Metallocenes

Vol. 7

METALLOCENES

Scope

Although the traditional usage of the term metallocene has encompassed sand-
wich structures containing two rings of ﬁve carbons each bound through all atoms
to a central metal atom, common parlance in the ﬁeld of oleﬁn polymerization
has expanded this deﬁnition to include structures having only one C

ring or

none. The term is now applied to all single-site catalysts (qv) (ie, all catalysts
having a single, well-deﬁned active-center structure), in some cases restricted to
complexes of the early- and middle-transition metals and lanthanides, to distin-
guish the catalysts from the recent families discovered by Brookhart and others
based on late-transition metals. For the purposes of this publication, we equate
the word metallocene with the family of compounds containing at least one C

ring (or analogous heterocycle) bound to an early-transition-metal or f-block atom
that, in the presence of a cocatalyst if required, effect the polymerization of
ethylene and other

α-oleﬁns. We will concentrate on those complexes, predomi-

nantly bis(cyclopentadienyl)zirconium(IV) compounds, that polymerize ethylene
and propylene with high efﬁciency.

Historical Background

Although the discovery of (C

)

Fe (ferrocene), ﬁrst reported by Kealy and

Paulson in 1951 (1) but recognized as a sandwich compound by Wilkinson and
co-workers (2), brought into being a new family of complexes that contain the
Cp

M (Cp

= C

) fragment, it was not until the discoveries of Sinn and Kaminsky

in the late 1970s that metallocenes, and speciﬁcally zirconocenes, were recognized

METALLOCENES

Vol. 7

as powerful oleﬁn polymerization catalyst precursors. (For a good discussion of
the succession of discoveries leading to the Kaminsky system, see Refs. 3) This
work brought together two strands of research: (1) the investigation of the alkyla-
tion and polymerization chemistry of Cp

= Ti, Zr) (4,5), which had begun

soon after the ﬁrst reports of titanocene dichloride; and (2) the observation that
small amounts of water added to the medium of certain Ziegler–Natta ethylene
polymerizations caused marked increases in the rate of enchainment. It was the
ﬁnding (by many accounts serendipitous) that large amounts of moisture, when
added to the conventional Ziegler–Natta cocatalyst trimethylaluminum, dramat-
ically improved the activity of mixtures of the aluminum alkyl with metallocenes,
that set into motion the wave of industrial and academic research in this area.

Although the activities of the new metallocene catalysts were quite remark-

able (

>1 × 10

g(PE)/mmol(Zr)), they would have probably remained a labora-

tory curiosity had they not exhibited a crucial characteristic: they were homoge-
neous. (Here we understand homogeneous to mean soluble and also having a single
active-site structure.) Single-site catalysts were known, of course, for oleﬁn poly-
merization. Catalysts based on vanadium had produced polymers with molecular
weight distributions (M

) of about 2, the value expected for ideal single-site

behavior in polymerization and chain transfer. However, the vanadium family of
single-site catalysts had aspects of its chemistry that made it less attractive to
academic work and also to commercialization. First, these catalysts are generally
mixtures of vanadium complexes of hard inorganic ligands such as halides, oxides,
or acetylacetonate groups, most if not all of which would be expected to be lost
upon reaction with a large excess of trialkylaluminum or chlorinated aluminum
alkyl compound. Because of this, the precursor to the active catalyst generally
lacks modiﬁable substituents that persist in the active complex, making group
substitution and structure–property trend analysis futile. The paramagnetism of
vanadium catalysts also precludes NMR spectroscopy, a key tool for characteriza-
tion. Also, because of the generally low activity and temperature sensitivity of the
vanadium-based single-site catalysts, use of these catalysts in modern large-scale
industrial settings has been difﬁcult, because such important processes as Union
Carbide’s UNIPOL or the slurry process of Phillips Petroleum have eliminated
ash removal, and also require fairly high reaction temperatures (40–110

◦

C) for a

commercially acceptable operation, as dictated by the economics of heat transfer
from the reactor. The early non–aluminoxane-based titanium catalysts (such as
Cp

TiCl

/(C

)

AlCl (4)) were homogeneous, but not extremely active.

In one stroke, the discovery of methylaluminoxane (the product of the par-

tial hydrolysis of trimethylaluminum) gave to researchers a system of interest to
both the synthetic chemist and the reaction engineer. Thus we have seen a ﬂower-
ing of chemical discovery and innovation in open and patent literatures alike, as
investigators have taken advantage of the stability of the Cp M linkage during
polymerization and have introduced substituents that change the active-site be-
havior in myriad ways. Driven as much by the desire for a degree of novelty that
would confer patent protection as by the thirst for knowledge, researchers have
extended the range of single-site catalysts far beyond those having Cp groups. The
rise of metallocenes has in large part legitimized the study of oleﬁn polymerization
among “serious” organometallic chemists, and in turn, has forced captains of in-
dustry to become more ﬂuent with the fundamentals of synthetic and mechanistic

Vol. 7

METALLOCENES

chemistry than seemed possible only decades ago. One suspects that the ramiﬁ-
cations of both changes will be felt for years to come, regardless of the ultimate
commercial acceptance of metallocene polymerization technology.

Properties of Metallocenes

Structure and Bonding.

Both bis- and mono-Cp complexes of early tran-

sition metals are marked by a pseudo-octahedral environment, which is evident
from the bond angle between the halide atoms of zirconocene dichloride, for ex-
ample, of 90

◦

. The angle formed from the metal atom and the two centroids of

the Cp rings in bis-Cp complexes is generally about 120

◦

, although this can be

raised or lowered, with the latter effect being more common, through the addi-
tion of bridging groups between the rings. Metallocenes useful for polymerization
generally have d

electronic conﬁgurations, and are thus diamagnetic. Complexes

containing Cp rings alone are usually weakly colored, often yellow–gold, but the
fusion of an aromatic ring to the Cp ring results in a much stronger coloration,
presumably because of L–M (

π–d

∗

) transitions.

The molecular orbitals involved in bonds between the Cp

M fragment and

the remaining, nonring ligands are shown in Figure 1, adapted from Lauher and
Hoffmann (6). In d

systems such as Cp

ZrCl

, the electrons ﬁlling these orbitals

are derived solely from the non-Cp ligands. Alteration of the metallocene geome-
try, or in the electronic properties of any of the ligands will obviously change the
energies of the molecular orbitals, with consequences for metal–alkyl reactivity.
Because of the lack of metal-based electron density, however, Group IVB metal-
locenes are not strongly inﬂuenced by acceptor orbitals on non-Cp ligands such as
the

∗

orbitals of ethylene or CO.

Chemical Reactivity.

Group IVB metallocenes are moderately air-stable

compounds unless (1) there is a highly basic carbanion attached, either in the form
of a highly alkylated Cp ring or a polycyclic C

-containing ligand, or (2) there is

an alkyl group attached directly to the metal; in the latter case the compound
is extremely water- and oxygen-sensitive, and often decomposes upon heating or
exposure to strong light. The hydrolysis products of Group IVB metallocene dichlo-
rides are generally HCl and zirconia, as well as a mixture of organic compounds.

Preparation.

Group IVB metallocenes and Lanthanocenes are convention-

ally prepared by ligand exchange between lithium or Grignard reagents and metal
chlorides (eqs. 1a and 1b). Often the metal chlorides are present as etherate com-
plexes; such complexes need to be prepared carefully, as the reaction is unusually
exothermic.

m Cp

′Li +

′

−m)

+ m LiX

(1a)

m Cp

′MgX +

′

−m)

′

+ m MgXX′

[X,X

′ = halide; M = transition metal; Cp′ = (substituted) cyclopentadienyl]

(1b)

Milder preparatory schemes include the reaction of silyl or stannyl cyclopen-

tadienides with metal halides, as shown in equation 2. (For the application of this

METALLOCENES

Vol. 7

(b)

(a)

Fig. 1.

(a) Orbitals responsible for metal–hydride bonding in Cp

TiH

. Left-hand orbitals

derive from Cp

, while the in-phase and out-of-phase combinations of the two H

−

1s orbitals are on the right. Coordinates deﬁned in (b). Reprinted with permission from
Ref. 6. Copyright (1976) American Chemical Society.

reaction to bridging ligands, see Ref. 7.) This method is especially well suited to
the preparation of titanocenes, as the use of a strongly reducing metal alkyl such
as LiCp is avoided. Another route to metallocenes involves the reaction of the
neutral mono- or bis-Cp ligand with a metal tetraamide (eq. 3). This pathway has
been demonstrated by Jordan (8) to allow the preparation of racemic bis(indenyl)
zirconium compounds in high selectivity. Aluminum complexes of bis(indene) lig-
ands will also undergo transmetallation reactions with zirconium tetraamides in
high yield (9).

+ MX

+ n XEMe

EMe

−m)

(2)

Vol. 7

METALLOCENES

+ M(NR′

)

+ n HNR′

−m)

= halide; M = transition metal; E = Si, Ge, Sn]

(3)

Toxicity.

There is little known of the effects on living organisms of metal-

locene precursors beyond those of the parent compounds Cp

MCl

= Ti, Zr).

Titanocene dichloride has been reported (10) to be a potentially useful anticancer
agent, while the zirconium analogue has been identiﬁed as a mutagen in at least
one toxicological report (11). Prudence dictates that extreme care should be exer-
cised with metallocenes, particularly those with polycyclic aromatic groups.

Cocatalysts for Polymerization

Aluminoxanes.

Health and Safety Considerations.

Solutions of MAO are generally non-

pyrophoric because of the low volatility of toluene. However, if toluene is removed,
a ﬁnely divided white powder remains that is extremely air- and water-reactive,
and will often spontaneously burst into ﬂame. Contact of MAO or its solutions with
water usually results in an immediate explosion, but caution must be maintained
even after the initial reaction, because, similar to other aluminum alkyl reagents,
MAO often forms a crust of alumina when hydrolyzed quickly. This crust tends to
protect the remaining alkyl, and disturbing it without care may cause a second
violent reaction.

Similar safety considerations apply to modiﬁed MAO, although when solvent

has been removed these materials are often glasses or viscous oils. Modiﬁed MAO
solutions in light alkanes such as isopentane may form dangerous mixtures with
air because of the volatility of the solvent and the presence of an intrinsic source
of ignition.

Methylaluminoxane.

Methylaluminoxane (MAO) is prepared from the con-

trolled hydrolysis (eq. 4) of trimethylaluminum (TMA), usually in toluene.

2n H

+ 2n CH

(4)

This is an extremely exothermic reaction and poses great hazards, as large

amounts of heat and gas must be safely removed. Early workers employed hy-
drated inorganic salts as a controllable method for addition of water, but it appears
that this procedure is not practiced commercially. The patent literature suggests
that the main suppliers of MAO (Albemarle, Akzo Nobel, and Witco) add water
directly, either by use of moist nitrogen or a marginally wet aromatic solvent. The
precautions necessary for the synthesis of MAO add to the cost of this material,
which suffers as well from the expense of TMA, which cannot be made by the
normal routes available for the higher trialkylaluminum compounds, such as the

METALLOCENES

Vol. 7

addition of oleﬁn to AlH

. MAO solutions normally have methyl/Al ratios greater

than 1, reﬂecting the fact that some TMA (about 30 mol% on an Al basis) must
be present for the solution to be stable. NMR analysis suggests that much of the
excess methyl content, at least, is present as TMA [as opposed to ((CH

)

Al)

O),

although it may be complexed to the MAO.

The function of the excess TMA is the focus of considerable controversy. To

it have been ascribed the roles of alkylating agent, ionizing agent, chain-transfer
agent, and deactivator. TMA can be removed, at least partially, by vacuum distil-
lation, leaving a toluene-soluble material, which remains a competent activator
for metallocene polymerization.

As noted above, the initial high activity metallocene catalysts involved zir-

conocenes and titanocenes mixed with methylaluminoxane. MAO is a poorly un-
derstood mixture of oligomers (generally with degree of polymerization of about
20), despite years of study. It can accurately be characterized by its proton NMR
spectra, its methyl content, and its average molecular weight, but proposed struc-
tures for speciﬁc MAO oligomers or subunits remain highly speculative. As will be
discussed in greater detail, the two actions of MAO are alkylation and activation
by Lewis acid methyl anion abstraction. The key to the activity of MAO-activated
metallocenes is the degree to which the anion formed from this process is relatively
inert and well separated from the metallocene cation, which has only 14 valence
electrons and tends to form bridging complexes such as shown in equation 5 and
1a with ordinary aluminum alkyls.

RCH CH

AlR

′

−AlR′

′

(5)

While the origin of the ability of MAO to effectively abstract methide and

remain loosely associated with the metallocene is not well understood, the most
convincing proposal comes from the work of Barron, in which structural studies
of analogous oligomeric t-butylaluminoxanes (12) and their exchange reactions
with Cp

Zr(CH

)

have led to the concept of “latent Lewis acidity,” which can

be paraphrased as a potential for the drum-like subunit of the structure of the
aluminoxane to reversibly open up a highly electrophilic tricoordinate aluminum
atom for methide abstraction, resulting in an anion of high stability and bulk,
factors that militate for a weak ion–pair interaction (eq. 6).

t-Bu

ZrMe

t-Bu

(6)

Vol. 7

METALLOCENES

C(9)

C(7)

C(8)

C(6)

C(5)

Al(2)

Al(1)

Fig. 2.

Structure of [Me

)Al(OAlMe

)]

dianion. Reproduced from Ref. 13 by permission

of The Royal Society of Chemistry.

The two crystal structures that come closest to shedding light on the actual

form of MAO are shown in Figures 2 and 3, structures for a dianion of a MAO
oligomer (13) and of a t-butylaluminoxane hexamer (14).

In practice, it has been found that MAO must be used at fairly high atomic

ratio of Al to transition metal, conventionally between 300 and 10,000, although
this is not necessarily the case when immobilized systems are used. For example,

C(22)

C(32)

C(3)

C(31)

C(33)

C(21)

C(2)

C(23)

O(2)

C(11)

C(1)

C(13)

C(12)

Al(3)

Al(1)

Al(2)

O(3)

O(1)

Fig. 3.

Crystal structure of (Al(t-Bu)O)

. Reprinted with permission from Ref. 14. Copy-

right (1993) American Chemical Society.

METALLOCENES

Vol. 7

[Zr]

Me-MAO

[Zr]

MAO

+ AlMe

Scheme 1.

most of the silica-supported metallocene/MAO catalysts reported in the patent
literature employ Al/Zr ratios of between 100 and 300. The high ratios required
have led to speculation that there exists a particular oligomer or family of
oligomers that perform the bulk of the activation. Experiments done by Sishta
and co-workers have shown (15) that a very low Al/Zr ratio—perhaps represent-
ing the reaction of one molecule of MAO per metallocene—is sufﬁcient to produce
NMR spectra similar to those for much higher ratios. This result hints that the
additional MAO is required either for scavenging of poisons, reactivation of met-
allocene in a dormant state, or agglomeration of MAO oligomers following activa-
tion of the metallocene, leading to larger anions and hence looser ion pairs. Recent
NMR experiments led by Brintzinger have suggested that the large MAO coun-
terion contains 150–200 aluminum atoms, and have given weight to the theory
that the main effect of increasing Al/Zr ratio is to shift the equilibrium between
coordinatively saturated zirconocene–trimethylaluminum dimers and ion pairs
with more accessible open sites (Scheme 1) in favor of the latter (16).

Modiﬁed Methylaluminoxane.

The fact that MAO is only soluble in aro-

matic solvents can present problems in handling; often a highly pyrophoric solid
precipitates when these solutions are mixed with other reagents. For this reason,
as well as to limit the consumption of expensive TMA, much effort was expended
in making MAO soluble in aliphatic solvents, principally by replacing some of the
methyl groups with higher alkyl substituents. The most commercially success-
ful reagents, known as modiﬁed methylaluminoxanes, contain up to 30 mol% of
isobutyl substitution. These complexes are made either by direct hydrolysis of a
mixture of TMA and higher aluminum alkyl compound, or by a number of more
exotic processes, and are extremely soluble in hexane or isopentane. Catalysts pre-
pared from modiﬁed MAO are of comparable activity to those activated by MAO.

A nonhydrolytic method of making MAO has been described (17); it appar-

ently results in an aluminoxane much reduced in residual TMA, which harms cer-
tain catalysts, yet of similar shelf life to conventional MAO. The procedure uses a
carbonyl compound as the initial oxygen source (eq. 7a, 7b 7c), then calls for alkox-
ide ligand fragmentation and secondary hydrolysis to make the ﬁnal product.

Al O

(7a)

(7b)

Vol. 7

METALLOCENES

−CH

O Al

(7c)

Discrete Activators.

MAO is a troublesome activator for several reasons.

More expensive on a molar basis than cocatalysts such as Al(C

)

used by

Ziegler–Natta catalysts, MAO seemed until recently to require a very large molar
ratio to transition metal, preventing metallocenes from competing with conven-
tional catalysts on a cost basis. From an academic perspective, MAO, an ill-deﬁned
mixture, hinders the detailed analysis of active-site structure. Not surprisingly,
many attempts have been made over the years to replace MAO, the most suc-
cessful of which have given us discrete activators based on perﬂuorinated aryl
groups.

There are three popular modes by which these cocatalysts ionize the metal-

locene (which must be alkylated by a different agent such as triisobutylaluminum):
Lewis acid extraction of the alkylide group by a neutral Lewis acid (eq. 8) or by
a carbocation (eq. 9) and protonolysis of the alkylide group (eq. 10). (Ionization
may also occur through direct reaction of the alkylated metallocene by a strong
oxidant such as [Cp

][B(C

)

−

] (18).)

(8)

B(C

)

−MeCPh

(9)

METALLOCENES

Vol. 7

B(C

)

−CH

−PhNMe

(10)

The ﬁrst pathway, extensively studied by Marks and co-workers, is a rapidly

reversible ionization (19). The second two routes, actually discovered prior to re-
action 8 by scientists at Exxon and Fina, are technically irreversible, as one of the
by-products is a kinetically unreactive alkane. (For protonation, see Ref. 20 and
for methyl transfer, see Ref. 21.) In practice, more than one equivalent of discrete
activator is usually employed. While the source of this requirement is still un-
known, candidates include reactions with the trialkylaluminum species (present
in excess as alkylating agents) that destroy the special activating power of the
borane in equation 8 or the inertness of the anions in equations 9 and 10. Another
possible culprit is the formation of dormant alkyl-bridged metallocene dimers
(22). In equation 10, one of the by-products is a trisubstituted amine that might
be expected to trap the coordinatively unsaturated cation formed in this reaction.
(N,N-Dimethylaniline is shown, but is not exclusively used; tri-n-butylamine is
another common component of the activator.) However, it does not appear to di-
minish catalytic activity. Variations on the structure of B(C

)

−

have been ex-

plored, including aluminum substitution for boron (23), silylation of the arene
(24), and the use of larger perﬂuoroaryl substitutuents (24,25). Not surpris-
ingly, activity rises with the size of the counterion. Use of the nonperﬂuorinated
aryl groups phenyl (26), p-ﬂuorophenyl (27), and 3,5-bis(triﬂuoromethyl)phenyl
gives a zwitterionic catalyst (eq. 11), a ﬂuorine-bridged ion pair (eq. 12), and
a weakly active titanocene catalyst at low temperature (28), respectively. (This
counterion has been used extensively by Brookhart and co-workers, especially
for the activation of single-site cyclopentadienylcobalt polymerization catalysts;
see Ref. 29.) In equation 12, activity appears to be high only for the metallocene
(C

(CH

)

Zr(CH

)

, whose steric congestion is probably responsible for prevent-

ing decomposition.

)

ZrMe

[Bu

NH][BPh

]

−2 CH

−NBu

)

B(C

)

(11)

B(p-C

[PhMe

NH][B(p-C

]

−CH

−PhNMe

)

ZrMe

(12)

Vol. 7

METALLOCENES

Other noncoordinating anions have been discovered that enable high activity

polymerization. Among these are the metallacarboranes [(C

)

−

= Fe,

Co, Ni), where catalyst ionization is by protic abstraction by an ammonium ion
(30).

