Lewis acid induced β-elimination from a nickelalactone: Efforts toward acrylate production from CO2 and ethylene

Article abstract of DOI:10.1021/om400025h

The Lewis acid tris(pentafluorophenyl)borane was found to rapidly promote ring-opening β-hydride elimination in a 1,1′-bis(diphenylphosphino) ferrocene (dppf) nickelalactone complex under ambient conditions. The thermodynamic product of nickelalactone ring-opening was characterized as (dppf)Ni(CH(CH₃)CO₂BAr^f₃), the result of β-hydride elimination and subsequent 2,1-insertion from a transient nickel(II) acrylate hydride intermediate. Treatment of (dppf)Ni(CH(CH₃)CO₂BAr^f₃) with a nitrogen-containing base afforded a diphosphine nickel(0) η²- acryl borate adduct. Formation of the diphosphine nickel(0) η²- acryl borate adduct completes a net conversion of nickelalactone to acrylate species, a significant obstacle to catalytic acrylate production from CO ₂ and ethylene. Displacement of the η²-acrylate fragment from the nickel center was accomplished by addition of ethylene to yield a free acrylate salt and (dppf)Ni(CH₂CH₂).

Full text of DOI:10.1021/om400025h

Article
pubs.acs.org/Organometallics
Lewis Acid Induced β‑Elimination from a Nickelalactone: Efforts
toward Acrylate Production from CO2 and Ethylene
Dong Jin,† Timothy J. Schmeier,‡ Paul G. Williard,† Nilay Hazari,‡ and Wesley H. Bernskoetter*,†
†Department of Chemistry, Brown University, Providence, Rhode Island 02912, United States
‡Department of Chemistry, Yale University, New Haven, Connecticut 06520, United States
S
* Supporting Information
ABSTRACT: The Lewis acid tris(pentafluorophenyl)borane
was found to rapidly promote ring-opening β-hydride
elimination in a 1,1′-bis(diphenylphosphino)ferrocene (dppf)
nickelalactone complex under ambient conditions. The
thermodynamic product of nickelalactone ring-opening was
characterized as (dppf)Ni(CH(CH3)CO2BArf3), the result of
β-hydride elimination and subsequent 2,1-insertion from a
transient nickel(II) acrylate hydride intermediate. Treatment of (dppf)Ni(CH(CH3)CO2BArf3) with a nitrogen-containing base
afforded a diphosphine nickel(0) η2-acryl borate adduct. Formation of the diphosphine nickel(0) η2-acryl borate adduct
completes a net conversion of nickelalactone to acrylate species, a significant obstacle to catalytic acrylate production from CO2
and ethylene. Displacement of the η2-acrylate fragment from the nickel center was accomplished by addition of ethylene to yield
a free acrylate salt and (dppf)Ni(CH2CH2).
INTRODUCTION
■
The utilization of CO2 as a feedstock for the production of
commodity chemicals potentially offers a more cost-effective
and renewable alternative to fossil fuel based carbon sources in
the chemical industry.1 Unfortunately, the kinetic and
thermodynamic stability of CO2 has limited its exploitation
thus far to a handful of commercial chemicals.2 One method to
surmount this stability is the reduction of CO2 via coupling to
other relatively high energy small molecules, a technique that is
already employed in the commercial production of urea and
polycarbonates from ammonia and epoxides, respectively.3 The
functionalization of CO2 with light olefins to produce α,β-
unsaturated carboxylic acids is yet another intriguing target for
this methodology, with potentially significant implications for
the manufacture of acrylates used in superabsorbent polymers,
elastomers, and detergents.4
Figure 1. Pioneering reports of CO2−ethylene coupling at transitions
metal complexes.5
that β-hydride elimination from computationally predicted
metalalactone intermediates is swift.5b,7,8a,9 Unfortunately, the
strong oxophilicity of molybdenum and tungsten has hampered
reductive acrylate removal using methods compatible with
catalysis, as proven examples of acrylate liberation require either
a harsh base (i.e., butyllithium) or a strong electrophile (i.e.,
iodomethane).7a,9
By contrast, removal of acrylate from nickel should be more
facile owing to its lower oxophilicity and reduction potential as
well as its greater tolerance of protic functional groups. The
principal obstacles for catalytic acrylate formation at nickel
seem to lie in the high pressures required for most oxidative
couplings (up to 40 bar) and a strong reticence for β-hydride
elimination from the nickelalactone intermediates. The
Transition metal promoted coupling of CO2 and ethylene
toward acrylate formation has been explored as an alternative to
currently used propylene oxidation technology since the
seminal reports of Hoberg and Carmona in the 1980s (Figure
1).5 These pioneering investigators independently pursued new
routes for CO2−ethylene coupling using zerovalent nickel6 and
group VI metals,7 respectively, though catalytic activity
remained elusive. Subsequent experimental and computational
mechanistic studies on these reactions suggest that the early
and late metal complexes likely share several common
intermediates on the desired catalytic pathway, but are
challenged by different steps in the proposed cycle (Figure
2).8 In the case of group VI metals, the oxidative coupling of
CO2 and ethylene appears relatively facile, occurring at ambient
temperature and pressures. The couplings at molybdenum and
tungsten have consistently afforded acrylate products, implying
Received: January 12, 2013
© XXXX American Chemical Society
A
dx.doi.org/10.1021/om400025h | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Recently Limbach and co-workers have circumvented these
barriers to β-hydride elimination by adding external bases such
as sodium tert-butoxide to diphosphine nickelalactone species,
which deprotonate the β-hydrogen directly without necessitat-
ing transfer of the hydride to nickel.10 This approach affords
sodium acrylate (NaCO2CHCH2) in good yield, and by
repeated sequential additions of CO2, ethylene, and base,
several equivalents of sodium acrylate may be obtained in one
reaction vessel. Unfortunately, the strong sodium base required
for the deprotonation is not compatible with the high CO2
pressure needed for nickelalactone formation, obviating
catalytic production under a constant set of reaction conditions.
