Ancillary Ligand and Base Influences on Nickel-Catalyzed Coupling of CO2 and Ethylene to Acrylate

Article abstract of DOI:10.1021/acs.organomet.9b00708

The coupling of CO₂ and ethylene to produce acrylates has been an area of increasing interest in recent years following a number of studies which have empirically improved catalytic turnover. Notably, the incorporation of moderately Br?nsted and Lewis basic sodium phenoxide salts, as well as zinc dust, and Lewis acidic lithium salts were found to facilitate acrylate formation in batch catalysis. Despite these advances, there has been limited investigation into the effect of the catalyst ancillary ligand and phenoxide base structure on catalytic performance. Here, a collection of 1,2-bis(dialkylphosphino)benzene and related diphosphine ligands were used to show that the influence of steric environs has a marked effect on turnover. Ancillary diphosphine ligands featuring at least two smaller alkyl substituents are needed for strong activity, while the oft-used benzene annulation of the diphosphine does not appear to be determinant in achieving high turnover values. Additionally, the investigation of a collection of substituted sodium phenoxide bases suggests that a subtle balance of basicity and steric factors must be satisfied to obtain optimal catalytic performance. These trends appear to result from competitive, deleterious nucleophilic reactions between base and CO₂ to produce carbonate and the need to maintain sufficient basicity and access to the metal coordination sphere to drive the endergonic CO₂-ethylene coupling reaction.

Full text of DOI:10.1021/acs.organomet.9b00708

pubs.acs.org/Organometallics
Article
Ancillary Ligand and Base Influences on Nickel-Catalyzed Coupling
of CO2 and Ethylene to Acrylate
Katherine B. Uttley, Kenichi Shimmei, and Wesley H. Bernskoetter*
Cite This: https://dx.doi.org/10.1021/acs.organomet.9b00708
Read Online
sı
*
Metrics & More
Article Recommendations
Supporting Information
ACCESS
ABSTRACT: The coupling of CO2 and ethylene to produce acrylates has been an area of
increasing interest in recent years following a number of studies which have empirically
improved catalytic turnover. Notably, the incorporation of moderately Brønsted and Lewis
basic sodium phenoxide salts, as well as zinc dust, and Lewis acidic lithium salts were found
to facilitate acrylate formation in batch catalysis. Despite these advances, there has been
limited investigation into the effect of the catalyst ancillary ligand and phenoxide base
structure on catalytic performance. Here, a collection of 1,2-bis(dialkylphosphino)benzene
and related diphosphine ligands were used to show that the influence of steric environs has a
marked effect on turnover. Ancillary diphosphine ligands featuring at least two smaller alkyl
substituents are needed for strong activity, while the oft-used benzene annulation of the
diphosphine does not appear to be determinant in achieving high turnover values.
Additionally, the investigation of a collection of substituted sodium phenoxide bases
suggests that a subtle balance of basicity and steric factors must be satisfied to obtain
optimal catalytic performance. These trends appear to result from competitive, deleterious
nucleophilic reactions between base and CO2 to produce carbonate and the need to maintain sufficient basicity and access to the
metal coordination sphere to drive the endergonic CO2−ethylene coupling reaction.
The first successful catalytic production of acrylate was
achieved by Limbach and co-workers using a diphosphine
nickel(0) complex in the presence of NaOtBu, resulting in a
turnover number (TON) of 10. However, the catalytic cycle
had to be divided into separate steps to prevent deleterious
reactions between CO2 and the strong base (Figure 1).8 The
INTRODUCTION
■
Anthropogenic emissions of greenhouse gases such as carbon
dioxide (CO2) are recognized to have significant adverse
effects on the environment and human health.1 Furthermore,
the increasing demand for petroleum-based materials as a
primary carbon source is unsustainable, given the finite nature
of these resources.2 The utilization of CO2 as a carbon
feedstock is a promising, sustainable alternative to petroleum
for the production of select commodity chemicals.3 However,
the thermodynamic and kinetic stability of CO2 has limited its
large-scale utilization as a carbon source.3 Notable examples of
industrial utilization involve coupling CO2 with other small
molecules such as amines, epoxides, and dihydrogen to yield
urea, carbonates, and formates, respectively.4 Another potential
method of CO2 utilization is its coupling with light olefins for
the production of acrylates, a valuable group of commodity
chemicals used in adhesives, fabrics, coatings, superabsorbent
polymers, and detergents.5
Investigations into transition-metal-mediated acrylate syn-
thesis from CO2 first found promise in the 1980s following
reports from Hoberg and Schaefer.6 These initial findings were
seminal in identifying a few zerovalent metal complexes
capable of forming a C−C bond between CO2 and ethylene.7
The resulting nickelalactone was first characterized in 1983,
but despite this progress, catalytic activity remained elusive for
several decades. Only recently have low-valent group 10 metal
species provided the first evidence for catalytic acrylate
formation from CO2−ethylene coupling.8
Figure 1. Proposed catalytic cyclic for CO2−ethylene coupling in the
presence of base.
