5
8
N M Bravaya, P M Nedorezova, V I Tsvetkova
2
0 ± 25) barely affects the activity of the system and the properties
g/K = 0.6. The D parameter varies in the 1 ± 10 range as the
of the elSBPP formed.7
analogue of the complex 16 with the methyl substituent at the
-position give rise to an amorphous low-molecular-mass PP with
very high polydispersity coefficients, M : M
= 4 ± 15.84 The
2, 78
Both rac- and meso-forms of the
propylene concentration changes from 1.2 to 11 mol litre
71
.
However, the authors of the model admit 80 that the same kinetic
dependence of the pentad composition on the monomer concen-
tration can be described qualitatively and quantitatively by the
1
w
n
Busico model,6
3, 64, 91
which takes into account the competition
elSBPP formed under the action of the complex 30 is close in
stereochemical composition to the polymer formed with the
between the chain growth and epimerisation processes with the
assumption that the AC have two active vacancies accessible for
monomer coordination and the polymer chain growth.
1
6 ± MAO catalyst system; however, regarding the melting point
and the macrotacticity index, it is close to the elSBPP prepared on
catalysts with C
symmetry.71 Complexes 31 ± 35 with mixed
1
Since the model of the `oscillating' complex is widely recog-
nised, several factors deserve attention, both those confirming the
proposed mechanism and those conflicting with it. The mecha-
nism in question is supported by the increase in the content of the
mmmm pentads with an increase in the monomer concentration,
which was observed experimentally for these complexes. The
mathematical model describing this mechanism provides a quan-
titative correspondence between the polymer microstructures
calculated and found experimentally depending on the polymer-
isation conditions (monomer concentration, temperatures) for
reliable values of model parameters. The X-ray diffraction data
ligands and different substituents are less active catalysts than
the bisindenyl complexes with identical ligands.85
The main channel of chain transfer by the catalyst systems
considered is the transfer of the b-hydrogen atom to the metal and
the coordinated monomer.80 An interesting feature of 2-arylin-
denyl systems is the clear-cut effect of the co-monomer, which was
identified in a study of propylene copolymerisation with ethylene
induced by the complexes 16 and 18.7
9, 86
The activity of the
catalyst system increases 2- to 8-fold even at a low ethylene : pro-
pylene molar ratio (0.05 ± 0.08) in the monomer mixture. The
increase in the activity is accompanied by a substantial (2 ± 3-fold)
increase in the molecular mass of the copolymer. The ethylene
content in the copolymer varies in the range of 15 mol % ±
5
the introduction of hydrogen. It has been assumed
co-monomer effect shows itself as the ability of ethylene to be
inserted and to `wake up' the AC after the formation of a
regiodefect of the polymer chain as a result of the monomer
for the complexes 16, 19 are also a forcible argument in favour of
the `oscillating' complex. Theoretical studies 9
2, 93
demonstrated
the presence of two energy minima for the transition state
corresponding to the rac- and meso-forms of the complex 16 and
stabilised by the p-interactions between the phenyl substituents
and the aromatic system of the indenyl ligands (p-stacking
interactions); the two minima are energetically equivalent
(DDH = 0.6 kcal mol7 ). Indeed, on passing to hydrogenated
0 mol %. The system activity also increases appreciably upon
that the
86, 87
1
2
,1-addition.
Studies of the complexes 16 and 18 showed 87 that they are
analogues of the complex 16, (2-CyInd)
(2-Ph-H Ind)(2-PhInd)ZrCl and (2-Cy-H
2
ZrCl
2
(Cy is cyclohexyl),
Ind) ZrCl , all other
4
2
4
2
2
more regiospecific than the isospecific EtInd
indenyl). The non-bridged complex 16 produces PP containing
.1 mol % ± 0.3 mol % of 2,1-added units. It is of interest
2
ZrCl
2
catalyst (Ind is
factors being the same, a substantial decrease in the PP isotacticity
has been noted: [mmmm] = 32%, 15%, 13% and 5%, respec-
tively.77
0
that `regioerrors' are detected only in isotactic sequences of
the polymer, i.e., they arise only in those cases where the `oscil-
lating' metallocene occurs in the isospecific state. This is
confirmed by the fact that no `errors' are found in the PP
synthesised by the polymerisation of propylene in the presence of
Direct investigation of isomerisation processes for `oscillating'
catalysts is of interest. The data of dynamic NMR for
(2-PhInd)
cyclopenta[l ]phenanthryl)
frequency of rotation of hapto-bonded ligands: 6800 s for the
2
HfBn
2
(see Ref. 73) and structurally related (2-phenyl-
2
ZrCl (see Ref. 94) indicate a very high
2
7
1
8
71
meso-Me
2
Si(2-PhInd)
complex produces PP with `errors' caused by the 3,1-addition.
