Copolymerizations of ethylene and α-olefins with supported piano-stool catalysts

Article abstract of DOI:10.1016/j.poly.2005.05.001

Recent developments in single-site catalysis (SSC) have led to many interesting discoveries in ethylene and α-olefin copolymerizations. Here, we report the synthesis of, and polymerizations with, supported piano-stool catalysts and their related analogues. CpTiCl₃, Cp*TiCl ₃, TMSCpTiCl₃, and (TMS)₂CpTiCl₃ were covalently bound to a poly(styrene-r-4-hydroxystyrene) copolymer, affording macroligated catalysts. Ethylene/1-octene and ethylene/styrene copolymerizations with each of these supported catalysts were attempted with cocatalysts including methyl aluminoxane, B(C₆F₅) ₃, [Ph₃C][B(C₆F₅)₄], and [PhN(CH₃)₂H][B(C₆F₅)₄]. To determine the effects that the macroligand and the substituents on the cyclopentadienyl ring have on polymer yield and activity, comonomer incorporation, polymer composition, molecular weight, polydispersity index, melting point, and crystallization behavior were investigated. We discovered that the catalysts have vastly different, yet interesting, effects on ethylene/1-octene (EO) and ethylene/styrene (ES) copolymerizations. For ethylene/1-octene copolymers, the activity of the supported catalysts and the amount of 1-octene incorporated into the copolymer is dependent upon the substituents on the cyclopentadienyl ring. The formation of EO copolymers using supported aryloxy-based cyclopentadienyl complexes can also be used as additional evidence for the presence of the tethered aryloxy moiety. The copolymer composition and the thermal properties of these copolymers were also studied. Ethylene/styrene copolymerizations produced polymers with varying characteristics specific to the catalyst employed for the polymerizations. Polymers synthesized with the supported CpTiCl₃ and TMSCpTiCl ₃ displayed both polyethylene and syndiotactic polystyrene melting points. Conversely, polymers synthesized with the supported Cp*TiCl ₃ displayed only a polyethylene melting point, while polymers prepared with the supported (TMS)₂CpTiCl₃ displayed only a syndiotactic polystyrene melting point. These findings suggest that copolymerization of ethylene and styrene is not possible with these catalysts. Analysis of the homopolymerizations of styrene and ethylene with Cp*TiCl₃ and (TMS)₂CpTiCl₃, respectively, shows significantly lower activities, leading to preferential polymerization of the more active monomer during copolymerization.

Full text of DOI:10.1016/j.poly.2005.05.001

Polyhedron 24 (2005) 1347–1355
www.elsevier.com/locate/poly
Copolymerizations of ethylene and a-olefins with supported
piano-stool catalysts
*
Jeanette F. Craymer, Rajeswari M. Kasi, E. Bryan Coughlin
Department of Polymer Science and Engineering, University of Massachusetts Amherst, 120 Governors Drive, Amherst, MA 01003-4530, USA
Received 19 January 2005; accepted 12 May 2005
Available online 18 July 2005
Abstract
Recent developments in single-site catalysis (SSC) have led to many interesting discoveries in ethylene and a-olefin copolymer-
izations. Here, we report the synthesis of, and polymerizations with, supported piano-stool catalysts and their related analogues.
CpTiCl3, Cp*TiCl3, TMSCpTiCl3, and (TMS)2CpTiCl3 were covalently bound to a poly(styrene-r-4-hydroxystyrene) copolymer,
affording macroligated catalysts. Ethylene/1-octene and ethylene/styrene copolymerizations with each of these supported catalysts
were attempted with cocatalysts including methyl aluminoxane, B(C6F5)3, [Ph3C][B(C6F5)4], and [PhN(CH3)2H][B(C6F5)4]. To
determine the effects that the macroligand and the substituents on the cyclopentadienyl ring have on polymer yield and activity,
comonomer incorporation, polymer composition, molecular weight, polydispersity index, melting point, and crystallization behav-
ior were investigated.
We discovered that the catalysts have vastly different, yet interesting, effects on ethylene/1-octene (EO) and ethylene/styrene (ES)
copolymerizations. For ethylene/1-octene copolymers, the activity of the supported catalysts and the amount of 1-octene incorpo-
rated into the copolymer is dependent upon the substituents on the cyclopentadienyl ring. The formation of EO copolymers using
supported aryloxy-based cyclopentadienyl complexes can also be used as additional evidence for the presence of the tethered aryloxy
moiety. The copolymer composition and the thermal properties of these copolymers were also studied.
Ethylene/styrene copolymerizations produced polymers with varying characteristics specific to the catalyst employed for the
polymerizations. Polymers synthesized with the supported CpTiCl3 and TMSCpTiCl3 displayed both polyethylene and syndiotactic
polystyrene melting points. Conversely, polymers synthesized with the supported Cp*TiCl3 displayed only a polyethylene melting
point, while polymers prepared with the supported (TMS)2CpTiCl3 displayed only a syndiotactic polystyrene melting point. These
findings suggest that copolymerization of ethylene and styrene is not possible with these catalysts. Analysis of the homopolymer-
izations of styrene and ethylene with Cp*TiCl3 and (TMS)2CpTiCl3, respectively, shows significantly lower activities, leading to
preferential polymerization of the more active monomer during copolymerization.
