Experimental study of speciation and mechanistic implications when using chelating ligands in aryl-alkynyl Stille coupling

Article abstract of DOI:10.1039/d0dt02335c

Neutral palladium(ii) complexes [Pd(Rf)X(P-L)] (Rf = 3,5-C6Cl2F3, X = Cl, I, OTf) with P-P (dppe and dppf) and P-N (PPh2(bzN)) ligands have chelated structures in the solid-state, except for P-L = dppf and X = Cl, were chelated and dimeric bridged structures are found. The species present in solution in different solvents (CDCl3, THF, NMP and HMPA) have been characterised by 19F and 31P{1H} NMR and conductivity studies. Some [Pd(Rf)X(P-L)] complexes are involved in equilibria with [Pd(Rf)(solv)(P-L)]X, depending on the solvent and X. The ΔH° and ΔS° values of these equilibria explain the variations of ionic vs. neutral complexes in the range 183-293 K. Overall the order of coordination strength of solvents and anionic ligands is: HMPA ? NMP > THF and I-, Cl- > TfO-. This coordination preference is determining the complexes participating in the alkynyl transmetalation from PhCCSnBu3 to [Pd(Rf)X(P-L)] (X = OTf, I) in THF and subsequent coupling. Very different reaction rates and stability of intermediates are observed for similar complexes, revealing neglected complexities that catalytic cycles have to deal with. Rich information on the evolution of these Stille systems after transmetalation has been obtained that leads to proposal of a common behaviour for complexes with dppe and PPh2(bzN), but a different evolution for the complexes with dppf: this difference leads the latter to produce PhCCRf and black Pd, whereas the two former yield PhCCRf and [Pd(CCPh)(SnBu3)(dppe)] or [Pd(CCPh)(SnBu3){PPh2(bzN)}]. This journal is

Full text of DOI:10.1039/d0dt02335c

Dalton
Transactions
View Article Online
View Journal
PAPER
Experimental study of speciation and mechanistic
implications when using chelating ligands in aryl-
alkynyl Stille coupling†‡
Cite this: DOI: 10.1039/d0dt02335c
Ana M. Gallego,
Jose M. Martin-Alvarez,
Marconi N. Peñas-Defrutos,
Guillermo Marcos-Ayuso,
and Pablo Espinet
Jesús M. Martínez-Ilarduya
*
Neutral palladium(II) complexes [Pd(Rf)X(P–L)] (Rf = 3,5-C6Cl2F3, X = Cl, I, OTf) with P–P (dppe and dppf)
and P–N (PPh2(bzN)) ligands have chelated structures in the solid-state, except for P–L = dppf and X = Cl,
were chelated and dimeric bridged structures are found. The species present in solution in different sol-
vents (CDCl3, THF, NMP and HMPA) have been characterised by 19F and 31P{1H} NMR and conductivity
studies. Some [Pd(Rf)X(P–L)] complexes are involved in equilibria with [Pd(Rf)(solv)(P–L)]X, depending on
the solvent and X. The ΔH° and ΔS° values of these equilibria explain the variations of ionic vs. neutral
complexes in the range 183–293 K. Overall the order of coordination strength of solvents and anionic
ligands is: HMPA ≫ NMP > THF and I−, Cl− > TfO−. This coordination preference is determining the com-
plexes participating in the alkynyl transmetalation from PhCuCSnBu3 to [Pd(Rf)X(P–L)] (X = OTf, I) in THF
and subsequent coupling. Very different reaction rates and stability of intermediates are observed for
similar complexes, revealing neglected complexities that catalytic cycles have to deal with. Rich infor-
mation on the evolution of these Stille systems after transmetalation has been obtained that leads to pro-
posal of a common behaviour for complexes with dppe and PPh2(bzN), but a different evolution for the
complexes with dppf: this difference leads the latter to produce PhCuCRf and black Pd, whereas the two
former yield PhCuCRf and [Pd(CuCPh)(SnBu3)(dppe)] or [Pd(CuCPh)(SnBu3){PPh2(bzN)}].
Received 1st July 2020,
Accepted 2nd August 2020
DOI: 10.1039/d0dt02335c
rsc.li/dalton
equilibria represented in eqn (1), of [Pd(Pf)X(PPh3)2] (Pf =
C6F5; X = Cl, Br, I, OTf) in THF, NMP (N-methyl-2-pyrrolidone),
Introduction
Pd-catalysed cross-coupling reactions are a most powerful C–C and HMPA (hexamethylphosphoramide). These are solvents
and C–heteroatom bond-forming tool.1,2 The catalysts used are with different coordinating ability, utilized in coupling reac-
either isolated complexes or complexes synthesized in situ tions where the oxidative addition of PfX to Pd0 species in the
from appropriate precursors. Often no further consideration catalytic cycle gives rise to [Pd(Pf)XL2].
on their exact nature in the flask is made. However, speciation
of the catalysts in solution is very important because, as shown
trans‐½PdðPfÞXðPPh3Þ2ꢀ þ nPPh3
for instance in our studies on the Stille process,3–6 it can be
determining of the transmetalation mechanism (cyclic, open,
or via solvent-coordinated cationic species) and can be critical
for the outcome of the processes.7 Some time ago we exam-
ined, in the context of Stille reactions, the speciation, due to
ð1Þ
solvent
ꢁ! Complexes in solution
Table 1 shows that, depending on X, n, and the solvent,
several species can coexist, with features leading to different
electrophilicity of the Pd centre, and opening alternative path-
ways to the next catalytic steps.
Chelating ligands look, in principle, ideal to favour a cis
coordination of the two groups to be cross-coupled, which is a
requirement for coupling, but in practice, they have also
IU CINQUIMA/Química Inorgánica, Facultad de Ciencias, Universidad de Valladolid,
47071 Valladolid, Spain. E-mail: espinet@qi.uva.es
†Dedicated to so many people working in the health services worldwide, who, at
the time this manuscript is being submitted for publication, continue bravely shown to have disadvantages. For instance, in the study of the
fighting to control the COVID-19 pandemic, too often with only limited media
coupling of PfOTf and CH2vCHSnBu3 catalysed by [Pd(Pf)
(OTf)(dppe)], our group observed the key intermediates of the
and risk for their own lives.
‡Electronic supplementary information (ESI) available: Synthesis and full
catalytic cycle, including the transmetalation product [Pd(Pf)
characterisation of the complexes, kinetic, computational and X-ray details; 56
(CHvCH2)(dppe)],8 but the reductive elimination step at room
pages. CCDC 1999841–1999845 and 2012676. For ESI and crystallographic data
in CIF or other electronic format see DOI: 10.1039/d0dt02335c
temperature was very slow because of the strongly chelating
This journal is © The Royal Society of Chemistry 2020
Dalton Trans.