Special Issue: Organometallic Chemistry for Enabling
Carbon Dioxide Utilization
Received: October 23, 2019
https://dx.doi.org/10.1021/acs.organomet.9b00708
© XXXX American Chemical Society
Organometallics XXXX, XXX, XXX−XXX
A

Organometallics
pubs.acs.org/Organometallics
Article
t
Scheme 1. Synthesis of Bu-phosphacycle and BenzP Derivatives
addition of exogenous base was crucial in allowing for the
deprotonation of the Ni−lactone intermediate and driving the
reaction thermodynamics to produce the more favorable
sodium acrylate product (ΔGrxn = −59 kJ mol−1) as opposed
to acrylic acid (ΔGrxn = +79 kJ mol−1).9 Subsequent reports
found that the utilization of weaker, less nucleophilic sodium
phenoxide bases which react reversibly with CO2 allowed for a
one-pot synthesis to yield sodium acrylate with a TON of
approximately 40.10 This innovation provided a considerable
improvement in the operation of the catalytic reaction, and
TONs have been further enhanced via the addition of zinc
dust, which presumably reduces off-cycle Ni(II) complexes and
maintains high concentrations of the catalytically active Ni(0)
species.10 Additionally, advancements in the coupling of CO2
and ethylene were aided by the discovery that Lewis acids can
promote β-hydride elimination from Ni−lactones by destabi-
lizing the nickel−carboxylate interaction. Collectively, all of
these studies have led to the employment of three additives
phenoxide base, zinc dust, and Lewis acidsto achieve
catalytic turnovers of greater than 400.11
Despite these substantial improvements in turnover,
increases of many orders of magnitude are still required to
become competitive with current industrial methods of
acrylate production. A rational approach to this goal in catalyst
development necessitates a better understanding of the
complex interactions of the three additives with the catalyst.
Herein, we report the development of a series of nickel
catalysts for the coupling of CO2 and ethylene to acrylate that
are used to gain insight into the structure−reactivity
relationships between catalyst architecture and phenoxide
substituents.
RESULTS AND DISCUSSION
■
To date, there have been relatively few investigations into the
relationship between the catalyst or base structure and catalytic
efficacy for CO2−ethylene coupling.8,10c The primary focus of
earlier investigations has been the empirical optimization of
reaction conditions and additives.8−11 One of the few
systematic structure−reactivity studies examined the role of
the hydrocarbon bridge on ancillary diphosphine ligands,
which suggested that C2 spacing groups are preferred for
CO2−ethylene coupling. In particular, rigid, nonrotating C2
bridges, such as ortho-substituted arenes, proved optimal for
acrylate formation.10 The most productive catalyst systems to
date are derived from the (R,R)-(+)-1,2-bis(tert-
butylmethylphosphino)benzene (BenzP) ligand, which bears
this feature. Additionally, the limited available catalytic data
indicate that CO2−ethylene coupling is best promoted by
ligands which bear electron-donating alkyl groups, in lieu of
more commonly available aryl phosphino groups. Given this
limited prior art, our laboratory targeted a family of alkyl-
substituted derivatives of the BenzP ligand in order to elucidate
the steric preferences for Ni(0)-catalyzed acrylate formation.
Racemic BenzP ligand and its derivatives were prepared by
the addition of methyl triflate to 7,8-diphosphabicyclo[4.2.0]-
octa-1,3,5-triene, 7,8-bis(1,1-dimethylethyl) (tBu-phospha-
cycle).14 The resulting triflate salt was treated with alkylated
Grignard reagent to yield the BenzP derivatives (Scheme 1a).
A single isomer was observed in each case, and the relative
configuration of the two tBu−P substituents is proposed to be
anti in analogy to several closely related species prepared via
this method.12 Attempts were made to further expand the
diversity of racemic BenzP ligands by altering the substituents
of the phosphacycle precursor. However, these attempts failed
B
https://dx.doi.org/10.1021/acs.organomet.9b00708
Organometallics XXXX, XXX, XXX−XXX

Organometallics
pubs.acs.org/Organometallics
Article
in our hands, as we were unable to isolate any variants of the
tBu-phosphacycle. Despite this limitation, new racemic BenzP
ligands were successfully prepared featuring isopropyl (1-(tert-
butylmethylphosphino)-2-(tert-butylisopropylphosphino)-
benzene (iPrBenzP)) and cyclohexyl (1-(tert-butylmethylphos-
phino)-2-(tert-butylcyclohexylphosphino)benzene (Cy-
BenzP)) phosphine substituents (Scheme 1a).
these conditions. Examination of the data from these
comparative catalytic experiments reveals some salient trends
between ancillary ligand sterics and TON. Alteration of the
top-performing BenzP by replacing one P−Me substituent
with a P−Cy substituent in (CyBenzP)Ni(COD) (entries 1
and 2) yielded a small but distinguishable drop in productivity.
Similar activity was observed by replacing the P−Me
substituent with a P−iPr group, where (iPrBenzP)Ni(COD)
afforded a TON of 60 (entry 3).
Zerovalent Ni complexes featuring the new BenzP
derivatives were prepared by dropwise addition of a THF
solution of iPrBenzP or CyBenzP to a THF solution of bis(1,5-
cyclooctadiene)nickel(0) (Ni(COD)2), affording golden col-
ored samples of (iPrBenzP)Ni(COD) and (CyBenzP)Ni-
(COD) for use in catalytic experiments (Scheme 1b). Both
complexes exhibited two doublet signals in the 31P NMR
spectrum, as expected for the inequivalent substituents on the
P atoms. The effectiveness of (iPrBenzP)Ni(COD) and
(CyBenzP)Ni(COD) in CO2−ethylene coupling was first
benchmarked by comparison with (dicyclohexylphosphino)-
ethane (dcpe)- and (dicyclohexylphosphino)propane (dcpp)-
ligated catalyst variants. Both dcpe and dcpp ligands have been
previously reported in Ni-catalyzed acrylate formation, though
interestingly under our conditions the C3-bridged dcpp proved
to be slightly more active.10c Initial reaction conditions for
CO2−ethylene coupling experiments were derived from
previously established protocols which include the presence
of reducing zinc dust, Lewis acid lithium iodide, and sodium 3-
fluorophenoxide (3-FPhONa) base.11 These conditions and
the resulting TONs for acrylate formation using a family of
BenzP derivatives as well as dcpe and dcpp are described in
Table 1.