ꢀ 2005 Elsevier Ltd. All rights reserved.
Keywords: Piano-stool catalysts; Single-site titanium catalyst; Ethylene/octene copolymer; Ethylene/styrene copolymer
1. Introduction
lytic (SSC) architectures [1–6]. For example, ethylene/
a-olefin copolymers (LLDPE) or ethylene/styrene (ES)
copolymers are produced using piano-stool and modi-
fied piano stool complexes, with uniform comonomer
incorporation, high molecular weight, and good activity.
These copolymers with alkyl or aryl branches have bet-
ter melt-flow properties, toughness and optical proper-
ties compared to high-density polyethylene (HDPE)
[7]. The introduction of aromatic groups in the ES
The copolymerization of ethylene and various como-
nomers (1-butene, 1-hexene, 1-octene, and styrene) is
made possible by the discovery of new single-site cata-
*
Corresponding author. Tel.: +1 413 577 1616; fax: +1 413 545
0082.
E-mail address: Coughlin@mail.pse.umass.edu (E.B. Coughlin).
0277-5387/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.poly.2005.05.001

1348
J.F. Craymer et al. / Polyhedron 24 (2005) 1347–1355
copolymer results in significant increase in viscoelastic,
mechanical properties and compatibility of ES copoly-
mers with LLDPE [7]. Additionally, ES copolymers con-
tain syndiotactic styrene dyads, which enhance the
thermal properties, as syndiotactic polystyrene is known
to have a Tm of 273 ꢁC.
Other discoveries put forth regarding piano-stool cat-
alysts include the effect that substitution of alkyl groups
on the cyclopentadienyl ring and of the halides on the
titanium catalyst center for various substituents has on
the copolymerization of ethylene and a-olefins. Most
significant to this work are NomuraÕs publications on
the aryloxy substituted CpTi(R)2OAr complexes [8–
13]. The authors found that by substituting one of the
halide groups on the Ti catalyst center with an aryloxy
group, polyethylene, poly(ethylene-co-1-butene), and
poly(ethylene-co-styrene) could be obtained with rea-
sonable activity, molecular weight, polydispersity, and
comonomer distribution [8–13].
All solvents used were obtained from a solvent manifold
[19]. Titanium tetrachloride, bis(trimethylsilyl)cyclo-
pentadiene, 2,5,5-tris(trimethylsilyl)-1,3-cyclopentadiene,
cyclopentadienyltitanium trichloride, pentamethyl-cyclo-
pentadienyltitanium trichloride, styrene, 4-acetoxy
styrene, AIBN, tetrahydrofuran, and hydrazine hydrate
were obtained from Aldrich. AIBN was recrystallized
from acetone, and styrene was distilled over calcium
hydride. Methyl aluminoxane (MAO, 30 wt% (4.7 M)
in toluene) was generously supplied by Albemarle. The
copolymerizations were performed using heavy-walled
glass reactors and a polymerization manifold [20].
The styrene-co-hydroxy styrene (HPS, macroligand)
and the supported aryloxy cyclopentadienyltitanium
trichloride complexes were synthesized as previously
reported [18].
2.2. Instrumentation
There is a significant commercial drive to combine
the inherent advantages of SSC catalysts (high activity,
control over monomer reactivity and polymer structure
control) with the attributes of supported catalyst tech-
nology (morphology and high bulk density control)
[14]. With this in mind, piano-stool catalysts have been
supported on a variety of substrates including, but not
limited to, lightly cross-linked polystyrene beads [15],
alumina or MgCl2 [14], butyl ethylmagnesium treated
silica [14], mesoporous silica [16], and (4-hydroxysty-
rene)-styrene (HPS) copolymers [17].
In this work, various piano-stool catalysts were re-
acted with a (4-hydroxystyrene) styrene copolymer
(HPS), in essence creating bulky aryloxy substituents
(macroligand). We have previously shown that the pres-
ence of the macroligand is vital for the production of bu-
tyl branched polyethylene produced during ethylene
homopolymerization [18]. More importantly, the sup-
port should be regarded as a part of the ligand–metal
environment. Hence, chemical and structural changes
to the support affect the overall performance of the pia-
no-stool complex bound to the HPS support [18]. In this
paper, the use of macroligated catalysts for the copoly-
merization of ethylene with 1-octene and styrene has
been explored. The effect of using the macroligated cat-
alysts with appropriate cocatalysts for ethylene/a-olefin
copolymerization on the molecular weight, branch con-
tent, melting temperature, degree of crystallinity, and
copolymer microstructure has been analyzed in detail.
13C NMR spectra of the polyolefins in tetrachloreth-
ylene-d2 were recorded at 150 MHz on an Avance DPX
600 using a quantitative pulse program for solution
NMR at 135 ꢁC. Gel permeation chromatography
(GPC) of the polyolefins was performed with a Polymer
Laboratories PL-220 high temperature GPC equipped
with a Wyatt MiniDawn (620 nm diode laser) high tem-
perature light scattering detector and refractive index
detector. Experiments were performed at 135 ꢁC using
1,2,4-trichlorobenzene as solvent, and results were based
on calibration with polystyrene standards. The dn/dc
value for polyethylene used was À0.11. Thermal analy-
ses were performed using a TA differential scanning cal-
orimeter (DSC) 2910 equipped with liquid nitrogen
cooling accessory unit. Experiments were performed un-
der a continuous nitrogen purge (50 mL/min) with a
heating and cooling scan rate of 10 ꢁC/min.
