Intriguing Behavior of an Apparently Simple Coupling Promoter Ligand, PPh2(p-C6H4-C6F5), in Their Pd Complexes

Article abstract of DOI:10.1021/acs.organomet.9b00460

Ligand PPh₂(p-C₆H₄-C₆F₅), L^HF, is an example of monodentate biphenyl phosphine that allows for cis coordination of two phosphines to Pd, as in complex cis-[PdPf₂Pd(L^HF)₂] (A) (Pf = C₆F₅). At 25 °C, complex A undergoes easy reductive elimination to decafluorobiphenyl and, competitively, isomerizes to trans-[Pd(C₆F₅)₂Pd(L^HF)₂] via functionally three-coordinate intermediates cis-[PdPf₂(L^HF)(S)] with the fourth position empty or weakly protected (S= THF, OH₂, or π-aryl). Unexpectedly, the direct reductive C₆F₅-C₆F₅ elimination is faster from the four-coordinate complex A than from the intermediates with only one strong L^HF. The reason for this is that two cis L^HF ligands play the role of a chelate with a large bite angle and some tetrahedral distortion. As a matter of fact, using L^HF in excess (Pd:L ? 1:2), a Pf-Pf coupling barrier ΔG^?_Pf-Pf = 23.1 kcal·mol^-1 is measured, which ranks its efficiency for coupling and formation of the corresponding Pd⁰ catalyst as better than XantPhos or PhPEWO-F and about the same as ^tBuBrettPhos. On the other hand, complex (μ-Cl)₂[Pd₂Rf₂(L^HF)₂] (B) (Rf = C₆F₃Cl₂ = 3,5-dichloro-2,4,6-trifluorophenyl), obtained by reaction of (μ-Cl)₂[Pd₂Rf₂(tht)₂] (tht = tetrahydrothiophene) with L^HF, presents in the ¹⁹F NMR COSY spectrum a very intriguing through-space coupling pattern of the F_ortho atoms of the C₆F₅ group in L^HF and the 3-5-C₆F₃Cl₂ group on Pd. The intermittent coupling mechanism proposed is based on the switching of π-π-stacking of C₆F₅ from one Ph group to another Ph group of L^HF, which gives rise to enantiomers at the chiral P atom. Rotation around the P-biphenyl bond under hindered rotation around the C-C₆F₅ bond produces the intriguing selective coupling observed.

Full text of DOI:10.1021/acs.organomet.9b00460

Article
pubs.acs.org/Organometallics
Cite This: Organometallics XXXX, XXX, XXX−XXX
Intriguing Behavior of an Apparently Simple Coupling Promoter
Ligand, PPh2(p‑C6H4−C6F5), in Their Pd Complexes
‡
María Perez-Iglesias, Rebeca Infante,‡ Juan A. Casares,* and Pablo Espinet*
́
́
IU CINQUIMA/Química Inorganica, Facultad de Ciencias, Universidad de Valladolid, 47011 Valladolid, Spain
S
* Supporting Information
ABSTRACT: Ligand PPh2(p-C6H4−C6F5), LHF, is an example of
monodentate biphenyl phosphine that allows for cis coordination of
two phosphines to Pd, as in complex cis-[PdPf2Pd(LHF)2] (A) (Pf =
C6F5). At 25 °C, complex A undergoes easy reductive elimination to
decafluorobiphenyl and, competitively, isomerizes to trans-[Pd(C6F5)2Pd-
(LHF)2] via functionally three-coordinate intermediates cis-[PdPf2(LHF)-
(S)] with the fourth position empty or weakly protected (S= THF, OH2,
or π-aryl). Unexpectedly, the direct reductive C6F5−C6F5 elimination is
faster from the four-coordinate complex A than from the intermediates
with only one strong LHF. The reason for this is that two cis LHF ligands play the role of a chelate with a large bite angle and
some tetrahedral distortion. As a matter of fact, using LHF in excess (Pd:L ≪ 1:2), a Pf−Pf coupling barrier ΔG‡Pf−Pf = 23.1 kcal·
mol−1 is measured, which ranks its efficiency for coupling and formation of the corresponding Pd0 catalyst as better than
t
XantPhos or PhPEWO-F and about the same as BuBrettPhos. On the other hand, complex (μ-Cl)2[Pd2Rf2(LHF)2] (B) (Rf =
C6F3Cl2 = 3,5-dichloro-2,4,6-trifluorophenyl), obtained by reaction of (μ-Cl)2[Pd2Rf2(tht)2] (tht = tetrahydrothiophene) with
LHF, presents in the 19F NMR COSY spectrum a very intriguing through-space coupling pattern of the Fortho atoms of the C6F5
group in LHF and the 3-5-C6F3Cl2 group on Pd. The intermittent coupling mechanism proposed is based on the switching of
π−π-stacking of C6F5 from one Ph group to another Ph group of LHF, which gives rise to enantiomers at the chiral P atom.
Rotation around the P−biphenyl bond under hindered rotation around the C−C6F5 bond produces the intriguing selective
coupling observed.
INTRODUCTION
■
Recently, we reported a procedure to evaluate the ability of
different ligands to facilitate difficult C−C couplings in Pd. It
uses the so-called couplimeter complex cis-[PdPf2(THF)2] (1)
(Pf = C6F5; THF = tetrahydrofuran), which allows ΔG‡
,
Pf−Pf
the activation energy of the coupling step producing Pf−Pf
upon addition of the ligand, to be measured experimentally.
This leads to ranking the ligands according to their
1
t
corresponding ΔG‡
values. Ligands BuXPhos (ΔG‡
Pf−Pf
Pf−Pf
= 22.3
= 21.8 kcal·mol−1 at 0 °C) and PhPEWO-F (ΔG‡
Pf−Pf
kcal·mol−1 at 25 °C) are among the most active ligands for
coupling at 0−25 °C, certainly more efficient than the typical
large bite angle ligand XantPhos (ΔG‡Pf−Pf = 24.2 kcal·mol−1 at
25 °C). The coupling product of the reaction, Pf−Pf, is only
hardly reactive; hence, complex 1 can be used as a convenient
precursor to generate [Pd0Ln] catalysts in situ.1
Figure 1. Structures of tBuXPhos, XantPhos, and the partially
fluorinated PhPEWO-F and PPh2(o-C6H4−C6F5) (LHF) ligands.
atom of the distal aryl group is located at a short distance of
the Pd atom; similarly, in some PEWO-F complexes, one
carbon atom of the olefin group is a short distance from the Pd
atom.7 These common features, suggestive of some similarity
of the respective EWO or biphenyl interactions with the Pd
center, call for more detailed examination.
