Nickel(II) dibromide complexes bearing bis(benzimidazolyl)amine and Bis(benzimidazolyl)pyridine ligands for ethylene oligomerizations

Article abstract of DOI:10.1007/s10562-011-0697-9

A series of bis(benzimidazolyl)amine and bis(benzimidazolyl)pyridine ligands and their respective nickel(II) dibromide complexes were synthesized and fully characterized. After activation of the nickel complexes with ethylaluminum sesquichloride showed productivity in excess of 10⁶ (g-oligomer)(mol-Ni)^-1h^-1bar^-1 towards ethylene oligomerization, producing butenes as major products. The bis(benzimidazolyl) pyridine nickel complexes showed higher activity and dimerization selectivity than corresponding bis(benzimidazolyl)amine nickel complexes. Graphical Abstract: [InlineMediaObject not available: see fulltext.]

Full text of DOI:10.1007/s10562-011-0697-9

Catal Lett (2011) 141:1608–1615
DOI 10.1007/s10562-011-0697-9
Nickel(II) Dibromide Complexes Bearing
Bis(benzimidazolyl)amine and Bis(benzimidazolyl)pyridine
Ligands for Ethylene Oligomerizations
•
Gyeong Mi Lee Vinu K. Appukuttan
•
•
Hongsuk Suh Chang-Sik Ha Il Kim
•
Received: 9 August 2011 / Accepted: 4 September 2011 / Published online: 21 September 2011
Ó Springer Science+Business Media, LLC 2011
Abstract
A
series of bis(benzimidazolyl)amine and
group of Brookhart with the development of highly active
catalysts for ethylene polymerization [3–5]. The important
progress of ethylene reactivity promoted by nickel com-
plexes is reflected by the recent review articles [6–11].
Both oligomerization and polymerization can be occurred
in the same batch with the mechanistic difference of chain
termination and propagation. Recent challenges in the
application of nickel complexes in ethylene oligomeriza-
tion are improving their catalytic activity and selectively
for the formation of a-olefins and controlling the polymer
properties of produced polymers. From an academic view,
the fundamental points in designing novel complexes as
good homogeneous catalysts are based on designing and
synthesizing suitable ligands to provide a conductive
environment. Since the initial reports of ethylene poly-
merization with non-metallocene-based catalysts, there has
been a significant focus on developing alternative variants
of the bis(imino)pyridine ligand with varying degrees of
success [10, 12–15]. One direction of exploration has been
the replacement of the central pyridine ring with alternative
heterocycles including pyrimidine, pyrazine, triazine, pyr-
role, carbazole, furan, or thiophene [16]. Subtle changes to
the ligand structure appear to have a dramatic effect upon
catalysis, as well as the ability of the ligand to ligate zir-
conium, chromium, iron or cobalt.
bis(benzimidazolyl)pyridine ligands and their respective
nickel(II) dibromide complexes were synthesized and fully
characterized. After activation of the nickel complexes with
ethylaluminum sesquichloride showed productivity in excess
of 106 (g-oligomer)(mol-Ni)-1 -1bar-1 towards ethylene
h
oligomerization, producing butenes as major products. The
bis(benzimidazolyl)pyridine nickelcomplexesshowedhigher
activity and dimerization selectivity than corresponding
bis(benzimidazolyl)amine nickel complexes.
Keywords Catalysis Á Ethylene oligomerization Á
Ligands Á Nickel complex Á Transition metal catalysts
1 Introduction
The modern chemical industry has much reliance on
a-olefins, and development of new catalysts for oligomer-
ization and polymerization of ethylene is of great interest
in both academic and industrial pursuits. One of the major
industrial processes in ethylene oligomerizations, the Shell
higher olefin process (SHOP), employs nickel complexes
as catalyst [1, 2]. Over the past decade, research into nickel
complexes as catalysts for olefin reactivity has attracted
much attention due to the substantial breakthrough by the
In general, for ligand oriented metal complexes for
ethylene oligomerization, ligands with no substantial
adjacent substituents, facilitating b-hydrogen transfer, are
preferred. In that aspect bis(benzimidazolyl)amine/pyridine
(Fig. 1) with a wide open coordination site can be an ideal
candidate. The transition metal complex of similar ligand
systems Co, Cr and V have been utilized for ethylene and
1,3-butadiene polymerizations [12–16]. In polar organic
solvents, tridentate aromatic and aliphatic ligands pos-
sessing three heterocyclic nitrogen donor atoms can
G. M. Lee Á V. K. Appukuttan Á C.-S. Ha Á I. Kim (&)
Department of Polymer Science and Engineering, The WCU
Centre for Synthetic Polymer Bioconjugate Hybrid Materials,
Pusan National University, Busan 609-735, Korea
e-mail: ilkim@pusan.ac.kr
H. Suh
Department of Chemistry and Chemistry Institute for Functional
Materials, Pusan National University, Pusan 609-735, Korea
123

Ethylene Oligomerizations by Nickel Complexes
1609
Fig. 1 Bis(benzimidazolyl)pyridyl (a) and bis(benzimidazolyl)amine (b) ligands and their geometry optimized structures [(c) and (d)] with
R = Me by density functional theory
displace solvent molecules and/or anions from the first
coordination sphere of divalent nickel salts, NiX2, to give a
plethora of [NiLX2] complexes (L = ligand). Moreover,
the bis(benzimidazole)amine/pyridine backbone is easy to
derivatize, which allows its facile incorporation into
extended segmental ligands for the design of electronically
modulated complexes. In order to utilize the potential of Ni
as catalyst for ethylene oligomerizations we are reporting
synthesis and characterization of a series of new nickel(II)
complexes bearing the tridentate bis(benzimidazol-
yl)amine/pyridine ligands. Ethylene oligomerizations were
performed at different reaction parameters for process
optimization.