3. General procedure for supported catalyst preparation
(HPS-Ti-1 to -4)
Synthesis of the support matrix poly(styrene-r-4-
hydroxystyrene (HPS)) and subsequent assembly of
the aryloxy piano stool complex on HPS was carried
out in the glovebox using precursor catalysts (1–4) as
previously reported [18,21]. The support contains 0.3
mol% hydroxystyrene, as determined by 1H NMR.
The supported catalysts were analyzed via elemental
analysis in order to determine the weight percent of
Ti, as well as the ratio of Cl:Ti. For all cases, it was
determined that the Cl:Ti ratio was between 1 and 2,
meaning that the most prevalent forms of the supported
catalysts were Form A and Form B (Fig. 1). Although a
Form C (containing three bonds to the macroligand) is
not ruled out, the authors believe that this form would
be sterically hindered. The reaction of the homogeneous
2. Experimental
2.1. Reagents and general techniques
All reactions were performed under nitrogen in
Schlenk glassware or in an inert atmosphere dry box.

J.F. Craymer et al. / Polyhedron 24 (2005) 1347–1355
1349
R2
R2
R1
R1
R5
Cl
R3
R4
R3
R4
R5
Form B
Ti
O
1. Monomers
2. Cocatalysts
Ti
Cl
Cl
O
O
R
R
Form
A
R = Hexyl
R = Phenyl
HPS-Ti-1, R1-5=H
HPS-Ti-2, R1-5=CH3
HPS-Ti-3, R1=Si(CH3)3, R2-R5=H
HPS-Ti-4, R1=R3=Si(CH3)3, R2=R4=R5=H
Fig. 1. Synthesis of ethylene/a-olefin random copolymers using supported aryloxy piano stool complexes.
catalysts with the support was observed to form a gel,
5. Synthesis of bis(phenoxy)(trimethylsilyl-
(cyclopentadienyl)) titanium(IV) chloride (6)
which further supports the crosslinked structure out-
lined in Fig. 1. Independent confirmation of the pro-
posed supported structure was obtained via 1H and
13C NMR. The 1H NMR displays a downfield shift
for the trimethylsilyl protons from d 0.2 for the precur-
sor of HPS-Ti-4 to d 0.4 ppm for HPS-Ti-4. Further
confirmation of the presence of Forms A and B was ob-
tained from the 13C NMR of HPS-Ti-4, which displays
two peaks in the trimethylsilyl region at d À0.1
(Form A) and d 1.5 ppm (Form B).
To a solution of 10.0 mL of toluene and (trimethylsi-
lyl)cyclopentadienyltitanium trichloride (3, 102 mg, 0.34
mmol), phenol (75 mg, 0.7 mmol) was added slowly with
constant stirring, and the reaction mixture was refluxed
for a period of 1 day. The solvent was removed in vacuo,
affording a red solid in 90% yield.
1H NMR (CDCl3): 0.3 (s, 9H, Si(CH3)3), 6.6–7.4
(m,14H, C6H5, C5H4).
13C NMR (CDCl3, CPD): 0.0, 115.7, 118.9, 121.5,
123.2, 126.3, 130.1, 133.7, 167.6.
4. Synthesis of mono(phenoxy)(trimethylsilylcyclo-
pentadienyl) titanium(IV) dichloride (5)
6. Copolymerization of ethylene and 1-octene
To a mixture of 10.0 mL of dichoromethane and
(trimethylsilyl)cyclopentadienyltitanium trichloride (3,
102 mg, 0.34 mmol), phenol (32 mg, 0.35 mmol) was
added slowly with constant stirring, The reaction was al-
lowed to proceed at room temperature for a period of
12 h. The solvent was removed in vacuo, affording an
orange solid (110 mg, 82% yield).
To a mixture of 10.0 mL of toluene and 1.7 mL of
30 wt% of MAO in toluene in a polymerization reactor,
38.2 mg of HPS-Ti-1 was added. To this system, 1-oc-
tene (5 mL) was added, and ethylene (60 psig) was sup-
plied for 15 min with stirring. The ethylene supply was
stopped and the reactor was vented to remove excess
ethylene, the reaction mixture was quenched with 10%
hydrochloric acid–methanol mixture and the polymer
was collected by filtration and dried in vacuo overnight
resulting in ethylene/1-octene copolymer (Table 1,
1H NMR (CDCl3): 0.4 (s, 9H, Si(CH3)3), 6.8 (t, 2H),
7.1 (t, 2H), 7.0–7.4 (m, 5H, C6H5).
13C NMR (CDCl3, CPD): À0.2, 118.8, 124.6, 123.9,
126.9, 128.6, 129.8, 137.5, 168.3.
