Investigations into asymmetric post-metallocene group 4 complexes for the synthesis of highly regioirregular polypropylene

Article abstract of DOI:10.1021/ma500453t

A series of asymmetric post-metallocene group 4 complexes based on a modular anilide(pyridine)phenoxide framework have been synthesized and tested for propylene polymerization activity. These complexes, upon activation with methylaluminoxane (MAO), produce highly regioirregular and stereoirregular polypropylene with moderate to good activities. Surprisingly, modification of the anilide R-group substituent from 1-phenethyl to benzyl or adamantyl did not significantly change the polymer microstructure as determined by ¹³C NMR spectroscopy. Although polymer molecular weights and polydispersities vary with propylene pressure, temperature, and activator, regio- and stereoirregularity were also found to be relatively insensitive to these variables. When the polymerization is conducted at 70 °C under dihydrogen, partial decomposition to a highly active catalyst that produces an isotactic microstructure occurs; the undecomposed catalyst continues to produce highly regioirregular and stereoirregular polypropylene under these conditions.

Full text of DOI:10.1021/ma500453t

Article
pubs.acs.org/Macromolecules
Investigations into Asymmetric Post-Metallocene Group 4
Complexes for the Synthesis of Highly Regioirregular Polypropylene
Rachel C. Klet,† Curt N. Theriault,‡ Jerzy Klosin,‡ Jay A. Labinger,*,† and John E. Bercaw*,†
†Arnold and Mabel Beckman Laboratories of Chemical Synthesis, California Institute of Technology, Pasadena, California 91125,
United States
‡Corporate R&D, The Dow Chemical Company, 1776 Building, Midland, Michigan 48674, United States
S
* Supporting Information
ABSTRACT: A series of asymmetric post-metallocene group 4
complexes based on a modular anilide(pyridine)phenoxide
framework have been synthesized and tested for propylene
polymerization activity. These complexes, upon activation with
methylaluminoxane (MAO), produce highly regioirregular and
stereoirregular polypropylene with moderate to good activities.
Surprisingly, modification of the anilide R-group substituent from
1-phenethyl to benzyl or adamantyl did not significantly change
the polymer microstructure as determined by 13C NMR
spectroscopy. Although polymer molecular weights and poly-
dispersities vary with propylene pressure, temperature, and
activator, regio- and stereoirregularity were also found to be
relatively insensitive to these variables. When the polymerization is conducted at 70 °C under dihydrogen, partial decomposition
to a highly active catalyst that produces an isotactic microstructure occurs; the undecomposed catalyst continues to produce
highly regioirregular and stereoirregular polypropylene under these conditions.
Scheme 1. Propylene Polymerization with Group 4
Anilide(Pyridine)Phenoxide (NNO) Complexes Yields
Stereoirregular and Regioirregular Polypropylene
1. INTRODUCTION
Polyolefins are one of the most important classes of synthetic
polymers.1 Decades after the discovery of Ziegler−Natta
catalysts for producing polyethylenes and polypropylenes,2
single-sited homogeneous catalytic systems were described. The
development of metallocene catalysts in the 1980s led to
significant advances in our understanding of how catalyst
structure affects the polymer microstructure.3 More recently,
the rise of “post-metallocene” olefin polymerization catalysts
has led to the realization of new polymer architectures and
significant innovations in living polymerization.4
We have reported group 4 complexes having triaromatic
dianionic (XLX) pincer ligands of the type anilide(pyridine)-
phenoxide (NNO), which upon activation with methylalumi-
noxane (MAO) polymerize propylene with good activities
(Scheme 1).5 These catalysts follow previous work in our group
to develop olefin polymerization catalysts based on triaromatic
dianionic (XLX) pincer ligands, including bis(phenolate)
ligands with pyridine (ONO), furan (OOO) and thiophene
(OSO) linkers, and bis(anilide)pyridine ligands (NNN).6−8
Significantly, the polypropylene produced by asymmetric NNO
complexes is highly regioirregular, while propylene produced by
complexes with symmetric triaromatic ligands is regioregular. In
fact, we have determined that as many as 30−40% of
enchainments may be inverted to a 2,1-insertion mode in
polypropylene produced by NNO complexes. This observed
lack of regiocontrol is very unusual for an early metal
polymerization catalyst and warrants further study to elucidate
its origin. Herein we report our studies aimed at revealing the
origins of the lack of regiocontrol exhibited by asymmetric
group 4 NNO complexes.
2. RESULTS AND DISCUSSION
Anilide R-Group Variations. Our initial ligand design L1
included a chiral 1-phenylethyl group on the anilide arm,
resulting in a C1-symmetric ligand and precatalyst (Figure 1).
The NNO ligand was designed to be easily variable at the
anilide R-group, and given the proximity of this group to the
Received: February 28, 2014
Revised: April 28, 2014
Published: May 14, 2014
© 2014 American Chemical Society
3317
dx.doi.org/10.1021/ma500453t | Macromolecules 2014, 47, 3317−3324

Macromolecules
Article
Figure 1. Ligand L1.
metal center, it was expected to influence the stereochemistry
of incoming α-olefin monomers. We reasoned one potential
origin of regioerrors could be the interactions of the incoming
monomer with the chiral group on the ligand arm. To probe
the effect of this group on regiocontrol, we sought to make Cs-
symmetric ligands having achiral anilide R-groups. Ligands L2
and L3, with benzyl and adamantyl groups, respectively, were
synthesized using synthetic procedures similar to that reported
for the synthesis of L1 (Scheme 2).5
Figure 2. Probability ellipsoid diagram (50%) of the X-ray structure 1.
Selected bond lengths (Å) and angles (deg): Zr(1)−O(1) =
1.9917(7), Zr(1)−N(1) = 2.2911(8), Zr(1)−N(2) = 2.1482(8),
Zr(1)−C(17) = 2.8470(9), Zr(1)−C(33) = 2.2913(10), Zr(1)−
C(34) = 2.5765(9), Zr(1)−C(40) = 2.2851(9); O(1)−Zr(1)−N(2) =
157.17(3), N(1)−Zr(1)−C(33) = 96.19(3), C(33)−Zr(1)−C(40) =
126.48(3), C(40)−Zr(1)−N(1) = 120.71(3), Zr(1)−C(33)−C(34) =
83.53(5), C(17)−N(2)−Zr(1) = 104.95(6).