The electron density on the biphenyl of PR2(biaryl) ligands
has been reported to strongly affect the rate of reductive
elimination from palladium(II) complexes. The frequent
The ability of RPEWO-F ligands to induce C−C coupling is
due to the presence of a strongly electron-withdrawing double
bond: it lowers the coupling activation energy by withdrawing
the increasing electron density on Pd as the reduction from
PdII to Pd0 progresses.2−6 In fact, the ligand PhPEWO-F,
shown in Figure 1, left, induces much faster coupling than the
related nonfluorinated PhPEWO-H, with H atoms instead of F
atoms, because its olefin function is a stronger EWO group.1
In X-ray diffraction structures of “three-coordinate”
palladium complexes with biphenyl phosphines, one carbon
Received: July 8, 2019
© XXXX American Chemical Society
A
DOI: 10.1021/acs.organomet.9b00460
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
presence of electron-donating groups on the distal arene is
detrimental for the reductive elimination rate.8 We hypothe-
sized that ligand fluorination at the distal aryl of the biphenyl
might increase their similarity with our PEWO-F ligands.
Although some partially fluorinated biaryl phosphines have
already been reported and their catalytic performance was not
particularly good when compared with that of nonfluorinated
analogues,9−13 these results are not necessarily informative
about the reductive elimination step. In this context, we
decided to synthesize and examine the behavior of a new
PR2(biaryl) ligand fluorinated at the distal aryl group, LHF
(Figure 1, right) and immediately set to estimate its coupling
Figure 2. (a) Most stable conformation of free LHF. (b) Less stable
conformation for free LHF, which can be preferred in complexes.
potential, measuring ΔG‡
with the standard protocol
Pf−Pf
reported in ref 1. Unexpectedly, LHF behaved very differently
from other biphenyl phosphines such as tBuXPhos or
diphosphines such as XantPhos that were previously submitted
to this protocol.1 In these, the addition of free ligand to cis-
[PdPf2(THF)2] above the L:Pd = 1:1 ratio did not alter the
coupling rate measured for L:Pd = 1:1. In contrast, for LHF,
successive additions of free ligand over the 1:1 ratio produced
progressively attenuated increases of the coupling rate. This
effect ceased for a L:Pd ratio well above 2:1, when the coupling
reaction reached was about 3-fold the initial one. This behavior
is compatible with the formation of a mixture of [PdPf2(LHF)]
and cis-[PdPf2(LHF)2], at variance with tBuXPhos or XantPhos,
which form only cis-[PdPf2(L)] in both cases. Yet, the
accelerating effect upon addition of free ligand is unusual on
an isolated coupling process: deceleration or no effect is
usually observed for PdII.14 This work aims at supporting and
explaining in detail the different behavior of LHF compared to
that of many other biarylphosphines.
stacking is observed, with Pfcentroid−Phcentroid distance = 3.654
Å. The less stable atropisomer, shown in Figure 2b, adopts a
conformation with the biaryl roughly opposite to the lone
electron pair of the phosphorus atom. In this conformation, the
two P···Fortho distances are 4.210 and 4.227 Å, minimizing the
through-space contribution to coupling. Note also in Figure 2b
the existence of a very efficient Pf−Ph π-stacking, with
Pfcentroid−Phcentroid distance = 3.426 Å. As far as this π-stacking
is operative, it makes the P atom temporarily chiral; if the Pf−
Ph π-stacking is moved to the other Ph group, it produces the
other enantiomer. This is relevant later in the LHF Pd
complexes.
Coordination of LHF to Palladium(II). Scheme 1 shows
the syntheses of complexes 2−5, prepared for the study. Other
complexes appearing or being formed in solution are discussed
later, as they are required.
Scheme 1. Synthesis of Complexes 2−5
RESULTS AND DISCUSSION
■
Synthesis of the Ligand. LHF was synthesized from 2′-
lithium-2,3,4,5,6-pentafluoro-1,1′-biphenyl and PPh2Cl and
was chemically characterized in solution by H, 19F, 31P, and
1
13C NMR as well as computationally. Its 19F NMR spectrum
shows the typical AA′XX′Z spin system found for symmetrical
C6F5 groups. The 31P{1H} NMR signal is a triplet by coupling
to the F1 and F5 ortho atoms (JF−P = 27 Hz) as shown by a
19F−31P HSQC correlation. This supports that a fast rotation
of the fluoraryl group around the C−C bond is taking place in
solution at 25 °C. Values of about 30 Hz are typical for
3J(Fortho−P) couplings (e.g., P(C6F5)3 or P(C6F5)2Ph),15,16 and
smaller values are found in other fluoroaryl phosphines such as
PhPEWO-F (3JF−P ≈ 15 Hz)17 and in fluoroalkylphosphines.18
Coupling constants 5J(19F−31P) are usually very small or
unobservable, but exceptions are dicyclohexyl(2′,6′-difluoro-
[1,1′-biphenyl]-2-yl)-phosphine (JF−P = 24 Hz) and LHF.11
The simplest explanation for these anomalous values is that
there is some “through-space” coupling contribution to
coupling, strongly dependent on the distance between
Ligand LHF (1 eq per palladium) reacts with PdCl2(NCMe)2
to produce the dimer (μ-Cl)2[PdCl(LHF)]2 (2). The use of
excess of LHF yields the monomer trans-[PdCl2(LHF)2] (3),
which is better obtained by the reaction of [PdCl2(1,5-COD)]
with 2 eq of LHF. Both complexes show in the IR spectrum the
typical ν(35Cl−Pd) stretching band of terminal Pd−Cl bonds
close to 360 cm−1 and the ν(37Cl−Pd) shoulder at 353 cm−1.
The dimer shows additional absorptions for bridging chlorides,
at lower wavenumbers (304 and 294 as well as 272 and 264
cm−1).
