Efficient slurry-phase homopolymerization of ethylene to branched polyethylenes using α-diimine nickel(II) catalysts covalently linked to silica supports

Article abstract of DOI:10.1021/ma061032v

α-Diimine nickel(II) complexes covalently linked to a silica support were obtained by the reaction of an amino- or hydroxy-functionalized α-diimine nickel complex with silica particles treated with trimethyl aluminum. These silica-supported aryl α-diimine nickel(II) precatalysts linked via an amino (1a) or hydroxy functionality (1b-c) at the p-aryl position, and precatalysts linked via one (2a) or two (2b) hydroxy functionalities in the alkyl backbone were used in slurry polymerizations of ethylene to produce branched polyethylenes. Activation of precatalysts with ethylaluminum sesquichloride was effective even at very low Al/Ni ratios of 48. Initial polymerizations, performed at 60°C and 150 psig ethylene, showed excellent productivities of up to 10.8 kg PE per gram of 1b (3.0 wt % nickel loading). Low productivities at temperatures above 60°C can in part be overcome by increase in ethylene pressure. For example, a productivity of 6.0 kg PE per gram of 1b (3.0 wt% nickel loading) was obtained at 80°C and 700 psig ethylene. The best polymerization productivities were obtained with trimethylalumlnum as a traceless silica-linker, however, tetrachlorosilane and trichloroborane are attractive alternatives. Branching densities can be controlled by choice of ortho-aryl substituents. No reactor fouling occurs with these supported catalysts.

Full text of DOI:10.1021/ma061032v

Macromolecules 2006, 39, 6341-6354
6341
Efficient Slurry-Phase Homopolymerization of Ethylene to Branched
Polyethylenes Using R-Diimine Nickel(II) Catalysts Covalently Linked
to Silica Supports
Henri S. Schrekker,*,† Vasily Kotov, Peter Preishuber-Pflugl, Peter White, and
Maurice Brookhart*
Department of Chemistry, UniVersity of North Carolina at Chapel Hill,
Chapel Hill, North Carolina 27599-3290
ReceiVed May 8, 2006; ReVised Manuscript ReceiVed July 17, 2006
ABSTRACT: R-Diimine nickel(II) complexes covalently linked to a silica support were obtained by the reaction
of an amino- or hydroxy-functionalized R-diimine nickel complex with silica particles treated with trimethyl
aluminum. These silica-supported aryl R-diimine nickel(II) precatalysts linked via an amino (1a) or hydroxy
functionality (1b-c) at the p-aryl position, and precatalysts linked via one (2a) or two (2b) hydroxy functionalities
in the alkyl backbone were used in slurry polymerizations of ethylene to produce branched polyethylenes. Activation
of precatalysts with ethylaluminum sesquichloride was effective even at very low Al/Ni ratios of 48. Initial
polymerizations, performed at 60 °C and 150 psig ethylene, showed excellent productivities of up to 10.8 kg PE
per gram of 1b (3.0 wt % nickel loading). Low productivities at temperatures above 60 °C can in part be overcome
by increase in ethylene pressure. For example, a productivity of 6.0 kg PE per gram of 1b (3.0 wt % nickel
loading) was obtained at 80 °C and 700 psig ethylene. The best polymerization productivities were obtained with
trimethylaluminum as a traceless silica-linker, however, tetrachlorosilane and trichloroborane are attractive
alternatives. Branching densities can be controlled by choice of ortho-aryl substituents. No reactor fouling occurs
with these supported catalysts.
Introduction
The demand for polyethylenes (PEs) is enormous due to
numerous end applications in bulk products.1-3 Transformation
of the PE industry into a multibillion dollar a year business4,5
started with the discovery of heterogeneous titanium catalysts
by Ziegler in the mid 1950s.6-8 Production of PEs is still
dominated by coordination-insertion catalysts based on early
transition metals. Coordination polymerization technologies of
supported titanium (Ziegler-Natta)8-10 and chromium (Phillips
and Union Carbide)11,12 catalysts are used for the production
of high-density polyethylene (HDPE), a linear semicrystalline
ethylene homopolymer (Tm ∼ 136 °C).13 Linear low-density
polyethylenes (LLDPEs) are random copolymers of ethylene
and an R-olefin exhibiting a range of Tm’s below 130 °C13
prepared using Ziegler-Natta,8 chromium,11 or metallocene
catalysts.4,14-17 An alternative technique for the synthesis of
LLDPE combines an ethylene oligomerization catalyst with a
catalyst that copolymerizes the in situ produced R-olefins with
ethylene.7,18-20 Such a dual catalyst system was described by
Phillips for the production of a hexene-ethylene copolymer.21
A disadvantage of this general approach is that R-olefin residues
can be present in the copolymer due to an incomplete incorpora-
tion. Low-density polyethylene (LDPE, Tm ) 100-115 °C)13
is a homopolymer exhibiting branches of variable length, which
is produced via a radical chain growth process at high temper-
atures and high ethylene pressures (15 000-45 000 psig ethylene
at 120-300 °C).22
Figure 1. General structure of R-diimine nickel(II) catalysts.
In general, ethylene insertion rates for late transition metal
catalysts are lower than those of early transition metal cata-
lysts23,24 and â-hydride elimination leading to chain transfer
competes with chain growth, resulting in the formation of short-
chain oligomers.25 This chemistry forms the basis of the Shell
higher olefin process (SHOP) for production of R-olefins from
ethylene using neutral nickel catalysts containing bidentate
oxygen-phosphorus ligands.26-28 Certain nickel(II) complexes
containing modified SHOP ligands show moderate activity for
forming linear PE.18
A breakthrough in Ni-catalyzed ethylene polymerization
chemistry came in 1995 when this group reported the discovery
of ortho-substituted aryl R-diimine nickel(II) and palladium-
(II) catalysts (Figure 1) for the synthesis of high-molecular-
weight branched PE.18,24,29-33 Following this discovery, col-
laborative efforts of the UNC and DuPont groups led to
expansion of this catalyst family, which is now part of the
Versipol polymerization catalyst technology from DuPont.34
* Corresponding authors. E-mail: schrekker@iq.ufrgs.br (H.S.S.);
mbrookhart@unc.edu (M.B.). Telephone: +55-51-3316-6284 (H.S.S);
+1-919-962-0362 (M.B.). Fax: +55-51-3316-7304 (H.S.S.); +1-919-962-
2476 (M.B.).
In the nickel and palladium diimine systems, the aryl groups
are oriented perpendicular to the ligand coordination plane, and
introduction of substituents in the ortho-aryl positions provides
steric bulk above and below the square planar metal center and
† Current address: Institute of Chemistry, Universidade Federal do Rio
Grande do Sul, Av. Bento Gonc¸alves 9500, Porto Alegre, RS, CEP 91501-
970, Brazil.
10.1021/ma061032v CCC: $33.50 © 2006 American Chemical Society
Published on Web 08/25/2006

6342 Schrekker et al.
Macromolecules, Vol. 39, No. 19, 2006
Scheme 1. Chain Walking (P ) growing polymer chain).18,29,45-47
thus retards chain transfer (Scheme 1).29,30 Chain growth is now
favored over chain transfer, which results in a dramatic change
from low- to high-molecular-weight PE.29 A similar ligand
design was incorporated into the bis(imino) pyridyl iron and
cobalt catalysts, which yield high-molecular-weight PE.18,35-39
Ethylene polymerization activity of the nickel catalysts is
extremely high as a consequence of the highly electrophilic
nature of the cationic nickel center, which lowers the migratory
insertion barrier of ethylene into the growing polymer chain.29,40
Cationic palladium(II) analogues are much less reactive due to
their reduced electrophilicity.18,29 The bis(imino) pyridyl iron
catalysts are interesting due to their thermal stability, and high
activities are obtained even at temperatures above 100 °C.35
However, ethylene polymerization with these catalysts affords
HDPE.35,41 A unique feature of the R-diimine nickel(II) and
palladium(II) catalysts is that branched PE is obtained in
ethylene homopolymerizations.18,29 Highly branched, completely
amorphous PE is normally obtained with the palladium cata-
lysts.29,42 R-Diimine nickel(II) catalysts produce PE with
branching levels of 5-100 branches per 1000 carbons depending
on ligand structure and reaction conditions, which corresponds
to melting points between 25 and 135 °C.29,30 The types of
branches obtained vary from mainly methyl to ethyl and higher
aliphatic chains, and branches on branches are observed.18,43,44
A “chain-walking” mechanism, which involves sequential
â-hydride elimination and reinsertion reactions, is responsible
for the branching (Scheme 1).18,29,45-47 In general, the PE
microstructure can be predictably modified by altering the steric
bulk of the ortho-aryl substituents, ethylene pressure, and
reaction temperature.30,48 Branching densities increase with an
increase in steric bulk of the ortho substituents, with an increase
in temperature and with a decrease in pressure. These unique
properties of the R-diimine nickel(II) catalysts make this
technology a potentially attractive alternative for the production
of LDPE. Branching levels go far beyond what is possible in
the free radical polymerizations, thus a wider range of PE
properties are available using these catalysts.
particles are frequently chosen as the support material due to
their excellent fragmentation properties, which allows growth
of the particles together with good control of the exotherm.8,49-51
A typical approach for generating the supported catalyst
involves, first, treatment of the silica particles with methyl
aluminoxane (MAO) to form surface-bound MAO, followed by
treatment of these particles with an early metal precatalyst.52
The MAO serves to convert the precatalyst to the active cationic
metal alkyl complex, which is anchored to the particle through
the ionic attraction to the surface-bound anions (electrostatic
interactions). Under ideal circumstances, as ethylene is taken
up, these particles grow and fragment; the morphology of the
original particle is reproduced and no reactor fouling occurs.
An attractive feature of this method is that no modification of
the catalyst is required for “attachment” to the support.
On the basis of patent disclosures and publications, translation
of this technique to the R-diimine Ni(II) catalysts is not entirely
satisfactory. Heterogenized catalysts were obtained by addition
of Ni(II) diimine complexes to MAO-treated silica53-55 or,
alternatively, by reaction of the diimine complexes with MAO
prior to addition to the silica support (Figure 2).56 The following
limitations of this technique were observed: (1) a limited loading
capacity of the nickel catalysts, (2) agglomeration of PE on the
reactor wall due to leaching of R-diimine nickel(II) catalysts
from the silica support, and (3) significantly reduced produc-
tivities when compared to the homogeneous polymerizations
(Figure 2). When performed under the same polymerization
conditions, the productivity per gram of nickel decreased from
250-450 kg PE for the homogeneous polymerization to 50 kg
PE for the silica-MAO supported catalysts. Furthermore, the
productivity per gram of supported catalyst is quite low, which
can in part be ascribed to catalyst deactivation as well as low
loading capacity of the silica using this technique. Large molar
excesses of MAO are used to support and activate the R-diimine
nickel(II) catalysts. Excess MAO might be responsible for the
reduction and, as a consequence, deactivation of the catalyst.57,58
However, little is known about the deactivation of these
catalysts.
Gas-phase and slurry-phase polymerization processes em-
ploying titanium or chromium catalysts are the principal
commercial methods for production of polyethylenes via metal
catalysis.8-12 Using nickel diimine catalysts in either of these
commercial processes presents challenges, in large part due to
the deactivation of these catalysts at high temperatures.18
Because slurry-phase polymerizations can be run at moderate
temperatures of 70-90 °C, this process is more attractive for
use with the nickel catalysts. Ideally a “drop-in” catalyst system
might be developed which could be utilized in existing slurry
facilities.
Covalent attachment of the R-diimine nickel(II) precatalyst
to a support would offer a choice of the alkyl aluminum
cocatalyst to be used and better control of the cocatalyst amount
(Figure 2). One way to achieve this is to first support the
R-diimine ligand and then treat this system with NiBr2(1,2-
dimethoxyethane) to form the supported R-diimine nickel(II)
dibromide precatalyst.59-61 Boussie used this method in which
diimine ligands were bound to a polystyrene support for the
discovery of new R-diimine nickel(II) catalysts in a combina-
torial fashion.62 Another example employing this method
involved the covalent attachment of diimines to silica via a
siloxane linkage.63 A general disadvantage of this strategy is
that the formation of the nickel(II) complex is not always
A common approach to developing catalysts for use in slurry
processes is to heterogenize effective homogeneous catalysts
by immobilizing them on solid supports.8,49-51 Mesoporous silica

