Accelerating ethylene polymerization using secondary metal ions in tetrahydrofuran

Article abstract of DOI:10.1039/c9dt04288a

We have prepared a new series of nickel phosphine phosphonate ester complexes that feature two metal-chelating polyethylene glycol (PEG) side arms. Metal binding and reactivity studies in polar solvents showed that they readily coordinate external cations, including alkali (Li⁺, Na⁺, K⁺), alkaline (Mg²⁺, Ca²⁺), transition (Sc³⁺, Co²⁺, Zn²⁺), post-transition (Ga³⁺), and lanthanide (La³⁺) metals. Although olefin polymerization reactions are typically performed in non-polar solvents, which cannot solubilize +2 and +3 metal cations, we discovered that our nickel catalysts could promote ethylene polymerization in neat tetrahydrofuran. This advance allowed us, for the first time, to systematically investigate the effects of a wide range of M⁺, M²⁺, and M³⁺ ions on the reactivity of nickel olefin polymerization catalysts. In ethylene homopolymerization, the addition of Co(OTf)₂ to our nickel-PEG complexes provided the largest boost in activity (up to 11-fold, 2.7 × 10⁶ g mol^-1 h^-1) compared to that in the absence of external salts. The catalyst enhancing effects of secondary metals were also observed in studies of ethylene and polar olefin (e.g., propyl vinyl ether, allyl butyl ether, methyl-10-undecenoate, and 5-acetoxy-1-pentene) copolymerization. Notably, combining either Co²⁺ or Zn²⁺ with our nickel complexes increased the rates of polymerization in the presence of propyl vinyl ether by about 5.0- and 2.4-fold, respectively. Although further studies are needed to elucidate the structural and mechanistic roles of the secondary metals, this work is an important advance toward the development of cation-switchable polymerization catalysts.

Full text of DOI:10.1039/c9dt04288a

Dalton
Transactions
View Article Online
View Journal | View Issue
PAPER
Accelerating ethylene polymerization using
secondary metal ions in tetrahydrofuran†
Cite this: Dalton Trans., 2019, 48,
17887
Dawei Xiao, Zhongzheng Cai and Loi H. Do
*
We have prepared a new series of nickel phosphine phosphonate ester complexes that feature two metal-
chelating polyethylene glycol (PEG) side arms. Metal binding and reactivity studies in polar solvents
showed that they readily coordinate external cations, including alkali (Li+, Na+, K+), alkaline (Mg2+, Ca2+),
transition (Sc3+, Co2+, Zn2+), post-transition (Ga3+), and lanthanide (La3+) metals. Although olefin polymer-
ization reactions are typically performed in non-polar solvents, which cannot solubilize +2 and +3 metal
cations, we discovered that our nickel catalysts could promote ethylene polymerization in neat tetra-
hydrofuran. This advance allowed us, for the first time, to systematically investigate the effects of a wide
range of M+, M2+, and M3+ ions on the reactivity of nickel olefin polymerization catalysts. In ethylene
homopolymerization, the addition of Co(OTf)2 to our nickel-PEG complexes provided the largest boost in
activity (up to 11-fold, 2.7 × 106 g mol−1 h−1) compared to that in the absence of external salts. The cata-
lyst enhancing effects of secondary metals were also observed in studies of ethylene and polar olefin
(e.g., propyl vinyl ether, allyl butyl ether, methyl-10-undecenoate, and 5-acetoxy-1-pentene) copolymeri-
zation. Notably, combining either Co2+ or Zn2+ with our nickel complexes increased the rates of polymer-
ization in the presence of propyl vinyl ether by about 5.0- and 2.4-fold, respectively. Although further
studies are needed to elucidate the structural and mechanistic roles of the secondary metals, this work is
an important advance toward the development of cation-switchable polymerization catalysts.
Received 5th November 2019,
Accepted 21st November 2019
DOI: 10.1039/c9dt04288a
rsc.li/dalton
To help streamline polyolefin synthesis, we are developing
switchable catalysts19 that could be easily fine-tuned by inter-
Introduction
Although polyolefins have been produced commercially since changing their secondary metal ions (Scheme 1).20–23 Unlike
the 1930s,1 achieving high precision in the polymerization redox switches that typically toggle between 2–3 different
process is still an ongoing challenge.2–4 Advances in late tran- modes,24–26 cation switches could access virtually unlimited
sition metal catalysis have led to the development of a diverse number of reactivity states since their differences in properties
assortment of nickel and palladium complexes that are (e.g., charge, size, Lewis acidity, redox behavior, coordination
capable of polymerizing both non-polar and polar olefins.5–9 geometry, etc.) could be exploited to alter catalysts in distinct
However, because the coordination insertion process is ways. One of the earliest examples of using secondary metals
complex, controlling how each olefin is incorporated and the to boost catalyst performance was work reported by Johnson,
total number of monomers in each polymer chain is difficult Brookhart, and coworkers.27 The researchers showed that
to achieve. Many creative strategies have been employed to nickel catalysts in the presence of Li+ were significantly more
obtain polyolefins with unique microstructures and compo- active in ethylene and hexyl acrylate copolymerization than in
sitions, such as the application of electronic and steric
tuning,10–13 hemi-labile ligands,14,15 or bimetallic active
sites.16,17 Given the ever growing demand for new materials
with new properties, there are numerous opportunities to
improve upon existing polymerization technologies.18
Department of Chemistry, University of Houston, 4800 Calhoun Rd., Houston, Texas,
77004, USA. E-mail: loido@uh.edu
†Electronic supplementary information (ESI) available. CCDC 1943648–1943650.
For ESI and crystallographic data in CIF or other electronic format see DOI:
10.1039/c9dt04288a
Scheme 1 Expanding the scope of secondary metal ions.
This journal is © The Royal Society of Chemistry 2019
Dalton Trans., 2019, 48, 17887–17897 | 17887

View Article Online
Paper
Dalton Transactions
its absence. Recently, Jordan and coworkers demonstrated that of a select group of metal salts tested, Zn2+ and Co2+ provided
secondary metals such as Li+ or Zn2+ could induce the self- the greatest catalyst enhancements in both ethylene homopoly-
assembly of palladium cages that are capable of polymerizing merization and ethylene and polar olefin copolymerization.
ethylene to high molecular weight polymers.28,29 Tonk’s These results suggest that cation-switchable catalysis could be a
research group has studied nickel–zinc heterobimetallics viable approach toward synthesizing different designer poly-
(Ni1–Zn30 and Ni2–Zn,31 Chart 1) and showed that they exhibit olefins from a general catalyst platform. We will discuss how
distinct reactivity from that of mononickel complexes that lack these studies have informed our ongoing catalyst design efforts
zinc. Our research group’s efforts to create cation-responsive and the challenges to be addressed in future work.
catalysts have led to the development of nickel phenoxyimine
(Ni3),32,33 nickel phenoxyphosphine (Ni4),34 and palladium
phosphine phosphonate ester (Pd1)35 complexes bearing poly-
Results and discussion
ethylene glycol (PEG) chelators for binding external metal ions
Synthesis of P,O-ligands and their nickel complexes
(Chart 1). We established through both solution and X-ray crys-
Our previous success in using phosphine phosphonate-PEG
ligands to prepare palladium catalysts led us to use this plat-
form for nickel.35,40 The ligands 3a (Ar = 2-methoxyphenyl)
and 3b (Ar = 2,6-dimethoxyphenyl) were synthesized according
to our standard procedures (see Experimental section). The P,
O-ligands were metallated by mixing with AgSbF6 and
[Ni(allyl)(Cl)]2 to furnish either 4a or 4b (Scheme 2).
tallographic studies that these complexes readily formed 1 : 1
nickel–alkali or palladium–alkali species. In general, we
observed that addition of alkali salts to the catalysts often led
to significant polymerization rate enhancements and changes
in their polymer molecular weight and branching. Unlike in
systems that utilize boranes as remote Lewis acid
activators,36,37 the secondary metals in our catalysts could be
easily removed (e.g., by the addition of crown ethers). Our
PEGylated complexes are also amenable to conventional
electronic and steric modifications, which means that many
degrees of tuning are possible to optimize the catalysts.33
Although we successfully demonstrated the beneficial
effects of secondary metal ions on polymerization,32–35 our
previous studies were limited to alkali (Li+, Na+, and K+) tetra-
aryl borate salts because they could be readily dissolved in
typical polymerization solvents, such as toluene, xylene, tri-
chlorobenzene, and dichloromethane. Because most +2, +3,
and higher charged cations are not soluble in hydrocarbon or
halogenated solvents, we could not incorporate non-alkali ions
into our catalysts (Scheme 1). In the present work, we have over-
come this solubility problem by developing nickel phosphine
phosphonate-PEG complexes that are active in tetrahydrofuran
(THF).38,39 This advance enabled us to broaden the scope of sec-
ondary metals to study, including alkaline, transition, post-tran-
sition, and lanthanide ions. Interestingly, we discovered that out
Chart 1 Crystallographically characterized heterobimetallic complexes
(Ni1–Zn, Ni2–Zn, Ni3–Na, Ni4–Na, and Pd1–Na) studied for their olefin
polymerization activity.
