Bis(imino)pyridyl iron and cobalt complexes: The effect of nitrogen substituents on ethylene oligomerisation and polymerisation

Article abstract of DOI:10.1039/b101173l

The synthesis and characterisation of 2,6-bis(imino)pyridyl iron and cobalt complexes [(2,6-(RN=CMe)₂C₅H₃N)-MCl₂] containing nitrogen substituents of the type R = NPh₂, NPhMe, NMe₂ or 2,5-dimethylpyrrolyl are described. These complexes, in combination with the co-catalyst MAO, give active catalysts for the oligomerisation or polymerisation of ethylene. The catalytic activity is strongly affected by the substituents R and the polymerisation conditions used. The polymer properties are also a function of the R substituents. With R = NPhMe or NMe₂, toluene soluble α-olefins are obtained, whereas the bulkier substituents (R = NPh₂ or 2,5-dimethylpyrrolyl) give low molecular weight solid polyethylene.

Full text of DOI:10.1039/b101173l

View Article Online / Journal Homepage / Table of Contents for this issue
Bis(imino)pyridyl iron and cobalt complexes: the effect of nitrogen
substituents on ethylene oligomerisation and polymerisation
George J. P. Britovsek,a Vernon C. Gibson,*a Brian S. Kimberley,b Sergio Mastroianni,a
Carl Redshaw,a Gregory A. Solan,a Andrew J. P. Whitea and David J. Williamsa
a Department of Chemistry, Imperial College, London, UK SW7 2AY.
E-mail: v.gibson@ic.ac.uk
b BP Chemicals Ltd., Sunbury Research Centre, Sunbury on Thames, Middlesex, UK TW16
7LN
Received 6th February 2001, Accepted 15th March 2001
First published as an Advance Article on the web 25th April 2001
The synthesis and characterisation of 2,6-bis(imino)pyridyl iron and cobalt complexes [(2,6-(RN᎐CMe) C H N)-
᎐
2
5
3
MCl2] containing nitrogen substituents of the type R = NPh2, NPhMe, NMe2 or 2,5-dimethylpyrrolyl are described.
These complexes, in combination with the co-catalyst MAO, give active catalysts for the oligomerisation or
polymerisation of ethylene. The catalytic activity is strongly affected by the substituents R and the polymerisation
conditions used. The polymer properties are also a function of the R substituents. With R = NPhMe or NMe2,
toluene soluble α-olefins are obtained, whereas the bulkier substituents (R = NPh2 or 2,5-dimethylpyrrolyl) give low
molecular weight solid polyethylene.
Introduction
The development of homogeneous organo-transition metal
polymerisation catalysts has enabled the synthesis of poly-
olefinic materials with unprecedented levels of control over
macromolecular structure. Although metallocene technology
has been in the vanguard of these developments,1–8 much recent
work has focused on the search for non-metallocene catalysts,
and indeed the armoury of available olefin polymerisation
catalysts now traverses the entire transition metal series.9
One of the latest additions to the collection of highly active
Fig. 1 Bis(imino)pyridyl (A) and bis(hydrazone)pyridyl complexes
(B).
ethylene polymerisation catalysts is a family of iron and cobalt
bis(imino)pyridyl catalysts of type A (Fig. 1), developed
independently by Brookhart, Bennett and ourselves.10,11 The
versatility of these catalysts, in particular the ease of catalyst
tuning by ligand variation, has already been amply demon-
strated. For example, bulky aryl substituents provide catalysts
for the polymerisation of ethylene that give high molecular
weight, strictly linear polyethylene (HDPE)12 whereas smaller
substituents lead to the oligomerisation of ethylene to
α-olefins.13,14 As part of our ongoing investigations regarding
the structure–activity relationships within this family of
catalysts,12,14,15 we report here new ethylene oligomerisation/
polymerisation catalysts based on iron() and cobalt() com-
plexes containing amine substituted bis(imino)pyridyl ligands
of type B as shown in Fig. 1. Some of the results described here
have been disclosed previously by us16 and Eastman Chemical
Company.17
The complexes are all highly coloured and paramagnetic with
magnetic moments in the range of 5.1–5.3 µB (Evans balance)
for the iron complexes and 4.6–4.8 µB for the cobalt complexes,
consistent with four and three unpaired electrons respectively.
1H NMR spectra show broad resonances in the range Ϫ60 to 80
ppm consistent with paramagnetic contact shifted resonances.
The complexes have been further characterised by MS, micro-
analysis and X-ray analysis (see Experimental section).
Crystals of 2a, 2c and 3a were grown by diffusion of pentane
into the respective saturated dichloromethane solution at room
temperature whilst crystals of 2b were obtained from a satur-
ated acetonitrile solution. The molecular structures of 2a–c and
3a have been determined. All four possess very similar gross
overall structures, with those of the iron and cobalt complexes
2a and 3a being essentially isostructural (but not isomorphous).
They all exhibit non-crystallographic Cs symmetry about a
plane containing the metal atom, the pyridyl nitrogen and the
two chloride ions (Figs. 2 and 3).
