Zirconocene-catalyzed dimerization of 1-hexene: Two-stage activation and structure-catalytic performance relationship

Article abstract of DOI:10.1016/j.catcom.2016.02.013

The dimerization of 1-hexene was catalyzed by ten zirconocenes. High catalytic productivity was achieved via a two-step activation process, namely, the treatment of the zirconocene with triisobutylaluminum (TIBA) followed by methylaluminoxane (MAO). The z

Full text of DOI:10.1016/j.catcom.2016.02.013

Catalysis Communications 79 (2016) 6–10
Contents lists available at ScienceDirect
Catalysis Communications
journal homepage: www.elsevier.com/locate/catcom
Short communication
Zirconocene-catalyzed dimerization of 1-hexene: Two-stage activation
and structure–catalytic performance relationship
a
a
a,b
Ilya E. Nifant'ev a,b, , Alexey A. Vinogradov , Alexander A. Vinogradov , Pavel V. Ivchenko
⁎
a
A. V. Topchiev Institute of Petrochemical Synthesis, RAS, Moscow, Russian Federation
M. V. Lomonosov Moscow State University, Department of Chemistry, Moscow, Russian Federation
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 30 January 2016
Received in revised form 12 February 2016
Accepted 13 February 2016
Available online 16 February 2016
The dimerization of 1-hexene was catalyzed by ten zirconocenes. High catalytic productivity was achieved via a
two-step activation process, namely, the treatment of the zirconocene with triisobutylaluminum (TIBA) followed
by methylaluminoxane (MAO). The zirconocene [(C5H4SiMe2)2O]ZrCl2 (10) exhibits a higher productivity and
selectivity to dimer formation than the unsubstituted (C5H5)2ZrCl2 (1) complex. For the ansa-[Z(C5H4)2]ZrCl2
complexes, the catalytic activity increases as the angle between the cyclopentadienyl rings decreases. The dimer-
ization selectivity of 10 reaches 94% when the reaction is performed using Et2AlCl as the chlorine source needed
to form the catalytic species. The possible mechanism of selective α-olefin dimerization is discussed.
© 2016 Elsevier B.V. All rights reserved.
Keywords:
Zirconocenes
Single-site catalysts
Coordination oligomerization
Hydrides
1. Introduction
zirconocene complexes (Scheme 2). The effects of the nature of the η5
ligand (comparison of complexes 1–3), various substituents on the Cp
The ability of zirconocene dichloride Cp2ZrCl2 to catalyze the
transformation of α-olefins (RCH = CH2, Scheme 1) to α-olefin di-
mers, RCH2CH2C(=CH2)R in the presence of a minor excess of
methylaluminoxane (MAO) was discovered in the 1980s [1] and was
studied by Bergman in the late 1990s [2,3]. Several aspects of
metallocene-catalyzed α-olefin dimerization have been reviewed by
Janiak [4,5].
ligand of unbridged zirconocene dichlorides (comparison of complexes
4 and 5), and the geometries of the Cp2Zr moiety (comparison of the
ansa-zirconocenes 6–10) on the catalytic productivity and selectivity
in α-olefin dimerization were evaluated experimentally. 1-Hexene
was used as the model substrate.
The influence of the structure of zirconocene catalyst on oligomeri-
zation in the presence of significant excess of MAO has been studied
earlier [6–8]. Although a thorough investigation of α-olefin dimeriza-
tion is important, a systematic, comparative study of the catalytic
performances of various zirconocene complexes in selective α-olefin
dimerization has not yet been performed. In addition to Cp2ZrCl2 (1)
[1–3,9], only (n-BuC5H4)2ZrCl2 [10], (i-PrC5H4)2ZrCl2 [11], (EtC5H4)2ZrCl2,
(tert-BuC5H4)2ZrCl2 [6,12] and (Me3SiC5H4)2ZrCl2 [12] have been
studied. The results indicate that monosubstituted zirconocene
dichlorides are less selective than 1 in this type of dimerization
reaction.
2. Experimental
2.1. Zirconocene dichloride preparation
Zirconocene dichlorides 2 [13], 3 [14], 4 [15], and 5 [16]; the bridged
ligands; and ansa-zirconocene complexes 6 [17,18], 7 [19], 9 [20], and
10 [21] were synthesized according to previously reported procedures.