An alternative method that has interested our laboratory is the
facilitation of β-hydride elimination by ring-opening of the
metalalactone. This approach could afford a transient three-
coordinate nickel species, which alleviates the ring strain and
orbital constraints. Herein we describe the synthesis, character-
ization, and pathway of formation for a ring-opened
diphosphine nickelalactone complex using a Lewis acid as
well as the implications of Lewis acid addition on base-
mediated acrylate formation.
Figure 2. Proposed catalytic cycle for production of acrylic acid from
CO2 and ethylene.
pressures typically required for nickel-promoted CO2−ethylene
coupling may be attained in autoclave reactors, but can spur
undesired side reactions, such as ligand degradation and
multiple CO2 insertions.6 The high CO2 pressure can also
interfere with later steps in catalytic acrylate production (vide
infra).10 Nevertheless, it appears that inducing β-hydride
elimination from nickelalactone complexes is the more
systematic barrier, as the few well-defined examples of
nickelalactone β-hydride elimination occur only with ancillary
ligand activation or the use of vast excesses of electrophile.11
The origin of stability to β-hydride elimination in square
planar nickelalactone species is probably the result of two
interrelated factors: the strain energy associated with distorting
the five-membered lactone ring and the absence of an accessible
low-lying orbital for a β-agostic interaction. Crystallographic
evidence from multiple nickelalactone structures indicates the
key C−H bonds β-to-the-metal are oriented away from the
nickel and would likely require significant twisting of the five-
membered ring to be brought within a covalent interaction
distance.6c,10,11a,c,12 However the observation of swift elimi-
nation from proposed metalalactones of molybdenum and
tungsten suggests that additional orbital constraints may also
contribute.5b,7,8a,9 Brookhart, Green, and others have shown
that β-hydride elimination in late transition metals occurs via
formation of an agostic intermediate, which necessitates an
empty orbital on the metal to accept electron donation from
the target C−H bond.13 As is illustrated in the classic slow
RESULTS AND DISCUSSION
■
The interconversion between a four-coordinate nickelalactone
species and the corresponding η2-acrylic acid complex is central
to the catalytic production of acrylates from the coupling of
CO2 and ethylene. However, to date, there is no evidence of
direct interconversion between these species. Given the
significance of this transformation, our laboratory sought to
leverage one of the few easily prepared ligand platforms known
to support isolable examples of both isomers on nickel, 1,1′-
bis(diphenylphosphino)ferrocene (dppf). Walther and co-
workers have previously demonstrated that the metalalactone
species (dppf)Ni(CH2CH2CO2) (A) and the η2-acrylic acid
complex (dppf)Ni(CH2CHCO2H) (B) may be prepared
independently via ligand substitution reactions (Figure 4).12b In
our hands, no interconversion between these two isomers was
detected in DMSO, THF, or arene solution at temperatures up
to 45 °C, above which decomposition of A occurred via
reductive decoupling of CO2 and ethylene. Similar to the
observations of other diphosphine nickelalactone compounds,10
associative (or associative interchange) mechanism for ligand
14
substitution at square planar d8-complexes, the vacant dx −y
2
2
orbital in the nickel(II) lactone is inaccessible to incoming
ligands (or β-hydrogens) due to the position of the existing
square planar ligands (Figure 3). Thus coordination of the β-
C−H bond in nickelalactone complexes would require use of a
higher energy empty orbital or significant ligand rearrangement.
Figure 3. Qualitative d-orbital splitting diagram for square planar
nickel(II) metalalactone.
Figure 4. Synthetic routes for (dppf)Ni(CH2CH2CO2) (A) and
(dppf)Ni(CH2CHCO2H) (B).
B
dx.doi.org/10.1021/om400025h | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
elimination from A may be induced by addition of a strong
external base, such as sodium tert-butoxide, in DMSO to afford
sodium acrylate (detected by NMR spectroscopy) and
(dppf)2Ni along with a quantity of nickel(0) particulates (eq
1).15However slightly weaker neutral bases such as 1,8-
resonances at 15.7 and 34.3 ppm in the 31P NMR spectrum.
The corresponding 1H NMR spectrum revealed multiplet
signals at 0.84 and 2.21 ppm in a 1:1 integration ratio. These
resonances were identified as CH2 units on the basis of 1H−13C
HSQC NMR data and are consistent with formation of the
Lewis acid activated metalalactone, complex 1 (Figure 6). The
orange-colored solution containing complex 1 persisted for
only 2−3 h at ambient temperature, after which complete
conversion to a subsequent nickel species was observed. The
final product, (dppf)Ni(CH(CH3)CO2BArf3) (2), also exhibits
two doublet resonances in the 31P NMR spectrum, now at 22.4
1
and 36.2 ppm. The H NMR spectrum again displays two
diazabicyclo[5.4.0]undec-7-ene (DBU) or even tert-
butyliminotri(pyrrolidino)phosphorane (BTPP) were ineffec-
tive at promoting elimination from A. Given the inherent
compatibility challenges of using strong anionic bases in
conjunction with CO2 in coupling reactions, we directed our
attention toward Lewis acid promoted methods of inducing β-
hydride elimination from nickelactone species.