Table 1
Ethylene/1-octene copolymerization using macroligated catalysts in conjunction with MAO
1-Octene incorporation (mol%)a,b
Mw (·10À3
)
PDIc
a
c
Entry
Catalysts
Al:Ti
Activity (kg/mol Ti h atm)
Tm (ꢁC)
1
2
3
4
5
6
7
a
HPS-Ti-1
HPS-Ti-2
HPS-Ti-3
511
665
1710
1613
660
48
17
42
10
5
116.4
119.8
105.1
123.3
134.9
114.1
115.3
2.5
2.5
3.8
2.0
268
177
407
501
n.d
400
110
broad
2.4
2.2
broad
n.d.
2.0
HPS-Ti-4
3d
5
780
765
60
71
3.0
3.0
6
1.9
DSC first scan.
Calculated from depression in melting point [26] and 13C NMR measurements.
Determined through GPC at 135 ꢁC in TCB using light scattering detectors calibrated with PS standards.
b
c
d
(Trimethylsilyl)cyclopentadienyltitanium(IV) chloride.

1350
J.F. Craymer et al. / Polyhedron 24 (2005) 1347–1355
Table 2
entry 1). For copolymerizations catalyzed with an
unsupported catalyst, conditions were identical to the
supported catalyst systems, however, only 3–5 mg of
the catalyst were used (Table 1, entries 5–7).
Homopolymerization of styrene
Entry Catalyst Activity
(kg/mol Ti h) (ꢁC) (J/g) (·10À3
a
Tm
DH
Mw
PDI
)
b
8
9
HPS-Ti-1
962
6
261
15
1.5
14
2.1
17
HPS-Ti-2
260
271
132
2.31
7. Homopolymerization of styrene
10
11
HPS-Ti-3 1946
260
271
135
52
2.34
2.35
To 50 mL round bottom flasks equipped with mag-
netic stirbars, 10.5 mL of distilled toluene, 25–30 mg
of the supported catalyst, and 1.7 mL of MAO were
added. (Al:Ti ratios for HPS-Ti-1 and HPS-Ti-2 were
600, while HPS-Ti-3 was 1000 and HPS-Ti-4 was
1400. The amount of MAO was increased for HPS-Ti-
3 and -4, as we knew that the homogeneous precursors
were much less active.) After sealing with septa, the
charged flasks were removed from the glovebox and
placed on magnetic stirrers. At zero minutes reaction
time, 3.0 mL (0.026 mol) of distilled styrene (which
had been loaded in syringes in the glovebox) were added
to each flask. If gelation did not occur within a few min-
utes, the reaction was allowed to proceed for 50–60 min.
The reaction mixture was quenched with 10% hydro-
chloric acid–methanol mixture, and the polymer was
collected by filtration and dried in vacuo overnight.
9.2
HPS-Ti-4
114
260
271
4.0
17
a
Second scan.
Chromatogram is multimodal.
b
Table 3
Copolymerization of ethylene and styrene results
a
Entry Catalyst
Activity
Tm DH Mw
PDI
(kg/mol Ti h atm) (ꢁC) (J/g) (·10À3
)
12
HPS-Ti-1 26.6
114 30.4 412
247
2.10b
2.2
13
14
HPS-Ti-2 18.0
HPS-Ti-3 4.6
117 38.7 341
2.42b
2.22c
128 24.8 478
255
259
6.2
15
HPS-Ti-4 2.7
2.2 340
1.95c
270 11.3
a
b
c
Second scan.
Chromatogram is multimodal.
Chromatogram contains a low molecular weight shoulder.
8. Copolymerization of ethylene and styrene
To a 12 fluid ounce glass reactor, 75 mL of distilled
toluene, 25–30 mg of supported catalyst, and 1.7 mL
of MAO were added. (Al:Ti ratios were the same as
for the styrene polymerization.) The reactor was then fit-
ted with the mechanical stirring apparatus and reactor
head. To a 25 mL side-port on the reactor head,
21.5 mL (0.188 mol) of distilled styrene were added.
The entire closed system was then removed from the
glove box and attached to a medium pressure line and
mechanical stirrer [20]. After flushing the system, the
reactor was pressurized to 60 psig with ethylene gas.
At this point, the styrene was added by back-pressuring
the side port of the reactor with nitrogen. The nitrogen
line was then detached from the reactor, and the excess
nitrogen in the reactor was vented out of the side port.
The reactor was then repressurized with ethylene, and
the reaction was allowed to proceed for 1.5 h. The reac-
tion mixture was quenched with 10% hydrochloric acid–
methanol mixture and the polymer was collected by
filtration and dried in vacuo overnight.
support matrix on the catalyst performance and the
properties of the resulting polyolefin [18]. Additionally,
we studied how the substituents on the cyclopentadienyl
ring affect both the amount of comonomer incorporated
into the copolymer and also the microstructure of the
copolymer. These effects on polymer yield, activity,
weight average molecular weight (Mw), polydispersity
index (PDI), melting point (Tm), and percent comono-
mer incorporation are outlined in Tables 1–3.
9.1. Ethylene/1-octene copolymers
During ethylene homopolymerizations, the supported
aryloxy complexes (HPS-Ti-1 to -4), in conjunction
with different cocatalysts (MAO, tris(pentafluorophe-
nyl)borane, dimethylanilinium borate, trityl borate), af-
ford linear or branched polyethylenes from an ethylene
only feed. The production of butyl branched polyethyl-
ene from an ethylene only feed has been attributed to
1-hexene produced from an ethylene trimerization pro-
cess triggered by the use of macroligated catalysts [18].