Commercially, Dow appears to have implemented activation by well-deﬁned

cocatalyst for its Insite solution-phase metallocene process. Aside from historical
priority, one reason for the persistent use of MAO is that the perﬂuorinated boron
compounds are difﬁcult to prepare, thus raising their expense. (One of the inter-
mediates, (C

)Li, must be handled at very low temperature to prevent violent

decomposition.) Another is that, unlike the MAO systems, catalysts activated by
discrete activators have no large excess of alkylaluminum to scavenge poisons
from the reaction medium, while common scavengers may hinder activity by con-
suming the expensive activator.

Other Methods of Activation.

Despite some early claims (31), the evidence

for metallocene activation by simple trialkylaluminum compounds has not ac-
cumulated. Activation by alternate aluminoxanes such as isobutylaluminoxane
(IBAO) has also been attested (32). The fact that these much cheaper methods
have not consigned MAO to the ash-heap of history speaks, perhaps, of some lim-
itations in these alternatives.

Some support materials can be rendered Lewis acidic enough to ionize dialkyl

metallocenes. Marks and co-workers have reported (33) that alumina dried at very
high temperatures can react at least to some small degree with both thorium-
and zirconium-based metallocene dimethyl species to yield active catalysts for
polyethylene. The resulting cationic metal center is believed to remain coordi-
nated to the surface through an Al–O–M Lewis acid/base linkage, at least prior
to exposure to ethylene. Hybrid surface/cocatalyst systems based on aluminum
alkyl-treated clays have been developed (34) in which the solid substrate appears
to play some role in promoting polymerization activity far beyond that expected
for non–methyl aluminoxane- or trialkylaluminum-activated catalysts.

Forms of Catalyst Delivery

Porous Silica Supports.

While single-site catalysts continue to be stud-

ied in solution polymerizations, especially in academic laboratories, the adap-
tation of metallocene technology to existing industrial processes has generally
entailed the immobilization of the homogeneous catalyst on a solid surface of
well-controlled morphology and porosity. Generally speaking, the support favored
is amorphous silica containing fairly high pore volume (

>1 cm

/g) and average par-

ticle diameter of about 50

µm. Early on it was discovered that the extremely high

atomic ratios of MAO to zirconium that had been needed to obtain high activity
in solution polymerization were unnecessary when the catalyst was supported.
(Excellent activities based on supported metallocene/MAO catalysts with Al/Zr
<100 are reported in Ref. 35.) This has greatly facilitated the commercialization
of metallocenes. Supporting the catalyst can have consequences for the reactivity
pattern of the metallocene (36), indicating that the aluminoxane, when covalently
bound to the silica surface through one or more Si–O–Al groups (each formed from
the silanolysis of an Al Me bond), displays different ion–pairing dynamics with

METALLOCENES

Vol. 7

HCCHBr

2 [C

]Li

Ind

(1) 2 n-BuLi

(2) ZrCl

(THF)

Fig. 4.

Overview of bridge-anchored ansa-metallocene synthesis (Ref. 43).

the metallocene relative to the solution cocatalyst. Industrial research has yielded
numerous examples of polymerization supports beyond silica, including alumina,
zeolites, clays, and organic polymers. Metallocene catalysts have been covalently
bound to polymer either through prior attachment of part of the ligand to a poly-
meric substrate (37), or through the incorporation of a pendent unsaturated group
on the ligand itself into the growing polymer (38). In both cases, excellent particle
morphology was claimed.

Alternate Modes of Catalyst Physisorption.

Other forms of catalyst

delivery to the reactor have been implemented. Catalyst solution has been injected
directly into ﬂuidized-bed gas-phase reactors (39), evidently creating a mist of in
situ spray-dried catalyst. Spray-dried catalyst has been prepared ex situ as well, in
an attempt to better control the morphology of the particles (40). Solid catalytically
active materials have also resulted from precipitation of MAO by aliphatic solvent
(41) and through a low conversion polymerization (prepolymerization) prior to
introduction to the large-scale polymerization medium (42).

Supports as Part of Ligand Set.

Solid supports bear functional groups

that have been utilized as part of the ligand framework of metallocenes. Soga
and co-workers covalently bound bridged metallocenes to silica through reaction
of the surface silanols with reactive carbon halide or silicon halide bonds on the
ligand bridge. The overview of the synthesis for one catalyst precursor is shown
in Figure 4, with the indenyl rings in pseudo-racemic orientation because the
high isotacticity of the polypropylene produced suggests a structural analogy to
rac-Et(Ind)

ZrCl

. Activities were reported to be greatly superior to the homoge-

neous analogues (43).

Families of Metallocene Catalysts

In this section, we will brieﬂy examine the salient features of the basic classes of
metallocene catalysts. Polymerization activity comparisons and the mechanisms
that govern polymerization will be treated in subsequent sections.

Group IVB Metallocenes.

Bis(cyclopentadienyl) Complexes.

Representatives of the three most-

intensively studied subfamilies of bis(Cp) metallocenes are 1, 2, and 3. The parent

Vol. 7

METALLOCENES

zirconocene (1) was the ﬁrst to be exploited, and soon substitution of the Cp rings
with alkyl groups and replacement of the Cp rings with indenyl and ﬂuorenyl
ligands was undertaken. Unbridged metallocenes of this sort exhibit a suite of
characteristics that have caused their application in the commercial arena to be in-
tense, but limited in scope. High activities are common to this family, and because
of the relative ease of precatalyst synthesis, they are well suited to large-scale
production of commodity grades of polyethylene. Facile chain transfer in the pres-
ence of

α-oleﬁn, however, renders these metallocenes nearly useless for making

very low density copolymers of ethylene. With rare exception (covered in section
Mixed Tacticity Polyoleﬁns), they cannot polymerize

α-oleﬁns such as propylene

with any degree of stereospeciﬁcity.

Comparative studies of unbridged metallocene catalysts have indicated that,

in general, alkylation of the Cp ring(s) causes an increase in polymerization activ-
ity, although substitution of all ﬁve carbons can lead to very low activities, as steric
congestion prevents monomer approach. Bis(indenyl) and bis(ﬂuorenyl)zirconium
complexes have received far less attention, and in the latter case, this is almost
certainly due to the greater lability of the ﬂuorenyl–metal interaction.

Aside from the ubiquitous n-alkyl, halide, and alkoxide groups, metallocenes

also bear carboxylate, triﬂuoromethanesulfonate (44), thiolate (45), and other
more exotic moieties in addition to the Cp groups. It is commonly believed that
these ligands are completely exchanged for alkyl groups originating on the alumi-
noxane in the case of MAO activation; however, this supposition does not imply
that the groups have no effect on oleﬁn polymerization as part of the counterion
complex.

Monoanionic analogues of the cyclopentadienyl group have been explored.

The bis(boratabenzene) complexes of Zr(IV) as in 4 have been shown by Bazan
(46) to be catalysts for the oligomerization of ethylene to oleﬁns, with the groups
attached to boron inﬂuencing selectivity to

α-oleﬁn. Tethering the two C

B rings

(see structure 5) tends to give polymer (47).

OEt

N(i-Pr)

METALLOCENES

Vol. 7

C(37)

C(32)

C(31)

C(11)

C(15)

C(36)

C(35)

C(34)

C(38)

C(25)

C(24)

C(23)

C(22)

Cl(2)

Cl(1)

C(14)

C(12)

C(13)

C(21)

C(33)

Fig. 5.

Crystal structure of 1,3-bis(cyclopentadienylmethyl)benzenezirconium dichloride.

Reproduced from Ref. 50 by permission of The Royal Society of Chemistry.

Bridged Metallocenes.

The addition of a linkage between the two Cp

rings has enabled the precise manipulation of the steric environment encountered
by reactive molecules such as

α-oleﬁns, and the development of metallocenes for

stereoregular oleﬁn polymerization has proceeded from this advance (vide infra).
The ﬁrst and most widely employed bridges are

[CR

]

= 1, 2) and SiR

= methyl, phenyl), which are easily incorporated into ligands by reaction of two

equivalents of Cp

−

with

α,ω-dibromoalkanes and dichlorosilanes, respectively.

Increasing the bridge length generally increases the centroid–metal–centroid
(Cp M Cp) angle, which can change polymerization characteristics. For exam-
ple, the Cp M Cp angle rises in the series (CH

)

TiCl

: 121

◦

= 1); 128

◦

= 2); 133

◦

= 3) (48). The CH

and Si(CH

)

bridges, by far the most

popular bridging moieties, give rise to similar Cp M Cp angles, but the ethylene
bridge allows a wide range of relative motion between the two halves of the ligand
(49), which may lead to decreased molecular weight and stereocontrol in oleﬁn
polymerization relative to the more rigid silylene-bridged analogues. Increases in
bridge length beyond that of 1,3-propanediyl generally lead to severe distortion, in
which the bridge is rotated toward the wedge containing the nonchelating ligands,
as in Figure 5 (50).

Bridging linkers also include R

E (E

= Ge, Sn) (51), as well as RB (52).

The Buchwald and Bercaw groups have reported (53,54) complexes in which

two bridges tie the Cp rings together. The substituents on the Cp rings change
the selectivity of oleﬁn polymerization, as discussed in under Mixed Tacticity
Polyobeﬁns.

The bridging linkage need not be strictly covalent. In 6, a bridged

bis(cyclopentadienyl)zirconocene is depicted in which the two Cp’s are connected
via a Lewis acid–Lewis base interaction between P and B atoms attached to the

Vol. 7

METALLOCENES

rings. A racemoid bis(indenyl) version gives highly isotactic PP at low tempera-
tures, indicating the bridge is quite robust under polymerization conditions (55).

Bridges containing heteroatoms such as nitrogen that can bind directly to

the transition metal have also been explored (56).

Siloxide (57) and amide (58) groups have been appended to Cp rings. Lewis

acids such as Al atoms in MAO may reversibly coordinate to the heteroatom in
these systems, leading to two distinct types of catalyst site and a bimodal distri-
bution of polymer MW (59).

One of the perceived deﬁciencies of the early ethylene-bridged metallocenes

in propylene polymerization was the relative lack of rigidity of the bridge. Several
attempts have been made in recent years to make highly rigid bridged metal-
locenes, for example by use of cyclic groups in the bridge, as in 7 (60). Other
uses of bulky bridges are to provide chiral centers, as in 8 (61), or to prevent the
formation of mesoid isomers during synthesis due to bridge–ring repulsions, as
in 9 (62).

t-Bu

THF

Polynuclear metallocene complexes may be produced through linking groups

in which the alkyl group provides the bridge 10 (63). Other links may be through
the Cp ring substituents, either indirectly, with a ferrocenyl bridge (64), or more
directly through hydrocarbon-only linkages as in 11 (65).

METALLOCENES

Vol. 7

The aim of many of these investigations has been to facilitate cooperative

behavior among the two or more catalytically active centers by forcing them to
remain in proximity to each other. In some cases, positive enhancements in copoly-
merization have been observed (66).

Mono(cyclopentadienyl) Complexes.

Metallocenes bearing a single cy-

clopentadienyl ring are not as widely used in polymerization as their bis-Cp
cousins, but they display a distinct reactivity that makes them particularly suited
to certain polymerizations. As with the bis-Cp metallocenes, they appear as
both bridged and unbridged varieties, with structures 12 and 13 portraying the
most prominent members of these respective families. The most successful of the
mono-Cp catalyst precursors are titanocenes, which are distinguished from zir-
conocenes in having stable Ti(III) structures, such as 14.

Mono-Cp complexes are generally prepared by the same routes as are the

bis-Cp versions (see under preparation), but in addition they may be synthesized
via comproportionation (Cp

+ MX

= 2 Cp

; X

= halide) (67) or by

reaction of Cp salts with titanium compounds having only one substituent that is
readily exchanged (68), such as ClTi(OR)

Unbridged Complexes.

Unbridged mono-Cp titanium complexes have

been explored for polyethylene, but have found more consistent application in
the area of polystyrene. In the latter case, reduction to Ti(III) appears to be nec-
essary prior to entry into the polymerization cycle (69). Unbridged mono-Cp zir-
conocenes have been much scarcer in the polymerization literature. In many in-
stances, the remaining substituents on the zirconium atom are divided into those
ligands that are believed to exchange with aluminum alkyl groups and those as-
sumed not to do so. The sole Cp ligand is thereby combined with a Cp analogue in
a pseudo–bis-Cp metallocene–type active site arrangement, with catalytic perfor-
mance typically intermediate between bis-Cp and bis-pseudo-Cp analogues. Cp

Vol. 7

METALLOCENES

analogues include benzamidinate (70), tris(pyrazolyl)borate (71), and the chelat-
ing alkoxide (2-pyridyl)dimethylmethoxide (72). In the case of titanocenes, the
strategy of mixing one Cp with a pseudo–Cp analogue has often led to the discov-
ery of systems that far outperform typical bis- or mono-Cp catalysts, especially in
the polymerization of ethylene and other

α-oleﬁns. These complexes are generally

distinguished by signiﬁcant steric bulk and a heteroatom with the potential for
binding to titanium through multiple bonds. Two interesting examples from thisr
subfamily are the phosphinimide complex 15 reported by Stefan and co-workers
(73) and the phenoxide complex 16 described by researchers Sumitomo at (74). In
both cases, activities are fairly good for ethylene/

α-oleﬁn copolymerization, and

the levels of incorporated comonomer are high as well.

Recently, Baird and co-workers have reported (75) examples of polymeriza-

tions by a simple mono-Cp titanium complex, (C

(CH

)

)Ti(CH

)

activated with

a Lewis acid (B(C

)

) that not only copolymerizes ethylene and

α-oleﬁns but

also induces polymerization of monomers normally associated with cationic poly-
merization such as isobutylene and vinyl ethers. Shaffer and Ashbaugh found (76)
that for isobutylene and

α-methylstyrene, the metal complex is an initiator rather

than a catalyst (if it even participates at all), but that a transition from cationic
to coordination polymerization occurs in styrene polymerization as temperature
is raised. Even if it merely functions as an initiator, however, these investigations
have revealed new polymerization systems based on anions such as [RB(C

)

]

−

= alkyl, C

) that are less prone to side reactions tending to limit the MW and

degree of polymerization of monomers like isobutylene at moderate temperatures
(T

> −80

◦

C).

t-Bu

i-Pr

Bridged Complexes.

The best-known bridged mono-Cp metallocene is the

so-called constrained-geometry catalyst precursor 12, a titanocene with a fairly
bulky C

(CH

)

moiety on one side and a large amido group on the other. Despite

this bulk, the short bridging unit, which enforces a Cp-centroid Ti N bite angle
of 107.6

◦

(77) (hence the sobriquet “constrained geometry”), ensures easy access

of monomers to the active center. The ligand was ﬁrst disclosed by Bercaw and
co-workers in its complex with scandium (data published in—not ﬁrst discloure—
Ref. 78), a molecule with limited ethylene polymerization activity. Soon there-
after, patent applications from workers at Dow Chemical (79) and Exxon Chem-
ical (80) appeared, claiming high activities for oleﬁn polymerization using 12. A
heteroatomic analogue of 12 and 17 has also been found (81) to have good poly-
merization behavior.

METALLOCENES

Vol. 7

t-Bu

The bridged amido–Cp titanocenes such as 12 have risen to prominence

presumably for the following features: (1) polymerization kinetics well suited
to solution-phase reaction at high temperatures (T

> 120

◦

C), the environment

of Dow’s commercial reactor; (2) high comonomer-incorporation ability, which
makes production of amorphous polyethylenes practical; and (3) ability to gen-
erate long-chain branches that result in superior melt processability for LLDPE
(see under Long-Chain Branching). Analogous complexes feature substituted in-
denyl or ﬂuorenyl groups in place of the C

(CH

)

unit, as well as bulky alkyl or

cycloalkyl substituents besides t-butyl on the nitrogen atom; any advantages in ac-
tivity or performance appear to be offset by the complexity and expense of catalyst
synthesis. A marked exception to this general trend is the heteroatom-substituted
complex 18, which shows (82) activities for ethylene–1-octene copolymerization far
greater than those of complex 12. The zirconium analogue of 12 and its relatives
do not seem to share the high comonomer response of their titanium versions and
are consequently much less widely employed. Also, titanium- and zirconium-based
compounds with a single carbon atom bridging group show lower activity than does
12 (83).

Other bridged metallocenes bearing one Cp ring have been described re-

cently. One structure, 19, reported by Marks and co-workers (84), has a direct
Cp phenoxide linkage. A series of Ti(IV) complexes, of which 20 (85) is an exam-
ple, were developed by Chien and co-workers, bearing donor atoms that, based on
NMR spectroscopic and polymerization data, coordinate to titanium prior to and
upon activation by MAO. Compound 14 shown above also includes a Lewis basic

Vol. 7

METALLOCENES

nitrogen bound to titanium, but in this case the bond appears in the precursor and
the oxidation state of titanium is

+3. Precursor 14, developed by DSM under the

name Lovacat (86), also produces rubbery ethylene–propylene copolymers, indi-
cating that the reactivity of higher

α-oleﬁns in titanium-based copolymerizations

with ethylene is similar in catalysts derived from Ti(III) and Ti(IV) precursors.
Whether the oxidation state of complexes like 12 as well as 14 is also less than
+4 in the active state has not been resolved.

Complexes with Open Pentadienyl Ligands.

Open-chain analogues of

the Cp ligand have only received limited attention. Compounds 21 (87) and 22
(88) have been featured in patent examples as moderately active polyethylene
catalysts after activation by [Cp

Fe][B(C

)

] and MAO, respectively.

t-Bu

Complexes with Dianionic Ligands.

In the search for metallocenes that

do not require ionization, several workers have targeted the synthesis of Group IV
complexes in which a dianionic ligand replaces a monoanionic Cp

of bis-Cp

met-

allocenes. Bazan has found good activity for TMM 23 (trimethylenemethane dian-
ion) and TBM 24, (tribenzylidenemethane dianion) complexes of zirconium (89).

Li(TMEDA)

(The degree of association of Li with the Zr–Cl functionality appears to be

dictated by the size of the dianionic substituent.) Although MAO is required for
high productivity, similar activity from an unstable solution of the prealkylated
neutral complex (C

(CH

)

)(TBM)ZrCH

led the authors to propose that a similar

neutral species is indeed the propagating site.

Carboranes were some of the ﬁrst dianionic ligands to replace Cp in met-

allocenes. The unstable monomeric complexes 25a–c were examined by Jordan
and co-workers for ethylene activity. While 25b and 25c have fair activity
for ethylene polymerization (

∼7 g/mmol(Zr)/h/atm for 25b) (90), the titanium

METALLOCENES

Vol. 7

analogue only dimerizes ethylene to 1-butene (91). The dianionic carborane
(2,3-(Si(CH

)

-2,3-C

) has been used opposite Cp in complex 26 (92).

a: M

= Ti

b: M

= Zr

c: M

= Hf

OEt

i-Pr

SiMe

Pr-i

Quan and co-workers (93) have synthesized the aminoborollide complex 27,

bearing yet another dianionic Cp analogue. The cyclooctatetraene dianion complex
28 has shown fair polymerization activity with MAO (94).

Group IIIB and f-Block Metallocenes.

The need for expensive cocata-

lysts has always hampered the development of metallocene catalysts for oleﬁn
polymerization. It was recognized early on that substitution of the Group IVB
metal with a lanthanide or Group IIIB element in the

+3, d

state would rep-

resent a cocatalyst-free analogue of the Kaminsky system. Active catalysts are
indeed obtained from bis-Cp lanthanocenes and yttrium- and scandium-based
congeners. The lutetium dimer 29 has an activity of

>7 kg/mmol(Lu)/h/atm for

ethylene polymerization in cyclohexane (95). (This remarkable compound can also
break the C H bonds of alkanes.)