Scheme 2. Synthesis of NNO Ligand Variants L2 and L3
Figure 3. Different binding modes of NNO ligands in trigonal-
bipyramidal metal complexes depending on the identity of the X-type
ligands.
As has been observed for other early metal dibenzyl
complexes,6,9,10 one of the benzyl groups in 1 strongly interacts
with Zr and is significantly bent toward the metal center to give
a Zr(1)−C(33)−C(34)ipso angle of 83.5° and a short Zr(1)−
C(34)ipso distance of 2.58 Å.
The molecular structure of 3 was also determined by single
crystal X-ray diffraction of crystals grown from slow vapor
diffusion of pentane into a concentrated dichloromethane
solution of 3 (Figure 4). The structure of 3 is very similar to
that obtained for 4 with distorted trigonal-bipyramidal
geometry about titanium and very similar bond lengths and
angles. Similar to 4, 3 appears to have an ipso interaction with a
short Ti(1)−C(17)ipso distance of 2.59 Å and a Ti(1)−N(2)−
C(17)ipso angle of 102.8°.
Activation of complexes (L2)ZrBn2 (1), (L2)TiCl2 (2), and
(L3)TiCl2 (3) with MAO in toluene or chlorobenzene under 5
atm of propylene at 0 °C resulted in formation of
polypropylene (PP). The activity, molecular weight, and
polydispersity index (PDI) of the polymers obtained are
shown in Table 1. Data for the previously reported compounds
(L1)ZrBn2 (5) and (L1)TiCl2 (4) are included for
comparison.5 As was observed previously for complexes
supported by ligand L1, Ti complexes are more active than
their Zr congeners with the NNO ligand system. In comparing
Metalation of the new NNO ligand variants L2 and L3 was
achieved through protonolytic release of either 2 equiv of
toluene with tetrabenzylzirconium or of 2 equiv of dimethyl-
amine with TiCl2(NMe2)2 to yield (L2)ZrBn2 1, (L2)TiCl2 2,
and (L3)TiCl2 3 (Scheme 3).
Crystals of 1 suitable for X-ray diffraction were grown from a
concentrated pentane solution at 35 °C (Figure 2). The crystal
structure of 1 is similar to the previously reported structure of
(L1)TiCl2 4.5 Both complexes have distorted trigonal-
bipyramidal geometry, and the anilide arm is noticeably
distorted out of the O−N(pyridine)−M plane. In the case of
(L1)TiCl2, the anilide and phenoxide arms of the meridional
ligand L1 coordinate in the equatorial plane to put the most π-
donating ligand (Cl) in the axial position to maximize the
potential for π-donation. In contrast, 1 has the anilide and
phenoxide arms in axial positions, since the other ancillary
ligands (benzyl groups) do not participate substantially in
metal−ligand π-bonding (Figure 3).
Scheme 3. Synthesis of Anilide(Pyridine)Phenoxide Zr and Ti Complexes 1−3
3318
dx.doi.org/10.1021/ma500453t | Macromolecules 2014, 47, 3317−3324

Macromolecules
Article
Figure 4. Probability ellipsoid diagram (50%) of the X-ray structure 3.
Selected bond lengths (Å) and angles (deg): Ti(1)−O(1) =
1.8293(11), Ti(1)−N(1) = 2.1989(12), Ti(1)−N(2) = 1.8638(11),
Ti(1)−Cl(1) = 2.3328(6), Ti(1)−Cl(2) = 2.3023(6), Ti(1)−C(17) =
2.5863(14); O(1)−Ti(1)−N(2) = 112.01(5), O(1)−Ti(1)−Cl(2) =
121.10(3), N(2)−Ti(1)−Cl(2) = 123.85(4), Cl(1)−Ti(1)−N(1) =
176.39(3), C(17)−N(2)−Ti(1) = 102.77(8).
Figure 5. 13C NMR spectra of PP from complex 4 (top), 2 (middle),
and 3 (bottom) at 120 °C in TCE-d2.
account for the very similar regio- and stereocontrol for the
different precatalysts. One hypothesis for catalyst modification
that could explain the nearly identical regioselectivity for the Ti
catalysts 2, 3, and 4 is anilide arm dissociation under
polymerization conditions, thus greatly reducing the anilide
R-group influence. Notably, bis(anilide)pyridyl polymerization
catalysts reported by our group have very large PDIs (4.9−
31.2) for propylene polymerization,8 which may result from the
instability of the Ti−N(anilide) linkages under polymerization
conditions; if the Ti−N bonds are susceptible to cleavage,
multiple active species may be obtained leading to a broad
molecular weight distribution and large PDIs. Admittedly, the
NNO polymerization catalysts reported here exhibit narrow
PDIs indicative of primarily one active species (Table 1). One
would thus need to further postulate that, if the Ti−N(anilide)
bonds of the NNO complexes are unstable, the active
polymerization catalysts are somehow stabilized by having a
phenoxide (rather than another anilide) moiety in the ligand
framework.
the Ti catalysts with three different anilide R-groups (1-
phenylethyl (4), benzyl (2), and adamantyl (3)), 4 was
observed to be the most active catalyst and gave the highest
molecular weight polymer, although the differences are small.
Notably, all of the polymers obtained have narrow PDIs (Mw/
Mn), indicating single-site catalysis. The PP from complexes 2
and 3 had no melting points, as expected for stereoirregular PP,
and had similar Tg values to those measured for the PP from
complexes 5 and 4.
13C NMR spectroscopy was carried out on the polypropy-
lenes obtained with precatalysts 1, 2, and 3. We were
particularly interested in comparing the microstructures of
the PPs obtained with the Ti catalysts with three different
anilide R-groups (achiral 2, achiral 3, and chiral 4). Surprisingly,
we observed essentially identical polymer microstructures for all
three PPs, as determined by 13C NMR spectroscopy (Figure 5).
