Synthesis, structure and polymerization behavior of tri- and dichloroaminopyridinato complexes of titanium

Article abstract of DOI:10.1016/j.ica.2004.11.014

Three 2-phenylaminopyridine (Ap) complexes of types ApTiCl₃ and Ap₂TiCl₂ were characterized with X-ray diffraction. The complexes were studied as polymerization catalysts using MAO as cocatalyst. In ethylene polymerization

Full text of DOI:10.1016/j.ica.2004.11.014

Inorganica Chimica Acta 358 (2005) 1061–1067
www.elsevier.com/locate/ica
Synthesis, structure and polymerization behavior of tri- and
dichloroaminopyridinato complexes of titanium
*
Markku Talja , Martti Klinga, Mika Polamo, Erkki Aitola, Markku Leskela¨
Laboratory of Inorganic Chemistry, Department of Chemistry, University of Helsinki, P. O. Box 55, A.I.V-aukio 1, Helsinki, FIN-00014, Finland
Received 24 August 2004; accepted 10 November 2004
Available online 7 February 2005
Abstract
Three 2-phenylaminopyridine (Ap) complexes of types ApTiCl3 and Ap2TiCl2 were characterized with X-ray diffraction. The
complexes were studied as polymerization catalysts using MAO as cocatalyst. In ethylene polymerization, the catalysts showed mod-
erate activities between 65 and 285 kg PE molꢀ1 Tiꢀ1 hꢀ1 and the molecular weight distributions were between 2.5 and 4.2, and the
molar masses were between 135 000 and 804 000 kg/mol.
Ó 2004 Elsevier B.V. All rights reserved.
Keywords: Crystal structures; Titanium; Aminopyridine; Amidopyridine; Aminopyridinato; Polyethylene
1. Introduction
Since 1993, we have had a particular interest in the
aminopyridinato complex family as potential a-olefin
polymerization catalysts [10].
Aminopyridinato complexes of transition metals
show a wide range of reactivities [1]. Synthesis and
exploration of this type of compounds have been the fo-
cus of a considerable amount of research in recent years
[2–5]. Recently, Kempe [6] published a microreview on
the coordination of aminopyridinato ligands.
The majority of aminopyridinato metal complexes
contain hard or borderline Lewis acids such as group
4 metals or Ru(II), although aminopyridinatos with soft
metals, such as 2-anilinopyridinate of Cu(I), have also
been proposed [7,8]. The chemistry of aminopyridinato
ligands resembles that of amidinato ligands, and both
the complexes have the same kind of coordination
spheres [9].
Previously, we have characterized several aminopy-
ridinato complexes of group 4–6 metals and in each
case a four-membered chelate ring was found [11].
Normally, aminopyridinatos prefer chelate structures
in which both nitrogens of the ligand coordinate to
the same metal center, but di- and multimetallic com-
pounds where the nitrogens of aminopyridinato ligands
coordinate to two different metal centers are also
known [1].
Amido complexes of early transition metals are com-
monly synthesized via the transmetallation reaction (salt
metathesis). Lithium reagents have traditionally been
used in the syntheses of many amido complexes. Unfor-
tunately, it is known that lithium aminopyridinatos
form the desired transition metal complexes with only
very low yields [13].
Cotton and co-workers [7] published the very first
aminopyridinato complex bis[2-(phenylamino)pyridi-
nato]bis(triphenylphosphino)ruthenium(II) in 1984.
Several research groups have used the amine elimi-
nation procedure to coordinate amido ligands to early
transition metals (group 4 metals) [5,14]. Although the
amine elimination is one of the best methods used, it
*
Corresponding author. Tel.: +358 9 191 50223; fax: +358 9 191
50198.
E-mail address: markku.talja@helsinki.fi (M. Talja).
0020-1693/$ - see front matter Ó 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.ica.2004.11.014

1062
M. Talja et al. / Inorganica Chimica Acta 358 (2005) 1061–1067
has some problems associated with complete removal
of the eliminated amine. Other well known methods
to synthesize amido complexes are the direct synthesis
[2], silyl chloride elimination [15], and alkane elimina-
tion [16]. In order to find a synthetic route to the tri-
chloroamidinato complex of titanium(IV), our
research has concentrated on HCl elimination as an
alternative reaction to the salt and amine elimination
strategies described above [17].
