Five-coordinate iron(II) complexes bearing tridentate nitrogen donor ligands as catalysts for atom transfer radical polymerisation

Article abstract of DOI:10.1016/j.poly.2004.09.001

Five-coordinate iron complexes containing tridentate nitrogen donor ligands have been evaluated for the atom transfer radical polymerisation of styrene. Catalyst activities decrease in the order of donor group: alkyl amine ≈ pyridine > alkyl imine ? aryl imine > aryl amine. The trend in polymerisation activity correlates with the redox potential of the corresponding complexes. Complexes derived from aryl-imine or -amine ligands afford catalytic chain transfer rather than ATRP catalysis. A series of five-coordinate complexes of iron(II) containing tridentate nitrogen donor ligands has been synthesised and evaluated for the atom transfer radical polymerisation of styrene. The molecular structures of two derivatives are described. Cyclic voltammetric studies show that the redox potential for the Fe(II)/Fe(III) couple strongly depends on the donor capacity of the complexing ligand, and styrene polymerisations reveal that ligands derived from alkyl amine or pyridine groups are the most active for ATRP, although slower than for their four-coordinate relatives due to unfavourable steric interactions. In general, catalyst activity decreases in the order of donor group: alkyl amine ≈ pyridine > alkyl imine ? aryl imine > aryl amine, with the aryl derivatives being almost completely inactive in ATRP. The trend in polymerisation activity correlates with the redox potential of the corresponding complexes. Complexes derived from aryl-imine or -amine ligands afford catalytic chain transfer rather than ATRP catalysis.

Full text of DOI:10.1016/j.poly.2004.09.001

Polyhedron 23 (2004) 2921–2928
www.elsevier.com/locate/poly
Five-coordinate iron(II) complexes bearing tridentate nitrogen donor
ligands as catalysts for atom transfer radical polymerisation
*
Rachel K. OꢀReilly, Vernon C. Gibson , Andrew J.P. White, David J. Williams
Department of Chemistry, Imperial College, Exhibition Road, South Kensington, London SW7 2AZ, UK
Received 27 May 2004; accepted 8 September 2004
Dedicated to Malcolm Green for his inspiring contributions to chemistry
Abstract
A series of five-coordinate complexes of iron(II) containing tridentate nitrogen donor ligands has been synthesised and evaluated
for the atom transfer radical polymerisation of styrene. The molecular structures of two derivatives are described. Cyclic voltam-
metric studies show that the redox potential for the Fe(II)/Fe(III) couple strongly depends on the donor capacity of the complexing
ligand, and styrene polymerisations reveal that ligands derived from alkyl amine or pyridine groups are the most active for ATRP,
although slower than for their four-coordinate relatives due to unfavourable steric interactions. In general, catalyst activity
decreases in the order of donor group: alkyl amine ꢀ pyridine > alkyl imine ꢁ aryl imine > aryl amine, with the aryl derivatives
being almost completely inactive in ATRP. The trend in polymerisation activity correlates with the redox potential of the corre-
sponding complexes. Complexes derived from aryl-imine or -amine ligands afford catalytic chain transfer rather than ATRP
catalysis.
ꢀ 2004 Elsevier Ltd. All rights reserved.
Keywords: Iron; Radical; Styrene; Catalyst; Polymerisation
1. Introduction
Since the process of ATRP is dependent upon the
reversible binding of a halogen atom (Scheme 1), it is
important that the metal complex has a relatively open
coordination sphere, especially in its reduced form,
and that neither the reduced nor the oxidised metal spe-
cies are prone to decomposition. This may, for example,
rule out three-coordinate complexes for many metals
unless ligands are chosen that would stabilise the low
coordination number. However, such ligands tend to
be very bulky and not necessarily pre-disposed towards
allowing the metal to accept another ligand.
Recently, we described a new family of four coordi-
nate iron-based catalysts stabilised by a-diimine ligands
[13,14]. These four-coordinate iron(II) complexes
showed evidence for accessible and reversible redox cou-
ples to the corresponding iron(III) species depending on
the nature of the imine substituents. The high catalytic
Atom transfer radical polymerisation (ATRP) allows
the preparation of a wide range of well-defined poly-
meric materials with controlled molecular weights and
complex architectures [1]. Following key reports by
Matyjaszewski and Sawamoto in the mid-1990s, of
highly efficient ATRP catalysts based on copper(I) and
ruthenium(II) respectively, a number of different me-
tal–ligand combinations have been investigated, includ-
ing systems based on Rh, Pd, Ni and Fe [2–12]. There
remains, however, a need to understand the factors
important for efficient ATRP catalysis.
*
Corresponding author. Tel.: +20 7594 5830; fax: +20 7594 5810.
E-mail address: v.gibson@imperial.ac.uk (V.C. Gibson).
0277-5387/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.poly.2004.09.001

2922
R.K. OꢀReilly et al. / Polyhedron 23 (2004) 2921–2928
Scheme 1. Generalised mechanism for atom transfer radical polymerisation of styrene.
efficiency of these catalysts may be related to the fact
that the reduced Fe(II) species possess a relatively
uncluttered tetrahedral geometry and are, therefore,
able to accept an additional halogen atom without any
severe steric conflict arising between the co-ligands. In
the five-coordinate oxidised form, the flexibility in the
coordination geometry (via pseudo rotation) may assist
the complex in achieving a coordination environment
suited to the dissociation of a halogen atom (certain
six-coordinate environments, e.g. d3 and d6 octahedral
systems, have reduced kinetic lability). Thus, the 4–5
coordination manifold might be anticipated to possess
important attributes for the efficient functioning of
ATRP catalysts.
