The role of bulky substituents in the polymerization of ethylene using late transition metal catalysts: A comparative study of nickel and iron catalyst systems

Article abstract of DOI:10.1016/S0020-1693(02)01293-8

A series of nickel(II) and iron(II) complexes of the general formula [LMX₂] containing bidentate (for M=Ni) and tridentate (for M=Fe) heterocycle-imine ligands L have been synthesized and characterized. Compared to the well-known α-diimine nickel and bis(imino)pyridine iron catalysts, these systems contain a bulky imine substituent on one side and a non-bulky N-heterocycle on the other. Depending on the ligand and the conditions used, either four- or five-coordinate complexes are obtained in the case of nickel. Iron complexes are generally five-coordinate, even with potentially tetradentate ligands. Activation of these precatalysts with MAO affords active catalyst systems for the oligomerization/polymerization of ethylene. Compared to α-diimine nickel and bis(imino)pyridine iron catalysts, both metal systems provide only half of the steric protection and consequently the catalytic activities and the degree of polymerization are significantly lower. Lower activities are attributed to a reduced stability of the active species under polymerization conditions, whereas the lower molecular weights are a result of increased β-H transfer rates. Variations within the heterocyclic component of the ligand reveal that both steric and electronic factors influence the polymerization behavior of these catalysts.

Full text of DOI:10.1016/S0020-1693(02)01293-8

Inorganica Chimica Acta 345 (2003) 279ꢁ291
/
www.elsevier.com/locate/ica
The role of bulky substituents in the polymerization of ethylene using
late transition metal catalysts: a comparative study of nickel and iron
catalyst systems
George J.P. Britovsek, Simon P.D. Baugh, Olivier Hoarau, Vernon C. Gibson *,
Duncan F. Wass, Andrew J.P. White, David J. Williams
Department of Chemistry, Imperial College of Science, Technology and Medicine, South Kensington, London SW7 2AY, UK
Received 13 June 2002; accepted 28 June 2002
Dedicated in honor of Professor Richard R. Schrock
Abstract
A series of nickel(II) and iron(II) complexes of the general formula [LMX2] containing bidentate (for Mꢀ
/Ni) and tridentate (for
MꢀFe) heterocycle-imine ligands L have been synthesized and characterized. Compared to the well-known a-diimine nickel and
/
bis(imino)pyridine iron catalysts, these systems contain a bulky imine substituent on one side and a non-bulky N-heterocycle on the
other. Depending on the ligand and the conditions used, either four- or five-coordinate complexes are obtained in the case of nickel.
Iron complexes are generally five-coordinate, even with potentially tetradentate ligands. Activation of these precatalysts with MAO
affords active catalyst systems for the oligomerization/polymerization of ethylene. Compared to a-diimine nickel and
bis(imino)pyridine iron catalysts, both metal systems provide only half of the steric protection and consequently the catalytic
activities and the degree of polymerization are significantly lower. Lower activities are attributed to a reduced stability of the active
species under polymerization conditions, whereas the lower molecular weights are a result of increased b-H transfer rates. Variations
within the heterocyclic component of the ligand reveal that both steric and electronic factors influence the polymerization behavior
of these catalysts.
# 2002 Elsevier Science B.V. All rights reserved.
Keywords: Nickel; Iron; Polymerization; Ethylene; Polyethylene
1. Introduction
emphasized the general characteristics believed to be
responsible for highly active catalysts, such as electron
deficiency and coordinative unsaturation at the metal
center [1]. More significantly, these systems have also
highlighted a relatively new factor that is now recog-
nized to be of central importance for obtaining high
activities and high molecular weight product in nickel
and iron catalyst systems: bulky aryl substituents. For
both types of catalysts, A and B, small aryl substituents
result in the formation of low molecular weight oligo-
The field of olefin polymerization catalysis has made
enormous advances during the last decade. In addition
to the well established ZieglerꢁNatta and metallocene
catalyst systems, several late transition metal based
catalysts have shown great potential for the polymeriza-
/
tion of ethylene [1ꢁ
A (Fig. 1) [4ꢁ6] and bis(imino)pyridine iron catalysts of
type B [7ꢁ12], when activated with MAO, are highly
/
3]. Nickel a-diimine catalysts of type
/
/
mers [13ꢁ18], whereas bulky aryl substituents give high
/
active for the polymerization of ethylene. In terms of
understanding the underlying principles governing ole-
fin polymerization, these late metal systems have re-
molecular weight polyethylene [5,11].
Unusual behavior of polymerization catalysts con-
taining bulky substituents has been noted previously in a
number of polymerization systems. Wu and Swift
reported in 1972 that, in the oligomerization of isoprene
with iron(II) catalysts containing iminopyridine ligands
* Corresponding author
E-mail address: v.gibson@ic.ac.uk (V.C. Gibson).
0020-1693/02/$ - see front matter # 2002 Elsevier Science B.V. All rights reserved.
PII: S 0 0 2 0 - 1 6 9 3 ( 0 2 ) 0 1 2 9 3 - 8

280
G.J.P. Britovsek et al. / Inorganica Chimica Acta 345 (2003) 279ꢁ291
/
metal ethylene polymerization catalysts, for example,
systems containing [N,O], [P,O] and [P,P] ligands (see
Fig. 2) [22ꢁ24].
/
As part of our ongoing research in the field of late
transition metal olefin polymerization catalysis
[11,18,25ꢁ28], we describe here the synthesis of a series
/
of hybrid ligands containing a bulky arylimino sub-
stituent on one side of the ligand and a relatively
unhindered heterocyclic donor on the other. These
ligands thus provide only half of the steric protection
of derivatives with bulky substituents on both donor
atoms. The catalytic ethylene oligo- and polymerization
properties of nickel and iron precatalysts, i.e. nickel
complexes with bidentate imino-heterocycle ligands of
types C and D and iron complexes with tridentate
ligands of type E have been investigated and compared
with the parent a-diimine nickel and bis(imino)pyridine
iron precatalysts, with the aim of gaining a better
understanding of the role and exceptional influence of
bulky aryl substituents in polymerization catalysts.
Fig. 1. Nickel and iron olefin polymerization precatalysts.
(ligands as in complex C, Fig. 1), the bulky 2,6-
dimethylphenyl derivative showed an exceptionally
high activity and selectivity for the cyclodimerization
of isoprene, which was noted at the time as ‘anomalous
behavior’ [19]. During the 1980s Keim reported nickel
catalysts for the polymerization of ethylene that fea-
tured moderately bulky groups such as SiMe3 and very
bulky supermesityl substituents (see Fig. 2) [20,21].
However, variations on the substituents were not
explored and the effect of bulky groups remained largely
unremarked upon. It was not until the mid-1990s, when
Brookhart and co-workers showed that nickel com-
plexes with bulky a-diimine ligands afford highly active
catalysts for the polymerization of ethylene, and that the
degree of polymerization and activity is highly depen-
dent on the substitution pattern of the aryl groups, that
the bulky substituent effect became more obviously
apparent. More recently, bulky aryl substituents have
been used to give other highly active late transition
2. Results and discussion
2.1. Synthesis of ligands
Bidentate imino-heterocycle ligands 1ꢁ4, shown in
/
Fig. 3, were synthesized by condensation of the parent
anilines and heterocyclic aldehydes or ketones according
to known procedures [29]. Pyrazine carboxaldehyde and
8-quinoline carboxaldehyde were prepared by selective
oxidation of the corresponding methyl precursors
[30,31].
The tridentate iminobipyridine ligand 5 was obtained
by condensation of 6-acetyl-2,2?-bipyridine [32] with 2,6-
Fig. 2. Nickel catalysts with bulky substituents used in olefin poly-
merization.
Fig. 3. Bidentate and tridentate nitrogen ligands used in this study.