2
ZrCl
2
, whereas the rac-analogue of this
former complex and *1610
s
for the latter one. The latter
88
zirconocene activated by MAO ensures (although with a low
activity) the formation of a low-molecular-mass PP similar in
stereoisomeric composition to the elSBPP prepared under the
action of the complex 16. The frequencies of ligand rotation about
the axis passing through the M atom and the centroid of the Cp
ring are several orders of magnitude greater than the frequency of
olefin insertion estimated from the maximum rate of monomer
A specific feature of the `oscillating' catalysts is an increase in
the content of the mmmm pentads following an increase in the
monomer concentration (see, for example, Table 3, rows 15 ± 17).
Quantitative analysis of this phenomenon is based on the fact that
the rate of polymerisation is proportional to the monomer
concentration, while the rate of isomerisation does not depend
on it and, since the content of isotactic pentads is determined by
the ratio of the rates of these reactions, the regularity of the
3 6
absorption, which equals 0.1 ± 10.0 insertions of C H per second
even with allowance for the low efficiency of the formation of AC.
Evidently, such a fast rotation should average the stereospecific
action of the AC and ensure the formation of PP with a narrow
MMD. No data on the rate of isomerisation in real catalyst
systems (MAO-activated metallocenes) are available. To explain
the observed contradiction, it has been suggested 94 that the chain
growth actually proceeds at a very high rate, which exceeds the
rate of rotation of the hapto-bonded ligands but during the period
of growth of the macromolecules, the AC pass many times into the
`dormant' state, perhaps, due to the formation of a contact ion
pair with the counter-ion.
polymer increases with an increase in the monomer concentration.
This has been observed in numerous experiments.6
7 ± 69, 73, 78
As
discussed above, an opposite dependence of the content of
isotactic pentads on the monomer concentration was found for
1
metallocenes with C symmetry.
A mathematical model has been proposed to describe the
oscillation mechanism of polymer chain growth.80, 89, 90 Computer
generation of PP macromolecules has been carried out with
variation of three parameters, namely, the stereoselectivity of the
isospecific active centre (a), the selectivity of the aspecific centre
(
b) and the g/K parameter, which specifies the ratio of the relative
reactivity to the stability for iso- and aspecific states (g = kpa/kpi
K = k /k ). Thus g/K > 1 corresponds to the predominant exis-
Note that only for the complexes 16 and 19, was the existence
of two energetically equivalent stereoisomers identified. Data
from X-ray diffraction analyses of a number of other complexes
which catalyse the formation of stereoblock PP indicate the
presence of only one stereoisomer, for example, the meso-form
,
1
2
tence of the isospecific AC. The dependence on the pentad
composition on the monomer concentration for definite a, b and
g/K values is determined by the parameter
of the complex 17,6 the rac-form of the complex 18
8
68, 73
and its
hafnium analogue 73 and the meso-form of the complex 21,
which does not differ from the complex 16 in catalytic properties.
Although 13C NMR spectroscopy is a unique tool for ana-
lysing the structure of the polymer chain and, correspondingly, the
catalytic properties of the AC, which allows one to evaluate its
stereo- and regiospecificity, the stereocontrol mechanisms and so
78
ꢀ
k =k k =k C H
pi
1
pa
2
3
6
D =
.
2
The best agreement between experimental and calculated
pentad compositions was found for a = 0.97, b = 0.56 and