The nature of the active species in bis(imino)pyridyl cobalt ethylene polymerisation catalysts

Article abstract of DOI:10.1039/b107490c

Studies on cobalt ethylene polymerisation catalysts bearing bis(imino)pyridine ligands strongly indicate that the activated species is not the anticipated cobalt(II) alkyl cation.

Full text of DOI:10.1039/b107490c

View Article Online / Journal Homepage / Table of Contents for this issue
The nature of the active species in bis(imino)pyridyl cobalt ethylene
polymerisation catalysts
Vernon C. Gibson,* Martin J. Humphries, Kilian P. Tellmann, Duncan F. Wass†, Andrew J. P.
White and David J. Williams
Department of Chemistry, Imperial College of Science, Technology and Medicine, Exhibition Road,
South Kensington, London SW7 2AY, UK
Received (in Cambridge, UK) 20th August 2001, Accepted 14th September 2001
First published as an Advance Article on the web 10th October 2001
Studies on cobalt ethylene polymerisation catalysts bearing
bis(imino)pyridine ligands strongly indicate that the acti-
vated species is not the anticipated cobalt(II) alkyl cation.
(mol. wt, PDI) when activated with MAO; the same active
species thus appears to be present in each case. These
observations lead us to conclude that the species produced upon
activation of these procatalysts with MAO is a cobalt(
I
) species
In recent years there has been much interest in late transition
metal olefin polymerisation catalysts. Recently, Brookhart and
Bennett, and ourselves have independently reported highly
active polymerisation1 and oligomerisation2 catalysts based on
iron and cobalt bearing bis(imino)pyridyl ligands. The activity
of the cobalt polymerisation catalyst family, whilst approx-
imately an order of magnitude less than that for their iron
relatives, is nevertheless highly respectable for late transition
metal olefin polymerisation catalysts. However, the natures of
the active sites in both the iron and cobalt systems remain little
understood. Here, we report our preliminary findings on the
cobalt system.
rather than cobalt(II).
The addition of ethylene to a sample of the cobalt(
I
) methyl
species 2 in toluene does not lead to the formation of polymer,
nor does its NMR spectrum change; 2 is, therefore, not a
polymerisation-active species.‡ The addition of Al2Me6 does
not activate 2 towards ethylene polymerisation, nor does it
cause any colour change, thus a stronger Lewis acid (e.g. MAO)
appears to be required for activation of the catalyst. As a
potential model for the activation process, 2 was treated with the
Lewis acid B(C6F5)3. This afforded a royal blue solution of
[LCo(N2)][MeB(C6F5)3] 4, thus providing a clue as to the
nature of the second colour change in the reaction of 1 with
MAO.
The activation of the pre-catalyst [LCoCl2] 1 {L
=
2,6-[CMeNN(2,6-C6H3Pri2)]2C5H3N} with methylaluminoxane
(MAO) produces an initial colour change from yellow–brown to
claret, with a subsequent colour change to blue. Addition of
ethylene to this solution causes a colour change to aquamarine,
concomitant with the onset of polymerisation.1 An interesting
observation is that if the activation with MAO is performed
under an argon atmosphere as opposed to dinitrogen, the
resulting solution is purple. If this solution is degassed and
dinitrogen is then admitted the colour of the solution changes to
blue. Degassing returns the original purple colour. Thus the
activated species appears to be capable of reversibly binding
dinitrogen.
Molecular structures of the cobalt( ) species 3§ and 4§ have
I
been determined. They are closely related, having approximate
C2v symmetry about an axis passing through the pyridyl
nitrogen and the cobalt centre (complex 4 is depicted in Fig. 1).
