Highly selective trimerization of ethylene with half-sandwich cyclopentadienyl and indenyl titanium complexes bearing pendant thienyl group

Article abstract of DOI:10.1016/j.molcata.2014.02.010

A series of half-sandwich titanium complexes bearing thienyl group [Cp(Ind)-bridge-thienyl]TiCl₃ (CS1-CS10) have been synthesized and show high selective ethylene trimerizaion to 1-hexene. The molecular structure of CS8 [Ind-C(cyclo-C₅H₁₀)-(5-Me-thienyl)]TiCl₃ was confirmed by X-ray. No intramolecular coordination interaction between the sulfur atom on the thienyl group and the titanium center could be observed in the solid state of these complexes. After activated with MAO, the complexes can effectively catalyze ethylene trimerization. For [Cp-bridge-thienyl]TiCl ₃/MAO system, the best productivity is obtained at 30 C; increasing the bulk of the substituent on the 5-position of the thienyl can improve the productivity for 1-hexene. CS6 [Cp-C(cyclo-C₅H₁₀)-(5- SiMe₃-thienyl)]TiCl₃ upon with MAO can make a productivity of 553 kg/(mol Ti-h) and 1-hexene selectivity of 86% at 30 C, 0.5 MPa ethylene pressure. For [Ind-bridge-thienyl] TiCl₃/MAO system, catalysts have more active and more tolerant of temperature comparing to corresponding Cp complexes. For example, CS8 can make the productivity of 697 kg/(mol Ti-h) and 1-hexene selectivity of 95% at 80 C, 0.5 MPa ethylene.

Full text of DOI:10.1016/j.molcata.2014.02.010

Journal of Molecular Catalysis A: Chemical 387 (2014) 20–30
Contents lists available at ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Highly selective trimerization of ethylene with half-sandwich
cyclopentadienyl and indenyl titanium complexes bearing pendant
thienyl group
Yanlu Zhang, Chen Wang, Tianzhi Wu, Haiyan, Ma. Jiling Huang∗
Laboratory of Organometallic Chemistry, East China University of Science and Technology, 130 Meilong Road, P.O. Box 310, Shanghai 200237, PR China
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of half-sandwich titanium complexes bearing thienyl group [Cp(Ind)-bridge-thienyl]TiCl3
(CS1–CS10) have been synthesized and show high selective ethylene trimerizaion to 1-hexene. The
molecular structure of CS8 [Ind-C(cyclo-C5H10)-(5-Me-thienyl)]TiCl3 was confirmed by X-ray. No
intramolecular coordination interaction between the sulfur atom on the thienyl group and the titanium
center could be observed in the solid state of these complexes. After activated with MAO, the complexes
can effectively catalyze ethylene trimerization. For [Cp-bridge-thienyl]TiCl3/MAO system, the best pro-
ductivity is obtained at 30 ◦C; increasing the bulk of the substituent on the 5-position of the thienyl can
improve the productivity for 1-hexene. CS6 [Cp-C(cyclo-C5H10)-(5-SiMe3-thienyl)]TiCl3 upon with MAO
can make a productivity of 553 kg/(mol Ti-h) and 1-hexene selectivity of 86% at 30 ◦C, 0.5 MPa ethylene
pressure. For [Ind-bridge-thienyl] TiCl3/MAO system, catalysts have more active and more tolerant of
temperature comparing to corresponding Cp complexes. For example, CS8 can make the productivity of
697 kg/(mol Ti-h) and 1-hexene selectivity of 95% at 80 ◦C, 0.5 MPa ethylene.
Received 12 October 2013
Received in revised form 1 February 2014
Accepted 8 February 2014
Available online 17 February 2014
Keyword:
Titanium
Thienyl
Cyclopentadienyl
Indenyl
Ethylene trimerization
© 2014 Elsevier B.V. All rights reserved.
1. Introduction
catalysts [16] and a series of self-activating catalysts from Gambat-
otta group [17–20].
Linear alpha olefins (LAOs) as ethylene oligomerization prod-
ucts have a variety of applications including as co-monomers
for the copolymerization with ethylene to generate linear low-
density polyethylenes (LLDPEs) and as versatile intermediates for
many industrially consumer products like plasticizers, lubricant,
detergents, and surfactants. Conventional full-range production
technology of ethylene oligomerization normally results in a wide
mixture distribution and don’t match the continually increasing
market demand. Thus, the transformation of a statistical ethy-
lene oligomerization process into selective tri- or tetramerization
appears highly desirable [1,2].
Up to now, several chemical elements including chromium, tita-
nium and tantalum as interesting catalysts have been reported for
selective trimerization or tetramerization of ethylene to ␣-olefins
[3–12]. Among all the catalysts systems, the central metal based
on chromium, is widely studied in recent years, such as Phillips’
Cr-pyrrolide catalysts [5], BP’s o-OMePNP catalysts [13], Sasol’s
PNP/SNS trimerization catalysts [14,15] and PNP tetramerization
Contrary to varies of Cr-catalysts, catalyst systems that are based
on titanium are relatively less. The arene-substituted cyclopen-
tadienyl (Cp) titanium catalysts firstly reported by Hessen and
co-workers in 2001 were the most attractive ethylene trimer-
ization catalysts (Chart 1, 1) [9,10]. Systematic research work
has been carried out on kinetic studies to optimize the reac-
tions and theoretical studies to give the mechanistic description
about this system [21,22]. The intrinsic high selectivity for ethy-
lene trimerization essentially depends on the hemilabile behavior
of the Cp-arene ancillary ligand. Other group IV metals: Zr and
Hf compounds (Chart 1, 2) with similar ligands were also tested,
but the result exhibit that the major product is polyethylene (PE)
[23]. Recently Kim has reported a kind of tetradentate trianionic
triethoxyamine titanatrate compound with the Cp-arene ligand
(Chart 1, 3), which can obtain 92.3 wt.% ethylene trimerization
selectivity and 1.03 × 104 kg/(mol Ti-h) 1-hexene productivity [24].
Suzuki also reported a series of new titanium complexes derived
form FI catalyst bearing phenoxy-imine ligand with pendant aryl-
OMe donors (Chart 1, 4). In the present of MAO, these complexes can
selectively transform ethylene to 1-hexene with unprecedented
high productivity (3.15 × 105 kg/((mol Ti-h); and more insight stud-
ies are just in progress [25].
∗
Corresponding author. Tel.: +86 21 64253519; fax: +86 21 64253519.
E-mail address: jlhuang@ecust.edu.cn (Ma.J. Huang).
http://dx.doi.org/10.1016/j.molcata.2014.02.010
1381-1169/© 2014 Elsevier B.V. All rights reserved.

Y. Zhang et al. / Journal of Molecular Catalysis A: Chemical 387 (2014) 20–30
21
Chart 1. Typical ethylene trimerization catalysts basted on group IV metals.
Following Hessen’s Cp-bridge-arene ligand system, Huang’s
group successfully developed a half-sandwich titanium complex
bearing pendent thienyl group (Chart 1, 5), which can serve effi-
ciently for ethylene trimerization [26]. But the complex with bridge
linked to the 3-position of the thienyl group (Chart 1, 6), is not effec-
tive for ethylene trimerization. The findings indicate that whether
the coordination of metal center with sulfur atom is an important
factor in the selective ethylene trimerization.
However, up to now no more detail information is dis-
cussed about the Cp-bridge-thienyl system. So it is necessary
to study the influence of different hemilabile coordination
situations on ethylene trimerization. For the system of [Cp-bridge-
thienyl]TiCl3/MAO, the elements that influence the coordination
include the bridge, which can determine the orientation of thienyl
group relative to the metal center, and the substituents on the
thienyl group especially at 5-position, which can give an direct
steric and electronic effect on the coordination between the sul-
fur and central metal. Besides them, regarding the diversity and
flexibility of indenyl ring, it is helpful to understand the differ-
ent steric and electronic effects on catalytic trimerization behavior
when replacing Cp ring with indenyl ring.
NMR was measured on a Brucker AVANCE-400 Hz spectrom-
eter using tetramethylsilane (TMS) as an internal standard and
employing CDCl3 as solvent dried over the standard drying agent
for a minimum of 24 h. Elemental analyses were performed on
an EA-1106 spectrometer. GC-MS was performed on 6890N-5975
instrument. GC was performed on Fuli 9790 instrument equipped
with Agilent J&W HP-5 GC column. IR spectra were recorded on a
Nicolet FTIR 5SXC spectrometer. Melting point were tested on XGW
x-4 micro melting point instrument.
2.2. Synthesis of complexes CS1–CS10
2.2.1. Synthesis of CS1: [ꢀ5-C5H4CMe2C4H3S]TiCl3
To a solution of 21 mmol of 2-thienyl lithium in 40 mL of diethyl
ether was added dropwise 2.2 g (21 mmol) of 6,6-dimethylfulvene.