The X-ray structure of 2 shows a coplanar arrangement of
their two square-planar halves.22 The LHF conformation is
somehow similar to the most stable conformation of the free
phosphine, at least in the solid state: the modest out-of-plane
steric hindrance of the chloro ligands allows for the distal
pentafluorophenyl group of the biphenyl to occupy the space
above and below the Pd coordination planes, lying almost
parallel to them (Figure 3). Bond distances and angles are
conventional, but it is worth noting the short distance from Pd
to the Cipso (C1) of Pf (3.142 Å). This kind of biphenyl
arrangement with similar distances has been studied for linear
nuclei,19−21 in addition to the J scalar coupling contribution.
5
In the palladium complexes of LHF (see below), this 19F−31P
coupling is not observed.
The structure of LHF in the gas phase has been calculated by
DFT (wB97XD/6-31G// wB97XD /SDD, details in Support-
ing Information (SI)). Its most stable atropisomer is shown in
Figure 2a and shows the biaryl group roughly oriented toward
the lone pair of the P atom. In this conformation, one of the
P···Fortho distances is as short as 3.305 Å, close enough to
display a through-space contribution to coupling. A Pf−Ph π-
B
DOI: 10.1021/acs.organomet.9b00460
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Figure 3. ORTEP diagram (50%) of (μ-Cl)2[Pd2Cl2(LHF)2] (2).
Bond distances (Å): Pd−P = 2.226; Pd−C1 = 3.141; Pd−Cl1 =
2.272; Pd−Cl2 = 2.411; Pd−Cl3 = 2.320.
Figure 4. ORTEP diagram (50%) of (μ-Cl)2[PdRf(LHF)]2 (4·CHCl3,
disordered CHCl3 omitted for clarity). Bond distances (Å): Pd−C1 =
1.995; Pd−P = 2.262; Pd−Cl1 = 2.407; Pd−Cl2 = 2.388. Blue arrows
point to π−π-stacking zones: Ph−Rf stacking distance between
centroids = 3.117 Å; Ph−Pf stacking distance between centroids =
3.212 Å. Above: Another view of the angular arrangement of the two
coordination square planes.
AuI complexes (note that AuI and PdII have almost identical
radii).23,24
The 31P{1H} NMR spectrum of 2 at 25 °C displays one
5
singlet (ν1/2 = 5 Hz), where no JF−P is observed. In the 19F
NMR spectrum, the two Fortho nuclei and the two Fmeta nuclei
are equivalent, meaning that there is fast rotation of the C6F5
group about the C−C bond. Obviously, this requires that the
Pd coordination plane and the distal Pf group, observed in
close parallel arrangement in the X-ray structure, get separated
in solution in order to allow for rotation around the P−C and
C−Pf bonds. The coalescence temperature is not still reached
down to −30 °C, suggesting that the rotational barriers in 2
must be very small.
Similar 19F NMR behavior is found for the monomeric trans-
[PdCl2(LHF)2] (3). The trans structure of 3 was assigned
based on the observation in the IR spectrum of νassim(Pd−35Cl)
and νassim(Pd−37Cl) bands at 360 and 351 cm−1, respectively.
The dimeric complex (μ-Cl)2[PdRf(LHF)]2 (4) (Rf =
C6F3Cl2, 3,5-dichloro-2,4,6-trifluorophenyl) is easily prepared
by reacting the precursor (μ-Cl)2[PdRf(tht)]2, (tht =
tetrahydrothiophene) with LHF (Pd:LHF = 1:1). It shows IR
bridging chloride bands at lower wavenumbers, at 306 and 296
as well as 284 and 271 cm−1. The X-ray structure of 4·CHCl3 is
shown in Figure 4. The complex displays two square-planar
units making a 135.8° dihedral angle at the (μ-Cl)2 edge. Each
square-planar complex has conventional Pd−Cl (2.388 Å),
Pd−P (2.262 Å), and Pd−C (1.995 Å) distances and angles
and no significant distortion from planarity.25 The Rf group
imposes steric constraints at the coordination plane zone,
which force a conformational change on LHF, pushing the
pentafluorophenyl ring away from the palladium coordination
plane. This different rearrangement from 2 does not alter
significantly the bond distances and angles at the coordination
plane, but the LHF coordinated ligand is involved in two
relatively strong π-stacking interactions: one intraligand Ph−Pf
at 3.117 Å between centroids, reminiscent of what happens in
the free ligand and one extraligand Ph−Rf at 3.212 Å.
Figure 5. Above: 19F-COSY of 4 in CDCl3 at 25 °C. Below: Effect of
selective irradiations on the deceptive F1 triplet.
very well established that the rotation of Rf groups about the
C−Pd bond is severely restricted in complexes with nonflat
ligands that are not easily dissociable (e.g., PPh3) cis-
coordinated relative to Rf.26 Hence, because the Pd
coordination plane of 4 is not a symmetry plane for the
molecule, the two F1 atoms should be anisochronous.
The 31P{1H} NMR spectrum of 4 shows a triplet (4JF−P = 11
Hz), as expected for a phosphine cis to Rf, provided that the
two F1 were equivalent. Consequently, a mechanism different
from the hindered rotation about the Rf−Pd bound must exist.
The 19F-COSY of 4 in CDCl3 at 25 °C (Figure 5) confirms
the internal F−F coupling between the fluorine nuclei of Pf
and the absence of internal F−F coupling in Rf. Additionally, it
reveals the existence of unexpected (looking at the X-ray
structure in Figure 4) through-space JF−F coupling (J ≈ 10 Hz)
between F1 and F3.21 This produces a deceptive triplet for the
F1 signal. Double irradiation experiments (19F{31P} and 19F-
{19F3}) reveal that this deceptive multiplicity is due to coupling
The 19F NMR spectrum of compound 4 (see Figure 5 for
spectra and F labeling) is surprising and intriguing. It shows
the expected number of signals for an Rf group with chemically
equivalent halves (two signals in a ratio of 2:1 for F1 and F2)
and for a Pf group with chemically equivalent halves (three
signals in a 2:2:1 ratio for F3:F4:F5). Even assuming that the
two Pd coordination planes of the dimer will fast average to
coplanar in the NMR time scale, the chemical equivalence of
the two F1 atoms in the Rf group is something unusual: It is
C
DOI: 10.1021/acs.organomet.9b00460
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
to the phosphorus (4JF−P = 11 Hz) and to only one F3 atom (J
≈ 10 Hz) of the Pf group (JF−F = 10 Hz) (note that
paradoxically both F3 atoms are spectroscopically equivalent).
