Polymer Testing 88 (2020) 106522

Contents lists available at ScienceDirect

Polymer Testing
journal homepage: http://www.elsevier.com/locate/polytest

Morphological characteristics and properties of TPS/PLA/cassava

pulp biocomposites
Phatcharin Jullanun a, Rangrong Yoksan a, b, *
a
Department of Packaging and Materials Technology, Faculty of Agro-Industry, Kasetsart University, Bangkok, 10900, Thailand
b
Center for Advanced Studies for Agriculture and Food, KU Institute for Advanced Studies, Kasetsart University, Bangkok, 10900, Thailand

A R T I C L E I N F O A B S T R A C T

Keywords: The effect of cassava pulp (CP) on morphological, tensile, and thermal properties of a thermoplastic cassava
Thermoplastic starch starch (TPS)/poly (lactic acid) (PLA) blend was investigated. TPS/PLA/CP biocomposites were manufactured by
poly(lactic acid) melt extrusion and then converted into specimens using an injection molding. The weight fraction of PLA to TPS/
Cassava pulp
CP was fixed at 40:60, whereas the final CP concentration in the composites was varied in the range of 4.4–22.1
Biocomposite
Extrusion
wt%. CP could act as a reinforcement for TPS/PLA blend to enhance its tensile strength up to 354% and Young’s
Microstructure modulus up to 722% when 22.1 wt% of CP was loaded and a nucleating agent for PLA as confirmed from the
reduced Tcc. In addition, TPS/PLA/CP composites showed a discrete phase structure (i.e., droplets in matrix)
when CP with lower concentration (i.e., 4.4 wt%, 8.8 wt%, and 13.3 wt%) was incorporated and a bicontinuous
phase structure (i.e., co-continuous) when higher concentration of CP (i.e., 17.7 wt% and 22.1 wt%) was
employed. The results suggest that TPS/PLA/CP biocomposites have potential to be used in the manufacturing of
injection-molded articles, particularly when biodegradability and renewability of the material are required.

1. Introduction of starch [4]. The obtained TPS can be converted into final products
using the same technologies/machines as the ones used for conventional
In recent years, as a result of the increasingly strict environmental plastics.
regulations, extensive research aiming to develop biodegradable mate­ Cassava starch is the fifth most abundant starch produced in the
rials based on agricultural resources to partially replace nonbiodegrad­ world, and it is manufactured in high quantity in Thailand [5]. During
able conventional plastics has been conducted. Based on an economic the production of cassava starch, cassava pulp (CP) or bagasse is ob­
and availability point of view, starch is one of the most promising tained as a byproduct at about 14–20 wt% [6], and it contains 50–60%
renewable materials. Starch is a polysaccharide composed of linear of residual starch (dry weight basis) and 14–22% of fibers (dry weight
amylose and branched amylopectin, and it is a semicrystalline polymer basis) [6–9]. A major application of CP is as a feedstock for solid-state
in which hydroxyl groups of starch form inter- and intramolecular fermentation to produce protein enrichment, aroma compounds, pig­
hydrogen bonds in crystalline regions causing high melting temperature. ments [10], and organic acids such as citric acid [10,11] and lactic acid
In general, granular starch is degraded before it melts because its [12]. The rest of CP is simply discarded as a wasted resource [13]. Due to
melting point is higher than decomposition temperature [1]; therefore, inexpensive, biodegradable, and rich fiber content material, CP has been
granular starch has been used only as a filler for low melting tempera­ used either as a reinforcement for solution-cast cassava starch film [14]
ture thermoplastics. Converting granular starch into thermoplastic ex­ or as a raw material for thermoplastic [15,16]. To the best of our
pands the utilization of starch in the plastic industry. Thermoplastic knowledge, only one research group (i.e., Teixeira et al. [15,16]) has
starch (TPS) is produced by a plasticization of starch with plasticizers, reported using CP as a raw material for producing TPS. They found that
such as glycerol, water, glucose, fructose, and sucrose [2], under the fibers (17 wt%) in CP could act as reinforcement for thermoplastic from
action of shear and high temperature [3]. During plasticization, plasti­ CP as confirmed from the increased tensile strength and Young’s
cizers play an indispensable role to form hydrogen bonds with starch, modulus. Unfortunately, TPS has several unfavorable properties
while destroying the original hydrogen bonds between hydroxyl groups including high moisture uptake [17,18], high shrinkage in the mold

* Corresponding author. Department of Packaging and Materials Technology, Faculty of Agro-Industry, Kasetsart University, Bangkok, 10900, Thailand.
E-mail address: [email protected] (R. Yoksan).

https://doi.org/10.1016/j.polymertesting.2020.106522
Received 21 November 2019; Received in revised form 22 March 2020; Accepted 27 March 2020
Available online 18 April 2020
0142-9418/© 2020 Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
P. Jullanun and R. Yoksan Polymer Testing 88 (2020) 106522

