Synthesis, characterization, and butadiene polymerization of iron(III), iron(II) and cobalt(II) chlorides bearing 2,6-bis(2-benzimidazolyl)pyridyl or 2,6-bis(pyrazol)pyridine ligand

Article abstract of DOI:10.1016/j.jorganchem.2011.11.025

Iron(III), iron(II) and cobalt(II) complexes bearing neutral N,N,N-tridentate ligand (Fe(III)L1, 1a; Fe(III)L2, 1b; Fe(III)L3, 1c; Fe(III)L4, 2a; Fe(III)L5, 2b; Fe(III)L6, 2c; Fe(II)L1, 3a; Fe(II)L2, 3b; Fe(II)L3, 3c; Fe(II)L4, 4a; Fe(II)L5, 4b; Fe(II)L6, 4c; CoL1, 5a; CoL2, 5b; CoL3, 5c; CoL5, 6b and CoL6, 6c) have been synthesized from the metal chlorides (FeCl₃, FeCl₂·4H₂O or CoCl₂) by treating the corresponding ligands (2,6-bis(benzimidazol-2-yl)pyridine, L1; 2,6-bis(1′-ethylbenzimidazol-2′-yl)pyridine, L2; 2,6-bis(1′-benzylbenzimidazol-2′-yl)pyridine, L3; 2,6-bis(pyrazol)pyridine L4; 2,6-bis(3-methylpyrazol)pyridine, L5 or 2,6-bis(3,5-dimethylpyrazol)pyridine, L6. The complexes are characterized by FTIR and elemental analyses. The structures of complexes 1a, 1b, 3a, 3b, 3c, 4a, 5a, 5b, 5c and 6c are further confirmed by X-ray crystallographic analyses. Six coordination iron(III) (1a, 1b and 1c) and iron(II) (3c*DMF) complexes adopt a distorted octahedral configuration with the equatorial plane formed by the three nitrogen atoms and one chlorine atom, while the iron(II) (3a, 3b and 4a) and cobalt complexes (5a, 5b, 5c and 6c) adopt a trigonal bipyramidal configuration with the equatorial plane formed by the pyridyl nitrogen atoms and the two chlorine atoms. All complexes are evaluated as precursors for 1,3-butadiene polymerization in the presence of cocatalyst in toluene at room temperature. Iron(III) and iron(II) bearing the same ligand show comparable catalytic performance. The catalytic activity and selectivity are significantly influenced by the ligand structure, with the latter being tunable within a wide range from cis-1,4 to trans-1,4. Cobalt complexes, in combination with either MAO or aluminum chlorides shown high activity, and high cis-1,4-selectivity irrespective of the structure of ligand backbone.

Full text of DOI:10.1016/j.jorganchem.2011.11.025

Journal of Organometallic Chemistry 702 (2012) 10e18
Contents lists available at SciVerse ScienceDirect
Journal of Organometallic Chemistry
journal homepage: www.elsevier.com/locate/jorganchem
Synthesis, characterization, and butadiene polymerization of iron(III),
iron(II) and cobalt(II) chlorides bearing 2,6-bis(2-benzimidazolyl)pyridyl
or 2,6-bis(pyrazol)pyridine ligand
,
,
,
a
Dirong Gong a, Xiaoyu Jia a b, Baolin Wang a b, Xuequan Zhang a , Liansheng Jiang
*
a Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, 5625 Renmin Street, Changchun 130022, PR China
b Graduate School of the Chinese Academy of Sciences, Beijing 100049, PR China
a r t i c l e i n f o
a b s t r a c t
Article history:
Iron(III), iron(II) and cobalt(II) complexes bearing neutral N,N,N-tridentate ligand (Fe(III)L1, 1a; Fe(III)L2,
1b; Fe(III)L3, 1c; Fe(III)L4, 2a; Fe(III)L5, 2b; Fe(III)L6, 2c; Fe(II)L1, 3a; Fe(II)L2, 3b; Fe(II)L3, 3c; Fe(II)L4, 4a;
Fe(II)L5, 4b; Fe(II)L6, 4c; CoL1, 5a; CoL2, 5b; CoL3, 5c; CoL5, 6b and CoL6, 6c) have been synthesized from
the metal chlorides (FeCl3, FeCl2$4H2O or CoCl2) by treating the corresponding ligands
Received 25 September 2011
Received in revised form
14 November 2011
Accepted 17 November 2011
(2,6-bis(benzimidazol-2-yl)pyridine,
L1;
2,6-bis(10-ethylbenzimidazol-20-yl)pyridine,
L2;
2,6-
bis(10-benzylbenzimidazol-20-yl)pyridine, L3; 2,6-bis(pyrazol)pyridine L4; 2,6-bis(3-methylpyrazol)
pyridine, L5 or 2,6-bis(3,5-dimethylpyrazol)pyridine, L6. The complexes are characterized by FTIR and
elemental analyses. The structures of complexes 1a, 1b, 3a, 3b, 3c, 4a, 5a, 5b, 5c and 6c are further
confirmed by X-ray crystallographic analyses. Six coordination iron(III) (1a, 1b and 1c) and iron(II)
(3c*DMF) complexes adopt a distorted octahedral configuration with the equatorial plane formed by the
three nitrogen atoms and one chlorine atom, while the iron(II) (3a, 3b and 4a) and cobalt complexes (5a,
5b, 5c and 6c) adopt a trigonal bipyramidal configuration with the equatorial plane formed by the pyridyl
nitrogen atoms and the two chlorine atoms. All complexes are evaluated as precursors for 1,3-butadiene
polymerization in the presence of cocatalyst in toluene at room temperature. Iron(III) and iron(II) bearing
the same ligand show comparable catalytic performance. The catalytic activity and selectivity are
significantly influenced by the ligand structure, with the latter being tunable within a wide range from
cis-1,4 to trans-1,4. Cobalt complexes, in combination with either MAO or aluminum chlorides shown
high activity, and high cis-1,4-selectivity irrespective of the structure of ligand backbone.
Ó 2012 Published by Elsevier B.V.
Keywords:
Catalyst
2,6-bis(2-benzimidazolyl)pyridyl
2,6-bis(pyrazol)pyridine
Butadiene
Polybutadiene
1. Introduction
the potent application of the resultant polymers, attention has been
particularly focused on the late transition metals such as iron and
The past half century has witnessed tremendous growth in
interest of the applicable catalysts for selective polymerization of
conjugated dienes (butadiene and isoprene) as their extensive and
diverse applications in automobile industry [1e5]. Therefore,
diversified catalyst recipes based on transition metals and lantha-
nides have been invented to afford polybutadiene and polyisoprene
with various enchainments [6e20]. Excellent activity and selec-
tivity, even the superior controllable features in terms of micro-
structures, molecular weight as well as molecular weight
distribution of the resultant polymers, have been achieved, though
mostly interested in fundamental researches and practical appli-
cation of the catalysts have not yet been demonstrated. Considering
cobalt-based catalysts, the advantage of which are manifold,
spanning from the ease of preparation and handling to the use of
low-cost metals with negligible environmental impact.
We have been interested in iron and cobalt catalyzed butadiene
polymerization over the past decades. Cobalt-based catalysts are
known as the versatile recipes for initiating cis-1,4, trans-1,4 and
1,2- polymerization of butadiene, among them, much attention is
paid to those for high cis-1,4-selectivity, because the resulted
polymer is the most important elastomer used for tires and other
elastic materials. Thus, a large number of systems [21e31] have
been screened for initiating cis-1,4 selective polymerization. To
understand the nature of active species of cobalt catalysts, effort
has also been paid to the well-defined cobalt-based catalysts as
well as exploring their catalytic performance of homogeneous
model systems. Comparatively, researches on iron catalytic system
have been limited mainly because of the low activity and poor
* Corresponding author. Tel.: þ86 431 8526 2303; fax: þ86 431 8526 2307.
E-mail address: xqzhang@ciac.jl.cn (X. Zhang).
0022-328X/$ e see front matter Ó 2012 Published by Elsevier B.V.
doi:10.1016/j.jorganchem.2011.11.025

D. Gong et al. / Journal of Organometallic Chemistry 702 (2012) 10e18
11
selectivity. The activity has been improved by using 1,10-
phenanthroline or 2,20-bipyridine as the supporting ligands, and
high selectivity (1,2-selectivity: 91%) can also be available at ꢀ78 ꢁC,
at expense of activity [32,33]. Recently, progresses have been made
since dialkylphosphite and phosphate are invented as an important
additive in iron-based catalyst in our group, with which syndio-
tactically enriched 1,2-polybutadiene (1,2-: 87e95%, pentad rrrr:
65e95%) has been obtained under benign polymerization condi-
tions [18,20,34e38]. To date, iron catalysts are mainly for 1,2-
selectivity, and those based on high 1,4-selectivity are quite few
[39e43]. Meanwhile, for some of cobalt, and most iron catalyst
recipes [18,20,21], the active species are generated in situ by mixing
the components (cobalt or iron compound, additive and cocatalyst),
therefore, little is known about the structure of precursor and
catalystepolymer correlation. Thus, investigation of new cobalt and
iron catalyst that is homogeneous, well-defined, straightforward
and high yielding in synthesis, and providing highly 1,4 orientated
catalysis has still been the subject.