View Article Online
Paper
Dalton Transactions
Table 1 Speciation of palladium PPh3 complexes under different
conditions
Results and discussion
Synthesis and characterisation in the solid state and in CDCl3
X
Solvent
n
Complexes in solution
solution of [Pd(Rf)X(P–L)] (X = Cl, I, OTf) complexes
Cl, Br, I THF
0
0
trans-[Pd(Pf)X(PPh3)2]
trans-[Pd(Pf)X(PPh3)2] + [Pd(Pf)(solv)(PPh3)2]X
in equilibrium
The synthesis of the Rf palladium complexes with the ligands
depicted in Fig. 1 is summarised in Scheme 1. Reacting (2 : 1)
the corresponding P–L ligand with (μ-Cl)2[Pd(Rf)(tht)]2 (1) (tht
= tetrahydrothiophene)16 affords [Pd(Rf)Cl(P–L)] (P–L = dppe,
2a; dppf, 2b; PPh2(bzN), 2c) in high yield. For dppe and
PPh2(bzN) the reactions are as shown in the Scheme 1.
However, in the reaction of 1 with dppf a small amount of a
yellow solid precipitates, while most of the product remains in
solution. The structures of the two competitively formed com-
plexes were solved, and revealed two structural isomers.
OTf
THF
I
NMP
NMP
HMPA
0
trans-[Pd(Pf)X(PPh3)2]
OTf
Cl
0, 2 trans-[Pd(Pf)(solv)(PPh3)2]X
0
trans-[Pd(Pf)(solv)(PPh3)2]X 62% + cis-[Pd(Pf)
(solv)2(PPh3)]X 18% + trans(Cl-Pf)-[Pd(Pf)X
(solv)(PPh3)] 20%
Cl, Br,
OTf
HMPA
2
trans-[Pd(Pf)(solv)(PPh3)2]X
dppe. In fact, complexes carrying too strong chelating ligands
frequently do not lead to good results in catalysis,9 but many
exceptions are also known, not only in Pd-catalysed reac-
tions,10 but also in cooperative Cu/Pd catalysed reactions, e.g.
the recently reported for allylboration of alkynes using dppf as
Pd ligand.11
The major and most soluble isomer is a monomer (ESI
Fig. 1‡), [Pd(Rf)Cl(dppf)] (2b), that displays cis-P–P coordi-
nation to Pd with a P–Pd–P bite angle of 100.55(3)°, slightly
larger than the average bite angle value reported for this
diphosphine (99°).17 The less soluble isomer (Fig. 2) is
dimeric, 2b-dim, and displays two bridging dppf ligands,
mutually trans and monodentate at each palladium.18
Halide metatheses of [Pd(Rf)Cl(P–L)] (2a–2c) with KI
(excess) in acetone/dichloromethane (v/v = 1 : 1) afford the
iodo complexes [Pd(Rf)I(P–L)] with P–L = dppe (3a), dppf (3b),
An interesting case of Stille catalysis is the Pd-catalysed
aryl–alkynyl cross coupling using alkynyl stannane, where the
behaviour of monodentate and chelating ligands drastically
differs in an unexpected direction: the monodentate phos-
phine PPh3 is very inefficient, but some (nor any) dipho-
sphines or P–N ligands make the process very efficient.12,13
A
thorough mechanistic study of the PPh3 case uncovered a
many-pathways reaction and revealed that the reason for the
partial frustration of PhCuC–Rf coupling (Rf = 3,5-C6Cl2F3) is
the formation of trans-[Pd(Rf)(CuCPh)(PPh3)2], which is
reluctant to isomerize to cis as required for coupling.14
The good results with some chelating ligands suggest that they
will enforce the required cis configuration. However, there
have not been any detailed mechanistic studies of this
process.15
In this paper we report the synthesis, structural study and
speciation in three different solvents (THF, NMP, and HMPA)
of [Pd(Rf)X(P–L)] complexes (X = halide or triflate) for the che-
lating ligands shown in Fig. 1, which include two frequently
used chelating diphosphines and one potentially hemilabile
P–N ligand. The speciation data help to understand the
Scheme 1 Synthesis of the complexes used in this work. Rf = 3,5-
C6Cl2F3.
mechanism of transmetalation,
a step common to the
different Pd cross-coupling catalytic cycles that is studied here
for the Stille coupling between [Pd(Rf)X(P–L)] (X = OTf, I) and
PhCuCSnBu3. The aryl Rf provides stability to the aryl com-
plexes, allows us to obtain single crystals, and enables mean-
ingful 19F NMR studies. All the experimental information is
detailed in the ESI.‡
Fig. 2 Molecular structure of (μ-dppf)2[Pd(Rf)Cl]2 (2b-dim). All hydro-
gen atoms, as well as solvent molecules, have been omitted for clarity.
Selected bond lengths (Å) and angles (°): Pd(1)–Cl(1) = 2.334(2), Pd(1)–P
(2*) = 2.3170(19), Pd(1)–P(1) = 2.361(2), Pd(1)–C(11) = 2.015(8).
Fig. 1 P–L ligands used in this work.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Dalton Transactions
Paper
or PPh2(bzN) (3c), as shown in Scheme 1. The X-ray structure square-planar complexes have high rotational barriers,20
of 3c (ESI Fig. 2‡) confirms that the two strong soft ligands (Rf chemical non-equivalence of the two Fo atoms should be
and P) choose cis positions (antisymbiotic behaviour).19
expected. However, the 19F NMR spectra at 298 K of the
In the reactions with AgOTf, the presence of adventitious reported complexes show chemical equivalence, except for the
water in the solvents, produced mixtures in solution of the tri- PPh2(bzN) complexes where broad signals are observed. Lower
flate complexes [Pd(Rf)(OTf)(P–L)] (4a–4c) and [Pd(Rf)(OH2)(P– temperatures lead to different degrees of broadening in the
L)](TfO) (5a–5c), also present in the solid state. These struc- other complexes. This suggests that the chemical equivalence
tures are supported in all cases by C, H, N analysis, 19F, 31P observed at 298 K is attained by fast conformational inversion
and 1H NMR spectra, and IR data: a broad band close to of the non-planar palladacycles formed upon P–L coordi-
3270 cm−1 is due to the ν(O–H) stretching absorptions of co- nation,21 which are plane-averaged in the NMR time scale at
ordinated water, and bands due to free (1300 cm−1) and co- temperatures depending on the complex.
ordinated triflate (1230, 1210, and 1190 cm−1) are also
observed. Heating the mixtures up to constant weight (about
Equilibria between [Pd(Rf)X(P–L)] (X = Cl, I, OTf) and [Pd(Rf)
(solv)(P–L)]X complexes in THF, NMP and HMPA
2–3 days) in an oven at 50 °C led to complete elimination of
H2O, giving pure [Pd(Rf)(OTf)(P–L)] complexes.
