Synthesis and characterization of a brush-like copolymer of polylactide grafted onto chitosan

Article abstract of DOI:10.1016/j.carres.2003.09.013

A brush-like poly(DL)-lactide grafted onto chitosan as the backbone was investigated. The graft copolymerization was carried out with triethylaluminum as catalyst in toluene at 70°C. It was found that a greater lactide content in the feeding ratio results in a higher grafting percentage. FTIR spectrometry, ¹H NMR, DSC scanning, and wide-angle X-ray scattering, respectively, are used to characterize these branch copolymers. A copolymer has a definite melting point when the molar feeding ratio of lactide to chitosan is more than 10:1, and the ΔH of the copolymers increases with the feed ratio of lactide to chitosan in feeding.

Full text of DOI:10.1016/j.carres.2003.09.013

Carbohydrate
RESEARCH
Carbohydrate Research 339 (2004) 845–851
Synthesis and characterization of a brush-like
copolymer of polylactide grafted onto chitosan
Y. Liu,* F. Tian and K. A. Hu
State Key Laboratory of Metal Matrix Composites, Shanghai Jiao Tong University, Shanghai 200030, PRChina
Received 17 June 2003; accepted 10 September 2003
Abstract—Abrush-like poly( DL)-lactide grafted onto chitosan as the backbone was investigated. The graft copolymerization was
carried out with triethylaluminum as catalyst in toluene at 70 ꢀC. It was found that a greater lactide content in the feeding ratio
results in a higher grafting percentage. FTIR spectrometry, 1H NMR, DSC scanning, and wide-angle X-ray scattering, respectively,
are used to characterize these branch copolymers. Acopolymer has a definite melting point when the molar feeding ratio of lactide
to chitosan is more than 10:1, and the DH of the copolymers increases with the feed ratio of lactide to chitosan in feeding.
ꢁ 2003 Published by Elsevier Ltd.
Keywords: Chitosan; Polylactide; Graft copolymer
1. Introduction
been paid to its biomedical, ecological, and industrial
application in the past decades.
Biodegradable materials are becoming increasingly
more important in the biomaterials field, particularly for
wound healing, tissue reconstruction, and controlled
drug delivery. The main advantage of biodegradable
over nonbiodegradable materials is the disappearance of
implanted foreign materials from the body as a result of
their biodegradation. Most synthetic biodegradable
materials such as polylactide and its copolymers with
polyglycolide and polycaprolactone have high initial
strengths, while natural products like chitin and its de-
rivatives are low in mechanical strength but exhibit ex-
cellent cell adhesion.1
Chitosan and its derivatives have been reported to be
useful for biomedical applications such as wound heal-
ing and dressings, drug delivery agents, anti-choles-
terolemic agents, blood anti-coagulants, anti-tumor
agents, and immunoadjuvants.2
Polylactide and its copolymers form a family of bio-
gradable and biocompatible polymers that have been
widely used in biomedical applications, such as ab-
sorbable sutures, sustained drug delivery systems, im-
plants for orthopedic devices and absorbable fibers.4;5
Early studies on polylactide showed that it was readily
resorbed by living tissues.5 However, the high crystal-
linity and low hydrophilicity of polylactide reduce its
degradation rate, which results in poorer soft tissue
compatibility.5;6 Apossible way to solve the problem is
by copolymerizing the lactide with other biological
materials such as chitin and its derivatives. By inducing
the molecular chains of chitin and its derivatives, the
copolymer may be able to be degraded faster than
polylactide, yet retain its initial strength.