Lewis Acid Induced Ring-Opening. Inspiration for this
approach was derived from observations by Rieger and Kuhn
that limited quantities of methyl acrylate may be obtained from
nickelalactone complexes by treatment with a large excess of
methyl iodide (Figure 5).11 In such systems it was proposed
intriguing resonances, an apparent doublet of doublets at 0.23
ppm and a complex multiplet at 2.03 ppm, in a 3:1 integration
1
ratio. Upon 31P decoupling of the H NMR spectrum the
resonance at 0.23 ppm collapsed to a doublet, indicating
coupling constants of JH,H = 7.0 Hz and JP,H = 7.8 Hz. Likewise
the multiplet at 2.03 ppm simplified to a quartet upon 31P
1
decoupling. The corresponding H−13C HSQC NMR spec-
1
trum indicates correlations between these H NMR peaks at
0.23 and 2.03 ppm and respective 13C NMR peaks at 12.36 and
34.58 ppm. The phase of the 1H−13C HSQC NMR correlations
are the same as those attributed to methine positions of the
dppf ligand.
The NMR spectroscopic data are consistent with the
presence of Ni-α-CH and Ni-β-CH3 moieties in 2. This
assignment was confirmed by single-crystal X-ray diffraction
experiments (Figure 7). The molecular structure of 2 exhibits a
mildly distorted square planar geometry about the nickel,
composed of two Ni−P bonds from the dppf ligand, a Ni−C
linkage, and a close Ni−O interaction to the carbonyl. The data
were of sufficient quality such that all hydrogens were located
and refined from the Fourier map, establishing the C(2) unit as
a methine and C(3) as a methyl. The Ni(1)−C(2) distance of
1.978(9) Å is comparable to numerous other reports of nickel−
carbon bonds in metallacyclic carboxylates.10,12 The Ni(1)−
O(1) and C(1)−O(1) bond lengths of 1.905(4) and 1.27(1) Å,
respectively, indicate a strong interaction between the carbonyl
oxygen and nickel as well as a modest reduction of the C−O π-
bond. Additionally, the Ni(1)−P(1) bond (2.237(2) Å) trans to
the carbon ligand is slightly elongated compared to the Ni(1)−
P(2) distance (2.135(2) Å) opposite the carbonyl oxygen
ligand, consistent with the expected trans influence trend.
Significantly, the formation of 2 likely proceeds via β-hydride
elimination from complex 1 to afford an unobserved nickel
Figure 5. Methyl iodide induced elimination from nickelalactone
complexes.
that electrophilic attack by methyl cation on the Ni−O bond
cleaves the five-membered ring to assist β-hydride elimination,
though the role of the iodide counteranion was also deemed
significant. In order to gain insight into the β-hydride
elimination reactivity of such transient three-coordinate nickel
species, nickelalactone complex A was treated with the potent
neutral Lewis acid tris(pentafluorophenyl)borane (BArf3) at
ambient temperature. When this reaction was monitored by
NMR spectroscopy, immediate formation of a new nickel
species was indicated by the appearance of two doublet
Figure 6. Pathways for formation of complex 2 and reversible β-hydride elimination.
C
dx.doi.org/10.1021/om400025h | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Figure 7. Molecular structure of (dppf)Ni(CH(CH3)CO2BArf3) (2) with ellipsoids at 30% probability. All hydrogen atoms and a cocrystallized
diethyl ether molecule are omitted for clarity. Full view (a) and nickel coordination core (b).
acrylate hydride intermediate (Figure 6). Subsequent 2,1-
insertion of the acryl borate ligand would then produce the
isolated compound 2. Similar 2,1-insertions of acrylates into
late transition metal hydrides have been observed by Brookhart
and others.16 This conversion of A to 2 provides a rare well-
defined example of β-hydride elimination from a nickelalactone
species and establishes the role of Lewis acids in promoting the
transformation. Intriguingly, the ring-opened species 2 may
alternatively be prepared by addition of BArf3 to the η2-acrylic
acid complex, B (Figure 6). Treatment of a benzene-d6 solution
of B with 1 equiv of Lewis acid results in quick consumption of
the starting material and formation of complexes 1 and 2 via a
short-lived intermediate (t1/2 ≈ 15 min). The intermediate was
characterized only by NMR spectroscopy and features a pair of
doublet peaks in the 31P NMR spectrum at 23.42 and 33.42
ppm. The 19F NMR resonances at −165.7, −160.2, and −134.8
ppm are significantly shifted from those of free BArf3 and are
comparable to those observed in 2, suggesting the complex
contains a coordinated BArf3 unit. In addition, the observation
to the chemical shift of the −OH proton in B, suggests that this
unstable intermediate is simply a borane adduct of the η2-acrylic
acid complex B (Figure 6). More significantly, Lewis acid
addition to B produces complexes 1 and 2 simultaneously, with
the mixture gradually shifting to solely 2 over 8 h. This
contrasts the sequential formation of 1 then 2 observed in BArf3
addition to A, indicating that these two synthetic reactions
enter the equilibrium process at different intermediates (Figure
6). Our observations are most consistent with the reaction of B
and BArf3 affording the unobserved nickel(II) acrylate hydride
species, which can then diverge to form complexes 1 and 2 with
competitive rates of 1,2- and 2,1-insertion, respectively. Over
time the reversible reaction equilibrates to the thermodynami-
cally more stable 2,1-insertion product.17
Deprotonation of Complex 2. Having successfully
induced β-hydride elimination from a stable nickelalactone
species using a Lewis acid, experimental efforts were turned
toward expelling acrylate from complex 2. Unlike the
nickelalactone species A, deprotonation of 2 by neutral organic
bases proved an effective method of accessing acrylate. Use of
1
of a broad peak at 8.45 ppm in the H NMR spectrum, similar
D
dx.doi.org/10.1021/om400025h | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
the sterically hindered phosphazene base, BTPP, resulted in the
bases for acrylate liberation may enhance the viability of
deprotonation techniques to overcome this barrier under high
CO2 pressure. Ideally, the use of CO2-compatible bases such as
carbonates in conjunction with mild Lewis acids or frustrated
Lewis pairs could allow for a practical catalytic production of
acrylate. Investigations toward this goal are still in progress in
our laboratories, with some limited stoichiometric success in
deprotonating complex 2 with cesium carbonate in tetrahy-
drofuran. However, these reactions produce low yields of free
acryl borates salts (15−25% by NMR spectroscopy) over seven
days at 50 °C.