In order to further probe the effect of coordination
geometry on the efficiency of ATRP catalysis, we
decided to target a series of five-coordinate Fe(II) chlo-
ride complexes bearing tridentate nitrogen ligands and
evaluate their performance in ATRP catalysis. These
complexes possess an additional nitrogen donor relative
to the (a-diimine)FeCl2 catalysts reported recently, and
would therefore be expected to afford lower oxidation
potentials and render them more electronically suited
to ATRP catalysis. However, these catalysts necessarily
have to operate by interconverting between five- and six-
coordinate species which may lead to increased steric
conflict in the oxidised form of the catalyst.
2. Results and discussion
2.1. Synthesis and characterisation
The series of tridentate [NNN] ligands, I–VI, contain-
ing a central pyridine or amine unit and two peripheral
imine or amine donors with either N-alkyl or N-aryl
substituents was synthesised (Fig. 1). Ligand I (PMD-
ETA) is commercially available, while II, III, V and
VI were synthesised using literature procedures[16–21].
Ligand IV was synthesized by a modification of a liter-
ature preparation using Si(OEt)4 and recovered as a yel-
low solid in 62% yield [22]. Ligands I–VI were reacted
with iron(II) dichloride to afford the corresponding
monochelate complexes (1–6) in good yields.
Complexation of I to iron(II) dichloride was first re-
ported in 1966, by Ciampolini and Speroni [23]. At that
time, 1 was the first reported high spin five-coordinate
iron(II) complex. In 1999, Calderazzo and co-workers
[24] structurally characterized 1 and reported a distorted
square pyramidal structure with a chlorine atom occu-
pying the apical position. Ligand II has previously been
Earlier work by Go¨belt and Matyjaszewski examined
tridentate bis(imino)pyridine and bis(amino)pyridine
ligands bearing N-octyl substituents, in conjunction with
copper(I) bromide and iron(II) bromide [15]. They
found that the bis(imino)pyridine ligand formed an ac-
tive catalyst in both systems, whereas, the
bis(amino)pyridine ligand formed an active catalyst only
in conjunction with copper(I) bromide. The use of bro-
mide as the transferable group is recognised to hold
advantages in terms of C–halogen bond strengths,
though the thermodynamic advantage may be off-set
by the increased steric demands of bromine atoms coor-
dinated to the metal centre. In this study we focus on a
series of metal–chloride derivatives containing tridentate
ligands with a view to obtaining an improved under-
standing of the factors influencing efficient ATRP catal-
ysis in iron-based systems.
Fig. 1. Ligands I–VI.

R.K. OꢀReilly et al. / Polyhedron 23 (2004) 2921–2928
2923
complexed to mid and late transition metals such as
ruthenium and zinc by the groups of van Koten and
Reinhoudt [18,25]. It was synthesised by reaction of
2,6-bis(bromomethyl)pyridine with dimethylamine and
complexed to iron(II) to afford 2, as a yellow microcrys-
talline solid. Single crystals of 2 were grown by slowly
cooling a saturated solution of 2 in hot tetrahydrofuran.
X-ray single crystal analysis revealed the structure
(which has molecular CS symmetry about a plane con-
taining the iron, the two chlorides and the pyridine
nitrogen atom) to have a distorted square pyramidal
geometry around the iron centre (Fig. 2). The four basal
atoms {Cl(1),N(1),N(7),N(8)} are coplanar to within
n(II) dichloride in diethyl ether and was isolated as a
dark red, microcrystalline solid. This complex was first
reported in 1969 and, based on magnetic and spectro-
scopic techniques, a five-coordinate geometry around
the iron centre was proposed [30]. We confirmed this
assignment by determining the solid state X-ray struc-
ture of crystals of 3 obtained by slow diffusion of an
equal volume of heptane into a saturated solution of 3
in dichloromethane. This study revealed 3 to have a
structure (Fig. 3) closely related to that of its zinc dibro-
mide analogue [31]. The geometry at iron can be consid-
ered as intermediate between square pyramidal, with
Cl(2) occupying the apical position, and trigonal bipy-
ramidal with N(1) and N(10) defining the major axis,
though being somewhat closer to the latter description.
Considered as square pyramidal, the metal centre lies
˚
˚
0.06 A with the iron lying ca. 0.68 A out of this plane
in the direction of the apical atom Cl(2); the trans-basal
angles are 136.1(2) and 147.0(2)ꢁ, and the apical–basal
angles range between 106.2(2) and 111.4(2)ꢁ. The Fe–
Cl bond lengths are noticeably asymmetric, with that
to the apical chloride being significantly shorter
˚
˚
ca. 0.58 A [0.64 A] out of the basal plane which is only
˚
˚
coplanar to within 0.18 A [0.30 A]. Viewed as trigonal
bipyramidal, the axial atoms subtend an angle of
147.21(14)ꢁ [150.2(2)ꢁ] at the metal, and the angles with-
in the equatorial plane range between 108.30(11) and
141.54(11)ꢁ [111.8(2) and 133.9(2)ꢁ]; the iron lies ca.
