See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/225099602

Kinetics of propylene and ethylene polymerization reactions with

heterogeneous ziegler-natta catalysts: Recent results

Article  in  Polymer Science Series A · January 2008

DOI: 10.1134/S0965545X08110023

CITATIONS READS

32 3,107

2 authors:

Yury V. Kissin Laura A. Rishina

Rutgers, The State University of New Jersey Russian Academy of Sciences
205 PUBLICATIONS   4,281 CITATIONS    57 PUBLICATIONS   575 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

New catalysts for ethylene and propylene polymerization View project

All content following this page was uploaded by Yury V. Kissin on 29 May 2014.

The user has requested enhancement of the downloaded file.

ISSN 0965-545X, Polymer Science, Ser. A, 2008, Vol. 50, No. 11, pp. 1101–1121. © Pleiades Publishing, Ltd., 2008.
Published in Russian in Vysokomolekulyarnye Soedineniya, Ser. A, 2008, Vol. 50, No. 11, pp. 1911–1934.

Kinetics of Propylene and Ethylene Polymerization Reactions

with Heterogeneous Ziegler–Natta Catalysts: Recent Results1
Y. V. Kissina and L. A. Rishinab
a Rutgers, The State University of New Jersey, Department of Chemistry and Chemical Biology,
610 Taylor Rd., Piscataway, NJ 08854-8087 USA
b N.N. Semenov Institute of Chemical Physics, Russian Academy of Sciences,
ul. Kosygina 4, Moscow, 119991 Russia
e-mail: [email protected]

Abstract—The article discusses recent results of kinetic analysis of propylene and ethylene polymerization
reactions with several types of Ti-based catalysts. All these catalysts, after activation with organoaluminum
cocatalysts, contain from two to four types of highly isospecific centers (which produce the bulk of the crystal-
line fraction of polypropylene) as well as several centers of reduced isospecificity. The following subjects are
discussed: the distribution of active centers with respect to isospecificity, the effect of hydrogen on polymeriza-
tion rates of propylene and ethylene, and similarities and differences between active centers in propylene and
ethylene polymerization reactions over the same catalysts. Ti-based catalysts contain two families of active cen-
ters. The centers of the first family are capable of polymerizing and copolymerizing all α-olefins and ethylene.
The centers of the second family efficiently polymerize only ethylene. Differences in the kinetic effects of
hydrogen and α-olefins on polymerization reactions of ethylene and propylene can be rationalized using a sin-
gle assumption that active centers with alkyl groups containing methyl groups in the β-position with respect to
the Ti atom, Ti–CH(CH3)R, are unusually unreactive in olefin insertion reactions. In the case of ethylene poly-
merization reactions, such an alkyl group is the ethyl group (in the Ti–C2H5 moiety) and, in the case of propy-
lene polymerization reactions, it is predominantly the isopropyl group in the Ti–CH(CH3)2 moiety.
DOI: 10.1134/S0965545X08110023

1 INTRODUCTION Flory exponential functions [5, 6]. Polymer materials

produced by a single type of active center in such mul-
Professor N.M. Chirkov was one of the pioneers of ticenter catalyst are called Flory components, and they
kinetic studies of olefin polymerization reactions with have characteristic asymmetric profiles in gel perme-
transition metal catalysts. This article describes a con- ation chromatograms; their molecular weight distribu-
tinuation of his work; it discusses recent results of the tion corresponds to the Mw/Mn ratio of 2.0 [1, 3–5].
kinetic analysis of propylene and ethylene polymeriza-
tion reactions with several types of Ti-based catalysts, Polymers produced with any heterogeneous Ti-based
a classical solid catalyst, δ-TiCl3, and supported cata- catalyst consist of at least four or five Flory compo-
nents. Several examples below demonstrate this mani-
lysts of the TiCl4/dibutyl phthalate/MgCl2. festation of the multicenter nature of solid Ziegler–
All solid Ti-based Ziegler–Natta catalysts, both the Natta catalysts.
early catalysts utilizing TiCl4 or TiCl3 and the modern (2) When any α-olefin is homopolymerized with
supported catalysts, contain several types of active cen- these catalysts, different types of active centers produce
ters. There are several manifestations of the multicenter macromolecules with different degrees of isotacticity
nature of the catalysts: [7–13]. Some of the centers are highly isospecific,
(1) All polymers and copolymers produced with whereas other centers produce sterically irregular (atac-
such catalysts have broad molecular weight distribu- tic) polymers. Modern supported Ziegler–Natta cata-
tions [1–4], in contrast to polymers prepared with sin- lysts for polymerization of propylene are designed to
gle-center metallocene catalysts that have narrow produce polymer mixtures with a greatly reduced frac-
molecular weight distributions [3, 4]. The molecular tion of atactic polymer.
weight of any polymer prepared with any heteroge- (3) When α-olefins and ethylene, or two different
neous Ziegler–Natta catalyst under stationary reaction α-olefins, are copolymerized with these catalysts, dif-
conditions (constant monomer concentration, tempera- ferent types of active centers produce copolymer mole-
ture, etc.) can be described by a combination of several cules of different compositions [14–18].
(4) In the course of polymerization reactions, active
1 The text was submitted by the authors in English. centers of different types are formed that decay at dif-

1101
1102 KISSIN, RISHINA

v, g/(g cat min) ing jackets, and several ports for the introduction of cat-
alysts, cocatalysts, and monomers. The experimental
(a)
40 details of the polymerization reactions and the analyti-
cal procedures of the polymer analysis, such as GPC
analysis, IR and NMR, can be found in references [4,
14, 16, 18–21].
Figure 1 shows the kinetics of two homopolymer-
20
ization reactions, those of propylene (a) and ethylene (b),
with this catalyst system at the same temperature, 80°C.
This example illustrates the most complex case of a cat-
alyzed polymerization reaction. This catalyst has low
compositional stereospecificity: propylene polymers it
produces contain fractions of different stereoregularity
(b) in comparable amounts. When used in ethylene/α-ole-
40 fin copolymerization reactions, the catalyst produces
copolymer fractions of different compositions [9]. Fig-
ure 1 shows that the active centers in the ethylene poly-
merization reaction are obviously formed at a much
lower rate and are more stable than the active centers in
20 the propylene polymerization reaction.
Figure 2 shows the data of the GPC analysis of the
polymers. Separation of the overall GPC curves into
Flory components (carried out as described in reference
[4]) showed the following (Table 1):
0 20 40 60
Time, min (1) The polypropylene has a broad molecular weight
distribution (Mw /Mn = 14.6). At least six Flory compo-
Fig. 1. Kinetics of homopolymerization reactions of nents are required to represent this GPC curve. They are
propylene (a) and ethylene (b) with δ-TiCl3–AlEt3 marked A to F in Table 1 in the order of increasing
catalyst system at 80°C. Monomer partial pressures molecular weight. In reality, the situation is even more
and concentrations: propylene, 0.62 MPa, 1.68 mol/l; complicated because the overall polymer mixture con-
ethylene, 0.45 MPa, 0.32 mol/l. tains macromolecules with widely different isotacticity.
This subject discussed in the next section.
(2) The polyethylene has a relatively narrow molec-
ferent rates. As a result, properties of polymers (such as ular weight distribution (Mw /Mn = 5.3). It can be repre-
their molecular weight distribution, stereo-composi- sented by five Flory components. They are marked I to
tion, copolymer composition, etc.) often vary with V in Table 1 in order of increasing molecular weight.
reaction time [15–18]. This means that the catalyst has, as a minimum, five dif-
(5) Active centers of different types can be poisoned ferent types of active centers. The real number of types
by different chemical compounds. This difference in of active centers in the catalyst can be higher: GPC
reactivity is the principal means by which polypropy- analysis does not distinguish active centers if they pro-
lene with a high degree of fractional isotacticity is pro- duce Flory components with similar weight-average
duced, by applying specific poisons of active centers molecular weights even if other properties of the cen-
(internal and external electron donor compounds). ters differ significantly.
Below, for brevity, the expressions “the group of
active centers producing a given Flory component,” for
COMPARATIVE ANALYSIS example, “the group of active centers producing the
OF HOMOPOLYMERIZATION REACTIONS Flory component C or the Flory component III,” are
OF PROPYLENE AND ETHYLENE replaced with the expressions “center C or center III”.
WITH A CLASSICAL ZIEGLER–NATTA
CATALYST
Similarities and differences between active centers DISTRIBUTION OF ISOSPECIFIC ACTIVE
in propylene and ethylene polymerization reactions CENTERS IN SUPPORTED ZIEGLER–NATTA
over Ti-based catalysts can be illustrated by carrying CATALYSTS
out polymerization reactions of the two monomers with All Ti-based Ziegler–Natta catalysts for polymer-
the same catalyst system, a combination of solid ization of propylene, both TiCl3 and supported catalysts
δ-TiCl3 and AlEt3 as a cocatalyst. The polymerization of the type TiCl4/MgCl2/internal electron donor, con-
reactions were carried out at a high pressure in metal tain a large variety of active centers, including highly
reactors equipped with propeller stirrers, external heat- isospecific centers, centers of moderate isospecificity,

POLYMER SCIENCE Series A Vol. 50 No. 11 2008

 KINETICS OF PROPYLENE AND ETHYLENE POLYMERIZATION REACTIONS 1103

and nearly completely aspecific centers. The isospeci-

ficity of a particular active center is determined, in a (a)
general sense, by the probability of isotactic linking of
adjacent monomer units in a polymer chain. Physical
separation of polymer components produced by each
type of active center is a difficult task. At the present
time, the most detailed investigation of the steric inho-
mogeneity of polypropylene produced with multicenter
catalysts is achieved by using two complementary
modern fractionation techniques, analytical or prepara-
tive Tref methods [22–32] and Crystaf [33–37].
The results of these analyses are mainly dependent on
the stereoregularity of polypropylene and are almost
independent of molecular weight [38, 39]. The Tref
technique reveals even minor differences in stereoregu-
larity of polymers prepared with similar catalysts [30,
31, 40–42].