formation of the η2-acrylate complex 3 (eq 2),over two days at
Reactivity toward Ethylene and Carbon Dioxide.
Having successfully induced acrylate formation from complex
2 by deprotonation using several bases, our focus shifted
toward closing a hypothetical catalytic cycle (Figure 8) by
ambient temperature. Complex 3 exhibited limited solubility in
hydrocarbon solvents, but proved modestly soluble in ethereal
and halogenated solvents. The 1H NMR spectrum of a
chlorobenzene-d5 solution of 3 displays resonances at 2.06,
3.16, and 3.49 ppm assigned as the vinylic protons of the bound
1
olefin. The assignments were confirmed by H−13C HSQC
NMR experiments, which indicate correlations to 13C chemical
shifts at 46.30 (CH2) and 52.45 (CH) ppm. These resonances
are analogous to those reported for the η2-acrylic acid complex
B.12b The H NMR spectrum of 3 also exhibits the expected
1
peaks for the conjugate acid of BTPP, including a broad N−H
resonance at 4.75 ppm. The 31P NMR spectrum completed the
characterization with two doublets at 19.8 and 29.6 ppm
assigned to the dppf ligand, as well as a singlet at 23.1 ppm
from [BTPPH]+. Addition of slightly weaker bases including
DBU to complex 2 also promoted some acrylate formation,
although in the case of this bicyclic amine, the η2-acrylate
complex was not formed as selectively. Addition of 1 equiv of
DBU to complex 2 over two days produced a mixture of the
free [H2CCHCO2BArf3]−[DBUH]+ salt,18 a complex anal-
ogous to 3, and some free dppf ligand.
The mechanism for deprotonation of complex 2 by nitrogen
base has not been examined experimentally, but could well
occur via β-hydride elimination to yield the unobserved
nickel(II) acrylate hydride species, then proton abstraction by
base. DuBois and co-workers have previously reported a pKa
value of 22.2 (acetonitrile) for a related nickel(II) hydride
complex, suggesting a transient nickel(II) acrylate hydride
species would be sufficiently acidic for deprotonation by
amines.19 Direct deprotonation of complex 2 cannot be
eliminated completely, but seems less likely, as abstraction of
the acidic α-C−H bond adjacent to the carbonyl would afford a
nickel carbene species, which must then rearrange to produce
the observed complex 3. Alternatively, deprotonation of the less
acidic β-C−H bond could directly yield complex 3, but this
proton would probably only be sufficiently acidic for amine
deprotonation if activated by nickel via a β-agostic interaction.
No such interaction is evident in the solid-state structure of 2
(Figure 7), making deprotonation of the unobserved nickel(II)
acrylate hydride species the leading hypothesis.
The observed deprotonation of 2 with BTPP and DBU
stands in contrast to the reactivity of nickelalactone A, which,
similar to other diphosphine nickelacycles,10 requires stronger
bases to induce elimination. The role of Lewis acid in
facilitating nickelalactone deprotonation by more mild bases
is quite significant to the larger challenge of catalytic acrylate
production from CO2 and ethylene. As discussed above, several
nickel compounds capable of coupling CO2−ethylene into
nickelalatone species at elevated pressures have been
reported.6,10 However, inducing elimination to produce acrylate
in a fashion compatible with the presence of excess CO2 has
remained a persistent barrier. The ability to use more mild
Figure 8. Hypothetical cycle for Lewis acid/base promoted coupling
of CO2 and ethylene to form acrylate.
coupling CO2 and ethylene to re-form the nickelalactone
species A. Though A is analogous to other nickelalactone
complexes prepared directly by CO2−ethylene coupling at
zerovalent nickel, the dppf-substituted congener has not been
prepared in this manner.12b Our initial studies into the
regeneration of A treated complex 3 with 2 atm each of
ethylene and CO2, which resulted in immediate formation of
the η2-ethylene complex (dppf)Ni(CH2CH2) (4). After a
further four days at ambient temperature, no evidence of
formation of the nickelalactone A or other activation of CO2
was observed by NMR spectroscopy. Heating the mixture at
temperatures up to 100 °C resulted only in degradation of 4.
The ethylene complex 4 was characterized by multinuclear
NMR spectroscopy and combustion analysis, as well as
independent preparation by addition of ethylene to a mixture
of Ni(COD)2 (COD = 1,5-cyclooctadiene) and the dppf ligand
(Figure 9).
Solutions of complex 4 are moderately stable (3−4 days) in
arene solvent under N2 or Ar atmospheres. The 31P NMR
spectrum of 4 in benzene-d6 displays a singlet at 24.9 ppm,
indicating a higher symmetry about the metal center than
complex 3. Along with other peaks from the dppf ligand, the 1H
NMR spectrum exhibits a slightly broadened resonance at 2.77
ppm, which correlates to a methylene 13C chemical shift of
1
46.24 ppm in the H−13C HSQC NMR spectrum. These data
and the 1H NMR integration are consistent with a single bound
E
dx.doi.org/10.1021/om400025h | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Figure 9. Synthetic pathways for the formation of complex 4.
ethylene molecule, analogous to reports of several other
diphosphine nickel(0) ethylene complexes.10,12b,20
system for CO2−ethylene to acrylate at nickel and related
metals.
