Macromolecules 2002, 35, 349-354 349

Metallocene-Catalyzed Propene Polymerization: From Microstructure to

Kinetics. 1. C2-Symmetric ansa-Metallocenes and the "Trigger"
Hypothesis

Vincenzo Busico,* Roberta Cipullo, Francesco Cutillo, and Michele Vacatello

Dipartimento di Chimica, Università di Napoli Federico II, Via Cintia, 80126 Naples, Italy
Received August 6, 2001; Revised Manuscript Received October 16, 2001

ABSTRACT: A microstructural method for the determination of the reaction order with respect to the
monomer in coordinated 1-alkene polymerizations is proposed. The method is applicable whenever
enantioselective monomer polyinsertion competes with an intramolecular process introducing stereodefects
of known and recognizable nature in the growing polymer chains at a rate independent of monomer
concentration, [M] (e.g., chain or site epimerization for C2-symmetric and Cs-symmetric ansa-metallocene
catalysts, respectively). Indeed, in such cases, the stereoregularity of the polymer obtained at a given
temperature results from the balance between the rates of the two competing processes, and is obviously
a function of [M], because so is the rate of polyinsertion, 〈vp〉. Therefore, from the way polymer
stereoregularity changes with [M], the corresponding change of 〈vp〉 can be inferred. Compared with the
conventional macroscopic approach based on measurements of monomer consumption, the microstructural
route is more reproducible, and allows one to discriminate between “authentic” and apparent higher-
order polyinsertion rate laws. In this paper, in particular, we show that the “trigger” hypothesis invoked
to explain the observed pseudo-second-order kinetics of propene polymerization promoted by C2-symmetric
ansa-metallocene catalysts is inconsistent with the experimental results and that an alternative
explanation assuming first-order kinetics at catalytic species which interconvert between a propagating
and a “resting” state is more plausible.

Introduction (viii) Catalyst productivity may depend, inter alia, on

The kinetics of Ziegler-Natta and related 1-alkene the concentration of the cocatalyst (usually an Al-alkyl
polymerizations1,2 is still poorly understood, for a dis- species also used as a scavenger). For methylalumoxane-
coragingly high number of reasons: activated metallocene catalysts, in particular, this
(i) The catalytic species are extremely reactive, and dependence can be dramatic (for reasons still waiting
easily deactivated by a variety of ubiquitous contami- to be clarified).2
nants, including water and oxygen. All this results in the fact that the so-called polym-
(ii) As a rule, and also due to point i, only a fraction erization rate, vp (i.e., the rate of monomer consumption)
of the transition metal species used as catalyst precursor is a macroscopic parameter which, in addition to being
becomes involved in catalysis, and all approaches pro- poorly reproducible and time-dependent, represents
posed to now for the determination of active site almost invariably a complicated average over a wide
concentrations (such as quenched-flow3-5 and radio- distribution of active sites with unknown absolute and
tagging6,7 techniques) have serious drawbacks which relative concentrations and as such is hard to interpret
make them potentially flawed. in molecular terms. Therefore, it is not surprising that
(iii) Most catalysts show a decay of activity with time, most fundamental questions on the polymerization
which is often extensive and fast.1,2a kinetics, including the nature of the rate-limiting step
(iv) The polymerization process is highly exothermic, and the molecularity of monomer insertion, are still
which easily results in a poor temperature control, unanswered.
particularly with heterogeneous systems. One of the most puzzling observations concerns the
(v) Heterogeneous catalysts always have several types apparent reaction order with respect to the monomer.
of active species with different productivities and se- For a long time after the initial discovery of TiCl3-based
lectivities.1 Homogeneous catalysts, in turn, may well catalysts, it was found (assumed?) that in the rate law
have one single active species, but only in a few special
cases are the two coordination sites involved in the 〈vp〉 ) -d[M]/dt ) 〈kp〉Cp*[M]R (1)
reaction equivalent.2
(vi) With the only exception of ethene, a 1-alkene (where [M] and Cp* are the concentrations of monomer
monomer can insert in a growing chain in (at least) four and propagating chains, respectively) the value of the
different modes, which results in an equal number of R-exponent issreasonablys1.1,13 Later observations,
metal-alkyl species widely differing in reactivity.8,9 however, pointed out that this is not always the case.
(vii) A variety of catalytic processes (including several In the mid-1980s, for instance, Pino et al. measured
pathways of chain transfer1,2 and a number of isomer- R values up to 1.3 for propene polymerization with a
izations of the growing polymer chain2,10,11 and/or of the number of MgCl2/TiCl4 supported “high-yield” cata-
catalytic species2,12) compete with monomer polyinser- lysts.14 Also moving from these findings, Ystenes
tion. proposed a mechanism according to which monomer
insertion would be “triggered” by a second monomer
* Corresponding author. E-mail: [email protected]. molecule.15 Much more recently, values of R close to 2
10.1021/ma011401h CCC: $22.00 © 2002 American Chemical Society
Published on Web 12/12/2001
350 Busico et al. Macromolecules, Vol. 35, No. 2, 2002

