Formation of Transient Anionic Metal Clusters in Palladium/Diene-Catalyzed Cross-Coupling Reactions

Article abstract of DOI:10.1002/chem.201902610

Despite their considerable practical value, palladium/1,3-diene-catalyzed cross-coupling reactions between Grignard reagents RMgCl and alkyl halides AlkylX remain mechanistically poorly understood. Herein, we probe the intermediates formed in these reactions by a combination of electrospray-ionization mass spectrometry, UV/Vis spectroscopy, and NMR spectroscopy. According to our results and in line with previous hypotheses, the first step of the catalytic cycle brings about transmetalation to afford organopalladate anions. These organopalladate anions apparently undergo S_N2-type reactions with the AlkylX coupling partner. The resulting neutral complexes then release the cross-coupling products by reductive elimination. In gas-phase fragmentation experiments, the occurrence of reductive eliminations was observed for anionic analogues of the neutral complexes. Although the actual catalytic cycle is supposed to involve chiefly mononuclear palladium species, anionic palladium nanoclusters [Pd_nR(DE)_n]^?, (n=2, 4, 6; DE=diene) were also observed. At short reaction times, the dinuclear complexes usually predominated, whereas at longer times the tetra- and hexanuclear clusters became relatively more abundant. In parallel, the formation of palladium black pointed to continued aggregation processes. Thus, the present study directly shows dynamic behavior of the palladium/diene catalyst system and degradation of the active catalyst with increasing reaction time.

Full text of DOI:10.1002/chem.201902610

A Journal of
Accepted Article
Title: Formation of Transient Anionic Metal Clusters in Palladium/
Diene-Catalyzed Cross-Coupling Reactions
Authors: Marlene Kolter and Konrad Koszinowski
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Chem. Eur. J. 10.1002/chem.201902610
Link to VoR: http://dx.doi.org/10.1002/chem.201902610
Supported by

10.1002/chem.201902610
Chemistry - A European Journal
FULL PAPER
Formation of Transient Anionic Metal Clusters in
Palladium/Diene-Catalyzed Cross-Coupling Reactions
Marlene Kolter and Konrad Koszinowski*[a]
Abstract:
Despite
their
considerable
practical
value,
process. Thus, highly complex and dynamic mixtures of
palladium species result, which have been aptly referred to as
cocktail-type systems.[7,8] Several examples suggest that
palladium aggregates display distinct reactivities. For instance,
isolated dimeric palladium complexes, such as [Pd2X2L2]
(X = Br,I; L = t-Bu3P) have been found to be not only efficient
precatalysts in Pd(0)-mediated reactions,[9–12] but also to react
directly with electrophilic substrates, such as aryl iodides.[13–17]
Similarly, isolated tri- and tetranuclear species have been
demonstrated to mediate the dehydrogenation of formic
acid[18-21] or C−C-coupling reactions.[22,23] Clearly, such
aggregates are of significant interest.
palladium/1,3-diene-catalyzed cross-coupling reactions between
Grignard reagents RMgCl and alkyl halides AlkylX remain
mechanistically poorly understood. Here, we probe the intermediates
formed in these reactions by a combination of electrospray-ionization
mass spectrometry, UV/Vis spectroscopy, and NMR spectroscopy.
According to our results and in line with previous hypotheses (Chem.
Soc. Rev. 2011, 40, 4937), the first step of the catalytic cycle brings
about transmetallation to afford organopalladate anions. These
organopalladate anions apparently undergo SN2-type reactions with
the AlkylX coupling partner. The resulting neutral complexes then
release the cross-coupling products via reductive eliminations. Using
gas-phase fragmentation experiments, we could observe the
occurrence of reductive eliminations for anionic analogues of the
neutral complexes. While the actual catalytic cycle supposedly
chiefly involves mononuclear palladium species, we also observed
the anionic palladium nanoclusters [PdnR(DE)n], n = 2, 4, 6 (DE =
diene). At short reaction times, the dinuclear complexes usually
predominated whereas at longer times, the tetra- and hexanuclear
clusters became relatively more abundant. In parallel, the formation
of palladium black pointed to continued aggregation processes. The
present study, thus, directly shows the dynamic behavior of the
palladium/diene catalytic system and the degradation of the active
catalyst with increasing reaction time.
The difficulty in elucidating the behavior of in-situ formed
palladium nanoclusters lies in their transient nature and the
scarcity of analytical methods capable of establishing their
chemical identity. Under typical conditions, the aggregation of
mononuclear palladium precursors occurs relatively fast and,
thus, limits the lifetime of individual nanoclusters.[23–26] Moreover,
the low-valent character of palladium nanoclusters supposedly
renders them susceptible to oxidation reactions. The transient
nature of in-situ formed nanoclusters excludes the use of
crystallographic
or
microscopic
methods
for
their
characterization. The small size of nascent nanoclusters
furthermore prohibits their analysis by dynamic light scattering.
While viable, spectroscopic techniques, in turn, promise to afford
only limited information on the molecular constitution of
palladium nanoclusters.
Introduction
Here, we report the observation of transient anionic nanoclusters
formed in palladium/diene-catalyzed cross-coupling reactions.
As Kambe and coworkers have shown, this catalytic system
works well for the coupling of alkyl halides with Grignard
reagents.[27–31] The postulated reaction mechanism starts with
the in-situ formation of a mononuclear bis-(µ3-allyl)-palladium
complex (Scheme 1).[27–31] The transfer of an organyl group from
the Grignard reagent onto the palladium center then affords an
anionic intermediate, which undergoes oxidative addition of an
alkyl halide in an SN2-type process. The resulting heteroleptic
Palladium-catalyzed transformations, such as cross-coupling
reactions, are among the most widely used tools in organic
synthesis and, as such, have been studied extensively. By now,
the mechanisms of numerous reactions mediated by neutral,
mononuclear palladium catalysts are thoroughly understood and
have become classical textbook knowledge.[1,2] It is also
generally recognized that palladium nanoparticles often exhibit
high catalytic activities.[3–6] The size distribution of these
nanoparticles can be reliably determined by microscopic
methods. In contrast, far less is known about the intermediate
size regime. The formation of nanoparticles from mononuclear
precursors must proceed via so-called nanoclusters, which
might also be released from nanoparticles in the reverse
neutral species releases the cross-coupling product in
a
reductive elimination and regenerates the bis-(µ3-allyl)-palladium
complex.
Using electrospray-ionization (ESI) mass spectrometry as our
chief analytical method, we provide the first experimental
evidence of the invoked anionic palladate complexes in the title
reactions. Besides the originally proposed mononuclear catalytic
intermediate, we also find related palladium nanocluster anions
in high signal intensities. ESI mass spectrometry offers the
unique advantage of affording direct information on their
[a]
M. Kolter, Prof. Dr. K. Koszinowski
Institut für Organische und Biomolekulare Chemie
Universität Göttingen
Tammannstraße 2, 37077 Göttingen (Germany)
E-mail: konrad.koszinowski@chemie.uni-goettingen.de
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