Macromolecules, Vol. 39, No. 19, 2006
Slurry-Phase Homopolymerization of Ethylene 6343
Figure 2. Comparison of homogeneous vs heterogenized R-diimine nickel(II) catalysts.
Scheme 2. General Approach for the Synthesis of Silica-Supported r-Diimine Nickel(II) Catalysts via a Covalent Attachment.
Scheme 3. Synthesis of r-Diimine Ligands with an Amino or
straightforward and will be even more complicated when
Hydroxy Functionality Attached to the Para-Aryl Position.
performed on a solid support. The determination of precise
nature of the supported catalysts is problematic.
Formation of the R-diimine nickel(II) complexes prior
to attachment to the support circumvents these problems and
allows a reliable strategy for the synthesis of supported
R-diimine nickel(II) catalysts. Such a strategy has pre-
viously been used for supporting SHOP-type oligomeriza-
tion catalysts.64 Recently, Herrmann reported bis(imino) pyridyl
iron(II) ethylene polymerization catalysts covalently linked to
silica prepared via hydrosilation of the alkenyl-functionalized
iron(II) complexes by a functionalized silica support.65 In work
patented by Repsol,66 Mendez Llatas et al., reported attachment
of [ArNdC(R1)-C(R2)dNAr]NiBr2 (Ar ) 2,6-Me2C6H3- or
2,6-i-Pr2C6H3-, R1 ) -CH3 or -CH2CH2CH2Si(OCH3)3,
R2 ) -CH2CH2CH2Si(OCH3)3) to mesoporous silica particles
via Si-O linkages formed from elimination of MeOH. When
activated with MAO, these catalysts exhibited low productivi-
ties in the range of 0.01-0.09 kg PE per gram of supported
catalyst at 45-60 °C. Previously, we reported a new methodol-
ogy for the covalent attachment of R-diimine nickel(II) com-
plexes to a silica support (Scheme 2).67,68 R-Diimine nickel(II)
complexes which contain an amino or hydroxy functionality
were reacted with trimethylaluminum (TMA)-treated silica to
for the production of LDPE and related lower-density polyeth-
afford the covalently linked catalysts. Activation of these
ylenes.
precatalysts with an alkylaluminum cocatalyst afforded a highly
productive ethylene polymerization system. Note that nearly the
same productivity per gram of nickel is observed as in the
Results and Discussion
homogeneous system, and an excellent productivity of 4.5 kg
PE per gram of supported precatalyst is obtained (Figure 2).
Herein we report our results regarding optimization of (1) the
polymerization procedure, (2) the R-diimine nickel(II) catalysts,
(3) the silica-linker, (4) the polymerization temperature, (5)
the ethylene pressure, and (6) the nickel load of the R-diimine
nickel(II) complex on the silica support in slurry polymeriza-
tions of ethylene. These detailed investigations were performed
to obtain better productivities at elevated temperatures of 80
°C or higher and to realize control of the polymer properties
with the goal of developing this technology to be competitive
Ligand Synthesis. The highly active silica-supported aryl
R-diimine nickel(II) catalysts previously reported by us con-
tained either an amino or hydroxy functionality at the para-
aryl position, which allowed a covalent attachment to the silica
support.67 Synthesis of the corresponding aryl R-diimine ligands
is achieved by the reaction of the functionalized aniline with
acenaphthenequinone (Scheme 3). In our initial studies, we
examined ortho,ortho-dimethyl-substituted anilines, as these are
easily accessible, which allowed the comparison of various
amino- and hydroxyl-substituted ligand systems. The sulfuric
acid-catalyzed condensation reaction of aniline 6a with acenaph-

6344 Schrekker et al.
Macromolecules, Vol. 39, No. 19, 2006
Scheme 4. Synthesis of r-Diimine Ligands with a Hydroxy
Functionality Attached to the Backbone.
thenequinone afforded amino-functionalized R-diimine 8a in a
moderate yield of 62%. Introduction of a hydroxyl functionality
was accomplished starting from bromoaniline 5. A palladium-
catalyzed Heck coupling with methylacrylate, followed by a
lithium aluminum hydride reduction of the ester and hydrogena-
tion with palladium on activated carbon, afforded hydroxy-
functionalized aniline 6b in an overall yield of 22%. This
strategy allows facile variation of the ortho-aryl substituents,
and diisopropylaniline 6c was synthesized in an overall yield
of 18% from aniline 3. The acid-catalyzed condensation reaction
of anilines 6b and 6c with acenaphthenequinone afforded the
hydroxy-functionalized R-diimine ligands 8b and 8c in 66 and
48% unoptimized yields, respectively.69
The synthesis of R-diimine ligands with a hydroxy function-
ality attached to the backbone is presented in Scheme 4.
Attachment of one or two hydroxyl tails to the dimethyl
backbone of R-diimine 9 is easily achieved by a proper choice
of the reaction conditions. Treatment of R-diimine 9 with lithium
diisopropylamine and 1 equivalent TMS-protected iodohexanol
introduces the first hydroxy-tail; the second one can be
introduced by a consecutive addition of lithium diisopropylamine
and another equivalent of TMS-protected iodohexanol. Depro-
tection after the first or second step with tetrabutylammonium
fluoride affords either monohydroxy R-diimine 10a or dihydroxy
R-diimine 10b in yields of 82 or 65%, respectively.
Synthesis NiBr2 Complexes. A previously described pro-
cedure was used for the synthesis of R-diimine nickel(II)
bromide complexes 11a-c and 12a,b (Scheme 5).30,67 No
protective groups for the amino and hydroxy functionalities are
needed, and reaction of the R-diimine ligands with NiBr2(1,2-
dimethoxyethane) resulted generally in excellent yields. Yields
of complexes 11c and 12b were somewhat low due to good
solubility of these complexes in the solvents used for purifica-
tion. Suitable crystals for X-ray analysis were obtained for
R-diimine nickel(II) bromide 12a (Figure 3). The X-ray analysis
confirms the assigned structure of 12a. As expected for the
dibromide, tetrahedral coordination of Ni(II) is observed. The
aryl groups lie perpendicular to the plane of R-diimine ligand,
and the ortho-methyl substituents are located above and below
this plane.
Figure 3. ORTEP view of R-diimine nickel(II) bromide 12a. Selected
interatomic distances (Å): Ni(1)-Br(1) 2.331, Ni(1)-Br(2) 2.338, Ni-
(1)-N(1) 2.021, N(1)-C(2) 1.304, C(2)-C(3) 1.477, C(3)-N(4) 1.360,
Ni(1)-N(4) 1.991. Selected angles (deg): Br(1)-Ni(1)-Br(2) 121.31,
N(1)-Ni(1)-N(4) 80.07, Ni(1)-N(1)-C(2) 115.03, N(1)-C(2)-C(3)
116.25, C(2)-C(3)-N(4) 112.70, C(3)-N(4)-Ni(1) 115.83, C(2)-
N(1--C(5) 123.38, C(3)-N(4)-C(22) 121.67. Selected torsion angles
(deg): C(2)-N(1)-C(5)-C(6) 94.31, C(2)-N(1)-C(5)-C(10) -86.97.
with a linker (Scheme 6). Grace Davison silicas XPO-2402 and
SP9-496 were treated with an excess of TMA, or SiCl4, or BCl3,
which results either in the release of methane or hydrogen
chloride. Reaction of the R-diimine nickel(II) complexes 11a-c
and 12a,b with these activated silica supports 14 afforded the
desired covalent attachment with release of methane with
support 14a or hydrogen chloride with the supports 14b,c
(Scheme 6). It is unknown whether one or both of the
functionalities are linked to the silica support in the case of
precatalysts 1a-c and 2b. High nickel loadings of up to 3.0 wt
% were achieved with this strategy in quantitative yields for
the R-diimine nickel(II) precatalysts 1a-c and 2a,b.
Determination of the nickel loading on the silica-supported
precatalyst is important for determining the productivity per
gram of nickel. Values reported in our previous communication
were found by an ICP-OES analysis of the silica-supported
precatalyst.67 Since this report, we have discovered that this
analytical technique resulted consistently in low estimates of
nickel loading. For instance, preparation of silica-supported
precatalyst 1b1 with a 1.0 wt % nickel loading by the treatment
of activated silica 14a with a dichloromethane solution of
R-diimine nickel(II) complex 11b resulted in complete decol-
oration of the brown dichloromethane solution while the silica
particles acquired a brown color. However, a nickel loading of
0.6 wt % was found by ICP-OES analysis. In this work, nickel
loadings were determined by analysis of the nickel content left
in the dichloromethane solution following supportation as
determined by ICP-OES. The nickel loadings determined were
in all instances only slightly below the theoretical maximum
nickel loading. (If some insoluble (ligand) NiBr2 remains in the
dichloromethane solution and is filtered off, a high estimate of
Synthesis of Precatalysts. Covalent attachment of R-diimine
nickel(II) complexes requires the treatment of the silica support
Scheme 5. Synthesis of r-Diimine NiBr2 complexes.

Macromolecules, Vol. 39, No. 19, 2006
Slurry-Phase Homopolymerization of Ethylene 6345
Scheme 6. Preparation of Silica-Supported r-Diimine Nickel(II) Precatalysts.
Table 1. Ethylene Slurry Polymerizations Employing Precatalysts 1b1 and 1b2
f
T
[°C]
time
[h]
kg PE/
g 1b1/2
kg PE/
g Ni
Tm
Mw
branches/
1000Cg
entry
precata
wt % Nib
cocatc
Al:Ni
procd
[°C]e
[g/mol]
PDI f
1
2
3
4
5
6
7
8
9
1b2
1b1
1b1
1b1
1b1
1b1
1b1
1b2
1b2
1b2
1b2
1.0
1.0
1.0
3.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
60
60
60
60
80
60
80
60
60
60
60
1
2
2
2
2
2
2
1
2
2
2
none
214
420
48
3
0
0
450h
430h
350h
120h
390h
170h
279
EASC
EASC
EASC
EASC
MADC
MADC
EASC
EASC
EASC
EASC
1b
1b
1b
1b
1b
1b
3
3
1a
2
4.5
4.3
10.8
1.2
3.9
1.7
2.8
3.3
2.8
3.0
126
n.d.
125
122
127
121
126
125
126
127
131k
n.d.
136k
74k
131k
73k
148k
155k
149k
157k
3.3
n.d.
3.4
3.5
3.4
3.7
5.4
5.2
8.2
6.0
13
n.d.
14
n.d.
n.d.
n.d.
n.d.
22
330
480
480
480
214
214
214
214
334
284
301
10
11
n.d.
n.d.
a Precatalyst. b Theoretical maximum nickel loading. c Cocatalyst: EASC ) Et3Al2Cl3; MADC ) Me2Al2Cl4. d Procedure. e Melting point determined by
1
DSC. f Molecular weight and polydispersity determined by GPC/SEC. g Total branching determined by H NMR. h Results previously reported were based
on the nickel loading found by ICP-OES analysis: entry 2: 750; entry 3: 720; entry 4: 530; entry 5: 200; entry 6: 650; entry 7: 280.
loading could result.) Here, we report the ethylene polymeri-
zation activities per gram of nickel based on the theoretical
maximum nickel loading that affords the minimum possible
activity values. The presumably low ICP-OES analyses (cf.
Table 1, entries 2-7) are responsible for the previously reported
higher activities per gram of nickel.
Ethylene Polymerization Studies Employing Precatalysts
1b1 and 1b2. The ethylene polymerization activities of pre-
catalysts 1b1 and 1b2 are summarized in Table 1. Slurry
polymerizations were performed in 100 mL of pentane under
similar conditions (10 mg of 1b1 or 1b2, 150 psig ethylene).
The cocatalyst, ratio of aluminum cocatalyst to nickel precata-
lyst, nickel loading, polymerization temperature, polymerization
time, and polymerization procedure were varied. No PE was
obtained when the polymerization was performed in the absence
of an alkylaluminum cocatalyst (entry 1), thus attachment of
R-diimine nickel(II) complex 11b to TMA-activated silica
support 14a occurs without a simultaneous activation of
precatalyst 1b2. This feature allows a free choice of the
alkylaluminum cocatalyst, and we pursued alternatives less
expensive than MAO.
R-Diimine nickel(II) precatalyst 1b1 showed very high
productivities when activated with an alkylaluminum halide.
Activation with ethylaluminum sesquichloride (EASC) afforded
excellent productivities in a 2 h polymerization run at 60 °C.70
Using 1b1 with a 1.0 wt % nickel loading resulted in