The poor performance of the (iPr4BenzP) ligand (Table 1,
entry 5) also offers important insight into the structure−
reactivity relationship. The net sterics of having four P−iPr
groups on the ligand may be considered similar to having a
combination of two P−tBu and two P−Me groups (i.e. BenzP),
but the TONs are quite different. One possible explanation is
the influence of exhibiting at least one (but preferably two)
sterically accessible quadrants created by a P−Me substituent
as seen in entries 1−3. These findings hint at a sterically
sensitive relationship that underlies BenzP-ligated Ni com-
plexes. Prior studies have also suggested that the rotationally
constrained sp2-linked C2-bridged diphosphines of the BenzP
derivatives may be significant to their strong catalytic
performance. However, a comparison of entries 4−6 shows
any effect from sp2- versus sp3-linked diphosphines is likely
secondary to steric effects from the phosphine substiutents.8,10
Overall, the data in hand indicate that preparation of sterically
smaller variants of BenzP or diphosphino(ethane) complexes
of nickel which maintain two sterically accessible quadrants
would be advantageous for acrylate production. Efforts to
prepare such species are underway in our laboratory (vide
supra).
Multiple prior studies of CO2−ethylene coupling have
demonstrated a profound effect from varying the identity of
the external base used to drive the catalytic reaction.10c,11
Limbach et al. demonstrated that switching from NaOtBu to
weaker, less nucleophilic sodium phenoxide bases was pivotal
in developing batch catalytic processes and improving
TONs.10e The lower propensity of many sodium phenoxide
bases to react with CO2 to produce carbonates was suggested
as the origin of the catalytic enhancement. The limited stability
of the aryl carbonate species thus prevents complete
deactivation of the required base, as was observed with
NaOtBu.
Despite indications that phenoxide bases are especially
successful in mediating acrylate formation, systematic inves-
tigations into the steric and electronic influence of these bases
in catalytic acrylate production have been limited.10,11 To
further probe relationships between base identity and acrylate
production, a collection of phenoxide bases was employed in
comparative catalytic experiments (Table 2). The sterically
hindered, strong bases sodium hexamethyldisilazide
(NaHMDS) and NaOtBu, which were previously investigated
in acrylate formation, were also included to benchmark the
results for the phenoxide bases.8 The pressure conditions used
in Table 2 were reoptimized from the diphosphine Ni(0)
comparison experiments (Table 1) and modified in order to
minimize the mass of base required for each trial. The
(BenzP)Ni(COD) catalyst was selected for the base
comparison study, owing to its strong performance in Table
1 and its relative ease of synthesis. The change in conditions
resulted in a modest increase in TON from 82 to 105 when 3-
FPhONa was used (Table 1, entry 1, and Table 2, entry 4). As
expected, the strong bases NaHMDS and NaOtBu (entries 7
Table 1. Comparison of Diphosphine Ni(0) Catalysts for
Acrylate Formation
a
b
entry
catalyst
TON
1
2
3
4
5
6
(BenzP)Ni(COD)
(CyBenzP)Ni(COD)
(iPrBenzP)Ni(COD)
(dcpp)Ni(COD)
(iPr4BenzP)Ni(COD)
(dcpe)Ni(COD)
82 7.4
65 7.2
60 12.5
22 5.1
12 2.8
6
3.3
a
Conditions: 0.1 mmol of catalyst, 40 mmol of 3-FPhONa, 10 mmol
of zinc, 5 mmol of LiI, 10 bar of C2H4, 15 bar of CO2, 25 mL of THF,
b
110 °C, 20 h. The TON was determined by 1H NMR analysis in an
average of two to four trials.
In accord with prior findings, the (BenzP)Ni(COD)
complex was considerably more effective than either (dcpp)-
Ni(COD) or (dcpe)Ni(COD) under these reaction con-
ditions, producing acrylate with an average of TON of 82 in
comparison to 22 and 6, respectively (Table 1, entries 1, 4, and
6). Likewise, (CyBenzP)Ni(COD) and (iPrBenzP)Ni(COD)
was more effective than the dcpe- and dcpp-ligated species,
affording average TONs of 65 and 60 (entries 2 and 3).
Additionally, a tetraisopropyl-substituted ligand variant, 1,2-
bis(diisopropylphosphino)benzene (iPr4BenzP), recently re-
ported by our laboratory yielded a TON of 12 (entry 5) under
C
https://dx.doi.org/10.1021/acs.organomet.9b00708
Organometallics XXXX, XXX, XXX−XXX

Organometallics
pubs.acs.org/Organometallics
Article
be seen in comparing the TON of 32 (entry 6) for sodium 2,6-
dichlorophenoxide (2,6-Cl2PhONa) to the TON of 111 (entry
3) for sodium 2-chloro-6-methylphenoxide (2-Cl-6-MePhO-
Na). These two phenoxides have relatively similar sterics but
show a considerable enhancement for the more basic 2-Cl,6-
MePhONa. However, this electronic effect is attenuated by
steric influences on comparing the intermediate TON of 88
obtained from the dramatically more basic sodium 2,6-di-tert-
butyl-4-methylphenoxide (2,6-(tBu)2-4-MePhONa) (entry 5).
We hypothesize that the immense steric protection of the basic
O site inhibits the ability of 2,6-(tBu)2-4-MePhONa to access
the nickel complex. A slight reduction of the steric
encumbrance in sodium 2,6-diisopropylphenoxide (2,6-
(iPr)2PhONa) (entry 1) provides a marked improvement
with a TON of 174. This base appears to have met a preferred
balance of being basic enough to deprotonate the Ni−lactone
or related intermediate (Figure 1), while providing enough
steric protection to limit nucleophilic attack on CO2, all
without obviating access to the nickel complex. Likewise, the
high performance of the 2-ClPhONa base (entry 2) could
originate from a subtle balance between limiting deactivation
via carbonate formation and providing sufficient basicity.