Previously, we have showed that direct reaction be-
tween 2-(phenylamino)pyridine and TaCl5 gives readily
the desired trichloroaminopyridinato complex of tanta-
lum [2]. In the case of titanium, the direct reaction be-
tween 2-(phenylamino)pyridine and TiCl4 did not
succeed [11]. Although several aminopyridinato com-
plexes have been reported, ApTiCl3 and Ap2TiCl2 com-
plexes have not been obtained earlier [2,11,12].
2.3. Synthesis of 2, Ti[PhNpy]2Cl2
To a solution of PhNHpy (1.7 g, 10 mmol) in dry
Et2O (25 ml) was added NaH (240 mg, 10 mmol) at
room temperature and stirred for 1 h. TiCl4 (0.55 ml,
5 mmol) was added in dry toluene (80 ml). After stirring
at ambient temperature for an additional 2 h, the solu-
tion was filtrated through Kieselguhr and the volume
was reduced under vacuum to ꢁ20 ml, and dry hexane
(20 ml) was added. Yield: 0.23 g, 5%. 1H NMR (CDCl3,
200 MHz): 7.87 (d, 2H), 7.49 (dt, 2H), 7.23 (dd, 2H),
7.12 (t, 4H), 7.02 (t, 4H), 6.67 (dd, 2H), 6.49 (d, 2H).
13C NMR: 165.8, 148.8, 142.2, 142.0, 128.9, 125.5,
121.4, 115.2, 105.0.
EI mass spectrum: m/z (relative intensity) 456 (M+,
17), 421 (26), 386 (2), 287 (15), 251 (7), 217 (12), 169
(100), 77 (50). Anal. Calc. for TiC22H18N4Cl2: C,
57.80; H, 3.97; N, 12.25; Ti, 10.47. Found: C, 57.60;
H, 4.15; N, 12.06; Ti, 10.66%.
In this paper, we describe the structures of tri- and
dichloroaminopyridinato complexes of titanium(IV)
and the dichloroaminopyridinato complex of tita-
nium(III) and their activity in ethylene polymerization.
2.4. Synthesis of 3, Ti[PhNpy]Cl2(THF)2
Addition of 2 equiv. of MeLi (5% in Et2O) (6.29 ml,
10 mmol) to TiCl4(THF)2 (1.67 g, 5 mmol) in dry Et2O
(10 ml) at ꢀ78 °C generated a red solution of the active
reagent TiMe2Cl2(THF)2. Subsequent addition of the
sodium aminopyridinato (0.01mol) in dry Et2O (25 ml)
at ꢀ78 °C resulted in an immediate evolution of gas.
After stirring at ambient temperature overnight, the
solution was filtrated through Kieselguhr and the vol-
ume was reduced under vacuum to ꢁ20 ml. Yield:
0.30 g, 14%. 1H NMR(CDCl3, 200 MHz): 7.79 (d,
1H), 7.39 (td, 1H), 7.25 (t, 2H), 7.06 (t, 2H), 6.91 (d,
1H), 6.84 (d, 1H), 6.37 (d, 1H), 3.7 (s, 8H), 1.79 (s,
8H); 13C NMR: 148.7, 142.8, 142.1, 141.9, 128.7,
125.4, 121.3, 115.0, 104.9, 68.3, 25.6. EI mass spectrum:
m/z (relative intensity) 431 (M+, 13), 396 (2), 359 (7), 287
(28), 251 (32), 169 (100), 118 (10), 77 (45). Anal. Calc.
for TiC19H24N2O2Cl2: C, 59.38; H, 6.56; N, 7.29.
Found: C, 59.80; H, 5.56; N 7.72%.
2. Experimental
2.1. General procedures
All the complexation reactions were done under ar-
gon using standard Schlenk techniques. Solvents were
dried and distilled before use, hexane, toluene and
Et2O over sodium and dichloromethane over LiAlH4.
NMR spectra were recorded at 200 MHz on a Varian
Gemini 200 spectrometer; CDCl3 as a solvent, TMS as
an internal standard. The elementary analysis determi-
nation was performed at Analytische Laboratorien, by
Professor Dr. H. Malissa und G. Reuter GmbH.
2.2. Synthesis of 1, {Ti[PhNpy]Cl3}2
Slow addition of PhNHpy (850 mg, 5 mmol) in
dichloromethane (20 ml) to a solution of TiCl4 (0.55
ml, 5 mmol) in dichloromethane (20 ml) at room tem-
perature, followed by stirring for one hour, filtration
through Kieselguhr, and removal of the solvent, led to
the dimer {Ti[PhNpy]Cl3}2. The product was obtained
as a fine dark powder and was crystallized from toluene
or CH2Cl2, affording it analytically pure as moisture
sensitive dark powder (1.57 g, 87%).
