Anchoring via covalent binding of a salicylaldimine-nickel complex in multi-walled carbon nanotubes and its application in ethylene oligomerization

Article abstract of DOI:10.1080/00958972.2020.1811241

Silylated-salicylaldimine nickel(II) (1) was synthesized with 3-aminopropyl-trimethoxysilane, salicylaldimine and NiCl₂·6H₂O, and carbon nanotube-supported salicylaldimine-nickel catalyst (2) was synthesized by anchoring 1 covalently to multi-walled carbon nanotubes. The chemical structure of the synthesized ligand was characterized by elemental analysis, ¹H NMR and FT-IR. Compounds were characterized by XRD, FT-IR, TGA and XPS and were used as precatalysts for ethylene oligomerization. The catalytic activity and the selectivity of C₆ olefin for 2 were much higher than those for 1 in ethylene oligomerization. Compared with other cocatalysts, the catalytic activity and selectivity to C₆ olefin for 2 were the highest using MAO as cocatalyst. When Al/Ni molar ratio was 700, at 25 °C, 30 min and pressure was 0.5 MPa, the catalytic activity of 5.85 × 10⁶ g·(mol Ni·h)^?1 and selectivity of 94.7percent toward C₆ olefin for 2 in cyclohexane were observed. Compared with 1, catalyst 2 had the good repeatability and thermal stability, with catalytic activity of 4.29 × 10⁶ g·(mol Ni·h)^?1 and the selectivity of C₆ olefin was 84.3percent after the third oligomerization reaction.

Full text of DOI:10.1080/00958972.2020.1811241

Journal of Coordination Chemistry
ISSN: 0095-8972 (Print) 1029-0389 (Online) Journal homepage: https://www.tandfonline.com/loi/gcoo20
Anchoring via covalent binding of a
salicylaldimine-nickel complex in multi-walled
carbon nanotubes and its application in ethylene
oligomerization
Cuiqin Li, Xue Zhai, Ming Cheng, Yang Li, Nian Li, Feng Li & Jun Wang
To cite this article: Cuiqin Li, Xue Zhai, Ming Cheng, Yang Li, Nian Li, Feng Li & Jun Wang
(2020) Anchoring via covalent binding of a salicylaldimine-nickel complex in multi-walled carbon
nanotubes and its application in ethylene oligomerization, Journal of Coordination Chemistry, 73:13,
1937-1953, DOI: 10.1080/00958972.2020.1811241
To link to this article: https://doi.org/10.1080/00958972.2020.1811241
Published online: 10 Sep 2020.
Submit your article to this journal
Article views: 22
View related articles
View Crossmark data
Full Terms & Conditions of access and use can be found at
https://www.tandfonline.com/action/journalInformation?journalCode=gcoo20

JOURNAL OF COORDINATION CHEMISTRY
2020, VOL. 73, NO. 13, 1937–1953
https://doi.org/10.1080/00958972.2020.1811241
Anchoring via covalent binding of a salicylaldimine-nickel
complex in multi-walled carbon nanotubes and its
application in ethylene oligomerization
Cuiqin Lia, Xue Zhaia, Ming Chenga,b , Yang Lia, Nian Lia, Feng Lia and
Jun Wanga
aProvincial Key Laboratory of Oil & Gas Chemical Technology, College of Chemistry & Chemical
b
Engineering, Northeast Petroleum University, Daqing, Heilongjiang, China; PetroChina Fushun
Petrochemical Company Research Institute, Fushun, Liaoning, China
ABSTRACT
ARTICLE HISTORY
Received 31 July 2019
Accepted 30 July 2020
Silylated-salicylaldimine nickel(II) (1) was synthesized with 3-ami-
nopropyl-trimethoxysilane, salicylaldimine and NiCl2ꢀ6H2O, and
carbon nanotube-supported salicylaldimine-nickel catalyst (2) was
synthesized by anchoring 1 covalently to multi-walled carbon
nanotubes. The chemical structure of the synthesized ligand was
characterized by elemental analysis, 1H NMR and FT-IR.
Compounds were characterized by XRD, FT-IR, TGA and XPS and
were used as precatalysts for ethylene oligomerization. The cata-
lytic activity and the selectivity of C6 olefin for 2 were much
higher than those for 1 in ethylene oligomerization. Compared
with other cocatalysts, the catalytic activity and selectivity to C6
olefin for 2 were the highest using MAO as cocatalyst. When Al/
Ni molar ratio was 700, at 25 ꢁC, 30 min and pressure was
0.5 MPa, the catalytic activity of 5.85 ꢂ 106 gꢀ(mol Niꢀh)ꢃ1 and
selectivity of 94.7% toward C6 olefin for 2 in cyclohexane were
observed. Compared with 1, catalyst 2 had the good repeatability
and thermal stability, with catalytic activity of 4.29ꢂ 106 gꢀ(mol
Niꢀh)ꢃ1 and the selectivity of C6 olefin was 84.3% after the third
oligomerization reaction.
KEYWORDS
Carbon nanotube; nickel
complexes; covalent
modification; ethylene
oligomerization
CONTACT: Feng Li
lifeng_dqpi@sina.com
Provincial Key Laboratory of Oil & Gas Chemical Technology,
College of Chemistry & Chemical Engineering, Northeast Petroleum University, Daqing, Heilongjiang, 163318, China
ß 2020 Informa UK Limited, trading as Taylor & Francis Group