The role of deleterious carbonate formation between
phenoxide base and CO2 was further probed by monitoring
the catalytic reaction over several time points. This was
particularly illustrative in the 3-FPhONa-mediated reaction,
where 19F NMR could be used to confirm the presence of
various phenoxide-derived species and the corresponding
carbonate is stable at ambient temperature.10d The results of
this series of catalytic trials are illustrated in Figure 2. The
production of acrylate appears to be swift over approximately
the first 10 h of the reaction but decreases noticeably after that
point. This decrease in catalytic activity over time appears to
coincide with the accumulation of carbonate, which would
deplete the concentration of base required to complete the
reaction. However, it is also possible that the loss in catalytic
activity is due to catalyst deactivation that simply allows for
Table 2. Comparison of Exogenous Bases for Ni-Catalyzed
Acrylate Formation
a
b
c
entry
base
pKa
TON
1
2
3
4
5
6
7
8
2,6-(iPr)2PhONa
2-ClPhONa
2-Cl-6-MePhONa
3-FPhONa
2,6-(tBu)2-4-MePhONa
2,6-(Cl)2PhONa
NaOtBu
11.10
8.56
9.71
9.29
12.20
6.79
19.20
26
174 9.2
158 9.5
111 2.1
105 9.9
88 8.5
32 9.9
d
7
d
NaHMDS
≤1
a
Conditions: 0.05 mmol of (BenzP)Ni(COD), 20 mmol of base, 5
mmol of zinc, 2.5 mmol of LiI, 20 bar of C2H4, 10 bar of CO2, 25 mL
b
c
of THF, 110 °C, 20 h. pKa listed for the conjugate acid.13 The TON
was determined by 1H NMR analysis in an average of two to six trials.
Data for a single trial.
d
and 8) provided only marginal performance in comparison to
3-FPhONa (entry 4), likely due to a deleterious side reaction
between these bases and CO2 (vide infra).
Comparing the monosubstituted phenoxides 3-FPhONa and
sodium 2-chlorophenoxide (2-ClPhONa) (Table 2, entries 2
and 4) revealed an improved TON of 158 for the Cl-
substituted variant. Further expanding the scope to disub-
stituted phenoxides (entries 1, 3, 5, and 6) produced a widely
varied set of TONs. Table 2 ultimately reveals that there is not
a direct relationship between the strength of the phenoxide
bases (see the pKa of conjugate acids; Table 2) and catalytic
turnover. Indeed, there appear to be counterbalancing effects
with respect to the characteristics of basicity and steric
hindrance. A limited tolerance for more basic phenoxides can
Figure 2. Plot of TON for acrylate production and total moles of 3-fluorophenol and 3-fluorophenylcarbonate vs time during a catalytic trial.
D
https://dx.doi.org/10.1021/acs.organomet.9b00708
Organometallics XXXX, XXX, XXX−XXX

Organometallics
pubs.acs.org/Organometallics
Article
carbonate formation to outcompete acrylate production. To
distinguish between these two hypotheses for activity loss, a
sequential base addition study was performed using the 2-
ClPhONa phenoxide. A reactor was charged with 400 equiv of
phenoxide base and allowed to react under the conditions
described in Table 2 for 20 h.
The reaction was then paused, and an additional 200 equiv
of base was added before the reaction resumed for an
additional 20 h. This resulted in a total TON of 245 for
acrylate formation. The enhanced turnover (in comparison to
158; Table 2; entry 2) upon addition of a second fraction of
base suggests that the nickel catalyst was still active after 20 h.
Thus, the loss of acrylate production observed in Figure 2
likely results from base depletion. This is further supported by
comparison to a catalytic trial in which all 600 equiv of 2-
ClPhONa was added at the outset to afford only 163
turnovers, a conversion indistinguishable from that observed
in Table 2.
The catalytic influence of steric and electronic effects from
both the nickel catalysts and the phenoxide bases is evident
from the results in Tables 1 and 2. However, there appears to
be a balancing of these effects, which suggests that conclusions
regarding the identity of the most effective phenoxide base for
acrylate production may not be universally applied. For
example, it is likely that certain “pairings” between catalyst
ancillary ligand structure and base could reorder the trends
observed in Table 2. To test for this possibility, a small set of
comparative catalysis trials were conducted using the sterically
and electronically dissimilar bases 2,6-(tBu)2PhONa and 3-
FPhNa, along with (CyBenzP)Ni(COD) and (BenzP)Ni-
(COD) (Table 3). When 2,6-(tBu)2PhONa was used instead
CONCLUDING REMARKS
■
The series of diphosphine nickel catalysts prepared and
investigated here illustrate the subtle and competing influences
of both electronic and steric effects in CO2−ethylene coupling
to acrylate. These influences are important in both the ancillary
phosphine ligand substituent as well as the substituent of the
phenoxide bases typically used to drive the thermodynamics of
acrylate production. The interplay of several factors
complicates establishment of the broad structure−reactivity
relationships required to rationally guide catalyst design.
However, several features of a ligand and base selection can
be gleaned from the data in hand. As diphosphine-supported
nickel complexes are among the most studied promotors of
CO2-derived acrylate formation, it is useful to consider the four
quadrants of the metal coordination sphere created by the four
phosphorus substituents. On the basis of the trends observed
here it appears that optimal catalysts maintain a relatively
unhindered approach in at least two of the quadrants.
Additionally, the activating influence of utilizing a benzene
annulated diphosphine structure appears to be secondary to
steric effects in comparison to other bulky diphosphine ligands
with more flexible two-carbon bridges. Finally, the effect of the
phenoxide base substituents should not be undervalued in
optimizing reactivity. Optimization experiments must compare
varied base electronic and steric effects for each new catalyst
structure. Clearly, further efforts to expand the basis set for
drawing structure−reactivity relationships are required to
strengthen the effect of rational design for this class of
catalysts, and such investigations are currently ongoing in our
laboratory.
