of 7
Catalyst
Editing
via Post-Synthetic
Functionalization
by
Phosphonium
Generation
and Anion
Exchange
for Nickel-Catalyzed
Ethylene/Acrylate
Copolymerization
Priyabrata
Ghana,
Shuoyan
Xiong,
Adjeoda
Tekpor,
Brad
C. Bailey,
Heather
A. Spinney,
Briana
S. Henderson,
and Theodor
Agapie
*
Cite
This:
https://doi.org/10.1021/jacs.4c03416
Read
Online
ACCESS
Metrics
& More
Article
Recommendations
*
Supporting
Information
ABSTRACT:
Rapid,
efficient
development
of homogeneous
catalysts
featuring
desired
performance
is critical
to numerous
catalytic
transformations
but
remains
a key
challenge.
Typically,
this
task
relies
heavily
on
ligand
design
that
is often
based
on
trial
and
error.
Herein,
we
demonstrate
a “catalyst
editing”
strategy
in Ni-catalyzed
ethylene/acrylate
copolymerization.
Specifically,
alkylation
of a
pendant
phosphine
followed
by
anion
exchange
provides
a high
yield
strategy
for
a large
number
of
cationic
Ni
phosphonium
catalysts
with
varying
electronic
and
steric
profiles.
These
catalysts
are
highly
active
in ethylene/acrylate
copolymerization,
and
their
behaviors
are
correlated
with
the
electrophile
and
the
anion
used
in late-stage
functionalization.
P
olyolefin
synthesis
has
been
driven
by
advances
in catalyst
development.
1
7
For
homogeneous
catalysts,
this
task
relies
heavily
on
ligand
design,
which
has
often
involved
leaps
in discovery
of privileged
ligands
followed
by
extensive
steric
and
electronic
tuning
of
the
promising
motif.
Beyond
electronic
and
steric
tuning,
strategies
focused
on
chain-
shuttling,
metal
metal
cooperativity,
metal-substrate
effect,
and
redox
control
have
also
been
reported.
8
19
Nonetheless,
on-demand
catalyst
discovery
or
optimization
remains
limited
because
of
challenges
in
predicting
catalyst
performance
a
priori
given
the
mechanistic
complexity
of
copolymerization.
Further,
the
tunability
of the
catalyst
design
is limited
by
the
synthetic
methods
accessible
for
introducing
desired
sub-
stituents.
1
Elaborated
ligand
structures,
in many
cases,
result
in
an
increased
number
of
synthetic
steps
and
cost
of
preparation.
20
22
Redox
changes
or
cation
addition
are
powerful
methods
to switch
catalyst
performance
but
typically
only
between
a limited
number
of catalyst
states.
10,23
25
Ethylene/polar
olefin
copolymerization
has
been
pursued
to
improve
polyolefin
material
properties
and
introduce
potential
degradability,
but
precise
catalytic
control
remains
a challenge
(Figure
1a,b).
3
5,20,26
50
Ni
catalysts
have
been
a recent
focus
because
of a variety
of advantages,
including
lower
cost,
polar
group
tolerance,
improved
catalytic
performance,
and
higher
temperature
stability.
4,19,20,34,35,44,45,51
62
Herein,
we
report
a
versatile
catalyst-editing
strategy
on
the
basis
of postsynthetic
functionalization
of
a phosphine
moiety
followed
by
anion
exchange
in a bisphosphinephenoxide-based
Ni
catalyst
system
active
for
ethylene/acrylate
copolymerization
(Figure
1c).
20,55
Our
approach
was
based
on
phosphonium
synthesis
via
the
reaction
of triarylphosphine
with
alkyl
halides.
The
precursor,
the
bisphosphine
phenol
(
POPH
),
can
be
prepared
in
four
steps
from
commercial
compounds.