Results and discussion
Synthesis of ligands and complexes
The geometries at the metal centres can be considered either
as distorted square pyramidal with one of the chloride ions
[Cl(1)] occupying the apical position, or as equally distorted
trigonal bipyramidal with the pyridyl nitrogen atom, the metal
centre and the two chlorides forming the equatorial plane. The
ligand geometries (excluding the terminal amino substituents)
are also, within statistical significance, the same. In all instances
Bis(hydrazone)pyridine ligands 1a–c were prepared from
2,6-diacetylpyridine and the appropriate hydrazine derivative,
using a standard condensation protocol.18 The synthesis of the
pyrrolyl substituted ligand 1d from 1-amino-2,5-dimethyl-
pyrrole19 required considerably longer reaction times to obtain
satisfactory yields. Complexes 2a–d and 3a–c were readily pre-
pared in good yields by reaction of FeCl2 or CoCl2 with the
appropriate ligand according to Scheme 1.
the imine C᎐N double bond character has been retained [C᎐N
᎐
᎐
in the range 1.281(7)–1.293(6) Å] there being no evidence for
DOI: 10.1039/b101173l
J. Chem. Soc., Dalton Trans., 2001, 1639–1644
This journal is © The Royal Society of Chemistry 2001
1639

View Article Online
Table 1 Selected bond lengths (Å) and angles (Њ) for 2a, 2b, 2c and 3a
2a [M = Fe]
2b [M = Fe]
2c [M = Fe]
3a [M = Co]
M–Cl(1)
M–Cl(2)
M–N(1)
M–N(8)
M–N(11)
2.2872(14)
2.2565(13)
2.085(4)
2.318(4)
2.283(4)
2.311(2)
2.275(2)
2.069(4)
2.279(4)
2.256(4)
2.299(2)
2.294(2)
2.079(4)
2.271(4)
2.285(4)
2.260(2)
2.2480(14)
2.025(4)
2.250(4)
2.219(4)
Cl(1)–M–Cl(2)
Cl(1)–M–N(1)
Cl(1)–M–N(8)
Cl(1)–M–N(11)
Cl(2)–M–N(1)
Cl(2)–M–N(8)
Cl(2)–M–N(11)
N(1)–M–N(8)
N(1)–M–N(11)
N(8)–M–N(11)
136.45(6)
95.68(11)
99.48(10)
95.06(10)
127.86(11)
94.58(10)
96.00(10)
73.65(14)
74.05(14)
145.63(14)
129.74(7)
108.18(11)
96.05(11)
95.70(11)
122.07(12)
96.80(11)
98.35(12)
73.72(14)
74.28(14)
147.93(14)
129.95(6)
111.37(12)
96.75(11)
97.52(12)
118.69(11)
96.93(12)
95.80(11)
73.8(2)
127.90(6)
105.89(11)
99.50(11)
99.86(11)
126.21(11)
93.94(11)
94.37(11)
74.49(14)
75.33(14)
147.58(14)
74.1(2)
147.7(2)
Fig. 2 The molecular structure of 2a (that of 3a is essentially
isostructural).
Scheme 1
delocalisation into the adjacent hydrazido moieties which have
distinctive single bond character [N–N 1.392(6)–1.423(6) Å].
The main differences between the four complexes are in the
geometries and patterns of bonding around the metal centres
and in the degree of pyramidalisation at the terminal hydrazido
nitrogen centres (the latter being most pronounced in 2c).
Specifically, in the three iron complexes, the degree of pyramid-
alisation at the terminal nitrogen centres (defined by the dis-
placement of the nitrogen out of the N–C–C plane) ranges
from 0.19 Å in 2a through 0.32 Å in 2b to 0.48 Å in 2c. In each
case the nitrogen lone pair is directed away from the iron centre,
thereby discounting any possible metal ؒ ؒ ؒ lone pair inter-
action. Furthermore, there are in all three iron structures sub-
stantial torsional twists about the N–N bonds (ranging from
42Њ in 2b to 57Њ in 2a) thus precluding any significant N(pπ)–
N(pπ) orbital overlap. As expected, the bonds to cobalt in 3a
are all shorter than their counterparts in the iron analogue 2a
(Table 1). In the three iron complexes the Fe–N(pyridyl) dis-
tances do not differ significantly. There are, however, some
Fig. 3 The molecular structure of 2c.
noticeable asymmetries in the Fe–N(imino) distances which are
most pronounced in 2a where the two distances are 2.283(4)
and 2.318(4) Å, and least pronounced in the dimethylamino
complex 2c [2.271(4) and 2.285(4) Å]. These differences are
probably related to the better σ-donating ability of dimethyl-
amino versus diphenylamino substituents, resulting in increased
electron density on the metal bound imine nitrogens. There are
similar asymmetries in the Fe–Cl bond lengths which are most
pronounced in 2b and least in 2c. These differences cannot be
readily explained in terms of simple trans effects, due to the
substantial departures from an idealised square pyramidal
geometry.
The most dramatic differences are in the Cl(1)–M–Cl(2) and
1640
J. Chem. Soc., Dalton Trans., 2001, 1639–1644

View Article Online
Table 2 Results of ethylene polymerisation runs using complexes 2a–d, 3a–c and 4
Pre-catalyst/
mmol
Co-catalyst
(eq.)
Activity/g mmolϪ1
h
Run a
Yield/g
Ϫ1 barϪ1
αg
Mw
Mn
Mw/Mn
h
h
1
2b
3
2a (0.02)
2a (0.02)
2b (0.02)
2c (0.02)
2d (0.01)
2d (0.0005)
4 (0.0006)
3a (0.08)
3b (0.08)
3c (0.08)
MAO (200)
MAO (100)
MAO (200)
MAO (200)
MAO (160)
MAO (100)
MAO (1000)
MAO (200)
MAO (200)
MAO (200)
14.3d
0.5/1.0e
0.9g
71.5
2700
900
750
700
3.6
1.3
473
4.4
2.3
635
3140
9340
<1
0.84
0.51
0.83
0.83
4
0.5g
5b
6c
7i
8
0.4/0.8e
15.7 f
56.5
900
900
242000
500
500
9600
1.9
1.9
25.3
Trace
9
10
1.9g
2.4
1
0.84
0.75
0.8g
a Tests performed in 1 L autoclave, 10 bar ethylene pressure, 35 ЊC, isobutane solvent, pre-catalyst dissolved in toluene, MAO scavenger, 60 min.
b Tests performed in a Schlenk flask, 0.75 bar ethylene pressure, 25 ЊC, toluene solvent (40 ml), 15 min. c Conditions as under a, except 50 ЊC and
triisobutylaluminium as scavenger. d Solid polymer; saturated chain ends: 18.5 (/1000 C), vinyl chain ends: 17.2 (/1000 C), determined by 13C NMR.
e Solid polymer/oligomers in g. f Solid polymer; saturated chain ends: 22.1 (/1000 C), vinyl chain ends: 21.0 (/1000 C), determined by 13C NMR.
g Determined by GC. h Determined by GPC. i Results for pre-catalyst 4 taken from ref. 12.