spectroscopy (FAB-MS) was recorded by using JMS-700,
JEOL instrument. 1H-NMR (300 MHz) and 13C-NMR
(75 MHz) spectra were recorded at 25 °C on a Varian
Gemini 2000 spectrometer in CDCl3 containing tetra-
methyl silane as standard. Fast atom bombardment mass
spectra (FAB-MS) were obtained from the Korea Basic
Science Institute on a JEOL JMS AX 505 spectrometer,
operated at 3 kV accelerating voltage, equipped with a
JEOL MS-FAB 10 D FAB gun operated at a 10 mA
emission current, producing a beam of 6 keV xenon.
Nitrobenzyl alcohol was used as matrix. Data acquisition
and processing were accomplished using JEOL Comple-
ment software. The oligomers were analyzed by gas
chromatography (GC) (HP6890), performed using a J&W
Scientific DB608 column (30 m 9 0.53 mm) with a flame
ionization detector (FID). The injector and detector tem-
peratures were kept constant at 250 °C. The program set
the initial temperature to 30 °C (held for 2 min) and fin-
ishing temperature to 250 °C (held for 10 min), with a
heating rate of 10 °C/min.
2 Experimental
2.1 Material
All manipulations involving air or moisture sensitive
compounds were carried out under a purified nitrogen
atmosphere. Toluene was distilled over Na/benzophenone
2.3 Synthesis of Ligands 1a–1c and 2a–2c
˚
and stored over molecular sieves (4 A). Tetrahydrofuran
used for the precatalyst synthesis was dried over calcium
˚
hydride and stored over molecular sieves (4 A). Polymer-
Synthetic procedures of ligands and Ni(II) complexes are
illustrated in Scheme 1. The preparation and characteriza-
tion of Cr [12, 13], Mn and Fe [17], and Cu [18] complexes
bearing bis(benzimidazol)amine/pyridine ligands are well
studied for ethylene oligo-/polymerizations [12, 13] and
DNA binding [19]. Previously reported literature proce-
dures were employed to synthesize compounds 1a and 2a
[12], 1b [17], 2b [13] and 2c [18].
ization grade of ethylene was purified by passing it through
columns of Fisher RIDOXTM catalyst and molecular sieve
˚
5 A/13X. Ethylaluminum sesquichloride (EASC) in tolu-
ene solution was obtained from Aldrich and used without
further purification. All other chemicals (Sigma Aldrich
Co.) used without further purifications unless otherwise
mentioned.
For the preparation of N-methyl-N,N-bis[(1-benzyl-1H-
benzimidazol-2-yl)methyl]amine (1c), to 1a (0.50 g,
1.72 mmol) in 15 mL of THF at 0 °C, sodium hydride
(85%) (0.20 g, 6.87 mmol) was added. After 1 h, benzyl
bromide (1.63 mL, 13.73 mmol) was added to the reaction
mixture. The temperature was slowly raised to 50 °C and
2.2 Characterization
Elemental analysis of metal complexes was carried out
using Various EL analyzer. Fast atom bombardment mass
123

1610
G. M. Lee et al.
Scheme 1 Synthesis of
complexes:
i methyliminodiacetic acid ii
pyridine-2,6 dicarboxylic acid
iii NaH/R-X (Me-I, Bn-Br) iv
(DME) NiBr2
4a: greenish yellow crystals in 90.0% yield. Anal. Calcd.