Vol. 7

METALLOCENES

Because of their larger ionic radii, lanthanide (hereafter taken to include

Y and Sc) atoms are usually bound to heavily substituted cyclopentadienyl
groups such as C

(CH

)

. Although ionization of the lanthanocene is unnecessary,

preparation of the appropriate metal hydride or metal alkyl that can initiate
polymerization is a matter of considerable art. Lanthanide alkyl complexes with
β-hydrogens tend to decompose, while hydride and chloride species readily dimer-
ize or form “-ate” complexes with lithium salts. In fact, not only the Cp-ligands
but also the metal-bound alkyl substituent are often required to be large (such
as

CH(Si(CH

)

) (96), or the complex is further coordinated by a Lewis base

molecule such as THF.

Lanthanocenes are even more reactive toward polar impurities than are

Group IVB metallocene catalysts, which makes commercial use an even more
daunting prospect. Another feature that limits the applicability of lanthanocenes
is that the catalysts generally show high selectivity for ethylene insertion over
larger oleﬁns. Polyethylene produced by these catalysts generally is of very high
density and MW. Larger oleﬁns can be homooligomerized, but at rather low rates.

The scandium complex 30 disclosed by Shapiro and co-workers (97), which

represents the ﬁrst application of the bridging amido–Cp ligand in metallocene
chemistry, is unusual in its moderate ability to polymerize larger oleﬁns such
as 1-pentene. A yttrium analogue initiates polymerization of t-butyl acrylate and
acrylonitrile at room temperature and below (98). In the bis-Cp systems, a bridging
group seems to improve reactivity toward

α-oleﬁns (99). The yttrocene dimer 31

produces highly isotactic polypropylene (P

mmmm

= 97%) (100).

t-Bu

SiMe

t-Bu

= PMe

t-Bu

Although actinide complexes usually disqualify themselves from industrial

utility by their radioactivity, the thorocene (C

(CH

)

Th(CH

)

displays high

activity for ethylene polymerization after activation (101). Like zirconocenes, this
family of metallocenes tends to remain on the

+4 manifold, and thus requires

ionization by an appropriate cocatalyst (in this case, a highly Lewis acidic surface
such as dehydroxylated alumina or MgCl

) before it is active.

By their highly Lewis acidic nature, lanthanocenes are particularly suited

to the polymerization of polar monomers such as acrylates and lactones (for a
review, see Ref. 102). The lack of signiﬁcant chain-transfer and -termination paths
allows the production of block copolymers, including the case in which acrylate
polymerization follows that of ethylene (103).

Group VB and VIB Metallocenes.

Cyclopentadienylniobium and

-tantalum compounds bearing amidinate ligands as in 32 (104) and 33 (105)
have been occasionally used as ethylene polymerization precursors, usually with

METALLOCENES

Vol. 7

MAO as activator. Generally, the turnover rates for these systems are only mod-
erate, with that for 33 being on the high side at 470 g(PE)/mmol(Ta)/h/atm.
Activities for 34 and the vanadium complex CpV(N-p-tolyl)Cl

were 0.2 and

8.3 g(PE)/mmol(M)/h/atm, respectively (106). Molybdenum- and tungsten-based
systems are similarly mediocre catalysts (106).

i-Pr

SiMe

i-Pr

The mono-Cp butadiene complexes 35 and 36 have been found to polymerize

ethylene at moderate rates as well (35.23 and 5.69 g/mmol(M)/h/atm, respectively,
at 20

◦

C, with 500 eq. MAO) (107). At low temperatures, the polymerizations ap-

pear to be living in nature (see under Living Polymerization).

Cyclopentadienylchromium Compounds.

Chromocene, a paramag-

netic and somewhat labile metallocene, forms a polyethylene catalyst after
chemisorption on silica (108). The reaction generates an equivalent of C

, and

the propagating site has been proposed (109) to be as shown in equation 13. The
characteristics of this ill-characterized catalyst are facile chain transfer to H

poor comonomer incorporation, and fairly broad molecular weight distribution
(MWD). This catalyst is still in commercial use.

−CpH

(13)

Theopold and co-workers have prepared unsupported mono-Cp ethylene

polymerization catalysts by use of very bulky cyclopentadienyl and alkyl ligands
on chromium; these catalysts do not require an activator. Precursors having both

Vol. 7

METALLOCENES

unbridged 37 (110) and bridged 38 (111) ligand sets yield polyethylene with broad
MWDs, indicating a considerable breadth in active-site structures.

SiMe

Poor propylene incorporation by 38 is reported, although dimerization to

2-methyl-1-pentene was observed. Complexes such as 39 bearing electron-pair
donor atoms, described by D¨ohring and co-workers (112), are active catalysts
for ethylene polymerization (activity for 39: 10.5 kg(PE)/mmol(Cr); conditions:
21

◦

C, 2 atm ethylene, 100 eq. MAO); polymerization of propylene and higher

oleﬁns is approximately two orders of magnitude slower. Activities of up to 15
kg(PE)/mmol(Cr)/h/100 psi(C

) have been recorded for analogous unbridged

mono-Cp chromium compounds such as (C

(CH

)

)Cr(CH

)

(pyridine) supported

on metal oxides and activated by isobutylaluminoxane (113). A noteworthy aspect
of this system is that the MWD of the polymer produced is relatively narrow (2)
< M

< 3).

Another interesting catalyst system is the boratabenzene complex 40 from

the Bazan group (114). This precursor, after activation by MAO, B(C

)

or [(C

)

C][B(C

)

], polymerizes ethylene with activities comparable to

ZrCl

B Me

General Trends in Metallocene Polymerization

Although activity comparisons are fraught with perils, as the conditions of poly-
merization, including catalyst concentration, purity of monomers, reaction time,
etc., greatly inﬂuence the lifetime and polymerization rates of catalysts, a few
conclusions may be drawn from selected surveys. In Table 1, Kaminsky and
co-workers compare (115) the activities of metallocenes with different metals and
cocatalysts.

Under identical conditions, metallocenes based on Hf are approximately

an order of magnitude less active than their zirconocene analogues, and ti-
tanocenes are likewise of low activity. (There are inportant exceptions to this
rule: rac-C

(Ind)

HfCl

has been shown to have activity comparable to

rac-C

(Ind)

ZrCl

; see Ref. (116).) Of the aluminoxanes, MAO is by far the

METALLOCENES

Vol. 7

Table 1. Ethylene Polymerization Activities for Aluminoxane-Activated Unsubstituted
Metallocenes 115

Catalyst

µmol

◦

Activity

MW/1000

Cocatalyst

TiCl

430

MAO

ZrCl

0.01

5000

122

MAO

ZrCl

1.5

>500

EAO

ZrMe

175

>500

(iBu

Al)

HfCl

0.03

490

MAO

Units: g(PE)/g(M)/h/bar.

best activator. On the basis of periodic trends, the metal–carbon bond strength
for the L

M(polymer) cation should be greater for M

= Hf, and this may explain

the sluggishness of the third-row systems. A similar argument could be used to
account for the much lower rates of chain transfer (in the absence of added H

)

in hafnocenes. In comparing the behavior of hafnocenes and zirconocenes, it is
important to determine the level of Zr impurity in the Hf starting material (about
1 mol% in unenriched samples). This is normally carried through the synthesis
and can lead to non–single-site behavior because of the much higher activity of
the zirconium analogue (117).

Bridged metallocenes tend to incorporate larger oleﬁns more readily than

their unbridged counterparts. This probably has less to do with accessibility of
the monomer than with electronic effects, although the fact that large-amplitude
motions are restrained in the ansa-metallocenes cannot be ignored.

Electron-withdrawing substituents seem to cause polymerization activity to

diminish, if steric effects are accounted for (118,119).

Effects of Additives on Polymerization

Compared with the effort spent in optimizing metallocene catalyst performance
through modiﬁcation of the ligand framework, tuning catalyst properties through
the addition of catalyst modiﬁers has been far less popular. However, cat-
alyst behavior is strongly inﬂuenced by the presence of certain agents, ei-
ther directly (at the transition metal) or indirectly (by modifying the cocata-
lyst/anion). There are some reports of synergistic effects from mixed MAO/discrete
activator cocatalyst systems (120). Modiﬁcation of the MAO cocatalyst with
Lewis bases such as Si(OC

)

, hindered phenols, and bulky amines is de-

scribed in an Exxon patent application (121), in which the catalyst derived from
(CH

)(C

)Si(C

(CH

)

)(N-t-C

)TiCl

+ MAO shows enhanced comonomer in-

corporation and altered chain-transfer rates in the presence of these agents. The
seemingly simple catalysts based on Cp

ZrCl

seem to show distinct two-site be-

havior depending on the cocatalyst, in which the degree of association between
the ions apparently dictates fundamental kinetic parameters such as monomer
selectivity (122).

In commercial reactor operation, one would like the ability to either re-

versibly or irreversibly deactivate the catalyst, depending on the type of reactor

Vol. 7

METALLOCENES

upset. Benzyl chloride and CO

have been claimed to be useful for the reversible

deactivation of metallocenes, with MAO added to reactivate the catalyst (123).
Mobil workers have discovered that CO

, among other agents, can change the

MW-response of MAO-activated metallocenes (124).

Polymerization Using Metallocenes

Most of the experimental and computational work on metallocene polymeriza-
tion has dealt with the classic bis-Cp zirconocene family, and therefore this sec-
tion will focus on this area. However, many of the reactions delineated here can
be expected to occur in polymerizations involving other early transition metal
(early-TM) based catalysts.

Activation.

The paths shown in equation 14 are generally accepted to

be that by which a metallocene precursor is transformed into a catalytic site in
aluminoxane-cocatalyzed polymerization, while equations 8–10 portray the three
mechanisms for activation using Lewis and Brønsted acids.

(O)

(CH

)

Zr(CH

)

+ Al

(O)

(CH

)

−2)

Zr(CH

)

+ Al

(O)

(CH

)

−2)

Zr(CH

)

+ Al

(O)

(CH

)

−1)

−

(14)

The ﬁrst step in equation 14 involves the exchange of the halide groups

attached to the metal atom with the alkyl groups bound to one or more aluminum
atoms. In equations 8–10, it is assumed that the metallocene is introduced into
the reactor as a dialkyl complex, or that it has been converted into such form
by a halide-exchange process involving an alkylating agent in the polymerization
environment.

The next step in both MAO-activated and MAO-free systems is believed to

involve the transfer of an alkylide group to a Lewis acid (eqs. 8 and 9) or a pro-
ton (eq. 10), which creates an electron-deﬁcient (14 e

−

) metal center with a site

for the potential coordination of an oleﬁn. The nature of the counterion strongly
affects the ability of the site to polymerize oleﬁns, and indeed may change the
active-site geometry enough to drastically change the selectivity of polymerization
as well as the chain-transfer rate. Cation–anion interactions have been reviewed
by Bochmann (125). The reactivity of anion-free cationic metallocenes has been
probed by ion-cyclotron resonance (ICR) spectroscopy (126).

One variation on the ionization of dialkyl metallocenes is shown in

equation 15, in which a butadiene complex, which may be predominantly M(IV)
or M(II) in nature, reacts with a Lewis acid to form a Zwitterion that is able to
further polymerize oleﬁns, as shown by Erker and co-workers (127).

METALLOCENES

Vol. 7

[M]

B(C

)

[M]

)

[M]

= Cp

(15)

CHR

trans state

(16)

The approach and insertion of an oleﬁn molecule may or may not pass

through a local minimum or coordination complex (ﬁrst in brackets in eq. 16);
recent theoretical work (128) indicates that the well, if it indeed exists, is very
shallow. The insertion of the new molecule into the growing chain is represented
in equation 13 as a structure intermediate between reactants and products. The
mechanism for this apparently concerted reaction does not involve the participa-
tion of metal-based electrons, and can be considered to be a Lewis acid–assisted
anionic attack of the zirconium alkyl (ie, the polymer chain) upon one end of a
carbon–carbon double bond. The concept of this reaction pre-dates metallocene
study, and is merely a variant of the Cossee–Arlman mechanism (129) routinely
invoked in Ziegler–Natta polymerization. Computational studies indicate (130)
that an

α-agostic interaction (131) provides much needed stabilization during the

process of insertion.

A major consequence of this pathway, also known as migratory insertion,

is that the growing chain sweeps from one side to another with every addition
of monomer. This is a generalization, for some authors have explained the loss
of stereoselection in metallocene polymerizations of propylene, for example, at
low monomer concentrations to the action of a “windshield wiper” isomerization,
in which the polymer chain and open coordination site switch places without
the beneﬁt of monomer insertion. At low insertion rates, this site inversion phe-
nomenon may become competitive with insertion and thus render ineffective any
substituent inﬂuences which differ between the two faces of the catalyst site. With
appropriate ligand design, different or enantiomeric steric environments may be
created for the two sides of the active site. This makes possible stereoselective
polymerization of propylene and higher

α-oleﬁns, as will be seen below.

The use of discrete activators was instrumental in elucidating the structure

and electronic nature of the active site for Group IVB metallocenes. A number of
Lewis base–stabilized metallocene cations such as 41 (132), 42 (133), and 43 (134)
were studied, chieﬂy by the Jordan group at the University of Iowa. Reactivity
is not completely shut down by the base, although polymerizations are sluggish.
Remarkably, agostic species such as 42 and 43, which are believed to be impor-
tant in chain transfer and propagation, respectively, can be characterized in such
systems.

Vol. 7

METALLOCENES

[M]

BPh

t-Bu

[M]

BPh

[M]

i-Pr

BPh

[M]

= (C

)

Hf; L

= PMe

[M]

= (MeCP)

Zr; L

= PMe

[M]

= CP

Zr; L

= MeCN

Deactivation.

One of the factors that complicates the quantiﬁcation of

active-site concentration (135) is the fact that metallocene cations are subject
to equilibria between catalytically active and inactive forms. In situations in
which intramolecular coordination of an arene group can occur, this process com-
petes with monomer coordination in styrene (136) and possibly oleﬁn polymer-
ization. Another dormant state invoked to explain catalyst decay is the dimeric
structure [Cp

Zr(CH

)(

µ-CH

)Zr(CH

)Cp

]

in which a methyl group bridges two

metallocene fragments. This has been characterized by NMR for the reaction of
Cp

Zr(

)

with MAO and other cocatalysts (136).

Ethylene–

α-Oleﬁn Copolymerization.

Linear low density polyethylene

(LLDPE), engineered to match the ﬂexibility of polyethylene made by free-radical
processes (LDPE), contains short-chain branches, which reduce the crystallinity of
the resin. The short-chain branches result from incorporation of

α-oleﬁns such as

1-butene, 1-hexene, or 1-octene (known collectively as comonomers) when added to
the reactive mixture. Traditional catalysts such as MgCl

/TiCl

/AlR

incorporate

these comonomers in a manner that is very inhomogeneous, that is, there is con-
siderable breadth in the distribution of comonomer content both within individual
chains and from chain to chain. The ﬁrst form of inhomogeneity is often referred
to as blockiness, and is best estimated by examination of the monomer sequence
distribution as determined by

C NMR spectroscopy. The second, intermolecular

type of inhomogeneity shows up in extraction experiments, and is particularly
well suited to TREF chromatography (temperature rising elution fractionation
chromatography, which separates polymer fractions on the basis of solubility as
solvent temperature rises.) The single-site nature of metallocenes gives rise to ex-
tremely narrow intra- and intermolecular distributions of short-chain branching.

Figures 6 and 7 compare the sequence distributions and TREF

chromatograms respectively for Ziegler–Natta- and metallocene-catalyzed
ethylene–1-hexene copolymerizations. The lack of blockiness in the metallocene
resin in Figure 6 is demonstrated by the low levels of homopolymer triads, whereas
intermolecular compositional homogeneity is shown by the narrow peak for the
metallocene resin in Figure 7. It should be noted that while the narrow intermolec-
ular distribution of comonomer content is a direct consequence of the single-site
nature of the catalyst, the absence of blockiness is a reﬂection of the relative mag-
nitudes of the rate constants for insertion for the two monomers and does not
speak to catalyst site diversity. The kinetics of insertion will be treated in the
following section.

Exceptions to the rule of narrow intermolecular distribution of comonomer

can and do occur. These often result from the activity of catalytic sites of dif-
ferent structure (ie, loss of single-site nature) or, less often, from mass-transfer

METALLOCENES

Vol. 7

0.5

0.4

0.3

0.2

0.1

PPP

PPE

EPE

PEP

PEE

EEE

Triad

Amount of tr

iad in mole fr

action

Ethylene content 63

±1mol%

Fig. 6.

A comparison of the triad distributions for three ethylene-propylene copolymers

of similar propylene contents made by three different catalysts: MgCl

/TiCl

/AlEt

(black);

VOCl

/Et

AlCl (hatched); and Cp

ZrCl

(white). Reprinted with permission from Ref. 138.

100

110

120

130

Temperature,

°C

eight fr

action, %

Fig. 7.

TREF chromatograms for metallocene (dashed line; density

= 0.922 g/mL) and

Ziegler–Natta (solid and dot-dashed lines; densities

= 0.921 and 0.930 g/mL, resp.) cat-

alyzed ethylene-1-hexene copolymers. Reprinted from Ref. 139. Copyright (1994), with per-
mission from Elsevier Science.

limitations that may occur between gas and catalyst solution, between gas and
growing polymer particle, or between solution and polymer particle, depending
on the process employed. Diffusional limitations only effect the distribution of
comonomer content if there is a considerable gradient in the relative or absolute
concentration of monomer(s) across a nonnegligible fraction of the diameter of the
growing particle or volume of solution. Other potential sources of heterogeneity

Vol. 7

METALLOCENES

]

CHR]

]

CHR]

Scheme 2.

include distributions of reactor residence time and heat history, imposed either
by the reactor design or by a wide range of catalyst particle sizes. Reports on the
control of comonomer content heterogeneity have been few.

The relative rates of oleﬁn insertion reactions lead to models for the reactiv-

ity of the catalyst site, and the complexity of any chosen model arises principally
from the degree to which the last monomers to be inserted affect the rate of sub-
sequent insertion. If the last inserted monomer is assumed to have no effect, then
one obtains two rate constants for insertion: one for reaction with ethylene and one
for reaction with

α-oleﬁn. The next level of discrimination, also called First-Order

Markov theory, posits that the last-inserted monomer inﬂuences the rate of inser-
tion of the next unit, yielding four rate constants (Scheme 2). Higher-level treat-
ments lead to many more rate constants in obvious extension. Metallocene copoly-
merizations display the tell-tale signs of First-Order (or higher) Markov behavior.
This is necessary, for example, to explain the tendency for ethylene–

α-oleﬁn copoly-

merizations to alternate between ethylene and comonomer insertions beyond the
proportion predicted by a random distribution. While First-Order copolymeriza-
tion has been the basis of many studies of metallocene-based copolymerization,
some researchers, most notably Fink (140), have argued for Second-Order Markov
kinetics. A persuasive and statistically meaningful ﬁt of copolymer unit sequence
data to a model with eight parameters is difﬁcult to achieve, however; so this level
of treatment is unlikely to be generally applied.

At the ﬁrst-order level, the preference of the metallocene for one monomer

over the other is quantiﬁed by the two reactivity ratios, r

(

= k

) and r

(

). Since ethylene is invariably more reactive than any other monomer, r

> 1

and r

< 1. Study of the First-Order Markov constants (Table 2) reveals important

correlations with metallocene and comonomer structure. As is to be expected, the
larger the comonomer, the higher the r

value, indicating a steric hindrance to

its insertion. Bridged metallocenes, such as structure 2, give catalysts with much
lower r

values than do unbridged metallocenes. In general, metallocenes copoly-

merize monomers with a tendency toward alternation of monomers, as indicated
by a value for r

less than 1.