Over the past decade, the use of neutral tridentate ligands has
witnessed a momentous advancement in olefin area. This ligand set
has led to success in olefin polymerization catalysis due to the
ability to manipulate catalytic activity, comonomer incorporation,
as well as molecular weight, and stereochemistry of the polymer
produced by choice of substituents, leading to rational catalyst
design. Especially, the impressive catalytic performance observed
in the 2,6-bis(imino)pyridyl ligated transition metal derivatives for
ethylene polymerization and oligomerization, indicates the signif-
icant scope for incorporating such kind of ligand into late transition
metal-based catalysts. As such, various analogous tridentate
ligands such as N^N^N [44e52], N^X^N (X ¼ C, O, S et al.) [53e59]
and X^N^X (X ¼ O, P, S et al.) [60,61] ligated complexes have also
been proved to effectively promote activity, improve selectivity,
and govern the molecular weight of products. The exceptional
performance of such ligand set in olefin polymerization may also be
relevance to butadiene area and catalytic performance of transition
metals bearing such set of ligands for butadiene polymerization,
however, is less established [12,13,15,62,63]. We have reported the
high active 2,6-bis(imino)pyridyl ligands ligated transition
complexes for selectivity-controlled polymerization of butadiene
[39,42,43], and found the selectivity could be manipulated within
a wide range, from high cis-1,4 (97.0%) to high trans-1,4 (95.4%) just
by either altering the metal center or modifying the ligand envi-
ronments. The exceptional performance of 2,6-bis(imino)pyridyl
prompts a more thorough investigation of other N,N,N-tridentate
ligand derivatives. In this manuscript, we present coordination
chemistry of new iron(II), iron(III), cobalt dichlorides supported by
2,6-bis(2-benzimidazolyl)pyridyl
or
2,6-bis(pyrazol)pyridine
ligands and their catalytic behaviors in butadiene polymerization.
The combination of tunable ligand environment and metal center
could also allow easy stereo and electronic modification, which are
both highly desirable in the study of ligand and metal effect on both
activity and selectivity, ultimately open a door to desired catalyst
recipes.
2. Results and discussion
2.1. Syntheses and characterization of ligands and complexes
2,6-bis(2-benzimidazolyl)pyridine (L1) could be synthesized
from condensation of 1,2-phenylenediamines and pyridine-
2,6-dicarboxylic acid in the presence of polyphosphoric acid (PPA)
at 240 ꢁC [52]. The 2,6-bis(2-benzimidazolyl)pyridine derivatives
(L2 and L3) could be readily prepared by N-alkylation of L1 in
acetone at or above room temperature. The stretching frequencies
of all ligands fall within the following ranges: v(NeH) is about
3055 cmꢀ1 for L1, v(C]C) is between 1584 and 1600 cmꢀ1, and v(C]N)
is between 1567 and 1572 cmꢀ1. Ligands (L4eL6) were prepared as
the reported methods [64]. The ligands synthesized were also
confirmed by elemental analysis and NMR spectra.
Complexes (1ae6c) were synthesized in high yields by reaction
of the metal chlorides (FeCl3, FeCl2$4H2O, CoCl2) with one equiva-
lent of the corresponding ligands (L1eL6, Scheme 1) in THF at room
temperature. The products were collected by filtration, washed
with diethyl ether and dried under vacuum. They are stable toward
oxidation and hydrolysis. The one-step complexation at room
temperature, giving high yields (85e91%), represents an additional
advantage. The elemental analyses results reveal that the compo-
nents of all complexes are in accord with the formula MtClxL. The IR
spectra of the free ligands show that the C]N stretching
frequencies appear at 1561e1577 cmꢀ1. In complexes, the C]N
stretching
vibrations
shift
toward
lower
frequencies
1667e1707 cmꢀ1 and are greatly reduced in intensity, which indi-
cated the coordination interaction between the imino nitrogen
atoms and the metal ions. In order to establish the coordination
mode of the complexes, complexes 1a, 1b, 2a, 3a, 3b, 3c, 5a, 5b, 5c
and 6c were selected for structural analysis by single-crystal X-ray
diffraction and the crystal structures of 1b, 3c and 5a were depicted
in Fig. 1, Fig. 2 and Fig. 3, respectively. Crystal data and collection
parameters were collected in Table 1 and Table 2, and bond
distances and angles were compiled in Table 3 and Table 4.
X-ray analysis revealed that the conformers of 1a and 1b (Fig. 1)
have essentially identical structures, both displaying a slightly
Scheme 1. Synthesis of 2,6-bis(2-benzimidazolyl)pyridine derivatives (L1eL3) and their complexes (1ae1c, 3ae3c and 5ae5c).

12
D. Gong et al. / Journal of Organometallic Chemistry 702 (2012) 10e18
Fig. 3. ORTEP view of the complex 5a, drawn at 35% of probability. Hydrogen atoms
were omitted for clarity.
Fig. 1. ORTEP view of the complex 1b, drawn at 35% of probability. Hydrogen atoms
were omitted for clarity.
in a meridional geometry, where FeeCl bond distances located at
trans position (FeeCl(3)) to pyridine are shorter than those at cis
position to pyridine (FeeCl(2) and FeeCl(3)) in both complexes. The
corresponding FeeCl bond distances in two complexes are
comparable, suggesting little effect of the N-substitutes. The planes
of the pyridyl rings are essentially coplanar to two benzimidazoyl
rings in the both complexes.
distorted octahedron geometry around the iron(III) center with the
equatorial plane defined by N1 and three chlorines. Complex 1a has
approximate C2v symmetry about the plane, while 1b displays Cs
symmetry. All the bond angles in the equatorial plane are very close
to
a
right angle (N(1)eFeeCl(1), 87.49(4)ꢁ; N(2)eFeeCl(2),
88.26(4)ꢁ; N(3)eFeeCl(1), 89.66(5)ꢁ; N(3)eFeeCl(2), 89.82(5)ꢁ).
The axial angle N(2)eFeeN(3) deviates significantly from linearity
(147.91(6)ꢁ 1a; 146.16(10)ꢁ 1b). The deviations of the metal ion
from the plane formed by the coordinated nitrogen atoms in 1a and
1b are 0.0786, and 0.0631 Å, respectively. The iron center is coor-
dinated to pyridyl nitrogen atom (N1), and sp2 hybridized nitrogen
atoms instead of sp3 nitrogen in the benzimidazole ring. The
FeeN(pyridyl) (FeeN1) bond distances of complexes are about
0.027e0.077 Å longer than the FeeN2 and FeeN3 bond distances,
sharply contrast to those found in the 2,6-bis(imino)pyridyl iron(II),
iron(III) and cobalt complexes [42,43], where the bond of
Complex 3c*DMF (Fig. 2), with the structural features similar to
1a, the coordination geometry around the iron center can also be
described as a distorted octahedron because of the coordination of
one DMF molecule, which is also observed in nickel and iron
complexes, where one methanol is coordinated [65,66]. Similarly,
the iron center is coordinated to nitrogen atom in pyridyl ring and
sp2 hybridized nitrogen atom instead of sp3 nitrogen atom in the
benzimidazole ring, forming two fused-membered rings with
NeFeeN angle of 73.36(10)ꢁ (N1eFe1eN2, N1eFe1eN3), in which
the iron atom lies slightly out of the equatorial plane by 0.0651 Å.
The two benzimidazole planes are both nearly coplanar to the
pyridyl plane with a dihedral angle of about 4.8ꢁ. The benzyl ring is
individually perpendicular to the pyridyl rings as angles of nearly
88.0ꢁ. As expected, the bond lengths of FeeN are significantly
different: the Fe1eN1 bond length (2.182(5) Å) is longer than those
of Fe1eN2 and Fe1eN3 (2.150(3) Å), in which the differences are
virtually identical to those seen in the above discussed iron(III)
analogs. However, there is an unexceptional asymmetry in the two
FeeCl linkages because of the presence of one DMF molecule, with
Fe1eCl(1) being about 0.1052 Å, significantly longer than Fe1eCl(2)
in two terminal Cl in trans-location, which is also observed in the
iron(III) complex.
FeeN(pyridyl)
is
unanimously
shorter
than
that
of
FeeN(iminoaryl), presumably due to the stronger interaction
between iron and benzimidazole. It is notable that different
Nesubstituents in the ligands also have certain influence on the
Fe1eN bond length, though the exact reason is unclear. The FeeN1
distance (2.1688(15) Å) in 1a is shorter than the bond (2.204(2) Å)
in the analogous 1b, while both FeeN2 (2.1385(16) Å) and FeeN3
(2.1419(16) Å) bonds of former complex are longer than those
(2.1271(18) Å) found in the latter analog. Three chlorine ligands are
In the solid state of complexes 3a, X-ray analysis revealed
displays a slightly distorted trigonal bipyramidal configuration,
with the iron atom surrounded by
a
tridentate 2,6-
bis(benzimidazolpyridine) and two terminal chlorides, where the
nitrogen atom (N(2)) of the pyridyl nitrogen and two chlorine
atoms compose an equatorial plane and the central iron atom
slightly deviates by 0.0567 Å from this plane. The three equatorial
angles N(1)eFeeCl(1), N(1)eFeeCl(2) and Cl(1)eFeeCl(2) are
145.46(7)ꢁ, 102.54(6)ꢁ and 111.96(4)ꢁ, respectively, and the axial
FeeN bonds subtend an angle of 144.67(9)ꢁ (N(2)eFeeN(3)). The
dihedral angle between the equatorial plane and the pyridyl plane
is 87.9ꢁ. The two benzimidazole planes are essentially coplanar to
pyridyl ring, and nearly perpendicular to the equatorial plane with
the dihedral angles of 89.5ꢁ and 89.1ꢁ. The iron center is coordi-
nated to nitrogen atom in pyridyl ring (N1) and sp2 hybridized
nitrogen (N2 and N3) atom, too. Iron complexes 3a and 3b shown
roughly identical structures, and differ mainly in the orientation of
the FeCl2 moiety. In complex 3b, the bond distance of FeeN1
Fig. 2. ORTEP view of the complex 3c, drawn at 35% of probability. Hydrogen atoms
were omitted for clarity.