have actually been reported for some of the new, well- Substituting Cp* from eq 3 in eq 2 gives
defined homogeneous metallocene catalysts.16 This
prompted Prosenc et al. to investigate the “trigger” 〈vp〉 ) 〈kp〉Cp*[M] ) 〈kp〉C*[M]2/{[M] + 〈kt,H〉/〈ki,H〉 +
hypothesis theoretically for a model metallocene cation,
and their DFT calculations did provide some basis for + 〈kt,M〉[M]/〈ki,M〉 + 〈kt,Al〉[Al]0.5/〈ki,Al〉} (4)
it, although with a number of caveats.17
On the other hand, less exotic explanations are well It is easy to realize that eq 4 simulates a second-order
possible. In particular, it should be noted that eq 1 chain propagation law when
assumes that the (inevitable) occurrence of chain initia-
tion and termination (transfer) processes has no effect ([M] + 〈kt,M〉[M]/〈ki,M〉) , (〈kt,H〉/〈ki,H〉 +
on the propagation rate, whichshoweverscannot be 〈kt,Al〉[Al]0.5/〈ki,Al〉) (5)
taken for granted. According to Tait,18 in addition to
chain propagation (assumed to obey a first-order rate whereas it reduces to a first-order kinetics when
law)
([M] + 〈kt,M〉[M]/〈ki,M〉) . (〈kt,H〉/〈ki,H〉 +
〈vp〉 ) -d[M]/dt ) 〈kp〉Cp*[M] (2)
〈kt,Al〉[Al]0.5/〈ki,Al〉) (6)
the following reactions should also be considered:
•Chain transfer: Intermediate conditions obviously result in apparent
reaction orders between 1 and 2.
〈vt,H〉 ) 〈kt,H〉Cp* (intramolecular β-H elimination) Conceptually similar kinetic schemes have been sub-
sequently proposed, e.g. by Richardson et al.,19 and more
〈vt,M〉 ) 〈kt,M〉Cp*[M] recently by Resconi et al.20 The latter authors, in
(monomer-assisted β-H elimination) particular, postulated the existence of a catalyst “rest-
ing” state, not (necessarily) related with chain transfer
〈vt,Al〉 ) 〈kt,Al〉Cp*[Al]0.5 and (re-) initiation, which would form from the propa-
(chain transfer to the Al-alkyl) gating state at a rate, 〈vf,rest〉, independent of [M] but
would then be returned to it by (slow) monomer inser-
•Chain initiation: tion:

