T-BuC5H43Nd: A triscyclopentadienyl rare earth compound as non-classical isoprene polymerization pre-catalyst

Article abstract of DOI:10.1016/j.jorganchem.2013.06.035

eptic tris(cyclopentadienyl) Cp′ ₃Nd (Cp′ = C ₅H₄t-Bu) complex has been unprecedentedly considered as potential pre-catalyst for isoprene polymerization. The X-ray structural analysis establishes a monomeric non-solvated nature for this compound. Upon activation with appropriate borate/aluminium co-catalyst combinations, Cp0 3Nd affords in good yields polyisoprene more than 95% cis-regular.

Full text of DOI:10.1016/j.jorganchem.2013.06.035

Journal of Organometallic Chemistry 743 (2013) 139e146
Contents lists available at SciVerse ScienceDirect
Journal of Organometallic Chemistry
journal homepage: www.elsevier.com/locate/jorganchem
(t-BuC5H4)3Nd: A triscyclopentadienyl rare earth compound as
non-classical isoprene polymerization pre-catalyst
*
Ines Rodrigues, Tan Yong Xue, Pascal Roussel, Marc Visseaux
Unité de Catalyse et Chimie du Solide (UCCS), UMR CNRS 8181, Université de Lille Nord de France, USTL-ENSCL, Bat C7, BP 90108, 59652 Villeneuve d’Ascq,
France
a r t i c l e i n f o
a b s t r a c t
Homoleptic tris(cyclopentadienyl) Cp03Nd (Cp0 ¼ C5H4t-Bu) complex has been unprecedentedly consid-
ered as potential pre-catalyst for isoprene polymerization. The X-ray structural analysis establishes a
monomeric non-solvated nature for this compound. Upon activation with appropriate borate/aluminium
co-catalyst combinations, Cp03Nd affords in good yields polyisoprene more than 95% cis-regular.
Ó 2013 Elsevier B.V. All rights reserved.
Article history:
Received 23 March 2013
Received in revised form
18 June 2013
Accepted 25 June 2013
Dedicated to Professor André Mortreux at
the occasion of his 70th birthday.
Keywords:
Rare earths
Neodymium
cis-Polyisoprene
Tris(cyclopentadienyl)
Polymerization catalysis
Stereoselective
1. Introduction
investigations of Wilkinson dedicated to the reactivity of the Cp3Ln
derivatives [6], it was shown that Bronsted acids react with tris(cy-
Tris(cyclopentadienyl) derivatives of the rare earths represent a
model of thermodynamically stable compounds, since the set of
three cyclopentadienyl ligands is known as particularly well suited to
complete the steric/electronic features of a rare earth metal [1]. De
facto, tris(cyclopentadienyl) complexes have often been observed as
the main e unexpected, but yet e isolated products of syntheses that
were intended to yield metallocene-like organolanthanides [2]. The
simplest representative in this frame is the Cp3Ln (Cp ¼ C5H5,
Ln ¼ rare earth metal), which was synthesized in the early times of
organometallic chemistry [3]. With the lightest “early” rare earths,
which are also the biggest ones, it was frequent to isolate adducts,
even with substituted cyclopentadienyls [4]. Due to their ability to
easily sublimate and their stability, such compounds were mainly
investigated as potential precursors for OMCVD [5]. Actually, because
of an apparent steric saturation, tris(cyclopentadienyl) derivatives
may be seen as compounds that are relatively inert toward the
activation of small molecules, which is generally considered as the
first stage of a catalytic process. However, after the pioneering
clopentadienyl) complexes to afford the corresponding bis(cyclo-
pentadienyl) derivatives and cyclopentadiene [7]. Later,
combinations of Cp3Ln compounds with NaH were described as
efficient for a number of organic reactions [8]. Putative [Cp2LnH]
active species, arising from the displacement of one Cp ring by NaH,
were postulated by the authors, on the basis of previous studies by
Schumann, who initially proposed the cleavage of tris(cyclopenta-
dienyl)lanthanide complexes with alkylating reagents to afford
alkylbis(cyclopentadienyl)lanthanides. This method was useful in
the early series since the latter complexes, prone to act as potential
catalysts, are not accessible via the classical alkylation reaction from
bis(cyclopentadienyl)lanthanide halides [9]. To our knowledge,
Evans has reported a few examples of polymerization induced with
tris(cyclopentadienyl) f-element complexes in the (C5Me5)3M series
with Nd, Sm, and U. But such compounds act as single component
initiators to promote, via h5 1 shift, the polymerization of ethylene,
eh
and this peculiar reactivity is attributed to overcrowding of the
complexes [10]. The same author showed that (C5Me5)3Sm could also
react for similar reasons with Lewis acids like borane or alkylalu-
minium, generally used to activate pre-catalysts in polymerization
reactions, to displace one pentamethylcyclopentadienyl ligand [11].
* Corresponding author. Tel.: þ33 3 20 33 64 83; fax: þ33 3 20 43 65 86.
E-mail address: marc.visseaux@ensc-lille.fr (M. Visseaux).
0022-328X/$ e see front matter Ó 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.jorganchem.2013.06.035

140
I. Rodrigues et al. / Journal of Organometallic Chemistry 743 (2013) 139e146
Since in general numerous lanthanides-based polymerization
ray analysis. The 1H NMR spectrum was compatible with the for-
mula of a non-solvated Cp03Y complex since only signals attributed
catalysts are made of lanthanidocenes and half-lanthanidocenes
[12], and considering that the synthesis of such derivatives may
eventually be hard to achieve because of frequently observed li-
gands scrambling [13], an alternative can thus consist in the use of
the tris(cyclopentadienyl) compounds as potential source of active
species.
In this study, we proposed to investigate the ability of tris(cy-
clopentadienyl) derivatives as pre-catalysts for isoprene polymer-
ization, in combination with an aluminum co-catalyst, and in the
presence of a boron based molecule, which is known to act as an
activator by alkyl abstraction. Two complexes of neodymium, an
element known to give rise to efficient catalysts vs. dienes poly-
merization [14], were chosen in this frame: Cp03Nd (1), and Cp3Nd
(2) (Cp0 ¼ C5H4t-Bu, Cp ¼ C5H5). The X-ray structure of 1, until now
not described in the literature, is briefly presented in a preliminary
section. The results in polymerization are then discussed in details,
including some mechanistic considerations.
to the ligand were present. 1H NMR (C6D6):
d
¼ 6.13 (4H, C5H4), 6.01
(4H, C5H4), 1.17 (18H, t-Bu). This compound was authenticated by a
bulk synthesis carried out from different precursors: Y(BH4)3(THF)3
(0.5 g, 1.43 mmol) and a three-fold excess of KCp0 (0.7 g, 4.37 mmol)
and in toluene (20 mL) at room temperature followed by extraction
with pentane. No crystals could be obtained but the compound
obtained after evaporation of the solvents had a honey-like aspect,
and it finally solidified in 24 h at room temperature in the glove box
(0.51 g, 79% yield). Its 1H NMR spectrum was the exact replication of
the previous one received from the synthesis carried out from YCl3.
