Another example of carborane based trianionic ligand: Syntheses and catalytic activities of cyclohexylamino tailed ortho-carboranyl zirconium and titanium dicarbollides

Article abstract of DOI:10.1016/j.jorganchem.2005.04.052

In situ reaction of Li[closo-1-Ph-1,2-C₂B₁₀H ₁₀] with 7-azabicyclo [4.1.0] heptane results in the formation of the disubstituted carborane, closo-1-Ph-2-(2′-aminocyclohexyl)-1,2-C ₂B₁₀H₁₀ (

Full text of DOI:10.1016/j.jorganchem.2005.04.052

Journal of Organometallic Chemistry 690 (2005) 6284–6291
www.elsevier.com/locate/jorganchem
Another example of carborane based trianionic ligand: Syntheses
and catalytic activities of cyclohexylamino tailed
ortho-carboranyl zirconium and titanium dicarbollides
a,
a
a
Zhu Yinghuai *, Shirley Lo Pei Sia , Fethi Kooli ,
a
Keith Carpenter , Richard A. Kemp
b,c
a
Institute of Chemical and Engineering Sciences Ltd., 1 Pesek Road, Jurong Island, Singapore 627833, Singapore
b
Department of Chemistry, University of New Mexico, Albuquerque, NM 87131-2609, USA
c
Advanced Materials Laboratory, Sandia National Laboratories, Albuquerque, NM 87106, USA
Received 13 January 2005; received in revised form 18 March 2005; accepted 18 April 2005
Available online 25 July 2005
Abstract
In situ reaction of Li[closo-1-Ph-1,2-C2B10H10] with 7-azabicyclo [4.1.0] heptane results in the formation of the disubstituted
carborane, closo-1-Ph-2-(20-aminocyclohexyl)-1,2-C2B10H10 (1), in 63% yield. Decapitation of (1) with potassium hydroxide in
refluxing ethanol produces the cage-opened nido-carborane, K[nido-7-Ph-8-(20-aminocyclohexyl)-7,8-C2B9H10]ꢀ (2), in 80% yield.
Deprotonation of the above monoanion with two equivalents of n-butyllithium followed by reaction with anhydrous MCl4 Æ 2THF
(M = Zr, Ti) provides d0-half-sandwich metallocarboranes, closo-1-M(Cl)-2-Ph-3-(20-r-(H)N-cyclohexyl)-2,3-g5-C2B9H9 (3 M = Zr;
4 M = Ti) in 53% and 42% yields, respectively. The reaction of Li[closo-1,2-C2B10H11] with 7-azabicyclo [4.1.0] heptane in THF
affords closo-1-(20-aminocyclohexyl)-1,2-C2B10H10 (5) in 59% yield. Immobilization of the carboranyl amino ligand (1) to an organic
support, Merrifieldꢀs peptide resin (1%), has been achieved by the reaction of the sodium salt of (5) with polystyryl chloride in THF
to produce closo-1-(20-aminocyclohexyl)-2-polystyryl-1,2-C2B10H10 (6) in 87% yield. Further reaction of the dianion derived from
(6) with anhydrous ZrCl4 Æ 2THF led to the formation of the organic polystyryl supported d0-half-sandwich metallocarborane,
closo-1-Zr(Cl)-2-(20-r-(H)N-cyclohexyl)-3-polystyryl-2,3-g5-C2B9H9 (7), in 38% yield. These new compounds have been character-
ized by elemental analyses, NMR, and IR spectra. Polymerizations of both ethylene and vinyl chloride with (3) and (7) have been
performed in toluene using MMAO-7 (13% ISOPAR-E) as the co-catalyst. Molecular weights up to 32.8 · 103 (Mw/Mn = 1.8) and
9.5 · 103 (Mw/Mn = 2.1) were obtained for PE and PVC, respectively.
ꢀ 2005 Elsevier B.V. All rights reserved.
Keywords: Carborane trianion; Half-sandwich metallocarborane; Ligand immobilization; Olefin polymerization
1. Introduction
this field is still one of the most important and primarily
interesting areas in polymer chemistry research and
development. A major goal is to identify new genera-
tions of catalysts for the a-olefin polymerization process.
Geometry constrained complexes, which contain at least
one g5-ligand, such as cyclopentadienyl (Cp), indenyl li-
gand, etc., with a terminal r-bonding group and group
3–6 transition metals, is a particular subgenus of Zie-
gler-Natta catalysts and belongs in the group of metallo-
cene catalysts [3–5]. Although metallocene catalysts
Although Ziegler-Natta polyolefin catalysts have
been well-studied for the polymerization of simple ole-
fins to obtain desired molecular weight polymers [1,2],
*
Corresponding author. Present address: ICES, Apply Cat. Ayer
Rajah Crescent, Singapore 139959, Singapore. Tel.: +6568744217; fax:
+6568734805.
E-mail address: zhu_yinghuai@ices.a-star.edu.sg (Z. Yinghuai).
0022-328X/$ - see front matter ꢀ 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2005.04.052

Z. Yinghuai et al. / Journal of Organometallic Chemistry 690 (2005) 6284–6291
6285
have the advantage of often appearing to be ‘‘single-
site’’ catalysts, giving a narrow range of molecular
weight polymers of similar molecular structure, most
of the known Ziegler-Natta catalyst systems fail to pro-
duce functionalized polyolefins directly. The reason for
this is that the early transition metal centers in these cat-
alysts are highly electrophilic, generally making it
impossible to use olefins containing polar functional
groups as monomers or co-monomers. Therefore, con-
ventional free radical polymerization is still used indus-
trially to produce a wide range of functionalized
polymers [6]. However, the commonly used free radical
process is unsuitable to prepare the polymers with pre-
cise molecular weight distribution. Thus, there is a need
to generate new single-site catalysts for the polymeriza-
tion of polar a-olefins in order to prepare homo- or
copolymers with narrow molecular weight distributions.