Thus, the equivalence mechanism has to be able to explain the
through-space F1−F3 coupling involving only one pair of atoms
at a time and still justify the chemical equivalence of the two
halves of the Pf and Rf groups. In other words: what kind of
fluxional process produces chemical equivalence of the two F3
and still does not produce coupling of each F1 to both F3? This
suggests that the process must be explained under restricted
rotation of Pf around the Pf−C bond or at least a higher
rotational barrier than the fluxionality leading to the through-
space couplings observed.
A closer inspection of the X-ray structure of complex 4
shows that the π-stacking interactions noted above make the P
atom a chiral center. The choice of one or the other Ph ring to
interact with the Pf group determines the P conformation. This
is already seen in the less stable atropisomer of free LHF, where
the lone pair plays the role of Pd in the complex (Figure 2b).
In the complex, the hindrance produced by the π−π-stacked
chiral conformation allows that only one F3 can get close to the
F1 at that side of the coordination plane. The NMR behavior
observed 19F can then be explained as associated with the
switching between enantiomers. The slippage of the Pf group
from Ph1 to Ph2, involving progressive weakening of the π−π-
stacking with Ph1 and strengthening of the π−π-stacking with
Ph2, occurs along a crowded pathway where rotation around
the Pf−C bond is restricted. In the new isomer, the Pf group is
at the other side of the coordination plane (rotation around
the Pd−P bond is forced in the same movement), and it is the
other F1/F3 pair that can get close to produce through-space
coupling. It is interesting to note the fact that the rotation
around the P−C(biphenyl) bond occurs with the restriction
that the Pf−C rotation cannot occur, whereas the rotation of
the phosphine around the P−Pd bond maintains the hindrance
to rotation of the Rf group around the Rf−Pd bond (Scheme
2).
4 with a small variation due to the presence of two LHF ligands:
The 31P NMR shows a triplet due to the coupling with two F1
atoms, and the F1 signals appear as deceptive quintuplets in the
19F NMR spectrum because of coupling of two P atoms and
coincidence of JF−P and JF−F values.
Reductive Elimination Reaction. The reductive elimi-
nation of [PdPf2(THF)2] (1) was studied in toluene by adding
different proportions of LHF and monitoring the formation of
Pf−Pf by 19F NMR. The behavior was anomalous to the case
t
of BuXPhos, where the coupling rate was independent of the
Pd:L used.1
The study required previously being able to identify in
solution some complexes by NMR: cis-[PdPf2(LHF)2] (6),
which could also be isolated and its structural details studied
by a combination of defective X-ray diffraction results asa base
for DFT calculation; trans-[PdPf2(LHF)2] (8), which was also
prepared as a stable solid; and a number of complexes with just
one LHF per Pd, some of them observable under appropriate
conditions, which are species considered functionally three-
coordinate.27 In our case, these can have the fourth
coordination position empty or weakly protected by
interaction with the distal aryl of the biphenyl or with an
easily displaceable molecule (THF, OH2, which are weak
ligands for soft PdII) so that their conversion to that empty
position is not needed or requires very little energy. This
means that for kinetic purposes, they can be approximated
globally, with acceptable approximation, as all being the same
species cis-[PdPf2(LHF)(S)] (7). In this respect, no assumption
needs to be made on whether they really dissociate to the real
three-coordinate or they couple directly from the tetracoordi-
nate species 7 with only one LHF ligand. The kinetic treatment
is discussed in detail in the Supporting Information.
Addition to complex 1 in toluene of excess LHF produces
immediately cis-[PdPf2(LHF)2] (6). This complex is unstable in
solution at room temperature (undergoing reductive elimi-
nation of Pf−Pf), but it can be crystallized from toluene at −35
°C as a colorless solid. The poor quality of the crystals
obtained affects the quality of the X-ray diffraction data and the
reliability of the structure of 6·toluene obtained. This defective
structure (suppressing the toluene molecule) was DFT
optimized in the gas phase, and the valuable geometrical
information obtained (coincident with the X-ray data except
for moderate variations in bond distances and angles) is shown
in Figure 6. The structure confirms that two LHF ligands can
occupy cis positions in Pd. The complex displays tetrahedrally
Scheme 2. Fluxional Process Observed for Complexes 4 and
a
5
a
For simplicity, only the Fortho atoms (F1 and F3) are represented.
Green bonds indicate allowed rotations, and red bonds indicate
restricted rotations.
As the spin−spin connection is not lost along the fluxional
movement because there is no phosphine dissociation, the
observed J ≈ 10 Hz is the time averaged value varying along
the fluxional process from a nondetermined maximum value to
zero. As far as we are aware, nothing like this racemization
mechanism associated with π−π-stacking slippage has been
reported before, although it must be taking place in many
related systems, being unobservable or passing unnoticed.
Complex 4 reacts with an excess of LHF to produce trans-
[PdRfCl(LHF)2] (5). The NMR spectra of 5, following the
same clockwork fluxional behavior, behave as just discussed for
Figure 6. DFT capped-stick structure of cis-[PdPf2(LHF)2], optimized
from a defective X-ray solution of 6·toluene (see SI). It shows some
tetrahedral distortion of the square plane induced by the biaryl
substituents on P.
D
DOI: 10.1021/acs.organomet.9b00460
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
distorted square-planar cis geometry (the two trans Pf−Pd−P
angles are 164.0°). A C2 axis makes the two Pf groups and the
two LHF phosphines symmetry equivalent. The conformation
of LHF shows the biphenyl with its Pf ring away from the
palladium plane, as expected. The repulsion between the two
bulky phosphine ligands forces a P−Pd−P angle of 104.8°, a
bit larger than the same angle in [PdRf2(XantPhos)]·2CH2Cl2
(103.5°).28 Thus, the effect of the two voluminous phosphines
forcing a large P−Pd−P angle is higher than that of XantPhos
(a typical chelate with a large bite angle). Consistently, the C−
Pd−C angle in 6 (81.21°) is smaller than 90° but a bit larger
than in the rigid complex with XantPhos (79.6°), because in
the case of XantPhos, the planar square plane is less distorted.