For C31H29N5NiBr2: C, 43.07; H, 2.47; N, 13.22. Found:
C, 43.26; H, 3.01; N, 13.76. FAB-MS (m/z): 450 [M-
Br]?.
stirred the reaction mixture over night. The mixtures were
then added to the water, the off-white precipitate formed
was filtered, washed extensively with water and one time
with cold methanol. The compound was dried at 50 °C
under vacuum to get a yield of 75.4% (0.61 g). Mp:
180–181 °C. 1H NMR (300 MHz, CDCl3, d, ppm): 2.43 (s,
3H), 3.84 (s, 4H), 5.06 (s, 4H), 6.74 (m, 4H), 7.19 (br m,
12H) 7.77 (m, 2H). 13C NMR (75 MHz, CDCl3, d, ppm):
151.13, 142.30, 135.96, 135.58, 128.80, 127.65, 125.89,
123.03, 122.22, 119.90, 109.84, 54.15, 46.53, 43.63. Anal.
Calcd. For C31H29N5: C, 78.95; H, 6.20; N, 14.ll85. Found:
C, 79. 08; H, 6.14; N, 14.78.
4b: greenish yellow crystals in 88.0% yield. Anal. Calcd.
For C31H29N5NiBr2: C, 45.21; H, 3.07; N, 12.55. Found:
C, 45.32; H, 3.21; N, 12.12. FAB-MS (m/z): 568 [M-
Br]?.
4c: greenish yellow crystals in 89.0% yield. Anal. Calcd.
For C31H29N5NiBr2: C, 55.82; H, 3.55; N, 9.86. Found:
C, 56.17; H, 3.70; N, 9.55. FAB-MS (m/z): 620 [M-Br]?.
2.5 Ethylene Oligomerizations
2.4 Synthesis of Nickel Complexes 3a–3c and 4a–4c
Ethylene oligomerizations were performed at 1.3 bar in a
250-mL round-bottomed flask kept at constant temperature
by a thermostat and equipped with a magnetic stirrer and a
thermometer, and the oligomer products were analyzed by
GC according to the literature procedures [19].
All complexes were synthesized using the same procedure.
An equimolar mixture of ligand (2.0 mmol) and (DME)-
NiBr2 (2.0 mmol) was stirred in 10 mL of THF for 12 h at
room temperature. The precipitated complex was washed
once with THF (20 mL) and three times with diethyl ether
(20 mL), and then dried under vacuum.
2.6 Computational Details
3a: light green crystals in 84.0% yield. Anal. Calcd. For
C31H29N5NiBr2: C, 40.05; H, 3.36; N, 13.74 Found: C,
40.53; H, 3.59; N, 13.58. FAB-MS (m/z): 430 [M-Br]?.
3b: light green crystals in 86.0% yield. Anal. Calcd. For
C31H29N5NiBr2: C, 42.42; H, 3.94; N, 13.02. Found: C,
42.86; H, 4.08; N, 12.86. FAB-MS (m/z): 458 [M-Br]?.
3c: light green crystals in 87.0% yield. Anal. Calcd. For
C31H29N5NiBr2: C, 53.95; H, 4.24; N, 10.15. Found: C,
53.58; H, 4.86; N, 10.57. FAB-MS (m/z): 610 [M-Br]?.
The quantum chemical calculations were carried out by
density functional theory (DFT) since it usually gives
realistic geometries, relative energies and vibrational fre-
quencies for transition metal compounds. All calculations
were performed with the DMol3 DFT code [20] as imple-
mented in Accelrys Materials StudioÒ 4.3 (http://www.
accelrys.com/) on a personal computer (PentiumÒ dual-
core CPU at 2.50 GHz, 4.00 gigabytes ram) operated with
123

Ethylene Oligomerizations by Nickel Complexes
1611
Microsoft Windows XP. The non-local generalized gradi-
ent approximation (GGA) functional by Perdew-Burke-
Ernzerhof (PBE) [21] was used for all geometry optimi-
zations. The convergence criteria for these optimizations
high selectivity towards a-olefins, and polyethylenes were
produced with extremely broad polydispersity. Variants of
the bis(imino)pyridine ligand where only one imino arm
has been replaced with benzimidazole have been reported
bound to iron, cobalt, and nickel, all of which successfully
initiated ethylene oligomerization [34–36]. The Gibson
group reported a mechanism study of ethylene oligomeri-
zation by chromium complex bearing 2,6-bis(1H-
benzo[d]imidazol-2-yl)pyridine [37]. More recently the
Sun group synthesized a series of [2,6-bis(2-benzimidaz-
olyl)pyridyl]chromium chlorides and tested for ethylene
oligo-/polymerizations [13]. When MAO was employed as
the cocatalyst, the chromium complexes showed high
activity for ethylene oligomerization and polymerization.