1938
C. LI ET AL.
1. Introduction
Nickel(II)-based catalyst systems, which have high catalytic activity and good tolerance
toward functional groups/polar monomers, have been a topic of significant interest in
ethylene polymerization and oligomerization [1–3]. Many studies demonstrated transi-
tion metal, organic ligands and cocatalysts play an important role in ethylene oligo-
merization [4, 5]. The structures of organic ligands and the type of cocatalysts have a
major effect on the selectivity of the desired olefins in oligo-/polymerization reactions.
Efforts have been devoted to studying the effects of the ligand structure of transition
metal complexes on the catalytic activity and the selectivity in oligo/polymerization
reactions [6–8]. In particular, salicylaldimine ligands have attracted attention because
the monoanionic ligand structures are easy to access and the o-containing aromatic
ring can be modeled by adding substituent groups with different steric hindrance [7,
9, 10]. Song et al. investigated steric hindrance of the ligand on the catalytic activities
for a series of bidentate salicylaldimine nickel-based catalysts in ethylene oligomeriza-
tion, and the results showed that an increase of the steric hindrance for the ligand
improved the catalytic activity and selectivity of oligomer [11]. Wang investigated the
alkyl chain length of the ligand backbone for hyperbranched salicylaldimine nickel
complexes on the catalytic activity and selectivity for longer-chain oligomers in ethyl-
ene oligomerization; the results showed that bidentate nickel complex with 1-octa-
þ
decyl as core in the ligand backbone produced mainly octenes (59.38%) and C10
olefins (29.01%) and exhibited high activities (2.23 ꢂ 105 g/(mol Niꢀh)) and good selec-
tivities for the longer-chain oligomers (36.35%–60.33%) [12].
However, the major drawback of the above homogeneous catalysts in ethylene
oligomerization is the difficulty in separating the catalysts, products and solvent.
Heterogeneous catalytic systems are promising methods for reducing the use of
organic solvents in easy separation of the reaction medium and easing the catalyst
recycling [13]. Several supports have been used for heterogenization of homogeneous
catalysts for ethylene oligomerization, including silica, alumina, silica–alumina, mesopo-
rous materials, carbon nanotubes (CNTs), MOFs, COFs [14–18]. Rossetto et al. synthe-
sized homogeneous catalysts with b-diimine-nickel complexes, SBA-15 and (3-
chloropropyl) trimethoxysilane as materials by covalent binding, showing activity of
7.2 ꢂ 103 hꢃ1 for ethylene oligomerization with 98% selectivity toward C4 products
[19]. Souza et al. grafted Ni(MeCN)6(BF4)2 onto [Si]-MCM-41 and [Si,Al]-MCM-41 and
investigated the catalytic activities of two homogeneous catalysts in ethylene oligo-
merization in the presence of AlEt3. The homogeneous catalysts exhibited high 1-
butene selectivity (63%) and high catalytic activity (3.20 ꢂ 103 hꢃ1). Resende prepared
silica–alumina catalysts with Ni nanoparticles and Ni-Al-SBA-15 as materials for ethyl-
ene oligomerization, and the catalyst had high ethylene conversion (85%) and C4
selectivity (74%) [20]. CNTs with high mechanical strength, high electrical and thermal
conductivity, and poor solubility in organic solvents have potential applications in
many fields. The chemical functionalization of CNTs with metal complexes will be
promising materials in ethylene polymerization [21, 22].
In our previous study, a series of homogeneous salicylaldimine nickel complexes were
synthesized and used as catalyst precursors for ethylene oligomerization; the nickel com-
plexes exhibited high activities for ethylene oligomerization and showed good selectivity

JOURNAL OF COORDINATION CHEMISTRY
1939
for longer-chain oligomers produced with MAO as cocatalysts [12, 23–25]. However, the
disadvantage of these homogeneous salicylaldimine nickel complexes is the difficulty in
separating the catalyst from the reaction products, enabling catalyst recycle. In this work,
we investigate the potential for developing a new heterogeneous catalyst via anchoring
with covalent bonds between the multi-walled CNTs with homogeneous salicylaldimine
nickel complex (1) and to study the effect of the support on the catalytic activity and the
selectivity of oligomers. Moreover, the repeatability and the thermal stability of the
recovered catalyst will also be investigated.
2. Experimental
2.1. Materials and general considerations
Multi-walled CNTs (>95%) (MWNTs) and 3-aminopropyl-trimethoxysilane were pur-
chased from Sigma-Aldrich (Beijing, China). NiCl2ꢀ6H2O and salicylaldimine were
obtained from Tianjin Guangfu Fine Chemical Research Institute (Tianjin, China).
Methylaluminoxane (MAO, 10 wt% in toluene), diethylaluminum chloride (DEAC,
25 wt% in toluene), ethyl aluminum dichloride (EADC, 25 wt% in n-hexane), triethyl alu-
minum (TEAL, 25 wt% in n-hexane) and triethyldialuminum trichloride (EASC, 50 wt%
in cyclohexane) were purchased from Sigma-Aldrich (China). All experiments were per-
formed under a nitrogen atmosphere using standard Schlenk techniques, and all sol-
vents were distilled by refluxing over sodium and benzophenone under nitrogen.
The chemical states of 1 and 2 were investigated with an X-ray photoelectron spec-
trometer (XPS, Thermofisher Scientific Company, K-Alpha). The binding energy was
calibrated using C1s ¼ 284.8 eV as a reference. The X-ray powder diffraction (XRD)
characterization was performed on a Rigaku D/Max-2200 diffractometer with Cu Ka
1
radiation (Rigaku, Japan). H NMR spectra were recorded in CDCl3 with tetramethylsi-
lane (TMS) as the internal standard on a Varian 400 MHz. SEM images were obtained
using a Zeiss scanning electron microscope (SIGMA) operated at an accelerating volt-
age of 5 kV. Elemental analyses were carried out in German Heraeus element analyzer.
IR spectra were recorded on a Nicolet FT-IR 750 infrared spectrometer using KBr pel-
lets. The nickel analysis of the complexes was calculated using an ICPS-7510 spectrom-
eter. The samples were prepared by treating 15 mg of the nickel complex with 2 mL of
HCl, 6 mL of HNO3 and 5 mL of HF, adding the mixture to polytetrafluoroethylene
autoclaves and subsequently using a digester for 24 h at 120 ꢁC. After cooling to room
temperature, the samples were diluted to 50 mL with deionized water. The thermogra-
vimetric (TGA) measurements were obtained in a STA 409 PC device from NETZSCH.
All samples (typically 5–7 mg) were heated under nitrogen flow (200 cm3/min) to
900 ꢁC at a rate of 10 ꢁC/min. Gas chromatography (GC) analyses were conducted on a
Fuli GC 9720 equipped with a flame ionization detector (FID) and a 50 m (0.2 mm i.d.,
0.5 lm film thickness) HP-PONA column.
2.2. Synthesis of silylated-salicylaldimine nickel(II) (1)
3-Aminopropyl-trimethoxysilane (2.76 g, 12.5 mmol) was dissolved in 50 mL dried etha-
nol under nitrogen, and salicylaldimine (1.53 g, 15.0 mmol) was added to the mixture