It has been reported that the dicarbollide ligand
[C2B9H11]2ꢀ is isolobal and isoelectronic with Cpꢀ,
and this ligand also coordinates with metal centers in
an g5-fashion by contributing six delocalized p-electrons
to form metallocarboranes [7–10]. Moreover, the carbo-
rane ligands are thermally stable and robust. The neu-
tral icosahedral carborane cage can be heated to
600 ꢁC without decomposition, and the C2B9-cage ions
can be recovered unchanged after several hours of heat-
ing in solution with aqueous acids and bases [11–13]. As
these compounds are isoelectronic with their respective
Cp analogues, one would expect similar, but polar
group tolerable, catalytic behaviors in the polymeriza-
tion of a-olefins. The coordination chemistry of carbo-
rane clusters derived from the carborane cages systems
has been widely studied, and substituted and unsubsti-
tuted dicarbollide metal complexes have been reported
[14–20]. However, to the best of our knowledge, the
study of carborane trianion-based metallocene ana-
logues is limited with only a few reported examples
[21–26]. Following up on our former study, herein we
report the synthesis and coordination chemistry
of another carborane-based trianionic ligand, ortho-
carboranyl dianion tailed with a cyclohexyl-bridged ter-
minal nitrogen anion (Nꢀ). The a-olefin polymerization
results catalyzed by its group 4 metal complexes in the
presence of co-catalyst MMAO-7 is also presented.
Fig. 1. CH2 bridged carborane trianions.
Reaction of the lithium salt of the phenyl-substituted
carborane monoanion with synthesized 7-azabicyclo
[4.1.0] heptane followed by hydrolysis and purification
produced closo-1-Ph-2-(20-aminocyclohexyl)-1,2-C2B10-
1
H10 (1) in 63% yield (Scheme 1). The H, 13C, and 11B
NMR spectra appear normal relative to other related
structures. In the IR spectrum a strong absorption at
m = 2565 cmꢀ1 is attributed to mBH. Further reaction of
(1) with potassium hydroxide in refluxing ethyl alcohol
led to the formation of the decapitated carborane mono-
anion K[nido-7-Ph-8-(20-aminocyclohexyl)-7,8-C2B9H10]
(2) in 80% yield after purification. Compound (2) easily
dissolves in polar solvents such as THF, DMSO, ace-
tone, DMF, ethyl alcohol, and the like. Compared with
its precursor (1), the 13C chemical shift of –CH–NH2 in
2 shifted upfield 2.40 ppm, and the corresponding IR
absorption of mBH blue shifted 31 cmꢀ1. After in situ
deprotonation with two equivalents of n-BuLi in
dry THF followed by reaction with MCl4 Æ 2THF
(pale yellow powder), cyclohexylamino tailed ortho-
carboranyl trianionic ligand coordinated group
4
complexes, closo-1-M(Cl)-2-Ph-3-(20-r-(H)N-cyclohexyl)-
2,3-g5-C2B9H9 (3, M = Zr; 4, M = Ti), were produced
in 53% and 42% yields, respectively. Both (3) and (4) dis-
solve in benzene and permit purification by precipitation
from corresponding benzene solution with pentane.
2. Results and discussion
We previously reported the in situ synthesis and coor-
dination chemistry of carborane trianions composed of
dicarbollide dianion and methylene-bridged terminal
heteroatom anion (Fig. 1) [24,25]. The disadvantages
of high air-sensitivity and inconvenient synthetic meth-
odology for these compounds made us look for a more
feasible method to prepare additional stable carboranyl
trianion-derived metal complexes.
Scheme 1. Synthesis of carborane trianions group 4 compounds.

6286
Z. Yinghuai et al. / Journal of Organometallic Chemistry 690 (2005) 6284–6291
When exposed to air, the pale orange color of a benzene
solution of (3) or (4) changed into a deep brown solution
after 1 day. This is slower than the corresponding
changes of known literature compounds [24,25]. The
half-lives of decomposition of (3) and (4) are about 11
and 7 h, respectively, based on in situ NMR. These val-
ues are longer than the corresponding literature ana-
logues under the same conditions (t1/2 = 5, 8 h,
respectively, for ligand b in Fig. 1 for the titanium and
zirconium complexes). This confirms that (3) and (4)
are more stable than methylene bridged analogues.
Compound (3) and (4) are also stable in polar solvents,
such as THF, diethyl ether, DMSO, DMF, and absolute
ethanol. There was no obvious change in their 1H NMR
spectra after dissolution in these solvents. It was noted
that after metallization, the 13C chemical shifts of
–CHNH– shifted dramatically upfield, 38.46 and
35.09 ppm for (3) and (4), respectively. This change sup-
ports the formation of M–N (M = Ti, Zr) bonds. Com-
pared with the precursor ligand (2), the 11B NMR
spectra of (3) and (4) shifted significantly downfield to
between d ꢀ5.77 and ꢀ25.30 ppm, is consistent with
the literature values [24].
intermediate. In situ reaction of the above intermediate
with ZrCl4 Æ 2THF produced polymer-supported carbo-
rane trianionic ligand coordinated zirconium complex
(7) with loading amounts of zirconium species up to
0.46 mmol/g (Merrifieldꢀs peptide resin). Carborane
cages in (6) and (7) exhibit typical mBH absorptions at
2576 and 2513 cmꢀ1, respectively. These data support
the successful immobilization of carborane cages to
Merrifieldꢀs peptide resin (1%).
