nearly the same for all halides and renders this
step quite unfavorable, but still in an acceptable range, as T-
shaped structures of Pd complexes are known characterized
intermediates.[35] An alternative to the T-shaped intermedi-
ate formation 17 is a coordination of the halide to the two
metal centers as shown in structure 10chc, which could be
computationally characterized as an intermediate. By a possi-
way emerges if one postulates an associative replacement of
the XÀ anion by the incoming vinyl ligand. The energy re-
quired for this variant can be obtained from the components
14ch and 2 (Figure 9) and amounts to 97.54 (X=Cl), 92.21
(X=Br), and 82.34 kcalmolÀ1 (X=I), which still represents
a very unfavorable process. From these findings we deduce
À
that a purely dissociative way under cleavage of the Pd X
bond is energetically too costly and may be intrinsically cou-
À
pled to the formation of the Au X bond.
As we successfully found a transmetalation pathway for
ble assistance of the solvent molecules or phosphine ligand,
ch
À
À
À
monodentate ligands and we ruled out the Pd X bond
the X Au PMe3 species is expelled from the complex 10 c
and the final product 3 is formed by the favorite ring closure
of the chelate ligand, enabling a lower-energy pathway. It
should be pointed out that the chelate ring opening leads to
a dramatic increase of conformational isomers making the
computational localization of transition states a demanding
task. As we could already demonstrate a successful transme-
breakage, we envisioned the unusual possibility of a chelate
opening to coordinate the organogold species according to
the monodentate mechanism (Figure 9, pathway III).
The creation of a free coordination site on Pd is proposed
due to the fact that a square-planar configuration is very
much preferred by PdII over any higher coordination
Chem. Eur. J. 2013, 00, 0 – 0
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
7
&
ÞÞ
These are not the final page numbers!

M. Pernpointner, A. S. K. Hashmi et al.
turing reagent did not influence
the transmetalation yields in
any way.[4,6c] Despite the fact
that Au does not exhibit
a strong tendency to change its
oxidation state during a reac-
tion, a quadratic planar com-
plex with an energetic mini-
mum
could
be
localized
(Figure 11). This quadratic co-
ordination is typical for AuIII
complexes indicating a higher
oxidation state, even if the oxi-
dation states are difficult to
assign in such bimetallic com-
À
pounds. However, the Au P
À
and Au Cl bond lengths are
quite elongated (2.601 and
2.455 ꢄ) and allow for the in-
terpretation of an oxidative in-
sertion of the vinyl gold species
À
into the Pd Cl bond. A corre-
sponding transition state incor-
porating structure 19 was not
found in the calculated pathway
and the relevance of structure
19 with a significant Au/Pd in-
teraction could therefore not be
perfectly established.
Finally,
a one-step process
(Figure 12) through a [3+2] s-
Figure 10. Calculated reaction profile for a transmetalation mechanism in the coordination sphere of Pd
through a chelate opening.
bond metathesis was sought,
skipping the explicit transmet
ACHTUNGTRENNUNGal-
ACHTUNGTRENNaNUG AHCTUNGTRENNGtUN ion step. Despite the fact that
talation pathway based on the easier to analyze monoden-
tate phosphine model, and here could demonstrate that
a chelate opening is energetically feasible (but entropically
disadvantageous), we see no need for a further detailed
analysis.
Further models for other potential transmetalation mecha-
nisms: Beside our transmetalation models outlined above,
we briefly mention some viable alternatives that were tested
by us numerically. At first we investigated a nucleophilic
attack of an iodine ion to the gold vinyl species producing
an “aurate” compound that might transfer the organic
group more easily because of the higher electron density.
This is in analogy to the initial assumption of an “ate”-com-
Figure 11. Structure of 19: A local-minimum on the potential energy sur-
face.
plex in the Suzuki reaction[36a] (even though it was later
shown that not the “ate”-complex but a Pd-hydroxide spe-
cies was the relevant transmetalation intermediate).[36b]
A
corresponding local minimum for an anionic trigonal-coordi-
nated gold species could, however, not be detected compu-
tationally attributing only little probability to this variation.