Oligomers were produced with high selectivity towards a-
olefins, and polyethylenes were produced with extremely
broad polydispersity.
consisted of threshold values of 2 9 10-5 Ha, 0.004 Ha/A
˚
˚
and 0.005 A for energy, gradient and displacement con-
vergence, respectively, while a self-consistent field density
convergence threshold value of 1 9 10-5 Ha was speci-
fied. DMol3 utilizes a basis set of numeric atomic func-
tions, which are exact solutions to the Kohn–Sham
equations for the atom [22]. A polarized split valence basis
set, termed double numeric polarized (DNP) basis set was
used. All geometry optimizations employed highly efficient
delocalized internal coordinates [23]. The use of delocal-
ized coordinates significantly reduces the number of
geometry optimization iterations needed to optimize larger
molecules compared with the use of traditional Cartesian
coordinates. Some of the geometries optimized were also
subjected to full frequency analyses at the same GGA/PBE/
DNP level of theory to verify the nature of the stationary
points.
A theoretical calculation was performed to find out the
minimum energy structure (Fig. 2) of all nickel complexes
(3a–3c, 4a–4c) bearing bis(benzimidazolyl)pyridine/amine
ligands. The final structure was obtained using a GGA/
PBE/DNP level of theory with the DMol3 DFT code. The
complexes were fully optimized and a few selected geo-
metric parameters, i.e. bond length and bond angles are
given in Table 1. From the optimized structures of the
complexes, it was observed that the Ni centers in 3a–3c lie
in a distorted tetragonal–pyramidal environment and those
in 4a–4c lie in a distorted trigonal–bipyramidal environ-
ment. The N1–Ni–Br1 bond angle is 169.338, 168.043 and
177.964 for the complexes 3a, 3b and 3c, respectively,
124.842, 127.298 and 117.838 for the complexes 4a, 4b
and 4c, respectively. According to the bond lengths of Ni–
N1, Ni–N2 and Ni–N3 of the complexes 3a, 3b and 3c, Ni–
3 Results and Discussion
3.1 Synthesis and Characterization of Complexes
Ligand 1a was prepared by condensation reaction of 1,2-
phenylenediamine and methyl iminodiacetic acid in eth-
ylene glycol at 190 °C [12]. Ligand 2a was also synthe-
sized in the same way using pyridine-2,6-dicarboxylic acid.
The ligands 1b, 1c, 2b, and 2c were obtained in high yields
by the simple N-alkylation using corresponding alkyl
halides via Na salts of 1a and 2a in THF [13].
3
˚
N has much longer length (3.087–4.454 A) than the other
Nickel complexes 3a–3c and 4a–4c were prepared in
good yields by the matallation with one equivalents of
(DME)NiBr2 in THF at room temperature (see Scheme 1).
The complexes 3a–3c could be envisaged as to originate
bis(benzimdazolyl)s from a methyl amine core and 4a–
c that from a pyridinyl core. All the complexes were
consistent with their elemental analyses and FAB-MS. The
2,6-bis(2-benzimidazolyl)pyridine ligand derivatives have
been reported to be coordinated to transition metals such as
zirconium, chromium, iron and cobalt to form various
metal complexes [13, 16, 24–33]. In the case of benzimi-
dazole-substituted pyridines, ligands of the type bis(benz-
imidazole)pyridine have been applied only to chromium,
where in combination with MAO or diethylaluminum
chloride (DEAC) they acted as highly effective initiators
for ethylene oligomerization [13]. When methylaluminox-
ane (MAO) was employed as the cocatalyst, the chromium
complexes showed high activity for ethylene oligomeriza-
tion and polymerization. Oligomers were produced with
two ones. Possibly the bis(benzimidazolyl)amine ligands
might be coordinated to nickel in a bidentate way, which is
quite different from bis(benzimidazolyl)pyridine nickel
complexes, 4a, 4b and 4c. The observed differences in the
bond angle and bond length between 3a–3c and 4a–4c
come from a basic discrepancy of the structure induced by
the different ligand backbone structures. In addition there
exist differences in the bond angle and bond length in a
same series of complexes, attributable to the difference in
the steric hindrance due to the presence of the different of
alkyl groups on the benzimidazolyl rings.