1940
C. LI ET AL.
dropwise at 40 ꢁC. The reaction mixture was stirred at 80 ꢁC for 3 h. The solvent and
the unreacted salicylaldimine were removed by rotary evaporation to obtain the sily-
1
ated-salicylaldimine ligand (L) as a yellow liquid. Yield: 94%. H NMR (CDCl3, d ppm): d
8.44 (s, 1H), 7.33 (d, J ¼ 9.0 Hz, 2H), 6.85 (d, J ¼ 8.3 Hz, 2H), 3.85 (q, J ¼ 7.0 Hz, 7H), 3.64
(dd, J ¼ 11.6, 4.8 Hz, 3H), 3.33 (dt, J ¼ 3.3, 1.6 Hz, 1H), 1.90–1.78 (m, 2H), 1.22 (dt,
J ¼ 14.0, 7.0 Hz, 11H), 0.71 (dd, J ¼ 9.5, 7.1 Hz, 2H). Anal. Calcd for C16H27NO4Si (325): C
59.08, H 8.31, N 4.31. Found: C 58.87, H 8.15, N 4.46. Anal. by ICP (atomic %): Si 9.02.
Si/N ¼ 2.02 (Calcd 8.62).
L (0.46 g, 1.4 mmol) and nickel chloride hexahydrate (1.04 g, 1.7 mmol) were dis-
solved in dried ethanol (20 mL) under nitrogen at room temperature. The reaction mix-
ture was stirred at 80 ꢁC for 12 h. The solid was collected by filtration and washed
with diethyl ether (3 ꢂ 20 mL). The solid was dried under vacuum at 50 ꢁC to obtain
silylated-salicylaldimine nickel(II) complex (1). Yield: 53%. Anal. Calcd for
C32H52N2O8Si2Ni (707): C 54.34, H 7.36, N 3.96. Found: C 54.18, H 6.97, N 3.83. Anal. by
ICP (atomic %): Si 7.71, Ni 8.49. Ni/N ¼ 2.01, Si/N ¼ 2.53 (Calcd: Si 7.92, Ni 8.30).
2.3. Synthesis of CNT-supported salicylaldimine-nickel complex (2)
The pristine MWCNTs were pretreated with concentrated HNO3 at 140 ꢁC for 12 h to
remove the amorphous carbonaceous materials and metals, and then washed with
deionized water and THF, respectively. The solid was dried under vacuum at 60 ꢁC for
12 h, yielding MWCNT-COOH [26]. MWCNT-COOH (0.68 g, 0.75 mmol) and 1 were
added to 20 mL dried ethanol, and the mixture was stirred at 80 ꢁC for 24 h under
nitrogen. The solid was separated by vacuum filtration using 0.22 lm Millipore polycar-
bonate membrane filter and subsequently washed with hot dried ethanol to remove
unreacted 1 [27]. The resulting solid was dried under vacuum at 50 ꢁC for 6 h. FT-IR
(KBr, cmꢃ1): m ¼ 3430 cmꢃ1 (–OH), m ¼ 1690 cmꢃ1 (C ¼ N). Anal. by ICP (atomic %): Si
3.51, Ni 1.68.
2.4. General procedure for ethylene oligomerization
Ethylene oligomerization using 1 and 2 was evaluated by the method reported [12].
The reaction was carried out in a 250-mL stainless steel autoclave equipped with a
magnetically driven mechanical stirrer, a thermocouple and a pressure gauge. The
autoclave was dried in an oven at 100 ꢁC for 6 h and was flushed with ethylene gas
three times at room temperature. The reaction temperature was controlled using a
thermostatic bath. When the desired reaction temperature was reached, the required
amount of the prepared catalyst suspension and the cocatalyst were injected into the
reactor under ethylene. The reaction mixture was stirred for 30 min at a temperature,
and the mixture was cooled to 0 ꢁC. The reaction was quenched by the addition of
HCl in ethanol, and the organic phase was taken to investigate the distribution of the
products by GC. The catalytic activity (105gꢀ(mol Niꢀh)ꢃ1), which are defined as the
mass of olefin converted per mole of precatalyst over time (in hours), varies by 10%,
and the catalytic activities were determined by at least two independent experiments
performed under each condition.

JOURNAL OF COORDINATION CHEMISTRY
1941
Scheme 1. Synthetic routes of 1 and 2.
Recycle experiments were performed in a 250-mL stainless steel autoclave equipped
with a magnetically driven mechanical stirrer, a thermocouple, a sampling cannula and
constant supply of ethylene. The reactor was dried in an oven at 100 ꢁC and was
flushed with ethylene gas three times at room temperature. The autoclave was
charged with 50 mL cyclohexane under ambient ethylene atmosphere at 30 ꢁC.
Cocatalyst (MAO) and 2 suspended in cyclohexene were added in the amount needed
to obtain an aluminum-to-nickel molar ratio (Al/Ni) of 700. After 30 min, the mixture
was cooled and the products were removed from the reactor through the sampling
cannula and analyzed by GC. The spent catalyst was filtered from the oligomerization
reaction mixture using a Millipore polycarbonate filter with 0.22 lm pore size, washed
in cyclohexane at 25 ꢁC for 1 h and dried under vacuum at 50 ꢁC for 12 h. Then, the
recycle experiment was accomplished by adding 50 mL fresh cyclohexane and a cer-
tain amount of cocatalyst as the above described method.
3. Results and discussion
3.1. Synthesis and characterization
We have previously reported the synthesis of a series of dendritic salicylaldimine
ligands with symmetric structure and their transition metal complexes [23, 28].
Silyated-salicylaldimine ligand (L) and its nickel(II) complex (1) (Scheme 1) were synthe-
sized according to the previously reported method [29]. The CNT-supported salicylaldi-
mine-nickel complex (2) (Scheme 1) was synthesized according to the reported
method by Titinchi [27]. In order to make the amino group of 3-aminopropyl-trime-
thoxysilane completely form Schiff base in the reaction, salicylaldimine was used in
excess, leaving unreacted salicylaldimine in the reaction mixture. The unreacted salicy-
laldimine was removed by rotary evaporation at high temperature. The unreacted
nickel chloride hexahydrate was removed by washing the product with ethyl ether in
the synthesis of 1. The structure of the silyated-salicylaldimine ligand (L) was