In both structures the metal coordination plane extends to
include the ligand backbone [from C(11) through C(4) to
C(23)], the maximum deviation from planarity being 0.025 Å in
3 and 0.080 Å in 4. The 2,6-diisopropylphenyl ring systems are
in both complexes oriented approximately orthogonally to the
coordination plane (by between 82 and 88°). There are no
significant differences between the equivalent Co–N bonds to
the tridentate ligand in the two structures, or in the CNN double
bond lengths. To our knowledge there are no structurally
characterised square planar cobalt complexes bearing three
nitrogen donors with either a chloride or a dinitrogen ligand in
the fourth coordination site reported in the literature. The Co–
The well-established mode of activation of both early and late
transition metal systems [e.g. group 4 metallocenes and Ni(a-
diimine) catalysts] involves, via the action of MAO, the
transformation of a metal dihalide fragment LnMCl2 into an
alkyl cation [LnM–R]+. If this activation pathway is followed by
the cobalt catalyst family, a 15-electron cobalt alkyl cation of
the type [LCoMe]+ would be anticipated. Indeed, calculations
have been reported based on cationic alkyl species of this type
being the active centres.3 We attempted to synthesise the
dimethyl complex [LCoMe2] in order to access a well defined
analogue of this hypothesised cobalt(II) alkyl cation. However,
the reaction between 1 and methyl Grignard reagents (Scheme
1) affords the reduced species [LCoMe] 2.
This led us to investigate the reduction of 1 by other means.
Clean reduction to the cobalt( ) chloride [LCoCl] 3 is achieved
I
upon reaction of 1 with a suspension of zinc powder in toluene
at 50 °C. Both 2 and 3 are claret in colour, and, as with all other
cobalt( ) species reported in this study, are highly air- and
I
moisture-sensitive. They are readily oxidised by most chlorin-
ated solvents and other oxidising agents.
The activation of 2 or 3 with MAO affords a blue solution,
which also becomes aquamarine on the addition of ethylene to
the flask, colour changes that are similar to those observed upon
treatment of 1 with MAO. 1–3 all yield identical polyethylene
† Present address: BP Research Centre, Sunbury on Thames, Middlesex,
UK TW16 7LN.
Scheme 1 Reagents and conditions: i, 2.5 MeMgBr, Et2O, 278 °C, 18 h; ii,
B(C6F5)3, toluene; iii, Zn, toluene, 50 °C, 15 h; iv, C2H4, toluene.
2252
Chem. Commun., 2001, 2252–2253
This journal is © The Royal Society of Chemistry 2001
DOI: 10.1039/b107490c

View Article Online
cobalt(II) pre-catalyst to cobalt(
I
), (ii) to form the cobalt( )
I
cation by chloride or methide group abstraction (MAO), and
(iii) to enhance the stability of the catalyst, and/or the rate of
polymerisation. The manner in which polymerisation is initi-
ated from the cobalt( ) cation is currently unclear. One
I
possibility is the reaction of a bound ethylene with a Lewis acid
4
{cf. the activation of [Cp2Zr(butadiene)] by B(C6F5)3 } to give
a zwitterionic cobalt(III) complex. Another possibility is
oxidative coupling of two molecules of ethylene to afford a
cobaltacyclopentane species, analogous to the commonly
postulated active site of the Phillips chromium catalyst system.5
A third possibility is C–H s-bond activation of the bound
ethylene to afford a cobalt(III) vinyl hydride species, a reaction
which has precedence in the chemistry of the group 9 metals.6
For the first two possibilities it is less clear how 5 would be
regenerated during the polymerisation. A common outcome of
all of these processes, however, is the oxidation of the cobalt( )
I
Fig. 1 The molecular structure of 4. Selected bond lengths (Å) and angles (°)
for 4; Co–N(1) 1.812(3), Co–N(7) 1.915(3), Co–N(9) 1.908(3), Co–N(35)
1.841(3), C(7)–N(7) 1.303(5), C(9)–N(9) 1.303(5), N(35)–N(36) 1.112(6);
N(1)–Co–N(7) 81.67(12), N(1)–Co–N(9) 81.68(12), N(1)–Co–N(35)
178.3(2), N(7)–Co–N(9) 163.31(12), N(7)–Co–N(35) 98.23(14), N(9)–Co–
N(35) 98.44(14), Co–N(35)–N(36) 175.4(5). For 3; Co–Cl 2.1807(10), Co–
N(1) 1.797(3), Co–N(7) 1.916(3), Co–N(9) 1.912(3), C(7)–N(7) 1.317(5),
C(9)–N(9) 1.322(5); Cl–Co–N(1) 179.15(10), Cl–Co–N(7) 98.89(9), Cl–
Co–N(9) 98.17(9), N(1)–Co–N(7) 81.35(14), N(1)–Co–N(9) 81.58(14),
N(7)–Co–N(9) 162.92(12).