〈vi,H〉 ) 〈ki,H〉Ci,H*[M] (into an initial Mt-H bond) 〈vf,rest〉 ) 〈kf,rest〉Cp*

〈vi,M〉 ) 〈ki,M〉Ci,M*[M] 〈vd,rest〉 ) 〈kd,rest〉Crest*[M]
(into an initial Mt-CnH2n-H bond)
This means that an additional, and possibly dominating,
〈vi,Al〉 ) 〈ki,Al〉Ci,Al*[M] 〈kf,rest〉/〈kd,rest〉 term should be added into eqs 3 and 4:
(into an initial Mt-alkyl bond)
Cp* ) C*/{1 + 〈kt,H〉/(〈ki,H〉[M]) + 〈kt,M〉/〈ki,M〉 +
Set + 〈kt,Al〉[Al]0.5/(〈ki,Al〉[M]) + 〈kf,rest〉/(〈kd,rest〉[M])} (3′)
C* ) Cp* + ∑jCi,j* 〈vp〉 ) 〈kp〉Cp*[M] ) 〈kp〉C*[M]2/{[M] + 〈kt,H〉/〈ki,H〉 +
(where C* is the total concentration of transition metal + 〈kt,M〉[M]/〈ki,M〉 + 〈kt,Al〉[Al]0.5/〈ki,Al〉 +
species that can have catalytic activity, Cp* the concen- 〈kf,rest〉/〈kd,rest〉} (4′)
tration of propagating ones, and Ci,j* that of transition
metal species originating from the jth chain transfer which makes the condition for (apparent) second-order
process and waiting for the first monomer insertion in kinetics (eq 5′ below)
order to initiate a new chain), the stationary-state
condition on Cp* ([M] + 〈kt,M〉[M]/〈ki,M〉) ,
d(Cp*)/dt ) 0 (〈kt,H〉/〈ki,H〉 + 〈kt,Al〉[Al]0.5/〈ki,Al〉 +
+ 〈kf,rest〉/〈kd,rest〉) (5′)
requires that
more likely to be matched (compared with eq 5). For
∑j〈vi,j〉 ) ∑j〈vt,j〉 propene polymerization promoted by metallocene cata-
lysts, it was tentatively proposed20b that the “resting”
and (plausibly) that species be an R-branched Mt-polymeryl bond formed
by growing chain isomerization,10 although very many
〈vi,H〉 ) 〈ki,H〉Ci,H*[M] ) 〈vt,H〉 ) 〈kt,H〉Cp* alternatives can be conceived as well.
〈vi,M〉 ) 〈ki,M〉Ci,M*[M] ) 〈vt,M〉 ) 〈kt,M〉Cp*[M] Even more recently, Schaper et al. analyzed propene
polymerization data for a number of ansa-zirconocene
〈vi,Al〉 ) 〈ki,Al〉Ci,Al*[M] ) 〈vt,Al〉 ) 〈kt,Al〉Cp*[Al]0.5 catalysts making use of a genetic algoritm21 and con-
cluded not only that one or more catalyst “resting” states
From this, one can immediately derive need to be included in the reaction scheme in order to
reproduce the experimental rate law but also that the
Cp* ) C*/{1 + 〈kt,H〉/(〈ki,H〉[M]) + possibility of monomer “trigger” cannot be ruled out.
To solve the impasse, we propose a microstructural
〈kt,M〉/〈ki,M〉 + 〈kt,Al〉[Al]0.5/(〈ki,Al〉[M])} (3) approach to the problem, applicable to cases in which
Macromolecules, Vol. 35, No. 2, 2002 Metallocene-Catalyzed Propene Polymerization 351

