Structural and kinetic studies of the polymerization reactions of ε-caprolactone catalyzed by (pyrazol-1-ylmethyl)pyridine Cu(ii) and Zn(ii) complexes

Article abstract of DOI:10.1039/c3dt51338f

The structural and kinetic studies of polymerization reactions of ε-caprolactone (ε-CL) using (pyrazolylmethyl)pyridine Cu(ii) and Zn(ii) complexes as initiators is described. Reactions of 2-(3,5- dimethylpyrazol-1-ylmethyl)pyridine (L1) and 2-(3,5-diphen

Full text of DOI:10.1039/c3dt51338f

Dalton
Transactions
View Article Online
View Journal | View Issue
PAPER
Structural and kinetic studies of the polymerization
reactions of ε-caprolactone catalyzed by (pyrazol-
1-ylmethyl)pyridine Cu(II) and Zn(II) complexes†
Cite this: Dalton Trans., 2013, 42, 10735
Stephen O. Ojwach,*a Teddy T. Okemwa,b Nelson W. Attandoha and
Bernard Omondic
The structural and kinetic studies of polymerization reactions of ε-caprolactone (ε-CL) using (pyrazolyl-
methyl)pyridine Cu(II) and Zn(II) complexes as initiators is described. Reactions of 2-(3,5-dimethylpyrazol-
1-ylmethyl)pyridine (L1) and 2-(3,5-diphenylpyrazol-1-ylmethyl)pyridine (L2) with Zn(Ac)2·2H2O or
Cu(Ac)2·2H2O produced the corresponding complexes [Zn(Ac)2(L1)] (1), [Cu(Ac)2(L1)] (2), [Zn(Ac)2(L2)] (3)
and [Cu2(Ac)4(L2)2] (4) respectively. Solid state structures of 1 and 4 confirmed that complexes 1 and 4
are monomeric and dimeric respectively and that L1 is bidentate in 1 while L2 is monodentate in 4.
X-band EPR spectra of 2 and 4 revealed that complex 2 is monomeric both in solid and solution state,
while the paddle-wheel structure of 4 is retained in solution. Complexes 1–4 formed active initiators in
the ring opening polymerization of ε-CL. Zn(II) complexes 1 and 3 exhibited higher rate constants of
0.044 h−1 and 0.096 h−1 respectively compared to rate constants of 0.017 h−1 and 0.031 h−1 observed for
the corresponding Cu(II) complexes 2 and 4 respectively at 110 °C. All the polymerization reactions follow
pseudo first-order kinetic with respect to ε-CL monomer. Initiator 1 showed first-order dependency on
the polymerization reactions and utilizes only one active site as the initiating group. The molecular
weights of the polymers produced range from 1982 g mol−1 to 14 568 g mol−1 and exhibited relatively
broad molecular weight distributions associated with transesterification reactions.
Received 22nd May 2013,
Accepted 4th June 2013
DOI: 10.1039/c3dt51338f
www.rsc.org/dalton
polymerization (ROP) of cyclic esters through a coordination–
Introduction
insertion pathway involving a metal alkoxide catalyst.3 This
Polyesters derived from renewable resources have attracted route has the potential to offer stereo-selectivity and control of
considerable attention over the last decades as biocompatible molecular weight of the polymers produced.4 Furthermore,
and biodegradable alternatives to petrochemical-based plas- kinetic studies provide detailed mechanistic information per-
tics.1 In particular, polycaprolactone (PCL) has found varied taining to this pathway5 including insight into effects of cata-
applications in biomedical and pharmaceutical fields.2 These lyst structure on polymerization activity.6
aliphatic polyesters are traditionally prepared via ring opening
To date, the design of well-defined catalysts that could
produce polyesters with desirable molecular weight and
narrow molecular weight distribution still remains a daunting
task. While tin-based systems7 show promising results in
terms of control of polymer molecular weight and good
activity, their toxicity limits their industrial appeal.8 Several
discrete complexes ranging from metallocenes9 to early tran-
sition metal complexes like Ti,10 Zr,11 Hf12 have been investi-
gated as catalyst initiators for the ROP of cyclic esters. Despite
encouraging results in these studies, numerous challenges
aSchool of Chemistry and Physics, University of KwaZulu-Natal, Pietermaritzburg
Campus, Private Bag X01 Scottsville, 3209, South Africa. E-mail: ojwach@ukzn.ac.za;
Fax: +27 (33) 260 5009; Tel: +27 (33) 260 5239
bDepartment of Chemistry, Maseno University, P. O. Box 333, Maseno, 4105, Kenya
cSchool of Chemistry and Physics, University of KwaZulu-Natal, Westville Campus,
Private Bag X54001, Durban, 4000, South Africa
†Electronic supplementary information (ESI) available: Supplementary Fig. S1–S7
represent graphs of the percentage conversions of ε-CL to PCL for catalysts
1–4 (S1), 1H NMR spectra for the determination of the rates of polymerization
(S2), plot of ln[CL]0/[CL]t vs. time for 1 at different concentrations (S3), plot of namely; catalyst stability, poor control of polymer microstruc-
ln[CL]0/[CL]t vs. time for 2 at different catalyst concentrations depicting the longer
ture to cost of the catalysts are hindering their commercializa-
tion. One group of complexes that have the propensity to
induction periods (S4); plot of kobs vs. [1] for the determination of threshold cata-
lyst concentration (S5) and GPC chromatograms for PCL obtained (S6–S8). CCDC
produce effective catalysts for the ROP polymerization of cyclic
numbers 828074 and 927444 contain the supplementary crystallographic data
esters are the Zn(II) and Cu(II) complexes.13 The suitability
for 1 and 4. For ESI and crystallographic data in CIF or other electronic format
see DOI: 10.1039/c3dt51338f
of these complexes emanates from their ease of synthesis,
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 10735–10745 | 10735

View Article Online
Paper
Dalton Transactions
stability and more significantly their biocompatibility. Recently,
The compounds were characterized by 1H NMR spec-
Zn(II) α-diiminate complexes were reported to portray both troscopy for 1 and 3, mass spectrometry, elemental analyses
control of polymer stereo-regularity and excellent activity in and single crystal X-ray crystallography for 1 and 4. 1H NMR
the ROP of lactides.14 This control of polymer microstructure spectra of complexes 1 and 3 showed a singlet peak at about
is largely influenced by the steric bulk of the ligand used, and 2.10 ppm, diagnostic of the acetate protons. In addition, a
as such careful design of the ligand motif could result in con- singlet peak observed at 5.46 ppm and 5.59 ppm in 1 and 3
trolled ROP of the cyclic esters.
respectively were assigned to the CH2 linker protons. Micro-
In this current contribution, we explore the potential use of analyses data of 1–4 were consistent with one metal atom per
(pyrazol-1-ylmethyl)pyridine Zn(II) and Cu(II) acetate complexes ligand unit as shown in Scheme 1, and also confirmed their
as catalyst initiators in the ROP of ε-CL. We envisage that, by purity. Mass spectra of 1–4 showed m/z peaks corresponding to
regulating the steric bulk on the ligand backbone, proper the fragments of the parent compounds. For example, the
control of polymer molecular weight and molecular weight dis- mass spectra of 1 and 2 showed base peaks at m/z = 310 and
tribution could be achieved. Detailed structural and kinetics 309 respectively, associated with the loss of one acetate ligand
studies have been performed in order to elucidate the effect of [M+ − Ac]. The absence of the molecular ions [M+] could origi-
catalyst structure and nature of active species on the polymeri- nate from the lower stability of the parent compounds under
zation behaviour of the complexes. These findings are herein the ionization conditions.15
discussed.