Elemental analysis found (calc.) for C27H39Y: C% 69.79 (71.68); H%
8.56 (8.63).
2.3. X-ray structure determination of 1
Purple single crystals suitable for X-ray diffraction of 1 were
grown from pentane. The crystals were carefully chosen using a
stereo zoom microscope supported by a rotatable polarizing stage.
The data were collected at room temperature on Bruker’s KAPPA
2. Material and methods
All operations were performed under dry argon by using
Schlenk techniques. The solvents were dried over sodium/benzo-
phenone ketyl, deoxygenated by distillation, and stored over mo-
lecular sieves (3A) in a glove box. Styrene (99% from Aldrich) and
isoprene (99% from Aldrich) were dried over calcium hydride,
distilled twice, and stored over molecular sieves (3A) in a glove box.
Triisobutylaluminum (Ali-Bu3), tris(pentafluorophenyl)borane B(C6
F5)3 and trityl tetra(pentafluorophenyl)borate [CPh3][B(C6F5)4]
were purchased from Aldrich and used as received. Dimethylphe-
nylanilinium tetra(pentafluorophenyl)borate [HNMe2Ph][B(C6F5)4]
was purchased from Strem and used as received. Cp3Nd was syn-
thesized and purified by sublimation according to the published
procedure [3]. NMR scale experiments were conducted in a glove
box in tubes equipped with a Teflon valve (YoungÔ). 1H and 13C
NMR spectra of organometallic complexes and of polyisoprenes
were recorded on a Bruker Avance 300 instrument at 300 K in C6D6
(7.16 ppm) and in CDCl3 (7.26 ppm, 77.16 ppm), respectively, with
reference to residual proton or carbon signal in brackets. When
necessary, deuterated THF (1.72 ppm) or toluene (2.09 ppm), were
used as NMR solvents for mechanistic studies. Size exclusion
chromatography of samples was performed in THF as an eluent at
40 ꢀC (1 mL minꢁ1) with a Waters SIS HPLC pump, a Waters 410
refractometer, and Waters Styragel columns (HR2, HR3, HR4, and
HR5E) calibrated with polystyrene standards.
APEX II CCD X8 with graphite monochromated Mo-Ka radiation
ꢀ
(0.71073 A). The crystals were glued to a thin glass fiber using inert
oil and afterward placed in the cold stream of an Oxford Cry-
osystems open-flow nitrogen cryostat with a nominal stability of
0.1 K. The intensity data were processed using Bruker’s suite of data
processing programs [16], and absorption corrections were applied
using SADABS [17]. The crystal structure was solved by charge
flipping method using SUPERFLIP [18] and the data was refined by
full matrix least-squares refinement on
F with anisotropic
displacement parameters for non-H atoms, using CRYSTALS [19].
Crystal data of 1: Molecular formula ¼ C27H39Nd1, Formula
weight
¼
507.85, Crystal system
¼
orthorhombic, space
ꢀ
ꢀ
group ¼ P212121, a ¼ 8.3916 (2) A, b ¼ 16.4447 (4) A, c ¼ 17.1450
3
ꢀ
ꢁ3
,
ꢀ
(5) A, V ¼ 2365.96 (11) A , T ¼ 100 K, Z ¼ 4, Dc ¼ 1.426 g cm
39,722 reflections measured, 5774 independent reflections, 5558
observed reflections [I > 3.0
fit ¼ 1.11.
s
(I)], R1_obs ¼ 0.0138, Goodness of
2.4. Isoprene polymerization
In a typical polymerization (run 9 in Table 1 is given here as an
example), a flask was charged, in a glove box, with the pre-catalyst
(Cp03Nd, 5.1 mg, 10
m
mol Nd). The solvent (toluene, 1 mL) was
added, followed by the co-catalyst (Al(i-Bu)3, 38 L, 150 mol), and
the boron activator ([CPh3][B(C6F5)4], 27.6 mg, 30 mol). At last, the
m
m
m
2.1. Synthesis of Cp03Nd, 1
monomer (isoprene, 3 mL, 30 mmol) was added with a syringe. The
polymerization was conducted outside the glove box, under an
inert atmosphere, at given monitored temperature and running
time. After that time, the viscous mixture was quenched with a few
drops of toluene/HCl and treated with methanol containing 2,6-di-
tert-butyl-4-methylphenol (ca 1.0 wt%) as a stabilizer. The resulting
isolated polymeric material was dried under vacuum at room
temperature to a constant weight. The yield was determined by
gravimetry (m ¼ 1.87 g, yield 91.8%). Stereoselectivities were
determined with 1H and 13C NMR when overlapping of signals as
previously described [20].
Complex 1 (2 g, 5.1 mmol) was synthesized according to the
published procedure [5], with the following modification: the
mixture of reactants was refluxed in THF for 48 h followed with
toluene extraction, evaporation of the toluene and addition of
pentane (30 mL). After filtration, the pentane solution was
concentrated to one third and let to crystallize at room tempera-
ture, yielding finally 1.90 g (73%) of deep blueered crystals. 1H NMR
(C6D6):
d
¼ ꢁ17.7 (18H, t-Bu); ꢁ2.9 (4H, C5H4); 27.6 (4H, C5H4) [15].
Elemental analysis found (calc.) for C27H39Nd: C% 63.79 (63.88); H%
7.73 (7.69).
2.5. NMR monitoring
2.2. Synthesis of Cp03Y
Reaction of 1 with trityl perfluorophenylborate then triisobu-
The same procedure carried out with yttrium (YCl3, 0.5 g,
2.55 mmol; KCp0 [2d], 1.3 g, 8.1 mmol) led to a brownish oily
compound (0.78 g, 67.6% yield) which solidified after a week at
room temperature. This solid was unfortunately not suitable for X-
tylaluminum in C6D6: 3 mg (5.9 mmol) of 1 and 11 mg (11.9 mmol) of
[CPh3][B(C6F5)4] were weighed in the glove box in an NMR tube
equipped with a Teflon valve. At the addition of C6D6 (0.4 mL) the
formation of an oily brown insoluble compound was observed

I. Rodrigues et al. / Journal of Organometallic Chemistry 743 (2013) 139e146
141
Table 1
Selected results of the isoprene polymerization with Cp03Nd (1) as pre-catalyst using different systems.
