Mechanistic Studies of Pd(II)-Catalyzed Copolymerization of Ethylene and Vinylalkoxysilanes: Evidence for a β-Silyl Elimination Chain Transfer Mechanism

Article abstract of DOI:10.1021/jacs.6b10462

Copolymerizations of ethylene with vinyltrialkoxysilanes are reported using both a “traditional” cationic Pd(II) aryldiimine catalyst, t-1 (aryl = 2,6-diisopropylphenyl), and a “sandwich-type” aryldiimine catalyst, s-2 (aryl = 8-tolylnaphthyl). Incorporation levels of vinyltrialkoxysilanes between 0.25 and 2.0 mol % were achieved with remarkably little rate retardation relative to ethylene homopolymerizations. In the case of the traditional catalyst system, molecular weights decrease as the level of comonomer increases and only one trialkoxysilyl group is incorporated per chain. Molecular weight distributions of ca. 2 are observed. For the sandwich catalyst, higher molecular weights are observed with many more trialkoxysilyl groups incorporated per chain. Polymers with molecular weight distributions of ca. 1.2-1.4 are obtained. Detailed NMR mechanistic studies have revealed the formation of intermediate π-complexes of the type (diimine)Pd(alkyl)(vinyltrialkoxysilane)⁺. 1,2-Migratory insertions of these complexes occur with rates similar to ethylene insertion and result in formation of observable five-membered chelate intermediates. These chelates are rapidly opened with ethylene forming alkyl ethylene complexes, a requirement for chain growth. An unusual β-silyl elimination mechanism was shown to be responsible for chain transfer and formation of low molecular weight copolymers in the traditional catalyst system, t-1. This chain transfer process is retarded in the sandwich system. Relative binding affinities of ethylene and vinyltrialkoxysilanes to the cationic palladium center have been determined. The quantitative mechanistic studies reported fully explain the features of the bulk polymerization results.

Full text of DOI:10.1021/jacs.6b10462

Article
pubs.acs.org/JACS
Mechanistic Studies of Pd(II)-Catalyzed Copolymerization of Ethylene
and Vinylalkoxysilanes: Evidence for a β‑Silyl Elimination Chain
Transfer Mechanism
Zhou Chen,† Weijun Liu,‡,§ Olafs Daugulis,*,† and Maurice Brookhart*,†,‡
†Center for Polymer Chemistry, Department of Chemistry, University of Houston, Houston, Texas 77204-5003, United States
‡Department of Chemistry, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina 27599-3290, United States
S
* Supporting Information
ABSTRACT: Copolymerizations of ethylene with
vinyltrialkoxysilanes are reported using both a “traditional”
cationic Pd(II) aryldiimine catalyst, t-1 (aryl = 2,6-diisopropyl-
phenyl), and a “sandwich-type” aryldiimine catalyst, s-2 (aryl =
8-tolylnaphthyl). Incorporation levels of vinyltrialkoxysilanes
between 0.25 and 2.0 mol % were achieved with remarkably little
rate retardation relative to ethylene homopolymerizations. In the
case of the traditional catalyst system, molecular weights
decrease as the level of comonomer increases and only one trialkoxysilyl group is incorporated per chain. Molecular weight
distributions of ca. 2 are observed. For the sandwich catalyst, higher molecular weights are observed with many more
trialkoxysilyl groups incorporated per chain. Polymers with molecular weight distributions of ca. 1.2−1.4 are obtained. Detailed
NMR mechanistic studies have revealed the formation of intermediate π-complexes of the type (diimine)Pd(alkyl)-
(vinyltrialkoxysilane)+. 1,2-Migratory insertions of these complexes occur with rates similar to ethylene insertion and result in
formation of observable five-membered chelate intermediates. These chelates are rapidly opened with ethylene forming alkyl
ethylene complexes, a requirement for chain growth. An unusual β-silyl elimination mechanism was shown to be responsible for
chain transfer and formation of low molecular weight copolymers in the traditional catalyst system, t-1. This chain transfer
process is retarded in the sandwich system. Relative binding affinities of ethylene and vinyltrialkoxysilanes to the cationic
palladium center have been determined. The quantitative mechanistic studies reported fully explain the features of the bulk
polymerization results.
ethylene homopolymerization.9−11 Various factors responsible
INTRODUCTION
■
for reduced activities have been enumerated.12 As an example,
Coordination−insertion copolymerization of olefins with polar
vinyl monomers potentially offers a powerful method for the
mechanistic studies of the cationic α-diimine Pd(II) system for
MA/ethylene copolymerization revealed that after 2,1 insertion
controlled, low cost synthesis of high-value functionalized
polyolefins.1 Although early transition-metal catalysts are
extensively used for the coordination polymerization of
nonpolar olefins such as ethylene and propylene, the use of
of MA and rapid chain-walking, a six-membered carbonyl
chelate intermediate is formed. This chelate is the catalyst
resting state which is stable and strongly favored over the
“chelate opened” ethylene complex required for further
propagation (Scheme 1).8 The development of catalysts for
copolymerization of ethylene with polar vinyl monomers which
exhibit commercially viable activities and molecular weights
remains a challenge for the field.
these catalysts for copolymerization with polar vinyl monomers
is limited due to their high oxophilicity.2 In contrast, late
transition metal catalysts display enhanced tolerance toward
heteroatoms3 and can copolymerize ethylene with a range of
polar vinyl monomers, such as methyl acrylate,4 vinyl acetate,5
acrylonitrile,6 and other common polar monomers.7 An early
example from our group showed that cationic α-diimine Pd(II)
complexes can copolymerize ethylene (and higher α-olefins)
with methyl acrylate (MA) to form highly branched
copolymers.4a,8 Neutral phosphine-sulfonate Pd(II) “Drent-
type” catalysts copolymerize ethylene and MA to produce linear
functionalized copolymers.9 A variety of other Drent-type
catalysts have been used for copolymerization of several polar
monomers with ethylene.9,10
Scheme 1. Proposed Mechanism of α-Diimine Pd(II)
Catalyzed Copolymerization of Ethylene and MA
Copolymerizations of polar vinyl monomers with ethylene
Received: October 5, 2016
generally exhibit much lower activities relative to those of
© XXXX American Chemical Society
A
DOI: 10.1021/jacs.6b10462
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Polyethylene (PE) that is functionalized with trialkoxysilane
(−Si(OR)3) groups can be cross-linked to form PEX-b, a tough
material that is widely used for coating wires and cables and in
manufacturing plastic piping.13 The cross-links are formed by
exposing the PE containing −Si(OR)3 groups to steam to form
− Si−O−Si− bridges. The − Si(OR)3 groups are introduced
into PE either through postpolymerization radical grafting of
vinylalkoxysilanes (CH2CHSi(OR)3) to PE or through
radical-initiated copolymerization of ethylene with vinylalkoxy-
silanes.14 Vinylalkoxysilanes are polar monomers that have
generally received little attention with respect to coordination−
insertion polymerization using late transition metals.15,16
Successful copolymerizations of ethylene with vinylalkoxy-
silanes using α-diimine Ni(II) and Pd(II) catalysts have been
reported by researchers at DuPont.17 A very significant feature
of these copolymerizations is that the copolymerization
activities are higher than those of the MA/ethylene
copolymerizations and comparable to those of ethylene
homopolymerizations. The high activities of these copolymer-
izations prompted us to conduct a detailed mechanistic study of
the Pd-based diimine systems with the goal of understanding
why these polar monomers are so effective and to assess the
potential for designing even more effective catalysts.