The same holds for an assumed radical mechanism due to
the experimental finding that the addition of a radical-cap-
a reaction pathway starting from a distorted initial geometry
derived from Figure 12 led to the desired products, an ap-
propriate transition state could not be found and is necessa-
ry for a sound argumentation. This is in accordance with the
experimental result that adding free phosphane ligACHTUNGTRENNUNGands in-
&
8
&
www.chemeurj.org
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Au/Pd-Mediated Cross-Coupling Reactions
FULL PAPER
depending on the counteranion employed. Whereas in case
of X=Cl, the oxidative addition is non-favored with a high
activation barrier (DG°(Cl) =32.3 kcalmolÀ1) the correspond-
À
ing oxidative addition into the C I bond proceeds, as ex-
pected, with a lower activation energy (DG°(I) =22.5 kcal
molÀ1). As the activation energy for the transmetalation pro-
cess only differs slightly between the different halogen coun-
teranions, the rate-determining step changes from the oxida-
tive addition (X=Cl) to the transmetalation (X=I;
DG°(I) =23.3 kcalmolÀ1), respectively.
Figure 12. A potential one-step concerted R-group transfer.
hibits the reaction, which should not occur in a fully con-
certed mechanism and further supports our transmetalation
mechanism outlined above in which phosphine pre-dissocia-
tion is necessary.
We analogously calculated the process for the chelate [Pd-
AHCTUNGTREG(NNNU dmpe)] system and summarize here the main differences.
As the bent 14 valence electron (VE) Pd0 species is more
activated towards oxidative addition compared with the
linear Pd0 species, the oxidative addition proceeds even in
the case of X=Cl, with a much lower activation energy
(DG°(Cl) =14.5 kcalmolÀ1).[37] With the bidentate ligand
system employed, both transmetalation events (X=Cl, I)
proceed through relatively high-energy pathways through
opening of the chelate ring and represent the rate-determin-
ing step (Figure 14).
Global energy profile for the cross-coupling process: A sim-
plified complete picture: After analyzing in detail the trans-
metalation step for two different ligand systems, we ana-
lyzed the full catalytic cross-coupling cycle, including oxida-
tive addition and reductive elimination to gain insight into
the rate-determining step of the complete catalysis cycle de-
pending on the electrophile and the ligand system employed
(Figures 13 and 14).
In summary, the transmetalation step from Au to Pd is in
both ligand cases an endergonic process, which is driven by
the highly exergonic reductive elimination event. Whereas
the ligand system on palladium most effectively changes the
In the case of monodentate phosphine ligands (Figure 13;
L=PMe3) we observe a change of the rate-determining step
Figure 13. Energy profile for the complete cross-coupling catalysis cycle with [PdACHTUNRGTNEG(UN Me3P)2].
Chem. Eur. J. 2013, 00, 0 – 0
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
9
&
ÞÞ
These are not the final page numbers!

M. Pernpointner, A. S. K. Hashmi et al.
Figure 14. Energy profile for the full cross-coupling catalysis cycle with [PdACTHNUTRGNEU(GN dmpe)].
activation barrier of the oxidative addition, it almost has no
effect on the reductive elimination.
ate to high yields.[6d,38] The reaction conditions published
were very similar to our conditions with one major differ-
ence: the organogold compounds had not been isolated in
a preceding synthetic step but prepared in situ from organo-
Experimental Results
lithium compounds and [AuClACTHUNGRETNNU(G Ph3P)]. Thus, one equivalent
of LiCl is present in the cross-coupling reactions from
ref. [6d]. Based on these insights, we repeated our experi-
ments adding stoichiometric amounts of various salts. Small
amounts of cross-coupling products were formed in the ab-
sence of any halide (Table 3, entries 7 and 8). In contrast,
the presence of halides in the reaction mixture immediately
increased the yields to 77–99% (Table 3, entries 9–12). The
type of halide source did not have a strong influence on the
yields. This fully confirms our computational results for the
transmetalation step, demonstrating 1) no significant differ-
ence between the halide counterions in the transmetalation
step and 2) a significant difference between the halides and
the oxygen-based counteranions. As shown in Table 2, the
thermodynamics of the transmetalation reaction are only
slightly endergonic in the case of X=Cl, Br, I (ꢀ1-4 kcal
molÀ1) but strongly endergonic for X=OTf (ꢀ14 kcal
molÀ1). The oxidative addition of the Pd0 species into
ArOTf leads to a cationic [PdAr(S)L2]+ complex, which can
The role of the counteranion: Based on our protocol for the
palladium-catalyzed cross-coupling reactions of aryl-,
heteroACHTUNGTRENNUNGaryl-, vinyl-, alkynyl-, benzyl-, or acyl halides with sto-
ichiometric amounts of an organogold compound, we
wanted to experimentally verify the computational results
on the counteranions in our outlined transmetalation mech-
anism and reinvest the role of triflates as electrophiles. A
series of experiments with phenyltriflate and 4-acetylphenyl-
triflate were performed (Table 3). Unfortunately, all at-
tempts to use aryl triflates in our cross-coupling protocol
failed (Table 3, entries 1–6).
