of 9
Acrylate-Induced
β
H Elimination in Coordination Insertion
Copolymerizaton Catalyzed by Nickel
Shuoyan Xiong, Alexandria Hong, Priyabrata Ghana, Brad C. Bailey, Heather A. Spinney, Hannah Bailey,
Briana S. Henderson, Steve Marshall, and Theodor Agapie
*
Cite This:
J. Am. Chem. Soc.
2023, 145, 26463−26471
Read Online
ACCESS
Metrics & More
Article Recommendations
*
Supporting Information
ABSTRACT:
Polar monomer-induced
β
-H elimination is a key
elementary step in polar polyolefin synthesis by coordination
polymerization but remains underexplored. Herein, we show that a
bulky neutral Ni catalyst,
1
Ph
, is not only a high-performance
catalyst in ethylene/acrylate copolymerization (activity up to
37,000 kg/(mol
·
h) at 130
°
C in a batch reactor, mol % tBA
0.3) but also a suitable platform for investigation of acrylate-
induced
β
-H elimination.
4
Ph
d
t
Bu
, a novel Ni alkyl complex
generated after acrylate-induced
β
-H elimination and subsequent
acrylate insertion, was identified and characterized by crystallog-
raphy. A combination of catalysis and mechanistic studies reveals
effects of the acrylate monomer, bidentate ligand, and the labile
ligand (e.g., pyridine) on the kinetics of
β
-H elimination, the role of
β
-H elimination in copolymerization catalysis as a chain-
termination pathway, and its potential in controlling the polymer microstructure in polar polyolefin synthesis.
INTRODUCTION
Polyolefins account for over half of global plastic production.
1
8
Coordination copolymerization of nonpolar and polar mono-
mers is of high interest as it can provide value-added functional
polyolefins with diverse but controlled material properties and
potential degradability.
2
7,9
36
However, industrial implemen-
tation of this process is limited by low activity [typically <1000
kg/(mol
·
h)] and thermal stability (typically <100
°
C) of
reported catalysts (Figure 1a), as well as the low-molecular
weight (MW) of the resulting copolymers.
10,17,19,37
40
In ethylene/polar monomer copolymerization, the inter-
mediate generated after polar monomer insertion is typically the
resting state of catalysis due to coordination of the enchained
polar group to a vacant site on the transition metal and
challenges of olefin insertion into a secondary metal-alkyl bond
(Figure 1b).
37,41
49
β
-H elimination from this intermediate is
thus a key elementary step controlling catalyst performance and
polymer microstructure.
37,49
For ethylene and
α
-olefin polymer-
ization,
β
-H elimination has been investigated intensively, and
relevant intermediates including
β
-agostic species have been
identified and characterized.
44,50
61
On the other hand, limited
studies have been reported of intermediates relevant to polar
monomer-induced
β
-H elimination occurring during cataly-
sis.
46,47,62,63
The only catalyst species resulting from a
β
-H
elimination event characterized by crystallography is an
intermediate generated after acrylate insertion and chain
walking with a cationic Pd complex.
64
Prior studies were also
carried out on catalyst systems that exhibit no or low reactivity in
Received:
September 30, 2023
Revised:
October 25, 2023
Accepted:
October 27, 2023
Published:
November 22, 2023
Figure 1.
(a) Examples of catalysts for copolymerization of ethylene
and polar monomers; (b) mechanism of coordination copolymeriza-
tion, the resting state, and example of termination event involving
β
-H
elimination [R, R
: H, alkyl or polymer chain; L: olefin, or the labile
ligand (e.g., pyridine); and red circle: polar group].
Article
pubs.acs.org/JACS
© 2023 The Authors. Published by
American Chemical Society
26463
https://doi.org/10.1021/jacs.3c10800
J. Am. Chem. Soc.
2023, 145, 26463
26471
This article is licensed under CC-BY 4.0
ethylene/polar
monomer
copolymerization
[e.g.,
activity
<20
kg/(mol
·
h)].
Nickel
catalysts
have
been
a recent
focus
in copolymerization
involving
polar
monomers
due to nickel’s
relatively
low cost
and
promising
performance.
17
19
Despite
the
high
interest,
β
-H
elimination
at nickel
has
not
been
thoroughly
studied,
potentially
due
to the
lack
of a suitable
catalyst
system
that
undergoes
facile
β
-H
elimination
while
still being
productive
in
copolymerization.
Herein
we report
highly
active
Ni phosphine
phenoxide
catalysts
and
their
β
-H
elimination
behavior.