G.J.P. Britovsek et al. / Inorganica Chimica Acta 345 (2003) 279ꢁ
/291
281
diisopropylaniline [33]. The bis(imino)bipyridine and
bis(imino)phenanthroline ligands 6 and 7 were prepared
by selective oxidation of the methyl precursors to give
6,6?-diformyl-2,2?-bipyridine [34] and 2,9-diformyl phe-
nanthroline [35], followed by condensation with the
appropriate anilines (2,4,6-trimethylaniline in the case of
6 and 2,6-diisopropylaniline for 7).
Table 1
˚
Selected bond lengths (A) and angles (8) for [(1a)NiBr2(MeCN)]
Bond lengths
NiÃ
NiÃ
NiÃ
/
N(23)
N(10)
Br(1)
2.045(4)
2.068(3)
2.4383(9)
NiÃ
NiÃ
C(9)ÃN(10)
/
N(3)
2.059(3)
2.4251(9)
1.261(5)
/
/Br(2)
/
/
Bond angles
N(23)ÃNiÃN(3)
N(3)ÃNiÃ
N(3)ÃNiÃ
N(23)Ã
N(10)Ã
/
/
169.6(2)
78.78(12)
92.29(9)
92.19(13)
105.11(9)
N(23)Ã
N(23)Ã
N(10)Ã
N(3)ÃNiÃ
Br(2)ÃNiÃ
/
NiÃ
NiÃ
NiÃ
/
N(10)
Br(2)
Br(2)
91.47(14)
91.23(12)
129.67(9)
93.78(9)
/
/
N(10)
Br(2)
/
/
/
/
/
/
2.2. Synthesis of nickel complexes
/
NiÃ
/Br(1)
/
/
Br(1)
/
NiÃ
/Br(1)
/
/
Br(1)
124.98(3)
Reaction of ligands 1ꢁ4 with (dme)NiBr2 in dichlor-
/
omethane produces orange nickel dibromide complexes
in high yield. Elemental analysis is in all cases consistent
with the stoichiometry (ligand)NiBr2. 1H NMR spectro-
scopy shows all complexes to be paramagnetic, giving
bridging bromide. The binding of this solvent molecule
also reflects the more open coordination environment
afforded by the less bulky iminopyridine ligand com-
pared to the a-diimine ligands.
broad contact-shifted peaks in the range ꢂ20 to 80
/
ppm, indicating that the complexes possess a tetrahedral
geometry as seen for (a-diimine)NiBr2 complexes [5].
Mass spectral data (FAB) show peaks due to both
fragments [(ligand)NiBr2]ꢃ and [(ligand)NiBr2]2ꢃ, sug-
gesting that dimeric complexes may have been formed.
Indeed, five-coordinate dimeric structures for complexes
[(1a)NiBr2]2 and [(1b)NiBr2]2 have been reported by
Laine et al. [36,37]. We have confirmed the dimeric
structure of [(1a)NiBr2]2 when crystallized from a
saturated dichloromethane solution, but we have also
found that the solid state structure of these iminopyr-
idine nickel complexes is affected by the type of ligand
and also by the solvent used for crystallization. Cooling
a saturated acetonitrile solution of complex [(1a)NiBr2]
yielded yellow/orange needles of a monomeric acetoni-
The X-ray structure of [(1a)NiBr2(CH3CN)] shows
the geometry at nickel to be slightly distorted trigonal
bipyramidal, the equatorial angles ranging between
105.11(9) and 129.67(9)8, the largest deviation from
1208 being associated with N(10) and Br(1); the trans-
axial angle is 169.6(2)8. The equatorial coordination
˚
plane is planar to within 0.048 A, and the nickel atom is
˚
displaced by 0.064 A out of the N(10)/Br(1)/Br(2) plane
in the direction of the coordinated acetonitrile ligand.
The geometry of the chelating ligand is essentially
unchanged from those observed in the closely related
dimeric complexes [(1a)NiBr2]2 and [(1a)NiCl2]2 having
˚
a planar (to within 0.007 A) chelate ring, a bite angle of
78.78(12)8, and with the 2,6-diisopropylphenyl ring
system oriented approximately orthogonally (ca. 818)
to this plane [37]. This latter orientation facilitates a pair
trile
adduct.
The
molecular
structure
of
[(1a)NiBr2(CH3CN)] is shown in Fig. 4, with selected
bond lengths and angles in Table 1. In this case, a
preference for a five-coordinate complex is found with
one acetonitrile molecule in the axial coordination site.
Apparently, acetonitrile is a stronger donor than a
of CÃ
/
Hꢀ ꢀ ꢀN(pp) interactions between the isopropyl
methine hydrogen atoms [on C(17) and C(20)] and
˚
N(10), the Hꢀ ꢀ ꢀN distances being 2.46 and 2.44 A,
respectively. The C(9)Ä
/N(10) bond length does not
differ significantly from its counterparts in the related
bromo- and chloro-dimers. The metal coordination
distances are unexceptional and comparable to those
in [(1a)NiBr2]2. The pyridyl rings of centrosymmetrically
related molecules p-stack to form discrete ‘dimers’ with
mean interplanar and ring centroidꢀ ꢀ ꢀcentroid separa-
˚
tions of 3.41 and 3.66 A, respectively.
By contrast, the solid state structure of the quinoly-
limine nickel complex [(3)NiBr2] crystallized from di-
chloromethane, shows a monomeric four-coordinate
complex with a tetrahedral geometry around the nickel
center. No tendency to form a dimeric complex is seen,
most likely a consequence of the larger in-plane steric
demands of this ligand, as seen for a-diimine ligands [5].
The tetrahedral geometry at nickel (Fig. 5) is severely
distorted in this case with angles ranging between
81.1(2) and 130.0(2)8, the acute angle being associated
with the bite of the chelating ligand, and the obtuse with
Fig. 4. The molecular structure of [(1a)NiBr2(MeCN)].

282
G.J.P. Britovsek et al. / Inorganica Chimica Acta 345 (2003) 279ꢁ291
/
netic nature of these complexes indicates a square planar
geometry, in contrast to their dibromide precursors.
Treatment of the dimethyl complexes with
[H(OEt2)2]BArF (BArFꢀ
[B(3,5-(CF3)2C6H3)4]ꢂ) re-
/
sults in the evolution of methane by protonation of
one methyl ligand and formation of the cationic
complexes as diethylether adducts in near quantitative
1
yield by H NMR spectroscopy. These complexes are
extremely air and moisture sensitive and decompose
slowly (half-life approximately 30 min) in d2-dichloro-
methane solution at room temperature, even under an
inert atmosphere. Again, the diamagnetism of these
complexes suggests a square planar geometry. Only one
of the two possible isomers, i.e. with heterocycle or
imine donor trans to the methyl group, was observed.
The structure is presumed to be the isomer with trans
heterocycle and methyl groups, as observed in palladium
analogues [38].
Fig. 5. The molecular structure of [(3)NiBr2].
Br(2) and N(14). The replacement of the pyridine moiety
by quinoline does not significantly effect the overall
ligand conformation, the 2,6-diisopropylphenyl ring
system still being oriented approximately orthogonally
(ca. 778) to the plane of the chelate ring; the associated
isopropyl methine hydrogenꢀ ꢀ ꢀN(14) distances are here
2.3. Synthesis and characterization of iron complexes
Reaction of the ligands 5ꢁ7 with iron dichloride in n-
/
˚
butanol or thf affords the corresponding iron complexes
in high yield. The compounds are sparingly soluble in
common organic solvents, preventing NMR analysis,
but have been characterized by mass spectrometry,
microanalysis and magnetic moment. These data sup-
port the formulation of the complex as LFeCl2. The
bulky aryl substituent prevents the formation of bis-
chelate complexes of the type [L2Fe][FeCl4], which have
been reported for smaller substituents [39,40]. Magnetic
moments are typical for high spin iron(II) complexes
with four unpaired electrons. Based on these experi-
mental data, complex [(5)FeCl2] is most likely a five-
coordinate complex, similar to the many structures we
have determined for bis(imino)pyridine iron and cobalt
complexes [11,18,26]. The bis(imino)bipyridine and
bis(imino)phenanthroline ligands 6 and 7 can coordinate
as tridentate, tetradentate or bridging bidentate ligands
2.37 and 2.47 A. The deviation from planarity of the
˚ ˚
chelate ring is here 0.037
[(1a)NiBr2(MeCN)]), though N(3)/C(4)/C(13)/N(14) are
A (cf. 0.007 A in
˚
˚
planar to within 0.005 A, the nickel atom lying 0.132 A
out of this C2N2 plane (i.e. there is a slight folding of the
ring). The CÄ/N double bond length is unchanged (Table
2). The metal coordination distances to the two nitrogen
atoms are both significantly shorter (by approximately
˚
0.05 A) than those seen in [(1a)NiBr2(MeCN)] as are the
two NiÃBr distances. A similar packing motif to that
seen in [(1a)NiBr2(MeCN)] is observed with pꢁp overlap
/
/
of the quinoline rings of centrosymmetrically related
pairs of molecules; the mean interplanar and ring
˚
centroidꢀ ꢀ ꢀcentroid separations are 3.45 and 3.62 A,
respectively.
Two well-defined cationic nickel alkyl complexes
[(1a)NiMe]ꢃ and [(2)NiMe]ꢃ have been synthesized in
two steps from the parent dibromide complexes accord-
ing to Scheme 1. Treatment of [(1a)NiBr2] and
[(2)NiBr2] with methylmagnesium bromide gives the
highly air and moisture sensitive dimethyl complexes
[(1a)NiMe2] and [(2)NiMe2] in good yield. The diamag-
[41ꢁ
two bands, one for the uncoordinated CÄ
(1635 cmꢂ1, cf. 1639 cmꢂ1 in the free ligand) and
another for the coordinated CÄ
N stretch at 1615 cmꢂ1
/43]. The IR spectrum of complex [(7)FeCl2] shows
/N stretch
/
,
indicating that the ligand acts as a tridentate ligand. The
IR spectrum of complex [(6)FeCl2] is less clear. A broad
absorption is observed at 1634 cmꢂ1, partly overlapping
with a strong band at 1592 cmꢂ1, which probably
obscures the absorption attributable to the coordinated
imine. Mass spectrometry and magnetic moment mea-
surements suggest the formulation LFeCl2. Crystallo-
graphic analysis of complex [(6)FeCl2] indeed supports
this formulation whereby only one imine donor coordi-
nates to the metal center (see Fig. 6).
Table 2
˚
Selected bond lengths (A) and angles (8) for [(3)NiBr2]
Bond lengths
NiÃ
NiÃ
/
N(14)
2.010(5)
2.3049(13)
1.266(8)
NiÃ
NiÃ
/
N(3)
2.014(5)
2.3538(13)
/
Br(2)
/Br(1)
C(13)Ã
Bond angles
N(14)ÃNiÃN(3)
N(3)ÃNiÃ
N(3)ÃNiÃ
/
N(14)
/
/
81.1(2)
117.3(2)
108.2(2)
N(14)Ã
N(14)Ã
Br(2)ÃNiÃ
/
NiÃ
NiÃ
/
Br(2)
Br(1)
Br(1)
130.0(2)
101.3(2)
113.74(5)
The geometry at iron in complex [(6)FeCl2] can
probably be best described as severely distorted trigonal
bipyramidal with N(2) and N(3) occupying the axial
/
/
Br(2)
Br(1)
/
/
/
/
/
/

