Formation of elastomeric polypropylene promoted by the dynamic complexes [TiCl2{N(PPh2)2}2] and [Zr(NPhPPh2)4]

Article abstract of DOI:10.1016/S0022-328X(00)00214-X

The homoleptic phosphinoamide complex [Zr(NPhPPh₂)₄] (1) and the bisamido complex [TiCl₂{N(PPh₂)₂}₂] (2) were prepared from ZrCl₄ and four equivalents of LiNPhPPh₂ and from TiCl₄ and one equivalent of [Li(THF)N(PPh₂)₂]₂. In the solid state, the four NPhPPh₂ ligands in 1 exhibit η² coordination. The ZrN₄P₄ fragment is highly symmetrical and almost of D₂ symmetry. Hence, the complex is chiral, and the two enantiomers cocrystallize in the asymmetric unit. In solution, 1 exhibits signals for the six-coordinate complex [Zr(η²-NPhPPh₂)₂(η ¹-NPhPPh₂)₂]. In the presence of methylalumoxane (MAO), 1 and 2 are active catalysts for the formation of high-molecular-weight elastomeric polypropylene. The formation of elastomeric polypropylene is a consequence of an epimerization mechanism of the last-inserted monomer, indicating no detachment of the growing polymer chain from the metal center during this process. Fractionation studies of all the elastomeric polymers show no atactic fractions. As corroboration for this mechanism, we have shown that these complexes are active catalysts for the isomerization and oligomerization of 1-octene, as well as for the rapid isomerization of allylbenzene to trans-methylstyrene.

Full text of DOI:10.1016/S0022-328X(00)00214-X

www.elsevier.nl/locate/jorganchem
Journal of Organometallic Chemistry 604 (2000) 116–125
Formation of elastomeric polypropylene promoted by the dynamic
complexes [TiCl2{N(PPh2)2}2] and [Zr(NPhPPh2)4]ꢀ
Olaf Ku¨hl a, Thomas Koch a, Fernando B. Somoza Jr. a, Peter C. Junk a,
Evamarie Hey-Hawkins a,*1, Dorit Plat b, Moris S. Eisen b,*2
a Institut fu¨r Anorganische Chemie der Uni6ersita¨t Leipzig, Johannisallee 29, D-04103 Leipzig, Germany
b Department of Chemistry and Institute of Catalysis Science and Technology, Technion–Israel Institute of Technology, Haifa 32000, Israel
Received 29 February 2000; received in revised form 16 April 2000
Abstract
The homoleptic phosphinoamide complex [Zr(NPhPPh2)4] (1) and the bisamido complex [TiCl2{N(PPh2)2}2] (2) were prepared
from ZrCl4 and four equivalents of LiNPhPPh2 and from TiCl4 and one equivalent of [Li(THF)N(PPh2)2]2. In the solid state, the
four NPhPPh2 ligands in 1 exhibit h2 coordination. The ZrN4P4 fragment is highly symmetrical and almost of D2 symmetry.
Hence, the complex is chiral, and the two enantiomers cocrystallize in the asymmetric unit. In solution, 1 exhibits signals for the
six-coordinate complex [Zr(h2-NPhPPh2)2(h1-NPhPPh2)2]. In the presence of methylalumoxane (MAO), 1 and 2 are active
catalysts for the formation of high-molecular-weight elastomeric polypropylene. The formation of elastomeric polypropylene is a
consequence of an epimerization mechanism of the last-inserted monomer, indicating no detachment of the growing polymer chain
from the metal center during this process. Fractionation studies of all the elastomeric polymers show no atactic fractions. As
corroboration for this mechanism, we have shown that these complexes are active catalysts for the isomerization and
oligomerization of 1-octene, as well as for the rapid isomerization of allylbenzene to trans-methylstyrene. © 2000 Elsevier Science
S.A. All rights reserved.
Keywords: Homoleptic zirconium phosphinoamide complex; Titanium bisamido dichloride complex; Molecular structure; Polymerisation of
propylene
1. Introduction
tial Ziegler–Natta catalysts. We and others have inves-
tigated a-olefin polymerization with bis(benzamidinate)
dichloride or dialkyl complexes of Group 4 metals (A)
as catalysts, which are obtained as a racemic mixture of
cis octahedral complexes with C2 symmetry [14–16].
When activated with methylalumoxane (MAO) or per-
fluoroborane compounds, these complexes were found
to be active catalyst precursors for the polymerization
of ethylene, propylene, 1-hexene and 2,5-hexadiene [12].
Major advances have been made during the last
decade in the design and synthesis of ‘well-defined’ and
‘single-site’ catalysts for the polymerization of olefins
[1–3]. The majority of these catalysts belong to the
metallocene ensemble, although homogeneous and new
heterogeneous complexes that contain one cyclopenta-
dienyl ring and (for example) a pendant amido ligand
are also well known and have been of major interest
[4–6]. However, recently complexes containing chelat-
ing ancillary ligands, such as dialkoxides, diamides, or
benzamidines [7–13], have received attention as poten-
ꢀ These results have been presented in part at the XVIII Interna-
tional Conference on Organometallic Chemistry, Munich, 16–22
August 1998; E. Hey-Hawkins et al., Book of Abstracts 1, AO 32.
1 *Corresponding author. Tel.: +49-341-9736151; fax: +49-341-
9739319; e-mail: hey@rz.uni-leipzig.de
2 *Corresponding author.
0022-328X/00/$ - see front matter © 2000 Elsevier Science S.A. All rights reserved.
PII: S0022-328X(00)00214-X

O. Ku¨hl et al. / Journal of Organometallic Chemistry 604 (2000) 116–125
117
Scheme 1.
In general, the tacticity of the polymers varies pre-
dictably with the structure of the metallocene catalysts.
Complexes with C2 or C1 symmetry are expected to
produce isotactic polyolefins, whereas those with C26
symmetry generate atactic polymers [1–3,17]. However,
when the polymerization of propylene with A is carried
out at atmospheric pressure, atactic polypropylene is
formed as an oil, instead of the expected isotactic
polymer [10,11]. The formation of atactic polypropy-
lene can be rationalized by an intramolecular epimer-
ization reaction of the growing polypropylene chain (P)
at the last-inserted monomer unit that is faster than the
stereoregular insertion of propylene (Scheme 1) [18–
21].