Single crystals suitable for X-ray crystallography analyses for
1 and 4 were grown by slow diffusion of hexane into a dichloro-
methane solution of the complexes. Data collection and struc-
ture refinement parameters are given in Table 1 while Fig. 1
and 2 show the molecular structures and bond parameters for
1 and 4 respectively. In the solid structure of 1, L1 adopts a
bidentate coordination mode. The acetate as a ligand displays
a wide variety of binding modes similar to carbonate and
nitrate ligands: it can act as monodentate, bidentate and aniso-
dentate. Due to this flexibility in its coordination behaviour,
we have used a literature method16 to deduce the denticity of
the acetate ligand in 1. The differences between the two M–Onitrate
bond distances (Δd) and M–O–C bond angles (Δθ) are
used to classify the denticity (Δd < 0.3 Å and Δθ < 14° for
bidentate; 0.3 Å < Δd < 0.6 Å and 14° < Δθ < 28° for anisoden-
tate (Δd > 0.6 Å and Δθ > 28° for monodentate). Using this
approach, one acetate ligand is clearly monodentate while the
other one is anisodentate (Δd = 0.6 Å and Δθ = 17°) in 1. The
Results and discussion
Synthesis and structural characterization of Zn(II) and Cu(II)
(pyrazol-1-ylmethyl)pyridine complexes 1–4
Reactions of 2-(3,5-dimethylpyrazol-1-ylmethyl)pyridine (L1)
and 2-(3,5-diphenylpyrazol-1-ylmethyl)pyridine (L2) with
Zn(Ac)2·2H2O produced the corresponding monometallic Zn(II)
complexes [Zn(Ac)2(L1)] (1) and [Zn(Ac)2(L2)] (3). Similarly
treatment of L1 and L2 with Cu(Ac)2·2H2O produced mono-
metallic and bimetallic complexes [Cu(Ac)2(L1)] (2) and
[Cu2(Ac)4(L2)2] (4) respectively (Scheme 1). Complexes 1–4 were
obtained in moderate to high yields (62%–81%). Complexes 1
and 3 were isolated as pale yellow solids, while 2 and 4 were
obtained as blue solids.
Scheme 1
10736 | Dalton Trans., 2013, 42, 10735–10745
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
Table 1 Crystal data collection and structural refinement parameters for 1 and 4
Parameters
1
4
Empirical formula
Formula weight
Temperature
Wavelength
Crystal system
Space group
C
15H19N3O4Zn
C50Cu2H46O8N6O8
986.01
100(2) K
0.71073 Å
Triclinic
ˉ
P1
370.70
100(2) K
0.71073 Å
Orthorhombic
Pna21
a/Å
b/Å
c/Å
α
14.7323(10)
8.1094(5)
13.5844(8)
90°
90°
90°
1622.93(18)
9.6202(3)
10.6016(3)
11.7735(3)
72.8800(10)°
75.3930(10)°
78.1930(10)°
1099.39(3)
1
β
γ
Volume (Å3)
Z
4
Density (calculated)
Absorption coefficient
F(000)
1.517 Mg m−3
1.489 Mg m−3
1.032 mm−1
510
1.536 mm−1
768
Crystal size (mm3)
Theta range for data collection
Index ranges
0.14 × 0.14 × 0.06
2.77 to 28.35°
−19 ≤ h ≤ 19; −10 ≤ k ≤ 10; −17 ≤ l ≤ 18
30 990
4040 [R(int) = 0.0474]
(=28.35°) 99%
0.9135 and 0.8137
Semi-empirical from equivalents
4040/4/213
0.41 × 0.27 × 0.15
1.85 to 28.52°
−12 ≤ h ≤ 12; −12 ≤ k ≤ 14; −15 ≤ l ≤ 15
27 490
5516 [R(int) = 0.0248]
(=28.52°) 98.8%
0.8606 and 0.6771
Semi-empirical from equivalents
5516/0/300
1.030
Reflections collected
Independent reflections
Completeness to theta
Max and min transmission
Absorption correction
Data/restraints/parameters
Goodness-of-fit on F2
Final R indices [I > 2σ(I)]
R indices (all data)
Largest diff. peak and hole
1.029
R1 = 0.0253, wR2 = 0.0574
R1 = 0.0300, wR2 = 0.0593
0.398 and −0.182 e Å−3
R1 = 0.0243, wR2 = 0.0667
R1 = 0.0260, wR2 = 0.0683
0.472 and −0.488 e Å−3
Fig. 1 Molecular structure of complex 1 showing both the four-coordinate (a) and five coordinate (b) geometries drawn with 50% probability thermal ellipsoids.
Hydrogen atoms are omitted for clarity. Selected bond lengths [Å] and angles (°): O(2)–Zn(1), 1.9653(8); O(3)–Zn(1), 1.9649(13); O(1)–Zn(1), 2.566(2); Zn–N(1),
2.066(2); Zn(1)–N(3); 2.102(2). O(3)–Zn(1)–O(2), 141.14(6); O(3)–Zn(1)–N(1), 96.89(7); O(2)–Zn(1)–N(1), 115.81(7); O(3)–Zn(1)–(N(3), 103.52(4); O(2)–Zn(1)–N(3),
97.21(7); N(1)–Zn(1)–N(3), 89.99(6).
geometry around the Zn atom in 1 could thus be best for O(3)–Zn–O(2) significantly deviate from 109° expected for a
described as distorted tetrahedral (Fig. 1a). Five-coordination tetrahedral geometry. This could originate from steric restric-
sphere to give a trigonal bipyramidal geometry is also likely in tions imposed by L1 and flexibility of the acetate ligand. The
the case of bidentate coordination of the anisodentate acetate Zn(1)–N(1) and Zn(1)–N(3) bond distances of 2.066(2) Å and
ligand (Fig. 1b). This fluxionality has the potential to influence 2.102(2) Å respectively are normal and comparable to those of
the catalytic behaviour of 1, where a given geometry exhibits related compounds in literature.17 The average Zn–Oacetate
different catalytic behaviour. The bond angles around the bond lengths of 1.965(2) Å fall within the expected range of
Zn(II) atom of between 89.99(6)° for N(1)–Zn–N(3) to 141.14(6)° Zn–Oacetate distances.18
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 10735–10745 | 10737

View Article Online
Paper
Dalton Transactions
Fig. 2 Molecular structure of the paddle wheel copper complex 4 drawn with 50% probability thermal ellipsoids. Hydrogen atoms are omitted for clarity. Selected
bond lengths [Å] and angles (°): O(1)–Cu(1), 1.9748(9); O(2)–Cu(1), 1.9875(9); O(3)–Cu(1), 1.9716(9); O(4)–Cu(1), 1.9665(9); Cu(1)–N(3), 2.2138(10). Cu(1)–Cu(1)#,
2.6532(3). O(3)–Cu(1)–O(4), 90.93(4); O(4)–Cu(1)–O(1), 168.06(4); O(3)–Cu(1)–O(1), 87.12(4); O(4)–Cu(1)–(N(3), 98.52(4); O(1)–Cu(1)–N(3), 93.34(4); O(2)–Cu(1)–
N(3), 89.85(4).