The reaction mixture was stirred overnight. 2.25 g (21 mmol) of
trimethylsilyl chloride in 10 mL THF was added. The mixture was
stirred overnight, and then poured into 100 mL of ice-water. The
water layer was extracted with 30 mL of Et2O, after which the com-
bined organic layers were rinsed with 100 mL of brine. The organic
phase was dried over MgSO4. After the solvents were evaporated in
vacuum, the residue was distilled to give C5H4(SiMe3)CMe2C4H3S
(LS1) in yield of 42% (2.3 g). 80–95 ◦C/0.2 mmHg. 1H NMR (400 MHz,
CDCl3) ı 7.12 (m, 1H, thienyl-H), 6.88 (m, 1H, thienyl-H), 6.73–6.82
(m, 1H, thienyl-H), 6.40 (m, 2H, Cp-H), 6.11 (m, 1H, Cp-H), 3.25 (s,
1H, Cp-H), 1.64 (s, 6H, CH3), −0.03 (m, 9H, Si-CH3).
Herein, to give a comprehensive understanding about the [Cp-
bridge-thienyl] TiCl3/MAO system, a series of new half-sandwich
Cp titanium complexes with different bridges and different sub-
stituents on the thienyl group and also corresponding indenyl
titanium complexes are synthesized and used for ethylene trimer-
ization to further discuss the influence of different hemilabile
coordination conditions between sulfur atom and central metal on
ethylene trimerization.
To a solution of 0.50 mL of TiCl4(4.6 mmol) in 20 mL of CH2Cl2,
cooled to 0 ◦C, was added dropwise a solution of 1.20 g (4.6 mmol)
of C5H4(SiMe3)CMe2C4H3S (LS1) in 10 mL of CH2Cl2. The mix-
ture was warmed to room temperature and stirred over night.
The methylene chloride was removed in vacuum, and the residue
was extracted with petroleum ether. Cooling the filtrate to −30 ◦C
yielded red crystals of CS1 (420 mg, 27%). mp: 57.3–59.1 ◦C. 1H
NMR (400 MHz, CDCl3) ı: 7.17 (m, 1H, thienyl-H), 6.91 (m, 1H,
thienyl-H), 6.83 (m, 1H, thienyl-H), 7.07 (t, J = 2.8 Hz, 2H, Cp-H),
2. Experimental
2.1. General considerations
1
All manipulations were carried out under a dry argon atmo-
sphere using standard Schlenk techniques. Solvents were purified
by distillation over sodium benzophenone (diethyl ether, tetrahy-
drofuran, toluene and n-hexane) and P2O5 (dichloromethane).
MAO (toluene solution, 1.53 M) was produced by Crompton. Poly-
merization grade ethylene was used without further purification.
6,6-dialkylfulvenes [27], Benzofulvenes [28], substituted thio-
phenyl lithium [29–31], 2-trimethylsilyl thiophene [31], and 5 [26]
were synthesized according to the literature.
6.87 (t, J = 2.8 Hz, 2H, Cp-H), 1.92 (s, 6H, CH3). 13C { H} NMR
(125 MHz, CDCl3) ı: 153.06, 152.96, 126.81, 124.25, 123.54, 123.38,
121.37, 39.66, 30.13. IR (KBr, cm−1): 3102, 2970, 2925, 1474, 1415,
1381, 1360, 1259, 1228, 1143, 1046, 847, 828, 758. Anal. Calc. for
C12H13Cl3STi:C, 41.96; H, 3.81; Found C, 42.50; H, 3.74.
2.2.2. Synthesis of complex CS2: {ꢀ5-C5H4CMe2C4H2S(Me)}TiCl3
The same procedure as described for LS1 was used. 28 mmol of
2-methylthienyl lithium, 3.00 g (28 mmol) of 6,6ꢀ-dimethylfulvene

22
Y. Zhang et al. / Journal of Molecular Catalysis A: Chemical 387 (2014) 20–30
and 3.10 g (28 mmol) of trimethylsilyl chloride were used to
give C5H4(SiMe3)CMe2C4H2S(Me) (LS2) in yield of 36% (3.50 g).
90–92 ◦C/0.1 mmHg. 1H NMR (400 MHz, CDCl3) ı 6.58 (d, J = 3.4 Hz,
1H, thienyl-H), 6.52 (m, 1H, thienyl-H), 6.48 (br, 1H, Cp-H), 6.38 (br,
1H, Cp-H), 6.13 (br, 1H, Cp-H), 3.25 (s, 1H, Cp-H), 2.40 (s, 3H, CH3),
1.59 (s, 6H, CH3), −0.04 (m, 9H, Si-CH3).
2.2.5. Synthesis of complex CS5:
{ꢀ5-C5H4C(cyclo-C5H10)C4H2S(CMe3)}TiCl3
The same procedure as described for LS1 was used. The
5-tert-butyl-2-thiophenyl lithium 1.75 g (12 mmol), 6,6-
pentamethylenefulvene 1.9 g (13 mmol) and Me3SiCl 1.7 mL
(1.44 g, 13 mmol) were used to give LS5 3.1 g (8.6 mmol, 72%)
dark red oil. 1H NMR (400 MHz, CDCl3) ı 6.56 (d, J = 3.6 Hz, 1H,
thienyl-H), 6.55 (d, J = 3.6 Hz, 1H, thienyl-H), 6.53 (d, J = 4.6 Hz, 1H,
Cp-H), 6.40 (m, 1H, Cp-H), 6.11 (m, 1H, Cp-H), 3.25 (s, 1H, Cp-H),
2.15∼1.44 (m, 10H, CH2), 1.33 (s, 9H, CH3), −0.06 (s, 9H, Si-CH3).
The same procedure as described for CS1 was used.
C5H4(SiMe3)C(cyclo-C5H10)C4H2S(CMe3) (LS5) 2.8 g (7.8 mmol)
and TiCl4 0.86 mL (1.48 g, 7.8 mmol) were used to give 0.75 g
(1.7 mmol, 22%) red crystals of CS5. mp: 104.2–105.1 ◦C. 1H NMR
(400 MHz, CDCl3) ı 6.99 (t, J = 2.7 Hz, 1.92H, Cp-H), 6.86 (m, 0.08H,
Cp-H), 6.78 (t, J = 2.7 Hz, 2H, Cp-H), 6.75 (d, J = 3.6 Hz, 1H, thienyl-H),
6.65 (d, J = 3.6 Hz, 1H, thienyl-H), 2.57–1.53 (m, 10H, CH2), 1.34
The same procedure as described for CS1 was used.
0.72 mL (6.5 mmol) of TiCl4 and 1.80 g (6.5 mmol) of
C5H4(SiMe3)CMe2C4H2S(Me) (LS2) to yield red crystals of CS2
(600 mg, 26%). mp: 69.1–70.3 ◦C. 1H NMR (400 MHz, CDCl3) ı
6.95 (d, J = 3.5 Hz, 1H, thienyl-H), 6.53 (d, J = 3.5 Hz, 1H, thienyl-H),
7.06 (t, J = 2.8 Hz, 2H, Cp-H), 6.86 (t, J = 2.8 Hz, 2H, Cp-H), 2.41
1
(s, 3H, CH3), 1.87 (s, 6H, CH3). 13C { H} NMR (125 MHz, CDCl3)
ı: 153.25, 150.71, 138.91, 124.74, 123.38, 123.14, 121.38, 39.70,
29.85, 15.26. IR (KBr, cm−1): 3105, 2965, 2923, 2855, 1476, 1445,
1380, 1360 m, 1257, 1229, 1145, 1067, 1046, 826, 796, 698. Anal.
Calc. for C13H15Cl3STi: C, 43.67; H, 4.23; found C, 43.50; H, 4.19.
1
(s, 9H, CH3). 13C { H} NMR (125 MHz, CDCl3) ı: 157.02, 155.67,
144.30, 124.96, 123.19, 121.13, 120.96, 43.83, 37.16, 34.50, 32.41,
25.33, 22.30. IR (KBr, cm−1): 2938, 2858, 1463, 1362, 1255, 835,
799, 763, 463, 416. Anal. Calc. for C19H25Cl3STi: C, 51.90; H, 5.73;
found C, 51.93; H, 5.49.