Herfert and co-workers resolved the copolymerization reactivity ratios for

two bridged metallocenes at the second-Order Markov level (140), as shown in
Table 3. The data demonstrate bridged metallocenes are somewhat sensitive to

METALLOCENES

Vol. 7

Table 2. Reactivity Ratios for Ethylene/Propylene Copolymerization by Various
Metallocenes

Catalyst

Temp,

◦

Reference

)

ZrCl

250

0.002

0.5

141

(MeCp)

ZrCl

—

141

ZrCl

0.015

0.72

141

ZrCl

16.1

0.025

0.40

142

rac-Me

(3-t-BuCp)

ZrCl

25.6

0.110

2.8

143

(Me

SiCp)

ZrCl

0.029

0.70

141

HfCl

20.6

0.074

1.52

142

(n-BuCp)

ZrCl

0.005

0.095

144

(n-BuCp)

ZrCl

2.7

—

145

TiCl

15.7

0.009

0.14

142

Ind

ZrCl

13.75

0.182

2.45

146

rac-C

(Ind)

ZrCl

6.61

0.06

0.40

147

rac-Me

Si(2-Me-4-Ph-Ind)

ZrCl

2.84

0.74

2.10

148

C(Cp)(Flu)ZrCl

1.3

0.2

0.26

149

Si(Me

Cp)(N-t-Bu)TiCl

1.35

0.82

1.10

148

Table 3. Effect of Comonomer Size on Copolymerization Parameters for Two
Bridged Metallocenes at 25

◦

C (141)

Metallocene

Comonomer

Propylene

4.1

3.9

0.153

0.065

1-Butene

7.9

6.8

0.085

0.017

1-Hexene

10.3

6.4

0.111

0.022

Propylene

3.4

2.2

0.270

0.153

1-Butene

3.6

2.9

0.210

0.041

1-Hexene

3.2

2.6

0.150

0.065

= Me

Si(Ind)

ZrCl

; B

= Me

C(Cp)(Flu)ZrCl

)

the size of the comonomer, but that this effect is ligand-dependent. There is also
a trend toward a greater degree of alternation of monomers as comonomer size
rises, probably due to increased steric repulsion preventing consecutive

α-oleﬁn

insertion.

In the case of terpolymerization of ethylene with two dissimilar

α-oleﬁns,

an apparent competition between the larger monomers causes the increase in
monomer concentration of one to cause a reduction in the incorporation of the
other (150).

Within the best-described family of catalysts, the bis-Cp zirconocenes, steric

congestion tends to reduce the ability to insert the bulkier

α-oleﬁn, as might be

expected. Burger and co-workers have proposed (151) a two-parameter method-
ology for quantifying the steric resistance to monomer coordination. These
parameters, denoted the coordination gap aperture angle and the coordina-
tion gap obliquity angle are demonstrated in Figure 8 for the bridged met-
allocene rac-(CH

)

Si(3-C

Cp)

ZrCl

. Electronic effects often dominate over

steric effects, especially when polycyclic aromatic groups such as ﬂuorenyl

Vol. 7

METALLOCENES

6 Å

3 Å

Obliquity =

−21°

Aperture = 91

Fig. 8.

Aperture and obliquity angles as deﬁned in Ref. 151. Reprinted with permission

of Wiley-VCH and the authors.

and benz[e]indenyl groups are considered. For example, in ethylene–1-octene
copolymerization, rac-(CH

)

Si(benz[e]indenyl)

ZrCl

/MAO incorporates more

comonomer than does rac-(CH

)

Si(Ind)

ZrCl

/MAO, despite the greater bulk of

the benzannelated ligand (152).

The reader will note that the regiochemistry for the insertion of oleﬁns in

Scheme 2 and elsewhere is presumed to be that which causes carbon 1 of the
α-oleﬁn to be bound to the metal; this is denoted 1,2-insertion, while the case in
which the monomer inserts with opposite arrangement is known as 2,1-insertion
or regioerror. While chain-end studies have shown 1,2-insertion to be the domi-
nant regiochemistry, misinsertion may exert an inﬂuence upon the rates of chain
transfer or ultimate polymer properties out of proportion to its frequency. Regio-
chemical inversion seems to be rather prevalent in polypropylene produced by
mono-Cp titanocene catalysts (153).

As is often the case with polyethylene catalysts, both single-site and oth-

erwise, polymerization rates rise rapidly as the level of comonomer rises. This
phenomenon, often called the comonomer boost, has generated some contention
among researchers, as it is unclear whether the effect represents a rate enhance-
ment for insertion based on the donor effect of comonomer, or whether it derives
from a physical effect, such as a lowering of the mass-transport barrier between
the reservoir of monomer and the catalyst site. The two causes are difﬁcult to
separate, but at least part of the effect is diffusional: transport of small molecules
through polyethylene is believed to occur solely through amorphous regions, and
so any reduction in bulk crystallinity, such as through the inclusion of short-chain
branching, should dramatically increase the diffusion rate.

Chain Termination.

In this section, we will cover how polymer chain

growth ends and a new chain begins. This process, alternately called chain trans-
fer or chain termination, leads to a distribution of molecular weights. These two
terms, often used interchangeably in oleﬁn polymerization, mean rather different
things in other areas of polymerization, in which the reactivity of a polymer chain

METALLOCENES

Vol. 7

end (eg, a radical) can either be transferred (eg by hydrogen atom migration) to a
new monomer or be quenched or terminated by reaction with a sacriﬁcial agent
(eg, a radical trap). An ideal single-site catalyst will produce a distribution of chain
lengths conforming to the Schulz–Flory model and this leads to the ratio between
weight-averaged and number-averaged molecular weights, or polydispersity in-
dex (PDI), being equal to two. The PDI tends to the limiting value of 2.0 so long as
(1) the rate of chain transfer is sufﬁciently large relative to the rate of initiation
(activation), and (2) there are no reactions that lead to branched polymer, eg, di-
oleﬁns or reactive polymer chain ends. In practice, the distribution in MW is also
subject to complications similar to those that affect the comonomer distribution
(see above), inevitably resulting in larger values for the PDI. Now we will take a
closer look at the processes involved in chain transfer, and will divide them into
two categories: those that are induced by the action of chain-transfer agents and
those that occur in their absence.

Endogenous Chain Termination.

Two processes are now claimed to rep-

resent the vast majority of chain-transfer events that happen without the inter-
vention of chain-transfer agents. The ﬁrst, known as

β-hydride elimination and

shown in equation 17, is a mechanism of considerable lineage in organometallic
chemistry.

[M]

(17)

Here, the extrusion of a growing polymer chain is effected by the transfer

to the metal atom of an equivalent of H

−

from the carbon atom

β to this metal.

Theoreticians lean toward the existence of a discrete intermediate along this re-
action coordinate, often designated the agostic complex, in which electron density
is donated by the C H bond undergoing the process of breaking. Agostic-type
nonclassical bonding arrangements have been observed experimentally in special
cases, but have been invoked for many aspects of early-TM polymerization, includ-
ing the process of monomer insertion. The strength, estimated to lie in the range
of 8–15 kcal/mol, of the agostic interaction is a testament to the electrophilic-
ity of this electron-poor, cationic metal center. The reaction is completed by the
dissociation of a polymer chain with an unsaturated chain end.

In copolymerizations, the double bond formed by

β-hydride elimination may

signify the nature of the last-inserted monomer. As shown in Scheme 3, the type
of unsaturated group reﬂects whether the monomer was ethylene or comonomer
(in this case, 1-hexene). In fact, the regiochemistry of the last insertion may also
be given by the placement of the double bond. In normal insertion, followed by
elimination, a pendent methylene group is observed, while regiomisinsertion will
generate an internal oleﬁnic group, whose regiochemistry is dictated by whether
the hydrogen that is transferred to metal originates on C

or C

of the original

α-oleﬁn (This generalized discussion ignores isomerization reactions which may
occur prior to chain release, and also does not treat the formation of trisubstituted
oleﬁnic chain ends, which have been reported.)

Vol. 7

METALLOCENES

H + H

CHRR

′

CHRR

′

H +

CHRR

′

H +

′

′RHC

Scheme 3.

(In this Scheme, wavy lines indicate both E and Z isomers are possible, but

in practice, only the trans isomer derived from path (a) is observed.) The corre-
spondence between unsaturation structure and last-inserted monomer does break
down: in certain documented cases (154), chain-end isomerization leads to a pre-
ponderance of trans-CH

CH CH CH

P chain-termini from ethylene homopoly-

merization.

Another type of chain-transfer reaction that has gained believers in recent

years is the reaction that we will call chain transfer to monomer (in this case, ethy-
lene) (155) (eq. 18). Chain transfer here superﬁcially resembles

β-hydride elimi-

nation, in that the same oleﬁnically terminated chains are formed from the same
monomers, but in this case the

β-hydride is transferred directly from the chain

to another monomer molecule. The reader will note that this process is the con-
ﬂation of

β-hydride elimination followed by insertion of monomer into the new

M H bond, and indeed it is difﬁcult to distinguish between these mechanisms
by labelling studies, for example. Unlike

β-hydride elimination, however, chain

transfer to monomer is a ﬁrst-order process in monomer concentration, and if
one assumes a rate for monomer insertion that is also ﬁrst order in monomer
concentration, then the number-average MW should be unaffected by monomer
concentration in the case of chain transfer to monomer, while it should rise in
proportion to monomer concertration in the case of

β-hydride elimination.

H R

(18)

Some new work on ethylene homopolymerization indicates that this mech-

anism may dominate, at least under certain conditions (156). In fact, in the case

METALLOCENES

Vol. 7

of “certain meso-bridged zirconocenes,” ethyl branching occurs, and it is believed
to derive from the reinsertion of a vinyl-terminated polymer chain after chain
transfer to ethylene (157). However, it appears certain that

β-hydride elimination

is a large component of chain transfer in the presence of comonomer. Generally,
the MW of metallocene copolymers falls sharply as comonomer content rises. This
is attributed to the higher reactivity of the

β-hydride group attached to a ter-

tiary as opposed to a secondary carbon atom. There is ample evidence that the
MW of copolymers rises as the overall monomer concentrations rise, even with
the ratio of monomer concentrations held ﬁxed and the comonomer content of the
polymer remaining nearly constant. The most parsimonious explanation for this
phenomenon is the existence of a unimolecular chain transfer process involving
loss of a chain containing comonomer as the last inserted unit.

A process that has been noted in lanthanide metallocene polymerizations

α-oleﬁns as well as in the polymerization of propylene by the highly sterically

congested metallocene (C

(CH

)

)ZrCl

/MAO (158) is that shown below (eq. 19), in

which an alkyl group attached to the

β-carbon of the polymer chain is transferred

to metal, resulting in the loss of a vinyl-terminated chain.

M CH

(19)

Transfer of side-chain substituents larger than methyl is unknown.
Endogenous chain-transfer rates commonly increase with temperature. The

empirical activation energy found for H-transfer termination is about 34 kcal/mol
estimated from relative barrier to trans-vinylene termination over propagation in
Ref. 159.

Exogenous Chain Termination.

Industrial Ziegler–Natta processes rou-

tinely use chain-transfer agents, predominantly hydrogen, for MW control. This
reaction (eq. 17), as well as others, such as chain transfer to a metal alkyl com-
pound such as diethylzinc or triisobutylaluminum, works well for metallocenes
also. This type of chain transfer liberates a molecule with a saturated chain
end. Hydrogen has also been reported to alter overall metallocene activity, ei-
ther upwards or downwards. The phenomenon of rate acceleration by H

may

be related to the reactivation of dormant catalyst sites that contain sterically
demanding metal–polymer environments, eg following 2,1-insertion of propylene
(160).

H +

(20)

Other chain-transfer processes that have been observed are chain transfer

to aluminum (eq. 21) and to silicon (eq. 22).

Vol. 7

METALLOCENES

M R

′ +

AlR

′

(21)

M H +

′

SiH

−x)

′

−x)

(22)

In an elegant adaptation of the latter reaction, Koo and Marks found (161)

that 1,3,5-trisilylbenzene can lead to long-chain–branched copolymers through
successive chain-transfer events (eq. 23).

SiH

TiMe

t-Bu

B(C

)

(23)

This technique was also applied to the production of block copolymers of

dissimilar subunits, using silyl-capped polymer chains as chain-transfer agents.
Both mechanisms for chain extension using silanes could in principle be mirrored
in similar reactions with polyfunctional alkylaluminum or boron compounds, but
the ensuing C (Al/B) linkages would be vulnerable to hydrolysis, although oxi-
dation of a borane-terminated polyethylene sample can lead to a growth point for
methacrylate polymerization, yielding a diblock copolymer (162).

Living Polymerization.

Polymerization in the absence of both chain transfer

and termination, or living polymerization, has value to the study of metallocene
reactivity, but little potential beyond very specialized applications. If initiation
of polymerization is rapid on the timescale of the polymerization run, PDIs will
approach 1.0, the limiting value representing monodispersity. Several examples
of metallocene-based living oleﬁn polymerization catalysts have appeared in the
literature (105,163).

Long-Chain Branching.

Processing metallocene polyethylene can be dif-

ﬁcult, especially if extrusion equipment optimized for high pressure–low density
LDPE is used. Not only is the polymer melt of relatively high viscosity for a desired
application because of the narrow molecular weight distribution, it does not un-
dergo shear thinning to the same extent as LDPE because of the lack of long-chain
branches, which tend to reduce a polymer’s radius of gyration. During the inves-
tigation of metallocene polymerization using catalysts with very low r

values for

copolymerization, generally bridged bis(indenyl)zirconocenes, researchers noted
rheological behavior indicative of long-chain branching (LCB) in polymer melts

METALLOCENES

Vol. 7

and solutions. The mechanism for LCB formation in metallocene systems has been
postulated as in equation 24, where reincorporation of a vinyl-terminated chain
end from a dissociated polymer molecule, followed by further monomer insertion
and ﬁnally chain loss, yields a polymer with three branches radiating from a me-
thine carbon. (For the branches to have rheological impact, it is believed that they
should be about the entanglement length of linear polyethylene, or about 3 kD.)

′

(24)

Reinking and co-workers have raised the possibility that at least some of the

LCB seen in metallocene-based polymers may arise from an unusual aliphatic
C H activation of dissociated (“dead”) polymer molecules (eq. 25) (As part of a
study of non-metallocene vanadium catalysts, see Ref. 164). In this study, a bridged
bis(indenyl) zirconocene was found to incorporate ethylene units into the heptane
solvent, while an unbridged version, (n-C

Cp)

ZrCl

, did not.

(25)

The reactions in both (equations 24 and 25 lead to long-chain branches of

the Y-type, with three arms of appreciable length. The higher the level of LCB
by these routes, the more one obtains polymer molecules of a bushy or dendritic
structure (Fig. 9a). Cyclic structures cannot result from equation 24, a fact that
reduces signiﬁcantly the danger of passing into a cross-linked or gelled condition.
Cross-linking is much more difﬁcult to avoid when LCB is engendered by the
copolymerization of an

α,ω-diene such as 1,9-decadiene with ethylene (eq. 26),

(CH

)

(26)

which results in an H-type or four-armed polymer node (Fig. 9b).

Because gelled or networked polyoleﬁns are almost completely infusible and

insoluble, even small amounts can adversely affect the appearance and mechani-
cal performance of a fabricated article.

Side Reactions.

Many of the reactions of metallocenes that occur as mi-

nor pathways during polymerization demonstrate the extreme electrophilicity of
the metal cation, in that fairly unreactive bonds may be broken. Generally, the

Vol. 7

METALLOCENES

(a)

(b)

Fig. 9.

Schematics of (a) a dendritic polymer (left) and (b) a four-armed star polymer.

occasion for such a process is the formation of a species that has a low rate of
monomer insertion, thus increasing the likelihood of lower-probability reactions.
One interesting reaction that may inﬂuence the rate of chain transfer in batch
polymerizations is shown in equation 27). In the ﬁrst step of this process, detected
by NMR (165) and by ion-cyclotron resonance (166), activation of one of the al-
lylic C H bonds causes the extrusion of one molecule of H

and the formation of

an allyl–Zr cation. This structure has been considered inactive for further prop-
agation, yet there is evidence (167) that return to the polymerization cycle can
occur, either through the reverse of the allylic-activation process or through the
direct insertion of ethylene into the Zr–allyl bond, leading to unsaturation in the
polymer backbone.

H H

[Zr]

≡ Ph

C(Cp)(Flu)Zr

[Zr]

H H

−H

[Zr]

polymerization

(27)

Researchers have also noted that metallocene catalysts can exchange Zr C

and C H bonds during propagation, leading to 1,3-enchainment of propylene, for
example, following regiomisinsertion (eq. 28) (168).

[M]

(28)

A number of very similar paths have been proposed (169) for the rearrange-

ment of polyethylene chains attached to (C

(CH

)

, another system in which

propagation is sterically hindered (albeit by the ligand in this case). Interestingly,

METALLOCENES

Vol. 7

[M]

H Me

[M]

Scheme 4.

both of the previous two mechanisms require the existence, though transient, of
a metal–oleﬁn complex.

In the isotactic polymerization of propylene mediated by bridged metal-

locenes, certain errors in polymer microstructure have been conclusively found
(170) to result from an elimination–reincorporation mechanism involving inser-
tion of a 1,2-disubstituted unsaturated polymer chain end as in Scheme 4.

Although the metal–Cp linkage is usually assumed to be inviolable in most

mechanistic treatments of metallocene polymerization, there is evidence that this
assumption is ﬂawed. The rac/meso ratios of certain bridged bis(indenyl) metal-
locenes have been shown to equilibrate in the presence of light (171), LiCl, or
MeLi (172). These processes require the rupture of the M–Cp bond to one half
of the ligand, the rotation of the dissociated C

ring, and its reassociation with

the metal. Titanocenes and Group III metals appear to be more susceptible than
zirconocenes to this process (although it is claimed to occur for the latter as well).
Whether this reaction is important to polymerization has not been investigated,
but it could affect catalyst quality, especially in isotactic polypropylene production.

Stereoselective Polymerization.

In the early 1980s, Ewen (173) and

Kaminsky (174,175) independently examined the polymerization behavior of met-
allocenes rac-Et(Ind)

TiCl

and rac-Et(4,5,6,7-IndH

)

ZrCl

(44) each with two

indenyl fragments linked by a dimethylene bridge.

It was found that polypropylene made by such complexes (cocatalyzed by

MAO) tended to be isotactic, as opposed to the atactic polypropylene that had been
hitherto made by metallocenes such as Cp

ZrCl

. The strong relationship between

ligand structure and polymer tacticity is considered to be the most secure proof of
the migratory insertion mechanism for polymerization. The lure of homogeneous
stereoselective oleﬁn polymerization is responsible for much of the diversity of
ligand structures within the bis(Cp) family.

Since metallocenes with ligands that protrude into the coordination zone in

a fashion represented in Scheme 5 by complex A give rise to isospeciﬁc poly-
merization and those of structure B lead to syndiospeciﬁc polymerization, re-
searchers generally assumed that the interactions causing stereoselection were

Vol. 7

METALLOCENES

Scheme 5.

those directly between ligand and incoming monomer. Catalyst sites, therefore, of
approximately C

symmetry (A) would then choose the same face of the prochiral

monomer regardless of which side of the catalytic site the monomer approached.
Sites with approximate C

symmetry (B) would cause a selection of alternately

re or si faces of the incoming monomer as the polymer chain switcheds sides.
Bercaw has recently produced evidence (54) that the steric interactions that favor
monomer face alternation in complexes of type B are those between the grow-
ing chain and the incoming monomer. (A similar mechanism for stereoregulation
had earlier been proposed by Guerra and co-workers (176) for complexes of C

symmetry on the basis of molecular mechanics models.)

In this mechanism (Scheme 6), the directing effect of the ligand structure

is upon the orientation of the polymer chain (lower dotted curve in Scheme),
which then in turn dictates the preferred orientation of the incoming oleﬁn. This
leads to the somewhat counterintuitive conclusion that the methyl group in propy-
lene is actually positioned toward the large ﬂuorenyl side of the ligand in all in-
sertions for Cp–ﬂuorenyl complexes. While polymer–monomer interactions were
advanced by Brintzinger and co-workers (177) as factors partially responsible
for stereoselection in isospeciﬁc polymerization some years earlier, it is only in
the case of syndiospeciﬁc polymerization that the hypotheses predict different
monomer orientations with respect to the ligand. Combined ab initio and molec-
ular mechanics calculations (178) had also led to the conclusion that the poly-
mer chain controls the orientation of monomer in isospeciﬁc polymerization by
(CH

)

Si(1,3-(CH

)

Cp)Zr(P)

Isotactic Polypropylene.