D. Gong et al. / Journal of Organometallic Chemistry 702 (2012) 10e18
13
Table 1
Crystal data and data collection parameters of iron complexes.
1a$2DMF
1b
3a$DMF
3b
3c$DMF
Formula
C25H25N7O2Cl3Fe
617.72
Monoclinic
P21/c
11.6132(4)
11.3896(4)
21.7169(7)
90
103.4360(10)
90
2793.87(16)
4
1.469
0.863
1268
C23H21N5Cl3Fe
529.65
Orthorhombic
Pnma
15.6479(7)
15.9379(7)
9.0570(4)
90.00
90.00
90.00
2258.77(17)
4
1.557
C22H18N6OCl2Fe
509.17
Monoclinic
P21/n
14.5862(6)
7.7262(3)
19.8126(9)
90.00
102.6230(10)
90.00
2178.83(16)
4
1.552
0.966
1040
C23H21N5Cl2Fe
494.20
Triclinic
C36H32N6OCl2Fe
691.43
Monoclinic
Cm
9.0444(7)
16.2526(13)
10.4674(8)
90.00
96.4450(10)
90.00
1528.9(2)
2
1.502
0.709
1040
Molecular weight
Crystal system
Space group
a (Å)
b (Å)
c (Å)
P-1
8.8979(6)
9.6977(7)
13.0379(9)
92.8820(10)
103.8960(10)
95.7280(10)
1083.43(13)
2
a
b
g
(deg)
(deg)
(deg)
V (Å3)
Z
Dcalcd (Mg/m3)
Absorp coeff (mmꢀ1
F(000)
1.515
0.964
508
)
1.045
1084
Crystal size (mm)
0.32 ꢂ 0.21 ꢂ 0.18
1.80 to 26.08
15227
0.33 ꢂ 0.23 ꢂ 0.18
2.56 to 26.09
11981
0.33 ꢂ 0.24 ꢂ 0.16
1.58 to 26.10
11775
0.33 ꢂ 0.21 ꢂ 0.15
1.61 to 26.02
6057
0.23 ꢂ 0.22 ꢂ 0.12
1.96 to 26.15
4235
q
Range (deg)
No. of reflns collected
No. of indep reflns
5530
2328
4322
4119
2408
(Rint ¼ 0.0218)
5530/0/347
1.034
(Rint ¼ 0.0290)
2328/0/155
1.015
(Rint ¼ 0.0273)
4322/0/291
1.098
(Rint ¼ 0.0110)
4119/0/282
1.054
(Rint ¼ 0.0196)
2408/2/216
1.045
No. of data/restraints/params
GOF on F2
R1 (I > 2sigma(I))
wR2
0.0303
0.0865
0.0329
0.1182
0.0424
0.1389
0.0296
0.0785
0.0411
0.0999
(2.1826(15) Å) is significantly longer than those of FeeN2
(2.1460(15) Å) and FeeN3 (2.1513(15) Å), while in the 3a, there is an
exceptional asymmetry, the corresponding distance of FeeN1 is
2.173(2) Å, shorter by about 0.018 Å than FeeN3 (2.191(2) Å), but
longer by 0.004 Aꢁ than FeeN2 bond (2.169(2) Å), which is also
remarkably different from the previous reported tridentate N3
ligand systems [39,43]. The two FeeCl bond lengths in 3b just show
a slight difference between the Fe1eCl2 (2.1513(15) Å) and Fe1eCl1
(2.3184(5) Å), but the two FeeCl bond lengths in 3a are substan-
tially different (2.2880(9) Å, 2.2880(9) Å). The oxidation state of
iron on FeeN and FeeCl bond distance are also observable: the
Fe(III)eN distances (1a) are all shorter than those corresponding
values found in the Fe(II) (3a), probably due to the much stronger
interaction between iron center and nitrogen atoms in 1a, while the
elongated Fe(III)eCl bond distance are possibly resulted from
a more congested environment around the iron center.
The cobalt complexes 5a (Fig. 3), 5b and 5c have an essentially
identical structure, differ slightly in the orientation of the CoCl2
moiety. The coordination geometry around the cobalt center is
a distorted trigonal bipyramid, similar to the five coordinated Fe(II)
analogs, in which the pyridine-nitrogen (N(1)) and two chlorine
atoms form the equatorial plane, with total value of three equato-
rial angles N1eCoeCl1, N1eCoeCl2 and Cl1eCoeCl2 in each
complex are equal to 360ꢁ, while the cobalt atom lies slightly out of
the equatorial plane by 0.0356 Å in 5a, 0.0407 Å in 5b and 0.0343 Å
in 3c. The axial positions are occupied by the other two sp2
nitrogen atoms N(2) and N(3). The N(2)eCoeN(3) moiety deviates
significantly from linearity (5a: N(2)eCo(1)eN(3) 146.00(7)ꢁ; 5b:
141.8(2)ꢁ; 5c: 144.21(7)ꢁ), a distortion as a consequence of satis-
fying the tridentate chelating constraints of the ligand. The equa-
torial planes of these two complexes are nearly perpendicular to
two benzimidazole rings, with average dihedral angles of 88.4ꢁ in
Table 2
Crystal data and data collection parameters of 4a and cobalt complexes.
4a
5a$DMF
5b
5c
6c$DMF
Formula
C11H9N5Cl2Fe
337.98
Monoclinic
P21/c
10.6816(8)
8.0290(6)
15.1952(12)
90
93.4260(10)
90
1300.85(17)
4
1.726
1.561
680
C22H20N6OCl2Co
514.27
Monoclinic
P21n
14.6172(6)
7.7043(3)
19.6714(8)
90.00
102.3420(10)
90.00
2164.10(15)
4
1.578
1.069
1052
C23H21N5Cl2Co
497.28
Triclinic
C33H25N5Cl2Co
621.41
Triclinic
C18H24N6OCl2Co
470.26
Triclinic
Molecular weight
Crystal system
Space group
a (Å)
b (Å)
c (Å)
P-1
P-1
P-1
8.902(5)
9.662(5)
13.109(7)
92.846(9)
103.611(9)
95.032(9)
1088.7(10)
2
8.7617(7)
12.4505(10)
13.6184(11)
71.9870(10)
89.4920(10)
81.5180(10)
1396.26(19)
2
8.197(6)
11.510(9)
11.856(9)
88.381(11)
75.101(12)
77.556(12)
1055.2(13)
2
a
b
g
(deg)
(deg)
(deg)
V (Å3)
Z
Dcalcd (Mg/m3)
Absorp coeff (mmꢀ1
F(000)
1.517
1.055
510
1.478
0.839
614
1.480
1.088
972
)
Crystal size (mm)
0.34 ꢂ 0.24 ꢂ 0.16
1.91 to 26.04
6966
2562 (Rint ¼ 0.0165)
2562/0/172
1.051
0.28 ꢂ 0.18 ꢂ 0.12
1.58 to 26.10
11615
4301(Rint ¼ 0.0258)
4301/0/299
1.105
0.33 ꢂ 0.25 ꢂ 0.23
1.60 to 24.97
6422
3785 (Rint ¼ 0.0761)
3785/0/282
0.948
0.33 ꢂ 0.21 ꢂ 0.18
1.57 to 26.17
7824
5506 (Rint ¼ 0.0186)
5506/0/370
1.004
0.33 ꢂ 0.21 ꢂ 0.15
1.78 to 25.99
5787
3999 (Rint ¼ 0.0634)
3999/0/259
0.967
q
Range (deg)
No. of reflns collected
No. of indep reflns
No. of data/restraint/params
GOF on F2
R1 (I > 2sigma(I))
wR2
0.0243
0.0671
0.0295
0.0834
0.0882
0.2354
0.0344
0.0837
0.0890
0.2422

14
D. Gong et al. / Journal of Organometallic Chemistry 702 (2012) 10e18
Table 3
Selected bond distances (Å) and angles (ꢁ) of iron complexes 1a, 1b, 4a, 3a, 3b and 3c.