In simultaneous DSC–TGA analysis, the residue
weight of (6) and (7) increases noticeably when com-
pared to that of Merrifieldꢀs peptide resin (1%), which
is completely combusted when heated to 600 ꢁC in air
(Fig. 2). Unfortunately, the amount of immobilized car-
borane cannot be determined by calculation from the
changes of the residual weight like other polymer analy-
ses, because both the starting material (5) and Merri-
fieldꢀs peptide resin (1%) completely combust in air at
900 ꢁC. In addition, compound (5) sublimes when
heated and this property also makes it impossible for
Immobilization of the new trianionic ligand has been
successfully achieved by the reaction of the sodium salt
of closo-1-(20-aminocyclohexyl)-1,2-C2B10H10, prepared
in situ from the deprotonation of (5) with sodium hy-
dride in THF at ꢀ30 ꢁC, with Merrifieldꢀs peptide resin
(1%) to produce closo-1-(20-aminocyclohexyl)-2-poly-
styryl-1,2-C2B10H10 (6) in 87% yield (Scheme 2). The
loading amount of ligand 1-(20-aminocyclohexyl)-1,2-
C2B10H10 reaches up to 1.28 mmol/g (Merrifieldꢀs
peptide resin (1%)), based on the produced sodium chlo-
ride. Decapitation of (6) with potassium hydroxide in
refluxing ethyl alcohol provided polymer supported
nido-carborane. This species can be further deproto-
nated by n-BuLi to form a corresponding trianionic
Fig. 2. TGA curves of Merrifieldꢀs peptide resin, 5, 6 and 7.
Scheme 2. Synthesis of polymer supported carborane trianion zirconium complex.

Z. Yinghuai et al. / Journal of Organometallic Chemistry 690 (2005) 6284–6291
6287
show difference vibrations. Having the isolobal concept
of carboranyl dianion with Cpꢀ in mind, we tentatively
assign the vibrations at 311, 337, and 376 in the far-IR
spectrum of (3) to mZr–NH, mZr–Cl and mZr-Cage stretching
vibrations, respectively. Compared with the relatively
smooth absorption curve of (6), (7) presented a similar
absorption model with free complex (3). This suggests
that the zirconium species in (7) has the same coordina-
tion fashion as the unsupported compound (3).
The catalytic properties of the complexes (3) and (7)
have been evaluated for the olefin polymerization
process using commercially available ethylene and vinyl
chloride in toluene with MMAO-7 co-catalyst. Table 1
summarizes the dual polymerization results of ethylene
and vinyl chloride with carborane trianionic ligand
coordinated zirconium complexes. It can be seen that
(a) the synthesized zirconium complexes are active a-ole-
fin polymerization catalysts whose catalytic efficiencies
demonstrate a higher activity for ethylene than vinyl
chloride; (b) under the same conditions, the homoge-
neous catalytic system is more active than corresponding
heterogeneous system, which may be caused by the
diminished diffusion of monomer into the interior pores
of the supported catalyst, or the result of fewer active
centers present in the heterogeneous variant [29]; and
(c) the obtained polymers have a well-defined molecular
architecture with relatively narrow molecular weight dis-
tributions (Mw/Mn are in the range 1.6–2.2).
To further elucidate the structure of the obtained
polyvinyl chloride (PVC), selected PVC samples pro-
duced from the homogeneous catalyst systems were re-
duced to saturated hydrocarbon polymer using
Bu3SnH and AIBN in THF according to the literature
[30,31]. Branched structures cannot be determined from
NMR spectra of PVC. The 13C NMR spectra of the re-
duced polymer no longer contained any PVC signals but
only showed the main peak of the CH2 groups (d
29.02 ppm), signals of remaining Bu3SnH, and no
detectable amounts of branches. This indicates that side
reactions leading to branches do not occur under these
polymerization conditions, and the resulting PVC is free
of structural defects.
Fig. 3. Far-IR spectra of 3, 6 and 7.
the quantitative determination of carborane attached to
Merrifieldꢀs peptide resin even under nitrogen atmo-
sphere by simultaneous DSC–TGA analysis, due to
unavailable reference. Hence, we determined the amount
of anchored carborane by titration of the sodium or lith-
ium chloride produced in the reactions with AgNO3 and
further confirmed the zirconium contents with ICP–MS
analysis.
To confirm the exact structure of the zirconium spe-
cies in the Merrifieldꢀs peptide resin (1%) supported tri-
anionic ligand coordinated zirconium dicarbollide (7),
the far-IR spectra of (3), (6) and (7) have been analyzed
(Fig. 3). The broad peaks in the solid-state 13C NMR
spectra of (6) and (7), as well as the pale yellow color
of (3) and (7), make it impossible to use either solid-state
NMR or Raman spectroscopy to analyze the detailed
coordination modes. However, far infrared techniques
have been developed that allow us to probe lower fre-
quencies. Much information related to the metal–ligand
bond stretching frequencies mM–L is available. The fun-
damental absorptions of mZr–N, mZr–Cl and mZr–Cp have
been assigned at 300, 310, and 360 cmꢀ1, respectively
[27,28]. Obviously, owing to the possibility of various
coupling stretching vibrations, it may be difficult to dis-
tinguish the fundamental metal–ligand stretching vibra-
tions from low-frequency infrared spectra. However, the
far-IR spectra of compounds (3), (6) and (7) clearly
The tacticity of the PVC obtained from the polymer-
ization of vinyl chloride with (3) was determined by the
Table 1
Results of ethylene and vinyl chloride polymerization by 3 and 7
Polymera
Activityb
Mw (·103/mol)c
Mw/Mn
Polyethylene
Polyvinyl chloride
47(A) 21(B)
9(A) 4(B)
32.8(A) 10.2(B)
8.8(A) 9.5(B)
1.8(A) 1.6(B)
2.2(A) 2.1(B)
a
A result is for (3) and B result is for (7), polymerization conditions: ratio of catalyst and co-catalyst ([Al]/[Zr]), 2000; solvent, toluene; tem-
perature, 50 ꢁC; pressure, 1.5 bar (polyethylene) and 1.2 bar (polyvinyl chloride); polymerization time, 2 h.