2.2.3. Synthesis of complex CS3:
[ꢀ5-C5H4CMe2C4H2S(SiMe3)]TiCl3
The same procedure as described for LS1 was used. 5 g
(21 mmol) of The 5-(trimethylsilyl)-2-thiophenyl lithium, 6,6-
dimethylfulvene 2.2 g (21 mmol) and 1.83 g (16.8 mmol) of
trimethylsilyl chloride was used to yield dark red oil of LS3 4.6 g
(13.7 mmol, 82%). 1H NMR (400 MHz, CDCl3) ı 7.02 (d, J = 3.4 Hz,
1H, thienyl-H), 6.85 (d, J = 3.4 Hz, 1H, thienyl-H), 6.49 (m, 1H, Cp-
H), 6.41 (m, 1H, Cp-H), 6.14 (m, 1H, Cp-H), 3.26 (s, 1H, Cp-H), 1.64
(s, 6H, CH3), 0.26 (s, 9H, Si-CH3), −0.04 (s, 9H, Si-CH3).
The same procedure as described for CS1 was used. 1.14 g
(6.0 mmol) of titanium tetrachloride and 2 g (6.0 mmol) of
C5H4(SiMe3)CMe2C4H2S(SiMe3) (LS3) were used to yield yellow
2.2.6. Synthesis of complex CS6:
{ꢀ5-C5H4C(cyclo-C5H10)C4H2S(SiMe3)}TiCl3
The same procedure as described for LS1 was used. The
5-(trimethylsilyl)-2-thiophenyl lithium 5 g (21 mmol), 6,6-
pentamethylenefulvene 3.2 g (22 mmol) and Me3SiCl 2.8 mL
(2.38 g, 22 mmol) were used to give 5.9 g (15.8 mmol, 72%) dark
red oil of LS6. 1H NMR (400 MHz, CDCl3) ı 7.01 (d, J = 3.4 Hz, 1H,
thienyl-H), 6.84 (d, J = 3.4 Hz, 1H, thienyl-H), 6.52 (d, J = 4.3 Hz, 1H,
Cp-H), 6.40 (d, J = 4.3 Hz, 1H, Cp-H), 6.12 (m, 1H, Cp-H), 3.25 (s, 1H,
Cp-H), 2.18–1.43 (m, 10H, CH2), 0.24 (s, 9H, Si-CH3), −0.07 (s, 9H,
Si-CH3).
crystals of CS3. Yield: 0.32 g (0.8 mmol, 13%). mp: 78.1–79.7 ◦C. 1
H
NMR (400 MHz, CDCl3) ı 7.07 (t, J = 2.7 Hz, 1.75H, Cp-H), 7.00 (d,
J = 3.4 Hz, 0.88H, thienyl-H), 6.99 (d, J = 3.4 Hz, 0.12H, thienyl-H),
6.94 (t, J = 2.7 Hz, 0.25H, Cp-H),6.85 (t, J = 2.7 Hz, 1.75H, Cp-H), 6.83
(d, J = 3.4 Hz, 0.88H, thienyl-H), 6.79 (d, J = 3.4 Hz, 0.12H, thienyl-H),
6.73 (t, J = 2.7 Hz, 0.25H, Cp-H),1.9 (s, 6H, CH3), 0.26 (s, 9H, Si-CH3).
The same procedure as described for CS1 was used.
C5H4(SiMe3)C(cyclo-C5H10)C4H2S(SiMe3) (LS6) 1 g (2.7 mmol)
and TiCl4 0.29 mL (0.51 g, 2.7 mmol) were used to give 0.38 g
(0.8 mmol, 30%) yellow crystals of CS6. mp: 81.8–82.4 ◦C. 1H
NMR (400 MHz, CDCl3) ı 7.07 (d, J = 3.3 Hz, 1H, thienyl-H), 7.00
(d, J = 3.3 Hz, 1H, thienyl-H), 6.97 (t, J = 2.7 Hz, 1.83H, Cp-H), 6.84
(t, J = 2.7 Hz, 0.17H, Cp-H), 6.76 (t, J = 2.7 Hz, 1.83H, Cp-H), 6.62
(t, J = 2.7 Hz, 0.17H, Cp-H), 2.61–1.25 (m, 10H, CH2), 0.26(s, 9H,
1
13C { H} NMR (125 MHz, CDCl3) ı: 158.46, 152.97, 139.29, 133.82,
124.87, 123.40, 121.43, 39.86, 30.12, 0.11. IR (KBr, cm−1): 2955,
1476, 1249, 981, 840, 804, 696, 422. Anal. Calc. for C15H21Cl3SSiTi:
C, 43.34; H, 5.09; found C, 43.96; H, 5.26.
1
Si-CH3). 13C { H} NMR (125 MHz, CDCl3) ı: 155.26, 153.36, 140.07,
134.23, 126.93, 123.19, 120.96, 44.04, 37.47, 25.33, 22.34, −0.05. IR
(KBr, cm−1): 2936, 2857, 1249, 1047, 985, 841, 827, 757, 416. Anal.
Calc. for C18H25Cl3SSiTi: C, 47.43; H, 5.53; found C, 47.40; H, 5.65.
2.2.4. Synthesis of complex CS4:
{ꢀ5-C5H4C(cyclo-C5H10)C4H2S(Me)}TiCl3
2.2.7. Synthesis of CS7: [ꢀ5-C9H6-C(cyclo-C5H10)-C4H3S]TiCl3
To a solution of 40 mmol of 2-thienyl lithium in 20 mL of
ether was added dropwise a solution of 7.8 g (40 mmol) of 6,6-
pentamethylenebenzofulvene in 45 mL diethyl ether at −30 ◦C. The
reaction mixture was warmed to room temperature and was stirred
for 24 h, and then was hydrolyzed with 100 mL ice-water. The water
layer was extracted with 50 mL light petroleum, and the combined
organic layers were dried with anhydrous MgSO4. The solvent was
removed under vacuum, and the residue was chromatographed on
the silicagel column with petroleum as eluent to give pure LS7 in
yield of 20% (2.23 g) mp: 84.5–85.3 ◦C. 1H NMR (400 MHz, CDCl3) ı
7.43–7.41 (m, 1H, Ind-H), 7.24–7.22 (m, 1H, Ind-H), 7.11–7.08 (m,
3H, thienyl-H, Ind-H), 6.89–6.87 (m, 2H, thienyl-H, Ind-H), 6.42 (t,
J = 2.0 Hz, 1H, Ind-H), 3.38 (d, J = 2.0 Hz, 2H, Ind-H), 2.50–2.42 (m,
2H, (CH2)5), 2.25–2.17 (m, 2H, (CH2)5), 1.70–1.60 (m, 4H, (CH2)5),
1.57–1.47 (m, 2H, (CH2)5).
The same procedure as described for LS1 was used. 0.94 g
(9 mmol) of 5-methyl-2-thiophenyl lithium, 1.3 g (9 mmol) of
6,6-pentamethylenefulvene and 0.97 g (9 mmol) of trimethylsilyl
chloride were used to yield dark red oil of LS4 2.7 g (8.5 mmol, 94%).
1H NMR (400 MHz, CDCl3) ı 6.62 (m, 1H, Cp-H), 6.57 (m, 2H, thienyl-
H), 6.45 (m, 1H, Cp-H), 6.16 (br, 1H, Cp-H), 3.29 (s, 1H, Cp-H), 2.44
(s, 3H, CH3), 2.14 (m, 10H, CH2), 0.00 (s, 9H, Si-CH3).
The same procedure as described for CS1 was used. 1.5 g
(7.9 mmol) of titanium tetrachloride and 2.5 g (7.9 mmol) of
C5H4(SiMe3)C(cyclo-C5H10)C4H2S(Me) (LS4) were used to yield
yellow crystals of CS4. Yield: 0.54 g (1.0 mmol, 16%). mp:
101.1–102.7 ◦C. 1H NMR (400 MHz, CDCl3) ı 6.91 (t, J = 2.7 Hz, 1.83H,
Cp-H), 6.78 (m, 0.17H, Cp-H), 6.72 (t, J = 2.7 Hz, 1.83H,Cp-H), 6.69
(m, 1H, thienyl-H), 6.60 (m, 0.17H, Cp-H), 6.54 (m, 1H, thienyl-H),
1
2.37 (s, 3H, CH3), 2.26–1.22 (m, 10H, CH2). 13C { H} NMR (125 MHz,
CDCl3) ı: 155.63, 145.62, 139.40, 125.50, 125.28, 123.20, 120.09,
43.86, 37.07, 25.31, 22.29, 15.39. IR (KBr, cm−1): 3098, 3075, 2930,
1466, 1444, 1261, 1255, 1040, 840, 668, 522, 455, 419. Anal. Calc.
for C16H19Cl3STi: C, 48.33; H, 4.82; found C, 48.72; H, 5.14.