The coupling promoter effect of ligands with large bite angles
is well-known. In theoretical calculations,14a not only the
reduction of the C−Pd−C angle but also a tetrahedral
distortion of the square plane are features observed in the
structural evolution during C−C coupling in case that the four
ligands are conserved during the reaction.29 Both features are
incipiently observed for cis-[PdPf2(LHF)2], whereas the
tetrahedral distortion effect is not observed in the complex
with XantPhos, and perhaps, this explains the significantly
higher coupling efficiency of 2LHF (ΔG‡Pf−Pf = 23.1 kcal·mol−1
isomerization of 6 to trans-[PdPf2(LHF)2] (8). The reductive
elimination behavior at 25 °C can be analyzed based on the
existence of these equilibria and the kinetic monitoring of the
processes in different LHF:Pd ratios. Data related to the
characterization of these equilibria at different temperatures are
discussed in the SI.
(1) With a large excess of LHF([LHF]:[PdPf2(THF)2] ≥
4:1), initially the only significant Pd species in solution is cis-
[PdPf2(LHF)2] (6, red dots) plus at least 2 eq of free LHF and 2
eq of free THF). The reductive elimination to Pf−Pf (blue
dots) is the only process observed at 25 °C. The
concentration/time plots (Figure 7 fit very well first order
at 25 °C, see later) compared to that of XantPhos (ΔG‡
=
Pf−Pf
24.2 kcal·mol−1 at 25 °C), in spite of the smaller C−Pd−C
angle defined by XantPhos in its ground state structure.
On the other hand, trans-[PdPf2(LHF)2] (8) was easily
prepared by reaction of cis-[PdPf2(SMe2)2]30 with a stoichio-
metric amount of LHF in dichloromethane and isolated as a
stable solid (details in SI). With the spectroscopic data of 1, 6,
and 8 in hand, the evolution of the reactions 1 + nLHF could be
studied.
The NMR spectra of mixtures of 1 and LHF in different
proportions, prepared and recorded at −20 °C in order to
prevent coupling, reveal the existence of the complexes and
equilibria shown in Scheme 3. Above 0 °C, two processes are
observed: the reductive elimination to Pf−Pf and the
Figure 7. Concentration/time plots and best least-squares fitting
(continuous line) for reductive elimination of 6 in toluene under
selected concentrations of LHF at 298 K. Red dots: cis-[PdPf2(LHF)2]
(6); blue dots: Pf−Pf. Starting conditions: (a) [1] = 0.02 M + [LHF
]
= 0.08 M (1:4), kobs = 7.7 × 10 −5 8 × 10 −8 s−1; (b) [1] = 0.018 M
+ [LHF] = 0.14 M (1:8), kobs = 7.44 × 10 −5 1 × 10 −7 s−1; (c) [1] =
−5
−7
0.018 M + [LHF] = 0.22 M (1:12), kobs = 7.36 × 10
4 × 10
s−1; (d) [1] = 0.018 M + [LHF] = 0.22 M (1:12) and [FC6H4I] =
0.054 M (1:3), kobs = 7.38 × 10−5 1.69 × 10−7 s−1.
Scheme 3. Equilibria Observed in Toluene Solution at −20
a
°C and Evolution to 8 at 0 °C
kinetics (kelim = 7.7 × 10−5 s−1, ΔG‡
= 23.1 kcal·mol−1 at
Pf−Pf
25 °C), and the observed rate does not change when
increasing the concentration of LHF, confirming that the
coupling is taking place on the four-coordinated complex,
induced by the large bite angle defined by the two phosphines
in cis.
Thus, in this proportion, LHF performs better to form the
corresponding Pd0 catalyst than XantPhos or PhPEWO-F and
about the same as tBuBrettPhos.2 The 31P spectra in toluene at
25 °C show independent signals for free LHF (δ = −11.1 ppm)
or coordinated to Pd0 (δ = 45.6 ppm), and integration
supports the composition Pd(LHF)3 for the Pd0 complex,
regardless of the excess of LHF used, as far as it is ≥3.
It is also worth noting that competing formation of PfH,
observed for some other ligands,1 was not detected in these
experiments. Furthermore, an experiment in the presence of 4-
fluoroiodobenzene as a reoxidant to prevent formation of
Pd(LHF)3 gave the same kelim, demonstrating that the low
a
Note that “dry” toluene contains a small amount of residual water
(about 5 pm).31 The initial complex 1 is represented as cis-
[PdPf2(THF)(S)] to include molecules where THF has been
substituted by water in the solvent. For details of the kinetic study,
see SI.
ΔG‡
observed is due to the destabilization of the ground
Pf−Pf
state structure with two LHF ligands and not the stabilization of
the Pd0 complex with an extra ligand.
E
DOI: 10.1021/acs.organomet.9b00460
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
(2) For [LHF]:[PdPf2(THF)2] = 2:1 or with the isolated
complex cis-[PdPf2(LHF)2] (6), the evolution shows more
complex kinetics. In addition to the coupling product, Pf−Pf,
the reaction produces also: (i) isomerization to trans-
[PdPf2(LHF)2] (8) by a dissociative mechanism,32 which is
more easily accessible for crowded complexes, and (ii) if 6 has
been made in situ from 1, and THF is present in the solution,
the reaction regenerates some complex [PdPf2(THF)2] (1).
The latter is as a consequence of the capability of palladium(0)
complex Pd(LHF)2 to sequester LHF, forming Pd(LHF)3, so that
some 1 is formed from 6 (Scheme 3). The values obtained
from the 19F integrals corresponding to the Pd0 complex with
respect to the internal standard suggest that the number of
coordinated phosphines is between 2 and 3. In the presence of
4-fluoroiodobencene as a reoxidant, the fast formation of trans-
[Pd(p-FC6H4)I(LHF)2] precludes formation of of Pd(LHF)3
and also prevents formation of 1 from 6.
process. The most plausible isomerization mechanism is via a
three-coordinate intermediate 7empty, which means that this
intermediate is energetically accessible from the other
functionally three-coordinate species and that the reductive
elimination from 7empty is not extremely fast. On the other
hand, the sum of contributions of all the functionally three-
coordinate species (the cis species with one LHF) produces (for a
ratio of 1:LHF < 1:1.5) an apparent reductive elimination rate
that is less than one-third the coupling rate from the cis species
with two LHF (6) (2.12 × 10−5 vs 7.7 × 10−5 s−1). More
detailed discussion is available in the SI).