The inability to produce the nickelalactone A from 3 in
NMR tube experiments prompted further investigations at
higher pressure, as multiple reports of CO2−ethylene coupling
at zerovalent nickel utilize pressures in excess of 20 bar.6,10 For
these larger scale experiments, isolated (dppf)Ni(COD) was
used as the source of zerovalent nickel to better approximate
procedures applied in previous nickelalactone syntheses.5a,6,10
Unfortunately, a range of reaction conditions with pressures up
to p(CO2)/p(C2H4) = 50 bar/13 bar at 57 °C for four days in
tetrahydrofuran failed to produce any detectable quantities of A
by NMR spectroscopy. Under these conditions, an approximate
2:1 ratio of complex 4 and (dppf)2Ni was observed for the
organometallic nickel products following pressure release and
extraction with dichloromethane. Although these high-pressure
tests portend little promise for the dppf platform in nickel-
catalyzed coupling of CO2 and ethylene for acrylates, the
technique of Lewis acid induced elimination from nickel-
alactones described here could well be applied to other ligand
platforms better optimized for CO2−ethylene coupling.
Investigations into this application as well as the scope of
viable Lewis acids for ring-opening and CO2-compatible bases
for acrylate extrusion are ongoing in our laboratories.
EXPERIMENTAL SECTION
■
General Considerations. All manipulations were carried out using
standard vacuum, Schlenk, cannula, or glovebox techniques. Ethylene
was purchased from Corp Brothers and stored over 4 Å molecular
sieves in heavy-walled glass vessels prior to use. Argon and nitrogen
were purchased from Corp Brothers and used as received. (dppf)Ni-
(CH2CH2CO2) (A) and (dppf)Ni(CH2CHCO2H) (B) were
prepared according to literature procedures.12b All other chemicals
were purchased from Aldrich, VWR, Strem, Fisher Scientific, or
Cambridge Isotope Laboratories. Volatile, liquid chemicals were dried
over 4 Å molecular sieves and distilled prior to use. Solvents were
dried and deoxygenated using literature procedures.21
1H, 13C, 19F, and 31P NMR spectra were recorded on Bruker DRX
400 Avance and 300 Avance MHz spectrometers. 1H and 13C chemical
shifts are referenced to residual solvent signals; 19F and 31P chemical
shifts are referenced to the external standards C6H5CF3 and H3PO4,
respectively. Probe temperatures were calibrated using ethylene glycol
and methanol as previously described.22 Unless otherwise noted, all
NMR spectra were recorded at 23 °C. IR spectra were recorded on a
Jasco 4100 FTIR spectrometer. X-ray crystallographic data were
collected on a Bruker D8 QUEST diffractometer. Samples were
collected in inert oil and quickly transferred to a cold gas stream. The
structures were solved from direct methods and Fourier syntheses and
refined by full-matrix least-squares procedures with anisotropic thermal
parameters for all non-hydrogen atoms. Elemental analyses were
performed at Robertson Microlit Laboratories, Inc., in Madison, NJ, or
Atlantic Microlab, Inc., in Norcross, GA.
CONCLUDING REMARKS
■
The potential economic and sustainability advantages of
producing acrylates by coupling CO2 with ethylene at nickel
have resulted in many independent investigations over the last
three decades. However the vast majority of these studies have
been frustrated by the inability of nickelalactone complexes to
undergo β-hydride elimination. In this work, a Lewis acid,
tris(pentafluorophenyl)borane, has been found to promote
rapid β-hydride elimination from an isolable nickelalactone
species, (dppf)Ni(CH2CH2CO2) (A), under ambient con-
ditions. The reversible β-hydride elimination ultimately results
in the formation of the thermodynamically stable 2,1-acryl
borate insertion product (dppf)Ni(CH(CH3)CO2BArf3) (2).
Interestingly, the Lewis acid activation renders 2 more facile
toward deprotonation by external base than the starting
nickelalactone species. Treatment of 2 with nitrogen-containing
bases formed either a free acrylate salt or an η2-acrylate
coordination complex with the nickel. The coordinated borate-
substituted acrylate may easily be substituted by ethylene, but
the (dppf)Ni platform proved resistant to CO2−ethylene
coupling. Nevertheless, the ability to both ring-open the
nickelalactone with a discrete Lewis acid and promote acrylate
liberation with more mild bases provides a key step toward
establishing a viable catalytic process that can withstand the
high CO2 pressure conditions typical of nickel-promoted
couplings. In addition, demonstration of a sequential Lewis
acid, base addition to enable acrylate liberation from a
nickelalactone fosters consideration of cocatalysts, such as
frustrated Lewis pairs, for promoting a functional catalytic
Observation of (dppf)Ni(CH2CH2CO2B(C6F5)3) (1). A 20 mL
scintillation vial was charged with 0.013 g (0.019 μmol) of
(dppf)Ni(CH2CH2CO2) (A), 0.010 g (0.019 μmol) of B(C6F5)3,
and approximately 2 mL of CH2Cl2. The deep orange solution was
stirred for 5 min, and the volatiles were removed under vacuum. The
resulting solid was then dissolved in C6D6 for NMR study. Identical
NMR spectra may be taken at 10 °C to slow the conversion of 1 to 2
1
for longer time scale experiments. H NMR (C6D6): δ 0.84 (m, 2H,
Ni-α-CH2), 2.21 (m, 2H, Ni-β-CH2), 3.36 (s, 2H, CpH), 3.62 (s, 2H,
CpH), 3.72 (s, 2H, CpH), 4.12 (s, 2H, CpH), 6.98−7.67 (Ph).