˚
˚
[2.291(2) A] than to its basal counterpart [2.341(2) A];
the Fe–N distances are unexceptional [21,24,26]. The
two five-membered chelate rings both have envelope
geometries with in each case the amino nitrogen lying
˚
˚
0.06 A [0.04 A] out of the equatorial plane in the direc-
tion of N(10). (The values in square parentheses refer to
the zinc dibromide analogue [31].) The five-membered
N(1)/N(7) chelate ring is almost planar with the metal
˚
out of the plane of the other four atoms [by ca. 0.57 A
˚
for N(7) and ca. 0.68 A for N(8)] which are coplanar
˚
to within 0.02 A. The only intermolecular packing inter-
˚
˚
action of note is a p–p stacking between adjacent pyridyl
rings with centroidꢂ ꢂ ꢂcentroid and mean interplanar sep-
lying ca. 0.09 A [<0.01 A] out of the plane of the other
˚
four atoms which are coplanar to better than 0.01 A. By
contrast, the N(7)/N(10) five-membered chelate ring has
a twisted envelope conformation with N(10) displaced
˚
arations of 3.72 and 3.55 A, respectively. Subsequent to
this determination, Grubbs and co-workers [27] re-
ported the room temperature (298 K) structure of the
same polymorph of 2 in a private communication; tak-
ing into account the 115 K difference in temperature,
the unit cells are the same.
Ligand III has recently been complexed to late tran-
sition metals such as copper and palladium [28,29].
Complex 3 was synthesised by reaction of III with iro-
˚
˚
by ca. 0.72 A [0.76 A] out of the plane of the other
˚
˚
atoms, which are coplanar to within 0.08 A [0.12 A].
The pattern of bonding around the metal centre is gen-
erally the same between the FeCl2 species 3 and its pre-
viously reported ZnBr2 analogue, though it is noticeable
that, whereas in the ZnBr2 complex the M–N(pyridyl)
and M–N(amino) distances are the same [2.245(8) and
˚
Fig. 2. The molecular structure of 2. Selected bond lengths (A) and
˚
Fig. 3. The molecular structure of 3. Selected bond lengths (A) and
angles (ꢁ); Fe–Cl(1) 2.341(2), Fe–Cl(2) 2.291(2), Fe–N(1) 2.121(6), Fe–
N(7) 2.262(6), Fe–N(8) 2.248(6), C(7)–N(7) 1.468(9), C(8)–N(8)
1.474(10), N(1)–Fe–N(8) 74.3(2), N(1)–Fe–N(7) 74.4(2), N(8)–Fe–
N(7) 136.1(2), N(1)–Fe–Cl(2) 106.2(2), N(8)–Fe–Cl(2) 106.4(2), N(7)–
Fe–Cl(2) 111.4(2), N(1)–Fe–Cl(1) 147.0(2), N(8)–Fe–Cl(1) 95.7(2),
N(7)–Fe–Cl(1) 94.1(2), Cl(2)–Fe–Cl(1) 106.84(9).
angles (ꢁ); Fe–Cl(1) 2.3034(12), Fe–Cl(2) 2.3361(13), Fe–N(1) 2.185(4),
Fe–N(7) 2.129(3), Fe–N(10) 2.236(4), C(7)–N(7) 1.258(6), N(7)–Fe–
N(1) 74.04(14), N(7)–Fe–N(10) 75.98(14), N(1)–Fe–N(10) 147.21(14),
N(7)–Fe–Cl(1) 141.54(11), N(1)–Fe–Cl(1) 93.79(10), N(10)–Fe–Cl(1)
101.55(10), N(7)–Fe–Cl(2) 108.30(11), N(1)–Fe–Cl(2) 102.39(10),
N(10)–Fe–Cl(2) 99.34(11), Cl(1)–Fe–Cl(2) 109.94(5).

2924
R.K. OꢀReilly et al. / Polyhedron 23 (2004) 2921–2928
˚
2.236(7) A, respectively], in 3 the M–N(pyridyl) bond
˚
length is noticeably shorter [2.185(4) A] than that to
Table 1
Cyclic voltammetry results for 1–6
˚
the amino nitrogen [2.236(4) A]. The Fe–N bond lengths
Eox (mV)
E1/2 (mV)
DEp (mV)
differ noticeably from those in, for example, (2-((2,6-di-
isopropylphenyl)iminomethyl)-6-(1-(2,6-di-isopropylphe-
nylamino)ethyl)pyridine-N,N0,N00)dichloroiron(II) and
other related complexes with the bond to the pyridine
nitrogen being longer and that to the imino nitrogen
shorter [25]. The Fe–Cl bond lengths in 3 are noticeably
asymmetric, with that to the pseudo-apical chloride
1
2
3
4
5
6
ꢄ370
ꢄ240
ꢄ190
ꢄ150
180
ꢄ280
ꢄ150
ꢄ100
ꢄ70
330
170
170
180
170
310
320
260
410
˚
Cl(2) being significantly longer [2.3661(13) A] than to
Complexes 1–4 all have low oxidation potentials
˚
its pseudo-basal counterpart Cl(1) [2.3034(12) A]. The
(ꢄ370 to ꢄ150 mV) and redox potentials (ꢄ280 to
ꢄ70 mV) indicating that they are easy to oxidise, unlike
5 and 6 whose high oxidation potentials indicate the
opposite. Thus, it would be predicted that 1–4 should
be active ATRP catalysts since they are easily oxidised
and possess an accessible and reversible one-electron re-
dox couple. In fact, 1–4 have lower redox potentials
than the previously reported N-alkyl diimine complexes
(E1/2 ca.ꢄ100 mV) and thus would be predicted to be
faster ATRP catalysts [13]. However, 5 and 6, which
are supported by ligands possessing N-aryl groups
have irreversible (DEp > 300 mV) and less accessible
(E1/2 > 300 mV) redox couples and thus would be pre-
dicted to have low activity in ATRP.
only intermolecular packing feature of note is a p–p
stacking of the pyridine rings of centrosymmetrically re-
lated pairs of molecules; the centroidꢂ ꢂ ꢂcentroid and
˚
mean interplanar separations are 3.54 and 3.38 A,
respectively. An analogous contact is seen in the zinc
dibromide species with centroid–centroid and mean
˚
interplanar separations of 3.56 and 3.36 A, respectively.
