Palladium(ii) agostic complex: Exchange of aryl?pd and alkyl?pd bonds

Article abstract of DOI:10.1021/om200451v

Cyclopalladation of 2-^tbutyl-6-(4-fluorophenyl)pyridine with palladium acetate cleanly gives cyclometalated complex 3 with an aryl?Pd bond. Replacement of the acetates by chloride gives the monomeric species 4 with a crystal structure confirming the presence of an agostic interaction from the alkyl group. Reaction of 3 with Na(acac) initially gives the monomeric acac complex 5 which maintains the aryl?Pd bond. Complex 5 shows fluxional behavior of the acac group, and this provides a pathway for its isomerization to complex 6 which has an alkyl?Pd bond; complex 6 was characterized crystallographically. Reaction of agostic 4 with triphenylphosphine gives an initial complex that maintains the aryl?Pd bond, but this complex isomerizes to crystallographically characterized 8 which has an alkyl?Pd bond.

Full text of DOI:10.1021/om200451v

Article
pubs.acs.org/Organometallics
Palladium(II) Agostic Complex: Exchange of Aryl−Pd and Alkyl−Pd
Bonds
Helen R. Thomas, Robert J. Deeth, Guy J. Clarkson, and Jonathan P. Rourke*
Department of Chemistry, Warwick University, Coventry, U.K. CV4 7AL
S
* Supporting Information
ABSTRACT: Cyclopalladation of 2-tbutyl-6-(4-fluorophenyl)pyridine with
palladium acetate cleanly gives cyclometalated complex 3 with an aryl−Pd
bond. Replacement of the acetates by chloride gives the monomeric species
4 with a crystal structure confirming the presence of an agostic interaction
from the alkyl group. Reaction of 3 with Na(acac) initially gives the
monomeric acac complex 5 which maintains the aryl−Pd bond. Complex 5
shows fluxional behavior of the acac group, and this provides a pathway for
its isomerization to complex 6 which has an alkyl−Pd bond; complex 6 was
characterized crystallographically. Reaction of agostic 4 with triphenylphos-
phine gives an initial complex that maintains the aryl−Pd bond, but this
complex isomerizes to crystallographically characterized 8 which has an
alkyl−Pd bond.
acetate (relative integral 3), six signals (each of relative integral
1) in the aromatic region, and one singlet for the butyl group
(relative integral 9).
We can postulate three likely structures for this product: one
with a monodentate acetate and an agostic interaction, 1, one
with a bidentate chelating acetate, 2, and one with a bridged
dimeric structure, 3, Scheme 1.
INTRODUCTION
■
t
The problem of realizing a selective and general method for the
activation of C−H bonds remains one of considerable topical
interest.1 Cyclometalation reactions provide one widely studied
route into the study of such reactions as the prior coordination
of a ligating group can bring C−H bonds into close proximity
to a metal.2 Agostic complexes, where a C−H bond interacts
directly with a metal center are seen as intermediates along the
pathway to full cleavage of the bond,3 and it is generally
recognized that a preference exists for the activation of aryl over
alkyl C−H bonds,4 though the balance between the two can be
quite fine.5
A survey of the Cambridge Crystallographic Database9
suggests considerable precedent for all three structural types.
It reveals 161 structures containing palladium coordinated with
a monodentate acetate (though admittedly none of these have
additional agostic interactions), 11 monomeric structures with a
bidentate acetate, and some 204 structures with bridging
acetates (105 of these are dimeric with a further N donor, and
an additional 30 are dimers with a P donor ligand). The dimeric
structure 3 has the two halves of the molecule related by a C2
rotation (but not an inversion): the bridging acetates fold the
two halves of the molecule back on each other, rather like an
open book, and this can lead to Pd−Pd interactions.10 This
Palladium is probably the metal most studied in this area,6
with recognition of its role in synthetic organic chemistry
coming in 2010 with the award of the Nobel Prize for
palladium catalyzed C−C bond formation. Recently Sanford
has used the cyclopalladation reaction to oxidatively function-
alize arenes,7 and Ritter has made extensive studies into the role
of dimeric palladium complexes in organic transformations.8
In this paper we present details of an aryl C−H bond
activation with palladium to give a new agostic complex with
the agostic interaction coming from an alkyl group. Subsequent
reactions convert this alkyl agostic interaction, at the expense of
the aryl−Pd bond, to give complexes with alkyl−Pd bonds.
t
arrangement would allow the butyl groups to be in a sterically
undemanding position pushed away from the central core, and
such dimeric open book structures are known to persist in
solution.11
NMR and mass-spectral data support the dimeric structure 3
but cannot be considered definitive. In particular, the peak of
highest intensity in the high-resolution ESI mass spectrum
corresponds to (dimer 3 − OAc)+, with other high intensity
peaks corresponding to (dimer-OAc-Me)+ and (dimer-2OAc +
O−H)+; a strong peak corresponding to (monomer-OAc)+ is
RESULTS AND DISCUSSION
■
Direct reaction of palladium acetate with 2-tbutyl-6-(4-
fluorophenyl)pyridine, in acetic acid, activates one of the sp2
C−H bonds of the fluorinated phenyl ring to give a single
cyclometalated species, cleanly, and in good yield. The room
temperature 1H NMR shows the product to contain one
Received: May 26, 2011
Published: October 12, 2011
© 2011 American Chemical Society
5641
dx.doi.org/10.1021/om200451v|Organometallics 2011, 30, 5641−5648

Organometallics
Article
Scheme 1
also present. Variable temperature NMR shows no fluxional
processes over the temperature range investigated (+30 to −64
°C): in particular there is no additional broadening of the tbutyl
resonances compared with the rest of the resonances on
reducing the temperature (contrast this with the case for the
agostic 4, see below). One-dimensional nuclear Overhauser
effect (NOE) experiments support a folded dimer: irradiation
resonances. However, even at −64 °C (the lowest temperature
t
we were able to go to before the solvent froze), the butyl
resonance had by no means separated out into the individual
resonances implied by an agostic structure. We were able to
grow crystals of 4, Figure 1, and gain considerable further
insight into its structure.
