Kinetico-mechanistic insights on the assembling dynamics of allyl-cornered metallacycles: The Pt-Npy bond is the keystone

Article abstract of DOI:10.1002/chem.201403467

The square-like homo- and heterometallamacrocycles [{Pd(η³-2-Me-C₃H₄)(Lⁿ)₂}₂{M(dppp)}₂](CF₃SO₃)₆ (dppp = 1,3-bis(diphenylphosphino)propane) and [

Full text of DOI:10.1002/chem.201403467

DOI: 10.1002/chem.201403467
Full Paper
&
Self-Assembly
Kinetico-Mechanistic Insights on the Assembling Dynamics of
Allyl-Cornered Metallacycles: The PtÀNpy Bond is the Keystone
Inmaculada Angurell,[a] Montserrat Ferrer,*[a] Albert Gutiꢀrrez,[a] Manuel Martꢁnez,*[a]
Mercꢂ Rocamora,[a] Laura Rodrꢁguez,[a] Oriol Rossell,[a] Yvonne Lorenz,[b] and
Marianne Engeser[b]
Abstract: The square-like homo- and heterometallamacrocy-
cles [{Pd(h3-2-Me-C3H4)(Ln)2}2{M(dppp)}2](CF3SO3)6 (dppp=1,3-
bis(diphenylphosphino)propane) and [{Pd(h3-2-Me-C3H4)-
(L1)2}2{M(PPh3)2}2](CF3SO3)6 [py=pyridine, M=Pd, Pt, Ln=4-
PPh2py (L1), 4-C6F4PPh2py (L2)] containing allyl corners were
synthesised by antisymbiotic self-assembly of the different
palladium and platinum metallic corners and the ambiden-
tate N,P ligands. All the synthesised assemblies displayed
a complex dynamic behaviour in solution, the rate of which
is found to be dependent on the electronic and/or steric
nature of the different building blocks. A kinetico-mechanis-
tic study by NMR line shape analysis of the dynamics of
some of these assemblies was undertaken in order to deter-
mine the corresponding thermal activation parameters. Both
an enhanced thermodynamic stability and slower dynamics
were observed for platinum-pyridine-containing species
when compared with their palladium analogues. Time-de-
pendent NMR spectroscopy in combination with ESI mass
spectrometry was used to study the exchange between the
assemblies and their building blocks, as well as that occur-
ring between different metallamacrocycles. Preliminary stud-
ies were carried out on the activity of some of the metalla-
macrocyclic compounds as catalytic precursors in the allylic
substitution reaction, and the results compared with that of
the monometallic allylic corner [Pd(h3-2-Me-C3H4)(L1)2]+.
Introduction
Even though the use of supramolecules as reactors has pro-
vided a significant number of amazing examples of activity
and selectivity,[3] an alternative approach by designing new
functional self-assembled supramolecular metallamacrocycles
is also possible; the final purpose being their direct participa-
tion in selected catalytic processes.[4] In particular, and given
the fact that the majority of the assembled complexes report-
ed so far have coordinatively inert metal centres, the incorpo-
ration of weakly coordinating ligands, such as olefins, seems to
be a good strategy to preclude the destruction of the supra-
molecular species during the catalytic transformations.
Since the dawn of supramolecular coordination chemistry in
the 1980s, an overwhelming number of discrete metallasupra-
molecular architectures with different combinations of metal
centres and ligands has been synthesised and characterised.[1]
Although the formation of complex and aesthetically pleasing
structures has been a goal by itself in many cases, the poten-
tial of self-assembled structures originates in the derived func-
tions of the species produced. In this respect, functionalisation
of these architectures, which is possible by employing func-
tion-specific ligands and/or metal centres in the assembly pro-
cess, should produce molecules having a potential use as cata-
lysts, sensors or in material sciences.[1f,l,2]
With this alternative strategy in mind, we have recently de-
scribed[5] the incorporation of allyl-palladium building blocks
such as {Pd(h3-2-Me-C3H4)}+, into well-defined, self-assembled
homo- and heterometalla architectures. In these, the typical di-
phosphane or diamine capping ligands have been substituted
in part by the 2-Me-C3H4 allyl group. It is known that allyl-palla-
dium moieties have a potential utility in processes that form
new CÀC or carbonÀheteroatom bonds,[6] either as precursors
of the catalytic species or as intermediates. These include allyl-
ic substitutions[7] and hydrovinylation reactions[8] that could be
incorporated to the metallamacrocyclic reactivity. Surprisingly,
despite these obvious reasons for the interest in these macro-
cyclic architectures, little work in the field has been carried
out.
[a] Dr. I. Angurell, Dr. M. Ferrer, Dr. A. Gutiꢀrrez, Prof. M. Martꢁnez,
Dr. M. Rocamora, Dr. L. Rodrꢁguez, Prof. O. Rossell
Departament de Quimica Inorgꢂnica
Universitat de Barcelona
Martꢁ i Franquꢃs 1-11, 08028 Barcelona (Spain)
Fax: (+34)934907725
E-mail: montse.ferrer@qi.ub.es
manel.martinez@qi.ub.es
[b] Dipl.-Chem. Y. Lorenz, Dr. M. Engeser
Kekulꢀ-Institut fꢄr Organische Chemie und Biochemie
Universitꢅt Bonn
Gerhard-Domagk-Str.1
53121 Bonn (Germany)
In our preliminary study[5] we found that the obtained allyl-
palladium metallamacrocycles, [{Pd(h3-2-Me-C3H4)(PPh2py)2}2{M-
(dppp}2](CF3SO3)6 (dppp=1,3-bis(diphenylphosphino)propane,
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201403467.
Chem. Eur. J. 2014, 20, 1 – 16
1
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
M=Pd, Pt), featured a rather uncommon dynamic behaviour
in solution that implied, in the case of the platinum-containing
species, a remarkable lability of the usually inert PtII square-
planar centres. Although the dynamic nature of coordination
bonds in self-assembly is widely recognised as a keystone of
the nature of supramolecular architectures, only few studies
that characterise in a systematic and quantitative way such ex-
changes in this kind of species have been reported.[9] It is clear
that a good knowledge of the general solution behaviour, as
well as detailed thermodynamic and/or kinetic data for the
molecular self-assembled systems, become especially impor-
tant when rationalising their application in different functions,
such as molecular recognition or catalytic processes among
others.
Results and Discussion
Synthesis and characterisation of the ligand 4-PPh2C6F4py
(L2) (py=pyridine)
With the aim of obtaining larger metallamacrocycles than
1aL1c and 1bL1c (Scheme 1), the new ambidentate ligand L2,
that is, 4-(4-diphenylphosphino)-tetrafluorophenylpyridine, was
synthesised by lithiation of the previously described[11] 4-
bromo-tetrafluorophenylpyridine by using nBuLi (1.6 m in
hexane) in a thf solution at À788C, followed by the addition of
diphenylchlorophosphane
at
the
same
temperature
(Scheme 2). A white crystalline solid was obtained, which
showed spectroscopic, mass spectrometric and analysis data
consistent with the proposed structure (see the Experimental
Section).
In our previous communication[5] we described the selective
self-assembly of the dppp derivatives 1a and 1b, the allyl-palla-
dium metallic block c, and the heteroditopic linker L1 to pro-
duce square like metallamacrocycles displaying two different
metal corners 1aL1c and 1bL1c (Scheme 1).
Scheme 2. Synthesis of ligand L2.
Synthesis and characterisation of the metallamacrocycles
As indicated above, our interest was centred on widening our
study of the metallamacrocycles depicted in Scheme 1 by
tuning the basicity and/or size of the different building blocks
in order to better understand the observed exchange process.
As a first option, two PPh3 ligands were used as ancillary li-
gands in the non-organometallic corner instead of the chelat-
ing ligand dppp. For the synthesis of a homometallic derivative
[{Pd(h3-2-Me-C3H4)(PPh2py)2}2{Pd(PPh3)}2](CF3SO3)6 (2aL1c), the
self-assembly reaction (Scheme 3, left) was performed under
the same conditions than those already used for the synthesis
of 1aL1c and 1bL1c.[5] Mixing [Pd(h3-2-Me-C3H4)(cod)](CF3SO3) (c,
cod=1,5-cyclooctadiene), [Pd(H2O)2(PPh3)2](CF3SO3)2 (2a) and
4-PPh2py (L1) in a molar ratio of 1:1:2 in CH2Cl2 at room tem-
perature produced the targeted metallamacrocycle 2aL1c as
a white solid in good yield. Alternatively, the same macrocycle,
that is, 2aL1c, can be obtained by reaction of the preformed
[Pd(h3-2-Me-C3H4)(4-PPh2py)2](CF3SO3) (L1c) with the bis(triphe-
nylphosphane) corner block 2a under the same conditions of
solvent and temperature (Scheme 3, right).
Scheme 1. Schematic representation of the self-assembly of the dppp deriv-
atives 1a and 1b, the allyl-palladium building block c and the linker L1.
In that paper, we provided a detailed characterisation of
these species, including X-ray crystal structure determination,
fluxional behaviour in different solvents and a kinetico-mecha-
nistic study of the non-rigidity of the 1aL1c compound. The
obtained thermal activation parameters were indicative of an
associatively activated exchange process that involved the
whole metallamacrocycle and some of its fragments. This ex-
change mechanism is in contrast with that based on the rota-
tion of the pyridine moiety of the ligand L1, postulated by sev-
eral authors.[9b,10]
For the synthesis of the analogous heterometallamacrocycle
2bL1c, namely [{Pd(h3-2-Me-C3H4)(PPh2py)2}2{Pt(PPh3)}2](CF3SO3)6,
a new strategy had to be used because the platinum acceptor
unit [Pt(H2O)2(PPh3)2](CF3SO3)2 (2b) could not be isolated in
solid state. Unlike for the palladium compound cis-[PdCl2-
(PPh3)2], treatment of cis-[PtCl2(PPh3)2] with silver triflate result-
ed in the formation of unidentified silver phosphane deriva-
tives; therefore, the acceptor corner {Pt(PPh3)2}2+ (2b) was gen-
erated in situ. Thus, equimolar quantities of the preformed allyl
palladium donor [Pd(h3-2-Me-C3H4)(4-PPh2py)2](CF3SO3) (L1c)
and the precursor cis-[PtCl2(PPh3)2] were mixed in dichloro-
In the present report, we expand the scope of the described
methodology to diverse homo- and heterometalla macrocyclic
structures with allyl-palladium {Pd(h3-2-Me-C3H4)} subunits. In
these structures, the basicity of the phosphane ligands at the
corners and/or the size of the resulting metallacycles have
been carefully tuned. The different dynamic processes involved
in the synthesis and stability of the prepared species have
been looked into, and a preliminary study about the activity of
some of these compounds as catalysts in the allylic alkylation
process has been undertaken.
&
&
Chem. Eur. J. 2014, 20, 1 – 16
www.chemeurj.org
2
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
Scheme 3. Self-assembly pathways for the generation of the different metallamacrocycles.
methane at room temperature and, after several minutes, two
equivalents of solid AgCF3SO3 were added with constant stir-
ring. After two hours and subsequent work up the new macro-
cycle 2bL1c was obtained as a yellow powder in good yield.
Similarly to the ones previously described for the macrocycles
1aL1c and 1bL1c,[5] the 31P{1H} NMR spectra of the new com-
pounds 2aL1c and 2bL1c showed two singlets that can be as-
signed to the P atoms of the PPh3 and 4-PPh2py ligands, re-
spectively. As expected, 195Pt satellites (1J(P,Pt)=3323 Hz) asso-
ciated with the PPh3 signals were found for the heterometallic
1
1
Pd/Pt macrocycle 2bL1c. The H NMR spectra were found to be
Figure 1. Temperature-dependent H NMR spectra of the a pyridine region
for the different metallamacrocycles.
solvent- and temperature-dependent; in CDCl3 and at 298 K
1
the H NMR spectra of 2aL1c and 2bL1c display rather broad
signals that sharpen and/or split on lowering the temperature.