The preparation of complex 5 was difficult due to the low thermal sta-
bility and oxidizability of the initial ligand, cyclopentadienyl benzene
11, which made its isolation and purification complicated. Therefore,
an alternative method for the synthesis of 11 was developed using the
Shapiro reaction. To synthesize 5 from 11, the previously published
procedure [16] was markedly modified. Likewise, the preparation of
zirconocene dichloride 8 included a newly developed synthesis
procedure for CpCMe2CMe2Cp 12. Specifically, 6,6-dimethylfulvene
was reacted with biphenyl sodium, an effective single-electron
reducing agent, at low temperature. The experimental details for
the synthesis of 5, 11 and 12 are described in the Supplementary
Information (SI).
This paper presents the first attempt to study the relationship be-
tween the zirconocene dichloride structure and its catalytic properties
in α-olefin dimerization using a series of bridged and unbridged
⁎
Corresponding author at: A. V. Topchiev Institute of Petrochemical Synthesis, RAS,
Moscow, Russian Federation.
E-mail address: inif@org.chem.msu.ru (I.E. Nifant'ev).
http://dx.doi.org/10.1016/j.catcom.2016.02.013
1566-7367/© 2016 Elsevier B.V. All rights reserved.

I.E. Nifant'ev et al. / Catalysis Communications 79 (2016) 6–10
7
the protocol described in [3] (0.05% of 1, AlMAO/Zr = 10, neat 1-
hexene). After reaction at 20 °C for 1 h, the reaction mixture contains
only the traces of the product. At 60 °C, the conversion reaches 6% in
1 h. Then, the reaction accelerates, and the 1-hexene conversion reaches
90% in 12 h (TOF of 300 h−1), which is consistent with previously re-
ported data [2,3]. Bergman showed that the induction period is reduced
and the integral catalyst productivity increases up to a TOF of 1800 h−1
when Cp2ZrHCl is used instead of 1 [3]. Assuming that zirconocene hy-
drides can be readily generated by the reaction of triisobutylaluminum
(TIBA) with zirconocene dichloride [22], a two-step catalyst preparation
procedure was developed. In the first step, the zirconocene dichloride
was treated with 20 eq. of TIBA to generate Zr–H species (20 min at
60 °C). Although this reaction proceeds in neat 1-hexene, no 1-hexene
dimerization is observed. In the second step, MAO (10 eq. relative to
the zirconocene pre-catalyst) was added to the reaction mixture, and
the exothermic reaction started immediately. After 1 h, the conversion,
selectivity, and TOF of this benchmark experiment are 76%, 84%, and
1300 h−1, respectively (run 1, Table 1).
To try and improve the dimerization selectivity, the AlMAO/Zr ratio
was varied in the experiments. The catalyst prepared with 1–2 eq. of
MAO is nearly inactive; however, increasing the AlMAO/Zr ratio substan-
tially leads to the formation of considerable amounts of oligomers in the
reaction mixture. For example, 1-hexene dimerization in the presence
of 100 eq. of MAO gives the hexene dimer with 73% selectivity (run 2,
Table 1).
Previously, it was reported that adding Et2AlCl to the reaction
mixture improves the selectivity of α-olefin dimerization [9]. The
effectiveness of this approach when used in conjunction with a catalyst
prepared by the two-step procedure was determined. The results show
that the dimerization reaction is slightly more selective in the presence
of 1 eq. of Et2AlCl (87% selectivity, run 3, Table 1) than in its absence.
However, increasing the amount of Et2AlCl does not further improve
the selectivity, and it results in a sharp decrease in the activity.
Scheme 1. α-Olefin dimerization [1–3].
2.2. Oligomerization of 1-hexene
2.2.1. Typical procedure for determining the catalytic activity in 1-hexene
dimerization
1-Hexene (25 mL, 200 mmol) and a 1 M TIBA solution in hexane
(2 mL, 2 mmol) were mixed in a two-necked flask prefilled with
argon, which was then placed in a thermostated bath with diethylene
glycol. After maintaining the external bath at 60 °C for 5 min, a solution
or suspension of zirconocene dichloride (0.1 mmol) in toluene (6 mL)
was added to the flask. After 20 min of stirring, a 1.5 M MAO solution
(0.66 mL, 1 mmol) was added to the mixture. Samples were removed
from the flask and analyzed by NMR and GC after 15 min, 30 min and
1 h.
