Bidentate iron complexes based on hyperbranched salicylaldimine ligands and their catalytic behavior toward ethylene oligomerization

Article abstract of DOI:10.1007/s11243-017-0137-9

A series of bidentate iron complexes based on hyperbranched salicylaldimine ligands were synthesized and characterized by spectroscopic and analytical methods. Upon activation?with methylaluminoxane (MAO), the complexes showed good activities [up to 8.17?×?10⁴ g/(mol Fe?h)] for ethylene oligomerization. Activation of the bidentate iron complex with a 1-octadecyl moiety in the ligand backbone (complex C3) with Et₂AlCl produced higher catalytic activity than C3 with MAO, although the selectivity for C₈₊ oligomers was lower. The choice of solvent and reaction parameters significantly affected both the activities and selectivities of these complexes. Under the conditions ([Fe]?=?5?μmol; temperature?=?25?°C; toluene?=?50?mL; time?=?30?min; ethylene pressure?=?0.5?MPa; MAO as cocatalyst), complex C3 gave high activity [7.46?×?10⁴ g/(mol Fe h)] with better selectivity for C₈₊ oligomers (26.58%). The catalytic activities and selectivities were also influenced by the ligand structure and choice of metal. The catalytic activities declined with increasing alkyl chain length of the ligand backbone. Compared to the nickel complex with 1-tetradecyl as core in the ligand backbone (C4), the iron complexes exhibited lower catalytic activities but the better selectivities for C₁₀₊ oligomers.

Full text of DOI:10.1007/s11243-017-0137-9

Transit Met Chem
DOI 10.1007/s11243-017-0137-9
Bidentate iron complexes based on hyperbranched salicylaldimine
ligands and their catalytic behavior toward ethylene
oligomerization
1
1
1
1
1
•
Cui-Qin Li
1
Jun Wang
•
Feng-Feng Wang
•
Rui Gao
•
Peng Sun
•
Na Zhang
Received: 21 November 2016 / Accepted: 20 March 2017
Ó Springer International Publishing Switzerland 2017
Abstract A series of bidentate iron complexes based on
hyperbranched salicylaldimine ligands were synthesized
and characterized by spectroscopic and analytical methods.
Upon activation with methylaluminoxane (MAO), the
Introduction
Ethylene oligomerization is an important catalytic process
for the production of linear a-olefins, which are extensively
used in the preparation of lubricants, plasticizers, deter-
gents and surfactants, as well as comonomers for the syn-
thesis of linear low-density polyethylene [1]. Initial
discovery of the so-called nickel effect by Ziegler leading
to ethylene polymerization [2] and oligomerization [3]
triggered intensive efforts to develop new homogeneous
late transition metal catalysts. Such catalysts can tolerate a
range of polar functional groups and even water, which
tends to poison traditional Ziegler–Natta/metallocene cat-
alysts. Development of iron and cobalt catalysts in partic-
ular was sparked by the independent reports [4–7] of
Gibson and Brookhart on bis(arylimino)pyridine iron and
cobalt as catalysts for ethylene oligomerization and poly-
merization, leading to highly linear products. Moreover,
the iron and cobalt complexes have high activities and
selectivities for a-olefins [5, 6].
Iron-based catalysts are gaining much attention in var-
ious organic transformations due to the abundance, low
toxicity and low cost of this metal. Following investiga-
tions of the iron active species [8], it is generally accepted
catalytic iron species should possess a 14e configuration
akin to that involved in metallocenes [9] and nickel-based
catalysts [10]. For this reason, there have been a number of
investigations into iron-based complexes bearing modified
bis(imino)pyridine derivatives [11]. Apart from iron(II)
complexes of bis(imino)pyridines ligands, other highly
active iron complexes have been explored through the
4
complexes showed good activities [up to 8.17 9 10 g/
(
mol Fe h)] for ethylene oligomerization. Activation of the
bidentate iron complex with a 1-octadecyl moiety in the
ligand backbone (complex C3) with Et AlCl produced
2
higher catalytic activity than C3 with MAO, although the
selectivity for C8? oligomers was lower. The choice of
solvent and reaction parameters significantly affected both
the activities and selectivities of these complexes. Under
the conditions ([Fe] = 5 lmol; temperature = 25 °C;
toluene = 50 mL;
time = 30 min;
ethylene
pres-
sure = 0.5 MPa; MAO as cocatalyst), complex C3 gave
4
high activity [7.46 9 10 g/(mol Fe h)] with better selec-
tivity for C8? oligomers (26.58%). The catalytic activities
and selectivities were also influenced by the ligand struc-
ture and choice of metal. The catalytic activities declined
with increasing alkyl chain length of the ligand backbone.
