A remarkable inversion of structure-activity dependence on imido N-substituents with varying co-ligand topology and the synthesis of a new borate-free zwitterionic polymerisation catalyst

Article abstract of DOI:10.1039/b514467a

Ethylene polymerisation productivities of tris(pyrazolyl)methane-supported catalysts [Ti(NR){HC(Me₂pz)₃}Cl₂] show a dramatically different dependence on the imido R-group compared to those of their TACN analogues, attributed to differences in fac-N₃ donor topology; when treated with AlⁱBu₃, the zwitterionic tris(pyrazolyl)methide compound [Ti(N-2-C₆H₄ ^tBu){C(Me₂pz)₃}Cl(THF)] also acts as a highly active, single site catalyst (TACN = 1,4,7-trimethyltriazacyclononane). The Royal Society of Chemistry 2006.

Full text of DOI:10.1039/b514467a

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
A remarkable inversion of structure–activity dependence on imido
N-substituents with varying co-ligand topology and the synthesis of a
new borate-free zwitterionic polymerisation catalyst{
Helen R. Bigmore,a Stuart R. Dubberley,a Mirko Kranenburg,b Sally C. Lawrence,a Andrew J. Sealey,a
Jonathan D. Selby,a Martin A. Zuideveld,c Andrew R. Cowleya and Philip Mountford*a
Received (in Cambridge, UK) 12th October 2005, Accepted 3rd November 2005
First published as an Advance Article on the web 5th December 2005
DOI: 10.1039/b514467a
metal. Well defined monoalkyl cations have also been established
for these systems.4
Ethylene polymerisation productivities of tris(pyrazolyl)-
methane-supported catalysts [Ti(NR){HC(Me2pz)3}Cl2] show
a dramatically different dependence on the imido R-group
compared to those of their TACN analogues, attributed to
differences in fac-N3 donor topology; when treated with AliBu3,
the zwitterionic tris(pyrazolyl)methide compound [Ti(N-2-
Michiue and Jordan have recently shown how tris(pyrazolyl)-
borate Group 4 complexes [M{HB(R3R5pz)3}Cl3] (II, Chart 1;
M 5 Ti or Zr; R3R5pz 5 substituted pyrazolyl) generate highly
active ethylene polymerisation catalysts with interesting properties
on activation with MAO.5 Versatile as they are, however,
pyrazolylborate ligands6 are not without potential problems,
notably facile cleavage of the B–N bonds, as found recently in
Lee and Jordan’s cationic zirconium benzyl derivative
[Zr{HB(Me2pz)3}(CH2Ph)2]+.7
t
C6H4 Bu){C(Me2pz)3}Cl(THF)] also acts as a highly active,
single site catalyst (TACN 5 1,4,7-trimethyltriazacyclononane).
There continues to be considerable interest in ‘‘post-metallocene’’
olefin polymerisation catalysis.1 However, efficient terminal imido-
supported Ziegler type polymerisation catalysts have been
noticeably lacking from Group 4, despite being well-established
for Groups 5 and 6.2 We recently reported the very highly active
titanium imido polymerisation catalysts [Ti(NR)(TACN)Cl2] (I,
Chart 1). High throughput parallel screening of ca. 50 such
catalysts under commercially-relevant conditions (T 5 100 uC,
MAO activation) found that bulky alkyl imido R-substituents
Tris(pyrazolyl)methane ligands, as exemplified by HC(Me2pz)3
(III, Chart 1),8 are the neutral analogues of the tris(pyrazolyl)bo-
rates.6 These readily available ligands have been little used in early
transition metal chemistry. Furthermore, only one report of their
use in olefin polymerisation catalysis is known, this being a
preliminary communication of [Sc{HC(Me2pz)3}(CH2SiMe3)3].9
Two reports of Group 4 tris(pyrazolyl)methane compounds have
appeared,10,11 but no structural authentication accompanied them
and neither were in the context of catalytic chemistry. Here we
report preliminary studies of a library of tris(pyrazolyl)methane-
supported titanium imido compounds, their evaluation under
laboratory and commercially-relevant conditions, and steps
towards new borate-free zwitterionic catalysts.