1a$2DMF
1b
2.204(2)
3a$DMF
3b
3c$DMF
4a
2.1539(14)
2.1833(14)
2.1860(15)
2.3269(5)
2.3048(5)
FeeN1
FeeN2
FeeN3
FeeCl1
FeeCl2
FeeCl3
2.1688(15)
2.1385(16)
2.1419(16)
2.3311(6)
2.3838(5)
2.2686(5)
73.93(6)
73.98(6)
147.91(6)
87.49(4)
89.66(5)
88.22(5)
177.31(5)
104.39(5)
107.70(5)
94.62(2)
89.82(5)
88.26(4)
85.20(4)
172.53(2)
92.73(2)
2.173(2)
2.169(2)
2.191(2)
2.2880(9)
2.3653(9)
e
2.1826(15)
2.1460(15)
2.1513(15)
2.3184(5)
2.3566(5)
e
2.182(5)
2.150(3)
2.150(3)
2.4871(15)
2.3819(16)
2.283(4) (FeeO1)
73.36(10)
73.36(10)
145.90(19)
85.89(13)
92.42(9)
92.42(9)
87.78(17) (N1eFeeO1)
85.76(9) (N3eFeeO1)
85.76(9) (N2eFeeO1)
173.67(14) (Cl1eFeeO1)
106.61(10)
2.1271(18)
2.1271(18)
2.3874(9)
2.3319(9)
2.2500(10)
73.08(5)
73.08(5)
146.16(10)
85.29(7)
88.86(5)
88.86(5)
178.69(7)
106.92(5)
106.92(5)
93.39(4)
e
N1eFeeN3
N1eFeeN2
N3eFeeN2
N1eFeeCl1
N3eFeeCl1
N2eFeeCl1
N1eFeeCl3
N3eFeeCl3
N2eFeeCl3
Cl1eFeeCl3
N3eFeeCl2
N2eFeeCl2
N1eFeeCl2
Cl1eFeeCl2
Cl3eFeeCl2
73.14(9)
73.73(9)
144.67(9)
145.46(7)
98.31(6)
101.98(7)
e
73.40(6)
73.52(6)
141.56(6)
158.00(4)
102.03(4)
101.27(4)
e
72.66(5)
72.22(5)
142.87(6)
139.08(4)
105.04(4)
93.75(4)
e
e
e
e
e
e
e
e
e
e
89.31(5)
89.31(5)
88.44(7)
173.73(3)
92.87(4)
100.24(6)
98.57(7)
102.54(6)
111.96(4)
e
103.68(4)
99.41(4)
97.29(4)
104.68(2)
e
98.41(4)
106.61(10)
179.50(13)
94.60(6)
91.73(13)
104.36(4)
109.69(4)
111.015(19)
e
(O1eFeeCl2)
e
5a, 83.7ꢁ in 5b and 83.7ꢁ in 5c. The two benzimidazole rings lie
almost coplanar to the pyridyl plane. The benzyl moiety in 3c is
almost orthogonal to the other rings, as observed in the iron(II) 2c.
The effect of substituents at sp3 nitrogen (N3 and N4) is also re-
flected in the CoeN and CoeCl bond lengths not in each complex
but also among three complexes. Following the same trend
observed in iron(II) complexes, CoeN1 distances in N-substituted
5b and 5c are slightly longer that CoeN1 and CoeN2, while in 3a,
that of CoeN1 is intermediate between two CoeN2 and CoeN3. The
corresponding CoeN distances among the three complexes are also
different and follow the trends, CoeN(1): 5b (2.170(6) Å) > 5a
(2.1358(18) Å) > 3c (2.1345(17) Å); CoeN(2): 5a (2.1360(17) Å) > 5b
(2.118(6) Å) > 5c (2.1060(19) Å); CoeN(2): 5a (2.1427(17) Å) > 5b
(2.122(6) Å) > 5c (2.1069(18) Å), while average CoeCl distance: 5b
(2.3245 Å) > 5a (2.3055 Å) > 5c (2.3031 Å) and Cl1eCoeCl2
changes in this way: 5a (111.56(3)ꢁ) > 5c (107.64(2)ꢁ) > 5b
(104.50(8)ꢁ). Clearly the introduction of electron withdrawing
substituents in the ortho-positions of the phenyl rings has
a pronounced effect on the structure, and this is also likely to affect
the catalytic activities. As compared with those seen in iron(II)
analogs (3a, 3b), CoeN and CoeCl (5a, 5b) are all shorted by
0.01e0.05 Å, most probably due to the reduced Co2þ radius, while
no significant variation regarding either NeMeN, NeMteCl or
CleMteCl is found.
2.2. Butadiene polymerization behaviors of complexes activated by
MMAO or aluminum chlorides
Recently, we have been engaged in the stereospecific polymer-
izations of 1,3-conjugated dienes [18,20,39,42,43], which is a very
important process in the chemical industry to afford products that
are among the most significant and widely used rubbers. Therefore
these complexes have been tested as precatalysts for butadiene
polymerization. The generated homogeneous catalysts showed
versatile catalytic performances and the selected polymerization
data were summarized in Table 5. As compared with catalytic
performance of the naked cobalt chloride, the attached ligands has
a significant role on promoting the catalytic activity as well as
improving the selectivity as evidenced from the remarkable
improved catalytic properties of the activated corresponding
complexes in butadiene polymerization.
Iron(III) complex 1a, in combination with MMAO exhibits high
activity and high selectivity, reaching monomer conversion of 91%
within 4 h at room temperature, affording polybutadiene with
80.2% of cis-1,4 and 11.7% of trans-1,4 regularity, however, switching
to activator AlEt2Cl, the formed system shows completely no
activity toward butadiene. It is evident to find the ligand effect also
plays a significant role in control of both activity and selectivity,
incorporation of either ethyl (1b) or benzyl group (1c), instead of
the hydrogen (1a) on the nitrogen atom of the benzimidazole
ligand leads to a sharply decrease in both selectivity and monomer
conversion. Applying catalyst 1b, only a small amount of polymer
(polymer yield, 15%) is obtained with cis-1,4 moderate selectivity of
55.4%. With N-benzyl substituted catalyst 1c, comparatively, only
trace polymer is obtained. Evidently, steric argument from N-
substituent could not rationalize the difference, as they are too far
to sterically affect the metal center. Tentatively, these results
suggest that NeH functionality, which could be deprotonated to
give anionic amide ligands when activated by MMAO, is assumed to
be essential for high activity, possibly selectivity as well with this
ligand system. The anionic amide ligands could be free or form
NeAl species (anionecation pair) which may act to increase cata-
lytic activity. This observation is also consistent with the previous
reports [52], although the mechanism is not clear yet. These
hydrido tridentate ligands thus offer a mix of donor characters that
provides the metal unique environment throughout the catalytic
cycle. These features could also offer an explanation to the rare cis-
Table 4
Selected bond distances (Å) and angles (ꢁ) of cobalt complexes 5a, 5b, 5c and 6c.
5a$DMF
5b
2.170(6)
2.118(6)
2.122(6)
2.309(2)
2.338(2)
73.7(2)
74.7(2)
141.8(2)
160.27(17)
101.90(18)
100.66(17)
103.77(17)
99.82(17)
95.21(17)
104.50(8)
5c
6c$DMF
CoeN1
CoeN2
CoeN3
CoeCl1
2.1345(17)
2.1360(17)
2.1427(17)
2.2825(6)
2.3284(6)
74.31(6)
2.1358(18)
2.1060(19)
2.1069(18)
2.2993(6)
2.3069(6)
74.79(7)
2.069(5)
2.162(5)
2.158(5)
2.282(2)
2.287(2)
74.41(18)
75.00(18)
149.40(18)
124.26(17)
100.69(16)
95.99(15)
96.52(16)
99.65(17)
121.26(16)
114.48(7)
CoeCl2
N1eCoeN3
N1eCoeN2
N3eCoeN2
N1eCoeC11
N3eCoeCl1
N2eCoeCl1
N3eCoeCl2
N2eCoeCl2
N1eCoeCl2
Cl1eCoeCl2
74.62(6)
73.95(7)
146.00(7)
147.02(5)
98.38(5)
100.13(5)
101.02(5)
98.33(5)
144.21(7)
155.47(5)
100.69(5)
100.48(5)
103.49(5)
97.12(5)
101.41(5)
111.56(3)
96.81(5)
107.64(2)

D. Gong et al. / Journal of Organometallic Chemistry 702 (2012) 10e18
15
Table 5
Polymerization of 1,3-butadiene with 2,6-bis(2-benzimidazolyl)pyridine or 2,6-bis(pyrazolyl)pyridine iron and cobalt complexes.
Entry
Complex
Al
Al/Fe (Co)
Yield (%)
Mn
Mw/Mn
Microstructure (%)
Cisꢀ1,4
Transꢀ1,4
1,2
1
2
3
4
5
6
7
8
1a
1a
1b
1c
2a
2b
2c
3a
3a
3b
3c
4a
4b
4c
5a
5a
5a
5b
5c
6b
6c
MMAO
AlEt2Cl
MMAO
MMAO
MMAO
MMAO
MMAO
MMAO
AlEt2Cl
MMAO
MMAO
MMAO
MMAO
MMAO
MMAO
AlEt2Cl
Al2Et3Cl3
Al2Et3Cl3
Al2Et3Cl3
Al2Et3Cl3
Al2Et3Cl3
200
50
91
0
153,000
e
2.7
e
80.2
e
11.7
e
8.1
e
200
200
200
200
200
200
50
200
200
200
200
200
200
50
15
trace
74
37
26
95
0
11
trace
78
38
31
32
100
100
88
85
82
91
94,000
e
3.3
e
55.4
e
34.3
e
10.3
e
55,000
115,000
141,000
142,000
e
1.9
2.2
2.1
2.5
e
1.5
91.2
44.5
46.5
12.1
e
7.3
30.1
32.1
6.7
e
25.4
21.4
81.2
e
9
10
11
12
13
14
15
16
17
18
19
20
21
88,000
e
3.1
e
57.2
e
30.9
e
11.9
e
42,000
101,000
144,000
69,000
102,000
114,000
89,000
76,000
114,000
125,000
1.6
2.5
2.0
4.0
2.3
2.6
2.9
3.5
2.2
2.1
1.5
90.4
44.5
47.0
0.6
0.5
0.3
0.1
0.3
1.0
0.7
8.1
28.4
33.7
5.8
3.6
2.5
2.1
3.3
2.4
2.5
27.1
19.3
93.6
95.9
97.2
97.8
96.4
96.6
96.8
60
60
60
60
60
Polymerization conditions: precursor, 10 m
mol; butadiene, 10 mmol; toluene, 5 mL; time, 4 h; temperature, 20 ꢁC.