G.J.P. Britovsek et al. / Inorganica Chimica Acta 345 (2003) 279ꢁ
/291
283
Scheme 1.
gonally (approximately 78 and 808). The CÄ
/
N distances
for the coordinated and non-coordinated imino moieties
˚
are 1.275(11) and 1.236(12) A, respectively, the shorter
length reflecting the absence of coordination of the
imino nitrogen atom. There is a small out of plane twist
(approximately 68) within the bipyridyl portion of the
ligand. The N(1)/N(3) containing chelate ring is planar
˚
to within 0.019 A whereas the N(1)/N(2) containing ring
˚
deviates by 0.038 A. The metal coordination distances
are overall very similar to those reported for 2,6-bis[1-
(2,6-diisopropylphenylimino)ethyl]iron(II) chloride [11],
the only significant difference being a lengthening by
˚
approximately 0.03 A of the two ‘axial’ FeÃ/N distances,
although the coordination geometry is noticeably dif-
ferent (distorted trigonal bipyramidal in [(6)FeCl2]
versus distorted square based pyramidal) [11]. The
packing of the molecules in this structure contains a
more extensive pattern of pꢁp stacking interactions than
/
was seen in either [(1a)NiBr2(MeCN)] or [(3)NiBr2]. The
bipyridyl rings of centrosymmetrically related pairs of
molecules mutually overlap (mean interplanar and ring
centroidꢀ ꢀ ꢀcentroid separations of approximately 3.53
Fig. 6. The molecular structure of [(6)FeCl2].
˚
sites and N(1), Cl(1) and Cl(2) forming the equatorial
plane. The equatorial angles range between 113.35(10)
and 131.9(2)8, and the transaxial angle is 148.5(3)8, the
deviation from linearity being a consequence of the two
chelate bite angles (Table 3). In common with the
structures of [(1a)NiBr2(MeCN)] and [(3)NiBr2] dis-
cussed above, [(6)FeCl2]) has approximate Cs molecular
symmetry, having a near planar ligand backbone with
the terminal mesityl ring systems oriented nearly ortho-
and 3.68 A, respectively), this stacking motif extending
to include the mesityl ring of the non-coordinated ligand
arm of the ligand which overlays the N(1) containing
pyridyl ring of a symmetry related complex (mean
interplanar and ring centroidꢀ ꢀ ꢀcentroid separations of
˚
approximately 3.48 and 3.78 A, respectively) to create a
continuous chain of pꢁp linked complexes.
/
3. Polymerization results
Table 3
˚
Selected bond lengths (A) and angles (8) for [(6)FeCl2]
3.1. Ethylene polymerization using nickel complexes
(L)NiBr2
Bond lengths
FeÃ
FeÃ
FeÃ
/
N(1)
Cl(2)
Cl(1)
2.108(7)
2.272(3)
2.323(3)
1.236(12)
FeÃ
FeÃ
N(3)ÃC(11)
/
N(3)
N(2)
2.266(8)
2.282(8)
1.275(11)
The ethylene polymerization activity of the nickel and
iron dihalide complexes was investigated using MAO as
co-catalyst. The activity values and polymer data are
reported in Table 4, together with data for the two
single-component [LNiMe]ꢃ catalysts. Alkylaluminium
halides (Et3Al2Cl3 or Et2AlCl) also yield active catalytic
systems and indicate the tolerance of these iminopyr-
idine catalysts to a wide variety of activating agents,
similar to a-diimine systems [44].
/
/
/
/
N(4)Ã
/
C(21)
Bond angles
N(1)Ã
N(3)Ã
N(3)Ã
N(1)Ã
Cl(2)Ã
/
FeÃ
FeÃ
FeÃ
FeÃ
FeÃ
/
N(3)
Cl(2)
N(2)
Cl(1)
74.8(3)
102.8(2)
148.5(3)
114.7(2)
113.35(10)
N(1)Ã
N(1)Ã
Cl(2)Ã
N(3)ÃFeÃ
N(2)ÃFeÃ
/
FeÃ
FeÃ
FeÃ
/
Cl(2)
131.9(2)
73.8(3)
99.6(2)
96.6(2)
94.4(2)
/
/
/
/N(2)
/
/
/
/
N(2)
/
/
/
/Cl(1)
/
/Cl(1)
/
/Cl(1)

284
G.J.P. Britovsek et al. / Inorganica Chimica Acta 345 (2003) 279ꢁ291
/
Table 4
Results of ethylene polymerization with nickel and iron precatalysts
Yield (g) Activity (g mmolꢂ1 hꢂ1 barꢂ1
)
Mw
Mn
Mw/Mn
a
d
d
Run Precatalyst (mmol)
MAO (mmol/equivalents)
a
1
2
3
4
5
6
7
8
[(1a)NiBr2] (10)
[(1b)NiBr2] (10)
[(2)NiBr2] (10)
[(3)NiBr2] (10)
[(4)NiBr2] (10)
[(1a)NiMe]ꢃ (10)
[(2)NiMe]ꢃ (10)
[(5)FeCl2] (6)
2/200
2/200
2/200
2/200
2/200
3.46
0.20
trace
1.3
690
40
540
300
310 1.7
140 2.1
a
a
a
a
b
b
c
B1
/
260
6
oligomers
0.06
1.1
110
60
350
2100
210 1.7
1000 2.1
0.6
17.1
e
e
1.2/200
570
oligomers
0.35
a
b
c
Toluene solvent, 1 bar, 30 min, 25 8C.
Dichloromethane solvent, 1 bar, 60 min, 0 8C.
Isobutane solvent, 5 bar, 1 h, 50 8C, scavenger MAO.
Determined by GPC at 135 8C.
d
e
Determined by GC.
The activity of the iminopyridine catalyst [(1a)NiBr2]
was found to be 690 g mmolꢂ1 hꢂ1 barꢂ1, which is
considerably lower than the activities seen for (a-
diimine)NiBr2 catalysts of type A where RꢀH and
/
Arꢀ
weight of the polyethylene product is also significantly
lower (Mwꢀ540 vs. 76 000) [4]. Compared to the
/
2,6-diisopropylphenyl (Fig. 1) [4]. The molecular
/
iminopyridine catalyst [(1a)NiBr2], the 6-methyl deriva-
tive [(1b)NiBr2] shows much reduced activity (40 g
mmolꢂ1 hꢂ1 barꢂ1, run 2) and a lower molecular
weight than for [(1a)NiBr2], in accord with the observa-
tions of Laine et al. [37]. The in-plane steric bulk of the
6-methyl group seems to play an important role in
determining the activity of this catalyst. To explore this
effect further, we also synthesized and tested complex
[(3)NiBr2], which contains a quinolyl instead of a pyridyl
moiety (run 4). This complex was previously reported as
inactive for the polymerization of ethylene [45]. How-
ever, under the conditions used here, we found that this
catalyst converts ethylene with good activity, but mainly
to butenes and hexenes, i.e. an even more dramatic
decrease in product molecular weight. Noteworthy, a
much lower activity has been observed for the 2,2?-
biquinoline derivative [46].
Fig. 7. Nickel complexes containing six-membered chelate rings.
precatalyst [(3)NiBr2] (vide infra). In addition, very low
activities have also been observed in the case of the
planar b-diketiminate nickel catalysts G, which also
feature a six-membered ring and bulky groups on each
nitrogen donor [48]. However, these systems form
neutral catalysts which are often less active than cationic
systems. The difference in activity is therefore most
likely to be the result of chelate ring size, whereby five-
membered chelates generate more active catalysts than
six-membered chelates.
The pyrazine derivative [(2)NiBr2] is sterically iden-
tical to the active precatalyst [(1a)NiBr2]. It therefore
seems surprising that this complex shows no polymer-
ization activity upon activation with MAO (run 3). It is
possible that the reduced basicity of the pyrazine
The 8-quinoline derivative [(4)NiBr2] was synthesized
to probe the effect of a larger chelate bite angle on the
polymerization activity. In this case, a six-membered
chelate ring is formed compared to the five-membered
rings in the other nickel complexes. A very low activity
of 6 g mmolꢂ1 hꢂ1 barꢂ1 was obtained with this
catalyst under standard conditions (run 5). Low activ-
ities have also been reported for related b-diimine nickel
catalysts of type F (Fig. 7) [47]. Whereas the difference
in catalytic activity between b- and a-diimine nickel
catalysts may be related to the lack of planarity and
delocalization in the six-membered ring (the ligand
adopts a boat conformation in F), this explanation can
be ruled out for type H as the 2-quinolyl imine ligand 4
is likely to be planar, as seen in the crystal structure of
nitrogen donor (pKaꢀ0.4 vs. 5.2 for pyridine) [58]
/
may lead to catalyst deactivation (via ligand dissocia-
tion). Another possibility is coordination of the remote
nitrogen donor of the pyrazine moiety to Lewis-acidic
aluminium centers of the co-catalyst, which could also
lead to catalyst deactivation. We therefore synthesized
well-defined nickel alkyl complexes that do not require
the presence of a Lewis acidic co-catalyst. Both single-
component catalysts [(1a)NiMe]ꢃ and [(2)NiMe]ꢃ
prove to be active ethylene polymerization catalysts
(runs 6 and 7), though [(1a)NiMe]ꢃ is less active than its
MAO-activated counterpart, possibly due to the absence
of excess MAO to act as a poison scavenger for the more
sensitive cationic alkyl system. The observation of
activity for [(2)NiMe]ꢃ provides strong support for