As an example, compound [{Pd(h3-2-Me-C3H4)(PPh2py)2}2{Pd-
(PPh3)}2](CF3SO3)6 (2aL1c), shows a broad single resonance at
298 K assigned to the a-pyridine protons, which splits into two
different peaks at approximately 240 K (Figure 1, top). The ob-
served low-temperature NMR pattern resembles the one found
for the related dppp derivative 1aL1c at 298 K (Figure 1,
The formation of the PPh3-containing metallamacrocycles,
2aL1c and 2bL1c was further supported by ESI Fourier trans-
form ion cyclotron resonance (FTICR) mass spectrometry as
shown in the Supporting Information (Figures S3 and S4). The
mass spectrum of 2aL1c displays a peak at m/z=1617.2 corre-
sponding to the doubly charged species [2aL1c-2(CF3SO3)]2+
which appears superimposed with that of the mono-charged
[Pd2(h3-2-Me-C3H4)(PPh3)2(4-PPh2py)2(CF3SO3)2]+
fragment
,
1
middle). On the other hand, the H NMR spectrum in CDCl3 of
the 2bL1c analogue shows different a and a’ pyridine proton
broad signals at 298 K (Figure 1, bottom), which fully sharpen
at approximately 280 K. However, in the more coordinating
medium acetone, only one set of sharp signals was observed
for the pyridine protons at room temperature (Figures S1 and
S2 in the Supporting Information). Thus, variable-temperature
(VT)-NMR studies clearly indicate that the palladium macrocy-
cle 2aL1c shows faster dynamics than the PtÀN bond-contain-
ing, heterometallic analogue 2bL1c. This is in agreement with
the relative inertness of PtII complexes, as shown in many nu-
cleophilic substitution processes.[12]
formed by fragmentation of the macrocyclic architecture. For
the platinum-containing compound 2bL1c, the peaks corre-
sponding to the loss of two (m/z=1706.2) and three (m/z=
1087.5) triflate anions from the compound are observed, as
well as a signal at m/z=2324.3 that corresponds to the triply
charged aggregation species [(2aL1c)2-3(CF3SO3)]3+, (arising
from the interaction of two metallamacrocycles and the loss of
three triflate anions).[13] In both spectra, additional signals were
observed, which are due to fragmentation processes occurring
in the electrospray source and which are in full accordance
with the structure of the macrocycles.
Chem. Eur. J. 2014, 20, 1 – 16
www.chemeurj.org
3
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
Once the changes in the phosphane capping ligand had
been studied, the effect on the solution dynamics of increasing
the macrocyclic size by changing the connecting ambidentate
ligand was pursued. The new heteroditopic ligand bearing one
tetrafluorobenzene group between the pyridine and the phos-
phorus atom, 4-PPh2C6F4py (L2) was utilised, whereas the basic-
ity and bite angle on the phosphane-containing corner was
maintained as in the ligand dppp. Additionally, apart from al-
lowing the creation of bigger assemblies, ligand L2 has differ-
ent electronic characteristics than its related counterpart L1.
The introduction of a perfluorinated group next to the pyridine
ring produces a decrease of the
showed only a moderate solubility in CDCl3, hampered the
study even more.
Fortunately, ESI-FTICR mass spectrometry strongly supports
the formation of 1aL2c and 1bL2c. The mass spectra of these
compounds show very informative peaks arising from the sub-
sequent loss of anions from the parent metallamacrocycles
(Figures 2 and S9 in the Supporting Information). As usually
observed,[14] the more stable platinum derivative 1bL2c under-
goes lesser fragmentation and doubly [1bL2c-2(CF3SO3)]2+
,
triply [1bL2c-3(CF3SO3)]3+ and quadruply [1bL2c-4(CF3SO3)]4+
charged species can be found in the mass spectrum (Figure 2).
basicity of the Npy atom, which
has to be considered when dis-
cussing the dynamics of the
system.
Ligand L2 efficiently self-as-
sembles with [M(H2O)2(dppp)]2+
[M=Pd (1a), Pt (1b)] and [Pd(h3-
2-Me-C3H4)]+ (c) building blocks
in dichloromethane at room
temperature, yielding the target-
ed homo-, that is, [{Pd(h3-2-Me-
C3H4)(PPh2C6F4Py)2}2{Pd(dppp)}2]-
(CF3SO3)6 (1aL2c), and heterote-
trametallic, that is, [{Pd(h3-2-Me-
C3H4)(PPh2C6F4Py)2}2{Pt(dppp)}2]-
(CF3SO3)6, (1bL2c) metallamacro-
cycles in good yields (Scheme 3,
left).
The NMR spectra of the larger
macrocyclic species L2 display
a more complex solution dy-
namics than found for its L1 ana-
logues described above. The
resonances in the 31P{1H} NMR
spectra assigned to the ligand
L2 (Figures S5 and S6 in the Sup-
Figure 2. ESI-FTICR mass spectrum of 1bL2c in acetone.
porting Information), for both
1aL2c and 1bL2c, are broad and
indicative of an exchange pro-
NMR evidences of metallacycle building block exchange
cess in solution involving the macrocyclic ring. Moreover, the
dppp signal appears as a sharp singlet, which is in accordance
with it being tightly bonded to the Pd or the Pt centres on the
As reported in our previous communication,[5] the 1H NMR
spectra of both 1aL1c and 1bL1c in CDCl3 at room temperature
revealed two sets of signals for each a and b proton pairs of
the pyridine rings, as well as four different peaks for the ali-
phatic protons of dppp (Figure S10 in the Supporting Informa-
1
NMR time scale. The H NMR spectra were found to be solvent
and temperature dependent. Although the pattern in acetone
solution is similar to that observed for the L1-derived metalla-
macrocycles (Figures S7 and S8 in the Supporting Information),
in CDCl3 solution at 298 K only broad and poorly defined sig-
nals for all the coordinated ligands were obtained, which is in
agreement with a very complex molecular movement in solu-
tion. Unfortunately, in this case, VT-1H NMR spectroscopy in
CDCl3 did not provide further structural details about the dy-
namics of the system. Poorly defined spectra were obtained at
temperatures in the 223–313 K range, and no trends could be
inferred from the splitting of the very broad signals. Moreover,
the fact that the fluorinated derivatives 1aL2c and 1bL2c
1
tion). When the H NMR spectra were recorded at higher tem-
peratures or in more coordinating solvents such as
[D3]nitromethane or [D6]acetone, a trend to a simpler NMR pat-
tern was observed. These changes imply a reversible cleavage
of MÀL1 bonds either at the phosphane corner {(dppp)MÀNpy}
(M=Pd, Pt ) or at the allyl fragment {(h3-2-Me-C3H4)PdÀP}
(Scheme 3, right and bottom). These dissociation processes
open up the possibility of envisaging an exchange between
the supramolecular species and their constituent building
blocks as a plausible alternative to a mere simple rotation
&
&
Chem. Eur. J. 2014, 20, 1 – 16
www.chemeurj.org
4
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
around the MÀNpy bond that does not account for the ob-
same L1c block did not show evidence of ligand exchange be-
cause cross peaks between the a pyridine proton signals of
1aL1c and L1c were not observed (Figure 3b) even at longer
mixing times (1000 ms) and at the highest recommended tem-
perature for CDCl3 (313 K). These observations are in line with
the higher inertness of the PtÀNpy bond when compared with
the PdÀNpy bond.[14c]
served NMR changes. With the aim of investigating these ex-
1
changes, we undertook a series of H–1H NOESY NMR experi-
ments where the metallamacrocycles 1aL1c and 1bL1c were
mixed in the NMR tubes with the appropriate stoichiometric
amount of one of their building blocks. A NOESY experiment
not only provides cross peaks due to NOE contacts but also
cross peaks between mutually exchanging positions and con-
sequently is a powerful technique for the investigation of
chemical exchange in multisite systems.
Interestingly, for the fluorinated metallacycle 1bL2c, ex-
change NOESY cross peaks (Figure 3c) are even observed
when the Pt-containing heterometallic assembly is mixed with
the appropriate corner L2c in CDCl3 at room temperature and
at low mixing time (500 ms). The establishment of a dynamic
equilibrium between the metallamacrocycles and their building
blocks is thus further supported, as well as an exchange rate
enhancement observed for the architectures derived from the
fluorinated linkers. The worse donor character of the linker L2
induces a higher acidity of the PtII centre, and thus enables
a faster associatively activated classical substitution process in
1bL2c.
In spite of the dynamical exchange, a 1:2 mixture of metalla-
macrocycle 1aL1c and the free corner L1c in CDCl3 showed sep-
arate a pyridine proton signals that displayed cross peaks in
the 1H–1H NOESY NMR spectrum (Figure 3a). It is thus clear
that there is a moderately fast exchange between the whole
assembly [{Pd(h3-2-Me-C3H4)(4-PPh2py)2}2{Pd(dppp)}2](CF3SO3)6
(1aL1c) and its corresponding piece [Pd(h3-2-Me-C3H4)(4-
1
PPh2py)2](CF3SO3) (L1c). Contrarily, the H–1H NOESY NMR spec-
trum of a mixture of the platinum analogue 1bL1c and the
The collected data do not rule out an alternative dynamic
process involving the reversible dissociation of the PdÀP bond
at the allylic Lnc corners.[15] In fact, the 1H NMR spectrum of
a 1:2 mixture of the L1c palladium corner [Pd(h3-2-Me-C3H4)(4-
PPh2py)2](CF3SO3) and free 4-PPh2py (L1) in CDCl3 shows
a single broad signal assigned to the pyridine a protons, im-
plying a very fast PdÀP dissociation equilibrium. Consequently,
1H–1H NOESY NMR experiments were conducted to study the
exchange between 1aL1c and 1bL1c assemblies and the build-
ing block [M(dppp)(4-PPh2py)2]2+ [M=Pd (1aL1), M=Pt (1bL1)]
whose synthesis and characterisation is described in the Exper-
imental Section. In spite of the fast equilibrium observed be-
tween the free 4-PPh2py (L1) and its palladium complex L1c, no
exchange cross peaks between the proton pyridine signals
were observed between the 1aL1c and 1bL1c derivatives and
the 1aL1 and 1bL1 blocks, respectively. These data confirm that
the reversible dissociation of the MÀNpy bond is the main pro-
cess involved in the dynamic behaviour specifically observed
in the reported metallamacrocycles.
In order to complement the exchange experiments indicated
above, CH2Cl2 solutions containing the palladium metallama-
crocycles 1aL1c or 1aL2c and the platinum corner 1b in a 1:2
molar ratio were analysed by 31P{1H} NMR spectroscopy. In
both cases, the 31P{1H} NMR spectra recorded immediately after
mixing the components, show a complete exchange of the
palladium fragment {Pd(dppp)} by the platinum fragment {Pt-
(dppp)} within the supramolecules (Scheme 4). Only the char-
acteristic signals of the heterometallacycles 1bL1c or 1bL2c, as
well as those of the free palladium corner 1a, were found. The
experiments confirmed not only the easy dissociation of the
PdÀNpy bonds, but also the remarkable thermodynamic stabili-
ty of the PtÀNpy bonds in the metallamacrocycle, responsible
for the observed preference for these heterometallic assem-
blies (see below).