1,4-selectivity, in striking contrast to the documented FeN3 systems
as well as the pyrazolyl iron complexes discussed herein, which
normally give high trans-1,4-selectivity [13,43].
coordination as well as the subsequent insertion. Thus, modifying
catalyst for better catalytic performance is difficult in this system,
but it may drop a clue on how to seek for suitable ligand to obtain
trans-1,4 selective iron catalyst: neutral tridentate N3 ligand, with
appropriate space for monomer access and insertion [13,43].
The corresponding iron(II) complexes 3ae3c and 4ae4c shows
very similar catalytic performance with 1ae1c and 2ae2c,
respectively, in terms of monomer conversion, polymer properties
under the same polymerization conditions, attributable to the same
cationic iron species generated from the two type of complexes,
though differ in the original oxidation states.
Besides the types of co-activators and ligands, the central metal
also exerts an obvious influence on the catalytic performances, and
controlling of selectivity is attainable by changing the metal center.
Cobalt complexes generally show higher activity and cis-1,4-
selectivity compared with iron analogs. Complex 5a, in combina-
tion with MMAO, exhibits moderate activity towards butadiene
polymerization, providing polybutadiene with cis-1,4 incorporation
about 94%, and Mn calculated to be about 80,000 with broad
molecular weight distribution. The monomer conversion appears
to be sensitive to the cocatalyst, while the selectivity difference is
less pronounced. Alkylaluminum chlorides are found to be the
better cocatalysts, for example, with diethylalumium chloride,100%
monomer conversion is achieved from complexes 3a, and cis-1,4-
selectivity is improved to be 96% as well, and ethyl aluminum
sesquichloride initiated catalyst also shows the comparable excel-
lent catalytic performance. For the N-alkylated (5b, N-ethyl) and
(5c, N-Benzyl), high cis-1,4 enchainment in the resultant poly-
butadiene is maintained, albeit with marginally lower monomer
conversion, 88% and 85% for 5b and 5c, respectively. The unam-
biguous ligand effect summarized in the cobalt complexes is similar
to the trend observed in iron analogs, but less pronounced.
We extend the research with the 2,6-bis(prazolyl)pyridine into
cobalt complexes. Treatment of either 6b or 6c with Al2Et3Cl3 leads
to active butadiene polymerization catalyst converting butadiene
to cis-1,4 polybutadiene with high monomer conversion. Neither
the activity nor the selectivity appears to be sensitive to the ligand
environment.
It is quite interesting to note the pyrazolyl iron complexes
display distinctive trans-1,4-selectivity compared to the benzimi-
dazolyl analogs, but consistent with the results observed from the
reported symmetry N3 ligands like bisiminopyridine and terpyr-
idine [13,43]. Here, metal environments of solvent free iron(II) 3b
and 4a are compared. The neutral bulky benimidazoly moiety may
hinder monomer delivery and coordination, being to blame for
poor selectivity as well as activity as already observed in bulky iron
bisiminopyridine systems [13,43]. Complex 4a is also demonstrated
to have more free space near active center than 3b, as evident from
the shorter FeeN1 distance as well as longer FeeN2 and FeeN3
bonds, and this unique structure may facilitate trans-coordination
and insertion taking place synchronously, which leads to trans-1,4
unit, and such synchronicity excludes any significant errors which
possibly resulted from isomerization of syn- to anti-conformation
in case the insertion rate is slower than coordination rate [4].
Fe(III)-based precatalysts 2ae2c reveal that an increase of steric
bulk near metal center results in decreases in monomer conversion
and increases in Mn. Replacement of hydrogen by the methyl group
on both 3- and 5- position of the pyrazole rings (precatalyst 2c,
Table 5, entry 7) results in a decrease in monomer conversion from
74% to 26% and an increase in Mn from 55,000 to 141,000. The
replacement of 3-position on pyrazole ring with a methyl group
(precatalyst 2b, Table 5, entry 6) yields intermediate values of
conversion and Mn. This could be explained that in this catalytic
system with MMAO, the reaction speed mostly depended on the
insertion of butadiene at the active species, so a more bulky group
caused slower insertion reaction and consequently lower monomer
conversion which might be attributed that the catalysts with less
steric bulkiness facilitate the reaction of propagation not the b-
hydrogen elimination reaction. The steric bulk exerts obvious
influence on the selectivity as well. The regularity of polybutadiene
derived from 2a is mainly trans-1,4 enchained, which is in agree-
ment with results observed for iron catalytic [13,43]. The methyl
derivatives 2c is less trans-1,4-selectivity, and 2b gives a lower
trans-1,4-selectivity than either the 2a or 2c. Clearly, steric open of
the active site is also a crucial factor in controlling butadiene
The selectivity of cobalt system is mainly dependent on the
coordination atoms on the ligands. Cobalt complexes supported by

16
D. Gong et al. / Journal of Organometallic Chemistry 702 (2012) 10e18
N, N, N tridentate ligands [12,39,42,62,63], O, Ne or O, Oebased
anionic ligands [25,26,29] are likely to give cis-1,4 structure. And,
placed at the calculated positions and were included in the struc-
ture calculation without further refinement of the parameters.
such
p-donating ligand maybe favor cis-1,4 coordination of buta-
diene with cobalt center followed by quick insertion to terminal
carbon of cobalt allylic polymer chain [24]. While for those sup-
4.3. Syntheses and characterization of ligands and complexes
ported by s-donating phosphine ligands, the mainly 1,2-selectivity
[19,24,28] is generated when the coordinated monomer preferen-
tially attacks the third carbon to the last.
The N, N, N-tridentate ligand reported here as well as previously
published [12,39,42,62,63] may also serve to design appropriate
new ligand framework for cis- and/or trans-1,4 selective polymer-
ization of butadiene.
2,6-Bis(benzimidazol-2-yl)-pyridine
(L1),
2,6-Bis(10-ethyl-
benzimidazol-20-yl)pyridine (L2) and 2,6-Bis(10-benzylbenzimida-
zol-20-yl)pyridine (L3) were prepared in according to the reported
methods.
4.3.1. 2,6-Bis(benzimidazol-2-yl)-pyridine iron(III) chloride (1a)
2,6-Bis(benzimidazol-2-yl)-pyridine (0.31 g, 1.0 mmol) and
anhydrous FeCl3 (0.17 g, 1.0 mmol) were added to a flask containing
THF (15 mL). The mixture was stirred at 20 ꢁC for 12 h, during which
time, a deep-red suspension was formed. Et2O was added and
a suspension formed. The precipitate was collected by filtration and
washed with 4 ꢂ 5 mL of heptane. The desired product (0.40 g,
83.3%) was obtained after dried in vacuo at 40 ꢁC.
3. Conclusion
Structurally well-defined iron and cobalt homogeneous catalyst
systems were developed and applied in polymerization of 1,3-
butadiene, polybutadiene with a broad range of controllable
microstructure, which could be realized by varying a combination
of the ligand, central metal and cocatalyst, is accessible.
IR (KBr, cmꢀ1): 3051 (vNeH), 1606 (vC]C), 1591, 1577 (vC]N), 1465
(
dNeH), 1321, 1147, 1022, 997, 764, 756 (dNeH). 1H NMR (400 MHz,
4. Experimental
CDCl3): d (ppm) 8.35 (2H, Hpyr), 8.19 (1H, Hpyr), 7.76 (4H, Hph), 7.32
(4H, Hph). Anal. Calcd. For C19H13N5FeCl3: C, 48.19; H, 2.77; N, 14.79.
Found: C, 49.41; H, 3.75; N, 12.47. MALDIeTof for C19H13N5Fe56Cl335
(relative ratio): (m/z) 437 (mþ, ꢀCl), 402 (mþ, ꢀ2Cl).
4.1. General methods
All manipulation related to air and/or moisture sensitive
compounds were carried out under a dry and oxygen-free argon
atmosphere by using Schlenk techniques or under a nitrogen
atmosphere in a glovebox. FeCl3, FeCl2$4H2O, CoCl2 as well as
pyridine-2,6-dicarboxylic acid, o-phenylenediamine, benzyl chlo-
ride and ethyl bromide were all purchased from Alfa Aesar. Meth-
ylaluminoxane (MMAO) was commercially available from
AkzoNoble. Toluene and tetrahydrofuran (THF) were freshly
distilled in the presence of sodium and benzophenone. Polymeri-
zation grade butadiene was supplied from Jinzhou Petrochemical
Corporation and purified by passing through four columns packed
with 4 Å molecular sieves and KOH. Ligands (L1eL6) were prepared
according to the methods [52,64]. Other chemicals were used as
received unless otherwise noted. 1H NMR (400 MHz) and 13C NMR
(100 MHz) were recorded on a Varian Unity spectrometer in CDCl3
or DMSO-d6 at room temperature. IR spectra were performed on
BRUKE Vertex-70 FIR spectrophotometer. Elemental analyses were
recorded on an elemental Vario EL spectrometer. The proportion of
1,2, cis-1,4 and trans-1,4 units of polymer were determined by IR
spectra, 1H NMR and 13C NMR [67e69]. The molecular weights (Mn)
and molecular weight distributions (Mw/Mn) of polymer were
measured at 30 ꢁC by gel permeation chromatography (GPC)
equipped with a Waters 515 HPLC pump, four columns (HMW 7
THF, HMW 6E THF ꢂ 2, HMW 2 THF) and a Waters 2414 refractive
index detector. THF was used as eluent at a flow rate of 1.0 mL/min.