10.1002/chem.201902610
Chemistry - A European Journal
FULL PAPER
Analogous low-valent nanoclusters were also observed when
isoprene
was
exchanged
for
buta-1,3-diene
(DEB,
Figures S3-S5) and when PhMgCl or BnMgCl were used as
transmetallating agent (Figures S6-S20). In contrast, trinuclear
species ([Pd3Bu(DEB)4], [Pd3H(DEB)4], and [Pd3Bu(DEB)3],
Figure S3) proved to be less abundant and stable.
In all cases, the relative signal intensities of the palladate ions
changed considerably over time (Figure 2 for the reaction of
[Pd2(dba)3] with PhMgCl and isoprene). When we probed
sample solutions immediately after their preparation, we could
observe the low-valent mononuclear complexes [PdR(DEI)2],
R = Bu and Ph, as well as [PdPh(DEI)] in low signal intensities
(Figures S21-S26). The mononuclear complexes were only
short-lived and completely disappeared after a few minutes. In
marked contrast, the relative signal intensity of the dinuclear and,
in particular, of the tetranuclear nanoclusters increased with time.
For solutions of [Pd2(dba)3], BnMgCl, and isoprene, even the
hexanuclear aggregate [Pd6Bn(DEI)6] could be detected after
longer reaction times (Figures S19 and S20). The absolute
signal intensities of the observed palladate ions decreased with
time. Besides palladates of the type [PdnR(DEI)n], we also
observed complexes [PdnR3(DEI)n] with palladium in higher
average oxidation states (Figures S6, S10, S16, S24). In
addition, hydride-containing species, such as [PdnH(DEI)n] and
[PdnRxH3−x(DEI)n], were also present in most cases (Figure 1,
Figures S3, S6, S10, S16, S19, S21).
Scheme 1. Mechanism of palladium-catalyzed cross-coupling reactions
between alkyl halides and Grignard reagents in the presence of buta-1,3-diene
as suggested by Kambe and coworkers.[29,30]
aggregation states. The presence of the observed palladium
nanoclusters and their further growth at longer reactions times
point to a considerably higher complexity of the catalytic system
than previously assumed. We complement our ESI-mass
spectrometric studies by gas-phase fragmentation experiments
as well as UV/Vis- and NMR- spectroscopic measurements.
Results
Species Formed by Transmetallation
Upon the analysis of a solution of the palladium(0) precatalyst
[Pd2(dba)3] (dba
= dibenzylideneacetone), n-BuMgCl, and
isoprene (DEI) in tetrahydrofuran (THF) by negative-ion mode
ESI mass spectrometry, the di- and tetranuclear organopalladate
anions [PdnBu(DEI)n], n = 2 and 4, were detected as main
species (Figure 1, Figures S1 and S2, Supporting Information).
Figure 2. Negative-ion mode ESI mass spectra of a solution of Pd2(dba)3
(1.5 mM), isoprene (DEI, 24 mM), and PhMgCl (12 mM) in THF after 0, 11, and
17 minutes after removal from the cooling bath (78 °C, a = [AlPh4],
b = [PdPh3(DEI)],
c = [Pd2Ph2(DEI)2H],
d = [Pd2Ph3(DEI)2],
e = [Pd4Ph(DEI)3]). [AlPh4] stems from aluminum impurities in the used
Grignard reagent.
To probe whether the palladate complexes preferentially
incorporate isoprene or buta-1,3-diene, we performed
competition experiments with the two compounds present in
equimolar amounts (together with Pd2(dba)3 and n-BuMgCl). In
the resulting negative-ion mode ESI mass spectra, the signal
intensities of the butadiene-containing palladates exceeded
those of the complexes bearing both dienes, whereas hardly any
.
Figure 1. Negative-ion mode ESI mass spectrum of a solution of Pd2(dba)3
(1.5 mM), isoprene (DEI, 24 mM), and n-BuMgCl (12 mM) in THF
(a = [Pd2H(DEI)2], b = [Pd2Bu2H(DEI)2]).
This article is protected by copyright. All rights reserved.

10.1002/chem.201902610
Chemistry - A European Journal
FULL PAPER
palladates
(Figures S27-S30).
binding
only
isoprene
were
detected
appearance of a new absorption maximum at λ = 450 nm along
with a broad absorption at wavelengths between 500 and
700 nm (Figure 3). At longer reaction times, these absorption
features decreased in intensity. Moreover, the precipitation of
palladium black was observed by the bare eye.
Control experiments with the palladium(II) precursors PdCl2, PdI2,
or Pd(OAc)2 combined with dienes and Grignard reagents did
not yield any detectable organopalladate complexes. Instead,
the resulting negative-ion mode ESI mass spectra were of only
rather low absolute signal intensity and were dominated by non-
transmetalated PdII complexes or heterobimetallic adducts, such
as [Pd2Cl3(DE)] and [MgPd2Cl5(DE)] (Figure S31). Even with
the Pd0 precursor Pd2(dba)3, stable signal intensities could only
be achieved when the precatalyst and the diene were allowed to
react for at least 30 min before the addition of the Grignard
reagent.
In further control experiments, we replaced the Grignard reagent
by organolithium and -zinc compounds, respectively. The
transmetallation of [Pd2(dba)3] with n-BuLi in the presence of
isoprene led to the formation of the palladate complexes
[PdnBu(DEI)n], n = 2, 4, 6, in high signal intensities. In addition,
the mononuclear complexes [PdBu(DEI)] and [PdBu(DEI)2] as
well as several heterobimetallic complexes, such as
[LiPd2Bu2(DEI)4] and [LiPd3Bu2(DEI)4] were observed
(Figures S32-S34). Like in the case of the reactions with the
Grignard reagents, the mono- and dinuclear palladates rapidly
disappeared whereas the tetra- and hexanuclear ones increased
in time (Figure S35). After a few minutes, the overall signal
intensity strongly decreased. No palladate complexes could be
detected when n-BuZnCl · LiCl was used as transmetallating
agent (Figure S36).
Figure 3. UV/Vis spectra of a solution of Pd2(dba)3 (1 mm) and isoprene
(16 mm), stirred at 278 K for 50 min, before (black) and after the addition of
8 eq. of n-BuMgCl after 1 min (red) and 120 min (blue) reaction time at 278 K.
To gain more information on the binding mode of the diene in
the palladate complexes, we conducted NMR-spectroscopic
experiments. When isoprene was added to the palladium(0)
precursor, the 1H-NMR signals of the former did not shift and
were only slightly broadened (Figure S49). Upon the addition of
PhMgBr to the sample solution, an immediate broadening of the
isoprene signals could be observed both in the 1H- and 13C-NMR
spectra (Figure 4, Figure S50).
Analysis of the sample solutions by positive-ion mode ESI mass
spectrometry showed abundant magnesium-containing cations,
but no palladium complexes (Figure S37). For also obtaining
insight into possibly formed neutral palladium species, we made
use of a charge-tagged diene. In this way, quasi-neutral species
incorporating this diene become amenable to ESI-mass
spectrometric analysis.[32,33] Upon the treatment of a solution of
[Pd2(dba)3] and (E)-buta-1,3-dien-1-yltriphenylphosphonium
bromide, (DEPBr), with n-BuMgCl, we detected, among several
palladium-free species, the complexes [PdBu(DEP)2] and
[PdBuH(DEP)(PPh3)] (Figures S38-S47, note that both of these
species bear a total charge of +1 resulting from the positive
charge of the DEP units and the anionic or neutral character of
the palladium center, as indicated in the given formulae). The
former can be considered the cationized analogue of the
mononuclear
palladate
anions
dicussed
true quasi-neutral
above.
[PdBuH(DEP)(PPh3)] corresponded to
a
species containing a free PPh3 moiety, which presumably
originated from the decomposition of the charge-tagged diene.
Among the observed palladium-free species, an ion with the