Scheme 1

enantioselective 1-alkene polyinsertion competes with Results and Discussion

an intramolecular process introducing stereodefects of For this investigation, we chose two well-known
known and recognizable nature in the growing polymer isotactic-selective Si-bridged zirconocenes with C2-
chains, at a rate independent of monomer concentration. symmetrysnamely, rac-Me2Si(2-Me-1-indenyl)ZrCl224
For propene polyinsertion, such a process can be, e.g., (I) and rac-Me2Si(2-Me-benz[e]-1-indenyl)ZrCl216b (II)s
chain epimerization, which generates rr-type stereode- among the very few for which reliable investigations of
fects in the isotactic chains growing at C2-symmetric catalyst productivity as a function of propene concentra-
ansa-zirconocene species,10 or site epimerization at Cs- tion have been reported in the literature, and an
symmetric ansa-zirconocene species, originating m-type apparent reaction order R close to 2 has been deter-
stereodefects in growing syndiotactic chains.2,12 At each mined. For I/MAO (MAO ) methylalumoxane), in
given temperature, the stereoregularity of the polymer particular, Wester et al. measured R ) 1.7 at 80 °C and
obtained results from the balance between the rates of 1.0 bar e p(C3H6) e 2.5 bar;16a according to Stehling et
the two competing processes and is obviously a function al., for II/MAO, it is also R ) 1.7 at 40 °C and 1.0 bar e
of [M], because so is the rate of polyinsertion. Therefore, p(C3H6) e 5.0 bar.16b
from the way polymer stereoregularity changes with For all known bis(indenyl) C2-symmetric ansa-zir-
[M], the corresponding change of 〈vp〉 can be inferred. conocene catalysts, the process of isotactic propene
In our opinion, this approach is highly advantageous, polyinsertion competes with a concurrent reaction of
compared with more conventional ones, for two main epimerization of the growing chain, which consists of
reasons: the racemization of a last-inserted monomeric unit, most
probably according to the mechanism shown in Scheme
Discrimination between an authentic and an apparent
1.10 This generates rr-type stereodefects, which sum up
higher-order polyinsertion rate law is possible. Indeed,
to those deriving from monomer insertion with the
if the “trigger” hypothesis15 is correct, one has to expect
“wrong” enantioface.
that polymer stereoregularity will be a function of [M]R,
The rate of chain epimerization, 〈vepim〉 is given by10f
whereas a linear dependence should always be observed,
also when R > 1 in eq 1, in case the (apparent) higher-
〈vepim〉 ) 〈kepim〉Cp* (7)
order kinetics has to be traced to the co-presence of
“resting” and propagating transition metal species (eqs
3′-5′).18,20 Therefore, if chain transfer and (re-) initiation processes
are not considered (which is licit when the analysis is
Determining the degree of stereoregularity of a poly- restricted to inner chain microstructure), the probability
(1-alkene) sample by means of 13C NMR22 is much more that, under certain experimental conditions, a given
reliable and reproducible than measuring a polymeri- monomeric unit was not subjected to epimerization can
zation rate from monomer consumption. In particular, be expressed as10f
properly run independent quantitative spectra of a given
polypropylene sample are virtually superimposable,
Pins ) 〈kins〉[M]β/(〈kins〉[M]β + 〈kepim〉) ) [M]β/([M]β +
once the accumulation time is long enough. Moreover,
catalyst deactivation phenomena, or possible changes 〈kepim〉/〈kins〉) (8)
of active site concentration (C*) with changing [M], are
not expected to be influent on the microstructure of the where β is the “true” reaction order of monomer inser-
polymer produced.10f tion, intended as an elementary process (which explains
In the present paper, we report the results of the why, in eq 8, we used the symbols 〈kins〉 and β, rather
proposed method, as applied to isotactic propene po- than 〈kp〉 and R of eq 1).
lymerization promoted by two representative C2-sym- The probability that both monomeric units of a given
metric ansa-zirconocenes, for which it had been reported diad did not undergo epimerization is obviously Pins2.
that R is close to 2 in eq 1.16 In a subsequent paper, we From this, it follows immediately that the fraction of
will present similar results for syndiotactic propene meso diads in the polymer, [m] (far from the chain ends),
polymerization in the presence of a Cs-symmetric ansa- depends on [M] according to the following simple
zirconocene catalyst; in this second case, we will show expression10f
how the microstructural approach can also be used to
measure differences in the activation parameters for the [m] ) [mo]Pins2 + 0.5(1-Pins2) )
catalytic processes in competition (in the considered ([mo] - 0.5)Pins2 + 0.5 )
case, polyinsertion and site epimerization), with impor-
tant mechanistic implications.23 ) ([mo] - 0.5){[M]β/([M]β + 〈kepim〉/〈kins〉)}2 + 0.5 (9)
352 Busico et al. Macromolecules, Vol. 35, No. 2, 2002

Table 1. 13C NMR Fraction of meso Diads, [m], as a Table 2. 13C NMR Fraction of meso Diads, [m], as a
Function of Propene Concentration, [C3H6], for Function of Propene Concentration, [C3H6], for
Polypropylene Samples Prepared at 80 °C in the Polypropylene Samples Prepared at 80 °C in the
Presence of Catalyst System I/MAOa,b Presence of Catalyst System II/MAOa,b
sample no. [C3H6] (M) [m] sample no. [C3H6] (M) [m]
1 0.062(6) 0.72(1) 1 0.050(17) 0.82(2)
2 0.098(6) 0.79 (1) 2 0.101(17) 0.89(2)
3 0.118(6) 0.78(1) 3 0.165(6) 0.91(1)
4 0.151(8) 0.82(1) 4 0.221(17) 0.92(1)
5 0.193(8) 0.85(2) 5 0.28(3) 0.924(2)
6 0.34(3) 0.91(1) 6 0.39(3) 0.940(6)
7 0.45(3) 0.908(4) 7 0.56(3) 0.953(2)
8 0.67(3) 0.931(6) 8 0.56(3) 0.95(1)
9 1.48(3) 0.949(6) 9 0.67(3) 0.960(6)
10 2.24(3) 0.956(9) 10 1.37(6) 0.962(4)
11 2.41(3) 0.960(4) 11 2.35(6) 0.970(4)
12 3.72(17) 0.959(8) a [Zr] ) (1-5) × 10-6 M; [Al]/[Zr] ) 1.5 × 104. b Experimental
a [Zr] ) 1.0 × 10-6 to 5.0 × 10-5 M; [Al]/[Zr] ) (3-20) × 103. uncertainty on last significant digit(s) in parentheses.
b Experimental uncertainty on last significant digit(s) in paren-

theses.