1
Figure 3. Section of the H–1H NOESY NMR spectra of a) a 1:2 mixture of
metallamacrocycle 1aL1c and the free corner L1c in CDCl3 at 298 K (mixing
time 500 ms), relevant cross peaks are indicated by arrows, b) a 1:2 mixture
of metallamacrocycle 1bL1c and the free corner L1c in CDCl3 at 298 K
(mixing time 1000 ms) and c) a 1:2 mixture of metallamacrocycle 1bL2c and
the free corner L2c in CDCl3 at 298 K (mixing time 500 ms), relevant ex-
change cross peaks are indicated by arrows.
Chem. Eur. J. 2014, 20, 1 – 16
www.chemeurj.org
5
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
observed intensity trend might
be due, at least partially, to dif-
ferences in the tendency toward
fragmentation. These differences
could be a consequence of the
different kinetic and/or thermo-
dynamic stability of the various
species. Thus, the intermediate
intensity of the mixed 1a1bL1c
macrocycle clearly indicates that
the existence of PtÀNpy bonds
has a significant stabilising effect
on these supramolecules. More-
Scheme 4. Representation of the results obtained by using 31P{1H} NMR spectroscopy to clarify the exchange be-
haviour of the platinum and palladium supramolecules.
Electrospray ionisation mass spectrometry studies of the ex-
change between the metallamacrocycles
over, the mass spectrum displays some additional peaks that
correspond to aggregation species arising from the interaction
of two different metallamacrocycles (1aL1c/1a1bL1c, 1aL1c/
1bL1c and 1bL1c/1a1bL1c) with the loss of two triflate counter
anions (Figure 4).
Electrospray mass spectrometry is a soft ionisation method
that has been widely applied for the characterisation of supra-
molecules.[16] Because ESI-MS is able to identify the presence of
multiple species simultaneously, it can be used to
provide at least qualitative information on the com-
position of complex mixtures,[9d,17] and consequently
it can be used to detect free components, reaction
intermediates, wrongly assembled structures and the
desired final thermodynamic assemblies that co-exist
until equilibrium is reached. The technique is then
very well suited for the characterisation of the dy-
namic block exchange processes as well as the sub-
sequent thermodynamic equilibration that appear on
the metallamacrocycles prepared. Thus, we carried
out a series of ESI-MS experiments with the aim of
determining the composition of the solutions ob-
tained by mixing two different assemblies or their
corresponding building blocks. All the experiments
were undertaken in acetone to provide reliable and
constant spray conditions in the electrospray ion
source.
The ESI-MS analysis of an equimolar mixture of the
metallacycles 1aL1c and 1bL1c reveals, together with
the characteristic peaks of the individual compounds
1aL1c (m/z=3159.1 [1aL1c-(CF3SO3)]+ and 1505.1
[1aL1c-2(CF3SO3)]2+) and 1bL1c (m/z=3336.2 [1bL1c-
(CF3SO3)]+ and 1593.1 [1bL1c-2(CF3SO3)]2+), further
signals at m/z=3248.1 and 1549.1 (Figure 4). Their
isotopic distribution perfectly fits with the calculated
pattern for these species resulting from the loss of
one and two triflate anions from the nonsymmetrical
1a1bL1c mixed assembly, that is, [{Pd(h3-2-Me-
C3H4)(4-PPh2py)2}2{Pd(dppp)}{Pt(dppp)}](CF3SO3)6.
As shown in Figure 4, the ESI-MS spectrum ob-
tained after 30 minutes of mixing shows a decrease
of the intensity of the signals on increasing the
number of palladium atoms in the assembly that is,
the most intense peak corresponds to Pd2Pt2 (1bL1c),
whereas the least intense peak is assigned to Pd4
(1aL1c). Because the stoichiometry needs equimolar
Figure 4. Section of the ESI-MS spectrum of a 1:1 mixture of 1aL1c and 1bL1c in acetone.
Top) Taken after 30 minutes after mixing. Bottom) After the mixture has reached
quantities of both precursors 1aL1c and 1bL1c in
equilibrium with the mixed 1a1bL1c macrocycle, the equilibrium.
&
&
Chem. Eur. J. 2014, 20, 1 – 16
www.chemeurj.org
6
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
It is interesting to note that, despite the fact that PtÀNpy
bond dissociation has been found to be slower than the disso-
ciation of the equivalent PdÀNpy bond (see above), the effec-
tive PtÀNpy bond dissociation has to take place to explain the
formation of the species 1a1bL1c from the mixture of the two
symmetric macrocycles 1aL1c and 1bL1c. Obviously, because of
the relatively slow de-coordination of the PtÀNpy bonds, the
formation of mixed 1a1bL1c takes place most probably with-
out a full dissociation of the supramolecular assembly into
mononuclear blocks. Nevertheless, the process demands the
complete de-coordination of a {Pt(dppp)} corner and long
times are required for equilibration (Figure 4, bottom). The
final speciation in solution thermodynamically compensates
the loss of two PtÀNpy bonds in 1bL1c by the change of two
PdÀNpy bonds in 1aL1c to two new PtÀNpy bonds in 1a1bL1c
(Scheme 5, bottom).
NMR evidence for the exchange between metallamacro-
cycles
In addition to the mass spectrometric characterisation indicat-
ed above, supportive time-dependent NMR studies were em-
ployed to characterise the exchange processes between the
metallamacrocycles. To this end, a 31P{1H} NMR spectrum was
recorded immediately after mixing in [D6]acetone equimolar
quantities of 1aL1c and 1bL1c. The spectrum showed in addi-
tion to the signals of the two equivalent phosphorus atoms of
the 4-PPh2py ligands attached to the {Pd(h3-2-Me-C3H4)}+ unit
in either 1aL1c or 1bL1c, a coupled (2J(P,P)simulated =34 Hz) pair
of doublets (centred at dsimulated =24.4 and 24.9 ppm, respec-
tively) partially overlapped with the former signals. The ob-
served second-order NMR pattern (Figure 5) agrees well with
the lower symmetry expected for the Pd3Pt metallacycle
1a1bL1c, where the phosphorus signals of the ligand L1 in the
asymmetric assembly 1a1bL1c are not chemically equivalent.
As expected from the ESI data (Figure 4), the intensity of these
doublets increases with time, reaching equilibrium after ap-
proximately 48 h (Figure 5).
To prove that the mixed Pd3Pt metallamacrocycle (1a1bL1c)
was not a species that results from the assembly of different
fragments in the ESI source, the ESI-MS spectrum of an ace-
tone solution of the symmetrical compound 1bL1c with the
palladium corner [Pd(H2O)2(dppp)](CF3SO3)2 (1a) in a 1:2 molar
The equilibrated molar ratio
1aL1c:1bL1c:1a1bL1c is roughly
1:1:2, which corresponds to
a minimum of energy of the
1
system. Time-dependent H NMR
spectra showed exactly the
same behaviour. After 48 h three
very close (Dd<0.01 ppm) sin-
glets centred at d=1.96 ppm
were observed, that were as-
signed to the methyl signals of
the allyl moiety for 1aL1c, 1a1
bL1c and 1bL1c, respectively. The
other proton signals are not af-
fected by the loss of symmetry
on going from 1aL1c and 1bL1c
to 1a1bL1c and appear superim-
Scheme 5. Reaction scheme for the formation of the mixed 1a1bL1c and 1a1bL2c mixed species.
posed with that of the symmetri-
cal 1aL1c and 1bL1c metallamac-
rocycles. NMR spectra corre-
ratio was recorded. Only the peaks corresponding to the start-
ing materials were observed, thus indicating that no exchange
between the Pt and Pd corners in the metallamacrocycles
takes place in the mass spectrometer. Furthermore, the experi-
ment proves the enhanced thermodynamic stability of the PtÀ
Npy bonds, whose de-coordination cannot be energetically
compensated in this case.
The respective ESI mass studies on the composition of mix-
tures obtained from the fluorinated L2 derivatives, that is,
1aL2c and 1bL2c, produced equivalent results. The formation
of the mixed assembly 1a1bL2c was confirmed by a signal at
m/z=1845.1, assigned to the doubly charged [1a1bL2c-
2(CF3SO3)]2+, arising from the loss of two triflate anions from
the mixed macrocyclic assembly (Figure S11 in the Supporting
Figure 5. Time-dependent 31P{1H} NMR spectra of a 1:1 mixture of 1aL1c and
1bL1c in [D6]acetone. The represented region corresponds to the PPh2 sig-
nals of the ligand L1 in the indicated metallamacrocyclic compounds.
Information).
Chem. Eur. J. 2014, 20, 1 – 16
www.chemeurj.org
7
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
sponding to the final equilibrium speciation were also ob-
tained immediately after mixing 1a, 1b, c and Ln (or Lnc) build-
ing blocks in the appropriate molar ratio, indicating that the
attainment of the equilibrium in this case does not require the
dissociation of already formed PtÀNpy bonds and the most
thermodynamically favourable composition is immediately
reached (Scheme 5). Once the NMR spectra indicated that the
equilibrium was reached, the solutions were analysed by ESI
mass spectrometry and, although this system does not allow
ESI-MS quantification as indicated above, a clear predominance
of the mixed metallamacrocycle 1a1bL1c was observed
(Figure 4, bottom).
The NMR study of mixtures obtained from the fluorinated L2
derivatives did not prove to be definitive about the formation
of mixed Pd3Pt 1a1bL2c metallamacrocycles. In addition to the
expected singlets assigned to dppp in the metallacycles, the
31P{1H} NMR spectra showed
served in the 1H NMR spectra has been conducted for com-
pounds 1aL1c, 1bL1c, 2aL1c and 2bL1c in different solvents and
with different amounts of added triflate anions. The solvent
and the triflate concentration have been screened to address if
mild coordination media can be involved in the substitution
processes indicated before. Similar studies of these effects on
systems with much better donor anions have already been re-
ported.[9b,c] Table 1 collects the relevant kinetic and thermal ac-
tivation parameters for the systems studied in solution at dif-
ferent temperatures, and Figure 6 shows a typical rate simula-
tion by using the gNMR software[19] as well as their corre-
sponding Eyring plots for the determination of DH° and DS°.
The analogous fluorinated compounds 1aL2c and 1bL2c have
not been studied, being always in the fast exchange regime re-
gardless of the temperature and/or the solvent used.
a broad signal at d=9.3 ppm
Table 1. Kinetic and thermal activation parameters determined for the dynamic behaviour observed for the
metallamacrocyclic systems studied (Schemes 1 and 3) in different solution media and 2ꢄ10À3 m macrocycle
concentration.
that can be assigned to the
overlapping of the four different
L2 phosphorus resonances aris-
ing from 1aL2c, 1bL2c and
1a1bL2c in fast exchange (Fig-
ure S12 in the Supporting Infor-
mation), agreeing with the en-
hanced lability found for the L2
systems. It is interesting to note
that, in this case, the mass spec-
tra did not show dependence
with time, which indicates that
the equilibrium is reached imme-
diately after mixing, which is in
accordance with the enhanced
mobility observed for the fluori-
nated L2 system.