Chitosan, the fully or partially deacetylated form of
chitin, the principal component of living organisms such
as fungi and crustaceans, contains 2-acetamido-2-deoxy-
b-
D
-glucopyranose and 2-amino-2-deoxy-b-D-glucopyr-
anose groups. This polymer is known to be nontoxic,
odorless, and biocompatible in animal tissues, as well as
being enzymatically biodegradable. Much attention has
Chitosan is degraded via the cleavage of the b-gly-
cosidic bond between the N-acetylglucosamine units.3
Since chitosan is more hydrophilic than polylactide, the
chitosan:polylactide copolymers may exhibit different
* Corresponding author. Tel.: +86-21-62933751; fax: +86-21-62822012;
e-mail: yliu806@263.net
0008-6215/$ - see front matter ꢁ 2003 Published by Elsevier Ltd.
doi:10.1016/j.carres.2003.09.013

846
Y. Liu et al. / Carbohydrate Research 339 (2004) 845–851
degrees of hydration depending on the ratio chito-
san:polylactide in the copolymers. In particular, hydra-
tion plays an important role in polymer degradation via
hydrolysis of the ester backbone.
carried out with a Perkin–Elmer DSC7 differential
scanning calorimeter under N2 purge, at a heating rate
of 10 ꢀC/min. Temperature and heating flow calibration
were performed with indium and gallium. The scanning
range was from )50–200 ꢀC.
Very little has been reported about the copolymer-
ization of polylactide and chitin or its derivatives. In this
work, we synthesized a kind of novel graft copolymer
with the natural polysaccharide chitosan as the main
chain and the artificial biopolymer poly(DL)-lactide as
the side chains using Et3Al as the catalyst in toluene.
The relationships of comonomers and the resulting
grafting copolymer microstructure and thermal proper-
ties are reported. Amechanism for the copolymerization
reaction is also advanced.
2.5. Wide-angle X-ray scattering (WAXS)
WAXS was recorded on a Rigaku Dmax-rC diffrac-
tometer, in which the high-intensity monochromatic Ni-
filtered CuKa radiation was generated at 40 kV and
100 mA.
2.6. Synthesis of lactide grafted to chitosan
Polymerizations were carried out under stirring for 24 h
in toluene at 70 ꢀC. DL-Lactide does not dissolve below
the temperature in toluene. Chitosan (5 mmol, 100 mesh)
was degassed for 1 h in vacuo below 1 mm Hg, and
added to the reactor. The chitosan was suspended in
toluene (40 mL). Et3Al (3.8 g, 5 mmol 15% solution in
toluene) was added dropwise via a syringe through a
rubber septum under vigorous stirring at ambient tem-
perature in an apparatus equipped with an oil valve to a
gas buret. After ethane evolution ceased, the reaction
was kept for one additional hour under stirring. The
lactide with various molar ratios to chitosan (2:1, 5:1,
6:1, 10:1, 20:1, 40:1) was transferred into the reactor
after it was placed under a vacuum below 1 mm Hg for
1 h. The reactor was placed into a preheated oil bath at
70 ꢀC, and the reaction was allowed to proceed for 24 h.
When the setting reaction time of 24 h elapsed, the re-
actor was cooled to room temperature. The mixture was
precipitated in alcohol (300 mL). The precipitated
product was filtered out with a sintered glass funnel. The
grafting copolymers were twice washed with EtOAc, and
then extracted in a Soxhlet extractor with the same
solvent for 8 h. All samples were dried in an oven at
40 ꢀC for 48 h in vacuo.
2. Experimental
2.1. Materials
DL-Lactide was synthesized by condensation polymer-
ization of DL-lactic acid in vacuum firstly, and then de-
composing of condensed oligomer into lactide at high
temperature, using nanometer ZnO powder as catalyst.
The raw lactide was purified by recrystallization from
dried EtOAc. Its melting point was between 124 and
126 ꢀC. Triethylaluminum in toluene (15%) was pur-
chased from Tokyo Kasei and used without further
purification. Chitosan was obtained by refluxing chitin
in 40% NaOH and for 4 h at 110 ꢀC, and then washing to
neutral with distilled water. The degree of deacetylation
(94%) and viscosity average molecular weight (1 · 106)
were determined by elemental analyses and ½gŠ ¼
6:589 Â 10À3Mv0:88, respectively. Toluene and EtOAc
were dried by refluxing over CaH2 and purified by dis-
tillation.
2.2. NMR spectroscopy
1H NMR spectra of grafted copolymers were recorded
with a Bruker AM 400 400 MHz spectrometer. The
spectrum of chitosan was recorded on a Gemini-2000 at
300 MHz. In both cases, CF3COOD–D2O were used as
solvent.
3. Results and discussion
3.1. Synthesis and characterization
2.3. FTIR spectroscopy
Functional aluminum alkoxides such as (C2H5)3Àq
-
Al(ORX)q where X can be halogen atoms, tertiary am-
ines, and double bonds have been proven to be very
efficient in the ring-opening polymerization of end-
reactive lactones, such as caprolactone7 and lactide.8
Compounds with a hydroxyl group reacting with Et3Al
could be used to obtain the alkoxides. There is a primary
hydroxyl group (C-6) and a secondary one (C-3) in
chitosan molecules, respectively. The Et3Al can react
with the hydroxyl groups and can also initiate the
polymerization of lactide.
FTIR spectra were obtained with a Bruker 55 spec-
trometer, using KBr pellets.
2.4. Differential scanning calorimetry
The melting temperature and the melting enthalpy
(DHm) were determined by differential scanning calori-
metry. The melting points were defined as the peak of
the endothermal curve. The DSC measurements were

Y. Liu et al. / Carbohydrate Research 339 (2004) 845–851
847
Table 1. The lactide grafted onto chitosan copolymer in toluene at 70 ꢀC with Et3Al as catalyst
Grafting percentage (%)a
FLA=Fchitosan
DH (cal/g)
T (ꢀC)
b
Molar ratio (lactide:chitosan)
Total yield (%)
2:1
5:1
50.4
38.1
32.8
43.8
26.7
15.2
43
109
136
338
408
462
0.96
2.44
3.76
7.56
9.12
6:1
10:1
20:1
40:1
2.3359
3.1186
3.3055
150.65
149.76
138.95
10.33
a
graft copolymer ðgÞ À chitosan ðgÞ
chitosan ðgÞ
Grafting percentage ð%Þ ¼
 100%.
161
.
72
bMolar composition in graft copolymer ¼ grafting percentage Â
The polymerization of lactide onto chitosan in various
molar ratios took place easily under the designated
conditions (see Experimental section). The results are
shown in Table 1.
ratio of lactide to chitosan in the feed made the band at
ꢀ1188 cmÀ1 rise, which means that the more lactide had
been grafted to chitosan.
The peak positioned ꢀ3440 cmÀ1 belonging to the
hydroxyl group, the amino group, and the amide of
chitosan did not move remarkably after the copoly-
merization, but its intensity became less with the in-
crease of lactide in the feeding ratio. We added an equal
molar ratio of Et3Al with chitosan, and did not hydro-
lyze the terminal products; thus this band is attributed
to the corresponding unreacted groups of chitosan. This
means that there are still a certain number of hydrogen
bonds that survived in the samples. It should be noticed
that the peaks ꢀ1635 and ꢀ1595 cmÀ1, which are as-
cribed to the amino groups in chitosan, changed into a
singlet and shifted to ꢀ1630 cmÀ1 after the reaction. The
same phenomenon has also been observed in our ex-
periments in which chitin and hydropropylchitin react
with Et3Al, and then polymerize with lactide under the
same conditions as in this study.
The grafting percentage and the amount of lactide
introduced to chitosan increase with the feeding molar
ratio are shown in Table 1. When the feeding molar
ratio of lactide to chitosan increased from 2:1 to 40:1,
the grafting percentage rose from 43% to 462%; mean-
while, the molar ratio of polylactide to chitosan in the
copolymer also rose from 0.96 to 10.33. This indicates
that the higher the concentration of the lactide in tolu-
ene, the higher the opportunity for the lactide to react
with chitosan reactive centers. The FTIR spectra
for some samples and chitosan itself are compared in
Figure 1.
The band that appeared in the copolymers at
ꢀ1758 cmÀ1 was assigned to the carbonyl group of the
branched polylactide. The methyl asymmetric defor-
mation of polylactide appears at ꢀ1475 cmÀ1. The
ꢀ1188 and ꢀ1215 cmÀ1 doublets observed in the co-
polymer are assigned to the symmetric C–O–C stretch-
ing modes of the ester group. There are two other peaks
at ꢀ1130 and ꢀ1045 cmÀ1 attributed to the methyl
rocking and C–CH3 stretching, vibration, respectively.
This evidence suggests that the lactide can indeed react
with chitosan with Et3Al as catalyst. Increasing of the
Dubois et al. have reported that primary amines and
Et3Al can initiate the polymerization of PCL, but the
final product is a secondary amide.9 When polycapro-
lactone terminated with a primary amino group initiates
the ring-opening copolymerization of the lactide, the
block copolymer also has a secondary amide structure.11
In our study, no obvious evidence for the secondary
Figure 1. The FTIR spectra for chitosan and graft copolymers: (a) chitosan, (b) 5:1, (c) 10:1, (d) 20:1.

848
Y. Liu et al. / Carbohydrate Research 339 (2004) 845–851
Figure 2. The 1H NMR spectra of chitosan and graft copolymers: (A) chitosan; (B) (a) 5:1, (b) 10:1, (c) 20:1, (d) 40:1.
amide could be found in the FTIR spectrum. Owing to
the fact that Et3Al cannot react with primary amines,9
the reacting mechanism needs to be further studied.
cannot only react with a hydroxyl group of x-primary
amines, but it can also catalyze the reaction between a
primary amine and caprolactone to get a secondary
amide.9 When polycaprolactone terminated with a pri-
mary amino group initiates the ring-opening copoly-
merization of the lactide, the block copolymer has also a
secondary amide structure.11 According to the same
mechanism, when the lactide is grafted onto chitiosan
using Et3Al as catalyst, Et3Al cannot only react with the
hydroxyl group in chitosan, but it can also catalyze the
reaction between a primary amine and lactide. Although
the two methyls existing in the lactide increase steric
hinderance, which makes its reactivity lower than that of
caprolactone, in some cases, they can initiate ring-
opening polymerization mutually to produce diblock
and triblock copolymers.12;13 So we assume lactide can
act with chisotan as in Scheme 1 to produce I, which has
a secondary amide structure under Et3Al as catalyst.
As we know, chitosan is a weakly basic polymer with
a pKb ¼ 6:3–7.0.14 The amide in I can carry out a
transfer reaction like that in Scheme 2, which is shown
under the amino of chitosan that survived as catalyst,
which also take place in PCL polymerization initiated by
an x-primary amine.15 This hypothesis can explain why
the intensity of the methine proton peak at d 4.3 is
stronger than the one at d 5.3 and why the singlet peak
appeared at ꢀ1630 cmÀ1 in the FTIR spectra. This
1
The H NMR spectra of the original chitosan in 1:1
CF3COOD–D2O and grafted copolymer of 5:1, 10:1,
20:1, and 40:1 in the same solvent are compared in
Figure 2Aand B, respectively. Chitosan shows a singlet
at d 3.15 and mutiplets at d 3.55–3.80 (H-3, H-4, H-5,
2H-6) and a small signal at d 4.4 (H-1) corresponding to
the ring methine protons. The singlet at d 1.95 is due to
the survival of the N-acetylglucosamine units of chitin.
The copolymers not only show the original signals of
chitosan, but also have the new peaks at d 4.3 and d 5.3
assigned to the terminal methine protons of the bran-
ched polylactide and repeat units of it in the chain, re-
spectively. Both signals are clearly separated into two
groups of protons. This implies that there are two dif-
ferent kinds of atoms bonding to the methine. The peak
at d 1.56 is attributed to the methyl of polylactide.10
Very different from that which other authors have
reported,10 the intensity of the peak at d 4.3 is stronger
than that of that peak at d 5.3. In the common 1H NMR
spectra of polylactide, the signal at d 5.3 is much greater
than that at d 4.3, and the molecular weight of poly-
lactide can be calculated from the ratio of the two signal
intensities. Indeed, in the reaction of Et3Al initiating the
ring-opening polymerization of caprolactone, the Et3Al

Y. Liu et al. / Carbohydrate Research 339 (2004) 845–851
849
Scheme 1.