Even with the Pd/XPhos system (XPhos=2-dicyclohexyl-
phosphino-2’,4’,6’-triisopropylbiphenyl), which is
a very
active catalyst in Suzuki–Miyaura coupling of aryl triflates,
no cross-coupling product was detected (Table 3, entry 2).
Methyl triflate (Table 3, entry 6) and phenyl triflimide also
gave no cross-coupling products. To our surprise, a palladi-
um-catalyzed cross-coupling reaction with aryl triflates was
published a few months after our own unsuccessful at-
tempts.[6d] Aryl triflates were coupled with various organo-
gold compounds and the products were obtained in moder-
À
only form the Au OTf bond in the transmetalation event;
an energetically non-favored process outlined in Table 2. In
À
the presence of a halide source a Pd X triflate–halide ex-
change leads to the neutral [PdCl(Ar)L2] complex,[39] which
&
10
&
www.chemeurj.org
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Au/Pd-Mediated Cross-Coupling Reactions
FULL PAPER
Table 3. Palladium-catalyzed cross-coupling of organogold compounds with aryl triflates.
Entry
R1
([equiv])
R2
([equiv])
Catalyst
([mol%])
Additive
Solvent
T
t
Yield
[%]
G
R
G
[8C]
[h]
1
2
Ph (1.5)
Ph (1.5)
N
E
T
–
–
MeCN
MeCN
RT
RT
24
4
0
0
3
4
4-AcPh (1.5)
4-AcPh (1.5)
G
[PdCl
E
–
–
MeCN
THF
60
60
11
10
0
0
5
6
4-AcPh (1.5)
Me
N
G
N
–
–
THF
RT
RT
4
0
0
MeCN
16
7
8
4-AcPh (1.0)
4-AcPh (1.0)
4-AcPh (1.0)
4-AcPh (1.0)
4-AcPh (1.0)
4-AcPh (1.0)
4-AcPh (1.0)
G
[PdCl
(PPh3)2] (5.0)
–
THF
THF
THF
THF
THF
THF
THF
RT
RT
RT
RT
RT
RT
RT
18
18
18
18
18
18
18
6
9
LiOTf
LiCl
9
95
90
99
97
77
10
11
12
13
LiBr
Bu4NCl
Bu4NBr
Bu4NI
allows a transmetalation mechanism involving a halide
double catalytic Pd/Au catalysis, as a non-activated non-cat-
À À
À
bridged Pd X Au transition state (Figure 7) that fully ex-
ionic Au X species would not undergo classical gold-cata-
lyzed transformations.
plains these experimental findings. At this stage it cannot be
ruled out that for the oxidative addition a halide effect re-
sults in an acceleration through the formation of anionic
palladate complexes.[40] But the complete experimental fail-
ure of transmetalation with the triflates can only be ex-
plained by the transmetalation not taking place, as the oxi-
dative addition of aryl triflates especially in the presence of
Pd/XPhos is known to be possible even without halides. It is
interesting to note that in the case of the Stille reaction, the
addition of LiCl improves the rate for the oxidative addition
but has a retarding effect on the transmetalation step,[33g,39]
an effect that is in stark contrast to the corresponding orga-
nogold transmetalation in which the driving force is generat-
In ref. [6d] alkyl gold species were also used in cross-cou-
pling reactions. Under our typical conditions, [Au-
AHCTUNGTRENNG(UN CH3)Ph3P] was not effective in the cross-coupling reaction.
This may be due to its insolubility in MeCN. However, in
THF the cross-coupling was successful. At room tempera-
ture after 18 h only 27% yield of the product was obtained
(Table 4, entry 1). A longer reaction time of three days gave
only a minor improvement to 29% (entry 2). Increasing the
temperature to 608C gave 65% after 18 h (Table 4, entry 3)
and 83% yield after three days (entry 4). Overall, compared
with other organogold compounds (vinyl-, aryl-, and also
À
ed through the formation of the Au X bond. Furthermore,
Table 4. Cross-coupling experiments with alkyl gold species.
it is in accord with our calculations that palladium-catalyzed
cross-coupling reactions of organogold-species with aryl bro-
mides and aryl chlorides are mainly effected by the ligand
system on palladium to undergo the oxidative addition pro-
cess (Figures 13 and 14), whereas the transmetalation pro-
ceeds equally well with different halides but not triflates. In
situ-exchange of triflate with a halogen counterion re-estab-
lishes the transmetalation step, but raises problems in the
Entry
T [8C]
t
Yield [%]
1
2
3
4
RT
RT
60
18 h
3 d
18 h
3 d
27
29
65
83
60
Chem. Eur. J. 2013, 00, 0 – 0
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
11
&
ÞÞ
These are not the final page numbers!