Compound
Solution medium
CDCl3
273k [sÀ1
]
DH° [kJmolÀ1
]
DS° [JKÀ1 molÀ1
]
1aL1c
7.3[a]
20[a]
7.6
1.4
1.4
(46Æ3)[a]
(26Æ1)[a]
(59Æ2)
(53Æ5)
(49Æ4
(À61Æ10)[a]
(À126Æ4)[a]
(À13Æ7)
CDCl3 +10-fold (NBu4)CF3SO3 added
CD3NO2
CDCl3
1bL1c
2aL1c
2bL1c
(À49Æ16)
(À64Æ16)
(À25Æ13)
(À123Æ10)
(À41Æ7)
CDCl3 +10-fold (NBu4)CF3SO3 added
CD3NO2
CDCl3
2.8
(58Æ4)
(19Æ3)
(41Æ2)
(21Æ2)
(22Æ2)
(33Æ3)
(28Æ2)
600
700
420
70
60
55
CDCl3 +10-fold (NBu4)CF3SO3 added
CD3NO2
(À126Æ7)
(À130Æ8)
(À91Æ9)
CDCl3
CDCl3 +10-fold (NBu4)CF3SO3 added
CD3NO2
(À110Æ9)
[a] Taken from reference [5].
Kinetic studies on the exchange dynamics
As indicated in the previous sections, the gross dynamics in so-
lution of the prepared metallamacrocycles is reflected by the
1H NMR signal behaviour of the a,a’ and b,b’ pyridine protons
of the L1 and L2 linkers. According to the X-ray structure of
1bL1c,[5] the four pyridine protons of these linkers should be
non-equivalent, each a and b proton being in a different rela-
tive position with respect to the methyl group of the allyl
ligand of the c unit (Figure 1).[5] These signals are usually rather
broad, when observed separately, and even coalesce to
a single set in several cases. This dynamic behaviour has al-
ready been observed in a wealth of situations, and involves
the macroscopic movement of the linkers on the metal centres
in solution.[17e,18] Although rotation of the MÀNpy bond can ac-
count for the dynamic behaviour on the NMR time scale,[9b,c,10]
this is not the case in the present situation and exchange pro-
cesses have been observed (see the previous sections) be-
tween assembled and free building blocks.
Figure 6. a) Line-shape analysis (red dashed line) of the resonances of the
1H NMR spectra (CDCl3, 500 MHz) of the a and a’ proton pairs in compounds
2aL1c (top) and 2bL1c (bottom). b) Eyring plots for the kinetic exchange pa-
rameters for the same compounds.
In view of the results obtained in the preliminary studies car-
ried out on 1aL1c, the kinetic monitoring of the dynamics ob-
&
&
Chem. Eur. J. 2014, 20, 1 – 16
www.chemeurj.org
8
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
The data in Table 1 correspond to the gross reversible disso-
ciation of 1a, 1b, 2a and 2b fragments from L1c blocks, as es-
tablished from the ESI-MS and VT-NMR experiments indicated
above and in our preliminary communication.[5] The process
necessarily has to involve the coordination of another Lewis
base present in the medium such as the solvent, the six triflate
anions already present in the species, or those specifically
added (10-fold, i.e., 60 anions per macrocycle). From the data
it is clear that at 273 K the overall exchange process measured
is faster for the palladium derivatives (5–10-fold); nevertheless,
the difference in the values does not agree with the expected
much larger lability of the PdÀN bond (105–106-fold) with re-
spect to the PtÀN analogue.[12b] Given the fact that the data
correspond to an effective intermolecular exchange of the MÀ
Npy bonds, it is clear that the already reported cooperative labi-
lisation in large self-assembled structures is of key importance
in the observed solution behaviour.[17e,20] As for the thermal ac-
tivation parameters obtained, all of them indicate a clear asso-
ciativeness of the process, especially for the PPh3 systems
(2aL1c and 2bL1c), were the activation entropies are more neg-
ative, whereas DH° has lower values. Probably, the much more
flexible environment around the metal centres in 2a and 2b,
as well as the worse donor character of the ligand, enable
a better transient association of an incoming ligand in the co-
ordination sphere of the metal.[12b] This seems, in fact, to be
the determinant factor that differentiates dppp from PPh3 mac-
rocyclic systems. In this respect, the fact that the L2 analogues
show a much faster dynamic exchange can also be associated
to the worse donor character of the fluorinated ligand.
take into account that ion pairing is an omnipresent feature in
these systems and its formation has already been held respon-
sible of simple first order kinetics for nitrate substitution-acti-
vated exchange processes,[9d] as well as in molecular recogni-
tion. In fact, there are examples where molecular recognition
has been found to be more important for palladium- than for
analogous platinum-containing architectures.[14d] Interestingly,
although the results for the dppp 1aL1c macrocycle indicate
that the gross exchange process increases its associative char-
acter (smaller DH° and more negative DS° values) on increas-
ing the concentration of triflate anions in the reaction
medium, the opposite is observed for the PPh3 2aL1c
analogue.
From the data shown in the previous sections, the only way
to come to terms with this difference demands that, depend-
ing on the metallamacrocycle, the rate-dominant process for
the exchange corresponds to different reactions. For 1aL1c (the
macrocycle with a more acidic and less hindered Pd centre)
the dominant reaction, on increasing the quantities of triflate
anions present, moves from the pyridine by triflate substitution
on the initial 1aL1 units (more dissociative) to the back coordi-
nation of a L1c ligand to the de-coordinated {(dppp)Pd-triflate}
fragments (more associative). This shift is not surprising in the
presence of large amounts of tributylammonium triflate. For
the macrocycle 2aL1c (with the less acidic and more hindered
Pd centre), large amounts of {(PPh3)2Pd-triflate} fragments are
expected even at low triflate concentrations, thus providing
a more associative character of the substitution even under
these conditions. As a result, the dominant operation of the
more dissociative substitution process on the {(PPh3)2Pd-Npy}
units at high triflate concentrations has to be related with the
sequestering of the above-mentioned {(PPh3)2Pd-triflate} frag-
ments by formation of dead-end outer-sphere complexes by
association to the increasing amounts of triflate anions.[22]
The values obtained for the same systems in nitromethane
solution, also collected in Table 1, do not show any significant
trend with respect to the data in chloroform, thus indicating
the relative innocence of both solvents in the process, despite
1
the small differences observed in the H NMR experiments. In
1
this respect, the fact that the H NMR spectra in [D6]acetone at
room temperature already show the coalescence of the pyri-
dine proton pairs is indicative of the much better donor char-
acter of this solvent (Gutmann’s DNN being 0.0 kcalmolÀ1 for
chloroform, 2.7 kcalmolÀ1 for nitromethane and 17 kcalmolÀ1
for acetone).[21]
Preliminary catalytic studies
Allyl PdII complexes stabilised with phosphine ligands are good
catalytic precursors for the catalysed allylic alkylation reac-
tion.[6] The activity and selectivity of the process depends on
several factors such as the structure of the allyl substrate, the
ligand or nucleophile and the nature of the solvent.[7] The re-
gioselectivity of the allylic alkylation of a monosubstituted al-
lylic substrate depends on the nature of the ligand used.
When a symmetrical diphosphine is used the substitution
occurs mainly at the terminal carbon atom leading to the
linear substituted product[23] but the use of sterically bulk
monophosphines or unsymmetrical ligands reverse the regio-
chemistry to the branched product.[23a,24] The metallamacrocy-
cles described in this paper represent promising accessible pal-
ladium allyl complexes with monodentate ligands, whereas
having properties associated to bidentate chelating systems. In
this respect, interesting reports dealing with the effect on cata-
lytic processes of metal complexes containing self-assembled
supramolecular bidentate ligands exist in the literature.[25]
In this line we have studied the effect on both the activity
and the regioselectivity of the allylic alkylation of the rac-3-ace-
With respect to the data obtained from the experiments car-
ried out in chloroform solution containing a 10-fold excess of
tetrabutylammonium triflate (60:1 triflate/macrocycle ratio),
they merit a special consideration. Any difference has to in-
volve a dominant interaction with the large amount of weakly
coordinating triflate anions present in the reaction medium. Al-
though for the platinum-containing systems 1bL1c and 2bL1c
no relevant changes are observed, for the palladium species
1aL1c and 2aL1c important differences are evident between
their thermal activation parameters at 10-fold excess of tetra-
butylammonium triflate (60:1 triflate/macrocycle ratio) and
those obtained under stoichiometric conditions (6:1 triflate/
macrocycle ratio) (ꢀ20 kJmolÀ1 for DH° and 70 JKÀ1 molÀ1 for
DS°, see Table 1). Clearly, the more labile and polarising char-
acter of the smaller palladium units in the macrocycle have to
be responsible for the difference between the observed behav-
iour of the platinum and palladium macrocycles. One should
Chem. Eur. J. 2014, 20, 1 – 16
www.chemeurj.org
9
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
toxy-1-phenyl-1-propene catalysed by the macrocycles 1aL1c
and 1bL1c. The reaction was performed by using the sodium
salt of dimethylmalonate as the nucleophile in CH2Cl2 at room
temperature and with 0.5 mol% of concentration level of cata-
lyst (Scheme 6). A blank study by using 1 mol% of the mono-
nuclear palladium corner L1c has also been conducted for com-
parative purposes; the results obtained are collected in
Table 2.
described by Hayashi et al.[23a] when the bidentate ligand dppe
was used instead of the monodentate PPh3 (89:11 ratio linear/
branched regioisomers).
Discussion
In view of the results collected so far, it is clear that all the
metallamacrocyclic assemblies prepared in this work are labile,
unlike other compounds found in the literature,[17e,26] despite
showing a high thermodynamic stability. Although for the Pd4
compounds, that is, 1aL1c, 2aL1c and 1aL2c, this is somehow
expected, for the Pd2Pt2 derivatives, that is, 1bL1c, 2bL1c and
1bL2c, a rather unusual intrinsic labilisation is observed, which
is reflected in the dynamic behaviour of different fragments at
room temperature. Taking into account that not even a salt-
mediated substitution acceleration is observed,[27] and that the
expected changeover from associative to dissociative activa-
tion for substitution reactions on PtII organometallic complexes
is not relevant in these complexes,[28] this is an extremely re-
markable feature. A similar effect has already been observed in
compounds containing good p-acceptor pyridine derivative li-
gands.[29] The presence of two phosphane donors in the coor-
dination sphere of the PtII centres probably creates a similar re-
inforced trans effect on the attached linkers L1 and L2, thus en-
abling its increased labile substitution behaviour.
Scheme 6. Reaction scheme of the studied catalytic reaction.
Table 2. Catalytic results of the allylic alkylation of rac-3-acetoxy-1-
phenyl-1-propene according to Scheme 6.[a]
Precursor
Conversion [%]
Branched/linear
L1c[b]
1aL1c[c]
1bL1c[c]
100
95
50
12:88
13:87
14:86
[a] Catalytic conditions: 1 mmol rac-3-acetoxy-1-phenyl-1-propene,
3 mmol NaCH(COOMe)2, 8 mL CH2Cl2 at RT for 1 h. Conversion and bran-
ched:linear ratio determined by GC. [b] 1ꢄ10À2 mmol catalytic precursor.
[c] 0.5ꢄ10À2 mmol catalytic precursor.
However, even with fairly uniform gross dynamic behaviour
of the pyridine a protons, the results reported indicate the ex-
istence of differences between the assembly/disassembly dy-
namics of the palladium and platinum metallamacrocyclic sys-
tems. Whereas NOESY NMR experiments conducted with the
palladium species 1aL1c indicate that the macrocyclic unit is in
effective exchange, at the magnetisation transfer time scale,
with free corner L1c added to the reaction medium, this is not
so for the platinum species 1bL1c. These differences necessi-
tate of a rather complex behaviour of the metallamacrocycles
in solution. The overall dynamics should distinguish between
the movement making the a pyridine protons of ligand L1
equivalent from that leading to the exchange of the L1c frag-
ment. These processes are collected as an example in
Scheme 7 for the L1 dppp derivatives; the first reaction at the
left side corresponds to the one related to the a pyridine pro-
tons exchange (Figure 1), whereas the one at the centre corre-
sponds to the exchange of the L1c and/or 1a or 1b fragments
(Figure 3).