where [mo] is the limiting value of [m] for Pins f 1 (i.e.,

under conditions in which the rate of chain epimeriza-
tion is negligible compared with that of monomer
insertion, and polymer stereoregularity is only depend-
ent on catalyst enantioselectivity).
Therefore, from 13C NMR measurements of [m] on a
series of polymers prepared with a certain catalyst at
variable [M] (under otherwise identical experimental
conditions), one can calculate in terms of eq 9 the best-
fitting values of the adjustable parameters [m°] (if not
known independently), 〈kepim〉/〈kins〉, and β.
This is what we did for the two investigated zircono-
cenes. We chose to operate at 80 °C, a temperature at
which the effect of growing chain epimerization on the
stereoselectivity is appreciable up to comparatively high
[M]. We explored the whole range of [M] from virtually
0 (actually, 0.03 mol/L in toluene, below which we were
unable to achieve an accurate propene pressure control)
up to 3-4 mol/L (above which our temperature control
became problematic). Upon doing this, we found it
impossible to keep a number of experimental param-
eters at fixed values, as would have been desirable. In
particular, the concentrations of Zr and Al, as well as
the polymerization time, had to be adjusted in order to
result in polymerization rates compatible with a tem-
perature control within (0.5 °C and in enough product
for the microstructural characterization. In conse-
quence, we could not make use of polymerization yields
to extract trends of average catalyst productivity vs [M]
(which was however out of our scope). On the other
hand, in previous investigations10a.f,g we had already
noted that, at least in the case of C2-symmetric ansa- Figure 1. Fraction of meso diads, [m], vs propene concentra-
zirconocene catalysts, polymer stereoregularity at fixed tion, [C3H6], for polypropylene samples prepared at 80 °C in
[M] is practically insensitive to changes in [Zr] and [Al] the presence of catalyst system I/MAO: (top) complete plot;
over wide ranges (5 × 10-6 M e [Zr] e 5 × 10-4 M; 0.025 (bottom) expansion in the region 0-1 M. Continuous and
M e [Al] e 0.25 M), even when catalyst productivity dashed best-fitting curves correspond to minimization under
cases a and b, respectively (see text).
strongly depends on them. This finding is not trivial,
and suggests that the said dependence should be traced (continuous curve); (b) only [m°] and 〈kepim〉/〈kins〉 were
to Cp*, rather than to 〈kp〉 (〈kins〉); preliminary quenched- optimized, whereas β was imposed to coincide with the
flow investigations performed in our laboratory seem literature value of the apparent reaction order R in eq
to be consistent with such interpretation.25 1.
Tables 1 and 2 list the polymerization runs carried We are well aware that the comparison is delicate and
out for the two catalyst systems, along with the [m] that appreciable differences between the two fits can
values measured by 13C NMR on the polymers obtained. be expected mainly at low [M], where the function is
The corresponding plots of [m] vs [M] are given in steeper (and more stringent is the need for an extremely
Figures 1 and 2. In each plot, two best-fitting curves accurate reaction control).26 In our opinion, however, the
are shown, corresponding to the following two cases: (a) number of data points and their experimental accuracy
all three adjustable parameters in eq 9 were optimized are high enough to draw significant conclusions. As a
Macromolecules, Vol. 35, No. 2, 2002 Metallocene-Catalyzed Propene Polymerization 353