6346 Schrekker et al.
Macromolecules, Vol. 39, No. 19, 2006
Table 2. Effect of Structural Modifications on the r-Diimine Nickel(II) Precatalysts
d
T
[°C]
time
[h]
kg PE/
g 1/2
kg PE/
g Ni
Tm
Mw
branches/
1000Ce
entry
precatalyst
wt % Nia
Al/Ni
procb
[°C]c
[g/mol]
PDId
1
2
3
4
5
6
7
8
9
1a
60
60
80
60
60
80
60
80
60
80
1.0
1.0
1.0
1.0
2.0
1.0
1.0
1.0
1.0
1.0
2
2
2
1
1
1
2
2
2
2
280
420
480
214
214
214
214
214
214
214
1b
1b
1b
3
3
3
3
3
3
3
3.3
4.5
1.2
1.1
1.7
0.2
330f
450f
120f
109
84
19
197
66
128
126
122
120
118
119
n.d.
n.d.
n.d.
n.d.
153k
131k
74k
458k
369k
278k
596k
262k
825k
415k
3.6
3.3
3.5
4.5
3.7
3.7
15
12
18
13
n.d.
13
n.d.
54
54
64
38
52
31
48
1b1
1b1
1c
1c
1c
2a1
2a1
2b
2.0g
0.7g
2.1g
0.9g
208
96
10
2b
a Theoretical maximum nickel loading. b Procedure. c Melting point determined by DSC. d Molecular weight and polydispersity determined by GPC/
SEC. e Total branching determined by 1H NMR. f Results previously reported were based on the nickel loading found by ICP-OES analysis: entry 1: 580;
entry 2: 750; entry 3: 200. g A small fraction of oligomers (0.2-0.4 kg per gram of 2a1 or 2b) was obtained in addition to high-molecular-weight PE with
precatalysts 2a1 and 2b.
productivities of 4.5 kg PE/g 1b1 and 450 kg PE/g of nickel
(entry 2). Remarkably, productivities remained high at Al/Ni
ratios as low as 48 (430 kg PE/g Ni, entry 3). From these results,
it is apparent that EASC is a very efficient activator and that
excess EASC does not induce catalyst deactivation. An increase
of the nickel loading from 1.0 to 3.0 wt % raised the productivity
per gram of 1b1 from 4.5 to 10.8 kg PE without a dramatic
loss of the productivity per gram of nickel (entry 4 vs entry 2).
Although methylaluminum dichloride (MADC) was found to
be less suitable as a cocatalyst at 60 °C (entry 6), the contrary
is true when raising the polymerization temperature to 80 °C
(entry 7 vs entry 5). Deactivation of homogeneous R-diimine
nickel(II) catalysts observed at elevated temperatures of 80 °C
is also seen with these supported catalysts.30 The combination
of precatalyst 1b1 with MADC afforded productivities of 1.7
kg PE/g 1b1 and 170 kg PE/g nickel (entry 7). Although
absolute activity is lower with MADC as seen at 60 °C, its
improved performance at 80 °C compared to EASC is probably
attributable to a slower catalyst deactivation.
heating period in the absence of ethylene might explain the
slightly reduced productivity of 2.8 kg PE/g 1b2 (entry 10).
Saturating the pentane solution prior to heating (procedure 2)
showed a small increase in productivity, which suggests that
ethylene prevents deactivation (entry 11).
Structural Modifications of the R-Diimine Nickel(II)
Precatalysts. Precatalysts 1a-c and 2a,b were tested in ethylene
slurry polymerizations to study the effect of structural modifica-
tions (Table 2). Polymerizations were performed in 100 mL of
pentane under similar conditions (10 mg of 1a-c or 2a,b, 1 or
2 h, 150 psig ethylene, EASC as cocatalyst, large excess of
EASC to nickel precatalyst, polymerization procedure 1b or 3).
The polymerization temperature and nickel loading were varied.
Entry 3 of Table 1 showed that EASC is a suitable cocatalyst
at low Al/Ni ratios; however, in these polymerizations, a larger
excess of EASC was used to guarantee a complete activation
of the precatalysts. Earlier, we reported that polymerizations
with catalysts supported through an amino or hydroxy func-
tionality directly attached to the para position of the aryl
R-diimine moiety were disappointing.67 However, employing a
spacer group between these functionalities (-(CH2)3- in the
case of -OH and -CH2(3,5-CH3,CH3-C6H2)- in the case of
-NH2) results in excellent productivities (Table 2). The reasons
for the requirement of a spacer are not clear, but it is instructive
to note that Herrmann and co-workers observed increasing
productivities for silica-attached bis(imino) pyridyl iron ethylene
polymerization catalysts with increased length of the carbon
chain used for attachment.65
Initially, 2 h polymerization runs were performed as this is a
common residence time in commercial polymerization pro-
cesses. However, the active species is almost completely
deactivated after 1 h, and productivity is only slightly higher
after 2 h (entries 8, 9). A common characteristic of these
R-diimine nickel(II) catalysts is their extreme activity followed
by a rapid decay due to deactivation processes.18
Three polymerization procedures were tested in our single-
batch reactor (entries 9-11). Productivity and PE properties
showed only minor differences upon changes in the procedure.
In procedure 3, precatalyst and cocatalyst were added to an
argon-filled reactor at room temperature, and the reactor was
immediately pressurized with 150 psig ethylene and brought to
temperature. This procedure was employed in order to prevent
catalyst deactivation in the absence of ethylene. A complication
is that some polymerization will take place while heating to
the required temperature. Procedures 1a (similar to 1b) and 2
were modified so that the reactor was heated to the polymeri-
zation temperature before pressurizing with ethylene. Now
polymerization takes place only at the desired temperature,
which allows a better comparison with commercial continuous
processes. Partial deactivation of the active species during the
Both ortho-methyl-substituted R-diimine nickel(II) precata-
lysts 1a and 1b1 afforded polyethylenes with quite similar
properties (melting point, average molecular weight, polydis-
persity, and branching). Modifications of the most active
precatalyst 1b1 containing ortho-methyl groups were pursued
in an effort to improve productivity and to influence the
polyethylene properties. Introduction of ortho-isopropyl sub-
stituents was easily achieved and precatalyst 1c was readily
generated, as noted above. Surprisingly, productivities of 1c
relative to 1b1 decreased dramatically at both 60 and 80 °C to
1.1 kg PE/g 1c and 0.2 kg PE/g 1c, respectively (entries 4 and
6). Polymerization with precatalyst 1c having a 2.0 wt % instead
of a 1.0 wt % nickel loading (entry 5 vs entry 4) resulted in a

Macromolecules, Vol. 39, No. 19, 2006
Slurry-Phase Homopolymerization of Ethylene 6347
beads. As a consequence, polymer growth would not effectively
fragment the silica particle, and an outer coating of PE could
develop that would inhibit polymer growth. Energy-dispersive
X-ray analysis of cross-sectioned precatalyst particles 1c with
a 2.0 wt % nickel loading was performed to get a better
understanding of the distribution of R-diimine complex 11c
(Figure 4). Mapping of Al shows a homogeneous distribution
of the trimethylaluminum linker used for the attachment of 11c.
The weight percentage of 2.0 wt % 11c allowed the successful
detection of nickel (mapping of the actual catalytic metal center
was not successful in previously reported EDX studies on silica-
supported metallocene polymerization catalysts, as its weight
percentage was below the detection limit52,71), and showed that
the hypothesis of a core-shell distribution of 11c in precatalyst
1c is incorrect. A homogeneous distribution of nickel was
observed within the silica particles after slicing of the beads.
SEM of the cross-sectioned particle of precatalyst 1c showed
the locations of the pores within the silica bead, which was also
observed by EDX of the elements Al, Si, and Ni (Figure 4).
Figure 4. Representive example of a cross-sectioned particle of
precatalyst 1c with a 2.0 wt % nickel loading. EDX mapping of the
elements Al-Si-Ni (high brightness reflects high Al-Si-Ni concen-
trations) and SEM.
The physical appearance of the polyethylene particles ob-
tained with precatalyst 1c were significantly different from those
observed with precatalysts 1b1/1b2. The PE particles were
analyzed by scanning electron microscopy (SEM), and the
resulting micrographs are shown in Figure 5a-c. Figure 5a
shows the free-flowing PE particles obtained with precatalyst
1b2 after isolation. Most of these particles are roughly spherical,
with a diameter of ∼675 µm. PE particles produced with
precatalyst 1c are somewhat agglomerated after isolation at room
temperature; agglomeration intensifies after drying at 70 °C.
SEM (Figure 5b) shows these particles are again roughly
spherical but smaller in diameter (∼425 µm). The particles
appear connected by traces of PE, which are not part of the
spherical particles, seemingly consistent with their more ag-
gregated nature.
similar increase, as was observed when precatalyst 1b1 was
supported with a higher nickel loading (Table 1, entry 4). These
lower productivities are in sharp contrast to the homogeneous
polymerizations.30 However, interestingly, the polyethylene
properties showed a similar trend, as is seen in the homogeneous
mode (lower melting temperature, higher average molecular
weight, and higher branching level of the isopropyl-substituted
system versus the methyl-substituted system). The fact that only
the productivity did not follow the “homogeneous” trend
stimulated us to investigate this behavior in further detail. These
polymerizations are characterized by a high ethylene uptake at
the start, which diminishes rapidly compared to the polymeriza-
tions with precatalyst 1b1. One possible explanation of this
reduced activity is that the larger R-diimine complex 11c does
not penetrate to the inner surfaces of the mesoporous silica
Figure 5. (a) SEM of PE produced in a 1 h polymerization at 60 °C and 150 psig ethylene with precatalyst 1b2 containing a nickel loading of 1.0
wt % (Table 1, entry 8). (b) SEM of PE produced in a 1 h polymerization at 60 °C and 150 psig ethylene with precatalyst 1c containing a nickel
loading of 1.0 wt % (Table 2, entry 4). (c) SEM of PE produced in a 2 h polymerization at 80 °C and 700 psig ethylene with precatalyst 1b2
containing a nickel loading of 1.0 wt % (Table 6, entry 7). (d) SEM of Grace Davison SP9-496 that is calcined at 500 °C.