sum formula of C48H49P2 was of potential interest because its
stoichiometry matches that expected for the addition product of
two cationic diene molecules and a butyl anion.
Figure 4. 1H NMR spectra (400 MHz, THF-d8) of a mixture of Pd2(dba)3
(25 mM) and isoprene (100 mM), (a) before and (b) after the addition of
PhMgBr at 298 K. The signals marked with arrows correspond to isoprene, the
signals at  = 1.73 and 3.58 ppm (♦) to THF.
UV/Vis spectroscopy showed that the band characteristic of
Pd2(dba)3 (absorption maximum at λ = 520 nm) did not change
after the addition of isoprene (Figure S48). The treatment with
n-BuMgCl led to the decrease of this band and to the
This article is protected by copyright. All rights reserved.

10.1002/chem.201902610
Chemistry - A European Journal
FULL PAPER
Unimolecular Reactivity of Organopalladates Resulting from
Transmetallation
Reactivity toward Organyl Halides
We next examined the reactivity of the in-situ formed palladates
toward propyl bromide as a prototypical substrate of cross-
coupling reactions. When PrBr was added to a solution of
[Pd2(dba)3], isoprene, and PhMgCl, [PdPh2Pr(DEI)2]as well as
the related hydride-containing complexes [PdPh2H(DEI)2] and
[PdPh2Pr(DEI)] were found (Figure 6). These intermediates
apparently resulted from the reaction of [PdPh(DEI)n], n = 1 and
2, with PrBr. Interestingly, no analogous higher aggregates were
observed. At longer reaction times, the propyl-containing
palladates vanished and the [PdnPh(DEI)n] nanoclusters, n = 2,
4, and 6, reappeared (Figures S68 and S69).
Upon fragmentation in the gas phase, the tetra- and hexanuclear
[PdnR(DE)n] nanoclusters preferentially lost neutral [Pd2(DE)2],
eq. (1) with n = 4 (for DEI and DEB) and 6 (for DEI) (Figures 5,
S51-S57 and Table S1). Further prominent fragmentation
pathways resulted in the loss of single diene molecules, eq. (2)
(Figures S51-S55, S57-S62).
[PdnR(DE)n]
[PdnR(DE)n]
 [Pdn−2R(DE)n−2][Pd2(DE)2]
 [PdnR(DE)n−1]DE
(1)
(2)
Butyl-containing palladates also released butene, eq. (3)
(Figures S51, S52, S55, S60, S63-S65).
[PdnBu(DE)n]  [PdnH(DE)n]C4H8
(3)
Figure 6. Negative-ion mode ESI mass spectrum of a solution of Pd2(dba)3
(1.5 mM), isoprene (DEI, 24 mM), PhMgCl (12 mM), and PrBr (12 mM) in THF
(a = [PdPh(DEI)], b = [PdPh2Pr(DEI)], c = [PdPh3(DEI)], d = [Pd2Ph(DEI)2]).
Figure 5. Mass spectrum of mass-selected [Pd6Bu(DEI)6] (DEI = isoprene)
and its fragment ions produced upon collision-induced dissociation (ELAB
=
10.0 eV, a = [Pd6H(DEI)4]).
When BnMgCl was used as the transmetallating agent, the
reaction
with
PrBr
afforded
[PdBn2Pr(DEI)2]
and
[PdBn2H(DEI)2], which were analogous to the intermediates
formed from the phenyl-containing palladates (Figures S70-S76).
In contrast to the former case, the reaction of the benzyl-
containing palladates now also yielded small quantities of the
dinuclear complex [Pd2Bn2Pr(DEI)2], thus pointing to the
reactivity of [Pd2Bn(DEI)2] toward PrBr. In addition, small
amounts of [PdBn2Pr(DEI)] and [PdBn2H(DEI)] were observed.
However, similar to the experiments without added PrBr, the
obtained ESI mass spectra were dominated by [PdBn3(DEI)n]
complexes, n = 0 – 3.
All of these fragmentation processes occurred both as primary
and consecutive reactions. In addition, we also observed
product ions that showed the incorporation of dioxygen.
Obviously, these species did not originate from fragmentation
processes, but from ion-molecule reactions with residual traces
of O2 present in the vacuum system of the mass spectrometer.
For the cationized palladate [PdBu(DEP)2], the main
fragmentation pathway was the loss of butene (Figure S64). In
addition, it also afforded several other fragment ions stemming
from ligand dissociation and ligand decomposition reactions. In
contrast, the formation of the above-mentioned dimerization
Upon the addition of PrBr to solutions of [Pd2(dba)3], isoprene,
and n-BuMgCl, the dinuclear complex [Pd2Bu2Pr(DEI)2] formed,
but exhibited only a short lifetime comparable to that of
[Pd2Bu(DEI)2] (Figures S77-S79). In contrast, the tetranuclear
cluster [Pd4Bu(DEI)4] reached its maximum signal intensity only
after the decline of the [Pd2Bu2Pr(DEI)2] product complex
(Figures S77-S79). The reaction with PrI also furnished
[Pd2Bu2Pr(DEI)2] (Figure S80), whereas that with PrCl or the
secondary alkyl bromide i-PrBr did not.

product C48H49P2 was not observed. Other related (DEP)-
containing
species,
such
as
[PdBuH(DEP)(PPh3)],
[PdBu(DEP)2(PPh3)], and [Pd(DEP)(PPh3)2], also showed the
elimination of butene and/or ligand dissociation and
decomposition reactions (Figures S65-S67).
This article is protected by copyright. All rights reserved.