M. Pernpointner, A. S. K. Hashmi et al.
were dried by a MB SPS-800 using drying columns. Preparation of air-
and moisture-sensitive materials was carried out in flame dried flasks
under an atmosphere of nitrogen using Schlenk techniques. Thin layer
chromatography (TLC) was performed using Polygram pre-coated plastic
sheets SIL G/UV254 (SiO2, 0.20 mm thickness) from Macherey–Nagel.
Yields have been determined by GC using n-dodecane as internal stan-
dard for quantification.
Hammond-type vinyl gold species), the alkyl species seemed
to be less reactive. These experiments are also consistent
with the calculated mechanism. Chen et al. could isolate the
II
I
II
I
À
À
structurally similar Pt Cu and Pt Au complexes with
a methyl group coordinated in the coordination sphere of
PtII.[41] This indicates that a mechanism based on pre-coordi-
nation even in the case of alkyl gold species is reasonable,
but less favored, which is nicely in accord to our theoretical
findings.
General procedure for the cross-coupling of organogold-compounds with
aryl triflates: Under a nitrogen atmosphere, the organogold compound
(1.0 equiv), aryl triflate (1.5 equiv), palladium catalyst, and (in some
cases) an additive (1.2 equiv) were dissolved in dry solvents and stirred
for the specified time. n-Dodecane (1.0 equiv) was added and the reac-
tion mixtures were analyzed by GC or GC-MS (Quantification by GC).
General procedure for the cross-coupling with methyl(triphenylphosphi-
Conclusion
ne)gold(I): [Au
ACHTUNGTRENNUNG
(22.9 mg, 100 mmol) and [PdCl CTHGNUTRENNUNG
À
In this work a new palladium-catalyzed C C-coupling reac-
solved in dry THF under nitrogen atmosphere. n-Dodecane (23.0 mL,
100 mmol) was added and the reaction mixture was stirred for three days
and analyzed by GC-MS.
tion with an organogold-compound as the transmetalation
reagent was theoretically investigated at the DFT/B3LYP
and partially at the MP2 level of theory. Several potential
transmetalation scenarios were analyzed with a monoden-
tate- and bidentate-ligand system and were fit into the con-
text of the full cross-coupling energy profile. We could
locate a transmetalation mechanism based on pre-coordina-
tion of the organogold-species followed by transmetalation
in the coordination sphere of palladium via a transition state
Acknowledgements
The authors thank Umicore AG & Co. KG for the generous donation of
gold salts. M.M.H. is grateful to the Fonds der Chemischen Industrie for
a Chemiefonds scholarship and the Studienstiftung des deutschen Volkes.
À
exhibiting very short intermetallic Au Pd contacts. This
transmetalation mechanism is in very good agreement with
the theoretical studies reported on the Stille reaction[33] and
very recent reports by Espinet et al.[16] Furthermore, by ana-
lyzing the full energy profile we could locate a change in the
rate-determining step depending on the counteranion and
the ligand system involved; in the case of X=Cl the oxida-
tive addition is the rate-determining step, whereas in the
case of X=I it is the transmetalation (with L=PMe3). Em-
ploying a chelating ligand system on palladium significantly
lowers the activation barrier for the oxidative-insertion pro-
cess and determines the transmetalation to be rate-deter-
[1] Metal-Catalyzed Cross-Coupling Reactions, 2nd ed. (Eds.: A. de
Meijere, F. Diederich), Wiley-VCH, Weinheim, 2004.
[2] a) M. H. Pꢃrez-Temprano, J. A. Casares, P. Espinet, Chem. Eur. J.
2012, 18, 1864–1884; b) O. F. Wendt, Curr. Org. Chem. 2007, 11,
1417–1433.
[3] a) T. Lauterbach, M. Livendahl, A. Rosellꢅn, P. Espinet, A. M.
Echavarren, Org. Lett. 2010, 12, 3006–3009; b) A. S. K. Hashmi, C.
Lothschꢁtz, R. Dçpp, M. Ackermann, J. D. B. Becker, M. Rudolph,
Christian Scholz, F. Rominger, Adv. Synth. Catal. 2012, 354, 133–
147; in the presence of a strong oxidant or a fluoride source, AuI is
capable of undergoing oxidative cross-coupling reactions without
the involvement of a second metal; for selected examples, see:
c) A. S. K. Hashmi, T. D. Ramamurthi, F. Rominger, J. Organomet.