Scheme 2 Synthesis of the nickel complexes and metallation with Mn+
ions. Metal ions studied in this work: Li+, Na+, K+, Mg2+, Ca2+, Co2+
,
Cu2+, Zn2+, Sn2+, Al3+, Sc3+, Ga3+, Bi3+, and La3+
.
17888 | Dalton Trans., 2019, 48, 17887–17897
This journal is © The Royal Society of Chemistry 2019

View Article Online
Dalton Transactions
Paper
Compound 4a was isolated as a dark red oil whereas 4b was an
orange solid. As standard catalysts, we also synthesized com-
plexes 6a and 6b that have ethoxy instead of PEG groups. The
X-ray crystal structure of 6a (Fig. S56†) indicates that the nickel
center is ligated by both P- and O-donors from the phosphine
phosphonate ester ligand (Ni–P = 2.20 Å and Ni–O = 1.92 Å)
and has an allyl group bound in an η3 coordination mode.
Secondary metal ion binding
Similar to their palladium analogues,35 our nickel complexes
readily formed adducts with LiBArF4, NaBArF4, and KBArF
4
(BArF − = tetrakis(3,5-bis(trifluoromethyl)phenyl)borate anion)
4
in non-polar solvents. The nickel–alkali binding behavior was
studied by H NMR spectroscopy in CDCl3. We observed that
Fig. 2 X-ray crystal structure of complex 4a-Na (ORTEP view, displace-
ment ellipsoids drawn at 30% probability level). Hydrogen atoms and
disordered THF/pentane molecules have been omitted for clarity.
1
the resonance for the allyl hydrogen atom at ∼5.7 ppm for
both 4a (Fig. S1†) and 4b (Fig. S2–S4) shifted upfield upon the
addition of alkali salts. Job plot41 analysis of the NMR data for
4a + Na+ and 4b + Na+ suggests that the optimal nickel :
sodium binding stoichiometry is 1 : 1 (Fig. 1, peak maxima at
χNi = 0.5). Interestingly, the 31P NMR spectra of 4a vs. 4a/Na+
(Fig. S9†) and 4b vs. 4b/Na+ (Fig. S10†) in CDCl3 showed peaks
with nearly identical chemical shifts, suggesting that the
sodium ions do not interact directly with the phosphine oxide
moiety. Solution binding studies conducted in THF instead of
CDCl3 showed NMR spectral shifts of the allyl group when
either Na+ or Zn2+ was added to 4a at 40–50 °C (Fig. S6–S8).
Unfortunately, we do not currently understand why the peak
shifts are temperature dependent.
To determine the structures of the nickel–sodium species
in the solid state, single crystals were grown by slow diffusion
of pentane into a THF/Et2O mixture containing the nickel
complexes and NaSbF6 (1 : 1) (crystals could not be grown
using NaBArF4). X-ray diffraction analysis of the 4a + Na+ crys-
tals showed a complex with the formulation [NiNa(allyl)(3a)
(SbF6)2] (4a-Na, Fig. 2). The nickel center has the expected dis-
torted planar geometry with Ni–P and Ni–O bond distances of
2.19 and 1.92 Å, respectively, similar to those in mononickel
6a (Fig. S56†). Additionally, four oxygen donors from the
pendent PEG chains (Na–O bond distances range from 2.40
−
to 3.00 Å) and two fluorine donors from two SbF6 anions
(Na–F = ∼2.30 Å) occupy the sodium coordination sphere.
Surprisingly, the phosphonate oxygen donor does not coordi-
nate to Na+ and the nickel–sodium separation is ∼6.68 Å,
which is much longer than those in our other heterobimetallic
structures (Chart 1).32,34,35 The X-ray structure of 4b + Na+
revealed a discrete complex with the composition [NiNa(allyl)
(3b)(SbF6)]SbF6 (4b-Na, Fig. S55†). The structure of 4b-Na is
similar to that of 4a-Na, except that one of the methoxy groups
in 4b-Na appears to have close contact with its nickel center
(Ni–Omethoxy = 2.61 Å, van der Waals distance is ∼3.55 Å 42).
Interestingly, the phosphonate oxygen donor also does not
interact with Na+, which is consistent with our 31P NMR data
(Fig. S9 and S10).
Solution studies were then performed to investigate
whether non-alkali metal ions could coordinate to the nickel
phosphine phosphonate-PEG complexes. When 1 equiv. of
−
Co(OTf)2 (OTf− = SO3CF3 triflate anion) was added to 4a in
CD3CN, the NMR peaks became significantly broadened due
to the paramagnetism of Co2+ (Fig. S11†). Interaction of Co2+
with 4a was further established by UV-vis absorption spec-
troscopy, which showed a bathochromic shift of the band at
∼400 nm to ∼410 nm and increase in peak intensity upon the
addition of excess Co(OTf)2 to a THF solution of the nickel
complex (Fig. S12†). These spectral changes suggest that
cobalt is capable of binding 4a but the exact structure of the
4a-Co complex is uncertain.
When diamagnetic Zn(OTf)2 was combined with 4a in
CD3CN, the allyl hydrogen peak at 5.7 ppm shifted downfield
when increasing amounts of Zn2+ were introduced (Fig. S5†).
The Job plot obtained from mixing 4a + Zn2+ in different ratios
suggests that the 4a-Zn complex has a 1 : 1 nickel : zinc stoi-
chiometry (Fig. 1, blue squares). Unfortunately, despite many
attempts to grow single crystals of either 4a-Co or 4a-Zn for
Fig. 1 Job plots for the nickel complexes with NaBArF (4a + Na+
=
=
4
black dots, 4b + Na+ = red triangles) in CDCl3 or Zn(OTf)2 (4a + Zn2+
blue squares) in CD3CN. The mole fraction χNi = [nickel]/([nickel] +
[Mn+]). The total concentration of nickel and secondary metal salts is
6 mM for all data points. The Job plot data for 4b + Li+ and 4b + K+ are
given in Tables S3 and S4,† respectively; they also show 1 : 1 binding.
This journal is © The Royal Society of Chemistry 2019
Dalton Trans., 2019, 48, 17887–17897 | 17889

View Article Online
Paper
Dalton Transactions
X-ray structural characterization, our efforts were not Zn2+ could bind 6a at multiple sites. As demonstrated by our
successful. polymerization results below, there appears to be a strong cor-
As shown in Scheme 2A, we propose that the heterobimetal- relation between the stability of the nickel–metal complexes
lic species above could adopt one of two structural motifs, and their catalyst activity (Fig. 3).
either an open (4a-M or 4b-M) or bridged (4a-M′ or 4b-M′)
form. It is also feasible that the two structural isomers could
exist in equilibrium. Although we have only obtained struc-
tures of the open form for nickel (e.g., 4a-Na in Fig. 2 and 4b-
Na in Fig. S55†), our previous studies showed that the phos-
phine phosphonate-PEG ligands could bridge two different
metal ions in the presence of chloride (e.g., Pd1–Na, Chart 1).
mixture of toluene and dichloromethane (24 : 1) and then
However, it is possible that in the absence of suitable external
Ethylene homopolymerization studies in toluene and
dichloromethane
To determine whether our nickel complexes are competent cat-
alysts, they were first tested in ethylene homopolymerization.
For these studies, the Ni complexes were dissolved in a
stirred under ethylene (200 or 400 psi) at 80 °C for 1 h
bridge anions, 4a-M/4b-M may be preferred over 4a-M′/4b-M′.
