Mechanistic details of the nickel-mediated formation of acrylates from CO2, ethylene and methyl iodide

Article abstract of DOI:10.1021/om400262b

Methyl iodide induces the stoichiometric cleavage of nickelalactones, which are key intermediates in the nickel-mediated reaction of CO₂ and alkenes to acrylates. Herein, we propose a modified and extended mechanism for this reaction on the basis of theoretical and experimental investigations for the bidentate P ligand 1,2-bis(di-tert-butylphosphino)ethane (dtbpe). The calculated elementary steps agree well with experimental findings: reaction barriers are reasonable and explain the facile liberation of acrylate from a nickelalactone by methyl iodide. We were able to isolate reactive intermediates and to verify the existence of proposed reaction pathways. Additionally, we have identified unproductive pathways leading to byproducts (e.g., propionates and catalytically inactive organometallic species). Although those side reactions can be suppressed to a certain extent, the strong binding of acrylate to nickel prevents a catalytic reaction, at least for the chosen ligand.

Full text of DOI:10.1021/om400262b

Article
pubs.acs.org/Organometallics
Mechanistic Details of the Nickel-Mediated Formation of Acrylates
from CO2, Ethylene and Methyl Iodide
Philipp N. Plessow,†,‡ Laura Weigel,§ Ronald Lindner,‡ Ansgar Schafer,† Frank Rominger,§
̈
Michael Limbach,*,‡,∥ and Peter Hofmann*,‡,§
†BASF SE, Quantum Chemistry, GVM/M-B009, Carl-Bosch-Strasse 38, D-67056 Ludwigshafen, Germany
‡CaRLa (Catalysis Research Laboratory), Im Neuenheimer Feld 584, D-69120 Heidelberg, Germany
§Ruprecht-Karls-Universitat Heidelberg, Organisch-Chemisches Institut, Im Neuenheimer Feld 270, D-69120 Heidelberg, Germany
̈
∥BASF SE, Synthesis & Homogeneous Catalysis, GCS/C-M313, Carl-Bosch-Strasse 38, D-67056 Ludwigshafen, Germany
S
* Supporting Information
ABSTRACT: Methyl iodide induces the stoichiometric cleavage
of nickelalactones, which are key intermediates in the nickel-
mediated reaction of CO2 and alkenes to acrylates. Herein, we
propose a modified and extended mechanism for this reaction on
the basis of theoretical and experimental investigations for the
bidentate P ligand 1,2-bis(di-tert-butylphosphino)ethane (dtbpe).
The calculated elementary steps agree well with experimental
findings: reaction barriers are reasonable and explain the facile
liberation of acrylate from a nickelalactone by methyl iodide. We
were able to isolate reactive intermediates and to verify the
existence of proposed reaction pathways. Additionally, we have
identified unproductive pathways leading to byproducts (e.g.,
propionates and catalytically inactive organometallic species).
Although those side reactions can be suppressed to a certain
extent, the strong binding of acrylate to nickel prevents a catalytic reaction, at least for the chosen ligand.
methyl acrylate and methyl propionate. The fate of the
organometallic species remained unclear. The equimolar
formation of unsaturated and saturated organic products is in
accord with a first mechanistic hypothesis (Scheme 1),14,15 as 1
equiv of H+ (as HI) is formed and may protonate nickel-
alactone A or related organometallic species.
INTRODUCTION
■
Carbon dioxide is an abundantly available C1 building block
and could become a cheap raw material1−3 for a limited number
of bulk chemicals, such as formates and acrylates. Whereas the
technology for formates is already mature,4 the development of
acrylates has been in its infancy since the seminal work of
Hoberg5 and Yamamoto6 in the 1980s. The endergonic nature
of the overall reaction and the high activation barrier for the
proposed β-hydride elimination from a nickelalactone (ΔG =
164 kJ/mol)7,8 makes the direct reaction of CO2 and ethylene
to give acrylic acid a difficult task to study.
The transition state TSA‑C for the transformation of A to C
has been computed as a four-membered cyclic arrangement
with an activation barrier of ΔG⧧ = 240 kJ/mol,14 which seems
unreasonably high. As this barrier is even higher than that
computed for a β-H elimination without MeI (ΔG⧧ = 164 kJ/
mol),7 to which it is compared by Kuhn et al., one has to expect
̈
We have recently overcome both limitations and assembled
the first catalytic reaction of CO2 and ethylene to give acrylates
by targeting Na acrylate as a commercially attractive salt of
acrylic acid and by cleaving the intermediate lactones by strong
bases bearing Lewis acidic cations.9 Both experimental and
theoretical data point to the fact that the nickelalactone is
deprotonated by a base instead of undergoing β-H elimination.
Already Yamamoto, Hoberg, and Walther et al. had found10
that certain electrophiles such as alkyl halides are capable of
cleaving nickelalactones, as did Puddephatt and Chatami et al.
for the less explored zircona-11 and platinalactones.12 More
the latter reaction to be clearly preferred over methylation.
Furthermore, such a high activation barrier is rather
inconsistent with experimental observations by these authors:
i.e., substantial product formation within hours at room
temperature. As derived from transition state theory, realistic
barriers are ca. 100 kJ/mol, and the explanation of Kuhn et al.
̈
invoking concentration effects14,15 certainly does not seem
reasonable for a barrier more than twice as large. Due to this
apparent inconsistency, we became interested in revisiting the
proposed mechanistic pathway and, to clarify the central
recently, Kuhn and Rieger et al.13,14 have shown in
̈
stoichiometric reactions that an excess of methyl iodide
(MeI) cleaves nickelalactones to give an equimolar mixture of
Received: March 28, 2013
© XXXX American Chemical Society
A
dx.doi.org/10.1021/om400262b | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
halide.28 All Gibbs free energies in solution are given at
standard conditions in kJ/mol and relative to the nickelalactone
A1. Open shell species have been calculated using the
unrestricted Kohn−Sham formalism (UKS). The connectivity
of transition states has been confirmed by a displacement along
the transition vector followed by steepest descent optimization.
Activation energies for dissociation of ligands from pseudote-
trahedral endergonic Ni intermediates for different functionals
and basis sets were in general found to be only slightly above
those intermediates (typically <10 kJ/mol). In the harmonic
approximation, activation Gibbs free energies were then often
calculated to be slightly negative. Therefore we set ΔG ≈ 0 in
these cases, assuming a negligible free energy barrier.
Scheme 1. Mechanism Proposed by Kuhn et al. for the
̈
Reaction of the Nickelalactone A with Methyl Iodide14
Nickelalactone Cleavage. Kuhn et al. proposed a one-
̈
step, concerted methylation with the aforementioned very high
barrier based on the TMEDA-ligated nickelalactone A (cf.
Scheme 1).14 We focused our computational and experimental
studies on species bearing the 1,2-bis(di-tert-butylphosphino)-
ethane (dtbpe)29 ligand. In contrast to the TMEDA ligand,
dtbpe results in the direct formation of the nickelalactone from
ethylene and CO2, which has already been described
theoretically and experimentally.9 We have also found the
transition state corresponding to TSA‑C (G⧧ = 271 kJ/mol);
however, we consider methylation via an SN2 mechanism more
likely (cf. Scheme 2). The activation barrier is well below 100
kJ/mol, which explains why the reaction occurs within minutes
or hours at room temperature. Coordination of iodide to nickel
occurs with a low barrier to give the iodide complex C1, whose
TMEDA-ligated counterpart C has been claimed as an
intermediate on the basis of in situ IR experiments.14 Note
that the calculated free energy required for heterolytic
dissociation of species such as C1 to B1 and iodide is not
very accurate due to the strong solvent effects required for
stabilization of the ions.
question, how a nickelalactone is cleaved by MeI through
experiment and theory.