(tBu, adamantyl and CMe2CH2 Bu) provided the highest activities
t
(.200 000 kg(PE) mol21 h21 bar21 for R 5 tBu) (PE =
polyethylene).3 None of the catalysts of type I with bulky aryl
i
t
R-groups 2,6-C6H3 Pr2, 2-C6H4CF3 or 2-C6H4 Bu showed any
significant activity (,100 kg(PE) mol21 h21 bar21). Type I
compounds with bulky alkyl R-groups are by far the most active
of all the imido-supported Ziegler polymerisation catalysts
reported to date; this study being the first comprehensive survey
of imido substituent–polymerisation activity relationships for any
ð1Þ
Eqn. 1 summarises the one pot synthesis (carried out in parallel
using semi-automated procedures) of the primary catalyst library
[Ti(NR){HC(Me2pz)3}Cl2] (1–22, 56–78% yield).{ The intermedi-
ates [Ti(NR)Cl2(NHMe2)2] were not isolated, but a number of
them or their homologues have been fully described elsewhere.12
All of the compounds 1–22 gave satisfactory high resolution
1
(accurate mass) MS and/or H NMR spectra. The compounds
Chart 1
showing the most promising polymerisation performance (3,10 6,
15, 16, 20–22: see Fig. 1) were fully characterised in the usual way.{
Poorly performing catalyst 9 was also scaled up. The catalytic
performance of a particular compound did not depend on the
method of synthesis.
aChemistry Research Laboratory, University of Oxford, Mansfield
Road, Oxford, UK OX1 3TA
bDSM Elastomers, Global R&D, PO Box 1130, 6160 BC, Geleen, The
Netherlands
cDSM Research, PO Box 18, 6160 MD, Geleen, The Netherlands
{ Electronic Supplementary Information (ESI) available: Characterising
data and details of the polymerisation experiments. See DOI: 10.1039/
b514467a
The primary library 1–22 was screened for homopolymerisation
of ethylene at room temperature using MAO co-catalyst (Al/
Ti 5 1500), and their productivities are illustrated in Fig. 1.{ For
436 | Chem. Commun., 2006, 436–438
This journal is ß The Royal Society of Chemistry 2006

View Article Online
also disfavour cation–anion and/or cation–AlMe3 association in
the catalyst system derived from 16 and MAO.4,15
The space filling representation of 16 (Fig. 2, right) suggests
that the pyrazolyl ring 3-methyl substituents play a key role in
the ‘‘self-constrained’’ geometry of 16 and the other aryl imido
complexes in Fig. 1. As a probe, selected homologues of 1–22
were prepared using the ring 3,5-non-substituted ligand
HC(Bupz)3 (Bupz 5 4-n-butylpyrazolyl), namely [Ti(NR)-
{HC(Bupz)3}Cl2] (R 5 tBu (23), R = 2,6-C6H3 Pr2 (24) or R =
i
2-C6H4CF3 (25)) (eqn. 2). The room temperature productivity
of 23 (150 kg(PE) mol21 h21 bar21) was not dissimilar to that of
3, but those of 24 and 25 were much reduced (300 and
260 kg(PE) mol21 h21 bar21, respectively) in comparison with
21 and 15. At 100 uC, 23 and 24 were, like 3 and 21, inactive.
Compound 25 produced PE (Mw 5 1 305 000; PDI 5 2.0) but
the productivity was only 1 770 kg(PE) mol21 h21 bar21 (cf.
133 700 kg(PE) mol21 h21 bar21 for 15). Therefore the ring
3-substituents are clearly important in correctly aligning the
imido-bound aryl rings. In support of this, the 1H NMR
spectrum of 24 showed no evidence for restricted rotation about
the Ar–NTi bond, even at 233 K (unlike that of 21, which shows
that this process is ‘‘frozen out’’ at room temperature—e.g.,
inequivalent iso-propyl group methine protons at 4.98 and
2.85 ppm).