RESULTS AND DISCUSSION
■
Copolymerization Reactions. As noted above, the
catalysts employed for the copolymerization reactions are the
cationic traditional isopropyl-substituted complex t-1a and
sandwich α-diimine Pd(II) complex s-2a (Scheme 2). These
complexes were synthesized following previously reported
procedures.20 Vinyltriethoxysilane (VTEoS) was chosen as a
model CH2CHSi(OR)3 monomer.
Scheme 2. Copolymerization of Ethylene and VTEoS Using
α-Diimine Pd(II) Complexes
While a number variations of aryl-substituted α-diimine
Pd(II) catalysts have been devised and investigated,18 studies
reported here are focused on the “traditional” catalyst, t-1,
bearing ortho-isopropyl groups18e,19 and the “sandwich”
catalyst s-2, incorporating 8-tolylnaphthyl imine moieties
(Figure 1).20 The traditional diimine system t-1 polymerizes
The results of copolymerizations using catalyst t-1a are
summarized in Table 1. Polymerizations were carried out in
dichloromethane at 100 psig ethylene, 22 °C with varying
concentrations of VTEoS. As with homopolymerization of
ethylene (entry 5), the polymers obtained are highly branched
with branching densities of ca. 105 branches/1000C.21 The
mole fraction incorporation of VTEoS into the polymer is
nearly proportional to the VTEoS concentration (Table 1,
entries 1−3). The molecular weights of the copolymers
produced are sensitive to the concentration of the VTEoS
monomer. For example, complex t-1a produced a copolymer
with Mn values of 7.0 × 103 g/mol at a VTEoS concentration of
0.52 M, 5.4 × 103 g/mol at 1.04 M VTEoS and 3.1 × 103 g/mol
at 1.56 M VTEoS (entries 1−3, Table 1).22 All Mw/Mn values
are ca. 2 indicating that the Mn values are chain-transfer limited.
This is further verified by entry 4 where doubling the
polymerization time relative to entry 1 produces a polymer
with the same Mn value. Olefinic units were detected, as the
polymer end groups establishing that β-elimination is
responsible for chain transfer. A calculation based on mole
percent incorporation and Mn values suggests that each chain
contains only ca. one VTEoS unit.23
An interesting feature, in contrast to most other ethylene/
polar vinyl monomer copolymerizations,12 is that the turnover
frequency (TOF) for the copolymerizations is not dramatically
reduced relative to the TOF for homopolymerization of
ethylene under similar conditions. For example, compare the
TOF of entry 1 (0.52 M VTEoS) of 362 h−1 with that of
homopolymerization (entry 5) of 456 h−1. The TOF drops
only slightly as VTEoS concentration increases (e.g., 362 h−1 at
0.52 M; 227 h−1 at 1.04 M).24
Figure 1. “Traditional” and “sandwich” Pd(II) α-diimine complexes.
ethylene to yield hyperbranched polyethylene at 20 °C with ca.
110 branches per 1000C and Mn values in the range of 104 g/
mol with molecular weight distributions of ca. 2.18e,19 The
recently developed “sandwich” α-diimine catalyst s-2 also
produces hyperbranched polyethylene (ca. 110 branches/
1000C) which exhibits very narrow molecular weight
distributions (Mw/Mn ca. 1.1).20 Comparison of associative
ligand exchange rates revealed that 8-tolyl groups in s-2 are far
more effective in shielding the axial coordination sites than the
ortho isopropyl groups of traditional catalyst t-1.20 This feature
results in significant reduction in the rate of chain transfer
relative to the rate of propagation and thus leads to the low
Mw/Mn values observed.
We present here a comprehensive account of the mechanistic
details of the copolymerization of ethylene with vinyl-
alkoxysilanes using catalysts t-1 and s-2. These studies provide
information regarding insertion barriers of these polar
monomers, their binding affinities relative to ethylene, and
the stability and properties of an intermediate chelate complex.
A chain transfer mechanism involving a unique β-silyl
elimination process has been identified.
The mechanistic studies described below will clarify the
mechanisms of chain transfer, the reason for one VTEoS per
chain, and the near insensitivity of the TOF to VTEoS. The
results of copolymerizations using sandwich catalyst s-2a are
also summarized in Table 1. TOFs are approximately a factor of
10 lower than those observed for analogous runs using
traditional catalyst t-1a, and this is consistent with previously
observed comparisons of homopolymerizations of ethylene
B
DOI: 10.1021/jacs.6b10462
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Table 1. Copolymerization of Ethylene and VTEoS Using Pd(II) Catalysts
a
b
b
b
c
c
entry
precatalyst
[VTEoS]
weight/g
TOF/h−1
incorp. (mol %)
branches/1000C
α-olefins (%)
Mn( × 10−3 g/mol)
Mw/Mn
1
2
3
t-1a
t-1a
t-1a
t-1a
t-1a
s-2a
s-2a
s-2a
s-2a
s-2a
s-2a
0.52 M
1.04 M
1.56 M
0.52 M
0
2.23
1.40
1.15
3.72
2.81
0.27
0.29
0.85
0.41
0.71
0.29
362
227
187
302
456
44
0.26
0.45
0.71
0.26
106
108
106
106
109
99
58
66
72
58
7.0
2.0
2.1
2.3
2.1
2.4
1.4
1.2
1.4
1.3
1.4
1.2
5.4
3.1
d
4
7.4
g
h
5
6
7
n.o.
n.d.
12.1
18.1
17.2
13.3
30.7
40.8
21.4
h
0.52 M
1.04 M
1.56 M
0.52 M
0.52 M
0
0.58
1.30
2.03
0.57
0.54
n.d.
h
47
98
n.d.
e
h
8
d
46
100
97
n.d.
h
9
33
n.d.
f
h
10
25
101
109
n.d.
g
h
11
47
n.o.
n.d.
a
b
1
Conditions: V(total) = 10 mL, 10 μmol of Pd catalyst, 22 h, 100 psig ethylene, dichloromethane, 22 °C. Determined by using H NMR
c
spectroscopy. The percentage (%) of α-olefin is the fraction of terminal olefin in all olefin groups. Molecular weight was determined by SEC in
trichlorobenzene at 140 °C. 44 h. 30 μmol of Pd catalyst. 100 h. Not observed. Not determined due to high molecular weights of the polymers.
d
e
f
g
h
with these two catalysts.20 As in the case of catalyst s-2a, highly
branched polymers are produced (ca. 100 branches/1000C).21
Remarkably, the TOFs observed in the copolymerization
runs using s-2a are nearly the same as that for the
homopolymerization (compare entry 11 for the homopolyme-
rization with entries 6 and 7). Molecular weight distributions, as
in the case for ethylene homopolymerizations using this
catalyst,20 are narrow (ca. 1.2−1.4 under these conditions)
and substantially less than 2; thus, Mn values have not reached
their chain-transfer limited values. Thus, even though turnover
frequencies are significantly less for catalyst s-2a versus those
for catalyst t-1a, significantly higher molecular weight polymers
can be generated (compare results in entries 1−4 vs 6−9). For
example, the molecular weight Mn of copolymer produced in
entry 10 by catalyst s-2a reached 40.8 × 103 g/mol with a
narrow distribution retained (Mw/Mn = 1.4).