2.5 mL (6 mmol) n-BuLi in n-hexane was added dropwise to a
solution of 1.68 g (6 mmol) of ␩5-C9H6-C(cyclo-C5H10)-C4H3S (LS7)
in 40 mL n-hexane at 0 ◦C. The reaction mixture was warmed to

Y. Zhang et al. / Journal of Molecular Catalysis A: Chemical 387 (2014) 20–30
23
room temperature and stirred overnight. Then 0.66 mL (6 mmol)
TiCl4 was added to the mixture at −78 ◦C, the dark red solution
was warmed to room temperature and was stirred overnight. The
solvent and unreacted TiCl4 were removed in vacuum, and the
residue was extracted with n-hexane. After cooling the filtrate
to −20 ◦C, the title compound as deep red crystals was obtained.
(390 mg, 15%). mp: 104.1–105.3 ◦C. 1H NMR (400 MHz, CDCl3) ı
7.95 (dd, J = 8.8 Hz, J = 1.0 Hz 1H, Ind-H), 7.70 (d, J = 8.8 Hz, 1H,
Ind-H), 7.47 (ddd, J = 8.4 Hz, J = 6.8 Hz, J = 1.0 Hz, 1H, Ind-H), 7.35
(ddd, J = 8.4 Hz, J = 6.8 Hz, J = 1.0 Hz, 1H, Ind-H), 7.20 (dd, J = 5.2,
J = 1.0 Hz, 1H, thienyl-H), 7.18–7.15 (m, 2H, Ind-H, thienyl-H), 7.00
(d, J = 3.6 Hz, 1H, Ind-H), 6.97 (dd, J = 5.2 Hz, J = 3.6 Hz, 1H, thienyl-H),
2.91 (dd, J = 13.8, J = 1.6 Hz, 1H, (CH2)5), 2.77 (td, J = 13.8, J = 3.6 Hz,
1H, (CH2)5), 2.68–2.54 (m, 1H, (CH2)5), 2.33–2.20 (m, 1H, (CH2)5),
6.99 (d, J = 3.6 Hz, 1H), 2.91 (d, J = 13.2 Hz, 1H), 2.76 (td, J = 13.2,
3.2 Hz, 1H), 2.68–2.58(m, 1H), 2.32–2.18 (m, 1H), 1.84–1.64 (m, 5H),
1
1.55–1.45 (m, 1H), 0.25 (s, 9H). 13C { H} NMR (125 MHz, CDCl3)
ı: 153.54, 149.69, 139.82, 133.66, 132.03, 129.10, 128.79, 128.62,
127.87, 127.68, 127.40, 120.18, 115.07, 45.32, 37.70, 36.94, 34.11,
25.46, 22.51, 22.34, 22.24, 14.07. IR (KBr, cm−1): 3116.2, 3097.6,
2932.0, 2856.3, 1636.8, 1618.2, 1449.2, 1259.9, 1084.1, 840.9, 754.6.
Anal. Calcd. for C22H27Cl3SSiTi.0.8 C5H12: C, 55.41; H, 6.55; found:
C, 55.79; H, 6.35.
2.2.10. Synthesis of CS10:
[ꢀ5-C9H6-C(cyclo-C5H10)-(3-MeC4H3S)]TiCl3
The same procedure as described for LS7 was used.
The 3-methyl-2-thiophenyl lithium 5 g (48 mmol) and 6,6-
pentamethylenefulvene 9.4 g (48 mmol) were used to give pure
LS10 2.4 g (17%). mp: 100.1–101.5 ◦C. 1H NMR (400 MHz, CDCl3)
ı 7.43 (d, J = 7.2, 1H, Ind-H), 7.14–7.01 (m, 4H, 3Ind-H, thienyl-H),
6.69–6.61 (m, 1H, thienyl-H), 6.55–6.38 (m, 1H, Ind-H), 3.41
(s, 2H, Ind-H), 2.56–2.45 (m, 2H, (CH2)5), 2.21–2.11 (m, 2H,
(CH2)5), 1.92–1.80 (m, 3H, thienyl-CH3), 1.75–1.56 (m, 5H, (CH2)5),
1.51–1.41 (m, 1H, (CH2)5).
1
1.87–1.65 (m, 4H, (CH2)5), 1.57–1.39 (m, 2H, (CH2)5). 13C { H} NMR
(125 MHz, CDCl3) ı: 149.49, 148.45, 132.02, 129.19, 128.76, 128.69,
127.85, 127.23, 126.66, 126.45, 124.78, 120.10, 115.01, 45.08, 37.74,
36.78, 25.45, 22.49, 22.18. IR (KBr, cm−1): 3116.9, 3098.1, 2933.7,
2857.7, 1447.7, 1426.8, 1261.2, 1228.6, 802.8, 747.5, 649.1, 521.1.
Anal. Calcd. for C19H19Cl3STi-0.25C6H14:C, 54.09; H, 4.98; found C,
53.67; H, 4.94.
The same procedure as described for CS7 was used. 1.7 g
LS10 (6 mmol), 2.5 mL, n-BuLi(2.4 mol/L) and 0.66 mL TiCl4 (1.14 g,
6 mmol) were used to give deep red crystals of CS10. (350 g, 13%).
mp: 123.2–124.4 ◦C. 1H NMR (400 MHz, CDCl3) ı = 7.75 (d, J = 8.8,
1H, Ind-H), 7.69 (d, J = 8.8, 1H, Ind-H),7.44–7.37 (m, 1H, Ind-H),
7.37–7.31 (m, 1H, Ind-H), 7.26–7.24 (m, 1H, thienyl-H), 7.19 (d,
J = 3.6, 1H, Ind-H), 7.12 (d, J = 5.2, 1H, Ind-H), 6.60 (d, J = 5.2, 1H,
thienyl-H), 3.07–2.99 (m, 1H, (CH2)5), 2.80–2.70 (m, 1H, (CH2)5),
2.69–2.60 (m, 1H, (CH2)5), 2.33–2.07 (m, 3H, (CH2)5), 1.90–1.73
2.2.8. Synthesis of CS8:
[ꢀ5-C9H6-C(cyclo-C5H10)-(5-MeC4H3S)]TiCl3
The same procedure as described for LS7 was used.
The 5-methyl-2-thiophenyl lithium 5 g (48 mmol) and 6,6-
pentamethylenefulvene 9.4 g (48 mmol) were used to give pure
LS8 4.2 g (30%). mp: 102.7–103.5 ◦C. 1H NMR (400 MHz, CDCl3) ı
7.47–7.41 (m, 1H, Ind-H), 7.36–7.31 (m, 1H, Ind-H), 7.16–7.01 (m,
2H, Ind-H), 6.72–6.68 (m, 1H, thienyl-H), 6.54 (s, 1H, thienyl-H),
6.42 (s, 1H, Ind-H), 3.39 (s, 2H, Ind-H), 2.47–2.37 (m, 5H, 2(CH2)5,
thienyl-CH3), 2.27–2.10 (m, 2H, (CH2)5), 1.75–1.46 (m, 6H, (CH2)5).
The same procedure as described for CS7 was used. 1.7 g
LS8 (6 mmol), 2.5 mL, n-BuLi(2.4 mol/L) and 0.66 mL TiCl4 (1.14 g,
6 mmol) were used to give deep red crystals of CS8 (1.4 g, 52%).
mp: 125.3–126.7 ◦C. 1H NMR (400 MHz, CDCl3) ı 7.99 (d, J = 8.8 Hz,
1H, Ind-H), 7.71 (d, J = 8.8 Hz, 1H, Ind-H), 7.53–7.45 (m, 1H, Ind-
H), 7.40–7.31 (m, 1H, Ind-H), 7.16 (d, J = 3.6 Hz, 1H, Ind-H), 7.01
(d, J = 3.6 Hz, 1H, Ind-H), 6.94 (d, J = 3.6 Hz, 1H, thienyl-H), 6.60 (dd,
J = 3.6, 0.8 Hz, 1H, thienyl-H), 2.86 (dd, J = 14.0, 2.8 Hz, 1H, (CH2)5),
2.73 (td, J = 12.8, 3.6 Hz, 1H, (CH2)5), 2.54(dd, 12.8, 3.6HZ, 1H,
(CH2)5), 2.40–2.37 (m, 3H, thienyl-CH3), 2.27–2.16 (m, 1H, (CH2)5),
1.85–1.65(m, 4H, (CH2)5), 1.59–1.45 (m, 1H, (CH2)5), 1.44–1.31 (m,
1
(m, 4H, (CH2)5), 1.70 (s, 3H, thienyl-CH3). 13C { H} NMR (125 MHz,
CDCl3) ı: 150.77, 144.45, 138.11, 133.84, 131.97, 129.21, 128.64,
127.57, 126.85, 123.39, 120.06, 113.84, 102.57, 44.47, 41.89, 37.93,
25.54, 23.09, 22.00, 15.70. IR (KBr, cm−1): 3117.1, 3098.2, 2932.1,
2857.9, 1635.5, 1448.9, 1228.3, 1062.7, 836.8, 844.1, 648.9. Anal.
