Polymer Degradation and Stability 90 (2005) 488e500

www.elsevier.com/locate/polydegstab

Water absorption and enzymatic degradation of

poly(lactic acid)/rice starch composites
G.H. Yew a, A.M. Mohd Yusof a, Z.A. Mohd Ishak a,*, U.S. Ishiaku b
a
School of Materials and Mineral Resources Engineering, Engineering Campus, Universiti Sains Malaysia, Seri Ampangan,
14300 Nibong Tebal, Penang, Malaysia
b
Advanced Fibro-Science, Kyoto Institute of Technology, Matsugasaki, Sakyo-ku, Kyoto 606-8585, Japan
Received 4 January 2005; received in revised form 24 March 2005; accepted 8 April 2005
Available online 23 May 2005

Abstract

Poly(lactic acid) (PLA) composites consisting of PLA, rice starch (RS) (0e50 wt%) and epoxidised natural rubber (ENR50) were
compounded by a twin-screw extruder and compression moulded into dumbbell specimens. Tensile tests were performed to
characterize the mechanical properties of the PLA/RS composites. Morphological studies were done on the tensile fractured surface
of the specimens by using scanning electron microscopy (SEM). Twenty weight percent of RS achieved a good balance of strength
and stiffness. Beyond 20 wt% loading of RS, the tensile strength and elongation at break of PLA decreased drastically. This may be
attributed to the agglomeration of RS, which could then act as stress concentrator. The incorporation of ENR50 increased the
tensile strength and elongation at break of the PLA/RS composites remarkably, owing to the elastomeric behaviour and
compatibilisation effects of ENR50. Interestingly, the morphology of PLA/RS composites transformed to a more ductile one with
the addition of ENR. The kinetics of water absorption of the PLA/RS composites conforms to Fick’s law. The Mm and D values are
dependent on the RS and ENR concentrations. The tensile properties of the PLA/RS composites deteriorated after water
absorption. The retention-ability and recoverability of the PLA/RS composites are relatively low, attributed to the hydrolysis of
PLA, degradation of the PLAeRS interface and leaching of the RS particles. In addition, the tensile properties of PLA/RS
composites decreased drastically upon exposure to enzymatic degradation. Extensive pinhole and surface erosion on the PLA/RS
composites indicate high degree of hydrolysis. Whilst the addition of ENR leads to some improvements in tensile properties,
nevertheless, it enhanced the biodegradability of the PLA/RS composites when exposed to water and a-amylase enzymatic
treatments.
Ó 2005 Elsevier Ltd. All rights reserved.

Keywords: Polylactic acid (PLA); Rice starch (RS); Epoxidized natural rubber (ENR); Water absorption; Enzymatic degradation

1. Introduction and are slowly being damaged as a result of the usage of

Advanced technology in the field of petrochemical
polymers has brought many advantages to mankind.
However, the ecosystems are now considerably disturbed
* Corresponding author. Present address (until November 30, 2005):
Institute fur Verbundwerkstoffe GmbH, Technische Universitat
Kaiserslautern, Erwin-Schrodinger Strasse, Geb 58, D-67663 Kaiser-
slautern, Germany. Fax: C49 631 2017198.
E-mail address: zarifin.ishak@ivw.uni-kl.de (Z.A. Mohd Ishak).
non-degradable materials in disposable items. The
development of biodegradable plastics can play a crucial
role in helping to solve the waste disposal problems.
Since waste reduction is important for future municipal
solid waste, a number of important issues need to be
addressed to ensure that new biodegradable plastics
industry develops [1].
Polylactic acid (PLA) is a degradable thermoplastic
polymer with excellent mechanical properties which is
now, beginning to be produced on a large scale from