4(Me) (Scheme 3). The displacement of extremely labile ether
with VTMoS is fast and complete at −78 °C. The H NMR
1
Scheme 3. Insertion of Vinyltrimethoxysilane into the Pd-
CH3 Bond of Complex t-4(Me)
signals of the bound olefinic protons appear upfield (δ = 5.24,
4.99, and 4.11 ppm) of those for free VTMoS, consistent with
formation of a η2-olefin complex rather than an O-bound
complex.
In contrast to copolymers produced by t-1a, copolymers
generated by s-2a exhibit more than one VTEoS monomer per
chain. For example, in entry 6, there are ca. 4 VTEoS units on
average per chain while for entries 7 and 10 there are ca. 8
VTEoS monomers on average per chain. In contrast to
copolymerizations using catalyst t-1a, it is clear that VTEoS
does not induce rapid chain transfer. Mechanistic studies
described below will illuminate the difference in behaviors of
these two catalytic systems and demonstrate that β-silyl
elimination is responsible for chain transfer using the traditional
catalyst.
Mechanistic Studies. Low Temperature NMR Studies of
Migratory Insertion Reactions of Vinyltrialkoxysilanes:
Barriers and Activation Parameters. The complexes employed
for mechanistic studies are the cationic α-diimine Pd(II)
complexes t-1b, t′-3b, and s-2b (Figure 2), which were
synthesized according to the reported procedures.25
Migratory insertion occurs at −40 °C yielding exclusively the
1,2-insertion product t-5(Me), which is a five-membered
chelate complex involving chelation of one −OMe group.
Complex t-5(Me) exhibits fluxional behavior due to C−Si bond
rotation and interchange of the coordinated −OMe group (δ =
2.95 ppm) with the two free −OMe groups (δ = 3.54 and 3.48
ppm). At −60 °C, the rotational barrier for this process was
measured via line broadening techniques (ΔG‡ = 11.5 kcal/
mol, −60 °C). In addition to the −OMe resonances noted
above, characteristic resonances of insertion product t-5(Me)
include signals at δ 1.37 ppm and −0.82 ppm for the
diastereotopic methylene protons of Pd−CH2, δ 1.97 ppm for
the Si−CH methine proton and δ 0.37 ppm for the methyl
group at C-β of the chelate.26 The first-order insertion reaction
was followed by 1H NMR spectroscopy at temperatures
between −37 and −25 °C. From these data, values of the
activation enthalpy (ΔH‡ = 17.4 2.2 kcal/mol) and activation
entropy (ΔS‡ = −4.1 9.1 eu) were obtained (Figure 3). The
modest negative value of the activation entropy is consistent
with an intramolecular migratory insertion reaction. At −33 °C,
the calculated free energy barrier (ΔG‡ = 18.4 kcal/mol) is only
slightly higher than that of ethylene migratory insertion in the
corresponding ethylene complex t-1c (ΔG‡ = 17.8 kcal/mol,
−33 °C).25 Since the barrier of subsequent ethylene insertions
does not differ significantly from the barrier of the first insertion
of ethylene into the Pd−CH3 bond,25 we believe it is not
unreasonable to use the insertion of VTMoS into a Pd−CH3
bond as a model for insertion of VTMoS into the growing
polymer chain.
In order to generate simplified NMR spectra, vinyl-
trimethoxysilane (VTMoS) was used in place of VTEoS.
Addition of 1 equiv of VTMoS to a CD2Cl2 solution of t-1b at
−78 °C leads to clean formation of the η2-olefin complex t-
Figure 2. Cationic (α-diimine)Pd(CH3)(OEt2)+ complexes.
C
DOI: 10.1021/jacs.6b10462
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Table 2. Migratory Insertion Barriers for Different
Vinylalkoxysilane and Ethylene Complexes
entry
olefin
olefin adduct T/°C k/ × 104 s−1 ΔG‡ (kcal/mol)
1
2
3
4
5
6
7
8
9
VTMoS
VTEoS
VTPoS
VDMEoS
VTMoS
C2H4
t-4(Me)
t-4(Et)
t-4(iPr)
t-8
−30
−30
−25
−30
−30
−33
−20
−20
−20
1.4
1.3
2.2
2.6
0.8
3.1
1.7
2.8
1.4
18.4
18.5
18.6
18.1
18.7
17.8
19.1
18.9
19.2
t′-6(Me)
t-1c
a
C2H4
s-2c
VTMoS
VTEoS
s-11(Me)
s-11(Et)
a
An identical barrier for this case has previously been reported by
Tempel et al.25
kcal/mol at −30 °C). The insertion barrier for VTPoS complex
t-4(iPr) was found to be 18.6 kcal/mol at −25 °C, ca. 0.2 kcal/
mol higher than that for VTMoS complex t-4(Me). The
insertion barrier for vinyldimethylethoxysilane (VDMEoS)
complex t-8 was determined to be 18.1 kcal/mol at −30 °C,
ca. 0.4 kcal/mol lower than that of VTEoS complex t-4(Et).
The migratory insertion barrier for VTMoS complex t′-6(Me),
which bears a less bulky ligand, was determined to be 18.7 kcal/
mol, which is only ca. 0.3 kcal/mol higher than that for t-
4(Me). Examining entries 1−5 shows there is a negligible effect
of −OR or ortho aryl substituents on barriers to migratory
insertion of vinylalkoxysilanes.
In comparison to catalyst t-1a, the activity of sandwich
catalyst s-2a is much lower as discussed above. The migratory
insertion barriers for sandwich ethylene and vinylalkoxysilane
complexes were thus investigated. Sandwich ethylene complex
s-2c was generated in situ by addition of 20 equiv of ethylene to
ether adduct s-2b (Scheme 5). At −40 °C, two broad signals at
Figure 3. Eyring plots for migratory insertion of VTMoS in t-4(Me)
and s-11(Me).