Chem. 2009, 694, 592–597; d) G. Zhang, Y. Peng, L. Cui, L. Zhang,
Angew. Chem. 2009, 121, 3158–3161; Angew. Chem. Int. Ed. 2009,
48, 3112–3115; e) G. Zhang, L. Cui, Y. Wang, L. Zhang, J. Am.
Chem. Soc. 2010, 132, 1474–1475; f) A. S. K. Hashmi, T. D. Rama-
murthi, M. H. Todd, A. S.-K. Tsang, K. Graf, Aust. J. Chem. 2010,
63, 1619–1626; g) T. Ball, G. C. Lloyd-Jones, C. A. Russell, Science
2012, 337, 1644–1648; h) X. C. Cambeiro, T. C. Boorman, P. Lu, I.
Larrosa, Angew. Chem. 2013, 125, 1825–1828; Angew. Chem. Int.
Ed. 2013, 52, 1825–1828.
[4] A. S. K. Hashmi, C. Lothschꢁtz, R. Dçpp, M. Rudolph, T. D. Rama-
murthi, F. Rominger, Angew. Chem. 2009, 121, 8392–8395; Angew.
Chem. Int. Ed. 2009, 48, 8243–8246.
[5] a) Y. Shi, K. E. Roth, S. D. Ramgren, S. A. Blum, J. Am. Chem. Soc.
2009, 131, 18022–18023; b) Y. Shi, S. D. Ramgren, S. A. Blum, Orga-
nometallics 2009, 28, 1275–1277.
[6] For palladium-catalyzed reactions employing stoichiometric amounts
of organogold species see the two reviews: a) P. Garcia, M. Malac-
ria, C. Aubert, V. Gandon, L. Fensterbank, ChemCatChem 2010, 2,
493–497; b) S. A. Blum, J. J. Hirner, Y. Shi, Acc. Chem. Res. 2011,
44, 603–613, and: c) A. S. K. Hashmi, R. Dçpp, C. Lothschꢁtz, M.
Rudolph, D. Riedel, F. Rominger, Adv. Synth. Catal. 2010, 352,
1307–1314; d) M. PeÇa-Lꢅpez, M. Ayꢆn-Varela, L. A. Sarandeses, J.
Pꢃrez Sestelo, Chem. Eur. J. 2010, 16, 9905–9909; e) D. Weber,
M. R. Gagnꢃ, Chem. Commun. 2011, 47, 5172–5174; f) K. E. Roth,
S. A. Blum, Organometallics 2011, 30, 4811–4813.
À
mining. Whereas the thermodynamics of the Au Pd trans-
metalation step are endergonic, the switch from a halogen
(X=Cl, Br, I) to an oxygen-based counteranion (OTf)
proved to dramatically increase the activation barrier for
a successful transmetalation event. In combination with the
performed experiments we could demonstrate that the
cross-coupling reaction is only accessible with aryl triflates
in the presence of additional halides sources (such as LiCl).
This clearly supports our theoretical work suggesting a tri-
flate–halogen exchange after oxidative addition followed by
our outlined mechanism. This is important information for
the further development of new Pd/Au dual-catalyzed reac-
tions because it highlights the problem of the use of weakly
coordinating counteranions together with the cross-coupling
strategy.
Experimental Section
General methods: All reagents and solvents were obtained from Fisher
Scientific, ABCR, Alfa Aesar, Sigma–Aldrich, or VWR and were used
without further purification unless otherwise noted. Absolute solvents
[7] Y. Shi, S. A. Blum, Organometallics 2011, 30, 1776–1779.
[8] J. J. Hirner, S. A. Blum, Organometallics 2011, 30, 1299–1302.
&
12
&
www.chemeurj.org
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Au/Pd-Mediated Cross-Coupling Reactions
FULL PAPER
[9] P. Cornell, Y. Shi, S. A. Blum, Organometallics 2012, 31, 5990–5993.
[10] L. Molinari, A. S. K. Hashmi, Organometallics 2011, 30, 3457–3460.
man, J. V. Ortiz, Q. Cui, A. G. Baboul, S. Clifford, J. Cioslowski,
B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. L.
Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Na-
nayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen,
M. W. Wong, C. Gonzalez, J. A. Pople, Gaussian, Inc., Wallingford
CT, 2004.
À
[11] For a more complete picture of various Au M transmetalations,
see: a) R. J. Cross, M. F. Davidson, A. J. McLennan, J. Organomet.