RESULTS AND DISCUSSION
■
Molecular structures were optimized at the BP86/def2-
SV(P)16−21 level of theory. Calculations were carried out
using the resolution of the identity approximation and
appropriate auxiliary basis functions22 with TURBOMOLE.23
Single-point energies at the M06/def2-TZVP//BP86/def2-
SV(P)24 level were computed with GAMESS-US.25 Gas-phase
Gibbs free energies were obtained within the usual rigid-rotator,
harmonic-oscillator approximation. Free energies in solution
(THF) were obtained by adding solvation free energies
calculated with COSMO-RS26,27 and the parametrization for
BP86/def-TZVP, which requires calculation of BP86/def-
TZVP COSMO and gas-phase energies. This approach proved
to be successful in benchmark calculations involving charge
separation including heterolytic dissociation of an organic
Methylation occurs most likely at the carbonyl oxygen of A1
and yields complex B1, which can be described best by two
resonance structures. Such cationic nickel species are well-
known in olefin polymerization catalysis.30 Furthermore,
methylated pallada-31 and platinalactones32 have been proposed
or observed, and various ring-substituted Pd and Ni analogues
of B1 with dtbpm (1,2-bis(di-tert-butylphosphino)methane) as
Scheme 2. Alternative Mechanistic Pathway for the Methylation of Nickelalactone A1 with MeI
B
dx.doi.org/10.1021/om400262b | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Scheme 3. Formation of the Methylated Nickelalactones B1 and B6 by Protonation of the π Complex D
Figure 1. Solid state (X-ray) molecular structures of B1, B6, and D, with ellipsoids drawn at the 50% probability level. H atoms, BArF anion, and
cocrystallized CH2Cl2 (B6) are omitted for clarity. B1 (two independent molecules): Ni−C3 = 1.926(9)/1.937(10) Å, Ni−O1 = 1.952(7)/1.930(6)
Å, Ni−P1 = 2.154(3)/2.150(3) Å, Ni−P2 = 2.254(3)/2.254(3) Å, C3−Ni−O1 = 83.1(4)/84.7(4)°, P−Ni−P = 91.62(11)/92.45(11)°. B6: Ni−P1
= 2.1761(15) Å, Ni−P2 = 2.2179(16) Å, Ni−O1 = 1.955(5) Å, Ni−C2 = 2.029(7) Å, Ni−C1 = 2.118(7) Å, C1−C2 = 1.418(13), C1−O1 =
1.294(14), P−Ni−P = 92.10(6)°, O1−Ni−C2 = 70.7(3)°. D: Ni−C1 = 1.940(2) Å, Ni−C2 = 1.989(2) Å, Ni−P2 = 2.1551(7) Å, Ni−P1 =
2.1832(7) Å, C1−C2 = 1.418(4) Å, P1−Ni−P2 = 93.65(3)°, C2−Ni−P1 = 118.91(8)°, C1−Ni−C2 = 42.31(10)°, C1−Ni−P2 = 105.63(8)°.
chelating ligand have been isolated and structurally charac-
terized (cf. the Supporting Information). Indeed, B1 is
accessible by protonation of the π complex (dtbpe)Ni(η2-
methyl acrylate) (D) with H(Et2O)2BArF in CH2Cl2 (Scheme
3).
Interestingly, at −78 °C in CD2Cl2 two so far unidentified
intermediates were observed first by 31P NMR, which at −30
°C slowly converted to the four- and five-membered nickela-
cycles B6-BArF (two doublets at δ 95.5 and 94.6 ppm) and B1-
BArF (two doublets at δ 82.4 and 78.9 ppm) favoring the four-
membered-ring structure B6-BArF (ratio B6/B1 = 2/1). Pure
complex B6-BArF was isolated by crystallization from a
CH2Cl2/pentane mixture at −60 °C in 87% yield. The
structure determination of B6 (Figure 1) displays an
interesting, somewhat unexpected molecular geometry in the
crystal. The four-membered nickelacycle is distinctly nonplanar.
The coordination mode of the organic O1−C1−C2 fragment at
Ni more closely resembles a metal-bound η3-oxallyl ligand
arrangement than a planar ring: the O1−C1−C2 plane is tilted
by 61° relative to the O1−Ni1−C2 plane, and the C2 atom has
almost sp2 geometry with an C1−C2−C3 angle of 116.8° and a
C1−C2 distance of 1.418 Å. The Ni−C1 distance is only 2.118
Å. Both resonance structures, oxallyl and distorted four-
membered ring, that can be used to describe the structure of
B6 are depicted in Scheme 3. For brevity, only the four-
membered-ring Lewis structure will be used to describe B6 in
C
dx.doi.org/10.1021/om400262b | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
the following. While methyl and methoxy groups are trans to
each other in the experimental crystal structure, all computed
free energies are given for the cis conformation, as according to
our calculations this geometry is more stable in solution (ΔG =
9 kJ/mol). The barrier for cis/trans rearrangement is low (ΔG⧧
< 40 kJ/mol), and the transition state for this isomerization
(ring inversion) turns out to be the planar structure (see
Scheme 5).
measurable impact on the final ratio B1/B6 (OTf, 2.5/1, BArF,
2/1; cf. Scheme 6). This is in agreement with the results of
Rieger et al., where the reaction with methyl triflate was found
to result in no free methyl acrylate.13
Similar experimental results have very recently been obtained
by Bernskoetter et al.:33 addition of a strong Lewis acid to
either a nickelalactone or an acrylic acid π complex led to the
formation of a four-membered-ring species structurally
analogous to B6, and a five-membered-ring species analogous
to B1 was detected as an intermediate.
The experimental equilibrium (Scheme 4) and rate constants
of the isomerization of cations B1 and B6 have been
Single crystals of B1-BArF were obtained by cooling the
Scheme 4. Equilibrium of B1-BArF and B6-BArF
reaction mixture before the rearrangement could take place and
−
by counterion exchange from OTf− to BArF (Figure 1). The
Ni−O bond of B1-BArF was longer than that in nickelalactone
A1 (1.889 vs 1.952 Å).9
Formation of a Methyl Acrylate π Complex by
Deprotonation. Lewis acid activated nickelalactones can be
deprotonated by strong bases.9,33 Furthermore, since B1 and
B6 form upon protonation of π complex D, the reverse reaction
is an obvious mechanistic pathway for acrylate formation.
Although D is not observed in the overall reaction with methyl
iodide, it could be an intermediate that releases methyl acrylate
afterward. Addition of 1 equiv of the moderately strong neutral
base NEt3 to the B1-BArF/B6-BArF equilibrium mixture in
CD2Cl2 yielded π complex D within minutes (by 31P NMR; cf.
Scheme 7).
Our calculations suggest the deprotonation of the cationic β-
H agostic complex B3 by any reasonably strong base, even a
second (neutral) nickelalactone (ΔG⧧ = 66 kJ/mol, ΔG = −20
kJ/mol relative to B1) is possible: cf. the transition state
structure in Scheme 8. In the computed transition state,
deprotonation occurs at the nonagostic hydrogen. The same
transition state of course applies to the reverse reaction, the
protonation of D. Another possibility would be direct
(de)protonation at nickel involving a hydride intermediate, as
proposed by Bernskoetter et al.33 This seems unlikely for the
studied dtbpe complexes because the hydrides easily rearrange
to the more stable agostic intermediates and the hydride is
shielded by the tert-butyl groups of dtbpe and by the
coordinated acrylate.
determined in CH2Cl2 by integration of the corresponding
31P NMR signals, assuming that the forward and backward
reactions are first order (k1 = 1.17 × 10−3 s−1, k−1 = 0.56 × 10−3
s−1, K = 0.478; cf. the Supporting Information).