Unfortunately, all the crystals we were able to grow had very
weak diffractions due to being multicomponent. At least four
components could be located with the twin option in
CrysAlis12 (ratio of twins 55/22/18/15 with overlaps of up
to 30%). There were high residual electron density peaks
around the palladium as no successful twin model could be
found; consequently the hydrogens on the methyl C(16) could
not be directly located, and these were placed at calculated
positions.
t
of the butyl group induces a small effect on the resonance of
the proton ortho to both the Pd and the F in the metalated
phenyl ring; such effects are not seen in complexes we know to
be monomeric.
Whatever the actual structure, the acetate complex reacts
with chloride (in the form of KCl) to give a chloride 4.
It is clear from the positions of the Pd, N(1), C(8), Cl(1),
and C(16) that, without any agostic interactions from the
hydrogens on C(16), the palladium would formally be a T-
shaped, three coordinate, 14e species. However, the closeness
of C(16) to the Pd (2.51 Å), a distance substantially shorter
than the sum of the van der Waals radii of the two (1.70 and
1.63 Å, respectively),13 indicates there definitely is an
interaction between the two. A comparison of the positions
of the heavy atoms of 4 with those of its platinum analogue
(where the hydrogens were directly located) shows their
positions to be very similar.5a In particular the Pt−C(agostic)
Compared with the acetate precursor, solution NMR data
does suggest a monomeric structure for 4, complete with an
agostic interaction. Variable temperature NMR does suggest
the beginnings of the freezing out of a fluxional process: as the
temperature of the sample is reduced, the resonance of the
tbutyl group broadens significantly more than the rest of the
Figure 1. X -ray crystal structure of 4; thermal ellipsoids at 50% probability. Note that due to poor crystal quality, the hydrogens on C(16) could not
be directly located. Selected bond lengths (Å) and angles (deg): Pd(1)−C(8) 1.953(8); Pd(1)−N(1) 2.018(7); Pd(1)−Cl(1) 2.308(2); Pd(1)−
C(16) 2.507(9); C(8)−Pd(1)−N(1) 83.0(3); C(8)−Pd(1)−Cl(1) 96.8(2); N(1)−Pd(1)−Cl(1) 179.1(2); C(8)−Pd(1)−C(16) 159.2(3); N(1)−
Pd(1)−C(16) 76.3(3); Cl(1)−Pd(1)−C(16) 103.9(2).
5642
dx.doi.org/10.1021/om200451v|Organometallics 2011, 30, 5641−5648

Organometallics
Article
distance of 2.472(4) in the platinum analogue (which actually
had a bifurcated dual agostic interaction) is close to the Pd−
C(16) distance of 2.507(9) Å seen in 4. It is impossible to say
for certain whether the agostic interaction in 4 is actually a
bifurcated one between two of the hydrogens on C(16) and the
Pd center, rather than the more common single C−H···Pd one,
but it must be a real possibility.
Molecules of 4 in the crystal are essentially flat, with only two
of the carbons, and the hydrogens, of the tbutyl group being out
of plane. These flat molecules pack in infinite stacks along the c
axis of the crystal related by the c glide. The cores are aligned
with all the Pd−Cl bonds facing in a similar direction, but each
molecule in the stack is alternating along the long axis of the
complex with a separation of 3.47 Å, a Pd−Pd distance of
3.572 0(3) Å, and a Pd−Pd−Pd angle of 152°, Figure 2.
2.78 and 2.71 Å,15 2.84 and 2.88 Å,16 2.71 and 2.83 Å,17 and
2.47 Å;18 our Pd···C distance is thus well within this range.
Reaction of acetate 3 with sodium acetylacetonate (acac) in
acetone cleanly yielded a single compound. Within this
1
compound, it is apparent from the H NMR that the phenyl
t
to palladium bond is maintained and that the butyl group is
not bonded to the palladium (two sets of three mutually
coupled signals of relative intensity 1, one set with additional
coupling from the fluorine, in the aromatic region, and a singlet
of relative intensity 9 in the aliphatic region). At low
temperatures, the resonances for the acac group are as
expected: two signals of relative integral 3 and one of relative
integral 1, and this allows us to assign the structure 5 to this
product, Scheme 2.
All other evidence for the structure of 5 points to a
monomeric square-planar structure, i.e., high-resolution mass-
spectral data shows no major peaks with m/z greater than the
sodiated mass ion. However, the structure of 5, as drawn, does
t
require the close proximity of one side of the acac to the butyl
group, and this might be problematic. In fact, we observe some
fluxional behavior for 5 with the resonance for the two methyl
groups of the acac interchanging in the 1H NMR. At 25 °C on a
400 MHz spectrometer, the resonances for the acac methyls are
barely visible (it would appear that this is actually the
coalescence temperature) but on cooling, two broad signals
are clear at 10 °C, and the signals are very sharp at −25 °C.
Conversely, on heating, a single broad peak is clear at +40 °C,
and there is some additional narrowing of the resonance by +60
°C (just below the boiling point of the solvent). Applying a
standard analysis, the barrier to the interconversion of the two
acac methyl resonances can be determined to be 60.7 ± 0.5 kJ
mol−1.
The interconversion of the acac methyls in the NMR
suggests either a very distorted coordination plane at the
palladium whereupon a twist of the acac (via an approximately
tetrahedral transition state) results in swapping the sides of the
acac or a pathway where one of the coordinating atoms of one
of the two chelating ligands dissociates from the palladium
before re-entering in a different position. Thermally induced
isomerisations of square planar palladium and platinum
complexes via tetrahedral transition states are currently
unknown19 (though photochemically induced reactions have
been reported20). In addition, tetrahedral complexes would
have unpaired electrons and thus would be paramagnetic, but
we observe no unexpected features in any of our NMR spectra
that would suggest such species. We therefore rule out an
isomerization that proceeds via a tetrahedral transition state.