Compared to the nickel complex with 1-tetradecyl as core
in the ligand backbone (C4), the iron complexes exhibited
lower catalytic activities but the better selectivities for
C10? oligomers.
&
Cui-Qin Li
dqpiliciqin@126.com
2
2
development of sp -nitrogen tridentate [12, 13] and sp -
nitrogen bidentate ligands [14]. Following on from these
studies, variation of the substituents has allowed access not
only to high catalytic activities, but also to the potential
1
Provincial Key Laboratory of Oil and Gas Chemical
Technology, College of Chemistry and Chemical
Engineering, Northeast Petroleum University, Daqing,
Heilongjiang, China
123

Transit Met Chem
commercialization of systems for ethylene oligomerization
and/or polymerization.
Chemical Research Institute. Generation 1.0 hyper-
branched macromolecules with 1-tetradecyl as core (R -
1
4
These considerations suggest that the structure of the
ligand plays a major role in producing effective catalysts.
One such ligand type is salicylaldimine, which has been
used to prepare effective nickel ethylene polymerization
catalysts [15]. The salicylaldimine ligand has the ability to
coordinate metals through hard nitrogen and oxygen donor
atoms, which leads to better stabilization of metal com-
plexes against reduction and usually good thermal stabili-
ties [16]. Our groups [17] have synthesized dendritic
salicylaldimine nickel-based complexes with 1.0 genera-
tion dendritic polyamide-amine as a bridging group and
investigated their potential for ethylene oligomerization.
Upon activation with methylaluminoxane (MAO), the
complex exhibited high activity and selectivity for C –C
14
1.0G), with 1-hexadecyl as core (R -1.0G) and with
16
1-octadecyl as core (R -1.0G) were prepared according to
1
8
the literature procedures [22]. The hyperbranched salicy-
laldimine ligands with R -1.0G as the backbone (L1), with
1
4
R -1.0G as the backbone (L2) and with R -1.0G as the
1
6
18
backbone (L3) were synthesized according to the literature
procedures [23]. The nickel complex with R -1.0G as the
1
4
backbone (C4) was synthesized through the reaction
between L1 and anhydrous nickel chloride according to the
method described in the literature [23]. FTIR spectra were
recorded on a Nicolet FTIR 750 infrared spectrometer
1
using KBr pellets. H NMR spectra were obtained using a
Varian 400 MHz spectrometer with CDCl as the solvent
3
and tetramethylsilane (TMS) as the internal standard. UV–
Vis spectra were determined using a UV-1700 UV–Vis
spectrophotometer. MS data were collected with a Bruker
Apex Ultra 70 FTMS using electrospray ionization (ESI) as
the ion source. GC analyses were conducted with a Fuli GC
9720 instrument equipped with flame ionization detector
(FID) and a 50-m (0.2 mm i.d., 0.5 lm film thickness) HP-
PONA column.
1
0
products. Following this study, we [18] have also synthe-
sized hyperbranched salicylaldimine nickel catalysts with
an octane alkyl group at one end. The catalytic activity
5
reached up to 5.59 9 10 g/(mol Ni h) with MAO as the
cocatalyst, and the main products were longer-chain oli-
gomers (C –C ). However, there are very few salicy-
1
0
18
laldimine iron catalysts [19, 20]. The few known examples
have been used in atom transfer radical polymerization
(
ATRP) [19]. Our group [20] have synthesized iron com-
Synthesis of the iron complexes
plexes with hyperbranched salicylaldimine ligands and
investigated the influence of the molecular cavity within
the catalyst structure on ethylene oligomerization. These
iron coordination complexes, when activated with MAO,
exhibited moderate activities in ethylene oligomerization.
Therefore, we have synthesized three bidentate iron com-
plexes based on hyperbranched salicylaldimine ligands and
investigated their properties as catalysts for ethylene
oligomerization. The catalytic reaction parameters, length
of alkyl chain of the ligand backbone and the choice of
metal center on their catalytic activities have been inves-
tigated in detail. Good catalytic activities toward ethylene
oligomerization were observed in the presence of MAO.