0141-3910/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.polymdegradstab.2005.04.006
 G.H. Yew et al. / Polymer Degradation and Stability 90 (2005) 488e500
fermentation of corn to lactic acid and subsequent
chemical polymerisation. Pure PLA can degrade to
carbon dioxide, water and methane in the environment
over a period of several months to 2 years, compared to
other petroleum plastics which need 500e1000 years [2].
Starch is of interest as a naturally occurring bio-
degradable material due to its cost effectiveness, abun-
dance, renewability, availability, thermo-processability
using conventional plastic processing equipment (extru-
sion, injection moulding), and its adequate thermal
stability for melt blending with synthetic plastics (i.e.
aliphatic polyester, non-biodegradable thermoplastics)
[3]. The incorporation of starch into plastics to enhance
their fragmentation and degradability in the environment
has generated considerable interest. The introduction of
low-cost starch as a filler into PLA would result in a fully
renewable biodegradable material [2].
Epoxidised natural rubber (ENR) is obtained from
natural rubber by replacing some of the double bonds
with epoxide units. ENR with 50 mol% epoxidation has
been shown to be polar with properties similar to those
of acrylonitrile butadiene rubber and butyl rubber [4]. It
has been reported that compatibility of ENR increases
as the ENR’s polarity increases [5]. The presence of the
oxirane groups in ENR introduced polarity and made
the rubber more versatile for rubbererubber or rubbere
plastic blending [6]. However, ENR vulcanizates exhibit
poor aging properties. The unstable nature of ENR is
believed to be due to the residual acidity that was
inherited from the natural rubber modification (epoxi-
dation) with peroxide. In long-term exposure to ambient
temperature (27e30
C) in the laboratory, poly(vinyl
chloride)/ENR blends deteriorated rather rapidly [7].
Addition of ENR into LLDPE/sago starch composites
had achieved a comparable mechanical properties and
remarkable degradability [8].
Water absorption behaviour of starch filled polylactic
acid composites has been studied by some researchers
[9e11]. The addition of starch increased the water
absorption of the PLA [9]. According to Wang et al. [2],
the higher water absorption after a long time would
probably lead to enhanced biodegradability. According
to Garlotta et al. [3], PLA is easily hydrolysed by
moisture, and its hydrolysis products are non-toxic to
human being. Water absorption may strongly affect the
organic materials behaviour, by producing changes in its
chemical and physical nature [11]. It is believed that, the
hydrophilicity of PLA (ester groups), ENR (oxirane
groups) and RS (hydroxyl groups) will increase the
water uptake. Thus, water, may act as a plasticiser, and/
or ‘‘hydrolysis agent’’ for the PLA matrix, and sub-
sequently deteriorate the properties (e.g. mechanical
properties) of PLA/RS composites. Therefore, the
degradability of the PLA/RS composites might be
accelerated in a moist environment by water during its
service life.
489
a-Amylase are endoamylases catalysing the hydro-
lysis of internal a-1,4-glycosidic linkages in the starch
in a random manner. The microbial a-amylase for
industrial purposes are derived mainly from Bacillus
licheniformis, Bacillus amyloliquefaciens and Aspergillus
oryzae [12]. a-Amylase catalyses the hydrolysis of starch
resulting in a significant release of maltose and leads to
formation of smaller maltosaccharide fragments [13,14].
Starch susceptibility to enzyme attack is influenced by
several factors, such as amylose and amylopectin
content, crystalline structure, granular structure, parti-
cle size and the presence of enzyme inhibitors. Hydro-
lysis pattern is an important factor for susceptibility to
a-amylase attack. Extensive surface erosion was shown
to indicate a high degree of hydrolysis, whereas less
surface erosion indicated less degradation [15]. Ishiaku
et al. [1] demonstrated that extensive enzyme activity
occurred on a polycaprolactone/sago starch film after
being subjected to enzyme amylase. The majority of the
sago starch granules were removed from the surface of
the film and leave behind transparent voids. Danjaji
et al. [16] reported that microbial activity occurred on
the enzyme amylase-treated LLDPE/sago starch sample,
which is indicated by the presence of holes and
agglomerates on the sample surface. Araújo et al. [17]
have investigated the enzymatic degradation of corn
starch/poly(ethylene-vinyl alcohol) copolymer blends
(SEVA-C), and interestingly, they have found that the
action of a-amylase leads to the starch degradation, and
part of the amorphous of SEVA-C blend was released
and degraded.
This paper focuses on the study of PLA composites
consisting of PLA, RS and ENR. The effect of RS
loading and ENR on the mechanical properties and
morphology of PLA/RS composites will be investigated.
Whilst the addition of ENR could lead to some
improvements of the mechanical properties of PLA/RS
composites, nevertheless, it could enhance the bio-
degradability of the composites when the composites
are being subjected to a moist environment (water).
Moreover, emphasis was laid on the moisture absorp-
tion studies (i.e. water uptake) and enzymatic tests of
PLA/RS composites. Effects of water absorption and
enzymatic attack on both mechanical and morpholog-
ical properties of PLA/RS composites were determined.
2. Experimental
2.1. Materials
Polylactic acid (PLA) was supplied by Mitsui
Chemicals Inc., Japan. Melt flow index (MFI) and
density of PLA were 64.8 g/10 min (at 190
C and
2.16 kg load) and 1.23 g/cm3, respectively. Rice starch
was obtained from TIGA Gajah Cho Heng Sdn. Bhd.,
490 G.H. Yew et al. / Polymer Degradation and Stability 90 (2005) 488e500

Malaysia. ENR50 was supplied by Rubber Research information about the functional groups and chemical
Institute of Malaysia (RRIM), Selangor. The designa- characteristics of the PLA/RS composites. FTIR spectra
tion of materials is shown in Table 1. were obtained from KBr pellets at room temperature.

2.2. Sample preparation 2.6. Water absorption

Melt compounding of the PLA/RS composites was
done in a twin-screw extruder (Berstorff). The extrusion
zone temperatures ranged from 180 to 195
C. Prior to
extrusion, PLA pellets and RS were dried by using
a dehumidifier (ConAir Franklin) at 70
C for 24 h. The
extrudates were pelletised with a Speecon 7200MA
pelletiser. The compounded samples were compression-
moulded in a Kao Tieh GoTech compression moulding
machine at 190
C using a pressure of 10 MPa for
12 min and cooled under pressure for 3 min. Moulded
samples were cut into dumbbell shaped specimens
according to ASTM D638.
Specimens were dried at 80
C in a vacuum oven until
a constant weight was attained prior to immersion in
water in a stainless steel water bath maintained at 30
C.
The water absorption tests were performed for 30 days.
Weight gains were recorded by periodic removal of the
specimen from the water bath and weighing on a balance
with a precision of 1 mg. The percentage gain at any
time t, Mt as a result of moisture absorption, was
determined by Eq. 3.1:
Mt ð%ÞZðWw  Wd Þ=Wd !100 ð3:1Þ

2.3. Melt flow index (MFI) and density
measurement
Melt flow index and density of various materials were
measured by using Melt Flow Indexer (at 190
C, load
2.16 kg) and density balance (model Precisa XT 220 A).
2.4. Tensile properties
Tensile testing was done according to ASTM D638,
by using an Instron testing machine. The crosshead
speed was set at 5 mm/min. Samples for tensile
measurements were conditioned at 30 G 2% relative
humidity (RH) for 24 h before testing. All these tests
were conducted at ambient temperature and an average
value of five repeated tests was taken for each
composition.
2.5. Fourier-transform infrared spectroscopy
where Wd and Ww denote, respectively, weight of dry
material (the initial weight of materials prior to exposure
to the water absorption) and weight of materials after
exposure to water absorption. The percentage equilib-
rium or maximum moisture absorption, Mm, was
calculated as an average value of several consecutive
measurements that shows no appreciable additional
absorption. The weight gain resulting from moisture
absorption can be expressed in terms of two parameters,
the diffusion coefficient or diffusivity, D, and the
maximum moisture content, Mm, as Eq. 3.2 [18]:
   
Mt 8 Dt
Z1  2 exp  2 p2 ð3:2Þ
Mm p h
where h is the thickness of the sample.
D can be calculated from the initial linear portion of
the absorption curve (i.e. slope of Mt vs t1/2).