Effects of Silane Modifiers on the Structure

of Crystalline (Isotactic) Fractions
The addition of alkyl- or arylalkoxysilanes to orga- (b)
noaluminum cocatalysts is the universally practiced
technique to increase the fraction of crystalline
polypropylene fraction produced with supported
TiCl4/dialkyl phthalate/MgCl2 catalysts. The primary
effect of the silane is a sharp decrease in the content of
the amorphous fraction, from 30–50% to less than 2–
4% [43–45]. This effect is mostly achieved by selective
poisoning of aspecific active centers. In addition, the
silanes affect both the molecular weight distribution
and the stereodistribution of the crystalline fractions.
Table 2 gives the results of resolution of GPC curves
of two crystalline fractions prepared with the same sup-
ported catalyst, TiCl4/dibutyl phthalate/MgCl2, at 80°ë.
One of the polymers was produced with Al–i-Bu3 as a
cocatalyst, and the second one, with the AlEt3–(cyclo-
C5H9)2Si(OMe)2 mixture at [Al] : [Si]molar = 8 [46]. 3 5 7
They afford a relatively straightforward explanation of log M
the silane effect. The silane significantly reduces the
relative activity of center A, and, to a smaller degree, Fig. 2. GPC curves of polypropylene (a) and polyeth-
center B. The weight-average molecular weight of the ylene (b) produced with δ-TiCl3–AlEt3 catalyst sys-
polymer nearly doubles as a result of adding silane tem at 80°C and their separation into Flory compo-
[46]. On the other hand, silane does not significantly nents.
affect the molecular weight of any Flory component,
and it does not affect the ratio between the contents of
components C and D. The same silane effect in propy- the fraction. A comparison of the GPC data in Table 2
lene polymerization with the TiCl4/DIBP/MgCl2 cata- and the NMR data in Table 3 strongly suggests that
lyst activated with an AlEt3/Ph2Si(OMe)2 mixture is Flory components A and B have lower stereoregularity
described in [4]. than components C and D. The results presented in [47]
Table 3 compares 13C NMR isotacticity parameters describe quantitative aspects of the distribution of isos-
of the crystalline fractions of the two polymers. Only pecific centers in five highly isospecific supported cat-
the most significant steric errors are listed, all other alyst systems. These catalysts represent three genera-
steric errors are present in amounts too low to be tions of supported polymerization catalysts [48]:
dependably measured. The overall effect of silane is
obvious: the modification not only increases the content Catalyst systems of the third generation: catalyst
of the crystalline fraction from 26 to 99%, but it also TiCl4/MgCl2/ethyl benzoate; cocatalyst AlEt3-ethyl
noticeably improves the average NMR isotacticity of p-ethoxy benzoate, [Al]:[ester] = 1.8 (1).

POLYMER SCIENCE Series A Vol. 50 No. 11 2008

1104 KISSIN, RISHINA

Table 1. Analysis of GPC data for polypropylene and polyethylene produced with δ-TiCl3–AlEt3 catalyst at 80°C [52] (see
Fig. 2)
Mw × 10–3
Polymer Mw /Mn Flory component Mw × 10–3 Fraction, %
(average)
Polypropylene 274.3 14.6 A 5.3 5.9
B 17.3 15.1
C 48.0 21.4
D 133.6 21.7
E 447.7 22.3
F 976.7 13.5
Polyethylene* 152.0 5.3 I 2.6 0.5
II 16.6 10.5
III 58.6 40.1
IV 151.7 32.4
V 468.6 16.6
* Hydrogen was used in the reaction (cH = 7.3 × 10–3 M) to reduce the molecular weight of the polymer and to make it completely soluble
in GPC analysis.

Table 2. Average molecular weights, Mw /Mn values, and results of GPC resolution for crystalline fractions of polymers pre-
pared at 80°C with TiCl4 /dibutyl phthalate/MgCl2 catalyst activated with different cocatalysts [46]
Cocatalyst Mw × 10–3 Mw /Mn Flory component Mw × 10–3 Fraction, %
Al–i-Bu3 568.960 5.87 A 104.670 39.8
B 323.890 38.0
C 971.380 15.7
D 3929.550 6.4
AlEt3–(cyclo-C5H9)2Si(OMe)2 1074.200 4.34 A ~90.000 4.7
B 430.750 47.0
C 1166.650 36.3
D 3708.600 12.0

Catalyst systems of the fourth generation: hydrogen. The molecular weights of the polymers and
TiCl4/MgCl2/diisobutyl phthalate–AlEt3; external elec- the contents of their three fractions, “amorphous” (sol-
tron donors (3,3,3-F3-Pr)(Me)Si(OMe)2 (2); (cyclo- uble in Ó-xylene at 25°ë), material of low crystallinity
C6H11)(Me)Si(OMe)2 (3), and (cyclo-C5H9)2Si(OMe)2 (soluble in Ó-xylene at 95°ë), and the highly crystalline
(4), all at a [Al]:[silane] ratio of 20. fraction (insoluble in Ó-xylene at 95°ë), are given in
Catalyst systems of the fifth generation: catalyst Table 4.
TiCl4/MgCl2/2-isopropyl-2-isopentyl-1,3-dimethox- The highly crystalline fractions were subsequently
ypropane, cocatalyst AlEt3 (5). fractionated by the analytical Tref method between
The polymerization reactions were carried out in 95°C and 120°ë at a rate of 3°C/h using o-xylene as a
liquid propylene at 70°ë for 1 h in the presence of solvent [30]. Figure 3 shows the Tref curves of all five

Table 3. Silane effect on average stereoregularity and some chain defects in crystalline fractions of polymers prepared at
80°C with TiCl4 /dibutyl phthalate/MgCl2 catalyst [46]
Steric-pentad content, %
Cocatalyst
[mmmm] [mmmr] [mmrr] [mmrm + rrmr] [rrrr] [mrrm]
Al–i-Bu3 89.2 3.55 3.78 0.63 0.37 1.63
AlEt3–(cyclo-C5H9)2Si(OMe)2 97.8 1.25 0.93 0.15 0.30 0.32

POLYMER SCIENCE Series A Vol. 50 No. 11 2008

 KINETICS OF PROPYLENE AND ETHYLENE POLYMERIZATION REACTIONS 1105

Table 4. Characteristics of propylene polymers prepared with different supported catalysts [47]

Generation Catalyst Content of fractions, %

External donor Mw × 10–3 Mw /Mn [ m 4 ]**
of catalyst system Am* L-cr* H-cr*
av

3 1 ethyl benzoate 485 8.3 5.1 19.2 75.7 0.973

4 2 (3,3,3-F3–Pr)(Me)Si(OMe)2 207 6.1 3.8 21.4 74.8 0.962
4 3 (cyclo-C6H11)(Me)Si(OMe)2 375 6.0 3.2 7.2 89.6 0.971
4 4 (cyclo-C5H9)2Si(OMe)2 356 8.9 2.4 4.6 93.0 0.987
3 5 – 259 6.4 2.4 5.9 91.7 0.972
Notes: * Fractions: Am—amorphous, L-cr—fraction of low crystallinity, H-cr—fraction of high crystallinity.
** The average [m4] value for the highly crystalline fraction, calculated from the Tref data.

highly crystalline fractions. The Tref analysis of of polypropylene prepared with three of the catalyst
polypropylene was based on the resolution of multi- systems [46].
component Tref curves into peaks of the constituent Examination of the data in Fig. 4 and Table 5 affords
elemental components represented by the Lorentz dis- several conclusions about the stereospecificity distribu-
tribution function [48]. As an example, Fig. 4 shows the tion of the active centers producing the highly crystal-
analytical Tref curve of the crystalline fraction pro- line polypropylene fractions. The Tref analysis shows
duced with system 4 and its resolution into elemental that the polypropylene fractions insoluble in o-xylene
components, each presumably produced by a single at 95°ë are by no means structurally uniform materials.
type of active center of a particular isospecificity. Each fraction contains several Tref components that
The number of components sufficient for the ade- differ in [m4]av values. Some of the Tref components,
quate representation of the experimental curves varies usually those eluting at the highest temperatures, are
from four to six, depending on the polymer. Table 5 dominant in the mixtures; other components are minor.
gives the peak positions of all the peaks, estimations of The [m4]av values for the dominant components in the
the [m4] values (the mmmm pentad content from 13C five polymer fractions are markedly different.
NMR data) for each component, and the fractions of the Based on the [m4]av values, all dominant Tref com-
components. The components are marked A to G in ponents can be separated into two groups:
order of decreasing [m4] values. These results correlate (1) The components of the first group (components
well with the experimentally measured by 13C NMR A and B in Table 5) have the highest [m4]av values, in
[m4]av values for the several xylene-insoluble fractions the range of 0.986 to 0.993. These are exceptionally

5
3 4
2 1

95 105 115 125

T, °C

Fig. 3. Tref curves of crystalline fractions of polypropylene prepared with catalyst systems of 3rd, 4th and 5th gener-
ations Curves 1–5 correspond to systems 1–5.

POLYMER SCIENCE Series A Vol. 50 No. 11 2008

1106 KISSIN, RISHINA

100 110 120 T, °C

Fig. 4. Resolution of Tref curve of the crystalline fraction of polypropylene prepared with supported catalyst system
TiCl4 /dibutyl phthalate/MgCl2–AlEt3 /(cyclo-C5H9)2Si(OMe)2.

high [m4]; the probability of isotactic chain propagation generations, but it is absent in the polymer produced by
for these centers is in the range of 0.997 to 0.999. Two the catalyst system of the fourth generation with the
such Tref components are produced by a diisobutyl external donor (3,3,3-F3-Pr)(Me)Si(OMe)2.
phthalate–based catalyst system containing (cyclo- In addition to several dominant components of the
C5H9)2Si(OMe)2 as an external electron donor and one highest stereoregularity, each highly crystalline
produced by the ethyl benzoate–based catalyst system. polypropylene fraction contains small amounts of com-
(2) The dominant Tref components of the second ponents of lower stereoregularity, components F and G,
group (components C, D, and E in Table 5) have the each less than 10%. These minor components have [m4]
[m4] values in the range of 0.960 to 0.970; the probabil- values in the range of 0.955 to 0.935 and the probability
ity of the isotactic chain propagation for these centers is of the isotactic chain propagation for these centers is
~0.995–0.996. ~0.987–0.990.
The data in Table 5 suggest that the dominant Tref The results presented in Table 5 causes us to reex-
components of different polymers eluting at very simi- amine the effects of organic electron donors in the sup-
lar temperatures are, in the first approximation, the ported Ziegler–Natta catalysts of this type. The average
same components, and that various catalyst systems [m4] value for a combined crystalline fraction of
differ mostly in the proportions between the compo- polypropylene, such as a polymer fraction insoluble in
nents. For example, Tref component C is the principal boiling n-heptane or in cold xylene, cannot be regarded
component in the polymer produced by the catalyst sys- as a fully representative parameter of the maximum ste-
tem of the third generation and one of the two principal reoregulating power of active centers in a particular cat-
components produced by catalyst systems of the fourth alyst. As the data in Fig. 4 and Table 5 show, all such
(external donor (cyclo-C6H11)(Me)Si(OMe)2) and fifth crystalline fractions (which in the past were regarded as

Table 5. Results of Tref curve resolution of highly crystalline polypropylene samples (see Table 4)

[m4] Fr., % [m4] Fr., % [m4] Fr., % [m4] Fr., % [m4] Fr., % [m4] Fr., % [m4] Fr., %
System
A B C D E F G