As depicted in Fig. 2, complex 4 exhibits a bimetallic complex19 of 2.0248 (3) Å and 1.991(4) Å respectively. The Cu–
paddle wheel conformation with an octahedral geometry Cu bond distance of 2.5316(6) Å in 4 is slightly shorter than
around the Cu(II) atom. Interestingly, L2 is monodentate, the average Cu–Cu bond lengths of 2.671(2) Å reported in
binding via the pyridine nitrogen atom, with the pyrazolyl literature.20
nitrogen atom uncoordinated. This contrasts the coordination
mode observed for L1 in 1. The four acetate ligands bridge the
EPR spectra of complexes 2 and 4
two Cu(II) centres to complete the octahedral arrangement.
The molecule has a crystallographic C2-symmetry axis passing
through C24 and C25 and bisecting the Cu–Cu bond thus ren-
dering many of the atoms symmetry equivalent. The bond
angles of 90.93(4)° and 91.55(4)° for O(4)–Cu(1)–O(3) and
O(1)–Cu(1)–O(2) respectively in 4 are close to the expected
bond angle of 90° for an octahedral geometry. This slight dis-
tortion could be due to reduced crowding around the Cu atom
as a result of the monodentate nature of L2. A closer examin-
ation of the structure reveals that the bulk of the ligand resides
away from the metal center. The average Cu–Npyridine and Cu–
Oacetate bond distances of 2.2138(10) Å and 1.9775(9) Å are
comparable to those reported for the 2-amino pyridine Cu
In order to understand the coordination environment of the
copper complexes both in solid state and in solution, X-band
EPR spectra of 2 and 4 were acquired in solid and methanol
solution at room temperature (Fig. 3). Both the solid and solu-
tion state EPR spectra of 2 (Fig. 3a) exhibited an axial (tetra-
gonal) g-tensor with a dx2−y2 ground-state doublet consistent
with monomeric Cu(II) complex.21a In solution state, the gz
signal of 2 shows evidence of hyperfine coupling of the single
electron to the Cu(II) nucleus while in the solid state, the
spectra does not exhibit copper hyperfine structure. This could
be attributed to very short spin-lattice relaxation times in the
solid state. Similar broad EPR spectra have been observed
for some copper carboxylate dimers.21c,d EPR spectra of 4 both
Fig. 3 X-band EPR spectra of (a) complex 2 and (b) complex 4 in solid and methanol solution at 298 K. 2 in solid: gz = 2.292; gy = 2.069; 2 in solution: gz = 2.174;
y = 2.087, A = 61. 4 in solution: gz = 2.207; gy = 2.151, gx = 2.0883. 4 in solid state: gz = 2.211; gy = 2.009, gx = 1.45.
g
10738 | Dalton Trans., 2013, 42, 10735–10745
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
Table 2 Summary of ε-CL polymerization data by complexes 1–4a
in solid and in methanol solution (Fig. 3b) display rhombic
tensor with the unpaired electron occupying the dx2−y2 orbital
typical of dimeric paddle-wheel structure.22 This confirms that
the dimeric structure of 4 is retained in solution, consistent
with stronger Cu–O bonds of carboxylates as opposed to
weaker Cu–Cl bridging bonds which results in dissociations of
Cu dimers when Cl is the bridging ligand.23 The solid state
EPR spectra of 4 shows evidence of hyperfine electron–Cu
nucleus coupling, while no hyperfine coupling is evident in
the solution spectra.21b,22 Smaller zero-field splitting (A = 61)
in addition to lower g values of 2 and 4 are in good agreement
with the existence of a pyridine nitrogen donor atom in the
copper coordination sphere.23 More significantly, solid state
EPR spectrum of 4 demonstrates that the two Cu atoms in the
dimer are not anti-ferromagnetically coupled (magnetically
dilute). The longer Cu–Cu bond distance of 2.5316(6) Å in 4
compared to a Cu–Cu bond length of 2.19 Å required to
achieve effective Cu–Cu metal interaction supports the EPR
data.24 The monomeric nature of 2 could be assigned to the
less steric requirements by L1 as opposed to the sterically
demanding L2 in 4. From this approach, we cannot rule out
the possibility of complex 3 existing as a dimer in the solid
state. So far attempts to obtain single crystals suitable for X-ray
analyses to confirm the solid state structures of 2 and 3 have
been unsuccessful.
Catalyst
Time (h)
Conversion (%)
Mwb
PDIb
IEc
1
24
32
48
43
68
94
1982
2413
2928
2.60
2.58
3.23
0.40
0.31
0.27
3
4
8
12
24
36
48
33
48
59
92
98
99
2121
2454
2845
3853
4111
4726
3.13
3.24
3.48
3.33
3.74
3.92
0.56
0.45
0.42
0.37
0.37
0.42
4
48
72
96
78
93
94
2749
3814
4652
2.84
3.52
3.96
0.31
0.36
0.43
1d
3d
48
72
32
82
2274
3413
3.12
3.17
0.63
0.37
12
24
48
45
73
98
3089
3338
4365
3.82
3.14
3.27
0.61
0.40
0.39
a Reaction conditions, [CL]0, 0.01 mol, temperature, 110 °C, bulk
polymerization. b Molecular-weight average and polydispersity index
(PDI) determined by GPC relative to polystyrene standards. c Initiator
efficiency (IE) = Mwexp/Mwcalc where Mwcalc = Mw(monomer) × [CL]0/[I] ×
[PCL/[CL]0 + Mw(chain-end
.
d Addition of second equivalent of
groups)
ε-CL without adding the initiator.
Polymerization of ε-caprolactone using 1–4 as initiators
Preliminary investigations of complexes 1–4 as catalyst
initiators in the ring opening polymerization (ROP) reactions
of ε-CL were performed at 110 °C in bulk using [M]/[I] ratio of
100 : 1. Under these conditions, all the complexes exhibited
significant catalytic activities within 24 h for 1 and 3 and 48 h
for 2 and 4 (Fig. S1†). Having established that complexes 1–4
form effective initiators in the ROP of ε-CL, detailed mechani-
stic and kinetics studies were performed in order to gain
insight into the nature of the active species, kinetics of the
reactions and influence of catalyst structure and reaction con-
ditions on catalyst activity and polymer properties.
Kinetics of ε-CL polymerization reactions
Kinetics of the ε-CL polymerization was investigated for com-
plexes 1–4 by monitoring the reactions using 1H NMR spec-
troscopy. Sampling was done at regular intervals and
percentage conversions of ε-CL to PCL determined by compar-
ing the intensity of the PCL signals at 4.0 ppm to that of the
ε-CL monomer at 4.2 ppm (Fig. S2†). A summary of the polymer-
ization data is given in Table 2. A plot of ln[CL]0/[CL]t versus
time gave a linear relationship consistent with a pseudo-first
order kinetics with respect to ε-CL for all the complexes
(Fig. 4). The kinetics of the ε-CL polymerization reactions thus
proceeded according to the simple pseudo first-order kinetics
with the respect to ε-CL as shown in eqn (1).
Fig. 4 First order kinetic plots of ln[CL]0/[CL]t vs. time for (a) Zn complexes 1
and 3 and (b) Cu complexes 2 and 4 in the bulk polymerization of ε-CL at
110 °C, [CL]0, 0.01 mol, [CL]0/[I] = 100.
d½CLꢀ
¼ k ½CLꢀ
ð1Þ
dt
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 10735–10745 | 10739

View Article Online
Paper
Dalton Transactions
where k = kp[I]x, kp = rate of chain propagation and I = initiator;
x = order of reaction.