10.1002/chem.201902610
Chemistry - A European Journal
FULL PAPER
In a control experiment, we added the aryl iodide ethyl-4- Discussion
iodobenzoate (ArI) to a solution of [Pd2(dba)3], isoprene, and
n-BuMgCl. Aryl halides are not prone to undergo SN2-type
reactions and therefore should not readily react according to the
mechanism postulated by Kambe and coworkers (Scheme 1).
The control experiment did not show a palladate intermediate
containing both a butyl and an aryl group, but instead resulted in
the detection of [PdArI2] and [PdArI2(DEI)] along with
[Pd2I3(DEI)] and purely inorganic anions (Figure S81). When the
amount of ArI was reduced to 0.4 equiv. (relative to n-BuMgCl),
[PdArBuH(DEI)2] was observed as the main species
Palladates in Different Aggregation States
The mononuclear complexes [PdR(DEI)2], R = Bu and Ph,
found in the present study have exactly the same stoichiometry
as the reactive intermediates that Kambe and coworkers have
proposed to result from the transmetallation of a palladium(II)
precursor.[29,30] Moreover, the corresponding cationized
analogue became detectable when the charge-tagged diene
was applied. The observation of this type of complex both in its
anionic and its cationized form points to its facile formation
under the given conditions.
(Figures S82-S89).
[PdArBuH(DEI)], [PdAr2Bu(DEI)], and [PdArBu2(DEI)] along
with and
[Pd2Bu(DEI)2], [Pd2Ar(DEI)2], [PdAr(DEI)2],
[PdAr(DEI)].
Other detectable species included
However, our time-dependent measurements show that the
mononuclear [PdR(DEI)2] complexes have only rather short
lifetimes and readily afford small nanoclusters [PdnR(DE)n], n =
2, 4, and 6. The predominance of nanoclusters with even
numbers of palladium centers as well as their gas-phase
fragmentation behavior implies that these species consist of
dimeric subunits. A survey of low-valent palladium clusters
reported in the literature reveals diverse structural motifs.
Moiseev has shown that tetrameric palladium clusters can
feature square, rectangular, or tetrahedral structures.[34] The
geometry of these clusters is controlled both by the ligand
environment as well as the formal oxidation state of the involved
palladium centers. In addition, linear[35,36] and butterfly-
shaped[21,37] structures are known as well.
To obtain additional information on possibly formed neutral
intermediates, we also probed solutions of [Pd2(dba)3], isoprene,
n-BuMgCl,
and
the
charge-tagged
substrate
(6-bromohexyl)triphenylphosphonium bromide ((RBr)Br). The
resulting positive-ion mode ESI mass spectra showed [RH],
RBr, (RBr)2Br, and very small amounts of the cross-coupling
product RBu (Figure S90). Increasing the quantity of the
Grignard reagent to 3 equiv. (relative to RBr) led to a higher
signal intensity of the cross-coupling product (Figure S91).
Negative-ion mode ESI afforded [PdBr3R]2 and, at higher
concentrations
of
n-BuMgCl,
[PdBrBuPhR]2
and
[PdBu2PhR]2as the only palladium-containing species
(Figures S92-S95). The latter two ions apparently originated
from a decomposition of the triphenylphosphonium group of the
charged tag.
Independent support for the formation of low-valent polynuclear
palladium complexes in the sample solutions comes from the
observation of a broad absorption band in the visible range.
Such an absorption at higher wavelengths is commonly ascribed
to the formation of palladium clusters.[24,25,38] Apparently, the
palladium nanoclusters continue to grow beyond [Pd6R(DE)6]
and furnish even larger clusters, whose m/z ratio lies outside of
the probed range, ultimately affording nanoparticles and metallic
palladium. This assumption is fully in line with the observed
precipitation of palladium black from the sample solutions.
Unimolecular Reactivity of Palladates Resulting from the
Reactions with Organyl Halides
Upon gas-phase fragmentation, the palladate complexes formed
from reactions with organyl halides preferentially underwent
reductive eliminations (Figures S96-S100). [PdPh2Pr(DEI)2] and,
to a very small extent, [PdPh2Pr(DEI)] reacted in this manner to
give only the homo-coupling product, eq. (4) with n = 1 and 2
(Figures S96 and S97). In contrast, the fragmentation of the
dinuclear complex [Pd2Bu2Pr(DEI)2] yielded both the cross- and
the homo-coupling products in similar amounts (Figure S98), eq.
(5a) and (5b), whereas that of [PdAr2Bu(DEI)] gave the cross-
coupling, eq. (6), but not the homo-coupling product
(Figure S99).
Binding Mode of the Diene
The ESI-mass spectrometric detection of palladate complexes
containing just one diene unit as well as the loss of a single
diene molecule upon the gas-phase fragmentation of the
[PdnR(DEI)n] ions is at odds with the dimerization of the diene
and the formation of bis-allylic complexes postulated by Kambe
and coworkers (Scheme 1). The signal broadening observed in
the NMR spectra indicates a fast exchange between palladium-
bound and free isoprene, but, again, does not give any evidence
of an allylic binding mode of the diene or its dimerization. We
also considered the possibility that the observed ion with the
[PdPh2Pr(DEI)n]
 [PdPr(DEI)n]Ph2
(4)
[Pd2Bu2Pr(DEI)2]
 [Pd2Bu(DEI)2]PrBu
 [Pd2Pr(DEI)2]Bu2
(5a)
(5b)
[PdBuAr2(DEI)]
 [PdAr(DEI)]ArBu
(6)

sum formula of C48H49P2 could have resulted from the coupling
Further notable fragmentation reactions released hydrocarbons
RH, eq. (7) with R = Ph, Bn, and Bu (Figures S96-S98, S101).
of a charge-tagged diene dimer and a butyl group brought about
by a reductive elimination (Scheme S1, left). However, the
failure of the cationized palladate [PdBu(DEP)2] to release

[PdR2Pr(DEI)n]
 [PdRPr(DEI)n−H]RH (7)

C48H49P2 upon gas-phase fragmentation provides strong
evidence against this hypothesis. Instead, C48H49P2 presumably
originated from the reaction of a butyl anion with one diene
This article is protected by copyright. All rights reserved.