2
31P{1H} NMR (C6D6): δ 15.7 (d, JP,P 15.8 Hz, 1P, PPh2), 34.3 (d,
2JP,P 15.8 Hz, 1P, PPh2). 19F NMR (C6D6): δ −166.60 (t), −161.47
(t), −135.77 (d).
(dppf)Ni(CH(CH3)(CO2B(C6F5)3) (2). A 20 mL scintillation vial was
charged with 0.096 g (0.142 μmol) of (dppf)Ni(CH2CHCO2H)
(B), 0.073 g (0.143 μmol) of B(C6F5)3, and approximately 5 mL of
toluene. The orange solution was stirred for one day, and the volatiles
were removed under vacuum. The resulting solid was washed with 2
mL of pentane, extracted with diethyl ether, and chilled at −35 °C to
afford 128 mg (76%) of 2 as orange crystals. Anal. Calcd for
C55H32BF15FeNiO2P2: C, 55.18; H, 2.69. Found: C, 55.50; H, 2.67. 1H
NMR (C6D6): δ 0.23 (dd, JH,H 7.0 Hz, JP,H 7.8 Hz, 3H, Ni-β-CH3),
2.03 (m, 1H, Ni-α-CH), 3.59 (s, 1H, CpH), 3.67 (s, 1H, CpH), 3.71
(s, 1H, CpH), 3.73 (s, 2H, CpH), 3.81 (s, 1H, CpH), 4.05 (s, 1H,
CpH), 4.35 (s, 1H, CpH), 6.91−7.11 (m, 12H, Ph), 7.34−7.39 (m,
2H, Ph), 7.51−7.59 (m, 4H, Ph), 7.67−7.70 (m, 2H, Ph). 31P{1H}
2
2
NMR (C6D6): δ 22.4 (d, JP,P 20.0 Hz, 1P, PPh2), 36.2 (d, JP,P 20.0
Hz, 1P, PPh2). 13C{1H} NMR (C6D6): δ 12.36 (Ni-β-CH3), 34.58
(Ni-α-CH), 73.36, 74.81, 75.04, 75.40, 76.18 (Cp), 128.66−129.15,
131.36, 133.03, 133.15, 133.85, 133.98, 133.12, 134.12, 134.24, 134.74,
F
dx.doi.org/10.1021/om400025h | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
134.86 (aryl) 181.20 (CO2). 19F NMR (C6D6): δ −166.60 (t),
−161.02 (t), −135.37 (d). IR (KBr): νCO = 1644 cm−1.
Weinheim, 2010. (e) Aresta, M. Dibenedetto. A. Industrial Utilization
of Carbon Dioxide. In Developments and Innovation in Carbon Dioxide
Capture and Storage Technology; Maroto-Valer, M. M., Ed.; Woodhead:
Cambridge, 2010; pp 377−410. (f) Feedstocks for the Future; Bozell, J.;
Patel, M. K., Eds.; ACS Symposium Series 921; American Chemical
Society: Washington, DC, 2006. (g) Tolman, W. B. Carbon Dioxide
Reduction and Uses As a Chemical Feedstock. Activation of Small
Molecules: Organometallic and Bioinorganic Perspectives; Wiley-VCH:
Weinheim, 2006; pp 1−35.
(2) (a) Quadrelli, E. A.; Centi, G.; Duplan, J.-L.; Perathoner, S.
ChemSusChem 2011, 4, 1194. (b) Cokoja, M.; Bruckmeier, C.; Rieger,
B.; Herrmann, W. A.; Kuhn, F. E. Angew. Chem., Int. Ed. 2011, 50,
8510.
(3) Sakakura, T.; Choi, J.-C.; Yasuda, H. Chem. Rev. 2007, 107, 2365.
(4) (a) Darensbourg, D. J. Inorg. Chem. 2010, 49, 10765. (b) Aresta,
M.; Dibenedetto, A. Dalton Trans. 2007, 2975. (c) Aresta, M.;
Dibenedetto, A. Catal. Today 2004, 98, 455. (d) Patil, Y.; Tambade, P.
J.; Jagtap, S. R.; Bhanage, B. M. Front. Chem. Eng. China 2010, 4, 213.
(5) (a) Hoberg, H.; Schaefer, D. J. Organomet. Chem. 1983, 251, C51.
(b) Alvarez, R.; Carmona, E.; Cole-Hamilton, D. J.; Galindo, A.;
Gutierrez-Puebla, E.; Monge, A.; Poveda, M. L.; Ruiz, C. J. Am. Chem.
Soc. 1985, 107, 5529.
(6) (a) Hoberg, H.; Ballesteros, A.; Sigan, A.; Jegat, C.; Barhausen,
D.; Milchereit, A. J. Organomet. Chem. 1991, 407, C23. (b) Hoberg,
H.; Ballesteros, A. J. Organomet. Chem. 1991, 411, C11. (c) Hoberg,
H.; Peres, Y.; Kruger, C.; Tsay, Y. H. Angew. Chem., Int Ed. Engl. 1987,
26, 771.
(7) (a) Alvarez, R.; Carmona, E.; Galindo, A.; Gutierrez, E.; Marin, J.
M.; Monge, A.; Poveda, M. L.; Ruiz, C.; Savariault, J. M.