We recently showed that the expected isotactic
polypropylene is formed (mmmm\99.5%; m.p.=
153°C) when the polymerization is carried out at high
concentration of monomer (i.e. in liquid propylene).
Apparently, a high propylene concentration allows
faster insertion of the monomer unit and suppresses the
epimerization reaction almost completely [22]. These
results corroborate the proposed epimerization mecha-
nism [23].
our knowledge, this is the first example of the rational
synthesis of an elastomeric polymer by using octahedral
early transition metal complexes.
2. Results and discussion
2.1. Synthesis and spectroscopic properties of 1 and 2
The phosphinoamines Ph2PN(H)Ph [26] and
Ph2PN(H)PPh2 [27] are readily deprotonated by BuLi
to give LiNPhPPh2 [28] and LiN(PPh2)2 [29], respec-
tively. Two mesomeric structures can be formulated for
these anions in which the negative charge is localized at
the N (phosphinoamide) or P atom (iminophosphide)
(R=Ph, PPh2). According to theoretical calculations
[30], the phosphinoamide structure is dominant.
A conceptual question is whether simple octahedral
or even tetrahedral complexes that have a dynamic
Lewis-basic pendant group that can donate an electron
lone pair to the highly electrophilic metal center are
suitable for the preparation of a unique type of poly-
mer. Thus, as shown in Scheme 2, a dynamic equi-
librium occurs between the tetrahedal and octahedral
configurations, whereby only a cis-octahedral complex
can induce the polymerization of an a-olefin, such as
propylene, allowing the production of an elastomer
(X=halide, E=a donor group with an electron lone
pair, R=C, N, P or other anionic bridging group).
In these types of dynamic processes, a plausible trans
isomer can be obtained, although it will have no cata-
lytic activity [1–3]. These isomers are unable to perform
the insertion of an a-olefin because the incoming
monomer unit is trans to the growing alkyl chain [7–9].
Here we present the synthesis of two early transition
metal complexes that contain pendant donor groups:
the homoleptic phosphinoamide complex [Zr(NPh-
PPh2)4] (1) and the bisamido complex [TiCl2{N(P-
Ph2)2}2] (2). In the presence of methylalumoxane
(MAO), these complexes are active catalysts for the
formation of elastomeric polypropylene [22–25]. To
Scheme 2.

118
O. Ku¨hl et al. / Journal of Organometallic Chemistry 604 (2000) 116–125
ppm) [26]. At lower temperature, the signal at −5.4
ppm sharpens and the ratio of the two signals changes
from ca. 1:1 at 253 K to ca. 1:3 at 223 K. Apparently,
at 253 K two of the four phosphino groups in 1 exhibit
h1 coordination in solution (Scheme 3).
Accordingly, similar chemical shifts are observed for
the phosphinoamine [Ph2PN(H)PPh2, 46 ppm (in
THF)] [27] and the Ti complex 2 (47 ppm, s). At room
temperature, two sets of signals are observed in the
31P-NMR spectrum: a major signal (90%) at 47.0 ppm,
corresponding to the tetrahedral symmetric complex,
and a minor set (10%) consisting of two small doublets
at 5.0 and −10.0 ppm and corresponding to a C2-sym-
metric octahedral configuration or to a complex of
higher symmetry. The 2D heterocosy spectrum shows
that the two doublets are correlated with each other
and indicates that, of the many complexes that conceiv-
ably be obtained, only one isomer was formed (other
possible structures, such as a combination of tridentate
and a monodentate ligands, would give rise to addi-
tional correlated sets of signals). At 223 K the intensity
of the signal at 47.0 ppm decreases, and the intensity of
the two doublets increases. Furthermore, small addi-
tional signals are observed, presumably due to the
formation of other mono-chelated and unsymmetrical
complexes. This process is fully reversible on warming
the sample to room temperature.
Scheme 3.
The possible octahedral complexes (I–IV) that can
be obtained with these systems are denoted in Scheme
5. As can be seen, compounds I, III and IV are
expected to exhibit the same type of signals as described
above (two correlated doublets). For complex II, each
phosphorus atom is different, and an additional pair of
two correlating doublets would be expected. Hence,
compound II was not obtained.
Scheme 4.
When ZrCl4 is treated with four equivalents of
LiNPhPPh2, the homoleptic phosphinoamide complex
[Zr(NPhPPh2)4] (1) is formed (Scheme 3). Similarly,
TiCl4 reacts with two equivalents of LiN(PPh2)2 to give
the bisamido complex [TiCl2{N(PPh2)2}2] (2) (Scheme
4). The related sulfonamido complexes of zirconium
and titanium, [Zr(NButSPh)4] and [TiCl2(NButSPh)2],
are also known [31].
Several transition metal [32,33] and lanthanide com-
plexes [34] with phosphinoamido ligands are known;
the anionic ligand can be terminal, chelating or
bridging.
At room temperature, two signals are observed in the
31P-NMR spectrum of 1 (24.7, s and −5.4 ppm, br, in
THF), which are shifted to high field by about 4–34
ppm relative to the phosphinoamine Ph2PN(H)Ph (28.6
Scheme 5.

O. Ku¨hl et al. / Journal of Organometallic Chemistry 604 (2000) 116–125
119
Fig. 1. 31P{1H}-NMR spectrum of complex 2 with MAO at room temperature.
Since the octahedral complex is obtained by a dy-
namic process, it seems reasonable to expect that both
isomers (III and IV) will be obtained as a mixture that
would give rise to two different sets of two doublets in
the 31P-NMR spectrum. To rule out the formation of
complexes III and IV and to support the exclusive
formation of the cis-octahedral complex with C2 sym-
metry, we treated complex 2 with MAO at room tem-
perature and monitored the changes in the 31P-NMR
signals. The reaction of MAO with any of the com-
pounds in Scheme 5 will lead to the formation of the
corresponding cationic methyl complexes (Scheme 2).
