Efficient synthesis and structural characterization of a post-metallocene α-olefin polymerization catalyst

Article abstract of DOI:10.1016/j.ica.2008.02.034

We describe the fast one-pot synthesis of a chelating diamido dichloride titanium compound through sequential addition of the protonated ligand and triethylamine to TiCl₄. The bisamido complex has been structurally characterized. A monoamido in

Full text of DOI:10.1016/j.ica.2008.02.034

Available online at www.sciencedirect.com
Inorganica Chimica Acta 362 (2009) 277–280
www.elsevier.com/locate/ica
Note
Efficient synthesis and structural characterization of a post-metallocene
a-olefin polymerization catalyst
R.M. Gauvin *, A. Arbaoui, E. Gautier, A. Mortreux, E. Berrier, G. Nowogrocki
Ecole Nationale Supe´rieure de Chimie de Lille, Baˆt. C7, BP 90108, 59652 Villeneuve d’Ascq Cedex, France
ˆ
CNRS, UMR 8181, Unite´ de Catalyse et de Chimie du Solide, Bat. C7, BP 90108, 59652 Villeneuve d’Ascq Cedex, France
Received 18 December 2007; received in revised form 11 February 2008; accepted 19 February 2008
Available online 26 February 2008
Abstract
We describe the fast one-pot synthesis of a chelating diamido dichloride titanium compound through sequential addition of the pro-
tonated ligand and triethylamine to TiCl4. The bisamido complex has been structurally characterized. A monoamido intermediate has
been isolated.
Ó 2008 Elsevier B.V. All rights reserved.
Keywords: Titanium; Amido ligand; Olefin polymerization; X-ray structure
1. Introduction
[8]. Further studies using this family of catalysts for poly-
merization and copolymerization have been undertaken
by Soga and co-workers, demonstrating the high potential
of such catalysts in the manufacturing of poly(a-olefins)
[11–13]. Chiral versions of these precatalysts have been
developed by Mair in view of their application in isoselec-
tive 1-hexene polymerization [14]. Thus, the development
of a simple, rapid synthesis of this type of catalyst precur-
sor is of interest. We describe here the synthesis and struc-
tural characterization of such a species following a simple
and fast procedure, which involves the diamine ligand pre-
cursor, TiCl4 and NEt3.
In the continuously evolving field of Ziegler–Natta ole-
fin polymerization, the development of new homogeneous
catalysts where the metallocene cyclopentadienyl ligands
are replaced by other functional groups, the so-called
‘‘post-metallocene catalysts”, has attracted considerable
attention over the recent years [1,2]. Among the explored
families, group 4 amido derivatives have been most partic-
ularly interesting. The potential of these compounds
started to be recognized in the mid 1990s [3–5]. McConville
and Schrock both made an impact in the field of post-
metallocene catalysts by disclosing amido-based systems
displaying outstanding performances in terms of activity
and living character for the polymerization of terminal ole-
fins [6–10]. More precisely, McConville showed that tita-
nium derivatives 1R and 2R bearing a bulky, chelating
bisamide ligand with a C3 bridge could polymerize 1-hex-
ene with high activity and in a living fashion (Scheme 1)
2. Results and discussion
The synthesis of 1R complexes reported by McConville
and co-workers comprises two steps. The ligand precursor
(LR)H2 is first deprotonated with MeLi and silylated into
[LR](SiMe3)2. This moisture sensitive diamine is isolated
and reacted with TiCl4 in refluxing xylenes for
12 h to afford 1R with concomitant release of SiMe3Cl
(Scheme 2). The overall reported yield is 75%, with a cumu-
lative reaction time of more than 24 h.
*
´
Corresponding author. Address: CNRS, UMR 8181, Unite de
Catalyse et de Chimie du Solide, Baˆt. C7, BP 90108, 59652 Villeneuve
d’Ascq Cedex, France. Tel.: +33 0 3 20 43 67 54; fax: +33 0 3 20 43 65 85.
E-mail address: regis.gauvin@ensc-lille.fr (R.M. Gauvin).