CONCLUSIONS
■
Ligand LHF is an example of monodentate biphenyl phosphine
with a size that still allows for cis coordination to Pd of two
phosphines. In the absence of sufficient extra free LHF, there
are equilibria of LHF dissociation or displacement by a weak
smaller ligand such as THF or OH2 or by the distal aryl of the
biphenyl of the phosphine that remains coordinated. When the
two phosphines LHF are coordinated in cis positions, the
interligand repulsions close the Pf−Pf angle to about 81.21°
and induce a tetrahedral distortion of the square-planar
coordination. These two distortions destabilize the ground
state, making the reductive elimination transition state more
accessible than expected. As a matter of fact, the use of LHF in
(3) Finally, for [LHF]:[PdPf2(THF)2] = n:1 with 2 > n > 1, a
mixture of complexes is formed in solution, which evolves as
shown in Figure 8: cis-[PdPf2(LHF)2] (6) is consumed
excess makes it very efficient (ΔG‡
= 23.1 kcal·mol−1),
Pf−Pf
better than XantPhos or PhPEWO-F and about the same as
tBuBrettPhos.1
The tetrahedral distortion caused by overcrowding seems to
be an important reason for the easier coupling. This suggests
that the design of chelating ligands (not necessarily bulky)
forcing tetrahedral distortions at Pd might be catalytically
productive. Other studies on structurally related ligands are in
progress.
Figure 8. Concentration/time plots (dots) and best least-squares
fitting (continuous lines) for the reductive elimination in toluene (0.5
mL) from [1] = 0.019 M + [LHF] = 0.030 M (1:1.5) at 298 K,
monitored by 19F NMR. Red dots: cis-[PdPf2(LHF)2] (6); purple dots:
cis-[PdPf2(LHF)(THF)] (7); green dots: trans-[PdPf2(LHF)2] (8);
blue dots: Pf−Pf; orange dots: [PdPf2(THF)2] (1). Colored lines
represent the kinetic simulation obtained by fitting the equations
optimized with different values for kobs from 6 and for the functionally
three-coordinate species as a whole.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.organo-
met.9b00460.
■
S
General methods. Synthesis of new compounds.
Determination of rate constants. Other studies in
solution. Computational section. Experimental proce-
dure for X-ray crystallography. NMR spectra. IR spectra
(PDF)
relatively fast, during the first 30 min, in which Pf−Pf is
formed slowly. Also, cis-[PdPf2(LHF)(THF)] (7), in equili-
brium with 6 and LHF, is consumed at a slower rate (Figure 8).
The main product of the reaction is trans-[PdPf2(LHF)2] (8),
and during the experiment, cis-[PdPf2(THF)2] (1) is also
formed. Knowing the rate constants for the formation of Pf−Pf
from 6 and from 1, it is possible to calculate kobs for the
Accession Codes
CCDC 1938395−1938396 contain the supplementary crys-
tallographic data for this paper. These data can be obtained
free of charge via www.ccdc.cam.ac.uk/data_request/cif, or by
emailing data_request@ccdc.cam.ac.uk, or by contacting The
Cambridge Crystallographic Data Centre, 12 Union Road,
Cambridge CB2 1EZ, UK; fax: +44 1223 336033.
reductive elimination from 7 (no assumption about the
−5
pathway was made); a value kobs = 2.3 × 10
s−1 was
obtained. Kinetic simulation visualizes how fast Pf−Pf would
be formed if the rates of the reductive elimination were
approximately identical for 6 and for 7 (black lines in Figure
8). The rate of formation of Pf−Pf would be noticeably faster
than observed experimentally.
In summary, the experimental observations fit well the
mechanistic Scheme 3 well and show that, in the absence of
excess LHF, which allows for LHF dissociation from 6, the
isomerization of 6 to 8 is faster than the reductive elimination
AUTHOR INFORMATION
Corresponding Authors
*Tel: (+34)983423231; E-mail: espinet@qi.uva.es. (P.E.)
*Tel: (+34)983184623; E-mail: casares@qi.uva.es; Web:
http://cinquima.uva.es. (J.A.C.)
■
ORCID
́
María Perez-Iglesias: 0000-0002-5391-264X
Rebeca Infante: 0000-0002-3754-4189
F
DOI: 10.1021/acs.organomet.9b00460
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Structural Analysis of Biarylphosphine Ligands in Arylpalladium
Bromide and Malonate Complexes. Organometallics 2017, 36, 129−
135. (h) Yadav, M. R.; Nagaoka, M.; Kashihara, M.; Zhong, R.-L.;
Miyazaki, T.; Sakaki, S.; Nakao, Y. The Suzuki−Miyaura Coupling of
Nitroarenes. J. Am. Chem. Soc. 2017, 139, 9423−9426.
(8) Arrechea, P. L.; Buchwald, S. L. Biaryl Phosphine Based Pd(II)
Amido Complexes: The Effect of Ligand Structure on Reductive
Elimination. J. Am. Chem. Soc. 2016, 138, 12486−12493.
(9) Leroux, F. R.; Bonnafoux, L.; Heiss, C.; Colobert, F.; Lanfranchi,
D. A. A Practical Transition Metal-Free Aryl-Aryl Coupling Method:
Arynes as Key Intermediates. Adv. Synth. Catal. 2007, 349, 2705−
2713.
Juan A. Casares: 0000-0001-8833-2066
Pablo Espinet: 0000-0001-8649-239X
Author Contributions
The manuscript was written through contributions of all
authors. All authors have given approval to the final version of
the manuscript.
Author Contributions
‡M.P.-I. and R.I. contributed equally.
Notes
The authors declare no competing financial interest.
(10) Olsen, E. P. K.; Arrechea, P. L.; Buchwald, S. L. Mechanistic
Insight Leads to a Ligand Which Facilitates the Palladium-Catalyzed
Formation of 2-(Hetero)Arylaminooxazoles and 4-(Hetero)-
Arylaminothiazoles. Angew. Chem., Int. Ed. 2017, 56, 10569−10572.