With III and IV, a set of two doublets should be
obtained for each compound, since two of the four
phosphorus atoms are magnetically equivalent. For
complex I, with a C2 symmetry, the phosphorus atoms
are inequivalent, and two sets of two doublets are
expected. As can be seen in Fig. 1, the 31P-NMR
spectrum of complex 2 with MAO exhibits, as expected,
four doublets (2JPP=278 Hz), supporting the exclusive
formation of the C2-symmetric cis-octahedral complex.
Interestingly, in the presence of MAO at room temper-
ature, isomer I is the major isomer present in solution
(minor amounts of the tetrahedral complex are also
observed), due to the donation of the phosphorus lone
pair of electrons to the cationic metal center.
symmetrical and almost of D2 symmetry. Hence, the
complex is chiral, and both enantiomers cocrystallize in
the asymmetric unit. The mirror image of molecule 2
differs from molecule 1 only in the orientation of the
phenyl groups. Similar structural features have been
observed in homoleptic lanthanide phosphinoamides
[34].
Homoleptic eight-coordinate complexes of Group 4
metals with four anionic chelating ligands are known
with six-, five-, four- and three-membered metallacycles
2.2. Molecular structure of 1
Colorless single crystals of 1 were obtained at room
temperature from diethyl ether. The compound crystal-
(
lizes in the triclinic space group P1 with two indepen-
dent molecules in the asymmetric unit (vide infra).
There is one disordered Et2O molecule in the unit cell
(O at 1/2 1 1/2). The four NPhPPh2 ligands exhibit h2
coordination (Fig. 2). The ZrN4P4 fragment is highly
Fig. 2. Molecular structure of 1. One of the two enantiomers in the
asymmetric unit is shown. Hydrogen atoms are omitted for clarity.

120
O. Ku¨hl et al. / Journal of Organometallic Chemistry 604 (2000) 116–125
(e.g. three-membered rings: pyrazolato complexes of Ti
[35], sulfenamido complex of Zr [31], hydroxylamido
complex of Ti [36] and Zr [36]). To date, no homoleptic
zirconium phosphinoamide complexes were known.
In 1960, Orgel suggested that in eight-coordinate
complexes of formula MA4B4 the two different donor
atoms A and B should occupy the two non-equivalent
positions in a dodecahedron [37]. Thus, in complexes
with d0 metal ions, the better p donor should occupy
the B coordination sites in a dodecahedron [38]. The
coordination polyhedron for an eight-coordinate transi-
tion metal complex can be determined by calculating
the angle of intersection between the planes of the two
interpenetrating BAAB trapezoids of an assumed do-
decahedron. This angle is then compared with that of
an idealized dodecahedron (90°) and of an idealized
square antiprism (74.4°) [39]. For 1, this angle is 86.4
(for molecule 1) and 86.2° (for molecule 2), and this
indicates a distorted dodecahedral coordination envi-
ronment (max. deviation of N, P from mean planes of
The ZrꢀN bond lengths of the ZrꢀNꢀP three-mem-
,
bered rings in 1 differ by ca. 0.02 A [mean ZrꢀN
,
,
2.177(5) A, range 2.168(5)–2.189(5) A], and the ZrꢀP
,
,
bond lengths by ca. 0.05 A [mean ZrꢀP 2.741(2) A,
,
range 2.711(2)–2.763(2) A] (Table 1). The ZrꢀN and
the ZrꢀP distances are longer than those observed for
homoleptic amido complexes (e.g. ZrꢀNterminal in
[Zr(NMe2)4]: 2.050(5)–2.104(5) A, ZrꢀN in [Zr{NH-
(2,6-Pr2C6H3)}4]: 2.027(7)–2.057(7) A) [41] and zir-
conocene phosphine complexes (e.g. ZrꢀP in [Cp2Zr-
(PMe3)2]: 2.650(2), 2.638(2) A; in [Cp2Zr(C2H4)(PMe3)]:
2.693(2) A) [42].
The structural data of the NPhPPH2 ligands [PꢀN
1.638(7)–1.682(6) A] are only slightly different from
,
i
,
,
,
,
,
those of the free ligand Ph2PN(H)Ph [PꢀN 1.696(3) A]
[34].
2.3. Catalytic polymerization studies
The polymerization of propylene was carried out
with precursors 1 and 2, which were activated by MAO
in toluene. To this mixture propylene was added under
a pressure that maintained it at all times as liquid
propylene. After a measured time interval, the reaction
was quenched by the addition of HCl–MeOH. Subse-
quently, the polymer was washed with water and ace-
tone, followed by multiple extractions with hot heptane
before drying the polymer under vacuum. In all cases,
no low-molecular-weight atactic polypropylene was ob-
served.
Table 2 shows the propylene polymerization results
obtained with the catalytic systems 1/MAO and 2/
MAO. The activity of the titanium complex 2 was
found to be much higher than that of the corresponding
zirconium complex 1 under the same experimental con-
ditions (entries 2 and 6). The polymerization reactions
were carried out in liquid propylene (highest monomer
concentration) to avoid, as much as possible, the
epimerization of the last-inserted unit, as found for
other octahedral complexes (Scheme 1), and to allow
rational design of an elastomeric polymer. The forma-
tion of an elastomeric polypropylene (no atactic high-
molecular-weight polymer was observed; vide infra) is
consistent with the proposed equilibrium between a
cis-octahedral and a tetrahedral active cationic catalyst,
as presented in Scheme 2.
,
trapezoids: 0.24–0.36 A). The bidentate phosphi-
noamido ligands span the m edges of the distorted
dodecahedron to give the mmmm [37b] stereoisomer.
This appears to be the preferred stereoisomer in com-
plexes with four bidentate ligands having a small bite
angle [40].