(Table 1). In general, all of the catalysts exhibited moderate
For comparison, we next investigated the metal binding be-
activity (∼105 g mol−1 h−1) and gave polyethylene (PE) with
havior of the conventional complex 6a, which lacks the two
moderate branching (13–22 per 1000 carbons) and low mole-
PEG chains present in 4a and 4b. In the non-coordinating
cular weight (Mn = ∼102 to 103). Consistent with established
trends,10,43 the bulkier catalysts 4b (entry 3) and 6b (entry 7)
were less active but provided PE with higher Mn than their
corresponding less bulky 4a (entry 1) and 6a (entry 5) com-
plexes, respectively. These data also indicated that the PEG
groups in 4a and 4b are not catalyst inhibiting since their non-
PEGylated counterparts gave similar results. All of the PE pro-
duced showed relatively narrow polydispersity (Mw/Mn typically
less than 2.0), suggesting that the nickel complexes are single
component catalysts.
solvent CDCl3, the addition of 1 equiv. of NaBArF to 6a led to
4
subtle but distinct changes to its 1H NMR spectrum
(Fig. S13†). For example, the allyl hydrogen peak at 5.72 ppm
shifted upfield to 5.65 ppm in the presence of Na+. These
changes are most likely due to binding of Na+ to one or more
of the oxygen donors in the phosphonate ester group of 6a.
Interestingly, when 1 equiv. of Zn(OTf)2 was added to 6a in
CD3CN, many additional NMR peaks appeared (Fig. S14†),
suggesting that more than one new species had formed. Since
6a does not have a well-defined pocket for secondary metals,
Next, we evaluated the effects of adding alkali salts to the
polymerization reactions in toluene/dichloromethane (24 : 1)
(Tables 1 and S6†). For the catalysts containing 2-methoxyphe-
nyl substituents (i.e., 4a and 6a), the presence of 1 equiv. of
LiBArF4, NaBArF4, or KBArF4 relative to nickel, resulted in up to
∼3.5-fold increase in catalyst activity. Interestingly, the PE pro-
ducts had similar morphologies regardless of whether
alkali salts were used. The activity enhancing effects of M+ is
tentatively attributed to electrophilic activation and/or greater
structural stability (vide infra).34,35 The addition of alkali ions
to the bulkier catalysts 4b and 6b, led to slight decreases in
polymerization rates. Because these catalysts bear sterically
demanding 2,6-dimethoxyphenyl substituents, the binding of
alkali ions to either 4b or 6b may cause excessive steric conges-
tion at the active site, thereby reducing polymerization
efficiency.
Fig. 3 Comparison of catalyst activity in ethylene homopolymerization
using either 4a or 6a in the presence of various metal triflate salts (THF
solvent). See Table S8† for detailed polymerization data.
Table 1 Ethylene homopolymerization data for 4a, 4b, 6a, and 6ba
Activity (×105 g mol−1 h−1
)
Branchesc (/1000 C)
Mn d (×103)
Mw/Mn
d
Entry
Complex
Salt
Polymer yield (g)
1
4a
4a
4b
4b
6a
6a
6b
6b
None
Na+
8.6
29.9
6.1
3.5
7.4
26.6
5.4
3.3
8.6
29.9
3.0
1.8
7.4
26.6
2.7
1.6
22
30
13
16
18
26
17
15
0.84
1.04
4.79
7.61
0.69
0.98
7.36
13.92
1.7
1.2
1.4
1.4
1.7
2.0
1.2
1.3
2
3b
4b
5
None
Na+
None
Na+
6
7b
8b
None
Na+
a Polymerization conditions: Ni catalyst (10 μmol), MBArF (10 μmol), ethylene (200 psi), 2 mL DCM, 48 mL toluene, 1 h at 80 °C.
4
b Polymerization conditions: Ni catalyst (20 μmol), MBArF (20 μmol), ethylene (400 psi), 2 mL DCM, 48 mL toluene, 1 h at 80 °C. c The total
number of branches per 1000 carbons was determined by 14H NMR spectroscopy. d Determined by GPC in trichlorobenzene at 150 °C.
17890 | Dalton Trans., 2019, 48, 17887–17897
This journal is © The Royal Society of Chemistry 2019

View Article Online
Dalton Transactions
Paper
Ethylene homopolymerization studies in tetrahydrofuran
found that adding metal halides to 4a typically reduced (e.g.,
Ca2+, Co2+) or inhibited (e.g., Mg2+, Al3+, Fe3+, Cu2+) ethylene
To expand our studies of secondary metal ions, we first had to homopolymerization (Table S7†). Interestingly, the presence of
identify a polymerization-compatible solvent that could dis- zinc chloride gave a 6.8-fold increase in activity (1.7 × 106
solve +2 and +3 metal salts. Li and coworkers showed that g mol−1 h−1) compared to that in its absence (2.5 × 105 g mol−1
nickel phenoxyphosphine catalysts are active in the presence h−1). Substituting ZnCl2 with either ZnBr2 or ZnI2, however,
of 2 M polar solvent additives, such as ethanol, diethyl ether, led to lower catalyst activity (∼1.7 × 105 g mol−1 h−1). It is gen-
acetone, ethyl acetate, water, acetonitrile, and dimethyl- erally observed that coordinating anions, such as halides,
formamide.39 There are only a few reports of coordination poison metal catalysts by occupying open coordination sites.48
insertion polymerizations performed in neat polar solvent, However, perhaps in the 4a/ZnCl2 complex, the chloride ions
specifically in water44–47 and THF.38 Coordinating/polar sol- may be involved in stabilizing the active nickel–zinc species.49
vents are typically avoided in polymerization reactions because In related studies, Tonks and coworkers found that NiZn com-
they could compete with olefin binding and thus, diminish plexes featuring bridging halides showed lower ethylene
catalyst efficiency. Motivated by the studies above, we pro- polymerization activity than their mononickel counterparts
ceeded to test our nickel catalysts in the presence of varying (Ni2–Zn, Chart 1).31 In contrast, NiZn complexes with non-
amounts of THF in toluene. To our surprise, the polymeriz- bridging halides (Ni1–Zn) showed increased ethylene polymer-
ation data showed that the nickel catalysts were relatively toler- ization rates compared to that of their parent nickel com-
ant of THF (Table 2). For catalyst 4a, the use of 4% THF plexes.30 Since we were unable to obtain crystallographic
(entry 2) led to no loss in activity, whereas the use of 20% THF characterization of 4a/ZnCl2, we are uncertain how the chlor-
(entry 3) led to only ∼1.3-fold reduction in activity compared to ide is incorporated into our nickel–zinc structures.
reactions conducted in the absence of THF (entry 1).
To avoid the use of halides, we screened a variety of metal
Interestingly, complex 4a in neat THF still gave a moderate triflate salts instead (Table S8†). When 4a was combined with
activity of 2.5 × 105 g mol−1 h−1 (entry 4). When K+ was added 1 equiv. of metal triflate in THF and then exposed to 200 psi of
to 4a, excellent activity was observed up to 20% THF (3.1 × 106 ethylene for 1 h at 80 °C, high polymerization activity was
g mol−1 −1, entry 7). As expected, in 100% THF (entry 8), the observed for K+, Zn2+, Mg2+, Ca2+, La3+, Sc3+, Ga3+, and Co2+
h
activity of 4a-K dropped about 3.8-folds compared to that in (from 3.7 × 105 to 2.7 × 106 g mol−1 h−1, entries 2–11).
0% THF (entry 5). Similar reactivity trends were observed for Polymerization rate enhancements of 10.6- and 4.8-fold were
the standard catalyst 6a in the presence and absence of alkali obtained using cobalt and zinc salts, respectively, relative to
salts (entries 10–18).
that of the control. The PE products produced in all reactions
The results above encouraged us to extend the scope of our were similar, with moderate branches (18 to 27 per 1000
cation studies. To survey secondary metals from different carbons), low molecular weight (Mn = <1 × 103), and relatively
groups of the periodic table, we chose a variety of salts from narrow polydispersity (Mw/Mn ≤ 2.1). Surprisingly, Cu2+ com-
the alkaline, transition, post-transition, and lanthanide metal pletely shut down the activity (Table S8,† entry 12), whereas
series. Using our standard reaction conditions in neat THF, we Al3+ (entry 13) and Bi3+ (entry 15) lowered the activity of 4a.