[19], relatively weak mechanical properties [17,18], and high viscosity Ltd., Thailand. Poly (lactic acid) (Ingeo™ Biopolymer 3052D) was ob­
[20]. tained from NatureWorks LLC, USA. Citric acid monohydrate (98%
Blending TPS with other biodegradable polymers such as poly purity) and potassium iodide (KI) were purchased from Ajax Finechem
(caprolactone) [19], poly (vinyl alcohol) [21] or poly (lactic acid) (PLA) Pty Limited, Germany. Chloroform (HPLC grade) was purchased from
[22] is an alternative to overcome the limitation of TPS and meanwhile RCI Labscan Limited, Thailand. Hydrochloric acid fuming (37%) was
retain its biodegradability. PLA is a biodegradable thermoplastic poly­ supplied by Merck, Germany. Iodine was purchased from Sigma-Aldrich,
ester derived from either condensation polymerization of L- or D-lactic USA. The glycerol used was a commercial grade product.
acid or ring-opening polymerization of lactides. PLA is of particular
interest due to good mechanical properties [22] and high transparency
[23]; however, the brittleness [22] and the relatively high price 2.2. Preparation of TPS/PLA/CP biocomposites
compared to the conventional plastics has constrained it uses. Therefore,
blending TPS with PLA is an interesting approach to overcome the Cassava starch, CP, glycerol, and citric acid were physically mixed in
drawbacks of TPS and to reduce the overall cost of the final PLA-based a ribbon mixer (Tong Hor Machine Lex Product, China). The obtained
products. However, the two polymers are incompatible upon their mixture was then melt blended in a twin-screw extruder with an L/D
mixing due to high interfacial tension between hydrophilic TPS and ratio of 40 (LTE-20-40, Labtech Engineering Co., Ltd., Thailand), using a
hydrophobic PLA when glycerol was used as a plasticizer for TPS barrel temperature profile in the range of 80–160 � C, a screw speed of
[24–26]. Chabrat, Abdillahi, Rouilly and Rigal [27], Teixeira, Curvelo, 280 rpm, and a material feed rate of 40–50 rpm, to obtain a TPS-based
Corr^ea, Marconcini, Glenn and Mattoso [16] and Wang, Yu, Chang and CP masterbatch. The extrudates of the masterbatch were cut into 2.5
Ma [28] reported that citric acid, an environmentally nonhazardous mm-length pellets using a pelletizer (LZ-120, Labtech Engineering Co.,
substance, could be used as a compatibilizer for PLA blended with Ltd., Thailand). A weight ratio of cassava starch to CP to glycerol of
thermoplastic wheat flour [27], thermoplastic cassava starch [16], 50:50:35 and citric acid concentration of 0.5 wt% were used to prepare
thermoplastic from CP [16], and corn starch [28]. the TPS-based CP masterbatch. TPS was also produced using the same
Teixeira, Curvelo, Corr^ea, Marconcini, Glenn and Mattoso [16] protocol as mentioned above; however, CP was not incorporated and
found that the blend of thermoplastic CP/PLA possessed higher tensile hence a weight ratio of cassava starch to glycerol was 100:35.
strength and modulus than the blend of thermoplastic cassava starch/­ The resulting TPS-based CP masterbatch, TPS, and PLA pellets were
PLA when the same weight fraction of TPS to PLA of 20:80 was used. In subsequently fed into a twin-screw extruder using a temperature profile
addition, fibers in CP also enhanced PLA crystallinity due to the nucle­ in the range of 85–175 � C, a screw speed of 280 rpm, and a material feed
ating effect. Until now, only a few publications have reported the uti­ rate of 8 rpm. Different weight ratios of PLA to TPS to TPS-based CP
lization of CP waste as raw material for TPS and only one article masterbatch of 40:60:0, 40:48:12, 40:36:24, 40:24:36, 40:12:48, and
reported about its blending with PLA [16]. In that case, CP was used as a 40:0:60 were used to prepare the TPS/PLA blend, TPS/PLA/CP1, TPS/
highly underutilized and inexpensive feedstock for making PLA/CP2, TPS/PLA/CP3, TPS/PLA/CP4, and TPS/PLA/CP5 composites,
fiber-reinforced TPS and PLA/TPS blends [16]. The materials were respectively. The obtained blend and composites possessed CP concen­
prepared using a twin-screw extruder and converted into test specimens trations of 0, 4.4, 8.8, 13.3, 17.7, and 22.1 wt% (Table 1) and fiber
using a hot press machine. Although CP was completely used to replace concentrations of 0, 0.84, 1.68, 2.52, 3.36, and 4.20 wt%, respectively.
cassava starch in TPS, its amount in the final blends was in the range of The extrudates of TPS/PLA blend and TPS/PLA/CP composites were
2.3–9.2 wt%, corresponding to 5–20 wt% of TPS plasticized with both pelletized into 2.5 mm-length pellets using a pelletizer. Regarding
glycerol and water. In addition, citric acid and stearic acid were used as Table 1, glycerol was used as a plasticizer for TPS based on both cassava
processing aids. Morphology by SEM, thermal properties by DSC, tensile starch and CP; therefore, its amount was constant at 15.48 wt% for all
property, and dynamic mechanical thermal property by DMTA of the TPS/PLA/CP composites. Citric acid was herein applied as a compati­
blends were discussed. Tensile strength and Young’s modulus of the bilizer for TPS and PLA phases because it has been reported to accelerate
blends decreased as a function of CP content. However, CP might pro­ dissolution of starch granules [3,29] and depolymerize PLA chains [30],
vide different effect to the blends with different composition, particu­ resulting in reduced viscosity of the blend and improved dispersion
larly a weight proportion of TPS to PLA. In addition, the phase structure between TPS and PLA [31]. Hence, the amount of citric acid was fixed at
of the blend should be intensively discussed. 0.3 wt% because the TPS/PLA weight proportion was constant.
Therefore, the objectives of the present work are to reduce the uti­ The pellets of TPS/PLA blend and TPS/PLA/CP composites were
lization of PLA by replacing with TPS produced from CP and cassava dried in a hot air oven at 50 � C overnight to remove the residual mois­
starch, which are cheaper, and to study the effect of CP and other ture and after that converted into dumbbell-shaped specimens by an
components on morphological, rheological, tensile, and thermal prop­ injection molding machine (Model SG50 M, Sumitomo Heavy Industries,
erties of TPS/PLA/CP biocomposites. The biocomposites were fabricated Co., Ltd., Japan) using a temperature profile along the extruder barrel in
in the presence of citric acid compatibilizer using melt processes (i.e., the range of 140–245 � C, a back pressure of 3%, and a cooling time of 14
extrusion and injection molding) with a high weight proportion of TPS s.
to PLA of 60:40 and various concentrations of CP, i.e., 4.4 wt%, 8.8 wt%,
13.3 wt%, 17.7 wt%, and 22.1 wt% based on the final composite. Table 1
Composition of TPS/PLA blend and TPS/PLA/CP composites.
2. Materials and methods Sample PLA (wt Cassava starch CP (wt Glycerol Citric acid
%) (wt%) %) (wt%) (wt%)
2.1. Materials TPS/PLA 40.00 44.22 0.00 15.48 0.30
TPS/PLA/ 40.00 39.80 4.42 15.48 0.30
Cassava starch (with total carbohydrate content of 88.0 � 0.04 wt%, CP 1
moisture content of 11.7 � 0.02 wt%, protein content of 0.2 � 0.002 wt TPS/PLA/ 40.00 35.38 8.84 15.48 0.30
CP 2
%, lipid content of 0.04 � 0.005 wt%, and ash content of 0.06 � 0.01 wt
TPS/PLA/ 40.00 30.96 13.27 15.48 0.30
%) was supplied by Tong Chan Registered Ordinary Partnership, CP 3
Thailand. CP (with starch content of 62.8 � 1.4 wt%, moisture content TPS/PLA/ 40.00 26.53 17.69 15.48 0.30
of 10.4 � 0.1 wt%, protein content of 3.8 � 0.07 wt%, lipid content of CP 4
0.1 � 0.02 wt%, ash content of 3.3 � 0.06 wt%, and fiber content of TPS/PLA/ 40.00 22.11 22.11 15.48 0.30
CP 5
18.9 � 0.2 wt%) was provided by Siam Modified Starch Corporation Co.,

2
P. Jullanun and R. Yoksan Polymer Testing 88 (2020) 106522

2.3. Characterization and property testing of TPS/PLA/CP biocomposites 2.3.7. Thermogravimetric analysis
Thermal stability of samples was analyzed by thermogravimetric
2.3.1. Melt flow index measurement analysis (TGA) using a TGA/DSC 1 STARe System Thermogravimetric
The melt flow index (MFI) of the sample was measured according to Analyzer (Greifensee, Switzerland) with TA evaluation software. The
ASTM 1238. The measurement was performed using an MFI2-203 sample (6–8 mg) was placed in a ceramic crucible and heated from 30 � C
(Custom Scientific Instruments, USA) at a temperature of 190 � C with to 600 � C at a heating rate of 10 � C/min under a nitrogen atmosphere
a load cell of 2.16 kg and a preheating time of 7 min. MFI was measured with a nitrogen flow rate of 50 mL/min.
in triplicate for each sample and reported in g/10 min as mean � SD (n
¼ 3). 2.3.8. Statistical analysis
The results were statistically analyzed using IBM SPSS Statistics
2.3.2. Rheological study Version 20 (Statsoft, Oklahoma), with analysis of variance (ANOVA) and
Rheological characteristics of the sample were determined using a Duncan’s test at significant differences at 5% level.
twin-bore capillary rheometer (Rosand RH7, Malvern, Germany). The
sample was loaded into a capillary rheometer barrel, in which the 3. Results and discussion
temperature was controlled at 155 � C, and then the sample was pushed
through a capillary using a plunger. The measurement was performed at Cassava starch used in this work consisted mainly of carbohydrate
a shear rate in the range of 10–10000 s 1. (88.0 wt%), attributing to amylose and amylopectin. Other minor
components were moisture, protein, lipid, and ash with contents of 11.7
2.3.3. Morphological observation by scanning electron microscopy wt%, 0.2 wt%, 0.04 wt%, and 0.06 wt%, respectively. Cassava starch
Microstructures of the sample were observed using a scanning elec­ granules are round or oval shape with a diameter in the range of
tron microscope (FEI Quanta 450, Thermo Fisher Scientific, Massachu­ 3.5–19.5 μm (Fig. S1Aa and Ba). On the other hand, CP, a byproduct of
setts, USA) at an accelerating voltage of 10 kV. A small piece of each cassava starch manufacturing, contained approximately 62.8 wt% of
sample was placed on a stub using a two-sided carbon tape and then starch and 18.9 wt% of fiber. High concentration of starch in CP came
dried at 40 � C in a vacuum oven for 6 h. After that it was coated with a from the inextricable starch trapped inside the cell of cassava root. The
thin layer of gold prior to SEM observation. other compositions of CP included 10.4 wt% of moisture, 3.8 wt% of
In order to determine the phase morphologies of the blend and protein, 0.1 wt% of lipid, and 3.3 wt% of ash. Fiber was the only
composites, selectively extracted samples were prepared. Cryofractured component of CP which made it different from cassava starch. From SEM
specimens of each sample were immersed in 10.2 N of hydrochloric acid image of CP, starch granules were observed as discrete particles with a
(HCl) solution at 25 � 2 � C for 8 h to remove TPS phase or chloroform at diameter in the range of 5.5–17.5 μm and interspersed on the surface of
25 � 2 � C for 48 h to remove PLA phase. The etched specimens were fibers (Fig. S1Ab). It should be pointed out that starch granules in CP
washed with ethanol (99%) and dried under reduced pressure at 35 � C showed narrower size distribution (Fig. S1Bb) than native cassava starch
for 4 h. granules (Fig. S1Ba). The diameter and length of the fibers presented in
CP were in the ranges of 14.5–94.5 μm and 24.5–324.5 μm, respectively
2.3.4. Tensile testing (Fig. S1Ab and S1Bc).
Dumbbell-shaped specimens of each sample were conditioned in a
desiccator containing a saturated solution of magnesium nitrate at 25 � 3.1. Melt flow ability of TPS/PLA/CP biocomposites
2 � C (52 � 2% RH) for 2 days prior to testing. The preconditioned
specimens were then tested according to ASTM D638 by a QC-506B2 MFI of the TPS/PLA blend and TPS/PLA/CP composites is shown in
Universal Testing Machine (Cometech Testing Machines Co., Ltd., Fig. 1. The TPS/PLA blend exhibited MFI of 36.7 g/10 min, while the
Taiwan) with a load cell of 50 kN. The test was performed using a MFI values of TPS/PLA/CP composites were in the range of 5.4–9.9 g/
crosshead speed of 50 mm/min, a grip separation of 115 mm, and a gage 10 min, suggesting that TPS/PLA/CP composites possessed poorer melt
length of 100 mm. At least three specimens were tested for each sample. flow ability than the TPS/PLA blend, probably because the fibers in CP
Thickness of the dumbbell-shaped specimens was measured by an ID-
C112BS micrometer (Mitutoyo, Japan) in the range of 3.280–3.355
mm. The average thickness of each dumbbell-shaped specimen was
calculated from five positions. Tensile strength (MPa), modulus of
elasticity (MPa), and elongation at break (%) were recorded as mean �
SD (n ¼ 3).