The ZrN4 fragment has a distorted tetrahedral ge-
ometry [NꢀZrꢀN range from 97.2(2) to 102.4(2)° and
from 131.7(2) to 132.0(2)°], and the ZrP4 fragment a
twisted rectangular arrangement [PꢀZrꢀP range from
75.77(5) to 108.39(5)° and 162.99(5) to 163.88(6)°]. The
sum of angles at the nitrogen atoms (CꢀNꢀP ca. 130°,
CꢀNꢀZr ca. 140°, PꢀNꢀZr ca. 90°, sum: ca. 360°)
indicates trigonal-planar coordination. Thus, the phos-
phinoamide ligand can be regarded as a 4 (N)+2 (P)
electron donor. As in related anionic phosphinoamido
complexes of lanthanides, the four NP groups can be
described as two N2P2 planes, which both form parallelo-
gramms and are staggered with respect to each other.
Complex 1 exhibits comparable structural parameters
to the zirconium sulfenamido complex [Zr(NButSPh)4],
which also possesses D2 symmetry and can be regarded
as a distorted dodecahedron in which the ZrN4 tetrahe-
dron is considerably flattened [ZrꢀN: 2.125(4)–2.147(4),
,
ZrꢀS: 2.774(2)–2.814(4) A, NꢀZrꢀN: 95.4(2)–104.7(2),
131.3(2)–132.0(2), SꢀZrꢀS: 79.4(1), 80.8(1), 101.2(1),
102.3(1), 165.3(1) and 165.8(1), ZrꢀNꢀS: 92.5(2)–
94.2(2), ZrꢀSꢀN: 49.1(1)–50.0(1), NꢀZrꢀS: 36.7(1)–
37.5(1)°] [31]. Different structural parameters are
observed for [Zr(ONMe2)4] [36]. The molecule has C2
symmetry, and the ZrO4N4 polyhedron can be de-
scribed as two interpenetrating strongly distorted MX4
tetrahedra (OꢀZrꢀO 90.5–131.5, NꢀZrꢀN 90.7–127.8°).
Here, the ZrꢀO(1) bond length is smaller than ZrꢀO(2)
The polymer microstructures of the elastomeric
polypropylene products were characterized by standard
13C-NMR analytical methods and measured at 135°C
in 1,2,4-trichlorobenzene [43]. Fig. 3 shows the 13C-
NMR spectrum of the elastomeric polymer obtained
under the conditions of entry 6 (Table 2). In the methyl
region it exhibits a high-intensity signal (35%) corre-
sponding to the mmmm pentad, and this rules out the
,
by ca. 0.05 A, and ZrꢀN(1) is larger than ZrꢀN(2) by
possibility
that
high-molecular-weight
atactic
,
ca. 1 A [36].
polypropylene was obtained. (All the extracted and

O. Ku¨hl et al. / Journal of Organometallic Chemistry 604 (2000) 116–125
121

122
O. Ku¨hl et al./Journal of Organometallic Chemistry 604 (2000) 116–125
Table 1
trum indicates polymer molecules with a combination
of atactic and isotactic blocks and a large domain of
isotactic pentads, as expected for elastomeric polymers
,
Selected bond lengths (A) and angles (°) in 1
Molecule 1
1
[24,25,45]. In the H-NMR spectrum, no vinylidene or
Zr(1)ꢀN(13)
Zr(1)ꢀN(11)
Zr(1)ꢀP(14)
Zr(1)ꢀP(12)
P(11)ꢀN(11)
P(13)ꢀN(13)
2.168(5)
2.173(5)
2.729(2)
2.744(2)
1.680(5)
1.666(5)
Zr(1)ꢀN(14)
Zr(1)ꢀN(12)
Zr(1)ꢀP(13)
Zr(1)ꢀP(11)
P(12)ꢀN(12)
P(14)ꢀN(14)
2.175(5)
2.181(5)
2.733(2)
2.762(2)
1.670(5)
1.646(6)
vinyl chain end groups were detected, suggesting an
aluminum chain transfer mechanism as the major ter-
mination step. Note that in Fig. 3, additional signals
appear in the methyl region, indicating an apparently
low resolution spectrum. A simple comparison with
other signals in the spectrum (25 and 48 ppm) shows
that this broadness is the result of different environ-
ments of the methyl protons besides the ten expected
isotactic-atactic signals (vide infra).
Interestingly, the activation of complex 1 by MAO
implies that two of the amido groups are exchanged for
methyl groups, followed by the removal of a methyl
group to give the corresponding cationic complex. We
have recently shown that in the reaction of simple
tetrahedral homoleptic tetraamido Group 4 complexes
with methylalumoxane (MAO), two of the four amido
groups are replaced by methyl ligands, with the subse-
quent formation of the active cationic catalysts for the
rapid polymerization of ethylene and the atactic poly-
merization of propylene [7,9].
Comparison of the relationships between activity,
co-catalyst concentrations and temperature shows that
in contrast to the octahedral benzamidinate complexes,
the catalytic activity increases with increasing
MAO:catalyst ratio and temperature. This behaviour is
believed to be a result of the low coordinative unsatura-
tion and increased electrophilicity of the metal center,
on increasing the MAO concentration, which favours
the coordination of the incoming monomer to the
active ‘cationic’ complex. Increasing the temperature
results in a faster process, and this suggests that the
cationic complexes are highly stable [6].
For both catalyst systems, the molecular weights of
the elastomeric polymers are similar with narrow poly-
dispersities. This result supports a polymerization
mechanism in which a single-site catalyst is responsible
for the formation of the elastomeric polypropylenes.
The results presented here show that octahedral
‘cationic’ complexes with C2 symmetry can induce the
formation of isotactic polypropylene, whereas tetrahe-
dral complexes of C26 symmetry lead to atactic
polypropylene. The rapid and dynamic interconversion
of these two structures leads to elastomeric polymers.