Table 2 Ethylene homopolymerization data for 4a and 6a in toluene and THF solvent mixturesa
Amount of THFc
Polymer yield (g)
Activity (×105 g mol−1 h−1
)
Branchesd (/1000 C)
Mn e (×103)
Mw/Mn
e
Entry
Complex
Salt
1b
2
3
4a
4a
4a
4a
4a
4a
4a
4a
4a
6a
6a
6a
6a
6a
6a
6a
6a
6a
None
None
None
None
K+
0%
4%
20%
100%
0%
8.6
8.7
6.8
8.6
8.7
6.8
22
21
19
21
28
32
31
24
24
18
16
17
19
22
24
27
21
19
0.84
1.14
0.86
0.70
1.12
0.84
0.77
0.83
0.95
0.69
1.64
0.78
0.90
0.73
0.87
0.61
0.95
1.01
1.7
1.6
1.3
1.1
1.2
1.7
2.2
1.9
1.2
1.7
1.2
1.5
1.2
1.7
1.6
1.6
1.2
1.7
4
2.5
2.5
5b
6
28.9
28.7
30.9
7.5
6.1
7.4
4.1
7.5
2.5
25.4
21.9
36.3
6.6
28.9
28.7
30.9
7.5
6.1
7.4
4.1
7.5
2.5
25.4
21.9
36.3
6.6
K+
4%
7
8
9
K+
20%
100%
100%
0%
K+
Na+
None
None
None
None
K+
10b
11
12
13
14b
15
16
17
18
4%
20%
100%
0%
K+
4%
K+
20%
100%
100%
K+
Na+
6.5
6.5
a Polymerization conditions: Ni catalyst (10 μmol), MBArF4 (10 μmol), ethylene (200 psi), toluene and THF in different ratios, 1 h at 80 °C. b These
reactions were performed in 2 mL DCM and 48 mL toluene. c The percentage of THF in Toluene. d The total number of branches per 1000
carbons was determined by 1H NMR spectroscopy. e Determined by GPC in trichlorobenzene at 150 °C.
This journal is © The Royal Society of Chemistry 2019
Dalton Trans., 2019, 48, 17887–17897 | 17891

View Article Online
Paper
Dalton Transactions
Control polymerization studies using the ligand 3a/Co(OTf)2 (entry 3) showed a 5.0- and 2.4-fold boost in catalyst activity,
or complex 4a/Ag(OTf) in the presence of 200 psi of ethylene respectively, compared to that of 4a alone (entry 1). The poly
in THF at 80 °C did not provide any polymer products. Many (ethylene-co-PVE) products were washed with chloroform to
factors could account for differences in the secondary metal remove any trace impurities and their identities were con-
ion effects, such as the coordination preferences of the PEG firmed by 1H (Fig. S44†) and 13C NMR (Fig. S45†) spectroscopy.
chelator for Mn+, solubility of the metal salts in THF, and/or The nickel–potassium complex 4a-K (entry 4) gave only a
nickel : Mn+ binding stoichiometry. Detailed studies to charac- modest 1.1-fold increase in catalyst activity compared to that
terize the various mixed metal complexes are needed to better of 4a only. As a control, reaction of ethylene and PVE in the
understand their structures and reactivity.
presence of only Co(OTf)2 (i.e., without 4a) did not yield any
The important role of the PEG chains in binding secondary polymers (entry 5). Interestingly, copolymerization of ethylene
metals was apparent by comparing the reactivity of 4a vs. 6a in and PVE using 6a was not enhanced in the presence of Co2+
THF (Fig. 3). For example, ethylene polymerization using 6a in (cf. entries 6 vs. 7).
combination with K+, Zn2+, Mg2+, Ca2+, La3+, Sc3+, Ga2+, or
Under similar reaction conditions, 4a itself was a poor cata-
Co2+ triflate had negligible effects on catalyst activity lyst for ethylene and ABE copolymerization (Table 3, entry 8).
(Table S8,† entry 16 vs. entries 17–24). Since THF solvent mole- However, moderate activities of 1.9 × 103 and 1.0 × 103 g mol−1
cules are coordinating, they are capable of competing with 6a h−1 were observed in the presence of Co2+ (entry 9) and Zn2+
for binding Mn+. These results indicate that without additional (entry 10) ions, respectively. The polymers obtained were con-
chelators to capture Mn+, 6a cannot form strong adducts with firmed to be poly(ethylene-co-ABE) by 1H NMR spectroscopy
external cations in polar solvents.
(Fig. S46†).
Examples of ethylene and vinyl or allyl ether copolymeriza-
tion by late transition metal catalysts are relatively rare
Ethylene and polar olefin copolymerization studies in
tetrahydrofuran
(Table S9†). A survey of the literature revealed that the most
The results above were exciting because they suggested that successful nickel-based systems are those ligated by imidazo
the more active 4a-M catalysts might be capable of catalyzing [1,5-a]quinolin-9-olate-1-ylidene (7),50 phosphine sulfonate
the polymerization of difficult to incorporate polar monomers, (8,51 1052), and diphosphazane monoxide (9).53 Interestingly,
including those that could form metallocyclic intermediates. our 4a-Co complex is so far the most active nickel catalyst for
In these investigations, 4a was mixed with 1 equiv. of a metal ethylene and vinyl ether copolymerization. For example, 4a-Co
salt and then exposed to ethylene and a polar olefin (e.g., PVE (activity = 3.4 × 104 g mol−1 h−1) is about 20-fold more active
= propyl vinyl ether, ABE = allyl butyl ether, MU = methyl-10- than 9 (1.7 × 103 g mol−1
h
−1) under optimized conditions.
undecenoate, AP = 5-acetoxy-1-pentene, and MA = methyl acry- However, our nickel–cobalt catalysts furnished copolymers
late) in THF for 2 h at 80 °C (Table 3). For ethylene and PVE with lower comonomer incorporation (0.2 mol% for 4a-Co vs.
copolymerization, we found that 4a-Co (entry 2) and 4a-Zn 2.0 mol% for 9) and slightly lower molecular weight (1.3 × 103
Table 3 Ethylene and polar olefin copolymerization data for 4a with metal salts in THFa
Comonomerc
(conc., M)
Polymer
yield (g)
Activity
Branchesd
(/1000 C)
Inc. f
(mol%)
e
Entry
Cat.
Salt
(×103 g mol−1 h−1
)
Mn e (×103)
Mw/Mn
1
2
3
4
5
6
7
8
4a
4a
4a
4a
none
6a
6a
4a
4a
4a
4a
4a
4a
4a
4a
4a
4a
4a
None
PVE (1.0)
PVE (1.0)
PVE (1.0)
PVE (1.0)
PVE (1.0)
PVE (1.0)
PVE (1.0)
ABE (0.5)
ABE (0.5)
ABE (0.5)
MU (0.5)
MU (0.5)
MU (0.5)
AP (0.5)
0.55
2.74
1.34
0.61
0
0.58
0.47
<0.01
0.15
0.08
0.58
2.94
2.44
0.73
5.45
2.17
0
6.9
21
25
27
23
—
18
18
—
23
25
20
23
23
21
28
21
—
—
1.43
1.29
1.14
1.26
—
0.74
0.82
—
3.35
1.50
1.61
2.15
1.94
2.23
1.78
1.82
—
0.21
0.17
0.24
0.25
—
0.17
0.14
—
0.21
0.33
0.38
0.69
0.74
0.31
0.20
0.29
—
1.2
1.3
1.2
1.8
—
1.5
1.4
—
2.4
1.5
1.3
1.5
1.3
1.6
1.4
1.2
—
Co(OTf)2
Zn(OTf)2
34.2
16.8
7.6
—
7.2
5.8
<0.1
1.9
KBArF
4
Co(OTf)2
None
Co(OTf)2
None
Co(OTf)2
Zn(OTf)2
None
Co(OTf)2
Zn(OTf)2
None
9
10
11b
12b
13b
14
15
16
17
18
1.0
14.5
73.5
61.0
9.1
68.1
27.1
—
Co(OTf)2
Zn(OTf)2
None
AP (0.5)
AP(0.5)
MA (1.0)
MA (1.0)
Co(OTf)2
0
—
—
—
—
a Polymerization conditions: Ni catalyst (40 μmol), metal salt (40 μmol), ethylene (400 psi), 50 mL of THF, 2 h at 80 °C. b Polymerization con-
ditions: Ni catalyst (20 μmol), metal salt (20 μmol). c Comonomer abbreviation: PVE = propyl vinyl ether, ABE = allyl butyl ether, MU = methyl-10-
undecenoate, AP = 5-acetoxy-1-pentene, MA = methyl acrylate. d The total number of branches per 1000 carbons was determined by 1H NMR spec-
troscopy. e Determined by GPC in trichlorobenzene at 150 °C. f Determined by 1H NMR spectroscopy.