The bis(indenyl) metallocenes that are most often

associated with isotactic polypropylene (iPP) and other isotactic polyoleﬁns are
shown in structures 45–49.

H P

Scheme 6.

METALLOCENES

Vol. 7

(naphthyl)

Synthesis of the complexes usually requires the separation (normally by

fractional crystallization) of the racemic isomer pair from the meso isomer, since
the latter leads only to atactic polymerization of propylene. The syntheses explored
by Jordan, described above in the section Preparation, tend to raise the overall
yield of racemic metallocene.

The resolution of the enantiomers of bridged bis(indenyl) metallocene dichlo-

rides has been accomplished by the replacement of both halides by one enantiomer
of binaphthol, followed by chromatography (179). Direct synthesis of enantiomer-
ically pure precursors via chiral epoxides has been reported (180). Polymerization
of

α-oleﬁns using such precursors does not lead to the production of apprecia-

bly chiral polymer due to the de facto mirror plane which exists in an isotactic
poly(

α-oleﬁn) of reasonably high degree of polymerization (Fig. 10).

The quest for iPP of commercially acceptable melting point (MP) has led to

preparation of bis(indene) ligands with increasingly elaborate substitution pat-
terns. MP appears to correlate positively with the run length of isotactic dyads
(stereoselective enchainment) as well as negatively with the level of regiomisin-
sertions. Substitution with small alkyl groups at the 2-position of the indene
causes the iPP to be much reduced in regioerrors, while increasing bulk at the

Fig. 10.

Vol. 7

METALLOCENES

Table 4. Bridged Bis(indenyl) Zirconocenes as Polypropylene Catalysts Ligand
Effects

Zr,

Catalyst

µmol Activity

/1000

◦

mmmm

rac-Me

Si(Ind)

ZrCl

190

138

81.7

rac-Me

Si(2-Me-4-iPrInd)

ZrCl

245

213

150

88.6

rac-Me

Si(4,5-benzoindenyl)

ZrCl

274

138

80.5

rac-Me

Si(2-Me-4,5-benzoindenyl)

ZrCl

403

330

146

88.7

rac-Me

Si(2-Me-4-PhInd)

ZrCl

755

729

157

95.2

rac-(Ph)(Me)Si(2-Me-4-PhInd)

ZrCl

553

778

157

95.1

rac-Me

Si(2-Me-5-PhInd)

ZrCl

188

139

78.1

rac-Me

Si(4-PhInd)

ZrCl

148

86.5

rac-Me

Si(2-Me-4-naphthylindenyl)

ZrCl

875

920

161

99.1

Reprinted with permission from Ref. 182. Copyright (1994) American Chemical Society.

Conditions: 70

◦

C, run time 1 h, liquid propylene, Al/Zr

= 15,000.

Units: kg(PP)/mmol(Zr)/h.

Percentage of sequences of 5 propylene units that are perfectly isotactic.

4-position has the effect of suppressing stereoerrors. Table 4 details the effects
of alkyl placement on indenyl ligands upon polypropylene properties, while Mise
and co-workers (181) have collected data on a large number of bis(Cp) complexes.

Recent work in the area of isospeciﬁc oleﬁn polymerization has challenged

the mindset that C

-symmetry is necessary for high isotacticity. For example,

compound 50 gives P

mmmm

between 80 and 91% (183). Another assumption which

has been overturned is that, in the bridged indenyl systems, substitution at the
3-position defeats the stereocontrol enforced by the fused benzene ring. Complexes
51 (184) and 52 (185) give some of the highest-melting iPP produced by metal-
locenes (up to 162

◦

C; R

= H).

ZrCl

t-Bu

= H, Me

Regioerrors have also been linked to elevated chain-transfer rates (186).

Ligand modiﬁcation with the objective of altering the electronic environment of the
propagating center has not been pursued systematically, although the inductive
effects of substituents may inﬂuence the stereoselectivity: Lee and co-workers
have shown electron-withdrawing groups to suppress stereoerrors in propylene
polymerization by rac-C

(Ind

)

ZrCl

/MAO (Ind

= substituted indenyl) (119).

Isotactic polymers of 1-oleﬁns higher/larger than propylene has been accom-

plished with metallocenes, as would be expected from the enhanced ability of these
catalysts to incorporate large comonomers in ethylene–oleﬁn copolymerization.

METALLOCENES

Vol. 7

Fig. 11.

Syndiotactic polypropylene.

Syndiotactic Polypropylene.

Syndiotactic polypropylene (sPP) is a crys-

talline homopolymer with alternating stereochemistry at each successive methine
carbon in the chain (Fig. 11).

Compared with iPP, sPP has lower modulus and melting point, making it less

useful for applications demanding strength. It does, however, appear to be superior
to iPP in such properties as clarity and ﬂexibility (187,188). Although catalysts
such as VCl

had been shown to produce sPP at low temperature, metallocenes

with C

symmetry such as 3 were the ﬁrst catalysts to make sPP under practical

conditions with high activity (189). The performance of some complexes in this
family is shown in Table 5. Note the effects of bridge substituents on polymer MW
(190).

Mixed Tacticity Polyoleﬁns.

The Waymouth group at Stanford has recently

unveiled (193) a catalyst family for the production of elastomeric polypropylene,
in which blocks of isotactic polymer are linked by segments of essentially atac-
tic material, resulting in a product with geographically separate crystalline and
amorphous domains, which is imbued with high ﬂexibility and recovery. The un-
derlying mechanism has been said to be as shown in Scheme 7, where an unbridged
complex oscillates between racemoid and mesoid conformers on a timescale inter-
mediate between that for monomer insertion and that for production of a single
polymer molecule.

However, recent microstructural investigations have led Busico and others

to advance an alternate mechanism (194), in which two enantiomeric racemoid
catalysts interconvert (Scheme 8) at a rate nearly equal to monomer insertion.

In any case, the relative amounts of isotactic and atactic sequences depend on

temperature, monomer concentration, transition-metal identity (Zr or Hf) as well
as upon the size of the 2-indenyl substituent. Chien (195) and Collins (196) have
also described metallocenes 53 and 54, respectively) which produce elastomeric
polypropylene.

Table 5. Comparison of Bridged Cp–Fluorenyl Metallocenes as Syndiotactic
Polypropylene Catalysts

Catalyst

Al/M

Act.

/1000

◦

rrrr

Reference

C(Cp)(Flu)ZrCl

13,000

117

134, 140

82.7

—

191

C(Cp)(Flu)HfCl

940

1.4

777

114

75.7

—

191

(Ph)(Me)C(Cp)(Flu)ZrCl

2,700

435

—

96.1

192

C(Cp)(Flu)ZrCl

2,700

519

—

95.9

192

Units of kg(PP)/mmol(M)/h.

Percentage of sequences of 5 propylene units which are perfectly syndiotactic.

Syndiotactic Index, as determined by 13C NMR.

Conditions: 50

◦

C, liquid propylene.

Conditions: 70

◦

C, liquid propylene.

Vol. 7

METALLOCENES

[Zr]

+ R

Chain transfer to borane

Borane copolymerization

[Zr]

Block copolymer formation

[O]

•

MMA

MeO

Graft copolymer formation

OOB R

MMA

(MMA)

[O]

= 9-borabicyclo[3.3.1]nonane (9-BBN)

•

Scheme 7.

Scheme 8.

In both of these cases the metallocenes have a highly asymmetric character.

The authors postulate a different kind of oscillation for the origin of the blockiness:
a competition between migratory insertion and polymer chain migration without
insertion. This is shown graphically for Collins’ system in Scheme 9.

Isotactic sequences are here due to successive insertions upon the same

side of the catalyst (rather than from alternate sides as is believed to occur in
C

-symmetric systems), which are made possible by “windshield-wiper” motions

of the chain relative to the catalyst. The dibridged catalyst 55 switches from be-
ing predominantly syndiospeciﬁc to being moderately isospeciﬁc as temperature
is raised and propylene concentration is lowered, probably due to a similar phe-
nomenon (54). The fact that the catalyst with the chiral substituent projecting

METALLOCENES

Vol. 7

PPCH

Isospecific state

Aspecific insertion k

Isospecific insertion

]

Aspecific state

−inv

inv

Scheme 9.

ical Society.)

into the wedge (as shown in 55) suppresses the strong syndiospeciﬁcity evi-
dent in the system with this substituent replaced by H is further evidence of a
missed-insertion mechanism, since one side becomes signiﬁcantly more crowded
than the other.

SiMe

i-Pr

t-Bu

A subtle modiﬁcation of the ligand sphere in Cp–ﬂuorenyl complexes, as in

structure 56, changes the stereospeciﬁcity of the catalyst enough so that hemiiso-
tactic polypropylene, rather than sPP, is produced (197). Apparently, the added
substituent on the Cp ring is enough to remove any enantiofacial selectivity from
one side of the site, but not enough to completely suppress monomer coordination
to one side, as in Scheme 9. This material, shown in Figure 12, can be thought of
as a syndio-isotactic stereoblock polymer.

Fig. 12.

Hemiisotactic polypropylene. (Wavy lines indicate methine units of indetermi-

nate chirality.)

Vol. 7

METALLOCENES

Catalysts such as 56 also lead to alternating isotactic ethylene–propylene

copolymers through a similar mechanism (198), since conditions can be arranged
such that ethylene incorporates predominantly on the more hindered catalyst side.

Other Monomers.

Dienes.

Dienes tend to reduce the polymerization activity of metallocenes,

and it is generally difﬁcult to incorporate more than a few mole pernts into
polyethylene. Monomers having two C C bonds can be incorporated into one
or two chains depending on the reactivity of the two unsaturated fragments. If
conjugated, the diene can be incorporated via 1,2- or 1,4-insertion, leaving one
double bond either in the polymer backbone or on a side chain. Conjugated dienes
are particularly potent catalyst poisons for bis-Cp metallocenes, although there
are occasional reports (199,200) of copolymerization with ethylene. Mono-Cp tita-
nium complexes, by contrast, are fairly competent for the homopolymerization of
1,3-butadiene or isoprene (201). Incorporation of butadiene leads in some cases to
the presence of trans-vinylidene groups and trans-1,2-cyclohexanediyl structures
(202) and in others to trans-1,2-cyclopentanediyl and 1,2-cyclopropanediyl groups.
(203). The oxidation state of titanium in the catalyst is (again) not known, but the
ability of the [C(C

)

][B(C

)

] activator to give high activity for polybutadiene

(1240 g(BR)/mmol(Ti)/h at 30

◦

C, giving about 80% cis and 15% vinyl microstruc-

ture) is suggestive of Ti(IV) character (204).

As mentioned in the section Long-Chain Branching, the presence of low lev-

els of

α,ω-dienes in the reaction mixture for ethylene polymerization can lead to

polyethylene with LCB or cross-linked PE.

An interesting type of polymerization—cyclopolymerization—occurs with

certain bridged metallocenes exposed to short

α,ω-dienes such as 1,5-hexadiene

(205,206). The homopolymerization, shown by (equation 29, involves the initial
1,2-insertion of the ﬁrst vinylic group followed by the 1,2-insertion of the second,
leading to cyclic groups.

[M]

(29)

The microstructure, shown in Figure 13 for the trans-isotactic polymer, is

deﬁned by two types of stereoselection, intra- and inter-ring, since the two types
of insertions in equation 29 differ chemically.

The structure of the metallocene inﬂuences the stereochemistry of acyclic in-

sertion, with isospeciﬁc metallocenes tending to select the same enantioface for the

METALLOCENES

Vol. 7

Fig. 13.

Trans-isotactic poly(methylene-1,3-cyclopentylene).

nonring insertion; aspeciﬁc and isospeciﬁc catalysts both favor trans insertion for
the ring-closing step. One intriguing property of polymer with a structure such as
shown in Figure 13 is that the polymer has no mirror plane (unlike poly(

α-oleﬁns)

as discussed in section Isotactic Polypsopylene), and hence can be optically active
even at high MW. This was demonstrated by using a single enantiomer of the
bridged racemic metallocene ethylenebis(

-4,5,6,7-tetrahydroindenyl)zirconium

dichloride, itself selected by reaction with a single enantiomer of 1,1-bi-2-naphthol.

The most industrially relevant family of dienes used in metallocene polymer-

izations is represented by 57, 58, and 59 (ethylidene norbornene, 1,4-hexadiene,
and dicyclopentadiene, respectively).

These dienes, along with ethylene and propylene, are used to make EPDM,

a sulfur-vulcanizable elastomer. Generally, EPDM contains 25–60 wt% propylene
and 1–5 wt% diene, and is completely amorphous. Metallocene catalysts have
inherent advantages in EPDM polymerization, the most important of which are
the ability to incorporate large amounts of

α-oleﬁn with ease and the avoidance of

long ethylene sequences that lead to crystallinity. Catalysts that appear to have
found commercial application in EPDM include mono-Cp catalysts such 12 (207)
and 14 (208).

Cyclooleﬁns.

The ability of metallocenes to controllably insert previously

recalcitrant monomers is further evidenced by the homo- and copolymerization of
cyclooleﬁns such as cyclopentene. Initially, enchainment of cyclopentene was as-
sumed (209) to proceed in normal 1,2-fashion, as shown in Figure 14a, and indeed
this appears to hold true in copolymerizations with ethylene. However, later work
(210) on homopolymers showed that isomerization following 1,2-incorporation is
obligatory, presumably due to steric considerations, yielding 1,3-inserted units
(Fig. 14b)

(a)

(b)

Fig. 14.

Polycyclopentene regiochemistries.

Vol. 7

METALLOCENES

Polar Comonomers.

Heteroatom-containing monomers have the poten-

tial to allow polyoleﬁns to ameliorate deﬁciencies in key product areas such
as printability and water permeability. Despite the well-deserved reputation
of metallocenes for sensitivity to polar functional groups, Waymouth and
co-workers have polymerized

α-oleﬁns bearing protected hydroxyl or dialky-

lamino groups on the terminal carbon, albeit at fairly low rates (211). The
very bulky metallocene (C

(CH

)

Zr(CH

)

combined with the discrete ac-

tivator [C

(CH

)

NH][B(C

)

] displayed the highest activities for poly

(5-(N,N-diisopropylamino)-1-pentene), apparently due to the inability of the nitro-
gen atom to coordinate to either transition metal or cocatalyst (212). Copolymer-
ization of ethylene and 4-allylanisole followed by deprotection led to the prepara-
tion of polyethylene-bearing phenol groups (213). Incorporation of polymer stabi-
lizers such as vinylic hindered amines (214) via copolymerization could potentially
defeat the phase separation and blooming, which often compromise the activity of
blended-in stabilizers.

Isobutylene.

Unlike conventional Ziegler–Natta and most metallocene cat-

alysts, the compound (CH

)

Si(C

(CH

)

)(NR)Ti(CH

)

= cyclododecyl; activa-

tors

= MAO or discrete ion pair) can copolymerize isobutene with ethylene, as

has been shown by an Exxon group (215). The isobutene units tend to be isolated,
eliminating a carbocationic addition mechanism.

Methylenecycloalkanes.

Strained

cycloalkanes

with

exo-methylene

groups have been homopolymerized as well as copolymerized with ethylene
by catalysts such as [(C

(CH

)

Zr(CH

)]

[(CH

)B(C

)

]

−

and analogous

Group IIIB and lanthanide metallocenes via the C C bond migration mechanism
shown in equation 30 (216).

[M]

(30)

Styrene (as Random Comonomer with Ethylene).

Both bis- and mono-Cp

Group IVB metallocenes copolymerize ethylene and styrene (217). The catalysts
are believed to be M(IV) in nature. The sequential enchainment of two styrene
units is generally disfavored (an exception being the isopropyl-bridged metal-
locenes, as reported in Ref. 218, even at styrene levels approaching 50 mol% in
the polymer, leading to nearly alternating ES copolymers. These copolymers have
properties that vary from thermoplastics to elastomers to crystalline or glassy
solids, depending on the level of styrene incorporation (218) (see Fig. 15).

Although the alternating ES copolymers are generally atactic, crystalline

= 145

◦

C) isotactic alt-ES (Fig. 16) results from copolymerization at

−25

◦

C in

the presence of rac-Et(Ind)

ZrCl

(2) activated by MAO.

Mixed Metallocene Catalysts.

By combining two or more metallocenes

in the (co)polymerization of oleﬁns, polymers of broadened MW distribution or
comonomer distribution may be obtained. In certain cases, it seems the polymer

METALLOCENES

Vol. 7

°C

Type E
Semicrystalline

Type M
Amorphous
rubbery

Type S
Amorphous
glasstomer

−80

−60

−40

−20

Styrene Content, wt%

,°

, %

Fig. 15.

Characteristics of ethylene–styrene copolymers made by INSITE technology.

From 218. Copyright c

(1998). This material is used by permission of Wiley-Liss, Inc.,

a subsidiary of John Wiley & Sons, Inc.

Fig. 16.

Isotactic alternating ethylene–styrene copolymer.

chains produced by one metallocene can serve as a comonomer for a second, thus
making LCB-containing polymer possible in cases where the two catalysts may
not be competent to produce it by themselves.

In a variation of this concept, metallocene catalysts can function in the

presence of oleﬁn oligomerization catalysts, whether the latter be metallocenes

Vol. 7

METALLOCENES

themselves, as in the case of Bazan’s bis(boratabenzene)zirconium catalyst (219),
or entirely different types of catalyst, as with the nickel-based single-component
catalyst 60 (220).

PPh

Polymerization of Monomers Other Than Alkenes

In this section we survey the use of metallocenes as catalyst precursors for poly-
merizations in which the predominant monomer is not ethylene or any

α-oleﬁn.

The subject of metallocenes as initiators for the cationic polymerization of vinyl
monomers is dealt with elsewhere. Except for styrene, true random copolymers of
the monomers in this section with alkenes are not produced because the mecha-
nisms for polymerization of the non-oleﬁns differ greatly from that operating in
Sinn–Kaminsky catalysis.

Styrene.

Mono-Cp complexes of titanium are especially useful in the syn-

diotactic polymerization of styrene (ﬁrst reported in Ref. 221, a material with
high melting point (

∼278

◦

C) and other useful features such as a rapid crystalliza-

tion rate (222). The catalysts are believed (223) to be cationic Ti(III) monoalkyl
species, which then undergo predominantly 2,1-insertions (Scheme 10). This is
true even in the case of systems in which the activator is the ﬂuorinated borane
B(C

)

. The mechanism of reduction is unclear, but appears to be accelerated

in the presence of styrene.

Catalyst activities can be as high as 38 kg(PS)/mmol(Ti)/mol(styrene)/h in

laboratory polymerizations, with syndiospeciﬁcities close to 99% (for the system

B(C

)

B(C

)

B(C

)

[Ti]

− 1) PhCH CH

B(C

)

= Me, CH

SiMe

; [Ti]

= (C

)Ti

Scheme 10.

METALLOCENES

Vol. 7

(CH

)

Ti(OCH

)

N, see Ref. 224). Solvent effects indicate that competition

of aromatic solvent with monomer for the open coordination site affects activity
(225).

Methyl

Methacrylate.

The

quasi–living

polymerization

methyl

methacrylate can be achieved using lanthanide (226) and Group IVB metal-
locene complexes, with the latter requiring organozinc, organoaluminum, or bo-
rate salt activators. At the heart of the chemistry is a metal-assisted Michael
addition, which can occur either through intra- (eq. 31) or intermolecular (eq. 32)
C C bond formation, the latter being more properly described as group-transfer
polymerization.

[Zr]

(31)

[Zr]

OR Me

[Zr]

= Cp

Zr; R

= Me, t-Bu.

(32)

In both cases, propagation requires the facile decomplexation of the poly-

mer chain carboxylate and its replacement by that of the monomer. Isotactic
PMMA results when a racemic bridged bis(indenyl) (227) or amidocyclopenta-
dienyl half-sandwich complex is employed (228); enantiomorphic site control in-
dicates that path (eq. 31) dominates in that case.

Lactones.