The rate constants for initiators 1–4 were extracted from
Fig. 4 and obtained as 0.044 h−1 (1), 0.017 h−1 (2), 0.096 h−1
(3) and 0.031 h−1 (4). Initiator 3 was thus the most active while
initiator 2 was the least active. Higher activities observed for
Zn initiators 1 and 3 in comparison to the Cu analogues 2 and
4 is consistent with literature reports.25 More evident was the
increase in catalytic activity with increase in steric bulk of the
pyrazolyl ligand in 1–4. For instance, replacing the Me groups
in 1 with the bulkier Ph groups in 3 resulted in a two-fold
increase in the rate of reaction from 0.044 h−1 to 0.096 h−1
respectively. This observation agrees with the reports of Silvernail
et al.26 In our case, the bimetallic nature of 4 could be
responsible for its higher activity compared to the monometal-
lic complex 2. As argued from EPR data, it is also possible that
complex 3 is bimetallic hence its greater catalytic activity than
the corresponding Zn complex 1.
Fig. 5 Linear plots of ln[kobs] vs. ln[1] polymerization of caprolactone at [CL]0
0.01 mol, 110 °C for the determination of order of reaction with respect to 1.
=
aggregation of the active sites during polymerization reac-
tions.29 Thus the overall rate law for ε-CL polymerization by 1
can be represented as shown in eqn (2). This rate law is con-
sistent with a mechanism involving a coordinative insertion at
a single Zn site. It is therefore apparent that only one Zn–Oacetate
in complex 1 (Fig. 1) acts as the initiating group. A closer
examination of ln[kobs] versus ln[2] showed the presence of an
induction period especially at low catalyst concentrations
(Fig. S4†). This behaviour has been largely attributed to
rearrangement of the coordinative aggregates in the
initiator.30,31 For 2, this may be associated with dissociation of
the Cu–Cu and Cu–OAc bonds prior to ε-CL coordination. Thus
attempts to determine the order of the reaction with respect to
initiator 2 were unsuccessful due to the longer induction
periods which rendered the plots non-linear (Fig. S4†).
Comparatively, the rate constants in the polymerization of
ε-CL for 1–4 are lower than some of the most active initiators
reported.27 A number of very active zinc catalysts are multi-
nuclear and contain alkoxides as the initiating groups. For
example, the trinuclear Zn complex reported by Chen et al.
exhibits a rate constant of 0.0508 s−1 in the polymerization of
ε-CL.27b Despite the relative low activities of 1–4, they were
found to be more active than the aluminium alkoxide catalyst
reported by Zhong et al.28 which displays apparent rate con-
stant of 0.067 day−1
.
Further kinetics was performed to assess the order of the
reaction with respect to initiators 1 and 3 and the overall rates
of reactions. This was done by carrying out the polymeri-
zation reactions at different catalyst concentrations (Fig. S3†)
d½CLꢀ
at constant concentration of ε-CL (Table 3). A plot of ln[kobs
]
0:8
¼ k ½CLꢀ ½1ꢀ
ð2Þ
versus ln[1] gave a linear relationship consistent with a first-
order dependency of reaction on 1 (Fig. 5). The order of the
reaction with respect to 1 was thus extracted from the plot of
ln[kobs] versus ln[1] and obtained as 0.8. Fractional orders of
reaction with respect to the initiator has been previously
observed29 and is believed to be due to complicated
dt
The number of active centers (n)
To determine the number of active sites in initiator 1, a plot of
the degree of polymerization (Xn) vs. [Cl]0/[1] at fixed conver-
sion was constructed (Fig. 6). Due to the linearity of the plot,
the average number of the initiating sites in 1 was determined
from the slope of the curve (0.788).32 Thus 1 has on average
Table 3 Effect of catalyst concentration, temperature and solvent on polymer-
ization kinetics of ε-CL using catalyst 1a
Entry [CL]0/[1] Conversionb (%) Kobs (h−1
)
Mwc
PDIc IEd
1
2
3
4
5
6
7
50
75
125
150
50e
50f
50g
90
85
92
94
99
99
80
0.063
0.052
0.037
0.010
0.189
0.026
0.022
4610 2.82 0.90
8008 2.96 1.10
10 696 3.23 0.81
14 568 3.25 0.91
2407 2.00 0.43
4193 2.42 0.74
3119 2.52 0.68
a Reaction conditions, [CL]0, 0.01 mol, temperature, 110 °C, bulk
polymerization. b Maximum conversions achieved. c Molecular-weight
average and polydispersity index (PDI) determined by GPC relative to
polystyrene standards. d Initiator efficiency (IE) = Mwexp/Mwcalc where
Fig. 6 Plot of degree of polymerization (Xn) of CL vs. [CL]0/[1] at fixed conver-
sion for the determination of number of active sties in 1. The inverse of the
slope 0.788 gives an average of 1.3 active sites.
Mwcalc = Mw(monomer) × [CL]0/[I] × [PCL/[CL]0 + Mw(chain-end
.
groups)
e Solvent, methanol. f Solvent, toluene. g Temperature, 90 °C.
10740 | Dalton Trans., 2013, 42, 10735–10745
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
Fig. 8 Plot of experimental and calculated Mw of PCL vs. % conversion for the
bulk polymerization of ε-CL by 3. [CL]0 = 0.01 mol, 110 °C, [3] = 0.0001 mol. The
empty rectangles show the theoretical Mw calculated from 1H NMR spectra
while the block rectangles show the experimental Mw determined by SEC.
Fig. 7 First order kinetic plots of ln[CL]0/[CL]t vs. time of the first and second
cycle experiments for 3, at 110 °C, [CL]0, 0.01 mol, [CL]0/[I] = 100. Equivalent
amount of ε-CL monomer was added in the second cycle without adding the
catalyst.
the use of toluene may reduce the concentrations of the reac-
1.3 (out of a possible two) active initiating sites per complex. tants hence lower rates of collisions of the molecules as com-
This data is in good agreement with the first-order dependency pared to the bulk reactions.
of the polymerization kinetics on 1 (Fig. 5). The findings that 1
Molecular weight and molecular weight distribution of
polycaprolactone (PCL)
does not utilize both the possible actives sites (two M–Oacetate
bonds) is in agreement with several literature reports.32b,c
The molecular weight and molecular weight distribution of
the polymers were determined by GPC and compared to the
theoretical values obtained from H NMR calculations (Tables
Stability of the initiators
1
To provide insight into the stability of these systems, a sequen-
tial two-stage polymerization of ε-CL was performed using 1
and 3 (Table 2, Fig. 7). Thus the first cycle ([CL]0/[I] = 100) was
allowed to proceed to completion (99%) and another 100 equi-
valent of ε-CL was added without adding the initiator ([CL]0/[I] =
200). For 3, rate constants of 0.096 h−1 and 0.085 h−1 were
reported in the first and second cycles respectively (Fig. 7).
This translates to 12% drop in the catalytic activity of 3 in the
second cycles and therefore confirms its relative stability.
2 and 3). Generally, low to moderate molecular weight poly-
mers between 1982 g mol−1 to 14 568 g mol−1 were obtained.
Consistent with living polymerization behaviour, molecular
weights increased with percentage conversion (Fig. 8). For
instance, molecular weights of 2121 g mol−1 and 4726 g mol−1
were obtained at 33% and 99% conversions respectively for 1.