17892 | Dalton Trans., 2019, 48, 17887–17897
This journal is © The Royal Society of Chemistry 2019

View Article Online
Dalton Transactions
Paper
for 4a-Co vs. 8.0 × 103 for 9). The low comonomer incorpor- polar comonomers tested are in the order MU > AP > PVE >
ation (<1 mol%) in our polymers suggests that not every chain ABE ≫ MA. In almost all cases, the addition of Co(OTf)2 to 4a
contains a polar olefin. It is possible that the incorporation of led to greater polymerization rate enhancements than the
polar monomers leads to chain capping/termination rather addition of Zn(OTf)2.
than copolymerization. On the basis of previous work,33 we
In reactions involving functional monomers, we discovered
expect that increasing the steric bulk of our catalysts would that solvents have minimal impact on polymerization rates.
increase their polymer molecular weights. In comparison to For example, when 4a was used in the polymerization of ethyl-
nickel, more examples are known for palladium catalysts that ene and PVE in toluene, the activity was measured to be 6.7 ×
are capable of catalyzing ethylene and ether olefin copolymeri- 103 g mol−1 h−1, which is similar to that when the reaction was
zation.54 They include palladium complexes bearing phos- performed in neat THF (activity = 6.9 × 103 g mol−1 h−1
,
phine sulfonate (11,55 13,56 1757), phosphine phosphine oxide Table 3, entry 1). These results suggest that when a monomer
(12,12 14,58 1659), and diimine (15)60 ligands. To date, it is still has greater coordinating ability than the solvent (e.g., PVE >
challenging to obtain catalysts that are fast (activity >104
g
THF) and is present in large excess, it outcompetes solvent
mol−1 h−1) and produce copolymers with high ether monomer molecules for catalyst binding. Thus, we believe that polar sol-
incorporation (>2 mol%) and molecular weight (>104).
vents are not detrimental to the polymerization of certain func-
It has been demonstrated that increasing the distance tional olefins and are generally underutilized for this
between the polar group and the double bond typically leads application.
to enhanced catalyst compatibility of polar olefins.15,16,61
These specialized monomers cannot form stable five- and six-
Ongoing challenges
membered ring intermediates that are believed to inhibit the Although our results are intriguing from the switchable cata-
polymerization process.62 Our studies showed that both MU lyst design standpoint, several important challenges remain.
and AP could be incorporated using 4a. In the absence of exter- First, it is difficult to predict a priori whether a secondary
nal metal salts, 4a provided poly(ethylene-co-MU) with about metal would enhance or diminish catalyst performance.
0.38 mol% with an activity of 1.4 × 104 g mol−1 h−1 (Table 3, Although the majority of cations tested led to improved catalyst
entry 11). When 1 equiv. of Co(OTf)2 (entry 12) or Zn(OTf)2 activity (e.g., Li+, Na+, K+, Mg2+, Ca2+, Co2+, Zn2+, Sc3+, Ga3+,
(entry 13) was added to 4a, the activity increased by about 5.1- and La3+), several did not (e.g., Cu2+, Al3+, Sn2+, Bi3+). We
and 4.2-fold, respectively. The MU incorporation increased hypothesize that cations in the latter category either did not
slightly with metal salts (e.g., 1.8-fold for Co2+ and 1.9-fold for bind to the nickel complex or formed mixed metal species that
Zn2+). Similarly, the addition of Co2+ or Zn2+ to 4a also led to are catalytically inactive. Furthermore, we cannot dismiss the
greater polymerization rates in the presence of AP. The presence possibility that the heterometallic species might precipitate
of Co2+ increased activity by 7.5-fold (entry 15), whereas the out of solution when activated due to poor solubility. We
presence of Zn2+ increased activity by 3.0-fold (entry 16).
propose that a more extensive set of secondary metals should
Catalyst 4a was found to be inactive for ethylene and methyl be screened to determine whether any reactivity trends could
acrylate copolymerization (Table 3, entry 17). The addition of be established. Preliminary studies comparing the ionic
Co(OTf)2 also had no beneficial effects (entry 18). Since poly- radius63 (Fig. S17A†) and Lewis acidity64 (Fig. S17B†) of the
ethylene was not obtained from either reactions, we concluded cations tested do not show any obvious correlation with
that MA is catalyst inhibiting.
polymerization activity (Table S10†). Detailed investigations of
Fig. 4 summarizes our polymerization data for 4a, 4a-Co, the active species, including their nuclearity and polymeriz-
and 4a-Zn. Our results showed that the relative reactivity of the ation mechanisms, should be performed to understand the
roles of secondary metals in catalysis.
Second, designing ligands that provide pre-organized het-
erobimetallic structures and have broad secondary metal ion
specificity is difficult to achieve. Although we wanted to assem-
ble bridged 4a-M′ complexes to promote metal–metal coopera-
tivity,65 the X-ray structures of 4a-Na and 4a-Na indicated that
they could adopt non-bridged motifs (Scheme 2A). Since the
two metals do not interact in these structures, the sodium ion
does not influence the electronic environment of nickel.
Perhaps the lack of metal–metal cooperativity explains why the
addition of cations do not affect the microstructures of our
polymers, unlike in our other systems.32 We believe that this
problem could be solved by making the PEG chains more rigid
so that the nickel–metal distances are shorter. To enhance the
binding affinity of external cations, it might be necessary to
Fig. 4 Comparison of catalyst activity in ethylene and polar olefin
copolymerization studies by 4a, 4a-Zn, and 4a-Co. Detailed polymeriz-
ation data are provided in Table 3. Abbreviations: PVE = propyl vinyl
replace PEG with other chelators that have greater metal ion
specificity.
ether, ABE = allyl butyl ether, MU = methyl-10-undecenoate, AP =
5-acetoxy-1-pentene.
This journal is © The Royal Society of Chemistry 2019
Dalton Trans., 2019, 48, 17887–17897 | 17893

View Article Online
Paper
Dalton Transactions
Third, despite demonstrating that our cation-switchable cat-
NMR spectra were acquired using JEOL spectrometers
alysts are capable of promoting ethylene polymerization in the (ECA-400, 500, and 600) and referenced using residual solvent
presence of polar olefins (Table 3), their catalyst activities are peaks. All 13C NMR spectra were proton decoupled. 31P NMR
still several orders of magnitude lower than those in ethylene spectra were referenced to phosphoric acid. For polymer
homopolymerization (Tables
1
and S6†). The molecular characterization, 1H NMR spectroscopy: each NMR sample
weights of the polar copolymers also tend to be lower than contained ∼20 mg of polymer in 0.5 mL of 1,1,2,2-tetrachlor-
those for homopolyethylene. These longstanding copolymeri- oethane-d2 (TCE-d2) and was recorded on a 500 MHz spectro-
zation problems have been attributed to stronger preferences meter using standard acquisition parameters at 120 °C. 13C
for σ- rather than π-monomer binding, tendency to undergo NMR spectroscopy: each NMR sample contained ∼50 mg of
β-X elimination (where X = usually acetate or halide), and back- polymer and 50 mM (8.7 mg) chromium acetylacetonate Cr
chelation by polar groups.62 We propose that through further (acac)3 in 0.5 mL of TCE-d2 and was recorded at 120 °C
catalyst engineering using our site-differentiated ligands,65 it (125 MHz). The samples were acquired using a 90° pulse of
might be possible to circumvent these undesired reaction 11.7 μs, a relaxation delay of 4 s, an acquisition time of 0.81 s,
pathways.
and inverse gated decoupling. The samples were preheated for
30 min prior to data acquisition. The carbon spectra were
assigned based on the chemical shift values reported in the lit-
erature.32 The NMR spectra of poly(ethylene-co-PVE),55 poly
(ethylene-co-ABE),52 poly(ethylene-co-MU),61 and poly(ethylene-
Conclusion
In summary, we have developed a new class of cation-respon- co-AP)61 matched those reported.
sive nickel catalysts that are active in neat tetrahydrofuran. For
High-resolution mass spectra were obtained from the mass
the first time, we were able to systematically evaluate and spectral facility at the University of Houston. Elemental ana-
compare the effects of M+, M2+, and M3+ secondary metals on lyses were performed by Atlantic Microlab.