In addition to VTMoS, several other vinylsilane monomers
were also investigated (Scheme 4). The η2-olefin complexes t-
Scheme 5. Insertion of Ethylene into the Pd−CH3 Bond of
Sandwich Complex s-2c
Scheme 4. Insertion of Vinylsilanes into the Pd−CH3 Bonds
of Traditional Vinylsilane Complexes
δ 4.07 and 3.95 ppm were observed for bound ethylene
together with a Pd−CH3 resonance at δ 0.25 ppm. The
insertion of ethylene into the Pd−CH3 bond is first-order, and
the rate was determined by monitoring the decrease of the
integral for the Pd−CH3 signal (ΔG‡ = 19.1 kcal/mol at −20
°C). The ethylene complex i-10 was obtained as the insertion
product. Characteristic resonances of complex i-10 were
observed at δ 1.75 ppm (broad multiplet, PdCH(CH3)2),
i
4(Et, Pr), t′-6(Me), and t-8 were generated in situ by
treatment of the olefins with ether adducts t-1b and t′-3b at
−78 °C. Insertion barriers for these η2-olefin complexes are
summarized in Table 2. VTEoS and vinyltrisopropoxysilane
(VTPoS) adducts, t-4(Et), t-4(iPr), display similar upfield
shifts of the methylidene olefinic protons. The migratory
insertions of these two complexes proceed in a similar manner
as that of VTMoS complex t-4(Me), resulting in the formation
of five-membered chelate complexes. The insertion barrier for
VTEoS complex t-4(Et) (ΔG‡ = 18.5 kcal/mol at −30 °C) is
similar to that for the VTMoS complex t-4(Me) (ΔG‡ = 18.4
3
0.73 ppm (d, JHH = 6.6 Hz, PdCH(CH3)2), and 0.31 ppm (d,
3JHH = 6.0 Hz, PdCH(CH3)2) similar to those reported for the
analogous isopropyl-substituted system.25 The isopropyl ethyl-
ene complex i-10 presumably forms by trapping a β-agostic
isopropyl complex, which forms via rapid isomerization of an
D
DOI: 10.1021/jacs.6b10462
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
initially formed β-agostic n-propyl complex. The isopropyl
ethylene complex i-10 gradually rearranges to n-propyl ethylene
complex n-10 (Scheme 5). This same isomerization/trapping
behavior has been found for the traditional diimine complex.25
Addition of 1 equiv of VTMoS to a CD2Cl2 solution of s-2b
at −78 °C leads to clean formation of two isomers of η2-
Scheme 7. β-Silyl Elimination Observed for Chelate
Complex t′-7(Me)
1
VTMoS complex s-11(Me) in a 3:2 ratio (Scheme 6). Key H
Scheme 6. Insertion of Vinylsilanes into the Pd−CH3 Bonds
protons. The ratio of complex t′-7(Me) and β-silyl elimination
product t′-13(Me) is ca. 10:1 at −25 °C. The addition of excess
acetonitrile to an NMR sample of t′-7(Me)/t′-13(Me) results
in the quantitative formation of free propylene (1H NMR: 5.80,
5.00, 4.89, and 1.67 ppm) and silyl acetonitrile complex t′-
14(Me) (Scheme 7). In contrast to the methyl-substituted
complex t′-7(Me), the more bulky isopropyl-substituted
complex t-5(Me) forms a much smaller fraction (<5%) of β-
silyl elimination product. Notably, no β-silyl elimination
product could be detected in the case of the sandwich complex
s-12(Me) even at 0 °C.
Chelate Opening with Ethylene and Chain Transfer.
Addition of 20 equiv of ethylene to the CD2Cl2 solution of
isopropyl-substituted complex t-5(Me) at −78 °C results in
rapid formation of a new complex proposed to be t-15 (Scheme
8). Warming from −78 to −60 °C, complex t-15 rapidly
of s-11
NMR resonances for the major isomer include signals at δ =
4.52, 4.44, and 3.56 ppm for the bound olefinic protons and δ =
0.44 ppm for the Pd−CH3 methyl group. Resonances
attributable to the minor isomer were observed at δ = 4.82,
4.70, 3.27 ppm for bound olefinic protons and δ = 0.52 ppm for
the Pd−CH3 methyl group. Upon warming to −40 °C, a 1:1
equilibrium ratio of isomers is generated. Similar to complex t-
4(Me), migratory insertion of VTMoS complex s-11(Me)
occurs in a 1,2 fashion to form a five-membered chelate
complex, but the case here is complicated by the fact the α-
diimine ligand possesses C2 symmetry and two isomeric five-
membered chelates can be formed. Only one chelate isomer
was fully distinguishable by NMR, due to overlapping and
Scheme 8. Chelate Opening of Complex t-5(Me) with
Ethylene
1
complex resonances. Key H NMR resonances of the major
chelate isomer include signals at δ 1.32 and −0.95 ppm for the
diastereotopic methylene protons of Pd−CH2, δ 1.61 ppm for
the Si−CH methine proton and δ 0.38 ppm for the methyl
group on the chelate. The migratory insertion reaction of the
VTMoS complex s-11(Me) was observed at −20 °C by
monitoring the disappearance of Pd−CH3 methyl resonances.
First-order kinetics were observed with kins = 2.8 × 10−4 s−1
(ΔG‡ = 18.9 kcal/mol). Figure 3 shows an Eyring plot for this
insertion reaction over the range −30 to −17 °C in CD2Cl2.
The activation enthalpy (ΔH‡) is 18.3 2.2 kcal/mol, while
the activation entropy (ΔS‡) is −1.1 8.6 eu. The near-zero
value of ΔS‡ is consistent with an intramolecular migratory
insertion. The insertion barrier (ΔG‡ = 18.9 kcal/mol, −20 °C)
is roughly the same as that for ethylene insertion (ΔG‡ = 19.1
kcal/mol, −20 °C) in complex s-2c and ca. 0.5 kcal/mol higher
than that of VTMoS insertion in the isopropyl complex t-
4(Me) (ΔG‡ = 18.4 kcal/mol, −30 °C). The corresponding
VTEoS complex s-11(Et) displays similar spectral features to
the VTMoS analogue s-11(Me), including the occurrence of
two isomers. At −20 °C, VTEoS inserts into the Pd−CH3 bond
with kins = 1.4 × 10−4 s−1 (ΔG‡ = 19.2 kcal/mol) to afford
chelate complexes. The insertion barrier is ca. 0.3 kcal/mol
higher than that for VTMoS complex s-11(Me) (ΔG‡ = 18.9
kcal/mol, −20 °C) and almost the same as that for sandwich
ethylene complex s-2c (ΔG‡ = 19.1 kcal/mol, −20 °C).
isomerizes to the linear silylpropyl ethylene complex t-16.
Characteristic resonances of complex t-16 were observed at δ
1.52 ppm (t, JHH = 8.4 Hz, PdCH2CH2CH2Si(OMe)3), 1.05
3
ppm (broad multiplet, PdCH2CH2CH2Si(OMe)3), and 0.17
3
ppm (t, JHH = 8.4 Hz, PdCH2CH2CH2Si(OMe)3). Upon
warming to −40 °C, further rearrangement of complex t-16 to
silylethyl ethylene complex t-17 was observed, along with
formation of ca. 1 equiv of free propylene (Scheme 8). The
formation of t-17 was clearly indicated by the presence of two
triplets at δ 1.74 (PdCH2) and δ 0.42 (SiCH2) ppm assigned to
the two sets of methylene protons.
The mechanism proposed for formation of t-16 and t-17 is
shown in Scheme 9. Loss of ethylene from complex t-15 (or
opening of chelate complex t-5(Me)) followed by β-silyl
elimination results in the formation of trimethoxysilyl η2-
propylene intermediate t-18. Intermediate t-18 undergoes a
2,1-migratory insertion to yield t-19.27 Subsequent “chain
walking” via intermediate t-20 yields the observed silylpropyl
ethylene complex t-16. If all steps are rapid and reversible, low
concentrations of intermediate t-18 can be intercepted by
ethylene to form t-22 and (observed) propylene. Insertion of
ethylene to the Pd-silyl bond of t-22 leads to a five-membered
chelate intermediate t-23. Reaction of ethylene (present in
Evidence for β-Silyl Elimination. At higher temperatures, β-
silyl elimination in methyl-substituted traditional complex t′-
7(Me) was observed forming a silyl propylene complex, t′-
13(Me) (Scheme 7).27 Characteristic resonances of bound
propylene in complex t′-13(Me) include δ 5.44, 4.51, and 4.21
ppm for the olefinic protons and δ 1.76 ppm for the methyl
E
DOI: 10.1021/jacs.6b10462
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Scheme 9. Proposed Mechanism of Formation of Complexes
t-16 and t-17
Scheme 11. Proposed Chain Transfer Process via a β-Silyl
Elimination Mechanism
excess) with intermediate t-23 opens the chelate to generate
the observed silylethyl ethylene complex t-17.
associative displacement of the unsaturated chain by ethylene
to yield the silyl ethylene complex, t-31, which can initiate
growth of a new chain as shown above. Species t-29 could
undergo 2,1 insertion to yield t-30, but ultimately, reversible β-
silyl elimination will reform t-29. Chain transfer via this
pathway will result in formation of copolymers with a terminal
olefin end group which matches well with the end group
analysis of these copolymers obtained in copolymerization
experiments (Table 1). This mechanism also accounts for the
incorporation of one −Si(OR)3 group per chain.