20
Reaction
of
POPH
with
methyl
iodide
in benzene
followed
by
anion
exchange
with
a
sodium
salt,
NaBAr
F
24
, affords
quantitative
formation
of
Received:
March
10,
2024
Revised:
June
10,
2024
Accepted:
June
24,
2024
Figure
1.
(a)
Depiction
of
inhibitory
effects
of
polar
monomers
in
metal-catalyzed
ethylene/polar
olefin
copolymerization.
(b)
Examples
of
Pd
and
Ni
catalysts
suitable
for
polar
polyolefin
synthesis.
(c)
Depiction
of the
catalyst
editing
strategy
in this
work
and
subsequent
metalation.
Communication
pubs.acs.org/JACS
© XXXX
The Authors.
Published
by
American
Chemical
Society
A
https://doi.org/10.1021/jacs.4c03416
J. Am. Chem.
Soc.
XXXX,
XXX,
XXX
XXX
This article is licensed under CC-BY-NC-ND 4.0
Downloaded via CALIFORNIA INST OF TECHNOLOGY on July 8, 2024 at 19:04:21 (UTC).
See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.
corresponding
phosphino(phosphonium)-phenol,
1a
(Figure
2a,
NMR
yield
> 99%,
isolated
yield:
97%),
as indicated
by
the
31
P{
1
H}
NMR
spectrum
featuring
two
characteristic
signals
for
phosphine
(
62.3
ppm)
and
phosphonium
(8.4
ppm).
63,64
Notably,
formation
of
bisphosphonium
species
was
not
observed,
potentially
because
of
the
low
solubility
of
monophosphonium
species
in nonpolar
solvents
that
hinders
further
reaction
upon
precipitation.
Several
alkyl
bromides
with
longer
alkyl
chains,
larger
steric
bulkiness,
or
electron-
withdrawing
substituents
were
also
screened,
which
led
to
the
near-quantitative
generation
of
phosphino-
(phosphonium)phenols
1b
1g
(Figure
2b).
For
Ni
phenoxide
catalysts,
tuning
the
steric
and
electronic
factors
on
the
“O”
side
has
shown
promise
for
increasing
catalytic
activity
and
stability,
though
those
ligands
required
multiple
additional
steps
in the
early
stages
of
ligand
synthesis.
20,53
The
current
approach
based
on
postsynthetic
functionalization
is a more
facile
strategy
to
increase
structural
diversity.
Given
the
commercial
availability
of
varieties
of
alkyl
halides
with
a
large
window
of steric
and
electronic
profiles,
PO
ligands
with
a wide
range
of steric
and
electronic
characteristics
on
the
O
side
are
readily
accessible.
In
addition,
customization
of
the
anion
is also
feasible
by
varying
the
Na
salt
employed.
For
example,
anion
exchange
with
NaOTf
after
the
addition
of methyl
iodide
generates
the
corresponding
phosphonium
triflate,
1a
(Figure
2a,b).
We
envisaged
that
the
difference
in anion
may
lead
to difference
in
cation
anion
interaction
and
consequently
varying
catalytic
behavior.
Metalation
of the
cationic
phosphino(phosphonium)-phenol
proligands
1a
g
and
1a
with
one
equivalent
of
Ni-
(py)
2
(CH
2
SiMe
3
)
2
in
benzene
or
THF
afforded
the
corre-
sponding
Ni
complexes
2a
g
and
2a
as orange
solids
(Figure
1a).
The
31
P{
1
H}
NMR
spectra
of these
complexes
display
two
characteristic
doublet
signals
for
two
inequivalent
phosphorus
centers
with
an
average
4
J
P,P
coupling
constant
of
13
Hz.
The
nickel-bound
phosphines
appear
at around
7
ppm,
which
is
consistent
with
reported
Ni
phosphine
phenoxide
com-
plexes.
20,34,53
The
phosphonium
groups
appear
in
the
same
range
(8
15
ppm)
as
observed
in
the
corresponding
proligands.