EXPERIMENTAL SECTION
■
Table 3. Comparison of Stereoelectronic Influences
between Base and Catalyst
General Considerations. All manipulations were carried out
using standard glovebox, Schlenk, cannula, or vacuum techniques.
Ethylene (dry) and carbon dioxide (laser-grade) were purchased from
Airgas and used as received. All other chemicals were purchased from
Fischer, Aldrich, Strem, VWR, or Cambridge Isotope Laboratories.
Solvents were dried and deoxygenated according to literature
procedures.14 The ligand precursor 7,8-diphosphabicyclo[4.2.0]octa-
1,3,5-triene, 7,8-bis(1,1-dimethylethyl) (tBu-phosphacycle) was pre-
pared as previously described.12 BenzP was prepared as previously
described and obtained as a racemic mixture.12 Nonvolatile solids
were dried under vacuum at 50 °C overnight. 1H, 13C, and 31P NMR
spectra were recorded on Bruker 300 MHz DRX and 500 MHz DRX
a
b
entry
catalyst
base
TON
1
2
3
4
(BenzP)Ni(COD)
(CyBenzP)Ni(COD)
(BenzP)Ni(COD)
(CyBenzP)Ni(COD)
3-FPhONa
3-FPhONa
2,6-(tBu)2PhONa
2,6-(tBu)2PhONa
82 7.37
70 0.71
33 3.79
49 4.24
spectrometers at ambient temperatures. H and 13C chemical shifts
1
were referenced to residual solvent signals; 31P chemical shifts were
referenced to an external standard of H3PO4. Probe temperatures
were calibrated using ethylene glycol and methanol as previously
described.15 High-pressure catalytic hydrogenation reactions were
performed using a Parr 5500 Series compact reactor with a 100 mL
glass insert.
a
Conditions: 0.1 mmol of catalyst, 40 mmol of base, 10 mmol of zinc,
5 mmol of LiI, 10 bar of C2H4, 15 bar of CO2, 25 mL of THF, 110 °C,
b
20 h. The TON was determined by 1H NMR analysis in an average
of two to four trials.
Preparation of 1,2-Bis(dialkylphosphino)benzene Com-
plexes. A 20 mL scintillation vial was filled with 757 mg (3.00
mmol) 7,8-diphosphabicyclo[4.2.0]octa-1,3,5-triene, 7,8-bis(1,1-di-
methylethyl) (tBu-phosphacycle), the solids were dissolved in 20
mL of dichloromethane, and the mixture was transferred to a 50 mL
sealed glass reaction vessel. The yellow solution was cooled to −196
°C, and 3.30 mmol of methyl triflate was added via vapor-phase gas
addition. The resulting light yellow solution was stirred at ambient
temperature for 4 h. The solvent was removed under vacuum, leaving
behind an oily, yellow residue. A 20−30 mL volume of Et2O was
added to the residue and cooled to −35 °C, at which point 3.15 mmol
of the appropriate alkyl Grignard reagent was added. The resulting
dark yellow solution was stirred at ambient temperature overnight.
The solution was filtered through Celite in a Pasteur pipet, and the
solvent was removed under vacuum. The resulting yellow solid was
of 3-FPhONa, the productivity of both catalysts decreased due
to the greater steric limitation of the base. Interestingly though,
the modest favorability for using (BenzP)Ni(COD) over
(CyBenzP)Ni(COD) in the presence of 3-FPhONa is inverted
with 2,6-(tBu)2PhONa. The precise origins for this inversion in
relative activity are not self-evident with the data in hand and
may involve participation from the Zn or LiI additives required
to achieve high conversion. Still, these observations serve as a
caution regarding comparisons of relative catalytic activities
when only a single base is used to benchmark the reaction.
E
https://dx.doi.org/10.1021/acs.organomet.9b00708
Organometallics XXXX, XXX, XXX−XXX

Organometallics
pubs.acs.org/Organometallics
Article
dissolved in pentane and filtered through Celite in a Pasteur pipet.
The pentane was removed under vacuum, leaving a thick oil that was
pale yellow to colorless and was used without further purification.
CH2), 37.46 (COD-CH2), 78.32 (COD-CH), 79.51 (COD-CH),
85.05 (COD-CH), 85.95 (COD-CH), 129.94 (Benz-CH), 130.05
(Benz-CH), 130.50 (Benz-CH), 130.60 (Benz-CH), 147.18 (Benz-
C), 148.30 (Benz-C).
i
Isolated yields of ligands: CyBenzP, 68.0%; PrBenzP, 74.5%.
Spectral Data for CyBenzP. 31P{1H} NMR (300 MHz, C6D6): δ
Preparation of Sodium Phenoxide Bases. A 20 mL
scintillation vial was filled with 3.60 g (150 mmol, 1 mol equiv) of
NaH and then quantitatively transferred to a 250 mL Schlenk flask via
a THF slurry (40 mL). In a separate vial, 158 mmol (1.05 mol equiv)
of freshly distilled liquid phenol was weighed out and then taken up
into a 20 mL syringe. In the case of solid 2,6-tBu2PhOH, the phenol
was first dissolved in 50 mL of THF prior to being taken up in the
syringe. The sodium hydride solution was cooled to 0 °C and the
phenol added dropwise under a positive flow of argon. The solution
was stirred overnight at ambient temperature, and then the solvent
was removed under vacuum and the resulting oily solid triturated with
20−40 mL off Et2O to remove residual THF. The solid product was
dissolved in Et2O and filtered through Celite. The Et2O was removed
under vacuum, and the solid was washed with cold pentane on a frit.
The remaining bases were then recrystallized in Et2O (with the
exception of 2,6-iPr2PhONa, which was recrystallized in THF layered
with pentane).