Ethylene complex t-17 in the presence of excess ethylene at
−25 °C undergoes chain growth along with formation of some
1-butene and minor amounts of higher α-olefins.24 Thus, the
picture that emerges for initiation and continued chain growth
is shown in Scheme 10. As the alkyl chain grows longer, the rate
at which chain-walking moves the silyl group to the β position
(necessary for β-silyl elimination) decreases relative to the rate
of ethylene insertion. Thus, elimination and formation of higher
α-olefins will rapidly decrease as the alkyl chain grows.
It is clear from the above results that β-silyl elimination
serves as the chain transfer mechanism and results in the low
molecular weight of copolymers if catalyst t-1 is used (Scheme
11). After chain growth via a series of ethylene insertions,
trapping of this species by the vinylsilane would yield a π-
complex of type t-27. 1,2-Insertion of the vinylsilane would
yield a five-membered chelate t-28 which can undergo β-silyl
elimination to generate t-29. Chain transfer is completed via
In the case on the sandwich system, we have established that
associative displacement of bound ethylene by free ethylene
from a sandwich Pd(II) alkyl ethylene complex is at least 5
orders of magnitude slower than the same displacement
reaction in the traditional catalyst system.20 This feature results
in polyethylene formed from these catalysts having very narrow
molecular weight distributions. On this basis, we would expect
that the rate of chain transfer via the β-silyl mechanism would
Scheme 10. Proposed Mechanism of Chain Growth in the Traditional Catalyst System
F
DOI: 10.1021/jacs.6b10462
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
be greatly diminished in these copolymerizations. Slow chain
transfer is supported by the observation that copolymers
formed from sandwich catalyst s-2a have very narrow molecular
weight distributions, copolymers can be generated with
considerably higher molecular weights than with t-1a, and
several vinyltrialkoxysilane units can be incorporated per chain.
Additional low temperature NMR experiments confirm this
hypothesis. Addition of 20 equiv of ethylene to a CD2Cl2
solution of sandwich chelate complex s-12(Me) at −78 °C
results in complete opening of the chelate to form two isomers
of the ethylene complex s-15 (Scheme 12). Again, the C2
Table 3. Average Rates for Subsequent Ethylene Insertions
initiating
complex
ΔG‡
entry
olefin
T/°C k/ × 104 s−1
(kcal/mol)
1
2
3
4
VTMoS
C2H4
t-17
t-1c
s-16
s-2c
−25
−28
0
3.1
5.6
5.8
6.7
18.4
a
17.9
VTMoS
C2H4
20.0
0
19.9
a
This barrier has also been measured to be ΔG‡ = 18.0 kcal/mol, −26
°C by Tempel et al.25
expected, nearly the same as that for ethylene homopolyme-
rization, ΔG‡ = 17.9 kcal/mol at −28 °C.25
Scheme 12. Chelate Opening with Ethylene in Sandwich
Complex s-12(Me)
For the sandwich complex s-16, the rate constant for
ethylene insertion was determined to be 5.8 × 10−4 s−1 at 0 °C
(ΔG‡ = 20.0 kcal/mol). This barrier is the same as that for
ethylene homopolymerization (ΔG‡ = 19.9 kcal/mol, 0 °C)
and ca. 1.6 kcal/mol higher than that of the isopropyl system t-
17 (ΔG‡ = 18.4 kcal/mol, −25 °C).
Relative Binding Constants. In the copolymerization of
ethylene and methyl acrylate using traditional α-diimine
catalysts, where ligand exchange is faster than insertion, the
incorporation ratio of ethylene vs MA into the copolymer is
determined by their relative insertion rates, their relative
binding affinities, and their relative concentrations in solution.8
Given that the insertion barriers of vinylalkoxysilanes are
comparable to those of ethylene, the relative binding affinities
of vinyltrialkoxysilanes and ethylene become a key factor in
determining the incorporation ratio aside from the concen-
tration difference in the solution of ethylene and the
vinylalkoxysilane.
symmetry of the ligand and the chiral center α to Si results in
formation of two isomers. Upon warming to −20 °C, the
silylpropyl ethylene complex s-16 was observed (in analogy
with t-16) which is readily identified by 1H NMR analysis. The
characteristic resonances of complex s-16 are seen at δ 0.93
ppm (PdCH2CH2CH2Si(OMe)3), −0.03 ppm (broad multip-
lets, PdCH2CH2CH2Si(OMe)3), and −0.62 ppm (t, 3JHH = 12.0
Hz, PdCH2CH2CH2Si(OMe)3). Complex s-16 reacts with
ethylene and alkyl chain growth ensues as evidenced by the
disappearance of the signals of s-16 and the decrease in the
ethylene resonance. A key point is that no free propylene was
detected even at 0 °C. Thus, while β-silyl elimination must occur
in converting s-15 to s-16 via a silyl propylene complex,
propylene is not displaced by ethylene but rather chain growth
via ethylene insertion ensues.28 These observations establish
that the rate of chain transfer via the β-silyl mechanism is very
greatly retarded in this sandwich catalyst system due to the very
slow associative displacement of olefins from any silyl olefin
complex intermediates.
Chain Propagation. Since chelate opening by ethylene is
extremely rapid and no observable chelate is detected in the
presence of ethylene, it is clear that alkyl ethylene complexes
are the catalyst resting states. The rate of migratory insertion of
ethylene controls the turnover frequency which will be zero-
order in ethylene. The turnover frequencies for the two
catalysts were determined by NMR spectroscopy by measuring
the decreasing signal for free ethylene in solution (Table 3).25
The numbers are consistent with values determined for
homopolymerizations of ethylene where the alkyl ethylene
complexes are the resting states. For the isopropyl-substituted
system t-17, the average rate of ethylene insertion was
measured at −25 °C and the corresponding activation barrier,
ΔG‡, was determined to be 18.4 kcal/mol, which is, as
For the isopropyl catalyst system, t-1, equilibrium constants
for the reaction depicted in Scheme 13 were measured by low-
Scheme 13. Relative Binding Affinities of Ethylene and
Vinylalkoxysilanes
temperature 1H NMR spectroscopy. Results show that ethylene
binds more strongly than vinyltrialkoxysilanes, presumably due
to steric factors. For VTMoS, the equilibrium constant (Keq) is
4.8 × 10−3 at −87 °C (ΔG = 2.0 kcal/mol), while, for
vinyltriethoxysilane (VTEoS), the equilibrium constant is
slightly smaller, Keq = 3.1 × 10−3, ΔG = 2.1 kcal/mol. In
comparison to the binding affinity of methyl acrylate where an
analogous Keq = 1.0 × 10−6 (−95 °C),8 the vinyltrialkoxysilanes
bind ca. 1000 times more strongly. Extrapolating the above data
to the temperature of the copolymerization reactions (22 °C)
and assuming ΔS = 0, the equilibrium constants calculated are
0.033 and 0.028, respectively, which favor ethylene but not to
such a high degree to exclude significant polar monomer
incorporation.