A solution of iron(II) chloride tetrahydrate (0.11 g,
0.56 mmol) in methanol (5 mL) was added dropwise to a
solution of L1 (0.30 g, 0.46 mmol) in methanol (15 mL)
under a nitrogen atmosphere, and the resulting mixture was
stirred at 25 °C for 24 h. Diethyl ether (150 mL) was then
added, giving a red precipitate. The solid was collected by
filtration, washed with cold ether (100 mL) and then dried
under vacuum to obtain the iron complex with R -1.0G as
1
4
the backbone (C1). Yield: 0.31 g (97%). FTIR (KBr
1
-
cm ): m (C=N) 1621 (s), m (C–O) 1306 (m), m (N–Fe) 613
(w). Anal. Calcd. for C H N FeO : C, 64.85; H, 8.16; N,
3
8
57
5
4
9.95. Found: C, 64.79; H, 8.28; N, 9.98%. ESI–MS (m/z):
?
?
?
7
03 [M] , 649 [M–Fe ? H] , 542 [M–Fe–C H O ? H] ,
7 5
?
52 [M–Fe–C H ? H] .
4
1
4 29
Experimental
The iron complex with R -1.0G as the backbone (C2)
16
was prepared according to the method described for C1
using L2 (0.31 g, 0.46 mmol) and iron(II) chloride
tetrahydrate (0.11 g, 0.56 mmol). Yield: 0.32 g (95%).
Materials and instrumentation
-
1
The reactions of air- and/or moisture-sensitive compounds
were performed under a nitrogen atmosphere using stan-
dard Schlenk techniques [21]. All solvents were of ana-
lytical grade and were dried and distilled prior to use.
Methylaluminoxane (10 wt% in toluene) and diethylalu-
minum chloride (25 wt% in toluene) were purchased from
Sigma-Aldrich. Toluene and methanol were provided by
the Tianjin Kermel Chemical Reagent Co., Ltd. Salicy-
laldehyde was obtained from the Tianjin Guangfu Fine
FTIR (KBr cm ): m (C=N) 1617 (s), m (C–O) 1311 (m), m
(N–Fe) 614 (w). Anal. Calcd. for C H N FeO : C, 65.65;
4
0
61
5
4
H, 8.40; N, 9.57. Found: C, 65.29; H, 7.84; N, 9.69%. ESI–
?
?
MS (m/z): 732 [M] , 678 [M–Fe ? H] , 571 [M–Fe–
?
C H O ? H] , 465 [M–Fe–C H ? H] .
?
7
5
15 31
The iron complex with R -1.0G as the backbone (C3)
8
1
was prepared according to the method described for C1
using L3 (0.32 g, 0.46 mmol) and iron(II) chloride
tetrahydrate (0.11 g, 0.56 mmol). Yield: 0.33 g (96%).
1
23

Transit Met Chem
-
1
FTIR (KBr cm ): m (C=N) 1619 (s), m (C–O) 1308 (m), m
N–Fe) 614 (w). Anal. Calcd. for C H N FeO : C, 66.39;
spectra of all three pro-ligands, in the region of
-
1619–1621 cm . Bands at 3300–3400 cm are assigned
1
-1
(
4
2
65
5
4
H, 8.62; N, 9.22. Found: C, 65.80; H, 8.45; N, 9.56%. ESI–
to the O–H functionalities of the salicylaldimine units. In
1
the H NMR spectra of the pro-ligands, the signal for the
?
MS (m/z): 760 [M] , 706 [M–Fe ? H] , 599 [M–Fe–
?
?
C H O ? H] , 488 [M–Fe–C H ? H] .
?
imine proton was observed around d 8.33–8.37 ppm. The
UV spectra of the hyperbranched salicylaldimine ligands in
methanol solution showed three absorption bands at 206,
7
5
15 31
General procedure for ethylene oligomerization
249 and 315 nm. The band at 206 nm can be assigned to
Ethylene oligomerization was carried out in a 250-mL
stainless steel autoclave equipped with a magnetic stirrer
and ethylene pressure control system. The reactor was
immersed in a bath previously set to the desired tempera-
ture and charged with ethylene after removing nitrogen gas
under vacuum. A typical reaction was performed by
introducing toluene (40 mL) and the required amount of
the R band of the C=O n ? p* transition, while the bands
around 249 and 315 nm are most probably due to p ? p*
transitions of the benzene rings and n ? p* transitions of
azomethine (C=N), respectively [24]. The mass spectra of
?