ethylene homo- and copolymerization. We discovered that a
Gel permeation chromatography (GPC) data were obtained
variety of alkali, alkaline, transition, post-transition, and using a Malvern high temperature GPC instrument equipped
lanthanide ions are capable of activating our nickel catalysts, with refractive index, viscometer, and light scattering detectors
some more so than others. In general, metal triflates are pre- at 150 °C with 1,2,4-trichlorobenzene, stabilized with 125 ppm
ferred over their metal halide salts since coordinating anions BHT, as the mobile phase. A calibration curve was established
could be catalyst inhibiting. Metal binding studies indicated using polystyrene standards in triple detection mode. All mole-
that Na+ and Zn2+ both bind to 4a/4b to form 1 : 1 species in cular weights reported are based on triple detection.
solution. The structures of the heterobimetallic 4a-Na and 4b-
Synthesis
Na complexes were determined by X-ray crystallography, which
showed that the nickel and sodium ions adopt open rather
Preparation of 4a. Inside the glovebox, a solution containing
than bridged structures. Among the select group of secondary 3a (200 mg, 0.33 mmol, 1.0 equiv.) and AgSbF6 (113 mg,
metals tested, Co2+ and Zn2+ provided the greatest catalyst 0.33 mmol, 1.0 equiv.) in 10 mL of CH2Cl2 was stirred for
activity enhancements. Most notably, the 4a-Co and 4a-Zn cata- 10 min at RT. Solid [Ni(allyl)Cl]2 (45 mg, 0.16 mmol, 0.5
lysts were able to significantly increase ethylene polymeriz- equiv.) was added in small portions. The reaction mixture was
ation rates in the presence of polar olefin (e.g., PVE, ABE, MU, stirred for an additional 3 h. The resulting red mixture was fil-
and AP). However, further studies are needed to understand tered through a pipet plug and then dried under vacuum to
the secondary metal effects. We anticipate that future research give a dark red oil. The product was recrystallized by dissolving
will lead to greater appreciation for the capabilities of cation- in CH2Cl2 and then layering with pentane to afford the final
switchable catalysts and offer new cost- and time-saving product as a dark red oil (289 mg, 0.31 mmol, 93.2%). 1H NMR
methods to synthesize designer polyolefins.
(CDCl3, 400 MHz): δ (ppm) = 7.69 (m, 1H), 7.66 (m, 1H), 7.56
(m, 3H), 7.14–7.08 (m, 3H), 6.97 (t, JHH = 7.8 Hz, 2H), 6.78 (m,
2H), 5.71 (m, 1H), 4.05 (m, 2H), 3.99 (m, 2H), 3.87 (s, 6H), 3.52
(m, 12H), 2.59 (br, 1H). 13C NMR (CDCl3, 125 MHz): δ (ppm) =
160.73 (d, JPC = 7.4 Hz), 134.87 (dd, JPC = 35.5, 13.5 Hz), 134.55,
134.44 (d, JPC = 4.9 Hz), 134.12, 133.74 (t, JPC = 14.8 Hz),
Experimental section
General procedures
Commercial reagents were used as received. All air- and water- 133.55, 131.09 (d, JPC = 13.5 Hz), 127.72 (dd, JPC = 188.5, 17.1
sensitive manipulations were performed using standard Hz), 121.75 (d, JPC = 8.6 Hz), 115.62, 115.24, 114.66, 111.71
Schlenk techniques or under a nitrogen atmosphere using a (d, JPC = 4.9 Hz), 71.79, 70.35, 69.36 (d, JPC = 6.1 Hz), 67.58 (d,
glovebox. Anhydrous solvents were obtained from an JPC = 6.1 Hz), 59.08, 56.06. 31P NMR (CDCl3, 162 MHz):
Innovative Technology solvent drying system saturated with δ (ppm) = 23.25 (d, JPP = 20.4 Hz), −1.97 (d, JPP = 20.6 Hz).
argon. High-purity polymer grade ethylene was obtained from
Preparation of 4b. Inside the glovebox, a solution containing
Matheson TriGas without further purification. The compounds 3b (100 mg, 0.15 mmol, 1.0 equiv.) and AgSbF6 (52 mg,
(2-bromophenyl)bis(2-methoxyphenyl)
phosphine35
and 0.15 mmol, 1.0 equiv.) in 10 mL of CH2Cl2 was stirred for
10 min at RT. Solid [Ni(allyl)Cl]2 (20 mg, 0.08 mmol, 0.5
[Ni(allyl)Cl]2 66 were prepared according to literature procedures.
17894 | Dalton Trans., 2019, 48, 17887–17897
This journal is © The Royal Society of Chemistry 2019

View Article Online
Dalton Transactions
Paper
equiv.) was added in small portions. The reaction mixture was 0.16 mmol, 85.9%). 1H NMR (CDCl3, 400 MHz): δ (ppm) = 7.66
stirred for an additional 3 h. The resulting red mixture was fil- (m, 1H), 7.58 (m, 1H), 7.51 (m, 2H), 7.44 (t, JHH = 8.4 Hz, 2H),
tered through a pipet plug and then dried under vacuum to 6.61 (dd, JHH = 8.2, 4.4 Hz, 4H), 5.67 (m, 1H), 3.83(m, 4H), 3.64
give a dark red oil. Upon the addition of pentane and after stir- (s, 12H), 2.86 (br, 2H), 2.30 (d, JHH = 13.2 Hz, 2H), 1.16 (t,
ring for ∼5 min, an orange solid formed. The product was JHH = 6.8 Hz, 6H). 13C NMR (CDCl3, 100 MHz): δ (ppm) =
recrystallized by dissolving in CH2Cl2 and then layering with 161.54 (d, JPC = 2.0 Hz), 138.16 (dd, JPC = 40.4, 12.7 Hz), 134.65
pentane to afford the final product as orange crystals (142 mg, (d, JPC = 13.6 Hz), 133.79, 132.72 (t, JPC = 8.8 Hz), 131.31 (d,
0.14 mmol, 94.5%). 1H NMR (CDCl3, 400 MHz): δ (ppm) = 7.79 JPC = 2.9 Hz), 129.63 (d, JPC = 14.6 Hz), 125.43 (dd, JPC = 181.9,
(m, 1H), 7.57 (m, 1H), 7.49 (m, 2H), 7.44 (t, JHH = 8.4 Hz, 2H), 19.4 Hz), 111.78, 105.58, 105.10, 104.56 (d, JPC = 2.9 Hz), 64.61
6.60 (dd, JHH = 8.2, 4.0 Hz, 4H), 5.66 (m, 1H), 3.95 (m, 2H), (d, JPC = 6.8 Hz), 55.95, 15.88 (d, JPC = 6.8 Hz). 31P NMR
3.84 (m, 2H), 3.64 (s, 12H), 3.50 (m, 12H), 3.35 (s, 6H), 2.86 (CDCl3, 162 MHz): δ (ppm) = 23.13 (d, JPP = 24.6 Hz), −21.86
(br, 2H), 2.30 (d, JHH = 13.2 Hz, 2H). 13C NMR (CDCl3, (d, JPP = 24.6 Hz). Anal. Calc. for C29H37F6O7P2SbNi: C, 40.79;
125 MHz): δ (ppm) = 161.53 (d, JPC = 2.9 Hz), 138.11 (dd, JPC
40.4, 12.7 Hz), 134.46 (d, JPC = 14.6 Hz), 133.78, 133.38 (t, JPC
=
=
H, 4.37. Found: C, 37.98; H, 4.39. *Similar elemental analysis
results were obtained from independently prepared samples. Trace
8.8 Hz), 131.42 (d, JPC = 3.9 Hz), 129.59 (d, JPC = 13.6 Hz), impurities that do not show by NMR spectroscopy may be present.