The total conversion and a 12:88 branched/linear ratio of
the alkylation product was obtained after one hour reaction
when the mononuclear L1c corner was used as catalytic precur-
sor. Similar results, both in activity and regioselectivity, have
been reported[23a] for the analogous compound [Pd(h3-
C3H5)(PPh3)2]+ prepared in situ: 99% conversion with a 9:91
ratio at À208C after twelve hours of reaction in thf with
NaCMe(COOMe)2 as nucleophile. Clearly, the presence of a dis-
tant nitrogen donor in the phosphine ligands does not have
any relevant effect on either the regioselectivity or the activity
of the alkylation process. On moving to the metallamacrocy-
cles described in this work, some differences are observed be-
tween the compounds 1aL1c and 1bL1c. Whereas the palladi-
um metallamacrocycle 1aL1c shows a similar activity to that of
the corner species L1c, the heterometallic palladium–platinum
assembly, 1bL1c, displays a lower activity, which parallels the
observed lower exchange rate of the allylic corner L1c with the
metallamacrocycle 1bL1c, with respect to the same process on
1aL1c. It seems clear that the fragment L1c in the metallacycle
is less prone to a nucleophilic attack or/and an oxidative addi-
tion of the substrate than in the unassembled form. Neverthe-
less, complete conversion is achieved after six hours of reac-
tion with both macrocycles. As indicated in Table 2, the same
regioselectivity was obtained for the mononuclear L1c and
both metallamacrocycles 1aL1c and 1bL1c. In our macrocyclic
systems the more rigid environment of the allyl palladium frag-
ment {Pd(h3-2-Me-C3H4)} when assembled into the metallama-
crocycle does not have any effect on the regioselectivity of the
catalytic process. The same sort of behaviour has already been
For the all-palladium macrocycles both processes are within
the NMR time scale, whereas for the platinum compound
1bL1c the exchange with the L1c fragment is not observed on
this time scale. Contrarily, for the closely related platinum de-
rivative 1bL2c the magnetisation-transfer process with the L2c
corner is observed, agreeing with an easier substitution pro-
cess on the 1bL2 fragment containing the poorer donor, that
is, the fluorinated ligand L2. Within the same context, the palla-
dium fragments 1a were found to exchange faster than the
equivalent platinum units 1b, which indicated the greater labil-
ity of the former within the metallamacrocycle, which leads to
the thermodynamic preference of the Pt-containing molecules.
&
&
Chem. Eur. J. 2014, 20, 1 – 16
www.chemeurj.org
10
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
to reach equilibrium. In this re-
spect, for the experiments car-
ried out on the much more
labile systems 1aL2c and 1bL2c,
a shorter equilibration time is
observed, which is in agreement
with the more labile nature of
the PtÀNpy bond of the units
1bL2.
Interestingly, the triflate con-
centration effects seen in the re-
versible
opening
process
(Scheme 7, left) can also be ra-
tionalised. Although for the
opening of the platinum-con-
taining metallamacrocycles only
the dissociation of the initial PtÀ
Npy bond is relevant and no
changes in the associativity are
expected, for the palladium-con-
taining macrocycles, an increase
of the triflate concentration has
to increase necessarily the
amount of fully disassembled
units 1a or 2a, thus producing
the effects related to the increas-
ing amounts of either {(dppp)Pd-
triflate} or {(PPh3)2Pd-triflate}
fragments indicated before, that
is, an increased associativity/dis-
Scheme 7.
Summarising, in Scheme 7 labilisation of the PtÀNpy bond to crimination character of the exchange process.
a level similar to the PdÀNpy analogue is indicated on the mac-
The results collected for the catalytic process studied also
rocyclic units. The dynamic process thus observed for the a,a’ agree with the general dynamic speciation behaviour observed
pyridine protons corresponds to the formation of an open in ESI-MS, magnetisation-transfer NMR and time-dependent
form on the cycle, which enables a free rotation of the struc- NMR experiments. Clearly, the opening and disassembling of
tures making them equivalent. Once the macrocycle has been the metallamacrocycles represent the keystone for the data
opened by the dissociation of one of the MÀNpy bond, the la- collected in Table 2, where the results are compared with
bility of the MÀNpy bonds is reduced considerably and no ex- those obtained for the corner L1c under the same conditions.
change evidences can be inferred from line shape analysis When the PdII allyl fragment L1c is tightly anchored to the mac-
NMR experiments. In a moderately fast time scale, for the rocycle, the activity of the alkylation process decreases. As
dppp palladium (1a) and/or L2 systems, further dissociation of a consequence, the larger catalytic activity corresponds to
the MÀNpy bonds can take place. The latter produces a magnet- compound 1aL1c where the formation of the open form of the
isation exchange with the respective free L1c or L2c fragments, macrocycle is faster and its disassembly to individual pieces
and the effective exchange of the palladium blocks by the more favoured.
platinum analogues, producing thermodynamically stable
mixed Pd/Pt derivatives, is observed (Scheme 7, right). The
Conclusion
diss
diss
final outcome being the consequence of the
rate constant ratio.
k / k
PdÀNpy PtÀNpy
The supramolecular metallamacrocycles 1aL1c, 1bL1c, 2aL1c,
The ESI-MS experiments carried out fully agree with the data 2bL1c, 1aL2c, 1bL2c and 1a1bL1c can be obtained through self-
collected by magnetisation transfer and/or line shape analysis assembly reactions of different combinations of building
NMR experiments. The favourable formation of new more inert blocks. The obtained metallamacrocycles are stereochemically
PtÀNpy bonds is observed from mixtures of the metallamacro- non-rigid and their dynamic process depends on the solvent,
cycles 1aL1c and 1bL1c producing mixed compounds 1a1bL1c the basicity of the ligands and/or the size of the metallamacro-
in a relatively slow process originated by the total disassembly cycles. The kinetico-mechanistic studies conducted corroborate
of a platinum unit 1b from the open form of the macrocycle that the pyridine dissociation of the metal is needed to explain
1bL1c (Scheme 7, centre). NMR experiments have also indicat- the line shape analysis of the NMR spectra. NOE magnetisation
ed that the process is slow and needs approximately 48 hours transfer experiments indicate that further dissociation into the
Chem. Eur. J. 2014, 20, 1 – 16
www.chemeurj.org
11
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
column chromatography (SiO2, ethyl acetate). Conversion and the
regioisomeric ratio (linear/branched) were determined by GC on an
Agilent HP-5 column.
building blocks has to take place in order to explain the full
comprehensive mechanistic assembly/disassembly processes.
The results confirm the slower kinetics of PtÀNpy versus PdÀ
Npy bond, despite its increased lability in the fully assembled
metallamacrocycle. In all cases, the platinum-containing assem-
blies show an enhanced thermodynamic stability that accounts
for the assembly of mixed asymmetric architectures on mixing
symmetrical analogues.
Synthesis of 4-PPh2C6F4py (L2): An hexane solution of nBuLi (1.6m,
1 mL) was added dropwise to a pre-cooled (À788C) solution of
BrC6F4py (465 mg, 1.52 mmol) in thf (60 mL). After 1 h of stirring,
the mixture was allowed to warm slowly to À408C. Then, the solu-
tion was cooled again to À788C and a solution of PPh2Cl (0.29 mL,
1.52 mmol) in thf was added dropwise. The mixture was allowed
to warm slowly to room temperature and concentrated to dryness
to give a yellow oil that was treated with CH2Cl2. The obtained sus-
pension was filtered and concentrated to dryness. Re-crystallisation
from ethanol of the solid residue gave a white crystalline solid in
65% yield (406 mg). 1H NMR (298 K, CDCl3): d=8.76 (d, J(H,H)=
6.1 Hz, 2H; Ha-py), 7.51–7.46 (m, 4H; Phortho), 7.42–7.39 ppm (m, 8H;
The introduction of allyl-palladium moieties in metallamacro-
cycles did not improve the catalytic activity or regioselectivity
of the corresponding active monometallic complex in the allyl-
ic substitution reaction. However, as a future prospect, we en-
visage the introduction of chiral building blocks near the cata-
lytic centres in the supramolecular species in order to widen
the scope of the strategy and induce stereoselectivity in the
studied catalytic process.
H
b-py, Phmeta+para); 31P{1H} NMR (298 K, CDCl3): d=À24.6 ppm (t,
3J(P,F)=37 Hz); 19F NMR (298 K, CDCl3): d=À127.6 (m, 2F; Fx),
À142.4 ppm (m, 2F; FA); IR (KBr): n˜ =1458, 1430, 970 cmÀ1; ESI(+)
m/z: 412.1 [M+H]+; calcd: 412.1; elemental analysis calcd (%) for
C23H14F4NP: C 67.16, H 3.43, N 3.40; found: C 66.98, H 3.41, N 3.43.
Experimental Section
Synthesis of [Pd(h3-2-Me-C3H4)(4-PPh2py)2](CF3SO3) (L1c): Solid
[Pd(h3-2-Me-C3H4)(cod](CF3SO3) (250 mg, 0.59 mmol) was added to
a solution of 4-PPh2py (314 mg, 1.19 mmol) in dichloromethane
(40 mL) at room temperature. After 15 min of stirring, the reaction
mixture was filtered, concentrated to 10 mL under vacuum and
precipitated with diethyl ether. A pale grey solid was obtained in
75% yield (375 mg). 1H NMR (298 K, CDCl3): d=8.41 (d, J(H,H)=
2.4 Hz, 4H; Ha-py), 7.52–7.37 (m, 20H; Ph), 6.99 (d, J(H,H)=2.06 Hz,
4H; Hb-py), 4.06 (s, 2H; Me-C3H4), 3.72 (s, 2H; Me-C3H4), 1.85 ppm (s,
3H; Me-C3H4); 1H NMR (298 K, [D6]acetone): d=8.49 (d, J(H,H)=
2.4 Hz, 4H; Ha-py), 7.62–7.47 (m, 20H; Ph), 7.12 (d, J(H,H)=5.2 Hz,
4H; Hb-py), 3.99 (s, 2H; Me-C3H4), 3.83 (t, J(H,P)=5.2 Hz, 2H; Me-
C3H4), 2.06 ppm (s, 3H; Me-C3H4); 31P{1H} NMR (298 K, CDCl3): d=
21.7 ppm (s); IR (KBr): n˜ =1615, 1480, 1438, 1262, 1150, 1110,
1031 cmÀ1; ESI(+): m/z: 687.1 [MÀCF3SO3]+; calcd: 687.1; elemental
analysis calcd (%) for C39H35F3N2O3P2PdS: C 55.96, H 4.21, N 3.35;
found: C 55.91, H 4.27, N 3.30.
All manipulations were performed under pre-purified N2 by using
standard Schlenk techniques. Infrared spectra were recorded on an
FT-IR 5700 Nicolet spectrophotometer. Solvents were dried by stan-
dard methods and distilled under argon or nitrogen immediately
prior to use, or alternatively from a solvent purification system (In-
novative Technologies). 31P{1H} NMR (d(85% H3PO4)=0.0 ppm),
1H NMR (d(TMS)=0.0 ppm) and 19F NMR (d(CFCl3)=0.0 ppm) spec-
tra were obtained on a Bruker DXR 250, a Varian-Inova 300,
a Varian Mercury 400, a Varian-Inova 500 or a Bruker DMX 500
spectrometer. ESI mass spectra were recorded on a Bruker APEX IV
Fourier-transform ion cyclotron resonance mass spectrometer with
a 7.05 T magnet and an Apollo electrospray (ESI) ion source
equipped or on a LC/MSD-TOF (2006) (Agilent Technologies)
(G1969A). Typically, acetone solutions of the metallacycles (0.1–
0.2 mm) were used with flow rates of 3–4 or 10 mLminÀ1 depend-
ing on the spectrometer used. Elemental analyses of C, H, N and S
were carried out at the Institut de Bio-Orgꢅnica in Barcelona. Line
shape analyses of exchanging signals were performed by using the
gNMR software package.[19] All the rate constants obtained for the
different systems under various conditions are collected in Table S1
in the Supporting Information.