The living polymerization nature of 1 was further augmented
by the observed increase in molecular weights with increase in
[CL]0/[1]. For example, an increase in the [CL]0/[1] from 50 to
125 resulted in a concomitant increase in molecular weight
from 4610 g mol−1 to 10 696 g mol−1 respectively (Table 3,
Effect of temperature and solvent on ε-CL polymerization
kinetics
The effect of temperature on the polymerization kinetics of entries 1–3). The highest initiator efficiency of 1.1 (110%) was
ε-CL by 1 was probed by comparing the activities at 60 °C, 90 °C obtained at [CL]0/[1] of 125. Despite the linear dependency of
and 110 °C at [CL]0/[1] of 50 (Table 3). At 60 °C, 1 showed very molecular weight on ε-CL conversion, the experimental mole-
low activity managing a paltry 25% conversion after 48 h. In cular weights were significantly lower than the theoretically
addition, the reactions were characterized by longer induction calculated values (Fig. 8). This was more evident at higher con-
periods leading to non-linear plots, hence the rate constant at versions where lower initiator efficiencies of 0.42 (42%) were
60 °C could not be evaluated. At 90 °C, the rate constant was reported. For example, at 95% conversion, the experimental
obtained as 0.022 h−1, three times lower than 0.063 h−1 molecular weight of 4726 g mol−1 was obtained, compared to
recorded at 110 °C. It is therefore apparent that 1–4 are only the theoretical value of 11, 294 g mol−1. In addition, the poly-
active at elevated temperatures. To understand the influence of mers exhibited relatively wide molecular weight distributions
solvent on the polymerization reactions, we compared the (2.00–3.48).
activities of 1 in bulk, methanol and toluene solvents. The kobs
The low initiator efficiencies of 1–4 were further confirmed
of 0.063 h−1 in the bulk polymerization was lower than by the two-stage polymerization reactions. As shown in
0.186 h−1 recorded in methanol solvent. This is consistent Table 2, addition of the monomer after completion of the first
with the formation of a metal-alkoxide (M-OEt), which is run resulted in a significant drop of molecular weight from
known to give active initiating groups.27b On the hand, use of 4726 g mol−1 (99%) to 3089 g mol−1 (45%); a clear indication
the non-coordinating toluene solvent resulted in a drop in kobs of growth of a new polymer chain. Indeed the maximum mole-
to 0.026 h−1. At this stage, it is unclear why reactions in cular weight of 4365 g mol−1 obtained in the second run is
toluene resulted in decreased activity. One hypothesis is that lower than 4726 g mol−1 reported in the first. This is in
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 10735–10745 | 10741

View Article Online
Paper
Dalton Transactions
contrast to the expected behaviour of living polymerization cata- the occurrence of transesterification reactions, ES-MS of the
lysts.26 For example, Silvernail et al. recorded an increase in polymers obtained after 4 h and 48 h were recorded (Fig. 9).
molecular weight of PCL from 4200 g mol−1 (run 1) to 10 400 The polymers showed m/z peaks corresponding to the formula
g mol−1 (run 2).26 These observations point to lack of controlled (nCL + 17) consistent with OH end groups. For example, the m/z
ε-CL polymerization by 1–4. A number of factors could be peak at 701 corresponds to {(114 × 6) + 17}. Significantly, the
responsible. One, the use of acetate ligand as the initiating spectrum of the crude products obtained after 48 h (Fig. 9b)
group has been reported to give low molecular weight and showed smaller peaks corresponding to molecular masses of
broad PDI in comparison to the alkoxide initiators.33 Indeed {(nCL + 1/2CL)} repeat units. As an illustration, the m/z peak at
polymers obtained in methanol solvent exhibited relatively 515 corresponds to {(4 × 114) + 57}. This confirms that inter-
narrow PDI of 2.00, indicating the presence of M-OCH3 initiat- molecular ester-exchange reactions do occur to some extent
ing group (Table 3, entry 5). The flexibility of the ligand back- especially at longer reaction times.31 The PDIs of 3.13 and 3.92
bone and varied coordination modes of the acetate ligands in reported after 4 h and 48 h respectively reinforces this
1–4 may also result in multiple active sites during the polymer- assumption.
ization process due to change in the molecular symmetry.34 As
The type of solvent used also affected the polymer weight
discussed vide supra, complex 1 (Fig. 1) can adopt either a tetra- and molecular weight distribution. A higher molecular weight
hedral or a trigonal bipyramidal geometry depending on the of 4193 g mol−1 was obtained in reactions performed in
coordination mode of the acetate ligand.
toluene compared to 2407 g mol−1 when methanol was used
A more possible route to the formation of low molecular as the solvent (Table 3, entries 5 and 6). This is consistent with
weight and broad molecular weight distributions of PCL the lower rate constant observed when toluene was used as the
obtained is via the transesterification reactions.30a,32,35 This solvent. As reported in literature, reduced catalytic activities
competes with ring open polymerization to give wide PDI as would decrease the number of polymer chains, thus increase
observed. The concerted low molecular weight and broad PDI molecular weight at a fixed conversion.36 This observations is
confirm the presence of both intramolecular (low molecular in good agreement with lower molecular weight of 3119 g
weights) and intermolecular (broad PDIs) transesterification mol−1 reported at 90 °C compared to 4610 g mol−1 at 110 °C
reactions. At higher retention times (Fig. S5–S7†) broad distri- (Table 3, entries 1 and 7). At higher temperatures, catalyst
butions were observed which could originate from the pres- activity was lower thus decreasing the number of active sites,
ence of both cyclic and linear oligomers. In order to confirm hence higher molecular weights of the resulted polymers.
Fig. 9 ES-MS of the crude PCL obtained from (a) catalyst 1; [CL]0/[1] = 100, time = 4 h; (b) catalyst 1, [CL]0/[1] = 100, time 48 h. The lager peaks corresponds to
(n(CL) + 17) indicating OH as the end group. The smaller peaks in (b) have masses corresponding to (n + 1/2CL) associated with repeating units from intermolecular
transesterification reactions.
10742 | Dalton Trans., 2013, 42, 10735–10745
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
Comparatively, initiators 3 and 4 bearing the bulky phenyl Electron paramagnetic resonance (EPR) spectra were recorded
groups on the pyrazolyl ring produced relatively higher mole- on a 9.1 GHz Bruker EMX/Premium-240 653 instrument.
cular weight PCL than the corresponding complexes 1 and 2,
[Zn(L1)Ac2] (1). To a solution of 2-(3,5-dimethylpyrazol-
containing the less sterically demanding methyl groups 1-ylmethyl)pyridine (L1) (0.15 g, 0.80 mmol) in methanol
(Table 2). For example at 96% conversions, molecular weights (10 mL) was added a solution of Zn(Ac)2·2H2O (0.15 g,
of 2928 g mol−1 and 4111 g mol−1 were obtained for 1 and 3 0.80 mmol) in methanol (10 mL) and the solution stirred for
respectively. This could be associated with enhanced chain 24 h. After the reaction period, the solvent was removed under
growth with increase in steric bulk.37 However, there was no vacuum to afford a white solid. Recrystallization of the crude
significant effect of the identity of the metal on polymer mole- product from
a dichloromethane–hexane solvent system
cular weight. For instance, at 99% conversion, 3 and 4 pro- afforded single-crystals suitable for X-ray analysis. Yield: 0.19 g
duced PCL with molecular weights of 4726 g mol−1 and 4652 g (63%). 1H NMR: (400 MHz, CDCl3): δ, 2.20 (s, 6H, CH3, Ac);
mol−1 respectively.