Esters cyclo-(CH

)

= 4, 5) have been shown by

Yamashita and co-workers to be efﬁciently polymerized by the lanthanocenes
(C

(CH

)

Sm((CH

))(THF) and [Cp

Yb(CH

)]

(229). Recent work of Evans

and co-workers has revealed that the highly reactive, divalent pentamethyl-
samarocene, optionally solvated by THF, which can effectively polymerize
ε-caprolactone by ring-opening (230), also copolymerizes ethylene carbonate and

ε-caprolactone (eq. 33) (231). Even when present at levels up to 23 mol% in the
copolymer, the carbonate units appear to be isolated, indicating a low level of
blockiness.

+ O

)

Sm(THF)

room temp., 2 d

68% yield

m/(m

+ n) = 0.8

1 : 1

(33)

Vol. 7

METALLOCENES

PhRHSi

[M]

PhH

SiHRPh

[M]

SiHRPh

[M]

PhRHSi

SiHRPh

PhSiH

[M]

= (Cp)(C

)Hf; X

= Cl; R = H or polymer chain.

Scheme 11.

Silanes.

Group IVB metallocenes catalyze the stepwise polymerization

(dehydrocoupling) of primary silanes, yielding H (SiHR)

H and n H

(232).

The mechanism appears to be strictly metathetical for such molecules as
(Cp)(C

(CH

)

)Hf(X) (R

= H, alkyl; X = H, alkyl, chloride), wherein the entire

process takes place on the neutral manifold (Scheme 11) (233).

A different pathway, probably involving the coupling of free silyl rad-

icals, appears to operate in polymerizations catalyzed by Group IVB com-
plexes (eg, Cp

ZrCl

) activated by noncoordinating ionizing agents such as

[(C

)

C][B(C

)

(n-C

)

] (n

= 0–2) following pretreatment with n-C

(234).

Graft and Block Copolymerization of Oleﬁns with Non-Oleﬁns.

The

compositional homogeneity and relatively narrow molecular weight distribu-
tion that are typical for metallocene copolymers present an opportunity for the
functionalization of the majority of polyoleﬁn chains with chains composed of
non-oleﬁnic units. In these cases, the secondary polymerizations are mediated
by catalysts or initiators that are unrelated to the polyoleﬁn catalyst. The reac-
tive groups at which anionic or free-radical polymerization can begin are either
derived from chain-transfer events during metallocene polymerization or from
oleﬁnic comonomers. It is often necessary to chemically transform the end groups
or side chains prior to the polymerization of the non-oleﬁn. The potential advan-
tages of this strategy, which has been thoroughly reviewed by Chung (235), lie
in the compatibilization of polyoleﬁns and other polymers (ie, homopolymers of
the non-oleﬁn blocks). Both graft and block copolymers of ethylene with methyl
methacrylate can be made through the oxidation of borane side-chain or terminal
groups, respectively, as shown in Scheme 12.

Miscellaneous.

Patten and Novak have described (236) the polymer-

ization of isocyanates using CpTiCl

(OCH

) as a single-component cata-

lyst. Unlike non-metallocene titanium catalysts, this species is active for the

METALLOCENES

Vol. 7

[Zr]

+ R

Chain transfer to borane

Borane copolymerization

[Zr]

Block copolymer formation

[O]

•

MMA

MeO

Graft copolymer formation

OOBR

MMA

(MMA)

[O]

= 9-borabicyclo[3.3.1]nonane (9-BBN)

•

Scheme 12.

polymerization of isocyanates with functional side chains such as 61. Acryloni-
trile (237) has been polymerized by lanthanocenes.

N C O

Reactions of Metallocenes Not Involving Polymerization

Transformations of small molecules using Group IVB metallocenes are important
in organic chemistry, and it is impossible to give more than the most superﬁcial
treatment here. Noncatalytic chemistry will be covered only brieﬂy.

A common theme in many reactions is the generation of an equivalent of

M through the reduction of a tetravalent precursor metallocene with a metal

alkyl such as n-butyllithium. The chemistry of oleﬁn complexes of Cp

M is charac-

terized by a facile interconversion between formally M(II) and M(IV) manifolds, as
shown in equation 34. Another central motif of metallocene chemistry, especially
that of titanocenes, is the accessibility of pathways connecting metallacycles and
metal alkylidene complexes, eg in alkene metathesis (eq. 35).

"Cp

(34)

Vol. 7

METALLOCENES

′

[Zr]

′

[Zr]

′

′OH

′X

′

′′

AlCl

1/n [R

′AlCl

]

′′C(O)Cl

[Zr]

= Cp

Zr; X

= Br, I;

Scheme 13.

(35)

Stoichiometric Reactions.

Hydrozirconation.

Of great use in preparative organic chemistry is the ad-

dition of the zirconium–hydride bond to alkenes and alkynes. Not only are the
alkylzirconium species thus formed useful in themselves, giving rise to a diverse
set of organic compounds depending on reagent used to destroy the Zr C bond
(238), they can undergo transmetallation reactions with chloroaluminum com-
pounds with retention of stereochemistry to yield more reactive alkylaluminum
moieties (239) (Scheme 13).

Coupling of Alkenes and Alkynes.

The highly reactive [Cp

Zr] fragment,

generated through trapping of an alkyne complex with Lewis base (240), forms
metallacycles with alkenes, alkynes, ketones, aldehydes, and nitriles (selected
reactions shown in Scheme 14).

Another means of generating zirconocene, via the decomposition of dialkyl

zirconocenes (241) can be used to produce bicyclic organic compounds from the
cyclization reaction followed by carbonylation (Scheme 15).

In both cases, acyclic ﬁnal products can be obtained through quenching

reactions such as aqueous workup or addition of elemental halogen. Alkene

METALLOCENES

Vol. 7

RCCR

′

Zr(H)(Cl)

[Zr]

′

MeLi

PMe

−MH

[Zr]

′

[Zr]

′

[Zr]

′

[Zr]

′

[Zr]

= Cp

PhCN

PMe

Scheme 14.

[Zr]

−BuH

Zr(n-Bu)

[Zr]

= Cp

SiMe

(CH

)

SiMe

(CH

)

[Zr]

SiMe

(CH

)

(1) CO

(2) H

I
I

Scheme 15.

complexes of zirconocene will also form carbon–carbon bonds with alkenes to give
trans-3,4-dialkylzirconacyclopentanes (eq. 36) with selectivity

>98%; these can be

reacted further, eg with CO or HCl to give cyclic and acyclic products, respectively
(242).

[Zr]

= Cp

Zr; R

= Et, n-hexyl

[Zr]

(36)

Metallocene complexes of o-benzyne will also add to alkenes or alkynes,

yielding 1-metallaindans and 1-metallaindenes, respectively (243). The zirconacy-
clopentadienes formed from the dimerization of alkynes are convenient synthons
for phospholes, siloles, and germoles (244).

Addition of styrenes to oleﬁns leads to 2-aryl-4-alkylzirconacyclopentanes.

Reductions.

Bromo- or chlorohydrocarbons may be efﬁciently dehalo-

genated by titanocene dichloride in the presence of metallic magnesium
(eq. 37) (245). The source of the hydrogen for the reaction appears to be the

Vol. 7

METALLOCENES

Cp rings of the titanocene, by way of the titanocene dimer, which has the
structure 62.

(1) Cp

TiCl

/Mg

(2) H

(37)

Noncatalytic metathesis of the carbenoid group of Cp

Ti CH

with organic

carbonyl oxygen atoms leads to formation of the corresponding exo-methylene
organic molecule. In this way, cyclohexanone is transformed into methylenecyclo-
hexane and benzaldehyde into styrene (246). The oxophilicity of titanium is again
called into service in the conversion of oxiranes, aldehydes, and esters (eq. 38) into
the corresponding alkanes by sodium-reduced titanocene dichloride (247).

TiCl

/Na

58%

70%

(38)

Titanocene has also been reported to reduce alkyl nitriles to alkanes (248).

Catalytic Reactions.

Hydrogenations.

Oleﬁns can be reduced to alkanes with good enantioselec-

tivity with either neutral Ti(III) (249) (eq. 39) or cationic Zr(IV) (250) metallocenes
bearing chiral bis(indenyl) ligands. Unbridged titanocenes and zirconocenes with
bulky chiral substituents attached to the Cp rings asymmetrically hydrogenate
1,1-disubstituted oleﬁns, though with somewhat lower e.e.’s (251).

THF, 2000 psig H

°C, 9 h

5% catalyst

94% yield, >99% e.e.

(39)

METALLOCENES

Vol. 7

Assisted Metallations of Unsaturated Hydrocarbons.

Titanocenes and zir-

conocenes catalyze the addition of M R (M

= B, Mg, Al; R = H, alkyl) bonds across

C C double bonds. The hydroboration of oleﬁns (eq. 40) is believed to involve the
species Cp

Ti(BH

)

(252).

n-C

+ LiBH

(2) NaOMe/H

(1) 5% Cp

TiCl

n-C

95 : 5

83%

(40)

The carbomagnesation of oleﬁns (eq. 41) is selective for terminal vinyl double

bonds (253).

n-C

+ Et

O, 0.5 h

°C

1% Cp

ZrCl

EtMg

n-C

65%

(41)

Alkynes may be converted to the corresponding substituted vinylalanes by

the action of zirconocene dichloride and trimethylaluminum (Eq. 42) (254). The
regioselectivity for Z-addition of the Al R bond is generally

>95%, but drops sig-

niﬁcantly if R is larger than the methyl group.

n-Bu

(1) 1 eq. Cp

ZrCl

/2 eq. AlMe

(2) H

n-Bu

98% Z

(42)

Titanocene diﬂuorides also catalyze the hydrosilation of imines, presumably

through the formation of the trivalent titanocene hydride Cp

Ti H. This reaction

can be made enantioselective, as shown in (equation 43) (255).

+ PhSiH

(1) 0.02% catalyst

°C, 12 h

(2) HCl/Et

95% yield, 99% e.e.

catalyst:

F + PhSiH

(43)

Alkene

and

Alkyne

Dimerization

and

Trimerization.

The

low-

valent mono-Cp zirconium compound CpZrMe(DMPE)

(DMPE

= 1,2-bis(di-

methylphosphino)ethane) catalyzes the dimerization of ethylene to 1-butene
at low frequency (3 t.o./d), a process wherein the 1,3-butadiene complex 64 is
presumed to be the catalyst (256). The clean dimerization of oleﬁns using the
Cp

ZrCl

/MAO (Al/Zr

= 1) catalyst has recently been reported (257). Recently,

Vol. 7

METALLOCENES

the monocyclopentadienyl titanium compound [(

-C

C(CH

)

]TiCl

has

been shown to generate 1-hexene from ethylene with fairly high selectivity (258).

PMe

2 R

2% (C

)

TiCl

/i-PrMgBr

O, 30

°C, 1—3 h

= alkyl, Ph,

OSiMe

, etc.]

−97%

(44)

The highly regioselective head–tail dimerization of alkynes has been accom-

plished using a Grignard-reduced titanocene dichloride (eq. 45) (259).

Metatheses.

Alkene metathesis can be catalyzed by the dinuclear metal-

lacycle Cp

Ti(

µ-CH

)(

µ-Cl)AlR

(Tebbe’s reagent) (260). Grubbs and co-workers

have postulated (261) that the reaction proceeds ﬁrst by the formation of a met-
allacyclobutane, followed by C C bond cleavage to give a carbene–oleﬁn complex
that can exchange oleﬁn through a dissociative process (Eq. 45).

t-Bu

Lewis Base

[Ti]

t-Bu

+ B

AlMe

[Ti]

−RCH

+RCH

(45)

Titanacyclobutanes can serve as catalysts for the living ring-opening

metathesis polymerization (ROMP) of cyclooleﬁns through analogous metathe-
sis chemistry (262) (Eq. 46). The driving force for polymerization is relief of ring
strain.

/octane

−65°C

"Cp

(46)

Oleﬁn metathesis reactions involving Groups V and VIB metallocenes have

also been thoroughly examined (263).

Allyl Coupling Reactions.

Zirconocenes assist in the attack of allylmagne-

sium species on allyl alkyl ethers (eq. 44) and other activated allylic compounds
(264).

METALLOCENES

Vol. 7

= allyl, Bu, Ph]

5 % Cp

ZrCl

O, 5 h, 30

°C

−88%

(47)

Alkylation.

Use of the chiral metallocene rac-Et(Ind)

Zr(CH

)

, activated

with the protic salt [HN(C

)

][B(C

)

], couples 1-hexene and 2-picoline with

some enantioselectivity (eq. 48), but at about one-tenth the turnover frequency of
the analogous Cp

Zr(CH

)

-derived catalyst (265).

n-Bu +

3 % S,S-C

(Ind)

ZrMe

, CH

, 50

°C

~0.1 t.o./h

n-Bu

∗

58% e.e.

(48)

Commercial Overview

Patenting Activity.

A large volume of patents has resulted from the glob-

ally intensive search for proﬁtable application of metallocenes. Starting modestly
in the mid-1980s, the rate of patenting activity rose rapidly through the early
1990s and has only recently begun to stabilize.

Four companies have historically led in the number of awarded patents,

Exxon Chemical, Hoechst Aktiengesellschaft and its successors (Targor, Elenac),
Mitsui Petrochemical, and Dow Chemical, as evinced by Figure 17. Other com-
panies have tried to stake out smaller, more focused domains of intellectual
property, with Fina (syndiotactic polypropylene), Idemitsu Kosan (syndiotactic
polystyrene), and Ticona (cyclooleﬁn copolymers) being notable examples. The vi-
cissitudes of the struggles over patent rights are such that nothing can be said
regarding the ultimate strength of any particular portfolio. However, the search
for freedom to operate in the ﬁeld of metallocenes has been a large factor in deter-
mining the direction and emphasis of research at many companies. The effects of
mergers and patent litigation on the climate for metallocene implementation has
been recently summarized (266).

Polyethylene.

The polyethylene market is rather conservative and

price-sensitive, especially in the commodity grades, so it is not surprising that
metallocene resins, which are more expensive than the Ziegler–Natta materials
they aim to replace, have made only small inroads here. The most promising area
of market penetration is in specialty grades, especially in the very low density
regime (d

< 0.910 g/cc), where the absence of oligomers and high-melting crys-

tallinity make for a clear, ﬂexible product with low tackiness and odor. These
products are well-suited for food packaging and cable insulation, and are begin-
ning to ﬁnd their way into markets hitherto dominated by plasticized PVC. Much
of the emphasis in metallocene research for polyethylene applications has been
directed toward adapting the polymer architecture so that the user may process
the molten resin in existing equipment. This is facilitated by use of catalysts that

Vol. 7

METALLOCENES

600

500

400

300

200

100

1980

1981

1982

1983

1984

1985

1986

1987

1988

1989

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

Fig. 17.

Patenting activity by company in metallocene-based oleﬁn polymerization (ap-

proximately all unique published applications by priority date as of May 2000).

Phillips,

BASF,

Hoechst,

Dow,

Mitsui,

Exxon,

Others.

give long-chain branched PE or by use of mixed catalyst systems to yield a polymer
with broadened MW distribution.

Metallocene catalysts have been successfully adapted to all commercial

polyethylene processes. Recently, companies have begun to join forces to avoid
patent disputes as well as to accelerate the pace of commercialization. Joint ven-
tures between Exxon and Union Carbide (Univation Technologies) and between
Dow and BP were formed to leverage popular process technologies (variants of
ﬂuidized-bed gas-phase reactor systems) to expand the market for metallocene
catalysts and products. The Dow–Union Carbide merger of 2001 could greatly
inﬂuence the rate of technology dissemination.

Polypropylene.

Metallocene iPP, as with metallocene PE, has made the

least headway in the commodity resin market, where the catalyst technology based
on Mg

and selectivity control agents already achieves the high isotacticity

(

>98%) and activity (<2 ppm Ti) metallocene catalysts have struggled to attain.

The areas in which metallocene catalysts have inherent advantages are in (1) non-
woven ﬁber applications, because of the low melt-viscosity associated with narrow
MW distribution resins, and (2) random copolymers with ethylene or butene, be-
cause of the low tendency to form long ethylene blocks. Metallocene iPP is also
beginning to displace polystyrene in injection-molding applications.

Fina and Mitsui Toatsu have combined development efforts in syndiotactic

polypropylene, but it is not clear where this material will have the greatest impact.

Elastomers.

Ethylene–propylene

elastomers

(EPR)

well

ethylene–styrene elastomers (Index) have been commercialized using Dow’s
constrained-geometry technology, with the former material being produced and
marketed out of the joint venture DuPont–Dow Elastomers. The low levels of crys-
tallinity possible with single-site catalysts make the advantages of metallocenes
over Ziegler catalysts much more apparent than in crystalline resins. Market
acceptance of these rubbers, along with the very low-density ethylene–1-octene

METALLOCENES

Vol. 7

copolymers made by Dex Plastomers (Exxon/DSM joint venture) and by Dow
(Afﬁnity grades), has been swift.

Cyclooleﬁn (co)Polymers.

Many metallocenes can incorporate bulky cy-

clooleﬁns into polyethylene, but only a few companies, notably Hoechst and its
successor Ticona, have attempted to sell the resulting materials. These glassy,
amorphous copolymers have high rigidity, chemical resistance, and clarity, which
predisposes them to be useful in optical applications and in pharmaceutical con-
tainers.

Polystyrene.

Dow and Idemitsu have worked together to bring syndio-

tactic polystyrene (Questra) to market, where it has certain advantages over ex-
isting PS products, chieﬂy due to its crystallinity, such as high heat-distortion
temperature.

Because of its popularity among both industrial and academic chemists, met-

allocene technology has been the subject of numerous reviews, some of which
have been cited in earlier sections. An excellent listing of metallocene catalyst
activities sorted by monomer system and catalyst structure has been collected
by Gupta and co-workers (267); the review of M¨ohring and Coville (268) also
stands out for its comparison of metallocene activities, and also thoroughly covers
non-ethylene polymerizations. Reddy and Sivaram have produced an overview
(269) that is especially strong in the area of aluminoxane structure and function.
The constrained-geometry catalyst family has been reviewed by McKnight and
Waymouth (270). A ﬁne overview of the state of metallocenes in industry has been
put together by Maier (271). A very good general reference for metallocene chem-
istry including stoichiometric reactions and Cp-equivalents has recently issued
(272). Useful collections of articles on the subject of metallocene polymerization
occasionally appear in review journals (273).

BIBLIOGRAPHY

1. T. J. Kealy and P. L. Paulson, Nature 168, 1039–1040 (1951).
2. G. Wilkinson, M. Rosenblum, M. C. Whiting, and R. B. Woodward, J. Am. Chem. Soc.

74, 2125–2126 (1952).

3. H. Sinn and W. Kaminsky, Adv. Organomet. Chem. 18, 99–149 (1980).
4. U.S. Pat. 2,827,446 (priority date Sept. 27, 1955), D. S. Breslow (to Hercules Powder).
5. G. Natta, P. Pino, G. Mazzanti, U. Giannini, E. Mantica, and M. Peraldo, Chim. Ind.

39, 19–20 (1957).

6. J. W. Lauher and R. Hoffmann, J. Am. Chem. Soc. 98, 1729–1742 (1976).
7. I. E. Nifant’ev and P. V. Ivchenko, Organometallics 16, 713–715 (1997).
8. G. M. Diamond, S. Rodewald, and R. F. Jordan, Organometallics 14, 5–7 (1995).
9. B. Thiyagarajan and R. F. Jordan, Organometallics 18, 5347–5359 (1999).

10. P. Koepf-Maier, Eur. J. Clin. Pharmacol. 47, 1–16 (1994).
11. E. Zeiger, B. Anderson, S. Haworth, T. Lawlor, and K. Mortelmans, Env. Mol. Mutagen.

11(12), 1–157 (1988).

12. C. J. Harlan, M. R. Mason, and A. R. Barron, Organometallics 13, 2957–2969 (1994);

C. J. Harlan, S. G. Bott, and A. R. Barron, J. Am. Chem. Soc. 117, 6465–6474 (1995).

13. J. L. Atwood and M. J. Zaworotko, J. Chem. Soc., Chem. Commun. 302–303 (1983).
14. M. R. Mason, J. M. Smith, S. G. Bott, and A. R. Barron, J. Am. Chem. Soc. 115,

4971–4984 (1993).