Fourier-transform infrared spectroscopy (FTIR; Perkin 2.7. Enzymatic tests

Elmer System 2000) was used to obtain some qualitative
Dumbbell specimens were weighed accurately to
Table 1 0.1 mg and placed in a bowl containing 3 g of a-amylase
Designation, composition, MFI and densities of materials in 3 l of distilled water. The bowl was heated externally
Designation Part (wt%) MFI (g/10 min) Density (kg/m3) in a water bath at 37
C for 14 days. This enzymatic test
PLA 100 64.8 1.23
method is adopted from Shah et al. [19]. The specimens
PLA/RS10 90/10 59.9 1.24 were then removed and rinsed with distilled water to
PLA/RS20 80/20 56.8 1.27 remove any traces of the a-amylase enzyme. These
PLA/RS30 70/30 50.3 1.28 specimens were placed in a desiccator for 24 h after
PLA/RS40 60/40 29.3 1.30 which they were weighed. In addition, samples without
PLA/RS50 50/50 27.3 1.31
PLA/RS20/E2.5 80/20/2.5 40.6 1.24
enzymatic treatment were also used as control (i.e.
PLA/RS20/E5 80/20/5 38.2 1.27 samples immersed in distilled water without a-amylase
PLA/RS20/E7.5 80/20/7.5 36.5 1.26 enzyme). Tensile tests were performed on both types of
PLA/RS20/E10 80/20/10 32.2 1.25 samples. Further, the samples’ surface of the specimens
Note: RS Z rice starch; E Z epoxidised natural rubber (ENR). was examined with SEM.
 G.H. Yew et al. / Polymer Degradation and Stability 90 (2005) 488e500 491

2.8. Morphological studies RS. This is in harmony with the results presented by
Ke and Sun [9]. According to Ke and Sun [9] and Ke
Scanning electron microscopy (SEM; Cambridge et al. [10], as the starch (i.e. dispersed phase) loading
Stereoscan 2000) was used to study the morphology increased, the effective cross-sectional area of PLA
and mode of failure of the PLA/RS samples. Test (i.e. continuous phase) was reduced, subsequently result-
samples were first sputter-coated with gold in order to ing in a decrease in tensile strength. However, it is
prevent electrical discharge during examination. interesting to note that the tensile strength of PLA/RS
composites achieved a relatively higher value at 20 wt%
RS loading. This may be attributed to a better
3. Results and discussions dispersion (or less agglomeration) of the RS in PLA
matrix, as well as better adhesion between RS and PLA.
3.1. Melt flow index (MFI) and density According to Park et al. [21], the tensile strength of
ionomer/starch blends increased with incorporation of
Table 1 shows the MFI and density values of PLA/ starch, which may be due to the polar interaction
RS composites. MFI of the RS filled PLA decreased as between starch and carboxyl groups of the ionomer. The
the filler loading increased. This is in agreement with the drops in tensile strength become more drastic at higher
most starch filled thermoplastics. Thakore et al. [20] filler loading for PLA/RS composites. This is owing to
reported that the MFI of low density polyethylene the fact that fillerefiller interaction becomes more
(LDPE)/modified starch blends is lower than that of pronounced than fillerematrix interaction with increas-
the neat LDPE. The incorporation of ENR further ing filler content. Accordingly, further addition of RS
decreased the MFI of PLA/RS composites. This may be (i.e. O20 wt%) could lead to agglomeration of starch
attributed to the elastomeric behaviour of ENR, which particles. This will give rise to stress-concentration
will resist flow. As expected, densities of the PLA/RS are points in the samples [22]. The drop in tensile strength
higher than that of pure PLA. is in accordance with the decrease in the effective cross-
sectional area for spherical particulate fillers. Further
3.2. Tensile properties evidence on the dispersion or agglomeration of starch
particles in the PLA matrix will be provided by SEM
3.2.1. Effects of RS content analysis, reported later in this section.
The variation of tensile modulus, tensile strength and The Nicolais and Narkis equation which relates the
elongation at break with the RS loading of PLA/RS tensile strength of the composites, sc to the tensile
composites is shown in Table 2. The tensile modulus of strength of the matrix, s0 is used to quantitatively
PLA was increased additively by the incorporation of identify whether any interaction (or adhesion) occurred
RS. Irrespective of the materials composition, all RS between the PLA matrix and RS:
filled PLA composites show higher modulus than pure
ÿ