0020-1693/$ - see front matter Ó 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.ica.2008.02.034

278
R.M. Gauvin et al. / Inorganica Chimica Acta 362 (2009) 277–280
Table 1
R
Crystal data and structure refinement details of [TiCl2 (N2C19H24)]
R
N
Empirical formula
Formula weight
Temperature (K)
Crystal system
Space group
Unit cell dimensions
C19H24N2Cl2Ti
399.20
100.0(3)
monoclinic
P21/c
cat. (1R or 2R)
activator
X
X
Ti
n
N
C4H9
C4H9
R
R
productivity upto
350000kgpol/molcat/h
1R: X = Cl
2R: X = Me
˚
a (A)
14.624(3)
7.696(2)
17.724(4)
90.00
105.168(3)
90.00
1925.2(7)
˚
b (A)
˚
c (A)
Scheme 1. Titanium bisamide precatalysts developed by McConville for
a-olefin polymerization.
a (°)
b (°)
c (°)
3
˚
Volume (A )
Z
4
R
R
N
R
N
Dcalc (Mg/m3)
1.377
0.725
832
R
R
R
SiMe3
SiMe3
R
Absorption coefficient (mmꢀ1
F(000)
Crystal dimensions (mm)
h range for data collection (°)
Index ranges
)
NH
1) 2.2 MeLi
TiCl4, 130°C
Cl
Cl
Ti
2) 2.2 SiMe3Cl
NH
R
N
R
N
R
0.35 ꢁ 0.25 ꢁ 0.19
2.38–25.99
ꢀ18 6 h 6 18
ꢀ9 6 k 6 9
ꢀ21 6 l 6 21
11605
3568 (0.057)
3568/0/289
1.070
R
R
[LR]H2
[LR](SiMe3)2
1R
Reflections collected
Scheme 2. Synthesis of complexes 1R reported by McConville.
Independent reflections (Rint
Data/restraints/parameters
Goodness-of-fit on F2
)
We considered in following a more direct route which
would not involve the isolation of an intermediate com-
pound. Thus, direct reaction of the [LMe]H2 ligand precur-
sor and TiCl4 was performed, and followed by
triethylamine addition (Scheme 3). Filtration of the reaction
mixture and recrystallization after stripping-off the volatiles
afforded 1Me as a red crystalline solid in a 84% yield.
Single crystals were grown from a toluene/pentane solu-
tion and were subjected to X-ray diffraction studies. The
Final R indices [I > 2r(I)]
R indices (all data)
Largest difference in peak and hole (e A
R1 = 0.0429, wR2 = 0.0883
R1 = 0.0731, wR2 = 0.0961
0.351 and ꢀ0.361
ꢀ3
˚
)
ORTEP representation of the solid-state structure of 1Me
is displayed in Fig. 1. Table 1 gathers crystal data and
structure refinement details.
The solid-state structure of 1Me is similar to that of the
dimethyl derivative 2Me described by McConville [6]. The
metal adopts a distorted tetrahedral configuration. The
chelating diamido ligand adopts a boat-type configuration.
As a consequence of the aryl groups’ orientation, the
methyl groups shield the metal above and below the N–
Ti–N plane. As expected from amido functions, the nitro-
gens are planar. The Ti1–N1 and Ti1–N2 bond distances
Me
NH
Me
Me
Me
Me
Cl
N
1) TiCl4
2) NEt3
Ti
Cl
Me
NH
Me
N
Me
˚
of 1.849(2) and 1.854(2) A, respectively, are typical of tita-
nium amido bonds [6,14,15]. The Ti1–Cl1 and Ti–Cl2 dis-
tances (2.2548(10) and 2.2755(9) A, respectively) are
[LMe]H2
1Me
˚
Scheme 3. New access route to complex 1Me
.
unremarkable. The N1–Ti–N2 chelating angle of
100.68(11)° is sharper than that of the dimethyl derivative
2Me of 102.6(3)° [6]. The same angle is also wider than in
the dichloride derivative of the more bulky bisisopropylaryl
ligand 1iPr (99.2(2)°) [6]. Accordingly, the Cl–Ti–Cl angle in
1Me (107.02(4)°) is slightly narrower than that of the iso-
propylaryl complex 1iPr (107.77(9)°).
C18
Cl1
C3
C2
C10
C5
C1
C13
C14
C12
In our procedure, the triethylamine is used to trap the
HCl released during the formation of the Ti–N bond, as
it could protonate a ligand’s nitrogen and prevent complete
reaction. The direct reaction between titanium tetrachlo-
ride and diols generating chelating alkoxide complexes
has been reported by Schaverien [16]. However, in this
case, in contrast to what we observe for the amido deriva-
tive 1Me, the use of a base to capture the released HCl gas is
N2
C4
N1
Ti1
C6
C7
C15
C9
C17
C11
C16
Cl2
C19
C8
Fig. 1. Molecular structure of 1Me at 50% probability level with hydrogen
atoms being omitted for clarity.