The best-fit values of [m°] and 〈kepim〉/〈kins〉 in Table 3

also deserve a comment. For each catalyst system,
largely different 〈kepim〉/〈kins〉 ratiossboth numerically
and dimensionallyswere obtained under cases a and b,
but this is mainly due to the different rate law used for
monomer insertion. More interesting, instead, is the fact
that the limiting stereoselectivity was also found to
depend slightly on the choice of the minimization
criterium and that for each catalyst a higher [m°] was
estimated in case a than in case b. Looking at the plots
in Figures 1 and 2, it can be seen that, at 80 °C, chain
epimerization lowers appreciably polymer stereoregu-
larity up to fairly high [M] and that the true catalyst
enantioselectivity can only be measured by extrapola-
tion in terms of eq 9. The result obviously depends on
the rate law of monomer insertion; in particular, setting
β ) R (case b)) leads one to underestimate [m°].
In conclusion, we believe that the microstructural
approach introduced in the present paper is an useful
tool for investigating the kinetics of isotactic propene
polymerization promoted by C2-symmetric ansa-metal-
locene catalysts. In the second part of this series, we
will show how the approach can be extended success-
fully to syndiotactic propene polymerization in the
presence of Cs-symmetric ansa-zirconocenes, and aimed
at investigating additional important mechanistic fea-
tures, such assin particularsthe effect of catalyst/
cocatalyst interaction on the dynamics of the catalytic
species.
Experimental Section
rac-Me2Si(2-Me-1-indenyl)2ZrCl2 was purchased from Boul-
der Scientific Co.; a sample of rac-Me2Si(2-Me-benz[e]-1-
Figure 2. Fraction of meso diads, [m], vs propene concentra- indenyl)2ZrCl2 was received from Prof. H. H. Brintzinger
tion, [C3H6], for polypropylene samples prepared at 80 °C in (Universität Konstanz).
the presence of catalyst system II/MAO: (top) complete plot; All polymerization experiments were carried out in a 2 L
(bottom) expansion in the region 0-1 M. Continuous and magnetically stirred stainless steel reactor (Brignole AU-2),
dashed best-fitting curves correspond to minimization under equipped with a glass vial holder-breaker. Dry toluene (Ald-
cases a and b, respectively (see text). rich, 0.5 L) and MAO (Witco GmbH, 10 wt % solution in
toluene) were introduced into the reactor, which was thermo-
Table 3. Best-Fit Valuesa of the Adjustable Parameters in stated at 80 °C and pressurized with propene at a given partial
Eq 9 for the Experimental Data in Tables 1 and 2 (See pressure.27 The reaction was started by breaking a glass vial
Text) and Corresponding Values of the χr2 Function containing the appropriate amount of solid catalyst (previously
catalyst system β 〈kepim〉/〈kins〉(Mβ) [m°] χr2 sealed under argon in a Vacuum-Atmospheres glovebox), and
allowed to proceed at constant propene concentration, making
I/MAO 0.94(7) 0.034(6) 0.972(6) 0.99 use of an electronic pressure controller (Brooks model 5866-
1.7b 0.0052(7) 0.945(7) 8.5
E) mounted on the feeding line. In case of polymerization runs
II/MAO 0.80(12) 0.023(7) 0.984(7) 1.2
1.7b 0.0030(8) 0.957(5) 9.1 performed at [C3H6] < 0.3 M, the composition of the gas phase
was also monitored by on-line Gas-Chromatographic analysis.
a Standard deviation (σ) in parentheses. b Imposed (see text). Catalyst and cocatalyst concentrations were adjusted in such
a way that monomer conversion (measured on-line with a
matter of fact, already on inspection of the plots it can Brooks 5850-E mass-flowmeter) never exceeded 1.0 g/min; this
be realized that case b results in a much worse fit. A enabled us to achieve a temperature control of the liquid phase
within (0.5 °C. After an appropriate time, the polymerization
quantitative confirmation comes from the values of the was stopped by quickly venting the reactor. The polymer was
reduced-χ2 function reported in Table 3: indeed, for each coagulated with acidified methanol, filtered, and vacuum-
catalyst system these are 1 order of magnitude lower dried.
in case a than in case b, and in the former case, they Quantitative 13C NMR spectra were recorded with a Varian
are ≈1, as they should be for a “good” model. VXR 200 spectometer operating at 50.3 MHz, on 100 mg/mL
polymer solutions in tetrachloroethane-1,2-d2 at 100-120 °C.
In the same Table 3, we also give the best-fit values Conditions: 5 mm probe; 76° pulse; acquisition time, 1.2 s;
of the three (case a)) or two (case b)) adjustable relaxation delay, 1.2 s; 10-15K transients. Peak integration
parameters. What is crucial to note for this investigation was carried out by means of full simulation, with the
is that, in case a, β turned out to be very close to 1 SHAPE2000 software package.28 The fractions of meso diads
(actually, lower than 1 by roughly 1σ for catalyst I, by were evaluated as explained in ref 10f; the confidence intervals
roughly 2σ for catalyst II). Therefore, we conclude that were estimated using the Monte Carlo simulation routine of
the rate of monomer insertion has a first-order depen- the CONFSTAT suite (Ver. 2.3 for Windows),29 accepting all
solutions of the enantiomorphic-sites model with a value of
dence on [M], and that the “trigger” hypothesis15 is not the χr2 function up to 1.2 times that at the absolute minimum.
consistent with our data. Conversely, an interpretation The best-fit of the experimental data in Tables 1 and 2 on
of the apparent second-order polymerization kinetics16 the basis of eq 9 was carried out with the curve-fitting routine
in terms of eqs 3′-5′18,20 may well hold. of the Microcal Origin (Ver. 5.0) software package.
354 Busico et al. Macromolecules, Vol. 35, No. 2, 2002