L1, L2 and L3 showed peaks that correspond to [M ? H]
at m/z 650.5, 678.5 and 706.5, respectively. These values
are in good agreement with the proposed compositions for
these hyperbranched salicylaldimines.
cocatalyst (MAO or Et AlCl) into the reactor under an
2
ethylene atmosphere. After 20 min, a toluene solution of
the catalyst (10 mL, [Fe] = 5 lmol) was injected into the
reactor under a stream of ethylene and then the reactor was
immediately pressurized. Ethylene was continuously sup-
plied in order to maintain the desired pressure. After
The bidentate iron complexes C1–C3 were prepared
from the pro-ligands L1–L3 by reaction with iron(II)
chloride (Scheme 1). All three complexes were isolated as
red solids. They are paramagnetic and could thus not be
characterized by NMR spectroscopy. However, other ana-
lytical techniques such as FTIR spectra (Fig. 1), UV
spectra (Fig. 2) and mass spectra (Fig. 3) confirmed their
formulations. Thus, the FTIR spectra of complexes C1–C3
showed shifts in the m(C=N) and m(C–O) stretching fre-
quencies compared to the free pro-ligands. In the case of
the m(C=N) band, the shift was from 1634 (pro-ligand) to
3
0 min, the reactor was cooled in an ice-water bath and
unreacted ethylene was vented. The total volume of the
products was determined and the reaction was quenched by
addition of 10% HCl in ethanol. The organic phase was
collected to investigate the distribution of the products by
GC.
-
1
1
619 cm (complex), while in the case of the m(C–O)
-
1
band the shift was from 1290 (pro-ligand) to 1308 cm
Results and discussion
(complex). These shifts are consistent with coordination of
the ligand to the metal via both the nitrogen and oxygen
donor sites. This was further confirmed by m(O–H) bands
Synthesis of the bidentate iron complexes
-
1
between 3400 and 3300 cm , which are observed in the
spectra of the pro-ligands but not those of the complexes.
The UV spectra of the iron complexes showed four
absorption bands. The three bands observed around 206,
234 and 255 nm are similar to what was observed in the
spectra of the pro-ligands, but shifted to lower wave-
lengths. The weak fourth band at 320 nm can be assigned
to forbidden d–d transitions of the metal and resembles
what has been seen for other salicylaldimine complexes as
We have previously reported the synthesis of a hyper-
branched salicylaldimine pro-ligand via a Schiff-base
condensation reaction of salicylaldehyde with the first-
generation hyperbranched PAMAM [23]. The same syn-
thetic approach was used to prepare the hyperbranched
salicylaldimine pro-ligands L1–L3 (Scheme 1). The pro-
gress of the reaction was monitored by FTIR using the
m(C=N) band. This band was clearly visible in the FTIR
Scheme 1 Synthesis of the hyperbranched salicylaldimine ligands and their bidentate iron complexes
123

Transit Met Chem
1
00
2.60%
C3
C2
8
6
0
0
47.04%
C1
40
40.35%
2
0
0
0
200
400
600
800 1000 1200
4
000 3500 3000 2500 2000 1500 1000 500
Temperature
( )
-1
cm
Fig. 4 TGA curve for the complex C3
Fig. 1 FTIR spectra of the bidentate iron complexes
based on the TGA curve. The first stage shows a mass loss
of 2.6%, up to 160 °C, and is attributed to the loss of
absorbed moisture. The next mass loss in the temperature
range of 160–445 °C corresponds to decomposition of the
ligand and loss of –C H O N; the observed mass loss of
4
C1
4
C2
3
C3
3
2
2
4 45 2
1
47.0% is in fair agreement with the calculated value of
50.0%. The final mass loss in the range of 445–780 °C
corresponds to complete decomposition of the ligand and
loss of ethylenediamine and benzene; the experimental
result (40.4%) is in good agreement with the calculated
value (40.6%). The residual mass remains almost constant
until 800 °C; the residual mass (9.5%) is consistent with
assignment of this residue to iron oxide (calc. 10.0%).
2
2
10 220 230 240 250 260 270
nm
1
0
2
00
300
400
500
nm
Fig. 2 UV spectra of the complexes
Ethylene oligomerization studies
380
675
760
338
C3
7
06
The oligomerization of ethylene catalyzed by these
bidentate iron complexes with various cocatalysts and in
various solvents has been systematically investigated.