Calcd. for C20H21TiCl3S–0.4 C5H12: C, 55.45; H, 5.46; found: C,
55.72; H, 5.48.
2.3. X-ray diffraction measurements
Single crystals of indenyl titanium complex CS8 was obtained
from saturated solutions of n-hexane at room temperature. The
CCDC No. is 896659. Diffraction data were collected on a Bruker
SMART 1000 CCD diffractometer using graphite monochromated
1
1H, (CH2)5). 13C { H} NMR (125 MHz, CDCl3) ı: 149.87, 145.82,
◦
˚
139.24, 132.06, 129.13, 128.83, 128.67, 127.84, 127.38, 126.27,
124.66, 120.11, 115.00, 45.12, 37.42, 36.48, 25.48, 22.49, 22.19,
15.36. IR (KBr, cm−1): 3117.7, 3098.3, 2933.4, 2857.6, 1634.9,
1447.7, 1228.3, 1063.9, 836.8, 802.9, 747.7, 649.1. Anal. Calcd. for
Mo K␣ radiation (ꢁ = 0.71073 A). All data were collected at 20 C
using the ␻-scan techniques. The structures of CS8 was solved by
direct method and refined using Fourier techniques. The absorp-
tion correction was applied based on SADABS [32]. Hydrogen
atoms were located and refined by the geometry method. All non-
hydrogen atoms were refined by full-matrix least-squares on F2
using the SHELXTL program package [33]. The cell refinement,
data collections, and reduction were done by Bruker SAINT [34].
The structure determination and refinement were performed by
SHELXS-97 [35] and SHELXL-97 [36], respectively. The crystal and
refinement data are collected in Tables 1 and 2.
C20H21TiCl3S: C, 53.66; H, 4.73; found: C, 54.10; H, 5.03.
2.2.9. Synthesis of CS9:
[ꢀ5-C9H6-C(cyclo-C5H10)-(5-Me3SiC4H3S)]TiCl3
The same procedure as described for LS7 was used. The
5-trimethylsilyl-2-thiophenyl lithium 5.8 g (36 mmol) and 6,6-
pentamethylenefulvene 7.1 g (36 mmol) were used to give pure LS9
3.1 g (24%). mp: 92.7–94.1 ◦C. 1H NMR (400 MHz, CDCl3) ı 7.48–7.40
(m, 1H, Ind-H), 7.35–7.27 (m, 1H, Ind-H), 7.17–7.10 (m, 2H, Ind-H),
7.06–7.01 (m, 1H, thienyl-H), 6.96–6.90 (m, 1H, thienyl-H), 6.42 (s,
1H, Ind-H), 3.40 (s, 2H), 2.55–2.44 (m, 2H, (CH2)5), 2.30–2.17 (m,
2H, (CH2)5), 1.75–1.45 (m, 6H, (CH2)5), 0.28 (s, 9H, thienyl-SiMe3).
The same procedure as described for CS7 was used. 1.3 g LS9
(3.6 mmol), 1.5 mL, n-BuLi(2.4 mol/L) and 0.39 mL TiCl4 (0.68 g,
3.6 mmol) were used to give deep red crystals of CS9. (870 g, 48%).
mp: 105.5–106.7 ◦C. 1H NMR (400 MHz, CDCl3) ı 7.96 (d, J = 8.8 Hz,
1H), 7.71 (d, J = 8.8 Hz, 1H), 7.51–7.44 (m, 1H), 7.39–7.32 (m, 1H),
7.18 (d, J = 3.2 Hz, 1H), 7.16 (d, J = 3.2 Hz, 1H), 7.09 (d, J = 3.6 Hz, 1H),
2.4. General procedure for the trimerization of ethylene
A 100 mL autoclave equipped with a magnetic stirrer was heated
under vacuum for one hour over 80 ◦C, then allowed to cool to
the required reaction temperature under ethylene atmosphere,
and charged with 20 mL of toluene. The autoclave was pressur-
ized to 0.5 MPa of ethylene pressure. Stirring for about 30 min
ensured that the solution was equilibrated and the required reac-
tion temperature was established. The ethylene pressure was then
released. Then catalyst and MAO in toluene solution were added,

24
Y. Zhang et al. / Journal of Molecular Catalysis A: Chemical 387 (2014) 20–30
Table 1
According to the literature [9,26], the Cp-based ligands are
Crystallographic data for compounds CS8.
prepared from the addition reaction of 6,6-dialkylfulvenes with
kinds of substituted thienyl lithium salts. The resulting cyclopen-
tadienyl lithium reacted with trimethylsilyl chloride to afford the
corresponding intermediates with trimethylsilyl groups. The vari-
ous half -sandwich titanium complexes are prepared via Me3SiCl
elimination upon reaction of the corresponding (trimethylsi-
lyl)cyclopentadienyl intermediates with TiCl4.
Formula
C20H21Cl3STi Mr
293(2) Wavelength [Å]
447.68
T [K]
0.71073
P2(1)
20.295(8)
90
90
4
Crystal system Monoclinic
a [Å]
Space group
9.514(4)
10.708(4)
94.872(5)
2060.1(14)
1.443
b [Å]
˛ [◦]
ꢂ [◦]
Z
c [Å]
ˇ [◦]
V [Å3]
ꢃcalcad [g cm−3
F(0 0 0)
Rint
Taking into account different steric and electron effects of
the indenyl group, we also synthesized a series of indenyl com-
plexes (Ind-bridge-thienyl) TiCl3. The synthetic route is shown in
Scheme 2. The synthesis of (Ind-bridge-thienyl) TiCl3 trying to use
the same synthetic route of Cp-based complexes by reacting Me3Si-
contained ligand with TiCl4 have failed. This is probably because of
the low activity of trimethylsilyl indenyl with bulk substituent on
the indenyl group. Actually, the indenyl complexes are synthesized
directly by the reaction of more active lithium salts of ligands with
TiCl4 in 1:1 molar ratio [37]. As shown in Scheme 2, benzofulvenes
firstly react with kinds of substituted thienyl lithium salts to give
corresponding substituted indenyl lithium salts. To remove unre-
acted benzofulvents and other impurities, the indenyl lithium salt
formed in the first step were not reacted with TiCl4 directly, instead
further hydrolysis and chromatographic purification were carried
out to afford the ligand LS7-LS10 in pure form, which were deproto-
nated with n-butyl lithium to give the corresponding lithium salts,
then react with TiCl4.
]
ꢄ [mm−1
]
0.907
920
0.0754
Reflections collected/unique 8390/3619
R1, wR2 (I > 2ꢅ) 0.0527, 0.1418
Table 2
Selected bond lengths (Å) and angles (◦) of complexes CS8.
Ti1-Cl1
Ti1-Cl3
Ti1-C2
Ti1-C4
Cl2-Ti1-Cl3
Cl3-Ti1-Cl1
Cl1-Ti1-C4
C1-Ti1-C5
C4-Ti1-C5
2.2174(13)
2.2396(12)
2.275(4)
Ti1-Cl2
Ti1-C1
Ti1-C3
2.2211(14)
2.368(4)
2.328(3)
2.413(3)
Ti1-C5
2.462(3)
101.07(5)
103.14(5)
100.22(8)
34.20(12)
34.12(11)
Cl2-Ti1-Cl1
Cl3-Ti1-C4
C2-Ti1-C5
C3-Ti1-C5
C4-C10-C16
105.61(5)
144.86(9)
57.72(12)
56.57(11)
106.7 (3)
respectively. The reactor was sealed only open for ethylene and
pressurized to required ethylene pressure. The ethylene pressure
was kept constant during trimerization. After the reaction was car-
ried out for 0.5 h, the reactor was quitted and the ethylene flow
was terminated. The autoclave was cooled to the room tempera-
ture and the ethylene pressure was released and 1 mL of ethanol
was injected immediately to stop the reaction. Heptane was added
as internal standard. The solution part was analyzed by GC/MS or
GC to determine the content of oligomers. The polymer was filtered,
washed with acidified ethanol (0.5% HCl), and dried under vacuum
at 60 ◦C to a constant weight.
3.2. Crystal structures of CS8
The molecular structures of CS8 was determined by single-
crystal X-ray diffraction analysis. The ORTEP drawings of its
molecular structure is given in Fig. 1. The complex shows the same
familiar piano-stool geometry as the half-sandwich Cp titanium
with pendant arene groups [37], where the thienyl substituent
adopt a conformation roughly perpendicular to the plane of the
indenyl ring, reverse direction of the metal.