In order to determine the effect of the macroligand on
the properties of LLDPE, ethylene and 1-octene copoly-
merizations were performed using the macroligated cat-
alysts (HPS-Ti-1–4) and MAO as the cocatalyst. It was
9. Results and discussions
We have designed synthetic schemes to tether piano-
stool complexes on hydroxy-functionalized polystyrene
supports (HPS) in order to study the influence of the

J.F. Craymer et al. / Polyhedron 24 (2005) 1347–1355
1351
observed that for the ethylene/1-octene copolymers,
HPS-Ti-3 and HPS-Ti-1 show higher activity
than HPS-Ti-2 and HPS-Ti-4. Also, HPS-Ti-3 affords
higher levels of 1-octene incorporation compared to the
other three macroligated catalysts. The copolymers
synthesized from these supported catalysts are of
moderately high molecular weights (Table 1, entries
1–4).
Another interesting feature is that the formation of
ethylene/octene-based copolymers using supported aryl-
oxy-based cyclopentadienyl complexes can also be used
as additional evidence for the presence of the tethered
aryloxy moiety. A previous control polymerization per-
formed with complex 3, MAO as the cocatalyst, and eth-
ylene and 1-octene in the monomer feed produced
HDPE (Table 2, entry 5). On the other hand, the use
of small molecular analogues 5 and 6 with MAO re-
sulted in the formation of ethylene/1-octene copolymers
(LLDPE) from ethylene and 1-octene copolymerization
(Table 1, entries 6 and 7, respectively).
For 1-octene incorporation, the trend followed HPS-
Ti-3 > HPS-Ti-2 > HPS-Ti-4. The trend in the molecu-
lar weight of the ethylene/1-octene copolymer was found
to be HPS-Ti-4 > HPS-Ti-3 > HPS-Ti-1 > HPS-Ti-2.
The trends observed in the activity and the molecular
weight of the copolymer produced from the macroli-
gated catalysts agree well with the results obtained by
Nomura and co-workers. These authors have shown
that in the case of non-bridged cyclopentadienyl aryloxy
9.2. Copolymer microstructure analysis of ethylene/octene
copolymers obtained from macroligated catalysts
In order to determine the copolymer composition of
the copolymers synthesized, 13C NMR of the copoly-
mers was performed at 135 ꢁC. The 13C NMR reso-
nances of the copolymer synthesized from complex
HPS-Ti-1 were analyzed. The spectrum is similar to that
reported by Cavagna [24,25], and hence low level of oc-
tene incorporations were expected. The nomenclature
for carbon assignments, monomer sequence distribution
and equations relating signal intensity to the sequences
as reported by Randall were used to identify the
branches in 13C NMR of the polyolefin [25]. The 13C
peak resonances were assigned as 14.19 (1s + 1B6,
CH3, EOE + EOO + OOE + OOO), 22.85 (2s + 2B6,
t
titanium species like BuCp, (CH3)5Cp, and (tBu)2Cp,
the Cp–Ti–O bond angle changes from 117.6ꢁ, 120.5ꢁ
and 119.3ꢁ [9,22,23]. For 1-hexene incorporation, the
activity of the substituents has been shown to follow
t
the trend: Bu Cp > (CH3)5Cp > (tBu)2Cp. On the other
hand, the molecular weight for the copolymer is highest
for (tBu)2Cp followed by the tBu and the (CH3)5Cp ana-
logues. Nomura and co-workers [9,22,23] have attrib-
uted the observed trends in activity and molecular
weight of the ethylene/hexene copolymer produced to
the rotational flexibility of the subsituents on the cyclo-
pentadienyl ring.
Ti
Cl
Cl
Ti
O
O
O
Cl
+ MAO + Ethylene + 1-Octene
γδ
αδ
αδ
γδ
γδ
αδ
αδ
γδ
δδ
δδ
δδ
βδ
6B6
δδ
βδ
βδ
βδ
4B4
5B6
3B4
4B6
2B6
2B4
29.99
3B6
1B6
+ +
δ δ
14.19
1s+1B6
1B4
34.42
αδ+ 6Β6
1-Octene=1.5 mol %
1-Hexene=1.0 mol%
30.01
δδ+4B6
34.04
4B4
22.85
2s+2B6
32.12
3s+3B6
27.10
βδ
27.20
βδ + 5B6
36
34
32
30
28
26
24
(ppm)
22
20
18
16
14
Fig. 2. 13C NMR composite pulse decoupling (CPD) spectrum of ethylene/octene copolymer synthesized from complex HPS-Ti-1 with MAO as the
cocatalyst.