(11) Milne, J. E.; Buchwald, S. L. An Extremely Active Catalyst for
the Negishi Cross-Coupling Reaction. J. Am. Chem. Soc. 2004, 126,
13028−13032.
(12) Mak, A. M.; Jong, H.; Robins, E. G.; Sullivan, M. B.; Lim, Y. H.;
Yang, Y.; Johannes, C. W. Mechanistic Insights and Implications of
Dearomative Rearrangement in Copper-Free Sonogashira Cross-
Coupling Catalyzed by Pd-Cy*Phine. Organometallics 2016, 35,
1036−1045.
ACKNOWLEDGMENTS
Financial support by the Spanish MINECO (CTQ2016-
80913-P) and the Junta de Castilla y Leon (JCyL VA051P17,
VA062G18 and M.P.-I. predoctoral grant) is gratefully
acknowledged. This paper is dedicated to Prof. Miguel A.
Ciriano (University of Zaragoza) on the occasion of his 70th
birthday.
■
́
REFERENCES
■
(1) Gioria, E.; delPozo, J.; Martínez-Ilarduya, J. M.; Espinet, P.
Promoting Difficult Carbon−Carbon Couplings: Which Ligand Does
Best? Angew. Chem., Int. Ed. 2016, 55, 13276−13280.
(2) delPozo, J.; Gioria, E.; Espinet, P. Olefin Insertion Versus Cross-
Coupling in trans-[Pd(Ar)X(AsPh3)2] Complexes (X = I, F, CF3)
Treated with a Phosphine-EWOlefin Ligand to Induce Ar−X
Coupling. Organometallics 2017, 36, 2884−2890.
(3) Gioria, E.; Martínez-Ilarduya, J. M.; García-Cuadrado, D.;
Miguel, J. A.; Genov, M.; Espinet, P. Phosphines with Tethered
Electron-Withdrawing Olefins as Ligands for Efficient Pd-Catalyzed
Aryl-Alkyl Coupling. Organometallics 2013, 32, 4255−4261.
(4) Williams, D. B. G.; Shaw, M. L. P-Alkene Bidentate Ligands: An
Unusual Ligand Effect in Pd-Catalysed Suzuki Reactions. Tetrahedron
2007, 63, 1624−1629.
(5) Luo, X.; Zhang, H.; Duan, H.; Liu, Q.; Zhu, L.; Zhang, T.; Lei,
A. Superior Effect of a π-Acceptor Ligand (Phosphine-Electron-
Deficient Olefin Ligand) in the Negishi Coupling Involving Alkylzinc
Reagents. Org. Lett. 2007, 9, 4571−4574.
(13) Baba, K.; Tobisu, M.; Chatani, N. Palladium-Catalyzed Direct
Synthesis of Phosphole Derivatives from Triarylphosphines through
Cleavage of Carbon-Hydrogen and Carbon-Phosphorus Bonds.
Angew. Chem., Int. Ed. 2013, 52, 11892−11895.
́
(14) (a) Perez-Rodríguez, M.; Braga, A. A. C.; García-Melchor, M.;
́
Perez-Temprano, M. H.; Casares, J. A.; Ujaque, G.; de Lera, A. R.;
́
Alvarez, R.; Maseras, F.; Espinet, P. C-C Reductive Elimination in
Palladium Complexes, and the Role of Coupling Additives. A DFT
Study Supported by Experiment. J. Am. Chem. Soc. 2009, 131, 3650−
3657. (b) Ozawa, F. Reductive elimination. Fundamentals of Molecular
Catalysis 2003, 3, 479. (c) Gillie, A.; Stille, J. K. Mechanisms of 1,1-
Reductive Elimination from Palladium. J. Am. Chem. Soc. 1980, 102,
4933−4941. (d) Ozawa, F.; Ito, T.; Yamamoto, A. Mechanism of
thermal decomposition of trans-diethylbis(tertiary phosphine)-
palladium(II). Steric effects of tertiary phosphine ligands on the
stability of diethylpalladium complexes. J. Am. Chem. Soc. 1980, 102,
6457−6463.
(15) Choukroun, R.; Lorber, C.; Lepetit, C.; Donnadieu, B.
Reactivity of [Cp2Ti(CO)2] and B(C6F5)3: Formation of the
Acylborane Complexes [Cp2Ti(CO)(η2-OCB(C6F5)3)] and [Cp2Ti-
(THF)(η2-OCB(C6F5)3)]. Organometallics 2003, 22, 1995−1997.
(16) Mehta, M.; Garcia de la Arada, I.; Perez, M.; Porwal, D.;
Oestreich, M.; Stephan, D. W. Metal-Free Phosphine Oxide
Reductions Catalyzed by B(C6F5)3 and Electrophilic Fluorophospho-
nium Cations. Organometallics 2016, 35, 1030−1035.
(6) Tuxworth, L.; Baiget, L.; Phanopoulos, A.; Metters, O. J.;
Batsanov, A. S.; Fox, M. A.; Howard, J. A. K.; Dyer, P. W. Phosphine-
Alkene Ligand-Mediated Alkyl-Alkyl and Alkyl-Halide Elimination
Processes from Palladium(II). Chem. Commun. 2012, 48, 10413−
10415.
(7) (a) Wolfe, J. P.; Tomori, H.; Sadighi, J. P.; Yin, J.; Buchwald, S.
L. Simple, Efficient Catalyst System for the Palladium-Catalyzed
Amination of Aryl Chlorides, Bromides, and Triflates. J. Org. Chem.
2000, 65, 1158−1174. (b) Christmann, U.; Vilar, R.; White, A. J. P.;
Williams, D. J. Synthesis of Two Novel Dinuclear Palladium(I)
Complexes and Studies of Their Catalytic Activity in Amination
Reactions. Chem. Commun. 2004, 4, 1294−1295. (c) Christmann, U.;
Pantazis, D. A.; Benet-Buchholz, J.; McGrady, J. E.; Maseras, F.; Vilar,
R. Experimental and Theoretical Investigations of New Dinuclear
Palladium Complexes as Precatalysts for the Amination of Aryl
Chlorides. J. Am. Chem. Soc. 2006, 128, 6376−6390. (d) Fors, B. P.;
Watson, D. A.; Biscoe, M. R.; Buchwald, S. L. A Highly Active
Catalyst for Pd-Catalyzed Amination Reactions. J. Am. Chem. Soc.