124.91 (dd, JPC = 184.9, 19.5 Hz), 111.86, 105.62, 105.13, 104.57
Metal binding studies with alkali ions
(d, JPC = 3.9 Hz), 71.84, 70.40, 69.37 (d, JPC = 6.9 Hz), 67.01 (d,
JPC = 5.8 Hz), 59.08, 55.98. 31P NMR (CDCl3, 162 MHz): The method of continuous variation (Job plot analysis) was
δ (ppm) = 23.58 (d, JPP = 25.9 Hz), −21.78 (d, JPP = 24.1 Hz). used to determine the binding stoichiometry of our nickel
Anal. Calc. for C35H49F6O11P2SbNi·(CH2Cl2)0.25: C, 41.37; H, complexes with alkali ions. To perform these experiments,
4.88. Found: C, 41.30; H, 4.95.
stock solutions of 4a (or 4b) (6 mM, 6 mL) and MBArF4 (6 mM,
Preparation of 6a. Inside the glovebox, a solution containing 15 equiv. Et2O to solubilize the salts, 6 mL, M = Li+, Na+, or K+)
5a (100 mg, 0.22 mmol, 1.0 equiv.) and AgSbF6 (75 mg, were prepared separately in CDCl3. Various amounts of each
0.22 mmol, 1.0 equiv.) in 10 mL of CH2Cl2 was stirred for stock solution were added to an NMR tube so that a total
10 min at RT. Solid [Ni(allyl)Cl]2 (30 mg, 0.11 mmol, 0.5 volume of 1 mL was obtained. Ten different NMR samples
equiv.) was added in small portions. The reaction mixture was were prepared, each containing a different ratio of 4a (or
stirred for an additional 3 h. The resulting red mixture was fil- 4b) : M. The samples were recorded at room temperature by 1H
tered through a pipet plug and then dried under vacuum to NMR spectroscopy. The hydrogen resonances centered at
give a dark red oil. Upon the addition of pentane and after stir- ∼5.7 ppm corresponding to the allyl group in 4a or 4b shift in
ring for ∼5 min, a yellow solid formed. The product was recrys- the presence of alkali ions. The changes in the 1H NMR
tallized by dissolving in CH2Cl2 and then layering with signals of Ha as a function of the mole fraction of the nickel
pentane to afford the final product as yellow crystals (160 mg, complexes are provided in Tables S1–S4.†
0.20 mmol, 91.6%). 1H NMR (CDCl3, 400 MHz): δ (ppm) = 7.85
(m, 1H), 7.69 (m, 1H), 7.58 (m, 3H), 7.14–7.08 (m, 3H), 6.98 (t,
Metal binding studies with zinc triflate
JHH = 7.2 Hz, 2H), 6.79 (m, 2H), 5.72 (m, 1H), 3.97 (m, 4H), The method of continuous variation (Job plot analysis) was
3.88 (s, 6H), 2.55 (br, 1H), 1.18 (t, JHH = 6.8 Hz, 6H). 13C NMR used to determine the binding stoichiometry of our nickel
(CDCl3, 100 MHz): δ (ppm) = 160.72 (d, JPC = 6.8 Hz), 135.08 complexes with zinc ions. To perform these experiments, stock
(dd, JPC = 36.0, 12.6 Hz), 134.64 (d, JPC = 4.9 Hz), 134.11, solutions of 4a (6 mM, 6 mL) and Zn(OTf)2 (6 mM, 6 mL) were
134.02, 133.88 (t, JPC = 7.8 Hz), 133.55 (dd, JPC = 5.8, 2.9 Hz), prepared separately in CD3CN. Various amounts of each stock
131.17 (d, JPC = 13.7 Hz), 127.85 (dd, JPC = 185.8, 17.6 Hz), solution were added to an NMR tube so that a total volume of
121.72 (d, JPC = 8.8 Hz), 115.68, 115.19, 114.56, 111.69 (d, JPC
=
1 mL was obtained. Ten different NMR samples were prepared,
3.9 Hz), 65.20 (d, JPC = 6.8 Hz), 56.05, 15.91 (d, JPC = 6.8 Hz). each containing a different ratio of 4a : Zn. The hydrogen reso-
31P NMR (CDCl3, 162 MHz): δ (ppm) = 22.91 (d, JPP = 21.1 Hz), nance centered at ∼5.7 ppm corresponding to the allyl group
1
−2.33 (d, JPP = 20.6 Hz). Anal. Calc. for C27H33F6O5P2SbNi: C, in 4a shift in the presence of zinc ions. The changes in the H
40.85; H, 4.19. Found: C, 40.60; H, 4.43.
NMR signals of Ha as a function of the mole fraction of 4a are
Preparation of 6b. Inside the glovebox, a solution containing provided in Tables S5.†
5b (100 mg, 0.19 mmol, 1.0 equiv.) and AgSbF6 (66 mg,
General ethylene polymerization
0.19 mmol, 1.0 equiv.) in 10 mL of CH2Cl2 was stirred for
10 min at RT. Solid [Ni(allyl)Cl]2 (26 mg, 0.10 mmol, 0.5 Inside the glovebox, the nickel complexes (10 μmol) and alkali
equiv.) was added in small portions. The reaction mixture was salts (10 μmol) were dissolved in 10 mL of toluene/DCM (8 : 2)
stirred for an additional 3 h. The resulting red mixture was fil- and stirred for 10 min. By visual inspection, the resulting
tered through a pipet plug and then dried under vacuum to nickel–alkali complexes appeared to be soluble in the reaction
give a dark red oil. Upon the addition of pentane and after stir- mixture. The mixture was sealed inside a vial using a rubber
ring for ∼5 min, an orange solid formed. The product was septum and brought outside of the glovebox. Under an atmo-
recrystallized by dissolving in CH2Cl2 and then layering with sphere of N2, the catalyst solution was loaded into a syringe.
pentane to afford the final product as orange crystals (139 mg, To prepare the polymerization reactor, 40 mL of dry toluene
This journal is © The Royal Society of Chemistry 2019
Dalton Trans., 2019, 48, 17887–17897 | 17895

View Article Online
Paper
Dalton Transactions
was added to an empty autoclave and preheated to the desired
temperature. The autoclave was purged with ethylene (20 psi)
for 1 min and then the catalyst solution was injected into the
autoclave via syringe. The reactor pressure was increased to
200 psi of ethylene and the contents were stirred vigorously for
1 h. To stop the polymerization, the autoclave was vented and
cooled in an ice bath. A solution of MeOH (100–200 mL) was
added to precipitate the polymer. The polymer was collected
by vacuum filtration, rinsed with MeOH, and then dried under
vacuum at 80 °C overnight. The reported yields are average
values of triplicate runs.
4 F. Yang and X. Li, J. Polym. Sci., Part A: Polym. Chem., 2017,
55, 2271–2280.
5 B. P. Carrow and K. Nozaki, Macromolecules, 2014, 47,
2541–2555.
6 L. Guo, W. Liu and C. Chen, Mater. Chem. Front., 2017, 1,
2487–2494.
7 C. Chen, Nat. Rev. Chem., 2018, 2, 6–14.
8 C. Tan and C. Chen, Angew. Chem., Int. Ed., 2019, 58, 7192–
7200.
9 N. E. Mitchell and B. K. Long, Polym. Int., 2019, 68, 14–26.
10 Z. Chen, M. Mesgar, P. S. White, O. Daugulis and
M. Brookhart, ACS Catal., 2015, 5, 631–636.
11 S. Dai, S. Zhou, W. Zhang and C. Chen, Macromolecules,
2016, 49, 8855–8862.
Ethylene and polar monomer copolymerization
Inside the glovebox, the nickel complexes (40 μmol) and metal
triflate (40 μmol) were dissolved in 10 mL of THF and stirred
for 10 min. The mixture was sealed inside a vial using a rubber
septum and brought outside of the drybox. To prepare the
polymerization reactor, 35–37 mL of dry THF was added to an
empty autoclave and preheated to the desired temperature.
The autoclave was purged with ethylene (20 psi) for 1 min and
then the polar comonomer was added first, followed by injec-
tion of the catalyst solution into the autoclave via syringe. The
reactor pressure was increased to 400 psi of ethylene and the
contents were stirred vigorously for 2 h. To stop the polymeriz-
ation, the autoclave was vented and cooled in an ice bath. A
solution of MeOH (100–200 mL) was added to precipitate the
polymer. The polymer was collected by vacuum filtration,
rinsed with MeOH, and then dried under vacuum at 80 °C
overnight. The poly(ethylene-co-PVE) products were also
washed with chloroform to remove any trace monomers or
poly(PVE) from the sample before characterization by GPC and
NMR spectroscopy. The reported yields are average values of
triplicate runs.
12 W. Zhang, P. M. Waddell, M. A. Tiedemann, C. E. Padilla,
J. Mei, L. Chen and B. P. Carrow, J. Am. Chem. Soc., 2018,
140, 8841–8850.