2.2.2. Reaction mixture analysis
The integrated intensities of the vinyl proton NMR signals of
1-hexene and the reaction products were compared. GC analysis was
performed using a capillary column and FID. The GC curves were
calibrated using 1-hexene dimer, trimer, and tetramer samples. For
experimental details, see the SI.
2.2.3. Optimization of the 2-butyl-1-octene preparation
Oligomerization was catalyzed by the activated 10 complex
(0.1 mmol) in 1-hexene (25 mL, 200 mmol) at 60 °C. The optimal
activation component ratio was determined experimentally to be
2 mmol of TIBA, 1 mmol of MAO, and 0.2 mmol of Et2AlCl. The reaction
time needed to achieve 100% conversion was 4 h (for 1H NMR spectra of
the reaction mixture see SI), and the isolated 2-butyloctene yield was
15.8 g (94%, distillation, b.p. 77–78 °C/8 Torr).
3.2. Comparison of 1-hexene oligomerization by different catalysts
3. Results and discussion
The catalytic performances of zirconocene complexes 1–10,
pre-treated using the two-step activation procedure, were studied at
defined conditions. The experimental results are summarized in Table 1.
Higher oligomers are predominantly formed in the presence of the
moderately active cyclopentadienyl-indenyl complex 2; the dimer
content is less than 40% (run 4, Table 1). Similarly, in the case of the
bis(indenyl) complex 3, higher 1-hexene oligomers (Mw = 3900 Da)
are formed (run 5, Table 1). In a previous study, considerable amounts
of higher oligomers were formed in the presence of di-tert-butyl
zirconocene dichloride 4, even with a minor excess of MAO [12]. The
results of this study are consistent with the previously published results
(run 6, Table 1). Zirconocene dichloride 5 also exhibits low selectivity to
the dimer, but it is significantly more active than 4 (run 7, Table 1).
Presumably, this difference is due to the additional stabilization of
the catalytic center in 5 by the π-donor properties of the phenyl
substituents.
3.1. Optimization of the catalytic procedure
Before starting the comparative study, appropriate conditions for the
catalytic experiments were determined. The long induction period
for α-olefin dimerization reactions catalyzed by Cp2ZrCl2/МАО was
considered. Initially, 1-hexene dimerization was performed following
The target compound 2-butyl-1-octene is the major product of
1-hexene oligomerization catalyzed by the bis-cyclopentadienyl ansa-
complexes 6–10 (runs 8–12, Table 1). The catalyst productivity,
selectivity to 2-butyl-1-octene formation, and 2-hexene content in the
reaction mixture depend substantially on the type and length of the
bridge between the cyclopentadienyl rings. Complexes 8–10 with the
longest bridges have the highest productivities, even higher than that
of zirconocene dichloride 1. However, of these metallocene catalysts,
only complex 10 catalyzes the formation of 2-butyl-1-octene with a
higher selectivity than 1, making it a promising catalyst for selective
dimerization.
Because the activity of 10 is higher than that of 1, it interacts with
Et2AlCl more effectively to improve the dimerization selectivity. When
Scheme 2. Zirconocene dichlorides studied in the dimerization of 1-hexene.

8
I.E. Nifant'ev et al. / Catalysis Communications 79 (2016) 6–10
Table 1
Product distribution in the zirconocene-catalyzed oligomerization of 1-hexene and selected geometric parameters of zirconocene dichlorides 1–10.
Run
Cat.
Product distribution
Selected geometric parameters (X-ray)
Conv., %
Dimer, wt.% in
products
Olig. wt.% in
products
2-Hexenes wt.%
in products
Dimer, wt.% in
reaction mixture
α
β
γ
d(Zr-Cp)
CGAa
Ref.