3. Result and discussion
3.3. Determination the structure of complexes by 1H NMR
3.1. Synthesis of the complexes (Cp-bridge-thienyl)TiCl3 and
(Ind-bridge-thienyl) TiCl3
A major point to determine the structure of the complexes is to
define whether the sulfur atom coordinate to the titanium. From
the X-ray determination of the CS8 in Fig. 1, it can be seen that the
sulfur atom do not coordinate with the metal in the solid state. In
addition to X-ray determination, it is also confirmed by 1H NMR
spectrum. Generally speaking, the chemical shift of the hydrogen
on the ␣-carbon of the heteroatom can be as a probe to determine
To study the influence of different hemilabile coordination
behavior on ethylene trimerization, catalyst precursors (Cp-bridge-
thienyl)TiCl3 with different bridges and different substituents on
the thienyl group are synthesized. The synthetic route is shown in
Scheme 1.
Scheme 1. The synthetic route of half-sandwich Cp titanium complexes with pendent thienyl group.

Y. Zhang et al. / Journal of Molecular Catalysis A: Chemical 387 (2014) 20–30
25
Scheme 2. The synthetic route of half-sandwich indenyl titanium complexes with pendent thienyl group.
By the same way, taking 5 and CS7 as examples, their chemical
shifts are shown in Fig. 3. The chemical shifts of the hydrogen on ␣-
carbon of the sulfur atom are 7.25 and 7.20 ppm, respectively, while
for their ligands, the chemical shift of the corresponding hydrogen
are 7.16 [26] and 7.11 ppm, respectively. The differences are only
0.09 ppm. So it can be drawn that the sulfur atom in these com-
plexes do not coordinate with the metal, just agree with the fact
from X-ray determination.
Another interesting phenomenon worth noting is that the 1H
NMR spectrums of Cp complexes (CS3, CS5 and CS6) with bulky
t
groups such as Me3Si and Bu groups in the thienyl moiety show
that each hydrogen in Cp moiety and thienyl moiety have two sets
of peaks (the proportion about 10:1). Variable temperature 1H NMR
experiment (SI, Fig. S1) show that increasing the temperature, the
ratio of the minor part will increase too. This means that in solu-
tion these complexes may have two different configurations which
probably from the low frequency of the single bond rotation sup-
pressed by the bulky group of these complexes.
Fig. 1. X-ray crystal structure of CS8. Thermal ellipsoids are shown at the 30% prob-
ability level.
3.4. Ethylene trimerization results and discussions
The
trimerization
products
catalyzed
by
[Cp(Ind)-
whether the heteroatom coordinate to the metal [38]. Due to the
change of the electronic inductive effect of the heteroatom, if the
heteroatom coordinates with the metal, the chemical shift of the
hydrogen of ␣-carbon of the heteroatom in the metal complexes
will move to the low field relative to that in the free ligand.
A series of half-sandwich titanium complexes with pendant
ether group [39–41] were reported by our group and firstly discov-
ered that the pendant oxygen atom is intramolecular coordination
with the titanium by X-ray determination. Take the compound CO1
(bridge)–thienyl]TiCl3/MAO system consist of two fraction: liquid
fraction and polyethylene. The liquid fraction as the oligomeriza-
tion product is analyzed by GC (Gas Chromatography) and GC/MS.
1-hexene is the main product (>99%) and only little C10 oligomers
are detected (<1%). Tables 3 and 4 show the trimerization results
by Cp- and Ind-based complexes/MAO system respectively, in
which the calculation of activity and selectivity for C10 oligomers
are ignored [26].
ꢀ
ꢀ
as a typical example [39], the chemical shift of Ha and Hb on
the ␣-carbon of the oxygen atom are 3.67 and 3.43 ppm respec-
tively (Fig. 2), while in the corresponding ligand, they are 2.81 and
2.70 ppm [41]. The differences are 0.86 and 0.73 ppm, respectively.
Fig. 3. The chemical shift change between the [Cp(ind)-bridge-thienyl]TiCl3 and the
Fig. 2. The chemical shift change of certain hydrogen between CO1 and LO1.
corresponding ligands.

26
Y. Zhang et al. / Journal of Molecular Catalysis A: Chemical 387 (2014) 20–30
Table 3
Catalytic ethylene trimerization with the (Cp-bridge-thienyl)TiCl3/MAO system.
Entry
Cat.
T (◦C)
P (MPa)
1-Hexene (mg)
Polymer (mg)
Prod.a
Select.b
1
2
CS1
CS2
30
30
0.5
0.5
41
55
24
30
55
73
63
65
3
4
5
6
0
30
30
80
0.5
0.5
0.8
0.5
364
222
412
34
32
64
160
62
102
296
549
45
92
78
72
35
CS3
5*
7
8
9
0
30
30
80
0.5
0.5
0.8
0.5
90
117
165
102
trace
17
22
120
156
220
136
> 95
87
88
10
206
33
11
12
13
14
0
30
30
80
0.5
0.5
0.8
0.5
59
187
368
67
2
78
249
490
90
97
89
93
53
24
26
60
CS4
CS5
CS6
15
16
17
18
0
30
30
80
0.5
0.5
0.8
0.5
24
207
317
89
10
55
111
115
32
275
423
118
71
79
74
44
19
20
21
22
0
30
30
80
0.5
0.5
0.8
0.5
364
415
524
115
32
66
79
50
485
553
698
153
92
86
87
70
Reaction condition: 1.5 ␮ mol catalyst; Al:Ti = 1000:1; 20 mL toluene solvent; 30 min run time.
a
Productivity: kg(1-hexene)/(mol Ti-h).
Selectivity: percentage of 1-hexene by mass in overall by mass.
Reported by [26].
b
*
Although the X-ray and 1H NMR spectrum show that there is
no coordination between the sulfur atom and the central metal
in the [Cp(Ind)-bridge-thienyl]TiCl3 complexes, after these com-
plexes were activated with MAO, the central metal was reduced
to low valence titanium, which may easily coordinated with sul-
fur atom. In Hessen’s ethylene trimerization system, it has proofed
by X-ray that low valence titanium can coordinate with arene
group; the variable temperature NMR shows that this coordination
is hemilabile [10]. The ethylene trimerization mechanism proposed
by Hessen presented that the hemilabile coordination of arene
group with low valence titanium is necessary to selective ethy-
lene trimerization [8]. Our previous studies are shown that the
hemilabile behavior of the sulfur atom as ␩1 coordination with low
valence titanium is essential for selective ethylene trimerization
[26]. So any factor that influence this hemilabile behavior will affect
the catalytic behavior.
low-valence metal. So the bridge moiety with different substituents
have a direct effect on the coordination between the sulfur atom
and the low valence titanium.
From Table 3, it can be seen that at 30 ◦C the activity and selec-
tivity of the catalysts 5, CS4 and CS6 with C(cyclo-C5H10) as bridge
group are remarkably higher than that of its analogue complexes
CS1, CS2 and CS3 with CMe2 as bridge group. This can be more
easily seen in Chart 2.
The experimental results indicate that the bulky bridging unit
is more favorable for the trimerization. The steric repulsive effect
of the “cyclo-C5H10” substituent could force the thienyl moiety to
move more closely to the titanium center than the CMe2, especially
in the state of cation. So a proper S-coordination will increase the
reactivity of the catalyst.
In previous studies, we reported that, in the preparation of
o-MeO-containing benzyl-substituted cyclopentadienyl titanium
complexes, when the bridging carbon is linked with ethyl or bigger
groups, the reaction always eliminates CH3Cl and gives titanoxacy-
cle complexes from the repulse force [42].
3.4.1. Ethylene trimerization results by
[Cp-bridge-thienyl]TiCl3/MAO system
In [Cp-bridge-arene]TiCl3/MAO system, Hessen and co-workers
reported the similar result that increasing the hindrance of the
bridge group can enhance the catalytic productivity [9]. But too
strong arene-metal coordination will restrain the arene “ring slip-
page” in the proposed rate-determining step. Our recent study
show that if the bridge group is bigger adamantly, the catalytic
activity will decrease [26]. So C(cyclo-C5H10) as bridge group may
be proper for hemilabile behavior to reach high trimerization pro-
ductivity, neither too small nor too big.
3.4.1.1.2. The influence by the substituent on the thienyl group.
The substituent at the 5-position of the thienyl group has both
electronic and steric effect on hemilabile behavior of sulfur-metal
coordination. From Table 3, it can very well that the substituent at
the 5-position of the thienyl group has a great impact on the activity
of ethylene trimerization.
As shown in Table 3, the [Cp-bridge-thienyl]TiCl3/MAO system
can effectively catalyze ethylene to 1-hexene. The productivity of 1-
hexene is above 105 kg/(mol Ti-h) and the selectivity of 1-hexene
can be as high as 97%, which are significantly influenced by the
structure of the Cp complexes and the experiment conditions.
3.4.1.1. Effect of the structure features of complexes on ethylene
trimerization.