Whereas there are small differences, all three are stereo- and
regioirregular.
These results suggest thatcontrary to our original
hypothesisthe anilide R-group does not affect the stereo-
or regiocontrol of the active polymerization catalyst. This
outcome is particularly puzzling considering the uniquely high
degree of 2,1-enchainments observed; in fact, no other early
metal polymerization catalysts have been reported to produce
polypropylene with such a large degree of regioerrors. Although
it is conceivable that the anilide R-group is in fact not close
enough to the incoming monomer to affect monomer
selectivity in the putative [(NNO)M−polymeryl/propylene]+
transition state for enchainment, we also considered the
possibility that, under the catalytic conditions, all three catalysts
are transformed to very similar active catalytic species to
Orthometalated CNO Polymerization Catalyst. In
considering the possibility of anilide arm dissociationperhaps
facilitated by MAOwe postulated that the arm could remain
uncoordinated or could rotate along the C(aryl)−C(aryl) bond
and possibly C−H activate meta to the C(aryl)−N(anilide)
bond (Scheme 4). Because studying the active catalyst in
solution was not feasible, we sought to synthesize model
complexes that upon activation with MAO would be analogous
to either a labile anilide arm or a C−H-activated anilide arm.
Group 4 orthometalated aryl−pyridine−phenoxide (CNO)
complexes are well-known and, in fact, have been used in
polymerizations of ethylene as well as ethylene/propylene
copolymerizations.11
Table 1. Propylene Polymerization Data for Complexes 1−5
a
precatalyst
precatalyst (μmol)
time (h)
solvent
yield PP (g)
activity (g PP (mol cat.)−1 h−1
)
Tg (°C)
Mw (g/mol)
Mw/Mn
5
4
4
4
1
2
2
2
3
7.6
9.2
1
toluene
PhCl
0.131
0.554
2.412
3.963
0.610
0.384
0.839
2.504
0.589
1.6 × 104
1.2 × 105
2.5 × 105
1.5 × 105
3.8 × 104
8.3 × 104
8.6 × 104
8.4 × 104
5.8 × 104
−8.77
−15.25
−14.40
−12.76
26 000
93 200
1.80
1.50
1.50
1.99
0.5
1
9.6
PhCl
147 000
401 000
9.1
3
PhCl
8.1
1
toluene
PhCl
9.3
0.5
1
−13.66
−13.54
−13.22
−15.36
80 200
133 000
197 000
91 500
1.47
1.55
2.38
1.35
9.8
PhCl
10.0
10.2
3
PhCl
1
PhCl
a
Polymerizations were carried out in 30 mL of liquid propylene with 1000 equiv of dry MAO in 3 mL of solvent at 0 °C for the time indicated.
3319
dx.doi.org/10.1021/ma500453t | Macromolecules 2014, 47, 3317−3324

Macromolecules
Article
Scheme 4. Potential Pathways for NNO Catalyst Modification upon Activation with MAO
Ligand L4 was synthesized as shown in Scheme 5. The 2-
bromo-6-(3,5-di-tert-butyl-2-(methoxymethoxy)phenyl)-
Scheme 5. Synthesis of Ligand L4
Figure 6. Probability ellipsoid diagram (50%) of the X-ray structure 6.
Selected bond lengths (Å) and angles (deg): Ti(1)−O(1) =
1.8649(4), Ti(1)−N(1) = 2.2132(4), Ti(1)−C(17) = 2.1352(5),
Ti(1)−C(27) = 2.1037(6), Ti(1)−C(28) = 2.6385(6), Ti(1)−C(34)
= 2.1135(6); O(1)−Ti(1)−C(17) = 153.81(2), C(27)−Ti(1)−C(34)
= 97.60(3), C(27)−Ti(1)−N(1) = 126.85(2), C(34)−Ti(1)−N(1) =
134.79(2), Ti(1)−C(27)−C(28) = 93.46(4).
pyridine synthon reported previously underwent Suzuki
coupling with commercially available o-tolylboronic acid to
form 2-(3,5-di-tert-butyl-2-(methoxymethoxy)phenyl)-6-(o-
tolyl)pyridine.5 Deprotection with acidic THF afforded the
desired CNO ligand L4.
regioregular PP (Figure 7). This result tentatively suggests
that the NNO complexes do not C−H activate to form CNO
polymerization catalysts.
Metalation of L4 was achieved by reaction with tetrabenzyl-
titanium to yield orthometalated (L4)TiBn2 6 (Scheme 6). An
Scheme 6. Synthesis of Ti Complex 6
Figure 7. 13C NMR spectrum of stereoirregular regioregular PP from
complex 6 at 120 °C in TCE-d2.
To further investigate the possibility of C−H activation, a
solution of 4 in chlorobenzene was activated with 50 equiv of
MAO in the presence of 1-hexene.12 The solution of precatalyst
4, MAO, and 1-hexene was stirred for 20 min and then
quenched with D2O. The organic layer was extracted and
analyzed by 1H NMR spectroscopy, which revealed the
formation of poly(1-hexene) and recovery of the intact ligand
L1 (Scheme 7). If C−H activation had occurred under the
polymerization conditions, deuterium incorporation into the 6-
position of the anilide aryl ring would be expected; however,
the ligand isolated after D2O quenching had only H at this
position (HRMS or 1H NMR). Moreover, the anilide 1-
phenylethyl R-group of the NNO ligand L1 was intact, ruling
out N−C bond cleavage (possibly by MAO) as another
potential pathway for catalyst modification. Finally, monomer
was not incorporated into the isolated ligand, as has been
X-ray quality crystal of 6 was grown from a 5:1 pentane/ether
solution at room temperature, which shows the expected
distorted trigonal-bipyramidal structure and bond lengths and
angles similar to those reported for crystal structures of other
(CNO)TiBn2 complexes (Figure 6).11a Notably, the Ti(1)−
C(27)−C(28)ipso angle for one of the benzyl groups is slightly
distorted at 93.5° and has a shortened Ti(1)−C(28)ipso distance
of 2.64 Å (compare to 123.6° and 3.17 Å), suggesting a weak
η2-ipso interaction between the benzyl group and Ti.