Organometallics 1989, 8, 2430. (b) Galindo, A.; Pastor, A.; Perez, P.;
Carmona, E. Organometallics 1993, 12, 4443. (c) Collazo, C.; del Mar
Conejo, M.; Pastor, A.; Galindo, A. Inorg. Chim. Acta 1998, 272, 125.
(8) (a) Bernskoetter, W. H.; Tyler, B. T. Organometallics 2011, 30,
520. (b) Graham, D. C.; Mitchell, C.; Bruce, M. I.; Metha, G. F.;
Bowie, J. H.; Buntine, M. A. Organometallics 2007, 26, 6784. (c) Papai,
I.; Schubert, G.; Mayer, I.; Besenyei, G.; Aresta, M. Organometallics
2004, 23, 5252. (d) Schubert, G.; Papai, I. J. Am. Chem. Soc. 2003, 125,
14847. (e) Sakaki, S.; Mine, K.; Hamada, T.; Arai, T. Bull. Chem. Soc.
Jpn. 1995, 68, 1873. (f) Sakaki, S.; Mine, K.; Taguchi, D.; Arai, T. Bull.
Chem. Soc. Jpn. 1993, 66, 3289. (g) Dedieu, A.; Ingold, F. Angew.
Chem., Int. Ed. Engl. 1989, 28, 1694.
[(dppf)Ni(η2-CH2CH−CO2B(C6F5)3)][HBTPP] (3). A 20 mL
scintillation vial was charged with 0.035 g (0.029 μmol) of
(dppf)Ni(CH(CH3)(CO2B(C6F5)3) (2), 9 μL (0.029 μmol) of
BTPP, and approximately 1 mL of benzene. The solution was stirred
for two days, resulting in precipitation of a yellow solid. The solid was
collected by filtration to afford 40 mg (91%) of 3 as a yellow powder.
The material may be extracted with THF if necessary to remove trace
nickel metal particulates. Anal. Calcd for C71H65BF15FeNiN4O2P3: C,
1
56.49; H, 4.34; N, 3.71. Found: C, 55.96; H, 4.48; N, 3.51. H NMR
(C6D5Cl): δ 0.97 (s, 9H, N-C(CH3)3), 1.39 (s, 12H, N-β-CH2), 2.06
(m, 1H, η2-CH2CH), 2.70 (s, 12H, N-α-CH2), 3.16 (m, 1H, η2-
CH2CH), 3.49 (m, 1H, η2-CH2CH), 3.79 (s, 1H, CpH), 3.89 (s,
1H, CpH), 3.96 (m, 1H, CpH), 3.99 (s, 1H, CpH), 4.02 (s, 2H, CpH),
4.47 (s, 1H, CpH), 4.62 (s, 1H, CpH), 4.75 (br, 1H, NH) 6.92−7.24
(m, 12H, Ph), 7.56−7.99 (m, 8H, Ph). 31P{1H} NMR (C6D5Cl): δ
19.8 (d, 2JP,P 22.8 Hz, 1P, PPh2), 23.1 (s, 1P, [HBTPP]+), 29.6 (d, 2JP,P
22.8 Hz, 1P, PPh2). 13C{1H} NMR (C6D5Cl): δ 26.01 (N-β-CH2),
30.93 (N-C(CH3)3), 46.30 (η2-CH2CH), 47.40 (N-α-CH2), 52.45
(η2-CH2CH), 70.09, 70.44, 72.62, 73.03, 73.46, 74.78, 74.90 (Cp),
127.55, 131.50, 131.94, 133.47, 135.05, 136.03, 136.38, 147.80, 149.37
(aryl) three aryl signals not located, 178.33 (CO2). 19F NMR
(C6D5Cl): δ −168.24(t), −164.42 (t), −134.00 (d). IR (KBr): νCO
=
1642 cm−1.
(dppf)Ni(CH2CH2) (4). A 50 mL flask was charged with 0.095 g
(0.345 μmol) of Ni(COD)2, 0.180 g (0.325 μmol) of dppf, and
approximately 7 mL of THF. The reaction mixture was stirred for 5
min and attached to a swivel filter frit, and the volatiles were removed
under vacuum. On a vacuum line, approximately 7 mL of Et2O was
transferred into the flask, and 7 equiv of ethylene (1470 Torr in 28.9
mL) was added to the apparatus via a calibrated gas bulb at −196 °C.
The resulting suspension was stirred at ambient temperature for 2 h.
The reaction mixture was filtered through the swivel filter frit at −78
°C to yield 0.146 g (70%) of 4 as a yellow powder. Anal. Calcd for
C36H32FeNiP2: C, 67.44; H, 5.03. Found: C, 66.87; H, 5.26. 1H NMR
(C6D6): δ 2.77 (s, 4H, C2H4), 3.93 (s, 4H, CpH), 4.30 (s, 4H, CpH),
7.03−7.40 (m, 12H, Ph), 7.78−7.81 (m, 8H, Ph). 31P{1H} NMR
(C6D6): 24.9 (s). 13C{1H} NMR (C6D6): δ 46.24 (η2-C2H4), 71.64,
74.87 (Cp), one Cp signal not located, 129.33, 134.61 (aryl), two aryl
signals not located.
(9) Wolfe, J. M.; Bernskoetter, W. H. Dalton Trans. 2012, 41, 10763.
(10) Lejkowski, M. L.; Lindner, R.; Kageyama, T.; Bodizs, G. E.;
Plessow, P. N.; Muller, I. B.; Schafer, A.; Rominger, F.; Hofmann, P.;
Futter, C.; Schunk, S. A.; Limbach, M. Chem.Eur. J. 2012, 18, 14017.