10.1002/chem.201902610
Chemistry - A European Journal
FULL PAPER
molecule followed by a consecutive addition to another diene
unit (Scheme S1, right). Thus, our experiments do not give any
indication for a palladium-mediated diene dimerization under the
present conditions. If such a dimerization takes place, it affects
ony a small fraction of the employed diene.
addition of the Grignard reagent) also points to slow exchange
reactions between the dba and diene ligands.
Among the different transmetalating reagents tested, n-BuMgCl
and n-BuLi show a similar performance whereas n-BuZnCl · LiCl
fails to produce detectable organopalladates. Presumably, this
difference reflects the lower reactivity of organozinc reagents.
Given that the latter are known to undergo highly efficient
transmetallation reactions with PdII centers in Pd-catalyzed
Negishi cross-coupling reactions,[42–45] the non-occurrence of
analogous processes in the present case points to a diminished
Lewis acidity of the palladium center. Such a diminished Lewis
acidity is to be expected for palladium in an oxidation state of 0.
The competition experiments show that the palladate complexes
with buta-1,3-diene are more stable than those with isoprene.
According to the results of Kambe and coworkers, the use of
buta-1,3-diene affords significantly better yields in cross-
coupling reactions than that of isoprene.[28] Apparently, the more
stable diene palladate complexes are catalytically more active.
Decomposition Processes
Catalytic Cycle and Role of Palladium Nanoclusters
The formation of hydride-containing complexes can be easily
rationalized for the use of n-BuMgCl as transmetallating agent
because the palladium complexes resulting from the transfer of
a butyl group can undergo β-hydride elimination (in the sample
solution or during the ESI process). Our gas-phase
fragmentation experiments directly showed the occurrence of
these reactions. Indeed, the propensity of alkylpalladium
complexes to undergo β-hydride eliminations is well known.[1]
However, the observation of [PdnPh2H(DE)m] (n, m = 1, 2) and
[PdnBn2H(DE)m] (n, m = 1–3) in the reactions with PhMgCl and
BnMgCl, respectively, suggests the operation of alternative
pathways that afford hydride-containing palladates. A possible
mechanism furnishing these products consists of an insertion of
the diene into an organyl-metal bond followed by the elimination
of a [(DE),RH] diene, in which one hydrogen atom has been
substituted by the organyl group R. Such a mechanism has
recently been proposed to explain the presence of similar metal-
hydride complexes observed in cobalt/diene-catalyzed cross-
coupling reactions.[39]
The observation of a manifold of anionic palladate complexes in
the present study suggests that anions are also involved in the
catalytic cycle, as proposed by Kambe and coworkers. While we
were not able to confirm the postulated formation of neutral
palladium bis-allyl species, we did detect the anions supposedly
resulting from these species by transmetallation (Scheme 2).
Our results strongly indicate that these anions contain not only
dimerized, but also – and presumably predominantly – separate
diene units (see above). As our experiments cannot directly
distinguish between the actual catalytically active intermediates
and the resting states of the catalyst, it could well be the case
that complexes containing dimerized dienes correspond to the
active species, but do not accumulate in significant
concentrations due to their higher reactivity. The anionic
intermediate then reacts with the alkyl halide in an SN2-type
process to afford a neutral complex, for which again the binding
of both dimerized and separate diene units has to be considered.
Although we could not observe this neutral complex directly, we
detected anionic analogues resulting from
a
further
transmetallation reaction, i.e., [PdR2Alkyl(DE)n]. In contrast, the
[PdArI2(DEI)n] species found in control experiments with ArI
substrates presumably originate from conventional oxidative
additions, which correspond to the typical reactions between
(neutral) Pd0 complexes and aryl halides.
Another type of decomposition process appears to be involved
in the formation of [PdnRxH3−x(DEI)n] complexes, whose average
oxidation states are increased relative to [PdnR(DE)n]. Possibly,
they resulted from the latter by the oxidation by residual traces
of oxygen present in the sample solution and/or the mass
spectrometer and consecutive transmetallation. The occurrence
of such processes would point to the high susceptibility of the
electron-rich diene-palladate complexes to oxidation reactions.
Alternatively, the complexes in higher oxidation states could also
originate from the reaction of the low-valent palladates with
remaining traces of RCl in the used Grignard reagents.[40]
The decrease of the signal intensity of the [PdR2Alkyl(DE)n]
intermediates at longer reaction times points to their
consumption in consecutive reactions, as expected for
a
catalytic process. As our gas-phase fragmentation experiments
directly demonstrate, these consecutive reactions correspond to
reductive eliminations. The presence of two identical organyl
groups R in the anionic complexes implies that the reductive
elimination can furnish both the desired cross-coupling and the
undesired homo-coupling product. Apparently, the branching
between the cross- and homo-coupling channels is controlled by
rather subtle effects and does not depend mainly on the
electronic properties of the organyl substituents, as had been
found for the reductive eliminations from other transition-metal
complexes.[46–49] This selectivity problem is avoided for reductive
eliminations from the neutral complexes, which only bear a
single R group. The high yields of the cross-coupling product
obtained with palladium/diene catalysts in synthetic studies
Influence of Palladium Precursor and Transmetallating
Agent
The inability of the palladium(II) precursors PdCl2, PdI2, or
Pd(OAc)2 to afford detectable organopalladate complexes upon
treatment with a diene and a Grignard reagent suggests that the
diene molecules do not react with the nascent Pd0 centers fast
enough to prevent the occurrence of competing processes under
the present conditions.[41] The need for relatively long incubation
times of the Pd2(dba)3 precatalyst and the diene (before the
This article is protected by copyright. All rights reserved.