2.3.5. X-ray diffraction analysis

The X-ray diffraction (XRD) pattern of the sample was recorded by a
Bruker D8 Advance Analyzer (Bruker, Bremen, Germany). Each sample
was scanned in a diffraction angle 2θ range of 5–40� using a scan rate of
0.02� /sec.

2.3.6. Differential scanning calorimetry analysis

Thermal properties of the sample were determined by differential
scanning calorimetry (DSC) using a model DSC 1, STARe system,
Mettler-Toledo (Greifensee, Switzerland). The sample was weighed
(6–8 mg) and packed in an aluminum pan. After that it was heated from
25 � C to 250 � C at a heating rate of 10 � C/min for the first heating scan,
held at 250 � C for 2 min, then cooled to 0 � C at a cooling rate of 10 � C/
Fig. 1. MFI measured at 190 � C with a load cell of 2.16 kg of (a) TPS/PLA
min and held at 0 � C for 2 min. Subsequently, it was reheated from 0 � C blend, (b) TPS/PLA/CP1 composite, (c) TPS/PLA/CP2 composite, (d) TPS/PLA/
to 250 � C at a heating rate of 10 � C/min. The measurement was per­ CP3 composite, (e) TPS/PLA/CP4 composite and (f) TPS/PLA/CP5 composite.
formed under a nitrogen atmosphere with a nitrogen flow rate of 50 mL/ Data are reported as mean � SD, n ¼ 3. Different lower-case letters indicate a
min. An empty aluminum pan was used as a reference. significant difference at p < 0.05 (Duncan’s new multiple range test).

3
P. Jullanun and R. Yoksan Polymer Testing 88 (2020) 106522

hindered the mobility of the polymer matrix, resulting in increased increasing fiber concentration resulting in increased viscosity. However,
apparent viscosity of the material. A similar result was found in the case when the shear rate increases, most of fibers locate closer to the tube
of Ecovio® biodegradable polyester reinforced with curaua fibers [32]. wall and align along the tube axis causing much lower fiber–fiber
It should be pointed out that the high melt flow ability of both the collision and consequently the increase in viscosity as a function of fiber
TPS/PLA blend as compared with those of neat TPS (1.32 g/10 min) and concentration is less pronounced at high shear rates. Fig. 2 also shows
neat PLA (8.03 g/10 min) came from the presence of citric acid. Wang, that shear viscosity of the TPS/PLA/CP composites slightly increased
Yu, Chang and Ma [28] revealed that citric acid could depolymerize with increasing CP concentration.
both starch and PLA chains in the PLA/thermoplastic dry starch blends Flow behavior index or power-law index (n) and power-law consis­
(50/50 w/w); as a result, viscosity of the blend decreased. tency (K) of the TPS/PLA blend and TPS/PLA/CP composites were
MFI values of TPS/PLA/CP composites gradually decreased as a determined and summarized in Table 2. Both TPS/PLA blend and TPS/
function of CP concentration. The reduction of flow ability of the com­ PLA/CP composites possessed n values lower than 1, suggesting that
posites might be explained by the stiffening of the material by CP these materials were pseudoplastic or shear thinning fluids, similar to
leading to more resistance to the flow of the TPS/PLA blend [33]. most polymer melts [34]. The value of n is used to describe the deviation
of flow behavior of a material from the Newtonian fluids. Here, the n
value of TPS/PLA blend was 0.45, while those of TPS/PLA/CP com­
3.2. Rheological properties of TPS/PLA/CP biocomposites
posites were in the range of 0.37–0.40, indicating that the flow behavior
of TPS/PLA/CP composites exhibited higher deviation from the New­
The plots between shear viscosity (η) and shear rate (_γ) determined
tonian shear rate or was more affected by the shear rate.
by a capillary rheometer of TPS/PLA blend and TPS/PLA/CP composites
In addition, Table 2 also shows that the TPS/PLA blend possesses K of
are presented in Fig. 2. Shear viscosity of both TPS/PLA blend and TPS/
3943 Pa sn, whereas the value of K increases to 6685–8229 Pa sn by
PLA/CP composites decreased with increasing shear rate, reflecting a
incorporating CP with a concentration in the range of 4.4–22.1 wt%,
typical shear-thinning behavior of the materials. This might be
suggesting the increased average viscosity of the TPS/PLA/CP
explained by the alignment or arrangement in the direction of applied
composites.
shear stress of polymer chain segments [34,35].
The above information from both MFI and rheological studies is
TPS/PLA/CP composites exhibited higher shear viscosity than TPS/
quite interesting for plastic converting industry. From an industrial
PLA blend, implying that CP obstructed the mobility of polymer chain
point of view, at high shear rate (102 104 s 1 [36,39]) the shear vis­
segments in the flow [36]. This result is in good agreement with the
cosity of the TPS/PLA/CP composites was close to that of the TPS/PLA
decreased MFI value. The difference in apparent viscosity between the
blend, suggesting that the power used to drive the injection molding
TPS/PLA blend and TPS/PLA/CP composites was more pronounced at
machine for the composites was not different from that for the blend
low shear rate and it became less when the shear rate was further
even though CP with a content of up to 22 wt% was incorporated. In
increased. This behavior was believed to be affected by CP fibers, which
addition, very little additional energy was required to manufacture the
were in the range of 0.84–4.2 wt%. According to the reports of Crowson
profile products using an extruder, which is operated at the intermediate
and Folkes [37], Crowson, Folkes and Bright [38] and George, Janard­
range of shear rate (101 103 s 1 [39]). In contrast, at low shear rate
han, Anand, Bhagawan and Thomas [36], fibers are randomly dispersed
(100 101 s 1 [39]) TPS/PLA/CP composites possessed higher viscosity,
in the absence of shear, and they tend to align themselves along the
measured either by an MFI analyzer or by a capillary rheometer, than
direction of shear when the force was applied, thus reducing the vis­
TPS/PLA blend, indicating that more power or force were consumed to
cosity. The degree of fiber alignment is relevant to the deformation rate,
produce compression molded product of TPS/PLA/CP composites.
i.e. fibers are less oriented at low shear rate. The collision of misaligned
fibers is possibly much higher than the aligned ones and it increases with
3.3. Morphological characteristics of TPS/PLA/CP biocomposites by SEM

Cryofractured surface morphologies of TPS/PLA blend and TPS/

PLA/CP composites were investigated (Fig. 3A). TPS/PLA blend without
citric acid exhibited phase separation and voids at the interfaces be­
tween two components (Fig. S2), suggesting immiscible and incompat­
ible blend. The compatibility between TPS and PLA phases was
improved by adding citric acid as no void at the interfaces was observed
(Fig. 3Aa).
In order to understand the phase structure of the materials, the
compatibility between PLA and TPS phases, and the distribution and size