Chem. 1984, 265, c37–c39; b) R. J. Cross, M. F. Davidson, J. Chem.
Soc. Dalton Trans. 1986, 411–414; c) M. Contel, M. Stol, M. A.
Casado, G. P. M. van Klink, D. D. Ellis, A. L. Spek, G. van Koten,
Organometallics 2002, 21, 4556–4559; d) M. I. Bruce, M. E. Smith,
N. N. Zaitseva, B. W. Skelton, A. H. White, J. Organomet. Chem.
2003, 670, 170–177; e) A. B. Antonova, M. I. Bruce, B. G. Ellis, M.
Gaudio, P. A. Humphrey, M. Jevric, G. Melino, B. K. Nicholson,
G. J. Perkins, B. W. Skelton, B. Stapleton, A. H. White, N. N. Zaitse-
va, Chem. Commun. 2004, 8, 960–961; f) M. Ferrer, L. Rodrꢇguez,
O. Rossell, J. C. Lima, P. Gꢅmez-Sal, A. Martꢇn, Organometallics
2004, 23, 5096–5099; g) M. I. Bruce, P. A. Humphrey, G. Melino,
B. W. Skelton, A. H. White, N. Zaitseva, Inorg. Chim. Acta 2005,
358, 1453–1468; h) M. Robitzer, I. Bouamaed, C. Sirlin, P. A. Chase,
G. van Koten, M. Pfeffer, Organometallics 2005, 24, 1756–1761;
i) C.-L. Chan, K.-L. Cheung, W. H. Lam, E. Chung-Chin Cheng, N.
Zhu, S. W.-K. Choi, V. W.-W. Yam, Chem. Asian J. 2006, 1, 273–286;
j) M. Stol, D. J. M. Snelders, H. Kooijman, A. L. Spek, G. P. M. van
Klink, G. van Koten, Dalton Trans. 2007, 2589–2593.
[20] a) P. Hohenberg, W. Kohn, Phys. Rev. B 1964, 136, B864–B871;
b) W. Kohn, L. J. Sham, Phys. Rev. 1965, 140, A1133–A1138.
[21] D. Figgen, G. Rauhut, M. Dolg, H. Stoll, Chem. Phys. 2005, 311,
227–244.
[22] K. A. Peterson, D. Figgen, M. Dolg, H. Stoll, J. Chem. Phys. 2007,
126, 124101.
[23] P. J. Stephens, J. F. Devlin, C. F. Chabalowski, M. J. Frisch, J. Phys.
Chem. 1994, 98, 11623–11627.
[24] a) C. Peng, H. B. Schlegel, Isr. J. Chem. 1993, 33, 449–454; b) C.
Peng, P. Y. Ayala, H. B. Schlegel, M. J. Frisch, J. Comput. Chem.
1996, 17, 49–56.
[25] a) K. Fukui, Acc. Chem. Soc. 1981, 12, 363–368; b) C. Gonzalez,
H. B. Schlegel, J. Chem. Phys. 1989, 90, 2154–2161; c) C. Gonzalez,
H. B. Schlegel, J. Phys. Chem. 1990, 94, 5523–5527.
[26] No large solvent effects could be detected experimentally; see
ref. [4], and, as we could not detect large dipole moments in our cal-
culations, we excluded a special calculation of solvent effects.
[27] a) D. F. McMillen, D. M. Golden, Annu. Rev. Phys. Chem. 1982, 33,
493–532; b) V. V. Grushin, H. Alper, Chem. Rev. 1994, 94, 1047–
1062.
[28] J. R. Brown, P. Schwerdtfeger, D. Schrçder, H. Schwarz, J. Am. Soc.
Mass Spectrom. 2002, 13, 485–492.
[29] We made a similar observation of no further improvement changing
from pVDZ to pVTZ, see: R. Dçpp, C. Lothschꢁtz, T. Wurm, M.
Pernpointner, S. Keller, F. Rominger, A. S. K. Hashmi, Organome-
tallics 2011, 30, 5894–5903.
[12] For representative reviews on gold catalysis, see: a) A. S. K. Hashmi,
G. Hutchings, Angew. Chem. 2006, 118, 8064–8105; Angew. Chem.
Int. Ed. 2006, 45, 7896–7936; b) A. Fꢁrstner, P. W. Davies, Angew.
Chem. 2007, 119, 3478–3519; Angew. Chem. Int. Ed. 2007, 46, 3410–
3449; c) A. S. K. Hashmi, Chem. Rev. 2007, 107, 3180–3211; d) Z.