Reaction factors such as the molar ratio of cocatalyst to
iron complex, the reaction temperature and pressure each
have an influence on the activity toward ethylene
oligomerization. Before carrying out parallel trials, we
assayed the standard deviations for catalytic activity and
oligomer distributions under the same conditions (Table 1).
In the past few years, numerous studies have revealed
the influence of cocatalyst on catalytic activity, stability,
488
2
61
2
5
99
6
78
732
C2
C1
465
571
87
3
69
3
33
6
50 703
542
452
m/z
1
00
200
300
400
500
600
700
800
Fig. 3 MS spectra of the complexes
Table 1 Error analysis for oligomerization reactions with catalyst C1
C1
Run 1
Run 2
Run 3
Average
SD
reported by Malgas-Enus et al. [15]. The MS of complexes
C1–C3 each showed a singly charged molecular ion peak,
at m/z 703, m/z 732 and m/z 760, respectively.
a
Activity
8.17
54.45
19.00
19.48
7.07
8.13
55.04
18.95
19.40
6.61
8.20
55.13
18.94
19.38
6.55
8.17
54.87
18.96
19.42
6.74
0.03
0.30
0.03
0.04
0.23
C
C
C
C
4
6
To examine the thermal stabilities of the complexes,
thermogravimetric analysis was carried out under a nitro-
gen atmosphere at a heating rate of 10 °C/min from
ambient temperature to 1200 °C. The TGA curve for
complex C3 is shown in Fig. 4. The thermal decomposition
process of this complex can be subdivided into three stages
8
8?
Reaction condition: [Fe] = 5 lmol; toluene = 50 mL; time = 30 -
min; Al/Fe = 500; T = 25 °C; ethylene pressure = 0.5 MPa
a
4
0 g/(mol Fe h)
1
1
23

Transit Met Chem
polymerization kinetic profile, polymer molecular weight
and stereoregularity in cationic transition metal-catalyzed
olefin polymerization processes [25]. In order to shed light
on the influence of cocatalysts on the catalytic behavior of
these bidentate iron complexes, we activated complex C3
the poor solubility of C3 in n-hexane and cyclohexane. It is
interesting to note that butenes (62.64%) were predominant
products for the system in toluene solvent. However,
longer-chain oligomers (53.01%) and hexenes (73.74%)
were the major products in cyclohexane and n-hexane
solvents, respectively. This trend is consistent with solvent-
induced variations in catalytic activities, since higher cat-
alytic activities are more likely to promote chain termina-
tion over chain propagation [27].
using MAO and Et AlCl in toluene solvent. Since the
2
function of the cocatalyst is to promote formation of the
active species in the presence of impurities such as mois-
ture and oxygen, an appropriate amount of cocatalyst is
required. It was clear that the choice of aluminum cocat-
alyst had a significant effect on both the catalytic activity
and product distribution (Table 2, entries 2, 13). When
activated with MAO, the complex showed moderate
The preliminary study was extended to investigate the
effects of temperature, Al/Fe molar ratio and reaction
pressure. The reaction temperature can affect the activity,
since ethylene oligomerization is a highly exothermic
reaction, and furthermore, the solubility of ethylene and
stability of catalytic species are also affected by tempera-
ture. The pre-catalyst C3 showed slightly lower
oligomerization activity at lower temperature (Table 2,
entry 1) compared to the result obtained at 25 °C (Table 2,
entry 2). This result can be associated with reduced solu-
bility of C3 in toluene at 15 °C generating lower amounts
of active species. Increasing the reaction temperature from
25 to 45 °C led to a loss in the catalytic activity (Table 2,
entries 2–4), which might be due to decomposition of the
active species and/or lower ethylene solubility at higher
temperature. The contents of the oligomeric proportions
were randomly changed, implying that the reaction tem-
perature did not rationally control the rate of chain prop-
agation to b-hydrogen elimination.
4
activity of up to 7.46 9 10 g/(mol Fe h) and afforded
more C8? oligomers (26.58%). The use of Et AlCl induced
2
higher catalytic activity, but with less selectivity for C8?
oligomers (19.23%). Thus, further studies of these com-
plexes for ethylene oligomerization were carried out with
MAO as cocatalyst.