Synthesis of [Pd(h3-2-Me-C3H4)(4-PPh2C6F4py)2](CF3SO3) (L2c):
Solid [Pd(h3-2-Me-C3H4)(cod)](CF3SO3) (50 mg, 0.12 mmol) was
added to a solution of 4-PPh2C6F4py (100 mg, 0.24 mmol) in di-
chloromethane (20 mL) at room temperature. After 15 min of stir-
ring, the reaction mixture was filtered, concentrated to 5 mL under
vacuum and precipitated with diethyl ether. A pale yellow solid
The compounds 4-PPh2py,[30] 4-BrC6F4py,[11] [M(H2O)2(dppp)]-
(CF3SO3)2 (M=Pd, Pt),[31] [MCl2(PPh3)2] (M=Pd, Pt),[32] [Pd(H2O)2-
(PPh3)2](CF3SO3)2,[33] [Pd(m-Cl)(h3-2-Me-C3H4)]2,[34] [Pd(h3-2-Me-C3H4)-
(cod)](CF3SO3),[35] and [{Pd(h3-2-Me-C3H4)(PPh2py)2}2{M(dppp)}2]-
(CF3SO3)6 [M=Pd (1aL1), Pt (1bL1)],[5] were prepared as described
previously. All other chemicals were commercial grade and were
used without further purification.
1
was obtained in 60% yield (81 mg). H NMR (298 K, CDCl3): d=8.71
(sbr, 4H; Ha-py), 7.48–7.42 (m, 32H; Ph, Hb-py), 4.06 (sbr, 2H; Me-
C3H4), 3.69 (sbr, 2H; Me-C3H4), 1.89 ppm (s, 3H; Me-C3H4);
31P{1H} NMR (298 K, CDCl3): d=6.3 ppm (s); 19F NMR (298 K, CDCl3):
d=À78.3 (s, 6F; CF3SO3), À125.0– À125.2 (m, 8F; Fx), À139.0–
À139.5 ppm (m, 8F; FA); IR (KBr): n˜ =1615, 1464, 1437, 1263, 1154,
1031 cmÀ1; ESI(+): m/z: 983.1 [MÀCF3SO3]+; calcd: 983.1; elemental
analysis calcd. (%) for C51H35F11N2O3P2PdS: C 54.05, H 3.11, N 2.47;
found: C 53.91, H 3.17, N 2.37.
General procedure for Pd-catalysed allylic alkylation of rac-3-
acetoxy-1-phenyl-1-propene: Reactions were carried out in
a Schlenk tube under N2 at room temperature. The PdII precursor
(0.01 mmol for the monomer complex L1c and 0.005 mmol for met-
allamacrocycles 1aL1 or 1bL1) was dissolved in CH2Cl2 (1 mL) and
rac-3-acetoxy-1-phenyl-1-propene (1 mmol) in CH2Cl2 (1 mL) was
added. After 15 min. stirring Na(CH(COOMe)2) (3 mmol) in CH2Cl2
(6 mL) was added. The mixture was stirred at room temperature
for the desired time (1 or 6 h). Then, the solution was diluted with
diethyl ether (20 mL) and washed with saturated ammonium chlo-
ride solution (3ꢄ10 mL) and water (3ꢄ10 mL). The organic phase
was dried over anhydrous Na2SO4 and filtered off, and the solvent
was removed under reduced pressure. Conversion was determined
Synthesis of [Pd(dppp)(4-PPh2py)2](CF3SO3)2 (1aL1): Solid 4-
PPh2py (30 mg, 0.11 mmol) was added to a solution of [Pd(H2O)2-
(dppp)](CF3SO3)2 (1a) (43 mg, 0.05 mmol) in dichloromethane
(15 mL) at room temperature. After 30 min of stirring, the reaction
mixture was concentrated to 5 mL under vacuum and precipitated
with diethyl ether. A creamy solid was obtained in 80% yield
1
(59 mg). H NMR (298 K, CDCl3): d=8.76 (d, J(H,H)=4.7 Hz, 4H; Ha-
py), 7.61 (sbr, 8H; Phdppp), 7.44–7.34 (m, 24H; Phdppp, Ph4-PPh py), 7.21–
2
7.17 (m, 8H; Ph4-PPh py), 6.77 (t, J(H,H)=4.9 Hz, 4H; Hb-py), 3.14 (sbr,
2
4H; P-CH2-), 2.30–2.17 ppm (m, 2H; C-CH2-C); 31P{1H} NMR (298 K,
1
by H NMR spectroscopy. Purification of the product was done by
CDCl3): d=5.4 (s, dppp), À4.4 ppm (s, 4-PPh2py); IR (KBr): n˜ =1482,
&
&
Chem. Eur. J. 2014, 20, 1 – 16
www.chemeurj.org
12
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
1436, 1253, 1157, 1100, 1029 cmÀ1
;
ESI(+): m/z: 522.1
6.80 (m, 48H; Ph, Hb-py), 3.50 (s, 4H; Me-C3H4), 2.23 (s, 4H; Me-
C3H4), 1.78 ppm (s, 6H; Me-C3H4); 31P{1H} NMR (298 K, [D6]acetone):
d=26.6 (s, PPh2py), 0.6 ppm (s, 1J(Pt,P)=3323 Hz, PPh3); IR (KBr):
n˜ =1616, 1470, 1438, 1258, 1157, 1031 cmÀ1; ESI(+): m/z : 2324.3
[MÀ2(CF3SO3)]2+; calcd: 522.1; elemental analysis calcd (%) for
C63H54F6N2O6P4PdS2: C 56.32, H 4.05, N 2.08. Found: C 56.46, H 4.07,
N 2.12.
[(2bL1)2À3(CF3SO3)]3+
;
calcd: 2324.3; 2280.2 [PtPd2(h3-2-Me-
calcd: 2279.2; 1706.2
Synthesis of [Pt(dppp)(4-PPh2py)2](CF3SO3)2 (1bL1): Solid 4-PPh2py
(30 mg, 0.11 mmol) was added to a solution of [Pt(H2O)2(dppp)]-
(CF3SO3)2 (1b) (47 mg, 0.05 mmol) in dichloromethane (15 mL) at
room temperature. After 30 min of stirring, the reaction mixture
was concentrated to 5 mL under vacuum and precipitated with di-
ethyl ether. A white solid was obtained in 85% yield (67 mg).
1H NMR (298 K, CDCl3): d=8.79 (d, J(H,H)=5.9 Hz, 4H; Ha-py), 7.63
C3H4)2(PPh3)2(4-PPh2py)3(CF3SO3)3]+;
[2bL1cÀ2(CF3SO3)]2+
,
[PtPd(h3-2-Me-C3H4)(PPh3)2(4-PPh2py)2-
(CF3SO3)2]+; calcd: 1705.7; 1442.1 [PtPd(h3-2-Me-C3H4)(PPh3)2(4-
PPh2py)(CF3SO3)2]+; calcd: 1442.1; 1131.2 [Pt(PPh3)2(4-PPh2py)-
(CF3SO3)]+; calcd: 1131.2; 1087.5 [2bL1cÀ3(CF3SO3)]3+
; calcd:
1087.5); 1065.1 [PtPd2(h3-2-Me-C3H4)2(PPh3)2(4-PPh2py)3(CF3SO3)2]2+
calcd: 1065.1; 687.1 [Pd (h3-2-Me-C3H4)(4-PPh2py)2]+; calcd: 687.1.
;
(sbr, 8H; Phdppp), 7.45–7.35 (m, 24H; Phdppp, Ph4-PPh py), 7.22–7.18 (m,
2
8H; Ph4-PPh py), 6.80 (t, J(H,H)=5.4 Hz, 4H; Hb-py), 3.24 (sbr, 4H; P-
2
CH2-), 2.27–2.17 ppm (m, 2 H; C-CH2-C); 31P{1H} NMR (298 K, CDCl3):
1
Self-assembly of [{Pd(h3-2-Me-C3H4)(4-PPh2C6F4py)2}2{Pd(dppp)}2]-
(CF3SO3)6 (1aL2c): A solution of [Pd(h3-2-Me-C3H4)(cod)](CF3SO3)
(10 mg, 0.02 mmol) in dichloromethane (5 mL) was added to a solu-
tion of [Pd(H2O)2(dppp)](CF3SO3)2 (1a) (20 mg, 0.02 mmol) in di-
chloromethane (5 mL) at room temperature and then a solution of
4-PPh2C6F4py (20 mg, 0.05 mmol) in dichloromethane (5 mL) was
added to this mixture. After 30 min. of stirring, the reaction mixture
was filtered, concentrated to 5 mL under vacuum and precipitated
with diethyl ether. A yellow solid was obtained in 75% yield
(36 mg). 1H NMR (298 K, [D6]acetone): d=9.23 (d, J(H,H)=5.0 Hz,
8H; Ha-py), 7.85–7.33 (m, 88H; Hb-py, Ph), 4.17 (d, J(H,P)=7.5 Hz, 4H;
Me-C3H4), 3.73–3.68 (m, 4H; Me-C3H4), 3.43 (sbr, 8H; P-CH2-), 2.53–
2.33 (m, 4H; C-CH2-C), 1.94 (s, 3H; Me-C3H4), 1.92 ppm (s, 3H; Me-
C3H4); 31P{1H} NMR (298 K, [D6]acetone): d=9.2 (sbr, 4-PPh2C6F4py),
8.0 ppm (s, dppp); 19F NMR (298 K, [D6]acetone): d=À78.1 (s, 6F;
CF3SO3), À124.5 (sbr, 8F; Fx), À139.1– À139.2 ppm (m, 8F; FA); IR
(KBr): n˜ =1616, 1464, 1437, 1259, 1158, 1030 cmÀ1; ESI(+): m/z=
1801.1 [(1aL2cÀ2(CF3SO3)]2+; calcd: 1801.1; 1295.0 [Pd2(h3-2-Me-
C3H4)2(4-PPh2C6F4py)2(CF3SO3)]+; calcd: 1295.0; 1234.5 [Pd3(h3-2-Me-
d=À3.8 (s, 4-PPh2py), À16.0 (s, J(Pt,P)=3051 Hz, dppp); IR (KBr):
n˜ =1481, 1438, 1256, 1153, 1100, 1031 cmÀ1; ESI(+): m/z : 1282.2
[MÀCF3SO3]+; calcd: 1283.2; 566.6 [MÀ2(CF3SO3)]2+; calcd: 566.6;
elemental analysis calcd (%) for C63H54F6N2O6P4PtS2: C 52.83, H 3.80,
N 1.96; found: C 52.96, H 3.78, N 1.95.
Self-assembly of [{Pd(h3-2-Me-C3H4)(4-PPh2py)2}2{Pd(PPh3)2}2]-
(CF3SO3)6 (2aL1c)
Method 1: A solution of [Pd(h3-2-Me-C3H4)(cod)](CF3SO3) (25 mg,
0.06 mmol) and 4-PPh2py (31 mg, 0.12 mmol) in dichloromethane
(10 mL) was added to a solution of [Pd(H2O)2(PPh3)2](CF3SO3)2 (2a)
(58 mg, 0.06 mmol) in dichloromethane (5 mL) at room tempera-
ture. After 30 min of stirring, the reaction mixture was filtered, con-
centrated to 5 mL under vacuum and precipitated with diethyl
ether. A white solid was obtained in 75% yield (79 mg).