6348 Schrekker et al.
Macromolecules, Vol. 39, No. 19, 2006
Figure 6. (a) DSC of PE produced in a 1 h polmerization at 60 °C and 150 psig ethylene with precatalyst 1b2 containing a nickel loading of 1.0
wt % (Table 1, entry 8). (b) DSC of PE produced in a 1 h polymerization at 60 °C and 150 psig ethylene with precatalyst 1c containing a nickel
loading of 1.0 wt % (Table 2, entry 4).
Table 3. 13C NMR Analysis of PE Produced in a 1 h Polymerization at 60 °C and 150 psig Ethylene with Precatalysts 1b2 (Table 1, Entry 8)
and 1c (Table 2, Entry 4) Containing a Nickel Loading of 1.0 wt %43
branches per 1000 CH2
precatalyst
methyl
ethyl
propyl
butyl
amyl
butyl+, EOC
amyl+, EOC
hexyl+, EOC
total branches
1b2
1c
13.5
36.8
1.9
4.1
0.6
1.9
0.6
1.2
0.3
1.0
1.5
4.5
0.7
3.0
0.9
1.7
17.6
47.4
Branching densities and types of branches for the PE’s
produced with the precatalysts 1b2 and 1c were determined by
13C NMR spectroscopy (Table 3). The branching densities
measured by 13C NMR agree with those determined by 1H NMR
and show ca. 20 branches/1000Cs for PE from 1b2 and ca. 50
branches/1000Cs for PE from 1c. Differential scanning calo-
rimetry (DSC) analysis showed interesting features for these
PE’s (Figure 6). PE from 1b2 with 20 branches/1000Cs shows
a fairly broad endotherm with a Tm of 126 °C (Figure 6a) and
is typical of solution-generated PE containing a similar number
of branches.30 The PE from 1c exhibits an extremely broad
endotherm, but a clear, prominent Tm at 120 °C (Figure 6b).
On the basis of solution results, a PE containing 50 branches/
1000Cs would be expected to exhibit a much lower Tm than
the observed 120 °C.30 A plausible explanation is that the
branches are not evenly distributed along these chains. This
could result if the nickel active centers become increasingly
buried in growing polymer and the effective ethylene concentra-
tion at these centers is reduced. On the basis of previous
results,18,30 lower effective concentrations of ethylene give rise
to increased branching, indicating that these polymers may be
tapered with the more linear segments, giving rise to the
prominent Tm at 120 °C. The more amorphous segments may
be responsible for agglomeration of these particles.
backbone rather than those containing the acenaphthyl group
in the backbone (see Figure 1). Productivities were significantly
lower (ca. 2×) compared to those observed with precatalyst 1b1
at both 60 and 80 °C (Table 2, entries 7-10). Productivities
and PE properties of PEs obtained with 2a1 and 2b are similar,
which suggests that a second hydroxyl functionality has no effect
on the performance of the catalyst. The increase in average
molecular weight for these precatalysts 2 with an alkyl-
substituted backbone compared to precatalysts 1 with a planar
aromatic acenaphthyl backbone is also observed with the
unsupported catalysts.29,30 The molecular weight distributions
of PE from precatalysts 2 are very broad compared to PE from
precatalysts 1.
Polymerizations with precatalysts 1b produced exclusively
high-molecular-weight PE particles; no formation of fines and
indications of catalyst leaching were observed nor was there
any reactor fouling. In the case of precatalyst 1c, while mainly
free-flowing PE particles were obtained, a small amount of PE
did adhere to the reactor wall and stirring blade. This seems
likely due to the low melting properties and more amorphous
nature of the PE.
In general, the same changes in the PE properties were
observed for precatalysts 1 and 2 upon increased polymerization
temperatures. Higher temperatures resulted in a decrease of the
average molecular weight together with a decrease of the melting
point due to increased branching densities. The same behavior
had been observed for the unsupported catalysts.30 However,
clear differences are observed for the polymer properties when
produced with supported or unsupported catalysts. Average
molecular weight, polydispersity, and melting point are in
general higher and branching densities lower for the supported
catalysts under similar conditions. These features have also been
observed for polymerizations using R-diimine nickel(II) catalysts
supported with other technologies.54,55 Reduced branching
suggests a reduced propensity for chain walking in supported
catalysts.
13C NMR analysis afforded more detailed information
1
compared to H NMR analysis, which only provides the total
number of branches.43,44 The 13C NMR analysis allowed us to
clarify the nature of the branches and the amount of each branch
in the PE chains (Table 3). Both PE’s obtained with precatalysts
1b2 and 1c contain the same types of branches in approximately
the same ratios, with the amount of each branch decreasing from
short to long chains. This is a consequence of the chain-walking
mechanism involved in the enchainment process, as has been
earlier documented18,29 and summarized above.
Covalent attachment in the aryl R-diimine nickel(II) precata-
lysts 2a1 and 2b was achieved using either one or two hydroxy-
tails in the backbone. These catalysts are more closely related
to the diimine catalysts containing methyl groups in the
Silica Support: Effect of Calcining Temperatures. R-Di-
imine nickel(II) complex 12a was supported on three different

Macromolecules, Vol. 39, No. 19, 2006
Slurry-Phase Homopolymerization of Ethylene 6349
Table 4. Influence of the Calcining Temperature of the Silica
Supporta
Table 5. Influence of the Silica Support Linkera
T
[°C]
kg PE/g
1b
kg PE/g
Ni
entry
precatalyst
linker
silica 13
[mmol OH/g]
T
[°C]
kg PE/g
2a
kg PE/g
Ni
1
2
3
4
5
6
1b2
1b2
1b3
1b3
1b4
1b4
Me3Al
Me3Al
SiCl4
SiCl4
BCl3
60
80
60
80
60
80
3.6
1.3
2.7
1.1
2.8
1.2
355
130
268
114
277
117
entry
precatalyst
1
2
3
4
5
6
2a1
2a1
2a2
2a2
2a3
2a3
1.36
1.36
1.25
1.25
1.07
1.07
60
80
60
80
60
80
2.0b
0.7b
1.5b
0.7b
1.8b
0.7b
197
66
151
66
184
66
BCl3
a Procedure 3.
a Procedure. b 0.2 to 0.4 kg oligomers per gram of precatalyst 2a isolated
for each polymerization.
screening and evaluation of catalysts. However, because com-
mercial slurry polymerizations are generally operated at higher
ethylene pressures (400-700 psig), we have examined these
systems in this pressure range. Slurry polymerizations were
performed using precatalysts 1b2 and 1c under similar condi-
tions (See Table 6: 10 mg of 1b2 or 1c, 1.0 wt % Ni, EASC,
Al/Ni ) 214, 2 h, 100 mL of pentane). Polymerization using
1b2 at 75 psig ethylene showed a productivity of 2.0 kg PE/g
1b2 at 60 °C (entry 1). Increasing ethylene pressure to 150 to
400 to 700 psig resulted in improved productivities of 2.8 to
4.4 to 4.8 kg PE/g 1b2, respectively (entries 2-4). This effect
of ethylene pressure was previously observed for homogeneous
R-diimine nickel(II) catalysts.30,72 Qualitative monitoring of
ethylene uptake suggests that catalyst lifetimes are somewhat
increased at higher ethylene pressures. A major challenge with
R-diimine nickel(II) catalysts is to achieve high productivities
at 80 °C, a temperature appropriate for commercial slurry
processes. Increased ethylene pressures in part address this
problem. The productivity of 1b2 at 80 °C was doubled when
the pressure was increased from 150 to 700 psig ethylene (entry
5 vs entry 7). Surprisingly, polymerization with ortho-isopropyl-
substituted precatalyst 1c at 700 psig ethylene (entry 8) showed
an excellent productivity of 2.2 kg PE/g 1c. This is a 10-fold
increase compared to the polymerization at 150 psig (Table 2,
entry 6). Furthermore, productivity is now comparable to that
of precatalyst 1b2 (2.7 kg PE/g precatalyst), which is in sharp
contrast to what was seen at 150 psig.
Free flowing PE particles were obtained in all polymerization
runs with precatalyst 1b2. The spherical morphology of Grace
Davison SP9-496 silica particles (Figure 5d) is reproduced in
the PE particles at 150 psig ethylene and 60 °C (Figure
5a).10,71,73,74 However, the morphology of the PE particles
produced at 700 psig ethylene and 80 °C (Figure 5c) is
somewhat different from that obtained at 150 psig ethylene.
Particle diameters are greater (compare Figure 5a and c), but
their somewhat irregular shapes suggest some particle fragmen-
tation during growth.
A clear trend was seen for the PE properties with an
increasing ethylene pressure. Melting temperature and average
molecular weight increased steadily going from 75 to 150 to
400 psig ethylene at both 60 and 80 °C. This trend was also
observed for the unsupported catalysts.30 However, this trend
was reversed in going from 400 to 700 psig as melting point
and average molecular weight decreased. This might indicate
that, at 700 psig, the local temperature in the growing PE
particles exceeds 80 °C and thus increases branching and
decreases Tm. Consistent with polymerizations in homogeneous
solutions,30 the isopropyl-substituted system 1c relative to 1b2
activated silica supports 14a. The Grace Davison SP9-496 silicas
used for the preparation of 14a were calcined at different
temperatures of 500, 600, and 700 °C. Calcination allows control
of the number of free hydroxy functionalities on silica 13,52
which decreases with an increasing calcination temperature from
1.36 mmol OH/g 13 at 500 °C to 1.25 at 600 °C to 1.07 at 700
°C. Slurry polymerizations with precatalysts 2a1, 2a2, and 2a3
were performed to study the effect of the calcination temperature
(Table 4). Polymerization runs of 2 h were performed under
the same conditions (10 mg of 2a, 1.0 wt % Ni, EASC, Al/Ni
) 214) in 100 mL of pentane at 150 psig ethylene. Similar
productivities of around 1.8 kg PE/g 2a were obtained at 60 °C
(compare entries 1, 3, 5), and the same productivities of 0.7 kg
PE/g 2a were observed at 80 °C (compare entries 2, 4, 6).
Apparently, 1.07 mmol of TMA-linked hydroxy functionalities
is sufficient to quantitatively support a 1.0 wt % nickel loading
of R-diimine nickel(II) complex 12a, and the presence of
additional SiOAlMe2 functionalities on precatalysts 2a1 and 2a2
do not effect the polymerization.
Silica-Support Linker. Grace Davison SP9-496 silica was
treated with the linkers trimethylaluminum, tetrachlorosilane,
and trichloroborane. Covalent attachment of R-diimine nickel-
(II) complex 11b to the linker of the silica supports 14a
(-AlMe2), 14b (-SiCl3), and 14c (-BCl2) afforded precatalysts
1b2, 1b3, and 1b4, respectively. The effect of the support linker
was tested under the same polymerization conditions (10 mg
of 1b, 1.0 wt % Ni, EASC, Al/Ni ) 214, 2 h) in 100 mL of
pentane at 150 psig ethylene (Table 5). Trimethylaluminum
(entries 1-2) was found to be an excellent linker in our initial
studies. Methane is released in both the attachment of the linker
to the silica support and upon attachment of R-diimine nickel-
(II) complex 11b (Scheme 6). The TMA-treated silica support
14a does not activate complex 11a in the grafting reaction. In
an attempt to find a less-expensive alternative to TMA, we
screened tetrachlorosilane and trichloroborane as linkers. These
linkers have in common that HCl is released upon attachment
to the support and attachment of complex 11b. Similar produc-
tivities were obtained with the -SiCl2- (1b3) and -BCl- (1b4)
precatalysts at 60 and 80 °C (compare entries 3, 5 to 4, 6). The
somewhat lower productivities obtained at 60 °C (entries 3 and
5) compared with that obtained with TMA precatalyst 1b2 (entry
1) might be ascribed to an incomplete removal of HCl. However,
all three precatalysts afforded similar productivities at 80 °C
(compare entries 2, 4, 6).
Effects of Variation of Ethylene Pressure. An ethylene
pressure of 150 psig was chosen as the standard pressure for