PLA. Recall that the modulus of RS is higher than that sc Zs0 1  1:21f2=3
of PLA. The increasing modulus with increasing of the
RS loading in PLA can be explained by the crystallinity, where f is volume fraction of the filler.
hydrogen bonding and stiffening effect of the starch Recall that this model is based on the assumption
granules. This is in agreement with an earlier work that the filler particle is spherical, the filler is distributed
reported by Garlotta et al. [3] and Ke et al. [10]. uniformly in the continuous phase, and that there is no
As expected, the tensile strength and elongation at adhesion between the filler particles and the matrix.
break of the PLA decreased with the incorporation of The Nicolais and Narkis plots (i.e. relative tensile
strength vs volume fraction plots) are shown in Fig. 1.
Table 2 The relative tensile strength of the PLA/RS composites
Tensile properties of PLA/RS composites decreased linearly as the RS fraction increased. The
Materials Tensile Tensile Elongation regression lines with slopes of 0.82 for RS filled PLA
modulus (GPa) strength (MPa) at break (%) composites were higher than 1.21 (i.e. the slope for
PLA 3.30 G 0.08 57.8 G 0.86 3.8 G 0.07 Nicolais and Narkis equation). This indicates that
PLA/RS10 3.50 G 0.08 40.9 G 0.77 2.6 G 0.06 adhesion occurred to some extent between the two
PLA/RS20 3.80 G 0.07 44.0 G 0.56 2.5 G 0.08 phases (i.e. PLA and RS particles), albeit the degree of
PLA/RS30 4.20 G 0.09 37.7 G 0.92 2.1 G 0.04
adhesion is inconclusive. Similar observations were also
PLA/RS40 4.40 G 0.09 33.6 G 0.68 1.9 G 0.05
PLA/RS50 4.50 G 0.08 30.6 G 0.78 1.7 G 0.07 reported by Ke and Sun [9] and Garlotta et al. [3]. Ke
PLA/RS20/E2.5 3.70 G 0.06 47.9 G 0.62 2.7 G 0.06 and Sun [9] suggested that adhesion between PLA and
PLA/RS20/E5 3.30 G 0.07 49.0 G 0.54 3.4 G 0.05 wheat starch probably existed, however, the interaction
PLA/RS20/E7.5 3.10 G 0.08 41.6 G 0.68 2.9 G 0.08 is not strong. The interaction between PLA and RS may
PLA/RS20/E10 2.80 G 0.06 37.2 G 0.79 2.6 G 0.07
be attributed to the possible hydrogen bonding that
492
1.2
1.0
Relative Tensile Strength
0.8
0.6
0.4
Nicolais & Narkis Equation (y = -1.21x + 1)
R2 = 1
0.2
0.0
0 0.1 0.2
Volume Fraction (φ2/3)
Fig. 1. Relative tensile strength of the predicted (Nicolais and Narkis
equation) and experimental data of the PLA/RS composites.
occurs between the carbonyl group (i.e. from ester
linkage) in PLA and the hydroxyl group in starch. A
proposed possible site for interaction between PLA and
RS is shown in Fig. 2. It is interesting to note that there
is a small amount of hydroxyl group (eOH) in the PLA
used in this study. The eOH groups may be due to
the terminal hydroxyl groups in the PLA main chain.
This claim is further supported by FTIR studies. Fig. 3
shows the FTIR spectra of the PLA and PLA/RS com-
posites. It can be seen that a relatively small peak
at approximately 3500 cm1 (characteristic absorption
region for OeH bond stretching deformation) is visible
for neat PLA. This seems to indicate the presence of
hydroxyl groups in the pure PLA. It is also interesting to
note that the characteristic peak of PLA (i.e. 3500 cm1)
has disappeared with the incorporation of RS, and the
peak was shifted to a broad peak similar to that of RS.
One may suggest that the disappearance of the PLA
peak is attributable to the overlapping of the peaks of
PLA and RS.
3.2.2. Effects of ENR
The effect of ENR50 on the tensile properties of
PLA/RS composites is shown in Table 2. From the
preceding discussion, 20 wt% possesses a good balance
of tensile properties, albeit elongation at break is
relatively low. It is noteworthy that the incorporation
of ENR (at 5 phr) into PLA/RS has slightly increased
the tensile strength and elongation at break of the
composites. However, the tensile modulus of PLA/RS
composites was decreased as the loading of ENR
increases. This may be attributed to the elastomeric
behaviour and compatibilisation effects of ENR. The
reduction in the tensile modulus of PLA/RS as in-
corporation of ENR can be attributed to the rubbery
behaviour of ENR. It is speculated that ENR (containing
ether group) could form favourable interaction with
G.H. Yew et al. / Polymer Degradation and Stability 90 (2005) 488e500
PLA and RS, presumably via hydrogen bonding. A
similar observation was reported earlier by Sharma et al.
[8]. A slight improvement in the tensile strength and
Experimental (y = -0.82x + 1.01)
elongation at break was achieved by the addition of
2
R = 0.98 6 wt% of ENR50 in the linear low density polyethylene
(LLDPE)/sago starch blends, which is attributed to the
improvement of the interaction between LLDPE and
sago starch. A proposed possible chemical interaction
sites between PLA, RS and ENR are presented in Fig. 2.
It is believed that hydrogen bonding could form between
the ester group of PLA and the oxirane group of ENR.
In addition, hydrogen bonding could also form between
0.3 0.4 0.5 0.6 0.7 0.8 RS (hydroxyl groups) with PLA and ENR. From Table 2,
it can be seen that further addition of ENR (i.e.
O5 phr) resulted in a reduction of the tensile strength.
It is interesting to note that the elongation at break of
PLA/RS composites was not much affected by ENR
content. This is in agreement with the earlier work done
by Sharma et al. [8]. A reduction in tensile strength was
observed at the loading of elastomer (e.g. ENR50) in
exceeded of 8 wt% in the LLDPE/sago starch systems.
3.3. Morphology
Fig. 4a shows the SEM micrograph of the tensile
fractured surface of pure PLA. According to Engel et al.
[23], ductile fractures can be arbitrarily divided into two
groups, the criterion being the lengths of the fibrils
remaining in the fracture plane. Polymer with high
plastic deformation has fracture-length of remaining
fibril O10 mm; polymer with low plastic deformation has
fracture-length of remaining fibrils 1e10 mm. Accord-
ingly, fractures producing fibrils less than 1 mm long are
categorized in brittle fracture. Thus, in the presence
studies, the fractured surface of PLA could be catego-
rized as ductile fractures. The fibril of the PLA in the
fractured surface plane is shown as the white circle in
Fig. 4a. Fig. 4b shows the SEM micrograph of the tensile
fractured surface of the PLA/RS (80/20) composites. The
RS particles were not completely covered by PLA phase
as some starch particles were aggregated. Garlotta et al.
[3] reported that the morphology of PLA/starch compo-
sites fractured surface contained numerous voids, intact
starch granules, interfacial debonding between starch
granules and matrix, and some fractured granules. In the
present study, although some voids could be observed
between the RS granules and PLA phase in the
composites, most of the starch particles were still kept
intact within the PLA matrix. Thus, this may hint that
there is some extent of interaction between PLA and RS,
due to the hydrophilicity and polarity of both materials.
Fig. 4c and d shows the SEM micrograph taken
from the tensile fractured surface of PLA/RS/ENR
composites containing 5 phr and 10 phr of ENR, respec-
tively. Interestingly, a more fibrillated morphology can
be observed for the PLA/RS/ENR composites as
 G.H. Yew et al. / Polymer Degradation and Stability 90 (2005) 488e500
Fig. 2. Proposed chemical interactions between PLA, RS and ENR.
compared to PLA/RS composites (cf. Fig 4b). The
extensive plastic deformation explains the slight im-
provement of the elongation at break of the PLA/RS/
ENR compared to PLA/RS composites.
Fig. 5a and b shows the SEM micrograph taken from
the cryogenically fractured (i.e. sample fractured in liquid
nitrogen) surface of PLA/RS20 and PLA/RS/E5 compo-
sites, respectively. It is interesting to note that, for the
PLA/RS20 composites, the RS particles are irregularly
shaped. However, there are still some starch granules
intact within the PLA matrix. This may again indicate
some extent of interaction between PLA and RS.
However, bigger gaps as revealed in the PLA/RS
composites (cf. Fig. 5a) imply relatively poor interaction
between the two phases. On the other hand, incorpora-
tion of ENR into PLA/RS enhanced the interfacial
bonding between PLA matrix and RS particles.
Compared to PLA/RS (Fig. 5a), the gap between RS
particles and PLA was relatively smaller and closer, this
may indicate that the wettability of the RS by PLA matrix
493
has been improved, based on the fact that the RS particles
appear to be bonded to the PLA matrix (cf. Fig. 5b). It is
interesting to note that in the PLA/RS20/E5 composites,
the RS particles are spherical in shape. Further analysis
needs to be performed to confirm this observation.
3.4. Water absorption
Fig. 6 shows the percentage moisture absorption
Mt of pure PLA and PLA/RS composites as a function
of t1/2 at immersion temperature of 30
C. An initial
linear relationship between Mt and t1/2 is observed in
each case, followed by saturation. This indicates that
the water uptake behaviour of PLA composites obeys
Fick’s law.
A rapid moisture uptake was observed for all the
specimens within the first few days of immersion. The
absorption of water is related to its rate of diffusion into
the composites. The Mt and D values are summarized in
494 G.H. Yew et al. / Polymer Degradation and Stability 90 (2005) 488e500