Runa
Activatorb
Co-catc (Al equiv)
[I]/[Ln]
T (ꢀC)d
t (h)
Yield (%)
3,4/1,4trans/1,4cise
Mn (PDI)f
1
2
3
4
5
6
7
1 TB
2 TB
2 TB
2 TB
2 TB
3 TB
3 TB
3 TB
3 TB
3 TB
5 TB
5 TB
5 TB
5 TB
5 TB
5 TB
5 TB
5 TB
3 B
5
10
10
10
10
15
15
15
15
15
15
15
25
25
25
25
25
e
1000
1000
1000
2000
2000
1000
3000
3000
3000
3000
1000
1000
1000
1000
3000
3000
1000
1000
1000
1000
1000
1000
3000
3000
50
50
RT
50
50
RT
RT
50
RT
50
50
50
50
RT
RT
50
50
e
20
20
96
5
20
72
60
12
72
12
20
20
5 h 30
60
44
17
17
11.0
90.9
9.0
18.2/81.8
15.8/8.0/76.2
16.5/83.5
20.3/2.2/77.5
19.9/80.1
16.8/6.7/76.5
17.9/14.6/67.5
19.3/80.7
14.5/5.5/80.0
16.8/2.4/81.9
3.9/8.4/87.7
9.7/35.5/54.8
6.8/3.3/89.9
3.3/1.2/95.5
6.1/1.5/92.4
3.6/1.3/95.1
8.9/91.1
e
19,800 (2.1)
e
40.4
62.6
36.0
30.0
64.3
91.8
94.8
14.7
16.8
88.5
75.2
60.9
85.6
64.9
Insi
32,000 (2.1)
27,000 (2.3)
e
e
8
e
9g
28,200 (4.4)
e
10g
11g
12
13g
14g
15g
16g
17g,h
18
19
20g
21
22j
23k
24g,k
e
e
14,200 (2.4)
59,100 (2.4)
30,000 (3.5)
e
17,000 (2.1)
e
1 min
20
20
36
20
e
15
15
15
10
10
10
50
50
RT
50
50
RT
78.2
87.3
26.8
0
2.1
71.2
5.1/10.1/84.9
4.9/7.7/87.4
4.0/6.0/90.0
e
24,200 (2.7)
18,700 (2.8)
30,600 (3.5)
e
3 B
3 B
2 TB
2 TB
2 TB
72
10 min
e
e
24.2/17.3/58.5
47,700 (3.8)
a
Toluene 1 mL; addition order: pre-cat/solvent/activator/co-cat/monomer.
b
c
TB ¼ trityl borate ([CPh3][B(C6F5)4]); B ¼ tri(perfluoro)phenylborane (B(C6F5)3).
Ali-Bu3 equivalents.
d
e
RT ¼ room temperature.
The 3,4/1,4 ratio was determined from 1H NMR integration. The cis/trans relative intensities were determined when low overlapping by 1H NMR. In case of overlap of the
Me signals, an average with data obtained from 13C NMR quantitative analysis was considered.
f
Determined from SEC with reference to PS standards.
Addition order: pre-cat/solvent/co-cat/activator/monomer.
g
h
Solvent ¼ pentane.
i
Insoluble.
j
Pre-catalyst Cp3Nd (2).
k
Pre-catalyst Cp03Y synthesized according to Scheme 1.
along with discoloration of the initial light blue tint of 1. 1H NMR
showed the complete consumption of 1, paramagnetic signals
possibility of clear interpretation:
d
ppm ¼ 12.5e13.7 (minor), ꢁ17.1
(major), ꢁ23.2 (major). Signals corresponding to Al(i-Bu)3 were
attributed to a Cp0 ligand attached to Nd:
d
ppm ¼ 11.8 (5.5H, br,
m), ꢁ18.9 (3.5H, s), ꢁ23.3 (9H, s), and diamagnetic signals
compatible with possible isomers of tert-BuC5H4CPh3:
observed as well as same as before.
Reaction of 1 with trityl perfluorophenylborate in THF-d8: 2.5 mg
(4.9 mmol) of 1 and 10 mg (10.8 mmol) of [CPh3][B(C6F5)4] were
d
ppm ¼ 7.40
(o-Ph), 7.05 (m, p-Ph), 6.45 (C5H4), 6.16 (C5H4), 6.09 (C5H4),
4.6 (allylic C5H4), 0.9 (tert-Bu). These diamagnetic signals were also
weighed in the glove box in an NMR tube equipped with a Teflon
valve. At the addition of THF-d8 (0.4 mL), a brown solution formed,
which was become solid within minutes, likely due to the poly-
merization of THF. A second try was done but the tube was immersed
in an acetone/carbonic ice mixture allowing to wait before NMR
analysis. The 1H NMR spectrum showed paramagnetic resonances:
1
found present in the H NMR spectrum of Cp03Y with [CPh3]
[B(C6F5)4] (see further). After recording of the 1H NMR spectrum of
1 with trityl perfluorophenylborate, a 10 equiv amount of Al(i-Bu)3
(15 mL, 59 mmol) was added in the NMR tube in the glove box. The
dark oil immediately dissolved, to give according to 1H NMR a new
neodymium compound as evidenced by the paramagnetic signals:
d
ppm ¼ 23.3,12.8,10.3, ꢁ0.3, ꢁ1.1, ꢁ8.7, and signals compatible with
the presence of tert-BuC5H4CPh3 isomers:
ppm ¼ 7.1e7.3 (Ph), 5.9e
6.6 (C5H4), 2.7e3.1 (allylic C5H4), 1.05e1.2 (tert-Bu).