A dissociative pathway could involve either chelate ligand,
but it is apparent that the group most likely to dissociate would
be the acac oxygen trans to the aryl group. This group will
experience the strongest trans influence, and in addition it is
Figure 2. Stacking arrangement of the heavy atoms of 4 in the crystal.
The Pd−Pd distance is 3.572 0(3) Å, and the Pd−Pd−Pd angle is
152°.
If we are careful to distinguish agostic and anagostic
interactions,3 actual agostic complexes of palladium(II) are
relatively rare. A survey of the Cambridge Structural Database9
shows only 10 published examples of T-shaped palladium
complexes with an additional agostic interaction completing the
coordination sphere of the palladium. Within these other
examples, the Pd···C distances are reported as 2.74−2.94 Å,14
t
rather sterically constrained by the butyl group. Acac ligands
Scheme 2
5643
dx.doi.org/10.1021/om200451v|Organometallics 2011, 30, 5641−5648

Organometallics
Article
Scheme 3
are known to be able to bond with only one oxygen bound to a
particular metal,21 so dissociation of the oxygen is not unlikely.
Furthermore, dissociation of this oxygen would give rise to a
transition state that could be stabilized by agostic interactions
22
from the butyl group; certainly we and others3 have seen
such a stabilization. For all these reasons, we believe the most
likely scenario is that depicted in Scheme 3.
t
Subsequently 5 isomerizes in solution on a time scale of
weeks to give 6, Scheme 2. Clear evidence for this
1
transformation comes from H NMR data whereupon there
are two signals in the aliphatic region (relative intensities 2 and
6) and two signals for the fluorinated phenyl ring (each of
1
relative intensity 2). At room temperature, the H resonances
of the acac methyls in 6 show no unusual features and variable
temperature NMR does not indicate any fluxional processes:
whatever process was interconverting the two sides of the acac
ligand in 5 is not operating in complex 6.
The transformation of 5 to 6 results in the replacement of a
strong aryl C−Pd bond by a weaker alkyl C−Pd bond, but this
unfavorable factor is presumably more than compensated for by
the relief of steric strain caused by the interaction of the acac
t
with the butyl group. A DFT optimization of the structures of
Figure 3. X-ray crystal structure of 6; thermal ellipsoids at 50%
probability. Note that the two hydrogens on C(16) were located in a
difference map and allowed to refine freely (but given thermal
parameters equal to 1.2 times that of the carbon to which they are
attached). Selected bond lengths (Å) and angles (deg): Pd(1)−C(16)
1.9950(18); Pd(1)−O(18) 2.0263(12); Pd(1)−N(1) 2.0567(14);
Pd(1)−O(20) 2.1390(12); C(16)−Pd(1)−O(18) 88.95(6); C(16)−
Pd(1)−N(1) 80.53(7); O(18)−Pd(1)−N(1) 169.36(5); C(16)−
Pd(1)−O(20) 175.70(6); O(18)−Pd(1)−O(20) 89.69(5); N(1)−
Pd(1)−O(20) 100.92(5).
5 and 6 suggests that the sp3 cyclometalated 6 is favored by
some 11 kJ mol−1.
Finally, we were able to grow crystals of 6 and solve the
structure, see Figure 3.
The crystal structure of 6 is relatively undistorted at the
palladium: the mean plane defined by Pd(1), N(1), C(16), and
the two oxygens is close to flat with none of these five atoms
being more than 0.06 Å away from it. With the fluorinated
phenyl ring twisted out of the plane with respect to the
pyridine, there does not appear to be any unfavorable steric
interactions with the acac group. We can therefore attribute for
the lack of any fluxional behavior in the NMR of 6 to this
uncomplicated arrangement.
relief of steric strain is sufficient to stabilize 7 with respect to
the unseen complex 9 by a greater amount than the unseen 9 is
stabilized with respect to agostic 4; reactions of this type have
been seen before.24
In a manner similar to that we have previously seen for the
platinum analogue of 4, reaction with 2 equiv of triphenyl-
phosphine rapidly leads to cleavage of the pyridine to palladium
bond with the formation of a complex containing two
triphenylphosphine ligands,23 Scheme 4. Evidence for a trans
arrangement of two PPh3 ligands comes from the 31P NMR (a
single resonance) and the 1H NMR (integrals show the ratio of
The equimolar mixture of 4 and 7 that results from the
addition of a single equivalent of triphenylphosphine to 4
transforms slowly (a time scale of days) to another new
palladium compound, complex 8, Scheme 5. The new complex
is a singly cyclometalated species with a single phosphine ligand
per palladium center, but it is apparent from the NMR spectra
t
that the new complex is cyclometalated via the butyl group,
t
PPh3 to Bu-pyridine to be 2:1).
rather than via the phenyl ring. Evidence for this formulation
comes from two sets of resonances (each of relative integral 2)
for the fluorophenyl ring, and two resonances (of relative
integral 2 [with coupling to phosphorus] and 6) in the aliphatic
region (rather than one resonance of integral 9).
We were able to crystallize 8 and solve the X-ray structure,
see Figure 4.
The coordination geometry of the palladium in 8 is
recognizably square-planar though rather distorted. Thus,
while the P−Pd−C angle of 91.5° is close to 90°, all other
In fact, 7 forms even when only 1 equiv of triphenylphos-
phine is added to 4, whereupon an equimolar mixture of 4 and
7 forms. We might have expected the reaction of 4 with a single
equivalent of PPh3 to give a complex that remains cyclo-
metalated via the phenyl ring, and this complex must be the
intermediate 9 in the reaction, but it would appear that steric
constraints render it unstable with respect to further reaction.