The computed isomerization pathway between B6 (η3-oxallyl
complex) and B1 (five-membered ring) involves the β-agostic
intermediates B3 and B5 (cf. Scheme 5). Following the
steepest descent path from B5 to B3 gives an apparent
minimum, the hydride complex B4. It is, however, only a very
flat area on the potential energy surface and rearranges
barrierless to B3.
The computed reaction free energy (−2.5 kJ/mol) and
activation barrier (92 kJ/mol) (both in CH2Cl2) compare very
well with the data obtained experimentally (−1.8 0.1 and 90
2 kJ/mol) in CH2Cl2. As the release of free methyl acrylate is
faster than the equilibration of the cyclic organometallic
intermediates B1 and B6, we consider the latter to be of
minor importance. Since protonation of (dtbpe)Ni(η2-methyl
acrylate) (D) instantaneously yielded a mixture of B1-BArF and
B6-BArF, we tried to obtain B1-OTf by methylation of
nickelalactone A1 with methyl triflate in CH2Cl2. Indeed, the
five-membered ring compound B1-OTf formed and within ca.
15 min partially rearranged to B6-OTf. The counterion had a
The deprotonation would yield not only π complex D but
also the protonated nickelalactone E1, which, as a strong acid,
is capable of inducing side reactions.
Scheme 5. Computed Isomerization Pathway between B6 and B1
D
dx.doi.org/10.1021/om400262b | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Scheme 6. Formation of the Five-Membered Methylated Nickelalactone B1 by Alkylation of Nickelalactone A1
Scheme 7. Deprotonation of Lewis Acid Activated Lactones B1-BArF/B6-BArF To Give the π Complex D
Scheme 8. Deprotonation of the β-Agostic Methylated Nickelalactone B3, Which Is in Rapid Equilibrium with B1 by the Neutral
a
Nickelalactone A1, and the Structure of the Transition State TSB3‑D
a
Note that the non-agostic hydrogen atom is abstracted. Hydrogen atoms that are distant from the reaction center and tert-butyl groups are omitted
for clarity. An analogous transition state likely exists for the protonation/deprotonation of β-agostic B5, which is in rapid equilibrium with B6.
Indeed, upon mixing of nickelalactone A1 and the
equilibrated B1-BArF/B6-BArF (2/1) mixture π complex D
formed (15%) within 20 min in CD2Cl2 at the cost of B1-BArF/
B6-BArF (by 31P NMR), in addition to the protonated
nickelalactone E1 and a new, so far unidentified species with
to be slightly endothermic in terms of absolute energy without
barrier for dissociation, it is slightly exergonic in terms of free
energy. Therefore, either direct deprotonation of C3 or
dissociation of HI is thermodynamically feasible. The most
favorable geometry for dissociation of methyl acrylate is C2.
The endergonic formation of the high-energy intermediate C2
has a very low barrier, and C2 readily dissociates methyl
acrylate to give F.
a
31P resonance at 87.9 ppm. We conclude that the
nickelalactone A1 is able to deprotonate B1/B6 to D; this
happens, however, in an isolated reaction with side/consecutive
reactions that are not clear at this point. Nevertheless, we will
discuss reactions caused by formal release of HI in terms of E1,
our best guess for a protonated species in solution.
Release of Methyl Acrylate from π Complex D.
Although D was not observed experimentally by Kuhn and
̈
Rieger et al.,13,14 it could be an intermediate which is formed by
formal release of HI or deprotonation starting from complexes
B1/C1 or isomers thereof. Methyl acrylate could have been set
free by formal exchange of D with methyl iodide, similar to a
β-H Elimination and Release of Methyl Acrylate from
C2−C4. β-H elimination from the neutral nickel iodide
complex C1 is feasible.14 We have identified several isomeric
transition states (Scheme 9), differing in the arrangement of the
hydrido and iodido ligands: i.e., C2 (cis), C4 (trans), or bound
end-on as hydrogen iodide (C3). The transition state favored
reaction described by Porschke et al., who observed release of
̈
ethylene from the ethylene π complex (dtbpe)Ni(η2-C2H4) and
methyl iodide within hours at room temperature.34 A computed
mechanism was not found to be favorable (one intermediate is
endergonic with ΔG ≈ 100 kJ/mol; cf. the Supporting
Information). Experimentally, in our hands the reaction of
MeI with D after 50 days at room temperature yielded only
by Kuhn et al. has a trans geometry.14
̈
Complexes C2−C4 can release methyl acrylate to form the
iodo hydride complex F, eliminate hydrogen iodide, or may be
deprotonated to give the methyl acrylate complex D (cf.
Scheme 9). Although elimination of molecular HI is computed
E
dx.doi.org/10.1021/om400262b | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Scheme 9. β-H Elimination from C1 and Consecutive Reaction Pathways
traces of free methyl acrylate and 24% of complex G,
of I, (dtbpe)NiCl, has been reported in the literature.35
According to our calculations, the corresponding dimer of I
with a triplet ground state (UKS: ⟨Sz⟩ = 1; ⟨S2⟩ = 2.034) is less
stable than the monomer in solution due to entropic effects
(ΔG ≈ 29 kJ/mol). However, a minor peak with the correct
mass for the dimer has been detected in the LIFDI mass
spectrum (m/z 1006.4).
determined by 31P NMR spectroscopy (Scheme 10).
Scheme 10. Attempted Liberation of Methyl Acrylate from π
Complex D
Formation of Methyl Propionate. We have identified two
possible pathways for the release of methyl acrylate from
nickelalactone A1 and MeI: i.e., via formal release of HI or
deprotonation resulting in the protonated nickelalactone E1 (or
other protonated species) and via β-H elimination, successive
liberation of methyl acrylate, and formation of complex F,
respectively. Experimentally, in addition to free methyl acrylate,
methyl propionate is formed as the product of a formal
protonation of nickelalactone A1 or its methylated derivatives.
Hydrogen could be transferred as a cation (proton), anion
(hydride), or radical. Our experimental data do not indicate a
hydride transfer, nor do our computational studies point to a
favorable hydride pathway (cf. the Supporting Information).
The formal release of 1 equiv of HI during the reaction could
cause the formation of methyl propionate by protonation. In
order to discuss the relative acidities of involved species, we
have computed reaction free energies, ΔGHI, for protonation of
Alternatively, our calculations suggest that (dtbpe)NiI2 (H)
reacts with D to free methyl acrylate and the Ni(I) species I (cf.