Complex 4 was synthesized by a similar route to that
reported in the literature for the N-aryl derivative, 5 [32].
After washing with diethyl ether and extraction into
tetrahydrofuran, 4 was isolated as a dark grey micro-
crystalline solid. The N-aryl pyridine(bis)imino and
pyridine(bis)amino complexes, 5 and 6, have both been
reported recently by this group and were synthesized
as described in the literature reports [32,33]. Both of
these complexes have previously been structurally char-
acterized and found to possess a five-coordinate dis-
torted square pyramidal geometry.
Complexes 1–6 are all air-stable paramagnetic solids,
with magnetic moments (determined by the Evansꢀ
NMR method) in the range 4.9–5.3 lB [34–36]. The
magnetic moments are consistent with the presence of
four unpaired electrons and a quintet ground state
(S = 2) for the iron(II) centres [37].
2.3. Polymerisation results
Complexes 1–6 were screened for the ATRP of sty-
rene (200 equiv.) at 120 ꢁC using 1-phenylethyl chloride
(1-PECl) as an initiator and the results are presented in
Table 2. 1–4 polymerised styrene in a controlled manner,
with linear kinetics and with molecular weights increas-
ing with monomer conversion [40]. Analysis of the pol-
ymer produced using 1–4 revealed the presence of a
1
halogen end group by H NMR spectroscopy and also
by microanalysis [for 2, Mn = 5100% Cl, found (calc.):
0.64 (0.70)]. The more reducing the complex, the lower
the redox potential, and the faster the ATRP catalyst
cf. 1 versus 3 (Fig. 4).
However, these complexes are slower ATRP catalysts
than their a-diimine counterparts (e.g. for 1,2-bis(cyclo-
2.2. Cyclic voltammetry
Matyjaszewski and co-workers [38] have investigated
the structure–activity relationship for a series of triden-
tate nitrogen donor ligands in copper-based ATRP. The
redox properties of these complexes were measured
using CV techniques and their electrochemistry was cor-
related to their activity as ATRP catalysts for styrene
and MMA polymerisation [39]. By comparison with
Matyjaszewskiꢀs findings in copper-based ATRP, it
would be predicted that the series of complexes should
exhibit the following order of reducing power:
Table 2
Results of styrene polymerisations using 1–6
kobs (hꢄ1
)
% Conv. in 24 h
Mn,calc
Mn
Mw/Mn
a
1
2
3
4
5
6
0.04
0.03
0.02
0.02
65
52
39
37
72
45
13700
10900
8200
15100
11800
8500
1.62
1.53
1.59
1.56
2.07
1.92
1 < 2 < 3 < 4 ꢃ 5 < 6.
Experimentally
determined
7800
15100
9500
8700
2900
electrochemical parameters for 1–6 are collected in Ta-
ble 1 and confirm this trend. Table 1 also shows the oxi-
dation potentials, Eox, for this series of complexes. Since
this parameter can be correlated with the ease of oxida-
tion of the complex, Eox should provide an indication of
activities in ATRP catalysis.
2100
Conditions: 120 ꢁC, bulk, [1-PECl]0:[catalyst]0:[styrene]0 = 1:1:200.
Control polymerisation of styrene at 120 ꢁC with no catalyst > 90%
conversion in 24 h, Mn ca. 200000.
a
Mn,calc is calculated Mn based on degree of conversion.

R.K. OꢀReilly et al. / Polyhedron 23 (2004) 2921–2928
2925
hexylimino)ethane iron(II) dichloride, kobs = 0.25 hꢄ1
,
cf. kobs = 0.04 hꢄ1 for 1), despite their significantly lower
redox potentials (for 1,2-bis(cyclohexylimino)ethane iro-
n(II) dichloride, E1/2 = ꢄ120 mV cf. E1/2 = ꢄ280 mV for
1) [13]. This can be attributed to the less favourable ster-
ic environment around the metal in these five-coordinate
complexes as compared with the four-coordinate tetra-
hedral a-diimine complexes. Thus, the rate of Cl atom
transfer from RCl to 1, for example, is slowed due to
steric crowding, despite 1 having more favourably dis-
posed redox chemistry. This may also explain the loss
of control and broadening of the polydispersity of the
polymers produced using these five-coordinate systems
compared to their a-diimine relatives.
Fig. 4. Plot of ln([M]0/[M]) vs. time for complexes 1 (ꢅ) and 3 (ꢆ).