113.5Њ cf. values that range between 104.5 and 105.8Њ in the
other complexes, this inclination being towards the pyridyl ring
of the tridentate ligand (Fig. 4). There are no short inter-
molecular approaches to Cl(1) in any of the four structures.
Oligomerisation and polymerisation results
Bis(imino)pyridyl iron (2a–d) and cobalt (3a–c) complexes, in
combination with the co-catalyst MAO, have been evaluated as
catalysts for the polymerisation of ethylene. General reaction
conditions are described in the Experimental section and the
polymerisation results are summarised in Table 2.
Under Schlenk-line conditions (25 ЊC and 0.75 bar) and
relatively short run times (15 minutes) complexes 2a and 2d,
containing the diphenylamine and the 2,5-dimethylpyrrolyl
substituents respectively, give a mixture of solid polyethylene
and oligomers (ratio 1 : 2) with comparable activites (see
Table 2, runs 2 and 5). Under high-pressure conditions (10 bar,
35 ЊC), the overall activity during one hour is considerably less
for complexes 2a–c than for the pyrrolyl substituted catalyst 2d.
The activity profiles20 indicate that all catalysts 2a–d give high
initial activities but that the catalyst lifetime of 2d is substan-
tially longer than for 2a–c, but not as long as for the previously
reported catalyst 4, containing 2,6-dimethylphenyl substituents
(see Fig. 1, A, where Ar = 2,6-Me2C6H3). For comparison, the
polymerisation results of 4, which were obtained under similar
conditions, are also included in Table 2 (run 7).12 Within the
catalyst series 2a–c, one possibility that could affect the catalyst
stability and activity is the availability of the nitrogen lone pair
of the NR2 substituents for donation, either to the imine nitro-
gen or to the metal centre. In either case, this would reduce the
electrophilicity of the metal centre, which would be expected to
lower the activity. However, in the solid state structures of 2a–c,
the pyramidalisation of the amino nitrogen and the orientation
of the NR2 groups indicate that lone pair donation to the imine
N {N(pπ)–N(pπ) overlap} is not strong (vide supra). The possi-
bility of the amino group binding to the metal centre can be
ruled out due to the constraints imposed by the ligand back-
bone. Another possibility is that, in the presence of MAO, it is
very likely that AlMe3 (or other aluminium centres in MAO)
will bind to the nitrogen lone pair, in a similar manner to that
Fig. 4 Views down the N(8) ؒ ؒ ؒ N(11) vector showing (a) the enlarge-
observed for hydrazone derivatives of “constrained geometry”
ment of the Cl(1)–Fe–Cl(2) angle and the tilting of the Fe–Cl(1) bond
catalysts, which also gave much lower activities.21,22 In the case
towards the pyridyl ring in 2a, and (b) the contraction of the Cl(1)–M–
of the 2,5-dimethylpyrrolyl catalyst derived from 2d, the pyr-
rolyl nitrogen lone pair is not available for donation to the imine
nitrogen due to the involvement in bonding within the pseudo-
aromatic pyrrole ring and, for the same reason, it is also not
available to donate to an external Lewis acid. Notwithstanding
these arguments, it is also possible that the comparable activi-
ties of 2d and 4 are mainly due to the steric resemblance of the
Cl(2) angle and the tilting of the M–Cl(1) bond away from the pyridyl
ring as seen in 2b, 2c and 3a (that of 2b is illustrated).
Cl(1)–M–N(1) angles in 2a, which are substantially larger (by
ca. 7Њ) and smaller (by ca. 13Њ) respectively, than in the others.
This difference can be visualised in terms of the inclination of
the “apical” M–Cl(1) vector to the basal plane, which in 2a is
J. Chem. Soc., Dalton Trans., 2001, 1639–1644
1641

View Article Online
2,5-dimethylpyrrole and the 2,6-dimethylphenyl substituents.
The cobalt precursors 3a–c show generally lower activities com-
pared to the iron analogues 2a–c, in line with the differences
in catalytic activity observed for iron versus cobalt bis(imino)-
pyridine systems.12
solvent. Mass spectra were obtained using either Fast Atom
Bombardment (FAB), Electron Impact (EI) or Chemical
Ionization (CI). IR spectra were recorded on a Perkin-Elmer
Spectrum GX1 System. Samples were prepared by evaporating
CH2Cl2 solutions of the ligands or complexes on NaCl plates.
Magnetic susceptibility studies were performed using an Evans’
balance.
All of the bis(imino)pyridyl iron and cobalt pre-catalysts
tested in this study give relatively low molecular weight prod-
ucts. Within the iron series 2a–c, at high pressure (10 bar), the
diphenyl derivative 2a gives solid polymer with a relatively low
molecular weight (Mw = 2700), whereas the Me/Ph and dime-
thyl derivatives 2b and 2c give toluene soluble oligomers. At
lower temperature and pressure, 2a gives a mixture of oligomers
and polymer (run 2). The oligomeric products have been shown
by GC/MS to consist of linear α-olefins following a typical
Schulz–Flory distribution. The distribution factors α are listed
in Table 2. The reason for the lower α-value of 2b compared to
2a is unclear at present, but these values are subject to a
relatively large error due to the very low activity of these
catalysts. In the case of bis(imino)pyridyl iron and cobalt
catalysts, the molecular weight of the polymer product is
strongly affected by the size of the aryl substituents. With this
in mind, it is rather surprising to see that the 2,5-dimethyl-
pyrrolyl derivative 2d produces low molecular weight poly-
ethylene (Mw = 900), whereas the 2,6-dimethylphenyl analogue
4, which is believed to possess comparable steric requirements,
gives high molecular weight polymer (Mw = 242000). This result
indicates that the size and shape of the substituent on the imino
nitrogen, allowing for the differences between six- and five-
membered rings, are not the only factors responsible for
the formation of high molecular weight polymers and that
electronic effects may be more important than realised to date.