6350 Schrekker et al.
Macromolecules, Vol. 39, No. 19, 2006
Table 6. Influence of the Ethylene Pressurea
entry
precatalyst
pressure [psig]
T [°C]
kg PE/g 1b2/1c
kg PE/g Ni
Tmb [°C]
Mwc [g/mol]
PDIc
1
2
3
4
5
6
7
8
1b2
1b2
1b2
1b2
1b2
1b2
1b2
1c
75
150
400
700
150
400
700
700
60
50
60
60
80
80
80
80
2.0
2.8
4.4
4.8
1.4
2.1
2.7
2.2
200
280
440
483
136
211
272
219
124
126
130
126
122
128
127
117
121k
149k
166k
127k
76k
119k
105k
338k
6.2
8.2
7.2
5.1
5.0
7.6
4.5
4.2
a Procedure 1a. b Melting point determined by DSC. c Molecular weight and polydispersity determined by GPC/SEC.
Table 7. Influence of the Nickel Loadinga
d
precatalyst
[mg]
pressure
[psig]
T
[°C]
kgPE/g
1b2
kg PE/g
Ni
Tm
Mw
entry
wt% Nib
Al/Ni
[°C]c
[g/mol]
PDId
1
2
3
4
5
6
10.0
10.0
10.0
5.0
5.0
5.0
1.0
1.0
3.0
3.0
3.0
3.0
214
214
143
143
143
143
400
700
400
400
700
700
80
80
80
80
80
90
2.1
2.7
1.9
4.9
6.0
3.9
211
272
63
164
200
130
128
127
123
124
125
125
119k
105k
61k
90k
94k
7.6
4.5
4.3
4.6
4.8
4.3
78k
a Procedure 1a. b Theoretical maximum nickel loading. c Melting point determined by DSC. d Molecular weight and polydispersity determined by GPC/
SEC.
under similar conditions yields PE with increased molecular
weight and decreased melting temperature.
supported with a 3.0 wt % nickel loading instead of 1.0 wt %
(entries 3-6 vs 1-2). It seems likely that this is due to the
large exotherm and heating of the particles above the solution
temperature. This proposal is strengthened by the observed
melting point of 123 °C and average molecular weight of 61k
in run 3, during which the solution temperature rose to 115 °C.
Variation of Nickel Loadings. Polymerization runs at 60
°C and 150 psig ethylene with precatalyst 1b1 showed an
increased productivity per gram of 1b1 of 4.5 to 10.8 kg PE by
increasing the Ni loading from 1.0 to 3.0 wt % (Table 1, entries
2 and 4). Considering the increase in productivity with increas-
ing ethylene pressure, we screened 1b2 with 3.0 wt % Ni loading
at pressures of 400 and 700 psig. Slurry polymerizations were
performed under similar conditions (EASC, Al/Ni ) 143, 2 h)
in 100 mL of pentane at 80 or 90 °C (Table 7). Polymerization
at 400 psig ethylene with 10 mg of precatalyst resulted in a
rapid temperature increase in the autoclave from 80 to 115 °C
as the cooling capacity was unable to control this extreme
exotherm. The PE produced was agglomerated, and the loss of
particle morphology suggests that the actual temperature in the
particles passed 115 °C. This exotherm no doubt deactivated
the catalyst and accounts for the lowered productivity relative
to the run using a 1.0 wt % Ni loading. When the precatalyst
loading was reduced from 10 to 5 mg, the temperature could
be controlled and maintained at ca. 80 °C (a slight exotherm to
86 °C was initially observed, entry 4). Now, productivity per
gram of 1b2 at 400 psig increased from 2.1 kg PE/g 1b2
(150psig) to 4.9 kg PE/g 1b2 with little loss of productivity on
a per-gram Ni basis. The best productivity at 80 °C was 6.0 kg
PE/g 1b2 obtained at 700 psig ethylene (entry 5). The value of
6.0 kg PE/g support at 80 °C conforms to requirements for
applications in commercial slurry reactors.75 Although reduced,
polymerization at 90 °C, 3.0 wt % Ni loading showed a
surprisingly good productivity of 3.9 kg PE/g 1b2 (entry 6).
These polymerization conditions not only affect productivity
but also the PE properties. Melting points and average molecular
weights were significantly lower when precatalyst 1b2 was
Conclusion
R-Diimine nickel(II) catalysts covalently linked to silica
supports are highly productive ethylene homopolymerization
catalysts. R-Diimine ligands with pendant amino and hydroxy
functionalities are easily prepared via standard synthetic pro-
cedures, and formation of the nickel(II) complexes was achieved
without the use of protecting groups. Reaction of these func-
tionalities with the linker of a silica support resulted in the
desired covalent attachment of the nickel(II) complexes. Tri-
methylaluminum, tetrachlorosilane, and trichloroborane were
identified as efficient linkers, with trimethylaluminum as the
preferred one. This approach enabled high catalyst loadings,
and activation of these supported precatalysts with alkylalumi-
num cocatalysts afforded highly productive ethylene polymer-
ization catalysts.
Development of this technology in slurry polymerizations
demonstrated the following: (1) Inexpensive ethylaluminum
sesquichloride is a very efficient cocatalyst even at low Al/Ni
ratios. (2) Polyethylene properties, including melting points,
molecular weights, and branching densities, can be varied by
modification of the ortho-aryl substituents. (3) Ethylene pressure
is a crucial parameter for optimization of productivity. Produc-
tivity increases steadily as ethylene pressure is raised from 150
to 700 psig, and at 700 psig, high productivities can be achieved
at temperatures up to 80 °C. (4) An increase in nickel loading
from 1.0 to 3.0 wt % results in a significant increase in

Macromolecules, Vol. 39, No. 19, 2006
Slurry-Phase Homopolymerization of Ethylene 6351
Materials. Polymer-grade ethylene (99.5%) and hydrogen were
purchased from National Speciality Gases. Ethylene was purified
over a 30% Cu/SiO2 catalyst and molecular sieves, and hydrogen
was used without purification. Et3Al2Cl3 (0.91 M in toluene), Me3-
Al (2 M in hexane), palladium acetate, tri-o-tolyl phosphine,
anhydrous triethylamine, methyl acrylate, lithium aluminum hy-
dride, 10% palladium on activated carbon, NiBr2(1,2-dimethoxy-
ethane), bromine, diisopropylamine, n-butyllithium (1.6 M in
hexane), tetrabutylammoniumfluoride (1 M in THF), sodium iodide,
chlorotrimethylsilane, tetrachlorosilane, trichloroborane, p-toluene-
sulfonic acid, 3, 5, 6a, and 7 were used as purchased from Aldrich
Chemical Co. CDCl3, CD2Cl2, and C6D5Br were purchased from
Cambridge Isotope Laboratories. Grace Davison silicas were used
as received. The compounds 1a,67 1b1,67 6b,67 8a-b,67 9,78 11a-
b,67 14a,67 and (6-iodo-hexyloxy)-trimethyl-silane79,80 were prepared
according to literature procedures. Toluene, dichloromethane,
diethyl ether, and pentane were passed through an activated alumina
column.81 Tetrahydrofuran was freshly distilled from sodium/
benzophenone.
3-(4-Amino-3,5-diisopropyl-phenyl)-propan-1-ol (6c). A modi-
fied literature procedure was used for the synthesis of 4, and the
spectral data were in agreement with those previously reported.82
Aniline 3 (35.0 g, 197 mmol, 1.00 equiv), bromine (11.2 mL, 217
mmol, 1.10 equiv), and 350 mL of acetic acid were used. Column
chromatography (flash silica, eluent; ethyl acetate/hexane ) 1:19)
afforded 4 in 35% yield (17.5 g, 68.2 mmol). The procedure
reported previously for the synthesis of 6b was used for the
synthesis of 6c from 4.67 Palladium acetate (138 mg, 0.60 mmol,
0.01 equiv), tri-o-tolyl phosphine (1.50 g, 4.90 mmol, 0.08 equiv),
and 60 mL of toluene were mixed under an argon atmosphere.
Anhydrous triethylamine (24.6 mL, 176 mmol, 2.87 equiv), 4 (15.7
g, 61.4 mmol, 1.00 equiv), and methyl acrylate (6.9 mL, 76.1 mmol,
1.24 equiv) were added in this order at 20 °C. The reaction mixture
was heated to 100 °C and stirred for 6 h at this temperature, after
which it was added to water. The water layer was extracted three
times with diethyl ether, and the organic layer was dried over
MgSO4. Removal of the solvent in vacuo resulted in the isolation
of the crude 3-(4-amino-3,5-diisopropyl-phenyl)-acrylic acid methyl
ester (20.6 g), which was used without further purification. This
mixture (20.6 g) was dissolved in 400 mL of THF, and lithium
aluminum hydride (5.97 g, 157 mol) was added at 0 °C. The
temperature was raised to 20 °C after complete addition, and the
reaction mixture was stirred for 30 min at this temperature. The
reaction mixture was poured into water. The water layer was
extracted with diethyl ether, and the collected organic fractions were
dried over MgSO4. The solvent was removed in vacuo. Column
chromatography (flash silica, eluent; ethyl acetate/hexane ) 1:2)
afforded a mixture of 3-(4-amino-3,5-diisopropyl-phenyl)-prop-2-
en-1-ol and 6c (8.41 g). The mixture was dissolved in 100 mL of
ethyl acetate, and palladium (10 wt %) on activated carbon (100
mg, 0.09 mmol) was added. Hydrogen was slowly purged through
the solution for 16 h at 20 °C. Removal of the palladium catalyst
and the solvent resulted in the isolation of 6c as a brown oil (7.67
g, 32.6 mmol). Aniline 6c was synthesized in an overall yield of
18% based on 3. Rf ) 0.59 (ethyl acetate/hexane ) 1:1); 1H NMR
(300.1 MHz, CD2Cl2, 298 K) δ: 6.84 (s, 2H), 3.70-3.54 (m, 4H),
2.91 (septet, J ) 6.9, 2H), 2.56 (t, J ) 7.8, 2H), 1.81 (tt, J ) 7.8,
J ) 6.6, 2H), 1.31 (bs, 1H), 1.24 (d, J ) 6.9, 12H).
productivity on a kg PE/g heterogeneous catalyst basis with only
minor decreases in productivity on a kg PE/g Ni basis. Using a
3.0 wt % Ni loading in combination with an ethylene pressure
of 700 psig, productivities of up to 6.0 kg PE/g precatalyst at
80 °C can be obtained. (5) No reactor fouling due to leaching
of the catalyst is observed.
These observations demonstrate productivities and temper-
ature and pressure ranges compatible with many commercial
slurry reactors used for PE production and suggest possible
utility of these supported nickel R-diimine catalysts in such
reactors.
Experimental Section
General Methods. All reactions involving oxygen- and/or water-
sensitive compounds were carried out under an argon atmosphere
in dried glassware using standard Schlenk, syringe, and septa
techniques. Argon was purified by passage through columns of
BASF R3-11 catalyst (Chemalog) and 4 Å molecular sieves.
Oxygen- and/or water-sensitive compounds were transferred or
manipulated in an argon-filled glovebox at room temperature. Thin-
layer chromatography was carried out on Whatman silica, 250 µm
layer, flexible plates for TLC, PE, SIL, G/UV. Flash chromatog-
raphy was performed using Scientific Adsorbents silica gel no.
02826-25, particle size 22-63 µm, pore size 60 Å.76 1H and 13C
NMR spectra were recorded on a Bruker Avance 300 MHz
spectrometer in CDCl3 or CD2Cl2. Chemical shifts are reported
relative to residual CHCl3 (δ 7.26 for 1H), residual CHDCl2 (δ 5.32
1
for H), CDCl3 (δ 77.00 for 13C), and CD2Cl2 (δ 53.7 for 13C).
Chemical shifts δ are quoted in parts per million (ppm), and
coupling constants J are given in hertz (Hz). Mass spectra were
recorded on a Micromass Quattro II triple quadrupole mass
spectrometer in ESI mode. Elemental analyses was performed by
Atlantic Microlab of Norcross, GA. X-ray diffraction data were
collected on a Bruker SMART 1K diffractometer using the Ω scan
mode (see Supporting Information for crystallographic information
file). High-temperature gel permeation chromatography (GPC/SEC)
was performed by DuPont (Wilmington, DE) in 1,2,4-trichloroben-
zene at 135 °C using a Waters GPCV2000-TCB dual/triple detector
instrument. A calibration curve was established with polystyrene
standards and universal calibration was applied using Mark-
Houwink constants for polyethylene (k ) 4.34 × 10-4; R ) 0.724).
Differential scanning calorimetry (DSC) data were recorded on a
Seiko Instruments DSC 220. Melting points were recorded in the
second heating run at a heating rate of 10 °C/min. 1H NMR spectra
of polyethylenes were recorded in C6D5Br at 120 °C. The formula
used to calculate branching was (CH3/3)/[CH + (CH2/2) + (CH3/
3)] × 1000 ) branches per 1000 carbons. CH3, CH2, and CH refer
to the integration obtained for the methyl, methylene, and methine
resonances, respectively. High-molecular-weight polyethylene was
obtained in all instances, which made correction for methyl end
groups unnecessary.77 13C NMR spectra of polyethylenes were
recorded in a 0.05 M Cr(acac)3 solution in C6D5Br with 15 wt %
polyethylene. Signals were assigned as described by Cotts.43 Nickel
contents of silica-supported nickel diimine precatalysts and the
residual nickel contents of the dichloromethane solutions used for
the support reaction were determined with inductively coupled
plasma spectrometry (ICP-OES) and were performed by DuPont
(Wilmington, DE). Scanning electron microscopy (SEM) of silica
and PEs was performed on a Hitachi S-4700 scanning electron
microscope. Scanning electron microscopy (SEM) of cross-
sectioned particles of silica-supported nickel diimine precatalysts
and the distribution of the elements Al-Si-Ni in these particles
was determined on a scanning electron microscope (SEM) equipped
with an energy-dispersive X-ray spectrometer (EDX), and was
performed by DuPont (Wilmington, DE). Precatalyst particles were
embedded in low-viscosity epoxy resin and cut with a razor blade
after cooling the sample in liquid nitrogen in order to obtain the
cross-sectioned particles. Charging of the cross-sectioned particles
was prevented by an additional carbon coating.
3-{4-[2-[4-(3-Hydroxy-propyl)-2,6-diisopropyl-phenylimino]-
2H-acenaphthylen-1-ylideneamino]-3,5-diisopropyl-phenyl}-
propan-1-ol (8c). A procedure reported previously in the literature
was used for the synthesis of 8c.67 Acenaphthenequinone (516 mg,
2.83 mmol, 1.00 equiv), 6c (2.00 g, 8.50 mmol, 3.00 equiv), and
concentrated aqueous H2SO4 (0.10 mL) were added to 60 mL of
toluene. This mixture was refluxed for 16 h. Column chromatog-
raphy (flash silica, eluent; ethyl acetate/hexane ) 1:1) afforded
R-diimine 8c as a yellow solid in 48% yield (828 mg, 1.34 mmol).
Rf ) 0.24 (ethyl acetate/hexane ) 1:1); 1H NMR (300.1 MHz, CD2-
Cl2, 298 K) δ: 7.89 (d, J ) 7.8, 2H), 7.36 (dd, J ) 7.8, J ) 7.8,
2H), 7.12 (s, 4H), 6.66 (d, J ) 7.8, 2H), 3.75 (dt, J ) 7.7, J ) 6.2,
4H), 2.94 (septet, J ) 6.9, 4H), 2.78 (t, J ) 7.7, 4H), 1.98 (m,