Activation of 6 with MAO in toluene under 5 atm of
propylene at 0 °C yielded PP. The activity of the complex was
measured to be 1.5 × 104 g PP (mol cat.)−1 h−1, an order of
magnitude less active than the NNO-type Ti polymerization
catalysts 2, 3, and 4. Importantly, investigation of the PP from 6
with 13C NMR spectroscopy revealed stereoirregular and
3320
dx.doi.org/10.1021/ma500453t | Macromolecules 2014, 47, 3317−3324

Macromolecules
Article
Scheme 7. Recovery of Intact Ligand L1 after Activation and Polymerization of 1-Hexene with Complex 4
observed for Hf pyridyl−amide catalysts.13 On the basis of
these experiments, we have tentatively ruled out (1) C−H
activation of the anilide arm to form a {(CNO)Ti} complex,
(2) N−C bond cleavage of the anilide R-group, and (3)
monomer insertion into M−ligand bonds to explain the nearly
identical regiocontrol observed for these NNO-type polymer-
ization catalysts.14
Propylene Polymerizations with Different Activators
at Variable Pressures and Temperatures. Our studies with
various post-metallocene polymerization catalysts up to this
point suggest that group 4 complexes with anilide(pyridine)-
phenoxide (NNO) ligands display a lack of regioselectivity for
α-olefin polymerization with the intact NNO ligand coordi-
nated and that this regioselectivity is inherent in the catalyst
structure in a manner not yet defined. We therefore examined
these catalysts systems under different polymerization reaction
conditions to investigate whether temperature, pressure, or
cocatalyst/activator had any effect on regioselectivity.
Titanium complex 4 was tested in a 1 L glass reactor at King
Fahd University of Petroleum and Minerals (KFUPM), which
allowed for carrying out propylene polymerizations at higher
temperatures (room temperature to about 50 °C) and higher
pressures (8−9 atm), compared to the standard Fisher−Porter
setup (0 °C, 5 atm). A polymerization run with 36 mg of
precatalyst 4 at 6 atm and 25 °C using 40 mL of toluene, ca.
100 mL of propylene, 1000 equiv of MAO, and 33 equiv of
triisobutylaluminum (TIBA) yielded very sticky polypropylene
(estimated activity of 106 g PP (mol cat.)−1 h−1). 13C NMR
spectroscopy of the polymer (Figure 8) revealed a somewhat
different, but still largely regioirregular microstructure than that
obtained under the conditions used previously (Table 1). Part
of the reason for this difference could be attributed to the
addition of TIBA in the KFUPM run. An additional run
therefore was carried out with 4 in chlorobenzene using
modified MAO (MMAO) at 0 °C and 5 atm. Indeed, with
MMAO as a cocatalyst, a sticky PP was obtained with activity
determined to be 1.0 × 105 g PP (mol cat.)−1 h−1. 13C NMR
spectroscopy on the PP from the reaction of 4/MMAO
revealed a microstructure very similar to that from the PP
synthesized at KFUPM with 4/MAO/TIBA (Figure 8). These
results suggest that the polymerization reaction is sensitive to
the aluminoxane activator but importantly show that the
regiorandom behavior of catalyst 4 is not significantly affected
by reaction temperatures between 0 and 25 °C.
GPC of the polymer obtained from 4/MAO/TIBA at
KFUPM revealed lower molecular weight (Mw = 4080 g/mol;
Tg = −26.11 °C) compared to the polymers obtained in the
Fisher−Porter setup with the same precatalyst under different
polymerization conditions (Table 1), although the PDI (2.45)
is still rather narrow. The GPC of PP from 4/MMAO run at 0
°C showed a bimodal distribution with a low molecular weight
peak of 3980 g/mol and a high molecular weight peak of
195 400 g/mol. The low molecular weight polymers observed
in polymerizations with 4/MAO/TIBA and 4/MMAO may be
due, at least in part, to free alkylaluminum present in the
system, as aluminum alkyls are known to act as chain-transfer
agents;15 only higher molecular weight PP was obtained when
dry MAO with minimal free trimethylaluminum (TMA) was
used (Table 1).
Ti complex 4 was also tested for propylene polymerization in
a 1.8 L stainless steel batch reactor. Polymerizations were run at
70 °C with 700 g of IsoparE, 150 g of propylene, and 50 psi of
hydrogen for 15 min. PMAO-IP or MAO was used as a
cocatalyst. These polymerizations yielded solid PP with
excellent activities of 2.1 × 106 g PP (mol cat.)−1 h−1 (4/
PMAO-IP) and 9.6 × 105 g PP (mol cat.)−1 h−1 (4/MAO)
(Table 2) and broad molecular weight distributions, Mw/Mn, of
18.6 and 20.9, respectively; however, the GPC traces show
trimodal distributions. Deconvolution of the GPC data for the
PP from 4/PMAO-IP reveals two low-Mw peaks of 320 and
1860 g/mol and a high-Mw peak of 85 880 g/mol. Similarly, the
deconvoluted GPC data for 4/MAO has two low-Mw peaks of
320 and 2360 g/mol and a high-Mw peak of 82 260 g/mol.
Most interestingly, unlike the PP produced by our catalysts
Figure 8. 13C NMR spectra of PP from complex 4/MAO/TIBA run at
25 °C at KFUPM (top), PP from 4/MMAO under conditions used in
Table 1 (middle), and PP from 4/dry MAO under conditions used in
Table 1 (bottom). Spectra were taken at 120 °C in TCE-d2.