Scheme 11). While I certainly has a doublet ground state
(UKS: ⟨Sz⟩ = 0.5; ⟨S2⟩ = 0.770), the singlet-optimized structure
of the planar complex H is only 3 kJ/mol lower in energy than
the triplet-optimized structure of H with pseudotetrahedral
geometry (computed energies are therefore given for the singlet
ground state). Indeed, H is capable of liberating methyl acrylate
1
from D, as experiments in CD2Cl2 indicate. H NMR shows
two broad signals at δ −8.6 and 10.4 ppm that point to the
presence of a paramagnetic species, which by mass spectrom-
etry we assign to be I (a signal with a mass/charge ratio
corresponding to the singly charged cation of I was observed at
m/z 502.9). The same species was detected in the overall
reaction of nickelalactone A1 with MeI. The μ2-Cl-bridged,
pseudotetrahedral, paramagnetic dimer of the chloride analogue
iodide to HI in THF. While the hydrido complex F (ΔGHI
=
162 kJ/mol) is not acidic, the protonated nickelalactone E1 is a
sufficiently acidic species in solution (ΔGHI = 60 kJ/mol) and
Scheme 11. Formation of Free Methyl Acrylate via Decomposition of π Complex D
F
dx.doi.org/10.1021/om400262b | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
a
Scheme 12. Formation of Methyl Propionate by Protonation
a
The energy of the first protonation step is given for protonation by species E1, which is assumed to serve catalytically in transferring the proton
generated, along with the methyl acrylate complex D. Consequently, all further free energies are independent of whether protonation is initiated by
E1, by other protonated species, or directly by HI.
could catalytically transfer a proton to the methylated
nickelalactone B1 (ΔGHI = 41 kJ/mol; conjugate base D).
Formation of methyl propionate and (dtbpe)NiI2 (H) from
nickelalactone A1 or intermediate C1 by protonation is
certainly exergonic, and the highest barriers in both processes
are ca. 100 kJ/mol (cf. Scheme 12). Free energies for the initial
protonation are given relative to E1, as the most likely acid in
solution. Protonation of nickelalactone A1 with HBArF resulted
first in the protonated isomer E1, which partially converts to
the unknown species at 87.9 ppm mentioned above and
propionate complex E3 after 1 h. After 24 h only E3 was
detected by 31P NMR, which is in agreement with the
thermodynamics predicted computationally (cf. Scheme 12).
Figure 2. Solid-state (X-ray) molecular structure of E3, with ellipsoids
The identity of E3 has been confirmed by synthesis on a
drawn at the 50% probability level. H atoms and the BArF anion are
separate route: starting from nickel(II) propionate, dtbpe was
omitted for clarity. E3 (two independent molecules): Ni−O =
added and NaBArF was used to change the counterion. The
complex was isolated upon crystallization from a CH2Cl2/
pentane mixture at −60 °C (cf. Figure 2). As already pointed
out, A1 and B1 did not react cleanly to give E3 and D but gave
a mixture of D, E1, and an unidentified product.
1.924(4)/1.935(4) Å, Ni−O2 = 1.938(4)/1.925(4) Å, Ni···C1 =
2.281(7)/2.275(7), Ni−P1 = 2.171(2)/2.166(2) Å, Ni−P2 =
2.167(2)/2.165(2) Å, O1−Ni−O2 = 67.87(2)/67.96(19)°, P1−Ni−
P2 = 90.46(7)/91.01(7)°.
Since, according to our calculations and experiments, radical
I originates most likely from the reaction of π complex D with
H, we also took into account radical pathways involving
complex F. H-radical transfer to C1 is thermodynamically
feasible; subsequent elimination of methyl propionate occurs
via a low barrier and is exergonic (cf. Scheme 13). Alltogether,
we consider the radical pathway the most favorable H transfer
from F.
Overall, the most likely mechanism is the formation of π
complex D by either direct deprotonation or dissociation of HI
after β-H elimination. On the basis of our computational data,
we consider the deprotonation scenario more likely; the β-H
elimination path, however, cannot be excluded, because the
most important quantities in this context, the reaction free
energies for heterolytic dissociation of iodide-containing
complexes to free iodide and cationic nickel complexes, are
not accurate enough. Experimentally, the deprotonation
mechanism has been confirmed in terms of its elementary
reaction, while β-H elimination cannot be ruled out. After β-H
elimination, release of methyl acrylate with formation of F is
less likely than elimination of HI but again cannot be ruled out.
Avoiding the Formation of Methyl Propionate and
Re-entering a Catalytic Cycle. Although we were not able to
definitely identify the major decomposition reaction yielding
methyl propionate, our results allow us to avoid these side
reactions: we were able to convert methylated nickelalactone
B1 quantitatively to π complex D by deprotonation with NEt3.
As most amine bases are rapidly quaternized with MeI, we
chose the weak and only partially soluble base Cs2CO3 to
reduce the formation of methyl propionate: indeed, in the
presence of Cs2CO3 in addition to propionate (19%) and
acrylate (31%) now significant amounts of π complex D (50%)
G
dx.doi.org/10.1021/om400262b | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Scheme 13. Formation of Methyl Propionate by H-Radical Transfer
Figure 3. Two favorable mechanisms for the formation of methyl acrylate: direct deprotonation of cationic intermediate B3 by a second
nickelalactone A1 (red pathway) and β-H elimination and subsequent elimination of methyl acrylate (black pathway). The relative stability of neutral
and dissociated ionic species is uncertain, due to charge separation and a different magnitude of solvent stabilization.
were formed. This was not the case in the absence of Cs2CO3
(68% acrylate, 32% propionate, 0% D).
According to our calculations, the final ligand exchange in D
from methyl acrylate to ethylene complex Oa crucial step to
close a hypothetical catalytic cycle (see Figure 3)via a
tetrahedral intermediate9 is endergonic (ΔG = 21 kJ/mol; see
Scheme 14). This endergonic free reaction energy adds to the
quantitative. This trend can be understood in terms of the
Dewar−Chatt−Duncanson model: sodium acrylate is a worse π
acceptor than methyl acrylate, which shows up in the higher
orbital energy of the unoccupied π* orbital that has the right
symmetry for back-bonding.
In conclusion, we have gathered theoretical and experimental
evidence that methylation of nickelalactones proceeds through
an SN2-type reaction with a low barrier. An equilibrium
between five- and four-membered-ring methylated, cationic
lactones via β-H elimination has been identified. This is similar
to the findings of Bernskoetter et al. for the reaction of
nickelalactones with Lewis acids,33 and we believe that most of
our mechanistic results for the alkylation apply to the reaction
with Lewis acids as well.
Scheme 14. Exchange of Methyl Acrylate in π Complex D
with Ethylene
Contrasting the postulated mechanisms in the literature by
Kuhn et al., an S 2 pathway explains the fast reaction of
̈
N
nickelalactones with MeI observed experimentally, even under
mild reaction conditions. Furthermore, we have identified
unproductive pathways leading to byproducts, such as methyl
propionate as well as Ni(I) and Ni(II) species (cf. Scheme 15,
red pathway). Release of methyl acrylate in the Ni-mediated
reaction of CO2, ethylene, and MeI is driven by this
decomposition to catalytically inactive species. Addition of
suitable bases suppresses the formation of byproducts to some
extent and leads instead to the formation of the methyl acrylate
π complex. The ligand exchange at nickel of methyl acrylate by
barrier of nickelalactone formation. This might be the reason
why reinitiation of nickelalactone formation from π complex D
has not yet been observed (cf. Figure 3). Nevertheless, the
barrier for exchange of methyl acrylate is still feasible (ΔG⧧ ≈
104 kJ/mol). The corresponding reaction of sodium acrylate
complexes is exergonic.9 The calculations agree with exper-
imental findings: substitution of ethylene by methyl acrylate
and replacement of sodium acrylate by ethylene is fast and
H
dx.doi.org/10.1021/om400262b | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Scheme 15. Hypothetic Catalytic Cycle and Side Reactions (in Red) Leading to the Formation of Methyl Acrylate and Methyl
Propionate
was covered by 0.5° ω scans in all cases. Intensities were corrected for
Lorentz and polarization effects, and an empirical absorption
correction was applied using SADABS36 on the basis of the Laue
symmetry of the reciprocal space. The structures were solved by direct
methods and refined against F2 with a full-matrix least-squares
algorithm using the SHELXTL (version 2008/4) software package.37
Hydrogen atoms were treated using appropriate riding models, unless
otherwise noted for the particular structures. Infrared spectra were
measured on Bruker FT-IR Equinox 55 and Tensor 27 spectrometers
as KBr pellets. Bis(di-tert-butylphosphino)ethane (dtbpe),29 (dtbpe)-
ethylene that is necessary for catalysis is endergonic, at least for
chelating bisphosphine ligands. This is the major hurdle for
catalytic formation of methyl acrylate. Future work should aim
at reducing the energy required for this ligand exchange under
the constraint of efficient nickelalactone formation from CO2
and ethylene. As in the in the case of sodium acrylate, the
compatibility of the base with reaction conditions remains
problematic.