Activity, kg polymer per mol catalyst per h per bar.
b
c
Molecular weight and molecular weight distribution of the polymers were determined by means of gel-permeation chromatography (GPC: Waters
150 ꢁC) at 145 ꢁC using 1,2,4-trichlorobenzene as a solvent. The weight average molecular weight and polydispersity index (Mw and Mw/Mn,
respectively) were calculated on the basis of polystyrene standards.

6288
Z. Yinghuai et al. / Journal of Organometallic Chemistry 690 (2005) 6284–6291
13C NMR spectra obtained in a mixed solvent of ben-
zene-d6 and o-dichlorobenzene from the area of each
split peak of the methine carbon [32]. The tacticity of
the PVC thus determined was: mm = 16%, mr = 47%,
rr = 37%, respectively, which are close to those obtained
with the conventional radical polymerization reported
(mm = 12%, mr = 49%, rr = 39%) [32].
zophenone until a blue color was obtained, and then dis-
tilled under argon just before use. N-butyllithium (1.6 M
in hexanes), sodium hydride (60% dispersion in mineral
oil), zirconium (IV) chloride bis(tetrahydrofuran), tita-
nium (IV) chloride bis(tetrahydrofuran). Merrifieldꢀs
peptide resin (1%) (contains 1.97 mmol Clꢀ/g, 100–200
mesh, cross-linked), ethylene, vinyl chloride, AIBN
and organic solvents were used as received from Aldrich.
Modified methylaluminoxane 7 (MMAO-7, 13% ISO-
PAR-E) was obtained from Akzo Chemicals Inc.
Ortho-carborane and 1-phenyl-ortho-C2B10H11 were
provided by Katchem Ltd. 7-Azabicyclo[4.1.0]heptane
was prepared according to the literature [35]. Tri(n-bu-
tyl)tin hydide (Bu3SnH) was synthesized according to
the literature [36]. 1H, 13C and 11B NMR were recorded
on a Bruker Fourier-Transform multinuclear spectrom-
eter at 400, 100.6 and 128.4 MHz, relative to external
Me4Si (TMS) and BF3 Æ OEt2 standards. Infrared (IR)
spectra were measured using a BIO-RAD spectropho-
tometer with KBr pellets technique and presented in
the sequence of signal strength as strong (s), middle
(m) and weak (w), and peak model as single (s), multiple
(m) and broad (br). Far-infrared spectra were recorded
on a Nicolet-7199B spectrophotometer with CsI pellets
technique. Elemental analyses were determined by a Per-
kin– Elmer 2400 CHN elemental analyzer. The TGA
analyses were carried out on an SDT 2960 Simultaneous
DSC–TGA analyzer. ICP analysis was determined by a
VISTA-MPX, CCD Simultaneous ICP–OES analyzer.
Although a few examples of vinyl chloride polymeri-
zation catalyzed by group 4 metallocene complexes in
the presence of cocatalysts have been described, the
mechanism studies are very limited and not definitive
[30,33]. On the other hand, zircollocene complexes also
have been reported to be inactive for vinyl chloride poly-
merization due to the rapid b-Cl elimination by the
alkylated metallozirconium intermediate formed from
insertion of vinyl chloride to in situ produced metallozir-
conium cation species [34]. Considering the decapitated
ortho-carborane dianion ([7-R-8-R0-C2B9H9]) is isoelec-
tronic with Cpꢀ [7–10], we anticipate the complexes will
show similar chemical properties with corresponding
metallozirconium analogues. Thus, it is reasonable for
us to propose the vinyl chloride polymerization is caused
by a radical polymerization process, initiated by the
carboranyl zirconium complexes in our conditions. This
consideration raises the intriguing question of whether it
might be possible to produce group 4 carborane based
trianion complexes more capable for polar homo- or
comonomer polymerization than those produced via a
living radical polymerization process. The detailed
mechanistic study is undergoing in our lab using in situ
NMR.
4.1. Synthesis of closo-1-Ph-2-(20-aminocyclohexyl)-1,2-
C2B10H10 (1)
3. Conclusions
A solution of 1.40 g (6.32 mmol) of 1-Ph-1,2-
C2B10H11 dissolved in a mixture of 60 ml diethyl ether
and 30 ml benzene was cooled to ꢀ78 ꢁC and 4.20 ml
(6.72 mmol) of n-BuLi (1.6 M in n-hexane) was added
with syringe. After addition and a half hour further stir-
ring, the mixture was warmed to room temperature
spontaneously and continued stirring for 5 h followed
by addition of 0.67 g (6.84 mmol) of 7-azabicy-
clo[4.1.0]heptane with a syringe at 0 ꢁC, then the reac-
tion was proceeded at room temperature for another
12 h before hydrolysis with 10 ml of water. The organic
phase was separated with a separating funnel, and the
aqueous phase was extracted with 2 · 40 ml diethyl
ether. The combined organic phase was dried with
MgSO4 and filtered, the filtrate was dried under reduced
pressure and the resulting residue was recrystallized with
n-hexane to obtain 1.26 g waxy solid (1) in yield 63%.
Analytic data: Calc. (Found) for C14H27B10N (1): C,
52.96 (53.01); H, 8.57 (8.54); N, 4.41 (4.38%). 1H
NMR (CDCl3, ppm), d = 7.35–6.80 (m, 5H, C6H5),
3.34 (m, 1H, CH–N), 2.90–0.64 (m, br, 21H, CH,
B10H10, 4CH2, NH2). 13C NMR (DMSO-d6, ppm),
d = 141.06, 130.93, 128.89 and 126.93 (C6H5), 87.09
A new carborane based trianionic ligand, composed
of a dicarbollide dianion and cyclohexyl bridged termi-
nal amino group, has been coordinated with group 4
metals via a convenient, three-step process from com-
mercially available 1-Ph-1,2-C2B10H11 and 7-azabicyclo
[4.1.0] heptane. The new ligand has been successfully
immobilized onto Merrifieldꢀs peptide resin (1%), which
in turn could be metallizated to give supported zirco-
nium metallacarboranes. Both free and anchored zirco-
nium carborane complexes were found to be efficient
olefin polymerization catalysts. Their activities are
greater for unsubstituted monomers than for substituted
ones such as vinyl chloride.