3.4.1.1.1. Effect of the bridge between Cp and thienyl group. It is
well know that the steric factor of the complexes usually affect the
catalytic behavior [9]. The bridge moiety between Cp and thienyl
group play a crucial role in the orientation of the thienyl group
relative to the metal centre and the degree of angle C(Cp)-C(bridge)-
C(thienyl). When the complexes are activated with MAO, the bridge
will “press” the thienyl group to a certain extent to approach the

Table 4
Catalytic ethylene trimerization with the [Ind-bridge -thienyl]TiCl3/MAO system.
Entry
23
Catalyst
T (◦C)
P (MPa)
1-Hexene (mg)
15
Polymer (mg)
trace
Prod.a
20
Selec.b
>98
0
0.5
S
24
25
26
30
30
80
0.5
0.8
0.5
58
80
121
7
21
15
77
107
161
89
79
88
Ti
Cl
Cl
Cl
CS7
27
0
0.5
84
trace
112
> 98
S
28
29
30
30
30
80
0.5
0.8
0.5
310
357
523
10
25
29
413
476
697
97
94
95
Ti
Cl
Cl
Cl
CS8
31
0
0.5
40
4
53
91
Si
S
32
33
34
30
30
80
0.5
0.8
0.5
246
336
302
15
74
158
328
448
403
94
82
66
Ti
Cl
S
CS9
35
0
0.5
8
trace
11
> 98
Ti
Cl
Cl
Cl
36
37
38
30
30
80
0.5
0.8
0.5
15
20
38
7
10
79
20
27
51
68
67
32
CS10
Reaction condition: 1.5 ␮ mol catalyst; Al:Ti = 1000:1; 20 mL toluene solvent; 30 min run time.
a
Productivity: in kg(1-hexene)/(mol Ti-h).
Selectivity: by mass in all.
b

28
Y. Zhang et al. / Journal of Molecular Catalysis A: Chemical 387 (2014) 20–30
Chart 2. The comparison of ethylene trimerization by Cp complexes with different bridge moieties.
Generally speaking, when the complexes combined with MAO,
From Table 3, it also can be seen evidently that the highest selec-
tivity of these complexes except for CS5 is obtained at 0 ◦C due to
few formation of PE. The amount of both 1-hexene and PE increase
with increasing temperature from 0 to 30 ◦C, but the selectivity
for 1-hexene is decrease slightly at 30 ◦C. The selectivity by all the
complexes decreases dramatically above 30 ◦C due to the increased
formation of by-product PE and decreased productivity of 1-hexene
at elevated temperature.
the substituent with strong electron-donating property is more
favorable to make sulfur coordinate with cationic titanium; the
substituent with big steric effect, however, is unfavorable to this
coordination. From Table 3, it is found that the productivity of 1-
hexene is increasing in the same order as the electron-donating
property and steric effect of the substituent at the 5-position of the
thienyl group increasing, H(5) < Me(CS4) < tBu(CS5). But it is noticed
that at 30 ◦C and 0.5 MPa of ethylene, the activity of complex CS6
is about 3 times higher than that of 5 and twice higher than CS5.
It indicates that in CS6, the trimethylsilyl group with bigger steric
effect than t-Bu is decisive for higher activity. It also can be seen
from Table 3 that when the bridge group is CMe2, the activity order
is CS1-H (entry 1) < CS2-Me (entry 2) < CS3-Me3Si (entry 4), just the
same as the bridge group with C(cyclo-C5H10).
3.4.1.3. The influence by the ethylene pressure. Generally, ethylene
pressure plays a positive role on trimerization activity. As shown in
Table 3, comparing with the catalytic behavior at 0.5 MPa (entries
4, 8, 12, 16 and 20), the higher productivity could be obtained with
increasing the ethylene pressure to 0.8 MPa (entries 5, 9, 13, 17
and 21). This increasing tread is just agreed with that of Hessen’s
Cp-bridge-arene system [9], probably due to the higher ethylene
solubility under higher ethylene pressure.
From Table 3, it is obvious that at 30 ◦C, the trimerization selec-
tivity by complexes with bridge group C(cyclo-C5H10) are higher
than that by CMe2, while the substituent on the thienyl group has
no definite relationship with the selectivity.
3.4.2. Ethylene trimerization results by
It is shown that a proper combination of electron and steric
effect with C(cyclo-C5H10) as bridge unit and SiMe3 as substituent
will get optimum hemialbile behavior which reach high catalytic
productivity.
[Ind-bridge-thienyl]TiCl3/MAO system
The indenyl group have different steric and electronic effects
from Cp group. When the indenyl ring replace the Cp ring, the
results of catalytic ethylene trimerization experiments with half-
sandwich indenyl titanium complexes bearing pendant thienyl
groups are shown in Table 4. After activated with MAO, these
indenyl complexes have more activities and higher selectivity for
ethylene trimerization comparing with corresponding Cp com-
plexes. At 80 ◦C and 0.5 MPa ethylene pressure, the productivity and
the selectivity of 1-hexene can be reached 6.9 × 105 kg/(mol Ti-h)
and 95% by mass in all, respectively. The trimerization results are
also greatly influenced by the substituents on the thienyl group and
experiment conditions. The effects by these factors are discussed
in details below.
3.4.1.2. The influence by the temperature. The temperature has a
dramatically effect on the productivity and selectivity of 1-hexene.
The highest productivity of all complexes in Table 3 is obtained at
30 ◦C. Then increasing the reaction temperature, the productivity
is decreased. Although the frequency of molecular motion is faster
at 80 ◦C than that at 30 ◦C, this also leads to weaken the coordi-
nation between sulfur atom and the cation metal which may be
favor for ethylene insertion to the polymer chain and increase the
PE content; another aspect, the active trimerization species may
gradually go to deactivation at 80 ◦C. At 0 ◦C, however, as the molec-
ular motion is slow, the coordination is tight and this make the
trimerization productivity lower than at 30 ◦C. So it is clear that the
most proper hemilabile behavior of sulfur atom and cation metal is
important.
3.4.2.1. The influence by the structure of indenyl complexes. The
introduction of indenyl ring increase the delocalization of the ␲
electrons and stabilize the central metal; at the same time, indenyl
ring has bigger steric effect than Cp ring which influence the cen-
tral metal to coordinate with other group. Therefore a proper

Y. Zhang et al. / Journal of Molecular Catalysis A: Chemical 387 (2014) 20–30
29
hemilabile coordination behavior needs from a proper steric hin-
drance. In Table 4, it can be seen a different trend from Cp system:
the best catalytic behavior (entries 27–30) are from CS8 with a
methyl on the 5-position of thienyl which is more active than
CS7 with H, and CS9 with bigger substituent trimethylsilyl (entries
31–34); in the Cp-thienyl complexes, CS6 with trimethylsilyl gives
the best results. Comparing to CS6, CS9 has a more hindrance role
on trimerization due to introducing the indenyl ring and this leads
to the decrease of the productivity. So, for the indenyl complexes,
CS8 with methyl substituent has the most proper hemilabile behav-
ior that favor for catalyzing ethylene trimerization and giving the
highest productivity and selectivity.
CS10 (entries 35–38) with a methyl on the 3-position of the
thienyl group gets very low activity and selectivity for 1-hexene.
This is probably due to the methyl too close to the bridge group
which limit the rotation of the thienyl group. In our recently studies,
indenyl complex with methyl at the ortho-position of the arene
group in the [Ind-bridge-arene]TiCl3/MAO system also show poor
catalytic behavior [37].
When indenyl is introduced into the ligand, because of
the change of the electronic and steric effect, the [Ind-bridge-
thienyl]TiCl3/MAO system has more optimal properties with Cp
based system. Firstly, indenyl based system is more tolerant of
temperature, they can catalyze ethylene trimerization with both
high productivity and 1-hexene selectivity even at 80 ◦C. CS8 gives
the highest productivity 697 kg/(mol Ti-h) and with 95% selectiv-
ity for 1-hexene at 0.5 MPa of ethylene and 80 ◦C. Secondly, the
trimethylsilyl on the 5-position of thienyl has more steric effect in
indenyl complex than in Cp complex, and this lead to the productiv-
ity catalyzed by CS9 is lower than that by CS8, CS10 with a methyl
on the 3-position of the thienyl group has more steric hindrance
effect and gets very low productivity and selectivity for 1-hexene.
Acknowledgments
We gratefully acknowledge financial support from the National
Natural Science Foundation of China (No. 20372022) and National
Basic Research Program of China (2005CB623801).
Appendix A. Supplementary data
3.4.2.2. The influence by the experiment conditions. From Table 4,
it can be seen that a marked character of trimerization results by
these indenyl complexes are that the productivity of 1-hexene is
increasing with the temperature rising during 0–80 ◦C. This is dif-
ferent from Cp-thienyl complexes in which the productivity will
descend at 80 ◦C (Table 3). This means the trimerization active
species of the indenyl complexes can keep stable even at 80 ◦C.