1942
C. LI ET AL.
Figure 1. FT-IR spectra of L, 1 and 2.
confirmed by FT-IR, 1H NMR and elemental analysis, and the structures of 1 and 2
were characterized by FT-IR, XRD, XPS and ICP.
3.2. Infrared spectra
The FT-IR spectra of L (Figure 1) indicated the ligand synthesized as the amidine band
at 3250 cmꢃ1 disappeared and the V(C ¼ N) stretch appeared at 1620–1640 cmꢃ1
.
Broad bands at 3400–3300 cmꢃ1 were due to V(O-H) of the salicylaldimine units.
Compared with spectra of L, the FT-IR spectra of the nickel complexes (Figure 1)
showed the absorption of O-H at 3400–3300 cmꢃ1 became weak and absorptions for
C-O appeared at 1320 cmꢃ1, shifted from high frequency to low frequency due to the
coordination of the ligand to the metal via both nitrogen and oxygen donors. IR spec-
tra of 2 showed a broad band at 3590 cmꢃ1, assigned to O-H stretches of MWCNTs
groups, and a peak at 1730 cmꢃ1 was assigned to C ¼ O stretch of MWCNTs groups,
which indicated that the COOH groups of MWCNTs were unreacted with 1.
1
3.3. H NMR spectra
1
The H NMR spectra showed the proton resonance of –CH¼N at 8.35 ppm. The aro-
matic protons of L showed asymmetric resonances at 6.85 and 7.92 ppm due to
–CH ¼ N and –OH attached to aryl ring. The elemental analysis and detailed analysis of
1H NMR spectra of L (Figure 2) are summarized in the synthesis section, in agreement
with FT-IR data to confirm the structure of L.
3.4. XRD analysis
Figure 3 shows the XRD patterns of 1, MWCNT-COOH and 2. For MWCNT-COOH, two
peaks appeared at 25ꢁ and 43ꢁ, which corresponded to the (002) and (100) planes of
graphite, respectively [30]. Comparing with MWCNT-COOH, two peaks appeared at

JOURNAL OF COORDINATION CHEMISTRY
1943
Figure 2. 1H NMR spectra of L.
Figure 3. XRD patterns for 1, MWCNT-COOH and 2.
around 25ꢁ and 43ꢁ in the XRD patterns of 2, which indicated that 1 treatments of
MWCNT-COOH do not destroy MWCNTs crystal structure [31].
3.5. XPS analysis
To study the coordination of Ni2þ to the frameworks, X-ray electron spectra (XPS)
were performed with 1 and 2. Figure 4 shows XPS spectra for 1 and 2. The first peak
at 531.9 eV (Figure 4(a,b)) was ascribed to oxygen single bonded to carbon (C-O), and

1944
C. LI ET AL.
Figure 4. XPS analyses for 1 (a) and 2 (b); N1s XPS analyses for 1 (c) and 2 (d); Ni2p XPS analyses
for 1 (e) and 2 (f); Si2p XPS analyses for 1 (g) and 2 (h).
the second peak at 283.0 eV (Figure 4(a,b)) was assigned to carbon–carbon single
bond (C-C) on the 3-aminopropyl-trimethoxysilane structure or the nanotube structure
[32]. The N1s peaks were observed at 397.8 and 399.5 eV for 1 and 2, respectively,
which indicated that the main peak of N1s line with binding energy was clearly shifted

JOURNAL OF COORDINATION CHEMISTRY
1945
Figure 4. Continued.
to a higher binding energy with formation of 2. Additional peak of N1s line of 2 with
higher binding energy (406.4 eV) appeared which might be due to breaking of the
weak coordination bond between Ni and N at high temperature.
In order to reveal the chemical state of Ni in 1 and 2, high-resolution spectra of
N1s and Ni2p line were analyzed. Figure 4e,f shows Ni2p spectra for 1 and 2. The
binding energies of the main Ni2p3/2 peak were observed at 854.5 and 854.8 eV for 1
and 2, respectively [18]. The Si2p peak for 1 was observed at 101.2 eV (Figure 4(g)),
which represents the bond of silicon with oxygen from 3-aminopropyl-trimethoxysi-
lane. The Si2p peak for 2 was observed at 102.1 eV (Figure 4(h)), which represents the
bond of silicon with oxygen originating from the MWCNTs(Si-O-CNT) [33].
3.6. SEM analysis
Scanning electron microscopy (SEM) was performed to evaluate the morphology of the
materials before and after anchoring of the complex. The micrograph corresponding to
MWCNT-COOH (Figure 5(a)) showed the thick and dispersive morphology of MWCNT-
COOH dried under vacuum at 80 ꢁC [32]. The micrograph corresponding to 2 (Figure 5(b))
showed the thin and entangled morphology of the MWCNT anchoring 1, indicating that 1
might be supported on single CNTs or different CNTs. To study the change of morphology
of 2 before and after catalyzing ethylene oligomerization, SEM was performed for the first
recycle catalyst 2 (Figure 5(c)) and the third recycle catalyst 2 (Figure 5(d)). After the first
oligomerization, there was little change in the frit size of 2 and the dispersive morphology
was observed for the first recycle catalyst 2 (Figure 5(c)), which was attributed to the disen-
tanglement of the MWCNT anchoring 1 because of the dispersion of solvents and the
ethylene oligomerization. After the third recycle of ethylene oligomerization, the morph-
ology of the catalyst 2 became more dispersive (Figure 5(d)) and there was little change in
the frit size of catalyst 2, compared with the first recycle catalyst 2.
3.7. TGA analysis
TGA and differential thermal analysis (DTA) curves of 1 and 2 under N2 from 30 to
900 ꢁC are shown in Figure 6. For 1, approximately 4.34% weight loss observed from

1946
C. LI ET AL.
Figure 5. SEM images of MWCNT-COOH (a), 2 (b), the first recycle of catalyst 2 (c) and the third
recycle of catalyst 2 (d).
Figure 6. TGA /DTA curves of 1 and 2.
30 to 120 ꢁC corresponded to water loss. The second weight loss (31.56 wt%) observed
from 130 to 473 ꢁC was attributed to decomposition of L. The third weight loss
(15.26 wt%) observed at 473–900 ꢁC corresponded to the step-by-step decomposition
of benzene and reduction of nickel chloride by carbon produced in the decomposition
of the ligand. The residual mass (48.84 wt%) of 1 remains almost constant until 900 ꢁC,
which corresponds to carbon and silicon produced in the decomposition of the ligand
and nickel reduced by carbon.