Li, C. Brouwer, C. He, Chem. Rev. 2008, 108, 3239–3265; e) A.
Arcadi, Chem. Rev. 2008, 108, 3266–3325; f) D. J. Gorin, B. D.
Sherry, F. D. Toste, Chem. Rev. 2008, 108, 3351–3378; g) A. Corma,
A. Leyva-Pꢃrez, M. J. Sabater, Chem. Rev. 2011, 111, 1657–1712;
h) J. Xiao, X. Li, Angew. Chem. 2011, 123, 7364–7375; Angew.
Chem. Int. Ed. 2011, 50, 7226–7236; i) M. Rudolph, A. S. K.
Hashmi, Chem. Soc. Rev. 2012, 41, 2448–2462; j) Modern Gold Cat-
alyzed Synthesis (Eds.: A. S. K. Hashmi, F. D. Toste), Wiley-VCH,
Weinheim 2012.
[13] a) L. A. Jones, S. Sanz, M. Laguna, Catal. Today 2007, 122, 403–406;
b) B. Panda, T. K. Sarkar, Chem. Commun. 2010, 46, 3131–3133;
d) B. Panda, T. K. Sarkar, Tetrahedron Lett. 2010, 51, 301–305.
[14] a) Y. Shi, S. M. Peterson, W. W. Haberaecker III, S. A. Blum, J. Am.
Chem. Soc. 2008, 130, 2168–2169; b) J. J. Hirner, K. E. Roth, Y. Shi,
S. A. Blum, Organometallics 2012, 31, 6843–6850.
À
[30] For some theoretical and experimental studies on Au X bond-disso-
ciation energies without phosphine ligands; see: a) C. Puzzarini,
K. A. Peterson, Chem. Phys. 2005, 311, 177–186; b) M. Guiche-
merre, G. Chambaud, H. Stoll, Chem. Phys. 2002, 280, 71–102; c) H.
Schwarz, Angew. Chem. 2003, 115, 4580–4593; Angew. Chem. Int.
Ed. 2003, 42, 4442–4454.
[31] A full calculation with triphenylphosphane was not feasible due to
the system size, but a modeling with PMe3 is a justified approach,
see: C. M. Krauter, A. S. K. Hashmi, M. Pernpointner, ChemCatCh-
em 2010, 2, 1226–1230.
[15] S. Ma, Z. Yu, J. Org. Chem. 2003, 68, 6149–6152.
[32] Espinet et al. demonstrated that a cis–trans-isomerization process in
the coordination sphere of PdII can be efficiently catalyzed by using
AuI complexes, see: A. L. Casado, P. Espinet, Organometallics 1998,
17, 3677–3683. We also calculated the transmetalation process start-
ing from the corresponding cis-[PdXPhL2] complex, which gave no
deeper insight and decided for the reason of clarity not to include
this.
[16] a) M. H. Pꢃrez-Temprano, J. A. Casares, ꢈ. R. de Lera, R. ꢈlvarez,
P. Espinet, Angew. Chem. 2012, 124, 5001–5004; Angew. Chem. Int.
Ed. 2012, 51, 4917–4920; b) J. delPozo, D. Carrasco, M. H. Pꢃrez-
Temprano, M. Garcꢇa-Melchor, R. ꢈlvarez, J. A. Casares, P. Espinet,
Angew. Chem. 2012, 125, 2245–2249; Angew. Chem. Int. Ed. 2013,
52, 2189–2193.
[17] For an overview, see: a) A. S. K. Hashmi, Angew. Chem. 2010, 122,
5360–5369; Angew. Chem. Int. Ed. 2010, 49, 5232–5241; b) L. P.
Liu, G. B. Hammond, Chem. Soc. Rev. 2012, 41, 3129–3139.
[18] M. Pernpointner, A. S. K. Hashmi, J. Chem. Theory Comput. 2009,
5, 2717–2725; b) M. Lein, M. Rudolph, S. K. Hashmi, P. Schwerdt-
feger, Organometallics 2010, 29, 2206–2210; c) A. S. K. Hashmi, M.
Pernpointner, M. M. Hansmann, Faraday Discuss. 2011, 152, 179–
184.
[19] Gaussian 03, Revision C.01, M. J. Frisch, G. W. Trucks, H. B. Schle-
gel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, J. A. Montgom-
ery, Jr., T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S.
Iyengar, J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani,
N. Rega, G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K.
Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda,
O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian,
J. B. Cross, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann,
O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, P. Y.
Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J. Dannenberg,
V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C. Strain, O.