The behaviour of the neutral complexes [Pd(Rf)X(P–L)] (X = Cl,
I, OTf) in THF, NMP, and HMPA is reported below. The NMR
spectra of the halo complexes [Pd(Rf)X(P–L)] (X = Cl, I) in THF,
are similar to those obtained in CDCl3. However the 19F NMR
of the triflate complexes 4b and 4c (not 4a) indicate the exist-
ence of entropy dependent equilibria of the neutral complexes
with [Pd(Rf)(THF)(P–L)](TfO) (P–L = dppf, 6b; PPh2(bzN), 6c),
which show strong temperature dependence (see Fig. 4 for
4b/6b). Participation of some adventitious water in these com-
plexes, in fast THF/OH2 exchange, is highly probable.
The X-ray structure of 5c (Fig. 3) shows bonds lengths and
angles typical for other complexes containing the PPh2(bzN)
ligand. The water ligand is involved in hydrogen bonding to
the triflate anions, giving rise to dimers in the crystal (ESI
Fig. 3‡).
In the non-coordinating solvent CDCl3 the neutral com-
plexes (2a–c, 3a–c, 4a–c) do not dissociate into cationic species
and their NMR spectra are consistent with the structures pro-
posed in Scheme 1. In significant presence of adventitious
water, OH2 competes with the triflate anion for the coordi-
nation position, and equilibria 4a–c/5a–c are established. The
31P{1H} spectra of the diphosphine complexes show two multi-
plets (1 : 1 ratio) due to P–P and P–Fo (ortho-fluorine) coup-
lings. The upfield signal corresponds to the P atom trans to Rf,
more shielded due to the high trans influence of this group.
For the PPh2(bzN) ligand, a triplet is observed in the 31P{1H}
spectra, due to coupling to the two Fo atoms of the fluoroaryl
group in cis (compound 4c), or a triplet of doublets when the
resolution is able to detect also coupling to Fp (compounds 2c
and 3c). All the NMR information is given in ESI.‡
The CF3 signal of OTf, and the Fp signals of Rf allow us to
observe 4b and 6b separately at low temperatures (Fig. 4). We
detect OTf/THF exchange and inversion of the 4b/6b relative
abundance in solution at higher temperatures (above 243 K).
For unambiguous assignment of the 4b and 6b signals, the
cationic complex [Pd(Rf)(THF)(dppf)](BF4) (6b*), with a non-
coordinating anion, was synthesized in situ at 183 K treating
2b with AgBF4 in dry THF. This reference complex proves that,
for triflate, the ionic complex is the one very predominant at
low temperature, but its percentage in the mixture diminishes
as the temperature increases. In addition, above 243 K all the
signals start to show broadening as symptom of slow neutral/
cationic exchange. At 293 K they appear as thin averaged
signals above coalescence, corresponding to a fast equilibrium
exchange. Plotting –ln K vs. 1/T in the low temperature range
(183–243 K), the thermodynamic parameters for the neutral/
In none of the structures studied is the Pd coordination
plane a symmetry plane. Since perhaloaryl groups in Pd
cationic equilibrium were calculated: ΔH° = –18.4
0.9
kJ mol−1; ΔS° = −77 4 J K−1 mol−1. The small value of ΔH°
Fig. 3 Molecular structure of 5c. Most hydrogen atoms, as well as
solvent molecules, have been omitted for clarity. Selected bond lengths
(Å) and bond angles (°): Pd(1)–P(1) = 2.2156(8), Pd(1)–N(1) = 2.162(3), Pd
(1)–O(1) = 2.133(3), Pd(1)–C(11) = 2.008(3), N(1)–Pd(1)–P(1) = 93.31(7).
Fig. 4 Variable temperature 19F NMR of a solution of [Pd(Rf)(OTf)
(dppf)] (4b) in dry THF as solvent.
This journal is © The Royal Society of Chemistry 2020
Dalton Trans.

View Article Online
Paper
Dalton Transactions
Table 3 Summary of the speciation observed in different solvents for
the complexes studied in this work. N = [Pd(Rf)X(P–L)]; CAT = [Pd(Rf)
(solv)(P–L)]+X−
makes the equilibrium highly entropy and temperature depen-
dent, as observed.22 With these data in hand, the equilibrium
constant and the neutral/ionic complex ratio in THF at any
temperature can be calculated. In a catalytic process this ratio
could be relevant in order to determine the preferred catalytic
mechanism. In the case of dppf, the Pd ratio 4b/6b is 92/8
(K323 = 0.085) for a solution 10−2 M in Pd at 323 K. The same
calculations for PPh2(bzN) affords 4c/6c = 96/4 molar ratio
(K323 = 0.043).
X
CDCl3
THF
NMP
HMPA
Cl
I
OTf
N
N
N
N
N
N
CAT
CAT
N + CATa
CAT
N + CAT(OH2)
N + CAT
a Only the cationic species is observed for X = I and P–L = PPh2(bzN).
For 4a in THF, only one species is observed at all tempera-
tures tested. This unexpected result prompted us to perform
conductivity measurements. The conductivity of [Pd(Rf)Cl
(dppe)] (2a) was also measured for comparison (Table 2,
entries 1–6). The molar conductivities for 4a and 4b–c are
similar and much lower than for a 1 : 1 electrolyte (entry 1),
suggesting the existence of similar neutral/ionic equilibria in
the three cases. Hence, we suggest that the different NMR
behaviour of 4a must be simply due to faster equilibration in
solution.
As for the speciation of [Pd(Rf)X(P–L)] (X = Cl, I, OTf) in
NMP and HMPA, these solvents are more strongly coordinating
than THF, so it can be assumed safely that their triflate
ligands in these solvents will be 100% in the ionic form, with
at least one solvent molecule coordinated: [Pd(Rf)(solv)(P–L)]
(TfO) (solv = NMP (7a–c), HMPA (8a–c)). In fact, the 19F NMR
spectra confirm that NMP and HMPA are able to displace
quantitatively the triflate at any temperature. These two sol-
vents differ in that NMP is not able to displace extensively any
halide from the metal coordination sphere, whereas the more
strongly coordinating HMPA is able to displace the chloride
ligand completely, and the iodide in a large extension in all
the [Pd(Rf)X(P–L)] complexes (complete for P–L = PPh2(bzN)).
The equilibrium constants, measured by 19F NMR in HMPA at
323 K with [Pd] = 10−2 M, are 0.057 (3a) and 0.106 (3b), corres-
ponding to the neutral-complex/ionic-complex ratios 13/87 and
8/92, respectively.