Thus displacement of the pyridine by a second PPh3 results in
t
the butyl pyridine moiety being able to rotate away, and this
5644
dx.doi.org/10.1021/om200451v|Organometallics 2011, 30, 5641−5648

Organometallics
Article
Scheme 4
Scheme 5
relatively small stabilization is presumably a function of the fact
that the favored isomer is still very much distorted away from
an ideal geometry.
Variable temperature NMR was used to investigate the
possibility of fluxional processes occurring within complex 8.
The signals that we anticipated showing effects are the CH2
protons (at room temperature they are a sharp doublet caused
by coupling to the 31P of the phosphine ligand) and the two
methyl groups. We reasoned that the distortions to the
coordination plane around the palladium would result in
sufficient asymmetry such that these two CH2 protons (and the
two methyl groups) would become inequivalent, if we were able
to freeze out the buckling motion that would equilibrate them
via a ring inversion. Reducing the temperature of the sample
(on a 400 MHz NMR spectrometer) did indeed result in the
broadening of the CH2 resonance giving a coalescence at
around −56 °C and separation into two broad peaks at −64 °C
(the lowest temperature we were able to reach in CDCl3
solution). Since we were unable to reach a true low-
temperature limiting spectrum, we can only estimate the final
separation of the two resonances and calculate a relatively
imprecise barrier to the inversion of the metallacycle of 40 ± 4
kJ mol−1.
It is pertinent to consider the mechanism of the conversion
of 5 to 6 and of the formation of 8. In both cases, an
unfavorable steric interaction is driving the formation of an
alkyl−Pd bond at the expense of an aryl−Pd bond. A crucial
step of the transformations will be the transfer of a hydrogen
from the alkyl group to the aryl group. We have previously5a,23
extensively studied the reaction that gives the platinum
analogue of 8 from the analogues of 4 and 7. We were able
to demonstrate that a hydrogen moves from the alkyl group to
the aryl group in an intramolecular transfer across the platinum,
in a process that is strongly reminiscent of the sigma-CAM
mechanism proposed by Perutz and Sabo-Etienne,26 and that
the most likely reaction route was one in which the rate
determining step was the isomerization of the analogue of
agostic 4 to the analogue of another agostic complex (i.e., 10,
with an interaction from the aryl ring). Assuming the reactions
with palladium go via similar intermediates, we propose
Scheme 6 as our favored route to the formation of 8.
Figure 4. X-ray crystal structure of 8; thermal ellipsoids at 50%
probability. Note that the two hydrogens on C(116) were directly
located in a difference map and allowed to refine freely (but given
isotropic displacement parameters equal to 1.2 times the equivalent
isotropic displacement parameter of C116). Selected bond lengths (Å)
and angles (deg): Pd(1)−C(116) 2.0305(17); Pd(1)−N(101)
2.1462(15); Pd(1)−P(201) 2.2369(5); Pd(1)−Cl(1) 2.4492(4)
C(116)−Pd(1)−N(101) 78.15(7); C(116)−Pd(1)−P(201)
91.53(6); N(101)−Pd(1)−P(201) 165.61(4); C(116)−Pd(1)−Cl(1)
160.33(6); N(101)−Pd(1)−Cl(1) 95.69(4); P(201)−Pd(1)−Cl(1)
97.250(17).
angles at the Pd center are between 6 and 20° away from ideal
values of 90 or 180° with the angles of the mutually trans
ligands being 165.6 and 160.3°. The Pd, N(101), Cl(1),
P(201), and C(116) atoms are by no means coplanar: the N,
Cl, P and C are, respectively, 0.250, 0.214, 0.316, 0.202 Å away
from the calculated mean plane defined by these atoms and the
Pd. The platinum analogue of 8 has a similarly distorted square-
planar geometry,23 with bond lengths that are slightly longer
than for related undistorted complexes.25
A DFT optimization suggests that the sp3 cyclometalated 8 is
more stable than its unseen sp2 cyclometalated isomer
(intermediate 9 in Scheme 4) by some 16 kJ mol−1. This
An initial estimate of the transition state between 4 and 10
was obtained from a linear transit starting from 4 where the
agostic C−H bond was lengthened from 1.1 to 3 Å. A full
5645
dx.doi.org/10.1021/om200451v|Organometallics 2011, 30, 5641−5648

Organometallics
Article
Scheme 6
transition state optimization (DZP in the gas phase) was
performed starting from the maximum in the linear transit path
to give the structure shown in Figure 5. This structure was
force for these reactions is a relief of steric strain caused by the
bulky butyl group.
t
EXPERIMENTAL SECTION
■
All chemicals were used as supplied, unless noted otherwise. All NMR
spectra were obtained on a Bruker Avance 300, 400, 500, or 600 MHz
spectrometer in CDCl3 or D6-acetone; 1H and 13C shifts are
referenced to external trimethylsilane (TMS), assignments being
made with the use of decoupling, NOE and the heteronuclear multiple
bond correlation (HMBC), heteronuclear single quantum correlation
(HSQC), distortionless enhancement by polarization transfer
(DEPT), and correlation spectroscopy (COSY) pulse sequences. 19F
chemical shifts are quoted from the directly observed signals and are
referenced to external CFCl3. All accurate mass spectra were run on a
Bruker MaXis mass spectrometer, and X-ray crystal structures were
collected on an Oxford Diffraction Gemini four-circle system with
Ruby CCD area detector. The synthesis of the starting 2-tbutyl-6-(4-
fluorophenyl)pyridine has been reported previously.5a
Figure 5. Calculated transition state between 4 and 10; the arrow
indicates the transition state vector.
confirmed as a first order saddle point with a single imaginary
frequency of −400 cm−1. The free energy for this process was
then calculated with TZ2P and COSMO corrections and
showed the process to be endoergic by 6.3 kJ mol−1 with a
barrier of 167 kJ mol−1. This barrier, as calculated, is an upper
limit but is sufficiently large to throw some doubt on the
probability of this route being viable. It is also important to
note that DFT calculations are not perfect and ours did not
include factors such as tunnelling, which will reduce the barrier,
maybe substantially.27 We must therefore be alert to other
possibilities such as a metal assisted acid−base reaction.