By extrapolation of these data to the temperature of the
copolymerization reaction (22 °C), we find good agreement
with the incorporation ratio observed in copolymerization of
ethylene with VTEoS. Eyring analyses have not been carried
out for the ethylene and VTEoS insertions, but assuming a
G
DOI: 10.1021/jacs.6b10462
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
similar ΔS‡ to that for VTMoS complex t-4(Me) (−4.1 9.1
eu) and using the measured ΔG‡ at −30 °C (18.5 kcal/mol) for
VTEoS complex t-4(Et) and the measured barrier for ethylene
insertion at −33 °C (17.8 kcal/mol), barriers for insertion of
VTEoS and ethylene at 22 °C can be estimated as ca. 18.7 and
18.0 kcal/mol, respectively. The estimated difference between
these two insertion barriers at 22 °C of ca. 0.7 kcal/mol
corresponds to a rate ratio kVTEoS/kethylene of ca. 0.3. Multiplying
this insertion rate ratio by the relative binding affinities and
taking into account the olefin concentrations in entry 1, Table 1
(i.e., [VTEoS] = 0.52 M, [Ethylene] = 1.39 M),29 a VTEoS
incorporation ratio of ca. 0.31 mol % is predicted, which agrees
reasonably well with that of 0.26 mol % observed.
As for the sandwich catalyst system, attempted measurement
of the equilibrium of s-2c and s-11(Me) (or s-11(Et)) at low
temperatures was unsuccessful due to the extremely slow
exchange rates and the fact that migratory insertions of these
complexes are faster than ligand exchanges. We can, however,
measure the relative kinetic trapping rates of ether complex s-
2b by vinyltrialkoxysilanes versus ethylene to yield complexes s-
2c and s-11(Me) or s-11(Et) as shown in Scheme 14. For
rates are significantly reduced relative to the ethylene
homopolymerizations. Studies reported here have exposed
numerous mechanistic features of the chain growth and
termination processes which together provide a rationale for
the behavior of these copolymerizations reactions. Key points
are summarized below:
(1) The traditional catalyst t-1a copolymerizes ethylene with
VTEoS to give highly branched functionalized poly-
ethylene. In comparison to copolymerization of ethylene
with methyl acrylate, the rate of copolymerization of
ethylene with VTEoS is much higher and comparable to
that of ethylene homopolymerization, but at the same
time, molecular weights of the copolymers are much
lower than those of the homopolymer.
(2) The cationic (α-diimine)PdMe(OEt2)+ complexes react
rapidly with vinyltrimethoxysilane (VTMoS) to form η2-
olefin complexes at −78 °C. Migratory insertion of
VTMoS into the Pd−CH3 bond occurs regioselectively
in a 1,2 fashion to yield five-membered chelates.
Insertion barriers of VTMoS were determined and
found to be only slightly higher than those for ethylene
insertions. Thus, slow insertion rates of the comonomer
are not a factor in retarding the levels of comonomer
incorporation.
Scheme 14. Relative Trapping Rates of Ethylene and
Vinyltrialkoxysilanes
(3) The five-membered chelates formed from VTMoS
insertion are very rapidly opened at −78 °C by ethylene
to form Pd(II) ethylene alkyl complexes. Equilibrium
between the chelates and the opened alkyl ethylene
complexes strongly favors the opened ethylene alkyl
species. Thus, copolymerization rates are not retarded by
chelate formation. This behavior is in contrast to MA
where the chelate generated from MA insertion is
favored over the open alkyl ethylene complex.
(4) For traditional catalyst t-1b, associative exchange of
olefins is rapid relative to insertion; thus, the fraction of
vinylsilane incorporation is determined by the relative
binding affinities of the two monomers coupled with
their relative insertion rates (Curtin−Hammett kinetics).
Under polymerization conditions, the ratio of insertion
rates, kVTEoS/kethylene, was estimated as ca. 0.3 and the
relative binding affinity of the vinyltrialkoxysilanes and
ethylene was ca. 0.03. By taking into account monomer
concentrations, the estimated VTEoS incorporation for
entry 1, Table 1 is 0.31 mol % in reasonable agreement
with the observed value of 0.26 mol %. This level of
incorporation is sufficient to effect cross-linking. In the
case of the sandwich complex, the exchange rates of
olefins are quite slow; thus, silane incorporation levels are
determined, at least in part, by relative rates of trapping
of the cationic β-agostic alkyl complexes. The relative
rates (ksilane/kethylene) of trapping the methyl diethyl ether
complex s-2b are ca. 0.3, which favor ethylene, and have
a smaller difference between silanes and ethylene than
the relative binding affinities.33 This feature results,
surprisingly, in a higher level of incorporation of the
vinyltrialkoxysilanes into the copolymers relative to the
traditional catalysts, even though the sandwich catalysts
possess more hindered coordination sites.
vinyltrimethoxysilane (VTMoS), the relative trapping rate
kVTMoS/kethylene is ca. 0.33 at −60 °C, while that of VTEoS is
essentially the same, ca. 0.34 at −60 °C. These relative trapping
rates are ca. 10 times greater than relative binding affinities for
the traditional complex at 22 °C, which might account for the
higher incorporation ratio (ca. 2 times) of VTEoS in the
copolymer produced by the “sandwich” complex s-2a relative to
complex t-1a, despite the fact that s-2a contains the more
crowded binding site.30,31 We cannot estimate what the
incorporation level of vinyltrialkoxysilanes should be as we do
not know what the trapping ratios of the bulkier β-agostic alkyl
intermediates32 will be relative to the ether complexes s-2b.
Furthermore, some displacement of the more weakly binding
vinylsilane may be competitive with insertion under the
polymerization conditions. But, nonetheless, slow olefin
exchange coupled with favorable kinetic trapping ratios
provides a potential rationale for the observation that greater
levels of the vinylsilanes can be incorporated into copolymers
produced using the bulkier sandwich catalysts.
CONCLUSIONS
■
The α-diimine Pd(II) catalysts t-1a and s-2a exhibit functional
group tolerance and allow for the copolymerization of ethylene
with vinyltrialkoxysilanes to yield highly branched copolymers
at rates similar to those for the homopolymerization of
ethylene. This behavior is in contrast to the copolymerization
of ethylene with methyl acrylate using these catalysts where
(5) For the traditional catalyst t-1a, each copolymer chain
produced contains one vinyltrialkoxysilane unit and the
copolymer molecular weights decrease as the concen-
tration of the vinyltrialkoxysilane increases. The Mw/Mn
H
DOI: 10.1021/jacs.6b10462
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Y. Macromolecules 2013, 46, 2870−2875. (g) Wang, X.; Wang, Y.; Shi,
X.; Liu, J.; Chen, C.; Li, Y. Macromolecules 2014, 47, 552−559.