All these observations were confirmed by the spectra
obtained in separate experiments where quantitative formation
of [Pd(Rf)(solv)(P–L)]+ was forced upon addition of AgBF4. The
differences observed between the halo complexes (X = Cl vs. I)
must be attributed to an important contribution of the sol-
vation energy of the anions (EsolvCl− > EsolvI−), overcoming the
behaviour expected from Pd–X bond energies (Pd–Cl > Pd–I).
The molar conductivities in NMP and HMPA (Table 2, entries
7–17) are in good agreement with the previous results, sup-
porting substantial displacement of I− from palladium only in
HMPA. This should open to halides ionic pathway in this
solvent, as it happens for triflates in THF. Some moderate
iodide displacement seems to occur in NMP, where a neutral/
ionic equilibrium could exist to a small extent.23
One could wonder whether these stronger solvents might
displace the N extreme of the ligand PPh2(bzN) from palladium,
but the observation of two (1 : 1) non-equivalent Fo signals in
both solvents suggests that this does not occur.24 A summary of
the results of this speciation study is given in Table 3.
Kinetic monitoring of transmetalation reactions in THF using
excess PhCuCSnBu3 and [Pd(Rf)X(P–L)] (X = OTf, I)
The reactions of [Pd(Rf)X(P–L)] with an excess of
PhCuCSnBu3 (9) were monitored in THF by 19F NMR
(eqn. (2)) for: (i) X = TfO, as an example of anion either ionized
in THF (Table 3) or expected to be easily substituted by the
triple bond of the reagent (hence expected to follow an ionic
transmetalation pathway); (ii) X = I as an example of co-
ordinated anion reluctant to facilitate ionization in THF. The
absence of RfI or RfOTf precludes the classical oxidative
addition leading, in the catalytic processes, to further
evolution.25
Table 2 Molar conductivities (ΛM) in THF, NMP and HMPA solutions at
298 K (rn or ri stand for reference of neutral or ionic 1 : 1 electrolytes,
respectively)
Entry
Solvent
Complex
ΛM (S cm2 mol−1
)
1(ri)
2
3
4
THF
THF
THF
THF
THF
THF
[Pd(Rf)(PPh3)(dppe)](TfO)
[Pd(Rf)(OTf)(dppe)] (4a)
[Pd(Rf)(OTf)(dppf)] (4b)
[Pd(Rf)(OTf){PPh2(bzN)}] (4c)
[Pd(Rf)I(dppe)] (3a)
19
4
4
4
0
5
6(rn)
[Pd(Rf)Cl(dppe)] (2a)
0
ð2Þ
7(ri)
8
9
10
11
NMP
NMP
NMP
NMP
NMP
NMP
[Pd(Rf)(PPh3)(dppe)](TfO)
[Pd(Rf)(OTf)(dppe)] (4a)
[Pd(Rf)I(dppe)] (3a)
[Pd(Rf)I(dppf)] (3b)
[Pd(Rf)I{PPh2(bzN)}] (3c)
[Pd(Rf)2(dppe)]
19
17
1
2
2
A Pd : Sn = 1 : 20 ratio was used in order to reproduce a fre-
quent catalytic proportion. Note that the products lacking Rf
are not observed in the 19F NMR monitoring, but can be fol-
lowed by 31P NMR. The results of these studies led to the
general pathways in Scheme 2.
12(rn)
0
13(ri)
14
15
16
17(rn)
HMPA
HMPA
HMPA
HMPA
HMPA
[Pd(Rf)(PPh3)(dppe)](TfO)
[Pd(Rf)I(dppe)] (3a)
[Pd(Rf)I(dppf)] (3b)
[Pd(Rf)I{PPh2(bzN)}] (3c)
[Pd(Rf)2(dppe)]
14
15
14
15
0
Transmetalation to [Pd(Rf)(OTf)(dppe)] (4a) in THF. When
[Pd(Rf)(OTf)(dppe)] (4a) and PhCuCSnBu3 (9) were mixed in
Dalton Trans.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Dalton Transactions
Paper
Scheme 3 COPASI equations for the kinetic simulations of the reaction
of [Pd(Rf)(OTf)(dppe)] (4a) with an excess of PhCuCSnBu3 (9) (Pd : Sn =
1 : 20) in THF.
Scheme 2 Products detected in the reactions of [Pd(Rf)X(P–L)] (X =
OTf, 4a–c; X = I, 3a–c) with an excess of PhCuCSnBu3 (9) (Pd : Sn =
1 : 20) in THF.
THF at 313 K, a fast transmetalation (instantaneous for the
standards of the monitoring) took place giving [Pd(Rf)
(CuCPh)(dppe)] (11a), which started to evolve slowly to the
initial reduction product [Pd(η2-PhCuCRf)(dppe)] (12a), and
the coupling product PhCuCRf (13), as plotted in Fig. 5. For a
long time, no black Pd is observed because, as shown in
Scheme 2, the formation of 13 is concomitant with the partici-
pation of a second molecule of the alkynylstannane 9, yielding
14a (observed by 31P{1H} NMR). Only at very long times black
Pd is formed by slow decomposition. The experimental evol-
ution was perfectly fitted (continuous lines) with COPASI soft-
ware,26 applying the equations in Scheme 3 (full details in
ESI‡). The adjusted rate constants for the competitive evol-
ution from 11a yield very precise values: (2.63 0.07) × 10−5
s−1 for k3, and (1.26 0.03) × 10−4 mol−1 L s−1 for k4.
Fig. 6 DFT optimised structure of the cation 10a+. Relevant distances
(Å): Pd–C1 = 2.269; Pd–C2 = 2.557; C1–C2 = 1.237; Pd–P1 = 2.374; Pd–
P2 = 2.290. Relevant angles (°): Sn–C1–C2 = 142.29; C1–C2–Cipso
=
176.48.
is favourable by ΔG = −6.5 kcal mol−1. The proposed complex
10a(TfO) could not be isolated, but the DFT optimised struc-
ture (wB97X-D level) of its cation 10a+ (with Me3Sn instead of
Bu3Sn) is shown in Fig. 6. Interestingly, the coordination
square plane of Pd is well defined by the two P atoms, the Rf
group, and C1 of the alkyne, all at reasonable bond distances,
but C2 is at long distance from Pd (2.557 Å). Moreover,
whereas the C1–C2 distance is consistent with a triple bond,
and the C1–C2–Cipso fragment is quite linear (176.48°), a
remarkable bending is observed for Sn–C1–C2 (142.29°). These
structural parameters are far from typical η2-alkyne coordination
and point out to much higher implication of C1 than C2 elec-
tron density in this asymmetric coordination. Natural Bonding
Orbital (NBO) and Second Order Perturbation (SOPT) analyses
on optimised geometry of 10a+ support this hypothesis.