JOURNAL OF COORDINATION CHEMISTRY
1947
Table 1. Effect of the type of co-catalysts on the catalytic activity and the product distributiona.
Oligomerization distribution (%)b
Activity 105gꢀ
(mol Niꢀh)–1
C4
C6
C8
C8
þ
Precatalyst
Cocatalyst
Solvent
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
MAO
DEAC
TEAL
EADC
EASC
MAO
DEAC
TEAL
EADC
EASC
MAO
DEAC
TEAL
EADC
EASC
Cyclohexane
Cyclohexane
Cyclohexane
Cyclohexane
Cyclohexane
Cyclohexane
Cyclohexane
Cyclohexane
Cyclohexane
Cyclohexane
Toluene
5.70
2.70
6.00
4.40
2.40
38.60
23.40
5.27
14.60
7.02
15.80
15.80
11.00
64.40
25.80
21.02
28.87
38.61
40.03
32.69
10.33
25.92
62.79
11.95
21.58
40.31
55.80
11.71
63.40
60.94
47.85
36.81
22.92
26.52
38.36
84.73
35.64
20.32
82.59
17.66
19.04
11.81
81.43
20.88
12.18
22.72
6.44
19.16
25.32
3.03
1.72
3.32
0.59
1.58
2.61
9.89
8.07
2.08
7.09
12.44
8.41
27.88
19.31
8.13
25.92
3.22
35.12
16.30
3.88
58.15
30.76
24.32
4.78
Toluene
Toluene
Toluene
Toluene
8.63
14.44
aConditions: 2 mmol precatalyst, 50 mL solvent, 500 Al/Ni, 25 ꢁC, 0.5 MPa. bDetermined by GC.
For 2, approximately 3.83% of weight loss from 30 to 120 ꢁC corresponded to water
loss [16]. The second weight loss (9.63 wt%) observed at 130–324 ꢁC was attributed to
decomposition of L. The third weight loss (16.56 wt%) was observed from 324 to
900 ꢁC, corresponding to step-by-step decomposition of benzene and reduction of
nickel chloride by carbon produced in decomposition of L. The residual mass
(69.98 wt%) of 2 remained almost constant until 900 ꢁC, which corresponded to carbon
and silicon produced in the decomposition of L, nickel reduced by carbon and carbon
of CNT; more remained than for 1.
3.8. The effect of cocatalyst on ethylene oligomerization
Many reports showed that cocatalyst composition had an impact on catalyst behavior
in ethylene oligomerization [34, 35]. Our report on bidentate nickel complexes bearing
hyperbranched salicylaldimine ligands also showed that oligomerization and
Friedel–Crafts alkylation reactions occurred with EADC as cocatalyst in toluene and the
oligomerization reaction only occurred with MAO as cocatalyst [12]. In the present
work, two new nickel complexes were used as precatalysts for ethylene reaction with
various cocatalysts to investigate the effect of the cocatalysts on the catalytic activity
of the new nickel complexes at an Al/Ni molar ratio of 500, 0.5 MPa ethylene pressure
and reaction time of 30 min (Table 1).
For 1, the cocatalyst has little effect on the catalytic activity in ethylene oligomer-
ization in cyclohexane and has a certain effect on selectivity. The main products were
C4, C6 and C8 þolefins with MAO and EADC as cocatalysts, and the main products were
C4, C6 and C8 olefins with DASC and DEAC as cocatalysts. However, there was wide
product distribution with TEAL as cocatalyst. Similar results were also observed in our
previous work [12].
However, the cocatalyst has a significant impact on the catalytic activity and the
product distribution in ethylene oligomerization in cyclohexane for 2. The order of
catalytic activities for the five cocatalysts with 2 as precatalyst in cyclohexane was
MAO > DEAC > EADC > EASC > TEAL. Using MAO as cocatalyst, 2 had the highest

1948
C. LI ET AL.
catalytic activity and the highest selectivity for C6 olefin in ethylene oligomerization in
cyclohexane, and the catalytic activity was the lowest and the selectivity for C4 olefin
was the highest using TEAL as cocatalyst. Compared with three cocatalysts, the add-
þ
ition of DEAC and EASC was formed the most C8 olefins. The results were different
from those of the homogeneous 1. A possible reason could be that the molecular size
of MAO is bigger than other cocatalysts, making it difficult for MAO molecules to
access inside CNTs and activating 2 loaded outside the CNTs to increase catalytic
activity. Using toluene as solvent, the order of catalytic activities for the five cocata-
lysts with 2 as precatalyst was EADC > EASC > MAO (DEAC) > TEAL. The cocatalysts
have a significant impact on the product distribution in ethylene oligomerization, and
the selectivity of C6 olefin was 81.34% for 2 in toluene with TþEAL as cocatalyst. Upon
activation with MAO, the main products were C4, C6 and C8 olefins. However, the
other three cocatalysts had little effect on the selectivity of 2 in ethylene
oligomerization.
Solvents also had a dramatic influence on catalytic activity and selectivity of 2 with
different cocatalysts. Upon activation with MAO and DEAC, the catalytic activity and
selectivity to C6-olefin in toluene were considerably lower than those of 2 in cyclohex-
ane. However, using TEAL, EADC and EASC as cocatalysts, the catalytic activities of 2
in toluene were higher than those of 2 in cyclohexane. Particularly with EADC as coca-
talyst, the catalytic activity of 2 in toluene was about five times as much as that of 2
in cyclohexane. When the solvent was changed from cyclohexane to toluene, decrease
of selectivity for C4 olefin and increase of selectivity for C6 olefin with TEAL as cocata-
lyst were observed. Increase of selectivity for C4 olefin with EADC and EASC as cocata-
lysts was also observed. There are two possible reasons: (1) CNTs have poor solubility
in organic solvents [16]. Chemical modifications for CNTs can improve the solubility of
CNTs in organic solvents; the solubility of functionalized CNTs is affected by the func-
tional groups of the CNTs surface and by the polarity of the solvent. Catalyst 2 has
higher solubility in toluene because of the higher polarity of toluene, which may cause
higher catalytic activity; (2) catalytic activity of 2 is also affected by the cocatalyst used
in ethylene oligomerization. The higher Lewis acidity of the cocatalyst results in higher
insertion rate and growth rate, which causes higher catalytic activity and higher select-
ivity of high carbon olefins [36]. The order of Lewis acidity for the five cocatalysts was
EADC > EASC > DEAC > TEAL (MAO). However, the solubility of cocatalyst in solvents
also affects the catalytic activity and selectivity of oligomer for 2. Higher solubility of
cocatalyst led to better activation of the cocatalyst on the precatalyst. The solubility of
the five cocatalysts in toluene was better than that of the five cocatalysts in cyclohex-
ane. The effects of the solvent and cocatalysts on the catalytic activity and the select-
ivity of 2 will be studied in our future work.
3.9. The effect of reaction conditions on ethylene oligomerization
Apart from the catalyst structure and the reaction system, the catalytic activity and
product selectivity for ethylene oligomerization were also influenced by reaction con-
ditions, such as catalyst mass, Al/Ni molar ratio, ethylene pressure, reaction