Table 2. Propylene Polymerization Data for 4/PMAIO-IP and 4/MAO
a
precatalyst
(mmol)
time
(h)
MAO
(equiv)
PMAO-IP
(equiv)
yield PP
(g)
activity
Tm
(°C)
Mw
(g/mol)
precatalyst
(g PP (mol cat.)−1 h−1
)
Tg (°C)
Mw/Mn
4
4
0.010
0.010
0.25
0.25
10 000
5.3
2.4
2.1 × 106
9.6 × 106
−12.8
−31.0
158.2
155.3
65 600
50 000
18.55
20.86
10 000
a
Polymerizations were carried out with 700 g of IsoparE, 150 g of propylene, and 50 psi of hydrogen at 70 °C for the time indicated.
3321
dx.doi.org/10.1021/ma500453t | Macromolecules 2014, 47, 3317−3324

Macromolecules
Article
under any other condition, the PP produced with 4/PMAO-IP
or 4/MAO had melting points of 158.2 and 155.3 °C, which is
in the range expected for isotactic PP (Table 2).
produce highly regioirregular (and stereoirregular) polypropy-
lene resulting from little apparent preference by these catalysts
for 1,2- or 2,1-insertions of propylene. Significantly, near-
regiorandom behavior is highly unusual for early metal
polymerization catalysts, which typically polymerize propylene
with a very high degree of (normally 1,2-) regiocontrol.
Subjecting the anilide(pyridine)phenoxide catalyst 4 to differ-
ent polymerization conditions, namely higher pressures of
propylene and higher reaction temperatures, revealed that the
catalytically active species that produces regioirregular PP
operates regardless of temperature or pressure but also that at
least one new polymerization species is formed at 70 °C under
dihydrogen, which, surprisingly, produces isotactic PP.
Indeed, 13C NMR spectroscopy on the polymers revealed
peaks indicative of isotactic PP (iPP) as well as peaks for
stereoirregular and regioirregular PP (Figures 9 and 10);
It is our belief that these results point to some form of
catalyst modification upon activation, which at this time
remains to be fully elucidated. Although these experiments
together do not provide a satisfying explanation of the unusual
polymerization behavior of group 4 anilide(pyridine)phenoxide
complexes, they represent a small contribution to our
understanding of the complex behavior of post-metallocene
catalysts. As recently noted by Busico, “the common belief that
“single-site” olefin polymerization catalysis is easily amenable to
rational understanding” does not hold true for post-metallocene
catalysts, and in fact, “it is clear that molecular catalysts are not
necessarily simple nor foreseeable.”17 Nonetheless, these results
importantly show that new discoveries are still possible in
established fields like early metal α-olefin polymerization
catalysis. Continued work in this area will undoubtedly lead
to new breakthroughs in post-metallocene catalysts for olefin
polymerization.
Figure 9. 13C NMR spectrum of PP from 4/PMAO-IP at 115 °C in
TCE-d2. Resonances for iPP are indicated with asterisks.
4. EXPERIMENTAL SECTION
General Considerations and Instrumentation. All air- and
moisture-sensitive compounds were manipulated using standard high-
vacuum and Schlenk techniques or manipulated in a glovebox under a
nitrogen atmosphere. Solvents for air- and moisture-sensitive reactions
were dried over sodium benzophenone ketyl and stored over
“titanocene”18 where compatible or dried by the method of Grubbs.19
TiCl2(NMe)2,20 ZrBn4, HfBn4,21 and (NNO)TiCl2 (4)5 were
prepared following literature procedures. Methylaluminoxane
(MAO) was purchased as a toluene solution from Albemarle and
was dried in vacuo at 150 °C overnight to remove free
trimethylaluminum before use. Propylene was dried by passage
through a column of activated alumina and molecular sieves.
Benzene-d6, toluene-d8, C6D5Cl, and 1,1,2,2-tetrachloroethane-d2
(TCE-d2) were purchased from Cambridge Isotopes. Benzene-d6
and toluene-d8 were dried over sodium benzophenone ketyl then
over titanocene. C6D5Cl was distilled from CaH2 and passed through a
plug of activated alumina prior to use. NMR spectra were recorded on
Varian Mercury 300, Varian INOVA 500, or Varian INOVA 600
spectrometers and referenced to the solvent residual peak. High-
resolution mass spectra (HRMS) were obtained at the California
Institute of Technology Mass Spectral Facility using a JEOL JMS-
600H magnetic sector mass spectrometer. Elemental analyses were
performed by Midwest Microlab LLC (Indianapolis, IN) or Robertson
Microlit Laboratories, Inc. (Ledgewood, NJ). X-ray quality crystals
were grown as indicated in the experimental procedures for each
complex. The crystals were mounted on a glass fiber with Paratone-N
oil. Data collection was carried out on a Bruker KAPPA APEX II
diffractometer with a 0.710 73 Å Mo Kα source. Structures were
determined using direct methods with standard Fourier techniques
using the Bruker AXS software package. In some cases, Patterson maps
were used in place of the direct methods procedure. Some details
regarding crystal data and structure refinement are available in the
Supporting Information. Selected bond lengths and angles are supplied
in the corresponding figures of the paper.
Figure 10. 13C NMR spectrum of PP from 4/MAO at 115 °C in TCE-
d2. Resonances for iPP are indicated with asterisks.
significantly, these results provide the first example of isotactic
PP from an anilide(pyridine)phenoxide type precatalyst.
Related phenoxide(pyridine)amido and phenoxide(thiazole)-
amido precatalysts have been reported that also give iPP under
slightly different polymerization conditions.16 Consistent with
the GPC data, the 13C NMR spectra suggest that more than
one type of polymer was made (presumably by different active
species). Comparison of the 13C NMR spectra for PP from 4/
PMAO-IP or 4/MAO to the PP from 4, 2, or 3 activated with
dry MAO shows identical regioirregular microstructures.
These results seem to indicate that at least one new species is
obtained from 4 under these polymerization conditions, which
polymerizes propylene with high stereo- and regioselectivity to
yield iPP. At the same time, however, the species which was
observed to yield regioirregular and stereoirregular PP at 0 or
22 °C is still active.