Method 2: A solution of [Pd(h3-2-Me-C3H4)(4-PPh2py)2](CF3SO3) (L1c)
(50 mg, 0.06 mmol) in dichloromethane (10 mL) was added drop-
wise to
a solution of [Pd(H2O)2(PPh3)2](CF3SO3)2 (2a) (58 mg,
0.06 mmol) in dichloromethane (5 mL) at room temperature. After
30 min of stirring the solution was concentrated to 5 mL under
vacuum and precipitated with diethyl ether. A white solid was ob-
tained in 80% yield (85 mg).
C3H4)2(dppp)2(4-PPh2C6F4py)2(CF3SO3)3]2+
; calcd: 1234.0); 1187.0
[Pd3(h3-2-Me-C3H4)2(dppp)(4-PPh2C6F4py)3(CF3SO3)2]2+: calcd: 1186.6;
1151.1 [1aL2cÀ3(CF3SO3)]3+
;
calcd: 1151.1; 983.1 [Pd(h3-2-Me-
C3H4)(4-PPh2C6F4py)2]+; calcd: 983.1; 667.0 [Pd(dppp)(CF3SO3)]+;
1H NMR (298 K, [D6]acetone): d=9.14 (d, J(H,H)=8.0 Hz, 8H; Ha-py),
7.76–6.98 (m, 108H; Ph, Hb-py), 3.87 (s, 4H; Me-C3H4), 2.29–2.26 (m,
4H; Me-C3H4), 1.94 ppm (s, 6H; Me-C3H4); 31P{1H} NMR (298 K,
[D6]acetone): d=26.6 (s, PPh2py), 26.3 ppm (s, PPh3); IR (KBr): n˜ =
1616, 1480, 1438, 1256, 1153, 1030 cmÀ1; ESI(+): m/z : 2192.2
[Pd3(h3-2-Me-C3H4)2(PPh3)2(4-PPh2py)3(CF3SO3)3]+; calcd: 2191.1);
1617.2 [2aL1cÀ2(CF3SO3)]2+, [Pd2(h3-2-Me-C3H4)(PPh3)2(4-PPh2py)2-
(CF3SO3)2]+; calcd: 1617.1; 1354.1 [Pd2(h3-2-Me-C3H4)(PPh3)2(4-
calcd: 667.0; 572.5 [Pd2(h3-2-Me-C3H4)2(4-PPh2C6F4py)2]2+
; calcd:
573.0.
Self-assembly of [{Pd(h3-2-Me-C3H4)(4-PPh2C6F4py)2}2{Pt(dppp)}2]-
(CF3SO3)6 (1bL2c): A solution of [Pd(h3-2-Me-C3H4)(cod)](CF3SO3)
(10 mg, 0.02 mmol) in dichloromethane (5 mL) was added to a solu-
tion of [Pt(H2O)2(dppp)](CF3SO3)2 (1b) (23 mg, 0.02 mmol) in di-
chloromethane (5 mL) at room temperature and then a solution of
4-Ph2C6F4py (20 mg, 0.05 mmol) in dichloromethane (5 mL) was
added to this mixture. After 30 min. of stirring, the reaction mixture
was filtered, concentrated to 5 mL under vacuum and precipitated
with diethyl ether. A yellow solid was obtained in 70% yield
(34 mg). 1H NMR (298 K, [D6]acetone): d=9.26 (d, J(H-H)=6.0 Hz,
8H; Ha-py), 7.89–7.36 (m, 88H, Hb-py, Ph), 4.18 (d, J(H,P)=10.0 Hz,
4H; Me-C3H4), 3.73–3.68 (m, 4H; Me-C3H4), 3.52 (sbr, 8H; P-CH2),
2.50–2.26 (m, 4H; C-CH2-C), 1.94 (s, 3H; Me-C3H4), 1.92 ppm (s, 3H;
Me-C3H4); 31P{1H} NMR (298 K, [D6]acetone: d=9.4 (sbr, 4-
PPh2C6F4py), À14.2 ppm (s, 1J(P,Pt)=3055 Hz, dppp); 19F NMR
(298 K, [D6]acetone): d=À78.7 (s, 6F; CF3SO3), À126.7 (sbr, 8F; Fx),
À141.0– À141.1 ppm (m, 8F; FA); IR (KBr): n˜ =1617, 1468, 1437,
PPh2py)(CF3SO3)2]+;
calcd:
1354.0;
1198.6
[Pd3(h3-2-Me-
C3H4)(PPh3)4(4-PPh2py)2(CF3SO3)3]2+; calcd: 1198.6; 779.1 [Pd(PPh3)2-
(CF3SO3)]+; calcd: 779.0) 685.2 [Pd(h3-2-Me-C3H4)(PPh3)2]+, [Pd(h3-2-
Me-C3H4)(PPh3)(4-PPh2py)]+, [Pd(h3-2-Me-C3H4)(4-PPh2py)2]+; calcd:
685.1, 686.1, 687.1.
Self-assembly of [{Pd(h3-2-Me-C3H4)(4-PPh2py)2}2{Pt(PPh3)2}2]-
(CF3SO3)6 (2bL1c):
A
solution of [Pd(h3-2-Me-C3H4)(4-PPh2py)2]-
(CF3SO3) (L1c) (50 mg, 0.06 mmol) in dichloromethane (15 mL) was
added to a solution of [PtCl2(PPh3)2] (47 mg, 0.06 mmol) in di-
chloromethane (15 mL) at room temperature. After a few minutes
of stirring AgCF3SO3 (31 mg, 0.12 mmol) was added and the reac-
tion mixture was stirred for two hours at room temperature light
protected. Then the solution was filtered through celite, concen-
trated to 5 mL under vacuum and precipitated with diethyl ether.
A yellowish solid was obtained in 86% yield (95 mg). 1H NMR
(298 K, [D6]acetone): d=9.18 (d, J(H,H)=6.0 Hz, 8H; Ha-py), 7.77–
7.05 (m, 108H; Ph, Hb-py), 3.93 (s, 4H; Me-C3H4), 2.29–2.26 (m, 4H;
Me-C3H4), 1.95 ppm (s, 6H; Me-C3H4); 1H NMR (220 K, CDCl3): d=
9.31 (s, 4H; Ha-py), 8.76 (s, 4H; Ha-py), 7.68–7.29 (m, 60H; Ph), 7.19–
1259, 1158, 1031 cmÀ1; ESI(+): m/z: 1890.1 [1bL2cÀ2(CF3SO3)]2+
;
calcd: 1890.1; 1295.0 [Pd2(h3-2-Me-C3H4)2(4-PPh2C6F4py)2(CF3SO3)]+;
calcd: 1295.0; 1210.1 [1bL2cÀ3(CF3SO3)]3+; calcd: 1210.1; 870.1
[1bL2cÀ4(CF3SO3)]4+
;
calcd: 870.3; 573.0 [Pd2(h3-2-Me-C3H4)2(4-
PPh2C6F4py)2]+; calcd: 573.0.
Chem. Eur. J. 2014, 20, 1 – 16
www.chemeurj.org
13
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
[12] a) B. Brusilowskij, E. V. Dzyuba, R. W. Troff, C. A. Schalley, Dalton Trans.
2011, 40, 12089–12096; b) M. L. Tobe, J. Burgess, Inorganic Reaction
Mechanisms, Longman, New York, 1999.
[13] a) C. A. Schalley, T. Muller, P. Linnartz, M. Witt, M. Schafer, A. Lutzen,
Chem. Eur. J. 2002, 8, 3538–3551; b) M. Ferrer, A. Gutierrez, L. Rodri-
guez, O. Rossell, E. Ruiz, M. Engeser, Y. Lorenz, R. Schilling, P. Gomez-Sal,
A. Martin, Organometallics 2012, 31, 1533–1545.
[14] a) K. S. Jeong, S. Y. Kim, U. S. Shin, M. Kogej, N. T. M. Hai, P. Broekmann,
N. Jeong, B. Kirchner, M. Reiher, C. A. Schalley, J. Am. Chem. Soc. 2005,
127, 17672–17685; b) M. Ferrer, A. Pedrosa, L. Rodriguez, O. Rossell, M.
Vilaseca, Inorg. Chem. 2010, 49, 9438–9449; c) T. Weilandt, N. L. Low, G.
Schnakenburg, J. Daniels, M. Nieger, C. A. Schalley, A. Lutzen, Chem. Eur.
J. 2012, 18, 16665–16676; d) M. Ferrer, A. Gutierrez, M. Mounir, O. Ros-
sell, E. Ruiz, A. Rang, M. Engeser, Inorg. Chem. 2007, 46, 3395–3406.
[15] Alternatively, the formation of the metallamacrocycles can also be ach-
ieved in good yields from the self-assembly of the building blocks indi-
cated in Scheme 3, bottom.
Acknowledgments
Financial support for this work was provided by the Ministerio
de Economꢁa y Competitividad (MINECO/FEDER) of Spain (Proj-
ects CTQ2012-31335, CTQ2012-37821-C02-01 and CT2010-
15292/BQU), and of the Deutsche Forschungsgemeinschaft
(SFB 624).
Keywords: allyl ligands · homogeneous catalysis · kinetico-
mechanistic studies · metallamacrocycles · self-assembly
[16] a) C. A. Schalley, Mass Spectrom. Rev. 2001, 20, 253–309; b) B. Baytekin,
H. T. Baytekin, C. A. Schalley, Org. Biomol. Chem. 2006, 4, 2825–2841;
c) M. Kogej, C. A. Schalley, Analytical Methods in Supramolecular Chemis-
try (Ed.: C. A. Schalley), Wiley-VCH, Weinheim, 2007; d) C. A. Schalley, A.
Springer, Mass Spectrometry and Gas Phase Chemistry of Non-Covalent
Complexes, Wiley, New York, 2009; e) L. Cera, C. A. Schalley, Chem. Soc.
Rev. 2014, 43, 1800–1812.
[17] a) C. G. Claessens, M. J. Vicente-Arana, T. Torres, Chem. Commun. 2008,
6378–6380; b) V. M. Cangelosi, T. G. Carter, L. N. Zakharov, D. W. John-
son, Chem. Commun. 2009, 5606–5608; c) Y. R. Zheng, H. B. Yang, K.
Ghosh, L. Zhao, P. J. Stang, Chem. Eur. J. 2009, 15, 7203–7214; d) J. S.
Mathieson, G. J. T. Cooper, A. L. Pickering, M. Keller, D. L. Long, G. N.
Newton, L. Cronin, Chem. Asian J. 2009, 4, 681–687; e) S. Sato, Y. Ishido,
M. Fujita, J. Am. Chem. Soc. 2009, 131, 6064–6065; f) B. R. Hall, L. E.
Manck, I. S. Tidmarsh, A. Stephenson, B. F. Taylor, E. J. Blaikie, D. A. Van-
der Griend, M. D. Ward, Dalton Trans. 2011, 40, 12132–12145; g) W.
Jiang, A. Schafer, P. C. Mohr, C. A. Schalley, J. Am. Chem. Soc. 2010, 132,
2309–2320.
[1] a) G. F. Swiegers, T. J. Malefetse, Chem. Rev. 2000, 100, 3483–3537; b) L.
Cronin, Annu. Rep. Prog. Chem., Sect. A: Inorg. Chem. 2004, 100, 323–
383; c) C. L. Chen, J. Y. Zhang, C. Y. Su, Eur. J. Inorg. Chem. 2007, 2997–
3010; d) E. C. Constable, Coord. Chem. Rev. 2008, 252, 842–855; e) M. W.