Characterization
of
2g
and
2a
by
single-crystal
X-ray
diffraction
(scXRD)
further
confirms
their
identities
as
cationic
Ni
phosphino(phosphonium)-phenoxide
complexes
featuring
a square
planar
geometry
and
a corresponding
outer-
sphere
anion
of BAr
F
24
or
OTf
(Figure
2c).
Next,
topographical
analysis
by
Cavallo’s
SambVca
2.1
program
was
conducted
with
2g
,
2a
,
and
POP-Ni
,
65,66
an
analogous
neutral
Ni
catalyst
derived
from
the
bisphosphine
phenol
proligand
(
1a
)
that
we
reported
previously.
20,54,55
Compared
with
POP-Ni
where
substituents
on
the
free
Figure
2.
(a)
Synthesis
route
for
phosphino-(phosphonium)phenol
proligands
1a
1g
and
1a
,
as well
as corresponding
nickel
catalysts
2a
2g
and
2a
(Ar
= dimethoxyphenyl).
(b)
Alkyl
halides
examined
in this
work,
as well
as corresponding
NMR
yields
(in
parentheses)
and
isolated
yields.
(c)
Solid-state
structures
of
2a
and
2g
(anion
and
hydrogen
atoms
omitted
for
for
clarity;
cyan:
Ni,
pink:
P,
red:
O,
blue:
N,
khaki:
Si).
(d)
Topographical
maps
and
corresponding
%
V
bur
of
POP-Ni
,
2a
,
and
2g
(see
the
Supporting
Information
section
6 for
more
details).
Journal
of the American
Chemical
Society
pubs.acs.org/JACS
Communication
https://doi.org/10.1021/jacs.4c03416
J. Am. Chem.
Soc.
XXXX,
XXX,
XXX
XXX
B
phosphine
provide
proximal
steric
hindrance
from
only
one
direction
on
the
“O”
side,
phosphonium
substituents
in both
2g
and
2a
occupy
both
bottom
right
and
top
right
spaces
of
the
Ni
coordination
sphere
(Figure
2d).
In
addition,
the
percentage
of buried
volume
(%
V
bur
) within
3.5
Å of the
Ni
center
for
2a
(49.0)
is slightly
higher
than
that
for
POP-Ni
(48.3),
while
that
for
2g
(51.0)
is significantly
higher.
This
trend
is consistent
with
the
relative
bulkiness
of
phosphine/
phosphonium
substituents
in these
three
complexes.
Overall,
the
topographical
analysis
presented
above
confirms
that
the
additional
alkyl
group
on
the
phosphonium
atom
can
impact
the
steric
profile
around
the
Ni
center.
The
nearly
quantitative
conversion
of
POPH
in
post-
synthetic
functionalization
and
of
1a
1g
in
metalation
prompted
investigation
of
one-pot
synthesis
of
Ni
phosphi-
no-(phosphonium)-phenoxide
complexes.
Indeed,
mixing
POPH
,
4-methylbenzyl
bromide,
and
NaBAr
F
24
in 1:1:1
ratio
followed
by
addition
of
one
equivalent
of
Ni-
(py)
2
(CH
2
SiMe
3
)
2
led
to
the
quantitative
formation
of
2e
.
Overall,
the
facile
and
selective
formation
of
cationic
Ni
complexes
combined
with
the
high
tunability
and
accessibility
of the
alkyl
bromide
and
potentially
the
anion
makes
precise
structural
editing
of this
type
of complexes
possible.
Catalytic
performance
of
these
cationic
complexes
was
examined
in an
ethylene/acrylate
copolymerization.
At
90
°
C
(Table
1, entries
1
17),
they
show
high
activity
[>100
kg/
(mol
·
h)]
in
producing
copolymers
with
moderate
tert
-butyl
alcohol
(tBA)
incorporation
(0.25%
1.13%).