precursor to cobalt(III). Thus, in accord with findings on other
cobalt polymerisation systems,7 and the apparent inactivity of
LCo( )R species, we err in favour of the active species being
I
cobalt(III). Investigations into these possibilities, and other
aspects of the chemistry of the cobalt system, will be reported in
due course.
BP Chemicals Ltd is thanked for financial support. Dr J.
Boyle and Dr G. Audley are thanked for NMR and GPC
measurements, respectively.
N2 bond length in 4 is comparable to those observed in
tetrahedral cobalt species. In 4 the shortest anion…cation
contact is an approximately orthogonal approach of 2.93 Å of
one of the pentafluorophenyl fluorine atoms to the pyridyl
nitrogen atom. The closest approach of a fluorine atom to the
cobalt centre is 3.5 Å.
The addition of ethylene to a toluene solution of 4 causes a
colour change to aquamarine, and polymer is formed, although
only with low activity (11 g mmol21bar21h21). The addition of
10 equiv. of Al2Me6 as co-catalyst increases the activity by an
order of magnitude (96 g mmol21 bar21 h21); the resultant
polyethylene is identical in molecular weight and molecular
weight distribution to that produced by the MAO activated
system.
Notes and references
‡ The longer chain Co(
)–alkyl homologues of 2, LCoR (where R = Et, Prn
I
or Bun) have been synthesised and they too are inactive towards ethylene
polymerisation.
§ Crystal data: for 3: C33H43N3ClCo, M = 576.1, monoclinic, space group
P21/n (no. 14), a = 8.819(1), b = 23.121(2), c = 15.744(2) Å, b =
101.08(1)°, V = 3150.7(7) Å3, Z = 4, Dc = 1.214 g cm23, m(Mo-Ka) =
6.54 cm21, T
refinement, R1
reflections [¡Fo¡
=
293 K; 4635 independent measured reflections, F2
=
0.046, wR2
=
0.100, 3253 independent observed
47°], 343 parameters. For 4:
>
4s(¡Fo¡), 2q @
¯
[C33H43N5Co][C19H3F15B], M = 1095.7, triclinic, space group P1 (no. 2),
a = 10.764(1), b = 15.797(1), c = 15.881(1) Å, a = 100.55(1), b =
102.86(1), g = 98.80(1)°, V = 2534.7(3) Å3, Z = 2, Dc = 1.436 g cm23
,
m(Cu-Ka)
= = 183 K; 8102 independent measured
35.1 cm21, T
reflections, F2 refinement, R1 = 0.055, wR2 = 0.128, 6057 independent
observed absorption corrected reflections [¡Fo¡ > 4s(¡Fo¡), 2q @ 126°], 668
parameters. CCDC reference numbers 169303 and 169304. See http:/
/www.rsc.org/suppdata/cc/b1/b107490c/ for crystallographic data in CIF or
other electronic format.
An important question then concerns how the propagating
species is formed from 4. The diamagnetic nature of the
cobalt( ) bis(imino)pyridyl system makes it amenable to study
I
by NMR spectroscopy. One useful feature of the 1H NMR
spectra is that the chemical shifts of the ligand resonances vary
considerably depending on the nature of the substitution at the
cobalt centre. The ketimine methyl resonance is particularly
sensitive in this respect, with the singlet being observed at d
21.14 in 2, d 0.05 in 3 and d 1.11 in 4, a trend which appears
to correlate with the electrophilicity of the cobalt centre.