4. Experimental
All synthetic procedures were carried out in inert
atmosphere with standard Schlenk techniques or glove
box. Tetrahydrofuran, diethyl ether, toluene, benzene,
pentane and n-hexane were heated over sodium and ben-

Z. Yinghuai et al. / Journal of Organometallic Chemistry 690 (2005) 6284–6291
6289
and 74.78 (Ccage), 56.00 (–CH–N), 35.99, 32.31, 25.40,
24.96, 23.06 (–CH–Ccage, 4CH2). 11B NMR (DMSO-
0 ꢁC, the mixture was stirred at room temperature for
3 days. After filtration and removal of all the solvents
under reduced pressure, the obtained residue was recrys-
tallized with a mixture of benzene/pentane (v:v = 2:1) to
give 1.21 g (3) in 53% yield. Analytic data: Calc.
(Found) for C14H25B9ClNZr (3): C, 38.98 (38.95); H,
5.84 (5.80); N, 3.25 (3.22). 1H NMR (DMSO-d6,
ppm), d = 7.23–6.68 (m, 5H, C6H5), 3.11 (m, 1H,
1
d6, ppm), d = ꢀ2.34 (1B, JBH = 153 Hz), ꢀ4.84 (1B,
1
1JBH = 141 Hz), ꢀ9.37 (2B, JBH = 93 Hz), ꢀ11.11 (6B,
1JBH = 160 Hz). IR (KBr pellet, cmꢀ1), 3548(m, s),
3208(s, s), 2936(s, s), 2859(s, s), 2565(vs, s), 1701(m, s),
1598(m, s), 1492(s, s), 1447(s, s), 1390(m, s), 1328(s, s),
1195(m, s), 1095(m, s), 1051(s, s), 1003(s, s), 963(m,
br), 900(m, s), 820(m, br), 760(m, s), 694(s, s), 551(w,
s), 490(w, s).
–CH–N), 2.07–0.25 (m, br, 19H, B9H9, –CH–Ccage
,
4CH2, NH). 13C NMR (DMSO-d6, ppm), d = 138.83,
127.75, 126.58 and 124.68 (C6H5), 75.84 and 62.61
(Ccage), 38.46 (–CH–N), 36.51, 27.36, 23.88, 22.14,
20.14 (–CH–Ccage, 4CH2). 11BNMR (C6D6, ppm),
4.2. Synthesis of K[nido-7-Ph-8-(20-aminocyclohexyl)-
7,8-C2B9H10] (2)
1
1
d = ꢀ25.30 (1B, JBH = 107 Hz), ꢀ16.71 (2B, JBH
=
1
1.86 g (5.85 mmol) of (1) was added to a clear solu-
tion of 2.00 g (32.08 mmol) potassium hydroxide in
65 ml of 95% ethanol with continuous stirring. When
all solids dissolved, the obtained solution was heated
to reflux for 20 h. The vessel was cooled to room tem-
perature, bubbled with carbon dioxide and filtered.
The filtrate was dried under reduced pressure and the
sticky residue was then extracted with anhydrous
THF. After filtration and drying in vacuum, 1.62 g of
colourless sticky residue (2) was obtained in 80% yield.
Analytic data: Calc. (Found) for C14H27B9KN (2): C,
48.63 (48.60); H, 7.87 (7.89); N, 4.05 (4.03%). 1H
NMR (DMSO-d6, ppm), d = 7.43–6.92 (m, 5H, C6H5),
3.29 (m, 1H, CH–N), 2.10–0.15 (m, br, 20H, B9H9,
–CH, 4CH2, NH2), ꢀ2.73 (br, 1H, BHbridge). 13C
NMR (DMSO-d6, ppm), d = 139.10, 127.97, 126.61
and 124.84 (C6H5), 85.34 and 72.10 (Ccage), 53.60
130 Hz), ꢀ9.02 (4B, JBH = 126 Hz), ꢀ6.27 (2B,
1JBH = 95 Hz). IR (KBr pellet, cmꢀ1), 3433(m, s),
2905(vs, s), 2854(s, s), 1722(m, s), 1623(s, s), 1490(s, s),
1438(s, s), 1388(m, s), 1315(s, s), 992(m, s), 803(m, br),
759(m, s), 688(s, s), 535(w, s).
4.4. Synthesis of closo-1-Ti(Cl)-2-Ph-3-(20-r-(H)N-
cyclohexyl)-2,3-g5-C2B9H9 (4)
A similar process of the preparation of (3) was
used to synthesize 0.97 g (4) in 42% yield from
2.06 g (5.96 mmol) (2), 7.65 ml (12.24 mmol) n-BuLi
(1.6 M in hexanes) and 2.05 g (5.95 mmol) Ti-
Cl4 Æ 2THF in 75 ml dry THF. Analytic data: Calcd.