6352 Schrekker et al.
Macromolecules, Vol. 39, No. 19, 2006
4H), 1.41 (bs, 2H), 1.20 (d, J ) 6.9, 12H), 0.98 (d, J ) 6.9, 12H);
13C{1H} NMR (100.6 MHz, CD2Cl2, 298 K) δ: 161.3 (CdN),
145.9 (q C), 141.1 (q C), 138.1 (q C), 135.4 (q C), 131.5 (q C),
130.0 (q C), 129.1 (CH), 128.2 (CH), 123.9 (CH), 123.4 (CH),
62.8 (CH2), 35.3 (CH2), 32.6 (CH2), 29.0 (CH), 23.2 (CH3), 23.1
(CH3); MS-ESI m/z (%): C42H53N2O2 [M + H]+ 617.0 (100%).
Anal. Calcd for C42H52N2O2: C, 81.78; H, 8.50; N, 4.54. Found:
C, 81.73; H, 8.50; N, 4.56.
yield (1.34 g, 2.18 mmol). Crystals suitable for X-ray analysis were
obtained by a slow diffusion crystallization of pentane into a
saturated dichloromethane solution. Anal. Calcd for C26H36Br2N2-
NiO: C, 51.10; H, 5.94; N, 4.58. Found: C, 51.80; H, 6.29; N,
4.54.
8,9-Bis-[(E)-2,6-dimethyl-phenylimino]-hexadecane-1,16-di-
ol Nickel Dibromide (12b). The same procedure was used as for
11c.30,67 R-Diimine 10b (1.18 g, 2.39 mmol, 1.00 equiv), NiBr2-
(dme) (730 mg, 2.39 mmol, 1.00 equiv), and 60 mL of dichlo-
romethane were used. NiBr2 complex 12b was isolated as a brown
solid in a 51% yield (0.87 g, 1.22 mmol). Anal. Calcd for C32H48-
Br2N2NiO2: C, 54.04; H, 6.80; N, 3.94. Found: C, 55.22; H, 7.32;
N, 3.97.
General Procedure for the Attachment of the Linker to the
Silica Support (14).67 Caution! Mechanical stirring with a stirring
bar should be avoided in the reactions involving silica as this
destroys the fragile silica particles. Grace Davison silica 13 (4.0 g)
was suspended in 20 mL of toluene, followed by the addition of
trimethylaluminum (6.0 mL, 2 M solution in hexane). The reaction
mixture was gently shaken manually at several intervals during a
period of 1 h. The solvent was removed, and 14a was washed three
times with 20 mL of toluene and one time with 20 mL of pentane.
Activated silica 14a was dried in vacuo for 16 h at 20 °C.
General Procedure for the Synthesis of Silica-Supported
Nickel Diimine Precatalysts (1-2).67 Caution! Mechanical stirring
with a stirring bar should be avoided in the reactions involving
silica as this destroys the fragile silica particles. Linker silica support
14 was suspended in dichloromethane, and a solution of the nickel
diimine complex 11-12 in dichloromethane was added at 20 °C.
The reaction mixture was gently shaken manually at several
intervals during a period of 1 h. The dichloromethane was removed
after 1 h by pipet and filtered through a FTPE 0.45 µm filter as a
sample preparation for the ICP-OES analysis. Silica-supported
precatalysts 1-2 were washed three times with dichloromethane
and dried in vacuo for 16 h at 20 °C. Amounts of silica, nickel
diimine complex, and dichloromethane used in preparing supported
catalysts via this procedure are given below.
8,9-Bis-[(E)-2,6-dimethyl-phenylimino]-decan-1-ol (10a). A
procedure reported previously in the literature was used for the
synthesis of 10a.83 R-Diimine 9 (4.0 g, 13.7 mmol, 1.00 equiv)
was dissolved in THF, and a THF solution of LDA (13.7 mmol,
1.00 equiv, prepared from diisopropylamine (2.20 mL, 15.7 mmol,
1.15 equiv) and n-butyllithium (8.55 mL, 13.7 mmol, 1.00 equiv,
1.6 M solution in hexane) was added at -50 °C. The color of the
reaction mixture turned from yellow to red. The reaction mixture
was slowly brought to 20 °C and stirred for an additional hour at
this temperature. The reaction mixture was cooled to -50 °C, and
(6-iodo-hexyloxy)-trimethylsilane (4.11 g, 13.7 mmol, 1.00 equiv)
was added at once. The reaction mixture was stirred for 16 h at 20
°C. The reaction was completed when the color turned from black
to bright yellow. Tetrabutylammoniumfluoride (14.0 mL, 14.0
mmol, 1.02 equiv, 1 M solution in THF) was added. The solvent
was removed in vacuo, and the isolated oily material was purified
by column chromatography (flash silica, ethyl acetate/hexane )
1:1). R-Diimine 10a was isolated as a bright-yellow oil in a 82%
1
yield (4.54 g). H NMR (300.1 MHz, CDCl3, 298 K) δ: 7.03 (m,
4H), 6.91 (m, 2H), 3.54 (m, 2H), 2.47 (m, 2H), 2.02 (s, 12H), 2.00
(s, 3H), 1.44 (m, 4H), 1.23 (m, 6H); 13C{1H} NMR (75.4 MHz,
CDCl3, 298 K) δ: 171.5, 167.5, 148.4, 147.9, 127.9, 124.6, 124.5,
123.1, 62.6, 53.7, 32.5, 29.7, 29.0, 28.8, 26.5, 25.3, 18.0, 17.9. Anal.
Calcd for C26H36N2O: C, 79.55; H, 9.24; N, 7.14. Found: C, 79.17;
H, 9.56; N, 7.14.
8,9-Bis-[(E)-2,6-dimethyl-phenylimino]-hexadecane-1,16-di-
ol (10b). The same procedure was used as for 10a.83 The only
difference was that the second substitution was performed before
deprotection with tetrabutylammoniumfluoride: first substitution:
R-diimine 9 (3.00 g, 10.3 mmol, 1.00 equiv), (6-iodohexyloxy)-
trimethylsilane (3.08 g, 10.3 mmol, 1.00 equiv), diisopropylamine
(1.44 mL, 10.3 mmol), and n-butyllithium (6.41 mL, 10.3 mmol,
1.6 M solution in hexane); second substitution: (6-iodohexyloxy)-
trimethylsilane (3.08 g, 10.3 mmol, 1.00 equiv), diisopropylamine
(1.44 mL, 10.3 mmol), n-butyllithium (6.41 mL, 10.3 mmol, 1.6
M solution in hexane), and tetrabutylammoniumfluoride (20.6 mL,
20.6 mmol, 2.00 equiv, 1 M solution in THF). The solvent was
removed in vacuo, and the isolated oily material was purified by
column chromatography (flash silica, ethyl acetate). R-Diimine 10b
Silica-Supported Precatalyst (1a): Theoretical Maximum Ni
Loading of 1.0 wt %. Linker silica support 14a (100 mg, Grace
Davison XPO-2402, calcined at 200 °C), nickel diimine complex
11a (17.5 mg), and 2 mL of dichloromethane were used. Precatalyst
1a was isolated as a brown solid. ICP-OES analysis: 0.56 wt %
Ni.
Silica-Supported Precatalyst (1b1): Theoretical Maximum
Ni Loading of 1.0 wt %. Linker silica support 14a (400 mg, Grace
Davison XPO-2402, calcined at 200 °C), nickel diimine complex
11b (56.4 mg), and 10 mL of dichloromethane were used. The
dichloromethane solution removed after 1 h was colorless. Pre-
catalyst 1b1 was isolated as a brown solid. ICP-OES analysis: 0.60
wt % Ni. Nickel loading based on the nickel content found in the
dichloromethane solution after the support reaction: 0.99 wt % Ni.
Silica-Supported Precatalyst (1b1): Theoretical Maximum
Ni Loading of 3.0 wt %. Linker silica support 14a (100 mg, Grace
Davison XPO-2402, calcined at 200 °C), nickel diimine complex
11b (58.7 mg), and 12 mL of dichloromethane were used. The
dichloromethane solution removed after 1 h had a red color.
Precatalyst 1b1 was isolated as a brown solid. ICP-OES analysis:
2.0 wt % Ni.
Silica-Supported Precatalyst (1b2): Theoretical Maximum
Ni Loading of 1.0 wt %. Linker silica support 14a (2.0 g, Grace
Davison SP9-496, calcined at 500 °C), nickel diimine complex (281
mg), and 25 mL of dichloromethane were used. The dichlo-
romethane solution removed after 1 h had a weak-yellow color.
Precatalyst 1b2 was isolated as a brown solid. Nickel loading based
on the nickel content found in the dichloromethane solution after
the support reaction: 0.99 wt % Ni.
1
was isolated as a bright-yellow oil in a 65% yield (3.31 g). H
NMR (300.1 MHz, CDCl3, 298 K) δ: 7.05 (m, 2H), 6.93 (m, 1H),
3.54 (m, 2H), 2.46 (m, 2H), 2.03 (s, 6H), 2.01 (s, 3H), 1.44 (m,
4H), 1.23 (m, 6H); 13C{1H} NMR (75.4 MHz, CDCl3, 298 K) δ:
171.1, 148.0, 127.9, 124.7, 123.0, 62.7, 32.5, 29.7, 29.2, 28.8, 26.3,
25.3, 18.1; Anal. Calcd for C32H48N2O2: C, 78.00; H, 9.82; N, 5.69.
Found: C, 77.75; H, 9.85; N, 5.66.
3-{4-[2-[4-(3-Hydroxypropyl)-2,6-diisopropyl-phenylimino]-
2H-acenaphthylen-1-ylideneamino]-3,5-diisopropyl-phenyl}-
propan-1-ol Nickel Dibromide (11c). A procedure reported
previously in the literature was used for the synthesis of 11c.30,67
R-Diimine 8c (250 mg, 0.41 mmol, 1.05 equiv) and NiBr2(dme)
(119 mg, 0.39 mmol, 1.00 equiv) were suspended in 10 mL of
dichloromethane. The reaction mixture was stirred for 16 h at 20
°C. The dichloromethane was removed in vacuo at 20 °C, and the
residual solids were washed with diethyl ether (3 × 5 mL). NiBr2
complex 11c was isolated as a brown solid in a 83% yield (270
mg, 0.32 mmol). Anal. Calcd for C42H52Br2N2NiO2: C, 60.39; H,
6.27; N, 3.35. Found: C, 58.54; H, 6.18; N, 3.17.
8,9-Bis-[(E)-2,6-dimethyl-phenylimino]-decan-1-ol Nickel Di-
bromide (12a). The same procedure was used as for 11c.30,67
R-Diimine 10a (1.00 g, 2.54 mmol, 1.00 equiv), NiBr2 (dme) (780
mg, 2.54 mmol, 1.00 equiv), and 60 mL of dichloromethane were
used. NiBr2 complex 12a was isolated as a brown solid in a 86%
Silica-Supported Precatalyst (1c2): Theoretical Maximum
Ni Loading of 3.0 wt %. Linker silica support 14a (200 mg, Grace
Davison SP9-496, calcined at 500 °C), nickel diimine complex 11b
(117 mg), and 20 mL of dichloromethane were used. The dichlo-
romethane solution removed after 1 h had a red color. Precatalyst