Cooke, D. Chartrand, G. S. Hanan, Coord. Chem. Rev. 2008, 252, 903–
921; f) A. Kumar, S. S. Sun, A. J. Lees, Coord. Chem. Rev. 2008, 252, 922–
939; g) B. H. Northrop, H. B. Yang, P. J. Stang, Chem. Commun. 2008,
5896–5908; h) E. Zangrando, M. Casanova, E. Alessio, Chem. Rev. 2008,
108, 4979–5013; i) B. H. Northrop, Y. R. Zheng, K. W. Chi, P. J. Stang, Acc.
Chem. Res. 2009, 42, 1554–1563; j) R. Chakrabarty, P. S. Mukherjee, P. J.
Stang, Chem. Rev. 2011, 111, 6810–6918; k) H. Amouri, C. Desmarets, J.
Moussa, Chem. Rev. 2012, 112, 2015–2041; l) T. R. Cook, Y. R. Zheng, P. J.
Stang, Chem. Rev. 2013, 113, 734–777.
[2] a) C. A. Schalley, A. Lutzen, M. Albrecht, Chem. Eur. J. 2004, 10, 1072–
1080; b) F. Wurthner, C. C. You, C. R. Saha-Moller, Chem. Soc. Rev. 2004,
33, 133–146; c) C. H. M. Amijs, G. P. M. van Klink, G. van Koten, Dalton
Trans. 2006, 308–327; d) S. J. Lee, W. B. Lin, Acc. Chem. Res. 2008, 41,
521–537; e) J. W. Steed, Chem. Soc. Rev. 2009, 38, 506–519.
[18] a) D. Fujita, A. Takahashi, S. Sato, M. Fujita, J. Am. Chem. Soc. 2011, 133,
13317–13319; b) S. J. Li, J. Y. Huang, T. R. Cook, J. B. Pollock, H. Kim,
K. W. Chi, P. J. Stang, J. Am. Chem. Soc. 2013, 135, 2084–2087.
[19] gNMR version 5.0, P. H. H. Budzelaar, Cherwell Scientific, 1995.
[20] a) E. S. Barrett, T. J. Dale, J. Rebek, J. Am. Chem. Soc. 2007, 129, 8818–
8824; b) A. V. Davis, D. Fiedler, M. Ziegler, A. Terpin, K. N. Raymond, J.
Am. Chem. Soc. 2007, 129, 15354–15363.
[21] C. Reichardt, Solvents and Solvent Effects in Organic Chemistry, 3edth ed-
Wiley-VCH, Weinheim, 2003.
[22] A. G. Lappin, Redox Mechanisms in Inorganic Chemistry, Elis Horwood,
New York, 1994.
[23] a) T. Hayashi, M. Kawatsura, Y. Uozumi, J. Am. Chem. Soc. 1998, 120,
1681–1687; b) R. J. van Haaren, P. H. Keeven, L. A. van der Veen, K. Gou-
bitz, G. P. F. van Strijdonck, H. Oevering, J. N. H. Reek, P. C. J. Kamer,
P. W. N. M. van Leeuwen, Inorg. Chim. Acta 2002, 327, 108–115.
[24] a) R. J. van Haaren, C. J. M. Druijven, G. P. F. van Strijdonck, H. Oevering,
J. N. H. Reek, P. C. J. Kamer, P. W. N. M. van Leeuwen, J. Chem. Soc.
Dalton Trans. 2000, 1549–1554; b) M. D. K. Boele, P. C. J. Kamer, M. Lutz,
A. L. Spek, J. G. de Vries, P. W. N. M. van Leeuwen, G. P. E. van Strijdonck,
Chem. Eur. J. 2004, 10, 6232–6246; c) S. C. Milheiro, J. W. Faller, J. Orga-
nomet. Chem. 2011, 696, 879–886.
[25] a) V. F. Slagt, M. Roder, P. C. J. Kamer, P. W. N. M. van Leeuwen, J. N. H.
Reek, J. Am. Chem. Soc. 2004, 126, 4056–4057; b) V. F. Slagt, P. W. N. M.
van Leeuwen, J. N. H. Reek, Dalton Trans. 2007, 2302–2310; c) P. E. Gou-
driaan, X. B. Jang, M. Kuil, R. Lemmens, P. W. N. M. Van Leeuwen, J. N. H.
Reek, Eur. J. Org. Chem. 2008, 6079–6092.
[3] a) K. Takaoka, M. Kawano, T. Ozeki, M. Fujita, Chem. Commun. 2006,
1625–1627; b) M. Yoshizawa, J. K. Klosterman, M. Fujita, Angew. Chem.
2009, 121, 3470–3490; Angew. Chem. Int. Ed. 2009, 48, 3418–3438;
c) J. L. Hou, D. Ajami, J. Rebek, J. Am. Chem. Soc. 2008, 130, 7810–7811;
d) M. D. Pluth, R. G. Bergman, K. N. Raymond, Acc. Chem. Res. 2009, 42,
1650–1659; e) D. Fiedler, D. H. Leung, R. G. Bergman, K. N. Raymond,
Acc. Chem. Res. 2005, 38, 349–358; f) T. H. Noh, E. Heo, K. H. Park, O. S.
Jung, J. Am. Chem. Soc. 2011, 133, 1236–1239.
[4] a) H. Jiang, A. Hu, W. Lin, Chem. Commun. 2003, 96–97; b) H. Wang, R.
Zhong, X. Q. Guo, X. Y. Feng, X. F. Hou, Eur. J. Inorg. Chem. 2010, 174–
178; c) J. Park, S. Hong, Chem. Soc. Rev. 2012, 41, 6931–6943.
[5] I. Angurell, M. Ferrer, A. Gutierrez, M. Martinez, L. Rodriguez, O. Rossell,
M. Engeser, Chem. Eur. J. 2010, 16, 13960–13964.
[6] H. Tsuji, Palladium Reagents and Catalysis, Wiley, New York, 2004.
[7] a) B. M. Trost, D. L. Van Vranken, Chem. Rev. 1996, 96, 395–422; b) B. M.
Trost, C. Lee in Catalytic Asymmetric Synthesis (Ed.: I. Ojima), Wiley-VCH,
Weinheim, 2000; c) L. Milhan, P. J. Guiry in Transition-Metal-Catalysed
Enantioselective Allylic Substitution in Organic Synthesis (Ed.: U. Kazmai-
er), Springer, Heidelberg, 2012.
[8] a) T. V. Rajan Babu, Chem. Rev. 2003, 103, 2845–2860; b) R. M. Ceder, A.
Grabulosa, G. Muller, M. Rocamora, Catal. Sci. Technol. 2013, 3, 1446–
1464.
[9] a) S. J. Park, D. M. Shin, S. Sakamoto, K. Yamaguchi, Y. K. Chung, M. S.
Lah, J. Hong, Chem. Eur. J. 2005, 11, 235–241; b) D. C. Caskey, T. Yama-
moto, C. Addicott, R. K. Shoemaker, J. Vacek, A. M. Hawkridge, D. C.
Muddiman, G. S. Kottas, J. Michl, P. J. Stang, J. Am. Chem. Soc. 2008,
130, 7620–7628; c) J. Vacek, D. C. Caskey, D. Horinek, R. K. Shoemaker,
P. J. Stang, J. Michl, J. Am. Chem. Soc. 2008, 130, 7629–7638; d) Y. R.
Zheng, P. J. Stang, J. Am. Chem. Soc. 2009, 131, 3487–3489; e) A. Gonza-
lez-Alvarez, A. Frontera, P. Ballester, J. Phys. Chem. B 2009, 113, 11479–
11489.
[10] a) G. Tarkanyi, H. Jude, G. Palinkas, P. J. Stang, Org. Lett. 2005, 7, 4971–
4973; b) M. Fuss, H. U. Siehl, B. Olenyuk, P. J. Stang, Organometallics
1999, 18, 758–769.
[11] M. Ferrer, M. Mounir, O. Rossell, E. Ruiz, M. A. Maestro, Inorg. Chem.
2003, 42, 5890–5899.
[26] C. J. Kuehl, S. D. Huang, P. J. Stang, J. Am. Chem. Soc. 2001, 123, 9634–
9641.
[27] M. Fujita, F. Ibukuro, K. Yamaguchi, K. Ogura, J. Am. Chem. Soc. 1995,
117, 4175–4176.
[28] a) U. Frey, L. Helm, A. E. Merbach, R. Romeo, J. Am. Chem. Soc. 1989,
111, 8161–8165; b) G. Alibrandi, G. Bruno, S. Lanza, D. Minniti, R.
Romeo, M. L. Tobe, Inorg. Chem. 1987, 26, 185–190; c) G. Alibrandi, D.
Minniti, L. M. Scolaro, R. Romeo, Inorg. Chem. 1989, 28, 1939–1943;
d) M. R. Plutino, L. M. Scolaro, A. Albinati, R. Romeo, J. Am. Chem. Soc.
2004, 126, 6470–6484.
&
&
Chem. Eur. J. 2014, 20, 1 – 16
www.chemeurj.org
14
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
[29] a) R. van Eldik, Advances in Inorganic Chemistry, Vol. 59, Academic Press,
New York, 2006; b) S. Hochreuther, S. T. Nandibewoor, R. Puchta, R. va-
n Eldik, Dalton Trans. 2012, 41, 512–522; c) T. Soldatovic, S. Jovanovic,
Z. D. Bugarcic, R. van Eldik, Dalton Trans. 2012, 41, 876–884.
[30] a) G. R. Newkome, D. C. Hager, J. Org. Chem. 1978, 43, 947–949; b) L.
Hirsivaara, M. Haukka, J. Pursiainen, J. Organomet. Chem. 2001, 633, 66–
68.
[33] Z. Qin, M. C. Jennings, R. J. Puddephatt, Inorg. Chem. 2001, 40, 6220–
6228.
[34] W. T. Dent, R. Long, A. J. Wilkinson, J. Chem. Soc. 1964, 1585–1588.
[35] D. A. White, J. R. Doyle, H. Lewis in Inorg. Synth., Vol. 13 (Ed.: F. A.
Cotton), Wiley, Hoboken, 1972, pp. 55–62.
[31] P. J. Stang, D. H. Cao, S. Saito, A. M. Arif, J. Am. Chem. Soc. 1995, 117,
6273–6283.
Received: May 9, 2014
[32] F. Dejong, J. J. Bour, P. P. J. Schlebos, Inorg. Chim. Acta 1988, 154, 89–
93.
Published online on && &&, 0000
Chem. Eur. J. 2014, 20, 1 – 16
www.chemeurj.org
15
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
FULL PAPER
&
Self-Assembly
I. Angurell, M. Ferrer,* A. Gutiꢀrrez,
M. Martꢁnez,* M. Rocamora, L. Rodrꢁguez,
O. Rossell, Y. Lorenz, M. Engeser
&& – &&
Kinetico-Mechanistic Insights on the
Assembling Dynamics of Allyl-
Cornered Metallacycles: The PtÀNpy
Bond is the Keystone
Platinum is the winner! PtÀNpy bonds
play a keystone role in the kinetics of
formation and solution behaviour of
several homo- and heterometallomacro-
cycles containing allyl-palladium cor-
ners. The exchange among different
combinations of adequate building
blocks and the supramolecular species
has been studied, and a preliminary
study of the role of theses assemblies as
catalysts in the allylic alkylation reaction
has been undertaken.
&
&
Chem. Eur. J. 2014, 20, 1 – 16
www.chemeurj.org
16
ꢃ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!