1
2
3
4
5
6
7
8
1
76
91
62
55
10
5
80
38
45
86
94
84.4
72.7
87.0
38.2
12.2
22.6
9.8
3.4
4.7
3.2
4.7
~0.3
5.0
b1
8.5
5.6
4.9
2.7
64.1
66.2
53.9
21.0
–
2
40
31.7
38.9
70.3
72.1
129.5
–
–
53.5
–
–
97.0
–
–
2.20
–
–
92
–
–
[23]
–
–
1b
1c
2
57.1
No data
128.3
128.7
129.3
116.7
125.4
125.0
131.2
3
4
5
6
7
8
9
Mw 3900d
40.0
59.3
54.2
56.0
71.1
60.1
55.9
51.2
94.7
94.3
95.2
100.3
98.0
97.5
95.7
98.4e
98.7
–
2.23
2.22
2.22
2.19
2.20
2.19
2.21
–
–
–
108
100
95
90
[24]
[25]
[16]
[26]
[19]
[27]
[20]
55.0
50.0
8.1
7.9
13.3
20.6
49.0
83.4
86.5
81.8
9
10
11
76.7
12
13
14
10
95
82
100
88.0
91.9
93.7
9.0
5.3
2.8
3.0
2.8
3.5
83.6
75.3
93.7
130.8
–
–
51.1
–
–
2.20
–
–
89
–
–
[28]
–
–
10c
10f
–
Catalyst activation procedure: 1) TIBA, AlTIBA:Zr = 20:1, 20 min; 2) MAO, AlMAO:Zr = 10:1.
Conditions: neat 1-hexene, 0.05 mol% Zr, 60 °C, 1 h.
a
Coordination gap aperture.
AlMAO:Zr = 100:1.
Et2AlCl:Zr = 1:1.
By GPC.
Two structurally independent molecules in the crystal cell.
Et2AlCl:Zr = 2:1, 4 h.
b
c
d
e
f
TIBA:MAO:Et2AlCl:10 = 20:10:2:1, the conversion and dimerization
selectivity reach 100% and 94%, respectively.
Zr–C bond of the alkylzirconocene species, followed by β-hydride trans-
fer to the monomer (M1, Scheme 3). Bergman proposed a different
mechanism for α-olefin dimerization catalyzed by 1 in the presence of
1 eq. of MAO [2]. Bergman's scheme accounts for the role of MAO as
an alkylating agent and as an anion moiety that stabilizes the active
species. It also assumes that the zirconocene hydride Cp2ZrH catalytic
species reacts successively with two alkene molecules to form an alkene
dimer after β-hydride transfer to Zr. Further research proved that chlo-
rine must be present in the reaction mixture to achieve high dimeriza-
tion selectivity; Bergman demonstrated that treating the olefin with
Cp2ZrMe2 and 1 eq. of MAO in the absence of Cl leads exclusively to
the formation of higher oligomer products [3].
Based on the two-step catalyst preparation described here, it is
suggested that the catalytic species is a μ-hydride Zr–Al cluster formed
by the interaction between the zirconocene dichloride and TIBA [22,
32–34]. The two simplest complexes of this type, effectively stabilized
by ClMAO [35], are the zirconium–aluminum hydride-chloride complex
A and dihydride complex B (Scheme 3, II). Clearly, the proportion of
the dihydride B should increase with increasing MAO concentration
because MAO captures Cl.
It is assumed that the zirconium–aluminum hydride-chloride complex
A catalyzes the α-olefin dimerization reaction (Scheme 3, Line III). On
the first stage n-alkyl complexes A1 (1,2-insertion) or sec-alkyl com-
plexes A1’ (2,1-insertion) are formed, the latter via β-hydride transfer
to Zr forms 2-hexenes, inert byproducts under current reaction condi-
tions. This process is irreversible. On the second stage isoalkyl complex
A2 is formed. Presumably, due to the presence of the μ-Cl in A, the Zr
center in A is less electrophilic and more closed than the Zr center in
B. The insertion of a third α-olefin molecule into A2 is less probable
than the β-hydride transfer to Zr, which leads to the predominant for-
mation of the α-olefin dimer. Thus, the μ-Cl induces the release of the
α-olefin dimer from A. The intermediate B is electrophilic and sterically
open enough to oligomerize an α-olefin. The observed increase in the
dimerization selectivity after treating the reaction mixture with Et2AlCl
3.3. Relationship between the catalyst activity and zirconocene dichloride
geometry
The key geometric parameters of the studied complexes are
summarized in Table 1. A comparison of the geometric parameters of
the ansa-zirconocene dichlorides 6–10 shows that the bridge type has
a minor influence on the CEN–Zr–CEN (α) and Cl–Zr–Cl (γ) angles.