10.1002/chem.201902610
Chemistry - A European Journal
FULL PAPER
suggest that under typical conditions the reductive elimination
occurs mainly at the stage of the neutral complexes. Indeed,
neutral transition-metal complexes in general are known to
undergo reductive eliminations more easily than the
corresponding anionic ones.[46,48]
Comparison to Nickel/Diene and Cobalt/Diene Catalysts
As Kambe and coworkers have demonstrated, not only
palladium/diene, but also nickel/diene[27] and cobalt/diene
catalysts[50] are effective in mediating alkyl-alkyl cross-coupling
reactions between alkyl halides and Grignard reagents.
Particularly the nickel-catalyzed reactions have been extensively
investigated. In the course of these investigations, Kambe and
coworkers
isolated
the
anionic
nickelate
complex
PhNi(η1,η3-C8H12)− from a solution of Ni(cod)2 (cod = cycloocta-
1,5-diene), PhLi, buta-1,3-diene, and 12-crown-4 in high
yields.[50] This complex contains two dimerized diene units and
has a structure analogous to that of the palladate anion shown in
Scheme 1. The apparently high tendency of nickel to mediate
the dimerization of the diene contrasts the behavior of the
palladates, which moreover showed only a lowered binding
affinity to the diene. This lowered binding affinity gives rise to
coordinately unsaturated complexes, which are prone to
aggregation reactions and, ultimately, the formation of palladium
black. In contrast, the mononuclear nickelate complexes
apparently are much more stable. Presumably, this higher
stability is one reason why cross-coupling reactions catalyzed by
nickel/diene have been proven synthetically more useful than
those catalyzed by palladium/diene.
While no intermediate from cobalt/diene-catalyzed reactions has
been isolated and structurally characterized, several in-situ
formed cobaltate anions have been identified by ESI mass
spectrometry under conditions very similar to those of the
present study, thus allowing for a direct comparison of the
obtained results. In the presence of a diene and a Grignard
reagent RMgCl, CoCl2 mainly afforded [CoHxR2x(DE)n] anions
(x = 0,1, n = 2–4).[39] These species differ from the palladate
anions in several aspects:
Scheme 2. Simplified catalytic cycle of a palladium-catalyzed cross-coupling
reaction between an alkyl halide and a Grignard reagent in the presence of
buta-1,3-diene. Species observed by ESI mass spectrometry (as shown or as
isoprene-containing analogues) are highlighted. Note that the present
experiments do not allow an unambiguous differentiation between different
isomers.
(i)
The [CoHxR2x(DE)n] complexes bear two organyl (or
hydride) substituents, whereas the
[PdnR(DE)n]
complexes contain only one. This difference reflects the
fact that cobalt has one valence electron less than
palladium.
The full catalytic system is significantly more complex than the
shown catalytic cycle based on mononuclear palladium
complexes. In particular, the role of the observed anionic
nanoclusters has to be clarified. Apart from the dinuclear
complexes [Pd2R2Pr(DEI)2], R = Bn and Bu, no aggregate
incorporating an organyl substituent stemming from the organyl
halide could be detected. This finding points to lower catalytic
activities of the nanoclusters in comparison with the
mononuclear complexes. For the palladium/diene-catalyzed
reaction between BuMgCl and PrBr, the time profiles of the
different anions detected by ESI mass spectrometry exclude the
direct participation of palladates [PdnR(DE)n] with n > 2 in the
catalytic process. Therefore, the main role of the palladium
nanoclusters seems to be the provision of a reservoir for the
reactive mononuclear complexes. The facile dissociation of the
small clusters upon gas-phase fragmentation suggests the
reversibility of the aggregation process. However, if the cluster
growth leads up to the formation of larger nanoparticles and,
eventually, palladium black, a deterioration of the catalyst
performance is to be expected. Thus, a highly dynamic nature of
the catalytic system results.
(ii) The cobaltate complexes incorporate a higher number of
diene units than their palladate counterparts. This
difference can be ascribed to a higher binding affinity of
the diene to cobalt than to palladium and/or a higher
tendency of the dienes to insert into Co−C than into Pd−C
bonds.
(iii) In contrast to the palladates, the cobaltates do not form
low-valent nanoclusters. The higher stability of the
mononuclear cobaltates is in accordance with their
supposedly higher binding affinity to the diene and their
reluctance to form coordinatively unsaturated species,
which then would aggregate.
Although palladium, nickel, and cobalt, thus, all give
organometallate anions upon their treatment with a Grignard
reagent and a diene, the exact nature of these complexes differs
considerably.
This article is protected by copyright. All rights reserved.