3. CONCLUSIONS
A series of asymmetric post-metallocene group 4 complexes
based on a modular anilide(pyridine)phenoxide (NNO)
framework have been synthesized and tested for propylene
polymerization activity. In most cases, the complexes were
found to polymerize propylene upon activation with MAO with
moderate to good activities. Interestingly, these complexes
3322
dx.doi.org/10.1021/ma500453t | Macromolecules 2014, 47, 3317−3324

Macromolecules
Article
(L2)ZrBn2 1. A 2 mL benzene solution of L2-H2 (66.5 mg, 0.143
mmol) was added to a 2 mL benzene solution of ZrBn4 (65.0 mg,
0.143 mmol) and stirred for 10 min under an inert atmosphere in the
glovebox. Benzene was removed in vacuo from the resulting yellow
solution to yield a yellow oil, which was redissolved in pentane and
pumped dry several times to remove residual toluene to reveal a yellow
toluene-d8) δ: 1.37 (s, 9H, C(CH3)3), 1.85 (s, 9H, C(CH3)3), 2.21 (s,
3H, Ar−CH3), 3.88 (d, J = 8.3 Hz, 2H, Ti−CH2), 4.15 (d, J = 8.3 Hz,
2H, Ti−CH2), 6.33−6.44 (m, 2H, aryl−CH), 6.54 (t, J = 7.7 Hz, 4H,
aryl−CH), 6.63−6.71 (m, 4H, aryl−CH), 6.82 (d, J = 4.7 Hz, 2H,
aryl−CH), 7.13 (d, J = 5.4 Hz, 1H, aryl−CH), 7.23 (t, J = 7.1 Hz, 1H,
aryl−CH), 7.37 (d, J = 2.4 Hz, 1H, aryl−CH), 7.69 (d, J = 2.4 Hz, 1H,
aryl−CH), 8.51 (d, J = 6.9 Hz, 1H, aryl−CH). 13C NMR (126 MHz,
C6D6) δ: 23.59 (tolyl-CH3), 30.99 (C(CH3)3), 31.84 (C(CH3)3),
34.66 (C(CH3)3), 35.80 (C(CH3)3), 92.42 (Ti-CH2), 119.61, 121.77,
123.32, 124.72, 125.70, 126.58, 127.75, 128.57, 129.33, 131.13, 132.54,
132.65, 133.00, 136.76, 137.81, 138.66, 142.08, 157.60, 158.15, 165.17,
204.42 (aryl-C). Anal. Calcd for C40H43NOTi (%): C, 79.85; H, 7.20;
N, 2.33. Found (1): C, 74.91; H, 6.99; N, 2.33. (2) C, 74.74; H, 6.86;
N, 2.32. (This compound is air- and moisture-sensitive, and despite
repeated attempts satisfactory %C analysis could not be obtained.)
Recovery of Ligand L1 from Small Scale Polymerization
Reaction with 4 and 1-Hexene. To a 20 mL vial in the glovebox
was added 1 mL of 1-hexene and 50 equiv (0.193 g) of dry MAO. The
1-hexene/MAO solution was stirred for 5 min, then a solution of 4
dissolved in 1 mL of PhCl was added to the vial, and the reaction was
stirred for 25 min at room temperature. The vial was then removed
from the glovebox, and 2 mL of D2O was added slowly, followed by 5
drops of concentrated HCl and 4 mL of D2O, which resulted in
decolorization of the dark red solution. The organic layer was extracted
with hexanes (3 × 4 mL), and the combined organics were dried over
magnesium sulfate and solvent removed in vacuo to reveal a pale
yellow solid. 44.9 mg (L1-D2 and poly(1-hexene)). MS (FAB+) m/z:
calcd for C33H38ON2 [M]+ 478.2984; found 478.3524.
General Polymerization Protocol. A high-pressure glass reactor
was charged with solid MAO (1000 equiv), and 2.3 mL of toluene
(distilled from “Cp2TiH2”) was added. The vessel was attached to a
propylene tank and evacuated, and propylene (∼30 mL) was
condensed in upon cooling to 0 °C. The appropriate precatalysts
was added as a solution (toluene or chlorobenzene, 0.7 mL) via
syringe. The reaction mixture was stirred vigorously at 0 °C for the
desired amount of time, excess propylene was (carefully) vented, and a
10% solution of HCl/MeOH (50 mL) was added slowly to quench the
reaction. The resulting mixture was transferred to an Erlenmeyer flask
and stirred at room temperature overnight. The precipitated polymer
was collected and washed with methanol (3 × 10 mL), evacuated to
remove solvent, further dried under high vacuum for 12 h, and
examined by NMR spectroscopy, GPC, and DSC. 13C NMR spectra
were acquired at 120 °C in tetrachloroethane, using a 2 s relaxation
delay with a 2.3 s acquisition time.
1
powder. (90.8 mg, 0.123 mmol, yield: 86%). H NMR (500 MHz,
toluene-d8) δ: 1.40 (s, 9H, C(CH3)3), 1.64 (s, 9H, C(CH3)3), 2.06 (d,
J = 10.2 Hz, 2H, Zr−CH2), 2.22 (d, J = 10.3 Hz, 2H, Zr−CH2), 4.85
(s, 2H, NCH2Ph), 6.64−6.74 (m, 7H, aryl−CH), 6.77−6.81 (m, 1H,
aryl−CH), 6.80−6.89 (m, 11H, aryl−CH), 7.07 (dd, J = 7.9, 1.6 Hz,
1H, aryl−CH), 7.25 (ddd, J = 8.5, 7.1, 1.6 Hz, 1H, aryl−CH), 7.35
(dd, J = 8.2, 1.2 Hz, 1H, aryl−CH), 7.40 (d, J = 2.3 Hz, 1H, aryl−CH),
7.60 (d, J = 2.2 Hz, 1H, aryl−CH). 13C NMR (126 MHz, toluene-d8)
δ: 30.90 (C(CH3)3), 32.24 (C(CH3)3), 34.97 (C(CH3)3), 36.14
(C(CH3)3), 55.96 (NCH2), 65.65 (ZrCH2), 121.70, 121.92, 122.33,
122.38, 123.75, 124.91, 125.00, 126.94, 127.35, 127.58, 127.92, 128.08,
128.67, 129.59, 131.74, 133.00, 138.49, 138.96, 140.76, 141.64, 142.29,
143.84, 156.24, 156.30, 157.26 (aryl-C). Anal. Calcd for C46H48N2OZr
(%): C, 75.06; H, 6.57; N, 3.81. Found (1): C, 68.19; H, 6.23; N, 3.79.