The same reaction was studied starting at 220 K, which
allowed us to observe and identify an initial intermediate
species
[Pd(Rf)(η2-PhCuCSnBu3)(dppe)](TfO),
10a(TfO),
formed by substitution of TfO−or THF by PhCuCSnBu3 (9), in
the equilibrium 4a/6a.27,28 The computed thermodynamics of
ligand substitution in 6a+ of THF by 9, in order to form 10a+,
Table 4 collects the two main donations from the alkyne
moiety to palladium. The acceptor orbital is common for both
interactions: the antibonding BD*(1) Pd–P2 with high palla-
dium contribution (sd electron density). The donors are
different. One contains π-electron density of the triple bond
(BD(3) C1–C2, 100% p) and accounts for the stronger
donation. The other is an occupied bonding orbital BD(1) C1–
Sn. The two have in common that they are polarised towards
the C1 atom (57% and 83% respectively), which leads to an
alkyne coordination much closer to η1 than to η2, very prone to
Fig. 5 19F NMR monitoring of the reaction of [Pd(Rf)(OTf)(dppe)] (4a)
with PhCuCSnBu3 (9) (Pd : Sn = 1 : 20) at 313 K. Continuous lines are the
result of COPASI fitting.
This journal is © The Royal Society of Chemistry 2020
Dalton Trans.

View Article Online
Paper
Dalton Transactions
Table 4 Selected Donor–Acceptor interactions and SOPT energies
(kcal mol−1) for 10a+
Donor/contribution
Acceptor/contribution
ESOPT
BD(3) C1–C2
57% C1; p(100%)
BD*(1) Pd–P2
71% Pd; s(48%) d(52%)
47.5
43% C2; p(100%)
29% P2; s(28%) p(72%)
BD(1) C1–Sn
83% C1; s(55%) p(45%)
BD*(1) Pd–P2
71% Pd; s(48%) d(52%)
23.0
17% Sn; s(16%) p(84%)
29% P2; s(28%) p(72%)
The atom labels are as in Fig. 6. BD and BD* stand, respectively, for
bonding and antibonding orbitals.
Fig. 7 Molecular structure of 11a. All hydrogen atoms, as well as
solvent molecules, have been omitted for clarity. Only one of the two
slightly different molecules of the asymmetric unit is showed. Selected
bond lengths (Å) and bond angles (°): Pd(1)–C(11) = 2.078(5); Pd(1)–P(1)
= 2.2882(14); Pd(1)–P(2) = 2.2804(15); Pd(1)–C(1) = 2.019(6); P(1)–Pd(1)–
P(2) = 85.84(5); C(1)–Pd(1)–P(2) = 96.76 (16); C(1)–Pd(1)–C(11) = 89.6(2);
P(1)–Pd(1)–C(11) = 93.87(13).
the formation of the σ Pd–C1 bond found in 11a, once the
alkynyl transmetalation has been completed.
It is worth commenting that the steric crowding between
the Rf ring and the PPh2 group should impose severe restric-
tions to rotation around the Pd–Rf bond, and the alkynyl
coordination perpendicular to the Pd coordination plane
should give rise to chemical non-equivalence of the two Fo.
Since this non-equivalence is not observed in the 19F NMR
spectra, equivalence must be achieved by fast rotation of the
alkyne around the Pd–C1, confirming the proposed lesser
implication of C2 in this alkyne coordination.
Already at the low temperature of 220 K, the alkynyl trans-
metalation is taking place slowly, leading to the σ-alkynyl
complex [Pd(Rf)(CuCPh)(dppe)] (11a). This transmetalation
sequence reminds that of vinyl stannane, which also starts by
coordination of the double bond of the vinyl group.29 Complex
11a was unambiguously characterised by NMR methods: its
19F NMR spectrum shows the resonances and chemical shifts
expected for a Rf–PdII group, with the Fo coupled to the trans
(4JP–F = 12.6 Hz) and cis (4JP–F = 3.3 Hz) 31P atoms. In the 31P
{1H} spectrum two signals at δ = 50.4 ppm (P trans to Rf) and δ
Fig. 8 31P{1H} NMR spectrum of the reaction between [Pd(dba)(dppe)]
(15a) and PhCuCSnBu3 (9) (Pd : Sn = 1 : 1), in THF, at 298 K.
2
= 55.2 ppm (P cis to Rf) are observed with JPP = 20.6 Hz. The
molecular structure of 11a could be solved (Fig. 7).
At 263 K, complex 11a starts to undergo reductive elimin-
ation and, in the presence of alkynyl stannane 9, [Pd0(η2-
PhCuCRf)(dppe)] (12a) and [PdII(CuCPh)(SnBu3)(dppe)] (14a) ponds to the phosphorous atom trans to the stannyl group. In
are formed, with concomitant release of PhCuCRf (13). The addition, the byproduct complexes [Pd(η2-PhCuC–CuCPh)
31P{1H} spectrum of 12a shows an AB spin system consistent (dppe)] (16a, red numbers),32 [Pd(dppe)2] (purple asterisk),33
with a trigonal-planar coordinated Pd0, with the triple bond of and unreacted 15a (green numbers) are observed. The unusual
the asymmetric alkyne lying in the P–Pd–P plane. At higher oxidative addition of 9 to Pd0 giving rise to 14a was already
temperature (295 K) more 13 is slowly released with increase of reported by Shirakawa and Hiyama.12,13
14a, which explains why only little black Pd is hardly observed.
Luckily, single crystals of the Pd–Sn bonded species 14a
At higher temperatures (see Fig. 5 for the full evolution at could be obtained from the reaction mixture. Its molecular
313 K), slow formation of more black Pd is observed with time. structure is shown in Fig. 9. Some structural aspects deserve
In order to help with structural assessment of the Pd–Sn comment: (i) the complex displays clearly a Sn–Pd bond with
bonded species from the NMR signals in the reaction mixture, intermetallic distance well below the sum of covalent radii
14a was independently prepared from [Pd(dba)(dppe)]30 (15a) (2.78 Å);34 (ii) the substantial elongation of the Pd(1)–P(1) dis-
and PhCuCSnBu3 (9).31 The resulting 31P{1H} NMR spectrum tance, almost 0.1 Å longer than Pd(1)–P(2), proves the large
of the latter reaction is shown in Fig. 8.
It displays for 14a two doublets (blue numbers) at 24.9 and small C(1)–Pd(1)–Sn(1) (78.2°) and large P(2)–Pd(1)–Sn(1)
trans influence of metal donors in a M–M′ bond;35 (iii) the
117
51.0 ppm (2JPP = 24.5 Hz). The tin satellites (2JP
= 1437 Hz (99.3°) angles may be related to steric hindrance issues pro-
Sn
2
119
and JP
= 1506 Hz) support that the upfield signal corres- voked by the bulky SnBu3 group.