R.M. Gauvin et al. / Inorganica Chimica Acta 362 (2009) 277–280
279
shorter time than with the reported procedure. The reaction
has been shown to proceed via a monoamido complex.
Me
NH
Me
Cl
Me
Me
Me
Me
Me
N
N
1) TiCl4
2) NEt3
NEt3
Cl
Ti Cl
Cl
Ti
Me
Me
3. Experimental
Cl
Me
Cl
N
H
NH
Me
N
H
Me
3.1. General procedures
[LMe]H2
3Me
1Me
All experiments were carried out under an argon atmo-
sphere in a M-Braun glove box or by using standard
Schlenk techniques. TiCl4 was purchased from Strem
Chemicals. Solvents were dried under nitrogen using con-
ventional reagents, degassed by freeze–pump–thaw cycles
and stored in the glove box over 3 A molecular sieves. 1,3-
Bis(2,6-dimethylanilino)propane ((LMe)H2) was prepared
according to the published procedures [6]. NMR analyses
were run on a Bruker Avance 300 (1H: 300 MHz, 13C:
75 MHz) at room temperature and referenced to SiMe4.
Elemental analyses were carried out at the Service d’Ana-
lyse, LSEO, University of Dijon, France. ORTEP drawings
were generated by ORTEP-3 1.074 [20]. Infrared spectra were
Scheme 4. Formation, proposed structure and reactivity of 3Me
.
not only unnecessary, but also detrimental to the course of
the reaction, since the formation of the alkoxide does not
occur when NEt3 is used.
Reaction between [LMe]H2 and titanium tetrachloride in
toluene affords a brown precipitate, which was separated
by filtration and characterized by NMR, IR, Raman and
elemental analysis. Both proton and carbon NMR indicate
the presence of a non-symmetric [L] framework. Further-
more, the presence of a TiN(Ar)CH2 moiety is evidenced
by characteristic 1H and 13C signals (triplet centered at
4.35 ppm, and singlet at 57.48 ppm, respectively). In addi-
´ ´
acquired on a Nicolet Protege 460 equipped with a Harrick
1
tion, a broad H NMR signal centered at 7.80 ppm (2H),
cell. Micro-Raman spectra were recorded at room tempera-
ture using the 531.95 nm second harmonic line of a
Nd:YAG laser. A 50X microscope objective was used to
focus the excitation beam (13.6 lm spot) and collect the
scattered light at the same time. The scattered light was col-
lected through a confocal hole (150 lm) by a nitrogen-
cooled CCD (Labram Infinity, Jobin Yvon). The Raman
spectra of the materials were collected under inert condi-
tions by means of a home-made sealed cell loaded under
Ar and equipped with a quartz window.
along with a deshielded NCH2 multiplet (3.73 ppm, com-
pared to 3.12 ppm for [L]H2) are indicative of an anilinium
fragment, ArNH2–CH2, which is further confirmed by the
presence of a 13C NMR signal at 50.62 ppm. The infrared
spectrum comprises a broad peak at 3150 cmꢀ1, assigned to
m
N–H in anilinium ðRÞðArÞNH2þ. This is consistent with the
formulation of 3Me as a monoamido compound resulting
from a single Ti-amido bond formation and where the thus
released HCl would be trapped by the second nitrogen
atom into an anilinium group (see Scheme 4). Closely
related [TiCl3(NRR0)] compounds have been prepared by
3.1.1. [Ti(L)Cl2] (1Me
)
Burger and Neese, from TiCl and the corresponding sily-
¨
A solution of (L)H2 (0.500 g, 1.85 mmol) in 25 mL of
toluene was added to a solution of TiCl4 (0.370 mg,
1.95 mmol) in 25 mL of toluene cooled at ꢀ78 °C. The
reaction mixture immediately turned red–brown, and a
brown precipitate formed. The cooling bath was removed
once the addition was complete, and after 15 min, 2 mL
NEt3 (14.3 mmol) were slowly added under vigorous stir-
ring. Once the reaction mixture had reached room-temper-
ature, the dark red supernatant was separated by cannula
transfer. The residual precipitate was washed with toluene
(2 ꢁ 10 mL). The volume of the combined toluene fractions
was reduced under vacuum. After cooling to ꢀ25 °C, a first
crop of (0.450 g) dark red crystalline 1Me was obtained. A
second batch (0.171 g) was obtained after further concen-
tration and cooling of the residual mother liquor (com-
bined yield 84%). Spectroscopic data matched those
reported in the literature [6]. Raman (cmꢀ1): 843 (s), 716
(m), 579 (s), 525 (m), 355 (m), 278 (w).