Reaction of with dimethylphenylanilinium tetra(penta-
fluorophenyl)borate in C6D6: 5.1 mg (10 mol) of 1 and 15.8 mg
(20 mol) of [HNMe2Ph][B(C6F5)4] were weighed in the glove box in
d
d
ppm ¼ 11.9 (minor), ꢁ12.7 (major), ꢁ23.2 (major). Signals cor-
responding to Al(i-Bu)3 were observed as well:
m), 1.02 (6H, d), 0.31 (2H, d).
d
ppm ¼ 1.95 (1H,
1
m
Reaction of
fluorophenylborate in C7D8: 4.5 mg (8.8
an NMR tube equipped with a Teflon valve. C6D6 (0.4 mL) and then
22 L (88
mol) of Al(i-Bu)3 were added in the glove box. 1H NMR
1
with triisobutylaluminum then trityl per-
m
mmol) of 1 was weighed in
an NMR tube equipped with a Teflon valve. As same as with [CPh3]
[B(C6F5)4], as soon as C6D6 (0.4 mL) was added, the formation of an
oily brown insoluble compound was observed along with discolor-
ation of the initial light blue tint of 1.1H NMR spectrum showed some
residual 1 accompanied with diamagnetic resonances that could be
m
m
analysis showed a little change of signals of 1 (by comparison with
the above experiment, though in C6D6), whereas the signals relative
to Al(i-Bu)3 were found shifted and broadened:
d
ppm ¼ 27.4 (Cp,
attributed to HCp0:
d
ppm ¼ 6e6.5, 2.8, 1.10e1.15, and to NMe2Ph:
Nd), ꢁ2.7 (Cp, Nd), ꢁ17.4 (tert-Bu, Nd), 2.12 (CH, Al), 1.14 (Me2, Al),
d
ppm ¼ 6.5e6.8, 2.45. The latter signals were quite broad, likely due
0.45 (CH2, Al). After subsequent addition of [CPh3][B(C6F5)4] (16 mg,
to coordination of the amine molecule to paramagnetic Nd atom.
Further extraction with THF-d8 showed signals typical of NMe2Ph in
this solvent: 6.67, 6.99, 2.91 [21], and a new compound was detected
which paramagnetic signals were close to what observed in the
17.4 mmol) all was soluble, no formation of an oily compound was
this time observed. Signals of 1 had totally disappeared, and they
were replaced by other paramagnetic signals, slightlydifferent from
the ones observed in the experiment described above, with no
reaction of 1 with trityl borate:
d
ppm ¼ 22.1, 14.9, ꢁ7.8.

142
I. Rodrigues et al. / Journal of Organometallic Chemistry 743 (2013) 139e146
Reaction of Cp03Y with successive amounts of trityl per-
fluorophenylborate in C6D6: to a light orange solution of Cp03Y
(4.5 mg, 10 mol) in C6D6 in an NMR tube equipped with a Teflon
valve, was added 1 equiv. [CPh3][B(C6F5)4] (9.2 mg, 10 mol). The
m
m
solution immediately changed to light yellow along with formation
of an oily phase at the bottom of the tube. 1H NMR showed the
complete consumption of initial Cp03Y, and signals attributed to
Cp0CPh3:
d
ppm ¼ 7.40 (o-Ph, 6H), 7.05 (m,p-Ph, 9H), 6.44 (C5H4,
vinyl 1H), 6.17 (C5H4, vinyl 1H), 6.09 (C5H4, vinyl 1H), 4.62 (C5H4,
allyl 1H), 0.9 (tert-Bu, 9H), and to a set of signals compatible with
the formula of a cationic yttrium derivative [Cp02Y][B(C6F5)4]:
d
ppm ¼ 5.97 (C5H4, 4H), 5.74 (C5H4, 4H), 0.82 (tert-Bu, 18H). After
addition of a second crop of [CPh3][B(C6F5)4] (9.2 mg, 10 mol), the
m
same compounds were present in solution, but the inferior oily
phase had increased. Addition of a third equiv [CPh3][B(C6F5)4]
(9.2 mg, 10 mmol) afforded a mixture in which the signals of the
putative cationic [Cp02Y][B(C6F5)4] were still observed, in company
with those of Cp0CPh3, but the resolution was seriously altered due
to large quantity of brown oil in the bottom of the tube.
3. Results and discussion
3.1. Synthesis of tris(cyclopentadienyl) complex Cp03Nd (1)
Fig. 1. Asymmetric uniꢀt showing the molecular structure of complex 1. Selected dis-
The title complex was synthesized according to metathetical
reaction between NdCl3(THF)2 and KCp0 (Scheme 1) as already
described [5], but in refluxing THF and followed by extraction with
toluene and evaporation to get the THF-free complex [4].
The crude neodymium product was recrystallized from pentane,
affording large dichroic reddish crystals of Cp03Nd free of solvent
molecule, according to 1H NMR [15] and elemental analysis. As far
as we are aware, it is the first time that this compound, already
known as THF-solvated lithium halide (bromide [2a] or chloride
[22]) is isolated without any additional ligand in its coordination
sphere.
ꢀ
tances [A] and angles [ ]: NdeCP (average) 2.50, CPeNdeCP (average) 119.88.
formal CN number of nine have been structurally characterized:
(C5H4CHTMS2)3Nd [26] and the recently described (C5H4TMS)3Nd
[27] (TMS ¼ SiMe3). Both complexes compare very well with 1 in
terms of structural features, particularly NdeCP distances and CPe
ꢀ
ꢀ
NdeCP angles (ca 2.50 A and 120 , respectively).
3.3. Isoprene polymerization using tris(cyclopentadienyl)
derivatives as pre-catalysts
3.2. X-ray structure of Cp03Nd (1)
The homoleptic Cp03Nd 1 was assessed as pre-catalyst for
isoprene polymerization. Due to the lack of coordinated THF, one
could expect some reactivity if at least one Cp0 ligand could be
displaced. The pre-catalyst was combined with a boron activator e
chosen among [HNMe2Ph]þ[B(C6F5)4]ꢁ (dimethylanilinium per-
fluorophenylborate designed here as HNB), [CPh3]þ[B(C6F5)4]e
Crystals of 1 are orthorhombic, space group P212121, with cell
ꢀ
ꢀ
dimensions of a ¼ 8.3916 (2) A, b ¼ 16.4447 (4) A and c ¼ 17.1450
ꢀ
(5) A (Fig. 1). In contrast to previously described unsolvated tris
[(monosubstituted)cyclopentadienyl] derivatives of neodymium,
the basic structural unit in this crystal is the discrete (C5H4t-Bu)3Nd,
(trityl
perfluorophenylborate,
TB),
or
[B(C6F5)3]
(per-
without (h1 5) bridging cyclopentadienyl with an external Nd, like
:h
fluorophenylborane, B) e which is well-known to afford efficient
isoprene polymerization catalysts with rare earth alkyls [28] or
borohydrides [29], and an aluminum co-catalyst (MMAO or Ali-
Bu3). Selected results are gathered in Table 1.