2.5. Polymerization experiments
Methylaluminoxane (30 wt%) was acquired from
Borealis Polymers Oy. Toluene (Lab Scan, analytical
grade) was refluxed with sodium and distilled under ar-
gon. Polymerization-grade ethylene was purchased from
AGA and used without further purification.
1H NMR (CDCl3, 200 MHz): 8.06 (d, 1H), 7.49 (td,
1H), 7.22–6.96 (m, 5H), 6.75 (dd, 1H), 6.29 (d, 1H).
13C NMR: 173.8, 145.6, 144.0, 142.7, 129.2, 125.5,
121.6, 114.6, 104.9. EI mass spectrum: m/z (relative
intensity) 322 (M+, 20), 287 (18), 251 (6), 169 (100),
153 (61), 118 (24), 77 (62). Anal. Calc. for
TiC11H9N2Cl3: C, 40.85; H, 2.80; N, 8.66. Found: C,
40.47; H, 2.85; N 7.71%.
Polymerization runs were performed at a constant
monomer concentration in a Buchi 1.0 l glass autoclave
¨
with a Julabo ATS-3 temperature control unit. Mechan-
ical stirring was applied at a stirring speed of 800 rpm.
During polymerization, the partial pressure of ethylene
was maintained constant by an electronic controlling
system. Ethylene consumption was measured with a cal-
ibrated mass flow meter and monitored on line together

M. Talja et al. / Inorganica Chimica Acta 358 (2005) 1061–1067
1063
Table 1
Crystal data, data collection and refinement parameters for compounds 1, 2, and 3
Data
1
2
3
Chemical formula
Formula weight
Crystal system
Space group
T (K)
C22H18Cl6N4Ti2
323.45
triclinic
C22H18Cl2N4Ti
457.20
triclinic
C19H25Cl2N2O2Ti
432.21
monoclinic
P21/a
193(2)
ꢀ
ꢀ
P1
193(2)
P1
193(2)
8.947(2)
13.676(4)
8.931(3)
91.69(3)
93.79(2)
78.43(2)
1068.1(5)
2
˚
a (A)
8.685(5)
9.992(5)
8.651(5)
112.517(5)
107.033(5)
80.785(5)
662.1(6)
2
20.077(5)
9.236(6)
22.473(7)
˚
b (A)
˚
c (A)
˚
a (A)
˚
b (A)
92.35(3)
˚
c (A)
3
˚
V (A )
4164(3)
8
Z
l (mmꢀ1
)
1.228
2567
2399
0.666
4128
3863
0.683
7672
7453
Number of data collected
Number of unique data
Rint
0.0353
0.0570
0.0625
0.0341
0.0526
0.0902
0.1346
0.0805
0.1669
Final R(jFj) for Fo > 2r(Fo)
Final R(F2) for all data
with the temperature and pressure inside the vessel using
TM
data sets were collected at ꢀ80 °C. Intensity data sets
were recorded on an automated four-circle Rigaku
AFC-7S diffractometer using graphite-monochroma-
the GENIE
computer program for data acquisition.
Toluene (250 ml) and the cocatalyst were introduced
to the argon-purged reactor. Once the polymerization
temperature had been reached, the vessel was charged
with ethylene to the appropriate pressure. Polymeriza-
tion was started by introducing 20 ml of the catalyst pre-
cursor solution into the reactor. Polymerization was
quenched by pouring the contents of the vessel into
methanol acidified with a small amount of HCl. The so-
lid polymer was filtered off, washed with methanol and
dried under vacuum.
˚
tized Mo Ka radiation (k = 0.71073 A). Reflection
intensities over background were collected using 2h–x
scans. The intensities of the three standard reflections,
recorded after every 200 intensity scans, showed only
standard fluctuations or minor decay (<2.5%) in all
measurements. The data set was compressed to reflec-
tion files with the TEXSAN software.
2.7. Structure determination and refinement
For the analysis of polymer molecular weights and
molecular weight distributions, a Waters 150-C GPC
chromatograph was used operating at 145 °C with
1,2,4-trichlorobenzene solvent and polystyrene calibra-
tion standards. Melting temperatures of preheated and
cooled samples were determined by differential scanning
calorimetry using a Perkin–Elmer DSC-2 calorimeter.