(3) (a) Younkin, T. R.; Connor, E. F.; Henderson, J. I.; Friedrich, S.
K.; Grubbs, R. H.; Bansleben, D. A. Science 2000, 287, 460−462.
(b) Diamanti, S. J.; Ghosh, P.; Shimizu, F.; Bazan, G. C.
Macromolecules 2003, 36, 9731−9735. (c) Zuideveld, M. A.;
values of these polymers are ca. 2 showing the Mn values
are chain-transfer limited. Low temperature NMR studies
established that the chelate intermediates undergo facile
β-silyl elimination with associative displacement of the
formed olefin by ethylene. Thus, insertion of
vinyltrialkoxysilane into a growing chain results in
chain transfer and initiation of a new chain via silane
insertion; each chain produced bears a −Si(OR)3 end
group. For the sandwich catalyst β-silyl elimination could
occur, but the formed olefin is not displaced. Thus, the
vinylsilane monomers do not induce rapid chain transfer,
high molecular weight polymers can be produced with
narrow molecular weight distributions, and many vinyl-
trialkoxysilane monomers can be incorporated into a
single chain.
Wehrmann, P.; Rohr, C.; Mecking, S. Angew. Chem., Int. Ed. 2004,
̈
43, 869−873.
(4) (a) Johnson, L. K.; Mecking, S.; Brookhart, M. J. Am. Chem. Soc.
1996, 118, 267−268. (b) Drent, E.; van Dijk, R.; van Ginkel, R.; van
Oort, B.; Pugh, R. I. Chem. Commun. 2002, 744−745. (c) Guironnet,
D.; Roesle, P.; Runzi, T.; Gottker-Schnetmann, I.; Mecking, S. J. Am.
̈
̈
Chem. Soc. 2009, 131, 422−423.
(5) Ito, S.; Munakata, K.; Nakamura, A.; Nozaki, K. J. Am. Chem. Soc.
2009, 131, 14606−14607.
(6) Kochi, T.; Noda, S.; Yoshimura, K.; Nozaki, K. J. Am. Chem. Soc.
2007, 129, 8948−8949.
Due to the high cost of palladium and the relative low
activities of these catalysts, they hold little industrial interest.
We are currently extending mechanistic and preparative studies
to nickel analogs that show much higher activities.
(7) (a) Luo, S.; Jordan, R. F. J. Am. Chem. Soc. 2006, 128, 12072−
12073. (b) Luo, S.; Vela, J.; Lief, G. R.; Jordan, R. F. J. Am. Chem. Soc.
2007, 129, 8946−8947. (c) Weng, W.; Shen, Z.; Jordan, R. F. J. Am.
Chem. Soc. 2007, 129, 15450−15451. (d) Runzi, T.; Frohlich, D.;
̈
̈
Mecking, S. J. Am. Chem. Soc. 2010, 132, 17690−17691. (e) Sui, X.;
Dai, S.; Chen, C. ACS Catal. 2015, 5, 5932−5937. (f) Schuster, N.;
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/jacs.6b10462.
■
Runzi, T.; Mecking, S. Macromolecules 2016, 49, 1172−1179.
̈
S
(8) Mecking, S.; Johnson, L. K.; Wang, L.; Brookhart, M. J. Am.
Chem. Soc. 1998, 120, 888−899.
(9) Nakamura, A.; Anselment, T. M. J.; Claverie, J.; Goodall, B.;
Jordan, R. F.; Mecking, S.; Rieger, B.; Sen, A.; van Leeuwen, P. W. N.
M.; Nozaki, K. Acc. Chem. Res. 2013, 46, 1438−1449.
Detailed experimental procedures, characterization data
for complexes t-1b, s-2b,c, t-4,5,8,9,16,17, t′-6,7, s-
1
(10) (a) Zhou, X.; Jordan, R. F. Organometallics 2011, 30, 4632−
4642. (b) Gott, A. L.; Piers, W. E.; Dutton, J. L.; McDonald, R.;
Parvez, M. Organometallics 2011, 30, 4236−4249. (c) Wucher, P.;
11,12,16; H NMR spectra of copolymer; and kinetic
plots (PDF)
Roesle, P.; Falivene, L.; Cavallo, L.; Caporaso, L.; Gottker-
̈
AUTHOR INFORMATION
Schnetmann, I.; Mecking, S. Organometallics 2012, 31, 8505−8515.
(d) Contrella, N. D.; Sampson, J. R.; Jordan, R. F. Organometallics
2014, 33, 3546−3555. (e) Zhang, Y.; Cao, Y.; Leng, X.; Chen, C.;
Huang, Z. Organometallics 2014, 33, 3738−3745. (f) Nakano, R.;
Nozaki, K. J. Am. Chem. Soc. 2015, 137, 10934−10937. (g) Jian, Z.;
■
Corresponding Authors
*mbrookhart@unc.edu
*olafs@uh.edu
ORCID
Falivene, L.; Wucher, P.; Roesle, P.; Caporaso, L.; Cavallo, L.; Gottker-
̈
Schnetmann, I.; Mecking, S. Chem. - Eur. J. 2015, 21, 2062−2075.
(11) (a) Carrow, B. P.; Nozaki, K. J. Am. Chem. Soc. 2012, 134,
8802−8805. (b) Long, B. K.; Eagan, J. M.; Mulzer, M.; Coates, G. W.
Angew. Chem., Int. Ed. 2016, 55, 7106−7110. (c) Tao, W.-j.; Nakano,
R.; Ito, S.; Nozaki, K. Angew. Chem., Int. Ed. 2016, 55, 2835−2839. (d)
A recent report showed that a cationic Pd(II) catalyst based on a
bidentate phosphine/phosphine oxide catalyst shows only a small drop
(ca. 3 times) in the rate of ethylene/methyl acrylate copolymerization
relative to ethylene homopolymerization: Mitsushige, Y.; Carrow, B.
P.; Ito, S.; Nozaki, K. Chem. Sci. 2016, 7, 737−744.
Zhou Chen: 0000-0003-4345-7070
Present Address
§Canon Nanotechnologies, Inc., Austin, TX 78758.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This research was supported by the National Science
Foundation (CHE-1010170 to M.B.) and the Welch
Foundation (Chair E-0044 to O.D. and Grant E-1893 to M.B.).
(12) Williams, B. S.; Leatherman, M. D.; White, P. S.; Brookhart, M.
J. Am. Chem. Soc. 2005, 127, 5132−5146.
(13) (a) Tsunaga, M.; Matsunami, S.; Sato, M.; Nakata, K. Plast.,
Rubber Compos. 1998, 27, 460−464. (b) Rizzo, P.; Baione, F.; Guerra,
G.; Martinotto, L.; Albizzati, E. Macromolecules 2001, 34, 5175−5179.
(c) Khonakdar, H. A.; Morshedian, J.; Wagenknecht, U.; Jafari, S. H.
Polymer 2003, 44, 4301−4309.
(14) (a) Scott, H. G. U.S. 3646155, 1972. (b) Narkis, M.; Tzur, A.;
Vaxman, A.; Fritz, H. G. Polym. Eng. Sci. 1985, 25, 857−862.
(15) Kawakami, T.; Ito, S.; Nozaki, K. Dalton Trans. 2015, 44,
20745−20752.