Fig. 1 Room temperature ethylene polymerisation productivities for
MAO-activated [Ti(NR){HC(Me2pz)3}Cl2] (1–22).{
the most productive catalysts, the PE molecular weights (Mw) were
in the range 334 000–1 515 000 with PDIs (polydispersity indices)
in the range 3.2–4.8 (except for 15 and 21, which had values of 5.9
and 7.0, respectively). Only aryl imido compounds with either two
ortho-substituents or one very bulky ortho-substituent were very
highly productive; the bulky alkyl imido compounds 3 and 6 had
reasonable productivities under these conditions but were
extremely poor in comparison to their aryl imido homologues.
Evaluation of the more promising leads 3, 15, 16 and 20–22,
together with compounds 9, 10 and 12, under more industrially-
relevant conditions (100 uC, 7 bar C2H4, 10 min, MAO,
Al/Ti 5 3200)13 gave extremely high productivity values of
133 700 kg(PE) mol21 h21 bar21 for 15 (Mw 5 343 000,
PDI 5 3.8) and 152 500 kg(PE) mol21 h21 bar21 for 16
(Mw 5 404 500, PDI 5 2.4). None of the other compounds
showed any significant activity, including the tert-butyl imide, 3.
Note, for comparison, the productivities of the TACN analogues
ð2Þ
t
of 15 and 16 (I, R 5 2-C6H4CF3 or 2-C6H4 Bu) under identical
conditions were three orders of magnitude lower (ca. 500 kg(PE)
mol21 h21 bar21). The dramatic inversion of the imido
substituent–polymerisation activity relationship for the
HC(Me2pz)3-supported catalysts, compared to the (apparently)
closely related TACN series I, is very striking. These results show
conclusively that simply choosing a ‘‘bulky’’ imido ligand
substituent is insufficient for optimal catalyst design.
t
Fig. 2 Molecular structure of [Ti(N-2-C6H4 Bu){HC(Me2pz)3}Cl2] (16):
Displacement ellipsoid plot (left) and space filling representation (right).
The ‘‘valued added’’ benefit of the ortho-substituted aryl imido
substituents in the catalyst family [Ti(NR){HC(Me2pz)3}Cl2] can
be traced to fundamental differences in fac-N3 co-ligand topology.
Thus while HC(Me2pz)3 ligands have clefts between the Me2pz
rings which are able to accommodate the aryl substituent of an
imido ligand, TACN has no such feature. This ‘‘self-constrained’’
feature of the HC(Me2pz)3-supported catalysts can readily be
appreciated by reference to the solid state structure of [Ti(N-2-
The family of compounds [Ti(NR){HC(Me2pz)3}Cl2] also form
the basis for the preparation of borate-free, zwitterionic tris(pyr-
azolyl)methide-supported catalysts. Electron deficient alkyl cations
are the accepted active species in Ziegler type olefin polymerisa-
tion.1,15 It is also well known that the use of very weakly
coordinating anions is essential for high polymerisation activity,
and considerable efforts have been made in developing such
species.15–17 Another approach has been to incorporate the
counteranion into the ligand framework in order to prevent close
association of the anion with the alkyl cation’s vacant site(s).
Borate-based systems are the most studied in this regard,18–22
t
C6H4 Bu){HC(Me2pz)3}Cl2] (16) (Fig. 2).{ The steric protection of
the imido ligand N lone pairs and/or potentially reactive TiLN
multiple bond14 by the favourably-oriented ortho-tert-butyl sub-
stituent is self-evident. This orientation of the tert-butyl group will
This journal is ß The Royal Society of Chemistry 2006
Chem. Commun., 2006, 436–438 | 437

View Article Online
though others have also been reported to a much lesser extent.23,24
Catalytically-competent, carbanion-based zwitterions have not
previously been disclosed.
family (analogues of the ubiquitous tris(pyrazolyl)borates) may
find applications in a number of other catalytic systems, especially
perhaps where the potential lability of the B–N bonds of
pyrazolylborates may hinder the development of robust systems.