End-group analysis of the polymers produced by 2a and 2d
Materials
1-Amino-2,5-dimethylpyrrole was prepared according to a pro-
cedure by Zimmermann,19 while 2,6-diacetylpyridine, MAO
(10% solution in toluene) and all hydrazines were purchased
from Aldrich Chemical Co. All other chemicals were obtained
commercially and used as received unless stated otherwise.
Preparations
2,6-Diacetylpyridinebis(N,N-diphenylhydrazone) (1a). 1,1-
Diphenylhydrazine hydrochloride (6 g, 0.027 mol) in dichloro-
methane (50 ml) was treated with gaseous ammonia for 30 min
at room temperature. After filtration, the solvent was removed
under vacuum to afford N,N-diphenylhydrazine as a viscous
oil (yield 4.7 g, 95%). To a solution of 2,6-diacetylpyridine
(1.67 g, 0.01 mol) in absolute ethanol (30 ml) was added
N,N-diphenylhydrazine (4.7 g, 0.025 mol). After addition of
a few drops of glacial acetic acid the solution was refluxed for
4 hours. Volatiles were removed under vacuum and the solid
was washed with cold ethanol and dried in a vacuum oven
1
(50 ЊC) overnight to give 5.9 g (92%) of 1a. H NMR data
(250 MHz, CDCl3, 298 K): δ 8.32 (d, 2H, 3J(HH) 7.9, Py-Hm),
7.75 (t, 1H, Py-Hp), 7.35–7.06 (m, 20H, Ar-H), 2.21 (s, 6H,
N᎐CCH ). MS (EI, m/z) 495 [M]ϩ. 13C NMR (100 MHz,
᎐
3
1
by H and 13C NMR has shown an equal amount of saturated
CDCl , 298 K): δ 161.9 (N᎐C), 155.0 (Ar-C), 148.2 (Py-C ),
᎐
3
o
and vinyl terminated ends (see Table 2). This analysis shows
that these low molecular weight polymers are α-olefins, gener-
ated by a termination pathway via β-H transfer, similar to
the oligomerisation catalysts 2b and 2c. No chain transfer to
aluminium is observed under these conditions, a transfer pro-
cess that only seems to become competitive with β-H transfer
when higher molecular weight polyethylene is produced, i.e.
when the rate of β-H transfer is sufficiently low.
136.3 (Py-Cp), 129.1 (Ar-C), 123.5 (Ar-C), 121.8 (Py-Cm),
121.0 (Ar-C), 17.8 (N᎐CCH ). IR (CH Cl film, NaCl, cmϪ1
)
᎐
3
2
2
3059 (br), 1588 (s), 1561 (m), 1488 (vs), 1451 (s), 1364 (w),
1270 (m), 1201 (s), 1179 (w), 1086 (w), 1049 (w), 843 (w), 819
(w), 750 (s), 629 (s). Elemental analysis for C33H29N5 (495.63)
found (required): %C = 79.5 (80.0), %H = 5.8 (5.9), %N = 13.9
(14.1).
In conclusion, we have shown that bis(imino)pyridyl iron and
cobalt dichloride complexes containing nitrogen substituents
can be readily synthesised and provide active catalysts for the
polymerisation of ethylene after activation with MAO. As seen
before, iron derived catalysts are, in general, much more active
than their cobalt analogues. Within the iron series tested here,
the catalytic activity obtained with pre-catalyst 2d containing
2,5-dimethylpyrrolyl substituents, is, under certain conditions
significantly higher than for 2a–c. The activity profile indicates
that the lifetime of the catalyst derived from 2d is longer than
for 2a–c but not as long as for the aryl substituted derivatives
such as 4. All iron and cobalt catalysts tested here give relatively
low molecular weight polyethylene or oligomers, with vinyl
end-groups in all cases.
2,6-Diacetylpyridinebis(N-methyl-N-phenylhydrazone) (1b).
Procedure as described for 1a, using 2,6-diacetylpyridine (1.67
g, 0.01 mol) and N-methyl-N-phenylhydrazine (3.05 g, 0.025
1
mmol). 1b was obtained in 83% yield (3.08 g). H NMR (250
3
MHz, CDCl3, 298 K): δ 8.27 (d, 2H, J(HH) 7.9 Hz, Py-Hm),
7.75 (t, 1H, Py-Hp), 7.35–7.05 (m, 10 H, Ar-H), 3.28 (s, 6H,
N-CH ), 2.21 (s, 6H, N᎐CCH ). 13C NMR (100 MHz, CDCl3,
᎐
3
3
298 K): δ 163.8 (N᎐C), 154.9 (Ar-C), 151.1 (Py-C ), 136.3 (Py-
᎐
o
Cp), 128.9 (Ar-C), 121.1 (Ar-C), 120.4 (Py-Cm), 115.7 (Ar-C),
42.9 (N–CH ) 15.8 (N᎐CCH ). MS (EI, m/z) 371 [M]ϩ. IR:
᎐
3
3
(CH2Cl2 film, NaCl, cmϪ1) 3067 (m), 2875 (br), 2798 (w), 1599
(vs), 1567 (vs), 1494 (vs), 1453 (vs), 1362 (s), 1286 (m), 1250 (m),
1185 (m), 1091 (vs), 996 (m), 879 (w), 837 (m), 819 (s), 788 (m),
753 (vs), 692 (vs), 647 (w), 520 (m). Elemental analysis for
C23H25N5 (371.48) found (required): %C = 74.9 (74.4), %H = 6.4
(6.8), %N = 18.0 (18.8).
Experimental
General
2,6-Diacetylpyridinebis(N,N-dimethylhydrazone) (1c). Pro-
cedure as described for 1a, using 2,6-diacetylpyridine (1.63 g,
0.01 mol) and N,N-dimethylhydrazine (1.6 ml, 0.021 mmol), 1c
All manipulations were carried out under an atmosphere of
nitrogen using standard Schlenk and cannula techniques or in a
conventional nitrogen-filled glove-box. Solvents were refluxed
over an appropriate drying agent, and distilled and degassed
prior to use. Elemental analyses were performed by the micro-
analytical services of the Department of Chemistry at Imperial
College, Medac Ltd. or SACS at the University of North Lon-
don. NMR spectra were recorded on a Bruker spectrometer at
250 MHz (1H), and 62.9 MHz (13C) at 293 K; chemical shifts
are referenced to the residual protio impurity of the deuterated
1
was obtained as a viscous oil in 78% yield (1.93 g). H NMR
(250 MHz, C6D6, 298 K): δ 8.19 (d, 2H, 3J(HH) 7.9 Hz, Py-Hm),
7.15 (t, 1H, Py-Hp), 2.53 (s, 12H, N(CH3)2), 2.50 (s, 6H,
N᎐CCH ). 13C NMR (100 MHz, C D , 298 K): δ 161.3 (N᎐C),
᎐
᎐
3
6
6
155.5 (Py-Co), 135.8 (Py-Cp), 120.5 (Py-Cm), 47.2 [N(CH3)2],
14.3 (N᎐CCH ). IR (CH Cl film, NaCl, cmϪ1): 2991 (s), 2956
᎐
3
2
2
(vs), 2858 (vs), 2820 (vs), 2776 (s), 1608 (w), 1569 (s), 1454 (vs),
1361 (s), 1256 (m), 1208 (w), 1148 (m), 1120 (s), 1100 (m), 1079
(m), 1014 (vs), 991 (m), 960 (m), 876 (m), 814 (vs), 742 (m), 639
1642
J. Chem. Soc., Dalton Trans., 2001, 1639–1644

View Article Online
(m). Elemental analysis for C13H21N5 (247.34) found (required):
1468 (m) 1450 (s), 1401 (w), 1361 (s), 1269 (m), 1191 (m), 1030
%C = 62.9 (63.1), %H = 8.4 (8.6), %N = 27.7 (28.3).
(vs), 1010 (m), 957 (m), 813 (vs), 745 (m), 557 (w). Elemental
analysis for C13H21N5FeCl2 (374.20) found (required):
%C = 41.4 (41.7), %H = 5.1 (5.6), %N = 18.3 (18.7). µeff (Evans
Balance) 5.27 µB.
2,6-Diacetylpyridinebis(2,5-dimethylpyrrolyl) (1d). A solution
of 1-amino-2,5-dimethylpyrrole (0.66 g, 6 mmol) and 2,6-
diacetylpyridine (0.49 g, 3 mmol) in 20 ml toluene was refluxed
for 5 days in the presence of molecular sieves (4 Å) and a drop
of glacial acetic acid. After filtration the solvent was removed
and the oily residue recrystallised from ethanol, yielding 1d as a
yellow solid (0.85 g, 82%). 1H NMR (250 MHz, CDCl3, 298 K):
δ 8.47 (d, 2H, 3J(HH) 7.9 Hz, Py-Hm), 7.94 (t, 1H, Py-Hp), 5.91
2,6-Diacetylpyridinebis(2,5-dimethylpyrrolyl)iron dichloride
(2d). A solution of 1d (190 mg; 0.55 mmol) in 10 ml THF was
added to a suspension of FeCl2 in THF. The mixture was stirred
for 3 hours at room temperature. The volume of the dark solu-
tion was reduced to 5 ml and the product precipitated with
diethyl ether. The product 2d was washed with diethyl ether and
dried in vacuo, yielding a dark green powder in 85% yield (222
(s, 4H, pyrrole CH), 2.37 (s, 6H, N᎐CCH ), 2.09 (s, 12H, pyrrole
᎐
3
CH3). 13C NMR (100 MHz, CDCl , 298 K): δ 176.1 (N᎐C),
᎐
3
1
153.5 (Py-Co), 137.2 (Py-Cp), 123.7 (Py-Cm), 123.5 (pyrrole
mg). H NMR (250 MHz, CD2Cl2, 298 K, broad singlets are
CCH ), 103.8 (pyrrole CH), 15.51 (N᎐CCH ), 11.82 (pyrrole
observed in each case): δ 66.9 (2H, Py-Hm), 48.3 (1H, Py-Hp),
᎐
3
3
CH3). MS (EI, m/z): 347 [M]ϩ, 253 [M Ϫ C6H8N]ϩ. IR (neat,
NaCl, cmϪ1): 3100 (w), 2976 (m), 2921 (s), 2891 (m), 2857
(w), 2235 (w), 1702 (w), 1613 (s), 1570 (s), 1514 (w), 1440 (s),
1391 (vs), 1370 (vs), 1290 (s), 1128 (m), 1020 (m), 911 (s), 822
(s), 737 (vs). Elemental analysis for C21H25N5 (347.47) found
(required): %C = 72.6 (72.6), %H = 7.4 (7.3), %N = 20.1
(20.2).
3.6–1.8 (16H, pyrrole CH and CH), Ϫ40.1 (6H, N᎐CCH ). MS
᎐
3
3
(FAB, m/z): 474 [M]ϩ, 438 [M Ϫ Cl]ϩ. IR (neat, cmϪ1): 3369 (br,
m), 2919 (m), 1580 (m), 1440 (w), 1387 (vs), 1267 (m), 1206 (m),
1019 (m), 811 (w), 737 (vs). Elemental analysis for C21H25N5-
FeCl2 (474.22) found (required): %C = 53.0 (53.2), %H = 5.4
(5.3), %N = 14.7 (14.8).
2,6-Diacetylpyridinebis(N,N-diphenylhydrazone)cobalt
2,6-Diacetylpyridinebis(N,N-diphenylhydrazone)iron dichlor-
ide (2a). A suspension of FeCl2 (0.184 g, 1.41 mmol) and 1a
(0.70 g, 1.41 mmol) was prepared in n-butanol (20 ml) which
was then heated at 90 ЊC for 30 min. The resultant dark product,
which precipitated from solution, was filtered off and washed
several times with diethyl ether (3 × 20 ml) and dried to afford
0.79 g (89%) of 2a. The complex could additionally be recrystal-
lised by slow diffusion of pentane into a saturated solution of
dichloride (3a). The procedure as for complex 2a, using 1a
(0.619 g, 1.25 mmol) and CoCl2 (0.162 g, 1.25 mmol) gave 3a as
a brown powder in 91% yield (0.71 g). The complex could
additionally be recrystallised by slow diffusion of pentane into
a saturated solution of 3a in dichloromethane to give brown
needles. 1H NMR (250 MHz, CD2Cl2, 298 K, broad singlets are
observed in each case): δ 75.5 (2H, Py-Hm), 0.1 (4H, Ar-Hp),
Ϫ0.7 (8H, Ar-H), Ϫ1.1 (1H, Py-H ) Ϫ10.4 (6H, N᎐CCH ),
᎐
p
3
1
2a in dichloromethane to give brown needles. H NMR data
(250 MHz, CD2Cl2, 298 K, broad singlets are observed in each
Ϫ24.5 (8H, Ar-H). IR (neat, cmϪ1): 3060 (w), 2963 (w), 1578
(s), 1521 (w), 1491 (vs), 1372 (w), 1269 (m), 1213 (m), 1064 (m),
1027 (w), 804 (w), 752 (m), 693 (m). MS (FAB, m/z) 589
[M Ϫ Cl]ϩ. Elemental analysis for C33H29N5CoCl2ؒCH2Cl2ؒH2O
(728.42) found (required): %C = 56.1 (56.1), %H = 4.1 (4.6),
%N = 9.5 (9.6). µeff (Evans Balance): 4.65 µB.
case): δ 72.1 (2H, Py-H ), 20.0 (6H, N᎐CCH ), 2.1 (8H, Ar-H),
᎐
m
3
1.6 (8H, Ar-H), 1.3 (4H, Ar-Hp), Ϫ8.7 (1H, Py-Hp). MS (FAB,
m/z) 621 [M]ϩ, 586 [M Ϫ Cl]ϩ. IR: (CH2Cl2 film, NaCl, cmϪ1
)
3059 (w), 1588 (vs), 1521 (w), 1491 (w), 1450 (w), 1369 (w), 1291
(m), 1269 (w), 1213 (m), 1083 (w), 1061 (w), 1023 (w), 752 (vs),
693 (vs). Elemental analysis for C33H29N5FeCl2ؒ0.75CH2Cl2
(686.08) found (required): %C = 59.1 (59.5), %H = 4.4 (4.5),
%N = 10.1 (10.2). µeff (Evans Balance): 5.15 µB.
2,6-Diacetylpyridinebis(N-methyl-N-phenylhydrazone)cobalt
dichloride (3b). The procedure as for complex 2a, using 1b
(0.464 g, 1.25 mmol) and CoCl2 (162 mg, 1.25 mmol) gave 3b as
a brown powder in 92% yield (0.58 g). The complex could be
recrystallised by slow diffusion of pentane into a saturated
2,6-Diacetylpyridinebis(N-methyl-N-phenylhydrazone)iron
dichloride (2b). The procedure as above for complex 2a, using 1b
(0.464 g, 1.25 mmol) and FeCl2 (152 mg, 1.2 mmol) gave 2b as a
brown powder in 88% yield (0.54 g). The complex could addi-
tionally be recrystallised from boiling acetonitrile to give brown
needles. 1H NMR (250 MHz, CD2Cl2, 298 K, broad singlets are
observed in each case): δ 69.5 (2H, Py-Hm), 28.6 (6H,
1
solution of 3b in dichloromethane to give brown needles. H
NMR (250 MHz, CD2Cl2, 298 K, broad singlets are observed
in each case): δ 79.5 (2H, Py-Hm), 2.9 (4H, Ar-H), 1.7 (2H,
Ar-Hp), Ϫ12.5 (4H, Ar-H), Ϫ14.1 (1H, Py-Hp), Ϫ19.1 (6H,
N᎐CCH ), Ϫ51.8 (6H, NCH ). MS (FAB, m/z): 500 [M]ϩ, 465
᎐
3
3
[M Ϫ Cl]ϩ. IR (neat, cmϪ1): 3044 (w), 2970 (w), 2895 (w), 1591
(s), 1528 (m), 1491 (vs), 1450 (m), 1368 (w), 1264 (m), 1097 (m),
1027 (m), 960 (w), 808 (w), 760 (m), 737 (m), 697 (m). Elemental
analysis for C23H25N5CoCl2ؒ1/3CH2Cl2 (529.65) found
(required): %C = 53.0 (52.9), %H = 5.0 (4.9), %N = 13.1 (13.2).
µeff (Evans Balance): 4.72 µB.
N᎐CCH ), 16.1 (6H, N(CH ) ), 13.4 (4H, Ar-H), 10.1 (Ar-H),
᎐
3
3 2
8.5 (Ar-Hp), Ϫ30.0 (1H, Py-Hp). MS (FAB, m/z): 497 [M]ϩ, 462
[M Ϫ Cl]ϩ. IR: (CH2Cl2 film, NaCl, cmϪ1): 3047 (m), 2974 (m),
2893 (w), 2808 (w), 1594 (vs), 1538 (s ), 1455 (s), 1371 (m), 1263
(vs), 1197 (s), 1100 (s), 1028 (s), 964 (m), 940 (w), 837 (w), 809
(m), 764 (s), 737 (vs), 698 (vs), 677 (w), 564 (w), 523 (m). Elem-
ental analysis for C23H25N5FeCl2 (498.24) found (required):
%C = 55.2 (55.5), %H = 5.3 (5.1), %N = 13.9 (14.1). µeff (Evans
Balance): 5.28 µB.
2,6-Diacetylpyridinebis(N,N-dimethylhydrazone)cobalt
dichloride (3c). The procedure as for complex 2a, using 1c
(0.370 g, 1.5 mmol) and CoCl2 (0.193 g, 1.5 mmol) gave 3c as a
brown powder in 62% yield (0.35 g). The complex could
additionally be recrystallised by slow diffusion of pentane into
a saturated solution of 3c in dichloromethane to give brown
needles. 1H NMR (250 MHz, CD2Cl2, 298 K, broad singlets are
observed in each case): δ 77.4 (2H, Py-Hm), Ϫ12.5 (1H, Py-Hp),
2,6-Diacetylpyridinebis(N,N-dimethylhydrazone)iron dichlor-
ide (2c). The procedure as for complex 2a, using 1c (0.370 g,
1.5 mmol) and FeCl2 (0.190 g, 1.5 mmol) gave 2c as a brown
powder in 88% yield (0.49 g). The complex could be recrystal-
lised by slow diffusion of pentane into a saturated solution of
Ϫ18.0 (6H, N᎐CCH ), Ϫ36.3 [12H, N(CH ) ]. MS (FAB, m/z):
᎐
3
3 2
1
2c in dichloromethane to give brown needles. H NMR (250
373 [M]ϩ, 338 [M Ϫ Cl]ϩ. IR (neat, cmϪ1): 3052 (w), 2925 (w),
2851 (w0, 1599 (w), 1543 (w), 1447 (br), 1361 (w), 1264 (vs),
1186 (w), 1027 (w), 808 (w), 741 (vs), 704 (m). Elemental anal-
ysis for C13H21N5CoCl2ؒ1/3H2O (383.18) found (required):
%C = 40.8 (40.8), %H = 5.3 (5.7), %N = 17.9 (18.3 ). µeff (Evans
Balance): 4.69 µB.
MHz, CD2Cl2, 298 K, broad singlets are observed in each case):
δ 69.4 (2H, Py-H ), 25.3 (6H, N᎐CCH ), 20.0 (12H, N(CH ) ),
᎐
m
3
3 2
Ϫ24.5 (1H, Py-Hp). MS (FAB, m/z): 373 [M]ϩ, 338 [M Ϫ Cl]ϩ.
IR (CH2Cl2 film, NaCl, cmϪ1): 3064 (w), 3000 (w), 2957 (m),
2879 (s), 28230 (w), 2794 (m), 1599 (m), 1571 (m), 1546 (m),
J. Chem. Soc., Dalton Trans., 2001, 1639–1644
1643

View Article Online
Crystallography
of aqueous hydrogen chloride. The solid PE was recovered by
filtration, washed with methanol (50 mL) and dried (vacuum
oven at 50 ЊC).
2a: C33H29N5Cl2Feؒ0.5CH2Cl2, M = 664.8, orthorhombic, Pbca
(no. 61), a = 15.877(2), b = 17.546(3), c = 22.725(4) Å, V =
6331(2) Å3, Z = 8, Dc = 1.395 g cmϪ3, µ(Cu-Kα) = 64.0 cmϪ1
,
T = 293 K, dark red platy needles; 4609 independent measured
reflections, F2 refinement, R1 = 0.050, wR2 = 0.113, 3062
independent observed absorption corrected reflections
[|Fo| > 4σ(|Fo|), 2θ ≤ 120Њ], 352 parameters.
Acknowledgements
BP Chemicals Ltd. are thanked for financial support.
Drs J. Boyle, C. Baker and G. Audley are thanked for NMR
and GPC measurements and Mr J. Barton for GC analyses.
Dr P. Maddox is thanked for helpful discussions.
2b: C23H25N5Cl2FeؒMeCN, M = 539.3, monoclinic, P21/c
(no. 14), a = 13.729(1), b = 13.277(1), c = 15.059(2) Å, β =
101.81(1)Њ, V = 2686.7(5) Å3, Z = 4, Dc = 1.333
g ,
cmϪ3
µ(Cu-Kα) = 65.2 cmϪ1, T = 293 K, deep red platy needles; 3690
independent measured reflections, F2 refinement, R1 = 0.053,
wR2 = 0.128, 2692 independent observed absorption corrected
reflections [|Fo| > 4σ(|Fo|), 2θ ≤ 120Њ], 284 parameters.
References
1 H. H. Brintzinger, D. Fischer, R. Mülhaupt, B. Rieger and R. M.
Waymouth, Angew. Chem., Int. Ed. Engl., 1995, 34, 1143.
2 M. Bochmann, J. Chem. Soc., Dalton Trans., 1996, 255.
3 W. Kaminsky, J. Chem. Soc., Dalton Trans., 1998, 1413.
4 P. C. Möhring and N. J. Coville, J. Organomet. Chem., 1994, 479, 1.
5 E. Albizzati and M. Galimberti, Catal. Today, 1998, 41, 159.
6 A. M. Thayer, Chem. Eng. News, 1995, Sept 11, 15.
7 K. B. Sinclair and R. B. Wilson, Chem. Ind., 1994, 857.
8 A. A. Montagna, R. M. Burkhart and A. H. Dekmezian,
CHEMTECH, 1997, December, 26.