1
−24.66 (d, J = 152.6 Hz), 1.19 (d, J = 152.6 Hz). H NMR (500
MHz, C6D6): δ 1.10 (m, 18 H, PtBu-CH3), 1.13 (s, 3 H, P-CH3), 1.48
(m, 10 H, Cy-CH2), 2.17 (m, 1H, PCy-CH), 7.13 (m, 2 H, Benz-
CH), 7.43 (m, 1 H, Benz-CH), 7.48 (m, 1 H, Benz-CH). 13C{1H}
NMR (500 MHz, C6D6): δ 6.19 (P-CH3), 26.77 (PCy-CH2), 27.36
(PCy-CH2), 27.72 (PtBu-CH3), 29.45 (PtBu-CH3), 30.72 (PtBu-C),
30.99 (PCy-CH2), 31.17 (PtBu-C), 31.43 (PCy-CH2), 31.68 (PCy-
CH2), 34.23 (PCy-CH), 132.02 (Benz-CH), 132.40 (Benz-CH),
145.64 (Benz-C), 146.95 (Benz-C).
i
Spectral Data for PrBenzP. 31P{1H} NMR (300 MHz, C6D6): δ
1
−25.44 (d, J = 151.2 Hz), 7.66 (d, J = 150.8 Hz). H NMR (300
MHz, C6D6): δ 0.96 (dd, J = 6.9, 13.1 Hz, 3 H, PiPr-CH3), 1.11 (m,
18 H, PtBu-CH3), 1.17 (m, 3 H, P-CH3), 1.27 (dd, J = 6.7, 13.9 Hz, 3
H, PiPr-CH3), 2,25 (m, 1 H, PiPr-CH), 7.12 (m, 2 H, Benz-CH), 7.39
(m, 2 H, Benz-CH). 13C{1H} NMR (500 MHz, C6D6) δ 6.11 (P-
CH3), 21.85 (PiPr-CH3), 22.02 (PiPr-CH3), 22.78 (PiPr-CH), 27.67
(PtBu-CH3), 29.17 (PtBu-CH3), 31.00 (PtBu-C), 132.23 (Benz-CH),
146.08 (Benz-C), 146.70 (Benz-C).
General Procedure for Catalytic Acrylate Production
Experiments. A stainless-steel autoclave reactor fitted with a glass
insert was loaded with the appropriate amount of catalyst (0.1 or 0.05
mmol; 1 mol equiv), sodium phenoxide base (400 mol equiv), zinc
dust (100 mol equiv), lithium iodide (50 mol equiv), and 25 mL of
THF. The reactor was sealed under an inert atmosphere and removed
from the glovebox. The contents of the reactor were stirred while the
reactor was pressurized with ethylene followed by carbon dioxide. The
reactor was heated to 110 °C for 20 h. Following the reaction period,
the vessel was removed from the heating element, cooled in an ice−
water bath for 30 min, and slowly vented to ambient pressure. The
reaction residue was extracted with D2O, and an internal standard of
sodium sorbate was added. The organic-soluble species were removed
by washing with Et2O. Production of the acrylate salt was quantified
Preparation of [1,2-bis(dialkylphosphino)benzene]Ni(COD)
Complexes. A 20 mL scintillation vial was filled with 196 mg (0.71
mmol) of Ni(COD)2, the solid was dissolved in 10 mL of THF, and
the mixture was stirred for 10 min in order to fully dissolve the solid.
In a separate vial, 0.71 mmol of 1,2-bis(dialkylphosphino)benzene
ligand was dissolved in 5 mL of THF and added dropwise to the
stirred nickel solution. The resulting dark brown solution was
transferred to a 50 mL sealed glass reaction vessel and stirred at 50 °C
overnight. The THF solvent was then removed under vacuum, and
the resulting dark brown oil was dissolved in pentane and filtered
through Celite in a Pasteur pipet. The filtrate was dried under
vacuum, and the resulting dark golden brown solid was recrystallized
in diethyl ether at −35 °C to afford the catalyst complex as a dark
golden brown solid. Isolated yields of catalysts: (CyBenzP)Ni(COD),
85%; (iPrBenzP)Ni(COD), 45%.
1
by integration of the H NMR spectrum.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/acs.organomet.9b00708.
■
Spectral Data for (CyBenzP)Ni(COD). 31P{1H} NMR (300 MHz,
sı
1
C6D6): δ 50.97 (d, J = 25.4 Hz), 71.81 (d, J = 25.4 Hz). H NMR
(500 MHz, C6D6): δ 1.01 (m, 18 H, PtBu- CH3), 1.24 (m, 7 H, PCy-
CH2), 1.42 (d, J = 4.5 Hz, 3 H, P-CH3), 1.53 (d, J = 11.0 Hz, 1 H,
PCy-CH2), 1.61 (d, J = 9.5 Hz, 1 H, PCy-CH2), 1.85 (d, J = 13.0 Hz,
1 H, PCy-CH2), 2.01 (apparent qt, J = 4.0, 12.8 Hz, 1 H, COD-CH2),
2.16 (m, 1 H, COD-CH2), 2.39 (m, 2 H, COD-CH2), 2.56 (apparent
qt, J = 3.8, 11.5 Hz, 1 H, PCy-CH), 2.81 (m, 2 H, COD-CH2), 2.90
(td, J = 8.0, 4.0 Hz, 1 H, COD-CH2), 3.13 (m, 1 H, COD-CH2), 4.29
(m, 1 H, COD-CH), 4.52 (m, 2 H, COD-CH), 4.93 (apparent
quintet, 1 H, COD-CH), 7.09 (m, 2 H, Benz-CH), 7.52 (m, 2 H,
Benz-CH). 13C{1H} NMR (500 MHz, C6D6): δ 8.78 (P-CH3), 26.69
(PCy-CH2), 27.66 (PtBu-CH3), 27. 52 (PCy-CH2), 27.85 (PCy-
CH2), 28.07 (COD-CH2), 29.03 (PtBu-CH3), 29.31 (COD-CH2),
30.06 (PCy-CH2), 33.24 (PtBu-C), 36.86 (COD-CH2), 37.59 (COD-
CH2), 38.24 (PCy-CH), 78.98 (COD-CH), 79.73 (COD-CH), 85.20
(COD-CH), 86.15 (COD-CH), 127.37 (Benz-CH), 127.84 (Benz-
CH), 129.78 (Benz-CH), 130.66 (Benz-CH), 147.43 (Benz-C),
148.24 (Benz-C).