JOURNAL OF COORDINATION CHEMISTRY
1949
Table 2. Effect of Al/Ni on the catalytic activity and the product distributiona.
Oligomerization distribution (%)b
Activity 105gꢀ
Precatalyst
Al/Ni
(mol Niꢀh)–1
C4
C6
C8
C8
þ
1
1
1
1
2
2
2
2
300
500
700
900
300
500
700
1000
3.91
5.70
8.64
12.32
14.00
38.60
58.50
76.10
27.98
21.02
18.02
15.57
41.01
10.33
2.60
38.07
47.85
57.87
63.76
26.72
84.73
94.65
92.37
10.00
22.72
19.72
16.84
3.08
1.72
0.38
23.95
8.41
4.39
3.83
29.19
3.22
2.37
0.33
6.66
0.64
aConditions: 2 mmol precatalyst, 50 mL cyclohexane, 30 min, 0.5 MPa, 25 ꢁC. bDetermined by GC.
temperature and reaction time. Catalysts 1 and 2 were investigated under different
reaction conditions to improve the catalytic performance.
3.9.1. Effect of Al/Ni
Generally, a higher Al/Ni ratio was found to be favorable for activity of 1 and 2 (Table
2), indicating that excess MAO was needed for activating 1 and 2 to form the active
species. A possible reason could be that increasing MAO concentration in the oligo-
merization system might increase the coordination strength of the ion pairs of 1 and
2, resulting in increase of catalytic activity [34, 37]. However, increasing MAO concen-
tration might increase the insertion rate of ethylene to the active species and the rate
of chain transfer to aluminum, which led to increase of the short carbon olefins (C4
and C6) and decrease of the long carbon olefin (C8þ). The selectivity of C4 and C6 ole-
fins increased with increasing Al/Ni ratio for catalysts 1 and 2 (Table 2). When the Al/
Ni ratio was 700, the catalytic activities of 1 and 2 were 8.6 ꢂ 105 gꢀ(mol Niꢀh)ꢃ1 and
5.85 ꢂ 106 gꢀ(mol Niꢀh)ꢃ1, respectively, and the contents of C6 olefin were 57.9% and
94.7%, respectively; 1 and 2 display high catalytic activity and high selectivity of C6
olefin at the Al/Ni ratio of 700 with MAO as cocatalyst. The activity of the heteroge-
neous 2 formed by 1 supported on the multi-walled CNTs exceeded the activity of 1
at the Al/Ni molar ratio of 300–1000. It is obvious that the reason for such significant
differences in catalytic activity and selectivity of olefin was that 2 had good dispersion
in the catalytic system and 1 was easy to form aggregates.
3.9.2. Effect of reaction pressure
For 1 and 2, the catalytic activity increases with increasing ethylene pressure (Table 3).
The ethylene pressure may increase ethylene concentration in the solvent, which in
turn enhances the insertion rate of ethylene and thus improves the catalytic activity
[38]. In the case of 1, the selectivity of C4 and C6 olefin increases with increasing ethyl-
ene pressure. However, for 2, an increase of ethylene pressure caused increase of the
selectivity of C4 olefin, and the selectivity of C6 olefin was 94.7% at ethylene pressure
of 0.5 MPa. Generally, an increase of ethylene pressure caused increase of the insertion
rate of ethylene to the active species for many homogeneous catalysts, which led to
increase of short carbon olefin (C4 and C6 olefins). Similar results of bidentate nickel
complexes bearing hyperbranched salicylaldimine ligands in ethylene oligomerization
using MAO as a cocatalyst were observed in our previous work [12]. However, for the
heterogeneous catalyst, the selectivity of oligomers was affected not only by the

1950
C. LI ET AL.
Table 3. Effect of reaction pressure on the catalytic activity and the product distributiona.
Oligomerization distribution (%)b
Pressure
(MPa)
Activity 105gꢀ
(mol Niꢀh)–1
C4
C6
C8
C8
þ
Precatalyst
1
1
1
1
2
2
2
2
0.1
0.5
0.7
1.0
0.1
0.5
0.7
1.0
3.40
8.64
12.15
18.02
23.10
25.86
1.86
2.60
9.83
47.44
57.87
62.84
72.31
92.47
94.65
86.85
85.76
20.59
19.72
10.07
1.07
5.52
0.38
19.82
4.39
3.99
0.76
0.15
2.37
3.19
0.86
10.00
13.76
58.50
58.50
61.26
64.40
0.13
1.38
12.00
aConditions: 2 mmol precatalyst, 700 Al/Ni, 50 mL cyclohexane, 30 min, 25 ꢁC. bDetermined by GC.
reaction parameters but also by the structure and property of the support. For 2
based on MWCNT, the porous and nano-tubular structure affects the selectivity of ole-
fin oligomerization toward the formation of C6 olefin. An increase of the ethylene
pressure caused increase of the insertion rate of ethylene, which resulted in more C4
and C6 olefins at the ethylene pressure of 0.1–0.5 MPa. A further increase of the ethyl-
ene pressure reduced the selectivity of C6 olefin for 2.
In addition, the activity and the C6 selectivity of the heterogeneous 2 formed by 1
supported on the multi-walled CNTs were higher than those of homogeneous 1 at
ethylene pressure of 0.1–1.0 MPa (Table 1). When the ethylene pressure was 0.5 MPa,
the catalytic activity was 5.85 ꢂ 106 gꢀ(mol Niꢀh)ꢃ1 and the C6 selectivity was 94.7% for
2, while the catalytic activity of 1 and the C6 selectivity were 8.6 ꢂ 105 gꢀ(mol Niꢀh)ꢃ1
and 57.9%, respectively. The results indicated the nickel complexes heterogenized on
carbon nanotubes via anchoring with covalent bonds showed higher catalytic activity
and the higher selectivity of C6 olefin compared with the corresponding homogeneous
catalyst.
3.9.3. Effect of reaction temperature on ethylene oligomerization
Generally, high reaction temperature is favorable for the catalytic activity. Catalysts 1
and 2 were tested at the reaction temperature from 15 to 45 ꢁC, and the results are
listed in Table 4. For 1 and 2, the activity increased with increasing temperature, and
the two catalysts exhibited higher activity at 35 ꢁC. However, a further increase in the
reaction temperature reduced the catalytic activity. This could be due to lower ethyl-
ene solubility at higher temperature. Moreover, the oligomerization reactions were
exothermic, and increasing the temperature led to decrease of the insertion rate of
ethylene [39]. The reaction temperature had a considerable impact on the selectivity
ofþthe two catalysts, and increasing temperature diminished the selectivities of C8 and
C8 olefins and enhanced the selectivities of C4 and C6 olefins at 15–45 ꢁC for 1. For 2,
þ
the selectivity of C8 and C8 olefins slightly increased with increasing temperature.
However, compared with 25 ꢁC, the selectivity of C6 olefin was lower. The results of
Table 4 showed that the activity and the C6 selectivity of 2 were higher than those of
1 at 15–45 ꢁC.
3.9.4. Effect of reaction time on ethylene oligomerization
To examine catalytic stability of 2 in ethylene oligomerization, the effect of reaction
time on the catalytic performance was investigated at the ethylene pressure of