G.J.P. Britovsek et al. / Inorganica Chimica Acta 345 (2003) 279ꢁ
/291
285
the co-catalyst interfering with the ligand, leading to
catalyst deactivation.
shows a much higher percentage of vinyl end groups
(25.3%) compared to run 1 (6.5%). An increase in
terminal unsaturation has important consequences for
applications of poly(isobutylene) materials, which often
require polymer chain-end functionalization. The 6-
methyl group of compound [(1b)NiBr2] clearly has a
profound effect on the polymer produced, presumably
by influencing the b-elimination/reinsertion mechanisms
(chain walking) which lead to chain isomerization and
the thermodynamically favored internal olefins. One
possibility is that the in-plane steric bulk of this
substituent disfavors the rotation of intermediate olefin
complexes, which is necessary for the chain walking
mechanism to operate. Since coordinated ethylene must
also rotate into a co-planar orientation to undergo
insertion this would also account for the low activity
observed for this catalyst.
3.2. Analysis of polyethylene produced with nickel
catalysts
In contrast to the high molecular weight polymer
formed by a-diimine nickel catalysts of type A (Rꢀ
/
H,
Arꢀ2,6-diisopropylphenyl) [4], the polymer materials
/
generated by iminopyridine nickel catalysts are essen-
tially low molecular weight, highly branched polyethy-
lenes. These highly branched materials resemble the
polyolefins formed by the polymerization of feedstocks
such as isobutylene [49,50], more than traditional grades
of polyethylene. Poly(isobutylene) has many applica-
tions and are among the most important compounds
currently employed as engine lubricant additives. The
molecular weights are only slightly higher than materials
normally considered to be oligomers and are typically
The material synthesized using the well-defined cation
[(1a)NiMe]ꢃ (run 6) closely resembles the polyethylene
obtained by MAO activation of the related dibromide
catalyst [(1a)NiBr2] (run 1), albeit with slightly lower
molecular weight. This is attributable to the differences
in temperature at which these experiments were per-
formed and/or the nature of the activation method/
anion used. Anion influences on polymer structure have
been reported for palladium diimine catalysts [47].
Comparison of the materials obtained from runs 6 and
7 (Table 4) shows a significant increase in molecular
weight using the pyrazyl-imine catalyst [(2)NiMe]ꢃ
compared to the iminopyridine catalyst [(1a)NiMe]ꢃ.
The material from run 7 also proves to be more linear.
As well as both polymerization runs being performed
under identical conditions, the well-defined cationic
catalysts have identical steric constraints. The only
difference between the two catalysts is the electronic
difference between the pyridyl and pyrazyl groups.
Thus, runs 6 and 7 show an example of electronic
influence over polymer structure. Electronic control in
polymerization/oligomerization has been reported for
neutral nickel catalysts supported by anionic pyridine
carboxylate and pyrazine carboxylate ligands. In this
case a change of mechanism is proposed to be respon-
sible [51]. Comparison of the pKa of pyridine and
pyrazine (5.2 and 0.4, respectively) [58] reveals the
additional nitrogen of the pyrazine heterocycle has an
electron withdrawing effect. It is possible that this
electron withdrawing effect influences (reduces) the
overall rate of b-H elimination, which would lead to
the observed higher molecular weight, and more linear,
material. Another possibility, which has been proposed
by Ziegler and co-workers for diimine systems, involves
electron withdrawing substituents enhancing a torsional
role of the bulky aryl groups, which aids chain
propagation and disfavors branching and chain transfer
[52]. By whatever mechanism it is asserted, electronic
control in late transition metal systems could offer a
characterized using GCꢁ
/
MS and SchulzꢁFlory analysis.
/
However, the materials described here contain a large
number of isomers for each Cn-fraction because of chain
branching and olefin isomerization. This hinders stan-
dard oligomer analysis by GC and requires polymer
characterization techniques (GPC and NMR), which
deal with averaged characteristics. Low molecular
weights are believed to be obtained because imino-
heterocycle ligands provide only half of the ortho aryl
steric protection compared to symmetric a-diimine
ligands. This reduced steric protection of the square
planar catalysts’ axial coordination sites is expected to
lead to higher rates of chain transfer compared to a-
diimine systems.
Runs 1 and 2 explore the differences in polymeriza-
tion behavior between pyridyl and 6-methyl pyridyl
derivatives with identical imine substitution patterns
[(1a)NiBr2] and [(1b)NiBr2]. The polymer from run 1 has
a molecular weight Mwꢀ540, which represents a
material with an average chain length of approximately
C35. The polyethylene obtained from run 2 is slightly
/
lower molecular weight; a value of Mwꢀ300 represents
/
a C20 material. The polydispersities (Mw/Mn) for both
materials are narrow and branching levels are similar. A
more detailed breakdown of the types of branching
present is problematic because polymer chain ends are
indistinguishable from branches, which leads to a large
overestimation in the amount of long (C6ꢃ) branches
/
for such low molecular weight materials. Examination
of the DEPT 90 spectra of the polymers reveals that by
far the largest CH signal is that assigned to methyl
branches which suggests that the predominant branch-
ing features are methyl groups.
The most dramatic difference between the materials is
found upon examination of the distribution of olefinic
groups (see Table 5). Although both polymers contain
predominantly internal olefins, the material from run 2

286
G.J.P. Britovsek et al. / Inorganica Chimica Acta 345 (2003) 279ꢁ291
/
Table 5
Polymer analysis results for runs 1,2 and 6
a
a
Run
Precatalyst
Branches
Me
Olefinic groups
Vinyl
Et
ÈBu
2-Alkene
È3-Alkene
1
2
6
[(1a)NiBr2]
[(1b)NiBr2]
[(1a)NiMe]ꢃ
7.3
9.7
8.9
1.8
3.4
2.9
8.7
11.0
10.9
6.5
25.3
3.6
35.2
32.7
37.8
50.0
34.3
47.3
a
Determined by 13C NMR. Values given are as % of total number of carbons.
useful additional handle to supplement the more usual
steric methods for influencing polymer microstructure.
4. Mechanistic considerations
In the case of nickel a-diimine catalysts, Brookhart
and co-workers have shown that the catalyst resting
state is a nickel alkyl olefin complex [5]. The rate
determining step in the catalytic cycle is the migratory
insertion of ethylene into the metal alkyl bond, and
therefore the rate of propagation for these catalyst
systems shows a zero order dependence on ethylene.
Ziegler has suggested that chain transfer via transfer to
monomer is suppressed by bulky substituents, which
destabilize the transition state and thereby increase the
activation barrier for this chain transfer process [52,54].
Thus, smaller substituents should lead to lower mole-
cular weight products, which is supported by our results
3.3. Ethylene polymerization using iron complexes
(L)FeCl2
The catalytic potential of the iron complexes contain-
ing tridentate ligands has been evaluated using MAO as
the co-catalyst [33,53]. From Table 4 (run 8) it can be
seen that, compared to the parent bis(imino)pyridine
iron catalyst which is highly active and produces high
molecular weight polyethylene (Act.ꢀ
/
5000 g mmolꢂ1
hꢂ1 barꢂ1, Mwꢀ
/500 000) [11], the imino bipyridine
derivative [(5)FeCl2] shows a lower activity of 570 g
mmolꢂ1 hꢂ1 barꢂ1 and produces very low molecular
weight oligomers. This complex shows similar behavior
to the previously reported oligomerization catalysts,
containing bis(imino)pyridine ligands with small aryl
substituents [18]. The product mixture consists of mainly
on the unsymmetrical ligandꢁ/nickel systems reported
here.
These same arguments may apply to bis(imino)pyr-
idine iron catalysts, although the resting state of the
catalyst is likely to be different. We have shown
previously that the rate of propagation for these
catalysts shows a first order dependence on ethylene
[11], which indicates that the rate determining step must
involve ethylene. We therefore propose that the catalyst
resting state is an iron alkyl species and that the rate
determining step is ethylene coordination to the iron
center. This would imply that catalysts containing less
sterically demanding ligands should provide easier
access to the metal center and thus should give higher
activities. However, reducing the steric protection of the
metal center can lead to more rapid decomposition of
the active species. Indeed, the activity profile of run 8
(ethylene flow vs. time) shows a much faster deactiva-
tion of the catalyst compared to typical bis(imino)pyr-
idine iron catalysts and we believe this to be responsible
for the lower activities found for [(5)FeCl2]. The rate of
chain transfer via transfer to monomer is first order in
ethylene and more bulky substituents will decrease this
rate, leading to higher molecular weight polyethylene
[11]. Reducing the sterics, either on both sides or only on
one side as described here, results in effectively only half
of the steric protection and therefore in lower activities
and lower molecular weight products.
1-butene and 1-hexene and follows a typical Schulzꢁ
/
Flory type distribution with a very low a-value of
0.35. Apparently, a big aryl substituent on one side of
the complex and the absence of steric protection on the
other side has an even more dramatic effect than having
a small aryl substituent on each side.
In the case of the complexes [(6)FeCl2] and [(7)FeCl2],
which contain a tridentate imino bipyridine and an
imino phenanthroline ligand with a pendant imino
substituent, respectively, no catalytic activity was ob-
served upon activation with MAO. It could be argued
that the pendant imino donor can coordinate to the
metal center in the activated species, thereby blocking a
coordination site on the metal that may be needed for
catalysis [42]. However, we have shown recently that
bis(imino)pyridine iron and cobalt complexes contain-
ing additional donors such as acetonitrile or thf do not
significantly affect the catalytic performance of these
catalysts [27]. We believe that, similar to the case of the
nickel complex [(1b)NiBr2] described earlier, this lack of
activity may be related to the in-plane steric bulk created
by the ortho substituent on the second pyridyl unit.