Among
these
catalysts,
2a
shows
the
highest
activity
[>1000
kg/(mol
·
h)]
and
produces
copolymers
with
the
highest
M
w
(up
to
85
000),
which
is potentially
related
to
the
small,
electron-donating
methyl
phosphonium
substituent.
Notably,
2a
shows
significantly
higher
activity
[1130
kg/(mol
·
h)]
compared
to
the
neutral
analogue,
POP-Ni
[660
kg/(mol
·
h)]
and
produces
copolymers
with
higher
M
w
(84.7
×
10
3
vs
55.1
×
10
3
) and
lower
acrylate
incorporation
(
0.6%
vs
2.2%)
under
identical
polymerization
conditions
(Entry
1 vs
27),
though
these
phenomena
may
be
interrelated
(i.e.,
the
higher
acrylate
incorporation
may
lead
to
lower
activity).
Nevertheless,
these
differences
demonstrate
significant
impacts
of phosphonium
vs phosphine
on
catalyst
performance.
While
producing
copolymers
with
similar
MW
and
acrylate
incorporation,
2b
and
2c
show
activity
lower
than
that
of
2a
(Entry
1 vs
4 vs
6).
Replacing
the
methyl
group
(
2a
)
by
a
pendant
ester
group
(
2d
)
leads
to
higher
PDI
(Entry
1 vs
7
and
2.4
vs
3.2).
Comparing
2a
and
2e
(entry
1 vs
10),
the
latter
features
a larger
proximal
benzyl
group
on
the
O side
and
produces
copolymers
with
significantly
lower
M
w
(
85
000
vs
39
000).
One
rationale
for
this
phenomenon
is that
bulky
substituents
on
the
O side
may
promote
β
-H
elimination
after
Table
1. Ethylene/Acrylate
Copolymerization
Results
a
entry
catalyst
[tBA]/M
T
/
°
C
act.
b
M
w
/10
3
PDI
% mol
tBA
T
m
/
°
C
1
2a
0.05
90
1130
84.7
2.5
0.59
123
2
2a
0.1
90
550
69.7
2.3
1.13
117
3
2b
0.025
90
1020
91.5
2.4
0.26
126
4
2b
0.05
90
470
85.0
2.7
0.60
122
5
2c
0.025
90
500
92.4
2.3
0.27
127
6
2c
0.05
90
240
80.7
2.3
0.55
122
7
2d
0.05
90
510
74.2
3.2
0.52
122
8
2d
0.1
90
220
59.5
2.9
1.03
117
9
2e
0.025
90
1050
42.0
2.6
0.25
124
10
2e
0.05
90
480
38.7
2.4
0.53
121
11
2e
0.1
90
240
37.2
2.4
0.99
117
12
2f
0.05
90
560
60.5
2.4
0.43
123
13
2f
0.1
90
300
52.1
2.2
0.83
120
14
2g
0.05
90
570
29.9
2.4
0.56
123
15
2g
0.1
90
280
28.7
2.4
0.84
119
16
c
2a
0.025
90
210
8.2
2.8
0.35
121
17
c
2a
0.05
90
120
9.0
4.1
0.76
117
18
2b
0.025
110
2190
41.2
2.1
0.29
124
19
2b
0.05
110
950
40.6
2.3
0.65
122
20
2c
0.025
110
1200
44.6
2.3
0.27
124
21
2c
0.05
110
530
41.0
2.2
0.52
122
22
2d
0.025
110
2000
40.3
2.7
0.40
124
23
2d
0.05
110
750
39.3
3.3
0.43
121
24
2e
0.05
110
790
19.8
2.0
0.6
120
25
2e
0.05
130
310
9.9
2.0
0.6
116
26
e
2e
d
0.054
110
1060
22.1
2.3
0.50
121
27
f
POP-Ni
g
0.05
90
660
55.1
2.2
2.11
111
28
POP-Ni
g
0.05
100
300
40.4
2.2
2.32
110
a
Unless
specified,
V
= 5 mL,
[Ni]
= 0.05
mM,
ethylene
pressure
= 400
psi,
and
solvent
is toluene.