Fig. 3. FTIR spectra of various PLA/RS composites.

Table 3. Both Mm and D values of PLA/RS composite content, Mm at 1%). The higher Mm and D values in the
are higher than that of neat PLA. Wang et al. [2] reported composites can be attributed to the hydrophilic nature of
that the water absorption of raw PLA levelled the RS by virtue of the presence of an abundance of
at approximately 1% (i.e. achieved maximum moisture hydroxyl groups which are available for interaction with

Fig. 4. SEM micrographs taken from the tensile fractured surface of (a) PLA, (b) PLA/RS20, (c) PLA/RS20/E5, and (d) PLA/RS20/E10.
 G.H. Yew et al. / Polymer Degradation and Stability 90 (2005) 488e500 495

6.0

Percentage of moisture absorption, Mt (%)

PLA
PLA/RS20
5.0 PLA/RS20/E5
PLA/RS20/E10

4.0

3.0

2.0

1.0

0.0
0 200 400 600 800 1000 1200 1400 1600 1800
Immersion time, t1/2 (s1/2)

Fig. 6. Effects of RS and ENR on the moisture uptake of PLA

composites during water absorption at 30
C.

In order to ascertain the extent of leaching of the

Fig. 5. SEM micrograph taken from the cryogenic-fractured surface of
(a) PLA/RS20 and (b) PLA/RS20/E5 composites.
the water molecules. Preechawong et al. [24] suggested
that starch is prone to moisture and water absorption due
to its hygroscopic nature. Water molecules may act as
a natural plasticiser for starch, which helps render starch
being flexible as compared to hard and rigid filler in its
completely dry state. Note that, the Mm value of PLA/RS
is approximately 5 times compared to that of the neat
PLA. It is expected that the exposed starch granules (e.g.
the starch particles at or near the specimens’ surface)
absorb moisture faster than those in the interior parts (i.e.
starch granules which are entrapped/coated by PLA
matrix). The water molecules could saturate the surface
of the PLA/RS composites easily and also penetrate into
the composites through voids, resulting in higher water
absorption in a short exposure time. Water absorption
may drop slightly as immersion time is increased, owing
to the fact that some starch particle was leached away
from the specimen [10]. On the other hand no significant
weight loss will imply that insignificant leaching has
taken place during the experiments [25].
starch particles from PLA/RS composites during water
absorption tests, the specimens were re-dried, and then
the weight loss of the composites was determined (cf.
Fig. 7). It can be seen in all the cases that the weight loss
of RS filled PLA composites is rather small i.e. about
0.8%. The weight loss is most probably due to the
leaching of the RS particles located in the specimen’s
surface.
From Table 3 it can be seen that Mm values of
the PLA/RS composites increased slightly with the
incorporation of ENR. This may be attributed to the
penetration of water molecules into the interfacial zone
i.e. between PLAeRS and PLAeENR. It is interesting
to note that the D value of the PLA/RS composites had
increased slightly in the presence of ENR. ENR is
a polar material that will enhance H-bond formation
with H2O. This indicates that the oxirane groups in the
ENR may facilitate water penetration.
Table 4 shows the tensile properties of PLA and
PLA/RS composites in as-received (AR), wet (W) and
re-dried (RD) states. It can be seen that both tensile
strength and modulus of PLA/RS decreased after being
subjected to water absorption. The low percentage
retention-ability in tensile strength indicates that water
absorption has caused a dramatic reduction in the
strength of the PLA/RS composites. This may probably
Table 3
The effect of RS and ENR on the diffusion constant (D) and maximum
moisture content (Mm) of PLA composites at immersion temperature
of 30
C
Materials Mm (%) D (!1013 m2/s)
PLA 1.0 5.60
PLA/RS20 5.0 6.97
PLA/RS20/E5 5.5 7.13
PLA/RS20/E10 5.7 7.36
496 G.H. Yew et al. / Polymer Degradation and Stability 90 (2005) 488e500

6.0 The action of water may have resulted in a further

maximum moisture absorption
weight loss due to leaching disruption of the adhesion between the RS granules and
PLA matrix. Subsequently, these will probably lead
Percentage of weight changes (%)