13 J. Gao, B. Yang and C. Chen, J. Catal., 2019, 369, 233–238.
14 C. J. Stephenson, J. P. McInnis, C. Chen, M. P. Weberski,
Jr., A. Motta, M. Delferro and T. J. Marks, ACS Catal., 2014,
4, 999–1003.
15 D. Zhang and C. Chen, Angew. Chem., Int. Ed., 2017, 56,
14672–14676.
16 M. R. Radlauer, A. K. Buckley, L. M. Henling and T. Agapie,
J. Am. Chem. Soc., 2013, 135, 3784–3787.
17 D. Takeuchi, Y. Chiba, S. Takano and K. Osakada, Angew.
Chem., Int. Ed., 2013, 52, 12536–12540.
18 P. D. Hustad, Science, 2009, 325, 704–707.
19 V. Blanco, D. A. Leigh and V. Marcos, Chem. Soc. Rev., 2015,
44, 5341–5370.
20 A. J. M. Miller, Dalton Trans., 2017, 46, 11987–12000.
21 T. Chantarojsiri, J. W. Ziller and J. Y. Yang, Chem. Sci.,
2018, 9, 2567–2574.
22 C. Yoo, H. M. Dodge and A. J. M. Miller, Chem. Commun.,
2019, 55, 5047–5059.
23 M. R. Kita and A. J. M. Miller, J. Am. Chem. Soc., 2014, 136,
14519–14529.
Conflicts of interest
24 J. M. Kaiser and B. K. Long, Coord. Chem. Rev., 2018, 372,
141–152.
There are no conflicts to declare.
25 C. Chen, ACS Catal., 2018, 8, 5506–5514.
26 J. Wei and P. L. Diaconescu, Acc. Chem. Res., 2019, 52, 415–
424.
Acknowledgements
27 L. Johnson, L. Wang, S. McLain, A. Bennett, K. Dobbs,
E. Hauptman, A. Ionkin, S. Ittel, K. Kunitsky, W. Marshall,
E. McCord, C. Radzewich, A. Rinehart, K. J. Sweetman,
Y. Wang, Z. Yin and M. Brookhart, in Beyond Metallocenes,
American Chemical Society, 2003, vol. 857, ch. 10, pp.
131–142.
We are grateful to the Welch Foundation (Grant No. E-1894 and
the National Science Foundation (Grant No. CHE-1750411) for
funding this research.
References
28 Q. Liu and R. F. Jordan, Organometallics, 2018, 37, 4664–
4674.
1 K. S. Whiteley, in Ullmann’s Encyclopedia of Industrial
Chemistry, Wiley-VCH Verlag GmbH & Co. KGaA, 2012, vol. 29 Q. Liu and R. F. Jordan, J. Am. Chem. Soc., 2019, 141, 6827–
29. 6831.
2 H. Makio, H. Terao, A. Iwashita and T. Fujita, Chem. Rev., 30 A. J. Smith, E. D. Kalkman, Z. W. Gilbert and I. A. Tonks,
2011, 111, 2363–2449. Organometallics, 2016, 35, 2429–2432.
3 Polyolefins: 50 Years After Ziegler and Natta II, 31 H.-C. Chiu, A. Koley, P. L. Dunn, R. J. Hue and I. A. Tonks,
ed. W. Kaminsky, Springer, Heidelberg, 2013.
Dalton Trans., 2017, 46, 5513–5517.
17896 | Dalton Trans., 2019, 48, 17887–17897
This journal is © The Royal Society of Chemistry 2019

View Article Online
Dalton Transactions
Paper
32 Z. Cai, D. Xiao and L. H. Do, J. Am. Chem. Soc., 2015, 137, 51 M. Chen and C. Chen, ACS Catal., 2017, 7, 1308–1312.
15501–15510.
52 J. Xia, Y. Zhang, J. Zhang and Z. Jian, Organometallics,
2019, 38, 1118–1126.
33 Z. Cai and L. H. Do, Organometallics, 2017, 36, 4691–4698.
34 T. V. Tran, Y. H. Nguyen and L. H. Do, Polym. Chem., 2019, 53 M. Chen and C. Chen, Angew. Chem., Int. Ed., 2018, 57,
10, 3718–3721.
3094–3098.
35 Z. Cai and L. H. Do, Organometallics, 2018, 37, 3874–3882.
36 A. M. Johnson, N. D. Contrella, J. R. Sampson, M. Zheng
and R. F. Jordan, Organometallics, 2017, 36, 4990–5002.
37 Z. J. A. Komon, X. Bu and G. C. Bazan, J. Am. Chem. Soc.,
2000, 122, 1830–1831.
38 P. Kenyon, M. Wörner and S. Mecking, J. Am. Chem. Soc.,
2018, 140, 6685–6689.
39 Y. Zhang, H. Mu, L. Pan, X. Wang and Y. Li, ACS Catal.,
2018, 8, 5963–5976.
40 N. D. Contrella, J. R. Sampson and R. F. Jordan,
Organometallics, 2014, 33, 3546–3555.
41 J. S. Renny, L. L. Tomasevich, E. H. Tallmadge and
D. B. Collum, Angew. Chem., Int. Ed., 2013, 52, 11998–12013.
42 S. S. Batsanov, Inorg. Mater., 2001, 37, 871–885.
54 W. Heyndrickx, G. Occhipinti, P. Bultinck and V. R. Jensen,
Organometallics, 2012, 31, 6022–6031.
55 S. Luo, J. Vela, G. R. Lief and R. F. Jordan, J. Am. Chem.
Soc., 2007, 129, 8946–8947.
56 J. Xia, Y. Zhang, X. Hu, X. Ma, L. Cui, J. Zhang and Z. Jian,
Polym. Chem., 2019, 10, 546–554.
57 Y. Ota, S. Ito, J.-i. Kuroda, Y. Okumura and K. Nozaki,
J. Am. Chem. Soc., 2014, 136, 11898–11901.
58 Y. Mitsushige, H. Yasuda, B. P. Carrow, S. Ito,
M. Kobayashi, T. Tayano, Y. Watanabe, Y. Okuno,
S. Hayashi, J. Kuroda, Y. Okumura and K. Nozaki, ACS
Macro Lett., 2018, 7, 305–311.
59 Y. Mitsushige, B. P. Carrow, S. Ito and K. Nozaki, Chem.
Sci., 2016, 7, 737–744.
43 D. Zhang, E. T. Nadres, M. Brookhart and O. Daugulis, 60 M. Li, X. Wang, Y. Luo and C. Chen, Angew. Chem., Int. Ed.,
Organometallics, 2013, 32, 5136–5143. 2017, 56, 11604–11609.
44 I. Göttker-Schnetmann, B. Korthals and S. Mecking, 61 C. Hong, X. Sui, Z. Li, W. Pang and M. Chen, Dalton Trans.,
J. Am. Chem. Soc., 2006, 128, 7708–7709. 2018, 47, 8264–8267.
45 A. Held and S. Mecking, Chem. – Eur. J., 2000, 6, 4623– 62 W. E. Piers and S. Collins, Compr. Organomet. Chem. III,
4629. 2007, 1, 141–165.
46 D. Zhang, D. Guironnet, I. Göttker-Schnetmann and 63 R. Shannon, Acta Crystallogr., Sect. A: Cryst. Phys., Diffr.,
S. Mecking, Organometallics, 2009, 28, 4072–4078. Theor. Gen. Crystallogr., 1976, 32, 751–767.
47 C. Boucher-Jacobs, B. Li, C. M. Schroeder and 64 I. Brown, Acta Crystallogr., Sect. B: Struct. Sci., 1988, 44,
D. Guironnet, Polym. Chem., 2019, 10, 1988–1992.
545–553.
48 D. Xiao and L. H. Do, Organometallics, 2018, 37, 254–260.
49 D. Xiao and L. H. Do, Organometallics, 2018, 37, 3079–3085.
65 Z. Cai, D. Xiao and L. H. Do, Comments Inorg. Chem., 2019,
39, 27–50.
50 W.-j. Tao, R. Nakano, S. Ito and K. Nozaki, Angew. Chem., 66 R. J. Hazlehurst, S. W. E. Hendriks, P. D. Boyle and
Int. Ed., 2016, 55, 2835–2839.
J. M. Blacquiere, ChemistrySelect, 2017, 2, 6732–6737.
This journal is © The Royal Society of Chemistry 2019
Dalton Trans., 2019, 48, 17887–17897 | 17897