Table 2
Power-law parameters of TPS/PLA blend and TPS/PLA/CP composites.
Sample Power-law index Consistency (K, Correlation coefficient
(n) Pa⋅sn) (R2)

TPS/PLA 0.447 � 0.007a 3942.6 � 138.0c 0.9941 � 0.0018

TPS/PLA/ 0.372 � 0.003cd 7722.5 � 600.4ab 0.9951 � 0.0004
CP1
TPS/PLA/ 0.365 � 0.009d 8228.9 � 346.9a 0.9982 � 0.0001
CP2
TPS/PLA/ 0.373 � 0.001cd 7463.7 � 45.0ab 0.9977 � 0.0001
CP3
TPS/PLA/ 0.387 � 0.007bc 6684.9 � 881.2b 0.9987 � 0.0004
CP4
TPS/PLA/ 0.398 � 0.012b 7251.2 � 520.2ab 0.9978 � 0.0003
Fig. 2. Relationship between shear viscosity (η) and shear rate ( ) of TPS/PLA
CP5
blend (●), TPS/PLA/CP1 composite ( ), TPS/PLA/CP2 composite ( ), TPS/
PLA/CP3 composite ( ), TPS/PLA/CP4 composite ( ), and TPS/PLA/CP5 Data are reported as mean � SD, n ¼ 2. Different superscript letters indicate a
composite ( ). significant difference at p < 0.05 (Duncan’s new multiple range test).

4
P. Jullanun and R. Yoksan Polymer Testing 88 (2020) 106522

Fig. 3. SEM micrographs at 10 kV and 3000 � of (A) cryofractured surface, (B) cryofractured surface extracted with chloroform, and (C) cryofractured surface
extracted with aqueous HCl solution (10.2 N) of (a) TPS/PLA blend, (b) TPS/PLA/CP1 composite, (c) TPS/PLA/CP2 composite, (d) TPS/PLA/CP3 composite, (e) TPS/
PLA/CP4 composite, and (f) TPS/PLA/CP5 composite.

of the dispersed phase, a selective extraction of all cryofractured samples The TPS/PLA blend showed a smooth cryofractured surface with
was performed using chloroform (CHCl3) and 10.2 N of aqueous HCl small spherical droplets (with a dimension of less than 1 μm) dispersed
solution as extraction solvents to remove PLA and TPS phases, respec­ throughout the matrix (Fig. 3Aa). Such small spherical droplets dis­
tively. The cryofractured surface morphologies of the etched sample appeared for the specimen extracted with CHCl3 (Fig. 3Ba), confirming
were examined and presented in Fig. 3B and C. that those droplets belonged to PLA. It should be noted that the TPS/PLA

5
P. Jullanun and R. Yoksan Polymer Testing 88 (2020) 106522

blend was completely dissolved in HCl solution, reflecting that TPS was rod-like structure (Fig. 3b d) and eventually with further increasing CP
a matrix of the blend. Therefore, the TPS/PLA blend was categorized as a concentration the morphology of the composites was changed to a co-
blend with discrete phase structure (i.e., droplets in matrix), in which continuous phase structure (Fig. 3e f) because of the enhanced melt
the minor component of PLA was finely dispersed in the TPS matrix. viscosity of TPS/citric acid, or in other words the decreased viscosity
Ayana, Suin and Khatua [40] also found that the TPS/PLA blend (60/40 ratio which was finally approaching 1 [44].
wt/wt) displayed a characteristic matrix-droplet model with droplets of Another possible reason for the compatibility improvement between
PLA dispersed in the TPS matrix. However, their work focused on the TPS and PLA by incorporating CP might also come from the increased
nanocomposites of TPS/PLA/clay. glycerol/starch weight ratio (Table 1), particularly when small amounts
In contrast, TPS/PLA/CP composites exhibited rougher cryofrac­ of CP (i.e., 4.4 wt%, 8.8 wt%, and 13.3 wt%) were loaded in which the
tured surfaces with different detailed characteristics according to the agglomeration of CP was negligible. However, at a higher concentration
amount of CP loaded (Fig. 3Ab f). Surprisingly, TPS/PLA/CP compos­ of CP (i.e., 17.7 wt% and 22.1 wt%) the morphological characteristics of
ites were incompletely soluble in aqueous HCl solution, suggesting that the TPS/PLA/CP composite were more affected by CP agglomeration
TPS was not only a component in the matrix of composites, but PLA than glycerol plasticization.
might also coexist. Fig. 3Ab c shows that PLA-rich dispersed phase Although Teixeira, Curvelo, Corr^ ea, Marconcini, Glenn and Mattoso
became smaller and more elongated with more uniform distribution in [16] also reported about the blends of PLA and TPS from CP (with
TPS-rich matrix when 4.4 wt% and 8.8 wt% of CP were loaded for TPS/ various weight fractions of PLA to TPS of 95:5, 90:10, 85:15, and 80:20),
PLA/CP1 and TPS/PLA/CP2 composites, respectively, implying the their morphological study was performed only at the cryofractured
improved interfacial adhesion between TPS and PLA phases. The change surfaces. However, the different weight proportions of PLA to TPS and
from discrete droplets of PLA dispersed phase to long ligament and to the content of other components, i.e., CP and glycerol plasticizer, pro­
smaller particles was also confirmed by SEM images of the specimens vided the composites with different phase structures. The investigation
extracted with CHCl3 (Fig. 3Bb c) and HCl solution (Fig. 3Cb c). A of phase structures of the materials is thus very important because they
similar observation was found when PLA-g-MA compatibilizer was have relationship with properties. Herein, the phase structure of the
added into the TPS/PLA blend [41]. This result suggested that the TPS/PLA/CP composites was intensively examined using a selective
compatibility between TPS and PLA phases was promoted by incorpo­ extraction with HCl and CHCl3. As discussed above, the phase structure
rating CP, probably due to the increased shear force induced during the of the composites significantly changed when CP with different contents
melt blending process, leading to better distribution of the dispersed was added.
phase. Raghu, Kale, Raj, Aggarwal and Chauhan [42] also found that
PLA40/MA-g-PLA10/TPS50 blend exhibited fine dispersion of TPS 3.4. Tensile properties of TPS/PLA/CP biocomposites
phase and good adhesion between TPS dispersed phase and PLA matrix
when wood fiber (20 wt%) was added. In addition, the presence of The neat PLA showed higher tensile strength and Young’s modulus,
particles has been reported to reduce the interfacial tension between two but lower elongation at break (65 MPa, 2482 MPa, and 4.5%, respec­
polymer phases [43]. tively) than the neat TPS (9 MPa, 399 MPa, and 7.4%, respectively).
It should be noticed that TPS/PLA/CP3 composite with 13.3 wt% of After blending PLA with TPS (in the absence of citric acid), the obtained
CP possessed a more homogenous cryofractured surface with less PLA TPS/PLA blend showed decreased tensile strength, Young’s modulus,
dispersed phase (Fig. 3Ad) as compared to TPS/PLA/CP1 and TPS/PLA/ and elongation at break to 41.4 MPa, 2043 MPa, and 2.7%, respectively
CP2 composites. In addition, both PLA and TPS phases were hardly as compared with PLA. The poorer tensile properties of the blend was
removed from the TPS/PLA/CP3 composite extracted with CHCl3 due to the effect of TPS [45]. However, in the presence of citric acid, the
(Fig. 3Bd) and HCl solution (Fig. 3Cd), respectively. The results reduction of tensile strength and Young’s modulus of the blend was
confirmed the enhanced compatibility between TPS and PLA phases. A more pronounced (5.2 MPa and 34.7 MPa, respectively), but its elon­
similar phenomenon was explained by Chabrat, Abdillahi, Rouilly and gation at break increased to 5.7%, most likely due to the improved
Rigal [27] and Wang, Yu, Chang and Ma [28] as they also found that PLA compatibility between TPS and PLA phases as confirmed from SEM
dispersed in the TPS/PLA blend was less extractable with chloroform images (Fig. S2 vs Fig. 3Aa). Chabrat, Abdillahi, Rouilly and Rigal [27]
when citric acid compatibilizer was incorporated. also reported that citric acid functioned as a compatibilizer for starch
Further increase in CP content to 17.7 wt% and 22.1 wt% caused and PLA by promoting their depolymerization. Fig. 4 shows tensile
phase separation with void formation at interphases between TPS and strength, Young’s modulus, and elongation at break of the TPS/PLA/CP
PLA of TPS/PLA/CP4 and TPS/PLA/CP5 composites (Fig. 3Ae f), composites as compared with those of the TPS/PLA blend. TPS/PLA/CP
probably due to the agglomeration of CP at high concentration. After composites showed significantly increased tensile strength and Young’s
extracting with both CHCl3 and HCl solution, larger PLA or TPS phases modulus to 18.1–23.6 MPa (248–354% increment) and 235.3–285.3
were removed, respectively (Fig. 3Be f and Fig. 3Ce f), indicating MPa (577–722% increment), but decreased elongation at break to
outstanding co-continuous phase of TPS and PLA. This might be a result 2.7–4.2% (26–53% reduction). The results indicated that CP could
of the viscosity ratio (ηdispersed phase/ηmatrix) close to unity [44]. It should improve tensile strength and stiffness, but decreased extensibility of the
be noted that MFI of TPS/citric acid is too high to measure at 190 � C with TPS/PLA blend, possibly due to the reinforcing effect by CP fibers, which
a load cell of 2.16 kg, implying very low viscosity due to the hydrolysis is the second most abundant composition of CP (19 wt%). A similar
of starch molecules catalyzed by citric acid [28]. However, when CP was result was reported by Teixeira, Curvelo, Corr^ea, Marconcini, Glenn and
loaded the MFI of TPS/citric acid decreased and the reduction of this Mattoso [16]; thermoplastic cassava bagasse/PLA (20/80) exhibited
value was more pronounced by increasing CP content (data not shown). 20% higher tensile strength and 40% higher modulus than thermoplastic
The result suggested that CP enhanced melt viscosity of TPS/citric acid starch/PLA (20/80). In addition, the incorporation of other fibers (e.g.,
and brought it close to that of PLA; therefore, the two coexisting phases sisal fibers, hemp fibers, and cellulose nanofibers) into TPS also pro­
interconnected throughout the whole blend volume. moted its rigidity [17,46].
From the above results, it was clearly seen that the morphology of Tensile strength, Young’s modulus, and elongation at break of TPS/
TPS/PLA/CP composites in the presence of citric acid was highly PLA/CP composites increased with increasing CP concentration in the
influenced by the viscosity ratio. As mentioned above, the melt viscosity range of 4.4–13.3 wt% (Fig. 4b d). The improvement of tensile
of TPS/citric acid was much lower than that of PLA; therefore, when strength, stiffness, and extensibility of these composites was explained
TPS/citric acid and PLA were blended, a discrete phase structure (i.e., by the enhanced reinforcing effect by CP fibers (0.84–2.52 wt%) and the
PLA droplets in TPS matrix) was observed (Fig. 3a). When CP concen­ improved compatibility between TPS and PLA phases as confirmed by
tration was increased, the PLA dispersed phase became an elongated SEM. Cao, Chen, Chang, Stumborg and Huneault [47] also reported that