G.J.P. Britovsek et al. / Inorganica Chimica Acta 345 (2003) 279ꢁ
/291
287
5. Experimental
concentrated and cooled to ꢂ
/
30 8C. Compound 2 was
isolated as large yellow blocks in 83% yield (2.0 g).
1H NMR (250 MHz, CDCl3, 298 K): d 9.52 (s, 1H, 3-
5.1. General
pyzÃ
CH), 7.22ꢁ
Hz, CH(CH3)2), 1.20 (d, 12H, Jꢀ
13C NMR (63 MHz, CDCl3, 298 K): d 161.5 (imineÃ
149.2, 148.1, 145.9, 144.2, 143.6, 137.1, 124.9, 123.2
C), 28.1 (CHMe2), 23.5 (CH(CH3)2). IR
/
H), 8.72ꢁ
/
8.62 (m, 2H, pyzÃ
H), 3.00 (sept, 2H, Jꢀ
6.7 Hz, CH(CH3)2).
C),
/
H), 8.38 (s, 1H, NÄ
/
All manipulations were carried out under an atmo-
sphere of nitrogen using standard Schlenk and cannula
techniques or in a conventional nitrogen filled glove
box. Solvents were refluxed over an appropriate drying
agent, and distilled and degassed prior to use. Elemental
analyses were performed by the microanalytical services
of the Department of Chemistry at Imperial College and
Medac Ltd. NMR spectra were recorded on a Bruker
spectrometer: 1H (250 MHz) and 13C (62.9 MHz) at 293
K; chemical shifts are referenced to the residual protio
impurity of the deuterated solvent. Polyethylene NMR
spectra were recorded on a JEOL GSX270 for the
/
7.12 (m, 3H, ArÃ
/
/
6.7
/
/
(pyzÃ
/
and ArÃ
/
(Nujol, NaCl, cmꢂ1): 1641(s), 1578(m), 1541(w),
1406(m), 1379(vs), 1365(s), 1349(m), 1321(w), 1300(w),
1266(w), 1255(w), 1179(s), 1162(m), 1108(w), 1052(m),
1016(s), 982(w), 967(w), 933(w), 876(m), 856(w), 847(w),
835(w), 807(w), 796(m), 755(m), 723(m), 708(w). MS
(EI, m/z): 267 [M]ꢃ. Elemental analysis for C17H21N3×
/
0.5H2O (276.38) found (required): C, 73.59 (73.87); H,
7.94 (8.02); N, 15.00% (15.20).
1
following nuclei shown (frequencies in parentheses): H
(270.05 MHz) and 13C (67.51 MHz). Mass spectra were
obtained using VG Autospec or VG Platform II Mass
Spectrometers for fast atom bombardment (FAB),
electron impact (EI) or chemical ionization (CI). Gel
permeation chromatographs (GPC) were obtained using
a Waters 150CV (columns supplied by Shodex (807, 806
and 804)) (BP Chemicals Ltd.) at 135 8C. IR spectra
5.2.2. 2-[(2,6-Diisopropylphenylimino)methyl]quinoline
(3)
To a solution of 2,6-diisopropylaniline (0.81 g, 4.58
mmol) and 2-quinolinecarboxaldehyde (0.72 g, 4.58
mmol) in 10 ml CH2Cl2 was added a drop of formic
acid and MgSO4. The mixture was stirred at room
temperature (r.t.) overnight. After filtration, the solvent
was removed and the product recrystallized from warm
C6H14. Compound 3 was isolated as a yellow solid in
77% yield (1.11 g).
were recorded on a PerkinꢁElmer Spectrum GX1
/
System FTIR spectrometer. Magnetic susceptibility
measurements have been performed using an Evans’
balance.
The synthesis of some pyridyl imine ligands (1aꢁ
/b,
1H NMR (250 MHz, CDCl3, 298 K): d 8.51 (s, 1H,
and 3) together with the corresponding Ni complexes
has been reported previously [28,36,38,55,56].
NÄ
3H, ArÃ
(d, 12H, Jꢀ
CDCl3, 298 K): d 163.4 (imineÃ
137.1, 136.7, 129.9, 129.1, 127.8, 124.5, 123.1, 118.3
C), 28.0 (CHMe2), 23.4 (CH(CH3)2).
/
CH), 8.46ꢁ
H), 3.04 (sept, 2H, Jꢀ
6.7 Hz, CH(CH3)2). 13C NMR (63 MHz,
C), 154.6, 148.5, 148.0,
/
7.60 (m, 6H, quinÃ
/
H), 7.25ꢁ7.17 (m,
/
/
/
6.7 Hz, CH(CH3)2), 1.22
[H(OEt2)2]BArF (BArFꢀ
/
[B(3,5-(CF3)2C6H3)4]ꢂ) [57],
/
pyrazine carboxaldehyde [30], 8-quinoline carboxalde-
hyde [31], 6-acetyl-2,2?-bipyridine and the precursors 2-
acetyl-6-bromopyridine and 2-(2?-methyl-1?,3?-dioxo-
lane-2?-yl)-6-bromopyridine [32] were prepared accord-
ing to literature procedures. The full procedure for the
synthesis of ligand 5 has been disclosed previously [33].
The precursors for ligands 6 and 7, 6,6?-diformyl-2,2?-
bipyridine [34] and 2,9-diformyl phenanthroline [35],
were synthesized by selective oxidation of the methyl
precursors. MAO (10% solution in C6H5CH3) was
obtained from Aldrich. All other chemicals were ob-
tained commercially and used as received unless stated
otherwise.
/
(quinÃ
/
and ArÃ
/
Elemental analysis for C22H24N2 (316.45) found (re-
quired): C, 83.67 (83.50); H, 7.64 (7.64); N, 8.67% (8.85).
5.2.3. 8-[(2,6-Diisopropylphenylimino)methyl]quinoline
(4)
An analogous procedure to that described for 3 was
followed using 2,6-diisopropylaniline (0.56 g, 3.2 mmol)
and 8-quinolinecarboxaldehyde (0.5 g, 3.2 mmol).
Compound 4 was isolated as a yellow solid in 80% yield
(0.8 g).
1H NMR (250 MHz, CDCl3, 298 K): d 9.73 (s, 1H,
5.2. Synthesis of ligands
NÄ
quinÃ
6.7 Hz, CH(CH3)2), 1.28 (d, 12H, Jꢀ
CH(CH3)2). 13C NMR (63 MHz, CDCl3, 298 K): d
160.6 (imineÃC), 150.3, 150.1, 147.1, 137.7, 136.0, 133.0,
130.9, 128.2, 127.3, 126.5, 124.0, 122.9, 121.3 (quinà and
/
CH), 8.98, 8.75, 8.20, 8.01, 7.75, 7.47 (m, 1H each,
H), 7.27ꢁ7.12 (m, 3H, ArÃH), 3.15 (sept, 2H, Jꢀ
6.7 Hz,
/
/
/
/
5.2.1. [(2,6-Diisopropylphenylimino)methyl]pyrazine
(2)
/
A discrete mixture of 2,6-diisopropylaniline (1.6 g, 9
mmol) and 2-pyrazinecarboxaldehyde (1.0 g, 9 mmol)
was stirred for 1 h. During this time the mixture became
viscous and heat was evolved. Pentane (approximately
50 ml) was added and the resultant yellow solution dried
over anhydrous Mg2SO4. The solution was then filtered,
/
/
ArÃ
/C), 28.0 (CHMe2), 23.5 (CH(CH3)2). IR (Nujol,
NaCl, cmꢂ1): 1733(w), 1678(w), 1626(s), 1608(m),
1592(w), 1573(m), 1499(m), 1463(s), 1379(s), 1361(m),
1317(m), 1265(w), 1254(m), 1243(m), 1209(w), 1186(w),

288
G.J.P. Britovsek et al. / Inorganica Chimica Acta 345 (2003) 279ꢁ291
/
1162(w), 1112(w), 1081(w), 1060(w), 1047(w), 1034(w),
936(w), 891(m), 832(m), 795(s), 770(w), 760(m), 745(m).
MS (EI, m/z): 317 [M]ꢃ. Elemental analysis for
suspension turned into a yellow solution. The product
precipitates upon cooling to r.t. Compound 7 was
isolated as yellow crystals in 72% yield (1.0 g).
C22H24N2×
/
0.25H2O (320.95) found (required): C, 82.42
1H NMR (CDCl3): d 8.75 (d, 4H, H3,4,7,8), 8.46 (d,
(82.33); H, 7.71 (7.69); N, 8.74% (8.73).
2H, H5,6), 7.98 (s, 2H, HCÄ
/
N), 7.1ꢁ
/
7.2 (m, 6H, ArH),
3.10 (sept, Jꢀ6.8 Hz, 4H, CH(CH3)2), 1.19 (d, Jꢀ
/
/6.8
5.2.4. 6-[1-(2,6-Diisopropylphenylimino)ethyl]-2,2?-
bipyridine (5)
To a solution of 6-acetyl-2,2?-bipyridine (0.45 g, 2.27
mmol) in warm EtOH (1.5 ml) was added glacial AcOH
Hz, 24H, CH(CH3)2). 13C NMR (CDCl3): d 163.1 (C Ä
N), 154.9, 148.1, 145.7, 137.2, 137.0, 130.2, 127.8, 124.6,
123.0, 120.6 (ArC and phenC), 27.9 (CH(CH3)2), 23.3
/
(CH(CH3)2). IR (neat, cmꢂ1): 1638 (s, n(CÄ
/
N)). MS
Me, 20]. Elemental
0.5H2O (563.79) found (re-
(five drops), 4 A mol sieves and 2,6-diisopropylaniline
(EI, m/z): 554 [Mꢃ, 100], 539 [Mꢃꢂ
/
˚
(0.53 g, 3.00 mmol). This mixture was heated to reflux
for 15 h, during which time a solid precipitated. The
mixture was cooled, filtered and the solid washed with
EtOH (3 ml) to give the product as a light yellow solid in
87% yield (0.71 g).
analysis for C38H42N4×
/
quired): C, 80.87 (80.96); H, 7.73 (7.69); N, 9.88% (9.94).
5.3. Synthesis of complexes
1H NMR (CDCl3): d 8.72 (dd, Jꢀ
pyrH), 8.60ꢁ8.55 (m, 2H, pyrH), 8.43 (d, Jꢀ
1H, pyrH), 7.96 (t, Jꢀ7.8 Hz, 1H, pyrH), 7.84 (t, Jꢀ
7.8 Hz, 1H, pyrH), 7.36ꢁ7.31 (m, 1H), 7.22ꢁ7.13 (m,
3H, ArH), 2.81 (sept, Jꢀ6.8 Hz, 2H, CH(CH3)2), 2.36
(s, 3H, H3CCÄN), 1.19 (d, Jꢀ6.8 Hz, 12H, CH(CH3)2).
N), 156.1, 155.7, 154.9,
/
0.9, 4.0 Hz, 1H,
/
/7.8 Hz,
/
/
5.3.1. 2-((2,6-Diisopropylphenylimino)methyl)pyridine
nickel dibromide [(1a)NiBr2]
/
/
/
Dichloromethane (approximately 30 ml) was added to
a mixture of 1a (0.45 g, 1.70 mmol) and (dme)NiBr2
(0.50 g, 1.60 mmol) in a Schlenk vessel. The mixture was
stirred for 18 h during which time an orange precipitate
was formed. This precipitate was isolated by filtration,
/
/
13C NMR (CDCl3): d 167.1 (C Ä
/
149.2, 146.6, 137.4, 136.9, 135.8, 123.8, 123.6, 123.0,
122.0, 121.2, 121.1 (ArC and pyrC), 28.3 (CH(CH3)2),
23.2, 22.9 (CH(CH3)2), 17.2 (H3CCÄ
3071, 2961, 2925, 2869, 1640, 1578, 1571, 1429. MS (EI,
m/z): 357 [Mꢃ, 65], 342 [Mꢃꢂ
Me, 100]. HRMS for
/
N). IR (drifts):
washed with 3ꢄ approximately 10 ml Et2O and dried in
/
vacuo (0.50 g, 64%). Crystals suitable for X-ray diffrac-
tion were grown from a saturated MeCN solution.
Elemental analysis for C18H19Br2N2Ni (481.87) found
(required): C, 44.43 (44.55); H, 4.34 (4.78); N, 5.79%
(5.77). IR data (Nujol, KBr, cmꢂ1): 1594(s), 1572(m),
1491(m), 1461(s), 1378(m), 1304(m), 1026(m), 813(m),
/
C24H27N3 Calc.: 357.2205. Found: 357.2219.
5.2.5. 6,6?-Bis[(2,4,6-trimethylphenylimino)methyl)-
2,2?-bipyridine (6)
To a solution of 2,4,6-trimethylaniline (0.21 g, 1.5
mmol) and 6,6?-diformyl-2,2?-bipyridine (0.13 g, 0.7
mmol) in 10 ml EtOH was added a drop of AcOH.
The mixture was refluxed for 3 h. After cooling down to
r.t. the solvent was removed in vaccuo and the yellow
solid was washed twice with C6H14. Compound 6 was
isolated as yellow powder in 50% yield (0.15 g).
772(s), 722(w). FABMS (m/z, for 79Br): 880 [2Mꢂ
/
Br]ꢃ,
480 [M]ꢃ. Crystal data for [(1a)NiBr2(MeCN)]×
/MeCN:
¯
567.0, triclinic, P1 (no. 2),
C20H25Br2N3Ni×
aꢀ7.243(2), bꢀ
88.43(1), bꢀ84.05(1), gꢀ
Zꢀ2, Dcalc
1.511 g cmꢂ3, m(Cu Ka)ꢀ
Tꢀ293 K, orange needles; 3698 independent measured
reflections, F2 refinement, R1ꢀ
0.041, wR2ꢀ0.104, 3172
independent observed absorption corrected reflections
[jFojꢀ4s(jFoj), 2u 51208], 263 parameters.
/
MeCN, Mꢀ
10.750(1), cꢀ
72.12(1)8, Vꢀ
/
˚
/
/
/
16.914(2) A, aꢀ
/
3
˚
/
/
/
1246.4(2) A ,
/
ꢀ
/
/
4.98 mmꢂ1
,
/
1H NMR (CDCl3): d 8.59 (dd, 2H, Jꢀ
5,5?), 8.43 (s, 2H, HCÄN), 8.31 (dd, 2H, Jꢀ
3,3?), 7.94 (t, 2H, Jꢀ7.8 Hz, H4,4?), 6.92 (s, 4H, ArH),
/
1.0, 7.8 Hz,
/
/
H
H
/
/
1, 7.8 Hz,
/
/
/
2.30 (s, 6H, CH3), 2.17 (s, 12H, CH3). 13C NMR
(CDCl3): d 163.8 (C Ä
/
N), 155.5, 154.2, 147.9, 137.5,
133.4, 128.8, 126.8, 122.6, 122.6, 121.1, 20.7 (CH3), 18.3
(CH3). MS (EI, m/z): 446 [Mꢃ, 100]. Elemental analysis
for C30H30N4 (446.04) found (required): C, 80.80
(80.68); H, 6.66 (6.77); N, 12.67% (12.55). IR (KBr,
5.3.2. ((2,6-Diisopropylphenylimino)methyl)pyrazine
nickel dibromide [(2)NiBr2]
An analogous procedure to that described for
[(1a)NiBr2] was followed, using 2 (0.43 g, 1.6 mmol)
and (dme)NiBr2 (0.50 g, 1.6 mmol). Compound
[(2)NiBr2] was isolated as an orange powder in 75%
yield (0.6 g).
cmꢂ1): 1644 (s, n(CÄ
/
N)).
5.2.6. 2,9-Bis((2,6-diisopropylphenylimino)methyl)-
phenanthroline (7)
To a solution of 2,6-diisopropylaniline (0.90 g, 5.1
mmol) and 2,9-diformyl phenanthroline (0.60 g, 2.5
mmol) in 30 ml EtOH was added a drop of formic acid.
The mixture was refluxed for 2 h. The initially white
IR data (Nujol, KBr, cmꢂ1): 1622, 1463, 1186, 1043,
807. FABMS (m/z, for 79Br): 822 [2Mꢂ
/
Br]ꢃ, 401 [Mꢂ
/
Br]ꢃ. Elemental analysis for C17H18Br2N3Ni (482.86)
found (required): C, 41.84 (42.02); H, 4.31 (4.36); N,
8.35% (8.64).