A study of the pre-catalyst 4 by 1H NMR spectroscopy
showed that addition of 1 equiv. of ethylene afforded a species
1 (a) G. J. P. Britovsek, V. C. Gibson, B. S. Kimberley, G. A. Solan, A. J.
P. White and D. J. Williams, Chem. Commun., 1998, 849; (b) B. L. Small,
M. Brookhart and A. M. A Bennett, J. Am. Chem. Soc., 1998, 120, 4049;
(c) G. J. P. Britovsek, M. Bruce, V. C. Gibson, B. S. Kimberley, P. J.
Maddox, S. Mastroianni, S. J. McTavish, C. Redshaw, G. A. Solan, S.
Strömberg, A. J. P. White and D. J. Williams, J. Am. Chem. Soc., 1999,
121, 8728.
2 (a) B. L. Small and M. Brookhart, J. Am. Chem. Soc., 1998, 120, 7143;
(b) G. J. P. Britovsek, S. Mastroianni, G. A. Solan, S. P. D. Baugh, C.
Redshaw, V. C. Gibson, A. J. P. White, D. J. Williams and M. R. J.
Elsegood, Chem. Eur. J., 2000, 6, 2221.
1
identifiable by its H NMR spectrum as the ethylene adduct
[LCo(h-C2H4)][MeB(C6F5)3] 5. The ketimine signal now
occurs at d 0.73, and a singlet resonance assignable to bound
ethylene is present at d 4.65 (cf. d 5.32 for free ethylene). The
ethylene is only weakly bound; if more ethylene is added to the
NMR tube then rapid exchange between free and bound
ethylene occurs, resulting in an averaged chemical shift for the
ethylene hydrogens. The ethylene can be seen to be consumed,
with polymer forming on the walls of the NMR tube. At the end
of the polymerisation, 5 remains.
3 P. Margl, L. Q. Deng and T. Ziegler, Organometallics, 1999, 18, 5701.
4 (a) B. Temme, G. Erker, J. Karl, H. Luftmann, R. Fröhlich and S. Kotila,
Angew. Chem., Int. Ed. Engl., 1995, 34, 1755; (b) for a review, see: ; G.
Erker, Acc. Chem. Res., 2001, 34, 309.
5 M. P. McDaniel, Adv. Catal., 1985, 33, 47.
6 See, for example: (a) P. O. Stoutland and R. G. Bergman, J. Am. Chem.
Soc., 1985, 107, 4581; (b) S. T. Bely, S. B. Duckett, D. M. Haddleton and
R. N. Perutz, Organometallics, 1989, 8, 748; (c) C. K. Ghosh, J. K.
Hoyano, R. Krentz and W. A. G. Graham, J. Am. Chem. Soc., 1989, 111,
5480; (d) R. S. Tanke and R. H. Crabtree, Inorg. Chem., 1989, 28, 3444;
(e) B. Papenfuhs, N. Mahr and H. Werner, Organometallics, 1993, 12,
4244; (f) E. Gutiérrez-Puebla, A. Monge, M. C. Nicasio, P. J. Pérez, M.
L. Poveda, L. Rey, C. Ruíz and E. Carmona, Inorg. Chem., 1998, 37,
4538.
There are several conclusions that can be drawn from these
observations. Firstly, the species that is initially formed upon
activation, and thus the species from which the initation of
polymerisation must occur, is a cobalt( ) cation, a species that
I
contains no cobalt–C(alkyl) bond. Secondly, the aluminium co-
catalyst appears to perform three functions: (i) to reduce the
7 G. F. Schmidt and M. Brookhart, J. Am. Chem. Soc., 1985, 107, 1443.
Chem. Commun., 2001, 2252–2253
2253