6
P. Jullanun and R. Yoksan Polymer Testing 88 (2020) 106522

Fig. 4. (A) Tensile strength, (B) Young’s modulus, and (C) elongation at break of (a) TPS/PLA blend, (b) TPS/PLA/CP1 composite, (c) TPS/PLA/CP2 composite, (d)
TPS/PLA/CP3 composite, (e) TPS/PLA/CP4 composite, and (f) TPS/PLA/CP5 composite. Data are reported as mean � SD, n ¼ 3. Different lower-case letters indicate
a significant difference at p < 0.05 (Duncan’s new multiple range test).

the reinforcing effect from the homogeneous dispersion of hemp cellu­

lose in starch matrix is a crucial phenomenon to improve both strength
and stiffness of the starch/hemp cellulose nanocrystals film.
It should be pointed out that when CP concentration was further
increased to 17.7 wt% and 22.1 wt%, elongation at break of TPS/PLA/
CP4 and TPS/PLA/CP5 composites dropped, although their tensile
strength and Young’s modulus slightly increased due to the reinforcing
effect. The reduction of extensibility might be a result of the co-
continuous phase structure of TPS and PLA, as confirmed by SEM,
leading to restricted mobility of TPS by PLA intervention.
It was interesting to note that the TPS/PLA (60/40) blend containing
CP obtained from the current work exhibited competitive tensile
strength with the TPS/PLA (20/80) blend containing CP with the same
concentration range (~9 wt%, the maximum concentration of CP
applied in the work of Teixeira et al.) prepared by Teixeira, Curvelo,
Corr^ea, Marconcini, Glenn and Mattoso [16], though the amount of PLA
was much lower. However, elongation at break and Young’s modulus of
the materials were depending on the weight proportion of TPS to PLA;
the higher proportion of TPS to PLA provided the material with the
higher elongation at break, but lower Young’s modulus.

3.5. Crystal type and crystallinity of TPS/PLA/CP biocomposites by XRD

Fig. 5. X-ray diffraction patterns of (a) TPS/PLA blend, (b) TPS/PLA/CP1
The TPS/PLA blend exhibited diffraction peaks at 2θ of 13.1� , 18.1� , composite, (c) TPS/PLA/CP2 composite, (d) TPS/PLA/CP3 composite, (e) TPS/
19.7� , and 24.2� (Fig. 5a). The peaks at 2θ of 13.1� and 19.7� belonged to PLA/CP4 composite, and (f) TPS/PLA/CP5 composite.
Vh-type crystal formed by single helical amylose with guest molecules
such as glycerol during extrusion or cooling [16,48], while the peak at mobility.
2θ of 18.1� was ascribed to Eh-type crystal [49,50]. The peak at 2θ of Herein, TPS/PLA/CP composites also showed higher peak intensity
24.2� was assigned to B-type crystal, which derived from the recrystal­ of B-type crystals than the TPS/PLA blend, implying that the retrogra­
lization of amylose and amylopectin chains during retrogradation, and it dation of TPS in the composites was more pronounced than that in the
took place quickly right after the extrusion process, as well as during blend because of the greater plasticizing effect.
storage, and was sensitive to hydration [51]. Moreover, all TPS/PLA/CP composites exhibited new peak at 2θ of
Considering XRD patterns of TPS/PLA/CP composites, only B-type 16.7� and the intensity of this peak increased as a function of CP con­
and Vh-type crystals were observed. The disappearance of Eh-peak at 2θ centration. This peak might be relevant to B-type diffraction of the ret­
~18.1� might be explained by the increased plasticization of TPS in the rograded TPS in the presence of cellulose fibers as reported by Martins,
TPS/PLA/CP composites as compared with that in the TPS/PLA blend Magina, Oliveira, Freire, Silvestre, Neto and Gandini [54]. They also
because of the decreased starch fraction (or increased glycerol/starch found that the intensity of this peak increased with fiber concentration
weight ratio) in the composites. Battegazzore, Bocchini, Nicola, Martini and inferred that fibers facilitated the retrogradation process of TPS to
and Frache [50] and van Soest, Hulleman, de Wit and Vliegenthart [49] increase its crystallinity. In contrast, some research groups reported that
reported that Eh-type crystal peak at 2θ ~18� appeared when the plas­ cellulosic fibers retarded the recrystallization of starch due to their as­
ticizing effect was pronounced. sociation via H-bond, resulting in reduced starch-starch chain interac­
It should be noticed that Vh-type crystallinity of the TPS/PLA/CP tion [55,56]. However, in the present work the plasticizing effect by
composites increased with increasing CP content because glycerol/ glycerol to favor starch retrogradation was more pronounced than the
starch weight fraction and lipid content increased. CP contained 0.1 wt formation of interaction between fiber and starch because glycerol/­
% of lipid, which is higher than cassava starch (0.04 wt%). Tang and starch weight proportion increased with CP concentration.
Copeland [52] reported that increasing the concentration of lipid led to
an increase in the amount of starch–lipid complexes. Lin and Tung [53] 3.6. Thermal property of TPS/PLA/CP biocomposites by DSC
also revealed that the increased glycerol content led to increased Vh-type
and B-type crystals due to the increasingly higher polymer chain Glass transition temperature (Tg), cold crystallization temperature