(Found) for C14H25B9ClNTi (4): C, 43.34 (43.31); H,
6.50 (6.47); N, 3.61 (3.63). 1H NMR (DMSO-d6,
ppm), d = 7.15–6.53 (m, 5H, C6H5), 3.25 (m, 1H,
(–CH–N), 34.31, 30.34, 24.32, 24.05, 22.62 (–CH–Ccage
,
–CH–N), 2.07–0.15 (m, br, 19H, B9H9, –CH–Ccage,
4CH2). 11B NMR (DMSO-d6, ppm), d = ꢀ36.50 (1B,
4CH2, NH). 13C NMR (DMSO-d6, ppm), d = 140.16,
127.83, 126.74 and 124.90 (C6H5), 76.20 and 62.35
(Ccage), 35.09 (–CH–N), 34.40, 28.45, 25.48, 23.37,
22.26 (–CH–Ccage, 4CH2). 11B NMR (C6D6, ppm),
d = ꢀ23.15 (1B, 1JBH = 117 Hz), ꢀ15.67 (2B,
1
1JBH = 145 Hz), ꢀ31.20 (1B, JBH = 104 Hz), ꢀ19.30
1
1
(1B, JBH = 110 Hz), ꢀ17.22 (2B, JBH = 114 Hz),
ꢀ14.29 (1B, 1JBH = 121 Hz), ꢀ9.90 (2B, 1JBH = 135 Hz),
1
ꢀ6.04 (1B, JBH = 143 Hz). IR (KBr pellet, cmꢀ1),
1
3504(m, br), 3210(m, s), 3061(w, s), 2933(s, s), 2856(m,
s), 2534(vs, s), 1599(m, s), 1445(s, s), 1379(m, s),
1195(m, s), 1092(m, s), 1045(s, s), 963(m, s), 903(m, s),
755(m, s), 701(m, s), 548(w, s).
1JBH = 125 Hz), ꢀ8.29 (4B, JBH = 121 Hz), ꢀ5.77
1
(2B, JBH = 107 Hz). IR (KBr pellet, cmꢀ1), 3488(m,
s), 3105(s, s), 3010(s, s), 2894(s, s), 2700(vs, s),
1680(s, s), 1518(s, s), 1401(s, s), 1299(w, br), 953(m,
br), 713(m, s), 457(s, s).
4.3. Synthesis of closo-1-Zr(Cl)-2-Ph-3-(20-r-(H)N-
cyclohexyl)-2,3-g5-C2B9H9 (3)
4.5. Synthesis of closo-1-(20-aminocyclohexyl)-1,2-
C2B10H11 (5)
1.83 g (5.28 mmol) of (2) was dissolved in 80 ml of
dry tetrahydrofuran, the obtained mixture was cooled
to ꢀ78 ꢁC and 6.93 ml (11.09 mmol) of n-BuLi (1.6 M
in hexanes) was added with syringe carefully. After addi-
tion, the mixture was kept reacting at that temperature
for 30 min before warmed to room temperature for
8 h. Stirring was then stopped and the reaction mixture
was allowed to cool to 0 ꢁC and 2.00 g (5.25 mmol)
ZrCl4 Æ 2THF was added. The colour of the mixture
changed quickly from pale yellow into red brown during
the addition process. After reaction about 30 min at
A similar process of the preparation of (1) was used
to produce 1.98 g (5) in 59% yield from 2.00 g
(13.87 mmol) ortho-carborane, 9.10 ml (14.56 mmol) n-
BuLi (1.6 M in hexanes) and 1.41 g (14.51 mmol) 7-aza-
bicyclo[4.1.0]heptane in 60 ml THF. Analytic data: Calc.
(Found) for C8H23B10N (5): C, 39.81 (39.79); H, 9.60
(9.53); N, 5.80 (5.77%). 1H NMR (DMSO-d6, ppm),
d = 3.30 (m, 1H, CH–N), 0.53–2.80 (m, br, ill-defined,
22H, B10H10, CH–Ccage, CcageH, 4CH2, NH2). 13C
NMR (DMSO-d6, ppm), d = 80.23 and 72.53 (Ccage),

6290
Z. Yinghuai et al. / Journal of Organometallic Chemistry 690 (2005) 6284–6291
52.36 (–CH–N), 35.07, 31.93, 25.25, 25.09, 24.61
(–CH–Ccage, 4CH2). 11B NMR (DMSO-d6, ppm),
saturated with carbon dioxide and filtered. The result-
ing solid was washed with deionized water (3 · 10 ml)
to remove any trace of KCl, and dried in high vac-
uum for two days before resuspended in 120 ml
THF. The mixture was cooled to ꢀ78 ꢁC and 6.5 ml
(10.40 mmol) n-BuLi in hexanes was added with con-
tinuous stirring. After 30 min at ꢀ78 ꢁC, the vessel
was warmed to room temperature spontaneously and
kept stirring for 1 day. Solvent was removed under
reduced pressure and the resulting residue was washed
with n-hexane (2 · 15 ml) to remove unreacted n-BuLi.
The residue was resuspended in 150 ml THF and
ZrCl4 Æ 2THF 1.10 g (2.89 mmol) was added at 0 ꢁC.
The reaction was continued at room temperature for
3 days before filtration. The collected solids was
washed with absolute ethyl alcohol (2 · 10 ml) to re-
move lithium chloride followed by dried in high vac-
uum for 2 days to give 2.40 g (7) (38% yield) as an
pale yellow solid. The collected alcohol solutions were
combined and dried in vacuum. The obtained residue
was dissolved in 5.0 ml deionized water and subjected
to titration with AgNO3 to determine the amount of
the produced LiCl as 0.246 g. The loading amount
of zirconium species reached up to 0.46 mmol (zirco-
nium species)/g (Merrifieldꢀs peptide resin). The result
is consistent with data derived from ICP analysis
which give 42.00 mg (zirconium)/g (Merrifieldꢀs pep-
tide resin (1%)). IR (KBr pellet, cmꢀ1): 3428(m, br),
3019(s, s), 2928(s, s), 2850(s, s), 2513(s, s), 1641
(s, s), 1525(s, s), 1490(s, s), 1408(m, s), 1358(s, s),
1161(s, s), 1088(m, s), 1010(s, s), 745(s, s), 701(s, s),
520(s, s).