Polymerization
was
stopped
after
consuming
a
set
amount
of ethylene
or at 1 h, whichever
came
first
(reaction
time:
13
60
min).
Each
entry
represents
multiple
replicated
runs.
In each
run,
30
121
mg
of copolymer
was
produced.
See
the
Supporting
Information
section
S6
for
detailed
procedures
and
Table
S3
for
original
catalytic
runs.
b
Activity
in kg/(mol
·
h).
c
Both
0.25
mL
of THF
and
4.75
mL
of toluene
were
added
as the
solvent.
d
A mixture
of
POPH
(22.56
μ
mol),
4-
methylbenzyl
bromide,
NaBAr
F
24
, and
py
2
Ni(CH
2
SiMe
3
)
2
in a 1:1:1:1
ratio
was
used
as
in situ
generated
2e
.
e
V
= 550
mL,
[Ni]
= 0.041
mM,
ethylene
pressure
= 430
psi,
[tBA]
= 0.054
M,
t
= 60
min.
f
Reported
in ref
31
as entry
7.
g
POP-Ni
:
a catalyst
first
reported
in ref
31,
prepared
from
ligand
POPH
and
py
2
Ni(CH
2
SiMe
3
)
2
.
Journal
of the American
Chemical
Society
pubs.acs.org/JACS
Communication
https://doi.org/10.1021/jacs.4c03416
J. Am. Chem.
Soc.
XXXX,
XXX,
XXX
XXX
C
monomer
insertion
and,
thus,
lead
to
decreases
in
MW.
Relatedly,
we
have
previously
reported
a P,O
Ni
catalyst
with
a rigid
aryl
group
on
the
O side,
which
produces
copolymers
with
a significantly
lower
MW
relative
to less
rigid
analogues,
and
a similar
case
is observed
with
cationic
complexes
herein.
20
Notably,
this
previous
example
requires
an
10-step
synthesis,
while
similar
steric
control
can
be
achieved
here
with
a
catalyst-editing
strategy
with
only
one
additional
step,
a
notable
step
economy.
Further
increasing
the
steric
hindrance
by
introducing
two
tert
-butyl
groups
on
the
phenyl
moiety
(
2e
vs
2g
)
leads
to
even
lower
M
w
(39
000
vs
30
000),
thereby
demonstrating
the
broad
range
of MW
tuning
with
this
facile
catalyst-modification
strategy.
Introduction
of
electron-withdrawing
substituents,
such
as
perfluoro-alkyl
or
aryl
group,
is known
to enhance
activity
or
polymer
MW
for
neutral
Ni
and
Pd
catalysts.
34,53,67,68
For
this
catalyst
system,
replacing
the
methyl
by
trifluoromethyl
(
2e
vs
2f
),
indeed,
leads
to a moderate
increase
of catalyst
activity
(by
20%)
and
significant
increase
of copolymer
MW
(by
50%),
albeit
with
a decrease
in
acrylate
incorporation.
These
differences
indicate
that
electronic
tuning
available
through
this
approach
also
impacts
catalyst
behavior.
Although
direct
interaction
between
the
cationic
catalyst
and
the
outer-sphere
anion
was
not
observed
in
solid-state
structures
of
2g
and
2a
,
the
cation
species
in
2a
and
2a
that
feature
the
identical
structure
exhibit
slightly
different
chemical
shifts
in
1
H,
13
C{
1
H},
and
31
P{
1
H}
NMR
spectra
(Figures
S1
S3
vs
S22
S24).
This
potentially
implies
differences
in the
cation
anion
interaction.
Notably,
switching
the
anion
leads
to
a dramatic
difference
in catalyst
perform-
ance.