The coordination gap aperture (CGA) concept [29,30] is widely used to
draw correlations between the geometry of ring-substituted
metallocene complexes and their catalytic activity. We measured the
CGA value for complexes 1 and 6–10, which did not contain additional
substituents in cyclopentadienyl rings (see Table 1) and showed that
CGA correlates with the dihedral angle β between the cyclopentadienyl
rings. An increase in the bridge length leads to a decrease in the dihedral
angle β between the cyclopentadienyl rings. This dihedral angle is
correlated to the catalytic activity of complexes 6–10; as β decreases,
the productivity of the metallocene catalyst increases. Of the
compounds studied, complex 10, which has the longest bridge, has
the highest productivity and selectivity in the 1-hexene dimerization
reaction. Complex 10 is conformationally flexible; another possible
reason of the high selectivity of catalytic particles derived from 10 is
the coordination of MAO to the bridging oxygen atom with “freezing”
of the catalyst geometry.
3.4. Mechanism of alkene dimerization
Two decades ago, Negishi found [31] that zirconocene dichloride 1
catalyzed α-olefin dimerization in the presence of Me3Al. However, in
the absence of МАО, the reaction proceeded very slowly, and the
selectivity was very low. Negishi proposed that the catalytic mechanism
involves the sequential insertion of two RCH = CH2 molecules into the

I.E. Nifant'ev et al. / Catalysis Communications 79 (2016) 6–10
9
Scheme 3. α-Olefin dimerization mechanism.
can be explained by the transformation of B into A. The formation of
Acknowledgments
higher oligomer products in the presence of Cp2ZrMe2 and 1 eq. of
MAO observed by Bergman could be due to the formation of the
Cl-free hydride-methyl complex C, which is analogous to complex B in
this study (Scheme 3, I).
Financial support by the Russian Science Foundation (grant no. 15-
13-00053) is gratefully acknowledged.
Thus, the α-olefin dimerization selectivity strongly depends on
the proportions of the catalytic species A and B. The proposed
mechanism undoubtedly requires additional experimental and
theoretical validation, which is currently being performed in our
laboratory.
Appendix A. Supplementary data
Supplementary data to this article can be found online at http://dx.
doi.org/10.1016/j.catcom.2016.02.013.
References
4. Conclusions
[1] L.H. Slaugh, G.W. Schoenthal, US4658078 (1987)
[2] J. Christoffers, R.G. Bergman, J. Am. Chem. Soc. 118 (1996) 4715–4716.
[3] J. Christoffers, R.G. Bergman, Inorg. Chim. Acta 270 (1998) 20–27.
[4] C. Janiak, F. Blank, Macromol. Symp. 236 (2006) 14–22.
[5] C. Janiak, Coord. Chem. Rev. 250 (2006) 66–94.
In this study, an effective two-step procedure for generating α-olefin
dimerization catalysts was developed, and the catalytic performances
of various zirconocene complexes in 1-hexene dimerization were
compared. It was determined that the catalytic activity of the ansa-
zirconocene dichlorides Z(η5-C5H4)2ZrCl2 increases as the length of Z
increases. This geometric change is accompanied by a decrease in the di-
hedral angle between the η5-cyclopentadienyl rings. The synthetically
accessible bis(cyclopentadienyl) complex 10, which contains an
extended–SiMe2OSiMe2–bridge, exhibits higher productivity and selec-
tivity to the dimerization product than the zirconocene dichloride 1. By
adding 2 eq. of Et2AlCl to the catalyst prepared from 10, 100% conver-
sion and 94% selectivity to the 2-butyl-1-octene product can be
achieved.
[6] C. Janiak, K.C.H. Lange, P. Marquardt, Macromol. Rapid Commun. 16 (1995)
643–650.
[7] C. Janiak, K.C.H. Lange, P. Marquardt, R.-P. Krüger, R. Hanselmann, Macromol. Chem.
Phys. 203 (2002) 129–138.
[8] C. Janiak, K.C.H. Lange, P. Marquardt, J. Mol. Catal. A Chem. 180 (2002) 43–58.
[9] K. Takeuchi, S. Fujikawa, US2011251445 (2011).