The choice of solvent can affect the catalytic activity
and product distribution, affecting both the monomer
concentration and the solubility of the catalyst itself [26].
With this in mind, the catalytic behavior of complex C3
was investigated in both cyclohexane and n-hexane and
compared with the corresponding results in toluene
(
Table 2, entries 2, 11–12). Under similar conditions, the
catalytic activities decreased in the order n-hexane \ cy-
clohexane \ toluene. These observations are attributed to
Table 2 Ethylene
oligomerization with catalyst
system C3
a
b
c
Entry
Al/Fe
T (°C)
P (MPa)
Activity
Oligomer distribution (%)
C4
C6
C8
C –C
10 18
1
2
3
4
5
6
7
8
9
500
500
500
500
200
700
1000
500
500
500
500
500
500
15
25
35
45
25
25
25
25
25
25
25
25
25
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.1
0.3
0.7
0.5
0.5
0.5
6.27
7.46
5.66
4.96
4.08
5.75
5.13
6.76
7.25
8.63
6.23
5.13
11.02
68.75
62.64
71.99
80.11
74.47
82.06
77.44
66.55
68.83
70.04
31.63
15.97
75.16
8.51
17.00
15.35
10.08
9.66
5.74
11.23
14.16
8.64
10.76
3.77
1.59
4.90
6.10
8.47
14.51
7.34
8.38
7.86
73.74
5.61
11.98
9.04
8.65
2.80
10.59
13.67
6.20
3.50
5.27
17.63
7.29
1
1
1
1
a
0
1
2
3
14.29
7.50
d
e
f
53.01
9.60
0.69
9.36
9.87
Reaction condition: [Fe] = 5 lmol; toluene = 50 mL; time = 30 min; MAO as cocatalyst
b
c
4
0 g/(mol Fe h)
1
Determined by GC
d
e
Cyclohexane as solvent (50 mL)
n-Hexane as solvent (50 mL)
f
2
Et AlCl as cocatalyst (500 equiv)
123

Transit Met Chem
The effects of the Al/Fe molar ratio on the system were
also investigated using C3. Increasing the Al/Fe molar ratio
in the range 200–1000 resulted in an initial increase and
then a gradual decrease (Table 2, entries 2, 5–7), with the
longer-chain oligomers (Table 3, entries 1, 4). Similar
results have been previously reported for nickel and iron
complexes of 2-(2-pyridyl) quinoxaline in ethylene
oligomerization reactions [30] and can be attributed to
electronic factors [31].
4
highest activity of 7.46 9 10 g/(mol Fe h) observed at an
Al/Fe molar ratio of 500 at 25 °C. Higher Al/Fe molar
ratios resulted in decreased catalytic activity, possibly due
to high accumulation of alkylaluminum impurities which
might cause catalyst deactivation [28]. Generally higher
We also noted that varying the alkyl substituent on the
ligand backbone from a tetradecyl to an octadecyl group
led to noticeable decreases in activities in all the catalytic
systems (Table 3, entries 1–3). For example, replacing the
tetradecyl group (C1) with a hexadecyl group (C2) resulted
Al/Fe molar ratios favored the formation of lower C oli-
4
4
4
gomers. For example, C proportions of 62.64 and 77.44%
4
in decreased activity from 8.17 9 10 to 7.98 9 10 g/
(mol Fe h), respectively, when using MAO as a cocatalyst
(Table 3, entries 1, 2). This may be due to bulkier ligands
hindering coordination of ethylene to the active metal
center. However, random data were observed for the dis-
tribution of oligomers produced by the different iron pre-
catalysts, suggesting that the degree of steric hindrance on
the iron pre-catalysts has little influence on product
distribution.
were observed at Al/Fe molar ratios of 500 and 1000,
respectively (Table 2, entries 2, 7). This trend could be
attributed to increased chain transfer to the cocatalyst or
greater chain termination due to increased catalytic activity
[
25].
The influence of ethylene concentration was studied by
varying ethylene pressure from 0.1 to 0.7 MPa using cat-
alyst C3 (Table 2, entries 2, 8–10). As expected, increasing
the ethylene pressure from 0.1 to 0.7 MPa led to increased
4
4
catalytic activity from 6.76 9 10 to 8.63 9 10 g/(mol Fe
h), respectively. The ethylene concentration also had a
significant effect on the product distribution. Increase in
ethylene pressure favored the formation of shorter-chain
oligomers (C and C ). Higher proportions of lower frac-
Mechanism of the catalytic oligomerization
Since late transition metal complexes exhibit very high
activities for ethylene polymerization and oligomerization,
researchers are keen to elucidate the catalytic mechanism.