5.0
4.0
3.0
2.0
1.0
0.0
PLA PLA/RS20 PLA/RS20/E5
Materials
Fig. 7. Weight gained due to moisture absorption and weight loss due
to leaching of the RS particles of the PLA/RS composites.
arise as a result of the combined effect of hydrolysis of
the PLA matrix, the degradation of the PLAeRS
interface and the leaching of the RS particles. Upon
re-drying, the strength of the PLA/RS composites is
not fully recovered, which may be attributed to the
hydrolysis of the PLA. Note that the tensile properties
(i.e. tensile strength and elongation at break) of PLA in
re-dried (RD) state are relatively lower than that of wet
(W) state. The tensile properties of PLA could not be
recovered significantly upon subjecting to water absorp-
tion which may be attributed to hydrolysis. PLA could
absorb water, resulting in hydrolysis of ester linkages,
breaking down long macromolecular chains [26].
According to Spychaj [27], degradation of polyesters can
take place by hydrolysis or by other solvolytic reaction.
The hydrolysis and solvolytic reactions of polyesters
consist of the cleavage of CeO bonds of the polymer
backbone (i.e. ester bond cleavage).
to the formation of additional voids in the PLA/RS
composites, which would then be filled with water (i.e.
water entrapment in the matrix). Upon drying, these
voids will act as stress concentrators, which can then
initiate matrix cracking, leading to reductions in both
stiffness and strength of the composites. Thus, the
presence of moisture at the PLAeRS interface further
weakened the adhesion of PLA/RS composites. Similar
observation has been reported by Danjaji et al. [16], for
the LLDPE/sago starch composites.
PLA/RS20/E10 Note that the incorporation of ENR into PLA/RS
composites further decreased the retention and re-
coverability of the tensile properties for the PLA
composites subjected to water absorption (cf. Table 4).
Recall that, the maximum moisture absorption (Mm)
value of PLA/RS increased in the presence of ENR. As
the PLA was hydrolysed by water, chain scissions and
backbiting of the main chains of PLA might occur.
Thus, the adhesion between the PLA and RS, as well as
the interfacial interaction between PLA and ENR will
be further weakened. Accordingly, the tensile properties
of the PLA/RS composites further deteriorate. Embrit-
tlement due to the degradation of the macromolecular
skeleton by hydrolysis and damage localized at the
matrixefiller interface may result in deterioration of the
properties of organic materials, e.g. starch [28].
SEM micrographs taken from the fractured surface
of PLA/RS20 and PLA/RS20/E5, highlight the effects of
water absorption on the morphology of the composites
in AR, W and RD states are shown in Figs. 8aec
and 9aec, respectively. It can be observed that there is
lack of plastic deformation and fibrillated structure in
the RD states revealed for both PLA/RS20 and PLA/
RS20/E5 composites, as compared to AS and W states.

Table 4
Tensile properties of PLA/RS/ENR composites in control (AR), wet (W) and re-dried (RD) states
Materials Tensile properties Control Water absorption tests
Wet Re-dried
PLA Tensile modulus (GPa) 3.3 2.9 (87.9) 3.2 [97.0]
Tensile strength (MPa) 57.8 54.3 (93.9) 38.7 [67.0]
Elongation at break (%) 3.8 3.2 (84.2) 2.0 [52.6]
PLA/RS20 Tensile modulus (GPa) 3.8 2.1 (55.3) 3.4 [89.5]
Tensile strength (MPa) 44.0 30.1 (68.4) 20.5 [46.6]
Elongation at break (%) 2.5 2.2 (88.0) 1.3 [52.0]
PLA/RS20/E5 Tensile modulus (GPa) 3.3 1.8 (54.5) 2.6 [78.8]
Tensile strength (MPa) 49.0 23.2 (47.3) 19.1 [39.0]
Elongation at break (%) 3.4 1.9 (55.9) 1.2 [35.3]
PLA/RS20/E10 Tensile modulus (GPa) 2.8 1.6 (57.1) 2.1 [75.0]
Tensile strength (MPa) 37.2 14.6 (39.2) 12.2 [32.8]
Elongation at break (%) 2.6 1.4 (53.8) 1.1 [42.3]
Values in parenthesis represent percentage retention of the tensile properties after water absorption at temperature Z 30
C. Values in square
brackets represent percentage recovery of the tensile properties after re-drying.
 G.H. Yew et al. / Polymer Degradation and Stability 90 (2005) 488e500 497

Fig. 8. (aec) SEM micrographs taken from the fracture surface of

PLA/RS20 composites (note: a Z AR; b Z W and c Z RD samples). Fig. 9. (aec) SEM micrographs taken from the fracture surface of
PLA/RS20/E5 composites (note: a Z AR; b Z W and c Z RD samples).