The Fukui function (FF) [38, 39] was also calculated to
determine the local reactivity in the complexes by using a
GGA/PBE/DNP level of theory with the DMol3 DFT code
(Table 2). Comparing the ratio of FF values for nucleophilic
attack to electrophilic attack (f?/f2-), bis(benzimidazol-
yl)amine Ni complexes display higher value than bis(benz-
imidazolyl)-pyridine Ni complexes. Therefore, for
electrophilic attack, the Ni center of the complex 3a–3c is
123

1612
G. M. Lee et al.
Fig. 2 Density functional
theory based geometry
optimized structure of the
complexes
Table 1 Some important bond lengths and bond angles estimated for the complexes 3a–3c and 4a–4c by geometry optimization
Parameters 3a 3b 3c 4a 4b
4c
˚
Bond length (A)
Ni–N1
Ni–N2
Ni–N3
Ni–Br1
Ni–Br2
C1–C2
2.707
1.937
4.454
2.327
2.357
1.481
2.081
2.088
1.967
3.087
2.331
2.353
1.480
1.894
1.909
1.909
2.443
2.433
1.439
1.892
1.896
1.891
1.914
2.455
2.428
1.445
1.934
3.919
2.326
2.343
1.482
1.912
1.892
2.441
2.444
1.445
Bond angle (degree)
N1–Ni–N2
N1–Ni–N3
N1–Ni–Br1
N1–Ni–Br2
N2–Ni–Br1
N3–Ni–Br2
Br1–Ni–Br2
C1–N2–Ni
C4–N3–Ni
82.757
46.899
169.338
93.255
95.218
91.195
90.685
107.263
33.477
82.320
51.836
168.043
93.323
95.264
93.323
91.888
114.725
51.836
82.660
63.311
177.964
91.560
90.084
93.775
90.025
113.400
90.495
81.229
81.234
81.443
81.336
81.462
81.402
124.842
129.241
94.763
127.298
126.895
94.897
117.838
134.143
93.384
95.754
95.611
96.294
105.919
114.358
114.354
105.805
115.164
114.432
108.014
115.177
114.212
more active than 4a–4c. For both series of complexes, the f?/
f2- value decreases as the alkyl group in benzimidazolyl
rings become bulky from hydride to benzyl. These electronic
parameters together with the structure of the complexes and
steric bulk will play important roles in deciding
oligomerization reactivity (vide infra). Modeling all the
NiBr2 complexes by ab initio calculations showed consid-
erable difference in the electro-static potential fitting (ESP)
charge distributions within these molecules. In general the
nickel atoms in the complexes bearing N-benzyl substituted
123

Ethylene Oligomerizations by Nickel Complexes
1613
Table 2 Some important parameters of the complexes 3a–3c and 4a–4c
Parameters
3a
3b
3c
4a
4b
4c
Optimized Ea
Binding Ea
–7586.96
-7665.45
-8126.67
-12.4343
0.027
-7660.74
-7739.26
-8200.90
-12.9789
-0.001
0.034
-7.7193
-8.6225
-7.9086
-8.8488
Fukui functionb: f?
Fukui functionc: f?
Fukui functionb: f2-
Fukui functionc: f2-
f?/f2-(Mullikan)
f?/f2-(Hirshfeld)
HOMO Ea
0.107
0.145
0.106
0.143
-0.003
0.012
-0.003
0.011
0.127
0.224
0.219
0.153
-0.009
0.034
-0.011
0.033
-0.006
0.139
0.237
0.231
0.249
0.478
0.484
0.177
0.333
0.273
0.167
0.612
0.619
0.510
0.353
0.333
0.245
-0.09570
-0.17340
-0.15220
-0.16819
-0.16573
-0.16201
Electro-static potential fitting (ESP) chargesd
Metal
Br1
Br2
N1
N2
N3
0.436
-0.376
-0.406
0.139
0.321
-0.342
-0.396
0.192
0.447
-0.400
-0.466
-0.302
-0.239
-0.399
0.217
-0.395
-0.391
-0.043
-0.278
-0.278
0.268
-0.394
-0.394
-0.007
-0.322
-0.340
0.406
-0.401
-0.415
-0.094
-0.409
-0.411
-0.594
-0.575
-0.423
-0.537
a
b
c
d
In Hartre
Fukui function for nucleophilic (f?) and electrophilic (f-2-) attacks in Mullikan charge
Fukui function in Hirshfeld charge
Notation of the type of atom is in Table 1
ligands in each system are more electropositive than other
complexes bearing N-methyl or non-substituted ligands, as a
result the electropositivity of nickel atom decreases in the
order of 3c [ 3a [ 4c [ 3b [ 4b [ 4a. However, on the
basis of these calculations the electronic differences among
the complexes appear to be less than are suggested by the
differences in reactivity towards ethylene.