No polymerization reaction took place with HNB (1, 2 or
3 equiv), whereas this boron-based molecule in particular has been
successfully used as activator for polymerization reactions [30]. We
in (C5H4Me)3Nd [23]. In the i-Bu series, the tri-substituted neo-
dymium derivative is isolated as a mono-THF adduct [24].
ꢀ
NdeCP distances are 2.493, 2.498, 2.520 A (CP is the centroid of
a cyclopentadienyl ligand). This is slightly smaller to what observed
in Cp03NdBrLi(THF)3 [2a] and Cp03NdClLi(THF)3 [22] (average value
ꢀ
ꢀ
2.568 A and 2.584 A, respectively), and reflects the absence of a
fourth coordinated ligand. CPeNdeCP are 119.24, 119.45, and
120.94ꢀ, which is superior to the values found for Cp03NdBrLi(THF)3
(116e117ꢀ) and Cp03NdClLi(THF)3 (115.7e118ꢀ) for the same rea-
sons. One can note that the CPeNdeCP in 1 are close to 120ꢀ, as
expected for a homoleptic complex having only three ligands and
no extra external Cp interaction. Surprisingly, among the early
lanthanides family and with the same Cp0 ligand, no other molec-
ular structure was determined [25]. However to our knowledge, in
the monosubstituted cyclopentadienyl series, two other scarce
examples of triscyclopentadienides of neodymium having like 1 a
observed in turn the production of polymer with
1 equiv
[CPh3]þ[B(C6F5)4]ꢁ and in the presence of 5 equiv Ali-Bu3 (run 1).
The choice of such catalytic conditions was guided from the liter-
ature and our previous experience in the field [31]. The yield was
low but the regio-selectivity observed (ca 82% 1,4-) was encour-
aging. With 2 equiv trityl borate and 10 equiv Ali-Bu3, quite
quantitative yield was obtained at 50 ꢀC (run 2) and the selectivity
was 84% 1,4- with 95/5 cis-/trans-ratio. At room temperature, very
low conversion was received (run 3) with similar selectivity.
Enhancing the monomer/catalyst ratio had weak effect upon the
result (runs 4, 5). It is, however, noteworthy that SEC traces of the
polymers for these runs displayed a shape that was nearly mono-
modal, indicative of the likely formation of one single active spe-
cies. An additional equivalent of (trityl borate/5 Ali-Bu3) was
Scheme 1. Synthesis of Cp03Nd (1) (Cp0 ¼ C5H4t-Bu).

I. Rodrigues et al. / Journal of Organometallic Chemistry 743 (2013) 139e146
143
necessary to obtain significant conversion at room temperature
(run 6), but the 1,4-cis rate was found that time less than 80%. The
same was obtained by increasing the monomer to catalyst ratio
(run 7). By raising the temperature to 50 ꢀC, some polymer was
received with good conversion (run 8). Interestingly, complete
conversion was received in 12 h at 50 ꢀC (run 9) and also at room
temperature upon 72 h reaction time (run 10) when inversing the
addition order of trityl borate/Ali-Bu3. This better activity may be
correlated with the initial solubility of the active species (see
Section 2.5). In all experiments carried out with this (1/3 borate/15
Al) combination, at least ca 15% 3,4-defects were still present in the
polymer.
Basically, we observed that five equivalents of trityl borate were
mandatory in order to get high cis-rates, but at the condition to
change the addition order (ie first Al, second borate): when main-
taining the quantity of 15 equiv Al, 87% cis-rate (with less than 4%
3,4-polyisoprene, run 11) was received, whereas when adding first
trityl borate and then Al-co-catalyst, a polymer containing a high
amount of trans-motives (ca 35%, run 12) was formed. This result
could be connected to steric hindrance in the coordination sphere
of the neodymium atom [32]. With 25 equiv Ali-Bu3, ie with the
previous Al/borate ratio of 5, and upon reverse addition (first Al,
second borate), high conversion along with ca 90% cis-rate was
observed (run 13) at 50 ꢀC. In the presence of 25 equiv Al but this
time at room temperature, highly cis-stereoregular polyisoprene
(95.5%, and cis/trans ratio close to 99%), of relatively high molecular
weight, was obtained (run 14). At high monomer to catalyst ratio
(3000, run 15), however, no improvement in terms of molecular
weights but broadening of the polymolecularity was noticed. At
50 ꢀC, the activity was higher than at room temperature, together
with high cis-selectivity that was maintained at a level superior to
95% (run 16). When toluene was replaced by pentane with the same
co-catalyst conditions, the catalyst was less efficient, with higher
amounts of 3,4-polymer (run 17).
In terms of comparison with the literature for other neodymium
catalysts [14,29a], the catalytic systems presented herein are only
reasonably active (up to ca 16,000 g PI/mol Nd/h, run 10); this is
likely related to the presence of the sterically demanding three Cp0
ligands set surrounding the neodymium atom in the pre-catalyst.
An experiment carried out with just trityl borate in the absence
of co-catalyst yielded, within minutes, an intractable material
typical of cationic polymerization affording cross-linked material
(run 18). With pentafluorophenylborane B(C6F5)3 as activator, and
with the same Al/B ratio of 5, good conversion was obtained at
50 ꢀC (runs 19e21). The reverse addition of co-catalyst/activator
had this time no impact upon the final result. Good cis-selectiv-
ities were received (85e90%) but the SEC traces showed broader
polymolecularities. Reactions conducted with Cp03Nd and MMAO
(100 and 500 equiv) failed to produce polyisoprene and they are not
included in Table 1. The “simple” Cp3Nd (2) was also assessed as
pre-catalyst, in the presence of trityl borate ([CPh3]þ[B(C6F5)4]ꢁ) as
boron activator, and Al(i-Bu)3 as co-catalyst, but the experiments
were not successful (run 22). This may be attributed to the low
solubility of complex 2 in the solvent of polymerization. Additional
preliminary polymerization experiments were also undertaken
with the yttrium analog of 1, Cp03Y, which was synthesized as
described in Scheme 1 and characterized by 1H NMR and elemental
analysis.1 We were able to obtain polyisoprene with high activity,
but with addition of Al-cocatalyst and then borate activator in this
order (runs 23, 24). The high amount of 3,4-defects (ca 25%) in the
isolated polymer may be connected to the smaller size of the rare
earth element, by comparison with neodymium, which favors
single diene coordination.