10.1002/chem.201902610
Chemistry - A European Journal
FULL PAPER
Information). All other organometallic reagents were purchased. The
concentrations of the organometallic reagents were determined by
Conclusions
iodometric
titration.[51]
(E)-Buta-1,3-dien-1-yltriphenylphosphonium
ESI mass spectrometry revealed the formation of anionic ate
bromide and (6-bromohexyl)triphenylphosphonium bromide were
synthesized in analogy to procedures reported in the literature.[52–54]
complexes
in
palladium/diene-catalyzed
cross-coupling
reactions. The transmetallation of [Pd2(dba)3] with a Grignard
reagent RMgCl or an organolithium compound RLi in the
presence of a diene furnished [PdnR(DE)n] species, n = 1, 2, 4,
and 6. These complexes resemble a mononuclear palladate
anion containing a dimerized diene moiety, which Kambe and
coworkers had postulated in analogy to a known intermediate
from nickel/diene-catalyzed reactions. However, the observed
palladates show a lower tendency to bind dienes and to mediate
the dimerization of the latter. The reduced diene affinity of the
palladates and, accordingly, their on average lower degree of
coordinative saturation explain why they undergo aggregation
reactions, which first afford nanoclusters and, eventually,
palladium black. While the preponderance of even-numbered
nanoclusters points to their formation from dinuclear subunits,
the present results do not permit a more definite determination
of their structures.
Sample solutions were prepared by dissolving Pd2(dba)3 in dry THF and
the addition of the diene followed by stirring at room temperature for
45 min. The solutions were then cooled down to 195 K before the
organometallic reagent and, in some of the experiments, the organyl
halide was added. The resulting solutions were analyzed immediately.
ESI Mass Spectrometry
Sample solutions were transferred into the ESI source of a micrOTOF-
Q II mass spectrometer (Bruker Daltonik) via gas-tight syringes at typical
flow rates of 8 µL min1. In the experiment probing the reaction of
[Pd2(dba)3], isoprene and n-BuMgCl with PrI, the sample solution was
kept in the cooled flask and transferred into the inlet line leading into the
ESI source by applying a slight overpressure of Ar to enable the
detection of the short-lived complex [Pd2Bu2Pr(DEI)2].[55,56] The ESI
source was operated at a voltage of 3500 V and with N2 as nebulizer gas
(flow rate of 5 L min−1) and drying gas (0.7 bar backing pressure,
temperature of 333 K). In simple MS1 experiments, all generated ions
with 50 ≤ m/z ≤ 3000 were allowed to pass the quadrupole mass filter of
the instrument. In MS2 fragmentation experiments, ions of interest were
mass-selected in the quadrupole-mass filter, accelerated to a kinetic
energy ELAB, and allowed to collide with N2 gas. Residual parent ions and
resulting fragment ions were then detected after passage through the
TOF analyzer. In most cases, accuracies of the measured m/z ratios of
≤ 25 ppm were obtained with an external calibration with a mixture of
CF3COOH and phosphazenes in H2O/CH3CN. In some cases, an internal
calibration proved necessary. To this end, the m/z scale was adjusted to
match the theoretical m/z ratio of an ion, whose identity could be
independently established with certainty (indicated for each experiment).
Theoretical exact m/z ratios and isotope patterns were calculated with
the DataAnalysis software package (Bruker Daltonik).
The observation of palladate anions bearing organyl substituents
stemming from the RMgCl or RLi reagents shows that for the
given catalytic cycle, as inferred by Kambe and coworkers, the
transmetallation step precedes that of the oxidative addition of
alkyl halides AlkylX. The latter apparently operates in an SN2-
type fashion and furnishes neutral intermediates, such as
[PdRAlkyl(DEI)n], which could not be detected in the present
experiments. However, the formation of these species is
suggested by the observation of related complexes
[PdR2Alkyl(DEI)n],
which
are
anionized
by
further
transmetallation reactions. Upon gas-phase fragmentation,
these anionic complexes predominantly undergo reductive
eliminations. Besides the desired cross-coupling products RAlkyl,
the reductive eliminations also yield homo-coupling products R2.
Such unwanted competing reactions are not to be expected for
the actual catalytic cycle because the putative neutral
intermediates [PdRAlkyl(DEI)n] cannot afford the homo-coupling
product.
UV/Vis Spectroscopy
Time-dependent UV/Vis-spectroscopic experiments were carried out with
a Cary60 instrument (Agilent Technologies). Sample solutions were
prepared and analyzed under argon atmosphere at 278 K in dry THF
using an Excalibur immersion probe (Hellma Analytics). Spectra were
recorded over a wavelength range of 350-1100 nm with a resolution of
5 nm.
The palladium nanoclusters have a lower tendency to react with
organyl halides than their mononuclear congeners. Hence, they
most likely do not correspond to the active catalytic
intermediates themselves, but affect the catalytic cycle indirectly
by tying up palladium atoms. The growth of the nanoclusters at
the expense of the mononuclear complexes renders the whole
catalytic system highly dynamic. Thus, the present case
constitutes another example of aggregation processes raising
the complexity of palladium catalysis to a formidable level.
NMR Spectroscopy
NMR spectra were recorded with a Bruker Avance III 400 instrument
operating at 400 MHz (1H) or 100 MHz (13C) or a Bruker avance III HD
300 instrument at 298 K. The chemical shifts were calibrated relative to
the solvent signals (THF-d8: 1.73 ppm (1H), 67.6 ppm (13C)).
Acknowledgements
We thank Dr. Friedrich Kreyenschmidt for the synthesis of (E)-
buta-1,3-dien-1-yltriphenylphosphonium
bromide
and
Experimental Section
(6-bromohexyl)triphenylphosphonium bromide and gratefully
acknowledge funding from the CaSuS (Catalysis for Sustainable
Synthesis) program (scholarship for M.K.).
General
To exclude moisture and oxygen, all experiments were carried out under
standard Schlenk conditions. THF was dried over sodium/benzophenone
and freshly distilled before use. PhMgBr in THF-d8 and BnMgCl were
synthesized according to literature procedures (see the Supporting
Keywords: Cluster compounds • Cross-coupling • Mass
spectrometry • Palladium • Reactive intermediates
This article is protected by copyright. All rights reserved.