In addition, voids were revealed on the tensile fracture PLA/RS composites. This is in agreement with the
surface of RD samples for these composites. Accord- observed embrittlement of the composites and explains
ingly, voids formation acts as stress-concentration the dramatic drop of the elongation at break values
points and further decrease the tensile properties of (cf. Table 4).
498 G.H. Yew et al. / Polymer Degradation and Stability 90 (2005) 488e500
3.5. Enzymatic tests
Table 5 shows the tensile properties of PLA/RS
composites after being subjected to enzymatic degrada-
tion. It can be seen that the tensile properties of PLA
and PLA/RS composites decreased upon exposure to
enzymatic degradation. Comparing the retention-ability
of the tensile properties between enzymatic and non-
enzymatic treated PLA, a further decrease of tensile
properties for PLA was observed for the former. This
seems to indicate that a-amylase enzymatic treatment
has little effect on the tensile properties of PLA. The
deteriorating effect of a-amylase on the PLA/RS is more
pronounced as evidenced by the significant reduction of
tensile properties of PLA/RS as compared to pure PLA
(cf. Table 5). a-Amylase are endoamylases which
catalyse the hydrolysis of internal a-1,4-glycosidic link-
ages in the starch in a random manner. The hydrolysis
of starch by a-amylase will further weaken the inter-
action between PLA and RS. In addition, leaching of
RS by a-amylase solution will result in the removal
of the RS from the surface of PLA (as evidenced by
the SEM micrographs shown in Fig. 10). This causes
a serious formation of voids which will act as stress
concentrators upon subjecting to tensile loads. Conse-
quently this will lead to the relatively low retention-
ability in tensile properties of PLA/RS. A similar
observation was earlier reported by Danjaji et al. [16]
for LLDPE/sago starch composites. Note that the
percentage drop in both tensile strength and elongation
at break of PLA/RS20/E5 is relatively higher than that
of PLA/RS20. Recall that the incorporation of ENR
into PLA/RS composites has further decreased the
tensile properties of PLA composites after being
subjected to water absorption. In our previous discus-
sion, it has been suggested that ENR could absorb more
water due to its polarity (oxirane groups). As a-amylase
solution contains distilled water, it is believed that
Table 5
Tensile properties of PLA/RS/ENR composites after being subjected to enzymatic treatments
Materials Tensile properties Control
PLA Tensile modulus (GPa) 3.3
Tensile strength (MPa) 57.8
Elongation at break (%) 3.8
PLA/RS20 Tensile modulus (GPa) 3.8
Tensile strength (MPa) 44.0
Elongation at break (%) 2.5
PLA/RS20/E5 Tensile modulus (GPa) 3.3
Tensile strength (MPa) 49.0
Elongation at break (%) 3.4
PLA/RS20/E10 Tensile modulus (GPa) 2.8
Tensile strength (MPa) 37.2
Elongation at break (%) 2.6
Values in parenthesis represent percentage retention of the tensile properties.
during enzymatic tests, PLA/RS20/E5 samples could
absorb more water/a-amylase solution than PLA/RS
composite does. This may facilitate the penetration of
a-amylase into the PLA/RS/E5 composites and
subsequently enhance the a-amylase attack on the RS.
Note that the retention-ability of tensile properties of all
enzymatic treated PLA/RS composites is lower than
that of PLA/RS that were subjected to distilled water i.e.
without a-amylase enzyme treatment (cf. Table 5). This
may indicate that PLA/RS composites are susceptible to
a-amylase enzyme treatment.
Fig. 10a and b shows SEM micrographs taken from
the sample surface of PLA/RS20 and PLA/RS20/E5
composites after being subjected to enzymatic degrada-
tion. Starch susceptibility to enzyme attack is influenced
by several factors, such as amylose and amylopectin
content, crystalline structure, granular structure, parti-
cle size and the presence of enzyme inhibitors. Hydro-
lysis pattern is an important factor for susceptibility to
a-amylase attack. Extensive surface erosion provides
a clear indication of the extent of hydrolysis taken place
in PLA/RS composites. According to Wang et al. [29],
the enzyme either erodes the entire/sections of the
granule surface or digests channels from selected points
on the surface toward the centre of the granule. Five
general patterns of attack have been identified, i.e. pin-
holes/pepper-potting, sponge-like erosion, many-medium
sized holes, distinct loci leading to single holes in
individual granules and surface erosion. Ishiaku et al. [1]
reported that extensive enzyme activity occurred in
a polycaprolactone/sago starch film after being sub-
jected to enzyme a-amylase. The majority of the sago
starch granules were removed from the surface of the
film leaving behind transparent voids. According to
Danjaji et al. [16], holes and agglomerates were observed
on LLDPE/sago starch samples after exposure to
a-amylase enzymatic tests, indicating the occurrence
of microbial activity. It is interesting to note that
Without enzyme Enzymatic degradation
3.0 (90.9) 2.8 (84.8)
56.1 (97.1) 48.5 (83.9)
3.5 (92.1) 2.6 (68.4)
2.8 (73.7) 2.2 (57.9)
35.8 (81.4) 23.2 (52.7)
2.2 (88.0) 1.1 (44.0)
2.5 (75.7) 2.0 (60.6)
30.5 (62.2) 21.9 (44.7)
2.3 (67.6) 1.0 (29.4)
2.4 (85.7) 1.9 (67.8)
28.7 (77.2) 13.6 (36.6)
2.0 (76.9) 0.9 (34.6)
 G.H. Yew et al. / Polymer Degradation and Stability 90 (2005) 488e500
Fig. 10. SEM micrographs taken from the surface of (a) PLA/RS20
and (b) PLA/RS20/E5 composites after being subjected to enzymatic
degradation.
the surface erosion of PLA/RS20/E5 (cf. Fig. 10b) is
relatively more severe than that of the PLA/RS20
composites (cf. Fig. 10a). More surface cracking and
pinhole could be observed in the PLA/RS20/E5 sample.
This is in agreement with the extent of deterioration of
tensile properties for PLA/RS20/E5 as compared to
PLA/RS20 composites (cf. Table 5).
4. Conclusions
Based on this work devoted to study the effect of rice
starch (RS) and epoxidized natural rubber (ENR) on the
mechanical, morphology, water absorption and enzy-
matic degradation behaviours of poly(lactic acid) (PLA)
composites, the following conclusions can be drawn:
1. Addition of RS increased tensile modulus of PLA
significantly. Twenty weight percent RS achieved
a good balance of tensile properties. The Nicolais
499
and Narkis model reveals that there is some degree
of interaction between PLA and RS. As expected,
elongation at break of the PLA composites decreased
with increasing RS. Transformation of the morphol-
ogy from brittle to more ductile was revealed by the