For the indenyl complexes with the electrophilic of the phenyl,
the titanium is more lack of electron, and more favor to coordinate
with sulfur atom. At low temperature, the coordination between
sulfur atom and the titanium cation is tighter, so the productivity is
low although obtaining a high selectivity relatively. While increas-
ing the temperature to 80 ◦C, the coordination is weak enough and
show the most proper hemilabile behavior for ethylene trimeriza-
tion. As for Cp complexes, the most proper hemilabile behavior for
ethylene trimerization is at 30 ◦C, while, at 80 ◦C, this coordination
is too loose to catalyze ethylene trimerization effectively, which
make the productivity decreased. This is also agreement with Hes-
sen’s “Cp-bridge-arene” system, in which the productivity declines
greatly at 80 ◦C [9]. The fact shows that the indenyl-thienyl com-
plexes exhibit higher temperature resistance.
Supplementary data associated with this article can be
found, in the online version, at http://dx.doi.org/10.1016/j.molcata.
2014.02.010.
References
[1] I.A. Husseina, T. Hameed, F.B.A. Sharkha, K. Mezghanib, Polymer 44 (2003)
4665–4672.
[2] T. Hussein, J. Hameed, Appl. Polym. Sci. 97 (2005) 2488–2498.
[3] D.S. Mcguinness, Chem. Rev. 111 (2011) 2321–2341.
[4] T. Agapie, Coord. Chem. Rev. 255 (2011) 861–880.
[5] R.M. Manyik, W.E. Walker, T.P. Wilson, J. Catal. 47 (1977) 197–209.
[6] J.W. Freeman, J.L. Buster, R.D. Knudsen, US 5856257 (1999).
[7] Y. Araki, H. Nakamura, Y. Nanba, T. Okanu, US 5856612 (1999).
[8] P.J.W. Deckers, B. Hessen, J.H. Teuben, Angew. Chem. Int. Ed. 40 (2001)
2516–2519.
[9] P.J.W. Deckers, B. Hessen, J.H. Teuben, Organometallics 21 (2002) 5122–5135.
[10] E. Otten, A.A. Batinas, A. Meetsma, B. Hessen, J. Am. Chem. Soc. 131 (2009)
5298–5312.
[11] C. Aneds, S.B. Harkins, S. Murtuza, K. Oyler, A. Sen, J. Am. Chem. Soc. 123 (2001)
7423–7424.
[12] R.A. Müller, H. Tsurugi, T. Saito, M. Yanagawa, S. Oda, K. Mashima, J. Am. Chem.
Soc. 131 (2009) 5370–5371.
[13] A. Carter, S.A. Cohen, N.A. Cooley, A. Murphy, J. Scutta, D.F. Wass, Chem. Com-
mun. (2002) 858–859.
[14] D.S. McGuinness, P. Wasserscheid, W. Keim, C. Hu, U. Englert, J.T. Dixon, C.
Grove, Chem. Commun. (2003) 334–335.
[15] D.S. McGuinness, P. Wasserscheid, W. Keim, D. Morgan, J.T. Dixon, A. Bollmann,
H. Maumela, F. Hess, U. Englert, J. Am. Chem. Soc. 125 (2003) 5272–5273.
[16] A. Bollmann, K. Blann, J.T. Dixon, F.M. Hess, E. Killian, H. Maumela, D.S.
McGuinness, D.H. Morgen, A. Meveling, S. Otto, M. Overett, A.M.Z. Slawin, P.
Wasserscheid, S. Kuhlmann, J. Am. Chem. Soc. 126 (2004) 1471214713.
[17] S. Licciulli, K. Albahily, V. Fomitcheva, I. Korobkov, S. Gambarotta, R. Duchateau,
Angew. Chem. Int. Ed. 50 (2011) 2346–2349.
Obviously, with the temperature increasing, the selectivity of
the 1-hexene catalyzed by CS9 and CS10 is declined, just similar to
that by Cp complexes. It is noteworthy that even at 80 ◦C, the selec-
tivity is still up to 95%, especially for CS8. This favorable outcome
is the result of introducing the indenyl ring replacing the Cp ring.
From Table 4, it also can be seen that raising the ethylene
pressure, all productivities catalyzed by indenyl complexes are
enhanced, the same as that by Cp complexes.
[18] I. Vidyaratne, G.B. Nikiforov, S.I. Gorelsky, S. Gambarotta, R. Duchateau, I.
Korovkov, Angew. Chem. Int. Ed. 48 (2009) 6552–6556.
[19] A. Jabri, C.B. Mason, Y. Sim, S. Gambarotta, T.J. Burchell, R. Duchateau, Angew.
Chem. Int. Ed. 47 (2008) 9717–9721.
[20] K. Albahily, Y. Shaikh, Z. Ahmed, I. Korobkov, S. Gambarotta, R. Duchateau,
Organometallics 30 (2011) 4159–4164.
4. Conclusions
In summary, a series of half-sandwich Cp(Ind)-bridge-thienyl
titanium complexes as catalytic precursors for catalyzing ethylene
trimerization have been synthesized. CS8 with methyl at the 5-
position of the thienyl group was determined by X-ray diffraction
analysis, which shows that in the solid state the sulfur atom in the
thienyl group do not coordinate with the metal.
In the present of cocatalyst MAO, the Cp-bridge-thienyl titanium
complexes show high activity and selectivity producing 1-hexene.
The complexes with the bridge group C[cylco-(CH2)5] are more
active than that with CMe2, and the complexes with bulk sub-
stituents on 5-position of thienyl are more active than that with
small substituents. The best activity is obtained at 30 ◦C.
[21] H. Hagen, Ind. Eng. Chem. Res. 45 (2006) 3544–3551.
[22] S. Tobisch, T. Ziegler, Organometallics 22 (2003) 5392–5405.
[23] J.A. Suttil, D.S. McGuinness, S.J. Evans, Dalton Trans. 39 (2010) 5278–5285.
[24] Y. Kim, J. Park, S. Yoon, S.B. Park, Trans. Met. Chem. 37 (2012) 439–444.
[25] Y. Suzuki, S. Kinoshita, A. Shibahara, S. Ishii, K. Kawamura, Y. Inoue, T. Fujita,
Organometallics 29 (2010) 2394–2396.
[26] J. Huang, T. Wu, Y. Qian, Chem. Commun. (2003) 2816–2817.
[27] K.J. Stone, D. Little, J. Org. Chem. 49 (1984) 1849–1853.
[28] H.G. Alt, M. Jung, G. Kehr, J. Organomet. Chem. 562 (1998) 153–181.
[29] L.I. Belen’kii, A.P. Yakubov, Tetrahedron 40 (1984) 2471–2477.
[30] M.A. Keegstra, L. Brandsma, Synthesis 11 (1988) 890–891.
[31] L. Albertin, C. Bertarelli, M.C. Gallazzi, S.V. Meille, S.C. Capelli, J. Chem. Soc.,
Perkin Trans. 10 (2002) 1752–1759.
[32] SADABS, Bruker Nonius Area Detector Scaling and Absorption Correction-
V2.05, Bruker AXS Inc., Madison, WI, 1996.

30
Y. Zhang et al. / Journal of Molecular Catalysis A: Chemical 387 (2014) 20–30
[33] G.M. Sheldrick, SHELXTL 5.10 for Windows NT, Structure Determination Soft-
ware Programs, Bruker Analytical X-ray Systems, Inc., Madison, WI, 1997.
[34] SAINT, Version 6.02, Bruker AXS Inc., Madison, WI, 1999.
[35] G.M. Sheldrick, SHELXS-97, Program for the Solution of Crystal Structures, Uni-
versity of Gottingen, Germany, 1990.
[37] Y. Zhang, H. Ma, J. Huang, J. Mol. Catal. A: Chem. 373 (2013) 85–95.
[38] M. Enders, R. Rudolph, H. Pritzkow, Chem. Ber. 129 (1996) 459–463.
[39] Q. Huang, Y. Qian, G. Li, Trans. Metal. Chem. 15 (1990) 483–485.
[40] J. Huang, Y. Zhang, Q. Huang, Y. Qian, Inorg. Chem. Commun. 2 (1992) 104–106.
[41] Y. Qian, G. Li, W. Chen, B. Li, X. Jin, J. Organomet. Chem. 373 (1989) 185–191.
[42] Y. Qian, J. Huang, J. Yang, A.S.C. Chan, W. Chen, X. Chen, G. Li, X. Jin, Q. Yang, J.
Organomet. Chem. 547 (1997) 263–279.
[36] G.M. Sheldrick, SHELXL-97, Program for the Refinement of Crystal Structures,
University of Gottingen, Germany, 1997.