By analogy, the transformation of the acac complex 5 to 6,
would then proceed via the intermediate we proposed in
Scheme 3, instead of 4. This agostic intermediate would be set
up to transfer a hydrogen across from the alkyl to aryl group,
followed by a recoordination of the second acac oxygen to give
the final product.
Synthesis of Palladium Acetate Complex 3. Palladium acetate
(0.209 g, 0.93 mmol, 2 equiv) and 2-tbutyl-6-(4-fluorophenyl)pyridine
(0.110 g, 0.48 mmol, 1 equiv) were refluxed in glacial acetic acid (25
mL) in air (2 h). During this time, there was an observable color
change from a red solution to a dark green mixture containing some
metallic palladium. The mixture was filtered to give a dark brown
filtrate. The solvent was removed, and the resultant brown solid was
dissolved in diethyl ether and chromatographed. The second fraction
yielded the product as a pale orange solid. Yield, 0.140 g, (0.36 mmol,
75%).
3
3
δH: 7.79 (1H, t, J 8 Hz, central py), 7.44 (1H, d, J 8 Hz, py
3
4
adjacent to ph), 7.31 (1H, dd, J 9 Hz, J(H−F) 5 Hz, ph, meta to F),
7.26 (1H, d, 3J 8 Hz, py, adjacent to tBu), 6.87 (1H, td, 3J(4H−H), (H−F)
9
4
3
Hz, J 2 Hz, ph, ortho to F), 6.85 (1H, dd, J(H−F) 8 Hz, J 2 Hz, ph,
t
ortho to Pd), 2.14 (3H, s, OAc), 1.64 (9H, s, Bu). δC: 183.8 (CO),
169.9 (py, adjacent to N, with tBu attached), 162.1 (py, adjacent to N,
1
with ph attached), 160.5 (d, J(C−F) 257 Hz, ph with F directly
attached), 145.0 (d, 3J(C−F) 6 Hz, ph, attached to Pd), 141.3 (d, 4J(C−F)
3
CONCLUSIONS
3 Hz, ph, attached to py), 139.6 (central py), 125.0 (d, J(C−F) 8 Hz,
■
ph, meta to F and Pd), 120.1 (py adjacent to py adjacent to tBu), 117.0
(py adjacent to ph), 116.9 (d, 2J(C−F) 21 Hz, ph, adjacent to F and Pd),
113.4 (d, 2J(C−F) 23 Hz, ph, adjacent to F), 39.6 (tBu nonprotonated),
30.5 (OAc Me), 23.5 (tBu Me). δF: −107.5. HR-MS (ESI): m/z
334.0224, calculated for C15H15F14N106Pd = (monomer-OAc)+
334.0226; 351.0489, calculated for C15H18F14N2106Pd = (monomer-
Cyclopalladation of 2-tbutyl-6-(4-fluorophenyl)pyridine with
palladium acetate has been shown to proceed cleanly to give a
cyclometalated complex with an aryl−Pd bond. This compound
is likely to have a dimeric structure with two bridging acetates
completing the square planar coordination. However, replace-
ment of the acetates by chlorides breaks up this dimer and gives
a monomeric species with an agostic interaction from the alkyl
group completing the coordination sphere. Subsequent
reactions give complexes that only temporarily maintain the
aryl−Pd bond: these complexes isomerize to more thermody-
namically stable complexes with alkyl−Pd bonds. The driving
OAc
+
NH3 )+ 351.0472; 685.0330, calculated for
C30H29F215N2O106Pd2 = (dimer-2OAc-H + O)+ 685.0334; 715.0427,
calculated for C31H31F215N2O2106Pd2 = (dimer-OAc-CH3)+ 715.0436;
729.0584, calculated for C32H3315N2F2O2106Pd2 = (dimer-OAc)+
729.0585.
Synthesis of Palladium Chloride Agostic Complex 4.
Palladium acetate complex 3 (50 mg, 0.13 mmol) was dissolved in
5646
dx.doi.org/10.1021/om200451v|Organometallics 2011, 30, 5641−5648

Organometallics
Article
Table 1. X-ray Data for Complexes 4, 6, and 8
complex
agostic 4
acac 6
PPh3 8
crystal form
yellow block
0.20 × 0.6 × 0.06
C15H15ClFNPd
370.13
colorless block
0.40 × 0.35 × 0.10
C20H22FNO2Pd
433.79
colorless block
0.35 × 0.25 × 0.10
C34H31Cl4FNPPd
751.77
dimensions/mm3
empirical formula
Mw
crystal system
monoclinic
P2(1)/c
8.915 7(4)
22.923 9(9)
6.941 5(3)
90
monoclinic
P2(1)/n
11.200 67(18)
14.281 2(2)
11.411 99(18)
90
monoclinic
P2(1)/c
space group
a/Å
19.257 4(4)
9.889 60(10)
18.584 7(4)
90
b/Å
c/Å
α/°
β/°
γ/°
105.875(5)
90
92.630 5(15)
90
116.374(10)
90
U/Å3
T/K
Z
1364.62(10)
100(2)
1823.53(5)
100(2)
3171.01(10)
100(2)
4
4
4
D
calc/Mg m−3
1.802
1.580
1.575
F(000)
μ(MoKα)/mm−1
736
880
1520
1.551
1.040
1.005
θ max/deg
30.22
30.55
30.14
refln measured
unique data
R1 [I > 2σ(I)]
wR2
12607
17506
18306
3689
5071
8229
0.0911
0.0243
0.0280
0.2545
0.0570
0.0632
data/rest/param
3689/0/ 175
5071/0/ 236
8229/0/ 385
glacial acetic acid (25 mL) together with excess KCl (0.5 g). The
mixture was stirred under reflux (30 min), with an almost immediate
color change from pale orange to pale yellow. The reaction solution
was cooled and filtered, and the solvent was removed to leave a pale
yellow solid. The solid was purified by column chromatography,
loading and eluting with diethyl ether: the product was the first
colored fraction. Yield: 40 mg (0.11 mmol, 85%).