Sn
Dalton Trans.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Dalton Transactions
Paper
Scheme 4 Kinetic model for the reaction of [Pd(Rf)I(dppe)] (3a) with an
excess of PhCuCSnBu3 (9). The inhibition of the conversion of 3a by 13
is proposed to occur via [Pd(Rf)(η2-PhCuCRf)(dppe)]I (Ialk), which is
experimentally unobservable.
Fig. 9 Molecular structure of 14a. All hydrogen atoms have been
omitted and Bu groups have been also simplified for clarity. Selected
bond lengths (Å) and bond angles (°): Pd(1)–Sn(1) = 2.5916(9); Pd(1)–C(1)
= 2.009(10); Pd(1)–P(1) = 2.362(2); Pd(1)–P(2) = 2.271(2); P(1)–Pd(1)–P(2)
= 85.99(8); C(1)–Pd(1)–Sn(1) = 78.2(3); C(1)–Pd(1)–P(1) = 96.1(3); Sn(1)–
Pd(1)–P(2) = 99.31(7).
intermediate 10a(I) (Scheme 2), and consequently slows down
the transmetalation. We confirmed experimentally that the
initial reaction rate falls dramatically in presence of added 13.
It is worth noting that the adjusted kinetic parameters for the
evolution of the system from 11a give values (2.11 0.02) ×
10−5 s−1 for k3 and (1.29 0.02) × 10−4 mol−1 L s−1 for k4, very
similar to the ones obtained in the previous fitting with the
OTf analogue (Fig. 5 and Scheme 3).
Transmetalation to [Pd(Rf)(OTf)(dppf)] (4b) in THF. The
reaction of [Pd(Rf)(OTf)(dppf)] (4b) with PhCuCSnBu3 (9) in
THF at 313 K is much faster than with dppe and leads immedi-
ately to the coupling product PhCuCRf (13) and abundant
black Pd. Monitoring the reaction at 295 K, to make the
different steps slower, we could observe some intermediates
(Fig. 11).
Although this fast process allowed only for acquisition of a
lower number of data during monitoring, the profile confirms
basically the same sequence of steps found for dppe in
Scheme 2. At 220 K, intermediate [Pd(Rf)(η2-PhCuCSnBu3)
(dppf)](TfO) (10b(TfO)) can be observed. Complex [Pd(Rf)
(CuCPh)(dppf)] (11b) is also observed, but [Pd(CuCPh)
(SnBu3)(dppf)] (14b) is not detected. Overall the data indicate
that the main qualitative difference occurs once 11b has been
formed. At variance with dppe, dppf is a wide bite angle
ligand,17 and efficiently induces cross coupling from 11b to
12b. In addition, it prevents oxidative addition by 9 to a puta-
tive 14b, too much crowded. Thus, the decomposition of 12b
leads directly and efficiently to 13 and black Pd.
Transmetalation to [Pd(Rf)I(dppe)] (3a) in THF. The kinetic
profile of the reaction of [Pd(Rf)I(dppe)] (3a) and
PhCuCSnBu3 (9), in THF at 313 K, is shown in Fig. 10 and
reflects clearly that this transmetalation is considerably slower
than for the triflato complex 4a. In the case of 3a, about 30%
of the initial complex is still unreacted after 16 h at 313 K.
Initially complex [Pd(Rf)(CuCPh)(dppe)] (11a) is the product
that is faster formed but its concentration starts to diminish
relatively early in favour of [Pd(η2-PhCuCRf)(dppe)] (12a) and
PhCuCRf (13) as the major products. In addition, [Pd
(CuCPh)(SnBu3)(dppe)] (14a) is also formed as evidenced by
31P{1H} NMR spectra.
The abnormal consumption profile of the initial complex
3a, which slows down not following a pseudo first order on Pd
(9 is in large excess), can be perfectly fitted (see Scheme 4 for
the kinetic model, and ESI Table 5‡ for all numeric data).
Clearly, the OTf− for 9 or I− for 9 ligand substitution equilibria
are determinant of the transmetalation rate, since k2 to com-
plete the transmetalation is fast. In the COPASI equations the
consumption of a second molecule of 9 per Pd, and the inhi-
bition introduced by product 13 competing with 9 for I− dis-
placement, produce much more effect than in the reaction
with triflate. This effect is detrimental for the formation of
Transmetalation to [Pd(Rf)I(dppf)] (3b) in THF. The reaction
of [Pd(Rf)I(dppf)] (3b) with PhCuCSnBu3 (9) at 313 K in THF
Fig. 10 19F NMR monitoring of the reaction between [Pd(Rf)I(dppe)]
(3a) and PhCuCSnBu3 (9) (Pd : Sn = 1 : 20) in THF at 313 K. Continuous
lines represent the kinetic fitting of data using COPASI software.
Fig. 11 19F NMR monitoring of the reaction between [Pd(Rf)(OTf)
(dppf)] (4b) and PhCuCSnBu3 (9) (Pd : Sn = 1 : 20) in THF at 295 K. In this
case continuous lines are just to help the eye.
This journal is © The Royal Society of Chemistry 2020
Dalton Trans.

View Article Online
Paper
Dalton Transactions
Fig. 14 19F NMR monitoring of the reaction between [Pd(Rf)I
{PPh2(bzN)}] (3c) and PhCuCSnBu3 (9) (Pd : Sn = 1 : 20) in THF at 313 K.
Fig. 12 19F NMR monitoring of the reaction between [Pd(Rf)I(dppf)]
(3b) and PhCuCSnBu3 (9) (Pd : Sn = 1 : 20) in THF at 313 K.
117
119
and 2JP
coupling constants (2JP
Fig. 9‡).36
≈ 15 Hz, see ESI
Sn
Sn
follows a pathway presumably similar to 3a. The evolution of
3b is slow because the I−/9 ligand exchange + transmetalation
step is slow. In fact, intermediate [Pd(Rf)(CuCPh)(dppf)] (11b)
was not detected because its formation rate is slower than its
coupling rate to [Pd(η2-PhCuCRf)(dppf)] (12b) and 13
(Fig. 12). Although 12b is an important early component of the
mixture, eventually PhCuCRf (13), released from slow
decomposition of 12b, becomes the main product. The
increasing concentration of alkyne 13 (putative 14b is not
formed) is in this case the only responsible for the abnormally
high deceleration of 3b conversion.
Transmetalation to [Pd(Rf)(OTf){PPh2(bzN)}] (4c) in THF.