Acknowledgment. The authors are grateful to Prof. (13) Natta, G.; Pasquon, I.; Giachetti, E.; Scalari, F. Chim. Ind.
Hans-Herbert Brintzinger, for kindly providing a sample (Milan) 1958, 40, 103.
(14) Pino, P.; Rotzinger, B.; von Achenbach, E. Makromol. Chem.,
of catalyst II, and to Witco GmbH (Bergkamen, FRG), Suppl. 1985, 13, 105.
for generous donations of MAO. Financial assistance (15) (a) Ystenes, M. J. Catal. 1991, 129, 383. (b) Ystenes, M.
from the Italian Ministry for the University (MURSTs Makromol. Chem., Macromol. Symp. 1993, 66, 71.
Progetti di Rilevante Interesse Nazionale, PRIN2000) (16) (a) Wester, T. S.; Johnsen, H.; Kittilsen, P.; Rytter, E.
is acknowledged. The 13C NMR characterization of all Macromol. Chem. Phys. 1998, 199, 1989. (b) Stehling, U.;
polypropylene samples was carried out at the “Centro Diebold, J.; Kirsten, R.; Roell, W.; Brintzinger, H. H.; Jun-
gling, S.; Muelhaupt, R.; Langhauser, F. Organometallics
di Metodologie Chimico-Fisiche”, University Naples 1994, 13, 964.
Federico II. (17) Prosenc, M. H.; Schaper, F.; Brintzinger, H. H. In Metalor-
ganic Catalysts for Synthesis and Polymerization; Kaminsky,
References and Notes W., Ed.; Springer-Verlag: Berlin, 1999; pp 223-235.
(18) Tait, P. In Comprehensive Polymer Science; Pergamon
(1) (a) Boor J. Ziegler-Natta Catalysts and Polymerizations; Press: Oxford, England, 1988; vol. 4, pp 533-573.
Academic Press: New York, 1979. (b) Kissin, Y. V. Isospecific (19) Richardson, D. E.; Alameddin, N. G.; Ryan, M. F.; Hayes, T.;
Polymerization of Olefins; Springer-Verlag: New York, 1985. Eyler, J. R.; Siedle, A. R. J. Am. Chem. Soc. 1996, 118, 11244.
(c) Moore, E. P. Polypropylene Handbook; Hanser: Munich, (20) (a) Resconi, L.; Corradini, P.; Fait, A.; Guerra, G. Macromol-
Germany, 1996. ecules 1999, 32, 2104. (b) Moscardi, G.; Resconi, L.; Cavallo,
(2) Reviews: (a) Brintzinger, H. H.; Fischer, D.; Muelhaupt, R.; L. Organometallics 2001, 20, 1918.
Rieger, B.; Waymouth, R. M. Angew. Chem., Int. Ed. Engl. (21) Schaper, F.; Brintzinger, H. H.; Kleinschmidt, R.; van der
1995, 34, 1143-1170. (b) Resconi, L.; Cavallo, L.; Fait, A.; Leek, Y.; Reffke, M.; Fink, G. In Organometallic Catalysts
Piemontesi, F. Chem. Rev. 2000, 100, 1253-1346. and Olefin Polymerization; Blom, R., Follestad, A., Rytter,
(3) (a) Keii, T.; Terano, M.; Kimura, K.; Ishii, K. Makromol. E., Tilset, M., Ystenes, M., Eds.; Springer-Verlag: Heidelberg,
Chem., Rapid Commun. 1987, 8, 583. (b) Keii, T.; Terano, Germany, 2001; pp 46-62.
M.; Kimura, K.; Ishii, K. In Transition Metals and Organo-
(22) Review: Busico, V.; Cipullo, R. Prog. Polym. Sci. 2001, 26,
metallics as Catalysts for Olefins Polymerization; Kaminsky,
443-533.
W., Sinn, H., Eds.; Springer-Verlag: Berlin, 1988; pp 3-12.
(c) Mori, H.; Terano, M. Trends Polym. Sci. 1997, 5, 314 and (23) For other papers on the connection between polymer micro-
refs. therein. structure and polymerization kinetics, see, e.g.: (a) Lin, S.;
(4) Busico, V.; Cipullo, R.; Esposito, V. Macromol. Rapid Com- Tagge, C. D.; Waymouth, R. M.; Nele, M.; Collins, S.; Pinto,