Structures were solved with the SHELXTL PC 4.1 pro-
gram package using direct methods that showed the
positions of the non-hydrogen atoms. Further refine-
ment with full-matrix least squares on F2 was carried
out with SHELXL-97 [19] using all collected reflections.
Non-hydrogen atoms were refined anisotropically.
Hydrogen atoms were introduced into calculated
positions.
2.6. X-ray crystallography
Table 1 lists a summary of crystallographic data for
all structurally characterized compounds. In each case,
a portion of air-sensitive but thermally stable crystals
was transferred to perfluoroether. The crystal selected
for the X-ray measurements was mounted on a glass fi-
ber using the oil-drop method [18]. All X-ray diffraction
3. Results
The ligand precursor, 2-phenylaminopyridine, was
treated with one equivalent of TiCl4 [16]. After evapora-
tion of the solvent, complex 1 was obtained as a dark
CH2Cl2
1h, rt
+
HCl
+
TiCl4
N
N
N
H
N
Cl Ti Cl
Cl
Scheme 1. Synthesis of trichloroaminopyridinato complex of Ti(IV) 1.

1064
M. Talja et al. / Inorganica Chimica Acta 358 (2005) 1061–1067
1)
2) ½TiCl4/-78˚C
NaH/Et2O/rt
N
N
N
N
Cl
Cl
Ti
N
H
N
Scheme 2. Synthesis of dichloroaminopyridinato complex of tita-
nium(IV) 2.
Cl
1) 2 MeLi/Hexane/-78˚C
THF
Cl
Cl
Ti
N
N
THF
2)
Et2O/NaAp/-78˚C
THF Ti THF
Cl
Cl
Fig. 1. Structural representation of 1. Hydrogen atoms are omitted for
clarity. The thermal ellipsoids correspond to 30% probability.
Cl
Scheme 3. Synthesis of dichloroaminopyridinato complex of tita-
nium(III) 3.
try. Four nitrogen atoms and two chloride atoms define
the co-ordination sphere of titanium. Compared to com-
plex 1, complex 2 has no chloro bridges, and the main
reasons for a distorted octahedral geometry are the
two aminopyridinato four-membered rings: N1–Ti–N7
and N14–Ti–N20 are 63.43(12)° and 63.19(11)°, respec-
tively. The wide Cl1–Ti–Cl2 angle 97.03(5)° compared
to 94.43° as that in Cp2TiCl2 implies a sterically unhin-
dered Ti center in the solid state [17].
Compound 3 crystallizes with two similar complex
molecules in the asymmetric unit (Fig. 3). The coordina-
tion sphere around the Ti atoms is pseudo-octahedral,
with two THF ligands in trans positions, and two chlo-
rides in cis positions. The coordination sphere of both Ti
atoms consists of two nitrogen atoms, two oxygen
purple solid material in good yield. Recrystallization
from toluene produced crystals suitable for X-ray mea-
surements (Scheme 1).
Attempts to replace another chloride of TiCl4 to the
aminopyridinato ligand in the titanium coordination
sphere failed by using the HCl elimination method.
However, the transmetallation synthesis yielded the met-
allated species 2 upon standing for a period of 5–6
weeks. The yield was 5% (Scheme 2).
Reduction of titanium(IV) occurred when we tried to
first methylate TiCl4(THF)2 by MeLi to get TiCl2Me2
and then let it react with sodium aminopyridinato
(NaAp) in diethyl ether (Scheme 3). The elemental anal-
ysis of the raw product indicated that the result of the
synthesis was not the expected Ti(IV)-complex. Recrys-
tallization of the mixture gave a few red-brown crystals
of TiApCl2(THF)2 3, and the X-ray structure and the
elemental analysis revealed the titanium(III) species 3.
In polymerization experiments, crystals of 3 were used
to ensure that the structure of the catalyst was correct.