Macromolecules, Vol. 39, No. 19, 2006
Slurry-Phase Homopolymerization of Ethylene 6353
1b2 was isolated as a brown solid. Nickel loading based on the
nickel content found in the dichloromethane solution after the
support reaction: 2.8 wt % Ni.
Silica-Supported Precatalysts (1b3 and 1b4): Theoretical
Maximum Ni Loading of 1.0 wt %. Linker silica support 14b or
14c (200 mg, Grace Davison SP9-496, calcined at 500 °C), nickel
diimine complex 11b (28.1 mg), and 10 mL of dichloromethane
were used. Precatalysts 1b3 and 1b4 were isolated as brown solids.
Silica-Supported Precatalyst (1c): Theoretical Maximum Ni
Loading of 1.0 wt %. Linker silica support 14a (300 mg, Grace
Davison SP9-496, calcined at 500 °C), nickel diimine complex 11c
(49.8 mg), and 10 mL of dichloromethane were used. The
dichloromethane solution removed after 1 h was colorless. Pre-
catalyst 1c was isolated as a brown solid. Nickel loading based on
the nickel content found in the dichloromethane solution after the
support reaction: 1.0 wt % Ni.
Silica-Supported Precatalyst (1c): Theoretical Maximum Ni
Loading of 2.0 wt %. Linker silica support 14a (100 mg, Grace
Davison SP9-496, calcined at 500 °C), nickel diimine complex 11c
(39.8 mg), and 10 mL of dichloromethane were used. The
dichloromethane solution removed after 1 h had a brown color.
Precatalyst 1c was isolated as a brown solid. ICP-OES analysis:
1.33 wt % Ni.
was reduced in vacuo to check for the presence of oligomeric
material.
Procedure 1b.67 Modifications compared to procedure 1a. A
solution of the cocatalyst in 100 mL of pentane was added to the
reactor, followed by addition of the supported precatalyst. All lines
to the reactor were closed, and the temperature was raised to the
required temperature within 2 min while stirring (350 rpm).
Procedure 2. The reactor was charged with toluene and heated
for 10 min at 100 °C. The hot toluene was removed, and the reactor
was dried under vacuum for 10 min at 100 °C. After cooling to
room temperature, the reactor was filled with ethylene. The reactor
was then charged with 50 mL of pentane. A suspension of the
supported precatalyst in 50 mL of pentane was cannula transferred
to the reactor, followed by Et3Al2Cl3 cocatalyst. The reactor was
immediately pressurized with ethylene to the desired pressure, and
the temperature was raised to the desired temperature within 5 min
while stirring (350 rpm). The reaction mixture was stirred (350
rpm) for the desired time at the desired temperature. The temper-
ature was controlled by cooling with water or heating with a heating
mantle. The reaction mixture was cooled to 15 °C, and 20 mL of
methanol was added. The polymer particles were isolated by
filtration and dried in an oven at 75 °C. The filtrate was reduced in
vacuo to check for the presence of oligomeric material.
Procedure 3. The reactor was charged with toluene and heated
for 10 min at 100 °C. The hot toluene was removed, and the reactor
was dried under vacuum for 10 min at 100 °C. After cooling to
room temperature, the reactor was filled with argon. The reactor
was charged with 50 mL of pentane. A suspension of the supported
precatalyst in 50 mL of pentane was cannula transferred to the
reactor, followed by Et3Al2Cl3 cocatalyst. A small vacuum was
applied, and the reactor was vented three times with ethylene. The
reactor was immediately pressurized with ethylene to the desired
pressure. The temperature was raised to the desired temperature
within 5 min while stirring (350 rpm), and the reaction mixture
was stirred (350 rpm) for the desired time at the desired temperature.
The temperature was controlled by cooling with water or heating
with a heating mantle. The reaction mixture was cooled to 15 °C,
and 20 mL of methanol was added. The polymer particles were
isolated by filtration, and dried in an oven at 75 °C. The filtrate
was reduced in vacuo to check for the presence of oligomeric
material.
Silica-Supported Precatalyst (2a1): Theoretical Maximum
Ni Loading of 1.0 wt %. Linker silica support 14a (100 mg, Grace
Davison SP9-496, calcined at 500 °C), nickel diimine complex 12a
(11.6 mg), and 10 mL of dichloromethane were used. The
dichloromethane solution removed after 1 h had a yellow color.
Precatalyst 2a1 was isolated as a green/brown solid. Nickel loading
based on the nickel content found in the dichloromethane solution
after the support reaction: 0.98 wt % Ni.
Silica-Supported Precatalyst (2a2): Theoretical Maximum
Ni Loading of 1.0 wt %. Linker silica support 14a (100 mg, Grace
Davison SP9-496, calcined at 600 °C), nickel diimine complex 12a
(11.6 mg), and 10 mL of dichloromethane were used. Precatalyst
2a2 was isolated as a deep-blue solid.
Silica-Supported Precatalyst (2a3): Theoretical Maximum
Ni loading of 1.0 wt %. Linker silica support 14a (100 mg, Grace
Davison SP9-496, calcined at 700 °C), nickel diimine complex 12a
(11.6 mg), and 10 mL of dichloromethane were used. Precatalyst
2a3 was isolated as a deep-blue solid.
Silica-Supported Precatalyst (2b): Theoretical Maximum Ni
Loading of 1.0 wt %. Linker silica support 14a (100 mg, Grace
Davison SP9-496, calcined at 500 °C), nickel diimine complex 12b
(13.7 mg), and 10 mL of dichloromethane were used. The
dichloromethane solution removed after 1 h was colorless. Pre-
catalyst 2b was isolated as a green solid. Nickel loading based on
the nickel content found in the dichloromethane solution after the
support reaction: 0.99 wt % Ni.
Ethylene Polymerization. Polymerizations were carried out in
a mechanically stirred 300 mL Parr reactor equipped with an electric
heating mantle controlled by a thermocouple in the reaction mixture.
The various procedures used for the polymerization are described
below.
Acknowledgment. We thank DuPont for financial support,
D. J. Adelman, D. R. Loveday, S. D. Ittel, M. F. Schiffhauer,
J. D. Citron, D. L. Kerbow, R. S. Schiffino, and A. M. A.
Bennett (all of DuPont) and Prof. B. Rieger (University of Ulm)
for helpful discussions, E. F. McCord (DuPont) for 13C NMR
analyses of PE samples, and E. Boyes (DuPont) for EDX
analyses. Grace Davison is gratefully acknowledged for gener-
ous gifts of the silicas. Fellowship support was provided to
P.P.P. by the Austrian Science Foundation (Erwin-Schroedinger-
Stipendium, J 1917-CHE). Financial support was provided to
H.S.S. by the Van Coeverden Adriani Foundation.
Supporting Information Available: X-ray crystallographic
information file for 12a. This material is available free of charge
via the Internet at http://pubs.acs.org.
Procedure 1a. The reactor was charged with toluene and heated
for 10 min at 100 °C. The hot toluene was removed, and the reactor
was dried under vacuum for 10 min at 100 °C. After cooling to
room temperature, the reactor was filled with argon. The reactor
was then charged with 50 mL of pentane, followed by Et3Al2Cl3
cocatalyst. A suspension of the supported precatalyst in 50 mL of
pentane was cannula transferred to the reactor. All lines to the
reactor were closed, and the temperature was raised to the required
temperature within 5 min while stirring (350 rpm). The reactor was
immediately pressurized with ethylene to the desired pressure after
the desired temperature was reached, and the reaction mixture was
stirred (350 rpm) for the desired time at the desired temperature.
The temperature was controlled by cooling with water or heating
with a heating mantle. The reaction mixture was cooled to 15 °C,
and 20 mL of methanol was added. The polymer particles were
isolated by filtration and dried in an oven at 75 °C. The filtrate
References and Notes
(1) Balsam, M.; Baum, P.; Engelmann, J. Macromol. Mater. Eng. 2001,
286 (4), A7-A12.
(2) Peacock, A. J. Polym. Mater. Sci. Eng. 1999, 80, 291-292.
(3) Peacock, A. J. Handbook of Polyethylene: Structure, Properties, and
Applications; Marcel Dekker: New York, 2000.
(4) Kaminsky, W.; Laban, A. Appl. Catal., A 2001, 222, 47-61.
(5) Wittcoff, H. A.; Reuben, B. G.; Plotkin, J. S. Industrial Organic
Chemicals; John Wiley & Sons: New York, 2004.
(6) Ziegler, K.; Holzkamp, E.; Breil, H.; Martin, H. Angew. Chem. 1955,
67, 426.
(7) Wilke, G. Angew. Chem., Int. Ed. 2003, 42, 5000-5008.
(8) Bohm, L. L. Angew. Chem., Int. Ed. 2003, 42, 5010-5030.

