which means that they contain a long spacer of methylene carbons between the
polymerizable double bond and the polar group. The spacer fades out the direct
electronic effect of the polar group on the double bond and makes chelate formation
after comonomer insertion improbable. Additionally, the effect of the polar groups
on the electron density of the double bond is weak over the spacer of many atoms.
Hence, the reactions of the polar group and the double bond can be considered as
competing, but fairly independent of each other. On this basis it was anticipated that
random-type copolymerization of these comonomers with ethylene and propylene
would be possible.
4.1.1 Copolymerization Behavior of Oxygen-Functional Alkenes
Several alcohol-, ether-, acid-, ester- and ketone-functional alkenes (Fig. 12) were
tested as comonomers in polymerization experiments [19, 21] A bridged zirconocene
complex rac-Et(Ind) 2 ZrCl 2 was selected as catalyst for the studies because it is a
relatively good copolymerization catalyst and capable of both ethylene and propylene polymerizations. MAO was used as cocatalyst. MAO and the comonomers were
pre-contacted for 15 min in the reactor just before the start of the polymerization.
The catalyst activity dropped significantly in the presence of every one of the
oxygen-functional comonomers. As depicted in Fig. 13, of the different alcohols the
more shielded ones (6, 8) induced the least decrease in activity in propylene
copolymerizations. The length of the spacer between the alcohol group and the
double bond had no effect on the decline in activity. Comparison of the alcohol- and
ether-functional comonomers (4 versus 12 and 13) in ethylene copolymerization
showed that the catalyst activity was on the same level for all three comonomers.
This means that methyl and trimethylsilyl groups are not good protecting groups for
oxygen atoms. Among the carbonyl-containing comonomers, the carboxylic acid
15 and the more shielded ester 17 performed best in terms of catalyst activity,
whereas the ketone comonomer 18 almost completely killed the catalyst.
In the studies of copolymerizations with propylene [19], it was found that a
longer spacer favors the copolymerizability of the oxygen-functional comonomer.
Double the amount of 10-undecen-1-ol 4 was incorporated into the polymer chain
compared with the amount of 5-hexen-1-ol 5. The comonomer 6 with methyl
branches at α-position to the vinyl group did not copolymerize at all. The
reactivities of carbonyl-functional comonomers 15 and 17 were about the same or
slightly higher than the reactivity of 10-undecen-1-ol. A similar trend was found in
the ethylene copolymerizations, where ether comonomers were incorporated to
about the same degree as 10-undecen-1-ol (Fig. 14) [21].
From the different reactions of those comonomers with MAO known from the
NMR studies [140, 141], it can be concluded that the formation of aluminum
alkoxides is not crucial for comonomer incorporation. However, nonfunctional
1-undecene exhibited the highest relative reactivity. This could be inferred to derive
from the smaller size of the comonomer, which does not react with aluminumcontaining cocatalyst species.
212
J. Seppa ¨la ¨ et al.
Précédent

- 219/371

Suivant