Complex 1 is a centrosymmetric dimer (Fig. 1). The
chloro bridges link the two titanium centers. Each tita-
nium atom is located in a distorted octahedral geometry
defined by the two nitrogen atoms of the aminopyridi-
nato and four chloride atoms as is in the case of the tet-
rachloro[2-(phenylamino)pyridinato]titanate(IV) anion
[2]. The distorted octahedral geometry is due to the nar-
row angles of the aminopyridinato four-membered me-
tal rings N1–Ti1–N7 = 64.13(12)° and Cl3–Ti1–Cl30 =
79.18(4)° of the Ti2Cl2 core. The Ti–N(pyridine), Ti–
N(amide), and Ti–Cl bond distances are as expected
[Ti1–N1 = 2.156(3); Ti1–N7 = 1.949(3); Ti1–Cl1 =
2.2311(14)]. The shorter distance of Ti1–N7 compared
to Ti1–N1 leads to a geometry where the angles N7–
Ti1–Cl1 and N7–Ti1–Cl3 are greater than respective
N1–Ti1–Cl1 and N1–Ti1–Cl3 angles.
Complex 2 has no symmetry elements (Fig. 2). The
titanium center also has a distorted octahedral geome-
Fig. 2. Structural representation of 2. Hydrogen atoms are omitted for
clarity. The thermal ellipsoids correspond to 30% probability.

M. Talja et al. / Inorganica Chimica Acta 358 (2005) 1061–1067
1065
Table 3
Selected bond lengths (A) and angles (°) for Ap2TiCl2 (2)
˚
Ti1–N1
Ti1–N7
Ti1–N14
T1–N20
Ti1–Cl1
Ti1–Cl2
2.135(3)
2.006(3)
2.203(3)
1.961(3)
2.2553(13)
2.3199(13)
N1–Ti1–N20
N7–Ti1–Cl1
N7–Ti1–Cl2
N7–Ti1–N14
N7–Ti1–N20
N14–Ti1–Cl1
N14–Ti1–Cl2
N14–Ti1–N20
N20–Ti1–Cl1
N20–Ti1–Cl2
Cl1–Ti1–Cl2
143.56(12)
97.45(10)
150.38(9)
84.31(12)
97.23(12)
161.61(8)
89.92(8)
N1–Ti1–Cl1
N1–Ti1–Cl2
N1–Ti1–N7
N1–Ti1–N14
113.80(9)
63.19(11)
98.48(9)
87.08(9)
63.43(12)
83.43(11)
106.00(10)
97.03(5)
Table 4
˚
Selected bond lengths (A) and angles (°) for ApTiCl2(THF)2 (3)
Ti1–N1
Ti1–N7
T1–O19
Ti1–O14
Ti1–Cl1
Ti1–Cl2
2.160(5)
2.113(5)
2.096(4)
2.100(4)
2.346(2)
2.324(2)
Ti2–N51
Ti2–N57
Ti2–O64
T2–O69
Ti2–Cl3
Ti2–Cl4
2.173(5)
2.070(5)
2.084(4)
2.141(5)
2.344(2)
2.348(2)
Fig. 3. Structural representation of 3. Hydrogen atoms are omitted for
clarity. The thermal ellipsoids correspond to 25% probability.
N1–Ti1–Cl2
N7–Ti1–Cl1
N1–Ti1–N7
O14–Ti1–O19
Cl1–Ti1–Cl2
157.98(15)
155.68(15)
62.06(19)
176.96(17)
108.18(8)
N51–Ti2–Cl4
N57–Ti2–Cl3
N51–Ti2–N57
O64–Ti2–O69
Cl3–Ti2–Cl4
158.17(14)
157.33(15)
62.77(19)
177.47(17)
106.93(8)
atoms, and two Cl atoms. In complex 3, the aminopyrid-
inato titanium four-rings have the smallest N–Ti–N an-
gles of the complexes 1–3, N1–Ti–N7 = 62.06(19)° and
N57–Ti2–N51 = 62.77(19)°.
Table 1 presents the crystal data and structural refine-
ment for the complexes 1–3. Results were in accordance
with the literature data on aminopyridinato complexes
[1,4,5,14,18]. The selected bond lengths and angles of
complexes 1–3 are shown in Tables 2–4.
tion trend because of the low molecular weight (Table
5). All the aminopyridinato complexes 1–3 resulted in
polymers with narrower molecular weight distributions
than the polymer obtained with TiCl4. The open struc-
ture of the complexes 1 and 3 could explain the high
molecular weights of the polymers. Elimination of the
polymer chain is not very probable if the monomer
can coordinate to the active metal center. In the case
of complex 2, due to steric reasons, the elimination of
the polymer chain was more probable than in the case
of 1 and 3.