G.J.P. Britovsek et al. / Inorganica Chimica Acta 345 (2003) 279ꢁ
/291
289
5.3.3. 2-((2,6-Diisopropylphenylimino)methyl)quinoline
nickel dibromide [(3)NiBr2]
Yield: 0.18 g (70%). 1H NMR (250 MHz, CD2Cl2, 298
K): d 8.86 (m, 1H, pyÃH), 8.40 (s, 1H, NÄCH), 8.06
(m, 1H, pyÃH), 7.78ꢁ7.64 (overlapping m, 1H each,
pyÃH), 7.36ꢁ7.20 (m, 3H, ArÃH), 3.22 (sept, 2H, Jꢀ
6.7 Hz, CH(CH3)2), 1.26 (d, 6H, Jꢀ6.7 Hz,
CH(CH3Me?), 1.12 (d, 6H, Jꢀ6.7 Hz, CHMe(CH3)?),
0.40, ꢂ0.14 (s, 3H each, NiÃ
CH3). 13C NMR (63 MHz,
CD2Cl2, 298 K): d 164.3 (imineÃC), 149.9, 148.8, 139.7,
139.3, 129.4, 127.4, 123.9, 123.4 (pyà and ArÃC), 28.2
(CHMe2), 22.9, 22.7 (CH(CH3)2), ꢂ0.1, ꢂ5.9 (NiÃC).
/
/
An analogous procedure to that described for
[(1a)NiBr2] was followed, using 3 (0.64 g, 2.02 mmol)
and (dme)NiBr2 (0.62 g, 2.02 mmol). Compound
[(3)NiBr2] was isolated as an orange powder in 66%
yield (0.71 g). X-ray quality crystals were grown from a
concd. CH2Cl2 solution.
/
/
/
/
/
/
/
/
/
/
/
IR data (Nujol, KBr, cmꢂ1): FABMS (m/z, for 79Br):
/
/
455 [Mꢂ
/
Br, 100]ꢃ, 374 [Mꢂ
/
2Br, 50]ꢃ. Elemental
/
/
/
analysis for C22H24Br2N2Ni (534.96) found (required):
C, 49.51 (49.40); H, 4.76 (4.52); N, 4.98% (5.24). Crystal
IR (Nujol, KBr, cmꢂ1): 1610(m), 1580(m), 1558(w),
1539(w), 1462(m), 1376(m), 1354(s), 1331(m), 1300(m),
1274(w), 1216(w), 1183(m), 1156(s), 1101(m), 1061(m),
1039(m), 1018(s), 981(w), 966(w), 944(w), 934(w), 800(s),
772(s), 721(m). MS (EI, m/z): 356 [M]ꢃ. Due to the
high air/moisture sensitivity of this compound, satisfac-
tory elemental analysis could not be obtained.
data for [(3)NiBr2]: C22H24Br2N2Ni, Mꢀ
clinic, P21/n (no. 14), aꢀ10.064(1), bꢀ18.239(4), cꢀ
99.90(1)8, Vꢀ2214.1(6) A , Zꢀ
1.605 g cmꢂ3, m(Mo Ka)ꢀ4.49 mmꢂ1, Tꢀ
293 K, deep orange/red prisms; 3857 independent
measured reflections, F2 refinement, R1ꢀ
0.055, wR2ꢀ
0.119, 2391 independent observed absorption corrected
reflections [jFojꢀ4s(jFoj), 2u 5508], 244 parameters.
/
535.0, mono-
/
/
/
3
˚
˚
12.244(1) A, bꢀ
/
/
/
4,
Dcalc
ꢀ
/
/
/
/
/
5.3.6. ((2,6-Diisopropylphenylimino)methyl)pyrazine
nickel dimethyl [(2)NiMe2]
/
/
An analogous method to that described for com-
pound [(1a)NiMe2] was followed using compound
[(2)NiBr2] (0.36 g, 0.74 mmol) and MeMgBr (0.5 ml of
3.0 M solution in Et2O, 1.5 mmol). Compound
[(2)NiMe2] was isolated as a highly air, moisture and
temperature sensitive blue/green solid in 60% yield (0.15
g).
IR (Nujol, KBr, cmꢂ1): 1620(w), 1600(m), 1596(m),
1571(w), 1506(w), 1461(s), 1410(m), 1378(s), 1365(s),
1327(m), 1302(w), 1177(w), 1147(w), 1109(w), 1033(m),
982(w), 957(w), 944(w), 933(w), 911(w), 849(w), 831(w),
806(m), 777(w), 745(w), 711(w). MS (FABꢃ, m/z): 356
[M]ꢃ. 1H NMR (250 MHz, CD2Cl2, 298 K): d 9.20 (m,
5.3.4. 8-((2,6-Diisopropylphenylimino)methyl)quinoline
nickel dibromide [(4)NiBr2]
An analogous procedure to that described for
[(1a)NiBr2] was followed, using 4 (0.32 g, 1.0 mmol)
and (dme)NiBr2 (0.31 g, 1.0 mmol). Compound
[(4)NiBr2] was isolated as an orange powder in 80%
yield (0.4 g).
IR data (Nujol, KBr, cmꢂ1): 1628(vs), 1603(vs),
1591(vs), 1580(s), 1516(m), 1464(vs), 1378(s), 1322(w),
1303(w), 1263(m), 1211(w), 1168(s), 1141(w), 1106(m),
1060(w), 1042(w), 1019(w), 1002(w), 986(w), 963(w),
928(m), 887(m), 861(w), 838(m), 804(vs), 781(s), 773(s),
1H, pyzÃ
/
H), 9.12 (m, 1H, pyzÃ
CH), 7.36ꢁ
6.7 Hz, CH(CH3)2), 1.22 (d, 6H,
/
H), 8.88 (m, 1H, pyzÃ
/
735(m), 724(w). FABMS (m/z, for 79Br): 986 [2Mꢂ
Br]ꢃ, [M]ꢃ.
532 Elemental analysis for
C44H48Br4N4Ni2×H2O (1087.93) found (required): C,
/
H), 8.46 (s, 1H, NÄ
/
/
7.18 (m, 3H, ArÃH),
/
2
3.10 (sept, 2H, Jꢀ
2Jꢀ
6.7 Hz, CH(CH3)Me?), 1.08 (d, 12H, Jꢀ
CHMe(CH3)?), 0.60, ꢂ0.24 (s, 3H each, NiÃ
CH3). 13C
NMR (63 MHz, CD2Cl2, 298 K): d 164.3 (imineÃC),
150.1, 146.4, 144.5, 139.8, 138.7, 128.1, 123.6 (pyzà and
ArÃC), 28.3 (CHMe2), 24.9, 22.6 (CH(CH3)2), 0.9, ꢂ7.6
(NiÃC). Due to the high air/moisture sensitivity of this
/
/
2
/
/6.7 Hz,
48.58 (48.58); H, 4.49 (4.63); N, 5.06% (5.15).
/
/
/
5.3.5. 2-((2,6-Diisopropylphenylimino)methyl)pyridine
nickel dimethyl [(1a)NiMe2]
To a stirred suspension of [(1a)NiBr2] (0.36 g, 0.74
/
/
/
/
mmol) in 20 ml Et2O at ꢂ
/
30 8C, MeMgBr (0.5 ml of
compound, satisfactory elemental analysis could not be
3.0 M solution in Et2O, 1.5 mmol) was added dropwise.
This was stirred for 2 h during which time the solution
was allowed to warm to 0 8C. After this time the
solution became a deep blue/green color. 1,4-Dioxane
(approximately 5 ml) was added and a fine precipitate of
obtained.
5.3.7. [2-((2,6-Diisopropylphenylimino)methyl)pyridine
nickel methyl diethyletherate]{tetrakis[3,5-
bis(trifluoromethyl)phenyl]borate} [(1a)NiMe]ꢃ
Diethyl ether (2 ml) was added to [(1a)NiMe2] (17.8
mg, 0.05 mmol) and [H(OEt2)2]BArF (BArFꢀ
(CF3)2C6H3)4]Ã) (50.6 mg, 0.05 mmol) at 0 8C in a
Schlenk vessel. A dark red solution formed almost
immediately which was stirred for 5 min. After this
time solvent was removed under reduced pressure to
yield the cationic complex as a dark red solid. All of this
product was taken up into approximately 1 ml CD2Cl2,
1,4-dioxane×
and the solvent removed under reduced pressure. The
resultant solid was washed with 3ꢄ approximately 10
ml C5H12, redissolved in 5 ml Et2O and cooled to
30 8C to yield highly air and moisture sensitive small
/MgBr2 formed. The solution was filtered
/
[B(3,5-
/
/
ꢂ
/
dark green blocks. Compound [(1a)NiMe2] was also
found to be temperature sensitive and decomposes
slowly in solution or solid state at ambient temperature.