For 5 and 6, which both possess N-aryl substituents,
the polymerisation did not proceed in a controlled man-
ner and low molecular weight polymers were produced
throughout, the molecular weight of which were inde-
pendent of the degree of conversion (Fig. 5). The poly-
mer produced had unsaturated end groups by 1H
NMR spectroscopy and zero halide content by micro-
analysis. Thus, complexes 5 and 6 do not polymerize sty-
rene by an ATRP mechanism, but rather chain
propagation is accompanied by catalytic chain transfer
(CCT). Here, the polystyrenyl radical reacts with the
Fe(II) catalyst to generate an organometallic Fe(III)
species which undergoes b-hydrogen elimination to
release the polymer chain bearing an unsaturated end-
group. This is presumed to occur in an analogous man-
ner to that proposed for the a-diimine catalyst system
shown in Scheme 2 [14]. The capacity of the aryl deriv-
atives to support CCT, we believe, is a result of a lower
spin (S = 3/2) ground state for the oxidised Fe(III) spe-
cies containing aryl substituents (the alkyl derivatives
are all high spin) affording stronger Fe–C bonds and
consequently more stable organometallic intermediates.
Fig. 5. Plot of Mn vs. conversion for 4 (m) and 5 (D).
3. Conclusions
Iron(II) dichloride complexes containing neutral tri-
dentate [NNN] ligands are found to be active ATRP cat-
alysts, and their activities correlate well with their
oxidation potentials. Although their electrochemical
properties suggest that they should be more efficient cat-
alysts than four-coordinate a-diimine iron systems, they
are in fact much slower. This is explained by the in-
creased steric hindrance at the metal for the oxidised
six-coordinate species, which prevents fast and facile
atom transfer from occurring. Thus the hypothesis made
Cl
nPh
Ph
.
N
N
R
R
n Ph
Ph
+
Fe
Cl
ATRP
Cl
Cl
N
N
R
R
Fe
Cl
Cl
N
N
R
R
N
N
n
Fe
R
R
Ph
+
Ph
Fe
Cl
n Ph
H
Cl
Ph
.
Cl
Cl
n Ph
Ph
β -H elimination
CCT
Scheme 2. ATRP vs. CCT pathways for a-diimine iron catalysts.

2926
R.K. OꢀReilly et al. / Polyhedron 23 (2004) 2921–2928
at the outset of this study, that a four/five-coordinate
equilibrium should be more favourable than a five/six-
coordinate equilibrium for efficient ATRP catalysis in
iron complexes appears to hold.
and the ampoule was then sealed. In all cases the cata-
lyst was soluble in the monomer solution. The ampoules
were heated in an oil bath, at 120 ꢁC. After stirring for
the allotted period of time, an aliquot (0.1 ml) was re-
moved and quenched with THF (1 ml). Conversion
was determined by integration of the monomer versus
polymer backbone resonances in the 1H NMR spectrum
of the crude product in CDCl3. After the reaction time,
usually 24 h, the contents of the ampoules were dis-
solved in THF. This solution was added dropwise to
an approximately 20-fold excess of rapidly stirred acid-
ified methanol (1% HCl v/v). The precipitate that
formed was filtered off and washed with methanol.
The precipitate was dried for 24 h in a vacuum oven
at 60 ꢁC. Samples were then analysed by GPC.
CV was performed using a sample vial (50 · 24 mm)
as a cell. The counter electrode was a section of 0.5 mm
platinum wire, the working electrode was a platinum
disk electrode and the reference electrode was a Ag/
AgCl electrode. Electrolyte ([nBu4N][PF6] (1.00 mmol))
and the complex under study (0.01 mmol) was loaded
into the cells before being degassed by evacuation and
admission of dinitrogen, acetonitrile (10 cm3) was then
added.
4. Experimental
4.1. General
All manipulations of water and/or moisture sensitive
compounds were performed by means of standard high
vacuum Schlenk and cannula techniques. Air sensitive
compounds were transferred to a nitrogen filled glove-
box and, unless stated otherwise, stored at room temper-
ature. Crystal data were collected on Siemens P4/PC or
P4/RA diffractometers. NMR spectra were recorded on
a Bruker AC-250 MHz, DRX-400 MHz or AM-500
MHz spectrometers. Infra-red spectra were obtained
with thin sample films on NaCl plates or as KBr discs
on a Perkin–Elmer 1710X FT-IR spectrometer. Mass
spectra were recorded on either a VG Autospec or a
VG Platform II spectrometer. Elemental analyses were
performed by the microanalytical services of the Chem-
istry departments of London Metropolitan University
and University College London. Magnetic susceptibili-
ties were determined by the Evans NMR method. Gel
permeation chromatography (GPC) was performed
using Viscotek Trisec software connected to a Knauer
differential refractometer. Samples were injected onto
two linear 10 lm columns using chloroform as elutant
(conventionally calibrated prior to use, using polysty-
rene standards (PSS Mainz)) at a flow rate of 1.0
cm3 minꢄ1 at room temperature. Cyclic voltammetry
(CV) was performed using a MacLab potentiostat oper-
ated by EChem 1.3.2. software. The working electrode
and reference electrode were purchased from Bioanalyt-
ical (ref: MF-2013 and RE-5B).
4.2. Preparation of 2,6-bis-[(dimethylamino)methyl]-
pyridine) (II)
Dimethylamine (6.0 g, 0.133 mol) was quickly added
to 30 ml of precooled (ca. 5 ꢁC) C6H6 after which a solu-
tion of 2,6-bis(bromomethyl)pyridine (3.51 g, 0.013 mol)
in 80 ml of C6H6 was added dropwise over 30 minutes
[12,13]. After stirring at room temperature for 30 min,
the reaction mixture was filtered under suction and the
residue was washed with Et2O. The C6H6 and Et2O frac-
tions were combined and reduced in vacuo to yield a
pale yellow oil. This oil was vacuum distilled at 60 ꢁC
and 8 mm/Hg to afford a colourless oil (2.45 g, 95%).