3.2 Ethylene Oligomerizations
The oligomerization rate of ethylene under constant pres-
sure using bis(benzimidazolyl)amine Ni (3a–3c) and
bis(benzimidazolyl)pyridine Ni (4a–4c) complexes com-
bined with EASC changes markedly with time, as shown in
Fig. 3. These kinetic curves can be divided into two parts:
i.e. induction period where the rate increases rapidly to a
maximum followed by a decay period during which the
rate decreases monotonously, even though the induction
and decay periods are somewhat changed depending on the
type of catalyst. Note that the preliminary screening tests
Fig. 3 Ethylene oligomerization rate profiles obtained by different
Ni(II) complexes combined with ethylaluminum sesquichloride
(EASC): a 3a, b 3b, c 3c, d 4a, e 4b, and f 4c. Oligomerization
conditions: catalyst = 5 lmol, [EASC]/[Ni] = 200, ethylene pres-
sure = 1.3 bar, toluene = 80 mL, T = 30 °C and time = 30 min
i
for the type of cocatalyst such as AlR3 (R = Me, Et, Bu),
AlEt2Cl and MAO show that EASC is the best cocatalyst
for Ni catalysts of this study. In fact all alkyl aluminum
cocatalysts, except EASC, gave only negligible activities.
Theoligomerizationresultsare summarizedinTable 3. All
the ethylene oligomerization reactions were performed
employing EASC as the cocatalyst and toluene as a solvent
under a semi-batch condition. The complexes derived from
pyridinewiththeir activity rangingfrom1.662to2.803 9 106
(g-oligomer)(mol-Ni)-1 -1bar-1 weremoreactivethanthose
h
derived from methylamine [1.086–1.622 9 106 (g-oligo-
mer)(mol-Ni)-1 -1
h
bar-1]. As a result the activity (as the
123

1614
G. M. Lee et al.
Table 3 Ethylene oligomerization results obtained by bis(benzimidazolyl)amine Ni (3a–3c) and bis(benzimidazolyl)pyridine Ni (4a–4c)
complexes combined with ethylaluminium sesquichloride (EASC)
Oligomer distribution (%)b
a
Entry
Catalyst
[Al]/[Ni]
T (8C)
Yield (g)
Rp,avg
C4
a-C4
C6
1
3a
3b
3c
4a
4b
4c
3a
3a
3a
3a
3a
3a
3a
200
200
200
200
200
200
50
30
30
30
30
30
30
30
30
30
30
10
50
70
3.53
5.40
3.81
5.56
5.27
9.11
3.91
6.32
3.53
2.47
0.62
1.56
0.51
1086
1622
1172
1711
1662
2803
1203
1945
1086
760
84.1
83.8
84.3
86.3
87.2
84.4
77.2
81.9
84.1
90.4
86.4
87.0
92.5
15.9
12.3
13.7
36.1
35.7
30.8
12.7
14.2
15.9
15.1
10.4
28.7
26.3
15.9
16.2
15.7
13.7
12.8
15.6
22.8
18.1
15.9
9.6
2
3
4
5
6
7
8
100
200
800
100
100
100
9
10
11
12
13
191
13.6
13.0
7.5
480
157
Oligomerization condition: catalyst = 5 lmol, PC2H4 = 1.3 bar, toluene = 80 mL, and time = 30 min
a
Average rate of polymerization over a polymerization period in kg (mol-Ni)-1h-1bar-1
b
Determined by GC
average rate of oligomerization over a period of oligomeri-
zation) decreases in the order of 4c [ 4a [ 4b [ 3b [
3c [ 3a. The activity changes might be arising from the
variation of electronic nature due to structural differences of
these two classes of complexes. The pyridyl rings in com-
plexes 4a–4c with their resonance effect possibly draw the
lone pair of electrons from N atom into the ring, thereby
reducing the availability of these lone pair of electrons to Ni
metal center in comparison with N-methyl groups in 3a–3c, in
which the positive inductive of methyl group makes the lone
pair electron more easily available for central metal. Note that
the f? and f2- values and their ratio f?/f2- (Table 2) of 3a–3c
complexes are larger than 4a–4c complexes and the higher
values of these parameters are not favorable to achieve high
activity. In addition to these electronic factors, the intrinsic
planar nature of the bis(benzimidazolyl)pyridine complexes
may facilitate the accessibility of the incoming monomers.
The ligand variation brought about by N-alkylation had
a significant effect on oligomerization. For complexes 3a–
3c, the methylated derivative (3b) yields the best activity of
1.622 9 106 (g-oligomer)(mol-Ni)-1h-1bar-1. In the case
of complexes 4a–4c, there is a strong effect on activity by
N-alkylation. As a result the complex 4c bearing benzyl
substituent gives very high activity, 2.803 9 106 (g-oli-
gomer)(mol-Ni)-1h-1bar-1, which is almost 2 times high
than that by complex 4a. There was a drastic decrease in
activity due to N-alkylation which is attributed to forma-
tion of anion amide group by the deprotonation of the free
N–H group by cocatalyst during activation [13].