1352
J.F. Craymer et al. / Polyhedron 24 (2005) 1347–1355
EOE + EOO + OOE + OOO), 27.10 (bd, OOEE +
EEOO), 27.20 (bd + 5B6, EOE), 29.99 (dd, EEE),
30.01 (dd + 4B6, EEE), 32.12 (3s + 3B6, EOE), 34.04
(4B4, EHE), 34.42 (ad + 6B6, EOEE + EEOE), and
38.01 (CH, EOO + OOE, methine), as shown in
Fig. 2. A quantitative relationship between signal inten-
sity to monomer sequences was established to quantify
the 1-octene incorporated in the copolymer [25]. The
1-octene content, calculated from the ratio of the sum
of the octene-centered sequences to that of all the se-
quences, was determined to be 1.5 mol%. The presence
of a 4B4 branch (butyl branch) was also observed. This
can be attributed to the 1-hexene formed by ethylene tri-
merization and further incorporation of 1-hexene into
the copolymer. The resonances 4B4 and (ad + 6B6) were
used to quantify the ratio of hexene (1.0 mol%) and oc-
tene (1.5 mol%) incorporated into the ethylene chain.
The microstructure of ethylene/octene copolymer
obtained from complexes HPS-Ti-2 and HPS-Ti-3
were also analyzed and assigned as d 14.22 (1s + 1B6,
EOE + EOO + OOE + OOO), 22.87 (2s + 2B6, EOE +
EOO + OOE + OOO), 26.95 (bd, OOEE + EEOO),
27.05 (bd + 5B6, EOE), 29.97 (dd, EEE), 32.21 (3s +
3B6, EOE), 34.51 (ad + 6B6, EOEE + EEOE), and
38.04 (CH, methine). The a-olefin incorporation in the
copolymers obtained from HPS-Ti-2 and HPS-Ti-3
were estimated to be 2.5 and 3.8 mol%, respectively, as
established by 13C NMR. These values are corroborated
by the observed depression in melting point. The micro-
structure of ethylene/octene copolymers from the
macroligated catalysts (HPS-Ti-1 to 3) is similar to the
microstructure of copolymers obtained from non-
bridged cyclopentadienyl aryloxy complexes [23].
does it have uniform molecular weight. This may be
due to leaching of the precursor catalyst from the sup-
ported catalyst, HPS-Ti-1. As the precursor to this sup-
ported catalyst (CpTiCl3) is known to produce
syndiotactic polystyrene in high yield and purity, the
macroligand form of the catalyst displays an obvious
and significant effect on the polymer properties.
Xu et al. [17] reported similar styrene polymeriza-
tion results with CpTiCl3, which had been supported
onto a polystyrene terpolymer containing hydroxy
and methoxy moieties. While our reaction conditions
are comparable, it is apparent that the melting point
data that we obtained for entry 8 corresponds to the
production of polystyrene with higher syndiotactic con-
tent. This further supports our argument that the use
of macroligated catalysts is an additional tuning
parameter in dictating the characteristics of the result-
ing polymer.
Additionally, entry 9 displays that the polystyrene
produced with HPS-Ti-2 was obtained with activity that
was two to three orders of magnitude lower than the rest
of the catalysts. It is widely known and accepted that the
precursor, Cp*TiCl3, produces syndiotactic polystyrene
in high yield and activity. This phenomenon further
illustrates the effect that the macroligand imparts on
the catalystÕs ability to polymerize styrene. Polymeriza-
tions with the borane cocatalyst system produced very
similar results with respect to catalyst activity, polymer
yield, and melting behavior, thereby ensuring this result
is neither an anomaly nor dependent upon the cocatalyst
system employed.
Polymerization of ethylene and styrene with the sup-
ported catalysts HPS-Ti-1 to -4 and MAO catalyst sys-
tem produced very interesting results, Table 3. As with
the styrene homopolymerization, very similar results
were obtained with the borane cocatalyst system. It is
evident that activity shows a marked decrease with the
introduction of the ethylene comonomer. Although no
trend is evident from yield and activity results with re-
spect to cyclopentyl substituents, it is apparent from
the DSC analysis that each supported catalyst produced
very different polymers. In entries 12 and 14, the data is
consistent with the presence of a reactor blend of poly-
ethylene and syndiotactic polystyrene. Entry 13, how-
ever, suggests that only polyethylene is produced. A
degree of crystallinity of 13.4% was calculated by divid-
ing the heat of fusion value for the polyethylene melting
by the total heat of fusion for a perfect polyethylene
crystal (293 J/g) [26]. Finally, data in entry 15 suggest
that only syndiotactic polystyrene is produced. This
can also be seen in the DSC chromatogram (Fig. 3).
Analysis of the homopolymerizations of ethylene and
styrene sheds some light on the polymer composition of
entries 13 and 15. For the homopolymerization of sty-
rene, HPS-Ti-2 is clearly a poor performer, as the activ-
ity of this polymerization is two to three orders lower
9.3. Ethylene/styrene copolymers
Homopolymerizations of styrene with the supported
catalysts HPS-Ti-1 to -4 and assorted cocatalysts
(MAO, tris(pentafluorophenyl)borane, dimethylanilin-
ium borate, trityl borate) were performed in order to
determine any differences between the catalysts and to
discern if they would yield syndiotactic polystyrene as
do the unsupported analogues. Table 2 shows that these
polymerizations did indeed yield syndiotactic polysty-
rene with varying activities, molecular weights and poly-
dispersities. Overall, polymerizations cocatalyzed with
MAO display higher yields than those polymerized with
a borane as the cocatalyst. The DSC results show that
runs with HPS-Ti-2, -3, and -4 contain the expected
melting temperature at 270 ꢁC, which is associated with
syndiotactic polystyrene. The polystyrene obtained with
HPS-Ti-1, however, lacks the melting transition around
270 ꢁC. Also, the GPC data indicates significantly lower
molecular weight and a multimodal distribution. The
data indicate that the polystyrene produced with cata-
lyst HPS-Ti-1 is neither completely syndiotactic nor

J.F. Craymer et al. / Polyhedron 24 (2005) 1347–1355
1353
12
13
14
15
Fig. 3. Differential scanning calorimetry chromatograms of entries 12–15 (before extraction).