JOURNAL OF COORDINATION CHEMISTRY
1951
Table 4. Effect of reaction temperature on the catalytic activity and the product distributiona.
Oligomerization distribution (%)b
Temperature
(ꢁC)
Activity 105gꢀ
(mol Niꢀh)–1
C4
C6
C8
C8
þ
Precatalyst
1
1
1
1
2
2
2
2
15
25
35
45
15
25
35
45
2.61
8.64
6.67
10.33
18.02
21.29
25.09
23.85
2.60
43.52
57.87
59.26
57.9
73.43
94.65
90.44
87.09
24.3
21.85
4.39
7.94
5.73
2.09
2.37
1.21
3.59
19.72
11.51
11.28
0.63
0.38
2.54
5.33
54.71
58.50
67.19
61.44
5.81
7.40
1.92
aConditions: 2 mmol precatalyst, 50 mL cyclohexane, 700 Al/Ni, 30 min, 0.5 MPa. bDetermined by GC.
Table 5. Effect of reaction time on the catalytic activity and the product distributiona.
Oligomerization distribution (%)b
Time
(min)
Activity 105gꢀ
Precatalyst
(mol Niꢀh)–1
C4
C6
C8
C8
þ
1
1
1
1
2
2
2
2
10
30
60
120
10
30
12.76
8.64
5.19
26.65
18.02
16.37
12.54
32.33
2.60
60.35
57.87
59.16
62.69
61.27
94.65
90.97
87.86
12.37
19.72
17.41
14.77
4.18
0.38
1.23
0.63
4.39
7.06
10.00
2.22
2.37
1.64
2.72
3.52
61.41
58.50
52.67
47.15
60
120
6.16
4.69
4.73
aConditions: 2 mmol precatalyst, 50 mL cyclohexane, 700 Al/Ni, 25 ꢁC, 0.5 MPa. bDetermined by GC.
0.5 MPa; the results are listed in Table 5. The catalytic activities of 1 and 2 and the
selectivity of C6 olefin increased with increasing reaction time from 10 to 60 min; fur-
ther increasing the reaction time led to decrease in selectivity of C6 olefin. However,
the catalytic activity and the selectivity of C6 olefin for 2 were far higher than those of
1 from 10 to 120 min and then decreased slightly with increasing of the reaction time
for 2, which indicated that the catalytic stability of heterogeneous 2 was better than
homogeneous 1.
3.10. Recycle reaction of 2 in ethylene oligomerization
In order to examine the reusability of 2, recycle reactions were performed, and the
results are listed in Table 6 (R1–R3). Compared with 2 before recovery, the Ni content
and the catalytic activity of 2 after recovery were lower. After the first recycle reaction,
the Ni content and the catalytic activity of 2 were 1.60% and 5.32 ꢂ 106 gꢀ(mol
Niꢀh)ꢃ1, respectively, attributed to the dissolution of free 1 in the oligomerization reac-
tion system. Less reduction in the recovery and the Ni content of 2 with increasing
number of recycle reactions were observed (Table 6). With increase of the recycle
times, the catalytic activity and the selectivity for C6 olefin slightly decreased, and the
selectivity of C4 olefin slightly increased. Moreover, the selectivity of 1-olefins had little
change with an increase of the recycle times. This indicated that 1 and 2 were active
and selective in the recycle reactions and is therefore promising for large-scale use.
The pore of the support gives the catalyst a stable structure, which makes the catalytic
activity stable and the recovery repeatability good.

1952
C. LI ET AL.
Table 6. Results of two recycle reactiona.
Ni content
Oligomerization distribution (%)b
of 2
(%)
Activity
105gꢀ(mol Niꢀh)–1
C4(1-C4)
C6(1-C6)
C8(1-C8)
C8þ(1-C8
)
þ
Recycle
Recovery (%)
R0
R1
R2
R3
–
1.68
1.60
1.56
1.52
58.5
53.2
47.6
42.9
2.6(98.5)
9.2(98.6)
14.2(98.2)
12.4(97.6)
94.7(95.7)
88.0(92.3)
83.7(91.6)
84.3(90.3)
0.4(99.8)
1.3(99.1)
1.1(98.9)
2.2(98.2)
2.4(92.5)
1.5(88.4)
1.0(88.0)
1.1(87.6)
85.7
84.9
84.2
aConditions: 2 mmol complex, 50 mL cyclohexane, 30 min, 700 Al/Ni, 0.5 MPa, 25 ꢁC. bDetermined by GC.
4. Conclusion
Silylated-salicylaldimine nickel(II) homogeneous catalyst 1 and heterogeneous catalyst
2 were obtained and characterized by spectroscopic and analytical methods. The cata-
lysts were used as precatalyst for ethylene oligomerization under activation with differ-
ent cocatalysts. The catalytic activity and the selectivity of 1 and 2 were affected not
only by the structure of the catalyst but also by the reaction system and the reaction
conditions. The catalytic activity and selectivity of C6 olefin for heterogeneous 2 were
much higher than those for homogeneous 1 in ethylene oligomerization. Upon activa-
tion with MAO, 1 and 2 have higher activity and selectivity of C6 olefin in cyclohexane.
The change of the selectivity could be attributed to the confinement effect of pore
channel of supported catalyst to increase stability of intermediate for C6 olefin. The
immobilization of the metal complexes on MWCNTs improved the catalytic perform-
ance. For heterogeneous 2, the addition of multi-walled CNT served as a support for
the complexes and the pore of the support increased the catalytic stability and high
catalytic activity. Other heterogeneous catalysts based on different supports and the
effect of different loading methods on the catalytic properties of the heterogeneous
catalysts will be studied in our further work.
Acknowledgment
The authors are grateful to State Key Lab of Inorganic Synthesis and Preparative Chemistry of
Jilin University for the characterization work.
Disclosure statement
No conflict of interest exits in the manuscript, and manuscript is approved by all authors for
publication.
Funding
This work was supported by the Heilongjiang Natural Science Foundation of China (E2018012)
for the financial support.
ORCID
Ming Cheng
http://orcid.org/0000-0001-7991-0258