(11) (a) Lee, S. Y. T.; Cokoja, M.; Drees, M.; Li, Y.; Mink, J.;
Herrmann, W. A.; Kuhn, F. E. ChemSusChem 2011, 4, 1275.
(b) Bruckmeier, C.; Lehenmeier, M. W.; Reichardt, R.; Vagin, S.;
Rieger, B. Organometallics 2010, 29, 2199. (c) Fischer, R.; Langer, J.;
Malassa, G.; Walther, D.; Gorls, H.; Vaughan, G. Chem. Commun.
2006, 2510.
ASSOCIATED CONTENT
* Supporting Information
Selected NMR spectral data; crystallographic information file
for complex 2. These data are available free of charge via the
Internet at http://pubs.acs.org.
■
S
AUTHOR INFORMATION
Corresponding Author
*E-mail: wb36@brown.edu.
■
(12) (a) Langer, J.; Walther, D.; Malassa, A.; Westerhausen, M.;
Gorls, H. Eur. J. Inorg. Chem. 2010, 275. (b) Langer, J.; Fischer, R.;
Gorls, H.; Walther, D. Eur. J. Inorg. Chem. 2007, 2257. (c) Langer, J.;
Walther, D.; Gorls, H. J. Organomet. Chem. 2006, 691, 4874.
(d) Langer, J.; Gorls, H.; Fischer, R.; Walther, D. Organometallics
2005, 24, 272. (e) Langer, J.; Gorls, H.; Gillies, G.; Walther, D. Z.
Anorg. Allg. Chem. 2005, 631, 2719. (f) Langer, J.; Fischer, R.; Gorls,
H.; Walther, D. J. Organomet. Chem. 2004, 689, 2952. (g) Walther, D.;
Liesicke, S.; Fischer, R.; Gorls, H.; Weston, J.; Batista, A. Eur. J. Inorg.
Chem. 2003, 4321. (h) Hipler, B.; Doring, M.; Dubs, C.; Gorls, H.;
Hubler, T.; Uhlig, E. Z. Anorg. Allg. Chem. 1998, 624, 1329.
(13) Brookhart, M.; Green, M. L. H.; Parkin, G. Proc. Natl. Acad. Sci.
2007, 104, 6908.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This material is based upon work supported by the National
Science Foundation under the Center for Chemical Innovation
“CO2 as a Sustainable Feedstock for Chemical Commodities”
(CHE-1240020).
REFERENCES
■
(1) (a) Renewable Raw Materials: New Feedstocks for the Chemical
Industry; Ulber, R.; Sell, D.; Hirth, T., Eds.; Wiley-VCH: Weinheim,
2011. (b) Vennestrom, P. N. R.; Osmundsen, C. M.; Christensen, C.
H.; Tarning, E. Angew. Chem., Int. Ed. 2011, 50, 10502. (c) Huang, K.;
Sun, C.-L.; Shi, Z.-J. Chem. Soc. Rev. 2011, 40, 2435. (d) Carbon
Dioxide As a Chemical Feedstock; Aresta, M., Ed.; Wiley-VCH:
(14) Langford, C. H.; Gray, H. B. Ligand Substitution Processes; W.A.
Benjamin, Inc.: New York, 1966.
(15) For characterization data of (dppf)2Ni see ref 12b.
(16) (a) Doherty, M. D.; Grant, B.; White, P. S.; Brookhart, M.
Organometallics 2007, 26, 5950. (b) Yi, C. S.; Liu, N. J. Organomet.
G
dx.doi.org/10.1021/om400025h | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Chem. 1998, 553, 157. (c) Brookhart, M.; Hauptman, E. J. Am. Chem.
Soc. 1992, 114, 4437.
(17) (a) Berkefeld, A.; Drexler, M.; Moller, H. M.; Mecking, S. J. Am.
Chem. Soc. 2009, 131, 12613. (b) Liang, L.-C.; Chien, P.-S.; Lee, P.-Y.
Organometallics 2008, 27, 3082.
(18) [H2 CCHCO2 BArf 3 ]− [DBUH]+ and [H2 C
CHCO2BArf3]−[BTPPH]+ salts were identified by NMR spectroscopy
against independently prepared salts synthesized through the reaction
f
of acrylic acid with base and BAr3 .
(19) Curtis, C. J.; Miedaner, A.; Ciancanelli, R.; Ellis, W. W.; Noll, B.
C.; Rakowski DuBois, M..; DuBois, D. L. Inorg. Chem. 2003, 42, 216.
(20) For a selection of other diphosphine nickel(0) monoethylene
complexes characterized by NMR spectroscopy see: (a) Jones, G. D.;
Vicic, D. A. Organometallics 2005, 24, 3821. (b) Garcia, J. J.; Arevalo,
A.; Brunkan, N. M.; Jones, W. D. Organometallics 2004, 23, 3997.
(c) Bennett, M. A.; Johnson, J. A.; Willis, A. C. Organometallics 1996,
15, 68. (d) Selingson, A. L.; Cowan, R. L.; Trogler, W. C. Inorg. Chem.
1991, 30, 3371. (e) Tolman, C. A.; Seidel, W. C.; Gerlach, D. H. J. Am.
Chem. Soc. 1972, 94, 2669.
(21) Pangborn, A. B.; Giardello, M. A.; Grubbs, R. H.; Rosen, R. K.;
Timmers, F. J. Organometallics 1996, 15, 1518.
(22) Sandstrom, J. Dynamic NMR Spectroscopy; Academic Press: New
̈
York, 1982.
H
dx.doi.org/10.1021/om400025h | Organometallics XXXX, XXX, XXX−XXX