than that of the other catalysts employed. For this rea-
son, it is expected that HPS-Ti-2 would preferentially
polymerize ethylene in copolymerization conditions with
styrene. Indeed, the DSC chromatogram of the polymer
obtained with HPS-Ti-2 in the presence of ethylene and
styrene monomers is consistent with the production of
polyethylene.
Similarly, HPS-Ti-4 displays a remarkably lower
activity for the homopolymerization of ethylene than
the other catalysts employed in this study [18]. Using
this information in conjunction with the styrene homo-
polymerization results from Table 2, it is possible to pre-
dict that HPS-Ti-4 would only polymerize styrene in
copolymerization conditions with ethylene. As pre-
dicted, the entry 15 data and the DSC chromatogram
display a preferential polymerization of styrene.
HPS-Ti-1 and HPS-Ti-3 display sufficiently high
activities in homopolymerizations of ethylene and sty-
rene, and therefore it is conceivable that they would pro-
duce either a copolymer or a reactor blend of two
homopolymers. Evidence of such a blend of homopoly-
mers was detected through DSC and NMR analyses.
Solvent
peak:
TCE
Pentad analysis
δ
A3-A5
44.3
A2
*
*
A4
y
x
A1
B1
A3
A5
36.9
27.6
*
*
*
*
B2
B3
B2
B3
B4
*
*
B2,
B3
25.0
B1
B4
A2
A1
140
130
120
110
100
90
80
70
(ppm)
60
50
40
30
20
10
0
Fig. 4. 13C NMR of entry 14.

1354
J.F. Craymer et al. / Polyhedron 24 (2005) 1347–1355
Sequential Soxhlet extractions in methyl ethyl ketone
microstructure of the copolymers was determined using
13C NMR and shows random incorporation of the oc-
tene comonomer. The formation of ethylene/1-octene
copolymers using supported aryloxy-based cyclopenta-
dienyl complexes can also be used as additional evidence
for the presence of the tethered aryloxy moiety, since the
precursor piano-stool complex affords HDPE during at-
tempted copolymerization. (Due to the incorporation of
the comonomer, the melting points (Tm) and the enthal-
py of fusion (DH) of the copolymers is depressed as
determined by DSC.)
We have also explored the effect of supported single-
site catalysts on styrene and ethylene/styrene polymer-
izations. Polymerizations with each of the supported
catalysts, (HPS-Ti-1 to -4) were carried out with MAO
and borane cocatalysts. The MAO and borane cocata-
lyst systems produced very similar results.
DSC analysis displayed that HPS-Ti-1 produced
polystyrene with lower syndiotactic content than the
precursor catalyst, indicating the presence of a strong
macroligand effect. DSC analysis also showed that poly-
mer composition varies greatly with the specific sup-
ported catalyst. Polymerizations of ethylene and
styrene via HPS-Ti-1 and HPS-Ti-3 display both poly-
ethylene and syndiotactic polystyrene melting behavior,
while polymerizations of ethylene and styrene via HPS-
Ti-2 and HPS-Ti-4 produce melting temperatures con-
sistent with polyethylene or syndiotactic polystyrene,
respectively. Based on homopolymerization data with
styrene and ethylene, the phenomenon of different cata-
lysts producing vastly different polymer compositions in
copolymerization conditions was proscribed to the fact
that certain supported catalysts preferentially polymer-
ize ethylene or styrene.
(MEK) and tetrahydrofuran (THF) were performed on
the entry 12 sample, in order to determine the relative
amounts of atactic polystyrene, random ethylene/sty-
rene copolymer, and polyethylene that were present.
These results indicate that there is a significant amount
of atactic polystyrene (MEK-soluble, >50%), random
ES copolymer (THF-soluble, 11%), and polyethylene
(THF-insoluble, 32%, Tm = 132 ꢁC). Entry 14 was ana-
lyzed via 13C NMR, as insufficient yield did not allow
for Soxhlet extraction. As shown in Fig. 4, significant
resonances in the NMR were consistent with syndiotac-
tic polystyrene and polyethylene. Peaks consistent with
the presence of an ES copolymer are of very low inten-
sity, suggesting that an insignificant amount of the
sample consists of random ES copolymer. The NMR
and DSC data of entry 14 indicate the presence of
two homopolymers: syndiotactic polystyrene and
polyethylene.
Due to insufficient yield quantities, Soxhlet extrac-
tions could not be performed on entries 13 or 15. While
DSC results of entry 15 are consistent with syndiotactic
polystyrene, DSC results of entry 13 cannot conclusively
determine whether or not a significant portion of the
sample contains atactic polystyrene.