The homo-copolymer (isotactic-atactic block copoly-
mer) can be easily tailored by changing the donor
strength of the dynamic pendant atoms, allowing the
formation of either short or large isotactic blocks be-
tween the atactic blocks. This type of polymer has been
obtained with non-bridged metallocene complexes in
which the rotational frequency of the substituted cy-
clopentadienyl ligands allows an exchange between C2
N(13)ꢀZr(1)ꢀN(14)
N(14)ꢀZr(1)ꢀN(11)
N(14)ꢀZr(1)ꢀN(12)
P(14)ꢀZr(1)ꢀP(13)
P(13)ꢀZr(1)ꢀP(12)
P(13)ꢀZr(1)ꢀP(11)
N(13)ꢀZr(1)ꢀP(14)
N(11)ꢀZr(1)ꢀP(14)
N(13)ꢀZr(1)ꢀP(13)
N(11)ꢀZr(1)ꢀP(13)
N(13)ꢀZr(1)ꢀP(12)
N(11)ꢀZr(1)ꢀP(12)
N(13)ꢀZr(1)ꢀP(11)
N(11)ꢀZr(1)ꢀP(11)
N(11)ꢀP(11)ꢀZr(1)
N(13)ꢀP(13)ꢀZr(1)
101.6(2)
132.0(2)
97.3(2)
100.93(6)
79.50(5)
162.99(5)
90.8(2)
99.6(2)
37.6(1)
129.6(1)
98.7(1)
92.2(2)
125.4(1)
37.5(1)
51.9(2)
52.5(2)
N(13)ꢀZr(1)ꢀN(11)
N(13)ꢀZr(1)ꢀN(12)
N(11)ꢀZr(1)ꢀN(12)
P(14)ꢀZr(1)ꢀP(12)
P(14)ꢀZr(1)ꢀP(11)
P(12)ꢀZr(1)ꢀP(11)
N(14)ꢀZr(1)ꢀP(14)
N(12)ꢀZr(1)ꢀP(14)
N(14)ꢀZr(1)ꢀP(13)
N(12)ꢀZr(1)ꢀP(13)
N(14)ꢀZr(1)ꢀP(12)
N(12)ꢀZr(1)ꢀP(12)
N(14)ꢀZr(1)ꢀP(11)
N(12)ꢀZr(1)ꢀP(11)
N(12)ꢀP(12)ꢀZr(1)
N(14)ꢀP(14)ꢀZr(1)
97.2(2)
131.7(2)
102.4(2)
163.88(6)
75.77(5)
108.39(5)
37.1(2)
128.0(2)
88.1(1)
100.1(2)
127.4(2)
37.5(1)
98.1(2)
94.8(2)
52.6(2)
52.9(2)
Molecule 2
Zr(2)ꢀN(23)
Zr(2)ꢀN(24)
Zr(2)ꢀP(21)
Zr(2)ꢀP(23)
P(21)ꢀN(21)
P(23)ꢀN(23)
2.172(5)
2.179(5)
2.711(2)
2.750(2)
1.648(5)
1.680(5)
Zr(2)ꢀN(22)
Zr(2)ꢀN(21)
Zr(2)ꢀP(22)
Zr(2)ꢀP(24)
P(22)ꢀN(22)
P(24)ꢀN(24)
2.176(5)
2.189(5)
2.739(2)
2.763(2)
1.672(5)
1.678(5)
N(23)ꢀZr(2)ꢀN(22)
N(22)ꢀZr(2)ꢀN(24)
N(22)ꢀZr(2)ꢀN(21)
P(22)ꢀZr(2)ꢀP(23)
P(22)ꢀZr(2)ꢀP(24)
N(23)ꢀZr(2)ꢀP(21)
N(24)ꢀZr(2)ꢀP(21)
N(23)ꢀZr(2)ꢀP(22)
N(24)ꢀZr(2)ꢀP(22)
N(23)ꢀZr(2)ꢀP(23)
N(24)ꢀZr(2)ꢀP(23)
N(22)ꢀZr(2)ꢀP(24)
N(24)ꢀZr(2)ꢀP(24)
P(21)ꢀZr(2)ꢀP(22)
N(21)ꢀP(21)ꢀZr(2)
N(23)ꢀP(23)ꢀZr(2)
131.3(2)
97.2(2)
N(23)ꢀZr(2)ꢀN(24)
N(23)ꢀZr(2)ꢀN(21)
N(24)ꢀZr(2)ꢀN(21)
P(21)ꢀZr(2)ꢀP(24)
P(23)ꢀZr(2)ꢀP(24)
N(22)ꢀZr(2)ꢀP(21)
N(21)ꢀZr(2)ꢀP(21)
N(22)ꢀZr(2)ꢀP(22)
N(21)ꢀZr(2)ꢀP(22)
N(22)ꢀZr(2)ꢀP(23)
N(21)ꢀZr(2)ꢀP(23)
N(23)ꢀZr(2)ꢀP(24)
N(21)ꢀZr(2)ꢀP(24)
P(21)ꢀZr(2)ꢀP(23)
N(22)ꢀP(22)ꢀZr(2)
N(24)ꢀP(24)ꢀZr(2)
104.7(2)
95.4(2)
131.9(2)
78.71(5)
108.30(5)
85.4(1)
37.4(1)
37.6(1)
92.7(1)
100.5(1)
127.8(1)
98.2(1)
97.4(1)
165.22(6)
52.6(2)
52.1(2)
101.4(2)
77.41(5)
160.97(6)
129.9(1)
101.9(1)
96.8(1)
126.6(1)
37.6(1)
90.9(1)
123.9(1)
37.4(1)
100.25(5)
53.8(2)
52.2(2)
Both molecules
PꢀC
1.807(9)
–1.854(6)
108.7(5)
–115.4(4)
125.4(2)
–132.2(2)
138.4(4)
–141.0(4)
NꢀC
1.406(7)
–1.447(8)
102.4(3)
–105.5(3)
128.7(4)
–131.8(4)
88.7(2)
NꢀPꢀC
CꢀPꢀZr
CꢀNꢀZr
CꢀPꢀC
CꢀNꢀP
PꢀNꢀZr
–90.7(2)
washed polymers obtained as shown in Table 2 show
similar behaviour, with intense mmmm pentad signals in
the 13C-NMR spectrum) [44]. Thus, the 13C-NMR spec-

O. Ku¨hl et al. / Journal of Organometallic Chemistry 604 (2000) 116–125
123
Table 2
Activity data for the polymerization of propylene promoted by complexes 1 and 2 activated by MAO in liquid propylene
Entry
Catalyst
Temperature (°)
Ratio MAO:complex
Activity g polymer/mol h
Mw
Mn
Mw/Mn
1
2
3
4
5
6
2
2
2
2
2
1
0
25
0
25
60
25
500
500
1000
1000
1000
500
890
1570
1450
4190
4750
80
145620
88240
1.65
1.57
2.40
163840
115390
104470
48010
and C26 symmetries [24,25] and recently with dual-side
bridged metallocene complexes [45].