7
P. Jullanun and R. Yoksan Polymer Testing 88 (2020) 106522

(Tcc), and melting temperature (Tm) of PLA in the TPS/PLA blend and broader and shallower single peak when CP concentration was
TPS/PLA/CP composites were determined by DSC. Herein, the result increased. The result suggested that PLA crystals were molten at lower
from the second heating scan after erasing the thermal history of the temperature, probably due to the imperfect PLA crystal formation
polymer was discussed as shown in Fig. 6. interfered by CP.
The second heating scan DSC thermogram of the TPS/PLA blend and
TPS/PLA/CP composites showed only the PLA information. The TPS/ 3.7. Thermal stability and decomposition temperature of TPS/PLA/CP
PLA blend possessed Tg of PLA at 44.5 � C, Tcc of PLA at 92.7 � C, and biocomposites by TGA
double peaks of Tm of PLA at 131.3 � C and 142.2 � C. The melting peak at
lower temperature belonged to less perfect crystals created during pre­ Thermal stability of the TPS/PLA blend and TPS/PLA/CP composites
vious cooling, whereas the melting peak at higher temperature was was investigated by TGA. Fig. 7A presents weight loss due to the
attributed to a more perfect crystalline structure of PLA [57,58]. It degradation of each component in the materials as a function of tem­
should be pointed out that the disappearance of Tm of TPS might be perature. The TPS/PLA blend showed four-step weight loss at 80–101
relevant to the insufficient time for recrystallization of TPS during �
C, 182–228 � C, 307–328 � C, and 331–369 � C (Fig. 7Aa) belonging to the
cooling [59], since starch possesses very large molecules of amylopectin evaporation of water [62,63], the loss of glycerol [64,65], the decom­
and amylose. position of TPS [14,66], and the decomposition of PLA [67], respec­
As compared with the TPS/PLA blend, the TPS/PLA/CP composites tively. Fig. 7Ba confirmed the decomposition temperatures (Td),
showed lower Tg, Tcc, and Tm of PLA at the temperature ranges of corresponding to the highest weight loss of such component in TGA
5.5–34.8 � C, 82.1–87.2 � C, and 107.8–132.5 � C, respectively. In addi­ thermogram, of TPS and PLA appeared at 316.2 � C and 345.3 � C,
tion, Tg, Tcc, and Tm of PLA shifted toward lower temperatures when CP respectively. The two distinct peaks in the DTG thermogram (Fig. 7Ba)
content was increased. The reduction of Tg of PLA in the composites indicated the immiscible and incompatible blend [68].
might be associated with the higher glycerol/starch weight fraction. Similar to the TPS/PLA blend, TPS/PLA/CP1 composite containing
Glycerol could migrate from TPS phase to PLA phase, resulting in higher 4.4 wt% of CP also exhibited four-step weight loss; however, its thermal
mobility of the PLA chain [60,61]. It should be pointed out that by stability became lower (Fig. 7Ab) and Td of TPS and PLA moved toward
adding CP, the broader cold crystallization peak of PLA and its shift each other (Fig. 7Bb). The reduction of thermal stability was more
toward lower temperature were observed, suggesting that cold crystal­ pronounced for the composites with higher content of CP (i.e., 8.8–22.1
lization of PLA in the composites took place over a wider range of wt%; Fig. 7Ac f). This might be attributed to the higher plasticizing
temperatures and faster than that in the blend. This result reflected that effect weakening the strong intermolecular bonds among polymer
CP might act as a nucleating agent for PLA. A similar result was observed chains. Similar behavior was observed in the case of glycerol plasticized
by Teixeira, Curvelo, Corr^ ea, Marconcini, Glenn and Mattoso [16]. By sugar palm starch; its thermal resistance decreased when glycerol con­
considering Tm of PLA in the TPS/PLA/CP composites, it not only shifted centration was increased [69]. In addition, the composites containing
toward lower temperature, but also transformed from double peaks to a CP content of 8.8–22.1 wt% exhibited three-step weight loss at 60–118
�
C, 150–268 � C, and 294–348 � C, assigning to the evaporation of water,
the loss of glycerol, and the decomposition of both TPS and PLA,
respectively (Fig. 7Ac f). Fig. 7Bc f also confirmed that TPS and PLA in
the composites decomposed at almost the same temperature as only one
decomposition peak appearing at 324–335 � C was observed. The
decomposition peaks of both TPS and PLA in the composites turned
sharper and combined to form a single peak when CP concentration was
increased. This might be a result of the increased glycerol/starch weight
fraction; thus, a higher amount of glycerol migrated from TPS phase to
PLA phase resulting in increased Td of TPS, but decreased Td of PLA.

4. Conclusions

The main goal of the current study was to reduce the utilization of
PLA by replacing with TPS produced from CP and cassava starch. TPS/
PLA/CP biocomposites were prepared by an extrusion and further con­
verted into dumbbell-shaped specimens by an injection molding. CP was
used to replace starch in the TPS preparation step and its concentration
was varied in the range of 4.4–22.1 wt%. PLA content was maintained at
40 wt%. Incorporation of CP resulted in reduced melt flow ability and
increased shear viscosity of the TPS/PLA blend. CP also facilitated a
good mixing between TPS and PLA phases as evidenced from the
morphological transformation from a discrete phase structure (i.e., PLA
droplets in TPS matrix in the case of TPS/PLA blend) to an elongated
rod-like structure when 4.4 wt%, 8.8 wt%, and 13.3 wt% of CP were
loaded, and eventually a co-continuous phase structure with further
increasing CP concentration to 17.7 wt% and 22.1 wt%. CP provided an
efficient reinforcing effect to the TPS/PLA blend as confirmed from the
significantly increased tensile strength and Young modulus, such as up
to 354% and 722%, respectively, for the biocomposite with the highest
CP concentration of 22.1 wt%. This reinforcing effect increased as a
function of CP concentration. However, extensibility of the TPS/PLA/CP
Fig. 6. DSC thermograms at second heating scan of (a) TPS/PLA blend, (b) biocomposites containing CP loaded in the range of 4.4–13.3 wt%
TPS/PLA/CP1 composite, (c) TPS/PLA/CP2 composite, (d) TPS/PLA/CP3 increased with increasing CP concentration due to the improved
composite, (e) TPS/PLA/CP4 composite, and (f) TPS/PLA/CP5 composite. compatibility, but it decreased for the biocomposites with high CP

8
P. Jullanun and R. Yoksan Polymer Testing 88 (2020) 106522

Fig. 7. (A) TGA and (B) DTG curves of (a) TPS/PLA blend, (b) TPS/PLA/CP1 composite, (c) TPS/PLA/CP2 composite, (d) TPS/PLA/CP3 composite, (e) TPS/PLA/CP4
composite, and (f) TPS/PLA/CP5 composite.