124.3 (d, 3J(C−F) 8 Hz, ph, meta to F and Pd), 121.4 (py, ortho to tBu),
2
116.9 (d, J(C−F) 18 Hz, ph, ortho to Pd and F), 114.9 (py, ortho to
ph), 112.1 (d, 2J(C−F) 19 Hz, ph, ortho to F), 99.9 (central acac), 38.5
(tBu nonprotonated), 31.4 (tBu Me), 27.5 (acac Me) ppm. δF: −110.5
ppm. HR-MS (ESI): m/z 456.0570, calculated for
C20H22F14NNaO2106Pd = (M + Na)+ 456.0583; 769.0897, calculated
for C35H37F215N2O2106Pd2 = (2M-acac)+ 769.0932.
Synthesis of sp3 Metalated Palladium acac Complex 6. A
solution of complex 5 in chloroform was left for 1 month at room
temperature, by which time much of the material had isomerized and
crystals of 6 had separated out.
δH: 7.78 (1H, t, 3J 8 Hz, central py), 7.42 (1H, dd, 3J(H−F) 9 Hz, 4J 3
Hz, ph, adjacent to Pd), 7.40 (1H, d, 3J 8 Hz, py, adjacent to tBu), 7.27
(1H, dd, 3J 9 Hz, 4J(H−F) 5 Hz, ph, meta to F and Pd), 7.17 (1H, d, 3J 8
3
Hz, py, adjacent to ph), 6.80 (1H, td, J(H−H), (H−F) 8 Hz, 4J 3 Hz, ph,
3
t
δH: 8.11 (2H, m, ph, meta to F), 7.76 (1H, t, J 8 Hz, central py),
adjacent to F), 1.44 (9H, s, Bu). δC: 169.2 (py, adjacent to N, with
3
4
3
4
tBu attached), 161.1 (py, adjacent to N, with ph attached), 160.7 (d,
1J(C−F) 258 Hz, ph with F directly attached), 148.0 (ph, directly
bonded to py), 142.5 (ph with Pd directly bonded), 139.1 (central py),
124.9 (d, 3J(C−F) 9 Hz, ph, meta to F and Pd), 122.6 (d, 2J(C−F) 21 Hz,
ortho to F and Pd), 120.5 (py, adjacent to ph), 116.8 (py, adjacent to
7.26 (1H, dd, J 8 Hz, J 1 Hz, py), 7.15 (1H, dd, J 8 Hz, J 1 Hz,
3
py), 7.10 (2H, t, J(H−F), (H−H) 8 Hz, ph, ortho to F), 5.51 (1H, s,
central acac), 2.39 (2H, s, CH2), 2.29 (3H, s, acac Me), 2.15 (3H, s,
acac Me), 1.61 (6H, s, Me2). δF: −113.2 ppm.
Synthesis of bis-PPh3 Palladium Complex 7. Triphenylphos-
phine (7.1 mg, 0.027 mmol, 2 equiv) was added to a solution of the
palladium agostic complex 4 (5 mg, 0.014 mmol) in chloroform. An
immediate color change from pale yellow to colorless was observed.
Yield: 10.3 mg (0.012 mmol, 85%).
2
tBu), 113.1 (d, J(C−F) 23 Hz, ph, ortho to F), 42.4 (tBu
nonprotonated), 28.5 (tBu Me) ppm. δF: −106.0 ppm. HR-MS
(ESI): m/z 334.0224, calculated for C15H15F14N106Pd = (M-Cl)+
334.0225; 352.0330, calculated for C15H17F14NO106Pd = (M-Cl +
H2O)+ 352.0325; 705.0140, calculated for C30H3035ClF215N2106Pd2 =
(2M-Cl)+ 705.0145.
δH: 8.43 (1H, d, 3J 8 Hz, py), 7.44 (1H, t, 3J 8 Hz, central py), 7.35
3
3
3
(12H, dd, J(H−P) 12 Hz, J 6 Hz, PPh3), 7.24 (6H, t, J 6 Hz, PPh3),
7.14 (12H, t, 3J 7 Hz, PPh3), 7.08 (2H, m, py and ph meta to F), 6.32
Synthesis of sp2 Metalated Palladium acac Complex 5.
Sodium acetylacetonate (48 mg, 0.40 mmol) was added to a solution
of complex 3 (50 mg, 0.13 mmol) in chloroform. The mixture was
refluxed overnight. After cooling, the solution was filtered and the
solvent removed to give a pale yellow solid. Yield: 46 mg (0.11 mmol,
85%).
3
4
(1H, dd, J(H−F) 9 Hz, J 2 Hz, ph ortho to Pd and F), 6.09 (1H, td,
3J(H−F), (H−H) 8 Hz, 4J 2 Hz, ph ortho to F), 1.11 (9H, s, Me3) ppm. δF:
−117.1 ppm. δP: 21.9 ppm. HR-MS (ESI): m/z 857.2062, calculated
for C51H45F14NP2105Pd = (M-Cl)+ 857.2502
Synthesis of sp3 Metalated PPh3 Palladium Complex 8.