10.1002/chem.201902610
Chemistry - A European Journal
FULL PAPER
[28] J. Terao, Y. Naitoh, H. Kuniyasu, N. Kambe, Chem. Lett. 2003, 32,
890–891.
[1]
[2]
C. Elschenbroich, Organometallchemie, Teubner, Stuttgart, 2003.
A. de Meijere, S. Bräse, M. Oestreich, Metal-Catalyzed Cross-Coupling
Reactions and More, Wiley-VCH, Weinheim, 2014.
J. Durand, E. Teuma, M. Gómez, Eur. J. Inorg. Chem. 2008, 2008,
3577–3586.
[29] J. Terao, N. Kambe, Acc. Chem. Res. 2008, 41, 1545–1554.
[30] N. Kambe, T. Iwasaki, J. Terao, Chem. Soc. Rev. 2011, 40, 4937–4947.
[31] T. Iwasaki, A. Fukuoka, W. Yokoyama, X. Min, I. Hisaki, T. Yang, M.
Ehara, H. Kuniyasu, N. Kambe, Chem. Sci. 2018, 9, 2195–2211.
[32] K. L. Vikse, M. A. Henderson, A. G. Oliver, J. S. McIndoe, Chem.
Commun. 2010, 46, 7412–7414.
[3]
[4]
A. Balanta, C. Godard, C. Claver, Chem. Soc. Rev. 2011, 40, 4973–
4985.
[5]
[6]
[7]
[8]
[9]
N. Yan, Y. Yuan, P. J. Dyson, Dalton Trans. 2013, 42, 13294–13304.
C. Deraedt, D. Astruc, Acc. Chem. Res. 2014, 47, 494–503.
V. P. Ananikov, I. P. Beletskaya, Organometallics 2012, 31, 1595–1604.
D. B. Eremin, V. P. Ananikov, Coord. Chem. Rev. 2017, 346, 2–19.
J. P. Stambuli, R. Kuwano, J. F. Hartwig, Angew. Chem. Int. Ed. 2002,
41, 4746–4748.
[33] M. A. Schade, J. E. Fleckenstein, P. Knochel, K. Koszinowski, J. Org.
Chem. 2010, 75, 6848–6857.
[34] I. I. Moiseev, J. Organomet. Chem. 1995, 488, 183–190.
[35] H. Kurosawa, J. Organomet. Chem. 2004, 689, 4511–4520.
[36] T. Murahashi, H. Kurosawa, Coord. Chem. Rev. 2002, 231, 207–228.
[37] K. Shimamoto, Y. Sunada, Chem. Eur. J. 2019, 25, 3761–3765.
[38] M. T. Reetz, M. Maase, Adv. Mater. 1999, 11, 773–777.
[39] F. Kreyenschmidt, K. Koszinowski, Chem. Eur. J. 2018, 24, 1168–1177.
[40] T. Gärtner, W. Henze, R. M. Gschwind, J. Am. Chem. Soc. 2007, 129,
11362–11363.
[10] T. J. Colacot, Platin. Met. Rev. 2009, 53, 183–188.
[11] M. Aufiero, F. Proutiere, F. Schoenebeck, Angew. Chem. Int. Ed. 2012,
51, 7226–7230.
[12] F. Proutiere, M. Aufiero, F. Schoenebeck, J. Am. Chem. Soc. 2012, 134,
606–612.
[41] Note that Kambe and coworkers have successfully used combinations
of PdII precursors, 1,3-dienes, and Grignard reagents for cross-coupling
reactions with alkyl (pseudo)halides, see [28].
[13] K. J. Bonney, F. Proutiere, F. Schoenebeck, Chem. Sci. 2013, 4, 4434–
4439.
[14] I. Kalvet, K. J. Bonney, F. Schoenebeck, J. Org. Chem. 2014, 79,
12041–12046.
[42] J. A. Casares, P. Espinet, B. Fuentes, G. Salas, J. Am. Chem. Soc.
2007, 129, 3508–3509.
[15] G. Yin, I. Kalvet, F. Schoenebeck, Angew. Chem. Int. Ed. 2015, 54,
6809–6813.
[43] B. Fuentes, M. García-Melchor, A. Lledós, F. Maseras, J. A. Casares,
G. Ujaque, P. Espinet, Chem. Eur. J. 2010, 16, 8596–8599.
[44] M. García-Melchor, B. Fuentes, A. Lledós, J. A. Casares, G. Ujaque, P.
Espinet, J. Am. Chem. Soc. 2011, 133, 13519–13526.
[16] M. Aufiero, T. Sperger, A. S.-K. Tsang, F. Schoenebeck, Angew. Chem.
Int. Ed. 2015, 54, 10322–10326.
[17] M. Aufiero, T. Scattolin, F. Proutière, F. Schoenebeck, Organometallics
2015, 34, 5191–5195.
[45] J. Li, L. Jin, C. Liu, A. Lei, Org. Chem. Front. 2014, 1, 50–52.
[46] A. Putau, H. Brand, K. Koszinowski, J. Am. Chem. Soc. 2012, 134,
613–622.
[18] I. Gauthron, J. Gagnon, T. Zhang, D. Rivard, D. Lucas, Y. Mugnier, P.
D. Harvey, Inorg. Chem. 1998, 37, 1112–1115.
[47] S. Weske, R. Schoop, K. Koszinowski, Chem. Eur. J. 2016, 22, 11310–
11316.
[19] D. Meilleur, D. Rivard, P. D. Harvey, I. Gauthron, D. Lucas, Y. Mugnier,
Inorg. Chem. 2000, 39, 2909–2914.
[48] S. Weske, R. A. Hardin, T. Auth, R. A. J. O’Hair, K. Koszinowski, C. A.
Ogle, Chem. Commun. 2018, 54, 5086–5089.
[20] D. Meilleur, P. D. Harvey, Can. J. Chem. 2001, 79, 552–559.
[21] D. Evrard, D. Meilleur, M. Drouin, Y. Mugnier, P. D. Harvey, Z. Anorg.
Allg. Chem. 2002, 628, 2286–2292.
[49] T. Parchomyk, K. Koszinowski, Chem. Eur. J. 2018, 24, 16342–16347.
[50] T. Iwasaki, H. Takagawa, S. P. Singh, H. Kuniyasu, N. Kambe, J. Am.
Chem. Soc. 2013, 135, 9604–9607.
[22] A. Sachse, M. John, F. Meyer, Angew. Chem. Int. Ed. 2010, 49, 1986–
1989.
[51] A. Krasovskiy, P. Knochel, Synthesis 2006, 890–891.
[52] P. L. Fuchs, Tetrahedron Lett. 1974, 15, 4055–4058.
[53] D. E. Morrison, J. B. Aitken, M. D. de Jonge, F. Issa, H. H. Harris, L. M.
Rendina, Chem. Eur. J. 2014, 20, 16602–16612.
[23] A. Leyva‐Pérez, J. Oliver‐Meseguer, P. Rubio‐Marqués, A. Corma,
Angew. Chem. Int. Ed. 2013, 52, 11554–11559.
[24] J. Wang, H. F. M. Boelens, M. B. Thathagar, G. Rothenberg,
ChemPhysChem 2004, 5, 93–98.
[54] D. Rasina, M. Otikovs, J. Leitans, R. Recacha, O. V. Borysov, I.
Kanepe-Lapsa, I. Domraceva, T. Pantelejevs, K. Tars, M. J. Blackman,
et al., J. Med. Chem. 2016, 59, 374–387.
[25] A. V. Gaikwad, G. Rothenberg, Phys. Chem. Chem. Phys. 2006, 8,
3669–3675.
[26] D. A. Alonso, C. Nájera, Chem. Soc. Rev. 2010, 39, 2891–2902.
[27] J. Terao, H. Watanabe, A. Ikumi, H. Kuniyasu, N. Kambe, J. Am. Chem.
Soc. 2002, 124, 4222–4223.
[55] K. L. Vikse, M. P. Woods, J. S. McIndoe, Organometallics 2010, 29,
6615–6618.
[56] K. L. Vikse, Z. Ahmadi, J. Luo, N. van der Wal, K. Daze, N. Taylor, J. S.
McIndoe, Int. J. Mass Spectrom. 2012, 323–324, 8–13.
This article is protected by copyright. All rights reserved.

10.1002/chem.201902610
Chemistry - A European Journal
FULL PAPER
Entry for the Table of Contents (Please choose one layout)
Layout 2:
FULL PAPER
M. Kolter, K. Koszinowski*
Page No. – Page No.
Formation of Transient Anionic Metal
Clusters in Palladium/Diene-
Catalyzed Cross-Coupling Reactions
A combination of ESI mass spectrometry, gas-phase fragmentation experiments,
UV/Vis and NMR spectroscopy found anionic nanoclusters in palladium-catalyzed
cross-coupling reactions with dienes as additives.
This article is protected by copyright. All rights reserved.