(2) C, 66.65; H, 6.08; N, 4.24. (This compound is air- and moisture-
sensitive, and despite repeated attempts satisfactory analysis could not
be obtained.)
(L2)TiCl2 2. A 3 mL benzene solution of L2-H2 (60.4 mg, 0.130
mmol) was added to a 3 mL benzene solution of TiCl2(NMe2)2 (26.9
mg, 0.131 mmol) and stirred for 10 min under an inert atmosphere in
the glovebox. Benzene was removed in vacuo from the resulting dark
red solution to yield a deep purple solid, which was triturated several
times with pentane to remove free dimethylamine (77.6 mg, 0.133
1
mmol, quantitative yield). H NMR (300 MHz, C6H5Cl) δ: 1.30 (s,
9H, C(CH3)3), 1.66 (s, 9H, C(CH3)3), 4.95 (s, 1H, NCH2), 6.12 (s,
1H, NCH2), 6.60−6.85 (m, 6H, aryl−CH), 7.02−7.12 (m, 2H, aryl−
CH), 7.17 (dd, J = 7.8, 1.0 Hz, 1H, aryl−CH), 7.25−7.34 (m, 1H,
aryl−CH), 7.48 (t, J = 8.0 Hz, 1H, aryl−CH), 7.56 (dd, J = 8.0, 1.5 Hz,
1H, aryl−CH), 7.63−7.72 (m, 3H, aryl−CH). 13C NMR (126 MHz,
C6D5Cl, −15 °C) δ: 31.41 C(CH3)3), 32.45 (C(CH3)3), 35.69
(C(CH3)3), 36.63 C(CH3)3), 117.29, 123.11, 123.98, 126.02, 128.18,
129.55, 130.64, 132.98, 138.48, 139.58, 139.67, 145.83, 153.52, 155.43,
159.48 (aryl-C). Anal. Calcd for C32H34Cl2N2OTi (%): C, 66.11; H,
5.89; N, 4.82. Found: C, 65.98; H, 6.06; N, 4.87.
(L3)TiCl2 3. A 3 mL benzene solution of L3-H2 (67.6 mg, 0.133
mmol) was added to a 3 mL benzene solution of TiCl2(NMe2)2 (27.5
mg, 0.133 mmol) and stirred for 10 min under inert atmosphere in the
glovebox. Benzene was removed in vacuo from the resulting dark red
solution to yield a light orange solid, which was triturated several times
with pentane to remove free dimethylamine (86.4 mg, 0.134 mmol,
ASSOCIATED CONTENT
■
1
quantitative yield). H NMR (500 MHz, C6D5Cl) δ: 1.30 (s, 9H,
S
* Supporting Information
C(CH3)3), 1.33−1.39 (m, Ad−CH2, 6H), 1.63−1.71 (m, 6H, Ad−
CH2), 1.78 (s, 9H, C(CH3)3), 1.79 (br s, 3H, Ad−CH), 7.24 (dd, J =
7.7, 1.0 Hz, 1H, aryl−CH), 7.36−7.43 (m, 2H, aryl−CH), 7.48−7.55
(m, 2H, aryl−CH), 7.58−7.62 (m, 1H, aryl−CH), 7.71 (q, J = 2.4 Hz,
2H, aryl−CH), 7.77 (dd, J = 8.3, 1.1 Hz, 1H, aryl−CH). 13C NMR
(126 MHz, C6D5Cl) δ: 29.93 (Ad-CH), 30.36 (C(CH3)3), 31.31
(C(CH3)3), 34.60 (C(CH3)3), 35.69 (C(CH3)3), 35.87 (Ad-CH2),
42.62 (Ad-CH2), 69.51 (Ad-quat), 122.09, 122.25, 123.34, 123.72,
127.83, 128.65, 130.31, 131.13, 132.03, 133.17, 134.16, 137.90, 139.15,
144.77, 152.18, 153.39, 158.07 (aryl-C). Anal. Calcd for
C35H42Cl2N2OTi (%): C, 67.21; H, 6.77; N, 4.48. Found (1): C,
66.53; H, 6.80; N, 4.20. (2) C, 66.37; H, 6.73; N, 4.36. (This
compound is air- and moisture-sensitive, and despite repeated
attempts satisfactory %C analysis could not be obtained.)
(L4)TiBn2 6. To a stirring slurry of L4-H2 (30.2 mg, 0.081 mmol) in
5:1 pentane/ether was added to a 3 mL solution of TiBn4 (33.4 mg,
0.081 mmol), and the resulting red solution was stirred for 10 min
under an inert atmosphere in the glovebox. The reaction solution was
passed through a pad of Celite to remove impurities and with 5:1
pentane/ether, and then solvent was removed in vacuo to yield a dark
red solid, which was triturated several times with pentane before being
redissolved in 5:1 pentane/ether and recrystallized by cooling in the
freezer (30.2 mg, 0.050 mmol, 62% yield). 1H NMR (300 MHz,
Experimental details including characterization of new com-
pounds and experimental procedures; crystallographic data.
This material is available free of charge via the Internet at
http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Authors
*E-mail jal@caltech.edu (J.A.L.).
*E-mail bercaw@caltech.edu (J.E.B.).
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The authors gratefully acknowledge the USDOE Office of Basic
Energy Sciences (DE-FG03-85ER13431 to J.E.B.) for funding.
An NSF-GRF to R.C.K. is gratefully acknowledged. We thank
Larry Henling, Dr. Michael Takase, and Dr. Wojciech Bury for
assistance with X-ray crystallography.