Thus, simple tetrahedral complexes with ancillary
ligands incorporating donating groups (with a lone pair
of electrons able to coordinate to the metal center)
allows the rational design of selective elastomeric
polypropylenes.
To corroborate that the epimerization mechanism is
responsible for the formation of elastomeric polypropy-
lene, we treated 1-octene with complex 1 activated by
MAO. Since the olefin is expected to polymerize much
slower than propylene, the epimerization mechanism, if
applicable, should induce the isomerization of a-olefins
to the corresponding internal olefins. A b-hydrogen
elimination is expected to occur either from the methyl
group at the b-position in the intermediate complex B
(Scheme 6, R=H, without causing any change in the
alkene), or from the b-position at the alkyl chain,
inducing the isomerization of the double bond.
3. Experimental
All manipulations of air-sensitive materials were per-
formed with the rigorous exclusion of oxygen and
moisture. Argon, propylene and nitrogen gases were
purified by passage through a MnO oxygen-removal
,
column and a Davison 4A molecular sieve column.
The isomerization of 1-octene promoted by complex
1 and MAO resulted in the quantitative formation of
trans-2-octene. The isomerization of 1-octene with com-
plex 2 leads to the formation of trans-2-octene (95%)
and small amounts (ꢀ5%) of cis-2-octene, (E)-3-
octene, (E)-4-octene and a myriad of saturated and
unsaturated dimers, trimers and tetramers (the latter
observed by GC-MS). The treatment of trans-2-octene
with 2/MAO resulted in slow isomerization to cis-2-
octene (up to 5% conversion after 16 h at 80°C), with
traces of the oligomeric compounds, indicating that the
process is fully in equilibrium.
The reactions of allylbenzene with catalysts 1 or 2
activated by MAO, induces selectively the formation of
trans-methylstyrene (Eq. (1)). For complex 1 the reac-
tion is extremely fast, producing 100% conversion at
room temperature in 15 min, whereas for complex 2,
77% yield was obtained after 16 h at 80°C [46].
Ether solvents (THF-d8) were distilled under argon
from sodium benzophenone ketyl. Hydrocarbon sol-
vents (toluene-d8, benzene-d6) were distilled under nitro-
gen from Na/K alloy. All solvents for vacuum line
manipulations were stored in vacuo over Na/K alloy in
resealable bulbs or freshly distilled prior to use. For 1,
the NMR spectra were recorded with an AVANCE
DRX 400 spectrometer (Bruker), for 2 on a Bruker AM
200 and a Bruker AM 400 spectrometer. 1H-NMR:
internal standard solvent, external standard TMS; 13C-
NMR: external standard TMS, internal standard sol-
vent; 31P-NMR: external standard 85% H3PO4. The IR
spectrum of 1 was recorded on a FTIR spectrometer
Perkin–Elmer System 2000 in the range 350–4000
cm−1. GC-MS experiments were conducted in a
GCMS (Finnigan Magnum) spectrometer. The melting
points were determined in sealed capillaries under ar-
gon and are uncorrected. Molecular weights were mea-
sured with a Waters GPCV 150 plus instrument
(1)
These isomerization results indicate that during poly-
merization the metal center is also able to migrate
through the growing polymer chain, inducing branch-
ing and additional environments for the methyl ligands,
as shown in Scheme 6 [47].
Scheme 6. Plausible isomerization pathway for 1-octene in the pres-
ence of complex 1.

124
O. Ku¨hl et al. / Journal of Organometallic Chemistry 604 (2000) 116–125
calibrated against polystyrene. LiNPhPPh2 [28] and
[Li(THF)N(PPh2)2]2 [29] were prepared according to
the literature procedure. TiCl4, ZrCl4, HN(SiMe3)2, n-
BuLi (solution in hexane) (all Aldrich), Ph2PCl (Strem)
are commercially available. a-Olefins (1-octene, 2-
octene, allylbenzene) were dried and stored over acti-
and immediately injecting a methanol–HCl mixture.
The polymeric product was collected by filtration,
washed with hexane and acetone and dried under vac-
uum for several hours.
3.4. Isomerization experiments
,
vated molecular sieves (4A), degassed, and freshly
vacuum-distilled.
In a typical procedure, an olefin (0.1 ml) was vacuum
transferred into a J-Young NMR tube containing 10
mg of catalyst and MAO (MAO:catalyst=400) in 0.6
ml of toluene-d8. The sealed tube was then heated in an
oil bath at 80°C. The organic products were vacuum
transferred (10−6 mmHg) to another J-Young NMR
tube and sealed, and both residue and volatiles were
3.1. [Zr(NPhPPh2)4] (1)
A solution of 4.27 g (10.0 mmol) LiNPhPPh2 in 50
ml diethyl ether was added dropwise to a stirred sus-
pension of 0.58 g (2.5 mmol) ZrCl4 in 100 ml diethyl
ether at −78°C. The solution was stirred overnight at
room temperature (r.t.). Then LiCl was filtered off and
the solvent partially removed in vacuo until ca. 40 ml of
the solution remained. Colourless crystals of 1 were
obtained at −30°C. Yield 65%. M.p. 180–182°C.
C72H60N4P4Zr (1196.34); Calc.: C, 72.31; H, 5.02; N,
4.67; P, 10.38; Found: C, 72.13; H, 5.07; N, 4.58; P,
10.51%. 1H-NMR (C6D6): l 6.69–7.74 (m, Ph). 31P-
NMR (THF, 283 K): l 24.2 (s), −5.4 (s).