Scheme 2.
assumption is supported by the nucleophilicity of the
amide and its reactivity toward the carbonyl group of
the ester under basic conditions.16
The integral intensity ratio between peak d 4.3 and
d 5.3, that is the ratio of the ending methine amount and
that in the chain of the polylactide, is determined by the
graft branch length. If the length of the polylactide is
shorter, the amount of methine in the chain is less, and
the peak intensity of the resonance at d 4.3 would be
stronger than that at d 5.3. In this study, owing to the
methine peak intensity, the d 4.3 resonance is much
greater than the peak of d 5.3; therefore, the graft
copolymerization between chitosan and lactide produces
predominantly short-branched polylactide. In the sam-
ples of 10:1, 20:1, and 40:1, the integral intensity ratio
between peak d 4.3 and d 5.3 is 2.24, 2.45, and 3.05,
respectively. The amount of short-branched polymer
increases with an increase in lactide content in the
feeding ratio.
Figure 3. The DSC curves of graft copolymers: (a) 10:1, (b) 20:1, (c)
40:1.
shifted to lower temperature as a greater content of
lactide is bonded in the copolymers, but the enthalpy
(DH) is raised.
3.2. Thermal properties of the grafting copolymers
There are a considerable number of intermolecular
and intramolecular hydrogen bonds that existed in
chitin and chitosan molecules, which makes their melt-
ing points higher than their decomposition points. If
chitosan were to be bonded with branched chains, the
integrity of its molecular structure will be broken. As all
hydroxyl and amino groups in chitosan were fully sub-
stituted with alkyl groups, that is, all hydrogen bonds in
it were destroyed, it still does not have a melting tran-
sition.17 This phenomenon is different with cellulose. In
It was found that samples with a feeding ratio of lactide
to chitosan of >10:1 have a melting transition in their
DSC curves. The melting enthalpies and peak temper-
atures of each of them are also shown in Table 1. The
DSCs are the curves illustrated in Figure 3. The melting
points of the copolymers decrease from 150.65 to
139.95 ꢀC with the content of polylactide increasing. The
enthalpies of the copolymers rise slowly from 2.3359 to
3.3055 cal/g. Namely the melting point (T) gradually

850
Y. Liu et al. / Carbohydrate Research 339 (2004) 845–851
Figure 4. WAXS patterns of chitosan and its graft copolymer: (a) chitosan, (b) 2:1, (c) 5:1, (d) 10:1, (e) 20:1, (f) 40:1.
the same case, its alkyl derivatives have clear melting
points.18 This suggests that chitosan has a more rigid
structure than cellulose, which is supported by the fact
that the persistence length of chitosan of 22 nm is much
greater than that of cellulose (11 nm) in LiCl
(5%) + N,N-dimethylacetamide mix solution.19 When
chitosan was grafted with polylactide, its original in-
tegrity was also destroyed. The hydrogen bonds that
existed in it were broken to a certain extent, which
makes intermolecular interactions decrease. On the
other hand, the ester bond is more flexible than that of a
hydrocarbon. When it is used as the branched chain of
chitosan, it can plasticize internally the rigid main chain
of chitosan. Meanwhile, the branched ester linkage is
more polar than that of a hydrocarbon, so there is a
stronger repulsive force among molecules of the bran-
ched ester than that existing among hydrocarbons,
which can be in favor of chitosan retaining its spatial
configuration.
3.3. Wide-angle X-ray scattering
Figure 4 shows the WAXS patterns as compared with
that of chitosan. The strongest reflection of chitosan
appears at 2h ¼ 19:84ꢀ. In the feeding ratio, the ratio of
lactide to chitosan is 2:1, the strongest reflection shifts to
2h ¼ 19:74ꢀ. In the case of 5:1 and 6:1, it is 2h ¼ 21:66ꢀ
and 21.44ꢀ, respectively. These results indicate the
crystalline patterns were different with chitosan, when
lactide was grafted onto it; the original crystallinity was
destroyed.
The strongest peak appears at about 2h ¼ 16:7ꢀ for
samples of 10:1, 20:1, 40:1, which changes more re-
markably than the ones mentioned above. In addition to
that, these three samples have almost the same d-spac-
ing, which means that graft copolymers have a similar
packing mode when the graft percentage is >338% (in
the feeding ratio 10:1).