In summary, there seems to be a notable macroligand
effect on the homopolymerization of styrene and the
copolymerization of ethylene and styrene. This is clearly
demonstrated in the decrease of the melting tempera-
tures for the polystyrene. For the copolymerizations, it
is evident that although the variety of substituents on
the cyclopentadienyl ring does not have a correlation
with the polymer composition, there is clearly another
effect that causes these catalysts to produce different
copolymer compositions. This effect is demonstrated
through the supported catalysts behavior in homopoly-
merizations of styrene and ethylene. As the results are
mirrored with the borane cocatalyst system, we believe
that they are true results and not merely anomalies.
Acknowledgements
This material is based on work supported by the Na-
tional Science Foundation under Grant No. CHE-
0113643, and the Army Research Laboratory Polymer
Materials Center of Excellence (DAAD19-01-2-0002
P00005). Central analytical facilities utilized used in
these studies were supported by the NSF-sponsored
Materials and Research Science and Engineering Center
on Polymers at UMass Amherst (DMR-0213695).
10. Conclusions
In order to study the influence of the support matrix
on the catalyst performance and the properties of the
resulting polyolefin, macroligated aryloxy titanium com-
plexes (HPS-Ti) were prepared. In this paper, the effect
of the macroligand on ethylene and a-olefin (1-octene or
styrene) copolymerization using aryloxycyclopentadie-
nyl macroligated catalysts were studied.
References
[1] R.W. Barnhart, G.C. Bazan, T. Mourey, J. Am. Chem. Soc. 120
(1998) 1082.
[2] S.D. Ittel, L.K. Johnson, M. Brookhart, Chem. Rev. 100 (2000)
1169.
[3] A.L. McKnight, R.M. Waymouth, Chem. Rev. 98 (1998) 2587.
[4] G. Melillo, L. Izzo, M. Zinna, C. Tedesco, L. Oliva, Macromol-
ecules 35 (2002) 9256.
LLDPE has been synthesized from ethylene and 1-oc-
tene feed using supported aryloxycyclopentadienyltita-
nium complexes. The activity of the supported
catalysts for copolymerization and the amount of 1-oc-
tene incorporated into the copolymer is dependent upon
the substituents on the cyclopentadienyl ring. The

J.F. Craymer et al. / Polyhedron 24 (2005) 1347–1355
1355
[5] C. Pellecchia, D. Pappalardo, G.J. Gruter, Macromolecules 32
(1999) 4491.
[16] K. Kageyama, J. Tamazawa, T. Aida, Science 285 (1999)
2113.
[6] G.J.P. Britovsek, V.C. Gibson, D.F. Wass, Angew. Chem. Int.
Ed. 38 (1999) 428.
[17] J. Xu, J. Ouyang, Z. Fan, D. Chen, L. Feng, J. Polym. Sci. A 38
(2000) 127.
[7] P.S. Chum, W.J. Kruper, M.J. Guest, Adv. Mater. 12 (2000)
1759.
[8] K. Nomura, N. Naga, M. Miki, K. Yanagi, A. Imai, Organo-
metallics 17 (1998) 2152.
[9] K. Nomura, N. Naga, M. Miki, K. Yanagi, Macromolecules 31
(1998) 7588.
[10] K. Nomura, K. Oya, T. Komatsu, Y. Imanishi, Macromolecules
33 (2000) 3187.
[11] K. Nomura, T. Komatsu, Y. Imanishi, Macromolecules 33 (2000)
8122.
[12] K. Nomura, K. Oya, Y. Imanishi, J. Mol. Catal. A 17 (2001) 127.
[13] K. Nomura, H. Okumura, T. Komatsu, N. Naga, Macromole-
cules 35 (2002) 5388.
[14] G.G. Hlatky, Chem. Rev. 100 (2000) 1347.
[15] S.B. Roscoe, J.M.J. Frechet, J.F. Walzer, A.J. Dias, Science 280
(1998) 270.
[18] R.M. Kasi, E.B. Coughlin, Macromolecules 36 (2003) 6300.
[19] A.B. Pangborn, M.A. Giardello, R.H. Grubbs, R.K. Rosen, F.J.
Timmers, Organometallics 15 (1996) 1518.
[20] G.S. Constable, R.A. Gonzalez-Ruiz, R.M. Kasi, E.B. Coughlin,
Macromolecules 35 (2002) 9613.
[21] J.M. Nasrullah, S. Raja, K. Vijayakumaran, R. Dhamodharan, J.
Polym. Sci. Polym. Chem. 38 (2000) 453.
[22] K. Nomura, T. Komatsu, M. Nakamura, Y. Imanishi, J. Mol.
Catal. A 164 (2000) 131.
[23] K. Nomura, T. Komatsu, Y. Imanishi, J. Mol. Catal. A 152
(2000) 249.
[24] F. Cavagna, Macromolecules 14 (1981) 215.
[25] J.C. Randall, J. Macromol. Sci.-Rev. Macromol. Chem. Phys.
C29 (1989) 201.
[26] H.Y. Chen, S.P. Chum, A. Hiltner, E. Baer, J. Polym. Sci. B 39
(2001) 1578.