1
1
d = ꢀ4.47 (1B, JBH = 127 Hz), ꢀ5.80 (1B, JBH
=
1
114 Hz), ꢀ9.11 (2B, JBH = 124 Hz), ꢀ9.74 (2B,
1
1JBH = 103 Hz), ꢀ10.62 (4B, JBH = 104 Hz). IR (KBr
pellet, cmꢀ1), 3450(s, s), 3089(s, s), 2937(s, s), 2859(s,
s), 2573(s, s), 2527(vs, s), 1640(s, sz), 1449(s, s),
1396(m, s), 1274(w, s), 1213(m, s), 1075(s, s), 1051(s,
s), 962(m, s), 915(m, s), 860(m, s), 727(m, s), 615(m,
br), 553(m, s), 506(m, s), 416(m, s).
4.6. Synthesis of closo-1-(20-aminocyclohexyl)-2-
polystyryl-1,2-C2B10H10 (6)
0.97 g (4.02 mmol) of (5) was added to a 250 ml
two-necked round bottom flask equipped with a mag-
netic stirring bar and dissolved with 100 ml THF.
The resulting solution was cooled to ꢀ30 ꢁC and
0.16 g (4.00 mmol) NaH was added carefully. After
addition, the mixture was kept reacting at that temper-
ature for 30 min followed by warming up to room tem-
perature spontaneously and reacting further 6 h.
Merrifieldꢀs peptide resin (1%), 2.00 g (contains
3.94 mmol Clꢀ), was added and kept stirring at room
temperature for two days. At the end of the reaction
process, the mixture was heated to reflux for 10 h be-
fore cooled to room temperature and quenched with
5.00 ml methanol. The solvents was then removed un-
der reduced pressure and the obtained crude product
was washed with deionized water (2 · 10 ml) and n-
hexane (2 · 20 ml) to remove any trace of NaCl and
starting
material
closo-1-(20-aminocyclohexyl)-1,2-
C2B10H11, and dried in high vacuum for two days to
give 2.72 g pale yellow solid closo-1-(20-amino-
cyclohexyl)-2-polymeryl-1,2-C2B10H10 (6) in 87% yield.
The collected aqueous solution was combined and con-
centrated to 5.0 ml and subjected to titration with
AgNO3 to detect the amount of the produced NaCl
as 0.204 g. The loading amount of aminocyclohexyl
carborane reached up to 1.28 mmol (carborane cage)/
g (Merrifieldꢀs peptide resin). IR (KBr pellet, cmꢀ1):
3430(m, s), 3013(s, s), 2916(s, s), 2576(vs, s), 1934(s,
s), 1800(m, s), 1611(s, s), 1485(s, s), 1422(m, s),
1364(s, s), 1304(m, s), 1252(s, s), 1178(m, s), 1150(m,
s), 1109(m, s), 1060(s, s), 1022(s, s), 905(s, s), 816(s,
s), 748(s, s), 690(s, s), 536(s, s).
4.8. Polymerization procedures
The polymerization of ethylene and vinyl chloride
catalyzed by compounds (3) and (7) was performed
for 2 h in 80 ml toluene in the presence of modified
methylaluminoxane 7 (MMAO-7, 13% ISOPAR-E) in
Schlenk tubes. The argon pressure inside the Schlenk
tubes was reduced by applying vacuum. Monomer
pressure was then applied to the tubes and the reactor
was adjusted to constant temperature and pressure.
During the polymerization process, 1 lmol catalyst (3
or 7) based on zirconium species and 2000 lmol
MMAO-7 were used to give a concentration ratio of
[Al]/[Zr] = 2000. The reaction temperature was 50 ꢁC
under 1.5 and 1.2 bar for polyethylene and polyvinyl
chloride, respectively. After polymerization process,
the reaction was carefully quenched with 10 ml mixture
of 10% HCl solution of MeOH. The polymer was then
precipitated with 150 ml methanol, collected by filtra-
tion, washed with MeOH (4 · 20 ml) and n-hexane
(2 · 20 ml) in sequence and dried at 60 ꢁC in high vac-
uum to a constant weight. Polymerization results are
summarized in Table 1.
4.7. Synthesis of closo-1-Zr(Cl)-2-(20-r-(H)N-
cyclohexyl)-3-polystyryl-2,3-g5-C2B9H9 (7)
2.27 g (contains carborane ligand 2.90 mmol) (6)
was suspended in a solution of 1.10 g (16.66 mmol)
potassium hydroxide in 150 ml of 95% ethanol, the
mixture was stirred at room temperature for 30 min
before refluxing for 2 days. The reaction mixture
was then cooled to room temperature, bubbled and

Z. Yinghuai et al. / Journal of Organometallic Chemistry 690 (2005) 6284–6291
6291
[10] G.C. Bazan, W.P. Schaefer, J.E. Bercaw, Organometallics 12
(1993) 2126.
4.9. Characterization of PVC
[11] G. Henrici-Olive, S. Olive, Angew. Chem., Int. Ed. 6 (1967) 790.
[12] H.A. Schroeder, Inorg. Macromol. Rev. 1 (1970) 45.
[13] A.M. Thayer, Chem. Eng. News 11 (1995) 15.
[14] R. Uhrhammer, Y.-X. Su, D.C. Swenson, R.F. Jordan, Inorg.
Chem. 33 (1994) 4398.
The reduction of PVC to the corresponding hydro-
carbon polymer was carried out by following process,
0.80 g of above produced PVC, 5.0 g (17.18 mmol)
Bu3SnH, 24 mg (1.46 · 10ꢀ4 mol) of AIBN, and 60 mL
of THF were added to a 100-mL glass autoclave, the
mixture was then degassed three time and filled with ar-
gon at last. After stirring for 8 h at 100 ꢁC, the reaction
solution was cooled to 80 ꢁC and added into 200 mL hot
methanol (62 ꢁC) and then cooled until room tempera-
ture. The product was collected by filtration and purified
by threefold precipitation from hot THF to MeOH. The
purified product was dried for 2 days in high vacuum to
a constant weight of 0.29 g (81%). The tacticity of
obtained PVC was analyzed from 13C NMR spectra
measured in the mixed solvent of benzene-d6 and
o-dichlorobenzene at 120 ꢁC.