3.4. Mechanistic investigations
As reminded above in the text, the reaction of a Bronsted acid
with a tris(cyclopentadienyl)lanthanide compound was published
in the early 1960’s [7]. Just recently, this strategy was utilized to
elaborate phenoxide complexes as initiators for lactide polymeri-
zation [33]. However, to our knowledge, such reactivity involving
tris(cyclopentadienyl) derivatives of the rare earths, had not been
exploited until now to promote the polymerization of conjugated
dienes [34]. Kaita et al. largely explored the potentialities in this
frame of Ln/Al bimetallics, in the bisCp* series, through borate
activation of alkyl precursors [35]. Our approach strongly differs
from Kaita’s one, since in our case the general idea was based on the
possibility to eliminate (at least) one h5 Cp-type ligand (here a Cp0
one) by direct protonation with an acidic borate activator. This type
of reactivity had already been reported by Ephritikhine ten years
ago in the case of a phospholyl abstraction (Scheme 2) [36].
We thus envisaged the three boron-based molecules e HNB, TB,
or B (see above) e as possible activators in our catalytic combina-
tions as described in Table 1. It is worth to be noted that the
abstraction of an h
5 Cp-type ligand (rather than an alkyl one in the
present case) by trityl borate was reported recently by Chen: a
mono(indenyl)bis(alkyl) scandium complex lost its indenyl ligand
by the reaction with TB to afford the corresponding Ph3Ceindene
and a dialkylscandium cation [37]. Tardif described on his side both
reactivity pathways, i.e. alkyl and ring abstraction, starting from
bisindenylsilylamido rare earths complexes with either anilinium
or trityl borate, giving rise to the elimination of an indenyl group
[38]. Finally, trityl borate was found capable to extract one C5Me4H
ligand from (C5Me4H)3Al [39]. We thus expected that CpR3Ln
(CpR ¼ Cp0 or Cp; Ln ¼ Nd or Y) would behave comparably (Scheme
3, boron activator ¼ HNB, TB, or B), to in fine generate an active
species by subsequent reaction with an alkylating reagent such as
AlR3.
The question of the displacement of a Cp-type ligand (noted
here as CpR) by an alkylating reagent could also be considered,
having in mind the above-mentioned article of Schumann [9]. Thus,
activation of a (CpR)3Ln pre-catalyst by an alkylaluminum reagent,
as depicted in Scheme 4, had to be taken into account in this regard,
despite the a priori steric saturation of a tris(cyclopentadienyl)
derivative, and as also confirmed by the weak NdeAl interaction
reported by Arnold when (C5Me5)Al is added to (C5H4TMS)3Nd [27].
On the other hand, comproportionation reactions that have
been observed in the Cp0 series, and which proceed through
bimolecular processes, account for a possible Cp0 “mobility” in
Cp03Ln compounds (Scheme 5) [2d].
We undertook a series of reactions at the NMR scale aiming at
identifying possible active species involved in the polymerization
process. When 1 was reacted with trityl borate (2 equiv) in C6D6, we
observed by 1H NMR the complete consumption of starting com-
pound 1 along with diamagnetic signals that could be compatible
with tert-BuC5H4CPh3 (see Experimental). Some paramagnetic
signals typical of a Cp0-supported Nd compound [15] were present,
but of low intensity due to poor solubility of the compound formed.
At the addition of Al(i-Bu)3 (10 equiv), all was soluble again,
1
To confirm this chemical formula, a synthesis of Cp03Y was carried out by ionic
Scheme 2. Activation of a Nd compound by protonation of a phospholyl ligand
metathesis from Y(BH4)3(THF)3. The isolated complex displayed the same 1H NMR
spectrum that the one isolated from YCl3.
induced by a borate activator (COT ¼ aromatic C8H8, P* ¼ aromatic tetramethylphos-
pholyl C4Me4P).

144
I. Rodrigues et al. / Journal of Organometallic Chemistry 743 (2013) 139e146
Scheme 3. Reaction of a CpR3Nd compound with a boron activator to generate an active species and the subsequent reaction with an alkylaluminum to generate the catalyst.
Scheme 4. Possible displacement of a Cp0 ring by AlR3 in 1.
presence of PhNMe2 [21], whereas a new compound was detected,
which paramagnetic signals were close to what observed in the
reaction of 1 with trityl borate.
Scheme 5. Comproportionation reaction from Cp03Sm affording mixed Cp0/BH4
complex.
Finally, the reaction of 1 with Al(i-Bu)3 (10 equiv) showed no
reaction at all, but signals of both reactants were slightly modified
in terms of chemical shifts, and Al signals were found broadened.
This accounts for a mechanism in which the Al-cocatalyst by itself is
not able to displace a Cp0 ligand, but a [Nd]/Al(i-Bu)3 interaction is
here clearly evidenced. The subsequent addition of trityl borate
(2 equiv) showed on the 1H NMR spectrum the disappearance of 1
with residual Al(i-Bu)3 (initially present in large excess), along with
apparition of paramagnetic signals, but their assignation to a
possible active species was not possible.
To summarize these NMR experiments, a pattern emerges
regarding paramagnetic data: one can observe in general two types
of resonances, one or two bigger signals in the range
[ꢁ10; ꢁ25 ppm] and smaller ones in the range [11e14 ppm]. These
data support the presence of a Cp0 ligand still attached to neo-
dymium. Similar reactions at the NMR scale conducted with
diamagnetic Cp03Y showed the complete consumption of the latter
when reacted with 1 equiv [CPh3][B(C6F5)4]. The signals of Cp0CPh3
showing among others the signals of free Al(i-Bu)3 and of a para-
magnetic species. When the reaction of 1 with trityl borate
(2 equiv) was carried out in deuterated THF at room temperature,
polymerization of THF took place within minutes, as already
observed by others [30c,31d,40]. This was clearly indicative of the
likely formation of a cationic species, which is known to rapidly
promote such polymerization. Another experiment was done again
but with immediate storage of the mixture at low temperature until
the 1H NMR spectrum was recorded. One could observe signals that
could be attributed to tert-BuC5H4CPh3 isomers and paramagnetic
resonances compatible (one big signal at high fields and several
others of less importance) with Cp0[Nd] species. The reaction of 1
with [HNMe2Ph][B(C6F5)4] (2 equiv) in C6D6 gave rise to the con-
sumption of most part of 1 and formation of HCp0 and free amine
[41], as depicted in Scheme 6. Extraction with THF-d8 confirmed the
Scheme 6. Reaction of Cp03Nd (1) with [HNMe2Ph][B(C6F5)4] to generate an active species.
1
Fig. 2. H NMR reaction of Cp03Y with one equiv. [CPh3][B(C6F5)4] showing the complete disappearance of the starting material (C6D6).