Selected NMR spectra, complete catalytic trial data, and
additional experiment descriptions (PDF)
AUTHOR INFORMATION
Corresponding Author
■
Wesley H. Bernskoetter − The University of Missouri,
Columbia, Missouri; orcid.org/0000-0003-0738-
5946; Email: bernskoetterwh@missouri.edu
Other Authors
Katherine B. Uttley − The University of Missouri,
Columbia, Missouri
Kenichi Shimmei − Sekisui Chemical Co. Ltd., Osaka,
Japan
Spectral Data for (iPrBenzP)Ni(COD). 31P{1H} NMR (300 MHz,
1
C6D6): δ 50.87 (d, J = 25.3 Hz), 77.66 (d, J = 25.3 Hz). H NMR
(500 MHz, C6D6): δ 0.83 (dd, J = 7.1, 14.2 Hz, 3 H, PiPr-CH3), 1.00
(m, 18 H, PtBu-CH3), 1.41 (d, J = 4.8 Hz, 3 H, P-CH3), 1.48 (dd, J =
7.0, 13.0 Hz, 3 H, PiPr-CH3), 2.14 (m, 2 H, COD-CH2), 2.39 (m, 2
H, COD-CH2), 2.67 (apparent ddt, J = 7.0, 11.5, 14.0 Hz, 1 H, PiPr-
CH), 2.79 (m, 2H, COD-CH2), 2.86 (apparent td, J = 8.0, 15.5 Hz, 1
H, COD-CH2), 3.09 (m, 1 H, COD-CH2), 4.28 (m, 1 H, COD-CH),
4.50 (m, 2 H, COD-CH), 4.83 (apparent quintet, 1 H, COD-CH),
7.08 (m, 2H, Benz-CH), 7.42 (m, 1H, Benz-CH), 7.50 (m, 1H, Benz-
CH). 13C{1H} NMR (500 MHz, C6D6): δ 8.76 (P-CH3), 20.10 (PiPr-
CH3), 21.53 (PiPr-CH3), 26.58 (PtBu-CH3), 27.62 (COD-CH2),
28.06 (PiPr-CH), 28.73 (COD-CH2), 33.24 (PtBu-C), 36.82 (COD-
Complete contact information is available at:
https://pubs.acs.org/10.1021/acs.organomet.9b00708
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The Curators of the University of Missouri are acknowledged
for financial support of this work. K.S. was supported by the
Sekisui Chemical Co., Ltd.
F
https://dx.doi.org/10.1021/acs.organomet.9b00708
Organometallics XXXX, XXX, XXX−XXX

Organometallics
pubs.acs.org/Organometallics
Article
P. G.; Hazari, N.; Bernskoetter, W. H. Effect of Sodium Cation on
Metallacycle β-hydride Elimination in CO2-Ethylene Coupling to
Acrylates. Chem. - Eur. J. 2014, 20, 3205−3211. (c) Hendriksen, C.;
REFERENCES
■
(1) IPCC, 2018: Global Warming of 1.5 °C − An IPCC special
report on the impacts of global warming of 1.5 °C above pre-
industrial levels and related global greenhouse gas emission pathways,
in the context of strengthening the global response to the threat of
climate change, sustainable development, and efforts to eradicate
poverty.
(2) (a) Organization of the Petroleum Exporting Countries. 2017
OPEC World Oil Outlook, October 2017. (b) Miller, R. G.; Sorrell, S.
R. The future of oil supply. Philos. Trans. R. Soc., A 2014, 372,
20130179.
(3) (a) Barbarossa, V.; Giuseppina, V.; Viscardi, R.; Gattia, D. M.
CO2 as a Carbon Source for Fuel Synthesis. Energy Procedia 2014, 45,
1325−1329. (b) Rahman, F. A.; Aziz, M. M. A.; Saidur, R.; Bakar, W.
A. W. A.; Hainin, M. R.; Putrajaya, R.; Hassan, N. A. Pollution to
Solution: Capture and Sequestration of Carbon Dioxide (CO2) and
its Utilization as a Renewable Energy Source for a Sustainable Future.
Renewable Sustainable Energy Rev. 2017, 71, 112−126.
(4) (a) Sakakura, T.; Choi, J. C.; Yasuda, H. Transformation of
Carbon Dioxide. Chem. Rev. 2007, 107, 2365−2387. (b) Aresta, M.;
Dibenedetto, A. In Industrial Utilization of Carbon Dioxide; Develop-
ments and Innovation in Carbon Dioxide Capture and Storage
Technology; Maroto-Valer, M. M., Ed.; Woodhead: Cambridge,
2010; pp 377−410.
̈
Pidko, E. A.; Yang, G.; Schaffner, B.; Vogt, D. Catalytic Formation of
Acrylate from Carbon Dioxide and Ethene. Chem. - Eur. J. 2014, 20,
12037−12040. (d) Manzini, S.; Huguet, N.; Trapp, O.; Paciello, R.
A.; Schaub, T. Synthesis of acrylates from olefins and CO2 using
sodium alkoxides as bases. Catal. Today 2017, 281, 379−386.