Anal. Calc. for C11H19N3: C, 68.4; H, 9.9; N, 21.7.
Found: C, 67.7; H, 9.5; N, 21.7%. dH (CDCl3) 7.50
(1H, t, J(HH) 8 Hz, 1Ar–H), 7.15 (2H, d, J(HH) 8
Hz, 2Ar–H), 3.46 (4H, m, 2CH2), 2.15 (12H, s, 4CH3);
dC (CD3COCD3) 159.6 (s, 2Ar–C), 137.3 (s, 1Ar–C),
121.5 (s, 2Ar–C), 66.5 (s, 2CH2), 45.9 (s, 4CH3); MS
(CI+) (m/z) 194 [M+H]+.
All solvents were dried by prolonged reflux over a
suitable drying agent under an atmosphere of nitrogen,
being freshly distilled prior to use. All solvents were de-
gassed prior to use unless otherwise stated. Styrene (St)
was purified by vacuum distillation and then stored un-
˚
der an inert atmosphere over 4 A molecular sieves at
ꢄ15 ꢁC. 1-PECl was synthesized according to literature
methods and purified by column chromatography [41].
All amines and anilines were freshly distilled before
use. All other reagents are commercially available and
were used without further purification. Compounds
1[23] 5[21] and 6[33] were prepared by literature
procedures.
All polymerisations were set up and performed under
an atmosphere of oxygen-free, dry dinitrogen using
standard Schlenk-line techniques. In an ampoule
equipped with a magnetic stirrer bar, were placed in or-
der, monomer, initiator and catalyst in various ratios
4.3. Preparation of 2,6-bis-(cyclododecylimino)pyridine
(IV)
A solution of 2,6-diacetylpyridine (5.0 g, 0.03 mol)
and cyclododecylamine (13.8 g, 0.075 mol) in Si(OEt)4
(14.05 g, 0.067 mol) with a few drops of concentrated
H2SO4 was placed in a round bottom flask and attached
to a side arm and air condenser which was stoppered
with a round bottom flask [18]. The solution was heated
overnight with stirring at 130 ꢁC, after which time etha-

R.K. OꢀReilly et al. / Polyhedron 23 (2004) 2921–2928
2927
nol (ca. 3 ml) had been distilled over from the reaction
mixture. The solid reaction mixture was dissolved in
CH2Cl2, washed with saturated solutions of NaHCO3
(2 · 10 ml) and brine (2 · 10 ml) and then dried over
MgSO4. The solvent was removed in vacuo to afford a
yellow solid. Washing this solid with cold petroleum
ether (20 ml) afforded ligand IV as a bright yellow solid
(9.17 g, 62%). Anal. Calc. for C35H55N3: C, 80.3; H,
11.2; N, 8.5. Found: C, 79.6; H, 11.1; N, 8.7%. IR/
cmꢄ1 (C@N) 1650s; dH (CDCl3) 8.03 (2H, d, J(HH) 8
Hz, 2Ar-H), 7.66 (1H, t, J(HH) 8 Hz, Ar–H), 3.82
(2H, m, 2CH), 2.41 (6H, s, 2CH3), 1.58–1.25 (44H, br
m, 22CH2); dC (CDCl3) 163.9 (s, 2Ar–C), 156.7 (s,
2Ar–C), 136.4 (s, Ar–C), 121.1 (s, 2Ar–C), 57.04 (s,
2CH), 40.35 (s, 2CH3), 24.7–21.3 (s, 22CH2); MS (CI+)
(m/z) 494 [M+H]+.
1117w, 1026w, 923w, 750w; MS (CI+) (m/z) 318
[M+NH4]+; leff = 4.96 lB.
4.6. Preparation of 2,6-bis-(cyclododecylimino)pyridine
iron(II) chloride (4)
To a suspension of FeCl2 (0.92 g, 0.007 mol) in
nBuOH (20 ml) was added dropwise, with stirring, IV
(4 g, 0.008 mol) also in nBuOH (10 ml) and the reaction
mixture stirred overnight at 90 ꢁC. After this time it was
allowed to cool and the solvent was then removed in va-
cuo at 60 ꢁC for 4 h. The residue was washed with Et2O
(3 · 10 ml) and then extracted into THF (20 ml). The
solution was reduced to ca. 10 ml and stored at ꢄ10
ꢁC overnight. After this time, dark grey crystals of 4
had formed which were isolated by filtration and dried
(3.30 g, 73%). Anal. Calc. for C33H55N3Cl2Fe: C, 63.9;
H, 8.9; N, 6.8; Cl, 11.4. Found: C, 63.7; H, 8.7; N,
6.8; Cl, 10.3%. IR/cmꢄ1 2938s, 2864s, 2362w, 1703m,
1585w, 1470m, 1446m, 1359w, 1292w, 1268w, 1205w,
1163w, 1075w, 1020w, 994w, 813w, 740w; MS (FAB+)
(m/z) 584 [MꢄCl]+; leff = 5.11 lB.