Farkas, D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B. Fores-
[33] a) V. Farina, B. Krishnan, J. Am. Chem. Soc. 1991, 113, 9585–9595;
b) C. Amatore, A. A. Bahsoun, A. Jutand, G. Meyer, A. N. Ntepe,
L. Ricard, J. Am. Chem. Soc. 2003, 125, 4212; c) R. Alvarez, O. N.
Faza, A. R. D. Lera, D. J. Cardenas, Adv. Synth. Catal. 2007, 349,
887–906; d) A. Nova, G. Ujaque, F. Maseras, A. Lledꢅs, P. Espinet,
J. Am. Chem. Soc. 2006, 128, 14571–14578; e) P. Espinet, A. M.
Echavarren, Angew. Chem. 2004, 116, 4808–4839; Angew. Chem.
Int. Ed. 2004, 43, 4704–4734; f) A. L. Casado, P. Espinet, J. Am.
Chem. Soc. 1998, 120, 8978–8985; g) A. L. Casado, P. Espinet, A. M.
Gallego, J. Am. Chem. Soc. 2000, 122, 11771–11782.
[34] For an overview, see: a) A. de Meijere, F. E. Meyer, Angew. Chem.
1994, 106, 2473–2506; Angew. Chem. Int. Ed. Engl. 1995, 33, 2379–
2411; b) I. P. Beletskaya, A. V. Cheprakov, Chem. Rev. 2000, 100,
3009–3066.
[35] R. Martin, S. L. Buchwald, Acc. Chem. Soc. 2008, 41, 1461–1473.
[36] a) A. A. C. Braga, N. H. Morgon, G. Ujaque, F. Maseras, J. Am.
Chem. Soc. 2005, 127, 9298–9307; b) C. Amatore, A. Jutand, G.
Le Duc, Chem. Eur. J. 2011, 17, 2492–2503.
[37] For studies analyzing the dependence of the ligand system on the
oxidative addition, see: a) S. Niu, M. B. Hall, Chem. Rev. 2000, 100,
Chem. Eur. J. 2013, 00, 0 – 0
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
13
&
ÞÞ
These are not the final page numbers!

M. Pernpointner, A. S. K. Hashmi et al.
353–406; b) W. A. Herrmann, C. P. Reisinger, M. Spiegler, J. Orga-
nomet. Chem. 1998, 557, 93–96; c) E. Zuidema, P. W. N. M. van
Leeuwen, C. Bo, Organometallics 2005, 24, 3703–3710.
[40] a) C. Amatore, M. Azzabi, A. Jutand, J. Am. Chem. Soc. 1991, 113,
8375–8384; b) C. Amatore, A. Jutand, A. Suarez, J. Am. Chem. Soc.
1993, 115, 9531–9541; for an overview, see: c) C. Amatore, A.
Jutand, Acc. Chem. Res. 2000, 33, 314–321.
[41] a) M.-E. Moret, D. Serra, A. Bach, P. Chen, Angew. Chem. 2010,
122, 2935–2939; Angew. Chem. Int. Ed. 2010, 49, 2873–2877; b) D.
Serra, M.-E. Moret, P. Chen, J. Am. Chem. Soc. 2011, 133, 8914–
8926.
[38] Recently, Gagne et al. were also unable to find suitable reaction
conditions for the cross-coupling of toluene triflate (TolOTf) with
a variety of palladium catalysts and observed redox incompatibilities
of the Pd/Au system, see ref. [6e].
[39] For a minireview on the role of chlorides in catalytic reactions, see:
a) A. Jutand, Appl. Organomet. Chem. 2004, 18, 574–582; b) A.
Jutand, A. Mosleh, Organometallics 1995, 14, 1810–1817.
Received: May 14, 2013
Published online: && &&, 2013
&
14
&
www.chemeurj.org
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Au/Pd-Mediated Cross-Coupling Reactions
FULL PAPER
A helping hand: For the transmetala-
tion from gold(I) to palladium(II),
halide ions play a crucial role (see
scheme). Without the formation of
a strong gold(I)–halogen bond (e.g.,
with sulfonate), the transmetalation
does not proceed.
Transmetalation
M. M. Hansmann, M. Pernpointner,*
R. Dçpp, A. S. K. Hashmi* . &&&& — &&&&
A Theoretical DFT-Based and Experi-
mental Study of the Transmetalation
Step in Au/Pd-Mediated Cross-
Coupling Reactions
Chem. Eur. J. 2013, 00, 0 – 0
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
15
&
ÞÞ
These are not the final page numbers!