2008, 130, 13552−13554. (e) Milner, P. J.; Maimone, T. J.; Su, M.;
(17) Gioria, E.; Martínez-Ilarduya, J. M.; García-Cuadrado, D.;
Miguel, J. A.; Genov, M.; Espinet, P. Phosphines with Tethered
Electron-Withdrawing Olefins as Ligands for Efficient Pd-Catalyzed
Aryl-Alkyl Coupling. Organometallics 2013, 32, 4255−4261.
(18) Murphy-Jolly, M. B.; Lewis, L. C.; Caffyn, A. J. M. The
Synthesis of Tris(Perfluoroalkyl)Phosphines. Chem. Commun. 2005,
4479−4480.
(19) Hierso, J. C. Indirect Nonbonded Nuclear Spin-Spin Coupling:
A Guide for the Recognition and Understanding of “through-Space”
NMR J Constants in Small Organic, Organometallic, and Coordina-
tion Compounds. Chem. Rev. 2014, 114, 4838−4867.
Chen, J.; Muller, P.; Buchwald, S. L. Investigating the Dearomative
̈
(20) Contreras, R. H.; Llorente, T.; Ducati, L. C.; Tormena, C. F.
Rearrangement of Biaryl Phosphine-Ligated Pd(II) Complexes. J. Am.
Chem. Soc. 2012, 134, 19922−19934. For recent examples, see:
(f) DeAngelis, A. J.; Gildner, P. G.; Chow, R.; Colacot, T. J.
Generating Active “L-Pd(0)” via Neutral or Cationic π-Allylpalladium
Complexes Featuring Biaryl/Bipyrazolylphosphines: Synthetic, Mech-
anistic, and Structure−Activity Studies in Challenging Cross-Coupling
Reactions. J. Org. Chem. 2015, 80, 6794−6813. (g) Goutierre, A.-S.;
Trinh, H. V.; Larini, P.; Jazzar, R.; Baudoin, O. Comparative
Revisiting NMR Through-Space J Spin−Spin Coupling Constants
FF
for Getting Insight into Proximate F—F Interactions. J. Phys. Chem. A
2014, 118, 5068−5075.
(21) The through-space coupling between fluorine nuclei is very
common. See refs 19 and 20.
(22) Aullon, G.; Ujaque, G.; Lledos, A.; Alvarez, S.; Alemany, P. To
Bend or Not To Bend: Dilemma of the Edge-Sharing Binuclear
G
DOI: 10.1021/acs.organomet.9b00460
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Square Planar Complexes of d8 Transition Metal Ions. Inorg. Chem.
1998, 37, 804−813.
́
́
́
(23) Perez-Galan, P.; Delpont, N.; Herrero-Gomez, E.; Maseras, F.;
Echavarren, A. M. Metal−Arene Interactions in Dialkylbiarylphos-
phane Complexes of Copper, Silver, and Gold. Chem. - Eur. J. 2010,
16, 5324−5332.
́
́
(24) Cordero, B.; Gomez, V.; Platero-Prats, A. E.; Reves, M.;
Echeverría, J.; Cremades, E.; Barragan, F.; Alvarez, S. Covalent radii
́
revisited. Dalton Trans. 2008, 2832−2838.
(25) Alvarez, S. A Cartography of the van Der Waals Territories.
Dalt. Trans. 2013, 42, 8617−8636.
́
(26) (a) Espinet, P.; Albeniz, A. C.; Casares, J. A.; Martínez-Ilarduya,
J. M. 19F NMR in Organometallic Chemistry. Applications of
Fluorinated Aryls. Coord. Chem. Rev. 2008, 252, 2180−2208.
(b) Alonso, M. A.; Casares, J. A.; Espinet, P.; Martínez-Ilarduya, J.
́
M.; Perez-Briso, C. The 3,5-Dichlorotrifluorophenyl Ligand, a Useful
Tool for the Study of Coordination Modes and Dynamic Behavior of
Complexes of Palladium and Platinum. Eur. J. Inorg. Chem. 1998,
1998, 1745−1753. (c) Casares, J. A.; Coco, S.; Espinet, P.; Lin, Y.-S.
Observation of a Slow Dissociative Process in Palladium(II)
Complexes. Organometallics 1995, 14, 3058−3067.
(27) Espinet, P.; Echavarren, A. M. The Mechanisms of the Stille
Reaction. Angew. Chem., Int. Ed. 2004, 43, 4704−4734.
(28) Gioria, E. Promoting Difficult Coupling Processes. Synthesis of
Phosphine-Olefin Hybrid Ligands, Reactivity and Reductive Elimi-
nation Studies. Doctoral Thesis, Universidad de Valladolid, 2018.
́
(29) Perez-Rodríguez, M.; Braga, A. A. C. C.; Garcia-Melchor, M.;
́
Perez-Temprano, M. H.; Casares, J. A.; Ujaque, G.; De Lera, A. R.;
́
Alvarez, R.; Maseras, F.; Espinet, P. C-C Reductive Elimination in
Palladium Complexes, and the Role of Coupling Additives. A DFT
Study Supported by Experiment. J. Am. Chem. Soc. 2009, 131, 3650−
3657.
́
́
́
(30) Uson, R.; Fornies, J.; Martínez, F.; Tomas, M. Mono- and Bi-
Nuclear Anionic Pentafluorophenyl Complexes of Palladium(II) and
Platinum(II). J. Chem. Soc., Dalton Trans. 1980, 888−894.
(31) This behavior was studied and reported by us in neat THF as
solvent and has been now verified in toluene: Salas, G.; Casares, J. A.;
Espinet, P. Enthalpy of Ligand Substitution in Cis Organopalladium
Complexes with Monodentate Ligands. Dalt. Trans. 2009, 79, 8413−
8420.
(32) Casado, A. L.; Casares, J. A.; Espinet, P. Mechanism of the
Uncatalyzed Dissociative Cis-Trans Isomerization of Bis-
(Pentafluorophenyl)Bis(Tetrahydrothiophene): A Refinement. Inorg.
Chem. 1998, 37, 4154−4156.
H
DOI: 10.1021/acs.organomet.9b00460
Organometallics XXXX, XXX, XXX−XXX