1
characterized by H- and 13C-NMR spectroscopy, GC-
MS and comparison with known compounds [46].
3.5. Crystallographic studies
,
Data of 1 (Mo–Ka, u=0.71073 A) were collected
with a Siemens CCD (SMART). All observed reflections
were used for determination of the unit cell parameters.
The structure was solved by direct methods (SHELXTL
PLUS [48]) and subsequent difference Fourier syntheses
and refined by full-matrix least-squares on F2 (SHELXTL
PLUS [48]). Restrictions for 1: C, N, P, and Zr atoms
anisotropic, H atoms located and refined isotropically.
3.2. [TiCl2{N(PPh2)2}] (2)
To a stirred solution of [Li(THF)N(PPh2)2]2 (2.0 g,
2.16 mmol) in 50 ml THF, was added TiCl4 (0.24 ml,
2.16 mmol) dropwise at r.t. The reaction mixture was
stirred overnight, the solvent removed in vacuum and
the residue was extracted with 30 ml of toluene. Hexane
was added to the filtrate to precipitate a brown solid,
which was recrystallized from dichloromethane at −
78°C. Yield 35%. M.p. 120°C (dec.). C48H40Cl2N2P4Ti
(887.5); Calc.: C, 64.96; H, 4.54; N, 3.16; Found: C,
Absorption correction with SADABS
.
Crystal data for 1·1/4Et2O: C72H60N4P4Zr·1/4Et2O,
M=1214.87, colourless crystals, triclinic, space group
(
P1 (no. 2), T=213(2) K, a=12.834(1), b=21.874(1),
,
c=23.975(1) A, h=86.81(1), i=77.53(1), k=
3
,
81.86(1)°, V=6503.0(2) A , Z=4, Dcalc=1.241 Mg
m
−3, F(000)=2522, v(Mo–Ka)=0.312 mm−1, 40265
reflections collected with 0.87BqB27.24°; of these,
24841 were independent; 1502 parameters, refinements
converge to R1=0.0714, wR2=0.2108 (for reflections
with I\2s(I)), R1=0.1325, wR2=0.2613. Even
though the outer shape and size of the crystals ap-
peared in good order, the data set obtained was of
rather low quality [Rint=0.0458]. Recrystallization and
several attempts to remeasure different crystals did not
result in a better data set.
1
64.03; H, 4.86; N, 3.01. H-NMR (THF-d8): l 7.0–8.2
(m, Ph). 31P-NMR (THF): l 47.0 (s).
3.3. Polymerization experiments
These experiments were conducted in a 100 ml
flamed round-bottom reaction flask attached to a high-
vacuum line. In a typical experiment 4.0 mg of the
catalyst 1 and 205 mg of MAO (Zr:Al=1:400) were
charged in the flask containing a magnetic stir bar. The
reaction vessel was connected to a high-vacuum line,
pumped down and back-filled three times with argon.
Then the flask was re-evacuated, and a measured quan-
tity of toluene (40 ml) was vacuum transferred into the
reaction flask from Na/K. Then gaseous propylene was
condensed into the vessel through a gas purification
column. The flask was warmed to r.t., and the pressure
was continuously measured by means of a manometer.
Rapid stirring of the solution was initiated, and after a
measured time interval (2 h) the polymerization was
quenched by releasing the excess pressure of propylene
4. Supplementary material
Crystallographic data (excluding structure factors)
for the structure reported in this paper have been
deposited with the Cambridge Crystallographic Data
Centre (1: CCDC no. 145086). Copies of this informa-
tion may be obtained free of charge from The Director,
CCDC, 12 Union Road, Cambridge, CB2 1EZ, UK
(Fax: +44-1223-336033;e-mail: deposit@ccdc.cam.ac.uk
or www: http://www.ccdc.cam.ac.uk).

O. Ku¨hl et al. / Journal of Organometallic Chemistry 604 (2000) 116–125
125
[22] V. Volkis, M. Shmulinson, C. Averbuj, A. Lisovskii, F.T. Edel-
mann, M.S. Eisen, Organometallics 17 (1998) 3155.
[23] M. Shmulinson, V. Volkis, D. Plat, E. Tish, M.S. Eisen, Elas-
tomeric polypropylenes can be obtained by modulating the pres-
sure during the course of the reaction with complexes of Group
4 or actinides containing acetylacetonate, azaallyl, oxazolines,
b-diketiminates, amidinates and diazapentadienyl ligands (in
preparation).
[24] G.W. Coates, R.M. Waymouth, Science 267 (1995) 217.
[25] Y. Hu, M.T. Krejchi, C.D. Shah, C.I. Myers, R.M. Waymouth,
Macromolecules 31 (1998) 6908 and refs therein.
[26] R.F. Hudson, R.J.G. Searle, F.H. Devitt, J. Chem. Soc. C (1966)
1001. (b) W. Wiegra¨be, H. Bock, Chem. Ber. 101 (1968) 1414.
[27] J.D. Woolins (Ed.), Inorganic Experiments, VCH, Weinheim,
1994, p. 145.
Acknowledgements
We gratefully acknowledge support from the Ger-
man-Israeli-Foundation grant no. I-142/95, the Fonds
der Chemischen Industrie, the Deutscher Akademischer
Austauschdienst (P.C.J.), the Funds for the Promotion
of Research at the Technion, the E. and J. Bishop
Research Fund administered by the Technion V.P.R.
Fund and to The Ministry of Science.
References
[28] M.T. Ashby, Z. Li, Inorg. Chem. 31 (1992) 1321.
[29] H. Schmidbaur, S. Lauteschla¨ger, F.H. Ko¨hler, J. Organomet.
Chem. 271 (1984) 173.
[30] G. Trinquier, M.T. Ashby, Inorg. Chem. 33 (1994) 1306.
[31] D.M. Hankin, A.A. Danopoulos, G. Wilkinson, T.K.N. Sweet,
M.B. Hursthouse, J. Chem. Soc. Dalton Trans. (1996) 1309.
[32] D. Fenske, M. Maczek, K. Maczek, Z. Anorg. Allg. Chem. 623
(1997) 1113 and refs therein.