The molecular weight of polymer was determined using the poly-
styrene calibration.
4.3.2. 2,6-bis(10-ethylbenzimidazol-20-yl)pyridine iron(III)Cl3 (1b)
The procedure as above in (4.3.1) using L2 and FeCl3 gave 1b as
a red powder in 85% yield.
IR (KBr, cmꢀ1): 2974, 2931, 1599 (vC]C), 1573 (vC]N), 1485, 1460,
1441, 1337, 1132, 770, 756. 1H NMR (400 MHz, CDCl3):
d (ppm) 8.37
(2H, Hpyr), 8.25 (1H, Hpyr), 7.76 (4H, Hph), 7.34 (4H, Hph), 1.32 (6H,
eCH3). Anal. Calcd. for C23H21N5FeCl3: C, 52.16; H, 3.40; N, 13.22.
Found: C, 52.24; H, 3.41; N, 12.70. MALDIeTof for C23H21N5Fe56Cl335
(relative ratio): (m/z) 493 (mþ, ꢀCl), 458 (mþ, ꢀ2Cl).
4.3.3. 2,6-bis(10-benzylbenzimidazol-20-yl)pyridine iron(III)Cl3 (1c)
The procedure as above in (4.3.1) using L3 and FeCl3 gave 1c as
a red powder in 85% yield.
IR (KBr, cmꢀ1): 1606 (vC]C), 1575 (vC]N), 1496, 1466, 1320, 763,
747, 668. 1H NMR (400 MHz, CDCl3):
d (ppm) 8.36 (2H, Hpyr), 8.21
(1H, Hpyr), 7.78 (2H, Hph), 7.49 (2H, Hph), 7.28 (4H, Hph), 7.16 (8H,
Hph), 6.81 (2H, Hph), 5.77 (4H, eCH2Ph). Anal. Calcd. for
C33H27N5FeCl3: C, 60.44; H, 4.15; N, 10.68. Found: C, 59.95; H, 4.08;
N, 9.95. MALDIeTof for C33H27N5Fe56Cl335 (relative ratio): (m/z) 617
(mþ, ꢀCl), 582 (mþ, ꢀ2Cl).
4.3.4. 2,6-bis(pyrazoyl)pyridine iron(III)Cl2 (2a)
The procedure as above in (4.3.1) using L4 and FeCl3 gave 2a as
a red powder in 86.1% yield.
IR (KBr, cmꢀ1): 1618, 1587, 1528, 1495, 1479, 1401, 1345, 1314,
1175, 1064, 977, 800, 779. 1H NMR (400 MHz, CDCl3):
d (ppm) 8.95
(2H, Hpyr), 8.14 (t,1H, Hpyr), 7.87 (2H, Hpyz), 7.82 (2H, Hpyz), 6.63 (2H,
Hpyz). Anal. Calcd. for C11H9N5Cl3Fe: C, 35.38; H, 2.43; N, 18.75.
Found: C, 37.08; H, 2.30; N, 17.58. MALDIeTof for C11H9N5Fe56Cl335
(relative ratio): (m/z) 337 (mþ), 302 (mþ, ꢀCl).
4.2. X-ray structure determinations
Crystals for X-ray analysis were obtained as described in the
experimental section. Data collections were performed at ꢀ88.5 ꢁC
on a Bruker SMART APEX diffractometer with a CCD area detector,
4.3.5. 2,6-bis(3-methylpyrazoyl)pyridine iron(III)Cl2 (2b)
The procedure as above in (4.3.1) using L5 and FeCl3 gave 2b as
a red powder in 92.4% yield.
using graphite monochromated Mo K radiation (
l
¼ 0.71073 Å). The
determination of crystal class and unit cell parameters was carried
out by the SMART program package. The raw frame data were
processed using SAINT and SADABS to yield the reflection data file.
The structures were solved by using SHELXTL program. Refinement
was performed on F2 anisotropically for all non-hydrogen atoms by
the full-matrix least-squares method. The hydrogen atoms were
IR (KBr, cmꢀ1): 3095, 3051, 1614, 1587, 1534, 1485, 1369, 1320,
1237, 1177, 1062, 990, 810, 788. 1H NMR (400 MHz, CDCl3):
d (ppm)
8.78 (2H, Hpyr), 8.05 (1H, Hpyr), 7.68 (2H, Hpyz), 6.42 (2H, Hpyz), 2.31
(6H, eCH3). Anal. Calcd. for C13H13N5FeCl3: C, 38.89; H, 3.26; N,
17.44. Found: C, 40.09; H, 3.70; N, 15.63. MALDIeTof for

D. Gong et al. / Journal of Organometallic Chemistry 702 (2012) 10e18
17
C13H13N5Fe56Cl335 (relative ratio): (m/z) 365 (mþ, ꢀCl), 330
(mþ, ꢀ2Cl).
4.3.12. 2,6-bis(3,5-dimethylpyrazoyl)pyridine iron(II)Cl2 (4c)
The procedure as above in (4.3.1) using L6 and FeCl2 gave 4c as
a yellow powder in 81.7% yield.
IR (KBr, cmꢀ1): 3100, 2925, 1613, 1562, 1479, 1393, 1363, 1309,
4.3.6. 2,6-bis(3,5-dimethylpyrazoyl)pyridine iron(III)Cl2 (2c)
The procedure as above in (4.3.1) using L6 and FeCl3 gave 2c as
a red powder in 91.7% yield.
1182, 1136, 1047, 985, 846, 788, 739. 1H NMR (400 MHz, CDCl3):
d
(ppm) 8.05 (1H, Hpyr), 7.67 (2H, Hpyr), 6.14 (2H, Hpyz), 2.21 (12H,
IR (KBr, cmꢀ1): 3108, 3001, 1613, 1591, 1561, 1483, 1472, 1419,
eCH3). Anal. Calcd. for C15H17N5FeCl2: C, 45.72; H, 4.35; N, 17.77.
Found: C, 44.89; H, 4.12; N, 18.41. MALDIeTof for C15H17N5Fe56Cl235
(relative ratio): (m/z) 394 (mþ, ꢀCl), 358 (mþ, ꢀ2Cl).
1393, 1362, 1315, 1302, 1182, 1138, 1051, 986, 813, 791, 741. 1H NMR
(400 MHz, CDCl3): d (ppm) 8.06 (1H, Hpyr), 7.67 (2H, Hpyr), 6.14 (2H,
Hpyz), 2.21 (12H, eCH3). Anal. Calcd. for C15H17Cl3N5Fe: C, 41.94; H,
3.99; N, 16.30. Found: C, 39.89; H, 3.72; N, 17.63. MALDIeTof for
C15H17N5Fe56Cl335 (relative ratio): (m/z) 393 (mþ, ꢀCl), 358
(mþ, ꢀ2Cl).
4.3.13. 2,6-bis(benzimidazol-2-yl)pyridine cobalt(II)Cl2 (5a)
The procedure as above in (4.3.1) using L1 and CoCl2 gave 5a as
a green powder in 85% yield.
IR (KBr, cmꢀ1): 3061(vNeH), 1607 (vC]N), 1691(vC]C), 1575, 1494,
1459 (dNeH), 1317, 1234, 1147, 996, 819, 744 (dNeH). Anal. Calcd. for
C19H13N5CoCl2: C, 51.73; H, 2.97; N, 15.87. Found: C, 51.24; H, 3.35;
N, 14.10. MALDIeTof for C19H13N5Co58Cl235 (relative ratio): (m/z) 405
(mþ, ꢀCl), 369 (mþ, ꢀ2Cl).
4.3.7. 2,6-bis(benzimidazol-2-yl)pyridine iron(II)Cl2 (3a)
The procedure as above in (4.3.1) using L1 and FeCl2 gave 3a as
a dark blue powder in 85% yield.
IR (KBr, cmꢀ1): 3195, 3046 (vNeH),1608 (vC]C),1577 (vC]N),1495
(
dNeH), 1493, 1443, 1317, 1237, 1147, 994, 820, 748 (dNeH). 1H NMR
4.3.14. 2,6-bis(10-ethylbenzimidazol-20-yl)pyridine
cobalt(II)Cl2 (5b)
The procedure as above in (4.3.1) using L2 and CoCl2 gave 5b as
a green powder in 85% yield.
(400 MHz, CDCl3): d (ppm) 8.35 (2H, Hpyr), 8.19 (1H, Hpyr), 7.76 (4H,
Hph), 7.32 (4H, Hph). Anal. Calcd. for C19H13N5FeCl2: C, 52.09; H,
2.99; N, 15.99. Found: C, 52.16; H, 2.98; N, 15.95. MALDIeTof for
C19H13N5Fe56Cl335 (relative ratio): (m/z) 437 (mþ), 402 (mþ, -Cl).
IR (KBr, cmꢀ1): 3020, 2987, 2935, 1600 (vC]C), 1573 (vC]N), 1512,
1485, 1460, 1443, 1349, 1335, 1154, 1094, 812, 762, 749. Anal. Calcd.
for C23H21N5CoCl2: C, 55.55; H, 4.26; N, 14.08. Found: C, 54.48; H,
4.31; N, 12.94. MALDIeTof for C23H21N5Fe56Cl235 (relative ratio):
(m/z) 461 (mþ, ꢀCl).