Typically, the catalysts 1–3 had a very good olefin up-
take during the very first minutes of the polymerization
process, and the activity was about 2000
kg PE molꢀ1 Tiꢀ1 hꢀ1 excluding the dissolution of ethyl-
ene. After that, the olefin consumption was nearly con-
stant and the overall activities in 1 h reaction were about
100–300 kg PE molꢀ1 Tiꢀ1 hꢀ1. The titanium(III) com-
plex 3 was more active than the titanium(IV) complexes
1 and 2. The activities of the complex-MAO systems 1
and 2 increased with temperature.
3.1. Polymerization studies
The polymerization activities of the catalyst precur-
sors 1–3 were studied in the ethylene polymerization test
reaction. The cationic catalysts were generated by the
reaction of the corresponding complexes with methyl-
aluminoxane (MAO). With MAO as a cocatalyst, the
catalysts 1–3 yielded polyethylene with relatively high
molecular weight, only run 5 did not fit the polymeriza-
Table 2
˚
Selected bond lengths (A) and angles (°) for APTiCl3 (1)
Ti1–N1
Ti1–N7
Ti1–Cl1
Ti1–Cl2
Ti1–Cl3
Ti1–Cl30
2.156(3)
1.949(3)
N7–Ti1–N1
N7–Ti1–Cl1
N7–Ti1–Cl2
N7–Ti1–Cl3
N7–Ti1–Cl30
Cl1–Ti1–Cl2
Cl1–Ti1–Cl30
Cl2–Ti1–Cl30
Cl3–Ti1–Cl1
Cl3–Ti1–Cl2
Cl3–Ti1–Cl30
64.13(12)
101.18(9)
99.91(9)
86.14(9)
149.45(9)
98.13(5)
90.92(5)
106.06(5)
169.60(4)
87.75(4)
79.18(4)
2.2311(14)
2.2357(15)
2.5506(15)
2.4134(13)
The melting temperatures of the obtained polymers
were typical for HDPE, except that run 7 was LDPE.
The high molecular weight distribution of the obtained
polymers (2.5–4.2) indicated that there was not a single
well-defined active species, but several undefined ones.
The increased polymerization pressure resulted in higher
molecular weights and narrower molecular weight
distributions.
N1–Ti1–Cl1
N1–Ti1–Cl2
N1–Ti1–Cl3
N1–Ti1–Cl30
94.08(9)
161.65(8)
82.39(8)
87.30(9)

1066
M. Talja et al. / Inorganica Chimica Acta 358 (2005) 1061–1067
Table 5
Results of ethylene polymerizations promoted by complexes 1–3 with MAOa
Activityc
Mw (g molꢀ1
)
Mw/Mn
b
Run
Catalyst
Tp (°C)
Pp (bar C2H4)
Yield (g)
Tm (°C)
1
2
3
4
5
6
7
8
9
a
1
60
60
60
80
80
80
60
60
60
5
4
3
5
4
3
5
5
5
2.72
2.04
1.78
3.78
3.54
1.30
1.90
5.70
11.12
134
134
134
134
134
136
124
135
133
136
102
89
804 000
39 0000
383 000
651 000
135 000
506 000
239 000
799 000
15 000
2.9
4.2
4.1
2.5
3.5
3.6
2.4
3.0
8.6
1
1
1
189
177
65
1
1
2
95
3
TiCl4
285
1112
Polymerization conditions: [Al]:[Ti] = 3000:1; ncat = 20 lmol; tp = 3600 s (run 9; tp = 1800 s).
b
Melting point determined by DSC. These values have been obtained from remelted samples at heating rate of 10 °C/min.
c
Activity of catalyst in kg PE molꢀ1 Tiꢀ1 hꢀ1
.
4. Discussion
merization activity. Decrease in electron density at the
amido nitrogen should unstabilize the aminopyridinato
complex and hence increase its reactivity.
In the case of complex 1, the formation of the salt,
2-(phenylamino)pyridinium tetrachloro[2-(phenylamino)-
pyridinato-N,N0]titanate(IV) (C11H11N2)-[TiCl4-(C11H9-
N2)] [11], can be avoided by adding the ligand dropwise
to the solution of TiCl4. Mechanistically, the pyridine part
of the aminopyridine moiety coordinates first to TiCl4 and
then the amine part reacts with the titanium center elimi-
nating HCl because there is no free base in the solution left
[21,22].
A possible explanation for the poor yield of the com-
plex 2 is that a significant amount of Ap3TiCl is formed
compared to Ap2TiCl2. A better yield (50%) can be
achieved by adding TiCl4 first in to a Schlenk tube
and then followed by the sodium aminopyridinato
ligand.