[15] C.K. Broder, A.E. Goeta, A.K. Howard, A.K. Hughes, A.L.
Johnson, J.M. Malget, K. Wade, J. Chem. Soc., Dalton Trans.
(2000) 3526.
[16] A.K. Saxena, J.A. Maguire, N.S. Hosmane, Chem. Rev. 97 (1997)
2421.
[17] Y.-X. Su, C.E. Reck, I.A. Guzei, R.F. Jordan, Organometallics
19 (2000) 4858.
[18] X. Bei, C. Kreuder, D.C. Swenson, R.F. Jordan, V.G. Young Jr.,
Organometallics 17 (1998) 1085.
[19] Z. Xie, Z. Liu, K. Chiu, F. Xue, T.C.W. Mak, Organometallics
16 (1997) 2460.
[20] X. Bei, V.G. Young Jr., R.F. Jordan, Organometallics 20 (2001)
355.
[21] J.-S. Park, D.-H. Kim, S.-J. Kim, J. Ko, S.H. Kim, S. Cho, C.-H.
Lee, S.O. Kang, Organometallics 20 (2001) 4483.
[22] (a) W. Kwong, H. Chan, Y. Tang, Z. Xie, Organometallics 23
(2004) 4301;
(b) Y. Wang, H. Wang, H.-W. Li, Z. Xie, Organometallics 21
(2002) 3311.
Acknowledgments
[23] J.-S. Park, D.-H. Kim, J. Ko, S.H. Kim, S. Cho, C.-H. Lee, S.O.
Kang, Organometallics 20 (2001) 4632.
We thank ICES in Singapore for support of the work.
We also acknowledge Professor Narayan S. Hosmane in
the Department of Chemistry and Biochemistry in
Northern Illinois University, and Professor John A.
Maguire in the Department of Chemistry, Southern
Methodist University, for their helpful discussions. San-
dia is a multiprogram laboratory operated by Sandia
Corporation, a Lockheed Martin Company, for the
US DOEꢀs National Nuclear Security Administration
under Contract DE-AC04-94-AL85000.
[24] Y. Zhu, K. Vyakaranam, J.A. Maguire, W. Quintana, F.
Teixidor, C. Vinas, N.S. Hosmane, Inorg. Chem. Commun. 4
(2001) 486.
[25] Y. Zhu, J.A. Maguire, N.S. Hosmane, Inorg. Chem. Commun. 5
(2002) 296.
[26] D.-H. Kim, J.H. Won, S.-J. Kim, J. Ko, S.H. Kim, S. Cho, S.O.
Kang, Organometallics 20 (2001) 4298.
[27] (a) R.J.H. Clark, C.S. Williams, Spectrochim. Acta 22 (1966)
1081;
(b) C.F. Robert, J.P. Thomas, Inorg. Chem. 7 (1968) 508.
[28] J. Zheng, Z. Xu, S. Chen, Beijing Daxue Xuebao, Ziran Kexueban
35 (1999) 855.
[29] P.J.T. Tait, M.G.K. Monteiro, M. Yang, J.L. Richardson,
Proceedings of MetCon 1996, Houston, TX, 12–13 June 1996
(available from Catalyst Consultants, P.O. Box 637, Spring House
PA 19477).
References
[1] U. Siemeling, Chem. Rev. 100 (2000) 1495.
[30] K. Endo, M. Saitoh, J. Polym. Sci. Part A: Polym. Chem. 41
(2003) 248.
[2] A.L. McKnight, R.M. Waymouth, Chem. Rev. 98 (1998) 2587.
[3] P.J. Shapiro, W.D. Cotter, W.P. Schaefer, J.A. Labinger, J.E.
Bercaw, J. Am. Chem. Soc. 116 (1994) 4623.
[31] W.H. Starnes Jr., R.L. Hartless, F.C. Schilling, F.A. Bovey,
Polym. Prep. (AM. Chem. Soc. Div. Polym. Chem.) 18 (1977) 499.
[32] M.W. Crowther, N.M. Szevernyi, G.C. Levy, Macromolecules 19
(1986) 1333.
[4] H.G. Alt, A. Reb, K. Kundu, J. Organomet. Chem. 628 (2001)
211.
[5] J. Klosin Jr., W.J. Kruper, P.W. Nickias, G.R. Roof, P.De.
Waele, K.A. Abbound, Organometallics 20 (2001) 2663.
[6] F. Stephen, in: Plastic: The Making of a Synthetic Century,
HarperCollins, New York, 1996.
[33] (a) K. Taeji, T. Uozumi, K. Soga, Polym. Prep. Jpn. 44 (1995) 2;
(b) T. Matsukawa, R. Kiba, T. Ikeda, Japan Patent 08208736;
T. Matsukawa, R. Kiba, T. Ikeda, Chem. Abstr. 125 (1996)
276902.
[7] D.J. Crowther, N.C. Baenziger, R.F. Jordan, J. Am. Chem. Soc.
113 (1991) 1455.
[34] R.A. Stockland Jr., R.F. Jordan, J. Am. Chem. Soc. 122 (2000)
6315.
[8] S. Saccheo, G. Gioia, A. Grassi, D.E. Bowen, R.F. Jordan, J.
Mol. Cat. A 128 (1998) 111.
[35] J. Christoffers, Y. Schulze, J. Pickardt, Tetrahedron 57 (2001)
1765.
[9] T. Doge, M.A. Curtis, J.M. Russell, M. Sabat, M.G. Finn, R.N.
Grimes, J. Am. Chem. Soc. 122 (2000) 10573.
[36] J. Szammer, L. Otvos, Chem. Ind. (London, United Kingdom) 23
(1988) 764.