[33] E. Lindner, M. Heckmann, R. Fawzi, W. Hiller, Chem. Ber. 124
(1991) 2171.
[34] T.G. Wetzel, S. Dehnen, P.W. Roesky, Angew. Chem. Int. Ed.
Engl. 111 (1999) 1155.
[1] W. Kaminsky, M. Arndt, Adv. Polym. Sci. 127 (1997) 144.
[2] M. Bochmann, J. Chem. Soc. Dalton Trans. (1996) 225.
[3] H.H. Brintzinger, D. Fischer, R. Mu¨lhaupt, B. Rieger, R.M.
Waymouth, Angew. Chem. Int. Ed. Engl. 34 (1995) 1143.
[4] G. Lanza, I.L. Fragala`, T.J. Marks, J. Am. Chem. Soc. 120
(1998) 8257 and refs therein.
[5] D.W. Carpenetti, L. Koppenbrug, J.T. Kupec, J.L. Petersen,
Organometallics 15 (1996) 1572.
[6] M. Galan-Fereres, T. Koch, E. Hey-Hawkins, M.S. Eisen, J.
Organomet. Chem. 580 (1999) 145.
[7] H. Mack, M.S. Eisen, J. Chem. Soc. Dalton Trans. (1998) 917
and refs therein.
[8] R. Baumann, W.M. Davis, R.R. Schrock, J. Am. Chem. Soc.
119 (1997) 3830.
[35] I.A. Guzei, A.G. Baboul, G.P.A. Yap, A.L. Rheingold, H.B.
Schlegel, C.H. Winter, J. Am. Chem. Soc. 119 (1997) 3387.
[36] N.W. Mitzel, S. Parsons, A.J. Blake, D.W.H. Rankin, J. Chem.
Soc. Dalton Trans. (1996) 2089.
[9] H. Mack, M.S. Eisen, J. Organomet. Chem. 525 (1996) 81.
[10] J. Richter, F.T. Edelmann, M. Noltemeyer, H.-G. Schmidt, M.
Shmulinson, M.S. Eisen, J. Mol. Catal. 130 (1998) 149.
[11] D. Walther, R. Fischer, H. Go¨rls, J. Koch, B. Schweder, J.
Organomet. Chem. 508 (1996) 13.
[12] D. Herskovics-Korine, M.S. Eisen, J. Organomet. Chem. 503
(1995) 307.
[13] J.C. Flores, J.C.W. Chen, M.D. Rausch, Organometallics 14
(1995) 1827 and refs therein.
[14] E. Mu¨ller, J. Mu¨ller, F. Olbrich, W. Bru¨ser, W. Knapp, D.
Abeln, F.T. Edelmann, Eur. J. Inorg. Chem. (1998) 87 and refs
therein.
[15] P.J. Stewart, A. Blake, P. Mountford, Organometallics 17 (1998)
3271.
[16] J.R. Hagardon, J. Arnold, Organometallics 17 (1998) 1355 and
refs therein.
[17] C. Averbuj, E. Tish, M.S. Eisen, J. Am. Chem. Soc. 120 (1998)
8640.
[37] (a) L.E. Orgel, J. Inorg. Nucl. Chem. 14 (1960) 136. (b) J.L.
Hoard, J.V. Silverton, Inorg. Chem. 2 (1963) 235.
[38] W.D. Bonds Jr., R.D. Archer, Inorg. Chem. 10 (1971) 2057.
[39] S.J. Lippard, B.J. Russ, Inorg. Chem. 7 (1968) 1686.
[40] D.G. Blight, D.L. Kepert, Inorg. Chem. 11 (1972) 1556.
[41] M.F. Lappert, P.P. Power, A.R. Sanger, R.C. Srivastava, Metal
and Metalloid Amides, Ellis Horwood, Chichester, West Sussex,
1980. (b) M.H. Chisholm, C.F. Hammond, J.C. Huffman, Poly-
hedron 7 (1988) 2515. (c) C.H. Zambrano, R.D. Profilet, J.E.
Hill, P.E. Fanwick, I.P. Rothwell, Polyhedron 12 (1993) 689.
[42] (a) H.G. Alt, C.E. Denner, U. Thewalt, M.D. Rausch, J.
Organomet. Chem. 356 (1988) C83. (b) I.B. Kool, M.D. Rausch,
H.G. Alt, M. Herberhold, B. Honold, U. Thewalt, J.
Organomet. Chem. 320 (1987) 37.
[43] J.L. Koenig, Spectroscopy of Polymers, American Chemical
Society, Washington, DC, 1992, p. 137.
[44] L. Resconi, R.L. Jones, A.L. Rheingold, G.P.A. Yap,
Organometallics 15 (1996) 998.
[45] U. Dietrich, M. Hackmann, B. Rieger, M. Klinga, M. Leskela¨, J.
Am. Chem. Soc. 121 (1999) 4348.
[46] C. Averbuj, M.S. Eisen, J. Am. Chem. Soc. 121 (1999) 8755.
[47] F.A. Bovey, P.A. Mirau (Eds.), NMR of Polymers, Academic
Press, San Diego, USA, 1996 (Chapter 3).
[18] V. Busico, R. Cipullo, L. Caporaso, G. Angelini, A.L. Segre, J.
Mol. Catal. Part A 128 (1998) 53 and refs therein.
[19] V. Busico, D. Brita, L. Caporaso, R. Cipullo, M. Vacatello,
Macromolecules 30 (1997) 3971.
[20] V. Busico, L. Caporaso, R. Cipullo, L. Landriani, G. Angelini,
A. Margonelli, A.L. Segre, J. Am. Chem. Soc. 118 (1996) 2105.
[21] M. Leclerc, H.H. Brintzinger, J. Am. Chem. Soc. 118 (1996)
9024 and refs therein.
[48] SHELXTL PLUS, Siemens Analyt. X-ray Inst. Inc. 1990. XS:
Program for Crystal Structure Solution, XL: Program for Crys-
tal Structure Determination, XP: Interactive Molecular Graph-
ics.
.
.