1 – – 0.986 7.6 0.978 71.2 – – 0.961 6.2 0.954 7.6 0.941 7.3
2 – – – – – – 0.969 28.3 0.965 51.7 0.948 12.0 0.935 8.0
3 – – – – 0.977 32.3 0.972 48.0 0.966 8.6 0.954 6.9 0.939 4.2
4 0.993 59.8 0.986 26.5 0.976 3.1 – – 0.963 3.4 0.951 6.3 0.940 0.9
5 – – – – 0.979 32.7 0.975 47.5 0.963 5.9 0.950 11.3 0.935 2.6

POLYMER SCIENCE Series A Vol. 50 No. 11 2008

 KINETICS OF PROPYLENE AND ETHYLENE POLYMERIZATION REACTIONS 1107

Table 6. Content of dominant components and [m4]av val- polymers, in Table 4) and the average isotacticity of the
ues for catalyst system 1–5 [48] dominant components in the same fractions could be
traced.
Catalyst system Dominant [m4]av of dominant
(Table 4) component, % component
1 78.8 0.976 KINETICS OF PROPYLENE
2 80.0 0.966 HOMOPOLYMERIZATION REACTIONS
WITH SUPPORTED CATALYSTS
3 80.3 0.974
4 86.6 0.991 Propylene polymerization reactions with the
5 80.2 0.977 δ-TiCl3–AlEt3 catalyst (see Figs. 1a and 2a) are an
inconvenient object for the kinetic research of olefin
polymerization reactions. This catalyst produces pro-
perfectly isotactic materials) contain noticeable pylene polymers with a low content of the crystalline
amounts of materials of reduced isotacticity, the minor isotactic fraction (the polymer fraction insoluble in
components F and G. The contents of these minor com- boiling n-heptane), ~70%, and the polymers prepared
ponents differ from polymer to polymer. It seems justi- with the catalyst are complex mixtures of macromole-
fied to introduce an additional parameter characterizing cules with widely varying degrees of isotacticity. Stud-
stereospecificity of active centers in Ti-based Ziegler– ies of propylene polymerization reactions with a highly
Natta catalysts, an average [m4] value for the main Tref isospecific catalyst system TiCl4/dibutyl phtha-
components in the crystalline fractions (the dominant late/MgCl2 activated with an AlEt3/PhSi(OEt)3 cocata-
isotactic component). These dominant components lyst mixture at [Al]:[Si] = 20 have significant advan-
(two Tref components in each polymer) constitute 80– tages from the kinetic standpoint [52]. Due to strong
85% of the material in the crystalline polypropylene poisoning of sterically aspecific active centers, this cat-
fractions. Their [m4]av value better represents the alyst system produces polypropylene with the isotactic-
effects of the electron donors on the performance of the ity index of ~95–97%. Therefore, the distribution of
active centers of the highest stereospecificity [46]. active centers with respect to the molecular weights of
the Flory components they produce becomes the domi-
The average isotacticity of the dominant compo- nant factor in comparison with their distribution with
nents ([m4]av value) is noticeably different for the poly- respect to stereospecificity. The principal results are
mers produced by different catalyst systems (Table 6). given in Table 7. Figure 5 shows the reaction kinetics in
System 4 produces the dominant component of the experiments 1 and 3. Similar to many published data
highest average isotacticity; systems 1, 3, and 5 have [16–19, 53, 54], the polymerization rate in the presence
similar dominant components; and system 2 produces of hydrogen is significantly higher. A mechanistic
the dominant component of the lowest isotacticity. The explanation of this effect [19] is discussed below.
isospecificity of the active centers producing the domi-
nant component reflects the effect of the steric bulk of The replacement of the δ-TiCl3–AlEt3 catalyst with
the external donor. The most effective external donors the highly isospecific TiCl4/dibutyl phthalate/MgCl2
R2Si(OMe)2 in the catalysts of the fourth generation catalyst results in significant narrowing of the molecu-
contain bulky alkyl groups R in the α position to the sil- lar weight distribution of polypropylene, mostly due to
icon atom [49, 50]. A possible reason for the advan- poisoning of aspecific centers: the Mw /Mn value
tages of these silanes is their weaker ability to form decreases from 14.6 to 6.5–5.0. Five types of active
complexes with organoaluminum cocatalysts, which centers, marked A through E in order of increasing
prevents removal of silanes from the catalyst surface molecular weight, describe the molecular weight distri-
[51]. No obvious correlation between the conventional bution of these polymers in a satisfactory manner.
fractionation data for the polypropylene samples (such Weight-average molecular weights and fractions of the
as the content of the highly crystalline fraction in the Flory components are listed in Table 8.

Table 7. Propylene homopolymerization reactions at 70°C with TiCl4 /dibutyl phthalate/MgCl2–AlEt3 /PhSi(OEt)3 catalyst
system, [Al] : [Si] = 20 [52]
cH × 10–3,
Exp. pPr , MPa pH /pPr cPr , mol/l Yield, g/h [Ti], ppm* Mw × 10–3 Mw /Mn
mol/l
1 0.42 0 1.30 0 12.5 22.0 412.9 6.48
2 0.40 0.052 1.24 1.24 13.3 4.1 76.0 4.94
3 0.37 0.130 1.15 2.87 18.1 5.3 59.5 5.44
* Ti content in polymers, measured with X-ray fluorescent spectroscopy method.

POLYMER SCIENCE Series A Vol. 50 No. 11 2008

1108 KISSIN, RISHINA

v, l/(g cat min) changes is, of course, a significant decrease in the aver-
2 age Mw and Mn values and a slight narrowing of the
molecular weight distribution (Table 6).

Correlation between Hydrogen Concentration

and the Molecular Weight of Polypropylene
1 In general, hydrogen is a very effective chain trans-
fer agent in the polymerization of all olefins with Zie-
gler–Natta and metallocene catalysts. The amount of
1 hydrogen required to reduce the molecular weight of
polypropylene to a commercially useful level is rela-
2 tively small, and the decrease in molecular weight cor-
relates with the hydrogen concentration, cH (or its par-
0 20 40 60 tial pressure, pH), according to the equation applicable
Time, min to olefin polymerization reactions in general [9]:

Fig. 5. Kinetics of propylene polymerization at 70°C

∑ R /(k c
i H
p Pr ) + ( k t c H )/ ( k p c Pr ),
with supported catalyst system TiCl4 /dibutyl phtha- 1/ν = t (1)
late/MgCl2–AlEt3 /PhSi(OEt)3 at 70°C. (1) pH/pPr = 0;
(2) pH/pPr = 0.13.
where ν is the average degree of polymerization, kp is
As the hydrogen partial pressure increases, molecu- the propagation rate constant, cPr is the propylene con-
lar weights of all Flory components decrease approxi- H
centration, k t is the rate constant of chain transfer with
mately in parallel [20], although the molecular weights
∑
i
of the two components with the highest Mw, D and E, hydrogen, and R t is the sum of rates of other chain
decrease to a greater extent than those of the first three transfer reactions, such as chain transfer with a mono-
components. The contents of each Flory component mer, with a cocatalyst, β-hydrogen elimination, etc.
remain approximately the same independently of the Equation 1 signifies that a linear correlation should
hydrogen concentration. The overall effect of these exist between a value of 1/ν and the cH/cPr ratio with
H
k t /kp as the slope. (If the cPr value is kept constant in a
Table 8. Hydrogen effect on molecular weights and con- given series of experiments, dependences 1/ν vs. cH or
tents of Flory components in propylene polymers produced vs. pH are often plotted).
at 70°C [20]. Experimental conditions are in Table 7
Equation 1 holds reasonably well at low cH values.
Exp. pH /pPr Component Mw × 10–3 Fraction However, when the relationships between 1/ν and cH
were tested in a broader cH range, significant deviations
1 0 A 16.7 4.1 from Eq. (1) were usually observed: the hydrogen
B 67.6 19.7 effect on the molecular weight at high cH was lower
C 197.8 37.9 than expected. To account for these deviations, the sec-
ond term in Eq. (1) is often replaced with a term
D 586.3 30.2 H
E 1205.0 8.1 k t (cH)0.5/(kpcPr) [8, 55]. This dependence was even
interpreted by some researchers as a manifestation of
2 0.052 A 3.7 3.6 chain transfer with atomic hydrogen [56–60]. Some
B 16.2 16.6 data [61] showed an even more peculiar effect: at very
C 47.4 47.5 high pH, the ν value does not depend on the hydrogen
concentration anymore. This effect is rarely observed
D 117.1 26.3
because polypropylene produced at such a high pH has
E 224.9 5.9 too low a molecular weight to be of any practical use.
3 0.130 A 2.9 5.3 To demonstrate the cH effect on the molecular
B 13.6 17.7 weight of polypropylene, three measures of the molec-
C 38.8 42.0 ular weight can be used: the weight- or the number-
average molecular weights, Mw or Mn, measured by
D 85.5 26.9 GPC; estimations of Mw from melt indices of the poly-
E 146.3 8.0 mers; and molecular weights of the Flory components.

POLYMER SCIENCE Series A Vol. 50 No. 11 2008

 KINETICS OF PROPYLENE AND ETHYLENE POLYMERIZATION REACTIONS 1109

(1/νn) × 103 (1/νn) × 103

2 8
(a)

1
4

0 4 8
(cç/cPr) × 102

Fig. 6. Effect of hydrogen concentration on molecular 20

weights of crystalline fractions of polypropylene pre- (b)
pared with system TiCl4 /dibutyl phthalate/MgCl2–
AlEt3 /PhSi(OEt)3 in coordinates of Eq. (1).