(16) (a) Amin, S. B.; Marks, T. J. J. Am. Chem. Soc. 2006, 128, 4506−
4507. (b) Liu, J.; Nomura, K. Macromolecules 2008, 41, 1070−1072.
(c) Nomura, K.; Kakinuki, K.; Fujiki, M.; Itagaki, K. Macromolecules
REFERENCES
■
(1) (a) Ittel, S. D.; Johnson, L. K.; Brookhart, M. Chem. Rev. 2000,
100, 1169−1203. (b) Boffa, L. S.; Novak, B. M. Chem. Rev. 2000, 100,
1479−1493. (c) Chung, T. C. Prog. Polym. Sci. 2002, 27, 39−85.
(d) Berkefeld, A.; Mecking, S. Angew. Chem., Int. Ed. 2008, 47, 2538−
2542. (e) Nakamura, A.; Ito, S.; Nozaki, K. Chem. Rev. 2009, 109,
5215−5244. (f) Boardman, B. M.; Bazan, G. C. Acc. Chem. Res. 2009,
42, 1597−1606. (g) Takeuchi, D. Dalton Trans. 2010, 39, 311−328.
(2) (a) Chung, T. C. Macromolecules 1988, 21, 865−869. (b) Kesti,
M. R.; Coates, G. W.; Waymouth, R. M. J. Am. Chem. Soc. 1992, 114,
9679−9680. (c) Terao, H.; Ishii, S.; Mitani, M.; Tanaka, H.; Fujita, T.
J. Am. Chem. Soc. 2008, 130, 17636−17637. (d) Yang, X.-H.; Liu, C.-
R.; Wang, C.; Sun, X.-L.; Guo, Y.-H.; Wang, X.-K.; Wang, Z.; Xie, Z.;
Tang, Y. Angew. Chem., Int. Ed. 2009, 48, 8099−8102. (e) Hong, M.;
Wang, Y.-X.; Mu, H.-L.; Li, Y.-S. Organometallics 2011, 30, 4678−
4686. (f) Chen, Z.; Li, J.-F.; Tao, W.-J.; Sun, X.-L.; Yang, X.-H.; Tang,
2008, 41, 8974−8976. (d) Zimmer, S.; Schobel, A.; Halbach, T.;
̈
Stohrer, J.; Rieger, B. Macromol. Rapid Commun. 2013, 34, 221−226.
(17) Johnson, L. K.; McLain, S. J.; Sweetman, K. J.; Wang, Y.;
Bennett, A. M. A.; Wang, L.; McCord, E. F.; Lonkin, A.; Ittel, S. D.;
Radzewich, C. E.; Schiffino, R. S. WO 2003044066, 2003.
I
DOI: 10.1021/jacs.6b10462
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(18) (a) Camacho, D. H.; Salo, E. V.; Ziller, J. W.; Guan, Z. Angew.
Chem., Int. Ed. 2004, 43, 1821−1825. (b) Cherian, A. E.; Rose, J. M.;
Lobkovsky, E. B.; Coates, G. W. J. Am. Chem. Soc. 2005, 127, 13770−
13771. (c) Meinhard, D.; Wegner, M.; Kipiani, G.; Hearley, A.; Reuter,
P.; Fischer, S.; Marti, O.; Rieger, B. J. Am. Chem. Soc. 2007, 129,
9182−9191. (d) Liu, J.; Chen, D.; Wu, H.; Xiao, Z.; Gao, H.; Zhu, F.;
Wu, Q. Macromolecules 2014, 47, 3325−3331. (e) Zou, W.; Chen, C.
Organometallics 2016, 35, 1794−1801.
(19) Johnson, L. K.; Killian, C. M.; Brookhart, M. J. Am. Chem. Soc.
1995, 117, 6414−6415.
(20) Allen, K. E.; Campos, J.; Daugulis, O.; Brookhart, M. ACS Catal.
2015, 5, 456−464.
(21) NMR analysis of these highly branched copolymers prepared
using the “traditional” catalyst t-1a and “sandwich” catalyst s-2a
establish a −CH2Si(OEt)3 (δ 0.56−0.65 ppm) end group and the
structure is in accordance with the copolymer produced by the
palladium catalyst in ref 17.
(22) Similar data of Pd-catalyzed copolymerization of ethylene (100
psig) and VTEoS [0.52 M] at 30 °C are reported in ref 17. The
molecular weight of the copolymer produced is ca. 8.0 × 103 g/mol,
which is comparable with the values reported in Table 1 of this report.
(23) The fact that these chains possess only a single −Si(OEt)3
group means that cross-linking will result in dimerization or 3- or 4-
armed stars, not a traditional crossed-linked network.
(24) As shown in Scheme 10, upon insertion of vinyl silane and prior
to polymer chain growth a distribution of low Mn α-olefins are formed.
Since these low molecular weight olefins are not isolated, their
formation may account for the small decrease in TOFs (based on mass
of isolated polymer) relative to the homopolymerizations. These linear
α-olefins (specifically, 1-hexene and 1-octene) were observed in the
copolymerization described in entry 3, Table 1.
(25) Tempel, D. J.; Johnson, L. K.; Huff, R. L.; White, P. S.;
Brookhart, M. J. Am. Chem. Soc. 2000, 122, 6686−6700.
(26) Small amounts (less than 10%) of an unidentified complex was
also observed during the insertion.
(27) LaPointe, A. M.; Rix, F. C.; Brookhart, M. J. Am. Chem. Soc.
1997, 119, 906−917.
(28) In copolymerization reactions using catalyst s-2a, nearly all silyl
groups in copolymer have a −CH2Si(OEt)3 structure. One mechanism
for this structure was proposed in analogy with the mechanism shown
in Scheme 11. After 1,2-insertion of VTEoS, the PdCH2CH(poly)-
Si(OEt)3 sandwich complex formed must undergo β-silyl elimination
to generate Pd(CH2CHpoly)(Si(OEt)3), followed by 2,1-insertion
to form PdCH(poly)CH2Si(OEt)3 (in analogy with complex t-30,
Scheme 11), which can undergo chain-walking or propagation.
(29) It is assumed that the solubility of ethylene in VTEoS is similar
to that in methylene chloride. The solubility of ethylene in methylene
chloride was determined to be 0.18 mol/L·atm in ref 8.
(30) Popeney, C. S.; Guan, Z. J. Am. Chem. Soc. 2009, 131, 12384−
12393.
(31) It is interesting to note that the kinetic preference of sandwich s-
2b for trapping vinyl trialkoxysilane relative to ethylene is 10 times
greater the thermodynamic preference of the less bulky traditional t-1b
for binding vinyl trialkoxysilane versus ethylene. One possible
explanation for this unexpected difference is that the transition state
for displacement of ether in the kinetic experiment comes early and
does not experience the degree of steric crowding that is felt when
olefins are fully bound.
(32) The β-agostic alkyl intermediates are initially formed from the
insertion of ethylene into the Pd(alkyl)(C2H4) complex under the
polymerization.
(33) It is likely relative trapping rates are smaller when palladium
complexes bearing an alkyl chain substituent larger (i.e., Pd(alkyl))
than methyl (i.e., Pd(CH3)) are trapped.
J
DOI: 10.1021/jacs.6b10462
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX