The use of Cu and Zn salicylaldimine complexes as catalyst precursors in ring opening polymerization of lactides: Ligand effects on polymer characteristics

Article abstract of DOI:10.1002/aoc.1728

A range of monomeric tetra-coordinate copper (II) and zinc (II) complexes based on N,O-bidentate salicylaldimine Schiff base ligands has been synthesized and characterized using various spectroscopic techniques. These complexes were then evaluated as init

Full text of DOI:10.1002/aoc.1728

Full Paper
Received: 11 April 2010
Revised: 11 August 2010
Accepted: 26 August 2010
Published online in Wiley Online Library: 11 November 2010
(wileyonlinelibrary.com) DOI 10.1002/aoc.1728
The use of Cu and Zn salicylaldimine complexes
as catalyst precursors in ring opening
polymerization of lactides: ligand effects
on polymer characteristics
Suraj Bhunoraa, Jane Mugob, Archana Bhaw-Luximona∗, Selwyn Mapolieb∗,
Juanita Van Wykc, James Darkwac and Ebbe Nordlanderd
A range of monomeric tetra-coordinate copper (II) and zinc (II) complexes based on N,O-bidentate salicylaldimine Schiff base
ligands has been synthesized and characterized using various spectroscopic techniques. These complexes were then evaluated
as initiators in ring-opening polymerization of lactides at both 70 ◦C and 110 ◦C. The effect of structural changes in the
complexes on the ability of these compounds to initiate lactide polymerization as well as the impact on the chemical and
physical characteristics of the polymers obtained indicate that the coordination geometry of the metal complex, M–O bond
length and substituents on the Schiff base ligand all play a role in the catalyst activity. Electronic factors were dominant in the
case of the copper complexes while steric factors prevailed in the case of Zn initiators. Both the Zn and Cu complexes exhibit
c
characteristics of living ring opening polymerization. Copyright ꢀ 2010 John Wiley & Sons, Ltd.
Supporting information may be found in the online version of this article.
Keywords: salicylaldimine complexes; copper Schiff base complexes; zinc Schiff base complexes; lactide polymerization; ring-opening
polymerization
Introduction
complex containing bulky substituents on the phenoxy and imino
rings (Fig. 1).[5]
Zhang et al. investigated the polymerization of ε-caprolactone
and DL-lactide using zinc and aluminum complexes containing
pyrazole substituents.[6] They reported that the zinc ethyl
complexes (Fig. 2A) displayed relatively high catalytic activity
in ε-caprolactone polymerization, giving high molecular weight
polymers (number average molecular weight, Mn = 58 000, 100%,
[monomer] : [initiator] = 200). It was, however, found that the
process proceeds via a nonliving mechanism, thus leading to poor
polydispersities. Onlycomplex2d(Fig. 2A)wasactiveforDL-lactide
polymerization but unfortunately it yielded only atactic polymers.
Hung et al. reported L-lactide polymerization using dimeric zinc
complexes with tridentate N,N,O Schiff base ligands (Fig. 3).[7]
These mediated controlled living polymerization.
Their biocompatibility and their degradation into nontoxic
components (water and carbon dioxide via the Krebs cycle)
make polylactides (PLAs) very useful for biomedical applications
such as controlled drug delivery systems.[1] Owing to their
mechanical properties, PLAs are widely used in surgery as sutures,
orthopedic applications, tissue engineering and biodegradable
internal fixation devices for repair of fractures to small bones and
joints.[2]
The most important and general way to prepare high molecular
weight PLAs is through ring opening polymerization (ROP). A
number of different metal initiators and catalysts have been used
in the ROP of lactides, including compounds of aluminum, lead,
tin, zinc and bismuth. Currently almost all commercial PLAs are
prepared using FDA-approved stannous octanoate as mediator.
Recently, catalysts based on Ca, Mg, Fe and Zn have received
increased attention because these elements can be metabolized
in the body.[3] Zinc lactate which shows catalytic characteristics
similar to stannous octanoate, has been proposed as a less toxic
catalyst substitute for PLA production.[4]
∗
Correspondence to: Archana Bhaw-Luximon, Department of Chemistry, Uni-
versity of Mauritius. E-mail: a.luximon@uom.ac.mu
Selwyn Mapolie, Department of Chemistry, University of Stellenbosch, South
Africa. E-mail: smapolie@sun.ac.za
´
a
Department of Chemistry, Faculty of Science, University of Mauritius, Reduit,
Mauritius
Our current interest is focused on zinc and copper metal com-
plexes of salicylaldimine ligands which remain largely unexplored
as catalysts for polylactides synthesis. Only a few reports concern-
ing such complexes as lactide polymerization catalysts exist in the
literature.[5–8]
b
c
Department of Chemistry, University of Stellenbosch, South Africa
Department of Chemistry, University of Johannesburg, South Africa
Chisholm et al., for example, have briefly described the polymer-
ization of DL-lactide with a three coordinate zinc phenoxy-imine
d
Inorganic Chemistry Research Group, Chemical Physics, Centre for Chemistry
and Chemical Engineering, Lund University, Sweden
c
Appl. Organometal. Chem. 2011, 25, 133–145
Copyright ꢀ 2010 John Wiley & Sons, Ltd.

S. Bhunora et al.
L-lactide under solvent-free conditions but only at fairly elevated
temperatures (160–180 ◦C), producing poly-L-Lactides (PLLAs) of
moderate molecular weights (Mn = 8000–11 000, [M] : [I] = 50)
and having narrow molecular weight distributions (polydispersity
index, PDI = 1.3–1.4 at 70–80% conversion).[8]
The relative success of phenoxy-imines in various polymeriza-
tion reactions is presumably due to the scope for suitable tuning of
the steric bulk and the electronics of the ancillary ligand and also
due to their easy synthetic accessibility, as the phenoxy-imines
are generally prepared via Schiff base condensation reactions.
Despite extensive utility of phenoxy-imines in many important
chemical transformations, their application in ROP of DL-lactide
largely remains unexplored.
Because of the relative paucity of reports on the use of zinc
and especially copper complexes as ring-opening polymerization
initiators, we embarked on a study of these complexes in such
processes. Here we report the synthesis and use of several Cu
and Zn phenoxy-imine complexes in DL-lactide polymerization.
These complexes offer the advantage of being stable in air,
unlike aluminum complexes, and are also easier to prepare. In
addition, the steric and electronic effects of substituents on the
phenoxy and imino rings on polymerization activity were also
investigated. A kinetic study of the ROP of DL-lactide was carried
out and correlated with the nature of the ligands complexed to
the metal atom, metal–oxygen bond length and geometry of the
complexes. In addition, an attempt was made to understand the
effect of the salicylaldimine ligands on stereoselectivity of the
DL-lactide polymerization and its influence on Tg of the resulting
polymers. The characteristics and morphology of the resulting
PLAs are also discussed.
Figure 1. Three-coordinated zinc phenoxide complex used as lactide
polymerization catalyst by Chisholm et al.[5]
.
Experimental
General Procedures
Ligands and metal complexes were synthesized using standard
Schlenk techniques under nitrogen using a dual vacuum/nitrogen
Schlenk line. The NMR spectra were recorded on a Varian Gemini
2000 spectrometer (1H at 200 MHz, 13C at 50.3 MHz) at room
temperature using tetramethylsilane as an internal standard.
The chemical shifts are reported in δ (ppm) and referenced
relative to residual proton signals for the NMR solvent. Infrared
spectra were recorded on a Perkin Elmer Paragon 1000 PC FT-
IR spectrophotometer as KBr pellets for solids or between NaCl
plates for oils. ESI-MS spectra were obtained on a Waters API Q-
TOF Ultima spectrometer calibrated with NaF. UV–Vis spectra
were recorded on a GBC UV-Vis S920 spectrophotometer as
dichloromethane solutions. Microanalyses were performed at
the University of Cape Town’s micro analytical laboratory. Size
exclusion chromatography was performed at the University of
Mauritius using a Polymer Standards Service (PSS) apparatus with
Figure 2. (A) Zinc ethyl; (B) zinc diphenolato complexes used for ROP of
lactide by Zhang et al.[6]
.
Copper complexes have been less frequently employed in
ring opening polymerization of DL-lactide. A recent report by
John et al. showed that phenoxy-ketimine copper complexes
of the type shown in Fig. 4 efficiently catalyzed the ROP of
Figure 3. NNO-tridentate Schiff’s base Zn complex used by Hung et al.[7]
.
c
wileyonlinelibrary.com/journal/aoc
Copyright ꢀ 2010 John Wiley & Sons, Ltd.
Appl. Organometal. Chem. 2011, 25, 133–145

Cu and Zn salicylaldimine complexes as catalyst precursors
Figure 4. Copper complexes used for ROP of L-lactide by John et al.[8]
.
Figure 5. Chemical structure of Cu and Zn complexes synthesized.
arefractiveindexdetector. Calibrationwasdoneusingpolystyrene
standards. THF was used as eluent at a flow rate of 1.0 ml min−1
photographs with a rotation angle of 2◦ were collected and
processed using the CrystalClear software package. Empirical cor-
rections were applied for the effects of absorption using the
REQAB program under CrystalClear. The structures were solved
by direct methods[14] and refined using full-matrix least-squares
calculations on F2 (SHELXL-97)[15] on all reflections, both programs
operating under the WinGX program package.[16]
Anisotropic thermal displacement parameters were refined for
all nonhydrogen atoms; the hydrogen atoms were included as a
riding contribution except for those of the amine groups, which
were located from difference maps and refined with fixed isotropic
thermal parameters. Structural illustrations have been drawn with
ORTEP-3 for Windows under WinGX.[17]
.
Scanning electron microscopy (SEM) samples were prepared for
analysisaccordingtothefollowingprotocol.Athinlayerofpolymer
was spread on a metal mount. The samples were then coated with
gold for 90 s in a 20 mV argon atmosphere on S150A sputter
coater. Analysis was done on a Leo 1430VP scanning electron
microscope. Differential scanning calorimetry was performed on
TA Instrument Q100 Differential Scanning Calorimeter. For DSC
analysis, each sample was weighed (7–8 mg). Tg and enthalpies
of fusion (ꢀHm) were measured. The DSC cell was purged with
nitrogen gas flow of 50 ml min−1. Experiments were performed
in aluminum hermetic pans using a heating and cooling rate of
5 ◦C min−1. Tm was determined by the first scan while Tg was
determined by the second scan.
Synthesis of Ligands
The salicylaldimine ligands were prepared via Schiff base conden-
sation of the appropriate 2-hydroxybenzaldehyde and aromatic
amine using a published procedure (Scheme 1).[9] The synthetic
procedure is described using the synthesis of HL1 as an example.
To a Schlenk tube containing salicylaldehyde (12 mmol) and
formic acid (0.5 ml) in methanol (15 ml) was added aniline
(16 mmol). The resulting orange-brown reaction mixture was
X-ray Crystallography
X-ray crystallography of some of the metal complexes were
performed on a Rigaku R-AXIS IIc diffractometer. Diffracted in-
tensities were measured at ambient temperature, using graphite-
monochromated Mo-Kα radiation (λ = 0.71073 Å) from a RU-H3R
rotating anode operated at 50 kV and 90 mA. Ninety oscillation
c
Appl. Organometal. Chem. 2011, 25, 133–145
Copyright ꢀ 2010 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/aoc

S. Bhunora et al.
Scheme 1. Synthesis of ligands (HL1 –HL7).
stirred at room temperature for approximately 15 h. During this
time a yellow solid precipitated from the solution and was isolated
by vacuum filtration, washed with cold methanol and dried under
vacuum. The salicylaldimine ligands were obtained as light yellow
to orange-yellow solids in yields of 65–93%. Full characterization
data for the ligands are given in Tables S1 and S2 in the Supporting
Information.
bis[N-(phenyl)-5-chloro-salicylaldiminato] copper(II), Cu(5)
Yield, 58%; m.p. >295 ◦C (dec); UV–vis. λmax (nm) CH2Cl2 soln: 274,
333, 372, 628, 657.
bis[N-(2,6-diisopropylphenyl)-5-chloro-salicylaldiminato] copper(II),
Cu(6)
Yield, 73%; m.p. >280 ◦C (dec); UV–vis. λmax (nm) CH2Cl2 soln: 276,
335, 375, 629, 662.
Synthesis of Copper Complexes Cu(1–6)
The general procedure for the preparation of the copper
complexes is given here. Copper acetate monohydrate (0.5 mmol)
and the appropriate Schiff base ligand (1 mmol) were placed
in a round-bottom flask, followed by methanol (20 ml). The
resulting reaction mixture was stirred under reflux for 4 h under
a nitrogen atmosphere. During this time a solid precipitated
from solution. The reaction mixture was cooled to 0 ◦C for
approximately 15 min and the solid isolated by vacuum filtration.
Complexes were recrystallized by slow diffusion of ethanol into
a concentrated dichloromethane solution of the complexes at
−4 ◦C. All complexes were obtained as dark green crystals.
bis[N-(2,6-diisopropylphenyl)-3,5-di-tert-butyl salicylaldiminato]
copper(II), Cu(7)
Yield, 51%; m.p. >260 ◦C (dec); UV–vis. λmax (nm) CH2Cl2 soln: 278,
326, 368, 620, 656.
Synthesis of bis[N-(2,6-diisopropylphenyl)-salicylaldiminato]
zinc(II), Zn(1)
ToaclearyellowsolutionofHL1 (1.0 g,3.6 mmol)inethanol(20 ml)
wasaddedzincacetate(0.39 g, 1.8 mmol)andthemixturerefluxed
for 2 h. During the reaction time, a yellow solid precipitated
from the solution. The solid was isolated by vacuum filtration.
The product was recrystallized using a mixture of CH2Cl2/EtOH
(1 : 2) at −4 ◦C, giving a cream-colored solid. Yield 0.55 g, 48%.
bis[N-(2,6-diisopropylphenyl)-salicylaldiminato]copper(II), Cu(1)
Yield, 62%; m.p. >300 ◦C (dec); UV–vis. λmax (nm) CH2Cl2 soln: 286,
328, 386, 630, 660.
FT-IR (cm−1, ATR): νC
= 1614 νC–O = 1277, ESI-MS [M +
N
H]+ = 626 m/z calculated for C38H44N2O2Zn. 1H-NMR (CDCl3
298 K) δ (in ppm) 1.19 [d, 12H, CH(CH3)2], 3.01 [m, 2H, CH(CH3)2],
6.98 (t, 2H, Ar); 7.09 (d, 2H,Ar); and 7.20 (m, 4H, Ar), 7.36 (m,
2H, Ar) 7.43 (m, 4H, Ar), 8.30 (s, 2H, CH N). 13C-NMR (CDCl3)
δ (in ppm), 177.01 (N CH), 173.39 (aromatic C–OH), 145.88
(aromatic C–N C), 137.01 (aromatic C2/6-iPr), 135.31 (C4/6 on
bis[N-(phenyl)-salicylaldiminato] copper(II), Cu(2)
Yield, 57%; m.p. >280 ◦C (dec); UV–vis. λmax (nm) CH2Cl2 soln: 277,
336, 376, 631, 661.
i
OH-containing aromatic ring), 126.31, (C3/5 on Pr containing
aromatic ring), 123.24, (C4- on iPr containing aromatic ring), 118.12
(C5OH-containingaromaticring),113.82(aromaticC–C N),27.44
[CH–(CH3)2], 22.21 [CH–(CH3)2]. Anal. calcd for C38H44N2O2Zn C,
72.91; H, 7.64; N, 4.47. Found: C, 72.47; H, 7.89; N, 4.45
bis[N-(phenyl)-3,5-di-tert-butyl salicylaldiminato] copper(II), Cu(3)
Yield, 73%; m.p. >290 (dec); UV–vis. λmax (nm) CH2Cl2 soln: 280,
320, 386, 632, 653.
bis[N-(2,6-diisopropylphenyl)-3-tert-butyl-salicylaldiminato]
copper(II), Cu(4)
Synthesis of bis[N-(2,6-diisopropylphenyl)-3-tert-butyl-sali-
cylaldiminato] zinc (II), Zn(4)
Yield, 62%; m.p. >270 (dec); UV–vis. λmax (nm) CH2Cl2 soln: 284,
To a clear yellow solution of HL4 (0.60 g, 1.78 mmol) in THF (10 ml)
328, 384, 633, 657.
was added NaH (0.045 g, 1.88 mmol). Immediately effervescence
c
wileyonlinelibrary.com/journal/aoc
Copyright ꢀ 2010 John Wiley & Sons, Ltd.
Appl. Organometal. Chem. 2011, 25, 133–145

Cu and Zn salicylaldimine complexes as catalyst precursors
wasobserved. Theyellowmixturewasstirredatroomtemperature
under nitrogen for 1 h. The reaction mixture was filtered under
nitrogen into another Schlenk tube and the solvent was removed
under vacuum. The residue (Na salt of ligand) was washed with
pentane (2 × 5 ml) and used without further purification in the
subsequent step. The sodium salt of the ligand and zinc acetate
(0.172 g, 0.78 mmol) were placed in a two-necked round-bottom
flask followed by ethanol (20 ml). The resulting yellow mixture
was stirred under nitrogen at room temperature for 4 h. During
this time, a yellow solid precipitated. The solid was filtered and
washedwithcoldmethanol.Recrystallizationwasperformedusing
a mixture of CH2Cl2/EtOH (1 : 2) at −4 ◦C. Yield 0.52 g, 89%. FT-IR
Spectral Characterization of Complexes
The copper complexes were characterized by IR and UV spec-
troscopy, elemental analysis and mass spectrometry (ESI). The zinc
complexes were characterized using similar techniques together
with 1H NMR spectroscopy. The main features of the IR spectra of
the copper complexes is the shift of theυ(C N) band from around
1615 cm−1 in the ligand to around 1602 cm−1 in the complexes.
Similar shifts are observed for the zinc complexes. In this case the
shift is from ∼1615 to ∼1598 cm−1. In all cases the complexation
is further confirmed by the disappearance of the υ(O–H) band
which normally occurs in the region of 2800–3000 cm−1. The
original υ(C–O) band of the phenoxy group in the free ligand
(1269–1279 cm−1) also shifts to higher frequencies on coordina-
tion to the metal.
Owing to the paramagnetic nature of Cu+2 complexes, we were
unable to study these complexes via NMR spectroscopy. The d10
Zn+2 complexes were, however, amenable to NMR studies, giving
clear 1H NMR spectra in CDCl3. The main features in the NMR
spectra of these complexes are the shift in the signal of the imine
proton to δ 8.30 ppm in Zn(1) and δ 8.31 ppm in Zn(4).
Complexes were also characterized using ESI mass spectrom-
etry. In all cases parent ions for the different complexes were
observed. The series of copper complexes showed similar frag-
mentation patterns. In the case of the copper complexes a typical
fragmentation pattern consisted of initial demetallation of the
complex producing the free ligand. The remaining fragmentation
pattern resembles that of the free ligand. The Zn complex Zn(1)
shows only a weak parent ion, with the remaining fragmentation
pattern being very similar to that of the free ligand. The ESI mass
spectrum of Zn(4) shows a fairly intense molecular ion at m/z 737.
There is very little fragmentation of the parent ion observed, which
seems to point to the relative stability of the complex under the
condition employed to carry out the mass spectral analysis.
The UV–vis spectra of the copper complexes show four bands
in the region 220–400 nm. The first two bands around 235 and
275 nm can be assigned to π –π∗ transitions while the third band
around 305 nm is due to a n–π∗ transition. A fourth band not
present in the ligand spectra can be assigned to a metal to ligand
charge (MLCT) transfer band which involves electrons in the n
orbital of the C N functionality. These spectra are similar to those
reported for analogous Cu+2 complexes. This seems to indicate
that we are dealing with distorted square planar geometries.
Full characterization data for the copper complexes are given in
Table 1.
(cm−1, KBr pellet): ν(C
= 1598, ν(C–O) = 1272. ESI-MS [M +
N)
H]+ = 739 m/z calculated for C46H60N2O2Zn.
1H-NMR (CDCl3 298 K) δ (in ppm) 1.19 [d, 12H, CH(CH3)2], 1.39
[s 18H, C(CH3)3], 3.03 [m, 2H, CH(CH3)2], 6.53 (2H, Ar), 6.94 (t, 6H,
Ar) 7.41 (m, 2H, Ar), 8.31 (s, 2H, CH N). 13C-NMR (CDCl3) δ (in
ppm), 175.92 (N CH), 172.65 (aromatic C–OH), 146.94, (aromatic
C–N C), 142.82, 141.90 (aromatic C2/6-iPr), 139.13, (aromatic
,
t
C3-tBu), 135.26, 132.78, (C4/6 on Bu containing aromatic ring),
127.12, (C3/5 on iPr containing aromatic ring), 124.69 (C4- on iPr
containing aromatic ring), 117.96 (C5 on tBu containing aromatic
ring), 114.85 (aromatic C-C N), 35.28 [C-(CH3)3], 28.46 [C-(CH3)3],
25.31 [CH-(CH3)2], 22.65 [CH-(CH3)2]. Anal. calcd for C46H60N2O2Zn
C, 74.8; H, 8.20; N, 3.80. Found: C, 75.27; H, 7.89; N, 3.45.
Polymerization
DL-Lactide (0.72 g, 0.005 mol) was dissolved in dioxane or
toluene (5 ml) in a Schlenk tube and the required amount of
catalyst/initiator was added. The reaction mixture was refluxed
at 70 or 110 ◦C (depending on the solvent used) under vacuum.
After the required reaction time, the solvent was removed and the
crude material was analyzed by 1H-NMR in CDCl3. The integration
values of the methine proton of the monomer (ICHmonomer ) and
that of polymer (ICHpolymer ) were used to calculate the percentage
conversion as shown in eqn (1).
ICH
polymer
% Conversion =
× 100
(1)
ICH
+ ICH
monomer
polymer
Polymer Characterization
Polymers were characterized using NMR spectroscopy, DSC anal-
ysis, scanning electron microscopy and size exclusion chromatog-
raphy.
X-Ray Structure of Complexes
Cu(4) and Zn(4) were further characterized by single crystal
X-ray diffraction. The crystallographic parameters for these two
complexes are given in Table 2 and the structures shown in Fig. 7.
It should be noted that the crystal structures for the complexes
Cu(1) and Cu(5) have previously been reported by our group
(Fig. 8).[9] X-ray crystallography of Cu(4) and Zn(4) revealed that
the Schiff base ligands bind to the metal in a bidentate fashion
with two ligand moieties interacting with the metal center.
The copper complex Cu(4) has a distorted square planar geometry
around the metal center. The structure for Cu(4) resembles that
of Cu(1) and Cu(5) reported recently.[9] As was the case for Cu(1),
complex Cu(4) also adopts a trans arrangement. This is most
likely due to the presence of the bulky iso-propyl groups on the
aromatic ring which is bonded to the imino nitrogen. This is unlike
the structure for an anlogous compound in which the N-aryl
Results and Discussion
Synthesis of Schiff-base Complexes
The Schiff-base complexes were prepared by the reaction of the
appropriate phenoxy-imine ligand with copper or zinc acetate
in a 2 : 1 mole ratio of ligand to metal precursors (Fig. 5). The
mononuclear complexes Cu(1) and Cu(5) have previously been
reportedbyuswhenusedasacatalystintheoxidationofphenol.[9]
The other complexes are new and were obtained as green
complexes for Cu and pale yellow solids for the Zn complexes. All
complexes were found to be stable but some decomposition is
observed over time when the complexes are dissolved in various
organic solvents. This is especially true of the zinc complexes.
c
Appl. Organometal. Chem. 2011, 25, 133–145
Copyright ꢀ 2010 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/aoc

S. Bhunora et al.
Table 1. Characterization data for copper(II) salicylaldiminato complexes Cu(1)–Cu(7)
Anal. found (calcd)
IR spectra (cm−1 b
)
Complex
Formula
M+ (calcd)a m/z
C
H
N
ν(C N)
ν(C–O)
Cu(1)
Cu(2)
Cu(3)
Cu(4)
Cu(5)
Cu(6)
Cu(7)
C38H44CuN2O2
624 (624.31)
455 (455.99)
680 (680.42)
736 (736.53)
c525 (525.89)
c694 (694.20)
848 (848.74)
72.90(73.11)
67.85(67.17)
74.57(74.14)
74.81(75.01)
58.41(58.49)
65.46(65.84)
73.62(73.32)
7.11(7.10)
4.17(4.42)
7.73(7.70)
8.17(8.21)
3.23(3.46)
6.12(6.11)
8.73(9.06)
4.45(4.49)
6.01(6.14)
4.13(4.12)
3.78(3.80)
5.54(5.34)
3.77(4.04)
2.53(2.97)
1602
1606
1609
1602
1606
1610
1610
1323
1327
1317
1318
1322
1322
1324
C26H20CuN2O2·1/2H2O
C
C
C
C
42H52CuN2O2
46H60CuN2O2
26H18Cl2CuN2O2·1/2H2O
38H42Cl2CuN2O2
C54H76CuN2O2·2H2O
a Recorded as ESI spectra unless otherwise stated.
b Recorded as nujol mulls between NaCl plates.
c Recorded as electron impact spectra.
Table 2. Crystallographic parameters
Cu(1)a
Cu(4)
Cu(5)a
Zn(4)
Empirical formula
Mr
C38H44CuN2O2
624.3
C46H60N2O2Cu
736.5
C26H18Cl2N2O2Cu
524.9
C46H60N2 O2Zn
738.3
Crystal system
Space group
a (Å)
Triclinic
P-1
Tetragonal
P-4
Monoclinic
C2/c
Triclinic
P-1
8.0725(16)
10.2332(9)
11.7022(3)
60.62(2)
76.39(3)
81.64(4)
818.2(1)
1
14.3631(10)
14.3631(10)
9.8445(7)
90
23.634(7)
9.337(2)
15.087(5)
90
10.9182(2)
12.2249(2)
16.0998(3)
90.0420(10)
103.0180(10)
97.9420(10)
2072.46(6)
2
b (Å)
c (Å)
α (deg)
β (deg)
γ (deg)
V (Å3)
90
137.343(7)
90
90
2030.9(2)
2
2256(1)
4
Z
D
calc (g cm−3
)
1.27
1.204
1.54
1.183
µ(MoKα) (cm−1
)
7.0
5.8
12.32
6.3
Temperature (K)
293(2)
100(2)
293(2)
100(2)
2θ range (deg)
4.0 < 2θ < 54.0
Empirical
3230/0/206
0.055
7.6 < 2θ < 56.4
psi-scan
4914/0/238
0.0326
7.6 < 2θ < 54.0
Empirical
2235/0/151
0.032
3.4 < 2θ < 50.6
psi-scan
7491/1/485
0.063
Absorption correction
Data/restraints/par.
R1 [I > 2σ(I)]
wR2 (all reflections)
Max/min residual electron density (e Å−3
0.119
0.0695
0.0909
0.0726
)
0.45; −0.86
0.336; −0.190
0.27; −0.35
0.386; −0.341
a Data from reference [9]
group is an unsubstituted phenyl group. The structure of the latter
compound reported by Repo et al. shows a cis arrangement of
N2O2 donor atoms.[10] These authors claim that in this case the
driving force for this arrangement is the ease of π –π stacking
of the N-phenyl rings. No such π –π stacking is observed in our
copper complexes, presumably due to the presence of the iso-
propyl substituents. The selected bond angles and bond lengths
for Cu(4) and Zn(4) together with those of the previously reported
structures,Cu(1)andCu(5),aregiveninTable 3.Fullcrystallograhic
data for the two new complexes are given in the Supporting
Information.
The zinc complex Zn(4) exhibits a distorted tetrahedral
geometry. The N-Zn–N angle in Zn(4) is 124.77◦ and is similar
to values obtained for analogous salicylaldimine complexes.[11]
The O–Zn–O angle in Zn(4) is 107.13◦ and is once again within
the same range of similar complexes. The two Zn–O bond lengths
are very similar, being around 1.93 Å, which corresponds well with
other Zn salicylaldimine complexes.
Preliminary Polymerization using DL-Lactide
The solubility of the complexes prepared was determined in
various solvents at different temperatures in order to determine
the initial polymerization conditions. Preliminary polymerization
reactions were thus carried out and the results are shown in
Table 4. As previously found for a range of other metal-alkoxide
and aryl-oxide complexes, the polymerization is assumed to most
probably occur via a coordination–insertion mechanism into an
Comparison of Cu(1) and Cu(4) with each other show that
the latter exhibits greater degree of distortion. This is clearly
seen when comparing the O–M–O bond angles. For Cu(4) this is
151.32◦ while for Cu(1) it is the expected bond angle of 180◦.
c
wileyonlinelibrary.com/journal/aoc
Copyright ꢀ 2010 John Wiley & Sons, Ltd.
Appl. Organometal. Chem. 2011, 25, 133–145

Cu and Zn salicylaldimine complexes as catalyst precursors
B
A
y = 0.0352x - 0.0911
R
[M]/[I] = 75
2 = 0.9977
y = 0.089x - 0.511
R2 = 0.876
[M]/[I] = 25
[M]/[I] = 75
[M]/[I] = 50
3
2
1.5
1
[M]/[I] = 25
[M]/[I] = 50
y = 0.0259x - 0.1653
2 = 0.9726
R
2
1
0
y = 0.026x + 0.003
R
2 = 0.8847
y = 0.0104x - 0.1172
2 = 0.9878
R
0.5
0
y = 0.0181x - 0.0382
R
2 =0.9369
0
20
40
60
80
100
120
0
10
20
30
40
Time (h)
Time (h)
6
4
2
0
C
[M]/[I] = 25
[M]/[I] = 50
[M]/[I] = 75
y = 0.0559x - 0.1248
R2 = 0.9456
y = 0.1255x - 0.9554
R
2 = 0.8838
y = 0.0332 - 0.7170
2 = 0.8934
R
0
20
40
60
Time (h)
80
100
Figure 6. Plot of ln[M]0/[M]t vs time for DL-lactide polymerization initiated using (A) Cu(4), (B) Cu(7), (C) Zn(1) at 70 ◦C in toluene.
M–O bond.[12a–d] Polymer end-group analysis via 1H NMR was
used to gain some insight into the polymerization mechanism.[12b]
An 1H NMR (CDCl3) spectrum of a sample of an oligolactide
exhibits two main signals at δ 1.6 ppm (d, CH3) and δ 5.2 ppm
(q, CH). A weak singlet located at δ 8.26 ppm corresponding to
the Schiff base’s imino group (HC N) was also observed in the
spectrum. This indicated that initiation of polymerization of DL-
lactide by the complexes occurred via insertion of one molecule
of lactide into the M–O bond of the catalyst, via acyl-oxygen
bond cleavage (Scheme 2). The result of this is the presence of
an imino end group in the growing oligomer. This process was
further confirmed by carrying out a reaction of DL-lactide with
Zn(1) in a 1 : 1 mole ratio and following the reaction via 1H NMR
spectroscopy. Over time signals are observed at 1.62, 5.2 and
8.26 ppm similar to those observed in the oligolactide’s spectrum
discussed above.
End-group analysis was also used to determine the molar mass
of polymers using eqn (2). A good relationship between this value
and the theoretical value further confirms living polymerization
via the coordination–insertion mechanism. For instance, Cu(4)
resulted in Mn(H NMR) compared with 4500 for Mn (theoretical)
of 5000 using [M] : [I] = 50 at 44% conversion. Similarly good
relationships were also obtained for other calculated values.
The initial polymerization results showed that only two copper
complexes, Cu(4) and Cu(7), and one zinc complex, Zn(1),
gave reasonable conversion of the monomer. These complexes
were thus used for further studies to establish the optimum
polymerization conditions. Reactions were performed at both 70
and 110 ◦C. In addition, kinetic studies using the more active metal
complexes were also conducted. This is discussed in more detail
in the next section.
Kinetic Studies
The kinetics of the ring opening polymerization at 70 ◦C, using
the two copper complexes Cu(4) and Cu(7) and as well as the
zinc complex Zn(1) as catalyst precursors, were investigated in
more detail. Linear plots were obtained for graphs of ln[M]0/[M]t
against time, where [M]0 is the initial monomer concentration,
while [M]t is the monomer concentration at time t, indicating
that the polymerization exhibited living characteristics (Fig. 6).
Linearity is observed for all [M] : [I] ratios, where [I] is the initiator
concentration. It can also be observed that the fastest reaction
rates are obtained at lower [M] : [I] ratios. An increase in [M] : [I]
ratios lead to a significant decrease in reaction rate. This is quite
marked for the copper complex, Cu(7).
The kp values for the copper and zinc complexes were
determined from the gradients of the plots of 1/t(ln[M]0/[M]t) vs
[I], where [I] is the initiator concentration (Table 5). Cu(4) showed
the lowest rate constant of polymerization while Zn(1) showed
the highest rate. This might be related to M–O bond strength. The
M–O bond length for Cu(4) is shorter than that of Zn(1) (Table 3).
The latter therefore has a weaker bond which should facilitate
monomer insertion into the M–O bond.
ICHpolymer /4
Mn(1H NMR) =
× 144
(2)
IHC N/2
whereICHpolymer isthepeakheightofmethineprotoninthepolymer
spectrum and IHC is the peak height of the HC N proton in
the polymer spectrum.
N
c
Copyright ꢀ 2010 John Wiley & Sons, Ltd.
Appl. Organometal. Chem. 2011, 25, 133–145
wileyonlinelibrary.com/journal/aoc

S. Bhunora et al.
A
B
Figure 7. X-ray structure of (A) Cu(4) and (B) Zn(4).
c
wileyonlinelibrary.com/journal/aoc
Copyright ꢀ 2010 John Wiley & Sons, Ltd.
Appl. Organometal. Chem. 2011, 25, 133–145

Cu and Zn salicylaldimine complexes as catalyst precursors
Table 4. Preliminary polymerization results using [M] : [I] = 50, [M] =
1 mol l−1
Temperature
(◦C)
Conversion
(%)b
Complex
Solventa
Time (h)
Zn(1)
Zn(4)
Cu(1)
Cu(2)
Cu(3)
Cu(4)
Cu(5)
Cu(6)
Cu(7)
Cu(3)
Cu(6)
Cu(7)
48
48
48
48
48
35
24
48
35
68
65
48
55
26
c
c
c
Toluene
70
55
18
c
80
Dioxane
THF
100
10
c
Dioxane
89
a Solvent used depended upon solubility of complex.
b Determined from 1H NMR.
c No polymerization observed.
ring, also showed low monomer conversion. The X-ray structure
of Cu(5) also showed a square planar geometry with a regular
arrangement of the phenyl groups and the molecule adopting a
cis configuration (Fig. 8). It is thought that the absence of bulky
substituents on the ligands makes the catalyst more susceptible
to deactivation via the dimerization of the metal centers. Such
dimerization often involves µ-oxo or µ-hydroxy bridges. On the
other hand sterically bulky groups around the metal prevent such
dimerization and thus prolong the catalyst life-time.[13] In addition
other sterically bulky groups in the vicinity of the metal center
retard any unwanted side reactions such as transesterification
reactions, thus increasing polymer yield.[18] They also prevent
catalyst deactivation via the formation of µ-oxo dimers.
Figure 8. X-ray structures of Cu(1) and Cu(5).[9]
.
Structure–Polymerization Relationship using X-ray
Crystallography
With a view to assessing whether the geometry imposed by the
four-coordinate ligand might account for polymerization activities
and selectivities, the crystal structures of some of the active and
inactive complexes were determined using single crystal X-ray
diffraction. Complexes Cu(1), Cu(2) and Cu(6) did not initiate
polymerization even when the reactions were carried out for
extendedperiodsat100 ◦C. Thesecomplexeshavenosubstituents
on the phenoxy moieties. X-ray studies showed that Cu(1) adopts
asquareplanargeometry(Fig. 7).Cu(5), withaClsubstituentinthe
para position but without any ortho substituent on the phenoxy
Cu(3) has di-substituted phenoxy rings with bulky t-butyl
groups at the ortho and para positions to the phenoxy moiety
but surprisingly no ROP of DL-lactide took place at 70 ◦C
while only 10% monomer conversion was observed when the
reaction is performed at 100 ◦C. Cu(3) also lacked substituents
on the imino bound aromatic ring. Van Wyk et al. have shown
that salicylaldimine ligands which do not contain substituents
on the imino bound aromatic rings tend to coordinate in a
cis arrangement.[9] This arrangement leads to increased steric
crowding around the metal center, unlike the trans arrangement,
Table 3. Selected bond lengths and bond angles obtained from X-ray data
Complex
∠N–M–O (deg)
∠O–M–O (deg)
M–O (Å)
1.999(2)
M–N (Å)
1.869(2)
Geometry
Cu(1)
91.39(9)
88.61(9)
93.13(5)
95.76(5)
93.56(6)
151.86(6)
119.99(5)
116.05(5)
93.76(5)
95.42(5)
180
Distorted square planar
Cu(4)
Cu(5)
151.42(7)
143.41(8)
88.07(8)
1.8959(11)
1.901(1)
1.9769(13)
1.985(1)
Distorted square planar
Distorted square planar
Zn(4)
107.15(5)
1.9280(12)
1.9294(11)
1.9837(13)
2.0088(13)
Distorted tetrahedral
c
Appl. Organometal. Chem. 2011, 25, 133–145
Copyright ꢀ 2010 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/aoc

S. Bhunora et al.
Scheme 2. Coordination–insertion mechanism for ring-opening polymerization of DL-lactide.
steric effects can be deduced from the fact that the kp value of
Cu(7) is three times faster than that of Cu(4).
Table 5. The kp values of DL-lactide polymerization using copper and
zinc complexes in toluene at 70 ◦C
The situation is reversed in the case of the zinc complexes.
Here a t-butyl substituent in the 3-position of the salicylaldimine
ligand retards the polymerization process. Steric effects are now
dominant. This is largely due to differences in the electronic
configuration of zinc (d10) as opposed to copper (d9). In the zinc
complexes the ability of the phenoxy O to act as a π-donor is
prevented by the fact that the metal has no empty d orbitals to
accept electrons from the phenoxy group. We thus postulate that,
in the case of the zinc complexes, steric factors now outweigh
electronic factors.
Complex
kp (mol−1 dm3 h−1
)
Cu(4)
Cu(7)
Zn(1)
0.869
2.777
2.839
whichforcesthephenylringsawayfromthemetalcenter,reducing
steric interaction and thus leading to higher polymerization
activities.[14]
As expected, X-ray crystallography data show that the copper
complexes have a distorted square planar geometry around
the metal center, while the zinc complexes showed a distorted
tetrahedral geometry [Fig. 8(b), Table 3]. DL-lactide can approach
the Cu metal center unhindered from either the bottom or top
of the coordination plane while the tetrahedral zinc complexes
provide a more hindered access. Structural data showed that the
inactive complex, Cu(1), has the shortest Cu–O bond length and
thus the strongest Cu–O bond. This would render the insertion of
the monomer into the Cu–O bond more difficult. X-ray analysis
thus confirms that electronic factors predominate in the case of
Cu while steric factor prevails for Zn.
It is interesting to note that Gibson et al. studied the effect of
phenoxy substituents of aluminum salen type complexes on the
rate and stereoselectivity of DL-Lactide polymerization.[20] They
found that electron-withdrawing groups attached to the phenoxy
ring gave an increase in polymerization rate whereas bulky ortho
substituentssloweddownthepolymerization.Thelatter,however,
increased the isotacticity of the resulting PDLLA. Gibson et al.
showed that the Al complexes adopt a five-coordinate trigonal
bipyramidal geometry at the metal center. Thus steric factors due
to the size of the ortho phenoxy substituent hinder the approach
of the lactide monomer to the aluminum center.
The most active copper complexes have ortho t-butyl sub-
stituents on the phenoxy ring and ortho isopropyl substituent on
the imino-bound ring. The enhanced activity ofCu(4) and Cu(7)
are most likely due to electronic factors rather than steric factors.
If the latter were dominant then one would have expected a
decrease in activity as the bulky tBu group in the 3 position would
have blocked access of the monomer to the catalyst active site,
thus retarding the polymerization process. The phenoxy oxygen is
most likely to act as a π-donor ligand. In this case an electron rich
substituent in the ortho position such as the tBu in the 3-position
of the phenoxy group will increase electron density on the phe-
noxy oxygen and enhance its π-donor ability, leading to a weaker
t
M–O bond. A comparison of the Cu–O bond in Cu(1) (no Bu
substituent) and Cu(4) (with tBu substituent) shows that the latter
has a longer and thus a weaker Cu–O bond (Table 3).
The ability of Cu(7) to mediate ROP of DL lactide at such high
conversionscanbeexplainedbythefactthatithastBusubstituents
at positions ortho and para to the hydroxyl group. This would lead
to a further increase in the π-donor ability of the phenoxy group,
thus further weakening the M–O bond and therefore leading to a
higher rate of monomer insertion into the M–O bond. In addition
the high degree of substitution on the aromatic rings helps protect
themetalsiteagainstundesiredsidereactionsaspreviouslyshown
by Kerton et al.[19] That electronic effects are more significant than
c
wileyonlinelibrary.com/journal/aoc
Copyright ꢀ 2010 John Wiley & Sons, Ltd.
Appl. Organometal. Chem. 2011, 25, 133–145

Cu and Zn salicylaldimine complexes as catalyst precursors
Table 6. 13C NMR characterization, [M] : [I] = 50, toluene 70 ◦C
Tetrad intensities (%)
Conversion
(%)
Time
(h)
Complex
ssi
sss
isi
iss
sis, sii, iis, iii
I
Isotactic[27,28]
0
0
0
0
0
0
0
8
25
30
16
16
0
0
0
3
75
70
84
73
[25]
Al(OiPr)3
75
94
96
16
6
–
[26]
HAPENAlOCH3
HAPENAlOiPr[26]
1.68
1.7
4
Cu(4)
Cu(7)
Zn(1)
Zn(1)
Zn(4)
98
86
70
94
99
96
68
72
96
96
0
0
0
0
4
0
0
24
27
30
30
24
0
0
76
73
70
70
62
1.31
1.54
1.85
2.28
2.54
0
0
0
0
3.4
6.6
in polymers with higher isotactic content than those obtained
from Zn complexes. It can be concluded that, while electronic
factors affect the rate of polymerization, steric factors influence
isotacticity of the polymers produced. It is also interesting to
note that previously reported Zn complexes resulted in atactic
PDLLAs.[6,7]
Polymers obtained using complexes Cu(4), Cu(7) and Zn(1) as
catalysts do not show the presence of the sss and ssi forbidden
tetrads which indicate a low percentage of transesterification
reactions even at high percentage conversion.[21–25] Complex
Zn(4), which adopts a highly distorted tetrahedral structure,
shows the presence of the forbidden tetrads. These results were
comparedwithaluminuminitiatorssynthesizedbyBhaw-Luximon
et al.[25,26] (Table 6) and the degree of isotacticity was found to be
between that of Al(OiPr)3 and aluminum Schiff base initiators.
Table 7. PDLA synthesized using Cu(4) complex in toluene at 70 ◦C
Conversion (%)
([M] : [I])
I
MnSEC (corrected)a
nb
25 (50)
98 (50)
30 (75)
47 (75)
1.25
1.31
1.21
1.19
5 800 (3 250)
12 500(7 000)
5 000 (2 800)
13 145 (7 360)
0.6
1.0
1.2
0.7
a After applying a correction factor of 0.56 for Mn.[30,31]
b Number of active sites calculated using Mn
.
SEC [29]
Characterization of Polymers
Stereochemistry of polymers
The stereochemistry of the polymers was determined using
Molecular weight determination of the polymers
1
13C{ H} spectroscopy. From the NMR spectra it can be concluded
that predominantly isotactic polymers were obtained using the
Cu and Zn complexes (Table 6), showing that the catalysts are
stereoselective. This result can be correlated with the presence of
bulky ortho phenoxy substituents as has also been reported by
Gibson et al.[20] for aluminum complexes. Cu complexes resulted
The molecular weights of the polymers were determined using
ambient temperature size exclusion chromatography (SEC). This
showed that the highest molecular weight polymers were
obtained in the case where Zn(1) was used as initiator while
narrower polydispersity index polymers were obtained with
A
B
Figure 9. SEC chromatograms of PDLLAs synthesized using (A) Zn and (B) Cu complexes in toluene at 70 ◦C at [M] : [I] = 50, [M] = 1M.
c
Appl. Organometal. Chem. 2011, 25, 133–145
Copyright ꢀ 2010 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/aoc

S. Bhunora et al.
Figure 10. DSC chromatogram recorded for a PDLLA synthesized using Cu(4) at [M] : [I] = 75.
Figure 11. SEM image of PDLLA synthesized using (A) Cu(4) and (B) Cu(7).
copper complexes as initiators (Tables 6 and 7). The occurrence
of transesterification reactions using zinc initiators was confirmed
from SEC traces (Fig. 9), which showed the presence of low molar
mass polymers. The values of the number average molecular
weights as obtained from SEC (MnSEC) were further used to
determine the number of active sites (n) for one of the complexes,
Cu(4), using eqn (3) in order to have a better insight into the
mechanism of polymerization (Table 7).[29] It was found that on
average only one M–O bond was active in polymerization.
of values obtained is directly related to the isotactic content
of the polymers, as shown by Urayama et al.[32] They showed
that an increase in isotacticity brings about an increase in Tg.
Indeed Tg determined here are higher than expected due to the
higher isotactic content, indicating that the polymers contain
stereoblocks. Cu complexes resulting in the higher isotactic
content also gave PDLLA with higher Tg.
The high stereoregularity of the PDLLAs was confirmed by
the presence of sharp melting points, Tm (130 ◦C), for polymers
having isotactic content >70% (Fig. 10).[33] The existence of a
melting transition is due to the formation of sufficiently long
stereosequences of D and L units. These stereoblocks give rise
to stereocomplexes between the chains. Spassky et al.[33–35] have
previously observed Tm values in the range 125–150◦C for PDLLAs
produced using aluminum Schiff base complexes as catalysts.
ꢀ
ꢁ
(M/I) × MM(DL-lactide)
n = %conversion
(3)
MSnEC
where n is the number of active sites; MM (DL-lactide) = molar
mass of DL-lactide = 144 g mol−1
.
Thermal properties of polymers produced
Polymer morphology
The thermal properties of the polymers were studied using
differential scanning calorimetry (DSC). PLLA with 100% L-content
and PDLLA with completely racemic and random sequence
have Tg at 62 and 45 ◦C respectively.[32] DSC chromatograms
of PDLLAs synthesized using copper and zinc complexes showed
the presence of Tg in the range 53–59 and 54–54.5 ◦C. The range
The morphology of the polymers produced were investigated
using scanning electron microscopy (SEM). The SEM images of the
purified PDLLAs were recorded. Different surface patterns were
observed, namely microporous [using Cu(4), Fig. 11A] and low
roughness homogeneous surfaces [using Cu(7), Fig. 11B].
c
wileyonlinelibrary.com/journal/aoc
Copyright ꢀ 2010 John Wiley & Sons, Ltd.
Appl. Organometal. Chem. 2011, 25, 133–145

Cu and Zn salicylaldimine complexes as catalyst precursors
Conclusions
[8] A. John, V. Katiyar, K. Pang, M. M. Shaik, H. Nanavati, P. Ghosh,
Polyhedron 2007, 26, 4033.
[9] J. L. Van Wyk, S. F. Mapolie, A. Lennartson, M. Håkansson, S. Jagner,
Naturforschung 2007, 62b, 331.
[10] J. U. Ahmad, P. J. Figiel, M. T. Ra¨isa¨nen, M. Leskela¨, T. Repo, Appl.
Catal. A Chem. 2009, 371, 17.
[11] D. J. Darensbourg, P. Rainey, Y. Yarbrough, Inorg. Chem. 2001, 40(5),
986.
[12] a) A. Hamitou, T. Ouhadi, R. Je´roˆme, Ph. Teyssie´, J. Polym. Sci.:
Polym. Chem. 1977, 15, 865; b) M. Wisniewski, PhD thesis Universite´
Pierre et Marie Curie (Paris VI), France, 1995; c) A. Le Borgne, M.
Wisniewski, N. Spassky, Polym. Preprint. 1995, 36(2), 217; d) M. Stolt,
A. So¨dergård, Macromolecules 1999, 32, 6412.
[13] Z.-W. Yang, Q.-X. Kang, H.-C. Ma, C-L. Li, Z.-Q Lei, J. Mol. Catal. A:
Chem. 2004, 213, 69.
Polymerization of DL-lactide has been successfully carried out
using novel Cu and Zn salicyaldimine complexes. We have shown
from X-ray data that the geometry of the complex influences the
kinetics and microstructure of the polymer. It was found that the
rate of polymerization was influenced by electronic factors for Cu
complexes and steric factors for zinc complexes. Predominantly
isotactic PDLLA was obtained which resulted in an increase of the
glass transition temperature (Tg) and the presence of a melting
point (Tm) for the polymers. Isotactic content was influenced by
the bulkiness of ligands coordinated to the metal center.
[14] A. Altomare, G. Cascarano, C. Giacovazzo, J. Appl. Crystallogr. 1993,
26(3), 343.
Acknowledgments
[15] SHELX97 (Release 97-2). G. M. Sheldrick, Institut fu¨r Anorganische
Chemie der Universita¨t, Tammanstrasse 4, D-3400 Go¨ttingen,
Germany, 1998.
[16] L. J. Farrugia, J. Appl. Crystallogr. 1999, 32, 837.
[17] L. J. Farrugia, J. Appl. Crystallogr. 1997, 30(5), 565.
[18] A. Pa¨rssinen, Dissertation, University of Helsinki, Finland, 2006.
[19] F. M. Kerton, A. C. Whitwood, C. E. Williams, Dalton Trans. 2004,
2237.
National Research Foundation, South Africa is acknowledged for
funding provided under the South African Regional Cooperation
Fund for Scientific Research and Technological Development
Scheme. The SIDA/NRF Swedish Research Links programme. is
also acknowledged. The authors also wish to thank PURAC for a
generous donation of DL-lactide.
[20] P. Hormnirun, E. L. Marshall, V. C. Gibson, R. I. Pugh, A. J. P. White,
Proc. Natl Acad. Sci. (USA) 2006, 103(42), 15343.
[21] M. Swarc, Adv. Polym. Sci. 1983, 49, 7.
[22] F. Chabot, M. Vert, S. Chapelle, P. Granger, Polymer 1983, 24(1), 53.
[23] H. R. Kricheldorf, I. Keiser-Saunders, Makromol. Chem. 1990, 191,
1057.
[24] H. R. Kricheldorf, I. Keiser-Saunders, C. Boettcher. Polymer 1995,
36(6), 1253.
[25] A. Bhaw-Luximon, PhD thesis, University of Mauritius, Mauritius,
Supporting Information
CCDC 771676 and 771677 contain the supplementary crystal-
lographic data for complexes Cu(4) and Zn(4). These data can
be obtained free of charge via http://www.ccdc.cam.ac.uk/conts/
retrieving.html, or from the Cambridge Crystallographic Data
Centre, 12 Union Road, Cambridge CB2 1EZ, UK; fax: (+44) 1223-
336-033; or via e-mail: deposit@ccdc.cam.ac.uk. Tables S1–S3 are
deposited as supporting information and can be found in the
online version of this article.
2001.
[26] D. Jhurry, A. Bhaw-Luximon, N. Spassky, Macromol.Symp. 2001, 175,
679.
[27] J. Coudane, C. Ustaritz-Peyret, G. Schwach, M. Vert, J. Polym. Sci.
Polym. Chem. 1997, 35(9), 1651.
[28] J. Kasperczyk, M. Bero, Polymer 2000, 41(1), 391.
[29] V. Simic, PhDthesis Universite´ Pierre et Marie Curie (Paris VI), France,
1998.
References
[30] H. Ma, J. Okuda, Macromolecules 2005, 38(7), 2665.
[31] J. Baran, A. Duda, A. Kowalski, R. Szymanski, S. Penczek, Macromol.
Symp. 1997, 123, 93.
[1] R. Langer, Science 1990, 249, 1527.
[2] D. J. Darensbourg, W. Choi, C. P Richers, Macromolecules 2007,
40(10), 3521.
[32] H. Urayama, S.-I. Moon, Y. Kimura. Macromol. Mater. Eng. 2003,
288(2), 137.
[3] L. Piao, M. Deng, X. Chen, L. Jiang, X. Jing, Polymer 2003, 44(8), 2331.
[4] Z. Zhong, P.J. Dijkstra, J. Feijen, J. Biomater. Sci., Polym. Edn. 2004,
15(7), 929.
[33] A. Bhaw-Luximon, D. Jhurry, N. Spassky, Polym. Bull. 2001, 44, 31.
[31] N. Spassky, C. Pluta, V. Simic, M. Thiam, W. Wisniewski, Macromol.
Symp. 1998, 128, 39.
[5] M. H. Chisholm, J. C. Gallucci, J.C. Huffman, Inorg. Chem. 2001,
40(19), 5051.
[35] N. Spassky, M. Wisniewski, C. Pluta, A. Le Borgne, Macromol. Chem.
Phys. 1996, 197(9), 2627.
[6] C. Zhang, Z-X. Wang, J. Organomet. Chem. 2008, 693(19), 3151.
[7] W. Hung, Y. Huang, C-C. Lin, J. Polym. Sci. Part A: Polym. Chem. 2008,
46, 6466.
c
Appl. Organometal. Chem. 2011, 25, 133–145
Copyright ꢀ 2010 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/aoc