38
39
NiI2,34 (dtbpe)Ni(CH2CH2COO),9 NaBArF, H(Et2O)2BArF and
nickel propionate40were prepared according to literature procedures.
NaBArF was azeotropically distilled with toluene as an alternative
drying procedure to remove the water. All other starting materials were
purchased in reagent grade purity from Acros, Aldrich, Fluka, or Strem
and used without further purification.
EXPERIMENTAL SECTION
■
General Comments. All manipulations were performed under an
inert atmosphere of dry argon in a glovebox (MBraun) or using
standard Schlenk techniques. Diethyl ether, n-hexane, tetrahydrofuran
(THF), and CH2Cl2 were dried using a MBraun SPS-800 solvent
purification system. Pentane was distilled over benzophenone sodium
ketyl. Deuterated solvents were purchased from Deutero GmbH.
CD2Cl2 was distilled over CaH2, and THF-d8 and benzene-d6 were
distilled over sodium. All solvents were degassed three times through
freeze−pump−thaw cycles prior to use and stored over molecular
sieves in Teflon valve ampules under argon. Methyl acrylate, NEt3, and
methyl iodide were distilled under argon prior to use. NMR spectra
were recorded on Bruker AC200−300, Bruker ARX250, and Bruker
DRX300, 500, and 600 spectrometers. The chemical shifts are given in
parts per million and are referenced to the nondeuterated signals of
the deuterated solvents used for 1H and 13C NMR and relative to 85%
H3PO4 for 31P NMR. Mass spectra were recorded on a JEOL JMS-700
spectrometer and a Bruker ApexQe Apollo II FT-ICR instrument.
Elemental analyses were performed at the “Mikroanalytisches
[(dtbpe)Ni(η2-methyl acrylate)] (D). [Ni(cod)2] (500 mg, 1.82
mmol) was suspended in Et2O (8 mL), methyl acrylate (235 mg, 2.73
mmol) was added, and the mixture was stirred until a red solution was
formed. A solution of dtbpe (637 mg, 2.00 mmol) in Et2O (5 mL) was
added to the reaction mixture, and the deep red solution was stirred
for 2 h while a yellow solid precipitated. The supernatant was removed
via filter cannula, and the precipitate was washed with pentane (2 × 5
mL). The resulting solid was dried under vacuum to yield D (670 mg,
80%) as a yellow powder. Anal. Calcd for C22H46O2P2Ni: C, 57.04; H,
10.01. Found: C, 56.29; H, 10.02. 1H NMR (300.13 MHz, CD2Cl2): δ
1.19 (dd, JH,H = 12.0 Hz, JP,H = 27.0 Hz, 18H, C(CH3)3), 1.23 (dd, JH,H
= 12.0 Hz, JP,H = 18.0 Hz, 18H, C(CH3)3), 1.63−1.76 (m, 4H,
PCH2CH2P), 1.81−1.89 (m, 1H, CH2), 2.09−2.16 (m, 1H, 
CH2), 2.73−2.82 (m, 1H, CHCOO), 3.45 (s, 3H, OCH3). 13C{1H}
NMR (75.56 MHz, CD2Cl2): δ 23.6−24.2 (m, PCH2CH2P), 30.5−
31.0 (m, C(CH3)3), 31.7 (dd, JP,C = 1.5 Hz, JP,C = 22.6 Hz, CH2),
34.8−35.3 (m, C(CH3)3), 40.2 (dd, JP,C = 1.5 Hz, JP,C = 14.3 Hz,
=CHCOO), 50.0 (s, OCH3), 176.9 (s, CO). 31P{1H} NMR (121.49
MHz, CD2Cl2): δ 87.5 (d, JP,P = 53.5 Hz), 95.5 (d, JP,P = 53.5 Hz). IR
Laboratorium der Chemischen Institute der Universitat Heidelberg”.
̈
For X-ray crystal structure analyses suitable crystals were mounted
with a perfluorinated polyether oil on nylon loops or micro mounts
and flash-frozen under a nitrogen stream at 200(2) K on the
goniometer head. Data collection was performed with Mo Kα
radiation (λ = 0.71073 Å) from a Incoatec IμS microsource on a
Bruker APEX-II diffractometer. A complete sphere in reciprocal space
I
dx.doi.org/10.1021/om400262b | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
(KBr): ν 2944, 2894, 2848, 1659, 1469, 1439, 1364, 1005, 936, 669,
454 cm−1. MS (LIFDI) m/z: 461.93 [M − H].
18.1 Hz, JP,C = 22.6 Hz, PCH2), 30.4 (d, JP,C = 4.5 Hz, C(CH3)3), 30.8
(d, JP,C = 3.0 Hz, C(CH3)3), 35.5 (d, JP,C = 10.6 Hz, C(CH3)3), 37.0
(dd, JP,C = 1.5 Hz, JP,C = 4.5 Hz, CH2COO), 38.3 (d, JP,C = 19.6 Hz,
C(CH3)3), 56.6 (s, OCH3), 118.0 (s, p-Ar), 125.2 (q, JF,C = 271.7 Hz,
CF3), 129.1−129.8 (m, m-Ar), 135.4 (s, o-Ar), 162.3 (q, JB,C = 49.8
Hz, ipso-CAr), 195.4 (d, JP,C = 12.1 Hz, CO). 31P{1H} NMR (121.65
MHz, CD2Cl2): δ 78.9 (d, JP,P = 3.6 Hz), 82.4 (d, JP,P = 3.6 Hz). IR
(KBr): ν 2963, 1615, 1469, 1379, 1355, 1278, 1132, 900, 682 cm−1.
MS (LIFDI) m/z: 463.1 [M − BArF].
Single crystals were obtained by slow evaporation of Et2O. Crystal
data: yellow crystal (polyhedron), dimensions 0.13 × 0.11 × 0.10
mm3, crystal system monoclinic, space group P21/n, Z = 4, a =
8.0740(9) Å, b = 22.595(3) Å, c = 13.9661(16) Å, α = 90°, β =
94.095(2)°, γ = 90°, V = 2541.3(5) Å3, ρ = 1.211 g/cm3, θmax = 27.15°,
mean redundancy 5.07 and completeness 99.9% for a resolution of
0.78 Å, 28964 reflections measured, 5631 unique reflection (R(int) =
0.0552), 4337 observed reflections (I > 2σ(I)), μ = 0.90 mm−1, Tmin
=
Single crystals were obtained from a mixture of CH2Cl2 and pentane
at −60 °C. Crystal data: yellow crystal (polyhedron), dimensions 0.07
0.89, Tmax = 0.92, 256 parameters refined, the three hydrogen atoms at
the coordinated double bond were refined isotropically, goodness of fit
1.03 for observed reflections, final residual values R1(F) = 0.037,
wR2(F2) = 0.081 for observed reflections, residual electron density
−0.27 to 0.40 e Å−3.