Triphenylphosphine 3.5 mg, 0.014 mmol, 1 equiv) was added to a
solution of the palladium agostic complex 4 in chloroform 5.0 mg,
0.014 mmol) and left to react for 3 days at room temperature. The
solvent was removed to give the product as a pale yellow solid. Yield
7.4 mg (0.012 mmol, 87%)
3
δH(263K): 7.72 (1H, t, 3J 8 Hz, central py), 7.43 (1H, dd, J 8 Hz,
3
4J 1 Hz, py, ortho to ph), 7.36 (1H, dd, J 8 Hz, 4J 1 Hz, py, ortho to
tBu), 7.29 (1H, dd, 3J 9 Hz, 4J(H−F) 5 Hz, ph, meta to F and Pd), 7.15
3
4
(1H, dd, J(C−F) 9 Hz, J 3 Hz, ph, ortho to Pd and F), 6.82 (1H, td,
3J(H−H), (H−F) 9 Hz, 4J 3 Hz, ph, ortho to F), 5.37 (1H, s, central acac),
2.09 (3H, s, acac Me), 1.96 (3H, s, acac Me), 1.70 (9H, s, Bu). δC:
187.3 (CO), 174.8 (py, with Bu attached), 164.1 (py, with Ph
3
δH: 8.01 (2H, m, ph meta to F), 7.70 (1H, t, J 8 Hz, central py),
t
3
3
4
7.56 (6H, m, J(H−P) 11 Hz, J 8 Hz, J 2 Hz, PPh3, ortho to P), 7.31
(10H, m, py and PPh3), 7.10 (3H, m, py and ph ortho to F), 1.82 (2H,
d, 3J(H−P) 4 Hz, CH2), 1.38 (6H, s, Me) ppm. δC: 173.2 (py with alkyl
t
1
attached), 160.3 (d, J(C−F) 254 Hz, ph, with F attached), 153.1 (d,
3J(C−F) 6 Hz, ph with Pd attached), 142.3 (Cd), 138.0 (central py),
attached), 163.8 (d, J(C−F) 254 Hz, ph with F attached), 159.2 (py
1
5647
dx.doi.org/10.1021/om200451v|Organometallics 2011, 30, 5641−5648

Organometallics
Article
3
with ph attached), 138.4 (central py), 134.5 (d, J(C−P) 11 Hz, PPh3),
(13) Bondi, A. J. Phys. Chem. 1964, 68, 441.
(14) Stambuli, J. P.; Incarvito, C. D.; Buhl, M.; Hartwig, J. F. J. Am.
Chem. Soc. 2004, 126, 1184.
(15) Yamashita, M.; Hartwig, J. F. J. Am. Chem. Soc. 2004, 126, 5344.
(16) Yamashita, M.; Takamiya, I.; Jin, K.; Nozaki, K. Organometallics
2006, 25, 4588.
(17) Watanabe, C.; Iwamoto, T.; Kabuto, C.; Kira, M. Angew. Chem.,
Int. Ed. 2008, 47, 5386.
1
131.9 (d, J(C−P) 52 Hz, ipso PPh3) 130.8 (ph meta to F), 130.1
(PPh3), 128.1 (d, J(C−P) 10 Hz, PPh3), 123.0 (py), 119.0 (py), 115.0
(d, J(C−F) 22 Hz, ph), 50.4 (nonprotonated alkyl), 47.9 (CH2), 32.4
(Me) ppm. δF: −112.5 ppm. δP: 35.5 ppm. HR-MS (ESI): m/z
596.1142, calculated for C33H30F14NP106Pd = (M-Cl)+ 596.1160.
X-ray Crystallographic Studies. X-ray data for complexes 4, 6,
and 8 are shown in Table 1.
4
3
DFT Calculations. All DFT calculations used the Amsterdam
density functional (ADF) code version 2008.01.28 The general features
available in ADF have been described.29 Here, we have used scalar
zero-order regular approximation (ZORA) relativistic corrections with
the OPBE functional30 and the supplied frozen-core, triple-ζ plus
polarization ZORA basis sets. Solvation effects were included via the
conductor-like screening model (COSMO) as implemented in ADF.
Default SCF and geometry optimization convergence criteria were
used.
(18) Walter, M. D.; White, P. S.; Brookhart, M. Chem. Commun.
2009, 6361.
(19) Tobe, M. L.; Burgess, J. Inorganic Reaction Mechanisms; Addison
Wesley Longman Limited: Harlow, U.K., 1999.
(20) Annibale, G.; Bonivento, M.; Canovese, L.; Cattalini, L.;
Michelon, G.; Tobe, M. L. Inorg. Chem. 1985, 24, 797.
(21) Alvarez, B.; Bois, C.; Jeannin, Y.; Miguel, D.; Riera, V. Inorg.
Chem. 1993, 32, 3783.
(22) (a) Crosby, S. H.; Clarkson, G. J.; Deeth, R. J.; Rourke, J. P.
Organometallics 2010, 29, 1966. (b) Crosby, S. H.; Clarkson, G. J.;
Deeth, R. J.; Rourke, J. P. Dalton Trans. 2011, 40, 1227.
(23) Crosby, S. H.; Clarkson, G. J.; Rourke, J. P. Organometallics
2011, 30, 3603.
(24) (a) Albrecht, M.; Dani, P.; Lutz, M.; Spek, A. L.; Koten, G. v. J.
Am. Chem. Soc. 2000, 122, 11822. (b) Rodriguez, G.; Albrecht, M.;
Schoenmaker, J.; Ford, A.; Lutz, M.; Spek, A. L.; van Koten, G. J. Am.
Chem. Soc. 2002, 124, 5127. (c) Crespo, M.; Granell, J.; Solans, X.;
Font-Bardia, M. J. Organomet. Chem. 2003, 681, 143. (d) Crespo, M.;
Solans, X.; FontBardia, M. J. Organomet. Chem. 1996, 509, 29.
(25) (a) Newman, C. P.; Casey-Green, K.; Clarkson, G. J.; Cave, G.