The following examples demonstrate deviations 10

from Eq. (1) in propylene polymerization reactions at
high cH in terms of average molecular weights. A series
of polymerization experiments was carried out at a con-
stant cPr value of 1.0 M [61], crystalline fractions of
polypropylene were separated, and their Mn values
0 10 20
were measured by GPC. Figure 6 shows the plot of 1/νn (pç/pPr) × 102
for the fractions (νn is the number-average degree of
polymerization) vs. cH /cPr. At low cH, the dependence is Fig. 7. Effect of hydrogen concentration on molecular
steep and nearly linear. Its slope corresponds to the weights of Flory component B (a) and C (b) in crystal-
H line fractions of polypropylene prepared with system
k t /kp value of ~0.36, indicating a strong chain-transfer
TiCl4 /dibutyl phthalate/MgCl2–AlEt3 /PhSi(OEt)3 in
effect of hydrogen. However, as cH increases further, coordinates of Eq. (1).
the molecular weight no longer decreases. A similar
dependence between 1/νw (from Mw values) vs. pH /pPr
was described in [53]. with different Ti-based supported catalysts [18] demon-
strate that the linear dependence between 1/νw and
The most straightforward approach to studies of the pH /pE values described by Eq. (1) holds in a broad range
hydrogen effect is via analysis of the shape of GPC
curves of polymers synthesized under stationary reac- of pH. Figure 8 shows plots of (I2)1/3.5 vs. pH /pE for two
tion conditions and plotting of the molecular weight of series of ethylene homopolymers prepared at different
each Flory component vs. pH /pPr in the coordinates of temperatures [(I2)1/3.5 is approximately proportional to
Eq. (1). Two of such plots, for Flory components B and 1/νw].
C, are shown in Fig. 7; the data for the other three com-
ponents show a similar trend. The plots duplicate the
dependences for average molecular weights of the same KINETICS OF ETHYLENE
polymers: the 1/νn value of each Flory component rap- HOMOPOLYMERIZATION
idly increases with pH /pPr in the low pH range, but the AND COPOLYMERIZATION REACTIONS
dependence levels off at high pH values. These results
indicate that the unusual cH effect on the average In order to compare several features of polymeriza-
molecular weight is not caused by any significant tion kinetics of ethylene and propylene, two types of
change in the contributions of different Flory compo- ethylene polymerization experiments were carried out
nents to the total polymer makeup but reflects a partic- with the same TiCl4/dibutyl phthalate/MgCl2 catalyst
ular kinetic effect inherent to each type of active center. (see Table 9). They include ethylene homopolymeriza-
It should be noted that the hydrogen effect on molecular tion reactions (Fig. 9), both with AlEt3 (experiment 1)
weight in ethylene homopolymerization reactions is and with the AlEt3/PhSi(OEt)3 cocatalyst mixture
different. Results of ethylene polymerization studies (experiment 3), and ethylene/1-hexene copolymeriza-

POLYMER SCIENCE Series A Vol. 50 No. 11 2008

1110 KISSIN, RISHINA

(I2)1/3.5 v, mol/(g cat min)

6
3 (a)

2
4

1
1

0 1 2 2
6 (b)
0 20 40 60
Time, min
4
Fig. 9. Comparison of ethylene (1) and propylene (2)
homopolymerization kinetics with the catalyst system
2 TiCl4/dibutyl phthalate/MgCl2–AlEt3 /PhSi(OEt)3
(experiment 2 in Table 7 and 3 in Table 9). Monomer
concentrations are 1.24 and 1.12 M, respectively.

0 2 4
pç/pE On the other hand, ethylene/α-olefin copolymerization
reactions in the presence of hydrogen are an important
Fig. 8. Effect of hydrogen concentration on molecular source of information on the polymerization kinetics,
weights of polyethylene in coordinates of Eq. (1) the molecular weight distribution, and the composi-
[(I2)1/3.5–1/νw]. Data for ethylene homopolymers pro- tional distribution of the copolymers.
duced with different Ti-based supported catalysts at Ethylene homopolymerization kinetics over
80°C (a) and 90°C (b).
TiCl4/dibutyl phthalate/MgCl2 catalyst with both cocat-
alysts (Fig. 9) resembles the kinetics of the same reac-
tion reactions (Fig. 10) with the same two cocatalysts tion with the δ-TiCl3–AlEt3 system (Fig. 1): the active
(experiments 2 and 4, respectively). species are formed relatively slowly but, after the max-
Ethylene homopolymerization reactions in the imum rate is reached, the centers remain nearly stable
absence of hydrogen usually result in the formation of for several hours. When pure AlEt3 is used as a cocata-
polymers with very high molecular weights that cannot lyst in ethylene polymerization reactions, the reaction
be reliably measured by GPC. Therefore, experiments kinetics is essentially the same as in Fig. 9 for the
1 and 3 were carried out strictly for kinetic purposes. AlEt3/PhSi(OEt)3 cocatalyst mixture although both the

Table 9. Ethylene (E) homopolymerization and ethylene/1-hexene (Hex) copolymerization reactions with TiCl4/dibutyl ph-
thalate/MgCl2 catalyst [52]

Yield × 10–3, c Hex

cop
, mol
Exp. Temp., °C pE , MPa cE , mol/l cH , mol/l cHex , mol/l M **
w × 10
–3 (M /M )**
w n
g/g cat %*
Cocatalyst: AlEt3
1 70 1.42 1.14 0 0 12.3 0
2 85 0.69 0.51 0.014 2.96 13.1 4.3 103 4.5
Cocatalyst: AlEt3 /PhSi(OEt)3 mixture at [Al] : [Si] = 20
3 70 1.40 1.12 0 0 5.58 0
4 85 0.69 0.51 0.014 2.96 7.25 2.1 144 4.0
Notes: * Content of 1-hexene in copolymer.
** Estimation from melt indexes and melt flow ratios [13].

POLYMER SCIENCE Series A Vol. 50 No. 11 2008

 KINETICS OF PROPYLENE AND ETHYLENE POLYMERIZATION REACTIONS 1111

v, g/(g cat min) row and can be effectively represented by four Flory
300 components (Table 10).
Comparison of Figs. 2 and 8 shows that elimination
of the distribution of active centers with respect to isos-
pecificity somewhat simplifies the search for corre-
200 spondence between the active centers that produce dif-
ferent Flory components in propylene and ethylene
polymerization reactions. Different types of active cen-
1 ters have very different abilities to copolymerize ethyl-
100 ene with α-olefins [14–18]. The centers IV and V copo-
lymerize α-olefins with ethylene very poorly and pro-
2 duce essentially ethylene homopolymers, whereas
centers I, II, and III easily produce ethylene/α-olefin
0 20 40 60 copolymers. This difference in the compositions of dif-
Time, min ferent copolymer constituents can be best demonstrated
by examining their crystallization fractionation (Crys-
Fig. 10. Kinetics of ethylene/1-hexene copolymeriza- taf) data.
tion reactions (experiments 2 and 4 in Table 9) with Figure 11 shows the Crystaf curve of an ethylene/1-
TiCl4 /dibutyl phthalate/MgCl2 catalyst activated with hexene copolymer prepared with the TiCl4/dibutyl
two cocatalysts, AlEt3 (1) and AlEt3 /PhSi(OEt)3 mix- phthalate/MgCl2–AlEt3 system at 85°ë (experiment 2
ture (2). in Table 9). The curve consists of two partially overlap-
ping features. The fraction precipitating in the 85–75°ë
reaction rate and the catalyst productivity are ~2 times range (its content is ~45%) has a very low degree of
higher (see Table 9). The reasons for the effect of branching; it mostly consists of Flory components IV
and V (Table 10). The very broad fraction precipitating
PhSi(OEt)3 are described below. at lower temperatures, from ~70°C to 30°C (~35%),
The kinetics of ethylene/α-olefin copolymerization represents the branched material produced by centers
reactions (Fig. 10) is quite different from the kinetics of III and II, whereas the fraction that remains in solution
ethylene homopolymerization: the active centers are even at 30°ë (~20%, not shown in the figure) is the
highly branched material produced by center I and, par-
formed much faster and the reaction rates are signifi- tially, center II. The difference in the copolymerization
cantly higher, but the catalyst becomes less stable. The abilities of different active centers was noticed long
molecular weight distribution of the ethylene/1-hexene time ago [10, 14, 62–64]; the above GPC and Crystaf
copolymer prepared with the TiCl4/dibutyl phtha- results assign the differences to particular populations
late/MgCl2–AlEt3 system (experiment 2) is quite nar- of active centers.

40 60 80 T, °C

Fig. 11. Crystaf curve for ethylene/1-hexene copolymer prepared with the TiCl4 /dibutyl phthalate/MgCl2–AlEt3 sys-
tem (experiment 2 in Table 9) and its separation into elemental components.

POLYMER SCIENCE Series A Vol. 50 No. 11 2008

1112 KISSIN, RISHINA

Table 10. Analysis of GPC data for ethylene/1-hexene co- the active centers, either in propylene polymerization
polymer (Table 9, experiment 2) produced with reactions (Figs. 1a and 12) or in ethylene/α-olefin
TiCl4 /dibutyl phthalate/MgCl2–AlEt3 system at 85°C [52] copolymerization reactions (Figs. 12 and 13). Centers
IV and V in ethylene/α-olefin copolymerization reac-
Component Mw × 10–3 Fraction, % tions are kinetically stable (see Fig. 15) merely because
I 4.6 ~1 these centers incorporate α-olefin molecules into poly-
mer chains very poorly.
II 15.4 12.1 Organosilicon modifiers (external organic donors),
III 47.1 42.0 such as PhSi(OEt)3 used in this study, poison centers I,
IV 127.5 34.7
II and III to different degrees [15], but they affect cen-
ters IV and V much less. When such poisons are intro-
V 364.4 10.3 duced in ethylene copolymerization reactions, they of
course decrease the overall catalyst activity (Fig. 10,
Table 9). In addition, these compounds, by selectively
Different types of active centers in ethylene/α-olefin poisoning centers I, II, and III, depress the overall abil-
copolymerization reactions are formed and decay at ity of the catalyst to copolymerize α-olefins with ethyl-
different rates. Figure 12 shows the approximate sepa- ene: the average 1-hexene content in the copolymer
ration of the overall kinetic curve in the ethylene/1-hex- prepared with PhSi(OEt)3 under the same conditions is
ene copolymerization reaction at 85°C with TiCl4/dibu- reduced by half (Table 9). Because centers I, II and III
tyl phthalate/MgCl2–AlEt3 system (experiment 2 in produce Flory components with lower molecular
Table 8) into kinetic curves of different active centers. weights (Table 10), such selective poisoning also
The procedure for this kinetic treatment is based on results in an increase in the average molecular weight
GPC analysis; it was described in detail in [3, 15, 16, of the copolymer.
18]. Centers IV and V usually exhibit a very stable
kinetic behavior, and they are unaffected by the pres-
ence of α-olefins. Centers I, II and III are formed sig- DIRECT COMPARISON OF ETHYLENE VS.
nificantly faster and decay faster as well. They are sig- THE PROPYLENE POLYMERIZATION KINETICS
nificantly activated in the presence of α-olefins (see the As mentioned above, an important subject in the
proposed mechanism below); this explains the differ- studies of the polymerization kinetics of ethylene and
ences in the kinetic profiles of ethylene homopolymer- α-olefins with any heterogeneous Ti-based catalyst is
ization reactions and ethylene/α-olefin copolymeriza- finding a correspondence between active centers A–E
tion reactions (Figs. 9 and 10). The main apparent rea- in propylene polymerization reactions and active cen-
son for instability of all active centers is the presence of ters I–V in ethylene polymerization reactions. In ethyl-
tertiary C–H bonds in the β-position to the Ti atom in ene polymerization reactions, centers IV and V, the cen-

v, g/(g cat min)

300

200

III
100
II
IV
I V

0 20 40 60
Time, min

Fig. 12. Approximate separation of the kinetic curve in ethylene/1-hexene copolymerization reaction with
TiCl4 /dibutyl phthalate/MgCl2–AlEt3 system (experiment 2 in Table 9) into kinetic curves of different active centers.