I. Rodrigues et al. / Journal of Organometallic Chemistry 743 (2013) 139e146
145
1
Fig. 3. H NMR reaction of Cp03Y with successive amounts of [CPh3][B(C6F5)4].
were clearly observed, along with those which could correspond to
a cationic yttrium derivative [Cp02Y][B(C6F5)4] (Fig. 2). Interestingly,
this compound was still present in solution in the presence of
additional amounts of [CPh3][B(C6F5)4] (Fig. 3). On the other hand,
the poorly soluble oily phase observed in the NMR tube, which
increased with the amount of trityl reagent added, was the prob-
able indication of the formation of cationic active species in our
polymerization mixtures.
framework [43]. This also resembles the dicationic active species
proposed by Okuda in the presence of excess borate activator [31c].
Further experiments aiming at completing this preliminary study for
a better understanding of the mechanism involved are in progress.
4. Conclusion
The homoleptic Cp03Nd complex (Cp0 ¼ t-BuC5H4) has been
considered as potential polymerization pre-catalyst on the basis of
its solvent-free molecular structure, despite the steric set of three
cyclopentadienyl ligands surrounding the metal. The X-ray struc-
ture of this compound is presented for the first time. When com-
bined with appropriate amounts of borate activators and aluminum
co-catalyst, Cp03Nd successfully afforded a highly selective catalyst
for the polymerization of isoprene. Up to 95.5% cis-regular polymer
was produced with good activity. This reactivity is tentatively
ascribed to the displacement of one or more Cp0 ligand, on the basis
of NMR experiments. Studies aiming at unequivocally establishing
the formation of active bisCp0 and monoCp0 active species in such
catalytic processes are in due course.
Up to now and despite our efforts, syntheses devoted to the
isolation of a cationic-like derivative that could arise from the reac-
tion of a trisCp0Ln complex with a borate activator have not been
successful yet, due to the too high reactivity in THF or low stability in
other solvents [41,42]. However, it seems established that, at least,
one Cp0 ligand is abstracted from the coordination sphere of the rare
earth in our catalytic combinations. The selectivity of the polymeri-
zation may be related to the number of Cp0 ligands that could be
displaced from the coordination sphere of the metal upon activation
with the boron compound. Indeed, a [Cp02Lnþ] active species would
probably lead to a polymer with significant amount of 3,4-motives,
which is generally accepted as arising from a single h
2 coordination
of the isoprene monomer [29a]. In general from Table 1, it appears
that quite high amounts of 3,4-defects are present in the poly-
isoprene prepared with 1e3 equiv borate activator. Considering the
high cis-selectivity and the much lower 3,4-defects quantity
observed in the presence of 5 equiv borate, one can postulate a
possible displacement of a second Cp0 ligand. Consequently, the
formation of a cationic half-sandwich [Cp0LnRþ] active species could
also be proposed in such conditions, in line with the well-known
high cis-selectivity of catalysts bearing such half-sandwich
In summary, we have demonstrated in this work that the acti-
vation of an a priori saturated and easily available compound can be
considered as a valuable approach to afford a catalyst for conju-
gated dienes polymerization.
Acknowledgments
The authors wish to thank Dr Frederic Hebrard for the synthesis
of Cp03Y compound, Aurelie Malfait for SEC analyses. “The Fonds

146
I. Rodrigues et al. / Journal of Organometallic Chemistry 743 (2013) 139e146
[16] Bruker Analytical X-ray Systems, Inc., Apex2, Version 2 User Manual (2006).
M86-E01078, Madison, WI.
[17] Siemens Industrial Automation, Inc., SADABS: Area-detector Absorption
Correction. Madison, WI (1996).
[18] L. Palatinus, G. Chapuis, J. Appl. Crystallogr. 40 (2007) 786.
[19] P.W. Betteridge, J.R. Carruthers, R.I. Cooper, K. Prout, D.J. Watkin, J. Appl.
Crystallogr. 36 (2003) 1487.
[20] M.J. Vitorino, P. Zinck, M. Visseaux, Eur. Polym. J. 48 (2012) 1289.
[21] S. Bambirra, D. van Leusen, A. Meetsma, B. Hessen, J.H. Teuben, Chem.
Commun. (2003) 522.
[22] J. Guan, Q. Shen, S. Jin, Y. Lin, Acta Crystallogr. C50 (1994) 891.
[23] J.H. Burns, W.H. Baldwin, F.H. Fink, Inorg. Chem. 13 (1974) 1916.
[24] W.A. Herrmann, R. Anwander, F.C. Munck, W. Scherer, Chem. Ber. 126 (1993)
331.
[25] The physical properties of (C5H4t-Bu)3Ce were claimed as nearly identical
with those of (Me3SiC5H4)3Ce, which was isolated as a base-free compound
isostructural with (C5H4SiMe3)3U as shown by X-ray powder diffraction
studies: A. Zalkin, J.G. Brennan, R.A. Andersen Acta Crystallogr. C44 (1988)
2104.
Européen de Développement Régional (FEDER)”, CNRS”, “Région
Nord Pas-de-Calais” and “Ministère de l’Education Nationale de
l’Enseignement Supérieur et de la Recherche” are acknowledged for
fundings of X-ray diffractometers.
Appendix A. Supplementary material
CCDC 908388 contains the supplementary crystallographic data
for this paper. These data can be obtained free of charge from The
Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/
data_request/cif.
Appendix B. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.jorganchem.2013.06.035.
[26] S. Al-Juaid, Y.K. Gun’ko, P.B. Hitchcock, M.F. Lappert, S.J. Tian, Organomet.
Chem. 582 (1999) 143.
[27] S.G. Minasian, J.L. Krinsky, J.D. Rinehart, R. Copping, T. Tyliszczak, M. Janousch,
D.K. Shuh, J. Arnold, J. Am. Chem. Soc. 131 (2009) 13767.
[28] S. Kaita, Z. Hou, Y. Wakatsuki, Macromolecules 32 (1999) 9078. and refs cited
therein.
References
[29] (a) Z. Zhang, D. Cui, B. Wang, B. Liu, Y. Yang, Struct. Bond. 137 (2010) 49;
(b) M. Visseaux, F. Bonnet, Coord. Chem. Rev. 255 (2011) 374.
[30] (a) S. Li, W. Miao, T. Tang, W. Dong, X. Zhang, D. Cui, Organometallics 27
(2008) 718;
[1] H. Schumann, J.A. Meese-Marktscheffel, L. Esser, Chem. Rev. 95 (1995) 865.