JOURNAL OF COORDINATION CHEMISTRY
1953
References
[1] D. Guironnet, T. Friedberger, S. Mecking. Dalton Trans., 41, 8929 (2009).
[2] S. Ito, Y. Ota, K. Nozaki. Dalton Trans., 41, 13807 (2012).
[3] T. Friedberger, P. Wucher, S. Mecking. J. Am. Chem. Soc., 134, 1010 (2012).
[4] S. Wang, W.H. Sun, C. Redshaw. J. Organomet. Chem., 751, 717 (2014).
[5] M.R. Barton, Y. Zhang, J.D. Atwood. J. Coord. Chem., 55, 969 (2002).
[6] C. Bianchini, G. Giambastiani, L. Luconi, A. Meli. Coord. Chem. Rev., 254, 431 (2010).
[7] V.C. Gibson, S.K. Spitzmesser. Chem. Rev., 103, 283 (2003).
~
[8] E. Rossetto, M. Caovilla, D. Thiele, R.F. de Souza, K. Bernardo-Gusmao. Appl. Catal. A: Gen.,
454, 152 (2013).
[9] D.S. McGuinness. Chem. Rev., 111, 2321 (2011).
[10] L. Bourget-Merle, M.F. Lappert, J.R. Severn. Chem. Rev., 102, 3031 (2002).
[11] L. Song, S. Song, J. Huang. Chin. J. Chem., 30, 1119 (2012).
[12] C.Q. Li, H.Y. Yu, Z.Y. Lin, F.F. Wang, N. Zhang, J. Wang. J. Coord. Chem., 70, 1303 (2017).
~
[13] E. Rossetto, B.P. Nicola, R.F. de Souza, K. Bernardo-Gusmao, S.B.C. Pergher. J. Catal., 323,
45 (2015).
ꢀ
[14] C. del Pozo, A. Corma, M. Iglesias, F. Sanchez. Organometallics, 29, 4491 (2010).
[15] A. Kermagoret, R.N. Kerber, M.P. Conley, E. Callens, P. Florian, D. Massiot, F. Delbecq, X.
ꢀ
Rozanska, C. Coperet, P. Sautet. J. Catal., 313, 46 (2014).
[16] S.M. Alshehri, T. Ahamad, A. Aldalbahi, N. Alhokbany. Adv. Polym. Technol., 35, 1 (2016).
[17] Z. Xu, D. Zhao, J.P. Chada, D.C. Rosenfeld, J.L. Rogers, I. Hermans, G.W. Huber. J. Catal.,
354, 213 (2017).
[18] E. Rozhko, A. Bavykina, D. Osadchii, M. Makkee, J. Gascon. J. Catal., 345, 270 (2017).
~
[19] E. Rossetto, B.P. Nicola, R.F. de Souza, S.B. Pergher, K. Bernardo-Gusmao. Appl. Catal. A:
Gen., 502, 221 (2015).
[20] R. D. Andrei, M. I. Popa, F. Fajula, V. Hulea. J. Catal., 323, 76 (2015).
[21] X. Dong, L. Wang, T. Sun, J. Zhou, Q. Yang. J. Mol. Catal. A: Chem., 255, 10 (2006).
[22] L. Zhang, E. Castillejos, P. Serp, W.H. Sun, J. Durand. Catal. Today, 235, 33 (2014).
[23] J. Wang, G. Yang, C.Q. Li, W.G. Shi, S.H. Wang. Chem. Pap., 68, 1532 (2014).
[24] J. Wang, N. Zhang, C.Q. Li, W.G. Shi, Z.Y. Li. J. Organomet. Chem., 822, 104 (2016).
[25] C.Q. Li, F.F. Wang, Z.Y. Lin, N. Zhang, J. Wang. Inorg. Chim. Acta, 453, 430 (2016).
[26] H. Kong, C. Gao, D. Yan. J. Am. Chem. Soc., 126, 412 (2004).
[27] S.J.J. Titinchi, H.S. Abbo. Catal. Today, 204, 114 (2013).
[28] J. Wang, P. Zhang, S. Chen, C.Q. Li, H.Y. Li, G. Yang. J. Macromol. Sci., Part A, 50, 163
(2013).
[29] J. Wang, Z.J. Fu, N. Zhang, C.Q. Li, W.G. Shi. J. Macromol. Sci., Part A, 54, 465 (2017).
[30] A. Ahmad, M.H. Razali, M. Mamat, F.S.B. Mehamod, K.A.M. Amin. Chemosphere, 168, 474
(2017).
[31] Z. Zhou, S. Wang, L. Lu, Y.X. Zhang, Y. Zhang. Compos. Sci. Technol., 68, 1727 (2008).
[32] H. Gaspar, C. Pereira, S.L.H. Rebelo, M.F.R. Pereira, J.L. Figueiredo, C. Freire. Carbon, 49,
3441 (2011).
[33] P.C. Ma, J.K. Kim, B.Z. Tang. Carbon, 44, 3232 (2006).
[34] D.S. McGuinness, A.J. Rucklidge, R.P. Tooze, A.M. Slawin. Organometallics, 26, 2561 (2007).
[35] F. Alam, L. Zhang, W. Wei, L. Wang, Y. Chen, C. Dong, T. Jiang. ACS Catal., 8, 10836
(2018).
[36] T. Agapie. Coord. Chem. Rev., 255, 861 (2011).
[37] A. Gollwitzer, T. Dietel, W.P. Kretschmer, R. Kempe. Nat. Commun., 8, 1226 (2017).
[38] T. Jiang, Y. Ning, B. Zhang, I. Li, G. Wang, J. Yi, Q. Huang. J. Mol. Catal. A: Chem., 259, 161
(2006).
[39] J.L. Rhinehart, L.A. Brown, B.K. Long. J. Am. Chem. Soc., 135, 16316 (2013).