Enolic Schiff base aluminum complexes and their catalytic stereoselective polymerization of racemic lactide

Article abstract of DOI:10.1002/chem.200701473

A series of enolic Schiff base aluminum(III) complexes LAlR (where L =NNOO-tetradentate enolic Schiff base ligand) containing ligands that differ in their steric and electronic properties were synthesized. Their single crystals showed that these complexes are five-coordinated around the aluminum center. Their coordination geometries are between square pyramidal and trigonal bipyramidal. Their catalytic properties in the solution polymerization of racemic lactide (rac-LA) were examined. The modifications in the auxiliary ligand exhibited a dramatic influence on the catalytic perfor_manee. Lengthening the backbone from C₂ alkylene to C, alkylene resulted in remarkable enhancement of both the stereoselectivity and the polymerization rate because of the increasing flexibility of the diimine backbone. Electron-withdrawing substituents in the diketone also highly improved the activity and the stereoselectivity. Among these complexes, 4b had the highest activity and the stereoselectivd the stereoselectivity. Among these complexes, 4b had the highest activity and the stereoselectivity owing to the C₃ alkylene backbone and the two gem-methyl groups on the middle carbon atom. The value of the polymerization rate constant (k_p) catalyzed by 4 b in 70°C was 1.90 Lmol^-1 min^-1, the activation energy of the polymerization (35.4 kJ mol^-1) was calculated according to the Arrhenius equation. Other factors that influenced the polymerization, such as the polymerization time, the temperature, and the monomer concentration, are also discussed in detail.

Full text of DOI:10.1002/chem.200701473

DOI: 10.1002/chem.200701473
Enolic Schiff Base Aluminum Complexes andTheir Catalytic
Polymerization of Racemic Lactide
ACHTRESUNG tereoselective
[a]
Xuan Pang,[a, b] Hongzhi Du,[a, b] Xuesi Chen,*[a] Xianhong Wang,[a] andXiabin Jing
Abstract: A series of enolic Schiff base
aluminum(III) complexes LAlR
mance. Lengthening the backbone
from C2 alkylene to C3 alkylene result-
ed in remarkable enhancement of both
the stereoselectivity and the polymeri-
zation rate because of the increasing
flexibility of the diimine backbone.
Electron-withdrawing substituents in
the diketone also highly improved the
activity and the stereoselectivity.
Among these complexes, 4b had the
highest activity and the stereoselectiv-
ity owing to the C3 alkylene backbone
and the two gem-methyl groups on the
middle carbon atom. The value of the
polymerization rate constant (kp) cata-
ACHTREUNG
(where L=NNOO-tetradentate enolic
Schiff base ligand) containing ligands
that differ in their steric and electronic
properties were synthesized. Their
single crystals showed that these com-
plexes are five-coordinated around the
aluminum center. Their coordination
geometries are between square pyrami-
dal and trigonal bipyramidal. Their cat-
alytic properties in the solution poly-
merization of racemic lactide (rac-LA)
were examined. The modifications in
the auxiliary ligand exhibited a dramat-
ic influence on the catalytic perfor-
lyzed
1.90 Lmolꢀ1 minꢀ1,
energy of the
by
4b
in
the
polymerization
708Cwas
activation
(35.4 kJmolꢀ1) was calculated accord-
ing to the Arrhenius equation. Other
factors that influenced the polymeri-
zation, such as the polymerization
time, the temperature, and the mono-
mer concentration, are also discussed
in detail.
Keywords: aluminum
· lactide ·
ring-opening polymerization · Schiff
bases · stereoselectivity
Introduction
promising biodegradable polyesters because of its renewable
resources.[3] PLA can be prepared by the ring-opening poly-
merization (ROP) of lactide (LA) or by direct condensation
of lactic acid. Lactide exists as three different stereoisomers:
l-lactide (l-LA), d-lactide (d-LA) and meso-lactide
(Scheme 1), so their polymers may have different chain con-
figurations (Scheme 2). the chain microstructures of PLA
are one of the most important factors that influence the
physical, mechanical and degradation properties of the poly-
mers.[4] Conventional catalysts such as aluminum trialkox-
ides[5] and tin(II) octanoate[6] do not prefer one of the three
different stereoisomers of LA, so the polymers obtained by
these catalysts contain random placements of lactide mono-
mers. The stereoregular polymers can be obtained contain
by using either the enantiopure l-LA or d-LA monomer.
Biodegradable polyesters have found a wide range of appli-
cations, such as sutures, bone fracture fixation devices, drug
controlled release carriers, tissue engineering scaffolds and
green plastics for wrapping materials, disposal containers
and fibers.[1] Because the ester linkage is easily hydrolyzed,
they can be converted into hydroxyl acid via degradation
and finally to CO2 and water. For these reasons, a lot of re-
search on the biodegradable polyesters has been carried out
in recent years.[2] Polylactide (PLA) is one of the most
[a] X. Pang, H. Du, Prof. Dr. X. Chen, Prof. Dr. X. Wang, Prof. X. Jing
State Key Laboratory of Polymer Physics and Chemistry
Changchun Institute of Applied Chemistry
Chinese Academy of Sciences, 5625 Renmin Street
Changchun 130022 (China)
Fax : (+86)431-852-65653
E-mail: xschen@ciac.jl.cn
[b] X. Pang, H. Du
Graduate School of Chinese Academy of Sciences
Beijing 100039 (China)
Supporting information for this article is available on the WWW
under http://www.chemeurj.org/ or from the author.
Scheme 1. Stereoisomers of lactides.
3126
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 3126 – 3136

FULL PAPER
Scheme 3. Ligand precursors explored for stereoselective ROP of lacti-
des.
might be a potential catalyst for ROP of rac-LA. We recent-
ly communicated two enolic Schiff base complexes acting as
the precursors for the polymerization of rac-LA by a chain
end controlled mechanism.[11] To expand our initial work,
we have investigated the effect of the complex architecture
on the lactide polymerization behavior. We tried to eluci-
date the relationship between the complex structure and the
catalytic performance of the enolic Schiff base aluminum
complexes. Our approach was based on fine-tuning of the
ancillary ligand and study of the effect on the catalytic reac-
tivity. Herein we systematically explore the synthesis, char-
acterization and stereoselective polymerization of the enolic
Schiff base aluminum complexes family.
Scheme 2. Microstructures of PLA.
Since stereocomplex polymers formed by an equivalent
mixture of PLLA and PDLA have many advantages such as
higher melting temperature (2308C).[7] Many efforts have
been made to obtain crystalline PLA via direct ring-opening
polymerization (ROP) of racemic lactide (rac-LA; i.e., a 1:1
mixture of l-LA and d-LA) by the stereoselective catalysts.
Stereoblock PLA can be obtained from rac-LA using the
aluminum complex of a chiral binaphthyl Schiff base origi-
nally proposed by Spassky et al.[8] A few other excellent
studies have attempted to elucidate the relationship between
the aluminum Schiff base complexes and stereoselectivity[9]
(Scheme 3). Among them, Coates,[9a] Feijen[9c,d] and Duda[9j]
used chiral salen-type Schiff base catalysts via the enantio-
morphic site control mechanism; Nomura[9e] and our
group[9g,h] used achiral salen-type Schiff base catalysts via
the chain-end control mechanism.
Results andDiscussion
Ligandsynthesis : The versatile ligand family (Scheme 4)
was easily synthesized from readily available starting materi-
als, namely primary diamines with acetylacetone, 1-benzoyl-
AHCTREUNG
acetone and 1,1,1-trifluoro-acetylacetone.[12]
Ligands 1, 2 and 7 had the identical aliphatic diimine
Enolic ligands could be obtained from the reaction of b-
diketone and diamine. Ideal Schiff base ligands and their
aluminum complexes can be obtained through keto–enol
tautomerism. To the best of our knowledge, few researches
on enolic Schiff bases and their metal complexes have been
reported in the rac-LA polymerization.[10] It is postulated
that aluminum complexes coordinated by enolic Schiff base
backbones but the different enol substituents: CH3 for 1,
C6H5 for 2 and CF3 for 7. Ligands 2, 3, 4 and 5 had identical
enol segments but the different diimine backbones:
-CH2CH2- for 2, -C HCH2CH2- for 3, -C HCCAHTRE(UNG CH3)2CH2- for
2
2
4 and -CH2C6H5CH2- for 5. Ligands 5 and 6 had the identi-
cal aryl diimine backbones but the different enol substitu-
ents: C6H5 for 5 and CH3 for 6.
Chem. Eur. J. 2008, 14, 3126 – 3136
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3127

X. Chen et al.
lying on the equatorial plane. The central Al atom was
about 0.029 Š above the equatorial plane formed by O2, N1
and C1 in the direction of N2 with an average compressed
axial O1-Al-N2 bond angle of 167.88(9)8 and equatorial O2-
Al-N1, C1-Al-N1, and O2-Al-C1 bond angles of 120.41(9),
120.56(11), and 118.96(11)8, respectively. The distances from
the Al atom to O1, O2, N1, N2 and C1 were 1.8881(18),
1.8108(19), 1.961(2), 2.081(2) and 1.949(3) Š, respectively.[11]
Compound 4a has an identical t value (0.78) with 3a. The
geometry of 4a (Figure 2) was analogous to 3a with an aver-
age compressed axial O2-Al-N1 bond angle of 167.78(6)8
and equatorial O1-Al-N2, O1-Al-C15, and N2-Al-C15 bond
angles of 120.60(6), 119.25(7) and 120.15(7)8, respectively.
The distances from the Al atom to O1, O2, N1, N2 and C15
were 1.8151(12), 1.8724(12), 2.0639(14), 1.9674(15) and
1.9885(18) Š, respectively. Central Al atom was about
0.001 Š above the equatorial plane formed by O1, N2 and
C15 in the direction of O2. The difference of the t values
between 2a and 3a was presumably due to the nature of
their diimine backbones, 3a had more flexible C3 backbone
than 2a (C2 backbone); the tbp geometry in 3a may avoid
the eclipse of methylene hydrogens. The similar results were
reported in salen aluminum systems.[13a] The substituent
groups in the middle carbon of the C3 backbone seemed to
have no obvious influence in comparison with the diimine
backbone length on the geometry as 3a and 4a had the
same t value.
Scheme 4. Ligands prepared for the stereoselective polymerization of lac-
tides. A) change of the diamine substituents, B) change of diketone sub-
stituents.
Further insight into the nature of real initiators of ROP
and the effect of the fifth coordinated group on geometry
came from the preparation and characterization of two
enolic Schiff base aluminum isopropoxides 2b and 4b via in
situ alcoholysis (Scheme 5). X-ray single crystal structure
analysis of 2b also showed a five-coordinated aluminum
center (Figure 3). The t value of 2b was 0.54. The geometry
around the Al atom was a distorted tbp geometry with O2
and N1 occupying the axial site and O1, N2 and O3 lying on
the equatorial plane. Central Al atom was about 0.088 Š
above the equatorial plane formed by O1, N2 and O3 in the
direction of O2 with an average compressed axial O2-Al-N1
bond angle of 165.50(7)8 and equatorial O1-Al-N2, O1-Al-
O3, and N2-Al-O3 bond angles of 133.27(6), 113.05(7) and
112.98(7)8. The distances from the Al atom to O1, O2, N1,
N2 and O3 were 1.8217(13), 1.8389(13), 1.9897(16),
1.9827(16) and 1.7443(14) Š. The large difference in the t
values between 2a (aluminum ethyl, t value was 0.09) and
2b (aluminum isopropoxide, t value was 0.54) manifested
that 2b was more distorted than 2a. The geometry of 4b
was distorted tbp with O1 and N2 occupying the axial site
and O2, N1 and O3 lying on the equatorial plane (Figure 4).
Central Al atom was about 0.046 Š above the equatorial
plane formed by O2, N1 and O3 in the direction of O1 with
an average compressed axial O1-Al1-N2 bond angle of
172.87(14)8 and equatorial O2-Al1-N1, O2-Al1-O3, and N1-
Al1-O3 bond angles of 120.31(14), 121.83(14) and
117.67(14)8, respectively. The distances from the Al1 atom
to O1, O2, N1, N2 and O3 were 1.861(3), 1.802(3), 1.967(3),
2.011(4) and 1.740(3) Š, respectively. The t value of 4b was
Complex formation andcrystal structure characterization :
Reaction of ligands 1–7 with stoichiometric AlEt3 in toluene
formed enolic Schiff base aluminum ethyl complexes 1a–7a,
respectively. All these complexes were isolated as solid pow-
ders. To estimate the relationship between geometric struc-
tures and their catalytic performance, single crystals of 2a,
3a, and 4a were measured (Table 1). X-ray diffraction data
showed that all three crystals were five-coordinated around
the aluminum center. The parameter t was used to describe
the degree of crystal distortion of a five-coordinated Schiff
base aluminum complex:[13] a perfect square-pyramidal ge-
ometry (sqp) corresponds to t=0, while perfect trigonal bi-
pyramidal geometry (tbp) corresponds to t=1.
The molecular structure of 2a with the t value of 0.09 was
shown in Figure 1. The geometry around the Al atom was
sqp geometry with the ethyl group lying on the axial posi-
tion and two nitrogen atoms and two oxygen atoms on the
basal position. Central Al atom was about 0.531 Š above
the N1-N2-O1-O2 mean plane with an average compressed
axial O1-Al-N2 bond angle of 150.34(15)8 and equatorial
O2-Al-N1, C23-Al-N1, and O2-Al-C23 bond angles of
144.91(15), 110.24(18), and 104.72(16)8, respectively. The
distances from the Al atom to O1, O2, N1, N2 and C23
were 1.831(3), 1.835(3), 1.982(5), 1.999(5) and 1.970(4) Š,
respectively. Compound 3a had a t value of 0.78, manifest-
ing that 3a adopted a distorted tbp geometry with O1 and
N2 occupying the axial site and O2, N1 and ethyl group
3128
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 3126 – 3136

Polymerization of Racemic Lactide
FULL PAPER
Table 1. Crystallographic data for 2a, 3a, 4a, 2b and 4b.[a]
meric behavior, we could not
conclude whether these com-
plexes retained their monomer-
ic structures in their solution-
state. Coates et al.[9a] reported
that the salen-type aluminum
isopropoxide, which had less
steric hindrance, had a dimer
character in the solution state,
though the dimer was inactive
in the lactide polymerization.
Recently, Lin et al.[14] reported
that the Schiff base zinc com-
plex showed dimer behavior in
the solid state, but in the solu-
tion state, a mixture of dimer
and monomer (include diaste-
reoisomer) was observed. In
order to determine the real
conformation of these com-
plexes in the solution state, we
investigated the 1H and 27Al
NMR spectra of these com-
2a
3a
4a
2b
4b
empirical formula
formula weight
T [K]
C24H27AlN2O2
402.46
187(2)
C25H29AlN2O2
416.48
187(2)
C27H33AlN2O2
444.53
C25H29AlN2O3
432.48
187.0(2)
0.71073
C70H86Al2N4O6
1133.39
187(2)
0.71073
monoclinic
Cc
22.4239(14)
13.4238(8)
21.9046(14)
90
102.7490(10)
90
187(2)
l [Š]
0.71073
monoclinic
P21/n
7.4501(6)
18.6477(16)
15.5635(14)
90
0.71073
monoclinic
C2/c
29.660(3)
7.5901(6)
22.3793(19)
90
0.71073
monoclinic
P21/n
9.9827(7)
11.1884(8)
21.7980(16)
90
96.1190(10)
90
2420.8(3), 4
1.220
crystal system
space group
a [Š]
b [Š]
c [Š]
triclinic
¯
P1
10.4661(11)
10.6923(11)
12.2855(13)
86.341(2)
66.2050(10)
64.0240(10)
1120.2(2), 2
1.282
a [8]
b [8]
98.712(2)
90
2137.2(3), 4
1.251
0.117
856
116.486(2)
90
4509.4(7), 8
1.227
0.113
1776
g [8]
V [Š3], Z
6431.0(7), 4
1.171
1calcd [Mgmꢀ3
]
m [mmꢀ1
(000)
]
0.110
952
0.120
460
0.099
2432
F
A
crystal size [mm]
0.36”0.11”0.10 0.41”0.15”0.06 0.50”0.21”0.15 0.53”0.13”0.03 0.27”0.25”0.19
q range [8]
limiting indices (hkl) ꢀ7 to 8
1.72 to 25.25
1.53 to 26.05
ꢁ36
1.88 to 26.00
ꢀ11 to 12
ꢁ13
1.83 to 25.50
ꢀ9 to 12
ꢁ12
1.78 to 26.04
ꢀ27 to 13
ꢁ16
ꢀ21 to 22
ꢁ18
ꢁ9
ꢀ21 to 27
12309
4434
ꢀ26 to 13
13302
ꢁ14
ꢀ26 to 27
17895
reflections collected
independent reflec-
11693
3855
6077
4766
4093
8412
tions
Rint
0.0455
0.0500
1.037
0.0229
1.036
0.0160
1.017
0.0299
1.009
goodness of fit on F2 1.015
1
plexes. The H NMR spectra of
final R1, wR2
[I>2s(I)]
0.0825, 0.1696
0.0613, 0.1305
0.0438, 0.1138
0.0431, 0.1061
0.0533, 0.1348
the ethyl complexes 1a–7a
showed one set of resonance
peak and the 27Al NMR spectra
of 1a–7a showed resonance
peak at about 36 ppm, indicat-
ing that all the ethyl complexes
R1, wR2 (all data)
0.1243, 0.1932
0.0981, 0.1489
0.0536, 0.1214
0.315, ꢀ0.197
0.0566, 0.1151
0.311, ꢀ0.210
0.0709, 0.1483
0.406, ꢀ0.227
largest diff. peak and 0.385, ꢀ0.404
0.353, ꢀ0.276
ꢀ3
hole [eŠ
]
[a] Selected bond lenths and angles are available in the Supporting Information.
Figure 1. Crystal structure of complex 2a.
Figure 2. Crystal structure of complex 3a.
0.85. Compound 4b showed more tbp geometry character
than 4a (t value was 0.78). From the crystal analysis, it was
believed that the fifth coordinated group had notable influ-
ence on the complexꢁs geometry.
retained their conformation with the five-coordinated mono-
1
meric Al center in the solution state.[15] The H NMR spec-
tra of the isopropoxide complexes 2b and 4b also showed
one set of resonance peak; additional 27Al spectra showed a
resonance peak at about 36 ppm. We could conclude from
Complexes structure characterization in solution: Although
the solid-state structures of these complexes showed mono-
Chem. Eur. J. 2008, 14, 3126 – 3136
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3129

X. Chen et al.
ꢀd½LAŠ
dt
ð1Þ
¼ kapp½LAŠ
The number-average molecular
weight (Mn) also followed a
linear relationship in monomer
Scheme 5. Synthetic pathway for the preparation of complexes.
conversion for both the aliphat-
ic diimine backbone complexes
(Figure 7) and aryl diimine
backbone complexes (Figure 8). The molecular control and
the low polydispersity indicated that the polymerization had
a characteristic of controlled propagation. To determine the
order in Al, kapp was plotted versus the concentration of Al
(Figure 9) using 4a. From this plot, kapp increased linearly
with the Al concentration, manifesting that the order in Al
was first-order too. Therefore, the polymerization of rac-LA
using 4a followed an overall kinetic equation of the form as
Equation (2), where kp was the polymerization rate con-
stants and kp =kapp/[Al].
ꢀd½LAŠ
ð2Þ
¼ kp½LAŠ½AlŠ
dt
We could extrapolate this equation to other complexes in
the rac-LA polymerization. Aluminum isopropoxides 2b
and 4b were also used as real initiators to investigate their
reactivities (Table 3). According to the polymerization data,
it was concluded that in the present of 2-propanol, alumi-
num ethyl complex 2a and aluminum isopropoxide 2b had
same reactivities, similar results were obtained for 4a and
4b. It was envisioned that the in situ alcoholysis reaction of
aluminum ethyl complex with 2-propanol were almost quan-
Figure 3. Crystal structure of complex 2b.
1
titative (Scheme 5). The H NMR spectrum of the LA oligo-
mers indicated that the polymer chains were end-capped
with an isopropyl ester and a hydroxyl group[11] and the
ring-opening occurred through a so-called coordination–in-
sertion mechanism[18] as described in Scheme 6. Because the
complexes in this paper had no chirality, it was presumed
that the polymerization followed a so-called chain-end con-
trol mechanism.[9e,19,20]
The influence of temperature on the polymerization rate
was investigated (Figure 10). The polymerization rate in-
creased with the increasing temperature, but the stereoselec-
tivity decreased. Determination of the stereochemical mi-
crostructures of PLA was achieved through inspection of the
Figure 4. Crystal structure of complex 4b.
1
methine region of homonuclear decoupled H NMR spectra
the spectral results that both 2b and 4b retained the five-co-
of the resultant polymers.[9a,c,11] The apparent polymerization
ordinated monomeric Al center as in the solid structure.
rate constant (kapp), polymerization rate constants (kp) and
Pm values are collected in Table 4. With these three pa-
[21]
Kinetic studies: The polymerization processes were investi-
gated by kinetic studies using 1a–7a. Polymerization data
were collected in Table 2. The data of conversion versus
time were collected for both the aliphatic diimine backbone
complexes (1a, 2a, 3a, 4a and 7a)[17] and aryl diimine back-
bone complexes (5a and 6a), respectively (Figures 5 and 6).
First-order kinetics in monomer was observed [Eq. (1)],
where kapp was the apparent polymerization rate constant.
rameters, the activation energy of the polymerization was
calculated by fitting them to the Arrhenius equation (kp
=
AeꢀE /RT): an activation energy Ea of 35.4 kJmolꢀ1 was de-
duced by plotting lnkp versus 1/T. The activation energy for
4a was much lower compared with that with tin(II) octa-
noate (70.9ꢁ1.5 kJmolꢀ1).[22]
a
3130
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 3126 – 3136

Polymerization of Racemic Lactide
FULL PAPER
Complex structure andstereo-
selective polymerization
Effect of the diimine backbone
length and substituent: Com-
plexes 2a, 3a, 4a and 5a have
an identical enol structure but
different diimine backbones.
Correspondingly, they displayed
different catalytic behaviors
(Table 2 entry 2, 3, 4 and 5).
Lengthening the backbone
from C2 to C3 gave rise to sig-
nificant increase in activity.
Compound 2a had C2 backbone
while 3a had C3 backbone; the
catalytic activity of 3a was
much higher than that of 2a.[11]
Apparently, the C3 backbone
had a positive effect on the
polymerization rate. Probably
the higher backbone flexibility
Figure 5. Kinetic plots of the rac-lactide conversion vs. the reaction time: a) complex 1a/2-propanol, [M]0/
[Al]=65; b) complex 4a/2-propanol, [M]0/[Al]=40; c) complex 4a/2-propanol, [M]0/[Al]=80; d) complex 4a/ was suitable for the metal coor-
2-propanol, [M]0/[Al]=110; e) complex 7a/2-propanol, [M]0/[Al]=65.
dination sphere to open the in-
serting monomer ring. To fur-
ther investigate the effect of
backbone on the catalytic activity, modification in the back-
bone of 3a was produced. The dimethyl-substituted complex
4a had the highest activity. The kp value for 4a
(1.90 Lmolꢀ1 minꢀ1) was much higher than those of 2a
(0.19 Lmolꢀ1 minꢀ1) and 3a (0.26 Lmolꢀ1 minꢀ1). Clearly the
gem-methyls in 4a were the reason for the remarkable
change in activity between 3a and 4a, presumably due to
ꢀ
the gem-methyls made the aluminum oxygen bond easier to
cleave when coordinated with the monomer resulting in
higher activity. Therefore, it could be concluded that the ali-
phatic diimine backbone length and substituent had remark-
able influence on the complexꢁs activity. Similar phenomen-
on was reported by Nomura[9l] and our group[9g] for the
salen-type catalysts: catalyst possessing 2,2-dimethyl-1,3-pro-
panediimine backbone was more active than that possessing
1,3-propanediimine backbone. Complex 5a had the lowest
stereoselectivity (Pm =0.58) among the four complexes. It
was assumed that because the aliphatic backbone was more
Figure 6. Kinetic plots of the rac-lactide conversion vs. the reaction time:
flexible than aryl analogues and therefore it could provide
more steric hindrance during the insertion of monomer.
Due to the planar aryl diimine backbone and the lack of
enough steric hindrance, aryl analogues would have lower
stereoselectivity.
a) complex 5a/2-propanol, [M]0/[Al]=56; b) complex 6a/2-propanol,
[M]0/[Al]=56.
Complexes 4a and 3a had the same t value (0.78), while
2a had 0.09. It manifested that 3a and 4a were closer to the
tbp geometry than 2a. By comparing the t values and the
stereoselectivity, maybe further deduction concerning the
geometry and the stereoselectivity could be made, that is,
the sqp geometry of a complex did not benefit the stereose-
lective polymerization of rac-lactides, and complexes con-
taining higher t values (closer to tbp geometry) would have
Figure 7. Plot of PLA Mn and polydispersity (Mw/Mn) as a function of
rac-lactide conversion using complex 4a/2-propanol, [M]0/[Al]=110.
Chem. Eur. J. 2008, 14, 3126 – 3136
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3131

X. Chen et al.
higher stereoselectivity. We temporarily assumed that the
complex containing the tbp geometry had positive synergic
effect to retain and extend the chirality caused by the chiral
monomer after the insertion. The tbp geometry effect made
the chiral recognition of incoming monomer more effective,
whereas the sqp geometry did not possess this character.[23]
A similar phenomenon was reported by Nomura[9l] and our
group[9g] for the salen-type catalysts.
Lin et al.[14] reported that the stronger metal alkoxide
bond would have a slower reaction rate in the ROP of lac-
tide compared with the Schiff base zinc complex. It was not
the case in our systems, since 2b and 4b had almost equal
aluminum alkoxide bone length [1.7443(14) for 2b and
1.740(3) for 4b] but their polymerization rate differed con-
siderably (Table 3).
Effect of the enol substituent: Complexes 1a, 2a and 7a had
the same diimine backbone, but different enol substituents.
Complex 7a had the highest activity and the highest stereo-
selectivity among the three complexes (Table 2, entries 1, 2
and 7). An electronic effect was obvious, with electron-with-
drawing substituents in the enol segments affording more
active aluminum centers, presumably a consequence of en-
hanced metal electrophilicities. Hence, the fluorine-substi-
tuted complexes typically exhibit greater polymerization ac-
tivities than do their counterparts bearing unsubstituted
enol segments. Similar conclusion were reported by Gib-
son[9k] that chloro substituents in the phenoxide unit had
higher activity and higher stereoselectivity than their di-
methyl analogues for salan-type catalysts. The increased se-
lectivity suggested that the electron-withdrawing groups did
not take a direct but a complicated way to affect the com-
plexꢁs behavior during the polymerization for both salan-
type catalysts and enolic complexes. The effect of substitu-
ent from methyl group to
Figure 8. Plot of PLA Mn and polydispersity (Mw/Mn) as a function of
rac-lactide conversion using a) complex 5a/2-propanol, [M]0/[Al]=56; b)
complex 6a/2-propanol, [M]0/[Al]=56.
Figure 9. kapp vs. the concentration of 4a/2-propanol initiator for the rac-
LA polymerization.
phenyl group was not remark-
able in comparison with the
diimine substituents (Table 2,
entries 1 and 2). Complexes 1a
and 2a had the similar activity
and stereoselectivity, and the
similar results were observed
for aryl backbone complexes
5a and 6a (Table 2, entries 5
and 6).
Stereochemistry of rac-LA: In
the case of random propaga-
tion (Bernoullian statistics),
the additions of different enan-
tiomeric monomers were inde-
pendent events, the different
rate constants were equal.
Since the ligands and com-
plexes reported were achiral,
and there was a preference for
isotactic addition during the
Scheme 6. Proposed mechanism for the polymerization of lactides using enolic Schiff base aluminum isoprop-
oxides.
3132
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 3126 – 3136

Polymerization of Racemic Lactide
FULL PAPER
[a]
Table 2. Polymerization data of rac-LA using complexes 1a–7a at 708Cin toluene.
rather than a Bernoullian statis-
tics in such a mode of propaga-
tion. PLA derived from rac-lac-
[e]
Entry Complex t [min] [M]0/[I] Conv.[b] [%]
M
nA
]
MnGPC[d] [10ꢀ3
]
PDI[d] kp
Pm
1
2
3
4
5
6
7
8
1a
2a
3a
4a
4a
5a
6a
7a
1927
1506
551
127
221
507
628
281
65
65
65
65
110
56
56
65
78
87
85
88
80
81
89
93
7.3
8.2
8.0
8.2
12.7
6.5
13.2
12.8
15.7
16.2
21.5
13.2
16.5
14.5
1.39
1.15
1.09
1.04
1.02
1.12
1.19
1.32
0.11 0.68
0.19 0.69 tide could exhibit up to five
0.26 0.75
1.90 0.78
1.90 0.78
0.36 0.58
0.41 0.57
0.92 0.75
tetrad sequences (mmm, mmr,
rmm, mrm, rmr) in relative
ratios determined by the ability
of initiators to control racemic
[r-diad] and meso [m-diad] con-
nectivity of the monomer units.
According to first-order Marko-
vian statistics, the probability
for meso linkages could be de-
termined as
7.2
8.7
[a] All polymerizations were carried out in toluene solution at 708C. [LA]0 =0.5 molLꢀ1. [b] Measured by 1H
NMR. [c] Calculated from the molecular weight of LA”[M]0/[I]”conversion + MwA(iPrOH). [d] Obtained
from GPCanalysis and calibrated against polystyrene standard. The true value of Mn could be calculated ac-
CTHREUNG
[16]
cording to formula Mn =0.58MnGPC
.
[e] kp(Lmolꢀ1 minꢀ1): Calculated from the relationship kp =kapp/[Al].
Pm ¼ km=ðkm þ krÞ
¼ kS=SS=ðkS=SS þ kS=RR
Table 3. Polymerization data of rac-LA using complexes 2b and 4b in toluene.[a]
Conv.[b] [%]
M
R
]
MnGPC[d] [10ꢀ3
]
PDI[d]
Pm
Þ
ð3Þ
Entry
Complex
t [min]
[M]0/[I]
¼ kR=RR=ðkR=SS þ kR=RR
Þ
1
2
3
4
5
6
2b
2b
4b
4b
4b
4b
1528
2140
96
140
230
335
65
75
40
65
110
150
90
93
93
90
85
90
8.4
10.1
12.8
8.4
12.8
19.5
13.5
14.7
23.5
16.3
23.5
36.8
1.14
1.20
1.03
1.05
1.03
1.04
0.68
0.66
0.78
0.78
0.78
0.78
where kS/SS and kR/RR were the
rate constants of homo-propa-
gation, kS/RR and kR/SS were the
rate constants of cross-propaga-
[a] All polymerizations were carried out in toluene solution at 708C, [LA]0 =0.5 molLꢀ1. [b] Measured by 1H
NMR. [c] Calculated from the molecular weight of LA”[M]0/[I]”conversion + MwA(iPrOH). [d] Obtained
from GPCanalysis and calibrated against polystyrene standard. The true value of Mn could be calculated ac-
CTHREUNG
tion. If kS/SS > kS/RR or kR/RR
>
[16]
kR/SS, the formation of isotactic
sequences were favored, other-
cording to formula Mn =0.58MnGPC
.
wise syndiotactic sequences
were formed. The following
equations could be deduced according to absolute reaction
rate theory:
Table 4. Kinetic results of rac-LA polymerization at different tempera-
tures using 4a/2-propanol in toluene, [M]/[I]=110.
T [8C]
kapp [minꢀ1
]
kp [Lmolꢀ1 minꢀ1
]
Pm
kS=SS ¼ kR=RR ¼ km ¼ ðkT=hÞexp½ðDS°,m=RÞꢀðDH°,m=RTފ
ð4Þ
70
90
110
0.0086
0.0141
0.0318
1.90
3.11
6.99
0.78
0.73
0.69
kR=SS ¼ kS=RR ¼ kr ¼ ðkT=hÞexp½ðDS°,r=RÞꢀðDH°,r=RTފ
ð5Þ
Further deduction of equation 6 could be obtained from
Equations (4) and (5):
Pm=ð1ꢀPmÞ ¼ km=kr
ð6Þ
¼ exp½ðDS°,mþDS°,rÞ=RꢀðDH°,mꢀDH°,rÞ=RTŠ
where (DS°,mꢀDS°,r) was the entropy difference between
Figure 10. Kinetics of the rac-LA polymerization using 4b at the reaction
homo-propagation
and
cross-propagation,
and
temperatures of a) 1108C ; b) 908C ; c) 708C , [M]/[Al]=110.
0
(DH°,mꢀDH°,r) was the enthalpy difference between homo-
propagation and cross-propagation. To determine the differ-
ROPs of rac-LA, we considered the chain-end control
mechanism.[9e,19,20] In such a reaction, the ligand and the
complex were achiral; the initiation reaction occurred with-
out any differentiation between the two enantiomers, and
the last unit in the growing polymer chain influenced which
enantiomer form of the monomers would incorporate next.
So we would prefer a first-order Markovian statistics[24–26]
ent values of (DS°,mꢀDS°
)
and (DH°,mꢀDH°,r),
,r
lnPm/A
(1ꢀPm) was plotted versus the 1/T (Figure 11). From
this plot, the entropy difference (DS°,mꢀDS°,r) of ꢀ26.6ꢁ
2.3 calKꢀ1 molꢀ1 and activation enthalpy difference
(DH°,mꢀDH°,r) of ꢀ12.7ꢁ0.8 kcalKꢀ1 molꢀ1 were obtained,
which may explain the preference of isotactic stereose-
quence.
Chem. Eur. J. 2008, 14, 3126 – 3136
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3133

X. Chen et al.
95.8 (CHCOH), 43.4 (NCH2), 28.6 (CH3COH), 18.4 ppm (CH3CN); ele-
mental analysis calcd (%) for C12H20N2O2: C64.26, H 8.99, N 12.49;
found: C64.52, H 9.20, N 12.38.12.
Ligand2 : Ligand 2 was obtained as a white crystalline solid in 80%
1
yield. H NMR (400 MHz, CDCl3): d = 7.86 (d, 4H, ArH), 7.42 (m, 6H,
ArH), 5.75 (s, 2H, CHCOH), 3.61 (d, 4H, NACHTRE(UNG CH2CH2)N), 2.09 ppm (s,
6H, CH CN); 13CNMR (100 MHz, CDCl 3): d = 188.2 (ArCOH), 164.7
3
(CH3CN), 140.1, 130.5, 128.1, 126.8 (ArC), 92.9 (CHCOH), 43.7 (NCH2),
19.1 ppm (CH3CN); elemental analysis calcd (%) for C22H24N2O2: C
75.83, H 6.94, N 8.04; found: C75.97, H 7.21, N 8.11.
Figure 11. Relationship between polymerization temperature and stereo-
chemistry of the resulting poly(rac-LA)s by using 4a/2-propanol.
Ligand3 : Ligand 3 was obtained as a white crystalline solid in 77%
AHCTREUNG
1
yield. H NMR (400 MHz, CDCl3): d = 7.87 (d, 4H, ArH), 7.43 (m, 6H,
ArH), 5.70 (s, 2H, CHCOH), 3.53 (m, 4H, N(CH2CH2CH2)N), 2.17 (s,
6H, CH CN), 2.04 ppm (t, 2H, N(CH2CH2CH2)N); 13CNMR (100 MHz,
3
CDCl3): d = 187.7 (ArCOH), 164.9 (CH3CN), 140.1, 130.3, 128.0, 126.7
Conclusion
(ArC),
92.3
(CHCOH),
39.8
(N(CH2CH2CH2)N),
29.9
(N(CH2CH2CH2)N), 19.1 ppm (CH3CN); elemental analysis calcd (%)
for C23H26N2O2: C76.21, H 7.23, N 7.73; found: C76.19, H 7.21, N 7.86.
We have synthesized a series of five-coordinated enolic
Schiff base aluminum complexes derived from b-diketone
and diamine. Their catalytic behavior in racemic-lactide
polymerization varied remarkably. A plausible mechanism
of polymerization was proposed. Complex 4b had the high-
est activity and stereoselectivity among all complexes. The
different performance of these complexes was attributed to
the different diimine backbone and diketone substituent
groups on auxiliary ligands. There was a combinatorial
effect of electronic and steric factors. Because the complexes
in this paper had no chirality, it was presumed that the poly-
merization followed a so-called chain-end control mecha-
nism. Researches toward the origin of the activity and ste-
reoselectivity of enolic Schiff base aluminum complexes are
currently in progress.
Ligand4 : Ligand 4 was recrystallized in hexane and obtained as a white
crystalline solid in 51% yield. 1H NMR (400 MHz, CDCl3): d = 7.85 (d,
4H, ArH), 7.41 (m, 6H, ArH), 5.70 (s, 2H, CHCOH), 3.30 (d, 4H,
N
N
C
ACHTREUNG
A
d
=
(ArCOH), 165.1 (CH3CN), 140.2, 130.3, 128.0, 126.7 (ArC), 92.3
(CHCOH), 50.4 (N(CH2C(CH3)2CH2)N), 35.6 (N(CH2C(CH3)2CH2)N),
23.5 (N(CH2C(CH3)2CH2)N), 19.3 ppm (CH3CN); elemental analysis
A
G
A
ACHTREUNG
A
ACHTREUNG
calcd (%) for C25H30N2O2: C76.89, H 7.74, N 7.17; found: C77.05, H
7.65, N 7.22.
Ligand5 : Ligand 5 was obtained as a yellow crystalline solid in 76%
1
yield. H NMR (400 MHz, CDCl3): d = 7.92 (d, 4H, ArH), 7.47 (m, 8H,
ArH), 7.26 (s, 2H, ArH), 5.72 (s, 2H, CHCOH), 4.60 (d, 4H,
NCH2C6H4CH2N), 2.11 ppm (s, 6H, CH3CN); 13CNMR (100 MHz,
CDCl3): d = 188.4 (ArCOH), 165.3 (CH3CN), 140.7, 139.0, 131.0, 130.0,
128.6, 127.6, 126.5, 125.9 (ArC), 93.1 (CHCOH), 47.3 (NCH2C6H4CH2N),
19.9 ppm (CH3CN); elemental analysis calcd (%) for C28H28N2O2: C
79.22, H 6.65, N 6.60; found: C78.96, H 6.56, N 6.46.
Ligand6 : Ligand 6 was obtained as a yellow crystalline solid in 72%
yield. 1H NMR (400 MHz, CDCl3): d = 7.35 (t, 1H, ArH), 7.21 (d, 2H,
ArH), 7.15 (s, 1H, ArH), 5.08 (s, 2H, CHCOH), 4.49 (d, 4H,
NCH2C6H4CH2N), 2.06 (s, 6H, CH3COH), 1.94 ppm (s, 6H, CH3CN);
Experimental Section
General: All experiments were carried out under argon using Schlenk
techniques. Starting materials for the synthesis of ligand 1–7 were pur-
chased from Aldrich Inc. and used without further purification. Toluene
was distilled from Na/benzophenone. Ethyl acetate and 2-propanol were
distilled from CaH2 under the protection of argon. rac-Lactide (Purac)
was purified by recrystallization from ethyl acetate and dried under
vacuum at room temperature (RT) before use. NMR spectra were re-
corded on Bruker AV 300m and Bruker AV 400m in CDCl3 at 258C.
Chemical shifts were given in parts per million from tetramethylsilane.
Gel permeation chromatography (GPC) measurements were conducted
13CNMR (100 MHz, CDCl 3): d
= 195.4 (CH3COH), 163.2 (CH3CN),
139.1, 129.6, 126.1, 125.4 (ArC), 96.2 (CHCOH), 46.8 (NCH2C6H4CH2N),
29.2 (CH3COH), 19.1 ppm (CH3CN); elemental analysis calcd (%) for
C18H24N2O2: C71.79, H 8.05, N 9.33; found: C71.92, H 7.89, N 9.20.
Ligand7 : Ligand 7 was obtained as a yellow crystalline solid in 63%
yield. 1H NMR (400 MHz, CDCl3): d = 5.43 (s, 2H, CHCOH), 3.68 (d,
(CH2CH2)N), 2.15 ppm (s, 6H, CH3CN); 13CNMR (100 MHz,
4H, NACHTREUNG
[D6]DMSO): d = 174.9 (CF3COH), 171.2 (CH3CN), 118.0 (CF3COH),
89.2 (CHCOH), 43.6 (NCH2CH2N), 18.7 ppm (CH3CN); 19F NMR
(400 MHz, CDCl3): d = ꢀ0.40 ppm (s, 6F); elemental analysis calcd (%)
for C12H14F6N2O2: C43.38, H 4.25, N 8.43; found: C43.62, H 4.38, N
8.60.
with a Waters 515 GPC with CHCl3 as the eluent (flow rate: 1 mLminꢀ1
,
at 358C). The molecular weights were calibrated against polystyrene (PS)
standards. Crystallographic data were collected on a Bruker APEX CCD
diffractometer with graphite-monochromated MoKa radiation (l=
0.71073 Š) at 187 K. The structure was refined by the full-matrix least-
squares method on F2 using the SHELXTL-97 crystallographic software
package. Anisotropic thermal parameters were used to refine all nonhy-
drogen atoms. Hydrogen atoms were located in idealized positions.
Complex synthesis (General Procedure): For enolic Schiff base aluminum
ethyl complexes 1a–7a, AlEt3 (0.2 mmol) in toluene (5 mL) was added
to the stirred 1 mL toluene solution of ligand precursors 1–7 (0.2 mmol)
at RT. The reaction was maintained at 808Cfor 12 h, and the reaction
mixture was then slowly cooled to RT. The toluene was removed under
vacuum. For enolic Schiff base aluminum isopropoxides, 2b and 4b were
prepared by mixing the corresponding ethyl complexes with 2-propanol
in toluene solution, the mixture was stirred at 808Cfor 10 h and then
slowly cooled to RT. The toluene was removed under vacuum. Crystals
suitable for X-ray diffraction were grown from a mixture of toluene and
pentane at ꢀ108C. The crystallographic data and the results of refine-
ments were summarized in Table 1.
Ligandsynthesis
(General Procedure):
A
solution of diamine
(0.1 molLꢀ1) in ethanol (50 mL) was added dropwise to a stirred solution
of diketone (0.2 molLꢀ1) in ethanol (50 mL). The reaction mixture was
refluxed for 14 h before cooling to RT. After removal of the solvent
under vacuum a crystalline solid was produced and purified by recrystal-
lization in ethanol.
Ligand1 : Ligand 1 was obtained as a white crystalline solid in 85%
yield. 1H NMR (400 MHz, CDCl3): d = 5.00 (s, 2H, CHCOH), 3.43 (d,
Complex 1a: Obtained as
(400 MHz, CDCl3):
(CH2CH2)N), 3.42 (m, 2H, NACHTREUNG
a
white solid in 95% yield. 1H NMR
5.02 (s, 2H, CHCOAl), 3.50 (m, 2H,
(CH2CH2)N), 2.03 (s, 6H, CH3COAl),
4H, N
A
d =
13CNMR (100 MHz, CDCl 3): d
=
N
A
3134
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 3126 – 3136

Polymerization of Racemic Lactide
FULL PAPER
1.96 (s, 6H, CH3CN), 0.85 (t, 3H, AlCH2CH3), ꢀ0.39 ppm (q, 2H,
AlCH2CH3); 13C NMR (100 MHz, CDCl3): d = 179.4 (CH3COAl), 171.4
(CH3CN), 98.8 (CHCOH), 45.7 (NCH2), 25.5 (CH3COAl), 21.3
(CH3CN), 13.6 (AlCH2CH3), 9.6 ppm (AlCH2CH3); elemental analysis
calcd (%) for C14H23AlN2O2: C60.41, H 8.33, N 10.06; found: C60.20, H
8.61, N 9.87.
(CH3CN), 138.6, 129.9, 128.1, 126.8 (ArC), 97.0 (CHCOAl), 62.7 (OCH-
ACHRTEGNU(CH3)2), 46.4 (NCH2), 27.5 (OCHACHTRE(UGN CH3)2), 22.5 ppm (CH3CN); elemental
analysis calcd (%) for C25H29AlN2O3: C69.43, H 6.76, N 6.48; found: C
69.60; H,6.82, N 6.35.
Complex 4b: Obtained as a yellow solid in quantitative yield. 1H NMR
(400 MHz, CDCl3): d = 7.83 (m, 4H, ArH), 7.29 (m, 6H, ArH), 5.74 (s,
Complex 2a: Obtained as
(400 MHz, CDCl3): d = 7.90 (d, 4H, ArH), 7.32 (m, 6H, ArH), 5.71 (s,
2H, CHCOAl), 3.53 (m, 2H, (CH2CH2)N), 3.45 (m, 2H,
(CH2CH2)N), 2.04 (s, 6H, CH3CN), 0.81 (t, 3H, AlCH2CH3),
a
yellow solid in 93% yield. 1H NMR
2H, CHCOAl), 3.94 (m, 1H, OCH
(CH3)2CH2)N), 3.35 (d, 2H, (CH2C
CH3CN), 1.04 (d, 6H, OCH(CH3)2), 1.00 (s, 3H, N
0.88 ppm (s, 3H, N(CH2C
(CH3)2CH2)N); 13CNMR (100 MHz, CDCl 3): d
= 172.5 (ArCOAl), 172.1 (CH3CN), 138.4, 129.8, 128.0, 126.9 (ArC), 97.4
(CHCOAl), 62.9 (OCH(CH3)2), 59.0 (N(CH2C(CH3)2CH2)N), 38.2
(N(CH2C(CH3)2CH2)N), 26.4 (N(CH2C(CH3)2CH2)N), 24.1 (OCH-
(CH3)2), 22.5 ppm (CH3CN); elemental analysis calcd (%) for
A
NACHTREUNG(CH2C-
A
N
A
ACHTREUNG
N
A
A
N
ACHTREUNG
N
A
A
ACHTREUNG
ꢀ0.38 ppm (q, 2H, AlCH2CH3); 13CNMR (75 MHz, CDCl 3): d = 173.6
(ArCOH), 172.3 (CH3CN), 138.8, 129.9, 128.0, 126.9 (ArC), 96.6
(CHCOAl), 46.4 (NCH2), 22.4 (CH3CN), 10.1 ppm (AlCH2CH3); elemen-
tal analysis calcd (%) for C24H27AlN2O2: C71.62, H 6.76, N 6.96; found:
C71.72, H 6.58, N 7.11.
A
N
ACHTREUNG
A
G
A
ACHTREUNG
AHCTREUNG
C28H35AlN2O3: C70.86, H 7.43, N 5.90; found: C70.59, H 7.12, N 5.75.
Complex 3a: Obtained as
(400 MHz, CDCl3): d = 7.77 (d, 4H, ArH), 7.24 (m, 6H, ArH), 5.79 (s,
2H, CHCOAl), 3.63 (b, 2H, N(CH2CH2CH2)N), 3.41 (b, 2H,
a
yellow solid in 96% yield. 1H NMR
Polymerization of rac-LA Under the protection of argon, the rac-LA
(22.4 mmol, 3.22 g), complexes 1a–7a, (0.30 mmol in 2 mL of toluene), 2-
propanol (0.30 mmol, in 5 mL of toluene), and toluene (38 mL) were
added to a dried reaction vessel equipped with a magnetic stirring bar.
The vessel was placed in an oil bath at 708C. Conversion of the monomer
was determined on the basis of H NMR spectroscopic studies. The poly-
mers were isolated by precipitation into cold methanol, then filtrated and
dried under vacuum at RT for 24 h.
N(CH2CH2CH2)N),
2.03
(s,
6H,
CH3CN),
1.91
(t,
2H,
N(CH2CH2CH2)N), 0.95 (t, 3H, AlCH2CH3), ꢀ0.06 ppm (q, 2H,
AlCH2CH3); 13CNMR (75 MHz, CDCl 3): d = 173.0 (CH3COAl), 172.3
(CH3CN), 138.7, 129.7, 127.9, 126.9 (ArC), 97.1 (CHCOAl), 49.4
(N(CH2CH2CH2)N), 26.2 (N(CH2CH2CH2)N), 22.06 (CH3CN), 10.4 ppm
(AlCH2CH3); elemental analysis calcd (%) for C25H29AlN2O2: C72.09, H
7.02, N 6.73; found C, 71.91, H 7.15, N 6.82.
1
CCDC 659939, 659940, 659941, 659942, and 659943, contain the supple-
mentary crystallographic data for this paper. These data can be obtained
free of charge from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif.
Complex 4a: Obtained as
(400 MHz, CDCl3): d = 7.93 (d, 4H, ArH), 7.36 (m, 6H, ArH), 5.84 (s,
2H, CHCOAl), 3.39 (s, 4H, (CH2C(CH3)2CH2)N), 2.14 (s, 6H,
CH3CN), 1.09 (s, 3H, N(CH2C(CH3)2CH2)N), 0.97 (s, 3H, N(CH2C-
a
yellow solid in 98% yield. 1H NMR
N
A
ACHTREUNG
G
A
ACHTREUNG
A
13CNMR (75 MHz, CDCl 3): d = 172.6, 171.8 (ArCOAl), 171.1, 170.2
Acknowledgements
(CH3CN), 139.4, 129.9, 128.2, 126.2 (ArC), 97.5 (CHCOAl), 58.2
(NACHTREUNG(CH2CACHTREUNG(CH3)2CH2)N), 38.4 (NAHCERTNU(G CH2CACHTERNG(U CH3)2CH2)N), 26.3 (NAHCTRE(UNG CH2C-
This project is financially supported by the National Natural Science
Foundation of China (Project No: 20574066), National Fund for Distin-
guished Young Scholar (No: 50425309), and the International Coopera-
tion fund of Science and Technology (Key project 2005DFA50290).
A
(AlCH2CH3); elemental analysis calcd (%) for C27H33AlN2O2: C72.95, H
7.48, N 6.30; found C, 72.63, H 7.85, N 6.05.
Complex 5a: Obtained as
(400 MHz, CDCl3): d = 7.84 (d, 4H, ArH), 7.36 (m, 8H, ArH), 7.24 (d,
2H, ArH), 5.76 (s, 2H, CHCOAl), 4.64 (d, 4H, NCH2ArCH2N), 2.35 (s,
[1] a) K. E. Uhrich, S. M. Cannizzaro, R. S. Langer, K. M. Shakesheff,
Chem. Rev. 1999, 99, 3181; b) D. J. Mooney, G. Organ, J. P. Vacanti,
R. Langer, Cell Transplant. 1994, 2, 203; c) J. L. Eguiburu, M. J. Fer-
nandez-Berridi, F. P. Cossio, J. San Roman, Macromolecules 1999,
32, 8252.
[2] a) A.-C. Albertsson, I. K. Varma, Biomacromolecules 2003, 4, 1466;
b) B. J. OꢁKeefe, M. A. Hillmyer, W. B. Tolman, J. Chem. Soc.
Dalton Trans. 2001, 2215; c) Z. Y. Zhong, P. J. Dijkstra, C. Birg, M.
Westerhausen, J. Feijen, Macromolecules 2001, 34, 3863; d) D. Take-
uchi, T. Nakamura, T. Aida, Macromolecules 2000, 33, 729; e) B. J.
OꢁKeefe, L. E. Breyfogle, M. A. Hillmyer, W. B. Tolman, J. Am.
Chem. Soc. 2002, 124, 4384.
[3] a) R. E. Drumright, P. R. Gruber, D. E. Henton, Adv. Mater. 2000,
12, 1841; b) E. Chiellini, R. Solaro, Adv. Mater. 1996, 8, 305.
[4] a) M. S. Reeve, S. P. McCarthy, M. J. Downey, R. A. Gross, Macro-
molecules 1994, 27, 825; b) J. R. Sarasua, ; R. E. Prud’homme, M.
Wisniewski, A. Le Borgne, N. Spassky, Macromolecules 1998, 31,
3895; R. E. Prud’homme, M. Wisniewski, A. Le Borgne, N. Spassky,
Macromolecules 1998, 31, 3895.
[5] a) M. Endo, T. Aida, S. Inoue, Macromolecules 1987, 20, 2982; b) A.
Duda, Z. Florjanczyk, A. Hofman, S. Slomkowski, S. Penczek, Mac-
romolecules 1990, 23, 1640; c) T. Aida, S. Inoue, Acc. Chem. Res.
1996, 29, 39; d) A. Kowalski, A. Duda, S. Penczek, Macromol.
Rapid Commun. 1998, 19, 567; e) B. T. Ko, C. C. Lin, Macromole-
cules 1999, 32, 8296.
6H, CH CN), 0.91 (t, 3H, AlCH2CH3), ꢀ0.21 ppm (q, 2H, AlCH2CH3);
3
13C NMR (100 MHz, CDCl3): d
= 177.8 (ArCOH), 173.5 (CH3CN),
138.3, 138.0, 129.5–124.9 (ArC), 98.0 (CHCOAl), 51.6 (NCH2C6H4CH2N),
22.6 (CH3CN), 9.5 (AlCH2CH3), 0.2 ppm (AlCH2CH3); elemental analy-
sis calcd (%) for C30H31AlN2O2: C75.29, H 6.53, N 5.85; found C75.40,
H 6.88, N 6.13.
Complex 6a: Obtained as
a
yellow solid in 92% yield. 1H NMR
(300 MHz, CDCl3): d = 7.28 (m, 1H, ArH), 7.26 (m, 2H, ArH), 7.09 (m,
1H, ArH), 4.89 (s, 2H, CHCOAl), 4.48 (d, 4H, NCH2ArCH2N), 2.03 (s,
6H, CH COAl), 1.92 (s, 6H, CH3CN), 0.90 (t, 3H, AlCH2CH3),
3
ꢀ0.27 ppm (q, 2H, AlCH2CH3); 13CNMR (100 MHz, CDCl 3): d = 179.8
(CH3COAl), 173.4 (CH3CN), 140.0, 128.7, 125.3, 124.7 (ArC), 99.8
(CHCOAl), 52.4 (NCH2C6H4CH2N), 25.7 (CH3COAl), 21.2 (CH3CN),
9.35 (AlCH2CH3), ꢀ0.71 ppm (AlCH2CH3); elemental analysis calcd (%)
for C20H27AlN2O2: C67.78, H 7.68, N 7.90; found: C68.07, H 7.44, N
7.52.
Complex 7a: Obtained as
(400 MHz, CDCl3):
a
yellow solid in 88% yield. 1H NMR
d
= 5.57 (s, 2H, CHCOAl), 3.64 (m, 2H,
NACHTREUNG(CH2CH2)N), 3.55 (m, 2H, NAHCTRE(UGN CH2CH2)N), 2.14 (s, 6H, CH3CN), 0.85
(t, 3H, AlCH2CH3), ꢀ0.34 ppm (q, 2H, AlCH2CH3); 13CNMR
(100 MHz, CDCl3):
d = 161.7 (CF3COH), 129.0 (CH3CN), 119.4
(CF3COH), 97.4 (CHCOH), 46.6 (NCH2CH2N), 21.8 (CH3CN), 9.3
(AlCH2CH3), 1.0 ppm (AlCH2CH3); elemental analysis calcd (%) for
C14H17AlF6N2O2: C43.53, H 4.44, N 7.25; found: C43.63, H 4.21, N 7.65.
Complex 2b: Obtained as a yellow solid in quantitative yield. 1H NMR
[6] a) A. Kowalski, A. Duda, S. Penczek, Macromolecules 2000, 33, 689;
b) H. R. Kricheldorf, I. Kreiser-Saunders, A. Stricker, Macromole-
cules 2000, 33, 702.
[7] a) Y. Ikada, K. Jamshidi, H. Tsuji, S. H. Hyon, Macromolecules
1987, 20, 904; b) H. Tsuji, Y. Ikada, Polymer 1999, 40, 6699.
(400 MHz, CDCl3): d = 7.34 (d, 4H, ArH), 7.17 (m, 6H, ArH), 5.79 (s,
2H, CHCOAl), 3.83 (sep, 1H, OCH
3.50 (m, 2H, N(CH2CH2)N), 2.05 (s, 6H, CH3CN), 0.97 ppm (d, 6H,
OCH
(CH3)2); 13C NMR (100 MHz, CDCl3): d = 173.3 (ArCOAl), 172.7
ACHRTUG(NE CH3)2), 3.69 (m, 2H, NAHCTRE(UGN CH2CH2)N),
R
ACHTREUNG
Chem. Eur. J. 2008, 14, 3126 – 3136
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3135

X. Chen et al.
[8] N. Spassky, M. Wisniewski, C. Pluta, A. Le Borgne, Macromol.
Chem. Phys. 1996, 197, 2627.
[17] Kinetic studies of 2a and 3a have been reported, see ref. [11].
[18] a) H. R. Kricheldorf, S. R. Lee, S. Bush, Macromolecules 1996, 29,
1375; b) W. M. Stevels, P. J. Dijkstra, J. Feijen, Trends Polym. Sci.
1997, 5, 300; c) P. Dubois, C. Jacobs, R. Jerome, P. Teyssie, Macro-
molecules 1991, 24, 2266; d) A. Kowalski, A. Duda, S. Penczek,
Macromolecules 1998, 31, 2114; e) A. Kowalski, A. Duda, S. Penc-
zek, Macromolecules 2000, 33, 689.
[19] A. Amgoune, C. M. Thomas, T. Roisnel, J.-F. Carpentier, Chem.
Eur. J. 2006, 12, 169.
[20] B. M. Chamberlain, M. Cheng, D. R. Moore, T. M. Ovitt, E. B. Lob-
[9] a) T. M. Ovitt, G. W. Coates, J. Am. Chem. Soc. 2002, 124, 1316;
b) C. P. Radano, G. L. Baker, M. R. Smith, J. Am. Chem. Soc. 2000,
122, 1552; c) Z. Y. Zhong, P. J. Dijkstra, J. Feijen, Angew. Chem.
2002, 114, 4692; Angew. Chem. Int. Ed. 2002, 41, 4510; d) Z. Y.
Zhong, P. J. Dijkstra, J. Feijen, J. Am. Chem. Soc. 2003, 125, 11291;
e) N. Nomura, R. Ishii, M. Akakura, K. Aoi, J. Am. Chem. Soc.
2002, 124, 5938; f) M. Wisniewski, A. LeBorgne, N. Spassky, Macro-
mol. Chem. Phys. 1997, 198, 1227; g) Z. H. Tang, X. S. Chen, Y. K.
Yang, X. Pang, J. R. Sun, X. F. Zhang, X. B. Jing, J. Polym. Sci. Part
A: Polym. Chem. 2004, 42, 5974; h) Z. H. Tang, X. S. Chen, X. Pang,
Y. K. Yang, X. F. Zhang, X. B. Jing, Biomacromolecules 2004, 5, 965;
i) H. Y. Ma, G. Melillo, L. Oliva, T. P. Spaniol, U. Englert, J. Okuda,
Dalton. Trans. 2005, 721; j) K. Majerska, A. Duda, J. Am. Chem.
Soc. 2004, 126, 1026; k) P. Hormnirun, E. L. Marshall, V. C. Gibson,
A. J. P. White, D. J. Williams, J. Am. Chem. Soc. 2004, 126, 2688;
l) R. Ishii, N. Nomura, T. Kondo, Polym. J. 2004, 36, 261.
[10] a) S. A. Schuetz, C. M. Silvernail, C. D. Incarvito, A. L. Rheingold,
J. L. Clark, V. W. Day, J. A. Belot, Inorg. Chem. 2004, 43, 6203;
b) S. A. Schuetz, V. W. Day, R. D. Sommer, A. L. Rheingold, J. A.
Belot, Inorg. Chem. 2001, 40, 5292.
[11] X. Pang, H. Z. Du, X. S. Chen, X. L. Zhuang, D. M. Cui, X. B. Jing,
J. Polym. Sci. Part A: Polym. Chem. 2005, 43, 6605.
[12] P. J. McCarthy, A. E. Martell, Inorg. Chem. 1967, 6, 781.
[13] a) D. A. Atwood, M. J. Harvey, Chem. Rev. 2001, 101, 37; b) M.-A.
Munoz-Hernandez, T. S. Keizer, P. Wei, S. Parkin, D. A. Atwood,
Inorg. Chem. 2001, 40, 6782.
[14] H. Y. Chen, H. Y. Tang, C. C. Lin, Macromolecules 2006, 39, 3745.
[15] R. Benn, A. Rufinska, H. Lemkuhl, E. Janssen, C. Kruger, Angew.
Chem. 1983, 95, 808; Angew. Chem. Int. Ed. Engl. 1983, 22, 779.
[16] J. Baran, A. Duda, A. Kowalski, R. Szymanski, S. Penczek, Macro-
mol. Rapid Commun. 1997, 18, 325.
kovsky, G. W. Coates, J. Am. Chem. Soc. 2001, 123, 3229.
2
[21] Pm is the probability of meso linkages, according to [mmm]=Pm
+
(1ꢀPm)Pm/2,
[mmr]=[rmm]=(1ꢀPm)Pm/2,
[rmr]=(1ꢀPm)2/2,
[mrm]=[(1ꢀPm)2 +P
(1ꢀPm)]/2.
mA
[22] D. R. Witzke, R. Narayan, J. J. Kolstad, Macromolecules 1997, 30,
7075.
[23] 3a and 4a had same t value and almost same Pm value; 2a con-
tained lower t value (0.09) and had lower Pm (see Table 2). Further
evidence came from the t values of real initiators 2b (t value 0.54)
and 4b (t value 0.85) and the Pm values of 2b (Pm value 0.69) and
4b (Pm value 0.78).
[24] a) K. A. M. Thakur, R. T. Kean, E. S. Hall, J. J. Kolstad, E. J.
Munson, Macromolecules 1998, 31, 1487; b) K. A. M. Thakur, R. T.
Kean, M. T. Zell, B. E. Padden, E. J. Munson, Chem. Commun.
1998, 17, 1913; c) M. H. Chisholm, S. S. Iyer, M. E. Matison, D. G.
McCollum, M. Pagel, Chem. Commun. 1997, 20, 1999.
[25] a) M. Wisniewski, A. LeBorgne, N. Spassky, Macromol. Chem. Phys.
1997, 198, 1227; b) J. E. Kasperczyk, Macromolecules 1996, 29, 3937.
[26] P. J. Hocking, R. H. Marchessault, Macromolecules 1995, 28, 6401.
Received: September 17, 2007
Published online: January 29, 2008
3136
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 3126 – 3136