290
G.J.P. Britovsek et al. / Inorganica Chimica Acta 345 (2003) 279ꢁ291
/
transferred to an NMR tube and the NMR spectrum
recorded immediately (ꢀ
95% by 1H NMR spectro-
scopy).
1H NMR (250 MHz, CD2Cl2, 298 K): d 8.86 (m, 1H,
pyÃH), 8.40 (s, 1H, NÄCH), 8.06 (m, 1H, pyÃH), 7.78ꢁ
7.64 (m, 10H, pyÃH and BArF ortho-H), 7.56 (4H,
BArF para-H), 7.36ꢁ7.20 (m, 3H, ArÃH), 3.42ꢁ3.20
(m, 6H, O(CH2CH3)2 and CH(CH3)2), 1.29ꢁ1.18 (m,
12H, O(CH2CH3)2 and CH(CH3)Me?), 1.12 (d, 6H,
CHMe(CH3)?), ꢂ0.30 (s, 3H, NiÃ
CH3). 13C NMR (63
MHz, CD2Cl2, 298 K): d (some signals obscured by
BArF resonances) 169.8 (imineÃC), 162.2 (q, JCB 37.5
Hz, BArF ipso-C), 149.9, 142.2, 141.0, 140.3, 131.5,
118.5 (pyà and ArÃC), 135.3 (BArF ortho-C), 129.3 (q,
JCF 31.3 Hz, BArF meta-C), 125.0 (q, JCF 272.4 Hz,
BArF CF3), 66.5 O(CH2CH3)2), 28.5 (CH(Me)2), 24.6,
22.6 (CH(CH3)2), 15.2 (O(CH2CH3)2), 3.4 (NiÃC).
5.4. 6,6?-Bis[(2,4,6-trimethylphenylimino)methyl]-2,2?-
bipyridine iron dichloride [(6)FeCl2]
/
To a solution of FeCl2 (25 mg, 0.20 mmol) in 10 ml
thf was added 1 equiv. of the imine ligand 6 (92 mg, 0.20
mmol). The mixture was stirred at r.t. overnight. The
solvent was removed in vacuo and the residue washed
/
/
/
/
/
/
/
/
/
with C5H12 (2ꢄ
68% yield (0.80 g).
MS (FAB, m/z): 572 [Mꢃ, 23], 537 [Mꢃꢂ
Cl, 100].
/
15 ml) to provide a green/brown solid in
/
/
/
Elemental analysis for C30H30Cl2FeN4 (573.35) found
(required): C, 62.46 (62.85); H, 5.50 (5.27); N, 8.95%
(9.77). IR (KBr, cmꢂ1): 1634 (broad). meff (Evans’ NMR
/
ꢀ
/
/
/
method)ꢀ
C30H30Cl2FeN4×
(no. 88), aꢀ23.420(4), cꢀ
16, Dcalc
1.346 g cmꢂ3, m(Mo Ka)ꢀ
mmꢂ1, Tꢀ
173 K, very dark olive green prisms; 5377
independent measured reflections, F2 refinement, R1ꢀ
0.081, wR2ꢀ0.174, 2537 independent observed reflec-
tions [jFojꢀ4s(jFoj), 2u 5478], 419 parameters.
/
5.0 BM. Crystal data for [(6)FeCl2]:
2CH2Cl2, Mꢀ743.2, tetragonal, I41/a
26.735(3) A, Vꢀ
ꢀ
/
ꢀ
/
/
/
˚
/
/
/
14664(3)
3
˚
/
A , Zꢀ
/
ꢀ
/
/0.88
/
/
/
5.3.8. [((2,6-Diisopropylphenylimino)methyl)pyrazine
nickel methyl diethyletherate]{tetrakis[3,5-
/
/
bis(trifluoromethyl)phenyl]borate} [(2)NiMe]ꢃ
An analogous method to that described for com-
pound [(1a)NiMe]ꢃ was followed using compound
[(2)NiMe2] (17.8 mg, 0.05 mmol) and [H(OEt2)2]BArF
5.5. 1,9-Bis[(2,6-diisopropylphenylimino)methyl]-
phenanthroline iron dichloride [(7)FeCl2]
90% by 1H NMR
To a solution of FeCl2 (48 mg, 0.38 mmol) in thf was
added 1 equiv. of the imine ligand 7 (0.21 g, 0.38 mmol).
The mixture was stirred at r.t. overnight. The solvent
was removed in vacuo and the residue washed with Et2O
(50.6 mg, 0.05 mmol) (yield ꢀ
spectroscopy).
1H NMR (250 MHz, CD2Cl2, 298 K): d 9.28 (m, 1H,
pyzÃH), 9.16 (m, 1H, pyzÃH), 8.94 (m, 1H, pyzÃH),
8.48 (s, 1H, NÄCH), 7.71 (s, 8H, BArF ortho-H), 7.56
(s, 4H, BArF para-H), 7.38ꢁ7.18 (m, 3H, ArÃH), 3.44ꢁ
3.20 (m, 6H, O(CH2CH3)2 and CH(CH3)2), 1.30ꢁ1.16
(m, 12H, O(CH2CH3)2 and CH(CH3)Me?), 1.10 (d, 6H,
CHMe(CH3)?), 0.02 (s, 3H, NiÃ
CH3). 13C NMR (63
MHz, CD2Cl2, 298 K): d (some signals obscured by
/
/
/
/
(2ꢄ
(0.19 g).
MS (FAB, m/z): 680 [Mꢃ, 40], 645 [Mꢃꢂ
IR (NaCl, cmꢂ1): 1635 (s, n(CÄ
N) uncoordinated) and
N) coordinated). Elemental analysis for
/15 ml) to provide a green/brown solid in 73% yield
/
/
/
/
/Cl, 100].
/
/
1615 (s, n(CÄ
/
/
C38H42Cl2FeN4 (681.54) found (required): C, 67.12
(66.97); H, 6.34 (6.21); N, 8.18% (8.22). meff (Evans’
BArF resonances) 165.8 (imineÃ
Hz, BArF ipso-C), 150.6, 146.8, 146.0, 139.8 (pyzÃ
ArÃC), 135.3 (BArF ortho-C) 129.3 (q, JCF
BArF meta-C), 125.0 (q, JCF 272.4 Hz, BArF CF3),
66.3 (O(CH2CH3)2), 28.2 (CHMe2), 24.8, 22.6
(CH(CH3)2), 15.2 (O(CH2CH3)2), 2.4 (NiÃC).
/
C), 162.2 (q, JCB
ꢀ
/
37.5
NMR method)ꢀ5.1 BM.
/
/
and
/
ꢀ31.3 Hz,
/
ꢀ
/
6. Supplementary material
/
Crystallographic data for the structural analysis have
been deposited with the Cambridge Crystallographic
Data Centre, CCDC Nos. 195300ꢁ195302 for com-
/
5.3.9. 6-[1-(2,6-Diisopropylphenylimino)ethyl]-2,2?-
bipyridine iron dichloride [(5)FeCl2]
To a solution of FeCl2 (0.12 g, 0.98 mmol) in n-BuOH
at 80 8C was added 1 equiv. of the imine ligand (0.35 g,
0.98 mmol). The mixture was stirred at 80 8C for 15 min
and allowed to cool to r.t. The solvent was removed in
pounds (1a)NiBr2, (3)NiBr2 and (6)FeCl2. Copies of this
information may be obtained free of charge from The
Director, CCDC, 12 Union Road, Cambridge, CB2
1EZ, UK (fax: ꢃ44-1223-336-033; e-mail: deposit@
/
ccdc.cam.ac.uk or www: http://www.ccdc.cam.ac.uk).
vacuo and the residue washed with Et2O (2ꢄ
provide a blue solid in 76% yield (0.36 g).
MS (FAB, m/z): 484 [Mꢃ, 15], 348 [Mꢃꢂ
/
15 ml) to
/
Cl, 100].
Acknowledgements
Elemental analysis for C24H27Cl2FeN3 (484.25) found
(required): C, 59.40 (59.53); H, 5.52 (5.62); N, 8.49%
Dr. Jane Boyle is thanked for NMR polymer analysis
and Dr. Gary Audley is thanked for GPC analysis. We
(8.68). meff (Evans’ balance)ꢀ5.5 BM.
/