POLYMER SCIENCE Series A Vol. 50 No. 11 2008

 KINETICS OF PROPYLENE AND ETHYLENE POLYMERIZATION REACTIONS 1113

v, g/(g cat min)

200 2

100

1
0 120 240
Time, min

Fig. 13. Kinetics of consecutive propylene (1) and ethylene (2) polymerization experiments at 70°C with the δ-TiCl3–
AlEt3 system. First stage cPr = 1.40 mol/l for 3 hours; Second stage cE = 0.13 mol/l for 20 min.

ters that polymerize ethylene nearly exclusively, are troduced into the reactor and the ethylene homopoly-
very stable (Fig. 12), whereas centers I–III (true copo- merization reaction resumed at virtually the same rate
lymerization centers) decay much faster. Comparison as before replacement with propylene.
of Fig. 12 with Figs. 1a and 9 (propylene polymeriza-
tion kinetics) invites the suggestion that centers IV and In several consecutive polymerization experiments,
V are completely inactive in propylene polymerization. propylene was first polymerized for long periods of
On the other hand, centers I–III are the same as centers time and was then replaced with ethylene. Figure 13
A–E in propylene polymerization reactions. gives one example. The δ-TiCl3–AlEt3 system was used
at 70°ë in the absence of hydrogen (which explains the
To evaluate the validity of this argument, several relatively low polymerization rate). The nearly com-
sequential ethylene and propylene homopolymeriza- pletely decayed catalyst, after replacement of propy-
tion reactions with both catalysts were carried out. In lene with ethylene, was capable of very high activity in
these experiments, one of the monomers was polymer- ethylene polymerization. In the second example [52],
ized for a significant period of time (2–3 h); it was then
removed from the reactor and replaced with another
monomer. This technique is often used to evaluate the
effects of a monomer change or the use of monomer
mixtures in polymerization kinetics [54, 65–73]. The
main rationale for using the sequential polymerization 3
technique for these purposes is the finding that that var-
ious centers differ greatly in stability (Fig. 12).
In one consecutive polymerization experiment, eth-
ylene was polymerized with the supported TiCl4/dibu-
tyl phthalate/MgCl2 catalyst at 90°ë for 3 hours at a
low ethylene partial pressure, 0.19 MPa, to allow cen- 2
ters I, II and III to decay. Then, in preparation for
replacement of ethylene with propylene, the tempera-
ture was decreased to 70°ë, resulting in a steady ethyl- 1
ene polymerization reaction at a rate two times lower
than at 90°ë. Then, after ethylene removal from the 9 8 7
reactor (by repeated purges with nitrogen), propylene ν × 10–2, cm–1
was introduced at pPr = 0.36 MPa. This propylene par-
tial pressure corresponds to a propylene concentration
of 1.1 M, which is seven times higher than the ethylene Fig. 14. IR spectra of impact resistant polypropylene
concentration in the first part of the experiment. Never- prepared with Ti-based supported catalyst (1), ran-
theless, no propylene polymerization activity was dom ethylene-propylene copolymer prepared with
detected for over 30 min. Finally, the viability of the vanadium-based catalyst (2) and the difference spec-
catalyst system was confirmed when ethylene was rein- trum (3).

POLYMER SCIENCE Series A Vol. 50 No. 11 2008

1114 KISSIN, RISHINA

1
15

12
4

3 11
2 5 8 14
9 10
13
6

16 18 20 22
Time, min

Fig. 15. Gas chromatogram (C11–C12 range) of ethylene/propylene co-oligomer produced with TiCl4 /dibutyl phtha-
late/MgCl2-AlEt3 system at 100°C. Peak assignment: (1) n-C10; (2) 5-Me-C10; (3) 4-Me-C10; (4) 2-Me-C10;
(5) 3-Me-C10; (6) 2-Me-1-C10; (7) n-C11; (8) 2,6-Me2-C10; (9) 2,9-Me2-C10; (10) 2,8-Me2-C10; (11) 4-Me-C11;
(12) 2-Me-C11; (13) 2-Me-1-C11; (14) 1-C12; (15) n-C12 .

the catalyst system TiCl4/dibutyl phthalate/MgCl2– mixture is chosen to produce a random ethylene/propy-
AlEt3/PhSi(OEt)3 was used for propylene polymeriza- lene copolymer containing ~50–60 wt % of ethylene.
tion in the presence of hydrogen at 70°ë during the first The resulting total product is usually viewed as an inti-
stage. These examples show that the catalysts retained mate mixture of highly crystalline polypropylene and
high activity in ethylene polymerization after their pro- an amorphous ethylene/propylene copolymer. How-
pylene polymerization centers A–E were nearly com- ever, the previous data indicate that the catalyst parti-
pletely deactivated. cles exiting the first reactor, after they polymerized pro-
pylene for ~1 h, contain two groups of active centers.
Polymerization reactions under conditions similar The first group includes active centers A–E, which are
to those employed in the latter experiments are widely still capable of propylene polymerization as well as eth-
used in industry for the synthesis of impact-resistant ylene/propylene copolymerization. However, the cata-
polypropylene resins. These materials are manufac- lysts also contain active centers of the second type (cen-
tured in two-reactor processes with supported Ti-based ters IV and V in ethylene polymerization reactions),
catalysts. Highly isotactic polypropylene is produced in which are capable of ethylene polymerization only. The
the first reactor for 60–80 minutes and then the polymer latter centers are silent during the first reaction stage
mixture containing the active catalyst is transferred to (polypropylene synthesis) but, when exposed to ethyl-
another reactor and exposed to an ethylene/propylene ene at the second stage, they produce nearly linear,
gas mixture [67]. The composition of this monomer highly crystalline polyethylene rather than an amor-

POLYMER SCIENCE Series A Vol. 50 No. 11 2008

 KINETICS OF PROPYLENE AND ETHYLENE POLYMERIZATION REACTIONS 1115

phous ethylene/propylene copolymer. Therefore, the comparisons of reactivity between ethylene and α-ole-
combined product should contain three different types fins based on reaction yields: the latter strongly depend
of polymer materials: isotactic polypropylene, a ran- on the duration of the experiment.
dom ethylene/propylene copolymer, and linear polyeth-
ylene. (3) The third important difference between ethylene
and propylene polymerization reactions is related to the
Examination of the IR spectrum of such impact-
resistant polypropylene proves this assertion (Fig. 14). effect of hydrogen on the polymerization kinetics:
The spectrum contains numerous bands of isotactic hydrogen increases the catalyst activity in propylene
polypropylene (~80 wt % of the total product), such as polymerization reactions, but it depresses the catalyst
the bands at 899, 841, and 809 cm–1 in the figure. The activity in ethylene polymerization reactions with the
spectrum also contains a group of overlapping bands in same catalyst.
the 730–720 cm–1 range characteristic of ethylene units
[74, 75]. The second spectrum shown in Fig. 14 is that
of a random ethylene/propylene copolymer (58 mol % Hydrogen Effects
of ethylene) prepared with a homogeneous vanadium- in Ethylene Polymerization Reactions
based catalyst, which is essentially a single-center cat-
alyst [4]. After the spectrum of the random copolymer Several studies of homopolymerization and copoly-
was subtracted from the spectrum of the “ethylene/pro- merization kinetics with different Ti-based Ziegler–
pylene rubber component” in the impact-resistant Natta catalysts [16–21, 78] showed a complex nature of
polypropylene, the difference spectrum (insert to the ethylene polymerization reactions. There are two dom-
figure) contained two narrow bands at 731 and 720 cm–1 inant factors that determine the kinetic behavior of
typical of crystalline polyethylene in the orthorhombic these catalysts. As discussed in the first part of the
form [76]. Thus, the nominal ethylene/propylene review, the first factor is the existence of several types
copolymer prepared with a Ti-based supported catalyst of active centers in the catalysts. The centers differ in
is, in reality, a mixture of two types of macromolecules, the molecular weights of the polymer fractions that
a true random copolymer produced by centers A–E they produce, the ability to copolymerize different ole-
(centers I–III in ethylene polymerization) and linear fins, kinetic stability, etc.
polyethylene produced with centers IV and V. Judging
by the band absorbances, these two materials are The second factor affecting the kinetic behavior of
present in the mixture in a ~1 : 1 ratio. In [77] a compli- Ti-based catalysts is related to the chemistry of tita-
mentary example is given: GPC analysis of ethyl- nium-alkyl groups. The kinetic studies [16, 17, 78]
ene/propylene copolymers prepared with the strongly suggest that Ti-based active centers with alkyl
TiCl4/(Al2O3–SiO2)–AlEt3 catalyst suggests the forma- groups containing methyl groups in the β-position with
tion of two materials, a “polyethylene” fraction with a respect to the Ti atom are unusually unreactive in olefin
high molecular weight and a true copolymer.
insertion reactions. In the case of ethylene polymeriza-
tion reactions, it is the Ti–C2H5 group, which can be
KINETIC PECULIARITIES OF ETHYLENE viewed as a growing polymer chain with one ethylene
POLYMERIZATION REACTIONS unit [16, 17, 78]. Strong support for this hypothesis was
WITH ZIEGLER–NATTA CATALYSTS given by detailed ethylene polymerization studies with
As Figs. 1 and 9 demonstrate, the kinetics of ethyl- scandocene complexes [79]: the Cp 2* Sc –C2H5 species
ene polymerization reactions has several peculiarities is 14 times less reactive in an ethylene insertion com-
compared to that of propylene polymerization reac-
tions. Three of them are most obvious: pared to the Cp 2* Sc –n-C3H7 species. The stability of
(1) Ethylene has the highest reactivity compared to the Ti–C2H5 group can be a result of a relatively strong
all α-olefins in catalytic polymerization reactions. β-agostic interaction between the hydrogen atoms of its
Judging by the values of reactivity ratios in copolymer- β-methyl group and the Ti atom. This type of β-agostic
ization reactions [9, 10], ethylene is at least five to ten interaction was experimentally demonstrated for sev-
times more reactive than propylene. However, when a eral zirconocene complexes [80, 81], including
direct comparison of catalyst productivities or reaction (MeCp)2Zr(C2H5)(PMe3)–, and was proposed in exten-
kinetics is performed, ethylene never exhibits the level sive theoretical studies of the Ti(C2H5)(PH3)2Cl2H
of reactivity expected from the copolymerization data.
complex [82, 83].
(2) The rate of propylene homopolymerization reac-
tions rapidly decreases with time, whereas ethylene The ethylene polymerization mechanism, which
polymerization is a nearly stable reaction for a period of takes into account stability and lability of the Ti–C2H5
several hours. These data emphasize the unreliability of group in ethylene polymerization, is shown below.