The behaviour of [Pd(Rf)(OTf){PPh2(bzN)}] (4c) reacting with
PhCuCSnBu3 (9) in THF is similar to the dppe complex 4a,
also in the non-observation of black Pd, although differences
in the relative rates of the different steps (different ΔΔG‡ for
the steps with PPh2(bzN) vs. the steps with dppe) produce dra-
matic graphical modifications in the profiles observed
(Fig. 13).
Transmetalation to [Pd(Rf)I{PPh2(bzN)}] (3c) in THF. For the
reaction of [Pd(Rf)I{PPh2(bzN)}] (3c) with PhCuCSnBu3 (9) in
THF at 313 K, the profile of the reaction obtained by 19F NMR
shows only the disappearance of 3c and the formation of 13
(Fig. 14). The fast rates following the slow transmetalation
prevent any observation in the 19F monitoring other than the
formation of 13.
Analysis of the transmetalation mechanism
Mechanistic proposals requiring dissociation of one coordinat-
ing end of the chelates can be discarded for the diphosphine
complexes because this is a high energy process, and also for
the PPh2(bzN) complexes since they behave similarly. The
most reasonable mechanisms for all the cases studied here are
open mechanisms, with transmetalations initiated by SN2
associative ligand substitution via pentacoordinate intermedi-
ates, with retention of the coordination site (Scheme 5).
The cases of cationic (6) or neutral (4 and 3) starting com-
plexes are represented. In just one step they all lead to the
same intermediate 10(X). As we have seen in our kinetic
studies, this ligand substitution is difficult for 3, and very easy
for 4 or 6. Finally, SN2 attack of X− (X = I, OTf) to the electro-
philic Sn centre leads to concerted evolution to 11 and
The transmetalation step on 4c giving rise to intermediate
[Pd(Rf)(CuCPh){PPh2(bzN)}] (11c) is only somewhat slower
than for 4a, but the C–C coupling to form 12c, and the sub-
sequent oxidative addition of 9 to give 14c, are considerably
faster. The reaction is complete after 4 hours (vs. 16 hours for
4a) and only the final products 13 (in the 19F NMR monitoring)
and 14c (in the 31P NMR spectrum) can be observed at 313 K.
A cis-P,Sn configuration of the phosphine and the stannyl
2
groups is safely assigned for 14c, based on the small JPSn
Scheme 5 Proposed mechanism for the transmetalation reaction,
starting on 3, 4, or 6. Above, the transition states of the initial ligand sub-
stitution (TS1 or TS2), and the subsequent transmetalation (TS3).
Fig. 13 19F NMR monitoring of the reaction between [Pd(Rf)(OTf)
{PPh2(bzN)}] (4c) and PhCuCSnBu3 (9) (Pd : Sn = 1 : 20) in THF at 313 K.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Dalton Transactions
Paper
XSnBu3. Support for this proposal is multiple: (i) the alkynyl leads to early formation of black Pd and PhCuCRf as the pro-
triple bond is a better donor than the vinyl double bond, ducts of the reaction. For the other ligands, in the circum-
which was reported to follow this kind of pathway by DFT cal- stances of absence of RfI, the alkynylstannane PhCuCSnBu3,
culations,29 hence alkynyl makes the ligand substitution which is in addition the alkynyl-transmetalating nucleophile,
easier; (ii) as discussed for Fig. 6, the involvement of BD(1) is additionally able to act as oxidant at the end of the reaction,
C1–Sn as donor NBO in intermediates 10+ enhances the elec- preventing for long time the formation of black Pd. The reac-
trophilicity of Sn and facilitates the mechanism proposed in tion products are PhCuCRf and [Pd(CuCPh)(SnBu3)
Scheme 5 for conversion to 11. The transmetalations are extre- {PPh2(bzN)}] or [Pd(CuCPh)(SnBu3)(dppe)].
mely fast with coordinated OTf or coordinated THF starting
This work shows that carrying out the study of successive
complexes, because they make Pd a more electrophilic centre, catalytic steps in an on-purpose frustrated cycle is a useful strat-
and because they are better leaving ligands than iodide. egy to understand problems associated to catalytic synthesis.
Overall this makes the reductive elimination to be rate limiting However, tin derivatives are slow transmetalating reagents, and
for them. The unequal capability for the ligands used to similar studies with faster nucleophiles may require monitor-
promote C–C bond formation is clear, revealing a sequence ing at lower temperatures.
dppf ≫ PPh2(bzN) > dppe according to the different times
needed to achieve total conversion of the coupling products
into PhCuCSnBu3 (less than 30 min at 295 K for dppf, 4 and Conflicts of interest
16 hours at 313 K for PPh2(bzN) and dppe respectively). For
the iodo derivatives the transmetalation turns out to be always
There are no conflicts to declare.
the rds.
Acknowledgements
Conclusions
We gratefully acknowledge financial support of the Spanish
MINECO (projects CTQ2016-80913-P and CTQ2017-89217-P) and
the Junta de Castilla y León (project VA062G18, UIC 176).
M. N. P.-D. thanks the Spanish MECD for an FPU scholarship.
The study of the speciation of some [Pd(Rf)X(P–L)] complexes
with chelating ligands and its subsequent reactivity with
PhCuCSnBu3 in the absence of an electrophile as oxidant,
offers a frustrated Stille process that allows to uncover aspects
that would remain hidden in a running cycle.
The speciation studies show that some neutral starting
complexes can be immediately transformed in cationic com-
plexes in an appropriate solvent, which will change its reactiv-
ity as has been discussed in previous studies.3 Particularly
interesting, this change of Pd species is possible not only from
starting complexes with weak anionic ligands (X− = TfO−) but
also for those with halides (X− = I−, Cl−) if a good solvating
solvent is chosen in order to compensate for dissociation
energy.
Notes and references
1 T. J. Colacot, New Trends in Cross-Coupling. Theory and
Applications, Royal Society in Chemistry, Cambridge, UK,
2015.
2 For a recent review for application of the Stille reaction in
the synthesis of natural products see: M. M. Heravi and
L. Mohammadkhani, J. Organomet. Chem., 2018, 869, 106–
200.
3 P. Espinet and A. M. Echavarren, Angew. Chem., Int. Ed.,
2004, 43, 4704–4734.
4 C. Cordovilla, C. Bartolomé, J. M. Martínez-Ilarduya and
P. Espinet, ACS Catal., 2015, 5, 3040–3053.
5 A. L. Casado and P. Espinet, J. Am. Chem. Soc., 1998, 120,
8978–8985.
All the reactions in THF start with ligand substitution of X−
or coordinated THF by the entering alkynyl triple bond, via
open alkynyl transmetalation. This ligand substitution step is
fast for the easy leaving TfO− or THF, and considerably slower
for I−.