× 0.07 × 0.06 mm3, crystal system triclinic, space group P1, Z = 4, a =
̅
12.9017(14) Å, b = 20.347(2) Å, c = 25.577(3) Å, a = 70.864(3)°, β =
86.826(3)°, γ = 74.241(3)°, V = 6100.5(11) Å3, ρ = 1.492 g/cm3, θmax
= 21.04°, mean redundancy 4.12 and completeness 100.0% for a
resolution of 0.99 Å, 54200 reflections measured, 13150 unique
reflections (R(int) = 0.1264), 6687 observed reflections (I > 2σ(I)), μ
= 0.53 mm−1, Tmin = 0.96, Tmax = 0.97, 1790 parameters refined,
goodness of fit 1.02 for observed reflections, final residual values R1(F)
= 0.075, wR2(F2) = 0.182 for observed reflections, residual electron
density −0.52 to 0.64 e Å−3.
[(dtbpe)Ni(CH(CH3)COOCH3)BArF] (B6-BArF). A solution of
H(Et2O)2BArF (437 mg, 431.6 μmol) in CH2Cl2 (4 mL) was added
at −78 °C to a solution of D (200 mg, 431.7 μmol) in CH2Cl2 (10
mL). The reaction mixture was stirred at this temperature for 15 min
and then warmed to room temperature over 3 h. The solvent was
evaporated, and the crude product of the reaction was crystallized at
−60 °C from a mixture of CH2Cl2 and pentane to yield B6-BArF (500
mg, 87%) as an orange solid. Anal. Calcd for C54H59BF24O2P2Ni: C,
Reaction of [(dtbpe)Ni(CH(CH3)COOCH3)BArF] (B6-BArF) and
NEt3. To a solution of B6-BArF (10 mg, 21.6 μmol) in CD2Cl2 (0.6
mL) was added NEt3 (2.1 μL, 21.6 μmol). The product in solution was
characterized by 31P NMR spectroscopy.
Reaction of [(dtbpe)Ni(CH2CH2COO)] (A1) with [(dtbpe)Ni-
(CH(CH3)COOCH3)BArF] (B6-BArF). B6-BArF (20 mg, 15.1 μmol)
and A1 (7 mg, 15.1 μmol) were dissolved in THF-d8 (0.6 mL), and
the products were characterized by 31P NMR spectroscopy.
Reaction of [(dtbpe)Ni(η2-methyl acrylate)] (D) with MeI. To
a solution of D (20 mg, 43.2 μmol) in benzene-d6 (0.6 mL) was added
MeI (27.0 μL, 432 μmol). The solution was stirred for several days at
ambient temperature, and the products were characterized by 31P
NMR spectroscopy.
1
48.86; H, 4.48; P, 4.67. Found: C, 48.57; H, 4.28; P, 4.64. H NMR
(300.51 MHz, CD2Cl2): δ 1.28 (d, JH,H = 19.0 Hz, 3H, CH3), 1.32 (dd,
JH,H = 5.1 Hz, JP,H = 13.5 Hz, 18H, C(CH3)3), 1.44 (dd, 18 H, JH,H
=
5.1 Hz, JP,H = 13.5 Hz, 18H, C(CH3)3), 1.83−2.13 (m, 4H,
PCH2CH2P), 3.22−3.25 (m, 1H, NiCH), 3.78 (s, 3H, OCH3), 7.56
(s, 4H, p-H), 7.72 (s, 8H, o-H). 13C{1H} NMR (150.93 MHz,
CD2Cl2): δ 15.1 (s, CH3), 20.6 (dd, JP,C = 6.0 Hz, JP,C = 22.6 Hz,
PCH2), 26.8 (dd, JP,C = 13.6 Hz, JP,C = 25.7 Hz, PCH2), 30.1−30.8 (m,
C(CH3)3), 37.2 (dd, JP,C = 15.1 Hz, JP,C = 27.2 Hz, C(CH3)3), 38.0
(dd, JP,C = 1.5 Hz, JP,C = 19.6 Hz, C(CH3)3), 41.4 (dd, JP,C = 9.1 Hz,
JP,C = 33.2 Hz, Ni-CH), 53.9 (s, OCH3), 118.1 (s, p-Ar), 125.2 (q, JF,C
= 273.18 Hz, CF3), 129.5 (q, JF,C = 30.19 Hz, m-Ar), 135.4 (s, o-Ar),
162.4 (q, JB,C = 49.8 Hz, ipso-CAr), 167.7 (s, CO). 31P{1H} NMR
(121.65 MHz, CD2Cl2): δ 94.6 (d, JP,P = 6.1 Hz), 95.5 (d, JP,P = 6.1
Hz). IR (KBr): ν 2969, 2872, 1611, 1495, 898, 744, 682, 498 cm−1. MS
(LIFDI) m/z: 463.1 [M − BArF].
Reaction of [(dtbpe)NiI2] (H) with [(dtbpe)Ni(η2-methyl
acrylate)] (D). H (20 mg, 31.8 μmol) and D (14.5 mg, 31.8 μmol)
were dissolved in CD2Cl2 (0.6 mL). The solution was stirred for 30
min and then subjected to characterization by 1H and 31P NMR
spectroscopy.
[(dtbpe)Ni(CH3CH2COO)BArF] (E3). Nickel propionate (30 mg,
145.1 μmol) and dtbpe (47 mg, 145.1 μmol) were dissolved in CH2Cl2
(10 mL), and the mixture was stirred for 10 min. NaBArF (129 mg,
145.1 μmol) was added to the mixture. The red solution was stirred for
10 min, and a white solid precipitated. The supernatant was removed
via filter cannula. Afterward the solvent was evaporated and the
precipitate was washed with pentane (2 × 5 mL). The resulting solid
was dried under vacuum to afford 145 mg (76%) of E3 as an orange
powder. Anal. Calcd for C54H59BF24O2P2Ni: C, 48.47; H, 4.37; P, 4.72.
Found: C, 48.44; H, 4.35; P, 4.76. 1H NMR (300.51 MHz, CD2Cl2): δ
1.01 (t, 3H, JH,H = 9.0 Hz, CH3), 1.51 (d, 36H, JP,H = 12.0 Hz,
Single crystals were obtained from a mixture of CH2Cl2 and pentane
at −60 °C. Crystal data: yellow crystal (needle), dimensions 0.15 ×
0.09 × 0.07 mm3, crystal system orthorhombic, space group Pbca, Z =
8, a = 18.8623(15) Å, b = 24.459(2) Å, c = 25.783(2) Å, α = 90°, β =
90°, γ = 90°, V = 11895.4(18) Å3, ρ = 1.482 g/cm3, θmax = 22.72°,
mean redundancy 11.76 and completeness 100.0% for a resolution of
0.92 Å, 97690 reflections measured, 8004 unique reflections (R(int) =
0.0801), 5536 observed reflections (I > 2σ(I)), μ = 0.50 mm−1, Tmin
=
0.93, Tmax = 0.97, 852 parameters refined, H2 at C2 was refined
isotropically, goodness of fit 1.02 for observed reflections, final residual
values R1(F) = 0.051, wR2(F2) = 0.115 for observed reflections,
residual electron density −0.38 to 0.48 e Å−3.