concentrations of 17.7–22.1 wt% owing to the more outstanding CP [3] Y. Jiugao, W. Ning, M. Xiaofei, The effects of citric acid on the properties of
thermoplastic starch plasticized by glycerol, Starch - St€ arke 57 (10) (2005)
agglomeration and large void formation at interfaces. Moreover, the
494–504.
incorporation of CP caused increased Vh-type and B-type crystallinity [4] S.H.D. Hulleman, F.H.P. Janssen, H. Feil, The role of water during plasticization of
(retrogradation) of TPS in the biocomposites, decreased Tg, Tcc, Tm, and native starches, Polymer 39 (10) (1998) 2043–2048.
Td of PLA in the biocomposites, and reduced thermal stability of the [5] J. Waterschoot, S.V. Gomand, E. Fierens, J.A. Delcour, Production, structure,
physicochemical and functional properties of maize, cassava, wheat, potato and
biocomposites due to the plasticizing effect from glycerol. CP also rice starches, Starch - St€arke 67 (1–2) (2015) 14–29.
played a role as a nucleating agent for PLA in the biocomposites. [6] A. Pandey, C.R. Soccol, P. Nigam, V.T. Soccol, L.P.S. Vandenberghe, R. Mohan,
Biotechnological potential of agro-industrial residues. II: cassava bagasse,
Bioresour. Technol. 74 (1) (2000) 81–87.
Data availability [7] M.P. Cereda, Caracterizaç~ ao dos resíduos da industrializaç~
ao da mandioca,
Resíduos da industrializaç~ ao da mandioca no Brasil, S~ao Paulo: Paulic�eia, 1994,
pp. 11–50.
The raw/processed data required to reproduce these findings cannot
[8] S.C. Stertz, Bioconvers~ ao da farinha de mandioca crua (manihot esculenta, crantz)
be shared at this time as the data also forms part of an ongoing study. por fungos do g^ enero rhizopus em fermentaç~ ao no estado s�olido, Universidade
Federal do Parana Curitiba, 1997.
Declaration of competing interest [9] L.P.S. Vandenberghe, C.R. Soccol, J.M. Lebeault, N. Krieger, Cassava wastes
hydrolysate an alternative carbon source for citric acid production by Candida
lipolytica, Int. Congr. Biotech 98 (1998).
The authors declare that they have no known competing financial [10] C.R. Soccol, L.P.S. Vandenberghe, Overview of applied solid-state fermentation in
interests or personal relationships that could have appeared to influence Brazil, Biochem. Eng. J. 13 (2) (2003) 205–218.
[11] L.P.S. Vandenberghe, C.R. Soccol, A. Pandey, J.M. Lebeault, Solid-state
the work reported in this paper. fermentation for the synthesis of citric acid by Aspergillus Niger, Bioresour.
Technol. 74 (2) (2000) 175–178.
[12] R.P. John, K.M. Nampoothiri, A. Pandey, Solid-state fermentation for l-lactic acid
CRediT authorship contribution statement
production from agro wastes using Lactobacillus delbrueckii, Process Biochem. 41
(4) (2006) 759–763.
Phatcharin Jullanun: Validation, Formal analysis, Investigation, [13] P. Panichnumsin, A. Nopharatana, B. Ahring, P. Chaiprasert, Production of
Data curation, Writing - original draft, Visualization. Rangrong Yok­ methane by co-digestion of cassava pulp with various concentrations of pig
manure, Biomass Bioenergy 34 (8) (2010) 1117–1124.
san: Conceptualization, Methodology, Resources, Writing - review & [14] A. Edhirej, S.M. Sapuan, M. Jawaid, N. Zahari, Preparation and characterization of
editing, Supervision, Project administration, Funding acquisition. cassava bagasse reinforced thermoplastic cassava starch, Fibers Polym. 18 (1)
(2017) 162–171.
[15] E.M. Teixeira, A.L. R� oz, A.J.F. Carvalho, A.A. Silva Curvelo, Preparation and
Acknowledgements characterisation of thermoplastic starches from cassava starch, cassava root and
cassava bagasse, Macromol. Symp. 229 (1) (2005) 266–275.
[16] E.M. Teixeira, A.A.S. Curvelo, A.C. Corr^ea, J.M. Marconcini, G.M. Glenn, L.H.
The authors wish to thank the National Science and Technology
C. Mattoso, Properties of thermoplastic starch from cassava bagasse and cassava
Development Agency (NSTDA), Thailand; the National Research Council starch and their blends with poly (lactic acid), Ind. Crop. Prod. 37 (1) (2012)
of Thailand (NRCT); and the Graduate School of Kasetsart University, 61–68.
Thailand, for the financial support. Appreciation is extended to Siam [17] J. Giron�es, J.P. L�
opez, P. Mutj�e, A.J.F. Carvalho, A.A.S. Curvelo, F. Vilaseca,
Natural fiber-reinforced thermoplastic starch composites obtained by melt
Modified Starch Co., Ltd., Thailand, for providing cassava pulp and processing, Compos. Sci. Technol. 72 (7) (2012) 858–863.
Mettler-Toledo (Thailand) Ltd. for obligingness on DSC and TGA [18] X.F. Ma, J.G. Yu, J.F. Kennedy, Studies on the properties of natural fibers-
measurements. reinforced thermoplastic starch composites, Carbohydr. Polym. 62 (1) (2005)
19–24.
[19] L. Averous, L. Moro, P. Dole, C. Fringant, Properties of thermoplastic blends:
Appendix A. Supplementary data starch–polycaprolactone, Polymer 41 (11) (2000) 4157–4167.
[20] M.K.A. Wahab, N. Othman, H. Ismail, Thermoplastic elastomers from high density
polyethylene/natural rubber/thermoplastic tapioca starch: effects of different
Supplementary data to this article can be found online at https://doi. dynamic vulcanization, Natural Rubber Materials 1 (2014) 242–264. Blends and
org/10.1016/j.polymertesting.2020.106522. IPNs, The Royal Society of Chemistry.
[21] S. Yoon, S. Chough, H.J. Park, Properties of starch-based blend films using citric
acid as additive. II, J. Appl. Polym. Sci. 100 (3) (2006) 2554–2560.
References [22] J.M. Ferri, D. Garcia-Garcia, A. Carbonell-Verdu, O. Fenollar, R. Balart, Poly(lactic
acid) formulations with improved toughness by physical blending with
[1] M.C. Li, U.R. Cho, Starch in rubber based blends and micro composites, in: P. thermoplastic starch, J. Appl. Polym. Sci. 135 (4) (2018) 45751–45759.
M. Visakh (Ed.), Rubber Based Bionanocomposites: Preparation, Springer [23] F. Carrasco, P. Pag� es, J. G�
amez-P�erez, O.O. Santana, M.L. Maspoch, Processing of
International Publishing, Switzerland, 2017, pp. 109–140. poly(lactic acid): characterization of chemical structure, thermal stability and
[2] E.M. Teixeira, A.L. Da R�
oz, A.J.F. Carvalho, A.A.S. Curvelo, The effect of glycerol/ mechanical properties, Polym, Degradation Stab 95 (2) (2010) 116–125.
sugar/water and sugar/water mixtures on the plasticization of thermoplastic
cassava starch, Carbohydr. Polym. 69 (4) (2007) 619–624.