4.4. Preparation of 2,6-bis-[(dimethylamino)methyl]-
pyridine) iron(II) chloride (2)
To a solution of FeCl2(THF)1.5 (0.58 g, 0.0025 mol)
in toluene (20 ml) was added dropwise, with stirring,
ligand II (0.48 g, 0.0025 mol) also in toluene (20 ml)
and the resulting suspension stirred overnight at room
temperature. The mixture was then cooled to 0 ꢁC
and filtered; the filtrate was reduced to ca. 10 ml and
stored overnight at ꢄ10 ꢁC, after which time crystals
of 2 had formed. These yellow crystals were isolated
by filtration and dried overnight in vacuo. This af-
forded 2 as a microcrystalline bright yellow solid
(0.67 g, 84%). Slow cooling of a solution of 2 in hot
THF afforded crystals suitable for X-ray analysis. Anal.
Calc. for C11H19N3Cl2Fe: C, 41.2; H, 6.0; N, 13.1.
Found: C, 41.2; H, 6.0; N, 13.0%. IR/cmꢄ1 3584m,
3397s, 2349w, 2282w, 1653m, 1559w, 1541w, 1507w,
1457w, 1073w, 666w; MS (CI+) (m/z) 319 [M]+;
4.7. X-ray crystallography
Crystal data for 2: C11H19N3Cl2Fe, M = 320.04,
96),
tetragonal,
P43212
˚
(no.
a = 10.5887(5),
3
˚
c = 26.080(3) A, V = 2924.1(4) A , Z = 8, Dc = 1.454
gcmꢄ3, l(Cu Ka) = 11.491 mmꢄ1, T = 183 K, yellow
platy prisms; 2142 independent measured reflections,
F2 refinement, R1 = 0.054, wR2 = 0.122, 1720 independ-
ent observed absorption-corrected reflections [|Fo| > 4r
(|Fo|), 2hmax = 120ꢁ], 155 parameters. The absolute struc-
ture of 1 was determined by a combination of R-factor
tests ½Rþ1 ¼ 0:0541; Rꢄ1 ¼ 0:1025ꢇ and by use of the
Flack parameter [x+ = ꢄ0.013(14)]. CCDC 239659.
Crystal data for 3: C10H15N3Cl2Fe Æ 0.5CH2Cl2,
M = 346.46, monoclinic, I 2/a (no. 15), a = 14.870(3),
leff = 5.05 lB.
4.5. Preparation of N,N-dimethyl-N0-pyridin-2-
ylmethylene-ethane-1,2-diamino iron(II) chloride (3)
b = 10.2360(8), c = 20.545(4) A, b = 104.216(11)ꢁ,
˚
V = 3031.4(9) A , Z = 8, Dc = 1.518 gcmꢄ3, l(Mo
3
˚
Ka) = 1.508 mmꢄ1, T = 293 K, very dark red/brown
needles; 2666 independent measured reflections, F2
refinement, R1 = 0.046, wR2 = 0.103, 2011 independent
observed absorption-corrected reflections [|Fo| > 4r
(|Fo|), 2hmax = 50ꢁ], 173 parameters. CCDC 239660.
To a suspension of FeCl2 (1.27 g, 0.01 mol) in Et2O
(20 ml) was added dropwise, with stirring N,N-di-
methyl-N0-pyridin-2-ylmethylene-ethane-1,2-diamine,
III (1.77 g, 0.01 mol) in Et2O (10 ml) [24]. The reaction
mixture was stirred overnight at room temperature and
then filtered and the filtrate reduced under reduced pres-
sure. The solid residue was extracted into CH2Cl2 and
then the solvent was removed. The resultant solid was
dried in vacuo and this afforded 3 as a purple microcrys-
talline solid (2.01 g, 66%). Slow diffusion of a solution of
3 in CH2Cl2with an equal volume of heptane afforded
crystals suitable for X-ray analysis. Anal. Calc. for
C10H15N3Cl2Fe: C, 39.5; H, 5.0; N, 13.8. Found: C,
39.1; H, 5.0; N, 13.2%. IR/cmꢄ1 3447m, 2966m,
2838w, 1615w, 1568m, 1507w, 1457w, 1279w, 1174w,
5. Supplementary material
Full crystallographic details have been deposited in
CIF format with the Cambridge Crystallographic Data
Centre. Copies of this information may be obtained free
of charge from The Director, CCDC, 12 Union Road,
Cambridge, CB2 1EZ, UK (fax: +44 1223 336033;
e-mail: deposit@ccdc.cam.ac.uk or www: http//
www.ccdc.cam.ac.uk).

2928
R.K. OꢀReilly et al. / Polyhedron 23 (2004) 2921–2928
[18] P. Molenveld, W.M.G. Stikvoort, H. Koojiman, A.L. Spek,
Acknowledgements
J.F.J. Engbersen, D.N. Reinhoudt, J. Org. Chem. 64 (1999) 3896.
[19] G. Zahrzewski, G.P. Speroni, Inorg. Chem. 7 (1968) 1034.
BP Chemicals is thanked for financial support (R.K.
OꢀR). Dr. Jane Boyle is thanked for NMR
measurements.
´
[20] F. Guerin, D.H. Mc Conville, J.J. Vittal, Organometallics 15
(1996) 5586.
[21] 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. Stro¨mberg, A.J.P. White, D.J. Williams, J. Am. Chem.
Soc. 121 (1999) 8728.
References
[22] B.E. Love, J. Ren, J. Org. Chem. 58 (1993) 5556.