GC analysis of the products revealed that both the cat-
alytic systems were highly selective towards butenes con-
stituting around 85% of the total product and the remaining
fractions were hexenes. Even though selectivity towards C4
olefins were almost similar level, pyridine derived system
showed better selectivity to 1-butene (30.8–36.1%) than
methyl amine system (12.3–15.9%). The catalyst activity is
comparable to some Ni(II) coordination catalysts with
N-donor ligands under similar conditions. Previously,
Brookhart reported that catalysts with only one substituent
on the N-ring for the a-diimine do not yield polymers but
instead produce oligomers [40]. The products are mainly
butenes and hexenes, with small/moderate amounts of
higher oligomers up to C10. The oligomer distribution
pattern is consistent with a nickel hydride species being the
active center, where fast elimination of b-hydride limits the
products to mostly butenes and hexanes; however, in a few
cases higher olefins are also observed in trace amounts.
In general the parameters such as cocatalyst concentration
and temperature play important roles in determining the reac-
tion profiles and product distribution. The results of ethylene
oligomerization gathered at different conditions are summa-
rized in Table 3. In order to investigate the effect of variation of
cocatalyst concentration, a series of reactions were performed
with increasing [Al]/[Ni] ratios (from 50 to 800) with a 3a/
EASC system. The activity of the catalyst increases from
1.203 9 106 to 1.945 9 106 (g-oligomer)(mol-Ni)-1 -1bar-1
h
by increasing [Al]/[Ni] ratio from 50 to 100; however the
activity decreases monotonously by further increase of [Al]/
123

Ethylene Oligomerizations by Nickel Complexes
1615
¨
¨
2. Keim W, Behr A, Limbacker B, Kruger C (1983) Angew Chem
[Ni] ratio. The decrease of activity in the presence of excess
amount of cocatalyst is usually observed in transition metal
catalyzed olefin oligomerizations. In transition metal catalyzed
oligomerization process, the active species are established as
cation–anion ion pair, formed through the activation of catalyst
precursors by the aluminum cocatalyst and reaction attributes
connected with the nature of the ion pairing. At higher con-
centration of cocatalyst, the ion pair could be stabilized by the
excess cocatalyst making monomer insertion difficult and
hence a reduced the activity [41]. On increasing the [Al[/[Ni]
from 50 to 800 favored formation of C4 fraction due to exten-
sive chain transfer to Al.
Int Ed Engl 22:503
3. Brookhart M, Johnson LK, Killian CM (1995) J Am Chem Soc
117:6414
4. Johnson LK, Mecking S, Brookhart M (1996) J Am Chem Soc
118:267
5. Killian CM, Temple DJ, Johnson LK, Brookhart M (1996) J Am
Chem Soc 118:11664
6. Ittel SD, Johnson LK, Brookhart M (2000) Chem Rev 100:1169
7. Mecking S (2001) Angew Chem Int Ed 40:534
8. Gibson VC, Spitzmesser SK (2003) Chem Rev 103:283
9. Speiser F, Braustein P, Saussine L (2005) Acc Chem Res 38:784
10. Zhang W, Zhang J, Sun WH (2005) Prog Chem 17:310
11. Bianchini C, Giambastiani G, Luconi L, Meli A (2010) Coord
Chem Rev 254:431
12. Gibson VC, and Tomov AK To BP Chemical Ltd WO
2004/083263
13. Zhang W, Sun WH, Zhang S, Hou J, Wedeking K, Schultz S,
To explore temperature effect oligomerizations were
performed at temperatures ranging from 10 to 70 °C using
3a/EASC system at [Al[/[Ni] = 200 (Table 3). The opti-
mum activity was observed at 30 °C. There is a drastic
reduction in activity on increasing temperature from 30 to
70 °C, demonstrating the deactivation of active species is
accelerated at elevated temperatures. In addition facilitated
chain transfer reactions at high temperatures increase the
production of C4 olefin: 92.5% at 70 °C.