2.23 (s, 3H, CH3, pz); 2.54 (s, 3H, CH3, pz); 5.46 (s, 2H, CH2);
3
5.93 (s, 1H, pz); 5.93 (s, 1H, pz); 7.30 (t, 1H, py, JHH = 7.2 Hz);
3
3
7.43 (d, 1H, py, JHH = 7.2 Hz); 7.94 (t, 1H, py, JHH = 7.6 Hz);
9.07 (d, 1H, py, JHH = 7.6 Hz). 13C NMR (CDCl3): d; 12.13;
3
Summary and perspective
14.8; 23.6; 54.1; 108.2; 122.8; 125.1; 138.7; 141.1; 152.3;
In conclusion, we have demonstrated that steric factors control 153.4; 177.1. ES-MS: m/z (%); 309.93 [M+ − Ac, 100]; 251.94
bimetallic or monometallic formation of (pyrazol-1-ylmethyl)- [M+ − 2Ac, 20]. Anal. Calc. For C15H19N3O4Zn: C, 48.60;
pyridine Zn(II) and Cu(II) acetate complexes. The ligands adopt H, 5.17; N, 11.33. Found: C, 48.31; H, 5.07; N, 11.74.
monodentate and bidentate coordination modes in the mono-
[Cu(L1)Ac2] (2). To a solution of L1 (0.15 g, 0.80 mol) in
metallic and bimetallic complexes respectively. Complexes 1–4 methanol (10 mL) was added a solution of Cu(Ac)2·2H2O
form active and stable catalysts for the ring opening polymeriz- (0.16 g, 0.80 mmol) in methanol (10 mL) and the blue solution
ation of ε-CL. The kinetics of the polymerization are pseudo- was stirred for 24 h. Removal of solvent under vacuum gave a
first order in both monomer. Initiator 1 showed first order deep blue solid material. Recrystallization of the crude product
dependency on the polymerization reactions and utilizes only from a dichloromethane–hexane solvent system afforded com-
one of the two possible active sites as the initiating group. The pound 2 as an analytically pure solid. Yield: 0.19 g (62%).
non-rigid nature of the catalyst structure affected the living ES-MS: m/z (%); 308.93 [M+ − Ac, 100]; 249.94 [M+ − 2Ac, 20].
nature and control of polymer molecular weight of 1–4. By Anal. Calc. For C15H19N3O4Cu: C, 48.84; H, 5.19; N, 11.39.
changing catalyst concentrations, type of solvent and tempera- Found: C, 48.61; H, 5.44; N, 11.18.
ture of the reactions, the polymer properties could be regu-
[Zn(L2)Ac2] (3). This compound was prepared according to
lated. From these results, we believe that through careful the procedure describe for 1 using 2-(3,5-diphenylpyrazol-
ligand design and choice of the initiating group, more efficient 1-ylmethyl)pyridine (L2) (0.10 g, 0.32 mmol) and Zn(Ac)2·2H2O
1
catalyst systems with better control of polymer microstructure (0.06 g, 0.32 mmol). Yield: 0.13 g (81%). H NMR: (400 MHz,
could be developed.
CDCl3): δ, 2.20 (s, 6H, CH3, Ac); 5.59 (s, 2H, CH2); 6.71 (s, 1H,
pz); 5.93 (s, 1H, pz); 7.05 (t, 1H, py, 3JHH = 7.6 Hz); 7.40 (d, 1H,
3
py, JHH = 7.2 Hz); 7.44 (m, 4H, ph); 7.86 (m, 2H, ph); 7.64 (m,
3
3
4H, ph); 7.70 (t, 1H, py, JHH = 7.6 Hz); 8.63 (d, 1H, py, JHH
=
Experimental section
General materials and methods
7.6 Hz). 13C NMR (CDCl3): d; 24.4; 58.2; 109.2; 122.8; 124.1;
125.6; 138.9; 139.1; 142.1; 151.9; 153.8; 178.3. ES-MS: m/z (%);
All air sensitive manipulations were performed under argon 433.88 [M+ − Ac, 100]; 373.88 [M+ − 2Ac, 20]. Anal. Calc. For
using standard Schlenk line techniques. Compounds 2-(3,5- C25H23N3O4Zn: C, 60.68; H, 4.68; N, 8.49. Found: C, 60.52; H,
dimethylpyrazol-1-ylmethyl)pyridine (L1) and 2-(3,5-diphenyl- 4.85; N, 8.78.
pyrazol-1-ylmethyl)pyridine (L2) were prepared following litera-
[Cu2(L2)2Ac4] (4). This complex was synthesized following
ture procedures.38 Toluene and hexane solvents were distilled the procedure adopted for compound 2 using L2 (0.10 g,
and dried from sodium–benzophenone mixture while dichloro- 0.32 mmol) and Cu(Ac)2·2H2O (0.06 g, 0.32 mmol). Recrystalli-
methane was distilled from phosphorous pentaoxide. zation of the crude product from a dichloromethane–hexane
Zn(Ac)2·2H2O, Cu(Ac)2·2H2O and other chemicals were obtained mixture afforded single-crystals suitable for X-ray analysis.
from Sigma-Aldrich and used as received. The monomer ε-CL Yield: 0.10 g (62%). ES-MS m/z (%); 308.93 [M+ − Ac, 100];
was purchased from Sigma-Aldrich, vacuum distilled and 249.94 [M+ − 2Ac, 20]. ES-MS; m/z (%); 432.88 [M+ − Ac, 100];
stored under inert conditions prior to use. NMR spectra were 372.88 [M+ − 2Ac, 20]. Anal. Calc. For C50H46N6O8Cu2: C,
recorded on a Bruker instrument at room temperature in 60.90; H, 4.70; N, 8.52. Found: C, 60.62; H, 4.75; N, 8.67.
CDCl3 (1H at 400 and 13C at 100 MHz). Chemical shifts are
X-ray crystallography
reported in δ (ppm) and coupling constants are measured in
Hertz (Hz). Elemental analyses were recorded on Flash 2000 Crystal evaluation and data collection of complex 1 and 4 were
thermoscientific analyser while mass spectrometry was performed on a Bruker APEXII Duo CCD diffractometer and
recorded on a micro-mass LCT premier mass spectrometer. the reflections were successfully indexed by an automated
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 10735–10745 | 10743

View Article Online
Paper
Dalton Transactions
indexing routine built in the APEXII program suite.39 A suc-
cessful solution by the direct methods of SHELXS97 and
WinGX32 provided all non-hydrogen atoms from the E-map.
All non-hydrogen atoms were refined with anisotropic displa-
cement coefficients.40 Detailed information are given as ESI.†
References
1 (a) F. Majoumo-Mbe, E. Smolensky, P. Lonnecke,
D. Shpasser, M. S. Eisen and E. Hey-Hawkins, J. Mol.
Catal. A: Chem., 2005, 240, 91; (b) A. P. Gupta and V. Kumar,
Eur. Polym. J., 2007, 43, 4053; (c) M. Vivas and J. Contreras,
Eur. Polym. J., 2003, 39, 43.
General procedure for bulk polymerization of ε-caprolactone
2 (a) L. Azor, C. Bailly, L. Brelot, M. Henry, P. Mobian and
S. Dagorne, Inorg. Chem., 2012, 51, 10876; (b) H. Tian,
Z. Tang, X. Zhuang, X. Chen and X. Jing, Prog. Polym. Sci.,
2012, 37, 237; (c) M. G. Davidson, C. T. O’Hara,
M. D. Jones, C. G. Keir, M. F. Mahon and G. Kociok-Kohn,
Inorg. Chem., 2007, 46, 7686.