C(CH3)3), 1.82 (d, 4H, JP,H = 9.0 Hz, PCH2CH2P), 2.16 (q, 2H, JH,H
=
[(dtbpe)Ni(CH2CH2COOCH3)BArF] (B1-BArF). Lactone A1 (200
mg, 446 μmol) was dissolved in CH2Cl2 (8 mL), and MeOTf (50.5
μL, 446 μmol) was added to the solution, which was immediately
cooled to −78 °C and stirred for 5 min. This mixture was added to a
cooled solution (−78 °C) of NaBArF (396 mg, 447 μmol) in CH2Cl2
(30 mL). The yellow solution was stirred for 10 min while a yellow
solid precipitated. The supernatant was removed via filter cannula, and
the solvent was evaporated; during these procedures the temperature
was not allowed to reach more than −40 °C to avoid isomerization.
The crude product of the reaction was crystallized at −60 °C from a
mixture of CH2Cl2 and pentane to yield B1-BArF (385 mg, 64%) as a
yellow solid. Anal. Calcd for C54H59BF24O2P2Ni: C, 48.86; H, 4.48; P,
4.67. Found: C, 48.76; H, 4.33; P, 4.61. 1H NMR (300.51 MHz,
CD2Cl2): δ 1.31 (bs, 2H, NiCH2), 1.39 (d, 18H, JP,H = 15.0 Hz,
C(CH3)3), 1.41 (d, JP,H = 12.0 Hz, 18H, C(CH3)3), 1.59−1.70 (m, 2H,
PCH2CH2P), 1.87−2.02 (m, 2H, PCH2CH2P), 2.45−2.53 (m, 2H,
CH2COO), 3.88 (s, 3H, OCH3), 7.56 (s, 4H, p-H), 7.72 (s, 8H, o-H).
13C{1H} NMR (150.93 MHz, CD2Cl2): δ 9.8 (q, JP,C = 33.2 Hz,
9.0 Hz, CH2COO), 7.55 (s, 4H, p-H), 7.71 (s, 8H, o-H). 13C{1H}
NMR (150.93 MHz, CD2Cl2): δ 7.9 (s, CH3), 22.9 (dd, JP,C = 16.6 Hz,
JP,C = 19.6 Hz, PCH2), 30.0 (s, C(CH3)3), 31.1 (s, CH2), 37.9 (dd, JP,C
= 7.6 Hz, JP,C = 9.0 Hz, C(CH3)3), 118.0 (s, p-Ar), 125.2 (q, JF,C
=
271.7 Hz, CF3), 129.1−129.8 (m, m-Ar), 135.4 (s, o-Ar), 162.3 (q, JB,C
= 48.3 Hz, ipso-CAr), 201.0 (s, COO). 31P{1H} NMR (121.65 MHz,
CD2Cl2): δ 101.6 (s). IR (KBr): ν 2972, 1612, 1567, 1485, 1356, 1279,
1131 cm−1. MS (LIFDI) m/z: 449.2 [M − BArF].
Single crystals could be obtained from a mixture of CH2Cl2 and
pentane at −35 °C. Crystal data: orange crystal (polyhedron),
dimensions 0.10 × 0.08 × 0.07 mm3, crystal system triclinic, space
group P1, Z = 4, a = 12.6924(5) Å, b = 18.3585(7) Å, c = 25.7549(10)
̅
Å, α = 86.316(1)°, β = 88.748(1)°, γ = 88.911(1)°, V = 5986.5(4) Å3,
ρ = 1.457 g/cm3, θmax= 22.72°, mean redundancy 2.95 and
completeness 100.0% for a resolution of 0.92 Å, 48015 reflections
measured, 16082 unique reflections (R(int) = 0.0648), 9177 observed
reflections (I >2σ(I)), μ = 0.49 mm−1, Tmin = 0.95, Tmax = 0.97, 1691
parameters refined, goodness of fit 1.02 for observed reflections, final
NiCH2), 18.7 (dd, JP,C = 6.0 Hz, JP,C = 19.6 Hz, PCH2), 27.1 (dd, JP,C
=
J
dx.doi.org/10.1021/om400262b | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
residual values R1(F) = 0.062, wR2(F2) = 0.130 for observed
reflections, residual electron density −0.44 to 0.96 e Å−3.
Reaction of [(dtbpe)Ni(CH2CH2COO)] (A1) with H(Et2O)2BArF.
A1 (20 mg, 44.6 μmol) and H(Et2O)2BArF (45 mg, 44.6 μmol) were
dissolved in CD2Cl2 (0.6 mL). The solution was characterized by 31P
NMR spectroscopy after 10 min, 1 h, and 1 day.
Nakamura, Y.; Yamamoto, A. J. Organomet. Chem. 1979, 171, 103−
112.
(7) Graham, D. C.; Mitchell, C.; Bruce, M. I.; Metha, G. F.; Bowie, J.
H.; Buntine, M. A. Organometallics 2007, 26, 6784−6792.
(8) Papai, I.; Schubert, G.; Mayer, I.; Besenyei, G.; Aresta, M.
Organometallics 2004, 23, 5252−5259.
Reaction of [(dtbpe)Ni(CH2CH2COO)] (A1) with MeI. (a) To a
solution of A1 (20 mg, 44.5 μmol) in THF-d8 (1 mL) were added
mesitylene (10.0 μL, 72.1 μmol, internal standard) and MeI (27.8 μL,
445 μmol). The solution was stirred for 24 h at ambient temperature
́
(9) Lejkowski, M. L.; Lindner, R.; Kageyama, K.; Bodizs, G. E.;
́
Plessow, P. N.; Muller, I. B.; Schafer, A.; Rominger, F.; Hofmann, P.;
Futter, F.; Schunk, S. A.; Limbach, M. Chem. Eur. J. 2012, 18, 14017−
̈
̈
14025.
1
and characterized by H and 31P NMR spectroscopy.
(10) (a) Hoberg, H.; Ballesteros, A.; Sigan, A.; Jegat, C.; Milchereit,
(b) To a solution of A1 (67.4 mg, 150 μmol) in THF-d8 (1 mL)
was added MeI (93 μL, 1.5 mmol). The solution was stirred for 12 h at
ambient temperature, evaporated to 0.2 mL, and precipitated with
Et2O (2 mL). I precipitated as a pale red solid (25 mg, 33%). Anal.
Calcd: C, 42.89, H, 8.00. Found: C, 42.07, H, 7.94%. MS (FAB+)
measured: 503.1 [C18H40P2NiI].
A. Synthesis 1991, 395−398. (b) Castano, A. M.; Echavarren, A. M.
̃
Organometallics 1994, 13, 2262−2268. (c) Schonecker, B.; Walther,
̈
D.; Fischer, R.; Nestler, B.; Braunlich, G.; Eibisch, H.; Droescher, P.
̈
Tetrahedron Lett. 1990, 31, 1257−1260. (d) Braunlich, G.; Fischer, R.;
̈
Nestler, B.; Walther, D. Z. Chem. 1989, 29, 416−417. (e) Fischer, R.;
Walther, D.; Braunlich, G.; Undeutsch, B.; Ludwig, W.; Bandmann, H.
̈
Reaction of [(dtbpe)Ni(CH2CH2COO)] (A1) with MeI and
Cs2CO3. To a suspension of A1 (20 mg, 44.5 μmol) and Cs2CO3 (58
mg, 178 μmol) in THF-d8 (1 mL) were added mesitylene (10.0 μL,
72.1 μmol) and MeI (27.8 μL, 445 μmol). The solution was stirred for
J. Organomet. Chem. 1992, 427, 395−407. (f) Yamamoto, T.; Sano, K.;
Osakada, K.; Komiya, S.; Yamamoto, A.; Kushi, Y.; Tada, T.