¨
Frohlich R, Song H (2006) Organometallics 25:1961
14. Appukuttan V, Zhang L, Ha C-S, Kim I (2009) Polymer 50:1150
15. Cariou R, Chirinos J, Gibson VC, Jacobsen G, Tomov AK,
Elsegood MRJ (2009) Macromolecules 42:1443
16. Tenza K, Hanton MJ, Slawin AMZ (2009) Organometallics 28:
4852 and the references therein
17. Adams H, Bailey NA, Crane JD, Fenton DE, Latour JM, Wil-
liams JM (1990) J Chem Soc Dalton Trans 1727
18. Xu X, Xi Z, Wanzhi C, Daqi W (2007) J Coord Chem 60:2297
19. Chandran D, Kwak CH, Oh JM, Ahn IY, Ha C-S, Kim I (2008)
Catal Lett 125:27
4 Conclusions
20. Delley B (2000) J Chem Phys 113:7756
21. Perdew JP, Wang Y (1992) Phys Rev B 45:13244
22. Delley B, Seminario JM, Politzer P (1995) Modern density
functional theory: a tool for chemistry. Elsevier, Amsterdam
23. Andzelm J, King-Smith RD, Fitzgerald G (2001) Chem Phys Lett
335:321
24. Breslow R, Hunt JT, Smiley R, Tarnowski T (1983) J Am Chem
Soc 105:5337
25. Wu FJ, Kurtz DM, Hagen KS, Nyman PD, Debrunner PG,
Vankai VA (1990) Inorg Chem 29:5174
26. Kimblin C, Allen WE, Parkin G (1995) J Chem Soc Chem
Commun 1813
27. Sorrell TN, Allen WE, White PS (1995) Inorg Chem 34:952
28. Chatlas J, Kaizaki S, Kita E, Kita P, Sakagami N, van Eldik R
(1999) J Chem Soc Dalton Trans 91
29. Winter JA, Caruso D, Shepherd RE (1988) Inorg Chem 27:1086
30. Bocarsly JR, Chiang MY, Bryant L, Barton JK (1990) Inorg
Chem 29:4898
Two classes of Ni(II) complex bearing bis(benzimidazol-
yl)amine and bis(2,6-benzimidazolyl)pyridine ligands with
varying N-alkyl substitutents were synthesized and character-
ized. From the optimized structures of the complexes using a
GGA/PBE/DNP level of theory with the DMol3 DFT code, the
Ni centers in bis(benzimidazolyl)amine and bis(2,6-benzimi-
dazolyl)pyridine ligands lie in a distorted tetragonal–pyramidal
and a distorted trigonal–bipyramidal environment, respec-
tively. All the Ni(II) complexes combined with EASC cocata-
lyst showed high activity towards ethylene to yield butenes
(around 85%) as major product along with hexenes. The Ni(II)
complexes bearing planar bis(2,6-benzimidazolyl)pyridine
ligands showed higher activity, dimerization and 1-butene
selectivity than those bearing bis(benzimidazolyl)amine
ligands. The modifications of ligand by N-alkylation had a
strong influence on oligomerization profiles. Higher oligo-
merization temperatures and cocatalyst ratio favored the pro-
duction C4 fraction and enhanced selectivity towards 1-butene.
´
31. Ceniceros-Gomeza AE, Barba-Behrensa N, Quiroz-Castro ME,
`
¨
Bernes S, Noth H, Castillo-Blum SE (2000) Polyhedron 19:1821
32. Wang L, Sun WH, Han L, Yang H, Hu Y, Jin X (2002) J Or-
ganomet Chem 658:62
33. Ma Z, Sun WH, Li ZL, Shao CX, Hu YL (2002) Chin J Polym Sci
20:205
34. Hao P, Zhang S, Sun WH, Shi Q, Adewuyi S, Lu X, Li P (2007)
Organometallics 26:2439
Acknowledgments This study was supported by grants-in-aid for
the World Class University Program (No. R32-2008-000-10174-0)
and the National Core Research Center Program from MEST (No.
R15-2006- 022-01001-0).
35. Sun WH, Hao P, Zhang S, Shi Q, Zuo W, Tang X (2007)
Organometallics 26:2720
36. Xiao L, Gao R, Zhang M, Li Y, Cao X, Sun WH (2009) Orga-
nometallics 28:2225
37. Tomov AK, Chirinos JJ, Jones DJ, Long RJ, Gibson VC (2005) J
Am Chem Soc 127:10166
38. Parr RG, Yang W (1984) J Am Chem Soc 106:4049
39. Thanikaivelan P, Padmanabhan J, Subramanium V, Ramasami T
(2002) Theor Chem Acc 107:326
40. Svejda SA, Brookhart M (1999) Organometallics 18:65
41. Chen EYX, Marks TJ (2000) Chem Rev 100:1391
References
1. Keim W, Kowaldt FH, Goddard R, Kru¨ger C (1978) Angew
Chem Int Ed Engl 17:466
123