3 B. J. O’Keefe, L. E. Breyfogle, M. A. Hillmyer and
W. B. Tolman, J. Am. Chem. Soc., 2002, 124, 4384.
4 (a) J. Cayuela, V. Bounor-Legaré, P. Cassagnau and
A. Michel, Macromolecules, 2006, 39, 1338; (b) J. Wu,
T.-L. Yu, C.-T. Chen and C.-C. Lin, Coord. Chem. Rev., 2006,
250, 602; (c) B. J. O’Keefe, M. A. Hillmyer and W. B. Tolman,
J. Chem. Soc., Dalton Trans., 2001, 2215; (d) K. M. Stridsberg,
M. Ryner and A.-C. Albertsson, Adv. Polym. Sci., 2002, 157, 41.
5 (a) H. Du, X. Pang, H. Yu, X. Zhuang, X. Chen, D. Cui,
X. Wang and X. Jing, Macromolecules, 2007, 40, 1904;
(b) C. Jérôme and P. Lecomte, Adv. Drug Delivery Rev., 2008,
60, 1056.
6 (a) L. E. Breyfogle, C. K. Williams, V. G. Young,
M. A. Hillmayer and W. B. Tolman, Dalton Trans., 2006,
928.
7 A. C. Albertsson and I. K. Varma, Adv. Polym. Sci., 2002,
157, 1.
8 (a) M. Oshimura, T. Tang and A. Takasu, J. Polym. Sci., Part
A: Polym. Chem., 2001, 49, 1210; (b) Y.-C. Liu, B.-T. Ko and
C.-C. Lin, Macromolecules, 2001, 34, 6196.
Bulk polymerization reactions were performed by introducing
an appropriate amount of the complex, depending on the
[CL]0/[I] ratio, in a Schlenk tube equipped with a magnetic
stirrer under argon. The monomer, ε-CL (1.14 g, 0.01 mol) was
then added via a gas tight syringe and the temperature set at
110 °C before the reactions were initiated. Kinetic experiments
were carried out by withdrawing samples at regular intervals
(approx. 0.2 mL) using a syringe and quickly quenched by
rapid cooling into NMR tubes containing CDCl3 solvent. The
1
quenched samples were analyzed by H NMR spectroscopy for
determination of polymerization of ε-CL to PCL. The percen-
tage conversion of [PCL]/[CL]0 × 100, where [CL]0 is the initial
concentration of monomer and [PCL] is the concentration of
the polymer at time t, was evaluated by integration of the
peaks for CL (4.2 ppm, OCH2 signal) and PCL (4.0 ppm, OCH2
signal) according to the equation [PCL]/[CL]0 = I4.0/(I4.2 + I4.0
)
where I4.2 is the intensity of the CL monomer signal at 4.2 ppm,
and I4.0 is the intensity of the PCL signal at 4.0 ppm for the
OCH2 protons. The observed rate constants, Kobs, were extracted
from the slopes of the lines of best-fit to the plots of ln[CL]0/
[CL]t vs. time. The polymers were purified by dissolving the
crude products in CH2Cl2 followed by addition of cold metha-
nol. A white precipitate was formed, which was isolated by filt-
ration and dried to constant weight prior to analyses by SEC.
Polymer characterization by size exclusion chromatography (SEC)
9 (a) D. Thomas, P. Arndt, N. Peulecke, A. Spannenberg,
R. Kempe and U. Rosenthal, Eur. J. Inorg. Chem., 1998,
1351; (b) P. Arndt, D. Thomas and U. Rosenthal, Tetra-
hedron Lett., 1997, 38, 5467; (c) P. Arndt, A. Spannenberg,
W. Baumann and U. Rosenthal, Eur. J. Inorg. Chem., 2001,
2885; (d) A. Asandei and G. Saha, Macromol. Rapid
Commun., 2005, 26, 626; (e) A. Touris, K. Kostakis,
S. Mourmouris, V. Kotzabasakis, M. Pitsikalis and N.
Hadjichristidis, Macromolecules, 2008, 41, 2426.
10 (a) S. K. Russell, C. L. Gamble, K. J. Gibbins, K. C. S. Juhl,
W. S. Mitchell, A. J. Tumas and G. E. Hofmeister, Macro-
molecules, 2005, 38, 10336; (b) Y. Kim and J. G. Verkade,
Macromol. Rapid Commun., 2002, 23, 917; (c) Y. Kim and
J. G. Verkade, Macromol. Symp., 2005, 224, 105; (d) Y. Kim,
G. K. Jnaneshwara and J. G. Verkade, Inorg. Chem., 2003,
42, 1437; (e) Y. Kim and J. G. Verkade, Organometallics,
2002, 21, 2395.
Polymer samples were dissolved in BHT stabilized THF (2 mg
mL−1). The sample solutions were filtered via syringe through
0.45 μm nylon filters before analyses. The SEC instrument con-
sists of a Waters 1515 isocratic HPLC pump, a Waters 717plus
auto-sampler, Waters 600E system controller (run by Breeze
Version 3.30 SPA) and a Waters in-line Degasser AF. A Waters
2414 differential refractometer was used at 30 °C in series with
a Waters 2487 dual wavelength absorbance UV/Vis detector
operating at variable wavelengths. Tetrahydrofuran (THF,
HPLC grade, stabilized with 0.125% BHT) was used as eluent
at flow rates of 1 mL min−1. The column oven was kept at
30 °C and the injection volume was 100 μL. Two PLgel
(Polymer Laboratories) 5 μm Mixed-C (300 × 7.5 mm) columns
and a pre-column (PLgel 5 μm Guard, 50 × 7.5 mm) were used.
Calibration was done using narrow polystyrene standards
ranging from 580 to 2 × 106 g mol−1. All molecular weights
were reported as polystyrene equivalents.
11 (a) L. Azor, C. Bailly, L. Brelot, M. Henry, P. Mobian and
S. Dagorne, Inorg. Chem., 2012, 51, 10876; (b) F. Gornshtein,
M. Kapon, M. Botoshansky and M. Eisen, Organometallics,
2007, 26, 497.
Acknowledgements
The authors would like to thank University of KwaZulu-Natal 12 A. J. Chmura, M. G. Davidson, C. J. Frankis, M. D. Jones
and Third World Academy of Sciences for financial support.
and M. D. Lunn, Chem. Commun., 2008, 1293.
10744 | Dalton Trans., 2013, 42, 10735–10745
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
13 (a) C.-Y. Sung, C.-Y. Li, J.-K. Su, T.-Y. Chen, C.-H. Lin and 25 M. Labet and W. Thielemans, Chem. Soc. Rev., 2009, 38,
B.-T. Ko, Dalton Trans., 2012, 41, 953; (b) I. dos Santos 3484.
Vieira and S. Herres-Pawlis, Eur. J. Inorg. Chem., 2012, 765; 26 C. M. Silvernail, L. Y. Yao, L. M. R. Hill, M. A. Hillmyer and
(c) T.-P. A. Cao, A. Buchard, X. F. Le Goff, A. Auffrant and W. B. Tolman, Inorg. Chem., 2007, 46, 6565.