[2] Selected examples: (a) S. Song, Q. Shen, S. Jin, Polyhedron 11 (1992) 2863;
(b) Z. Wu, Z. Xu, X. You, X. Zhou, X. Wang, J. Chen, Polyhedron 13 (1994) 379;
(c) S. Wang, Y. Yu, Z. Ye, C. Qian, J. Organomet. Chem. 464 (1994) 55;
(d) D. Baudry, A. Dormond, B. Lachot, M. Visseaux, G. Zucchi, J. Organomet.
Chem. 547 (1997) 157.
[3] J.M. Birmingham, G. Wilkinson, J. Am. Chem. Soc. 76 (1954) 6210.
[4] S.D. Stults, R.A. Andersen, A. Zalkin, Organometallics 9 (1990) 115.
[5] W.A. Herrmann, R. Anwander, M. Kleine, K. Ofele, J. Riede, W. Scherer, Chem.
Ber. 125 (1992) 2391.
[6] J.M. Birmingham, G. Wilkinson, J. Am. Chem. Soc. 78 (1956) 42.
[7] R.E. Maginn, S. Manastyrskyj, M. Dubeck, J. Am. Chem. Soc. 85 (1963) 672.
[8] (a) C. Qian, D. Zhu, Y. Gu, J. Organomet. Chem. 401 (1991) 23;
(b) C. Qian, D. Zhu, D. Li, J. Organomet. Chem. 430 (1992) 175.
[9] H. Schumann, G. Jeske, Angew. Chem. Int. Ed. Engl. 24 (1985) 226.
[10] (a) W.J. Evans, K.J. Forrestal, J.W. Ziller, Angew. Chem. Int. Ed. Engl. 36 (1997)
774;
(b) R. Litlabø, M. Enders, K.W. Tornroos, R. Anwander, Organometallics 29
(2010) 2588;
(c) L. Zhang, Y. Luo, Z. Hou, J. Am. Chem. Soc. 127 (2005) 14562;
(d) M. Visseaux, M. Mainil, M. Terrier, A. Mortreux, P. Roussel, T. Mathivet,
M. Destarac, Dalton Trans. (2008) 4558.
[31] (a) Y. Luo, S. Fan, J. Yang, J. Fang, P. Xu, Dalton Trans. 40 (2011) 3053;
(b) Z. Jian, D. Cui, Z. Hou, X. Li, Chem. Commun. 46 (2010) 3022;
(c) S. Arndt, K. Beckerle, P.M. Zeimentz, T.P. Spaniol, J. Okuda, Angew. Chem.
Int. Ed. 44 (2005) 7473;
(d) F. Bonnet, C. Da Costa Violante, P. Roussel, A. Mortreux, M. Visseaux, Chem.
Commun. (2009) 3380.
[32] G. Du, Y. Wei, L. Ai, Y. Chen, Q. Xu, X. Liu, S. Zhang, Z. Hou, X. Li, Organome-
tallics 30 (2011) 160.
[33] K. Nie, X. Gu, Y. Yao, Y. Zhang, S. Qi, Dalton Trans. 39 (2010) 6832.
[34] Evans showed that (C5Me5)3Nd can be protonated by HNEt3BPh4 to yield the
(b) W.J. Evans, J.M. Perotti, S.A. Kozimor, T.M. Champagne, B.L. Davis,
G.W. Nyce, C.H. Fujimoto, R.D. Clark, M.A. Johnston, J.W. Ziller, Organome-
tallics 24 (2005) 3916.
corresponding cation. This reactivity is attributed to the “
reactivity of the (C5Me5)3Ln complex”, see Ref. [10].
h
1-C5Me5 alkyl-like
[11] W.J. Evans, K.J. Forrestal, M.A. Ansari, J.W. Ziller, J. Am. Chem. Soc. 120 (1998)
2180.
[12] Selected references: (a) J. Gromada, J.F. Carpentier, A. Mortreux, Coord. Chem.
Rev. 248 (2004) 397;
[35] (a) S. Kaita, Z. Hou, M. Nishiura, Y. Doi, J. Kurazumi, A.C. Horiuchi,
Y. Wakatsuki, Macromol. Rapid Commun. 24 (2003) 179;
(b) S. Kaita, M. Yamanaka, A.C. Horiuchi, Y. Wakatsuki, Macromolecules 39 (4)
(2006) 1359.
[36] S.M. Cendrowski-Guillaume, G. Le Gland, M. Lance, M. Nierlich,
M. Ephritikhine, C. R. Chim. 5 (2002) 73.
[37] X. Xu, Y. Chen, J. Sun, Chem. Eur. J. 15 (2009) 846.
[38] O. Tardif, S. Kaita, Dalton Trans. (2008) 2531.
[39] S.J. Lee, P.J. Shapiro, B. Twamley, Organometallics 25 (2006) 5582.
[40] Y. Luo, J. Baldamus, Z. Hou, J. Am. Chem. Soc. 126 (2004) 13910.
[41] M. Zimmermann, K.W. Tornroos, R. Anwander, Angew. Chem. Int. Ed. 47
(2008) 775.
(b) Z. Hou, Bull. Chem. Soc. Jpn. 76 (2003) 2253;
(c) Y. Nakayama, H. Yasuda, J. Organomet. Chem. 689 (2004) 4489;
(d) P.M. Zeimentz, S. Arndt, B.R. Elvidge, J. Okuda, Chem. Rev. 106 (2006) 2404.
[13] See for example: (a) R. Anwander, M.G. Klimpel, H.M. Dietrich, D.J. Shorokhov,
W. Scherer, Chem. Commun. (2003) 1008;
(b) S. Arndt, J. Okuda, Chem. Rev. 102 (2002) 1953;
(c) H.J. Heeres, A. Meetsma, J.H. Teuben, R.D. Rogers, Organometallics 8 (1989)
2637.
[14] L. Friebe, O. Nuyken, W. Obrecht, Adv. Polym. Sci. 204 (2006) 1.
[15] The complex had previously been synthesized in situ for the sake of NMR
characterization but it was not isolated: M. Visseaux, A. Dormond, D. Baudry
Bull. Soc. Chim. Fr. 130 (1993) 173.
[42] L. Zhang, M. Nishiura, M. Yuki, Y. Luo, Z. Hou, Angew. Chem. Int. Ed. 47 (2008)
2642.
[43] Z. Hou, Y. Luo, X. Li, J. Organomet. Chem. 691 (2006) 3114.