6354 Schrekker et al.
Macromolecules, Vol. 39, No. 19, 2006
(9) Fink, G.; Muelhaupt, R.; Brintzinger, H. H. Ziegler Catalysts: Recent
Scientific InnoVations and Technological ImproVements; Springer:
Berlin, 1995.
(10) Galli, P.; Vecellio, G. Prog. Polym. Sci. 2001, 26, 1287-1336.
(11) Weckhuysen, B. M.; Schoonheydt, R. A. Catal. Today 1999, 51, 215-
221.
(48) Kunrath, F. A.; Mota, F. F.; Casagrande, O. L., Jr.; Mauler, R. S.; de
Souza, R. F. Macromol. Chem. Phys. 2002, 203, 2407-2411.
(49) Hlatky, G. G. Chem. ReV. 2000, 100, 1347-1376.
(50) Chien, J. C. W. Top. Catal. 1999, 7, 23-36.
(51) Kristen, M. O. Top. Catal. 1999, 7, 89-95.
(52) Fink, G.; Steinmetz, B.; Zechlin, J.; Przybyla, C.; Tesche, B. Chem.
ReV. 2000, 100, 1377-1390.
(53) Mackenzie, P. B.; Moody, L. S.; Killian, C. M.; Lavoie, G. G. (Eastman
Chemical Company). Patent PCT WO9962968, priority date Jan 22,
1998.
(54) Simon, L. C.; Patel, H.; Soares, J. B. P.; De Souza, R. F. Macromol.
Chem. Phys. 2001, 202, 3237-3247.
(55) AlObaidi, F.; Ye, Z.; Zhu, S. Macromol. Chem. Phys. 2003, 204,
1653-1659.
(56) Vaughan, G. A.; Canich, J. A. M.; Matsunaga, P. T.; Gindelberger,
D. E.; Squire, K. R. (Exxon Chemical Patents, Inc.). Patent PCT
WO9748736, priority date June 17, 1996.
(12) Weckhuysen, B. M.; Wachs, I. E.; Schoonheydt, R. A. Chem. ReV.
1996, 96, 3327-3349.
(13) Reitz, W. In Handbook of Polyethylene: Structure, Properties, and
Applications; Peacock A. J., Eds.; Marcel Dekker: New York, 2001;
Vol. 16, pp 737-738.
(14) McKnight, A. L.; Waymouth, R. M. Chem. ReV. 1998, 98, 2587-
2598.
(15) Coates, G. W. Chem. ReV. 2000, 100, 1223-1252.
(16) Scheirs, J.; Kaminsky, W. Metallocene-Based Polyolefins; John Wiley
& Sons: New York, 2000; Vol. 1.
(17) Scheirs, J.; Kaminsky, W. Metallocene-Based Polyolefins; John Wiley
& Sons: New York, 2000; Vol. 2.
(57) Peruch, F.; Cramail, H.; Deffieux, A. Macromolecules 1999, 32, 7977-
7983.
(18) Ittel, S. D.; Johnson, L. K.; Brookhart, M. Chem. ReV. 2000, 100,
(58) Arthur, S. D.; Teasley, M. F.; Kerbow, D. L.; Fusco, O.; Dall’Occo,
T.; Morini, G. (E. I. Du Pont de Nemours & Co. and Basell
Technology Company B.V.). Patent PCT WO0168725, priority date
March 9, 2000.
(59) Bansleben, D. A.; Connor, E. F.; Grubbs, R. H.; Henderson, J. I.;
Younkin, T. R.; Nadjadi, A. R., Jr. (Cryovac, Inc.). Patent PCT
WO0056786, priority date March 17, 1999.
(60) Bansleben, D. A.; Connor, E. F.; Grubbs, R. H.; Henderson, J. I.;
Younkin, T. R.; Nadjadi, A. R., Jr. (Cryovac, Inc.). Patent PCT
WO0056787, priority date March 17, 1999.
(61) Haddleton, D. M.; Radigue, A.; Kukulj, D.; Duncalf, D. (University
of Warwick). Patent PCT WO9928352, priority date Dec 2, 1998.
(62) Boussie, T. R.; Murphy, V.; Hall, K. A.; Coutard, C.; Dales, C.; Petro,
M.; Carlson, E.; Turner, H. W.; Powers, T. S. Tetrahedron 1999, 55,
11699-11710.
1169-1203.
(19) AlObaidi, F.; Ye, Z.; Zhu, S. J. Polym. Sci., Part A: Polym. Chem.
2004, 42, 4327-4336.
(20) Galland, G. B.; Quijada, R.; Rojas, R.; Bazan, G.; Komon, Z. J. A.
Macromolecules 2002, 35, 339-345.
(21) Pettijohn, T. M.; Reagen, W. K.; Martin, S. J. (Phillips Petroleum
Co.). U.S. Patent 5,331,070, priority date July 19, 1994.
(22) Mark, H. F.; Bikales, N. M.; Overberger, C. G.; Menges, G.
Encyclopedia of Polymers Science and Engineering; John Wiley &
Sons: New York, 1990.
(23) Musaev, D. G.; Morokuma, K. Top. Catal. 1999, 7, 107-123.
(24) Leatherman, M. D.; Svejda, S. A.; Johnson, L. K.; Brookhart, M. J.
Am. Chem. Soc. 2003, 125, 3068-3081.
(25) Skupinska, J. Chem. ReV. 1991, 91, 613-48.
(26) Keim, W.; Hoffmann, B.; Lodewick, R.; Peuckert, M.; Schmitt, G.;
Fleischhauer, J.; Meier, U. J. Mol. Catal. 1979, 6, 79-97.
(27) Keim, W.; Schulz, R. P. J. Mol. Catal. 1994, 92, 21-33.
(28) Vogt, D. In Applied Homogeneous Catalysis with Organometallic
Compounds; Wiley-VCH: Weinheim, 1996; Vol. 1, pp 245-258.
(29) Johnson, L. K.; Killian, C. M.; Brookhart, M. J. Am. Chem. Soc. 1995,
117, 6414-15.
(30) Gates, D. P.; Svejda, S. A.; Onate, E.; Killian, C. M.; Johnson, L. K.;
White, P. S.; Brookhart, M. Macromolecules 2000, 33, 2320-2334.
(31) Mecking, S. Coord. Chem. ReV. 2000, 203, 325-351.
(32) Schmid, M.; Eberhardt, R.; Klinga, M.; Leskelae, M.; Rieger, B.
Organometallics 2001, 20, 2321-2330.
(63) Mendez Llatas, L.; Munoz-Escalona, L. A.; Campora Perez, J.;
Carmona Guzman, E.; Lopez Reyes, M. (Repsol Quimica S.A.)
European Patent EP 1134225, priority date March 13, 2000.
(64) Nesterov, G. A.; Zakharov, V. A.; Fink, G.; Fenzl, W. J. Mol. Catal.
1991, 69, 129-136.
(65) Kaul, F. A. R.; Puchta, G. T.; Schneider, H.; Bielert, F.; Mihalios,
D.; Herrmann, W. A. Organometallics 2002, 21, 74-82.
(66) Mendez Llatas, L.; Munoz-Escalona, L. A.; Campora Perez, J.;
Carmona Guzman, E.; Lopez Reyes, M. (Repsol Quimica S.A.).
European Patent EP 1134236, priority date Mar 13, 2000.
(67) Preishuber-Pflugl, P.; Brookhart, M. Macromolecules 2002, 35, 6074-
6076.
(68) Preishuber-Pflugl, P. (E. I. Du Pont de Nemours & Co. and University
of North Carolina at Chapel Hill). Patent PCT WO02079276, priority
date March 28, 2001.
(69) Interestingly, both the E,E- and E,Z-isomer of R-diimine 8b were
observed at room temperature by NMR spectroscopy. Rates and the
mechanism of interconversion of these isomers can be determined by
variable temperature NMR analysis. An NMR study of the dynamic
behavior of R-diimine ligands will be reported separately.
(70) Teasley, M. F.: (E. I. Du Pont de Nemours & Co.). Patent PCT
WO0022007, priority date Oct 13, 1999.
(33) Camacho, D. H.; Salo, E. V.; Ziller, J. W.; Guan, Z. Angew. Chem.,
Int. Ed. 2004, 43, 1821-1825.
(34) Brookhart, M. S.; Johnson, L. K.; Killian, C. M.; Arthur, S. D.;
Feldman, J.; McCord, E. F.; McLain, S. J.; Kreutzer, K. A.; Bennett,
M. A.; Coughlin, E. B.; Ittel, S. D.; Parthasarathy, A.; Tempel, D. J.
(E. I. Du Pont de Nemours & Co. and University of North Carolina
at Chapel Hill). Patent PCT WO9623010, priority date April 3, 1995.
(35) Small, B. L.; Brookhart, M.; Bennett, A. M. A. J. Am. Chem. Soc.
1998, 120, 4049-4050.
(36) Birtovsek, G. J. P.; Gibson, V.; Kimberley, B. S.; Maddox, P. J.;
McTavish, S. J.; Solan, G. A.; White, A. J. P.; Williams, D. J. Chem.
Commun. 1998, 7, 849-850.
(37) Bennett, A. M. A. (E. I. Du Pont de Nemours & Co.). Patent PCT
WO9827124, priority date Dec 17, 1996.
(38) Britovsek, G. J. P.; Gibson, V. C.; Wass, D. F. Angew. Chem., Int.
Ed. 1999, 38, 428-447.
(39) Gibson, V. C.; Spitzmesser, S. K. Chem. ReV. 2003, 103, 283-315.
(40) Svejda, S. A.; Johnson, L. K.; Brookhart, M. J. Am. Chem. Soc. 1999,
121, 10634-10635.
(41) Britovsek, G. J. P.; Bruce, M.; Gibson, V. C.; Kimberley, B. S.;
Maddox, P. J.; Mastroianni, S.; McTavish, S. J.; Redshaw, C.; Solan,
G. A.; Stroemberg, S.; White, A. J. P.; Williams, D. J. J. Am. Chem.
Soc. 1999, 121, 8728-8740.
(42) Tempel, D. J.; Johnson, L. K.; Huff, R. L.; White, P. S.; Brookhart,
M. J. Am. Chem. Soc. 2000, 122, 6686-6700.
(43) Cotts, P. M.; Guan, Z.; McCord, E.; McLain, S. Macromolecules 2000,
33, 6945-6952.
(71) Zheng, X.; Smit, M.; Chadwick, J. C.; Loos, J. Macromolecules 2005,
38, 4673-4678.
(72) AlObaidi, F.; Ye, Z.; Zhu, S. Polymer 2004, 45, 6823-6829.
(73) Knoke, S.; Ferrari, D.; Tesche, B.; Fink, G. Angew. Chem., Int. Ed.
2003, 42, 5090-5093.
(74) McKenna, T. F.; Soares, J. B. P. Chem. Eng. Sci. 2001, 56, 3931-
3949.
(75) Ha, K.-S.; Yoo, K.-Y.; Rhee, H.-K. J. Appl. Polym. Sci. 2001, 79,
2480-2493.
(76) Still, W. C.; Kahn, M.; Mitra, A. J. Org. Chem. 1978, 43, 2923-
2925.
(77) Daugulis, O.; Brookhart, M. Organometallics 2002, 21, 5926-5934.
(78) Tom Dieck, H.; Svoboda, M.; Greiser, T. Z. Naturforsch., B: Chem.
Sci. 1981, 36, 823-832.
(79) McCloskey, J. A.; White, E. V. J. Org. Chem. 1970, 35, 4241-4244.
(80) Hon, Y.-S.; Chang, F.-J.; Lu, L.; Lin, W.-C. Tetrahedron 1998, 54,
5233-5246.
(81) Pangborn, A. B.; Giardello, M. A.; Grubbs, R. H.; Rosen, R. K.;
Timmers, F. J. Organometallics 1996, 15, 1518-1520.
(82) Lu, Z.-l.; Mayr, A.; Cheung, K.-K. Inorg. Chim. Acta 1999, 284, 205-
214.
(83) Kim, Y. H.; Kim, T. H.; Lee, B. Y.; Woodmansee, D.; Bu, X.; Bazan,
G. C. Organometallics 2002, 21, 3082-3084.
(44) Galland, G. B.; De Souza, R. F.; Mauler, R. S.; Nunes, F. F.
Macromolecules 1999, 32, 1620-1625.
(45) Guan, Z.; Cotts, P. M.; McCord, E. F.; McLain, S. J. Science 1999,
283, 2059-2062.
(46) Moehring, V. M.; Fink, G. Angew. Chem. 1985, 97, 982-984.
(47) Schubbe, R.; Angermund, K.; Fink, G.; Goddard, R. Macromol. Chem.
Phys. 1995, 196, 467-478.
MA061032V