An
intermediate,
4
Ph
d
t
Bu
, generated
from
a putative
Ni-hydride,
was
characterized
by X-ray
crystallography.
These
results
provide
insights
into
how
catalyst
design
impacts
catalyst
activity,
copolymer
Mw,
and
chain-end
functionality
in polar
polyolefin
synthesis.
RESULTS
AND DISCUSSION
Catalyst
Design,
Preparation,
and Characterization.
The
nickel
complexes
generated
after
β
-H
elimination
of a
copolymer
chain
are
expected
to be highly
reactive
toward
further
insertion
or catalyst
decomposition
reactions.
Previous
mechanistic
studies
have
identified
several
catalyst
deactivation
pathways
starting
from
inter-
and
intramolecular
interactions
axial
to the
nickel
center.
45
47,65
67
To
stabilize
reactive
intermediates,
a catalyst
design
strategy
targeting
large
axial
shielding
was
chosen
(Figure
2a).
Increasing
proximal
steric
hindrance
has also
shown
promise
in improving
catalytic
activity
and
thermal
stability
in Ni catalysts
supported
by anionic
PO
ligands.
22,37,40,68
73
Two
neutral
Ni complexes,
1
Me
and
1
Ph
,
were
synthesized
as single-component
catalysts
for ethylene/
acrylate
copolymerization
and precursors
for the investigation
of
β
-H
elimination
(Figure
2a).
Structural
characterization
by
single-crystal
X-ray
diffraction
(scXRD),
in combination
with
topographical
steric
analysis
by Cavallo’s
SambVca
2.1,
74,75
confirms
that
axial
positions
of the Ni center
in both
complexes
are covered
from
both
the top and bottom
directions
(Figures
2b
and Figure
S1).
Notably,
the phenoxy
group
in
1
Ph
also
provides
steric
shielding
extending
to the side
of the O side,
while
the
methoxy
group
in
1
Me
provides
steric
shielding
only
on the P
side.
High-Temperature
Ethylene/Acrylate
Copolymeriza-
tion.
Both
1
Me
and
1
Ph
are highly
active
in ethylene/acrylate
copolymerization
reactions
(Table
1, entries
1
9).
The
bulkier
catalyst,
1
Ph
, shows
significantly
higher
activity
than
1
Me
but
produces
copolymers
with
lower
tBA
incorporation
(e.g.,
entry
2
vs 6), consistent
with
structure
performance
relationships
of Ni
catalysts
reported
previously.
22,34,40
The
optimized
reaction
temperature
for
1
Me
is 90
°
C (entry
1 vs 2, 3 vs 5) while
1
Ph
is
significantly
more
active
at 110
°
C than
that
at 90
°
C (entry
7 vs
9). The
latter
is also
in contrast
with
the optimized
reaction
temperatures
for other
reported
Pd and
Ni catalysts,
typically
ranging
between
50 and 90
°
C (Figure
S13).
22,40,49,70
An activity
of
33,000
kg/(mol
·
h)
was
achieved
at 110
°
C in a batch
reactor
(entry
10),
demonstrating
a
10 times
increase
compared
to the state-of-art
activity
of Ni phosphine
phenoxide
catalysts,
with
a similar
level
of acrylate
incorporation
(0.3%).
34
Overall,
1
Ph
features
significantly
improved
activity
and
thermal
stability
compared
to reported
catalysts
(Supporting
Informa-
tion
Section
S6, Figures
S12
and
S13).
At 130
°
C,
1
Ph
shows
an
activity
of
37,000
kg/(mol
·
h)
in a batch
reactor
(entry
8),
albeit
with
low
tBA
incorporation
(0.3
mol
%). To the best
of
our
knowledge,
this
is the first
reported
example
of ethylene/
acrylate
coordination
copolymerization
at >110
°
C.
These
results
show
promise
for potential
practical
applications
as low
catalyst
activity,
low
catalyst
thermal
stability,
and
low
copolymer
MW
are
three
major
limitations
to industrial
implementation.
17,76
Identification
of
β
-H Elimination
and Subsequent
Acrylate
Insertion.
With
these
two highly
active
and thermally
robust
catalysts,
tBA
insertion
and
subsequent
reactions
were
investigated.
Treatment
of
1
Ph
with
excess
tBA
(ca.
15 equiv)
results
in a color
change
from
yellow
to red.