9
P. Jullanun and R. Yoksan Polymer Testing 88 (2020) 106522

[24] N. Wang, J.G. Yu, P.R. Chang, X.F. Ma, Influence of formamide and water on the [46] M. Hietala, A.P. Mathew, K. Oksman, Bionanocomposites of thermoplastic starch
properties of thermoplastic starch/poly(lactic acid) blends, Carbohydr. Polym. 71 and cellulose nanofibers manufactured using twin-screw extrusion, Eur. Polym. J.
(1) (2008) 109–118. 49 (4) (2013) 950–956.
[25] N. Wang, J. Yu, X. Ma, Preparation and characterization of compatible [47] X. Cao, Y. Chen, R. Chang, M. Stumborg, A. Huneault, Green composites reinforced
thermoplastic dry starch/poly(lactic acid), Polym. Compos. 29 (5) (2008) with hemp nanocrystals in plasticized starch, J. Appl. Polym. Sci. 109 (6) (2008)
551–559. 3804–3810.
[26] O. Martin, L. Av� erous, Poly(lactic acid): plasticization and properties of [48] J.J.G. van Soest, P. Essers, Influence of amylose-amylopectin ratio on properties of
biodegradable multiphase systems, Polymer 42 (14) (2001) 6209–6219. extruded starch plastic sheets, J. Macromol. Sci. A. 34 (9) (1997) 1665–1689.
[27] E. Chabrat, H. Abdillahi, A. Rouilly, L. Rigal, Influence of citric acid and water on [49] J.J.G. van Soest, S.H.D. Hulleman, D. de Wit, J.F.G. Vliegenthart, Crystallinity in
thermoplastic wheat flour/poly(lactic acid) blends. I: thermal, mechanical and starch bioplastics, Ind. Crop. Prod. 5 (1) (1996) 11–22.
morphological properties, Ind. Crop. Prod. 37 (1) (2012) 238–246. [50] D. Battegazzore, S. Bocchini, G. Nicola, E. Martini, A. Frache, Isosorbide, a green
[28] N. Wang, J.G. Yu, P.R. Chang, X.F. Ma, Influence of citric acid on the properties of plasticizer for thermoplastic starch that does not retrogradate, Carbohydr. Polym.
glycerol-plasticized dry starch (DTPS) and DTPS/poly(lactic acid) blends, Starch - 119 (2015) 78–84.
St€
arke 59 (9) (2007) 409–417. [51] P. Myll€arinen, A. Buleon, R. Lahtinen, P. Forssell, The crystallinity of amylose and
[29] N. Wang, X. Zhang, N. Han, J. Fang, Effects of water on the properties of amylopectin films, Carbohydr. Polym. 48 (1) (2002) 41–48.
thermoplastic starch poly(lactic acid) blend containing citric acid, J. Thermoplast. [52] M.C. Tang, L. Copeland, Analysis of complexes between lipids and wheat starch,
Compos. Mater. 23 (1) (2010) 19–34. Carbohydr. Polym. 67 (1) (2007) 80–85.
[30] E. Chabrat, H. Abdillahi, A. Rouilly, L. Rigal, Influence of citric acid and water on [53] C. Lin, C. Tung, The preparation of glycerol pseudo-thermoplastic starch (GTPS)
thermoplastic wheat flour/poly(lactic acid) blends. I: thermal, mechanical and via gelatinization and plasticization, Polym-Plast. Technol. 48 (5) (2009) 509–515.
morphological properties, Ind. Crop. Prod. 37 (1) (2012) 238–246. [54] I.M.G. Martins, S.P. Magina, L. Oliveira, C.S.R. Freire, A.J.D. Silvestre, C.P. Neto,
[31] N.F. Zaaba, H. Ismail, A review on tensile and morphological properties of poly A. Gandini, New biocomposites based on thermoplastic starch and bacterial
(lactic acid) (PLA)/thermoplastic starch (TPS) blends, Polymer-Plastics Technology cellulose, Compos. Sci. Technol. 69 (13) (2009) 2163–2168.
and Materials 58 (18) (2019) 1945–1964. [55] W. Xia, G. Fu, C. Liu, Y. Zhong, J. Zhong, S. Luo, W. Liu, Effects of cellulose, lignin
[32] F. Harnnecker, D.S. Rosa, D.M. Lenz, Biodegradable polyester-based blend and hemicellulose on the retrogradation of rice starch, Food Sci. Technol. Res. 20
reinforced with curau� a fiber: thermal, mechanical and biodegradation behaviour, (2) (2014) 375–383.
J. Polym. Environ. 20 (1) (2011) 237–244. [56] S. Cui, M. Li, S. Zhang, J. Liu, Q. Sun, L. Xiong, Physicochemical properties of
[33] P. Shah, R. Prajapati, P. Singh, Enrichment of mechanical properties of maize and sweet potato starches in the presence of cellulose nanocrystals, Food
biodegradable composites containing waste cellulose fiber and thermoplastic Hydrocolloids 77 (2018) 220–227.
starch, EJAET 4 (4) (2017) 282–286. [57] A.N. Frone, S. Berlioz, J.F. Chailan, D.M. Panaitescu, Morphology and thermal
[34] K. Hamad, M. Kaseem, F. Deri, Rheological and mechanical characterization of properties of PLA-cellulose nanofibers composites, Carbohydr. Polym. 91 (1)
poly(lactic acid)/polypropylene polymer blends, J. Polym. Res. 18 (6) (2011) (2013) 377–384.
1799–1806. [58] C. Fonseca, A. Ochoa, M.T. Ulloa, E. Alvarez, D. Canales, P.A. Zapata, Poly(lactic
[35] W. Sinthavathavorn, M. Nithitanakul, B.P. Grady, R. Magaraphan, Melt rheology acid)/TiO(2) nanocomposites as alternative biocidal and antifungal materials,
and die swell of PA6/LDPE blends by using lithium ionomer as a compatibilizer, Mater. Sci. Eng. C Mater. Biol. Appl. 57 (2015) 314–320.
Polym. Bull. 63 (1) (2009) 23–35. [59] J. Stagner, V. Dias Alves, R. Narayan, A. Beleia, Thermoplasticization of high
[36] J. George, R. Janardhan, J.S. Anand, S.S. Bhagawan, S. Thomas, Melt rheological amylose starch by chemical modification using reactive extrusion, J. Polym.
behaviour of short pineapple fibre reinforced low density polyethylene composites, Environ. 19 (3) (2011) 589–597.
Polymer 37 (24) (1996) 5421–5431. [60] H. Li, M.A. Huneault, Crystallization of PLA/thermoplastic starch blends, Int.
[37] R.J. Crowson, M.J. Folkes, Rheology of short glass fiber-reinforced thermoplastics Polym. Process. 23 (5) (2008) 412–418.
and its application to injection molding. II. The effect of material parameters, [61] P. Müller, B. Imre, J. Bere, J. M�ocz�
o, B. Puk�
anszky, Physical ageing and molecular
Polym. Eng. Sci. 20 (14) (1980) 934–940. mobility in PLA blends and composites, J. Therm. Anal. Calorim. 122 (3) (2015)
[38] R.J. Crowson, M.J. Folkes, P.F. Bright, Rheology of short glass fiber-reinforced 1423–1433.
thermoplastics and its application to injection molding I. Fiber motion and [62] K.M. Dang, R. Yoksan, Development of thermoplastic starch blown film by
viscosity measurement, Polym. Eng. Sci. 20 (14) (1980) 925–933. incorporating plasticized chitosan, Carbohydr, Polymer 115 (2015) 575–581.
[39] C.Y. Khor, Z.M. Ariff, F.C. Ani, M.A. Mujeebu, M.K. Abdullah, M.Z. Abdullah, M. [63] F.M. Pelissari, M.V. Grossmann, F. Yamashita, E.A. Pineda, Antimicrobial,
A. Joseph, Three-dimensional numerical and experimental investigations on mechanical, and barrier properties of cassava starch-chitosan films incorporated
polymer rheology in meso-scale injection molding, Int. Commun. Heat Mass Tran. with oregano essential oil, J. Agric. Food Chem. 57 (16) (2009) 7499–7504.
37 (2) (2010) 131–139. [64] J.B. Olivato, J. Marini, E. Pollet, F. Yamashita, M.V. Grossmann, L. Averous,
[40] B. Ayana, S. Suin, B.B. Khatua, Highly exfoliated eco-friendly thermoplastic starch Elaboration, morphology and properties of starch/polyester nano-biocomposites
(TPS)/poly (lactic acid)(PLA)/clay nanocomposites using unmodified nanoclay, based on sepiolite clay, Carbohydr. Polym. 118 (2015) 250–256.
Carbohydr. Polym. 110 (2014) 430–439. [65] M.Z.B. Yunos, W.A.W.A. Rahman, Effect of glycerol on performance rice straw/
[41] M.R.A. Moghaddam, S.M.A. Razavi, Y. Jahani, Effects of compatibilizer and starch based polymer, J. Appl. Sci. 11 (13) (2011) 2456–2459.
thermoplastic starch (TPS) concentration on morphological, rheological, tensile, [66] P. Boonprasith, J. Wootthikanokkhan, N. Nimitsiriwat, Mechanical, thermal, and
thermal and moisture sorption properties of plasticized polylactic acid/TPS blends, barrier properties of nanocomposites based on poly(butylene succinate)/
J. Polym. Environ. 26 (8) (2018) 3202–3215. thermoplastic starch blends containing different types of clay, J. Appl. Polym. Sci.
[42] N. Raghu, A. Kale, A. Raj, P. Aggarwal, S. Chauhan, Mechanical and thermal 130 (2) (2013) 1114–1123.
properties of wood fibers reinforced poly(lactic acid)/thermoplasticized starch [67] M.M.F. Ferrarezi, M. de Oliveira Taipina, L.C. Escobar da Silva, M.C. Gonçalves,
composites, J. Appl. Polym. Sci. 135 (15) (2018) 46118–46128. Poly(ethylene glycol) as a compatibilizer for poly(lactic acid)/thermoplastic starch
[43] M. Barghamadi, M. Karrabi, M.H.R. Ghoreishy, S. Mohammadian-Gezaz, Effects of blends, J. Polym. Environ. 21 (1) (2012) 151–159.
two types of nanoparticles on the cure, rheological, and mechanical properties of [68] V. Mittal, T. Akhtar, N. Matsko, Mechanical, thermal, rheological and
rubber nanocomposites based on the NBR/PVC blends, J. Appl. Polym. Sci. 136 morphological properties of binary and ternary blends of PLA, TPS and PCL,
(25) (2019). Macromol. Mater. Eng. 300 (4) (2015) 423–435.
[44] G.M. Jordhamo, J.A. Manson, L.H. Sperling, Phase continuity and inversion in [69] M. Sanyang, S.M. Sapuan, M. Jawaid, M. Ishak, J. Sahari, Effect of plasticizer type
polymer blends and simultaneous interpenetrating networks, Polym. Eng. Sci. 26 and concentration on tensile, thermal and barrier properties of biodegradable films
(8) (1986) 517–524. based on sugar palm (arenga pinnata) starch, Polymers 7 (6) (2015) 1106–1124.
[45] H.M. Heidemann, M.E.R. Dotto, J.B. Laurindo, B.A.M. Carciofi, C. Costa, Cold
plasma treatment to improve the adhesion of cassava starch films onto PCL and
PLA surface, Colloids Surf. Physicochem. Eng. Aspects 580 (2019).

10