4.3.8. 2,6-bis(10-ethylbenzimidazol-20-yl)pyridine iron(II)Cl2 (3b)
The procedure as above in (4.3.1) using L2 and FeCl2 gave 3b as
a dark blue powder in 85% yield.
IR (KBr, cmꢀ1): 2933, 2872, 1600 (vC]C), 1570 (vC]N), 1507, 1460
(
dNeH), 1484, 1460, 1441, 1417, 1349, 1335, 1135, 1095, 873, 810, 761,
748. Anal. Calcd. for C23H21N5FeCl2: C, 55.90; H, 4.28; N, 14.17. 1H
NMR (400 MHz, CDCl3): (ppm) 8.37 (2H, Hpyr), 8.25 (1H, Hpyr), 7.76
4.3.15. 2,6-bis(10-benzylbenzimidazol-20-yl)pyridine
cobalt(II)Cl2 (5c)
The procedure as above in (4.3.1) using L3 and CoCl2 gave 5c as
a green powder in 85% yield.
d
(4H, Hph), 7.34 (4H, Hph), 1.32 (6H, eCH3). Found: C, 55.21; H, 4.29;
N,14.44. MALDIeTof for C23H21N5Fe56Cl235 (relative ratio): (m/z) 493
(mþ), 458 (mþ, ꢀCl).
IR (KBr, cmꢀ1): 1596 (vC]C), 1568 (vC]N), 1498, 1471, 1436, 1335,
815, 781, 749, 731, 695. Anal. Calcd. for C33H27N5CoCl2: C, 63.58; H,
4.37; N, 11.23. Found: C, 63.98; H, 4.19; N, 10.82. MALDIeTof for
C33H27N5Co58Cl325 (relative ratio): (m/z) 585 (mþ, ꢀCl), 549
(mþ, ꢀ2Cl).
4.3.9. 2,6-bis(10-benzylbenzimidazol-20-yl)pyridine iron(II)Cl2 (3c)
The procedure as above in (4.3.1) using L3 and FeCl2 gave 3c as
a dark blue powder in 85% yield.
IR (KBr, cmꢀ1): 3024 (vNeH), 1603 (vC]C), 1449, 1470, 1442, 1157,
4.3.16. 2,6-bis(3-methylpyrazoyl)pyridine cobalt(II)Cl2 (6b)
The procedure as above in (4.3.1) using L5 and CoCl2 gave 6b as
a blue powder in 95.1% yield.
861, 734, 696. 1H NMR (400 MHz, CDCl3):
d (ppm) 8.36 (2H, Hpyr),
8.21 (1H, Hpyr), 7.78 (2H, Hph), 7.49 (2H, Hph), 7.28 (4H, Hph), 7.16
(8H, Hph), 6.81 (2H, Hph), 5.77 (4H, eCH2Ph). Anal. Calcd. for
C33H27N5FeCl2: C, 63.89; H, 4.39; N, 11.29. Found: C, 63.63; H, 4.48;
N, 11.59. MALDIeTof for C33H27N5Fe56Cl235 (relative ratio): (m/z) 617
(mþ), 582 (mþ, ꢀCl).
IR (KBr, cmꢀ1): 3103, 3074, 1616, 1582, 1542, 1484, 1385, 1316,
1241, 1175, 1061, 983, 795, 778, 790. Anal. Calcd. for C13H13N5CoCl2:
C, 42.30; H, 3.55; N, 18.97. Found: C, 42.00; H, 3.81; N, 19.31.
MALDIeTof for C13H13N5Co58Cl325 (relative ratio): (m/z) 333
(mþ, ꢀCl).
4.3.10. 2,6-bis(pyrazoyl)pyridine iron(II)Cl2 (4a)
The procedure as above in (4.3.1) using L4 and FeCl2 gave 4a as
a yellow powder in 76.9% yield.
4.3.17. 2,6-bis(3,5-dimethylpyrazoyl)pyridine cobalt(II)Cl2 (6c)
The procedure as above in (4.3.1) using L6 and CoCl2 gave 6c as
a blue powder in 91.1% yield.
IR (KBr, cmꢀ1): 1619, 1587, 1526, 1478, 1400, 1344, 1311, 1179,
1077, 1059, 968, 782. Anal. Calcd. for C11H9N5FeCl2: C, 39.09; H,
2.68; N, 20.72. Found: C, 38.48; H, 2.52, N, 21.63.; MALDIeTof for
C11H9N5Fe56Cl235 (relative ratio): (m/z) 337 (mþ), 302 (mþ, ꢀCl).
IR (KBr, cmꢀ1): 3100, 2971, 1613, 1590, 1563, 1480, 1393, 1363,
1309, 1181, 1136, 1048, 983, 842, 788, 739. Anal. Calcd. for
C15H17N5CoCl2: C, 45.36; H, 4.31; N, 17.63. Found: C, 46.18; H, 4.12;
N,18.04. MALDIeTof for C15H17N5Co58Cl235 (relative ratio): (m/z) 361
(mþ, ꢀCl).
4.3.11. 2,6-bis(3-methylpyrazoyl)pyridine iron(II)Cl2 (4b)
The procedure as above in (4.3.1) using L5 and FeCl2 gave 4b as
a yellow powder in 92.4% yield.
4.4. Procedure for butadiene polymerization
IR (KBr, cmꢀ1): 3104, 1615, 1582, 1540, 1483, 1382, 1371, 1316,
1241, 1175, 1062, 984, 795, 776. 1H NMR (400 MHz, CDCl3):
d
(ppm)
A typical procedure for the polymerization is as follows (entry 1
in Table 5): a toluene solution of butadiene (5 mL, 2.0 mol/L) was
added to a moisture free ampere bottle preloaded with complex 1a
(5.3 mg, 0.01 mmol), then MMAO (2.0 mmol) was injected to
initiate the polymerization at 20 ꢁC. After 4 h, methanol was added
to the system to quench the polymerization. The mixture was
8.78 (2H, Hpyr), 8.04 (1H, Hpyr), 7.68 (2H, Hpyz), 6.41 (2H, Hpyz), 2.30
(6H, eCH3). Anal. Calcd. for C13H13N5FeCl2: C, 42.66; H, 3.58; N,
19.13. Found: C, 43.46; H, 3.23; N, 19.73. MALDIeTof for
C13H13N5Fe56Cl325 (relative ratio): (m/z) 365 (mþ, ꢀCl), 330
(mþ,ꢀ2Cl).

18
D. Gong et al. / Journal of Organometallic Chemistry 702 (2012) 10e18
[28] G. Ricci, A. Forni, A. Boglia, T. Motta, J. Mol. Catal. A: Chem. 226 (2005)
235e241.
[29] K. Endo, T. Kitagawa, K. Nakatani, J. Polym. Sci. Part A: Polym. Chem. 44 (2006)
4088e4094.
[30] M. Gippn, Eng. Chem. Prod. Res. Dev. 1 (1962) 32e39.
[31] C. Longgivae, R. Castelli, G.F. Groce, Chim. Ind. 43 (1961) 625e630.
[32] H.E. Swift, J.E. Bozik, C.-Y. Wu, J. Catal. 17 (1970) 331e340.
[33] Z.-Y. Zhang, H.-J. Zhang, H.-M. Ma, Y. Wu, J. Mol. Catal. 17 (1982) 65e76.
[34] C. Bazzini, A. Giarrusso, L. Porri, Macromol. Rapid Commun. 23 (2002)
922e927.
poured into a large quantity of methanol containing 2,6-di-tert-
butyl-4-methylphenol (1.0 wt%) as a stabilizer. Filtered and dried
under vacuum at 40 ꢁC, polybutadiene was resulted at a constant
weight (0.49 g, 91%).
Acknowledgment
Authors appreciate financial supports from National Science and
Technology Infrastructure Program (2007BAE14B01-06) and The
Fund for Creative Research Groups (50621302).
[35] G. Ricci, D. Morganti, A. Sommazzi, R. Santi, F. Masi, J. Mol. Catal. A: Chem.
(2003) 204e205 287e293.
[36] S. Luo, Iron-based catalyst composition and process for producing syndiotactic
1,2-polybutadiene, US Patent, 2001, 6284,702.
[37] J. Lu, Y.-M. Hu, X.-Q. Zhang, J.-F. Bi, W.-M. Dong, L.-S. Jiang, B.-T. Huang, J. Appl.
Polym. Sci. 100 (2006) 4265e4269.
Appendix A. Supplementary material
[38] Y.-M. Hu, W.-M. Dong, L.-S. Jiang, X.-Q. Zhang, Y. Guo, L.-H. Cao, Chin. J. Catal.
8 (2004) 664e668.
[39] D.-R. Gong, B.-L. Wang, C.-X. Bai, J.-F. Bi, F. Wang, W.-M. Dong, X.-Q. Zhang, L.-
S. Jiang, Polymer 50 (2009) 6259e6264.
[40] G. Natta, L. Porri, A. Mazzei, Chim. Ind. 41 (1959) 116e122.
[41] S. Milione, C. Cuomo, C. Capacchione, C. Zannoni, A. Grassi, A. Proto, Macro-
molecules 40 (2007) 5638e5643.
[42] D.-R. Gong, B.-L. Wang, H.-G. Cai, X.-Q. Zhang, L.-S. Jiang, J. Organomet. Chem.