Monitoring
of the
1
H and
31
P{
1
H} NMR
spectra
confirmed
the consumption
of
1
Ph
. One
broad
resonance
appears
in
31
P{
1
H} NMR
spectra
over
time
(Figures
S14
and
S17),
and
four
new
resonances
were
observed
in the
1
H NMR
spectra
in a
1:9:9:9
ratio
(Figures
S15,
S16
and
S18):
one
new
doublet
in the olefinic
region
(
δ
5.8 ppm),
one
in the upfield
region
corresponding
to a Me
3
Si-
containing
species
(
δ
0 ppm),
and
two
t
BuO-
resonances
(
δ
1.2
1.5
ppm).
These
results
suggest
reactivity
with
two
acrylates
and
the
generation
of a new
olefinic
species.
A
combination
of
1
H
1
H COSY
NMR
and gas chromatography
mass
spectrometry
analysis
revealed
the identity
of the internal
olefin
as
d
t
Bu
IO
Si
(Figure
3a, Figures
S18
S22).
Furthermore,
1
H,
31
P{
1
H},
and
1
H
1
H COSY
NMR
analysis
suggests
the
identity
of the
other
species
as
4
Ph
d
t
Bu
, which
is most
likely
generated
via tBA
insertion
into
a Ni hydride
complex
(
3
Ph
,
Figures
3a, S23
and
S24).
Structural
Characterization
of 4
Ph
d
t
Bu
.
Despite
numerous
attempts,
bulk
isolation
of pure
4
Ph
d
t
Bu
as a solid
was
not
successful.
The
complex
decomposes
quickly
at room
temper-
ature,
both
under
vacuum
and
in solution.
Nevertheless,
single
crystals
of
4
Ph
d
t
Bu
were
obtained
from
tBA
insertion
experiments
with
1
Ph
in the presence
of tBA
and
excess
pyridine
(ca.
five
equiv),
and the scXRD
structure
of
4
Ph
d
t
Bu
is shown
in Figure
3b.
To the best
of our
knowledge,
this
is the first
crystallographic
Figure
2.
(a) Depiction
of the steric
profile
of the catalyst
system
and
two
new
catalysts
in this work.
(b) Solid-state
structures
of
1
Me
and
1
Ph
with
thermal
ellipsoid
representations
emphasizing
the
metal
coordination
sphere
and
the steric
profile
of groups
along
the axial
positions.
Journal
of the American
Chemical
Society
pubs.acs.org/JACS
Article
https://doi.org/10.1021/jacs.3c10800
J. Am. Chem.
Soc.
2023,
145, 26463
26471
26464
characterization
of an intermediate
generated
after
polar-
monomer-induced
β
-H
elimination
relevant
to Ni-catalyzed
polar
polyolefin
synthesis.
The
Ni(1)
C(1)
distance
in
4
Ph
d
t
Bu
[2.030(5)
Å] is longer
than
that
in
1
Ph
[1.949(2)
Å] or in
reported
Ni complexes
resulting
from
tBA
insertion
into
a metal
alkyl
moiety
[1.972(8)
2.003(8)
Å].
37,49
This
comparison
suggests
a weakened
Ni
alkyl
bond
in
4
Ph
d
t
Bu
, potentially
due to
steric
repulsion
induced
by the bulky
phenoxy
and
t
Bu groups.
These
steric
interactions
may
also
promote
facile
β
-H
elimination
in crowded
intermediate
2
Ph
d
t
Bu
.
Kinetic
Studies
of Acrylate-Induced
β
-H Elimination.
Identification
of the
internal
olefin
d
t
Bu
IO
Si
and
4
Ph
d
t
Bu
, in
combination
with
in situ
1
H and
31
P{
1
H} NMR
monitoring,
established
a kinetic
profile
of the reactions
with
tBA
(Figure
3c).
The
concentration
of
4
Ph
d
t
Bu
is roughly
equal
to that
of
d
t
Bu
IO
Si
during
the course
of the reaction
and
the two
putative
intermediates,
2
Ph
d
t
Bu
and
3
Ph
, were
not observed,
indicating
that
acrylate
insertion
(step
1) is rate
determining
in this reaction.
In
contrast,
2
Ph
nBu
and
2
Ph
Me
were
observed
as the intermediates
in analogous
reactions
with
n
butyl
acrylate
(
n
BA)
and
methyl
acrylate
(MA),
indicating
that
these
two
acrylates
feature
faster
rates
of initial
insertion
(step
1) and
a lower
tendency
for
β
-H
elimination
after
acrylate
insertion
compared
to tBA
(Figure
3c
e).