We thank the EPSRC, DSM Elastomers, DSM Research and
Millennium Pharmaceuticals Ltd. for support, Albermarle for
samples of MAO, Mr. Bing Wang for the polymerisation
experiments with [Ti(NtBu){C(Me2pz)3}Cl(THF)] and Mr. Dan
Cowell for assistance with the library synthesis.
It has recently been shown that reaction of the free HC(Me2pz)3
ligand with MeLi in THF gives the zwitterionic lithium salt
…
[C(Me2pz)3Li(THF)], containing no direct C Li bond because of
the steric protection afforded by the pyrazolyl ring 5-substituents.25
In an analogous way, 3 was apically deprotonated in THF to give
structurally characterized [Ti(NtBu){C(Me2pz)3}Cl(THF)] (3-
zwitt).10 No catalytic or other reaction chemistry of this compound
has been described and no other tris(pyrazolyl)methide chemistry
has been established.
Notes and references
{ Crystal Data for 16?2(CH2Cl2): C28H39Cl6N7Ti, Mw 5 734.28, mono-
clinic, P21/n, a 5 10.2075(2), b 5 24.0696(3), c 5 14.0799(2) s,
Polymerisation of ethylene with 3-zwitt in the presence of MAO
at 100 uC proceeded with relatively poor activity (510 kg(PE)
mol21 h21 bar21) to afford PE with a broad and featureless
molecular weight distribution. Guided by the results in Fig 1, we
therefore turned to a zwitterionic derivative of the successful ortho-
tert-butyl phenyl imide, 16-zwitt. The THF-stabilised compound
b 5 90.9514(4)u, U 5 3458.82(9) s3, Z 5 4, T 5 150 K, m 5 0.741 cm21
,
5159 reflections I . 3s(I), Rint 5 0.084, R 5 0.0502, Rw 5 0.0592. CCDC
286735. For crystallographic data in CIF or other electronic format see
DOI: 10.1039/b514467a
1 V. C. Gibson and S. K. Spitzmesser, Chem. Rev., 2003, 103, 283.
2 P. D. Bolton and P. Mountford, Adv. Synth. Catal., 2005, 347, 355.
3 N. Adams, H. J. Arts, P. D. Bolton, D. Cowell, S. R. Dubberley,
N. Friederichs, C. Grant, M. Kranenbrug, A. J. Sealey, B. Wang,
P. J. Wilson, A. R. Cowley, P. Mountford and M. Schro¨der, Chem.
Commun., 2004, 434.
t
[Ti(N-2-C6H4 Bu){C(Me2pz)3}Cl(THF)] (16-zwitt) was synthesised
according to eqn. 3 in 64% yield.
4 P. D. Bolton, E. Clot, A. R. Cowley and P. Mountford, Chem.
Commun., 2005, 3313.
(3)
5 K. Michiue and R. F. Jordan, Organometallics, 2004, 23, 460.
6 S. Trofimenko, Scorpionates. The Coordination Chemistry of
Polypyrazolylborate Ligands, Imperial College Press, London, 1999.
7 H. Lee and R. F. Jordan, J. Am. Chem. Soc., 2005, 127, 9384.
8 H. R. Bigmore, S. C. Lawrence, P. Mountford and C. S. Tredget,
Dalton Trans., 2005, 635.
9 S. C. Lawrence, B. D. Ward, S. R. Dubberley, C. M. Kozak and
P. Mountford, Chem. Commun., 2003, 2880.
10 S. C. Lawrence, M. E. G. Skinner, J. C. Green and P. Mountford,
Chem. Commun., 2001, 705.