(e) Huguet, N.; Jevtovikij, I.; Gordillo, A.; Lejkowski, M. L.; Lindner,
R.; Bru, M.; Khalimon, A. Y.; Rominger, F.; Schunk, S. A.; Hofmann,
P.; Limbach, M. Nickel-Catalyzed Direct Carboxylation of Olefins
with CO2: One-Pot Synthesis of α,β-Unsaturated Carboxylic Acid
Salts. Chem. Eur. J. 2014, 20, 16858−16862.
(11) Hopkins, M. N.; Shimmei, K.; Uttley, K. B.; Bernskoetter, W.
H. Synthesis and Reactivity of 1,2-Bis(di-iso-propylphosphino)-
benzene Nickel Complexes: A Study of Catalytic CO2 − Ethylene
Coupling. Organometallics 2018, 37, 3573−3580.
(12) Reynolds, S. C.; Hughes, R. P.; Glueck, D. S.; Rheingold, A. L.
Synthesis, Reactivity, and Resolution of a C2 − Symmetric, P
Stereogenic Benzodiphosphetane, a Building Block for Chiral
Bis(phosphines). Org. Lett. 2012, 14, 4238−4241.
(13) Serjeant, E. P., Dempsey, B. Ionization Constants of Organic
Acids in Aqueous Solution. International Union of Pure and Applied
Chemistry (IUPAC); Pergamon Press: New York, NY, 1979; IUPAC
Chemical Data Series No. 23.
(5) (a) Yu, B.; Diao, Z. F.; Guo, C. X.; He, L. N. Carboxylation of
Olefins/Alkynes with CO2 to Industrially Relevant Acrylic Acid
Derivatives. Journal of CO2 Utilization 2013, 1, 60−68. (b) Wang, X.;
Wang, H.; Sun, Y. Synthesis of Acrylic Acid Derivatives from CO2 and
Ethylene. Chem. 2017, 3, 211−228.
(14) Pangborn, A. B.; Giardello, M. A.; Grubbs, R. H.; Rosen, R. K.;
Timmers, F. J. Safe and Convenient Procedure for Solvent
Purification. Organometallics 1996, 15, 1518−1520.
̈
(15) Sandstrom, J. Dynamic NMR Spectroscopy; Academic Press:
London, 1982.
(6) (a) Hoberg, H.; Schaefer, D. Nickel(0)-induzierte C-C-
verknupfung zwischen alkenen und kohlendioxid. J. Organomet.
̈
Chem. 1982, 236, C28−C30. (b) Walther, D.; Dinjus, E. Aktivierun
̈
von Kohlendioxid an Ubergangsmetallzentren; Die matallarings-
chlußreaktion zwischen Komplexrumpf. Z. Chem. 1982, 22, 228−
229. (c) Walther, D.; Dinjus, E.; Sieler, J.; Thanh, N. N.; Schade, W.;
̈
et al. Aktivierung von CO2 an Ubergangsmetallzentren: Struktur und
̈
Reaktivitat eines C-C Kopplungsproduktes von CO2 und 2,3-
Dimethylbutadien am elektronenreichen Nickel(0). Z. Naturforsch.,
B: J. Chem. Sci. 1983, 38, 835−840. (d) Hoberg, H.; Schaefer, D.;
Burkhart, G.; Kruger, C.; Romao, M. Nickel(0)-induzierte C-C-
verknupfung kohlendioxid und alkinen sowie alkenen. J. Organomet.
̈
Chem. 1984, 266, 203−224. (e) Walther, D.; Dinjus, E.; Sieler, J.;
Andersen, L.; Lindqvist, O. Aktivierung von Kohlendioxid an
̈
Ubergangsmetallzentren: Metallaringschluss mit Dicyclopentadien
am elektronenreichen Nickel(0) − komplexrumpf als topo − und
stereoselektive Reaktion. J. Organomet. Chem. 1984, 276, 99−107.
(f) Hoberg, H.; Peres, Y.; Milchereit, A. C-C-Verknupfung von
̈
̈
Alkenen mit CO2 an Nickel(0); Herstellung von Zimtsaure aus Styrol.
J. Organomet. Chem. 1986, 307, C38−C40.
(7) (a) Hoberg, H.; Schaefer, D. Nickel(0)-induzierte CC-
Verknu
oder di-substitutierten Alkenen. J. Organomet. Chem. 1983, 251,
ger, C.; Tsay, Y. H. A 1-oxa-
̈
pfung zwischen Kohlendioxid und Ethylen sowie mono-
C51−C53. (b) Hoberg, H.; Peres, Y.; Kru
̈
2-nickela-5-cyclopentanone from Ethane and Carbon Dioxide:
Preparation, Structure, and Reactivity. Angew. Chem., Int. Ed. Engl.
1987, 26, 771−773.
́
(8) 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. The First Catalytic Synthesis of
an Acrylate from CO2 and an Alkene − A Rational Approach. Chem. -
Eur. J. 2012, 18, 14017−14025.
(9) Graham, D. C.; Mitchell, C.; Bruce, M. I.; Metha, G. F.; Bowie, J.
H.; Buntine, M. A. Production of Acrylic Acid through Nickel-
Mediated Coupling of Ethylene and Carbon Dioxide − A DFT Study.
Organometallics 2007, 26, 6784−6792.
(10) (a) Jin, D.; Schmeier, T. J.; Williard, P. G.; Hazari, N.;
Bernskoetter, W. H. Lewis Acid Induced β-elimination from a
Nickelalactone: Efforts Toward Acrylate Production from CO2 and
Ethylene. Organometallics 2013, 32, 2152−2159. (b) Jin, D.; Williard,
G
https://dx.doi.org/10.1021/acs.organomet.9b00708
Organometallics XXXX, XXX, XXX−XXX