15. C. Sishta, R. M. Hathorn, and T. J. Marks, J. Am. Chem. Soc. 114, 1112–1114 (1992).

Vol. 7

METALLOCENES

16. D. E. Babushkin and H.-H. Brintzinger, J. Am. Chem. Soc. 124, 12869–12873 (2002);

and references therein.

17. U.S. Pat. 5,728,855 (priority date Oct. 19, 1995), G. M. Smith and D. B. Malpass (to

Akzo Nobel NV).

18. Eur. Pat. Appl. 0 468 651 A1 (priority date July 3, 1990), R. E. LaPointe (to Dow).
19. X. Yang, C. L. Stern, and T. J. Marks, J. Am. Chem. Soc. 113, 3623–3625 (1991); X.

Yang, C. L. Stern, and T. J. Marks, J. Am. Chem. Soc. 116, 10015–10031 (1994).

20. Eur. Pat. Appl. 0 277 004 A1 (priority date Jan. 30, 1987), H. W. Turner (to Exxon).
21. U.S. Pat. 5,155,080 (ﬁled Oct. 10, 1989), M. J. Elder, A. Razavi, and J. A. Ewen (to

Fina Technology).

22. M. Bochmann and S. J. Lancaster, J. Organomet. Chem 497, 55–59 (1995).
23. Y.-X. Chen, C. L. Stern, and T. J. Marks J. Am. Chem. Soc. 119, 2582–2583 (1997).
24. L. Jia, X. Yang, C. L. Stern, and T. J. Marks, Organometallics 16, 842–857 (1997).
25. L. Li and T. J. Marks, Organometallics 17, 3996–4003 (1998).
26. G. G. Hlatky, H. W. Turner and R. R. Eckman, J. Am. Chem. Soc. 111, 2729–2731

(1989).

27. A. D. Horton and A. G. Orpen, Organometallics 10, 3910–3918 (1991).
28. M. Bochmann and A. J. Jaggar, J. Organomet. Chem. 424, C5–C7 (1992).
29. M. Brookhart, B. Grant, and A. F. Volpe Jr., Organometallics 11, 3920–3922 (1992).
30. G. G. Haltky, R. R. Eckman, and H. W. Turner, Organometallics 11, 1413–1416 (1992).
31. L. Resconi, S. Bossi, and L. Abis, Macromolecules 23, 4490–4491 (1990); M. Kaminaka

and K. Soga, Polymer 33, 1105–1107 (1992).

32. Eur. Pat. Appl. 0 643 079 A2 (priority date Sept. 13, 1993), M. Galimberti and E.

Albizzati (to Spherilene); World Pat. Appl. 95/27717 A1 (priority date Apr. 6, 1994),
C. E. Bishop, R. L. Jones, K. Raman, V. A. Dang, L. Resconi, T. Dall’Occo, and M.
Galimberti (to Spherilene).

33. T. J. Marks, Acc. Chem. Res. 25, 57–65 (1992).
34. EP 0 683 180 A2 (priority date May 18, 1994), Y. Suga, Y. Maruyama, E. Isobe, Y.

Uehara, Y. Ishihama, and T. Sagae (to Mitsubishi Chemical); EP 0 849 288 A1 (priority
date Dec. 20, 1996), S. Hamura, H. Yasuda, T. Yoshida, and M. Sato (to Tosoh).

35. U.S. Pat. 4,752,597 (priority date Feb. 19, 1987), H. W. Turner (to Exxon).
36. U.S. Pat. 5,756,416 (priority date Feb. 25, 1998), E. P. Wasserman, S. C. Kao, and F.

J. Karol (to Union Carbide).

37. M. Stork, M. Koch, M. Klapper, K. M ¨

ullen, H. Gregorius, and U. Rief, Macromol.

Rapid Commun. 20, 210–213 (1999).

38. U.S. Pat. 5,854,363 (priority date Jan. 8, 1997), M. Jung, H. G. Alt, and B. M. Welch

(to Phillips Petroleum).

39. U.S. Pat. 5,317,036 (priority date Oct. 16, 1992), R. C. Brady, III, F. J. Karol, T. R.

Lynn, R. J. Jorgensen, S.-C. Kao, and E. P. Wasserman (to Union Carbide).

40. U.S. Pat. 5,648,310 (priority date Dec. 23, 1993), E. P. Wasserman, M. W. Smale, T. R.

Lynn, R. C. Brady III, and F. J. Karol (to Union Carbide).

41. U.S. Pat. 4,923,833 (priority date Dec. 27, 1986), M. Kioka, and N. Kashiwa (to Mitsui

Petrochemical).

42. EP 0 447 071 A1 (priority date Mar. 12, 1990), J.-C. A. Bailly, and C. J. Chabrand (to

BP Chemicals).

43. K. Soga, H. J. Kim, and T. Shiono, Macromol. Chem. Phys. 195, 3347–3360 (1994).
44. G. A. Luinstra, J. Organomet. Chem. 517, 209–215 (1996).
45. R. Fandos, M. Lanfranchi, A. Otero, M. A. Pellinghelli, M. J. Ruiz, and P. Terreros,

Organometallics 15, 4725–4730 (1996).

46. J. S. Rogers, G. C. Bazan, and C. K. Sperry, J. Am. Chem. Soc. 119, 9305–9306 (1997).
47. A. J. Ashe III, S. Al-Ahmad, X. Fang, and J. W. Kampf, Organometallics 17, 3883–3888

(1998).

METALLOCENES

Vol. 7

48. J. A. Smith, J. von Seyerl, G. Huttner, and H. H. Brintzinger, J. Organomet. Chem.

173, 175–185 (1979); and Refs. 7 and 8 within.

49. F. Piemontesi, I. Camurati, L. Resconi, D. Balboni, A. Sironi, M. Moret, R. Zeigler,

and N. Piccolrovazzi, Organometallics 14, 1256–1266 (1995).

50. C. Qian, J. Guo, C. Ye, J. Sun, and P. Zheng, J. Chem. Soc., Dalton Trans. 3441–3446

(1993).

51. W. A. Herrmann, M. J. A. Morawietz, and H.-F. Herrmann, F. K ¨

uber, J. Organomet.

Chem. 509, 115–117 (1996).

52. A. J. Ashe III, X. Fang, and J. W. Kampf, Organometallics 18, 2288–2290 (1999); and

references therein.

53. R. B. Grossman, J.-C. Tsai, W. M. Davis, A. Guti´errez, and S. L. Buchwald,

Organometallics 13, 3892–3896 (1994).

54. T. A. Herzog, D. L. Zubris, and J. E. Bercaw, J. Am. Chem. Soc. 118, 11988–11989

(1996).

55. K. A. O. Starzewski, W. M. Kelly, A. Stumpf, and D. Freitag, Angew. Chem. Int. Ed.

38, 2439–2443 (1999).

56. C. Qian, Z. Xie, and Y. Huang, J. Organomet. Chem. 323, 285–294 (1987).
57. R. Leino, H. Luttikhedde, C.-E. Wil´en, R. Sillanp ¨a ¨a, and J. H. N ¨asman,

Organometallics 15, 2450–2453 (1996).

58. H. Plenio and D. Burth, J. Organomet. Chem. 519, 269–272 (1996);

H. J. G. Lut-

tikhedde, R. P. Leino, C.-E. Wil´en, J. H. N ¨asman, M. J. Ahlgr´en, and T. A. Pakkanen,
Organometallics 15, 3092–3094 (1996).

59. C. M ¨

uller, D. Lilge, M. O. Kristen, and P. Jutzi, Angew. Chem. Int. Ed. 39, 789

(2000).

60. L. Atovmyan, S. Mkoyan, I. Urazowski, R. Broussier, S. Ninoreille, P. Perron, and B.

Gautheron, Organometallics 14, 2601–2604 (1995).

61. B. Rieger, J. Organomet. Chem. 428, C33–C36 (1992).
62. J. P. Mitchell, S. Hajela, S. K. Brookhart, K. I. Hardcastle, L. M. Henling, and J. E.

Bercaw, J. Am. Chem. Soc. 118, 1045–1053 (1996).

63. T. Takahashi, K. Kasai, N. Suzuki, K. Nakajima, and E. Negishi, Organometallics 13,

3413–3414 (1994).

64. M. Mitani, K. Oouchi, M. Hayakawa, T. Yamada, and T. Mukaiyama, Macromol. Chem.

Phys. 197, 1815–1822 (1996).

65. S. J ¨

ungling, R. M ¨

ulhaupt, and H. Plenio, J. Organomet. Chem. 460, 191–195 (1993).

66. L. Li, M. V. Metz, H. Li, M.-C. Chen, T. J. Marks, L. Liable-Sands, and A. L. Rheingold,

J. Am. Chem. Soc. 124, 12725–12741 (2002).

67. K. Mach, V. Varga, and H. Antropiusov ´a, J. Organomet. Chem. 333, 205–215 (1987).
68. A. Kucht, H. Kucht, S. Barry, J. C. W. Chien, and M. D. Rausch, Organometallics 12,

3075–3078 (1993).

69. G. Xu and S. Lin, Macromolecules 30, 685–693 (1997).
70. L. R. Sita and J. R. Babcock, Organometallics 17, 5228–5230 (1998); and references

therein.

71. U.S. Pat. 4,870,042 (priority date Oct. 8, 1987), T. Kohara, and S. Ueki (to Toa Nenryo

K.K).

72. S. Doherty, R. J. Errington, A. P. Jarvis, S. Collins, W. Clegg, and M. R. J. Elsegood,

Organometallics 17, 3408–3410 (1998).

73. D. W. Stephan, J. C. Stewart, F. Gu´erin, R. E. v. H. Spence, W. Xu, and D. G. Harrison,

Organometallics 18, 1116–1118 (1999).

74. K. Nomura, N. Naga, M. Miki, K. Yanagi, and A. Imai, Organometallics 17, 2152–2154

(1998).

75. Q. Wang, R. Quoyum, D. J. Gillis, M.-J. Tudoret, D. Jeremic, B. K. Hunter, and M. C.

Baird, Organometallics 15, 693–703 (1996).

Vol. 7

METALLOCENES

76. T. D. Shaffer and J. R. Ashbaugh, J. Polym. Sci, Part A: Polym. Chem. 35, 329–344

(1997).

77. J. L. Petersen, private communication, 1999.
78. P. J. Shapiro, E. Bunel, W. P. Schaefer, and J. E. Bercaw, Organometallics 9, 867–869

(1990).

79. Eur. Pat. Appl. 0 416 815 A (priority date Aug. 31, 1989), J. C. Stevens, G. F. Schmidt,

P. N. Nickias, R. K. Rosen, G. W. Knight, and S. Lai (to Dow Chemical).

80. Eur. Pat. Appl. 0 420 436 A (priority date Sept. 13, 1989), J. A. M. Canich (to Exxon

Chemical).

81. World Pat. Appl. 98/41529 (priority date Mar. 19, 1997), R. E. Spence, X. Gao, L. Koch,

S. J. Brown, and D. G. Harrison (to Nova Chemicals).

82. J. Klosin, W. J. Kruper Jr., P. N. Nickias, G. R. Roof, P. De Waele, and K. A. Abboud,

Organometallics 20, 2663–2665 (2001).

83. K. Kunz, G. Erker, S. D¨oring, S. Bredeau, G. Kehr, R. Fr¨ohlich, Organometallics 21,

1031–1041 (2002).

84. Y.-X. Chen, P.-F. Fu, C. L. Stern, and T. J. Marks, Organometallics 16, 5958–5963

(1997).

85. J. C. Flores, J. C. W. Chien, and M. D. Rausch, Organometallics 13, 4140–4142

(1994).

86. World Pat. Appl. WO96/13529 (priority date Oct. 31, 1994), J. A. M. van Beek, G. H.

J. van Doremaele, G. J. M. Gruter, H. J. Arts, and G. H. M. R. Eggels (to DSM).

87. World Pat. Appl. 96/08497 A1 (priority date Sept. 12, 1994), D. R. Wilson, D. R. Nei-

thamer, P. N. Nickias, and W. J. Kruper Jr. (to Dow Chemical).

88. World Pat. Appl. 97/10248 A1 (priority date Sept. 11, 1995), J. A. Ewen, R. W. Strozier,

R. L. Jones Jr., and M. J. Elder (to Montell).

89. G. Rodriguez and G. Bazan, J. Am. Chem. Soc. 119, 343–352 (1997).
90. D. J. Crowther, N. C. Baenziger, and R. F. Jordan, J. Am. Chem. Soc. 113, 1455–1457

(1991).

91. C. Kreuder, R. F. Jordan, and H. Zhang, Organometallics 14, 2993–3001 (1995).
92. N. S. Hosmane, Y. Wang, H. Zhang, J. A. Maguire, E. Waldh¨or, W. Kaim, H. Binder,

and R. K. Kremer, Organometallics 13, 4156–4158 (1994).

93. R. W. Quan, G. C. Bazan, A. F. Kiely, W. P. Schaefer, and J. E. Bercaw, J. Am. Chem.

Soc. 116, 4489–90 (1994).

94. Eur. Pat. Appl. 0 672 676 A2 (priority date Mar. 15, 1994), F. G. Cloke (to BP Chemi-

cals).

95. P. L. Watson and G. W. Parshall, Acc. Chem. Res. 18, 51–56 (1985).
96. G. Jeske, H. Lauke, H. Mauermann, P. N. Swepston, H. Schumann, and T. J. Marks,

J. Am. Chem. Soc. 107, 8091–8103 (1985).

97. P. J. Shapiro, E. Bunel, W. P. Schaefer, and J. E. Bercaw, Organometallics 9, 867–869

(1990).

98. K. C. Hultzch, T. P. Spaniol, and J. Okuda, Angew. Chem. Int. Ed. 38, 227–230 (1999).
99. G. Jeske, L. E. Schock, P. N. Swepston, H. Schumann, and T. J. Marks, J. Am. Chem.

Soc. 107, 8103–8110 (1985).

100. E. B. Coughlin and J. E. Bercaw, J. Am. Chem. Soc. 114, 7606–7607 (1992).
101. P. J. Toscano and T. J. Marks, J. Am. Chem. Soc. 107, 653–659 (1985); D. Hedden and

T. J. Marks, J. Am. Chem. Soc. 110, 1647–1649 (1988).

102. H. Yasuda and E. Ihara, Bull. Chem. Soc. Jpn. 70, 1745–1767 (1997).
103. H. Yasuda, M. Furo, H. Yamamoto, A. Nakamura, S. Miyake, and N. Kibino, Macro-

molecules 25, 5115–5116 (1992).

104. C.-T. Chen, L. H. Doerrer, V. C. Williams, and M. L. H. Green, J. Chem. Soc., Dalton

Trans. 967–974 (2000).

105. J. M. Decker, S. J. Geib, and T. Y. Meyer, Organometallics 18, 4417–4420 (1999).

METALLOCENES

Vol. 7

106. WO94/01471 (priority date July 1, 1992), D. J. Crowther, R. A. Fisher, J. A. M. Canich,

G. G. Hlatky, and H. W. Turner (to Exxon).

107. K. Mashima, S. Fujikawa, Y. Tanaka, H. Urata, T. Oshiki, E. Tanaka, and A. Naka-

mura, Organometallics 14, 2633–2640 (1995).

108. U.S. Pat. 3,709,853 (priority date Apr. 29, 1971), G. L. Karapinka (to Union Carbide).
109. F. J. Karol, G. L. Karapinka, C. Wu, A. W. Dow, R. N. Johnson, and W. L. Carrick, J.

Polym. Sci., Part A-1 10, 2621–2637 (1972).

110. R. A. Heintz, S. Leelasubcharoen, L. M. Liable-Sands, A. L. Rheingold, and K. H.

Theopold, Organometallics 17, 5477–5485 (1998).

111. Y. Liang, G. P. A. Yap, A. L. Rheingold, and K. H. Theopold, Organometallics 15,

5284–5286 (1996).

112. A. D¨ohring, J. G¨ohre, P. W. Jolly, B. Kryger, J. Rust, and G. P. J. Verhovnik,

Organometallics 19, 388–402 (2000).

113. U.S. Pat. 5,418,200 (priority date Mar. 29, 1991), M. J. Carney and D. L. Beach (to

Chevron).

114. J. S. Rogers, X. Bu, and G. C. Bazan, J. Am. Chem. Soc. 122, 730–731 (2000).
115. W. Kaminsky and H. Hahnsen, Adv. Polyoleﬁns (Proc. Amer. Chem. Soc. Intl. Symp.

1985) 361–371 (1987).

116. K. Heiland and W. Kaminsky, Makromol. Chem. 193, 601–610 (1992).
117. A. Toyota, T. Tsutsui, and N. Kashiwa, J. Mol. Catal. 56, 237–247 (1989).
118. N. Piccolrovazzi, P. Pino, G. Consiglio, A. Sironi, and M. Moret, Organometallics 9,

3098–3105 (1990).

119. I.-M. Lee, W. J. Gauthier, J. M. Ball, B. Iyengar, and S. Collins, Organometallics 11,

2115–2122 (1992).

120. W. Michiels, A. Mu ˜

noz-Escalona, Macromol. Symp. 97, 171–183 (1995).

121. World Pat. Appl. 93/13140 (priority date Dec. 23, 1991), J. A. M. Canich (to Exxon).
122. H. Katayama, H. Shiraishi, T. Hino, T. Ogane, and A. Imai, Macromol. Symp. 97,

109–118 (1995).

123. U.S. Pat. 5,474,961 (priority date Sept. 12, 1991), R. Schlund, M. Kersting, and K.-D.

Hungenberg (to BASF).

124. World Pat. Appl. 95/13385 (priority date Nov. 8, 1993), S. Cheruvu, F. Y.-K. Lo and S.

C. Ong (to Mobil).

125. M. Bochmann, J. Chem. Soc., Dalton Trans. 255–270 (1996).
126. D. E. Richardson, N. G. Alameddin, M. F. Ryan, T. Hayes, J. R. Eyler, and A. R. Siedle,

J. Am. Chem. Soc. 118, 11244–11253 (1996).

127. J. Karl, G. Erker, and R. Fr¨ohlich, J. Organomet. Chem. 535, 59–62 (1997).
128. T. K. Woo, L. Fan, and T. Ziegler, Organometallics 13, 432–433 (1994).
129. E. J. Arlman and P. Cossee, J. Catal. 3, 99–104 (1964).
130. C. Janiak, J. Organomet. Chem. 452, 63–73 (1993).
131. M. Brookhart, M. L. H. Green, and L.-L. Wong, Prog. Inorg. Chem. 36, 1–124 (1988).

W. Piers and J. E. Bercaw, J. Am. Chem. Soc. 112, 9406–9407 (1990).

132. S. L. Borkowsky, N. C. Baenziger, and R. F. Jordan, Organometallics 12, 486–495

(1993).

133. Y. W. Alelyunas, Z. Guo, R. E. LaPointe, and R. F. Jordan, Organometallics 12, 544–553

(1993).

134. Z. Guo, D. C. Swenson, and R. F. Jordan, Organometallics 13, 142–1432 (1994).
135. T. K. Han, Y. S. Ko, J. W. Park, and S. I. Woo, Macromolecules 29, 7305–7309.
136. J. C. Flores, J. S. Wood, J. C. W. Chien, and M. D. Rausch, Organometallics 15,

4944–4950 (1996).

137. I. Tritto, R. Donetti, M. C. Sacchi, P. Locatelli, and G. Zannoni, Macromolecules 30,

1247–1252 (1997).

138. J. Koivum ¨aki, Acta Polytechn. Scand. 227, 1–50 (1995).

Vol. 7

METALLOCENES

139. J. A. Parker, D. C. Bassett, R. H. Olley, and P. Jaaskelainen, Polymer 35, 4140–4145

(1994).

140. N. Herfert, P. Montag, and G. Fink, Makromol. Chem. 194, 3167–3182 (1993).
141. J. A. Ewen, in T. Keii, K. Soga, eds., Catalytic Polymerization of Oleﬁns, Kodansha,

Tokyo, 1986, p. 271.