4
lated amine with the release of SiMe3Cl [17]. Based on the
low-frequency region spectrum, the authors proposed a
polynuclear structure for such compounds. Indeed, the
Raman spectrum of 3Me comprises features similar to those
of these monoamido species: a strong intensity line at
302 cmꢀ1 (terminal Ti–Cl) and several low intensity signals
at 267 and 240 (br) cmꢀ1 (bridging TiCl). This tends to
indicate that 3Me exists as polynuclear entities. The rather
low wavenumber of the N–H vibration (3150 cmꢀ1) may
be indicative of some degree of hydrogen bonding, the
intra- or intermolecular nature being undetermined with-
out X-ray structure diffraction data [18,19].
The addition of excess triethylamine to a suspension of
3Me affords quantitatively (NMR) the bisamido derivative
1
Me: the formation of the second titanium amido function
results from trapping of two equivalents of HCl by NEt3.
1H NMR studies confirm the formation of NEt3HCl. Note-
worthy, this method could not be successfully extended to
the more sterically crowded (LiPr)H2 derivative, which did
not react under analogous reaction conditions.
In conclusion, we have shown that the synthesis of a post-
metallocene catalyst precursor can be carried out following
a very simple procedure under mild conditions in a much
3.1.2. X-ray structure of 1Me
A red irregular crystal of the studied complex was
mounted on a Bruker Smart-Apex II-CCD automatic dif-
fractometer equipped with a CCD detector, and used for
data collection. X-ray intensity data were collected

280
R.M. Gauvin et al. / Inorganica Chimica Acta 362 (2009) 277–280
with graphite-monochromated MoKa radiation (k =
0.71073 A) at the temperature of 100.0(3) K, with the x scan
Appendix A. Supplementary material
˚
mode. The full Ewald sphere reflections were collected up to
2h = 52°. The unit cell parameters were determined from
least-squares refinement of the setting angles of 259 strong
reflections. Details concerning crystal data and refinement
are given in Table 1. Lorentz, polarization, and numerical
absorption corrections were applied. The structure was
solved by direct methods [21–23]. All the non-hydrogen
atoms were refined anisotropically using full-matrix least-
squares on F2. The hydrogen atoms were found from differ-
ence Fourier synthesis and refined with individual isotropic
temperature factors equal to 1.2 (or 1.5 for methyl groups)
times the value of the equivalent temperature factor of the
parent atom. SHELXTL chain of programs was used for all
the calculations [24]. Atomic scattering factors were those
incorporated in the computer programs.
CCDC 666363 contains the supplementary crystallo-
graphic data for 1Me. These data can be obtained free of
charge from The Cambridge Crystallographic Data Centre
via www.ccdc.cam.ac.uk/data_request/cif. Supplementary
data associated with this article can be found, in the online
version, at doi:10.1016/j.ica.2008.02.034.
References
[1] V.C. Gibson, S.K. Spitzmesser, Chem. Rev. 103 (2003) 283.
[2] G.J.P. Britovsek, V.C. Gibson, D.F. Wass, Angew. Chem., Int. Ed.
Engl. 38 (1999) 428.
[3] J.A.M. Canich, H.W. Turner, WO-A 92/12162, 1992.
[4] A.D. Horton, J. de With, Chem. Commun. (1996) 1375.
[5] S. Tinkler, R.J. Deeth, D.J. Duncalf, A. McCamley, Chem. Commun.
(1996) 2623.
[6] J.D. Scollard, D.H. McConville, N.C. Payne, J.J. Vittal, Macromol-
ecules 29 (1996) 5241.
[7] J.D. Scollard, D.H. McConville, J. Am. Chem. Soc. 118 (1996) 10008.
[8] J.D. Scollard, D.H. McConville, J.J. Vittal, N.C. Payne, J. Mol.
Catal. A 128 (1998) 201.
[9] R. Baumann, W.M. Davis, R.R. Schrock, J. Am. Chem. Soc. 119
(1997) 3830.
[10] L.-C. Liang, R.R. Schrock, W.M. Davis, D.H. McConville, J. Am.
Chem. Soc. 121 (1999) 5797.
[11] J. Jin, S. Tsubaki, T. Uozumi, T. Sano, K. Soga, Macromol. Rapid
Commun. 19 (1998) 597.