The melting point of a polymer is determined by the
ratio between melting enthalpy and entropy, namely
T ¼ DH=DS. The DH differences among the samples
10:1, 20:1, 40:1 are not significant, but they have the
tendency to increase. Meanwhile this melting point de-
creases. This means that DS must be increasing. With
the content of lactide increasing in the feeding ratio, the
amount of short-branched chain rises in the copolymer
(see discussion in 1H NMR sections); therefore, the
conformation entropy must also increase, and the
melting points have a tendency to drop. The dropping
range for short side chains from 10:1 to 20:1 is less than
from 20:1 to 40:1; meanwhile, the decreasing extent of
the melting point of the copolymer obeys the same
trend.
4. Conclusions
Anew type of copolymer lactide grafted onto chitosan
was synthesized, and the grafting copolymerization
could be carried out well in toluene at 70 ꢀC. The de-
rivatives of chitosan have clear melting transitions in the
DSC scans when the molar feeding ratio lactide to
chitosan is >10:1.
References
1. Tominhate, K.; Ikata, Y. Biomaterials 1997, 18, 567–575.
2. Teng, W. L.; Khor, E.; Tan, T. K.; Lim, L. Y.; Tan, S. C.
Carbohydr. Res. 2001, 332, 305–316.

Y. Liu et al. / Carbohydrate Research 339 (2004) 845–851
851
3. Mi, F.-L.; Lin, Y.-M.; Wu, Y.-B.; Shyu, S.-S.; Tsai, Y.-H.
Biomaterials 2002, 23, 3257–3267.
4. Barakat, I.; Dubois, P.; Jerome, R.; Teyssie, P.; Goethals,
E. J. Polym. Sci.: Part A: Polym. Chem. 1994, 32, 2099–
2110.
5. Chen, X.; McCarthy, S. P.; Gross, R. A. Macromolecules
1997, 30, 4295–4301.
6. Grijpma, D. W.; Kroeze, E.; Wijenhuis, A. J.; Pennings,
A . J.Polymer 1993, 34, 1496–1503.
12. InÕt Veld, P. J. A.; Velner, E. M.; Van De Witte,
P.; Hamhuis, J.; Dijkstra, P. J.; Feijen, J. J. Polym.
Sci.: Part A: Polym. Chem. 1997, 35, 219–
226.
13. Song, C. X.; Feng, X. D. Macromolecules 1984, 17, 2764–
2767.
14. Krajewska, B. React. Funct. Polym. 2001, 47, 3441–3448.
15. Degee, P.; Dubois, P.; Jerome, R.; Teyssie, P. Macromole-
cules 1992, 25, 4242–4248.
7. Dubois, P.; Jerome, R.; Teyssie, P. Macromolecules 1991,
24, 977–981.
16. Shafer, J. A.; Morawetz, H. J. Org. Chem. 1963, 28, 1899–
1901.
8. Barakat, I.; Dubois, P.; Jerome, R.; Teyssie, P. J. Polym.
Sci.: Part A: Polym. Chem. 1993, 31, 505–514.
9. Dubois, P.; Degee, P.; Jerome, R.; Teyssie, P. Macromole-
cules 1992, 25, 2614–2618.
10. Arvanitoyannis, I.; Nakayama, A.; Kawasaki, N.; Ya-
mamoto, N. Polymer 1995, 36, 2271–2279.
11. Tian, D.; Dubois, P.; Jerome, R.; Teyssie, P. Macromole-
cules 1994, 27, 4134–4144.
17. Zong, Z.; Kimura, Y.; Takahashi, M.; Yamane, H.
Polymer 2000, 41, 899–906.
18. Yamagishi, T.; Fukuda, T.; Miyamoto, T.; Yakoh, Y.;
Takashina, Y.; Watanabe, J. Liq. Cryst. 1991, 10, 467–
473.
19. Terbojevich, M.; Cosani, A.; Conio, G.; Marsano,
E.; Bianchi, E. Carbohydr. Res. 1991, 209, 251–
260.