However, due to limitations resulting from the paramag-
netism of such iron complexes, their active species are not
as well understood as are those of the a-diimine palladium
catalysts. Considering the results of this study and previous
reports, it is possible to deduce the mechanism of
oligomerization at the iron metal center (Fig. 6). On
4
6
tion oligomers at higher pressures are consistent with
increased catalytic activity leading to rapid chain termi-
nation [29].
Influence of catalyst structure on ethylene oligomerization
The influence of complex structure on both the catalytic
activity and product distribution was also studied. Under
the optimized conditions for the catalytic system employ-
ing C3, the bidentate nickel complex C4 (Fig. 5) and all
three bidentate iron complexes C1–C3 were investigated
for ethylene oligomerization. Most notable was the role of
the metal atom in regulating catalytic performance. Com-
pared with the nickel analogue C4, the iron complex C1
displayed lower catalytic activity and higher selectivity for
reaction with a cocatalyst (MAO or Et AlCl), the salicy-
2
laldimine iron complex is transformed into an active spe-
cies, often involving interactions between the metal and a
hydride (or alkyl) and ethylene molecule [32]. The iron
hydride species can be generated by a variety of standard
organometallic reactions including chain propagation and
Table 3 Ethylene oligomerization studies with C1–C4 in using
MAO as cocatalyst
a
b
c
Entry
Complexes Activity
Oligomer distribution (%)
C
4
C
6
C
8
C10–C18
1
2
3
4
a
C1
C2
C3
C4
8.17
7.98
54.45 19.00 19.48
7.07
35.54 25.26 25.82 13.08
62.64 10.76 15.35 11.23
7.46
13.50
52.10
9.30 32.63
5.97
Reaction
condition:
[M] = 5 lmol;
toluene = 50 mL;
ethylene
time = 30 min;
pressure = 0.5 MPa
Al/M = 500;
T = 25 °C;
b
4
0 g/(mol M h)
1
c
Fig. 5 Structure of the bidentate nickel complex C4
Determined by GC
1
23

Transit Met Chem
Fig. 6 Proposed mechanism of catalysis by the bidentate iron complexes in ethylene oligomerization
b-hydrogen elimination from an intermediate iron alkyl
and oxidative addition of a Lewis acid to a zero-valent iron
species [33, 34]. The type of cocatalyst can affect product
distribution. With MAO as cocatalyst, the better selectivity
for longer-chain oligomers might be due to the fact that the
chain propagation is faster than b-hydrogen elimination.
moderate, their superiority lies in their higher selectivities
for longer-chain oligomers. In future work, we hope to
obtain new oligomerization iron complexes based on
hyperbranched salicylaldimine ligands modulating struc-
tural and electronic characteristics by the nature of the
substituents on the salicylaldehyde moieties in order to
improve the catalytic activities.
Acknowledgements This work was supported by the National Nat-
ural Science Foundation of China (No. 21576048) and Petroleum
Innovation Foundation of China (No. 2014D-5006-0503) for the
financial support.
Conclusions
A series of bidentate iron complexes based on hyper-
branched salicylaldimine ligands were synthesized and
characterized. In the presence of either MAO or Et AlCl,
2
the bidentate iron complex C3 had high catalytic activities
for ethylene oligomerization. MAO was found to be a more
References
1
2
3
. Skupinska J (1991) Chem Rev 91:613–648
. Ziegler K, Gellert HG (1950) Liebigs Ann Chem 567:195–203
. Ziegler K, Holzkamp E, Martin H, Breil H (1955) Angew Chem
effective cocatalyst than Et AlCl. Among various solvents,
2
toluene was found to be most effective for these bidentate
iron complexes. The catalytic activity for ethylene
oligomerization and selectivity for oligomers could be
easily controlled by altering the reaction conditions such as
the temperature, Al/Fe molar ratio and reaction pressure.
The catalytic activity and selectivity were also significantly
influenced by the ligand structure and choice of metal. The
catalytic activities of these bidentate iron complexes
decreased with increasing alkyl chain length, and the alkyl
chain also had a great influence on the selectivity for
ethylene oligomerization. Compared with the nickel ana-
logue C4, the iron complex C1 displayed lower catalytic
activity and higher selectivity for longer-chain oligomers.