[12] C.-H. Ahn, M. Tahara, T. Uozumi, J. Jin, S. Tsubaki, T. Sano, K.
Soga, Macromol. Rapid Commun. 21 (2000) 385.
[13] T. Uozumi, S. Tsubaki, J. Jin, T. Sano, K. Soga, Macromol. Chem.
Phys. 202 (2001) 3279.
[14] F. Heatley, F.S. Mair, R.G. Pritchard, R.J. Woods, J. Organomet.
Chem. 690 (2005) 2078.
[15] B. Tsuie, D.C. Swenson, R.F. Jordan, J.L. Petersen, Organometallics
16 (1997) 1392.
[16] A. van der Linden, C.J. Schaverien, N. Meijboom, C. Ganter, A.G.
Orpen, J. Am. Chem. Soc. 117 (1995) 3008.
[17] H. Burger, H.J. Neese, Z. Anorg. Allg. Chem. 365 (1969) 243.
¨
[18] For intramolecular N–H ꢂ ꢂ ꢂ Cl–M hydrogen bonding, see: M.
Gandelman, A. Vigalok, L.J.W. Shimon, D. Milstein, Organometal-
lics 16 (1997) 3981, and papers quoted in Ref. 16 of this article.
[19] For intermolecular N–H ꢂ ꢂ ꢂ Cl–Ti hydrogen bonding, see: I.A. Guzei,
C.H. Winter, Inorg. Chem. 36 (1997) 4415.
[20] L.J. Farrugia, J. Appl. Crystallogr. 30 (1997) 565.
[21] SMART Software Users Guide, version 5.0, Bruker Analytical X-ray
Systems, Inc., Madison, WI, USA, 1999.
[22] SAINT Software Users Guide, version 6.0, Bruker Analytical X-ray
Systems, Inc., Madison, WI, USA, 1999.
[23] G.M. Sheldrick, SADABS, Bruker Analytical X-ray Systems, Inc.,
Madison, WI, USA, 1999.
3.1.3. [Ti(LH2)Cl4] 3Me
To a solution of 34 mg of TiCl4 (1.8 ꢁ 10ꢀ4 mol) in tol-
uene (2 mL) was added dropwise at ꢀ20 °C a solution of
50 mg (L)H2 (1.77 ꢁ 10ꢀ4 mol) in toluene (2 mL). A brown
precipitate formed immediately, which was separated by fil-
tration and washed with pentane (2 ꢁ 2 mL) to afford
1
75 mg of a brown solid (89% yield). H NMR (CDCl3):
7.87 (s, br, 2H, NH2), 7.37 (m, 1H, CH Ar), 7.21 (m, 5H,
2
CH Ar), 4.35 (t, JH–H = 7.5 Hz, 2H, Ti–N–CH2), 3.73
(m, 2H, ArNH2–CH2), 2.46 (s, 6H, ArMe), 2.37 (m, 2H,
CH2CH2CH2), 2.27 (s, 6H, ArMe). 13C NMR (CDCl3):
143.61, 132.68, 130.88, 130.79, 130.59, 130.12, 129.89,
129.56 (C Ar), 57.48 (TiNCH2), 50.62 (ArNH2–CH2),
25.14 (CH2CH2CH2), 18.92 (ArMe), 17.48 (ArMe). IR (dif-
fuse reflectance, cmꢀ1): 3150 (w), 3058 (w), 2988 (w), 2968
(w), 2925 (w), 1573 (s), 1468 (m), 1383 (m), 1314 (w),
1262.33 (w), 1193 (w), 1169 (m), 1122 (w), 1094 (m), 1030
(m), 993 (m), 944 (m), 911 (m), 902 (m), 844 (m), 826
(m), 783 (s), 744 (m), 708 (m), 677 (m), 666 (m), 653 (m),
615 (m), 600 (m), 575 (m), 562 (m). Raman (cmꢀ1): 953
(m), 904 (s), 850 (s), 761 (w), 715 (s), 676 (w), 616 (s),
538 (s), 524 (m), 514 (m, br), 302 (s), 267 (w), 242 (br,
w). Anal. Calc for C19H26Cl4N2Ti: C, 48.34; H, 5.55; N,
5.93. Found: C, 45.95; H, 5.36; N, 5.80%.
Acknowledgements
`
We thank the CNRS, the ENSCL and the Ministere de
[24] G.M. Sheldrick, SHELXTL, version 5.10, Bruker Analytical X-ray
Systems, Inc., Madison, WI, USA, 1999.
l’Education Nationale et de la Recherche for their financial
support.