W. V.; Errington, W.; Rourke, J. P. Dalton Trans. 2007, 3170.
(b) Mamtora, J.; Crosby, S. H.; Newman, C. P.; Clarkson, G. J.;
Rourke, J. P. Organometallics 2008, 27, 5559.
ASSOCIATED CONTENT
* Supporting Information
■
S
DFT optimized coordinates and energies for 4, 5, 6, 8, 9, 10
and the TS in Figure 5; CIF files; and complete sets of
structural data for 4, 6, and 8. This material is available free of
charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*E-mail: j.rourke@warwick.ac.uk.
■
ACKNOWLEDGMENTS
■
(26) Perutz, R. N.; Sabo-Etienne, S. Angew. Chem., Int. Ed. 2007, 46,
2578.
(27) Schreiner, P. R.; Reisenauer, H. P.; Ley, D.; Gerbig, D.; Wu, C.-
We are grateful for support from Advantage West Midlands
(AWM) (partially funded by the European Regional Develop-
ment Fund) for the purchase of a high-resolution mass
spectrometer and the XRD system that was used to solve the
crystal structures of 4, 6, and 8.
H.; Allen, W. D. Science 2011, 332, 1300.
(28) Baerends, E. J. B., A.; Bo, C.; Boerrigter, P. M.; Cavallo, L.;
Deng, L.; Dickson, R. M.; Ellis, D. E.; Fan, L.; Fischer, T. H.; Fonseca
Guerra, C.; van Gisbergen, S. J. A.; Groeneveld, J. A.; Gritsenko, O. V.;
Harris, F. E.; van den Hoek, P.; Jacobsen, H.; van Kessel, G.; Kootstra,
F.; van Lenthe, E.; Osinga, V. P.; Philipsen, P. H. T.; Post, D.; Pye, C.
C.; Ravenek, W.; Ros, P.; Schipper, P. R. T.; Schreckenbach, G.;
Snijders, J. G.; Sola, M.; Swerhone, D.; te Velde, G.; Vernooijs, P.;
Versluis, L.; Visser, O.; van Wezenbeek, E.; Wiesenekker, G.; Wolff, S.
K.; Woo, T. K.; Ziegler, T. ADF 2008.01; Scientific Computing and
Modelling NV: Free University, Amsterdam, The Netherlands, 2008.
(29) Velde, G. t.; Baerends, E. J. J. Comput. Phys. 1992, 99, 84.
(30) (a) Perdew, J. P.; Burke, K.; Ernzerhof, M. Phys. Rev. Lett. 1996,
77, 3865. (b) Handy, N. C.; Cohen, A. J. Mol. Phys. 2001, 99, 403.
REFERENCES
■
(1) (a) Labinger, J. A.; Bercaw, J. E. Nature 2002, 417, 507.
(b) Bergman, R. G. Nature 2007, 446, 391.
(2) Albrecht, M. Chem. Rev. 2009, 110, 576.
(3) Brookhart, M.; Green, M. L. H.; Parkin, G. Proc. Natl. Acad. Sci.
U.S.A. 2007, 104, 6908.
(4) Goldman, A. S.; Goldberg, K. I. In Activation and Functionaliza-
tion of C-H Bonds; Goldberg, K. I., Goldman, A. S., Eds.; ACS
Symposium Series, Vol. 885; American Chemical Society: Washington,
DC, 2004.
(5) (a) Crosby, S. H.; Clarkson, G. J.; Rourke, J. P. J. Am. Chem. Soc.
2009, 131, 14142. (b) Garner, A. W.; Harris, C. F.; Vezzu, D. A. K.;
Pike, R. D.; Huo, S. Chem. Commun. 2011, 47, 1902.
(6) Lyons, T. W.; Sanford, M. S. Chem. Rev. 2010, 110, 1147.
(7) (a) Deprez, N. R.; Sanford, M. S. J. Am. Chem. Soc. 2009, 131,
11234. (b) Lanci, M. P.; Remy, M. S.; Kaminsky, W.; Mayer, J. M.;
Sanford, M. S. J. Am. Chem. Soc. 2009, 131, 15618. (c) Ye, Y. D.; Ball,
N. D.; Kampf, J. W.; Sanford, M. S. J. Am. Chem. Soc. 2010, 132,
14682.
NOTE ADDED AFTER ASAP PUBLICATION
■
In the version of this paper published on October 12, 2011, the
DFT data mentioned in the paragraph describing the
Supporting Information were missing. The version that appears
as of November 7, 2011, has the data correctly deposited as
Supporting Information.
(8) (a) Powers, D. C.; Geibel, M. A. L.; Klein, J.; Ritter, T. J. Am.
Chem. Soc. 2009, 131, 17050. (b) Powers, D. C.; Ritter, T. Nat. Chem.
2009, 1, 302. (c) Powers, D. C.; Xiao, D. Y.; Geibel, M. A. L.; Ritter, T.
J. Am. Chem. Soc. 2010, 132, 14530. (d) Chuang, G. J.; Wang, W. K.;
Lee, E.; Ritter, T. J. Am. Chem. Soc. 2011, 133, 1760.
(9) Fletcher, D. A.; McMeeking, R. F.; Parkin, D. J. Chem. Inf.
Comput. Sci. 1996, 36, 746.
(10) Bercaw, J. E.; Durrell, A. C.; Gray, H. B.; Green, J. C.; Hazari,
N.; Labinger, J. A.; Winkler, J. R. Inorg. Chem. 2010, 49, 1801.
(11) Evans, P. G.; Brown, N. A.; Clarkson, G. J.; Newman, C. P.;
Rourke, J. P. J. Organomet. Chem. 2006, 691, 1251.
(12) Crysalis, Agilent Technologies Ltd.: Yarnton, U.K., 2010.
5648
dx.doi.org/10.1021/om200451v|Organometallics 2011, 30, 5641−5648