addition of ENR into PLA/RS composites. This was
attributed to the rubbery behaviour as well as the
compatibilisation effects of the ENR.
2. The kinetics of water absorption of the PLA/RS
composites conforms to Fick’s law of diffusion. The
Mm and D values are dependent on the RS and ENR
content. The tensile properties of the PLA/RS
composites deteriorated after water absorption.
The retention-ability and recoverability of the
PLA/RS is very low, owing to the hydrolysis of
PLA and degradation of the PLAeRS and PLAe
ENR interfaces, and leaching of the RS particles.
3. The tensile properties of PLA/RS composites de-
creased drastically upon exposure to enzymatic
degradation. Further evidence provided by SEM
showed extensive pinhole and surface erosion on the
PLA/RS composites indicating a high degree of
hydrolysis.
4. The addition of ENR improved the tensile proper-
ties of PLA/RS composites. Interestingly, ENR also
plays an important role in enhancing the degrad-
ability of the PLA/RS when the composites are
subjected to water and a-amylase enzymatic treat-
ments, attributed to its polarity (the presence of
oxirane groups).
Acknowledgements
The authors would like to thank The Ministry of
Science, Technology and Innovation (MOSTI), Malaysia
for the IRPA grant (Grant Number: 6012913/IRPA
Grant). We also thank Dr. Kitagawa (Kyoto Municipal
Research Institute, Japan) and Mitsui Chemicals Inc.
Japan for supplying PLA used in this study.
References
[1] Ishiaku US, Pang KW, Lee WS, Mohd Ishak ZA. Mechanical
properties and enzymatic degradation of thermoplastic and
granular sago starch filled poly(3-caprolactone). Eur Polym J
2002;38:393e401.
[2] Wang H, Sun XZ, Seib P. Strengthening blends of poly(lactic
acid) and starch with methylenediphenyl diisocyanate. J Appl
Polym Sci 2001;82:1761e7.
[3] Garlotta D, Doane W, Shogren R, Lawton J, Willett JL.
Mechanical and thermal properties of starch-filled poly(D,L-lactic
acid)/poly(hydroxy ester ether) biodegradable blends. J Appl
Polym Sci 2003;88:1775e86.
[4] Mohd Ishak ZA, Abu Bakar A, Ishiaku US, Hashim AS,
Azahari B. An investigation of the potential of rice husk ash as
a filler for epoxidized natural rubber-II. Fatigue behavior. Eur
Polym J 1997;33(1):73e9.
500 G.H. Yew et al. / Polymer Degradation and Stability 90 (2005) 488e500
[5] Hashim AS, Ong SK, Jessy RS. A general review of recent
developments on chemical modification of NR. Nat Rubber
2002;28(4th quarter):3e11.
[6] Hashim AS, Kohjiya S. Preparation and properties of epoxidized
natural rubber. Kautschuk Gummi Kunststoffe 1993;3:208e13.
[7] Ishiaku US, Mohd Ishak ZA. In: Shonaike GO, Simon GP,
editors. Polymer blends and alloys: developments in poly(vinyl
chloride)/epoxidized natural rubber blends. New York: Marcel
Dekker; 1999. p. 663e718.
[8] Sharma N, Chang LP, Chu YL, Ismail H, Ishiaku US, Mohd
Ishak ZA. A study on the effect of pro-oxidant on the thermo-
oxidative degradation behaviour of sago starch filled polyethylene.
Polym Degrad Stab 2001;71:381e93.
[9] Ke TY, Sun XZ. Physical properties of poly(lactic acid) and
starch composites with various blending ratios. Cereal Chem
2000;77(6):761e8.
[10] Ke TY, Sun XZ, Seib P. Blending of poly(lactic acid) and starches
containing varying amylose content. J Appl Polym Sci 2003;89:
3639e46.
[11] Pavlidou S, Papaspyrides CD. The effect of hygrothermal history
on water absorption and interlaminar shear strength of glass/
polyester composites with different interfacial strength. Compo-
sites Part A 2003;34:1117e24.
[12] Konsula Z, Liakopoulou-Kyriakides M. Hydrolysis of starches
by the action of the a-amylase from Bacillus subtilis. Process
Biochem 2004;39(11):1745e9.
[13] Zajoncová L, Jı́lek M, Beranová V, Pec P. A biosensor for the
determination of amylase activity. Biosens Bioelectron 2004;20(2):
240e5.
[14] Heitmann T, Weazig E, Mersmann A. Characterization of three
different potato starches and kinetics of their enzymatic hydrolysis
by an a-amylase. Enzyme Microb Technol 1997;20:259e67.
[15] Zhang T, Oates CG. Relationship between a-amylase degradation
and physico-chemical properties of sweet potato starches. Food
Chem 1999;65:157e63.
[16] Danjaji ID, Nawang R, Ishiaku US, Ismail H, Mohd Ishak ZA.
Degradation studies and moisture uptake of sago-starch filled linear
low-density polyethylene composites. Polym Test 2002;21:75e81.
[17] Araújo MA, Cunha AM, Mota M. Enzymatic degradation of
starch-based thermoplastic compounds used in prostheses:
identification of the degradation products in solution. Biomaterials
2004;25:2687e93.
[18] Mohd Ishak ZA, Berry JP. Hygrothermal aging studies of short
carbon fiber reinforced nylon 6.6. J Appl Polym Sci 1994;51:
2145e55.
[19] Shah PB, Bandopadhyay S, Bellare JR. Environmentally degrad-
able starch filled low density polyethylene. Polym Degrad Stab
1994;47:165e73.
[20] Thakore IM, Desai S, Sarawade BD, Devi S. Studies on
biodegradability, morphology and thermo-mechanical properties
of LDPE/modified starch blends. Eur Polym J 2001;37:151e60.
[21] Park HM, Lee SR, Chowdhury SR, Kang TK, Kim HK,
Park SH, et al. Tensile properties, morphology, and biodegrad-
ability of blends of starch with various thermoplastics. J Appl
Polym Sci 2002;86:2907e15.
[22] Nawang R, Danjaji ID, Ishiaku US, Ismail H, Mohd Ishak ZA.
Mechanical properties of sago starch-filled linear low density
polyethylene (LLDPE) composites. Polym Test 2001;20:167e72.
[23] Engel L, Klingele H, Ehrenstein GW, Schaper H. In: An atlas of
polymer damage e surface examination by SEM. Carl Hanser
Verlag: Wolfe Sc. Books; 1981. p. 136e211 [Welling MS, Trans.].
[24] Preechawong D, Peesan M, Supaphol P, Rujiravanit R. Charac-
terization of starch/poly(3-caprolactone) hybrid foams. Polym
Test 2004;23:651e7.
[25] Mani R, Bhattacharya M. Properties of injection moulded blends
of starch and modified biodegradable polyesters. Eur Polym J
2001;37:515e26.
[26] Copinet A, Bertrand C, Govindin S, Coma V, Couturier Y.
Effects of ultraviolet light (315 nm), temperature and relative
humidity on the degradation of polylactic acid plastic film.
Chemosphere 2004;55:763e73.
[27] Spychaj T. In: Fakirov S, editor. Handbook of thermoplastic
polyesters: chemical recycling of PET: methods and products.
Weinheim: Wiley-VCH; 2002. p. 1251e90.
[28] Merdas I, Thominette F, Tcharkhtchi A, Verdu J. Factors
governing water absorption by composite matrices. Compos Sci
Technol 2002;62:487e92.
[29] Wang WJ, Powell AD, Oates CG. Pattern of enzyme hydrolysis in
raw sago starch: effects of processing history. Carbohydr Polym
1995;26:91e7.