Compared
with
2a
,
catalyst
2a
featuring
weakly
coordinating
OTf
instead
of
BAr
F
24
as
the
anion
is much
less
active
and
produces
copolymers
with
significantly
lower
MW
and
higher
PDI
(Entry
1 vs 17).
One
rationale
is that
the
OTf
anion
is more
coordinating
than
BAr
F
24
and
competes
with
olefins
for
binding
to
the
nickel
center
and,
thus,
slows
catalysis
and
promotes
chain
transfer.
Further
screening
of
other
anions,
as well
as mechanistic
elucidation
of this
effect,
is
currently
ongoing.
Notably,
a catalyst
system
can
be
generated
in situ
from
a
1:1:1:1
mixture
of
POPH
,
4-methylbenzyl
bromide,
NaBAr
F
24
,
and
Ni(py)
2
(CH
2
SiMe
3
)
2
and
shows
high
activity
[1010
kg/
(mol
·
h)]
in
ethylene/acrylate
copolymerization
at
110
°
C
(Table
1, entry
26).
These
results
highlight
the
versatility
of
the
reported
postsynthetic
ligand
modification
strategy.
With
this
one-pot
procedure,
parallel
generation
and
screening
of a
large
number
of
new
catalysts
can
be
accessed
via
high-
throughput
methods.
Encouraged
by
the
accessibility
and
efficiency
of steric
and
electronic
tuning
with
this
postsynthetic
strategy,
we
further
explored
ethylene/acrylate
copolymer
synthesis
under
varying
acrylate
concentrations
and
temperatures
(Figure
3, Table
1).
Catalysts
2b
2e
show
50%
120%
higher
activity
at 110
°
C
than
at 90
°
C
(Entries
19,
21,
23,
and
24
vs 4, 6, and
7, Table
1),
thereby
demonstrating
higher
thermal
stability
than
typically
reported
optimized
temperatures
for
most
Ni
catalysts
(<100
°
C).
Catalyst
2e
is even
active
in copolymerization
at
130
°
C
(Entry
25,
Table
1),
though
the
ethylene
uptake
data
suggest
decomposition
over
time
(Figures
S27
and
S28).
Specifically,
analogous
neutral
catalyst,
POP-Ni
,
shows
60%
lower
activity
at 100
°
C
than
at 90
°
C
(Entry
27
vs 28,
Table
1).
This
scenario
suggests
that
converting
the
free
phosphine
to
phosphonium
is beneficial
for
the
thermal
stability.
By
tuning
acrylate
concentration
and
reaction
temperature,
ethylene/acrylate
copolymers
with
a wide
range
of
M
w
and
acrylate
incorporation
are
accessible
(Figure
3).
Micro-
structural
analysis
of
the
resulting
ethylene/acrylate
copoly-
mers
reveals
that
these
cationic
catalysts
produce
mostly
linear
copolymers
with
a small
amount
of methyl
branching
(Table
S4).
In
contrast,
POP-Ni
produces
copolymers
with
virtually
no
methyl
branching.
This
difference
further
demonstrates
significant
impacts
of
phosphonium
versus
phosphine
on
Figure
3.
Correlation
between
structure
of cationic
Ni
catalysts
and
polymer
characteristics
[
M
w
, polydispersity
index
(PDI),
incorporation
ratio
of
acrylate,
and
number
of acrylate
units
per
chain]
and
correlations
between
catalyst
structures
and
polymer
characteristics.
T
= 90
°
C
(filled
shapes)
or
110
°
C
(empty
shapes);
[tBA]
= 0.025
M
(triangles),
0.05
M
(diamonds),
or
0.1
M
(circles).
See
Table
1 for
catalysis
runs
and
detailed
polymerization
conditions.
Journal
of the American
Chemical
Society
pubs.acs.org/JACS
Communication
https://doi.org/10.1021/jacs.4c03416
J. Am. Chem.
Soc.
XXXX,
XXX,
XXX
XXX
D