While the catalytic activities of these systems are
6
7:541–547
4. Small BL, Brookhart M, Bennett AMA (1998) J Am Chem Soc
20:4049–4050
5
6
1
. Small BL, Brookhart M (1998) J Am Chem Soc 120:7143–7144
. Britovsek GJP, Gibson VC, Kimberley BS, Maddox PJ,
McTavish SJ, Solan GA, White AJP, Williams DJ (1998) Chem
Commun 7:849–850
7
8
. Gibson VC, Redshaw C, Solan GA (2007) Chem Rev
1
07:1745–1776
. Fern a´ ndez I, Trovitch RJ, Lobkovsky E, Chirik PJ (2008) Orga-
nometallics 27:109–118
9. Alt HG, Licht EH, Licht AI, Schneider KJ (2006) Coord Chem
Rev 250:2–17
0. Simon LC, Mauler RS, De Souza RF (1999) J Polym Sci A:
Polym Chem 37:4633–4656
1
123

Transit Met Chem
1
1
1
1
1. Bianchini C, Giambastiani G, Luconi L, Meli A (2010) Coord
Chem Rev 254:431–455
2. Xiao LW, Gao R, Zhang M, Li Y, Cao XP, Sun WH (2009)
Organometallics 28:2225–2233
3. Sun WH, Hao P, Li G, Zhang S, Wang WQ, Yi JJ, Asma M, Tang
N (2007) J Organomet Chem 692:4506–4518
4. Xiao TPF, Zhang S, Kehr G, Hao X, Erker G, Sun WH (2011)
Organometallics 30:3658–3665
5. Malgas-Enus R, Mapolie SF (2014) Inorg Chim Acta 409:96–105
6. Eaves R, Parkin S, Ladipo FT (2007) Inorg Chem 46:9495–9502
7. Wang J, Zhang P, Chen S, Li CQ, Li HY, Yang G (2013) J Macro
Sci Part A 50:163–167
23. Li CQ, Yu HY, Lin ZY, Wang FF, Zhang N, Wang J (2017) J
Coord Chem. doi:10.1080/00958972.2017.1293822
24. Tas E, Kilic A, Konak N, Yilmaz I (2008) Polyhedron
27:1024–1032
25. Chen EY, Marks TJ (2000) Chem Rev 100:1391–1434
26. Chandran D, Kwak CH, Oh JM, Ahn IY, Ha CS, Kim I (2008)
Catal Lett 125:27–34
27. Doherty MD, Trudeau S, White PS, Morken JP, Brookhart M
(2007) Organometallics 26:1261–1269
28. Abbo HS, Titinchi SJ (2013) Molecules 18:4728–4738
29. Killian CM, Johnson LK, Brookhart M (1997) Organometallics
16:2005–2007
1
1
1
1
1
2
2
2
8. Wang J, Yang G, Li CQ, Shi WG (2014) Chem J Chin Univ
3
30. Shao CX, Sun WH, Li ZL, Hu YL, Han LQ (2002) Catal Com-
mun 3:405–410
5:1536–1540
9. O’Reilly RK, Gibson VC, White AJP, Williams DJ (2003) J Am
Chem Soc 125:8450–8451
31. Laine TV, Lappalainen K, Liimatta J, Aitola E, L o¨ fgren B,
Leskel a¨ M (1999) Macromol Rapid Commun 20:487–491
32. Jolly PW (1982) In: Wilkinson G, Stone FGA, Abel EW (eds)
Comprehensive organometallic chemistry, vol 8. Pergamon Press,
Oxford, p 615
33. Carlini C, Isola M, Liuzzo V, Galletti AMR, Sbrana G (2002)
Appl Catal A: Gen 231:307–320
34. Kim I, Kwak CH, Kim JS, Ha CS (2005) Appl Catal A: Gen
0. Zhang N, Wang SH, Song L, Li CQ, Wang J (2016) Inorg Chim
Acta 453:369–375
1. Shriver DF (1986) The manipulation of air-sensitive compounds.
McGraw-Hill, New York
2. Wang J, Li CQ, Zhang SY, Sun F, Ge TJ (2008) Chin Chem Lett
1
9:43–46
2
87:98–107
1
23