POLYMER SCIENCE Series A Vol. 50 No. 11 2008

1116 KISSIN, RISHINA

–Polymer +H2

–C2H6
+H2
+nC2H4

Ti H +C2H4 Ti CH2 Ti C2H5 Ti (C2H4)n C2H5

H CH2

–Polymer
+C2H4
–C2H4

+α-alkene
Kinetic mechanism of ethylene polymerization reactions [16, 17, 78].

This mechanism differs from the standard ethylene ethylene polymerization reactions, ~70–75%, in a
polymerization mechanism [7–9] in three main features: “temporarily sleeping” state. Because the frequency of
(1) The stabilized Ti–C2H5 bond exists in equilib- the Ti–H bond formation (followed by generation of the
rium with the uncomplexed Ti–C2H5 bond capable of Ti–C2H5 group) is greatly increased in the presence of
ethylene insertion. hydrogen, addition of hydrogen always decreases eth-
(2) The Ti–C2H5 bond can undergo β-hydrogen ylene polymerization rates [16–18]. When, in ethyl-
elimination and produce the Ti–H bond. ene/α-olefin copolymerization reactions, an α-olefin
molecule is inserted into the Ti–H bond in the primary
(3) When any α-olefin molecule is inserted into the fashion and, thus, bypasses the stage of the stable Ti–
Ti–H bond, the reaction bypasses the stage of the stabi- C2H5 bond, the reaction rates increase compared to
lized Ti–C2H5 bond. For example, if the α-olefin in the
those in ethylene homopolymerization reactions under
last reaction in the scheme is 1-hexene, its insertion into the same conditions [16–18].
the Ti–H bond produces the Ti–C6H13 moiety that can
be viewed as a growing polymer chain containing three
ethylene units, Ti–(CH2–CH2)3–H.
THE HYDROGEN ACTIVATION EFFECT
These additions to the standard reaction scheme pro- IN PROPYLENE POLYMERIZATION REACTIONS
vide plausible explanations for most peculiarities of ethyl-
ene polymerization reactions with Ti-based catalysts [16, The hydrogen activation effect in propylene poly-
17, 78, 84]. The feature of ethylene polymerization reac- merization reactions is well known in the literature [54,
tions most pertinent to the present subject is the effect of 56, 85–96]. This activation is reversible: the removal of
hydrogen on the ethylene polymerization kinetics. hydrogen reduces the polymerization rate, and the addi-
The Ti–C2H5 group is formed in ethylene polymer- tion of hydrogen in the course of an established poly-
ization reactions each time after ethylene insertion into merization reaction results in immediate activation of
the Ti–H bond, as well as after each step of the chain the catalyst [54, 86, 88, 92, 95 96]. Two hypotheses can
transfer with the monomer. Modeling of kinetic esti- explain this effect, and both are based on the stability of
mates [84] suggests that the relative stability of the Ti– the Ti-alkyl species with a methyl group in the β-posi-
C2H5 group keeps a large fraction of active centers in tion with respect to the Ti atom.

POLYMER SCIENCE Series A Vol. 50 No. 11 2008

 KINETICS OF PROPYLENE AND ETHYLENE POLYMERIZATION REACTIONS 1117

The first hypothesis explaining the hydrogen activa- Reactivity of the Mt–CH(CH3)R (Mt = Ti, Zr) bond
tion effect focuses on the regioselectivity of Ti-based in the propylene insertion reaction is very low. In the
catalysts [93, 97–99]. Active centers of Ziegler–Natta absence of other chain transfer reactions, such a center
and metallocene catalysts are occasionally capable of remains inactive until the β–H elimination reaction
the secondary insertion (2,1-insertion) of a propylene takes place:
molecule into the metal-carbon bond, resulting in the
formation of a dormant active center with a polymer Ti–CH(CH 3 )–CH 2 –CH 2 –CH(CH 3 )–Polymer
chain bearing a methyl group in the β-position to the Ti (3)
atom: Ti–H + CH 3 –CH=CH –CH 2 –CH(CH 3 )–Polymer
Ti–CH 2 –CH(CH 3 )–Polymer + CH 3 –CH=CH 2
(2) However, this dormant center is reactivated by
Ti–CH(CH 3 )–CH 2 –CH 2 –CH(CH 3 )–Polymer hydrogen:

Ti–CH(CH 3 )–CH 2 –CH 2 –CH(CH 3 )–Polymer + H 2

(4)
Ti–H + CH 3 –CH 2 –CH 2 –CH 2 –CH(CH 3 )–Polymer

Indeed, formation of the n-butyl group in Reaction (4) conditions. Most Ti-based Ziegler–Natta catalysts do
was observed by 13C NMR in polymers prepared with not produce polymer chains with inverted monomer
different Ti-based catalysts [95–98, 100–102]. Such units: propylene polymers produced with such catalysts
chain ends are abundant in polypropylene prepared have negligible fractions of head-to-head links, as
with isospecific metallocene catalysts [100–106]. reviewed in reference [9]. This was confirmed once
The second type of a Ti-alkyl species with a methyl more in the study with a supported Ti-based catalyst
group in the β-position to the Ti atom is the Ti– [19]. A slow primary insertion reaction of propylene
CH(CH3)2 group. This group can be formed in two into the Ti–iso-C3H7 bond (leading to the 2,3-dimethyl
reactions: end group) was observed only for aspecific active cen-
The chain transfer reaction with propylene in which ters [96].
a propylene molecule is coordinated to the Ti atom in The dormant Ti–iso-C3H7 center either decomposes
the secondary orientation: by means of β-H elimination with the expulsion of pro-
Ti–CH2–CH(CH3)–Polymer + CH3–CH=CH2 pylene (cf. scheme) or it reacts with a cocatalyst. How-
(5) ever, if hydrogen is present, the Ti–iso-C3H7 group rap-
Ti–CH(CH3)2 + CH2=C(CH3)–Polymer idly reacts with it with the restoration of the Ti–H bond
Secondary propylene insertion into the Ti–H bond and propane generation:
(the latter is generated in the chain transfer reaction Ti–CH(CH3)2 + H2 Ti–H + CH3–CH2–CH3 (8)
with hydrogen):
Indeed, formation of propane and significant con-
Ti–H + CH3–CH=CH2 Ti–CH(CH3)2 (6) sumption of hydrogen in propylene polymerization
Reaction (6) proceeds in parallel with the com- reactions with some Ti-based catalysts are well known
monly accepted primary propylene insertion into the [54, 105]. In turn, the Ti–H bond formed in Reaction 8
same bond, can insert a propylene molecule in the primary fashion
Ti–H + CH2=CH–CH3 Ti–CH2–CH2–CH3 (7) and thus reactivates the polymerization center. This
Reaction (6) signifies that the Ti–H bond is less regi- reaction sequence; as well as Reactions 2–4, provides a
oselective in propylene insertion compared to the Ti–C reasonable explanation of the hydrogen activation
bond in chain growth reactions. effect in propylene polymerization reactions (Fig. 5).
The main assumption in the second kinetic mecha-
nism of the hydrogen activation effect is that the Ti– Identification of the Ti–CH(CH3)2 Group
iso-C3H7 species produced in Reactions 5 and 6 exhib-
its the same significant stabilization as the Ti–C2H5 As mentioned above, primary propylene insertion
species in the scheme. The Ti–iso-C3H7 species can be into the Ti–C bond in the Ti–CH(CH3)2 group gener-
viewed as a dormant active center with a polymer chain ated in Reactions 5 and 6,
containing one propylene unit in the secondary orienta-
tion, similarly to the Ti–CH(CH3)R group formed in Ti–CH(CH3)2 + CH2=CH–CH3
Reaction 2. The fate of the Ti–iso-C3H7 species (9)
depends on the nature of the catalyst and on reaction Ti–CH2–CH(CH3)–CH(CH3)2,

POLYMER SCIENCE Series A Vol. 50 No. 11 2008

1118 KISSIN, RISHINA

is very difficult, both due to stability of the Ti–iso-C3H7 Ti–CH2–CH(CH3)–(CH2–CH2)n–CH(CH3)2 + H2