3323
dx.doi.org/10.1021/ma500453t | Macromolecules 2014, 47, 3317−3324

Macromolecules
Article
REFERENCES
(1) Hustad, P. D. Science 2009, 325, 704−707.
■
(2) Natta, G.; Pino, P.; Corradini, P.; Danusso, F.; Mantica, E.;
Mazzanti, G.; Moraglio, G. J. Am. Chem. Soc. 1955, 77, 1708−1710.
(3) (a) Brintzinger, H. H.; Fischer, D.; Mulhaupt, R.; Rieger, B.;
̈
Waymouth, R. M. Angew. Chem., Int. Ed. 1995, 34, 1143−1170.
(b) Coates, G. W. Chem. Rev. 2000, 100, 1223−1252. (c) Resconi, L.;
Cavallo, L.; Fait, A.; Piemontesi, F. Chem. Rev. 2000, 100, 1253−1346.
(4) (a) Domski, G. J.; Rose, J. M.; Coates, G. W.; Bolig, A. D.;
Brookhart, M. Prog. Polym. Sci. 2007, 32, 30−92. (b) Anderson-Wile,
A. M.; Edson, J. B.; Coates, G. W. In Complex Macromolecular
Architectures; John Wiley & Sons (Asia) Pte Ltd.: Singapore, 2011; pp
267−316.
(5) Klet, R. C.; VanderVelde, D. G.; Labinger, J. A.; Bercaw, J. E.
Chem. Commun. 2012, 48, 6657−6659.
(6) Agapie, T.; Henling, L. M.; DiPasquale, A. G.; Rheingold, A. L.;
Bercaw, J. E. Organometallics 2008, 27, 6245−6256.
(7) Golisz, S. R.; Bercaw, J. E. Macromolecules 2009, 42, 8751−8762.
(8) Tonks, I. A.; Tofan, D.; Weintrob, E. C.; Agapie, T.; Bercaw, J. E.
Organometallics 2012, 31, 1965−1974.
(9) Bassi, I.; Allegra, G.; Scordamaglia, R.; Chioccola, G. J. Am. Chem.
Soc. 1971, 93, 3787−3788.
(10) Giesbrecht, G. R.; Whitener, G. D.; Arnold, J. Organometallics
2000, 19, 2809−2812.
(11) (a) Chan, M. C.; Kui, S. C.; Cole, J. M.; McIntyre, G. J.; Matsui,
S.; Zhu, N.; Tam, K. H. Chem.Eur. J. 2006, 12, 2607−2619.
(b) Tam, K.-H.; Lo, J. C.; Guo, Z.; Chan, M. C. J. Organomet. Chem.
2007, 692, 4750−4759. (c) Tam, K.-H.; Chan, M. C.; Kaneyoshi, H.;
Makio, H.; Zhu, N. Organometallics 2009, 28, 5877−5882. (d) Liu, C.-
C.; So, L.-C.; Lo, J. C.; Chan, M. C.; Kaneyoshi, H.; Makio, H.
Organometallics 2012, 31, 5274−5281. (e) Lo, J. C.; Chan, M. C.; Lo,
P.-K.; Lau, K.-C.; Ochiai, T.; Makio, H. Organometallics 2013, 32,
449−459.
(12) We have separately demonstrated that 4 polymerizes 1-hexene
to make stereoirregular and regioirregular poly(1-hexene): Klet, R. C.
Ph.D. Thesis, California Institute of Technology, 2013.
(13) (a) Boussie, T. R.; Diamond, G. M.; Goh, C.; Hall, K. A.;
LaPointe, A. M.; Leclerc, M. K.; Murphy, V.; Shoemaker, J. A.; Turner,
H.; Rosen, R. K.; Stevens, J. C.; Alfano, F.; Busico, V.; Cipullo, R.;
Talarico, G. Angew. Chem., Int. Ed. 2006, 45, 3278−3283. (b) Froese,
R. D.; Hustad, P. D.; Kuhlman, R. L.; Wenzel, T. T. J. Am. Chem. Soc.
2007, 129, 7831−7840.
(14) The conclusions reached from the results of these experiments
assume that the majority of the precatalyst is activated with MAO and
participates in catalysis prior to addition of water.
(15) (a) Resconi, L.; Bossi, S.; Abis, L. Macromolecules 1990, 23,
4489−4491. (b) Chien, J. C. W.; Wang, B. P. J. Polym. Sci., Part A:
Polym. Chem. 1988, 26, 3089−3102. (c) Chien, J. C. W.; Razavi, A. J.
Polym. Sci., Part A: Polym. Chem. 1988, 26, 2369−2380. (d) Chien, J.
C. W.; Wang, B. P. J. Polym. Sci., Part A: Polym. Chem. 1990, 28, 15−
38.
(16) The polymerization conditions were as follows: toluene, 100 psi
of propylene, 75 °C, and [PhNMe2H]+[B(C6F5)4]− was used as the
activator in combination with diisobutylaluminum as an alkylator. See:
Diamond, G. M.; Hall, K. A.; LaPointe, A. M.; Leclerc, M. K.;
Longmire, J.; Shoemaker, J. A.; Sun, P. ACS Catal. 2011, 1, 887−900.
(17) Busico, V.; Cipullo, R.; Pellecchia, R.; Rongo, L.; Talarico, G.;
Macchioni, A.; Zuccaccia, C.; Froese, R. D. J.; Hustad, P. D.
Macromolecules 2009, 42, 4369−4373.
(18) Burger, B. J.; Bercaw, J. E. ACS Symp. Ser. 1987, 357, 79−115.
(19) Pangborn, A. B.; Giardello, M. A.; Grubbs, R. H.; Rosen, R. K.;
Timmers, F. J. Organometallics 1996, 15, 1518−1520.
(20) Benzing, E.; Kornicker, W. Chem. Ber. 1961, 94, 2263−2267.
(21) Felten, J. J.; Anderson, W. P. J. Organomet. Chem. 1972, 36, 87−
92.
3324
dx.doi.org/10.1021/ma500453t | Macromolecules 2014, 47, 3317−3324