[23] M. Ciampolini, G.P. Speroni, Inorg. Chem. 5 (1966) 45.
[24] F. Calderazzo, U. Englert, G. Pampaloni, E. Vanni, C.R. Acad.
Sci. Paris (1999) 311.
[1] (a) For recent reviews see K. Matyjaszewski, J. Xia, Chem. Rev.
101 (2001) 2921;
(b) M. Kamigaito, T. Ando, M. Sawamoto, Chem. Rev. 101
(2001) 3689.
[25] I. Del Rio, R.A. Gossage, M.S. Hannu, M. Lutz, A.L. Spek, G.
van Koten, Organometallics 18 (1999) 1097.
[2] J.S. Wang, K. Matyjaszewski, J. Am. Chem. Soc. 117 (1995) 5614.
[3] M. Kato, M. Kamigaito, M. Sawamoto, T. Higashimura,
Macromolecules 28 (1995) 1721.
[26] B.L. Small, M. Brookhart, A.M.A. Bennett, J. Am. Chem. Soc.
120 (1998) 4049.
ˆ ´ ´ ˆ
[4] P. Lecomte, I. Drapier, P. Dubois, P. Teyssie, R. Jerome,
[27] M.W. Day, B.D. Ward, R.H. Grubbs, Private Communication,
CCDC OCUJOL (2001).
Macromolecules 30 (1997) 7631.
[5] G. Moineau, C. Granel, P. Dubois, P. Teyssie, R. Jerome,
Macromolecules 31 (1998) 542.
´
´ ˆ
[28] M. Costa, P. Pelagatti, C. Pelizzi, D. Rogolino, J. Mol. Catal. A
178 (2002) 21.
ˆ ´ ´
[6] C. Granel, P. Dubois, R. Jerome, P. Teyssie, Macromolecules 29
[29] P.S. Mukherjee, T.K. Maji, A. Escuer, R. Vicente, J. Ribas, G.
Rosair, F.A. Mautner, N.R. Chaudhuri, Eur. J. Inorg. Chem.
(2002) 943.
(1996) 8576.
[7] T. Ando, M. Kamigaito, M. Sawamoto, Macromolecules 30
(1997) 4507.
[30] P. Bamfield, R. Price, R.G.J. Millar, J. Chem. Soc. A 9 (1969)
1447.
[8] K. Matyjaszewski, M. Wei, J. Xia, N.E. McDermott, Macromol-
ecules 30 (1997) 8161.
[31] G. Zakrzewski, E.C. Lingafelter, Inorg. Chim. Acta 4 (1970) 251.
[32] G.J.P. Britovsek, V.C. Gibson, B.S. Kimberley, P.J. Maddox,
S.J. McTavish, G.A. Solan, A.J.P. White, D.J. Williams, Chem.
Commun. (1998) 849.
[9] K. Matyjaszewski, S. Coca, S.G. Gaynor, M. Wei, B.E. Wood-
worth, Macromolecules 30 (1997) 7348.
[10] M. Teodorescu, S.G. Gaynor, K. Matyjaszewski, Macromolecules
33 (2000) 2335.
[33] G.J.P. Britovsek, V.C. Gibson, S. Mastroianni, D.C.H. Oakes,
C. Redshaw, G.A. Solan, A.J.P. White, D.J. Williams, Eur. J.
Inorg. Chem. (2001) 431.
[11] T. Sarbu, K. Matyjaszewski, Macromol. Chem. Phys. 202 (2001)
3379.
[12] V.C. Gibson, R.K. OꢀReilly, D.F. Wass, A.J.P. White, D.J.
Williams, J. Chem. Soc., Dalton Trans. (2003) 2824.
[13] V.C. Gibson, R.K. OꢀReilly, D.F. Wass, W. Reed, A.J.P. White,
D.J. Williams, Chem. Commun. (2002) 1850.
[14] V.C. Gibson, R.K. OꢀReilly, D.F. Wass, A.J.P. White, D.J.
Williams, Macromolecules 36 (2003) 2591.
[34] D.F. Evans, J. Chem. Soc. (1959) 2003.
[35] D.F. Evans, G.V. Fazakerley, R.F.J. Phillips, J. Chem. Soc. A
(1971) 1931.
[36] D.H. Grant, J. Chem. Ed. 72 (1995) 39.
[37] B.N. Figgis, J. Lewis, in: J. Lewis, R.G. Wilkins (Eds.), Modern
Coordination Chemistry, Interscience, New York, 1960, p. 400.
[38] K. Matyjaszewski, B. Go¨belt, H.-J. Paik, C.P. Horwitz, Macro-
molecules 34 (2001) 430.
[15] B. Go¨belt, K. Matyjaszewski, Macromol. Chem. Phys. 201 (2000)
1619.
[16] M. Newcomb, J.M. Timko, D.M. Walba, D.J. Cram, J. Am.
Chem. Soc. 99 (1977) 6392.
[39] J. Qui, K. Matyjaszewski, L. Thouin, C. Amatore, Macromol.
Chem. Phys. 201 (2000) 1625.
[17] B.A. Markies, P. Wijkens, J. Boersma, H. Kooijman, N. Veldman,
A.L. Spek, G. van Koten, Organometallics 13 (1994) 3244.
[40] A. Goto, T. Fukuda, Prog. Polym. Sci. 29 (2004) 329.
[41] D. Landini, F. Rolla, J. Org. Chem. 45 (1980) 3527.