species and for steric reasons. However, one can expect (13)
that ethylene insertion into the Ti–CH(CH3)2 bond, Ti–H + (CH3)2CH–(CH2–CH2)n–CH(CH3)2,
although relatively slow, is nevertheless possible:
The oligomer molecules that can serve as identifiers of
Ti–CH(CH3)2 + CH2=CH2 Reaction 6 are (CH3)2CH–(CH2–CH2)n–CH(CH3)2
(10) molecules, isoalkanes with even carbon atom numbers
Ti–CH2–CH2–CH(CH3)2 and with isopropyl groups at both chain ends. Unam-
biguous identification of such products in oligomer
This reaction regenerates the active center in a sense mixtures by GC is based on the studies of ethylene/pro-
that it makes it available for further chain growth reac- pylene co-oligomers produced with Ni-based catalysts
tions, either with ethylene or with propylene. Indeed, [19, 107–109].
addition of a small quantity of ethylene to an estab- Several ethylene/propylene copolymerization
lished propylene polymerization reaction results in an experiments were carried out with both δ-TiCl3 and
increase of the propylene consumption rate [54]. To TiCl4/dibutyl phthalate/MgCl2 catalysts activated with
prove the existence of Reactions 5, 6 and 10, one has to AlEt3 at temperatures ranging from 85 to 105°C at very
identify a particular “starting” end of a polymer chain high hydrogen/monomer ratios [19]. GC analysis of the
in ethylene/α-olefin copolymers. In the case of ethyl- liquid samples collected after the experiments showed
ene/propylene copolymerization, this starting chain end that they all contained significant amounts of oligo-
is Polymer–CH2–CH2–CH(CH3)2. When the copoly- mers. Nearly every peak in the chromatograms was
merization reactions are carried out under a high hydro- identified. Figure 15 shows part of the chromatogram of
gen concentration, this chain end is formed overwhelm- oligomers produced with the TiCl4/dibutyl phtha-
ingly in a sequence of steps starting with the formation late/MgCl2–AlEt3 system at 100°ë. Most GC peaks in
of the Ti–H bond (in chain transfer with hydrogen) fol- Fig. 15 are assigned to oligomer structures expected for
lowed by the secondary insertion of a propylene mole- the standard co-oligomerization scheme when the pre-
cule into the Ti–H bond (Reaction 6) and by ethylene dominant chain transfer reaction is that with hydrogen
insertion (Reaction 10). Identification of such chain [19] and the standard propylene insertion reaction into
ends cannot be achieved by studying 13C NMR spectra the Ti–H bond is the primary insertion (Reaction 7).
of the copolymers because they contain “finishing” They include hydrogenated ethylene homo-oligomers,
chain ends of the identical structure produced in two linear alkanes with even carbon atom numbers
consecutive reactions, primary propylene insertion into (n-decane and n-dodecane in the figure), respective
a growing polymer chain with the last ethylene unit, α-olefins (1-decene and 1-dodecene), hydrogenated
ethylene/propylene co-oligomers with one propylene
Ti–CH2–CH2–Polymer + CH2=CH–CH3 unit, linear and branched alkanes with odd and carbon
(11) atom numbers (n-undecane, 2-methyldecane), respec-
Ti–CH2–CH(CH3)–CH2–CH2–Polymer, tive vinylidene olefins (2-methyl-1-decene and
2-methyl-1-undecene), small amounts of hydrogenated
followed by the chain transfer reaction with hydrogen: ethylene/propylene co-oligomers with two propylene
units, etc. The presence of 2,9-dimethyldecane in the
Ti–CH2–CH(CH3)–CH2–CH2–Polymer + H2 ë12 range (Fig. 15), 2,7-dimethyloctane in the C10
(12) range, and 2,11-dimethyldodecane in the C14 range
Ti–H + (CH3)2CH–CH2–CH2–Polymer gives support to the reaction sequence that starts with
the formation of the Ti–H bond followed by the forma-
However, gas chromatography (GC) analysis of low tion of the Ti–CH(CH3)2 species in Reaction 6. Analy-
molecular weight fractions of the copolymers, ethyl- sis of the oligomers formed in the ethylene/1-butene
ene/propylene co-oligomers, affords differentiation and ethylene/4-methyl-1-butene copolymerization
between the starting chain ends formed in Reactions 5, reactions [19] provided additional GC-based proofs of
6 and 10 and the ending chain ends formed in Reactions the secondary insertion of α-olefins into the Ti–H bond.
11 and 15. Several such options are discussed in [19]. Relative yields of various ethylene/propylene oligo-
In the case when the ethylene/propylene copolymeriza- mers were used to evaluate the probability of secondary
tion reactions are carried out at relatively high temper- propylene insertion into the Ti–H bond (Reaction 6)
atures and at high hydrogen concentrations (mostly, in compared to its primary insertion (Reaction 7) [19].
order to increase the yields of oligomers), the most con- The results of this evaluation are given in Table 11 as
venient identification option is the following. If a co- the ratios of the rate constants for primary and second-
oligomer chain starts in Reaction 6, includes several ary insertions of a propylene molecule into the Ti–H
ethylene insertion steps, one primary propylene inser- bond (regioselectivity of the Ti–H bond). These estima-
tion step, and is released in chain transfer with hydro- tions show that regioselectivity of propylene insertion
gen, into the Ti–H bond is quite poor: the secondary propy-

POLYMER SCIENCE Series A Vol. 50 No. 11 2008

 KINETICS OF PROPYLENE AND ETHYLENE POLYMERIZATION REACTIONS 1119

Table 11. Estimation of regioselectivity of propylene inser- The proposed mechanism activation of propylene
tion reaction into Ti–H bond polymerization reactions in the presence of hydrogen
can also be applied to the explanation of several other
Regioselectivity,
Catalyst Temp., °C features of propylene polymerization reactions: their
kprimary /ksecondary
activation in the presence of AlR2H [110] (it leads to
δ-TiCl3 85 <2.6–3.2 the Ti–H species) and the activation effect of ethylene,
105 <3.7–4.0 which is due to Reaction 10 [54].
TiCl4 /DBP/MgCl2 95 <3.5–4.0
100 <2.2–3.0 REFERENCES
1. Y. V. Kissin, in Proceedings of International Sympo-
sium on Advances in Olefin, Cycloolefin and Diolefin
lene insertion at 80–100°ë for both Ti-based catalysts Polymerization, Lyon, 1992, p. 6.
is, at best, only three to four times less probable than the 2. V. V. Vickroy, H. Schneider, and R. F. Abbott, J. Appl.
primary insertion. Polym. Sci. 50, 551 (1993).
In formal kinetic terms, the proposed mechanism of 3. Y. V. Kissin, Makromol. Chem., Macromol. Symp. 66,
the hydrogen activation effect in propylene polymeriza- 83 (1993).
tion due to the more frequent formation of Ti–H bonds 4. Y. V. Kissin, J. Polym. Sci., Part A: Polym. Chem. 33,
is equivalent to an increase in the number of “working” 227 (1995).
active centers in a catalyst in the presence of hydrogen. 5. P. J. Flory, J. Am. Chem. Soc. 62, 1561 (1940).
In chemical terms, there is not a large difference 6. P. J. Flory, Principles of Polymer Chemistry (Cornell
between the two explanations of the hydrogen activa- Univ. Press, New York, 1953), p. 568.
tion effect in propylene polymerization reactions, those 7. Stereoregular Polymers and Stereospecific Polymeriza-
based on Reactions 5 and 6, and the commonly tions, Ed. by G. Natta and F. Danusso (Pergamon,
accepted mechanism, Reactions 2–4. In the case of Oxford, 1967), Vol. 1, p. 52.
Reactions 5 and 6, the hydrogen effect is due to a rela- 8. T. Keii, Kinetics of Ziegler–Natta Polymerization
tively high formation probability and stability of the Ti– (Kodansha, Tokyo, 1972).
CH(CH3)2 group whereas, in the case of Reactions 2–4, 9. Y. V. Kissin, Isospecific Olefin Polymerization with
the effect is due to the stability of a growing polymer Heterogeneous Ziegler–Natta Catalysts (Springer, New
chain with the last inverted monomer unit, Ti– York, 1985), Chap. 3, p. 221.
CH(CH3)CH2CH2CH(CH3)–Polymer. 10. B. A. Krentsel, Y. V. Kissin, V. I. Kleiner, and L. L. Sto-
tskaya, in Polymers and Copolymers of Higher α-Ole-
The relative significance of these two reaction fins (Hanser, Munich, 1997), Chap. 8, p. 243.
routes hinges on the differences in the formation prob- 11. K. Y. Choi and W. H. Ray, J. Macromol. Sci., Rev. Mac-
abilities of both types of dormant centers. The Ti– romol. Chem. Phys. 25, 57 (1985).
CH(CH3)2 group is formed in two reactions, chain 12. K. S. Minsker, M. M. Karpasas, and G. E. Zaikov,
transfer to a propylene molecule in the secondary orien- J. Macromol. Sci., Rev. Macromol. Chem. Phys. 27, 1
tation (Reaction 5) and secondary propylene insertion (1987).
into the Ti–H bond (Reaction 6). According to the data 13. J. J. A. Dusseault and C. C. Hsu, J. Macromol. Sci., Rev.
in Table 11, the probability of Reaction 6 leading to the Macromol. Chem. Phys. 33, 103 (1993).
Ti–CH(CH3)2 group is less than three to four times 14. T. E. Nowlin, Y. V. Kissin, and K. P. Wagner, J. Polym.
lower than the probability of primary propylene inser- Sci., Part A: Polym. Chem. 26, 755 (1988).
tion into the same bond leading to the Ti–CH2CH2CH3 15. Y. V. Kissin, Makromol. Chem., Macromol. Symp. 89,
group (Reaction 7). At the present time, there is no reli- 113 (1995).
able information about the relative rate of Reaction 5 16. Y. V. Kissin, R. I. Mink, T. E. Nowlin, and A. J. Brando-
vs. that of Reaction 6, but, anyway, Reaction 5 lini, Top. Catal. 7, 69 (1999).
increases the formation probability of the Ti–CH(CH3)2 17. Y. V. Kissin, R. I. Mink, T. E. Nowlin, and A. J. Brando-
group even further. On the other hand, the probability lini, in Metallorganic Catalysts for Synthesis and Poly-
of the secondary propylene insertion into the Ti–C bond merization, Ed. by W. Kaminsky (Springer, Berlin,
in a growing polymer chain (Reaction 2) is quite low. 1999), p. 60.
The frequency of this event can be approximately esti- 18. Y. V. Kissin, R. I. Mink, and T. E. Nowlin, J. Polym.
mated from the content of n-butyl groups in polypropy- Sci., Part A: Polym. Chem. 37, 4255 (1999).
lene chains that are formed in Reaction 4 [95, 97, 98, 19. Y. V. Kissin and L. A. Rishina, J. Polym. Sci., Part A:
100–102]. Available 13C NMR data for the δ-TiCl3- Polym. Chem. 40, 1353 (2002).
based catalyst system [102] show that the n-butyl 20. Y. V. Kissin, L. A. Rishina, and E. I. Vizen, J. Polym.
groups are observable only in fractions of polypropy- Sci., Part A: Polym. Chem. 40, 1899 (2002).
lene generated by centers of a reduced isospecificity 21. Y. V. Kissin, J. Polym. Sci., Part A: Polym. Chem. 39,
and are almost completely absent from the highly iso- 1681 (2001).
tactic fraction. 22. L. Wild, Adv. Polym. Sci. 98, 1 (1991).