In the absence of the natural oxidative addition reagent
that restarts the cycle (RfI in this case), the evolution of the
process can be monitored, observing very different influences
of X and the ligand on the evolution and rates of the ligand
substitution, transmetalation, and reductive elimination pro-
cesses. The formation of cationic species facilitates the ligand
substitution and, consequently the transmetalation, due to
increased electrophilicity of the Pd centre. In THF, the stronger
ligand I− always makes the evolution of the system slower.
6 A. L. Casado, P. Espinet and A. M. Gallego, J. Am. Chem.
Soc., 2000, 122, 11771–11782.
7 For a recent review about solvent effects in Pd chemistry
see: J. Sherwood, J. H. Clark, I. J. S. Fairlamb and
J. M. Slattery, Green Chem., 2019, 21, 2164–2213.
8 A. L. Casado, P. Espinet, A. M. Gallego and J. M. Martínez-
Ilarduya, Chem. Commun., 2001, 339–340.
9 A. Fihri, P. Meunier and J.-C. Hierso, Coord. Chem. Rev.,
2007, 251, 2017–2055.
An interesting effect of the chelating ligand is observed for 10 (a) Y. Y. Belosludtsev, R. K. Bhatt and J. R. Falck,
dppf. Because it is a wide bite ligand, it promotes fast coupling
to PhCuCRf of the transmetalated intermediate and hinders
oxidative addition of the Pd0 byproduct by PhCuCSnBu3. This
Tetrahedron Lett., 1995, 36, 5881–5882; (b) A. S. Gajare,
R. S. Jensen, K. Toyota, M. Yoshifuji and F. Ozawa, Synlett,
2005, 144–148; (c) C. M. Crawforth, S. Burling,
This journal is © The Royal Society of Chemistry 2020
Dalton Trans.

View Article Online
Paper
Dalton Transactions
I. J. S. Fairlamb, A. R. Kapdi, R. J. K. Taylor and 23 For NMP conductivity values see: A. K. Banerjee,
A. C. Whitwood, Tetrahedron, 2005, 61, 9736–9751.
11 J. Mateos, E. Rivera-Chao and M. Fañanás-Mastral, ACS
Catal., 2017, 7, 5340–5344.
12 E. Shirakawa, H. Yoshida and T. Hiyama, Tetrahedron Lett.,
1997, 38, 5177–5180.
A. J. Layton, R. S. Nyholm and M. R. Truter, J. Chem. Soc. A,
1969, 2536–2543.
24 The formation of cis-[Pd(Rf)(solv)2{PPh2(bzN)-κ1-P}] should
produce Fo equivalence.
25 For a recent example of the Stille reaction using alkynyl-
stannanes see: A. S. Levashov, D. S. Buryi,
O. V. Goncharova, V. V. Konshin, V. V. Dotsenko and
A. A. Andreev, New J. Chem., 2017, 41, 2910–2918.
26 S. Hoops, S. Sahle, R. Gauges, C. Lee, J. Pahle, N. Simus,
M. Singhal, L. Xu, P. Mendes and U. Kummer,
Bioinformatics, 2006, 22, 3067–3074.
13 E. Shirakawa and T. Hiyama, J. Organomet. Chem., 1999,
576, 169–178.
14 M. H. Pérez-Temprano, A. M. Gallego, J. A. Casares and
P. Espinet, Organometallics, 2011, 30, 611–617.
15 An interesting kinetic study of the oxidative addition step,
in the context of the catalytic cycle, is available, but it does
not provide much information on the intricate transmetala- 27 For cationic PdII complexes with coordinated alkynes (par-
tion and reductive elimination steps: B. Crociani,
S. Antonaroli, L. Canovese, P. Uguagliati and F. Visentin,
Eur. J. Inorg. Chem., 2004, 732–742.
ticularly Me3Si–CuC–SiMe3) see: A. M. LaPointe and
M. Brookhart, Organometallics, 1998, 17, 1530–1537.
28 This proposal is supported by several observations: the Fp
signal appears at low-field, which is characteristic of cat-
ionic compounds; 31P{1H} NMR shows that the two P
atoms are inequivalent; besides, this compound evolves to
the transmetalation compound 11a.
16 P. Espinet, J. M. Martínez-Ilarduya, C. Pérez-Briso,
A. L. Casado and M. A. Alonso, J. Organomet. Chem., 1998,
551, 9–20.
17 To see bite angle values: M.-N. Birkholz, Z. Freixa and
P. W. N. M. van Leeuwen, Chem. Soc. Rev., 2009, 38, 1099– 29 A. Nova, G. Ujaque, F. Maseras, A. Lledós and P. Espinet,
1118. J. Am. Chem. Soc., 2006, 128, 14571–14578.
18 For X-ray crystal structures with bridging dppf see: 30 C. Amatore, G. Broeker, A. Jutand and F. Khalil, J. Am.
G. Bandoli and A. Dolmella, Coord. Chem. Rev., 2000, 209,
161–196.
19 R. G. Pearson, Inorg. Chem., 1973, 12, 712–713.
20 (a) J. A. Casares, P. Espinet, J. M. Martínez-Ilarduya and
Chem. Soc., 1997, 119, 5176–5185.
31 An excess of 10 eq. of 9 per palladium was added to a solu-
tion of [Pd(dppe)(dba)] in THF at room temperature. After
5 min the 31P{1H} NMR shows the formation of 14a.
Y.-S.
Lin,
Organometallics,
1997,
16,
770–779; 32 H. S. Huh, Y. K. Lee and S. W. Lee, J. Mol. Struct., 2006,
(b) J. A. Casares, S. Coco, P. Espinet and Y.-S. Lin,
Organometallics, 1995, 14, 3058–3067.
21 I. R. Butler, W. R. Cullen, T.-J. Kim, S. J. Rettig and
J. Trotter, Organometallics, 1985, 4, 972–980.
22 In general, one might expect the ionic complex side (two com-
ponents) to be entropically favoured at higher temperature vs.
789, 209–219.
33 E. G. Hope, R. D. W. Kemmitt and A. M. Stuart, J. Chem.
Soc., Dalton Trans., 1998, 3765–3770.
34 B. Cordero, V. Gómez, A. E. Platero-Prats, M. Revés,
J. Echeverría, E. Cremades, F. Barragán and S. Alvarez,
Dalton Trans., 2008, 2832–2838.
the neutral complex one (one component), but the opposite is 35 M. N. Peñas-Defrutos, C. Bartolomé, M. García-Melchor
observed, according to the negative value of entropy. This indi- and P. Espinet, Angew. Chem., Int. Ed., 2019, 58, 3501–3505.
cates that solvation of the ionic complex is a major contri- 36 With these small coupling constants the 117Sn and 119Sn
bution that compensates the formation of two particles. satellites cannot be observed separately.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2020