Consequently,
acrylate
insertion
(step
1) and
β
-H
elimination
(step
2) are
differentiable
in the
kinetic
profile
(Figure
3d,e),
allowing
direct
quantitative
kinetic
studies
of
β
-H
elimination
to elucidate
the
mechanism.
3
Ph
was
still
not
observed,
suggesting
that
acrylate
reinsertion
after
β
-H
elimination
(step
3) is faster
than
β
-H
elimination
(step
2) in
these
cases.
Next,
quantitative
kinetic
studies
of
β
-H
elimination
were
performed
with
n
BA
and
MA
under
otherwise
identical
conditions.
After
full consumption
of
1
Ph
, decay
of
2
Ph
nBu
or
2
Ph
Me
is representative
of
β
-H
elimination
(step
2). Notably,
linear
relationships
were
observed
in the log plot
for the decay
of
relative
concentrations
of
2
Ph
nBu
or
2
Ph
Me
over
time
(Figure
3f
g), consistent
with
pseudo
-first-order
kinetics.
Comparing
n
BA
with
MA,
β
-H
elimination
induced
by the
former,
bulkier
monomer,
features
a >50%
faster
rate
constant
(
k
2
). This
scenario
suggests
that
β
-H
elimination
(step
2) is faster
from
the
insertion
product
derived
from
the larger
monomer
than
from
the smaller
one.
In contrast,
the initial
insertion
(step
1) is fastest
with
MA,
the smallest
monomer
among
MA,
n
BA,
and
tBA,
as
evidenced
by the faster
decrease
of
1
Ph
in the reaction
with
MA
compared
to those
with
n
BA and tBA.
Consequently,
reaction
of
1
Ph
and
tBA,
the bulkiest
monomer
among
these
three,
features
the slowest
rate
of acrylate
insertion
(step
1) and the fastest
rate
of subsequent
β
-H
elimination
(step
2), leading
to
2
Ph
d
t
Bu
and
3
Ph
not being
observed
during
the tBA
reaction.
This
generation
of
4
Ph
d
t
Bu
free
of intermediates
is critical
for its isolation.
Furthermore,
the
rate
constant
of acrylate-induced
β
-H
elimination
(step
2) was
measured
with
1
Ph
under
varying
acrylate
and pyridine
concentrations
(15 or 50 equiv
of MA,
0
5
equiv
of pyridine)
to investigate
the
influence
of acrylate
monomer
and
labile
ligand,
pyridine.
MA
instead
of BA was
selected
to ensure
better
differentiation
of acrylate
insertion
(step
1) and
β
-H
elimination
(step
2). A linear
relationship
was
observed
between
pseudo
-first-order
rate
constant
of MA
insertion
(
k
1
, step
1) and
the reverse
of pyridine
concentration
(1/[py]),
consistent
with
the reported
mechanism
for acrylate
insertion.
37
For
acrylate-induced
β
-H
elimination
(step
2), a
near
linear
correlation
was
observed
between
the inverse
of a
pseudo
-first-order
rate
constant
(1/
k
2
) and
the
pyridine
concentration
([py],
Figure
3e).
These
observations
are
consistent
with
pyridine
being
involved
in the
β
-H
elimination
step
in this
model
system,
likely
through
a dissociative
process
(step
2, Figure
3a; Detailed
discussion:
Supporting
Information
section
S9).
We have
previously
reported
that
the labile
ligand
(i.e.,
pyridine)
also
affects
the rate
of catalyst
initiation
and chain
propagation
during
polymerization
catalysis
by neutral
Ni
catalysts,
suggesting
that
either
the resting
state
of this catalysis
is
a pyridine-coordinated
species
or pyridine
binding
to metal
is
competitive
with
olefin
coordination.
37,49,77
Together,
these
results
highlight
that
the effect
of the labile
ligand
(i.e.,
pyridine)
is a critical
consideration
in designing
neutral
Ni catalysts
for
polar
polyolefin
synthesis.
Notably,
k
2
does
not depend
on [MA]
(Table
S4, Figures
S49
and
S50),
indicating
that
acrylate
is not
involved
in this
portion
of the
mechanism.
Therefore,
a
mechanism
involving
β
-H-transfer
to olefin
is inconsistent.
78
80
EffectoftheCatalyst
Structure
onAcrylate-Induced
β
-
H Elimination
and Competing
Catalyst
Deactivation.