142. J. C. W. Chien and D. He, J. Polym. Sci., Part A: Polym. Chem. 29, 1585–1593 (1991).
143. M. Galimberti, F. Piemontesi, O. Fusco, and I. Camurati, Macromolecules 31,

3409–3416 (1998).

144. F. J. Karol, S.-C. Kao, E. P. Wasserman, and R. C. Brady, New J. Chem. 21, 797–805

(1997).

145. C. Lehtinen, P. Stark, and B. L¨ofgren, J. Polym. Sci, Part A: Polym. Chem. 35, 307–318

(1997).

146. C. Lehtinen and B. L¨ofgren, Eur. Polym. J. 33, 115–120 (1997).
147. H. Dr¨ogem ¨

uller, K. Heiland, and W. Kaminsky, in W. Kaminsky and H. Sinn,

eds., Transition Metals and Organometallics as Catalysts for Oleﬁn Polymerization,
Springer-Verlag, Berlin, 1998, p. 303.

148. M. Galimberti, N. Mascellani, and F. Piemontesi, I Camurati, Macromol. Rapid Com-

mun. 20, 214–218 (1999).

149. A. Zambelli, A. Grassi, M. Galimberti, R. Mazzocchi, and F. Piemontesi, Makromol.

Chem., Rapid Commun. 12, 523–528 (1991).

150. J. V. Sepp ¨al ¨a, J. Koivum ¨aki, and X. Liu, J. Polym. Sci., Part A, Polym. Chem. 31,

3447–3452 (1993).

151. P. Burger, K. Hortmann, and H.-H. Brintzinger, Makromol. Chem., Macromol. Symp.

66, 127–140 (1993).

152. M. J. Schneider, J. Suhm, R. M ¨

ulhaupt, M.-H. Prosenc, and H.-H. Brintzinger, Macro-

molecules 30, 3164–3168 (1997).

153. K. Soga and D. H. Lee, Makromol. Chem. 193, 1687–1694 (1992).
154. R. Blom and I. M. Dahl, Macromol. Chem. Phys. 200, 442–449 (1999).
155. T. K. Woo, P. M. Margl, T. Ziegler, and P. E. Bl¨ochl, Organometallics 16, 3454–3468

(1997).

156. J. B. P. Soares, J. D. Kim, and G. L. Rempel, Ind. Eng. Chem. Res. 36, 1144–1150

(1997).

157. L. Izzo, L. Caporaso, G. Senatore, and L. Oliva, Macromolecules 32, 6913–6916

(1999).

158. L. Resconi, F. Piemontesi, G. Franciscono, L. Abis, and T. Fiorani, J. Am. Chem. Soc.

114, 1025–1032 (1992); see also refs. in footnotes 13–16 of this citation.

159. K. Thorshaug, J. A. Støvneng, E. Rytter, and M. Ystenes, Macromolecules 31,

7149–7165 (1998).

160. A. Carvill, I. Tritto, P. Locatelli, and M. C. Sacchi, Macromolecules 30, 7056–7062

(1997).

161. K. Koo, and T. J. Marks, J. Am. Chem. Soc. 121, 8791–8802 (1999).
162. G. Xu, and T. C. Chung, J. Am. Chem. Soc. 121, 6763–6764 (1999).
163. H. Hagihara, T. Shiono, and T. Ikeda, Macromolecules 31, 3184–3188 (1998);

Fukui, M. Murata, and K. Soga, Macromol. Rapid Commun. 20, 637–640 (1999).

164. M. K. Reinking, G. Orf, and D. McFaddin, J. Polym. Sci., Part A: Polym. Chem. 36,

2889–2898 (1998).

165. A. D. Horton, Organometallics 15, 2675–2677 (1996).
166. C. S. Christ Jr., J. R. Eyler, and D. E. Richardson, J. Am. Chem. Soc. 112, 596–607

(1990).

167. E. P. Wasserman, E. Hsi, and W.-T. Young, ACS Div. Polym. Chem., Polym. Prepr 39(2),

425–426 (1998);

S. Lieber, M.-H. Prosenc, and H.-H. Brintzinger, Organometallics

19, 377–387 (2000).

100

METALLOCENES

Vol. 7

168. V. Busico, R. Cipullo, J. C. Chadwick, J. F. Modder, and O. Sudmeijer, Macromolecules

27, 7538–7543 (1994).

169. K. Thorshaug, J. A. Støvneng, E. Rytter, and M. Ystenes, Macromolecules 31,

7149–7165 (1998).

170. V. Busico, and R. Cipullo, J. Am. Chem. Soc. 116, 9329–9330 (1994); J. C. Yoder, and

J. E. Bercaw, J. Am. Chem. Soc. 124, 2548–2555.

171. F. R. W. P. Wild, L. Zsolnai, G. Huttner, and H. H. Brintzinger, J. Organomet. Chem.

232, 233–247 (1982); W. Kaminsky, A.-M. Schauwienold, and F. Freidanck, J. Mol.
Catal. A: Chem. 112, 37–42 (1996).

172. J. C. Yoder, M. W. Day, and J. E. Bercaw, Organometallics 17, 4946–4958 (1998).
173. J. A. Ewen, J. Am. Chem. Soc. 106, 6355–6364 (1984).
174. U.S. Pat. 4,769,510 (priority date Nov. 27, 1984), W. Kaminsky, K. K ¨

ulper, M.

Buscherm¨ohle, and H. L ¨

uker (to Hoechst)

175. W. Kaminsky, Angew. Makromol. Chem. 145/146, 149–160 (1986).
176. G. Guerra, L. Cavallo, G. Moscardi, M. Vacatello, and P. Corradini, J. Am. Chem. Soc.

116, 2988–2995 (1994).

177. W. R¨oll, H.-H. Brintzinger, B. Rieger, and R. Zolk, Angew. Chem., Int. Ed. Engl. 29,

279–280 (1990).

178. H. Kawamura-Kuribayashi, N. Koga, and K. Morokuma, J. Am. Chem. Soc. 114,

8687–8694.

179. F. R. W. P. Wild, J. Zsolnai, G. Huttner, and H.-H. Brintzinger, J. Organomet. Chem.

232, 233–247 (1982);

S. Collins, B. A. Kuntz, and Y. Hong, J. Org. Chem. 54,

4154–4158 (1989); B. Chin, and S. L. Buchwald, J. Org. Chem. 61, 5650–5651 (1996).

180. G. Jany, R. Fawzi, M. Steimann, and B. Rieger, Organometallics 16, 544–550 (1997).
181. T. Mise, S. Miya, and H. Yamazaki, Chem. Lett. 1853–1856 (1989).
182. W. Spaleck, F. K ¨

uber, A. Winter, J. Rohrmann, B. Bachmann, M. Antberg, V. Dolle,

and E. F. Paulus, Organometallics 13, 954–963 (1994).

183. Eur. Pat. Appl. 0 962 462 A2 (priority date June 1, 1998), M. D. Rausch, E. J. Thomas,

S. Bettonville, and D. Grandﬁls (to Solvay).

184. S. Miyake, Y. Okumura, and S. Inazawa, Macromolecules 28, 3074–3079 (1995).
185. L. Resconi, D. Balboni, G. Baruzzi, C. Fiori, S. Guidotti, P. Mercandelli, and A. Sironi,

Organometallics 19, 420–429 (2000).

186. L. Resconi, F. Piemontesi, I. Camurati, D. Balboni, A. Sirone, M. Moret, H. Rychlicki,

and R. Zeigler, Organometallics 15, 5046–5059 (1996).

187. E. Shamshoum and D. Rauscher, MetCon ’93, Proceedings of meeting, May 26–28

1993, Houston, Texas (Catalyst Consultants), 173–187;

188. T. Shiomura, SPO ’96, Proceedings of Sixth Intl. Bus. Forum on Specialty Polyoleﬁns,

Sept. 25–27 1996, Houston, Texas (Schotland), 53–67.

189. J. A. Ewen, R. L. Jones, A. Razavi, and J. D. Ferrara, J. Am. Chem. Soc. 110, 6255–6256

(1988).

190. A. Razavi, L. Peters, and L. Nafpliotis, J. Mol. Catal. A: Chem. 115, 129–154 (1997).
191. U.S. Pat. 4,892,851 (priority date July 15, 1988), J. A. Ewen and A. Razavi (to Fina

Technology).

192. Eur. Pat. 0 387 690 B1 (priority date Mar 11, 1989), A. Winter, J. Rohrmann, M.

Antberg, V. Dolle, and W. Spaleck (to Hoechst).

193. G. W. Coates and R. M. Waymouth, Science (Washington, D.C.) 267, 217–219 (1995).
194. V. Busico, R. Cipullo, W. Kretschmer, G. Talarico, M. Vacatello, and V. V. A. Castelli,

Macromol. Symp. 189, 127–141 (2002).

195. D. T. Mallin, M. D. Rausch, Y.-G. Lin, S. Dong, and J. C. W. Chien, J. Am. Chem. Soc.

112, 2030–2031 (1990).

196. A. M. Bravakis, L. E. Bailey, M. Pigeon, and S. Collins, Macromolecules 31, 1000–1009

(1998).

Vol. 7

METALLOCENES

101

197. U.S. Pat. 5,036,034 (priority date Oct. 10, 1989), J. A. Ewen (to Fina Technology).
198. W. Fan, M. K. Leclerc and R. M. Waymouth, J. Am. Chem. Soc. 123, 9555–9563 (2001).
199. World Pat. Appl. 88/04672 (priority date Dec. 19, 1986), H. C. Welborn Jr. (to Exxon).
200. W. Kaminsky and M. Schlobohm, Makromol. Chem., Macromol. Symp. 4, 103–118

(1986).

201. C. Pellecchia, Metallocenes ’95, Proceedings of Intl. Congress on Metallocene Polymers,

Apr. 26–27 1995, Brussels, Belgium (Schotland), 75–85; and references therein.

202. M. F. Llauro, C. Monnet, F. Barbotin, V. Monteil, R. Spitz, and C. Boisson, Macro-

molecules 34 2001, 6304–6311.

203. S. Pragliola, G. Milano, G. Guerra, and P. Longo, J. Am. Chem. Soc. 124, 3502–3503

(2002).

204. S. Ikai, J. Yamashita, Y. Kai, M. Murakami, T. Yano, Y. Qian, and J. Huang, J. Mol.

Catal. A: Chem. 140, 115–119 (1999).

205. World Pat. Appl. 88/04673 (priority date Dec. 19, 1986), R. G. Austin and H. C. Welborn

Jr., (to Exxon).

206. G. W. Coates and R. M. Waymouth, J. Am. Chem. Soc. 115, 91–98 (1993).
207. Chem. Week 159(19), 15 (1997).
208. Chem. Market. Rep. 249(26), 3, 10 (1996).
209. W. Kaminsky and R. Spiehl, Makromol. Chem. 190, 515–526 (1989).
210. S. Collins and W. M. Kelly, Macromolecules 25, 233–237 (1992).
211. M. R. Kesti, G. W. Coates, and R. M. Waymouth, J. Am. Chem. Soc. 114, 9679–9680

(1992).

212. U. M. Stehling, K. M. Stein, M. R. Kesti, and R. M. Waymouth, Macromolecules 31,

2019–2027 (1998).

213. D.-J. Byun, K.-Y. Choi, and S. Y. Kim, Macromol. Chem. Phys. 202, 992–997 (2001).
214. C.-E. Wil´en, M. Auer, J. Strand´en, J. H. N ¨asman, B. Rotzinger, A. Steinmann, R. E.

King III, H. Zweifel, and R. Drewes, Macromolecules 33, 5011–5026 (2000).

215. T. D. Shaffer, J.-A. M. Canich, and K. R. Squire, Macromolecules 31, 5145–5147 (1998).
216. L. Jia, X. Yang, A. M. Seyam, I. D. L. Albert, P.-F. Fu, S. Yang, and T. J. Marks, J. Am.

Chem. Soc. 118, 7900–7913 (1996).

217. C. Pellecchia and L. Oliva, Rubber Chem. Technol. 72, 553–558 (1999).
218. H. Chen, M. J. Guest, S. Chum, A. Hiltner, and E. Baer, J. Appl. Polym. Sci. 70,

109–119 (1998).

219. R. W. Barnhart and G. C. Bazan, J. Am. Chem. Soc. 120, 1082–1083 (1998).
220. C. Denger, U. Haase, and G. Fink, Makromol. Chem., Rapid Commun. 12, 697–701

(1991).

221. N. Ishihara, T. Seimiya, M. Kuramoto, and M. Uoi, Macromolecules 19, 2465–2466

(1986).

222. N. Tomotsu, N. Ishihara, T. H. Mewman, and M. T. Malanga, J. Mol. Catal. A: Chem.

128, 167–190 (1998).

223. A. Grassi, A. Zambelli, and F. Laschi, Organometallics 15, 480–482 (1996);

T. H.

Newman and M. T. Malanga, J. Macromol. Sci., A: Pure Appl. Chem. 34, 1921–1927
(1997).

224. Y. Kim, E. Hong, M. H. Lee, J. Kim, Y. Han, and Y. Do, Organometallics 18, 36–39

(1999).

225. G. Xu, Macromolecules 31, 586–591 (1998).
226. H. Yasuda, H. Yamamoto, K. Yokota, S. Miyake and A. Nakamura, J. Am. Chem. Soc.

114, 4908–4910 (1992).

227. H. Deng, T. Shiono, and K. Soga, Macromolecules 28, 3067–3073 (1995); S. Collins,

D. G. Ward, and K. H. Suddaby, Macromolecules 27, 7222–7224 (1994).

228. H. Nguyen, A. P. Jarvis, M. J. G. Lesley, W. M. Kelly, S. S. Reddy, N. J. Taylor, and S.

Collins, Macromolecules 33, 1508–1510 (2000).

102

METALLOCENES

Vol. 7

229. M. Yamashita, Y. Takemoto, E. Ihara, and H. Yasuda, Macromolecules 29, 1798–1806

(1996).

230. W. J. Evans and H. Katsumata, Macromolecules 27, 2330–2332 (1994).
231. W. J. Evans and H. Katsumata, Macromolecules 27, 4011–4013 (1994).
232. C. Aitken, J. F. Harrod, and E. Samuel, J. Organomet. Chem. 279, C11–C13 (1985).
233. H.-G. Woo, J. F. Walzer, and T. D. Tilley, J. Am. Chem. Soc. 114, 7047–7055 (1992).
234. V. K. Dioumaev and J. F. Harrod, Organometallics 16, 2798–2807 (1997).
235. T. C. Chung, Prog. Polym. Sci. 27, 39–85 (2002).
236. T. E. Patten and B. M. Novak, Makromol. Chem., Macromol. Symp. 67, 203–211 (1993).
237. J. Ren, J. Hu, and Q. Shen, Yingyong Huaxue (Chin. J. Appl. Chem.) 12, 105–106

(1995). [Chem. Abs. 123:170382].

238. J. Schwartz and J. A. Labinger, Angew. Chem. Int. Ed. Engl. 15, 333–340 (1976).
239. D. B. Carr and J. Schwartz, J. Am. Chem. Soc. 101, 3521–3531 (1979).
240. S. L. Buchwald, B. T. Watson, and J. C. Huffman, J. Am. Chem. Soc. 109, 2544–2546

(1987).

241. D. R. Swanson, C. J. Rousset, and E. Negishi, J. Org. Chem. 54, 3521–3523 (1989).
242. D. R. Swanson, C. J. Rousset, E. Negishi, T. Takahashi, T. Seki, M. Saburi, and Y.

Uchida, J. Org. Chem. 54, 3521–3523 (1989).

243. S. L. Buchwald and R. B. Nielsen, Chem. Rev. 88, 1047–1058 (1988).
244. P. J. Fagan, W. A. Nugent, and J. C. Calabrese, J. Am. Chem. Soc. 116, 1880–1889

(1994).

245. T. R. Nelsen, and J. J. Tufariello, J. Org. Chem. 40, 3159–3160 (1975).
246. F. N. Tebbe, G. W. Parshall, and G. S. Reddy, J. Am. Chem. Soc. 100, 3611–3613 (1978).
247. E. E. van Tamelen, and J. A. Gladysz, J. Am. Chem. Soc. 96, 5290–5291 (1974).
248. E. E. van Tamelen, H. Rudler, and C. Bjorklund, J. Am. Chem. Soc. 93, 7113–7114

(1971).

249. R. D. Broene, and S. L. Buchwald, J. Am. Chem. Soc. 115, 12569–12570 (1993). [Er-

ratum: J. Am. Chem. Soc. 118, 9458 (1996)].

250. M. V. Troutman, D. H. Appella, and S. L. Buchwald, J. Am. Chem. Soc. 121, 4916–4917

(1999).

251. L. A. Paquette, M. R. Sivik, E. I. Bzowej, and K. J. Stanton, Organometallics 14,

4865–4878 (1985).

252. K. Isagawa, H. Sano, M. Hattori, and Y. Otsuji, Chem. Lett, 1069 (1979).
253. U. M. Dzhemilev, and O. S. Vostrikova, J. Organomet. Chem. 285, 43–51 (1985).
254. D. E. van Horn, and E. Negishi, J. Am. Chem. Soc. 100, 2252–2254 (1978).
255. X. Verdaguer, U. E. W. Lange, M. T. Reding, and S. L. Buchwald, J. Am. Chem. Soc.

118, 6784–6785 (1996).

256. Y. Wielstra, S. Gambarotta, and M. Y. Chiang, Organometallics 7, 1866–1867 (1988).
257. J. Christoffers and R. G. Bergman, Inorg. Chim. Acta 270, 20–27 (1998).
258. P. J. W. Deckers, B. Hessen, and J. H. Teuben, Angew. Chem. Int. Ed. 40, 2516–2519

(2001).

259. M. Akita, H. Yasuda, and A. Nakamura, Bull. Chem. Soc. Jpn. 57, 480–487 (1984).
260. F. N. Tebbe, G. W. Parshall, and D. W. Ovenall, J. Am. Chem. Soc. 101, 5074–5075

(1979).

261. J. B. Lee, K. C. Ott, and R. H. Grubbs, J. Am. Chem. Soc. 104, 7491–7496 (1982).
262. L. R. Gilliom and R. H. Grubbs, J. Am. Chem. Soc. 108, 733–742 (1986).
263. R. H. Grubbs, in G. Wilkinson, F. G. A. Stone, E. W. Abel, eds., Comprehensive

Organometallic Chemistry, Vol. 8, Pergamon Press, Oxford, 1982, pp. 499–551.

264. A. G. Ibragimov, E. V. Gribanova, L. M. Khalilov, L. M. Zelenova, and U. M. Dzhemilev,

Zh. Org. Khim. 21, 259–265 (1985).

265. S. Rodewald and R. F. Jordan, J. Am. Chem. Soc. 116, 4491–4492 (1994).
266. Chem. Week 162(6), 35–37 (2000).

Vol. 7

MICROCELLULAR PLASTICS

103

267. V. K. Gupta, S. Satish, and I. S. Bhardwaj, J. Macromol. Sci., Rev. Macromol. Chem.

Phys. C34, 439–514 (1994).

268. P. C. M¨ohring and N. J. Coville, J. Organomet. Chem. 479, 1–29 (1994).
269. S. S. Reddy and S. Sivaram, Prog. Polym. Sci. 20, 309–367 (1995).
270. A. L. McKnight and R. M. Waymouth, Chem. Rev. 98, 2587–2598 (1998).
271. R.-D. Maier, Kunststoffe 89, 120–132 (Eng. transl. 45–52) (1999).
272. A. Togni and R. L. Halterman, eds., Metallocenes: Synthesis Reactivity Applications,

Wiley-VCH, Weinheim, 1998.

273. J. A. Gladysz, ed., Chem. Rev. 100, 1164–1682 (200).

RIC

AUL

ASSERMAN

Dow Chemical Company

METHACRYLIC ESTER POLYMERS.

See Volume 3.

Wyszukiwarka

Podobne podstrony:

więcej podobnych podstron