[10] Y.V. Kissin, F.C. Schwab, J. Appl. Polym. Sci. 111 (2009) 273–280.
[11] W.J. Heilman, Y.H. Jois, A.R. De Kraker, W. Song, US2011178348 (2011).
[12] S. Fujikawa, K. Yokota, M. Okano, M. Tsuji, US2011207977 (2011).
[13] M.A. Schmid, H.G. Alt, W. Milius, J. Organomet. Chem. 514 (1996) 45–49.
[14] N. Piccollrovazzi, P. Pino, G. Consiglio, A. Sironi, M. Moret, Organometallics 9 (1990)
3098–3105.
[15] A.I. Licht, H.G. Alt, J. Organomet. Chem. 684 (2003) 91–104.

10
I.E. Nifant'ev et al. / Catalysis Communications 79 (2016) 6–10
[16] M.P. Thornberry, N.T. Reynolds, P.A. Deck, F.R. Fronczek, A.L. Rheingold, L.M. Liable-
Sands, Organometallics 23 (2004) 1333–1339.
[17] I.E. Nifant'ev, P.V. Ivchenko, M.V. Borzov, J. Chem. Res. (1992) 162–163 (S).
[18] I.E. Nifant'ev, P.V. Ivchenko, Organometallics 16 (1997) 713–715.
[19] C.S. Bajgur, W.R. Tikkanen, J.L. Petersen, Inorg. Chem. 24 (1985) 2539–2546.
[20] K.H. Thiele, Ch. Schliessburg, K. Baumeister, K. Hassler, Z. Anorg. Allg. Chem. 622
(1996) 1806–1810.
[27] M. Buhl, G. Hopp, W. Philipsborn, S. Beck, M.H. Prosenc, U. Rief, H.H. Brintzinger,
Organometallics 15 (1996) 778–785.
[28] S. Ciruelos, T. Cuenca, P. Gomez-Sal, A. Manzanero, P. Royo, Organometallics 14
(1995) 177–185.
[29] C. Janiak, U. Versteeg, K.C.H. Lange, R. Weimann, E. Hahn, J. Organomet. Chem. 501
(1995) 219–234.
[30] C. Janiak, K.C.H. Lange, U. Versteeg, D. Lentz, P.H.M. Budzelaar, Chem. Ber. 129
[21] J. Graeper, G. Paolucci, R.D. Fischer, J. Organomet. Chem. 501 (1995) 211–218.
[22] D.E. Babushkin, V.N. Panchenko, M.N. Timofeeva, V.A. Zakharov, H.H. Brintzinger,
Macromol. Chem. Phys. 209 (2008) 1210–1219.
(1996) 1517–1529.
[31] D.Y. Kondakov, E. Negishi, J. Am. Chem. Soc. 117 (1995) 10771–10772.
[32] S.M. Baldwin, J.E. Bercaw, H.H. Brintzinger, J. Am. Chem. Soc. 130 (2008)
17423–17433.
[23] J.Y. Corey, X.-H. Zhu, L. Brammer, N.P. Rath, Acta Crystallogr. Sect. C: Cryst. Struct.
Commun. 51 (1995) 565–567.
[24] T. Repo, M. Klinga, I. Mutikainen, Y. Su, M. Leskelä, M. Polamo, Acta Chem. Scand. 50
(1996) 1116–1120.
[33] K.P. Bryliakov, E.P. Talsi, N.V. Semikolenova, V.A. Zakharov, J. Brand, C. Alonso-
Moreno, M. Bochmann, J. Organomet. Chem. 692 (2007) 859–868.
[34] L.V. Parfenova, P.V. Kovyazin, I.E. Nifant'ev, L.M. Khalilov, U.M. Dzhemilev, Organo-
metallics 34 (2015) 3559–3570.
[25] R.A. Howie, G.P. McQuillan, D.W. Thompson, G.A. Lock, J. Organomet. Chem. 303
(1986) 213–220.
[35] M.S. Kuklin, J.T. Hirvi, M. Bochmann, M. Linnolahti, Organometallics 34 (2015)
[26] T. Koch, S. Blaurock, F.B. Somoza, A. Voigt, R. Kirmse, E. Hey-Hawkins, Organometal-
lics 19 (2000) 2556–2563.
3586–3597.