mun. 1999, 20, 116. J. C. J. Am. Chem. Soc. 2000, 122, 11275. (b) Nele, M.;
(5) Liu, Z.; Somsook, E.; Landis, C. R. J. Am. Chem. Soc. 2001, Muqtar, M.; Xin, S.; Collins, S.; Dias, M. L.; Pinto, J. C.
123, 2915. Macromolecules 2001, 34, 3830. (c) Monaco, G. Macromol-
(6) Mejzlik, J.; Lesna, M.; Kratochvila, J. Adv. Polym. Sci. 1986, ecules 2001, 34, 4406.
81, 83 and references therein. (24) Spaleck, W.; Kuber, F.; Winter, A.; Rohrmann, J.; Bachmann,
(7) Busico, V.; Guardasole, M.; Margonelli, A.; Segre, A. L. J. B.; Antberg, M.; Dolle, V.; Paulus, E. F. Organometallics
Am. Chem. Soc. 2000, 122, 5226. 1994, 13, 954.
(8) Corradini, P.; Busico, V.; Guerra, G. In Comprehensive (25) Busico, V.; Cipullo, R.; Di Lena, F. Results to be published.
Polymer Science; Pergamon Press: Oxford, England, 1988; (26) In ref 10f,g, we used a similar approach to estimate, from
vol. 4, pp 29-50. the dependence of [m] on [M], the ratio of specific rates 〈kepim〉/
(9) (a) Busico, V.; Cipullo, R.; Corradini, P. Makromol. Chem. 〈kins〉 and its temperature dependence for three C2-symmetric
1993, 194, 1079. (b) Busico, V.; Cipullo, R.; Corradini, P. ansa-zirconocenes. However, the best-fit in terms of eq 9 was
Makromol. Chem., Rapid Commun. 1993, 14, 97. (c) Busico, carried out for case b only, i.e., using as adjustable param-
V.; Cipullo, R.; Chadwick, J. C.; Modder, J. F.; Sudmeijer, O. eters 〈kepim〉/〈kins〉 and [m°], and setting instead β ) R. Due to
Macromolecules 1994, 27, 7538. the fact that the investigated catalysts had R values only
(10) (a) Busico, V.; Cipullo, R. J. Am. Chem. Soc. 1994, 116, 9329. slightly higher than 1, and also that the data points were
(b) Leclerc, M.; Brintzinger, H. H. J. Am. Chem. Soc. 1995, lower in number and accuracy compared with those in the
117, 1651. (c) Busico, V.; Cipullo, R. J. Organomet. Chem. present study, we were satisfied by the fit and did not try to
1995, 497, 113. (d) Busico, V.; Caporaso, L.; Landriani, L.; verify whether a better one could be achieved by minimizing
Angelini, G.; Margonelli, A.; Segre A. L. J. Am. Chem. Soc. on β as well.
1996, 118, 2105. (e) Leclerc, M. K.; Brintzinger, H. H. J. Am. (27) For the relationship between propene partial pressure and
Chem. Soc. 1996, 118, 9024. (f) Busico, V.; Brita, D.; Capo- concentration in the liquid phase, see ref 10f, p 3976.
raso, L.; Cipullo, R.; Vacatello, M. Macromolecules 1997, 30, (28) Vacatello, M. SHAPE2000. Università Naples Federico II (for
3971. (g) Busico, V.; Cipullo, R.; Caporaso, L.; Angelini, G.; inquiries: vacatello@ chemistry.unina.it)
Segre, A. L. J. Mol. Catal., Part A (Chem.) 1998, 128, 53. (29) Vacatello, M. CONFSTAT (Ver. 2.3 for Windows). Università
(11) Rieger, B.; Mu, X.; Mallin, D. T.; Rausch, M.; Chien, J. C. W. Naples Federico II (for inquiries: vacatello@
Macromolecules 1990, 23, 3559. chemistry.unina.it).
(12) Ewen, J. A.; Jones, R. L.; Razavi, A.; Ferrara, J. D. J. Am.
Chem. Soc. 1988, 110, 6255. MA011401H