G.J.P. Britovsek et al. / Inorganica Chimica Acta 345 (2003) 279ꢁ
/291
291
[27] G.J.P. Britovsek, V.C. Gibson, S.K. Spitzmesser, K.P. Tellmann,
A.J.P. White, D.J. Williams, J. Chem. Soc., Dalton Trans. (2002)
1159.
are grateful for financial support from EPSRC and BP
Chemicals Ltd.
[28] P.-L. Bres, V.C. Gibson, C.D.F. Mabille, W. Reed, D.F. Wass,
R.H. Weatherhead, BP Amoco, WO 98/49208, 1998.
[29] R.E. Rulke, J.G.P. Delis, A.M. Groot, C.J. Elsevier, P.W.N.M.
References
¨
van Leeuwen, K. Vrieze, K. Goubitz, H. Schenk, J. Organomet.
Chem. 508 (1996) 109.
[1] G.J.P. Britovsek, V.C. Gibson, D.F. Wass, Angew. Chem., Int.
Ed. Engl. 38 (1999) 428.
[30] H. Rutner, P.E. Spoerri, J. Org. Chem. 28 (1963) 1898.
[31] F. Seyhan, J. Org. Chem. 22 (1957) 217.
[32] E.C. Constable, F. Heirtzler, M. Neuburger, M. Zehnder, J. Am.
Chem. Soc. 119 (1997) 5606.
[2] S.D. Ittel, L.K. Johnson, M. Brookhart, Chem. Rev. 100 (2000)
1169.
[3] S. Mecking, Angew. Chem., Int. Ed. Engl. 40 (2001) 534.
[4] L.K. Johnson, C.M. Killian, M. Brookhart, J. Am. Chem. Soc.
117 (1995) 6414.
[33] S.P.D. Baugh, V.C. Gibson, BP Chemicals Ltd., WO 00/58320,
2000.
[5] D.P. Gates, S.A. Svejda, E. Onate, C.M. Killian, L.K. Johnson,
˜
[34] G.R. Newkome, H.-W. Lee, J. Am. Chem. Soc. 105 (1983) 5956.
[35] C.J. Chandler, L.W. Deady, J.A. Reiss, J. Heterocycl. Chem. 18
(1981) 599.
P.S. White, M. Brookhart, Macromolecules 33 (2000) 2320.
[6] L.K. Johnson, C.M. Killian, S.D. Arthur, J. Feldman, E.F.
McCord, S.J. McLain, K.A. Kreutzer, M.A. Bennett, E.B.
Coughlin, S.D. Ittel, A. Parthasarathy, D.J. Tempel, M.S.
Brookhart, DuPont, WO 96/23010, 1996.
[36] T. Laine, M. Leskela, Borealis, WO 00/18776, 2000.
¨
[37] T.V. Laine, U. Piironen, K. Lappalainen, M. Klinga, E. Aitola,
M. Leskela, J. Organomet. Chem. 606 (2000) 112.
¨
[38] S. Plentz Meneghetti, P.J. Lutz, J. Kress, Organometallics 18
(1999) 2734.
[7] B.L. Small, M. Brookhart, A.M.A. Bennett, J. Am. Chem. Soc.
120 (1998) 4049.
[8] B.L. Small, M. Brookhart, Polym. Prep. (Am. Chem. Soc., Div.
Polym. Chem.) 39 (1998) 213.
[39] D. Onggo, D.C. Craig, D. Rae, H.A. Goodwin, Aust. J. Chem. 44
(1991) 331.
[9] A.M.A. Bennett, DuPont, WO 98/27124, 1998.
[10] G.J.P. Britovsek, V.C. Gibson, B.S. Kimberley, P.J. Maddox, S.J.
McTavish, G.A. Solan, A.J.P. White, D.J. Williams, Chem.
Commun. (1998) 849.
[40] H.A. Goodwin, D.W. Mather, Aust. J. Chem. 27 (1974) 2121.
[41] A. Angeloff, J.-C. Daran, J. Bernadou, B. Meunier, Eur. J. Inorg.
Chem. (2000) 1985.
[42] S. Ameerunisha, J. Schneider, T. Meyer, P.S. Zacharias, E. Bill,
G. Henkel, Chem. Commun. (2000) 2155.
[11] G.J.P. Britovsek, M. Bruce, V.C. Gibson, B.S. Kimberley, P.J.
Maddox, S. Mastroianni, S.J. McTavish, C. Redshaw, G.A.
Solan, S. Stromberg, A.J.P. White, D.J. Williams, J. Am. Chem.
¨
[43] R. Ziessel, A. Harriman, J. Suffert, M.-T. Youinou, A. De Cain,
J. Fischer, Angew. Chem., Int. Ed. Engl. 36 (1997) 2509.
[44] K.R. Kumar, S. Sivaram, Macromol. Chem. Phys. 201 (2000)
1513.
Soc. 121 (1999) 8728.
[12] G.J.P. Britovsek, B.A. Dorer, V.C. Gibson, B.S. Kimberley, G.A.
Solan, BP Chemicals Ltd., WO 99/12981, 1999.
[13] C.M. Killian, L.K. Johnson, M. Brookhart, Organometallics 16
(1997) 2005.
[45] A. Koppl, H.G. Alt, J. Mol. Catal. 154 (2000) 45.
¨
[46] T.-J.J. Kinnunen, M. Haukka, T.T. Pakkanen, T.A. Pakkanen, J.
Organomet. Chem. 613 (2000) 257.
[14] S.A. Svejda, M. Brookhart, Organometallics 18 (1999) 65.
[15] S. Stromberg, M. Oksman, L. Zhang, K. Zetterberg, Acta Chem.
¨
[47] J. Feldman, S.J. McLain, A. Parthasarathy, W.J. Marshall, J.C.
Calabrese, S.D. Arthur, Organometallics 16 (1997) 1514.
[48] L.K. Johnson, M.A. Bennett, S.D. Ittel, L. Wang, A. Parthasar-
athy, E. Hauptman, R.D. Simpson, J. Feldman, E.B. Coughlin,
DuPont, WO 98/30609, 1998.
Scand. 49 (1995) 689.
[16] B.L. Small, M. Brookhart, J. Am. Chem. Soc. 120 (1998) 7143.
[17] M.S. Brookhart, B.L. Small, DuPont/UNC, WO 99/02472, 1999.
[18] G.J.P. Britovsek, S. Mastroianni, G.A. Solan, S.P.D. Baugh, C.
Redshaw, V.C. Gibson, A.J.P. White, D.J. Williams, M.R.J.
Elsegood, Chem. Eur. J. 6 (2000) 2221.
[49] M. Sawamoto, Prog. Polym. Sci. 16 (1991) 111.
[50] J.E. Puskas, G. Kaszas, Rubber Chem. Technol. 69 (1996) 462.
[51] S.Y. Desjardins, K.J. Cavell, J.L. Hoare, B.W. Skelton, A.N.
Sobolev, A.H. White, W. Keim, J. Organomet. Chem. 544 (1997)
163.
[19] C.-Y. Wu, H.E. Swift, J. Catal. 24 (1972) 510.
[20] W. Keim, R. Appel, A. Storeck, C. Kruger, R. Goddard, Angew.
¨
Chem., Int. Ed. Engl. 20 (1981) 116.
[21] W. Keim, R. Appel, S. Gruppe, F. Knoch, Angew. Chem., Int.
Ed. Engl. 26 (1987) 1012.
[52] L. Deng, T.K. Woo, L. Cavallo, P.M. Margl, T. Ziegler, J. Am.
Chem. Soc. 119 (1997) 6177.
[53] G.M. Lenges, DuPont, WO 00/68280, 2000.
[54] L. Deng, P.M. Margl, T. Ziegler, J. Am. Chem. Soc. 119 (1997)
1094.
[22] C. Wang, S. Friedrich, T.R. Younkin, R.T. Li, R.H. Grubbs,
D.A. Bansleben, M.W. Day, Organometallics 17 (1998) 3149.
[23] V.C. Gibson, A. Tomov, A.J.P. White, D.J. Williams, Chem.
Commun. (2001) 719.
[55] T.V. Laine, M. Klinga, M. Leskela, Eur. J. Inorg. Chem. (1999)
¨
959.
[56] A. Koppl, H.G. Alt, Phillips Petroleum Company, WO 99/49970,
¨
[24] N.A. Cooley, S.M. Green, D.F. Wass, K. Heslop, A.G. Orpen,
P.G. Pringle, Organometallics 20 (2001) 4769.
[25] G.J.P. Britovsek, V.C. Gibson, S. Mastroianni, D.C.H. Oakes, C.
Redshaw, G.A. Solan, A.J.P. White, D.J. Williams, Eur. J. Inorg.
Chem. (2001) 431.
1999.
[57] M. Brookhart, B. Grant, A.F. Volpe, Organometallics 11 (1992)
3920.
[26] G.J.P. Britovsek, V.C. Gibson, B.S. Kimberley, S. Mastroianni,
C. Redshaw, G.A. Solan, A.J.P. White, D.J. Williams, J. Chem.
Soc., Dalton Trans. (2001) 1639.
[58] A.R. Katritzky, A.F. Pozharskii (Eds.), Handbook of Hetero-
cyclic Chemistry, 2nd ed., Elsevier, Amsterdam, 2000, p. 177.