696 (2011) 1584e1590.
CCDCnumbers
749520
(1a),
740170(1b),
836391(3a),
740445(3b), 740175(3c), 740454(4a), 740172(5a), 749517(5b),
740171(5c) and 740173(6c ) contains the supplementary crystallo-
graphic data for this paper. , . These data can be obtained free of
charge from The Cambridge Crystallographic Data Centre via www.
ccdc.cam.ac.uk/data_request/cif.
[43] D.-R. Gong, X.-Y. Jia, B.-L. Wang, F. Wang, C.-Y. Zhang, X.-Q. Zhang, L.-S. Jiang,
W.-M. Dong, Inorg. Chim. Acta 373 (2011) 47e53.
[44] B.L. Small, M. Brookhart, J. Am. Chem. Soc. 120 (1998) 7143e7144.
[45] B.L. Small, M. Brookhart, A.M.A. Bennett, J. Am. Chem. Soc. 120 (1998)
4049e4050.
Appendix. Supplementary material
Supplementary data related to this article can be found online at
doi:10.1016/j.jorganchem.2011.11.025.
[46] D. Takeuchi, R. Masuura, S. Park, K. Osakada, J. Am. Chem. Soc. 129 (2007)
7002e7003.
[47] Y.-F. Chen, R.-F. Chen, C.-T. Qian, X.-C. Dong, J. Sun, Organometallics 22 (2003)
4312e4321.
[48] D.M. Dawson, D.A. Walker, M. Thornton-Pett, M. Bochmann, J. Chem. Soc.
Dalton Trans. (2000) 459e466.
References
[1] Z.-Q. Shen, Inorg. Chim. Acta 140 (1987) 7e14.
[49] V.C. Gibson, C. Redshaw, G.A. Solan, Chem. Rev. 107 (2007) 1745e1776.
[50] G.J.P. Britovsek, M. Bruce, V.C. Gibson, B.S. Kimberley, P.J. Maddox,
S. Mastroianni, S.J. McTavish, C. Redshaw, G.A. Solan, S. Strömberg,
A.J.P. White, D.J. Williams, J. Am. Chem. Soc. 121 (1999) 8728e8740.
[51] C. Bianchini, G. Mantovani, A. Meli, F. Migliacci, F. Laschi, Organometallics 22
(2003) 2545e2547.
[52] W.-J. Zhang, W.-H. Sun, S. Zhang, J.-X. Hou, H.-B. Song, Organometallics 25
(2006) 1961e1969.
[53] L.-Y. Wang, W.-H. Sun, L.-Q. Han, Z.-L. Li, Y.-L. Hu, C.-X. He, C.-H. Yan,
J. Organomet. Chem. 650 (2002) 59e64.
[54] G.J.P. Britovsek, V.C. Gibson, O.D. Hoarau, S.K. Spitzmesser, A.J.P. White,
D.J. Williams, Inorg. Chem. 42 (2003) 3454e3465.
[55] J.S. Fossey, C.J. Richards, Organometallics 21 (2002) 5259e5264.
[56] S. Nückel, P. Burger, Organometallics 19 (2000) 3305e3311.
[57] W.J. Hoogervorst, C.J. Elsevier, M. Lutz, A.L. Spek, Organometallics 20 (2001)
4437e4440.
[2] Z.-Q. Shen, J. Ouyang, Handbook of the physics and chemistry of rare earth, in:
K. Gschneidner, L. FlemingJr. (Eds.), Ch. 61 (Rare Earth Coordination Catalysts
in Stereospecific Polymerization), Elsevier, Amsterdam, 1987.
[3] W. Kuran, Principle of Coordination Polymerization, John Wiley and Sons Ltd.,
New York, 2001.
[4] L. Friebe, O. Nuyken, W. Obrecht, Adv. Polym. Sci. 204 (2006) 1e154.
[5] A. Fischbach, R. Anwander, Adv. Polym. Sci. 204 (2006) 155e281 (and refer-
ences therein).
[6] S. Arndt, K. Beckerle, P.M. Zeimentz, T.P. Spaniol, J. Okuda, Angew. Chem. Int.
Ed. 44 (2005) 7473e7477.
[7] S. Kaita, Z.-M. Hou, Y. Wakatsuki, Macromolecules 34 (2001) 1539e1541.
[8] W. Gao, D.-M. Cui, J. Am. Chem. Soc. 130 (2008) 4984e4991.
[9] A. Miyazawa, T. Kase, K. Hashimoto, Macromolecules 37 (2004) 8840e8845.
[10] R. Cariou, J. Chirinos, V.C. Gibson, G. Jacobsen, A.K. Tomov, M.R.J. Elsegood,
Macromolecules 42 (2009) 1443e1444.
[11] A. Fischbach, C. Meermann, G. Eickerling, W. Scherer, R. Anwander, Macro-
molecules 39 (2006) 6811e6816.
[12] V. Appukuttan, L. Zhang, C. Ha, I. Kim, Polymer 50 (2009) 1150e1158.
[13] Y. Nakayama, Y. Baba, H. Yasuda, K. Kawakita, N. Ueyama, Macromolecules 36
(2003) 7953e7958.
[58] W.J. Hoogervorst, A.L. Koster, M. Lutz, A.L. Spek, C.J. Elsevier, Organometallics
23 (2004) 1161e1164.
[59] J.M. Vila, M. Gayoso, T. Pereira, T.M. Lopez, A. Fernández, J.M. Ortigueira,
J. Organomet. Chem. 506 (1996) 165e174.
[60] S. Al-Benna, M.J. Sarsfield, M. Thornton-Pett, D. Ormsby, P.J. Maddox, P. Bres,
M. Bochmann, J. Chem. Soc. Dalton Trans. (2000) 4247e4257.
[61] J. Zhang, M. Gandelman, D. Herrman, G. Leitus, L.J.W. Shimon, Y. Ben-David,
D. Milstein, Inorg. Chim. Acta 359 (2006) 1955e1960.
[62] V. Appukuttan, L. Zhang, J.Y. Ha, D. Chandran, B.K. Bahuleyan, C.-S. Ha, Il Kim,
J. Mol. Catal. A: Chem. 325 (2010) 84e90.
[14] W.J. Evans, D.G. Giarikos, J.W. Ziller, Organometallics 20 (2001) 5751e5758.
[15] E. Colamarco, S. Milione, C. Cuomo, A. Grassi, Rapid Commun. 25 (2004) 450e454.
[16] L.-H. Cao, W.-M. Dong, L.-S. Jiang, X.-Q. Zhang, Polymer 48 (2007) 2475e2480.
[17] G. Ricci, M. Battistella, Macromolecules 34 (2001) 5766e5769.
[18] D.-R. Gong, W.-M. Dong, J.-C. Hu, X.-Q. Zhang, L.-S. Jiang, Polymer 50 (2009)
2826e2829.
[63] R. Cariou, J.J. Chirinos, V.C. Gibson, G. Jacobsen, A.K. Tomov, G.J.P. Britovsek,
Andrew J.P. White, Dalton Trans. 39 (2010) 9039e9045.
[64] A.R. Karam, E.L. Catarì, F. López-Linares, G. Agrifoglio, C.L. Albano, A. Dìaz-
Barrios, T.E. Lehmanna, S.V. Pekerar, L.A. Albornoz, R. Atencio, T. González,
H.B. Ortega, P. Joskowics, Appl. Catal. A 280 (2005) 165e173.
[65] R. Gao, L.-W. Xiao, X. Hao, W.-H. Sun, F.-S. Wang, Dalton Trans. (2008)
5645e5651.
[66] S.-Y. Jie, S. Zhang, W.-H. Sun, X.-F. Kuang, T.-F. Liu, J.-P. Guo, J. Mol. Catal. A:
Chem. 269 (2007) 85e96.
[67] S. Tobisch, R. Taube, Organometallics 18 (1999) 5204e5218.
[68] S. Tobisch, Acc. Chem. Res. 35 (2002) 96e104.
[19] Y.C. Jang, P.S. Kim, H. Lee, Macromolecules 35 (2002) 1477e1480.
[20] D.-R. Gong, W.-M. Dong, Y.-M. Hu, J.-F. Bi, X.-Q. Zhang, L.-S. Jiang, Polymer 50
(2009) 5980e5986.
[21] Y.C. Jang, P.S. Kim, H.Y. Jeong, H. Lee, J. Mol. Catal. A: Chem. 206 (2003) 29e36.
[22] D.C.D. Nath, T. Shiono, T. Ikeda, Appl. Catal. A 238 (2003) 193e199.
[23] D.C.D. Nath, T. Shiono, T. Ikeda, Macromol. Chem. Phys. 204 (2003) 2017e2022.
[24] Z.-G. Cai, M. Shinzawa, Y. Nakayama, T. Shiono, Macromolecules 42 (2009)
7642e7643.
[25] K. Endo, N. Hatakeyama, J. Polym. Sci. Part A: Polym. Chem. 39 (2001)
2793e2798.
[26] D. Chandran, C.H. Kwak, C.S. Ha, I. Kim, Catal. Today 131 (2008) 505e512.
[27] P. Cass, K. Pratt, T. Mann, B. Laslett, E. Rizzardo, R. Burford, J. Polym. Sci. Part A:
Polym. Chem. 38 (2000) 1083e1089.
[69] A.R. O’Connor, P.S. White, M. Brookhart, J. Am. Chem. Soc. 129 (2007)
4142e4143.