9 G. J. P. Britovsek, V. C. Gibson, D. F. Wass, Angew. Chem., Int. Ed.,
1999, 38, 428.
10 B. L. Small, M. Brookhart and A. M. A. Bennett, J. Am. Chem.
Soc., 1998, 120, 4049.
11 G. J. P. Britovsek, V. C. Gibson, B. S. Kimberley, P. J. Maddox,
S. J. McTavish, G. A. Solan, A. J. P. White and D. J. Williams, Chem.
Commun., 1998, 849.
12 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.
13 B. L. Small and M. Brookhart, J. Am. Chem. Soc., 1998, 120, 7143.
14 G. J. P. Britovsek, S. Mastroianni, G. A. Solan, S. P. D. Baugh,
C. Redshaw, V. C. Gibson, A. J. P. White, D. Williams and M. R. J.
Elsegood, Chem. Eur. J., 2000, 6, 2221.
2c: C13H21N5Cl2Fe, M = 374.1, monoclinic, P21/n (no. 14),
a = 8.909(1), b = 17.133(1), c = 11.054(1) Å, β = 94.07(1)Њ,
V = 1683.0(2) Å3, Z = 4, Dc = 1.476 g cmϪ3, µ(Cu-Kα) = 101.1
cmϪ1, T = 293 K, dark brown prisms; 2491 independent meas-
ured reflections, F2 refinement, R1 = 0.051, wR2 = 0.118, 1883
independent observed absorption corrected reflections [|Fo| >
4σ(|Fo|), 2θ ≤ 120Њ], 191 parameters.
3a: C33H29N5Cl2Coؒ1.5CH2Cl2, M = 752.8, tetragonal, I41/a
(no. 88), a = 20.698(2), c = 33.803(3) Å, V = 14482(2) Å3,
Z = 16, Dc = 1.381 g cmϪ3, µ(Mo-Kα) = 8.74 cmϪ1, T = 293 K,
deep red needles; 6333 independent measured reflections,
F2 refinement, R1 = 0.058, wR2 = 0.118, 3823 independent
observed reflections [|Fo| > 4σ(|Fo|), 2θ ≤ 50Њ], 390 parameters.
CCDC reference numbers 158046–158049.
See http://www.rsc.org/suppdata/dt/b1/b101173l/ for crystal-
lographic data in CIF or other electronic format.
General polymerisation procedures
(a) High-pressure tests. A 1 litre stainless steel reactor was
baked out under a nitrogen flow for at least 1 hour at >85 ЊC
and subsequently cooled to the temperature of polymerization.
Isobutane (0.5 L) and trialkylaluminium (triisobutyl-
aluminium) were introduced into the reactor and stirred at reac-
tion temperature for at least 1 hour. Ethylene was introduced by
over-pressure and the difference between the total pressure and
the initial pressure (isobutane and nitrogen: ca. 10 bar) is the
pressure quoted in Table 2. The catalyst solution in toluene
was then injected under nitrogen. The reactor pressure was
maintained constant throughout the polymerisation run by
computer controlled addition of ethylene. The polymerisation
time was 60 min. Runs were terminated by venting off volatiles
and the reactor contents were isolated, washed with aqueous
HCl, methanol and dried in a vacuum oven at 50 ЊC.
15 G. J. P. Britovsek, V. C. Gibson, S. Mastroianni, D. C. H. Oakes,
C. Redshaw, G. A. Solan, A. J. P. White and D. Williams, Eur. J.
Inorg. Chem., 2001, 431.
16 G. J. P. Britovsek, B. A. Dorer, V. C. Gibson, B. S. Kimberley and
G. A. Solan, BP Chemicals Ltd., World Pat., WO 99/12981, 1999;
G. J. P. Britovsek, B. A. Dorer, V. C. Gibson, B. S. Kimberley and
G. A. Solan, Chem. Abstr., 1999, 130, 252793.
17 L. S. Moody, P. B. Mackenzie, C. M. Killian, G. G. Lavoie,
J. A. Ponasik Jr., A. G. M. Barrett, T. W. Smith and J. C. Pearson,
Eastman Chemical Company, World Pat., WO 00/50470, 2000;
L. S. Moody, P. B. Mackenzie, C. M. Killian, G. G. Lavoie,
J. A. Ponasik Jr., A. G. M. Barrett, T. W. Smith and J. C. Pearson,
Chem. Abstr., 2000, 133, 208316.
18 J. D. Curry, M. A. Robinson and D. H. Busch, Inorg. Chem., 1967, 6,
1570.
19 W. Flitsch, U. Krämer and H. Zimmermann, Chem. Ber., 1969, 102,
3268.
20 Catalyst activities, which are commonly expressed in g mmolϪ1 hϪ1
barϪ1 give an average value for the activity during one hour. It
should be noted however, that catalyst activities are rarely constant
over an extended period of time and, for a meaningful comparison
of catalyst activities, the kinetic profile (ethylene uptake over time)
needs to be taken into account.
(b) Schlenk-flask ethylene polymerisation tests. The pre-
catalyst was dissolved in toluene (40 ml) and MAO (10 wt% in
toluene) added to produce an orange solution. The Schlenk
tube was placed in a water bath, purged with ethylene and the
contents magnetically stirred and maintained under ethylene
(0.75 bar over-pressure) for the duration of the polymerisation
(15 min). The polymerisation was terminated by the addition
21 S. C. Yoon, B.-J. Bae, I.-H. Suh and J. T. Park, Organometallics,
1999, 18, 2049.
22 J. T. Park, S. C. Yoon, B.-J. Bae, W. S. Seo, I.-H. Suh, T. K. Han and
J. R. Park, Organometallics, 2000, 19, 1269.
1644
J. Chem. Soc., Dalton Trans., 2001, 1639–1644