11 A. Sa´nchez-Me´ndez, G. F. Silvestri, E. de Jesu´s, F. J. de la Mata,
J. C. Flores, R. Go´mez and P. Go´mez-Sal, Eur. J. Inorg. Chem., 2004,
3287.
12 N. Adams, H. R. Bigmore, T. L. Blundell, C. L. Boyd, S. R. Dubberley,
A. J. Sealey, A. R. Cowley, M. E. G. Skinner and P. Mountford, Inorg.
Chem., 2005, 44, 2882.
13 10 minute polymerisation run times were used in order that the intrinsic
catalyst productivities would not be masked by mass transport
limitation effects.
14 N. Hazari and P. Mountford, Acc. Chem. Res., 2005, 38, 839.
15 M. Bochmann, J. Organomet. Chem., 2004, 689, 3982.
16 E. Y.-X. Chen and T. J. Marks, Chem. Rev., 2000, 100, 1391.
17 J. Zhou, S. J. Lancaster, D. A. Walker, S. Beck, M. Thornton-Pett and
M. Bochmann, J. Am. Chem. Soc., 2001, 123, 223.
18 M. Bochmann, S. J. Lancaster and O. B. Robinson, J. Chem. Soc.,
Chem. Commun., 1995, 2081.
19 Y. Chen, G. Wu and G. C. Bazan, Angew. Chem., Int. Ed., 2005, 44,
1108.
20 G. Erker, Chem. Commun., 2003, 1469.
21 C. C. Lu and J. C. Peters, J. Am. Chem. Soc., 2002, 124, 5272.
22 W. E. Piers, Chem.–Eur. J., 1998, 4, 13.
23 H. Y. Kim, H. T. Kim, N. Y. Kim, E. S. Cho, B. Y. Lee, D. M. Shin
and Y. K. Chung, Organometallics, 2003, 22, 1503.
24 P. J. Bailey, D. Loron˜o-Gonza´lez and S. Parsons, Chem. Commun.,
2003, 1426.
Activation of 16-zwitt with MAO at 100 uC (Al/Ti 5 3200) gave
a very highly active catalyst with a productivity of 131 000 kg(PE)
mol21 h21 bar21. In contrast to the PE made with 16 under
identical conditions, that produced with 16-zwitt was bimodal with
a low (78 wt%, Mw 5 62 570, PDI y 2.3) and high (22 wt%,
Mw 5 1 380 000, PDI y 1.6) molecular weight fraction, showing
that the catalytic species derived from 16-zwitt is different to that
derived from 16. However, the activation of 16-zwitt with AliBu3
(Al/Ti 5 3200, all other parameters remaining unchanged) gave an
excellent productivity of 146 700 kg(PE) mol21 h21 bar21 and PE
with a unimodal molecular weight distribution (Mw 5 409 000,
PDI 5 2.5). Assuming that the Lewis acidic AliBu3 acts both as an
alkylating and THF abstracting agent, the catalytically-active
species in this system is likely be a charge-neutral, zwitterionic alkyl
t
cation of the type [Ti(N-2-C6H4 Bu){C(Me2pz)3}R] (R 5 growing
polymeryl chain).
In conclusion, we have demonstrated that varying the co-ligand
topology can dramatically invert structure–activity relationships in
imido-supported polymerisation catalysts. These catalysts, some of
which are highly productive under industrially relevant conditions,
are the first in Group 4 to contain the so far under-exploited
HC(Me2pz)3 ligand. Although a neutral donor (benefiting from no
formal ionic contribution to the metal–ligand bonding), its binding
to the metal is clearly sufficient to withstand harsh catalytic
reaction conditions. The high activity of the robust zwitterionic
precatalyst 16-zwitt suggests that the tris(pyrazolyl)methide ligand
25 F. Breher, J. Grunenberg, S. C. Lawrence, P. Mountford and
H. Ru¨egger, Angew. Chem., Int. Ed., 2004, 43, 2521.
438 | Chem. Commun., 2006, 436–438
This journal is ß The Royal Society of Chemistry 2006