POLYMER SCIENCE Series A Vol. 50 No. 11 2008

1120 KISSIN, RISHINA

23. L. Wild, T. R. Ryle, D. C. Knobeloch, and I. R. Peat, 51. M. Harkonen, J. V. Seppala, R. Chujo, and Y. Kogure,
J. Polym. Sci., Part B: Polym. Phys. 20, 44 (1982). Polymer 36, 1499 (1995).
24. M. Kakugo, T. Miyatake, Y. Naito, and K. Mizunuma, 52. Y. V. Kissin, J. Polym. Sci., Part A: Polym. Chem. 41,
in Transition Metal Catalyzed Polymerization: Ziegler– 1745 (2003).
Natta and Metathesis Polymerization, Ed. by R. P. Quirk 53. R. Spitz, C. Bobichon, and A. Guyot, Makromol. Chem.
(Cambridge Univ. Press, Cambridge, 1988), p. 624. 190, 707 (1989).
25. F. Defoor, G. Croeninckx, P. Schouterdan, and B. Van 54. R. Spitz, P. Masson, C. Bobichon, and A. Guyot, Mak-
der Heijden, Polymer 33, 3878 (1992). romol. Chem. 190, 717 (1989).
26. J. B. P. Soares and A. E. Hamielec, J. Appl. Polym. Sci. 55. G. Natta and L. Pasquon, Adv. Catal. 11, 1 (1959).
36, 1639 (1995).
56. G. Guastalla and U. Giannini, Makromol. Chem., Rapid
27. I. Mingozzi, G. Cecchin, and G. Morini, Int. J. Polym. Commun. 4, 519 (1983).
Anal. Charact. 3, 293 (1997).
57. M. Terano, M. Yamahiro, and H. Mori, in Metallorganic
28. M. Zhan, D. T. Lynch, and S. E. Wanke, J. Appl. Polym. Catalysts for Synthesis and Polymerization, Ed. by
Sci. 75, 960 (2000). W. Kaminsky (Springer, Berlin, 1999), p. 658.
29. F. M. Mirabella, J. Polym. Sci., Part B: Polym. Phys. 39, 58. H. Mori, T. Iizuka, K. Tashino, and M. Terano, Macro-
2800 (2001). mol. Chem. Phys. 197, 895 (1996).
30. J. C. Chadwick, G. Morini, G. Balbontin, et al., Macro- 59. H. Mori, T. Iizuka, K. Tashino, and M. Terano, Macro-
mol. Chem. Phys. 202, 1995 (2001). mol. Chem. Phys. 198, 2499 (1997).
31. J. C. Chadwick, Macromol. Symp. 173, 21 (2001). 60. K. Soga and T. Shiono, Polym. Bull. (Berlin) 8, 261
32. Y. V. Kissin, F. M. Mirabella, and C. C. Meverden, (1982).
J. Polym. Sci., Part A: Polym. Chem. 43, 4351 (2005). 61. E. I. Vizen, L. A. Rishina, L. N. Sosnovskaya, et al., Eur.
33. B. J. Monrabal, J. Appl. Polym. Sci. 52, 491 (1994). Polym. J. 30, 1315 (1994).
34. J. B. P. Soares and A. E. Hamiliec, Macromol. Theory 62. E. I. Vizen and Yu. V. Kissin, Vysokomol. Soedin.,
Simul. 4, 305 (1995). Ser. A 11, 1774 (1969).
35. B. Monrabal, Macromol. Symp. 110, 81 (1996). 63. K. Soga, R. Ohnishi, and T. Sano, Polym. Bull. (Berlin)
36. C. De Rosa, F. Auriemma, C. Spera, et al., Macromole- 7, 547 (1982).
cules 37, 1441 (2004). 64. Y. V. Kissin, Isospecific Olefin Polymerization with Het-
37. Y. V. Kissin and H. A. Fruitwala, J. Appl. Polym. Sci. erogeneous Ziegler–Natta Catalysts (Springer, New
106, 3872 (2007). York, 1985), Chap. 4, p. 360.
38. H. Hosoda, Polym. J. (Tokyo) 20, 383 (1988). 65. R. Spitz, P. Masson, C. Bobichon, and A. Guyot, Mak-
romol. Chem. 169, 1043 (1988).
39. P. M. Kamath and L. Wild, Polym. Eng. Sci. 6, 213
(1966). 66. D. C. Calabro and F. Y. Lo, in Transition Metal Cata-
lyzed Polymerization, Ziegler–Natta and Metathesis
40. J. Xu, L. Feng, S. Yang, et al., Eur. Polym. J. 34, 431 Polymerization, Ed. by R. P. Quirk (Cambridge Univ.
(1998). Press, New York, 1988), p. 729.
41. B. Liu, T. Nitta, H. Nakatani, and M. Terano, Macro- 67. J. A. Debling and W. H. Ray, J. Appl. Polym. Sci. 81,
mol. Chem. Phys. 204, 395 (2003). 3085 (2001).
42. G. Morini, E. Albizatti, G. Balbontin, et al., Macromol- 68. G. C. Han-Adebekun, M. Hamba, and W. H. Ray,
ecules 29, 5770 (1996). J. Polym. Sci., Part A: Polym. Chem. 35, 2063 (1997).
43. P. C. Barbe, L. Noristi, and M. A. Schexnayder, in 69. N. M. Gultseva, T. M. Ushakova, A. M. Aladyshev,
Advances in Polyolefins, Ed. by T. C. Cheng and B. Sey- et al., Polym. Bull. (Berlin) 29, 639 (1992).
mour (Plenum, New York, 1987), p. 295.
70. J.-G. Wang, W.-B. Zhang, and B.-T. Huang, Makromol.
44. K. Soga and T. Shiono, in Transition Metal Catalyzed Chem., Macromol. Symp. 83, 245 (1992).
Polymerizations: Ziegler–Natta and Metathesis Poly-
merizations, Ed. by R. P. Quirk (Cambridge Univ. Press, 71. V. Pasquet and R. Spitz, Makromol. Chem. 194, 451
Cambridge, 1988), p. 266. (1993).
45. Y. Hu and J. C. W. Chien, J. Polym. Sci., Part A: Polym. 72. B. K. Vindstad, K. A. Solli, and M. Ystenes, Makromol.
Chem. 26, 2003 (1988). Chem., Rapid Commun. 13, 471 (1992).
46. Y. V. Kissin, R. Ohnishi, and T. Konakazawa, Macro- 73. T. S. Wester and M. Ystenes, Macromol. Chem. Phys.
mol. Chem. Phys. 205, 284 (2004). 198, 1623 (1997).
47. Y. V. Kissin, J. C. Chadwick, I. Mingozzi, and G. Morini, 74. G. Natta, A. Valvassori, F. Ciampelli, and G. J. Maz-
Macromol. Chem. Phys. 207, 1344 (2006). zanti, J. Polym. Sci., Part A-1 3, 1 (1965).
48. E. Albizzati, U. Giannini, G. Collina, et al., in Polypro- 75. Y. V. Kissin, Adv. Polym. Sci. 15, 91 (1974).
pylene Handbook, Ed. by E. P. Moore (Hanser, New 76. S. Krimm, Adv. Polym. Sci. 2, 51 (1960).
York, 1996), Chap. 2, p. 11. 77. A. G. Rodionov, N. M. Domareva, A. A. Baulin, et al.,
49. J. V. Seppala, M. Harkonen, and L. Luciani, Makromol. Vysokomol. Soedin., Ser. A 23, 1560 (1981).
Chem. 190, 2535 (1989). 78. Y. V. Kissin, R. I. Mink, T. E. Nowlin, and A. J. Brando-
50. A. Proto, L. Oliva, C. Pellecchia, et al., Macromolecules lini, J. Polym. Sci., Part A: Polym. Chem. 37, 4281
23, 2904 (1990). (1999).

POLYMER SCIENCE Series A Vol. 50 No. 11 2008

 KINETICS OF PROPYLENE AND ETHYLENE POLYMERIZATION REACTIONS 1121

79. B. J. Burger, M. F. Thompson, W. D. Cotter, and 95. G. D. Bukatov, V. S. Goncharov, and V. A. Zakharov,
J. E. Bercaw, J. Am. Chem. Soc. 112, 1566 (1990). Macromol. Chem. Phys. 196, 1751 (1995).
80. R. F. Jordan, P. K. Bradley, N. C. Baenziger, and 96. J. C. Chadwick, Macromol. Symp. 172, 21 (2001).
R. La Pointe, J. Am. Chem. Soc. 112, 1289 (1990). 97. J. C. Chadwick, A. Miedema, and O. Sudmeijer, Mac-
81. R. F. Jordan, R. La Pointe, P. K. Bradley, and romol. Chem. Phys. 195, 167 (1994).
N. C. Baenziger, Organometallics 8, 2892 (1989). 98. S. Kojoh, M. Kioka, N. Kashiwa, et al., Polymer 36,
82. S. Obara, N. Koga, and K. Morokuma, J. Am. Chem. 5015 (1995).
Soc. 106, 4625 (1984). 99. H. Mori, K. Tashino, and M. Terano, Macromol. Chem.
83. S. Obara, N. Koga, and K. Morokuma, J. Organomet. Phys. 196, 651 (1995).
Chem. 270, C33 (1984).
100. K. Imaoka, S. Ikai, M. Tamura, et al., J. Mol. Catal. 82,
84. Y. V. Kissin, Macromol. Theory Simul. 11, 67 (2002). 37 (1993).
85. G. D. Bukatov, V. S. Goncharov, V. A. Zakharov, et al., 101. V. Busico, R. Cipullo, and P. Corradini, Makromol.
Kinet. Katal. 35, 392 (1994). Chem., Rapid Commun. 12, 15 (1992).
86. A. M. Aladyshev, O. R. Isichenko, P. M. Nedorezova, 102. T. Hayashi, Y. Inoue, R. Chujo, and T. Asakura, Macro-
et al., Vysokomol. Soedin., Ser. A 33, 1708 (1991). molecules 21, 2675 (1988).
87. A. Guyot, R. Spitz, and D. Journaud, Stud. Surf. Sci.
Catal. 89, 43 (1994). 103. A. Carvill, I. Tritto, P. Locatelli, and M. C. Sacchi, Mac-
romolecules 30, 7056 (1997).
88. L. A. Rishina, E. I. Vizen, L. N. Sosnovskaya, and
F. S. Dyachkovsky, Eur. Polym. J. 30, 1309 (1994). 104. A. Carvill, L. Zetta, G. Zannoni, and M. C. Sacchi,
Macromolecules 31, 3783 (1998).
89. J. J. C. Samson, P. J. Bosman, G. Weickert, and
K. R. Westerterp, J. Polym. Sci., Part A: Polym. Chem. 105. J. C. Randall, C. J. Ruff, J. C. Vizzini, et al., in Metal-
37, 219 (1999). lorganic Catalysts for Synthesis and Polymerization,
Ed. by W. Kaminsky (Springer, Berlin, 1999), p. 601.
90. I. W. Parsons and T. M. Al-Turki, Polym. Commun. 30,
72 (1989). 106. G. Moscardi, F. Piemontezi, and L. Resconi, Organo-
91. J. B. P. Soares and A. E. Hamielec, Polymer 37, 4607 metallics 18, 5264 (1999).
(1996). 107. Y. V. Kissin and G. P. Feulmer, J. Chromatogr. Sci. 24,
92. M. Kioka and N. Kashiwa, J. Macromol. Sci., Chem. 53 (1986).
28, 865 (1991). 108. Y. V. Kissin, J. Chromatogr. Sci. 24, 278 (1986).
93. J. C. Chadwick, G. Morini, E. Albizzatti, et al., Macro- 109. Y. V. Kissin, J. Polym. Sci., Part A: Polym. Chem. 27,
mol. Chem. Phys. 197, 2501 (1996). 605, 623 (1989).
94. J. C. W. Chien and T. Nozaki, J. Polym. Sci., Part A: 110. L. A. Rishina, E. I. Vizen, L. N. Sosnovskaya, and
Polym. Chem. 9, 505 (1991). I. L. Dubnikova, Kinet. Katal. 37, 421 (1996).

POLYMER SCIENCE Series A Vol. 50 No. 11 2008

View publication stats