Organometallics 1990, 9, 2396−2403.
1
2 h, and the resulting compounds were characterized by H and 31P
(11) Yamashita, K.; Chatami, N. Synlett 2005, 919−922.
(12) Aye, K. T.; Colpitts, D.; Ferguson, G.; Puddephatt, R. J.
Organometallics 1988, 7, 1454−1456.
NMR spectroscopy.
CCDC 927636 (B1), 927637 (B6), 927638 (D), and 927639 (E3)
contain supplementary crystallographic data for this paper. These data
can be obtained free of charge from The Cambridge Crystallographic
Data Centre via www.ccdc.cam.ac.uk/ data_request/cif.
(13) Bruckmeier, C.; Lehenmeier, M. W.; Reichardt, R.; Vagin, S.;
Rieger, B. Organometallics 2010, 29, 2199−2202.
(14) Lee, S. Y. T.; Cokoja, M.; Drees, M.; Li, Y.; Mink, J.; Herrmann,
W. A.; Kuhn, F. K. ChemSusChem 2011, 4, 1275−1279.
̈
(15) Drees, M.; Cokoja, M.; Kuhn, F. E. ChemCatChem 2012, 4,
̈
ASSOCIATED CONTENT
■
1703−1712.
S
* Supporting Information
(16) Weigend, F.; Ahlrichs, R. Phys. Chem. Chem. Phys. 2005, 7,
3297−3305.
Text, figures, tables, and CIF files giving additional
experimental details, characterization data, details of the
calculations, and crystallographic data for B1, B6, D, and E3.
This material is available free of charge via the Internet at
http://pubs.acs.org.
(17) Dirac, P. A. M. Proc. R. Soc. London, Ser. A 1929, 123, 714−733.
(18) Becke, A. Phys. Rev. A 1988, 38, 3098−3100.
(19) Perdew, J. P.; Wang, Y. Phys. Rev. B 1992, 45, 13244−13249.
(20) Vosko, S. H.; Wilk, L.; Nusair, M. Can. J. Phys. 1980, 58, 1200−
1211.
(21) Perdew, J. P. Phys. Rev. B 1986, 33, 8822−8824.
(22) Weigend, F. Phys. Chem. Chem. Phys. 2006, 8, 2051−2057.
(23) TURBOMOLE V6.3 2011, a development of University of
Karlsruhe and Forschungszentrum Karlsruhe GmbH, 1989−2007,
TURBOMOLE GmbH, since 2007; available from http://www.
turbomole.com.
(24) Zhao, Y.; Truhlar, D. G. Theor. Chem. Acc. 2008, 120, 215−241.
(25) Schmidt, M. W.; Baldridge, K. K.; Boatz, J. A.; Elbert, S. T.;
Gordon, M. S.; Jensen, J. H.; Koseki, S.; Matsunaga, N.; Nguyen, K. A.;
Su, S.; Windus, T. L.; Dupuis, M.; Montgomery, J. A. J. Comput. Chem.
1993, 14, 1347−1363.
(26) Eckert, F.; Klamt, A. COSMOtherm, Version C3.0, Release 12.01;
COSMOlogic, Leverkusen, Germany, 2012.
(27) Eckert, F.; Klamt, A. AlChE J. 2002, 48, 369−385.
(28) Deglmann, P.; Schenk, S. J. Comput. Chem. 2012, 33, 1304−
1320.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: ph@oci.uni-heidelberg.de (P.H.); michael.limbach@
basf.com (M.L.).
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
P.N.P., R.L., M.L., and P.H. work at CaRLa of Heidelberg
University, being cofinanced by Heidelberg University, the state
of Baden-Wurttemberg, and BASF SE. The support of these
̈
institutions is gratefully acknowledged. We are indebted to
Professor Peter Kundig, CaRLa Fellow 2013, for careful
̈
proofreading of the final manuscript.
(29) Benn, R.; Jolly, P. W.; Joswig, T.; Mynott, R.; Schick, K. P. Z.
Naturforsch., B 1986, 41B, 680−691.
REFERENCES
(30) Leatherman, M. D.; Svejda, S. A.; Johnson, L. K.; Brookhart, M.
J. Am. Chem. Soc. 2003, 125, 3068−3081.
■
(1) Cokoja, M.; Rieger, B.; Herrmann, W. A.; Kuhn, F. E. Angew.
̈
́
(31) Aresta, M.; Pastore, C.; Giannoccaro, P.; Kovacs, G.;
Chem., Int. Ed. 2011, 50, 8510−8537.
́
Dibenedetto, A.; Papai, I. Chem. Eur. J. 2007, 13, 9028−9034.
(2) Huang, K.; Sun, C.-L.; Shi, Z.-J. Chem. Soc. Rev. 2011, 40, 2435−
(32) Ganis, P.; Orabona, I.; Ruffo, F.; Vitagliano, A. Organometallics
1998, 17, 2646−2650.
2452.
(3) Peters, M.; Kohler, B.; Kuckshinrichs, W.; Leitner, W.;
̈
(33) Jin, D.; Schmeier, T. J.; Williard, P. G.; Hazari, N.; Bernskoetter,
W. H. Organometallics 2013, 32, 2152−2159.
Markewitz, P.; Muller, T. E. ChemSusChem 2011, 4, 1216−1240.
̈
(4) (a) Tanaka, R.; Yamashita, M.; Nozaki, K. J. Am. Chem. Soc. 2009,
131, 14168−14169. (b) Schaub, T.; Paciello, R. A. Angew. Chem., Int.
Ed. 2011, 50, 7278−7282.
(34) Bach, I.; Goddard, R.; Kopiske, C.; Seevogel, K.; Porschke, K.-R.
̈
Organometallics 1999, 18, 10−20.
(35) Mindiola, D.; Waterman, R.; Jenkins, D. M.; Hillhouse, G. Inorg.
Chim. Acta 2003, 345, 299−308.
(5) Hoberg, H.; Peres, Y.; Kruger, C.; Tsay, Y. H. Angew. Chem., Int.
̈
Ed. Engl. 1987, 26, 771−773.
(6) (a) Ishizu, J.; Yamamoto, T.; Yamamoto, A. Bull. Chem. Soc. Jpn.
1978, 51, 2646−2650. (b) Yamamoto, T.; Ishizu, J.; Komiya, S.;
(36) Sheldrick, G. M. Program SADABS 2008/1 for absorption
correction; Bruker Analytical X-ray Division, Madison, WI, 2008.
K
dx.doi.org/10.1021/om400262b | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
(37) Software package SHELXTL 2008/4 for structure solution and
refinement: Sheldrick, G. M. Acta Crystallogr. 2008, A64, 112−122.
(38) Reger, D. L.; Wright, T. D.; Little, C. A.; Lamba, J. J. S.; Smith,
M. D. Inorg. Chem. 2001, 40, 3810−3814.
(39) Brookhart, M.; Grant, B.; Volpe, A. F. Organometallics 1992, 11,
3920−3922.
(40) Bickley, R. I.; Edwards, H. G. M.; Gustar, R.; Rose, S. J. J. Mol.
Struct. 1991, 248, 237−250.
L
dx.doi.org/10.1021/om400262b | Organometallics XXXX, XXX, XXX−XXX