In complex 3, the titanium(III), TiCl3(THF)2 proba-
bly formed before the sodium aminopyridinato was
added in to the solution. We propose that the reduction
of titanium occurs via a reductive elimination step; the
reaction involves methylation of TiCl4(THF)2 (1), then
the reductive elimination of TiMe2Cl2(THF)2 (2), and
then the Ti(II)-species reduces the Ti(IV)-species (3).
Kempe has shown that titanium and zirconium
aminopyridinatos are not configurationally stable in
solution because the pyridine nitrogen is not very
strongly coordinated to the titanium center [20], and
that would increase the amount of active sites of the cat-
alysts 1–3 and explain the broader polydispersity. The
molecular weight distribution should be influenced by
the bond length of the titanium aminopyridinato ligand.
The shorter the bond length, the better is the polydisper-
sity, because the aminopyridinato ligand is closer to the
active titanium center. The molecular weight distribu-
tion increases inversely with bond length between the
titanium center and aminopyridinato ligand. On the
contrary, the activity of the catalyst increases with bond
length between the titanium center and aminopyridinato
ligand. Evidently, the optimum situation is a compro-
mise between the activity and the polydispersity where
we have the active titanium center at the proper distance
from the aminopyridinato ligand.
In previous aminopyridinato studies [5,20,24], ethyl-
ene polymerization activities have been reported to be
between 10 and 20 kg PE molꢀ1 Tiꢀ1 hꢀ1 barꢀ1. In our
case, the activity was between 19 and 57
kg PE molꢀ1 Tiꢀ1 hꢀ1 barꢀ1. It should be noticed that
we have used a higher polymerization pressure and tem-
perature than the other research groups. Thus, our poly-
dispersities of polyethylene are significantly narrower
than those reported earlier, below 5 and well above 20,
respectively, and the molar masses were also higher,
maximum 804 000 g molꢀ1 in our study, and 326 000
g molꢀ1 in the literature reports [5,14,20,24].
TiCl4ðTHFÞ þ 2MeLi ! TiMe2Cl2ðTHFÞ þ 2LiCl
2
2
ð1Þ
ð2Þ
TiMe2Cl2ðTHFÞ ! TiCl2ðTHFÞ þ EtH
2
2
TiCl2ðTHFÞ þ TiMe2Cl2ðTHFÞ
2
2
! TiCl3ðTHFÞ þ TiMe2ClðTHFÞ
ð3Þ
2
2
The ethylene polymerization activity of TiCl4 com-
pared to the corresponding aminopyridinato titanium
catalyst 1–3 could be explained by Lewis acidity [23].
The aminopyridinato ligand is a more electron donating
and sterically demanding ligand than chloride. There-
fore, the Lewis acidity is greater for titanium tetrachlo-
ride than for the complexes 1–3. Electron withdrawing
groups on the amido nitrogen should increase the poly-
5. Supplementary materials
Crystallographic data (excluding structure factors)
for the structure reported in this paper have been depos-
ited with the Cambridge Crystallographic Data Centre:

M. Talja et al. / Inorganica Chimica Acta 358 (2005) 1061–1067
1067
[10] K. Hakala, B. Lo¨fgren, M. Polamo, M. Leskela¨, Macromol.
Rapid Commun. 18 (1997) 635.
CCDC 241710–CCDC 241712. Copies of the data can
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 http://
www.ccdc.cam.ac.uk).
[11] M. Polamo, M. Leskela¨, Acta Crystallogr. C52 (1996) 2975.
[12] (a) M. Polamo, Acta Crystallogr. C52 (1996) 2977;
(b) M. Polamo, Acta Crystallogr. C52 (1996) 2980;
(c) M. Polamo, M. Leskela¨, Acta Chem. Scand. 51 (1997) 69;
(d) M. Polamo, M. Leskela¨, Acta Chem. Scand. 51 (1997)
449;
(e) M. Polamo, I. Mutikainen, M. Leskela¨, Acta Crystallogr.
C53 (1996) 1036;
Acknowledgments
(f) M. Polamo, M. Leskela¨, Acta Chem. Scand. 51 (1997) 709.
[13] (a) E. Benzing, W. Kornicker, Chem. Ber. 94 (1961) 2263;
(b) R. Kempe, A. Spannenberg, S. Brenner, Z. Kristallogr. – NCS
211 (1996) 567.