Effects
of the
catalyst
structure
on acrylate-induced
β
-H
Table
1. Ethylene/Acrylate
Copolymerization
Results
entry
a
catalyst
T
/
°
C
[tBA]/M
activity
b
M
w
/10
3
Đ
% mol
tBA
T
m
/
°
C
1
1
Me
70
0.05
750
120.0
2.6
2.3
113
2
1
Me
90
0.05
1550
73.3
2.4
1.5
115
3
1
Me
90
0.10
720
47.0
2.2
3.4
106
4
1
Me
90
0.15
410
35.4
2.3
4.8
99
5
1
Me
110
0.10
440
17.8
2.4
2.9
107
6
1
Ph
90
0.05
21,000
38.5
2.3
0.3
126
7
1
Ph
90
0.10
9700
32.9
2.4
0.7
123
8
1
Ph
90
0.15
5700
30.0
2.3
1.0
120
9
1
Ph
110
0.10
17,800
26.0
2.4
0.7
123
10
c
1
Ph
110
0.054
33,000
28.4
2.2
0.3
127
11
c
1
Ph
110
0.108
14,000
24.9
2.2
0.6
125
12
c
1
Ph
130
0.054
37,000
15.6
2.6
0.3
127
a
Entries
1
9:
copolymerization
in high-throughput
parallel
pressure
reactors.
[Ni]
= 0.05
mM,
ethylene
pressure
= 400
psi, toluene
solvent,
V
= 5
mL.
Polymerization
was
stopped
after
consuming
a set amount
of ethylene,
and
each
entry
represents
multiple
replicated
runs.
See
Supporting
Information
Section
S3 for detailed
procedures
and
Table
S2 for original
catalytic
runs.
b
kg/(mol
·
h).
c
Entries
10
12:
copolymerization
in a batch
reactor:
V (solvent)
= 550
mL,
[Ni]
= 0.043
mM,
ethylene
pressure
= 430
psi,
t
= 3.5 min
(entry
6), 6.5 min
(entry
7), or 3 min
(entry
8), ethylene
consumption
= 40 g. See
Supporting
Information
Section
S3 for detailed
procedures.
Journal
of the American
Chemical
Society
pubs.acs.org/JACS
Article
https://doi.org/10.1021/jacs.3c10800
J. Am. Chem.
Soc.
2023,
145, 26463
26471
26465
elimination
and
subsequent
reactions
were
investigated
by
comparing
the reaction
of tBA
with
1
Ph
and with
1
Me
. Initial
tBA
insertion
for the latter
is faster
by almost
an order
of magnitude
[0.042(1)
min
1
vs 0.0061(1)
min
1
, Figures
S47
vs S48],
while
Figure
3.
(a) Generation
of the internal
olefin
(
IO
Si
) and the acrylate-inserted
species
via a three-step
pathway.
(b) Solid-state
structure
of
4
Ph
d
t
Bu
. (c
e) Kinetic
profiles
of reaction
of tBA,
nBA,
and MA
with
1
Ph
. (f,g)
pseudo
-first-order
kinetics
of acrylate-induced
β
-H
elimination
revealed
by log plots
for the decay
of
2
Ph
R
. Note:
Colors
in (c
h)
correspond
to coloring
of structures
in (a):
purple
for the starting
catalyst
complex,
yellow
for the
acrylate-inserted
Ni
alkyl
complex,
red for the
β
-H
elimination
product,
and green
for the acrylate-inserted
Ni
H
complex.
Condition
for (c
g):
[Ni]
= [
1
Ph
]
t
=0
= 0.0118
M, [py]
= 0, [acrylate]
= 0.177
M, solvent:
C
6
D
6
,
V
(total)
= 0.5 mL,
T
= 25
°
C. (h) Plot
of the reverse
of first-order
rate constants
of
β
-H
elimination
[1/
k
(step
2), or 1/
k
2
] vs [py]/[Ni
t
=0
] for
1
Ph
. Condition
for h): [Ni]
= 0.0118
M, [py]
= 0039
0.059
M, [MA]
= 0.59
M, solvent:
C
6
D
6
,
V
(total)
= 0.5 mL,
T
= 25
°
C. See Supporting
Information
Section
S7 and
S8 for details.
Journal
of the American
Chemical
Society
pubs.acs.org/JACS
Article
https://doi.org/10.1021/jacs.3c10800
J. Am. Chem.
Soc.
2023,
145, 26463
26471
26466