C. K. Williams, Inorg. Chem., 2012, 51, 2157; (d) V. Poirier, 27 (a) C. K. Williams, L. E. Breyfogle, S. Kyung Choi, W. Nam,
T. Roisnel, J.-F. Carpentier and Y. Sarazin, Dalton Trans.,
2011, 40, 523; (e) L. F. Sanchez-Barba, A. Garces,
J. Fernandez-Baeza, A. Otero, C. Alonso-Moreno, A. Lara-
V. G. Young Jr., M. A. Hillmyer and W. B. Tolman, J. Am.
Chem. Soc., 2003, 125, 11350; (b) H.-Y. Chen, B.-H. Huang
and C.-C. Lin, Macromolecule, 2005, 38, 5400.
Sanchez and A. M. Rodriguez, Organometallics, 2011, 30, 28 Z. Zhong, P. J. Dijkstra and J. Feijen, J. Am. Chem. Soc.,
2775; (f) R. H. Platel, A. J. P. White and C. K. Williams, 2003, 125, 11298.
Inorg. Chem., 2011, 50, 7718; (g) J. Wu, Te.-L. Yu, C.-T. Chen 29 A. Kowalski, J. Libsiszowski and S. Penzcze, Macro-
and C.-C. Lin, Coord. Chem. Rev., 2006, 250, 602 and refer-
ences therein.
14 (a) R. R. Gowda and D. J. Chakraborty, Mol. Catal. A: Chem.,
2010, 333, 167; (b) M. Shen, W. Huang, W. Zhang, X. Hao,
W.-H. Sun and C. Redshaw, Dalton Trans., 2010, 39, 9912;
(c) A. Arbaoui, C. Redshaw, M. R. J. Elsegood, V. R. Wright,
molecules, 2001, 33, 1964.
30 (a) P. Dubois, C. Jacobs, R. Jerome and P. Teyssie, Macro-
molecules, 1991, 24, 2266; (b) T. Ouhadi, C. Stevens and
P. Teyssie, Makromol. Chem., Suppl., 1975, 1, 191;
(c) V. J. Shiner, D. Whittaker and V. P. Fernandez, J. Am.
Chem. Soc., 1963, 85, 2318.
A. Yoshizawa and T. Yamato, Chem.–Asian J., 2010, 5, 621; 31 C. K. Williams, L. E. Breyfogle, S. K. Cho, W. Nam,
(d) N. Nomura, A. Akita, R. Ishii and M. Mizuno, J. Am.
Chem. Soc., 2010, 132, 1750.
V. G. Young, M. A. Hillmyer and W. B. Tolman, J. Am.
Chem. Soc., 2003, 125, 11350.
15 (a) S. O. Ojwach, B. Letitia, I. A. Guzei, J. Darkwa and 32 (a) W. M. Stevels, M. J. Ankone, P. J. Dijkstra and J. Feijen,
S. F. Mapolie, Organometallics, 2009, 28, 2127; (b) S. O. Ojwach,
I. A. Guzei and J. Darkwa, J. Organomet. Chem., 2009, 694,
1393.
Macromolecules, 1996, 29, 6132; (b) B. M. Chamberlain,
B. A. Jazdzewski, M. Pink, M. A. Hillmyer and
W. B. Tolman, Macromolecules, 2000, 33, 3970;
(c) S. J. MClain, T. M. Ford and N. E. Drysdale, Polym. Prepr.
(Am. Chem. Soc., Div. Polym. Chem.), 1992, 33, 463.
16 G. Parkin, Chem. Rev., 2004, 104, 699.
17 (a) J. Börner, U. Flörke, K. Huber, A. Döring, D. Kuckling
and S. Herres-Pawlis, Chem.–Eur. J., 2009, 15, 2362; 33 B. M. Chamberlain, B. A. Jazdzewski, M. Cheng,
(b) S. O. Ojwach, G. S. Nyamato, B. Omondi, J. Darkwa and
A. O. Okoth, J. Coord. Chem., 2012, 65, 298.
D. R. Moore, T. M. Ovitt, E. B. Lobkovsky and G. W. Coates,
J. Am. Chem. Soc., 2001, 123, 3229.
18 F. A. Allen, Acta Crystallogr., 2002, B58, 380.
19 (a) M. Casarin, C. Corvaja, C. D. Nicola, D. Falcomer,
L. Franco, M. Monari, L. Pandolfo, C. Pettinari and
34 (a) P. J. Shapiro, W. P. Schaefer, J. A. Labinger and
J. E. Bercaw, J. Am. Chem. Soc., 1994, 116, 4623;
(b) R. M. C. Waymouth, Science, 1995, 267, 217.
F. Piccinelli, Inorg. Chem., 2005, 44, 6265; (b) N. Judas, Acta 35 (a) H. R. Kricheldorf, J. M. Jonte and M. Berl, Makromol.
Crystallogr., Sect. E: Struct. Rep. Online, 2005, 61, 2217.
20 (a) C. Harding, V. McKee and J. Nelson, J. Am. Chem. Soc.,
Chem., Suppl., 1985, 12, 25; (b) H. R. Kricheldorf, T. Mang
and J. M. Jonte, Macromolecules, 1984, 17, 2174.
1991, 113, 9684; (b) J. Casanova, G. Alzuet, J. Latorre and 36 (a) T. M. Ovitt and G. W. Coates, J. Polym. Sci., Part A:
J. Borra’s, Inorg. Chem., 1997, 36, 2052; (c) R. Sarma,
A. K. Boudalis and J. B. Baruah, Inorg. Chim. Acta, 2010,
363, 2279.
Polym. Chem., 2000, 38, 4686; (b) G. Odian, Principles of
Polymerization, John Wiley
pp. 535–540.
& Sons, New York, 1991,
21 (a) S. J. Brown, X. Tao, D. W. Stephan and P. K. Mascharak, 37 (a) D. J. B. Temple, Organometallics, 1998, 17, 2290;
Inorg. Chem., 1986, 25, 3377; (b) C. P. Pradeep,
P. S. Zacharias and S. K. Das, J. Chem. Sci., 2005, 117, 133;
(b) S. A. Svejda, L. K. Jonhson and M. Brookhart, J. Am.
Chem. Soc., 1999, 121, 10634.
(c) A. Ozarowski, I. B. Szymanska, T. Muziol and 38 (a) P. J. Steel, A. A. Watson and D. A. House, Inorg. Chim.
J. Jezierska, J. Am. Chem. Soc., 2009, 131, 10279;
(d) A. Ozarowski, Inorg. Chem., 2008, 47, 9760.
Acta, 1987, 130, 167; (b) S. O. Ojwach, I. A. Guzei, J. Darkwa
and S. F. Mapolie, Polyhedron, 2007, 26, 851.
22 D. L. Reger, A. Debreczeni and M. D. Smith, Inorg. Chem., 39 Bruker-AXS, Madison, Wisconsin, USA, 2009.
2012, 51, 1068.
40 (a) G. M. Sheldrick, Acta Crystallogr., 2008, A64, 112;
(b) SADABS Area-Detector Absorption Correction, Siemens
Industrial Automation, Inc., Madison, WI, 1996; (c) SAINT
Area-Detector Integration Software, Siemens Industrial Auto-
mation, Inc., Madison, WI, 1995.
23 S. J. Brown, X. Tao, T. A. Wark, D. W. Stephan and
P. K. Mascharak, Inorg. Chem., 1988, 27, 1581.
24 G. F. Kokoszka, M. Linzer and G. Gordon, Inorg. Chem.,
1968, 7, 1730.
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 10735–10745 | 10745