This work was supported by the Finnish National
Technology Agency (TEKES) and Borealis Polymers
Oy. Ms. Hedvig Byman–Fagerholm (Lic. Techn.) is
thanked for carrying out the GPC measurements.
[14] H. Fuhrmann, S. Brenner, P. Arndt, R. Kempe, Inorg. Chem. 35
(1996) 6742.
[15] F. Guerin, D.H. McConville, N.C. Payne, Organometallics 15
(1996) 5085.
References
[16] Y.X. Chen, T.J. Marks, Organometallics 16 (1997) 3649.
[17] A. Clearfield, D.K. Werner, C.H. Saldarriaga–Molina, R. Ropal,
Can. J. Chem. 53 (1975) 1622.
[1] (a) F.A. Cotton, E.A. Hillard, C.A. Murillo, X. Wang, Inorg.
Chem. 42 (2003) 6063;
[18] T. Kottke, D. Stalke, J. Appl. Crystallogr. 26 (1993) 615.
[19] G.M. Sheldrick, SHELXL-97, Program for the Refinement of
Crystal Structures, University of Go¨ttingen, Germany, 1997.
[20] (a) C. Morton, P. OÕShaughnessy, P. Scott, Chem. Commun.
(2000) 2099;
(b) J. Ashenhurst, L. Braccaleon, S. Gao, W. Liu, H. Schmider,
S. Wang, G. Wu, Q.G. Wu, Organometallics 17 (1998) 5334;
(c) M.L. Cole, P.C. Junk, New J. Chem. 27 (2003) 1032.
[2] M. Polamo, M. Leskela¨, J. Chem. Soc., Dalton Trans. (1996)
4345.
(b) R. Kempe, P. Arndt, Inorg. Chem. 35 (1996) 2644.
[21] J.C. Flores, M.D. Rausch, US Patent 5502128, 1996.
[22] G.B. Barlin, W. Pfleider, J. Chem. Soc. B (1971) 1425.
[23] (a) K. Oita, T.D. Nevitt, J. Polym. Sci. 43 (1960) 585;
(b) P. Sobota, J. Ejfler, S. Szafert, K. Szczegot, W. Sawka-
Dobrowolska, J. Chem. Soc., Dalton Trans. (1993) 2353;
(c) M.L. Cooper, J.B. Rose, J. Chem. Soc. (1959) 795;
(d) A. Malatesta, Can. J. Chem. 37 (1959) 1176;
(e) D. Ludlum, A. Anderson, C. Ashby, J. Chem. Soc. 80 (1958)
1380;
[3] R. Kempe, Angew. Chem. 112 (2000) 478.
[4] C. Jones, P.C. Junk, S.G. Leary, N.A. Smithies, Inorg. Chem.
Commun. 6 (2003) 1126.
[5] I. Westmoreland, I.J. Munslow, P.N. OÕShoughnessy, P. Scott,
Organometallics 22 (2003) 2972.
[6] R. Kempe, Eur. J. Inorg. Chem. (2003) 791.
[7] A.R. Chakravarty, F.A. Cotton, E.S. Shamshoum, Inorg. Chim.
Acta 86 (1984) 5.
[8] J.M. Seco, M.J.G. Garmendia, E. Pinilla, M.R. Torres, Polyhe-
dron 21 (2002) 457.
(f) R.J. Kern, H.G. Hurst, J. Polym. Sci. (1960) 272;
(g) M.K. Skalli, A. Markovits, C. Minot, A. Belmajdoub, Catal.
Lett. 76 (2001) 7;
[9] (a) J. Richter, F.T. Edelmann, M. Noltemeyer, H.-G. Schmidt,
M. Shmulinson, M.S. Eisen, J. Mol. Cat. A 130 (1998) 149;
(b) J.R. Hagadorn, J. Arnold, J. Chem. Soc., Dalton Trans.
(1997) 3087;
(h) R.J.H. Clark, A.J. McAlees, J. Chem. Soc. A (1970) 2027;
(i) R.J.H. Clark, M.A. Coles, J. Chem. Soc., Dalton Trans. (1972)
2454.
(c) J.R. Hagadorn, J. Arnold, Organometallics 17 (1998) 1355;
(d) S. Hao, K. Feghali, S. Gambarotta, Inorg. Chem. 36 (1997)
1745.
[24] M. Oberthur, P. Arndt, R. Kempe, Chem. Ber. 129 (1996)
1087.